Advances in Agronomy, Volume 97

CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors’ contributions begin. M. Ashraf (45) Depa...

Author: Donald L. Sparks

261 downloads 995 Views 9MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors’ contributions begin.

M. Ashraf (45) Department of Botany, University of Agriculture, Faisalabad 30840, Pakistan H. R. Athar (45) Institute of Pure and Applied Biology, Bahauddin Zakariya University, Multan, Pakistan E. Ben-Dor (321) Department of Geography and Human Environment, Tel Aviv University, Tel Aviv, Israel 69978 J. Bouma (175) Wageningen University and Research Centre, Wageningen, The Netherlands P. Cannavo (131) INH, De´partement Geńie Agronomique, UR EPHOR, F-49045 Angers, France S. Chabrillat (321) GeoForchungsZentrum (GFZ) Potsdam, Section 1.4: Remote Sensing, Potsdam, Germany Julian J. C. Dawson (1) Institute of Biological & Environmental Sciences, School of Biological Sciences, University of Aberdeen, Aberdeen AB24 3UU, Scotland, United Kingdom; and Environment and Hydrology, School of Geosciences, University of Aberdeen, Aberdeen AB24 3UF, Scotland, United Kingdom J. A. M. Dematteˆ (321) Department of Soil Science, Saõ Paulo University, ESALQ, Piracicaba, Brazil J. A. de Vos (175) Wageningen University and Research Centre, Wageningen, The Netherlands Changming Fang (1) Coastal Ecosystems Research Station of Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, Shanghai 200433, China P. J. C. Harris (45) Faculty of Business, Environment and Society, Coventry University, Coventry, United Kingdom ix

x

Contributors

G. B. M. Heuvelink (175) Wageningen University and Research Centre, Wageningen, The Netherlands J. Hill (321) Remote Sensing Department, Faculty of Geography/Geosciences, Trier University, Trier, Germany T. R. Kwon (45) The National Institute of Agricultural Biotechnology, Suwon 441-707, Korea John B. Moncrieff (1) Atmospheric and Environmental Science, Edinburgh EH9 3JN, Scotland, United Kingdom Mustafa Pala (273) International Center for Agricultural Research in the Dry Areas (ICARDA), Aleppo, Syria V. Parnaudeau (131) INRA, UMR 1069 Sol-Agronomie-Spatialisation, F-35042 Rennes, France Jeff S. Piotrowski (111) Division of Biological Sciences, University of Montana, 507 Health Sciences, Missoula, Montana 59812 R. Reau (131) INRA, UMR 211 Agronomie, F-78850 Thiverval Grignon, France S. Recous (131) INRA, UR 1158 Agronomie, F-02000 Laon, France Matthias C. Rillig (111) Freie Universita¨t Berlin, Institut fu¨r Biologie, D-14195 Berlin, Germany John Ryan (273) International Center for Agricultural Research in the Dry Areas (ICARDA), Aleppo, Syria Murari Singh (273) International Center for Agricultural Research in the Dry Areas (ICARDA), Aleppo, Syria Pete Smith (1) Institute of Biological & Environmental Sciences, School of Biological Sciences, University of Aberdeen, Aberdeen AB24 3UU, Scotland, United Kingdom S. Sommer (321) European Commission—DG Joint Research Centre, Institute for Environment and Sustainability, Ispra, Italy

Contributors

xi

M. P. W. Sonneveld (175) Wageningen University and Research Centre, Wageningen, The Netherlands J. J. Stoorvogel (175) Wageningen University and Research Centre, Wageningen, The Netherlands R. G. Taylor (321) School of Biological, Earth and Environmental Science, University of New South Wales, Sydney, Australia Benno P. Warkentin (239) Department of Crop and Soil Science, Oregon State University, Corvallis, Oregon 97331 M. L. Whiting (321) Department of Land Air Water Resources, University of California, Davis, California

PREFACE

Volume 97 contains eight excellent reviews on various aspects of the crop and soil sciences. Chapter 1 is a timely review on the impact of global warming on soil organic carbon including discussions on warming trends, factors impacting the response of soil carbon to global warming and development of models to predict this response, and research on the temperature sensitivity of soil organic carbon pools. Chapter 2 covers strategies for improving salt tolerance, a major environmental constraint to crop productivity globally. This review discusses strategies for growing plants on saline soils including seed priming, exogenous application of organic chemicals such as plant growth regulators to plants that are under salinity stress, and the use of plant selection and breeding. Chapter 3 deals with succession of arbuscular mycorrhizal fungi including patterns, causes, and considerations in organic agriculture. Chapter 4 is a comprehensive treatise on modeling dynamics to assess environmental impacts of cropped soils. Chapter 5 is a thought-provoking and timely review on scientists’ role in multiscale land use analysis from a Dutch communities perspective. A number of case studies at both the farm and regional levels are discussed. Chapter 6 is an historical perspective on soil structure with discussions covering periods from the pre-Renaissance up to modern times. Chapter 7 deals with the implications of long-term cereal-based rotation trials in the Mediterranean region on cropping sustainability, including a synthesis of long-term trials at ICARDA. Chapter 8 is a timely review on the application of imaging spectrometry to soils with a discussion of fundamental aspects of the technique as well as its applications to various soil conditions and properties. I thank the authors for their fine contributions. DONALD L. SPARKS University of Delaware

xiii

C H A P T E R

O N E

Impact of Global Warming on Soil Organic Carbon Pete Smith,* Changming Fang,† Julian J. C. Dawson,*,‡ and John B. Moncrieff § Contents 1. Introduction 1.1. Scope of the review 1.2. Recent trends in global temperature 1.3. Why is the response of soil carbon response to global warming important? 2. Factors Affecting the Response of SOC to Global Warming 2.1. The balance of carbon inputs to, and outputs from, the soil 2.2. Increasing decomposition rate under global warming 2.3. Global and regional trends in changes in NPP and SOC loss 2.4. Overall impact—transient versus equilibrium effects 3. Temperature Sensitivity of SOC Pools 3.1. Summary of the debate on temperature sensitivity 3.2. Apparent versus actual temperature sensitivity 3.3. Q10 and beyond 4. Methods for Measuring Soil Responses to Global Warming 4.1. Soil respiration measurements in the laboratory 4.2. Soil respiration measurements in the field 4.3. Manipulation methods in the field 4.4. Eddy covariance for total net ecosystem exchange 5. Approaches to Modeling Soil Responses to Global Warming 5.1. Soil decomposition in models 5.2. Carbon input to the soil in models

* {

{

}

2 2 3 5 6 6 7 9 10 12 12 14 16 18 18 19 22 25 26 26 29

Institute of Biological & Environmental Sciences, School of Biological Sciences, University of Aberdeen, Aberdeen AB24 3UU, Scotland, United Kingdom Coastal Ecosystems Research Station of Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, Shanghai 200433, China Environment and Hydrology, School of Geosciences, University of Aberdeen, Aberdeen AB24 3UF, Scotland, United Kingdom Atmospheric and Environmental Science, Edinburgh EH9 3JN, Scotland, United Kingdom

Advances in Agronomy, Volume 97 ISSN 0065-2113, DOI: 10.1016/S0065-2113(07)00001-6

#

2008 Elsevier Inc. All rights reserved.

1

2

Pete Smith et al.

6. How Will Soil Carbon Respond the Global Warming? 6.1. New perspectives 6.2. Sensitive/vulnerable regions and soils 6.3. Reducing the vulnerability of soil C to the impacts of global warming 7. Conclusions References

29 29 30 31 32 33

Soils contain a stock of carbon that is about twice as large as that in the atmosphere and about three times that in vegetation. Small losses from this large pool could have significant impacts on future atmospheric carbon dioxide concentrations, so the response of soils to global warming is of critical importance when assessing climate carbon cycle feedbacks. Models that have coupled climate and carbon cycles show a large divergence in the size of the predicted biospheric feedback to the atmosphere. Central questions that still remain when attempting to reduce this uncertainty in the response of soils to global warming are (1) the temperature sensitivity of soil organic matter, especially the more recalcitrant pools; (2) the balance between increased carbon inputs to the soil from increased production and increased losses due to increased rates of decomposition; and (3) interactions between global warming and other aspects of global change, including other climatic effects (e.g., changes in water balance), changes in atmospheric composition (e.g., increasing atmospheric carbon dioxide concentration) and land-use change. In this chapter, we review trends in warming, factors affecting the response of soil carbon to global warming, evidence on the balance between changes in production and soil organic matter decomposition, recent research on the temperature sensitivity of soil organic carbon pools, methods for measuring soil responses to global warming, approaches to modeling soil responses to global warming, regions/ecosystems likely to be most vulnerable to future warming, and available technologies to reduce vulnerability of soil carbon to the impacts of future global warming.

1. Introduction 1.1. Scope of the review Climate change is no longer a subject for debate. The global climate is changing and there is now little doubt that humans have contributed significantly to climate change (IPCC WGI, 2007). Climate change manifests in many ways including changes in temperature, precipitation patterns, and seasonality. In this review, we will focus on the responses of soil organic carbon (SOC) to increased temperature, commonly termed global warming. Changes in precipitation amount and pattern are not dealt with in detail in this review, but will be considered where they affect soil responses to increasing temperature.

3

Impact of Global Warming on Soil Organic Carbon

1.2. Recent trends in global temperature 1.2.1. Trends in global temperature since preindustrial times IPCC Working Group I (IPCC WGI, 2007) states that warming of the climate system is unequivocal and presents evidence of this from increases in global average air and ocean temperatures. The temperature rise is predominantly because of increases in atmospheric greenhouse gas emissions, dominated by carbon dioxide (CO2). Current atmospheric concentrations of CO2 far exceed preindustrial values with increases since 1750 due primarily to emissions from fossil fuel use, agriculture, and land-use changes (IPCC WGI, 2007). 1.2.2. More recent trends in temperature 1.2.2.1. Global trends Eleven of the last 12 years rank among the 12 warmest years in the instrumental record of global surface temperature (since 1850). The 100-year linear trend (1906–2005) is 0.74 C (0.56–0.92). The linear rate of warming averaged over the last 50 years [0.13 C (0.10–0.16) per decade] is nearly twice that for the last 100 years (IPCC WGI, 2007), showing that the rate of global warming is increasing. Figure 1 (from IPCC WGI, 2007) shows change in temperature (relative to the 1961–1990 long-term average) between 1850 and 2000 (from IPCC WGI, 2007). 1.2.2.2. Regional trends There are different regional trends in temperature increase. Figure 2 shows how temperature has changed in different global regions over the last century (IPCC WGI, 2007). This may have significant impacts on soil carbon. For example, temperatures at the top of the permafrost layer have increased by up to 3 C since the 1980s in the Arctic. The maximum area covered by seasonally frozen ground has decreased by about 7% in the Northern Hemisphere since 1900 (IPCC WGI, 2007).

Global average temperature

0.0

14.0

−0.5

13.5

(⬚C)

14.5

1850

1900

1950

Temperature (⬚C)

0.5

2000

Figure 1 Changes in global average temperature 1850–2005 shown as a change ( C) relative to the 1961–1990 long-term average (from IPCC WGI, 2007).

0.5 0.0

1900

1950 Year

1900 Temperature anomaly (⬚C)

0.0 2000

South America 1.0 0.5

1950 Year Africa

2000

1.0 0.5

Asia 1.0 0.5 0.0

0.0 1900

1950 Year

2000

0.0 1900

1950 Year

1950 Year

1900 Temperature anomaly (⬚C)

0.5

1.0

Temperature anomaly (⬚C)


1.0



North America

Europe

2000

2000 Australia

1.0 0.5 0.0 1900

1950

2000

1.0 0.5 0.0

1900

1950 Year

2000

1.0 0.5 0.0

1900

1950 Year

2000


Global Land

Global Temperature anomaly (⬚C)


Year

Global Ocean 1.0 0.5 0.0

1900

1950 Year

2000


5

1.3. Why is the response of soil carbon response to global warming important? Globally, soils contain about 1500 Pg (1 Pg ¼ 1 Gt ¼ 1015 g) of organic carbon (Batjes, 1996), about three times the amount of carbon in vegetation and twice the amount in the atmosphere (IPCC WGI, 2001). The annual fluxes of CO2 from atmosphere to land [global net primary productivity (NPP)] and land to atmosphere (respiration and fire) are each of the order of 60 Pg C year1 (IPCC WGI, 2001). During the 1990s, fossil fuel combustion and cement production emitted 6.3 1.3 Pg C year1 to the atmosphere, while land-use change emitted 1.6 0.8 Pg C year1 (IPCC WGI, 2001; Schimel et al., 2001). Atmospheric C increased at a rate of 3.2 0.1 Pg C year1, the oceans absorbed 2.3 0.8 Pg C year1 with an estimated terrestrial sink of 2.3 1.3 Pg C year1 (IPCC WGI, 2001; Schimel et al., 2001). Soil carbon pools are smaller now than they were before human intervention. Historically, soils have lost between 40 and 90 Pg C globally through cultivation and disturbance (Houghton, 1999; Houghton et al., 1999; Lal, 1999; Schimel, 1995). The size of the pool of SOC is large compared to gross and net annual fluxes of carbon to and from the terrestrial biosphere. Figure 3 (IPCC WGI, 2001) shows a schematic diagram of the global carbon cycle. Small changes in the SOC pool could have dramatic impacts on the concentration of CO2 in the atmosphere. The response of SOC to global warming is, therefore, of critical importance. One of the first examples of the potential impact of increased release of terrestrial C on further climate change was given by Cox et al. (2000). Using a climate model with a coupled carbon cycle, Cox et al. (2000) showed that release of terrestrial carbon under warming would lead to a positive feedback, resulting in increased global warming. Since then, a number of coupled climate carbon cycle (so called C4) models have been developed. However, there remains considerable uncertainty concerning the extent of the terrestrial feedback, with the difference between the models amounting to differences in the atmospheric CO2 concentration of approximately 250 ppm by 2100 (Friedlingstein et al., 2006). This is of the same order as the difference between fossil fuel carbon emissions under the IPCC SRES emission scenarios (IPCC SRES, 2000). It is clear that better quantifying the response of terrestrial carbon, a large

Figure 2 Changes in continental- and global-scale decadal surface air temperature for 1906–2005, relative to the corresponding average for the 1901–1950 period, compared with model simulations. Lines indicate observed changes and are dashed where spatial coverage is less than 50%. Lower grey bands show the 5–95% range for 19 simulations from 5 climate models using only natural forcings and upper grey bands show the 5–95% range for 58 model simulations from 14 climate models using both natural and anthropogenic forcings. The changes shown are unadjusted model output in regions where observations are available (IPCC WGI, 2007).

6

Pete Smith et al.

A 0.4

Atmosphere [730]

120

Vulcanism < 0.1

Soil [1500]

Plants [500]

90

[Weathering 0.2] Total river transport 0.8

Land

0.6

[DOC 0.4] [Weathering 0.2]

Fossil organic carbon

Ocean [38000]

Rock carbonates

Sediment

Geological reservoirs B

Atmosphere Land uptake 1.9 Land use change 1.7 Fossil fuel burning 5.3

Land

5.4 Ocean uptake 1.9 Cement production 0.1

Ocean Fossil organic carbon

Rock carbonates

Sediment

Geological reservoirs

Figure 3 The global carbon cycle for the 1990s (Pg C): (A) the natural carbon cycle (DOC, dissolved organic carbon) and (B) the human perturbation (redrawn from IPCC WGI, 2001; as in Smith, 2004).

proportion of which derives from the soil, is essential for understanding the nature and extent of the earth’s response to global warming.

2. Factors Affecting the Response of SOC to Global Warming 2.1. The balance of carbon inputs to, and outputs from, the soil The level of SOC in a particular soil is determined by many factors including climatic factors (e.g., temperature and moisture regimes) and edaphic factors (e.g., soil parent material, clay content, cation exchange capacity; Dawson and Smith, 2007). For a given soil type however, SOC stock can also vary, the stock being determined by the balance of net carbon


7

inputs to the soil (as organic matter, OM) and net losses of carbon from the soil (as CO2, dissolved organic carbon and losses through erosion). Carbon inputs to the soil are largely determined by the land use, with forest systems tending to have the largest input of carbon to the soil (inputs all year round) and often this material is also the most recalcitrant, decomposing less readily. Grasslands also tend to have large inputs, although the material is often less recalcitrant than forest litter and the smallest input of carbon is often found in croplands that have inputs only when there is a crop growing and where the carbon inputs are among the most labile (Guo and Gifford, 2002; Smith, 2004). The smaller net input of carbon to the soil in croplands also results from removal of biomass in the harvested products, and can be further exacerbated by crop residue removal, by tillage that increases SOC loss by breaking open aggregates to expose protected organic carbon to weathering and microbial breakdown (Dawson and Smith, 2007), and also by changing the temperature regime of the soil. The rate of carbon input to the soil is also related to the productivity of the vegetation growing on that soil, measured by NPP. NPP varies with climate, land cover, species composition, and soil type (Falge et al., 2002). Moreover, NPP shows seasonal variation due to its dependence on light and temperature, for example broadleaf temperate forests are highly productive for part of the year only (Malhi et al., 2002). Over varying time periods, a proportion of NPP enters the soil as OM either via plant leachates, root exudates, or by decomposition of litter and fragmented plant structures ( Jones and Donnelly, 2004; Ostle et al., 2000), where it is converted back to CO2 and CH4 via soil (heterotrophic) respiration processes. The remaining C is termed net ecosystem production (NEP). However, other processes such as harvest, fire, and insect damage also remove C that when combined with riverine export and the respiratory (both autotrophic and heterotrophic) processes counterbalance the terrestrial CO2 input from gross primary production (GPP). Climate and land use are the main causes of temporal and spatial fluctuations between these opposing fluxes and can determine whether the terrestrial ecosystem is a source or sink for C (Cao et al., 1998; Dawson and Smith, 2007). Very small carbon inputs to the soil, as dissolved organic and inorganic carbon, comes from wet, dry, and occult (fog and cloud) deposition.

2.2. Increasing decomposition rate under global warming The balance between the input of C via photosynthesis and losses by respiration is a key aspect of soil functioning that determines whether a soil is a C source or sink (Freeman et al., 2004a; Gorham, 1991). The rate of decomposition of above- and belowground vegetative matter and litter to soil OM, and further decomposition/mineralization to DOC and CO2, respectively, is dependent on a number of environmental factors, including temperature, moisture, plant residue composition, and the capacity of the

8

Pete Smith et al.

soil to stabilize soil OM (Blanco-Canqui and Lal, 2004; Coleman and Jenkinson, 1996; Martens, 2000; Moore et al., 1998). These parameters in turn affect the activity of the soil decomposer community comprising microorganisms, fungi, and soil invertebrates. Temperature, in particular, is a major factor regulating decomposition rate; it influences the observed seasonal patterns of higher soil CO2 concentrations in the summer compared with the colder winter months (Castelle and Galloway, 1990; Hope et al., 1994; Jones and Mulholland, 1998). The temperature sensitivity of soil OM decomposition, however, remains a topic for debate (Davidson and Janssens, 2006; see Section 3.1). Although plants tend to be composed of similar C compounds, that is celluloses, starches, proteins, lipids, and phenolic compounds, the proportions of each depends on plant species and age (Cheshire and Chapman, 1996; Martens, 2000). As decomposition occurs, plant residues are reduced in size to soil OM as light or particulate OM fractions (Post and Kwon, 2000). Humic substances are the major components of soil OM, described as ‘‘complicated mixtures of biologically transformed organic debris’’ (Hayes and Clapp, 2001). They are insoluble but contain nonhumic components such as hydrocarbons, esters, acids, polysaccharides and are associated with soil minerals as well as the more recalcitrant plant residues (Hayes and Clapp, 2001). Physical and chemical protection of soil OM from microbial action further decreases decomposition rates, which are dependent on the ease with which organic C can be encapsulated within stable aggregates. The quantity and quality of plant residues determine the amount of soil C found in such aggregates. The physicochemical interactions between soil OM and its mineral constituents and ‘‘the mechanisms responsible for the dynamics of aggregate formation and stability,’’ with respect to C decomposition in soils are not, however, well-understood or well-quantified (Basile-Doelsch et al., 2005; Blanco-Canqui and Lal, 2004). The three main methods of organic C stabilization that have been proposed are (1) microaggregation (53–250 mm) formation within macroaggregates, (2) physical binding with clay and silt particles, and (3) biochemical formation of recalcitrant soil OM compounds (Bronick and Lal, 2005; Jones and Donnelly, 2004; Six et al., 2002). Decomposition is reduced in alkaline soils because precipitation of hydroxides, oxides, phosphates, and inorganic C as carbonates enhance the formation of aggregates. The formation of secondary carbonates is influenced by concentrations of soil organic C, calcium, and magnesium ions (Bronick and Lal, 2005). Light fractions and particulate OM that do not bind within aggregates, however, generally remain more susceptible to microbial decomposition (Six et al., 2002). More labile OM such as plant detritus and microbial biomass decomposes more rapidly, but this usually comprises only 3–5% of total OM ( Jones and Donnelly, 2004). The majority of the soil OM occurs in the micro- and macroaggregates that form around the light fractions, through binding of C


9

to mineral particles to form heavy fractions of organic C; these have residence times of decades (Post and Kwon, 2000). Disturbance increases breakdown of macroaggregates that ‘‘diminishes C stabilization in newly formed microaggregates within macroaggregates’’ (Six et al., 2004). Material that has been physically or biochemically protected can be extremely recalcitrant with mean turnover times of hundreds to thousands of years ( Jenkinson, 1990; Post and Kwon, 2000). In peatland and wetland systems, the rate of microbial-mediated decomposition is reduced by low temperatures, anaerobic conditions, poor substrate quality, acidic pH, and low O2 and nutrient availability, the combination of which tends to stabilize soil OM (Hobbie et al., 2001). Under anaerobic conditions, DOC rather than gaseous forms of C and N tends to be the primary end product of decomposition, although the proportions change as conditions become increasingly aerobic (Freeman et al., 2004b). Enzymatic inhibition, for example hydrolases which are involved in decomposition of labile OM to DOC and CO2, is increased by the prevalence of phenolic compounds common in peaty soils. However, the activity of phenol oxidases that can specifically break down complex compounds to less recalcitrant humic materials in the presence of O2 (Duran and Esposito, 2000) is reduced under anaerobic conditions, hence reducing the OM decomposition rate by the hydrolase enzymes (Freeman et al., 2001a, 2004a). This ‘‘enzymic latch’’ mechanism has been suggested as a major control on OM decomposition in peatland and wetland soils. However, the enzymic latch is still ultimately determined by other environmental factors (e.g., climate change, land use, disturbance, pollution), which influence anaerobicity within peatlands and hence the rate of C decomposition. The rate at which OM decomposes affects the rate at which C is lost from the soil. Decomposition rate varies considerably, but it is ultimately determined by the accessibility of OM to the microbial community, and hence the potential ability of a soil to retain C originally incorporated from the atmosphere. The decomposition/mineralization of OM ultimately results in losses of soil C as CO2 and CH4 to the atmosphere, as POC and DOC by erosion processes, and as losses of dissolved, particulate, and gaseous C to surface waters (Dawson and Smith, 2007).

2.3. Global and regional trends in changes in NPP and SOC loss As described in Sections 2.1 and 2.2, the overall response of soil carbon to global warming will depend on the balance of increased carbon inputs to the soil due to increased plant productivity (which will tend to increase SOC stocks) and the increasing rate of decomposition at warmer temperatures (which will tend to decrease SOC stocks). If warming-induced C emissions from soils exceed vegetation growth, soils could become sources of

10

Pete Smith et al.

atmospheric CO2 (Davidson and Janssens, 2006). It is not yet clear how these opposite effects will balance, either globally or for specific regions. Although there is some evidence that terrestrial ecosystems in middle and high latitudes of the Northern Hemisphere have functioned as carbon sinks over the past 20 years (Schimel et al., 2001), different studies from around the world report different trends in SOC stocks over the last decades, with a few authors suggesting that observed SOC losses may be due to climate change. Bellamy et al. (2005), for example, suggest a link to climate change to explain their observed mean loss of topsoil SOC of 0.6% year1, between 1978 and 2003 in England and Wales. However, Smith et al. (2007c) reported modeling results that suggested that, at most, only about 10–20% of recently observed soil carbon losses in England and Wales could possibly be attributable to climate warming. Janssens et al. (2003) suggest that the United Kingdom and Europe as a whole are a net CO2 sink, and other studies in nonagricultural areas of the United Kingdom suggest that topsoil SOC in British woodlands (Kirkby et al., 2005), from 1648 plots randomly located in 103 woods and sampled in 1971 and 2000–2003, show no significant change in SOC over 30 years (slight increase of þ0.38% over 30 years; þ0.01% year1). Other repeated sampling studies in Europe have shown contrasting results, with some showing loss of SOC (e.g., for Flemish cropland soils; Sleutel et al., 2003), attributed to changing manure application practices, and others showing no loss of SOC in Danish croplands (Heidmann et al., 2002) and Austrian soils (Dersch and Boehm, 1997). Elsewhere in the world, increases in SOC have been reported over recent decades. Liao et al. (2008) report increases in cropland SOC stocks over the last two decades in China’s Jiangsu province, but here the most likely explanation is a large contribution from a change in paddy field management over this period. Small-scale laboratory and field experiments and modeling studies suggest that climate change is likely to induce soil carbon loss from northern ecosystems (Goulden et al., 1998; Melillo et al., 2002), but little evidence comes from large-scale observations.

2.4. Overall impact—transient versus equilibrium effects Ecosystems respond to, and acclimate to, climate change in complicated ways that all affect SOC dynamics. Enhanced SOC decomposition rate and changed NPP are only two components of this response (Kirschbaum, 2006; Trumbore, 2006). Other important changes include shifts in plant species or vegetation type (Bachelet et al., 2001), adaptation of the soil microbial community (Sjursen et al., 2005), and changes in soil properties or processes (Rasmussen et al., 2006), which may occur over much longer time periods than changes in SOC decomposition and NPP in response to climate change. The short- to medium-term response, as reflected in


11

experimental studies and field measurements, may be only a transient response of SOC under the compound effects of many factors. An equilibrium response of SOC to global warming is the net difference between two equilibrium states of ecosystem, regardless of the length of time required to reach the new equilibrium after climate change. The behavior of soil C cycling during any transition period may be different from the estimated equilibrium responses because of the different rates at which SOC components respond, and other prevailing environmental conditions. In general, the transient response of SOC to climate change is expected to be rateregulated, but the equilibrium response is dominated by changes in NPP (Davidson and Janssens, 2006) and others related to soil capacity to retain SOC (Rasmussen et al., 2006; Van Groenigen et al., 2006). Experimentally, warming soil has similar effects on SOC decomposition as global warming does. Average increase in soil respiration was about 20% by elevating soil temperature by 0.3–6.0 C (average 2.4 C) in soil-warming experiments carried out worldwide, having a Q10 of about 2.1 (Rustad et al., 2001). As a general trend, the warming effect on SOC decomposition was significant at the early stages of warming, for example the SOC decomposition is temperature-sensitive at the beginning and will decline with ongoing of warming (Eliasson et al., 2005; Luo et al., 2001). This decline in the temperature sensitivity was interpreted as an acclimatization of SOC decomposition under the high temperature (Luo et al., 2001). Model analysis of the temporal variation in soil heterogenic respiration after soil warming indicated that the change in SOC decomposition over time by elevated temperature can be explained by the change of SOC components, without the need of invoking assumptions of change in temperature sensitivity or acclimation of SOC decomposers (Eliasson et al., 2005; Kirschbaum, 2004). Although some soil-warming experiments have been continued for longer than 10 years, soils may be still in a transition, considering the long turnover time of soil recalcitrant C. Differing from global warming, manipulated warming experiments are not always coupled with other changes which alter the NPP of ecosystem, depending on vegetation types, climatic zones, and warming methods. If NPP is not increased, or increased less than soil respiration by warming, and experiments are carried on for enough time, the difference in soil respiration between heated and unheated soils will be likely to decrease along with warming, and SOC in heated soil may be significantly lower than unheated soil. However, if NPP is increased at the same level as respiration by soil warming, soil respiration may stay at a higher rate than in unheated soil for long time. Results from soil-warming experiments may be equivocal, and may only have limited use in predicting SOC response to long-term global warming. In forest soils, the C pool with a greatest turnover rate is the forest floor where fresh organic materials are stocked and the most pronounced temperature change is expected compared with the underlying soil. The forest

12

Pete Smith et al.

floor is expected to respond to climate change differently from the mineral soil. There might be a great increase in C release from the forest floor during the early stages of climate change in northern forest soils because of the large amount of OM in the layer and the possibly different Q10s of decomposition of soil OM in the forest floor and mineral soil at various stages. If a soil is a C source, it takes a long time for SOC to reach a new equilibrium, especially in cold regions (Trumbore, 1997) because of the slow decomposition rate of SOC. On the other hand, if a soil is a C sink with respect to climate change, it may reach a new SOC equilibrium more rapidly because NPP-related carbon inputs to the soil respond quickly to climate change.

3. Temperature Sensitivity of SOC Pools 3.1. Summary of the debate on temperature sensitivity The temperature sensitivity (or dependence) of litter and SOC decomposition is important because it determines how much the feedback from the expected warmer climate may be depending on atmospheric CO2 concentration. Despite the fact that various studies have been reported to investigate the temperature sensitivity of SOC decomposition and its relationship with environment, there is still no consensus on an agreed temperature sensitivity, largely due to omission of the consideration of important confounding and feedback effects (Davidson and Janssens, 2006; Kirschbaum, 2006). The temperature sensitivity of SOC decomposition remains a topic for debate (Briones et al., 2007; Czimczik and Trumbore, 2007; Davidson and Janssens, 2006; Pare et al., 2006), with some studies suggesting that recalcitrant C is not sensitive to temperature variation (Giardina and Ryan, 2000), others suggesting that nonlabile OM is more sensitive to temperature than labile pools (Fierer et al., 2005, 2006; Knorr et al., 2005), or that recalcitrant and labile pools have a similar temperature sensitivity (Conen et al., 2006; Fang et al., 2005). The term of temperature sensitivity of SOC decomposition is a prime case for confusion (Kirschbaum, 2006). By definition, the temperature sensitivity of SOM decomposition (often referred to as Q10, the factor by which the respiration rate differs for a temperature interval of 10 C) is the change in the decomposition rate of SOM to variable temperature under otherwise constant conditions (Fang and Moncrieff, 2001). Other terms, for example the short-term temperature sensitivity (Kirschbaum, 2006) and the actual temperature sensitivity, have been used on a similar basis of assumption. On the other hand, the term of apparent temperature sensitivity defines the relationship of field measured soil respiration or SOC decomposition rate against the seasonal temperature or the mean annual temperature change (Reichstein et al., 2005). Another term of the sensitivity of SOC to long-term climate change has also been used by different researchers,


13

referring either the response of SOC stock or its turnover rate/time to climate change (Giardina and Ryan, 2000) or against the gradient of MAT (Raich and Schlesinger, 1992). All of these terms referred to different phenomena and may have different values, for example Q10 value, but have been often used interchangeably. Current debate on the temperature sensitivity is at least partly due to the interchanged use or comparison of these terms. The temperature dependence of SOC decomposition is the result of a number of processes that effectively contributes to the rate of mineralization (A˚gren and Wetterstedt, 2007). Three different mechanisms in the environment interact to determine the temperature response: (1) the rate at which decomposers take up substrate at their surface, (2) the rate by which substrate diffuses up to the surface of the decomposer, and (3) the rate at which substrate is made available. The mechanisms are characterized by activation energies for each process and the temperature response is mainly determined by one or two of the mechanisms (A˚gren and Wetterstedt, 2007; Davidson and Janssens, 2006). Giardina and Ryan (2000) compiled the stable isotopic C data from both field observations and laboratory incubations and concluded that the turnover of old SOC or the SOC in mineral soil is insensitive to temperature, suggesting that the decomposition of labile C is more sensitive than recalcitrant C to temperature variation. Giardina and Ryan may have inadequately interpreted the field data of stable isotopic C by neglecting the fact that a small fraction of labile C contributes an important part to the turnover of the total SOC (Davidson and Janssens, 2006). Their interpretation of field data is correct only when the fluxes of C input to the soil are similar across sites along the gradient of MAT. Giardina and Ryan’s interpretation on incubation data may be also inappropriate due to the difference of substrate supply and microbes under different temperatures in a long-term incubation (Fang et al., 2005). In a recent incubation study of eight european forest soils, Rey and Jarvis (2006) found that the temperature sensitivities of the rate constant of the labile fractions tended to be higher than those for the rate constant of the recalcitrant fractions, but the difference was not statistically significant. On the basis of a theoretical consideration of SOC components and the dynamics of its decomposition, the recalcitrant C was suggested have a ˚ gren and high activation energy, that is high-temperature sensitivity (A Wetterstedt, 2007; Davidson and Janssens, 2006). This was supported by some recent experimental studies (Fierer et al., 2005, 2006; Leifeld and Fuhrer, 2005). Despite that, analytically partitioning of SOC into C pools of different recalcitrance to decomposition is still impossible, the overall resistance of SOC to decomposition can be approximated qualitatively under few circumstances, such as the resistance increases along with the progress of soil incubations; a low production of CO2 is often associated with a high resistance of SOC under similar environment conditions; and the SOC in a region of high MAT is commonly more resistant that that of low MAT.

14

Pete Smith et al.

In above experimental studies, the overall temperature sensitivity of SOC decomposition has been observed to increase along with incubation time and associated with adding know resistant C compounds (Fierer et al., 2005), negatively related to CO2 production (Leifeld and Fuhrer, 2005) and to MAT across sites (Fierer et al., 2006), suggesting the higher temperature sensitivity of recalcitrant C pools. It has been suggested that the conclusion by Knorr et al. (2005) from a modeling analysis, that the resistant C is more sensitive to temperature change, was largely due to a assumptions in the model that both recalcitrant and labile pools have a same decomposition rate under the reference temperature (Fang et al., 2006). On the other hand, the temperature sensitivity of resistant C pool was reported not significantly different from that of labile pool in both laboratory incubation and field investigation (Conen et al., 2006; Czimczik and Trumbore, 2007; Fang et al., 2005). By comparing the stable isotopic signature of SOC and respired CO2 a few decades after a shift between C3 and C4 plants, Conen et al. (2006) reported that the differences in temperature sensitivity between young and old carbon are negligible under field conditions and that feedbacks of the carbon cycle on climate change are driven equally by young and old SOC.

3.2. Apparent versus actual temperature sensitivity To determine the actual temperature sensitivity of SOC decomposition or soil respiration requires a constant environment condition, except temperature. Such a requirement can only be approximated in laboratory incubation experiments. Although the soil incubation was thought to provide the best and least biased estimate of the temperature dependence of SOC decomposition (Kirschbaum, 2004, 2006), different methods of incubation have brought different uncertainties in estimated temperature sensitivity. In most incubation experiments, soil samples were incubated separately at differently constant temperatures. The rate of respired CO2, remaining mass of SOC, or the time of a same percentage of C loss at different temperatures were used to estimate the response of SOC decomposition to temperature change. However, the temperature sensitivity estimated with all of these methods could have been biased (Fang et al., 2005). For field measurements, fitting soil respiration rate against the diurnal variation of temperature may approximate the actual temperature sensitivity of soil respiration. However, this estimate of temperature sensitivity may be biased by variations in photosynthesis (Larsen et al., 2007). In published field experiments, Q10s were commonly estimated by relating observed soil respiration rate to temperature over a period of 1 year or a whole growing season. In this case, Q10 included not only the response of soil respiration to temperature but also the compound effect of changes in


15

root and microbial biomass, available substrate for respiration as well as soil moisture content during the period. Multiple confounding factors influence soil respiration in the field that often hampers a correct separation and interpretation of the different environmental effects on respiration (Reichstein et al., 2005). From the aspect of substrate supply, the environmental constraints that can temporarily or indefinitely affect apparent temperature sensitivities of decomposition may include physical protection, chemical protection, drought, flooding, and freezing. The affinity of enzyme to substrate will affect the temperature sensitivity (Davidson and Janssens, 2006). Factors influencing the rate constant of soil biological processes also affect the apparent temperature sensitivity. Despite that there is no systematic impact of soil moisture on the apparent temperature sensitivity of SOC decomposition for the full moisture regime, it is clear that, in dry soil, increasing soil moisture will increase the temperature sensitivity of SOC decomposition (Gaumont-Guay et al., 2006; McCulley et al., 2007; Rey et al., 2005). Soil temperature affects significantly both SOC decomposition and its temperature sensitivity. As a general trend, a higher temperature is always associated with a higher respiration rate and a lower Q10 value (Niklin´ska and Klimek, 2007). However, the heterogeneity of soil temperature within the soil profile (Reichstein et al., 2005) may bias the apparent temperature sensitivity if it is estimated based on measurement on the soil surface. The apparent temperature sensitivity of SOC decomposition in the field may be further complicated by the ‘‘rhizosphere priming effect,’’ in which the presence of live roots may accelerate or suppress the decomposition of SOC. SOC mineralization in the rhizosphere of cottonwood was observed unresponsive to seasonal temperature changes due to the strength of the rhizosphere priming effect (Bader and Cheng, 2007). In alpine and subalpine soils, the magnitude and seasonal pattern of temperature sensitivity of respiration were observed to be closely related to the characteristics of microbial communities (Lipson, 2007; Monson et al., 2006; Trasar-Cepeda et al., 2007). Temperature sensitivities derived from different studies and their affecting factors can be schematized as in Fig. 4. At global scale, the long-term relationship between SOM and temperature is much less significant than the short-term or the actual temperature sensitivity of SOM decomposition in terms of Q10 value, clearly suggesting that some unknown processes/ mechanisms are important in affecting the response of soil C to long-term climate change. If we use the top-down relationship between the turnover rate/soil C stock and MAT for simulation with current bottom-up models, the response of SOM decomposition to global warming will be underestimated.

16

Pete Smith et al.

Others Soil C stocks + modeling

Soil NPP

Radioactive C across sites

Q10 against shortterm temperature

Resistant C Radioactive C at a single site

Plant growth Soil warming

SOM

Mass loss in incubation Field litter bag

Microbes

Q10 against MAT

Field respiration

Global data of soil respiration

Moisture Respiration in incubation

Temperature

Labile C

1

10

102

103

Year

Figure 4 Schematic diagram of temperature sensitivities derived from various methods and their affecting factors. Methods are positioned by involved timescale and possible affecting factors. Dashed ellipses are methods for estimating SOM turnover rate only. Solid ellipses are methods for determining the turnover rate of SOC or respiration rate as well as estimating Q10 value. The horizontal span of each ellipse indicates the temporal scale of the method. The upper boundary of ellipses indicates the possible maximum factors or the critical factor affecting the resultant turnover rate or Q10 value. The bottom of ellipses suggests measured processes be dominated by different SOM pools. SOM is divided into two pools in terms of quality, with the dark grey showing resistant, and white showing the labile pool. The light grey region between indicates soil processes being affected by both resistant and labile pools.

3.3. Q10 and beyond Globally, Q10 values of soil respiration obtained from field observation and laboratory soil incubation are similar, with global means at around 2.3 (Lenton and Huntingford, 2003). Despite that soil incubation was thought to provide the best estimate of temperature dependence of soil OM decomposition (Kirschbaum, 2004, 2006) and that field estimated Q10 may be biased by confounding factors (Reichstein et al., 2005), it is interesting that there seems to be no significant difference between the global means of field and laboratory estimated Q10s. A possible explanation is that, under field conditions, different factors may offset each other’s influence on the temperature dependence of soil respiration. However, the explanation may be not applicable at site scale. By fitting a mechanistic decomposition model to a global dataset of SOC, optimizing the model’s temperature and moisture dependencies to


17

best match the observed global distribution of SOC, Ise and Moorcroft (2006) fund a much smaller Q10 values (about 1.3) at global scales than that measured from short-term experiments both in the laboratory and in the field (Lenton and Huntingford, 2003) and that used in many terrestrial ecosystem models that directly apply temperature sensitivity from smallscale studies. Raich et al. (2002) estimated the Q10 of soil respiration rate measured across-site in short period against MAT at around 1.5, which was close to that by Ise and Moorcroft (2006). It is noteworthy that Q10 values of both Ise and Moorcroft and Raich et al. were somewhat related to the longterm response of soil respiration or SOC decomposition to temperature change. At present, the knowledge to link the so-called short-term temperature dependence of SOC decomposition or soil respiration to a long-term one is still lacking. We do not know whether there is a temperature dependence of SOC decomposition on the long-term climate change, as presently termed the long-term temperature dependence may merely be a short-term Q10 biased by confounding factors. Because of the limitation of methodology, it is still impossible to reliably estimate the long-term Q10 of SOC decomposition to global warming. If there is a long-term Q10 and it is significantly smaller than the short-term one, the magnitude of the soil decomposition feedback on the rate of global climate change will be less sensitive to increases in temperature as presently predicted. Regarding the temperature dependence and the resistance of SOC pools to decomposition, published studies have not provided unequivocal evidence. While some studies suggest that the recalcitrant C pool may be more sensitive to temperature change (Fierer et al., 2005, 2006; Leifeld and Fuhrer, 2005), several other studies reported that no significant difference of temperature dependence between resistant and labile C pools (Conen et al., 2006; Czimczik and Trumbore, 2007; Fang et al., 2005; Rey and Jarvis, 2006) in both field observations and laboratory incubation experiments. The soils under coniferous trees were commonly thought to have a lower quality of SOC. However, coniferous soils are not more resistant than deciduous forests to increasing their specific rates of soil heterotrophic respiration on warming (Pare et al., 2006). A difficulty in estimating the relationship between Q10 of decomposition and SOC component is how to partition soil heterotrophic respiration into different SOC components. As the concepts of recalcitrant and labile C pool are biofunctional ones, no analytical method can be used to separate these pools from each other. Recalcitrant and labile C pools can only be inferred from experimental data by fitting some kind of conceptual models. Current confusion on the temperature dependence of SOC decomposition can be partly due to uncertainties in SOC pool partitioning. Using the change in stable isotope composition in transitional systems from C3 to C4 vegetation, Conen et al. (2006) distinguished the temperature sensitivity of carbon differing several decades in age. This approach allowed the authors to

18

Pete Smith et al.

identify C release from old carbon, avoiding unnatural conditions of longterm incubations and debatable curve-fitting. However, this approach can only be carried out when strict site conditions are met. Until new methods are developed and applied to partition SOC pools and to identify the shortand long-term response of SOC, the debate on how the SOC decomposition will respond to warming will continue. It may be correct that there are already enough large number of observations in the global dataset and simply adding more individual studies similar as present ones will not likely to lead to great new insight (Kirschbaum, 2006). In the absence of a consensus on the temperature sensitivity of decomposition of a large fraction of soil C stocks, the significance of a positive feedback between climate change and SOC decomposition continues to be debated (Davidson and Janssens, 2006). Only after we have understood these mechanisms/processes, linked them with our current understanding from short-term studies, and properly integrated these mechanisms into models, can we reliably predict how soil C stocks will change under global warming.

4. Methods for Measuring Soil Responses to Global Warming Given the relative size of the soil carbon stock and its likely significance to global carbon balance under different warming scenarios, it is interesting to reflect that the measurement of soil respiration remains difficult and that uncertainties associated with these measurements must feed inevitably into our uncertainty of global carbon balance. Although measurements of soil respiration date back nearly a century or so, it is recognized the difficulty lies in the nature of the complex soil medium itself, being composed of a range of organomineral particles and aggregates and that contains numerous organisms with differing physiological characteristics. Not only that but soil properties vary temporally and spatially, both in the horizontal and in the vertical and there are several biogenic sources of CO2 efflux. This extreme variability explains why even current methods of measuring soil respiration are not settled on any particular methodology, although we are getting better at establishing the uncertainties associated with the methods. Several recent reviews have discussed these issues in detail (e.g., Kuzyakov, 2006; Ryan and Law, 2005; Subke et al., 2006).

4.1. Soil respiration measurements in the laboratory Most of the earliest determinations of soil CO2 efflux were made under laboratory conditions and they involved the measurement of CO2 released from, or oxygen uptake by, known quantities of soil samples or undisturbed


19

soil cores incubated under controlled conditions. The simplest procedure for these measurements is that a known quantity of soil sample and the alkaline absorbent held by an open container are enclosed in a container, and soil respiration rate was obtained by the amount of CO2 trapped in the alkali during a given period. One of the earliest laboratory studies was by Neller (1922) who estimated the CO2 evolved from a soil core contained in a flask as it diffused into a bell jar and then was drawn into a bottle where CO2 was absorbed by an alkali solution. CO2 production was determined by titrating the alkali solution at the end of absorption. Romell (1932) summarized the early studies on CO2 exchange and in principle, such methods differ little from present day practice (Nakayama, 1990), although methods of determining CO2 concentration may use analysis by infrared gas analyzers or gas chromatographs. The big advantage of experimentation in the laboratory is of course the ability to control and manipulate the environmental drivers of soil CO2 efflux and many determinations can be made to help understand the physical and biological basis of the exchange processes. Incubation of samples at a range of temperatures can be used to cover a range of possible soil-warming scenarios quite efficiently (Amador and Jones, 1995; Nicolardot et al., 1994; Oberbauer et al., 1992). The main drawback of laboratory techniques is that the soil conditions are modified in comparison to the field situation. Gas diffusion may be changed in the soil sample as a result of crushing the natural structure of the soil; oxygen concentration may differ from the original environment in the field and the disturbance caused by sampling, sieving, and distributing the soil causes a flush of OM mineralization (Blet-Charaudeau et al., 1990).

4.2. Soil respiration measurements in the field Field measurements of soil CO2 efflux are usually made by enclosing a known area of soil, cleared of green vegetation, and isolating this patch from the atmosphere by chambers of differing designs. The CO2 evolved from the soil is measured quantitatively by one of three different methods known as static absorption, enrichment, and dynamic (Singh and Gupta, 1977), the latter two methods also being known as the non-steady-state and steady-state methods, respectively (Livingston and Hutchinson, 1995). 4.2.1. Static absorption This is one of the earliest methods of measuring soil CO2 efflux. The method uses an alkali solution (KOH or NaOH) contained in a dish and placed within an isolating chamber for (usually) several hours to days in period. The CO2 efflux is a function of the gain in CO2 trapped in the absorbent, the exposing time, and the area underneath the chamber. Various designs of chamber have been employed for surface CO2 measurements (e.g., PVC pipe of 152 mm i.d. by Cowling and Maclean, 1981) and

20

Pete Smith et al.

the method has been extended to measure profiles of CO2 efflux at different depths in the soil (Campbell and Frascarelli, 1981). The use of a granular absorbent (Edwards, 1982) made the technique more convenient and well-suited to be deployed under natural communities where spatial heterogeneity in soil efflux may be high and a large number of samples are required to capture this natural variability (Yim et al., 2002). Not surprisingly, the method has been examined critically to determine the absolute accuracy of the methodology. In general, it is recognized that the precision of the static absorption method depends on a number of factors including the physical dimensions of the chamber [Monteith et al. (1964) suggested the chamber cover an area at least 600 cm2 to avoid edge effects; Edwards (1982) recommended a 5% minimum ratio of the surface area of absorbent to the area enclosed by the chamber] and the concentration and volume of the alkali (Gupta and Singh, 1977). It is also likely that because the absorption of CO2 by alkali is temperature-dependent, then this is a further correction that needs to be applied to the technique (Edwards and Sollins, 1973). In general, it has been found that the absorption method underestimates the rate of soil CO2 efflux as measured by one of the other chamber methods, and it is best although of as a means of obtaining a relative rate of soil respiration rather than as a measure of the absolute rates of exchange (Ewel et al., 1987). 4.2.2. Dynamic (or steady-state) chambers In a dynamic chamber method, air is continually circulated through a chamber and the gas analyzer in a closed loop; the efflux of CO2 from the soil (Sr) covered by the chamber is obtained as a function of the difference in CO2 concentration between air entering and leaving the chamber:

Sr ¼ Dc

f A

ð1Þ

where Dc is the difference in CO2 mass fraction in the incoming and outgoing air streams, f is the gas flow rate through the chamber, and A is the surface area covered by the chamber (Nakayama, 1990). The difference in CO2 concentration is usually measured by infrared gas analysis (IRGA). Much has been written about ensuring that pressure differences between inside and outside the chamber are eliminated; if air is blown into the chamber, an overpressure within the chamber will be established and the natural efflux of CO2 from the soil will be diminished; conversely, drawing air out of the chamber will induce relative negative pressure in the chamber and an increase in the soil efflux rate. Kanemasu et al. (1974) showed that the measured CO2 efflux was about an order of magnitude larger when air was drawn out of a chamber (DP ¼ 2.5 pa) compared to when air was blown


21

in (DP ¼ þ1 pa). Fang and Moncrieff (1996) suggested that it was necessary to keep the pressure difference to within 0.2 pa with the dynamic chamber method for reliable measurements to be made but this is recognized to be practically quite difficult to achieve. A number of designs have been published recently that seek to overcome some of the difficulties in the dynamic method. 4.2.3. Enrichment (or non-steady-state) methods If a closed chamber is placed on the soil, the concentration of CO2 respired from the soil will build up inside the chamber and this enrichment can be used to estimate the efflux from the soil. This method is the basis of many of the successful commercial designs on the market today. The soil efflux can be expressed by

Dc V Sr ¼ Dt A

ð2Þ

where Dc is the CO2 concentration increment in the chamber in the time interval Dt, V is the volume of air within the chamber, and A is the soil surface area covered by the chamber. The CO2 content of a sample taken at discreet intervals can be measured by alkali absorption (Raich et al., 1990), by gas cromatography (GC) (Castro et al., 1994, Crill, 1991), by IRGA (Parkinson, 1981), or by mass spectrometer (Clymo and Pearce, 1995). From equation, it can be seen that chamber dimensions must be known accurately. In addition, as samples are withdrawn from the chamber, it is important to replace the sampled air by an equal volume of air so as to avoid any artifacts of pressure imbalance with in the chamber as an underpressure will draw more gas from the soil to compensate. The latter point has been addressed by the major manufacturers of such systems based on the designs by Parkinson (1981) and Norman et al. (1992). The paper by Norman et al. also recommended that a narrow piece of vent tube be located on one of the walls of the closed chamber to minimize the pressure differential between inside and outside; if the tube is narrow, the diffusive pathway is small and no CO2 leakage from the chamber should occur. This conclusion is supported by the more recent review by Davidson et al. (2002), who also conclude that when pressure differentials are kept small (say 0.1 pa), then errors in flux estimates are around 15%. They further conclude that for typical chambers of 10–20 cm height, most non-steady-state methods on typical soils will underestimate fluxes by about 15%. The use of a vent to equalize pressure inside and outside a closed chamber is not without controversy; however, as Conen and Smith (1998) argued that wind blowing over the vent induced a Venturi effect, which actually caused air to come into the chamber from the soil thus increasing the efflux. The effect of

22

Pete Smith et al.

changing wind speed outside, even a vented chamber on soil efflux inside, the chamber remains difficult to predict and it is argued that any new design of chamber needs to be tested thoroughly for such possible sources of error. A new design of vent tube described by Xu et al. (2006) suggests that it is possible to eliminate the impact of fluctuating wind speed outside the chamber on soil efflux measurements (a result also suggested by Bain et al., 2005). Closed chambers also use a fan to mix the air within the chamber so that a representative subsample can be taken and various designs of fan and mixing arrangements exist so as not to disturb the soil boundary layer (Welles et al., 2001). 4.2.4. Sampling uncertainties Assuming that the technical limitations of any chamber-based system could be identified and corrected for, one remaining limitation of such systems is their inability to be in more than one place at a time! The spatial heterogeneity of the soil and thus the efflux is the elephant in the room. Even with seemingly uniform soils, the efflux may vary by a factor of two over a few meters. Manufacturers are bringing out systems which have multiplexed chambers (up to 16 auto-opening and closing chambers in some instances) attached to a central analysis unit and this undoubtedly helps. But just how many chambers are required to estimate the mean flux and its standard error? Figure 5 shows the number of chamber measurements that would be required to meet a certain measurement of uncertainty as a function of the nonuniformity of the measurement site (expressed here as a coefficient of variation where a value of 10% would represent the kind of soil typically used in agronomic research plots). For such a site, about 20 samples would be required to get a measurement uncertainty of 5%. Conen and Smith (2000) and Rayment (2000) argued that the variation in air-filled porosity of soils was a crucial determinant of the accuracy of closed chambers because it alters the effective chamber volume and the more coarse the soil, the greater the underestimate in the efflux. This conclusion is supported by experiments using a laboratory setup with a constant CO2 generator situated below chambers of different design sitting on artificial soil media with varying air-filled porosities (Butnor et al., 2005; Widen and Lindroth, 2003).

4.3. Manipulation methods in the field Autotrophic and heterotrophic respiration is likely to respond differently to environmental drivers and thus it is desirable to be able at times to separate out the two sources of respired soil CO2 (Subke et al., 2006). Different designs of manipulation experiment have been designed to tease apart the total efflux signal; some of the methods are destructive, others treat the soil in situ (Kuzyakov and Larionova, 2005).

23


Number of chambers required as a function of surface heterogeneity and for an acceptable error 1000 5% 10% 20% 50% 100% 100

10

1 0

5

10

15

20

25

30

Measurement uncertainty (%)

Figure 5 The number of chamber measurements that would be required to meet a certain measurement of uncertainty as a function of the nonuniformity of the measurement site (expressed as a coefficient of variation where a value of 10% would represent the kind of soil typically used in agronomic research plots).

4.3.1. Component integration This method involves separating the roots from the soil and then measuring respiration separately from each (accurately weighed) component. The basic assumption is that the weighted sum of the respiration rates measured after this separation is the same as if the soil was intact (Hanson et al., 2000). 4.3.2. Trenching This method effectively excludes roots from the soil under the soil chamber after a trench is dug around the perimeter of the chamber (Edwards and Norby, 1998; Hanson et al., 2000). This method has the advantage that there is less disturbance to trees outside the trenched plots and weather conditions are unaltered; however, there remain questions about how to correct for dead roots that remain within the trenched plot (leading to an overestimation of heterotrophic respiration) and how to allow for the reduction in evapotranspiration, soil moisture, and hence microbial biomass over extended periods (Ngao et al., 2007).

24

Pete Smith et al.

4.3.3. Girdling Removing the tree cambium effectively cuts off the supply of photosynthates to roots and the effect can quickly appear as a rapid decrease in soil efflux (Ho¨gberg et al., 2001; Subke et al., 2004). The beauty of these studies is that they reveal the timescale for coupling between input and output of C in ecosystems (Baldocchi et al., 2006; Stoy et al., 2007). 4.3.4. Chamber design Heinemeyer et al. (2007) used a multiplexed soil chamber system to estimate the fraction of the total CO2 efflux coming from autotrophic or heterotrophic respiration. They used a novel mesh collar design to isolate the three main soil CO2 efflux components: root, extraradical mycorrhizal hyphal, and soil heterotrophic respiration. In a pine forest in the United Kingdom, they found that the mycorrhizas were responsible for about 25% of the total soil CO2 respired, roots contributed about 15%, and soil heterotrophs contributed the remainder. This experiment also showed that the depth to which soil collars are inserted has a significant bearing on the measured soil efflux given that most soil collars are inserted about 5 cm into the soil, cutting any roots and micorrhizal hyphae in this layer and leading to underestimates in the fluxes (Raich and Nadelhoffer, 1989). This is likely to be true of many previously published measurements of soil efflux. 4.3.5. Isotope methods The 13C or 14C respired in soil CO2 can be used to distinguish between autotrophs and heterotrophs in soil given that they discriminate differently against isotopes of carbon and that the source of respired CO2 reflects its origins, for example photosynthates in the case of roots. Methods using isotope discrimination can be based on measuring the natural abundance of these isotopes in soil or by injecting known quantities of the isotopes into the system and monitoring its evolution some time later. 4.3.5.1. Natural abundance Plants that are C3 photosynthesiers have a 13C isotopic signature of about 26%; C4 plants have about 12%. By growing C4 plants in a soil that previously grew only C3 plants, for example, Rochette et al. (1999) showed that root respiration contributed about 43% of total soil respiration in a maize crop at the height of the growing season. 4.3.5.2. Labeling For C3 plants growing on C3-based OM in the soil, the d13C difference may be too small to use the natural abundance technique and an alternative is to label the plant with isotopically distinct air that can be traced in time (Meharg, 1994). Isotopes of carbon can be applied to plants either in a single one-off pulse or in a continuously throughout the lifetime of the plant as in FACE experiments (Paterson et al., 1997). The

25


isotopic signature of the air going into the plant will be the same as that coming from the roots and will be different from that coming from the soil heterotrophs (Ekblad and Hogberg, 2000).

4.4. Eddy covariance for total net ecosystem exchange Soil chambers are the most direct method of measuring soil efflux but suffer from limited spatial representativeness and in some configurations, cannot be left for extended periods on the ground because they may alter the soil itself by chamber effects. The micrometeorological method of eddy covariance can measure the net exchange of CO2 over areas typically of several hundred square meters and the instrumentation is sufficiently robust and low-powered that it can be used for extended periods of seasons to years (Baldocchi, 2003). The technique involves measuring the rapid fluctuations in vertical windspeed with a sonic anemometer and correlating them with simultaneous measurements of CO2 concentration as measured by a fastresponse infrared gas analyzer (Moncrieff et al., 1997). Figure 6 shows that the scale of observation for eddy covariance fluxes is very different from those of the soil chambers discussed previously. Eddy fluxes measure Ecosystem “lid”

Eddy fluxes Photosynthesis

Autorophic respiration

Leaf chamber Leaves

Stems

Litter

Translocation

Litterfall Litter traps

Heterotrophic respiration

Soil chamber

Roots

Decomposition

Soil biota

Soil organic matter

Figure 6 Eddy covariance sensors sit at a level in the atmosphere where they measure the net exchange of gross primary photosynthesis and heterotrophic and autotrophic respiration. The dashed lines in this figure show the carbon fluxes once they reach a hypothetical lid over the canopy and measured by eddy covariance. The triple lined boxes indicate different measurement methods at different scales of observation.

26

Pete Smith et al.

NEP only; to tease apart net ecosystem exchange (NEE) into GPP or autotrophic or heterotrophic respiration, ancillary measurements are required, including those from soil chambers. The uncertainties associated with eddy covariance are well known and it has proved itself to be a most valuable method for obtaining NEE over regional scales. In the context of this discussion, eddy covariance measurements of CO2 flux at night ought to be comparable to soil chamber measurements of CO2 efflux provided atmospheric turbulence is sufficient to mix the air from the ground surface up to the level where the eddy covariance sensors are. When this is true on windy nights, the two techniques together can be used to derive response surfaces relating soil efflux to soil temperature, a vital component in modeling carbon balance of ecosystems. In some studies, the agreement is less good even when data is rejected for low wind speed conditions (when nocturnal advection may be present at a site, thus removing CO2 before it can be carried past the eddy covariance sensors) and differences of up to 20–30% can be seen (e.g., Janssens et al., 2001; Lavigne et al., 1997).

5. Approaches to Modeling Soil Responses to Global Warming Models are tools that can be used to simulate the combined impact of many different factors on the target output. In this case, where SOC is the output of interest, models can be used to assess how changes in decomposition rate, C inputs, and other changes will affect SOC. Confidence in model outputs comes with the demonstration that they can replicate observed trends. In a comprehensive comparison of nine models of SOC turnover, Smith et al. (1997) tested the models against 12 datasets from long-term experiments representing different land-use, soil type, climate zone, and land management. Through such evaluation, the applicability and limitations of models can be assessed. Where models can simulate SOC trends across a range of environmental conditions, they can be applied to future scenarios to project likely changes in SOC. A number of such SOC models exist, as described below.

5.1. Soil decomposition in models There are a number of approaches to modeling SOC decomposition including: (1) process-based multicompartment models, (2) models that consider each fresh addition of plant debris as a separate cohort which decays in a continuous way, and (3) models that account for C and N transfers through various trophic levels in a soil food web. These approaches are described in more detail below.


27

5.1.1. Process-based models Most models are process-based, that is they focus on the processes mediating the movement and transformations of matter or energy and usually assume first-order rate kinetics (Paustian, 1994). Early models simulated SOC as one homogeneous compartment ( Jenny, 1941). Some years later, twocompartment models were proposed (Beek and Frissel, 1973; Jenkinson, 1977) and, as computers became more accessible, multicompartment models were developed (McGill, 1996; Molina and Smith, 1998). Of the 33 SOC models currently represented within the Global Change and Terrestrial Ecosystems and Soil Organic Matter Network database (Smith et al., 1996, 2001, 2002), 30 are multicompartment, process-based models. Each compartment or SOC pool within a model is characterized by its position in the model’s structure and its decay rate. Decay rates are usually expressed by first-order kinetics with respect to the concentration (C ) of the pool

dC ¼ kC dt

ð3Þ

where t is the time. The rate constant k of first-order kinetics is related to the time required to reduce by half the concentration of the pool when there is no input. The pool’s half-life [h ¼ (ln 2)/k] or its turnover time (t ¼ 1/k) are sometimes used instead of k to characterize a pool’s dynamics: the lower the decay rate constant, the higher the half-life, the turnover time, and the stability of the organic pool. The flows of C within most models represent a sequence of carbon going from plant and animal debris to the microbial biomass, then to soil organic pools of increasing stability. Some models also use feedback loops to account for catabolic and anabolic processes and microbial successions. The output flow from an organic pool is usually split. It is directed to a microbial biomass pool, another organic pool, and, under aerobic conditions, to CO2. This split simulates the simultaneous anabolic and catabolic activities and growth of a microbial population feeding on one substrate. Two parameters are required to quantify the split flow. They are often defined by a microbial (utilization) efficiency and stabilization (humification) factor that control the flow of decayed C to the biomass and humus pools, respectively. The sum of the efficiency and humification factors must be inferior to one to account for the release of CO2. A thorough review of the structure and underlying assumptions of different process-based SOM models is available (Molina and Smith, 1998). 5.1.2. Cohort models describing decomposition as a continuum Another approach to modeling SOC turnover is to treat each fresh addition of plant debris into the soil as a cohort (McGill, 1996). Such models consider one SOM pool that decays with a feedback loop into itself.

28

Pete Smith et al.

˚ gren, 1995), for example, is represented by a singleQ-SOIL (Bosatta and A rate equation. The SOC pool is divided into an infinite number of components, each characterized by its ‘‘quality’’ with respect to degradability as well as impact on the physiology of the decomposers. The rate equation for the model Q-SOIL represents the dynamics of each SOC component of quality q and is quality dependent. Exact solutions to the rate equations are ˚ gren, 1994). obtained analytically (Bosatta and A 5.1.3. Food-web models Another type of model simulates C and N transfers through a food web of soil organisms (Paustian, 1994; Smith et al., 1998); such models explicitly account for different trophic levels or functional groups of biota in the soil de Ruiter and Van Faassen, 1994; de Ruiter et al., 1993, 1995; Hunt et al., 1984, 1987, 1991). Some models have been developed which combine an explicit description of the soil biota with a process-based approach (McGill et al., 1981). Food-web models require a detailed knowledge of the biology of the system to be simulated and are usually parameterized for application at specific sites. 5.1.4. Temperature as a factors controlling decomposition in SOC models Rate ‘‘constants’’ (k) are constant for a given set of biotic and abiotic conditions. For nonoptimum environmental circumstances, the simplest way to modify the maximum value of k is by multiplication with a reduction factor m—ranging from 0 to 1. Environmental factors considered by SOC models include temperature, water, pH, nitrogen, oxygen, clay content, cation exchange capacity, type of crop/plant cover, and tillage (Molina and Smith, 1998). We focus here on temperature. Many studies show the effect of temperature on microbially mediated transformations in soil, either expressed as a reduction factor or the Arrhenius equation, but the assumption that SOC decomposition was temperature-dependent was challenged by a study suggesting that old SOM in forest soils does not decompose more rapidly in soils from warmer climates than in soils from colder regions (Giardina and Ryan, 2000). Other studies, however, suggest that old SOC is not more resistant than younger pools of SOC (Grant, 1991; Sierra and Renault, 1996), and that old SOC is not less sensitive to temperature (Fang et al., 2005), with Knorr et al. (2005) suggesting that old SOC is more temperature-sensitive than young SOC. Smith (2001, 2002) reviewed 33 current models for SOC decomposition. Models represent temperature sensitivity in different ways. Burke et al. (2003) reviewed how the temperature sensitivity in different biogeochemical models was represented. They showed great variation among models in the representation of temperature sensitivity, represented by the Q10 response, with some models such as DAYCENT showing rapidly increasing


29

Q10 at lower temperatures, others such as CENTURY and RothC with a more gradual increase in Q10 at lower temperatures, and others, such as the 2003 versions of pNET, TEM, and Biome-BGC, with a fixed Q10 of 2 at all temperatures. Davidson and Janssens (2006) reviewed the subject of temperature sensitivity of SOC decomposition, and Section 3 covers this area is some detail.

5.2. Carbon input to the soil in models There are many different SOC models, with some such as RothC (Coleman and Jenkinson, 1996) and ICBM (Andreń and Ka¨tterer, 1997) simulating only SOC transformation. These models do not simulate carbon inputs to the soil. Instead, C inputs to the soil are a user input and are either estimated from measurement or from fitting to initial SOC levels (Coleman and Jenkinson, 1996). Other SOC models (e.g., CENTURY, EPIC, DNDC; McGill, 1996) simulate the whole ecosystem. These models have plant growth components with grow biomass and feed carbon (through roots, exudates, debris, etc.) to the soils components of the model. The models containing the greatest feedbacks between climate and soil C turnover are the coupled climate carbon cycle (C4) models (Friedlingstein et al., 2006). These models not only have coupling between soil C and plant C inputs, but plants and soils also respond to, and feedback to, the climate via the atmospheric CO2 concentration. C4 models tend to have a simpler representation of soils than model dedicated soils models or ecosystem models ( Jones et al., 2005). Although C4 models can be used to examine climate impacts on soil C in a fully coupled way ( Jones et al., 2005), ecosystem models can also simulate climate impacts (without climate feedback) through use of climate scenario driving data (e.g., Muller et al., 2007), and dedicated soils models can also be used, derive C inputs from other models or data sources (e.g., Smith et al., 2005, 2006, 2007d,e). Smith et al. (2005), using the RothC model, showed that changes in plant productivity in Europe would likely counterbalance increased decomposition due to global warming to 2080.

6. How Will Soil Carbon Respond the Global Warming? 6.1. New perspectives As discussed in Section 2 , the ultimate response of soil carbon to global warming remains uncertain, with different directions of response reported in different regions of the world, and many factors confounding the attribution of any observed change to climate warming. Terrestrial systems

30

Pete Smith et al.

(including the soils) are currently a sink of 2.3 1.3 Pg C year1 (Section 1.3). Some models predict that in the future climate change will cause the current land carbon sink to switch to a land carbon source (Cox et al., 2000; Lenton, 2000; White et al., 1999), whereas others predict a land carbon sink that persists (Cramer et al., 2001). Models differ considerably in their predictions of future terrestrial carbon storage (Cramer et al., 2001). The soils components of these models can be tested against historical data (Section 4) to evaluate their performance, but more recently, earth system behavior over thousands of years has been used to test model assumptions. Lenton and Huntingford (2003) constructed a simple model of vegetation and soil and tested various temperature sensitivities of, among other things, plant and soil respiration. They compared the terrestrial carbon sink observed since the Last Glacial Maximum to test which figures in the range of temperature sensitivities of soil respiration found in the literature could explain the observations, and found that the upper limit of reported sensitivities (Q10 ¼ 3.63) underpredicted observed carbon storage (Lenton and Huntingford, 2003). Such studies, using earth system observations/reconstructions over many thousands of years, allow long-term soil carbon responses to be tested, facilitating greater confidence in predictions of future responses to global warming. Our tools for studying the response of soil carbon pools to climate change have improved greatly over recent years. Models have become more realistic and are being tested thoroughly to improve confidence in their predictions (Cramer et al., 2001; Smith et al., 1997). Further, model pools can now be matched to measurable fractions (Zimmermann et al., 2007), providing new opportunities for model initialization and for testing the performance of models, component by component. Incubation techniques for studying temperature sensitivity have greatly improved (Fang et al., 2005) and field techniques for studying (e.g., girdling, trenching) and measuring (e.g., respiration, eddy covariance, isotopic techniques) fluxes of carbon have greatly improved. Further, projections of climate change have improved (IPCC WGI, 2007), providing better datasets with which to examine climate change impacts. One of the most significant advances in recent years is the full coupling of the biospheric carbon cycle into climate models that previously relied only on the physics of the climate system. Through these models, feedbacks between global warming and soil carbon responses to warming can be studied.

6.2. Sensitive/vulnerable regions and soils Because the net response of soil C carbon to global warming will depend on the balance of increased C inputs to the soil, and increased C losses from the soil due to increased decomposition (Section 2), and there are likely to be regional difference in warming, soil types and vegetation types, certain


31

regions/ecosystems are likely to be more vulnerable than others. High latitude regions are thought to be particularly vulnerable for a number of reasons. First, high latitudes are projected to experience some of the greatest warming (Mitchell et al., 2004). Second, at cooler temperatures, especially for those at around freezing point, a change in temperature has a much greater impact than an increase in temperate regions. This is particularly true of permafrost soils in the taiga and tundra that hold around 500 Pg C, and could lose this carbon rapidly under warming (Zimov et al., 2006). The release of this huge stock of carbon to the atmosphere could significantly enhance further climate warming (Section 2; Zimov et al., 2006). Peatlands and wetlands are also expected to be particularly sensitive to climate warming. Peatlands also hold vast stores of carbon, with estimates ranging from approximately 200 Pg C (Post et al., 1982) to 860 Pg C (Bohn, 1976) with most estimates in the region of around 500 Pg C (e.g., Gorham, 1991; Houghton et al., 1985). Plant productivity in peatlands is relatively low compared to the large carbon losses that can occur when peatlands are drained and/or cultivated, with yearly losses of C of 0.8–8.3 t C ha1 year1 (Lohila et al., 2004; Maljanen et al., 2001, 2004; Nyka¨nen et al., 1995), and losses occurring for over a hundred years (Lohila et al., 2004). If increasing temperature increases evapotranspiration (either coupled with or independently from changes in precipitation), peatlands would be expected to dry as well as for decomposition rate to increase. Given the vast stocks of carbon in peatland soils, losses of carbon will likely far exceed any increase in plant productivity arising from increases in temperature. Other sensitive regions include currently hot arid areas. Although the soils in these areas tend to be low in carbon (Batjes, 1996), complete failure of plant growth in a warmer world would switch off all inputs of carbon to the soil, enhancing desertification and degradation. Even partial loss of vegetation integrity could make soils more vulnerable to degradation through other agents such as grazing and cultivation. Because the soil C stocks are not as high in these regions, total potential C losses are lower, but these regions cover large areas of the arid tropics/subtropics.

6.3. Reducing the vulnerability of soil C to the impacts of global warming Carbon stocks in the soil can be increased in managed ecosystems by optimizing ‘‘best management practices.’’ There have been numerous reviews of management to increase soil carbon stocks (e.g., Lal, 2004; Lal et al., 1998; Smith, 2008; Smith et al., 2007a). Increased carbon stocks in the soil increases soil fertility, workability, water holding capacity, and reduces erosion risk. Increasing soil carbon stocks can thus reduce the vulnerability of managed soils to future global warming (Smith, 2008). Management practices effective in increasing SOC stocks include improved plant

32

Pete Smith et al.

productivity (through nutrient management, rotations, improved agronomy), reduced/conservation tillage and residue management, more effective use of organic amendments, land-use change (crops to grass/trees), set-aside, agroforestry, optimal livestock densities, and legumes/improved species mix (Smith, 2008). While these measures have the technical potential to increase SOC stocks by about 1–1.3 Pg year1 (Lal, 2004; Smith et al., 2007a), the economic potentials for SOC sequestration estimated by Smith et al. (2007a) were 0.4, 0.6, and 0.7 Pg C yea1 at carbon prices of 0–20, 0–50, and 0–100 US$ t CO2-equivalent1, respectively. A small loss of C from permafrost or peatlands could offset this potential sequestration, but the increase in SOC engendered by improved management is expected to also reduce vulnerability of the soils to future SOC loss under global warming. As such, soil carbon sequestration can, in many respects, be regarded as a ‘‘win–win’’ and a ‘‘no regrets’’ option (Smith and Powlson, 2003; Smith and Trines, 2007; Smith et al., 2007b).

7. Conclusions Soils contain a stock of carbon that is about twice as large as that in the atmosphere and about three times that in vegetation. Small losses from this large pool could have significant impacts on future atmospheric CO2 concentrations, so the response of soils to global warming is of critical importance when assessing climate carbon cycle feedbacks. Models that have coupled climate and carbon cycles show a large divergence in the size of the predicted biospheric feedback to the atmosphere. Central questions which still remain when attempting to reduce this uncertainty in the response of soils to global warming are (1) the temperature sensitivity of soil OM, especially the more recalcitrant pools; (2) the balance between increased carbon inputs to the soil from increased production and increased losses due to increased rates of decomposition; and (3) interactions between global warming and other aspects of global change including other climatic effects (e.g., changes in water balance), changes in atmospheric composition (e.g., increasing atmospheric CO2 concentration), and land-use change. Our tools for studying the response of soil carbon pools to climate change have improved greatly over recent years. Models have improved and have become more realistic and laboratory and field techniques have given us new tool with which to study the response of soil carbon to global warming. Projections of climate change have improved the full coupling of the biospheric carbon cycle into climate models will allow the feedbacks between global warming and soil carbon responses to warming to be studied. Among the most vulnerable ecosystems are the northern peatlands and other regions either containing large carbon stocks, or within the


33

permafrost zone, and those in arid regions currently on the verge of desertification. A number of possible technologies to begin to mitigate these worst impacts are available, mainly in managed systems. These technologies, which promote soil carbon sequestration, will also help to mitigate climate change itself (by reducing atmospheric CO2 concentrations) and are cost competitive with mitigation options available in other sectors. Some warming will occur and it is important that humans adapt management practices to cope with this change, but soils also provide a great opportunity, along with a raft of other measures, to slow that rate of warming. Identifying the ‘‘win–win’’ options that deliver both adaptation and mitigation, and finding ways to implement these measures, remains one of our greatest challenges for this century.

REFERENCES ˚ gren, G. I., and Wetterstedt, J.A˚.M. (2007). What determines the temperature response of A soil organic matter decomposition? Soil. Biol. Biochem. 39, 1794–1798. Amador, J. A., and Jones, R. D. (1995). Carbon mineralization in pristine and phosphorusenriched peat soils of the Florida everglades. Soil. Sci. 159, 129–141. Andreń, O., and Ka¨tterer, T. (1997). ICBM: The introductory carbon balance model for exploration of soil carbon balances. Ecol. Appl. 7, 1226–1236. Bachelet, D., Neilson, R. P., Lenihan, J. M., and Drapek, R. J. (2001). Climate change effects on vegetation distribution and carbon budget in the United States. Ecosystems 4, 164–185. Bader, N. E., and Cheng, W. (2007). Rhizosphere priming effect of Populus fremontii obscures the temperature sensitivity of soil organic carbon respiration. Soil. Biol. Biochem. 39, 600–606. Bain, W. G., Hutyra, L., Patterson, D. C., Bright, A. V., Daube, B. C., Munger, W. J., and Wofsy, S. C. (2005). Wind-induced error in the measurement of soil respiration using closed dynamic chambers. Agric. Forest Meteorol. 131, 225–232. Baldocchi, D. D. (2003). Assessing ecosystem carbon balance: Problems and prospects of the eddy covariance technique. Global Change Biol. 9, 479–492. Baldocchi, D., Tang, J., and Xu, L. (2006). How switches and lags in biophysical regulators affect spatial-temporal variation of soil respiration in an oak-grass savanna. J. Geophys. Res. Atmos 111. G02008. doi: 02010.01029/02005JG 000063. Basile-Doelsch, I., Amundson, R., Stone, W. E. E., Masiello, C. A., Bottero, J. Y., Colin, F., Masin, F., Borschneck, D., and Meunier, J. D. (2005). Mineralogical control of organic carbon dynamics in a volcanic ash soil on La Reúnion. Eur. J. Soil Sci. 56, 689–703. Batjes, N. H. (1996). Total carbon and nitrogen in the soils of the world. Eur. J. Soil Sci. 47, 151–163. Beek, J., and Frissel, M. J. (1973). ‘‘Simulation of Nitrogen Behaviour in Soils.’’ Pudoc, Wageningen, The Netherlands. Bellamy, P. H., Loveland, P. J., Bradley, R. I., Lark, R. M., and Kirk, G. J. D. (2005). Carbon losses from all soils across England and Wales 1978–2003. Nature 437, 245–248. Blanco-Canqui, H., and Lal, R. (2004). Mechanisms of carbon sequestration in soil aggregates. CRC Crit. Rev. Plant Sci. 23, 481–504. Blet-Charaudeau, C., Muller, J., and Laudelout, H. (1990). Kinetics of carbon dioxide evolution in relation to microbial biomass and temperature. Soil Sci. Soc. Am. J. 54, 1324–1328.

34

Pete Smith et al.

Bohn, H. L. (1976). Estimate of organic-carbon in world soils. Soil Sci. Soc. Am. J. 40, 468–470. Bosatta, E., and A˚gren, G. I. (1994). Theoretical analysis of microbial biomass dynamics in soils. Soil Biol. Biochem. 26, 143–148. Bosatta, E., and A˚gren, G. I. (1995). Theoretical analyses of the interactions between inorganic nitrogen and soil organic matter. Eur. J. Soil Sci. 76, 109–114. Briones, M. J. I., Ostle, N. J., and Garnett, M. H. (2007). Invertebrates increase the sensitivity of non-labile soil carbon to climate change. Soil. Biol. Biochem. 39, 816–818. Bronick, C. J., and Lal, R. (2005). Soil structure and management: A review. Geoderma 124, 3–22. Burke, I. C., Kaye, J. P., Bird, S. P., Hall, S. A., McCulley, R. L., and Sommerville, G. L. (2003). Evaluating and testing models of terrestrial biogeochemistry: The role of temperature in controlling decomposition. In ‘‘Models in Ecosystem Science’’ (C. D. Canham, J. J. Cole, and W. K. Lauenroth, Eds.), pp. 225–253. Princeton University Press, Princeton, NJ. Butnor, J. R., Johnsen, K. H., and Maier, C. A. (2005). Soil properties differently influence estimates of soil CO2 efflux from three chamber-based measurement systems. Biogeochemistry 73, 283–301. Campbell, J. A., and Frascarelli, L. (1981). Measurement of CO2 evolved from organic soil at different depths in situ. Can. J. Soil Sci. 61, 137–144. Cao, M., Gregson, K., and Marshall, S. (1998). Global methane emission from wetlands and its sensitivity to climate change. Atmos. Environ. 32, 3293–3299. Castelle, A. J., and Galloway, J. N. (1990). Carbon dioxide dynamics in acid forest soils in Shenandoah National Park, Virginia. Soil Sci. Soc. Am. J. 54, 252–257. Castro, M. S., Peterjohn, W. T., Melillo, J. M., Steudler, P. A., Gholz, H. L., and Lewis, D. (1994). Effects of nitrogen fertilization on the fluxes of N2O, CH4 and CO2 from soils in a Florida slash pine plantation. Can. J. For. Res. 24, 9–13. Cheshire, M. V., and Chapman, S. J. (1996). Influence of the N and P status of plant material and of added N and P on the mineralization of C from C-14-labelled ryegrass in soil. Biol. Fertil. Soils 21, 166–170. Clymo, R. S., and Pearce, D. M. E. (1995). Methane and carbon dioxide production in, transport through, and efflux from a peatland. Philos. Trans. R. Soc. Lond., Ser. A 351, 249–259. Coleman, K., and Jenkinson, D. S. (1996). RothC-26.3—A model for the turnover of carbon in soil. In ‘‘Evaluation of Soil Organic Matter Models Using Existing, Long-Term Datasets’’ (D. S. Powlson, P. Smith, and J. U. Smith, Eds.), NATO ASI Series I, Vol. 38, pp. 237–246. Springer-Verlag, Heidelberg, Germany. Conen, F., and Smith, K. A. (1998). A re-examination of closed flux chamber methods for the measurement of trace gas emissions from soils to the atmosphere. Eur. J. Soil Sci. 49, 701–707. Conen, F., and Smith, K. A. (2000). An explanation of linear increases in gas concentration under closed chambers used to measure gas exchange between soil and the atmosphere. Eur. J. Soil Sci. 51, 111–117. Conen, F., Leifeld, J., Seth, B., and Alewell, C. (2006). Warming mineralises young and old soil carbon equally. Biogeosciences 3, 515–519. Cowling, J. E., and Maclean, S. F. (1981). Forest floor respiration in a black spruce taiga forest ecosystem in Alaska. Holarct. Ecology 4, 229–237. Cox, P. M., Betts, R. A., Jones, C. D., Spall, S. A., and Totterdell, I. J. (2000). Acceleration of global warming due to carbon-cycle feedbacks in a coupled climate model. Nature 408, 184–187. Cramer, W., Bondeau, A., Woodward, F. I., Prentice, I. C., Betts, R. A., Brovkin, V., Cox, P. M., Fisher, V., Foley, J. A., Friend, A. D., Kucharik, C., Lomas, M. R., et al.


35

(2001). Global response of terrestrial ecosystem structure and function to CO2 and climate change: Results from six dynamic global vegetation models. Global Change Biol. 7, 357–373. Crill, P. M. (1991). Seasonal patterns of methane uptake and carbon dioxide release by a temperate woodland soil. Global Biogeochem. Cycles 5, 319–334. Czimczik, C. I., and Trumbore, S. E. (2007). Short-term controls on the age of microbial carbon sources in boreal forest soils. J. Geophys. Res. Biogeosci 112, G03001. doi: 10.1029/2006JG000389. Davidson, E. A., and Janssens, I. A. (2006). Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature 440, 165–173. Davidson, E. A., Savage, K., Verchot, L. V., and Navarro, R. (2002). Minimizing artefacts and biases in chamber-based measurement of soil respiration. Agric. Forest Meteorol. 113, 21–37. Dawson, J. J. C., and Smith, P. (2007). Carbon losses from soil and its consequences for land management. Sci. Total Environ. doi: 10.1016/j.scitotenv.2007.03.023. de Ruiter, P. C., and Van Faassen, H. G. (1994). A comparison between an organic matter dynamics model and a food web model simulating nitrogen mineralization in agroecosystems. Eur. J. Agron. 3, 347–354. de Ruiter, P. C., Van Veen, J. A., Moore, J. C., Brussaard, L., and Hunt, H. W. (1993). Calculation of nitrogen mineralization in soil food webs. Plant Soil 157, 263–273. de Ruiter, P. C., Neutel, A.-M., and Moore, J.C (1995). Energetics and stability in belowground food webs. In ‘‘Food Webs, Integration of Patterns and Dynamics’’ (G. A. Polis and K. O. Winemiller, Eds.), pp. 201–210. Chapman and Hall, New York. Dersch, G., and Boehm, K. (1997). Beitra¨ge des Bodenschutzes zum Schutz der Atmospha¨re ¨ sterreich’’ (W. E. H. Blum, E. Klaghofer, und des Weltklimas. In ‘‘Bodenschutz in O A. Loechl, and P. Ruckenbauer, Eds.), pp. 411–432. Bundesamt und Forschungszentrum ¨ sterreich (in German). fuer Landwirtschaft, Vienna, O Duran, N., and Esposito, E. (2000). Potential applications of oxidative enzymes and phenoloxidase-like compounds in wastewater and soil treatment: A review. Appl. Catal. B-Environ. 28, 83–99. Edwards, N. T. (1982). A timesaving technique for measuring respiration rates in incubated soil samples. Soil Sci. Soc. Am. J. 846, 1114–1116. Edwards, N. T., and Norby, R. J. (1998). Below-ground respiratory responses of sugar maple and red maple saplings to atmospheric CO2 enrichment and elevated air temperature. Plant Soil 206, 85–97. Edwards, N. T., and Sollins, P. (1973). Continuous measurement of carbon dioxide evolution from partitional forest floor components. Ecology 54, 406–412. Ekblad, A., and Hogberg, P. (2000). Analysis of delta-13C of CO2 distinguishes between microbial respiration of added C4-sucrose and other soil respiration in a C3-ecosystem. Plant Soil 219, 197–209. Eliasson, P. E., McMurtrie, R. E., Pepper, D. A., Stro¨mgren, M., Linder, S., and ˚ gren, G. I. (2005). The response of heterotrophic CO2 flux to soil warming. Global A Change Biol. 11, 167–181. Ewel, K. C., Cropper, W. P., and Gholz, H. L. (1987). Soil CO2 evolution in Florida slash pine plantation. I. Changes through time. Can. J. Forest Res. 17, 325–329. Falge, E., Baldocchi, D., Tenhunen, J., Aubinet, M., Bakwin, P., Berbigier, P., Bernhofer, C., Burba, G., Clement, R., Davis, K. J., Elbers, J. A., Goldstein, A. H., et al. (2002). Seasonality of ecosystem respiration and gross primary production as derived from FLUXNET measurements. Agric. Forest Meteorol. 113, 53–74. Fang, C., and Moncrieff, J. B. (1996). An improved dynamic chamber technique for measuring CO2 efflux from the soil surface. Funct. Ecol. 10, 297–305.

36

Pete Smith et al.

Fang, C., and Moncrieff, J. B. (2001). The dependence of soil CO2 efflux on temperature. Soil. Biol. Biochem. 32, 168–178. Fang, C., Smith, P., Moncrieff, J. B., and Smith, J. U. (2005). Similar response of labile and resistant soil organic matter pools to changes in temperature. Nature 433, 57–59. Fang, C., Smith, P., and Smith, J. U. (2006). Is resistant soil organic matter more sensitive to temperature than labile organic matter? Biogeosciences 3, 65–68. Fierer, N., Craine, J. M., McLauchlan, K., and Schimel, J. P. (2005). Litter quality and the temperature sensitivity of decomposition. Ecology 86, 320–326. Fierer, N., Colman, B. P., Schimel, J. P., and Jackson, R. B. (2006). Predicting the temperature dependence of microbial respiration in soil: A continental-scale analysis. Global Biogeochem. Cycles 20, GB3026. doi: 10.1029/2005GB002644. Freeman, C., Ostle, N. J., and Kang, H. (2001a). An enzymic ‘latch’ on a global carbon store. Nature 409, 149. Freeman, C., Fenner, N., Ostle, N. J., Kang, H., Dowrick, D. J., Reynolds, B., Lock, M. A., Sleep, D., Hughes, S., and Hudson, J. (2004a). Export of dissolved organic carbon from peatlands under elevated carbon dioxide levels. Nature 430, 195–198. Freeman, C., Ostle, N. J., Fenner, N., and Kang, H. (2004b). A regulatory role for phenol oxidase during decomposition in peatlands. Soil Biol. Biochem. 36, 1663–1667. Friedlingstein, P., Cox, P., Betts, R., Bopp, L., Von Bloh, W., Brovkin, V., Cadule, P., Doney, S., Eby, M., Fung, I., Bala, G., John, J., et al. (2006). Climate-carbon cycle feedback analysis: Results from the (CMIP)-M-4 model intercomparison. J. Clim. 19, 3337–3353. Gaumont-Guay, D., Black, T. A., Griffis, T. J., Barr, A. G., Jassal, R. S., and Nesic, Z. (2006). Interpreting the dependence of soil respiration on soil temperature and water content in a boreal aspen stand. Agric. Forest Meteorol. 140, 220–235. Giardina, C. P., and Ryan, M. G. (2000). Evidence that decomposition rates of organic carbon in mineral soil do not vary with temperature. Nature 404, 858–861. Gorham, E. (1991). Northern peatlands: Role in the carbon cycle and probable responses to climatic warming. Ecol. Appl. 1, 182–195. Goulden, M. L., Wofsy, S. C., Harden, J. W., Harden, J. W., Trumbore, S. E., Crill, P. M., Gower, S. T., Fries, T., Daube, B. C., Fan, S. M., Sutton, D. J., Bazzaz, A., et al. (1998). Sensitivity of boreal forest carbon balance to soil thaw. Science 279, 214–217. Grant, R. F. (1991). A technique for estimating denitrification rates at different soil temperatures, water contents and nitrate concentrations. Soil. Sci. 152, 41–52. Guo, L. B., and Gifford, R. M. (2002). Soil carbon stocks and land use change: A meta analysis. Global Change Biol. 8, 345–360. Gupta, S. R., and Singh, J. S. (1977). Effect of alkali concentration, volume and absorption area on the measurement of soil respiration in a tropical sward. Pedobiologia 17, 233–239. Hanson, P. J., Edwards, N. T., Garten, C. T., and Andrews, J. A. (2000). Separating root and soil microbial contribution to soil respiration: A review of methods and observations. Biogeochemistry 48, 115–146. Hayes, M. H. B., and Clapp, C. E. (2001). Humic substances: Considerations of compositions, aspects of structure and environmental influences. Soil. Sci. 166, 723–737. Heidmann, T., Christensen, B. T., and Olesen, S. E. (2002). Changes in soil C and N content in different cropping systems and soil types. In ‘‘Greenhouse Gas Inventories for Agriculture in the Nordic Countries’’ (S. O. Petersen and J. E. Olesen, Eds.), Report 81, pp. 77–86. Ministry of Food, Agriculture and Fisheries, Danish Institute of Agricultural Sciences, Foulum, Denmark. Heinemeyer, A., Hartley, I. P., Evans, S. P., Carreira de La Fuentes, J. A., and Ineson, P. (2007). Forest soil CO2 flux: Uncovering the contribution and environmental responses of ectomycorrizas. Global Change Biol. 13, 1786–1797.


37

Hobbie, E. A., Weber, N. S., and Trappe, J. M. (2001). Mycorrhizal vs. saprotrophic status of fungi: The isotopic evidence. New Phytol. 150, 601–610. Ho¨gberg, P., Nordgren, A., Buchmann, N., Taylor, A. F. S., Ekblad, A., Ho¨gberg, M. N., Nyberg, G., Ottoson-Lo¨fvenius, M., and Read, D. J. (2001). Large-scale forest girdling shows that current photosynthesis drives soil respiration. Nature 411, 789. Hope, D., Billett, M. F., and Cresser, M. S. (1994). A review of the export of carbon in river water – fluxes and processes. Environ. Pollut. 84, 301–324. Houghton, R. A. (1999). The annual net flux of carbon to the atmosphere from changes in land use 1850 to 1990. Tellus 50B, 298–313. Houghton, R. A., Boone, R. D., Melillo, J. M., Palm, C. A., Woodwell, G. M., Myers, N., Moore, B., and Skole, D. L. (1985). Net flux of carbon-dioxide from tropical forests in 1980. Nature 316, 617–620. Houghton, R. A., Hackler, J. L., and Lawrence, K. T. (1999). The US carbon budget: Contributions form land-use change. Science 285, 574–578. Hunt, H. W., Coleman, D. C., Cole, C. V., Ingham, R. E., Elliott, E. T., and Woods, L. E. (1984). Simulation model of a food web with bacteria, amoebae, and nematodes in soil. In ‘‘Current Perspectives in Microbial Ecology’’ (M. J. Klug and C. A. Reddy, Eds.), pp. 346–352. American Society for Microbiology, Washington, DC. Hunt, H. W., Coleman, D. C., and Ingham, R. E. (1987). The detrital food web in a shortgrass prairie. Biol. Fertil. Soils 3, 57–68. Hunt, H. W., Trlica, M. J., and Redente, E. F. (1991). Simulation model for the effects of climate change on temperate grassland ecosystems. Ecol. Modell. 53, 205–246. IPCC SRES (2000). ‘‘Special Report on Emissions Scenarios.’’ Cambridge University Press, Cambridge, UK. IPCC WGI (2001). ‘‘Climate Change: The Scientific Basis.’’ Cambridge University Press, Cambridge, UK. IPCC WGI (2007). ‘‘Climate Change 2007: The Physical Science Basis.’’ Cambridge University Press, Cambridge, UK. Ise, T., and Moorcroft, P. R. (2006). The global-scale temperature and moisture dependencies of soil organic carbon decomposition: An analysis using a mechanistic decomposition model. Biogeochemistry 80, 17–231. Janssens, I. A., Kowalsky, A. S., and Ceulemans, R. (2001). Forest floor CO2 fluxes estimated by eddy covariance and chamber-based model. Agric. Forest Meteorol. 106, 61–69. Janssens, I. A., Freibauer, A., Ciais, P., Smith, P., Nabuurs, G.-J., Folberth, G., Schlamadinger, B., Hutjes, R. W. A., Ceulemans, R., Schulze, E.-D., Valentini, R., and Dolman, H. (2003). Europe’s terrestrial biosphere absorbs 7–12% of European anthropogenic CO2 emissions. Science 300, 1538–1542. Jenkinson, D. S. (1977). Studies on the decomposition of plant material in soil. V. J. Soil Sci. 28, 424–434. Jenkinson, D. S. (1990). The turnover of organic carbon and nitrogen in soil. Philos. Trans. R. Soc. Lond. B 329, 361–368. Jenny, H. (1941). ‘‘Factors of Soil Formation. A System of Quantitative Pedology.’’ McGraw-Hill, New York. Jones, M. B., and Donnelly, A. (2004). Carbon sequestration in temperate grassland ecosystems and the influence of management, climate and elevated CO2. New Phytol. 164, 423–439. Jones, J. B., and Mulholland, P. J. (1998). Carbon dioxide variation in a hardwood forest stream: An integrative measure of whole catchment soil respiration. Ecosystems 1, 183–196. Jones, C., McConnell, C., Coleman, K., Cox, P., Falloon, P., Jenkinson, D., and Powlson, D. (2005). Global climate change and soil carbon stocks; predictions from

38

Pete Smith et al.

two contrasting models for the turnover of organic carbon in soil. Global Change Biol. 11, 154–166. Kanemasu, E. T., Powers, W. L., and Sij, J. W. (1974). Field chamber measurements of CO2 flux from soil surface. Soil. Sci. 118, 233–237. Kirkby, K. J., Smart, S. M., Black, H. I. J., Bunce, R. G. H., Corney, P. M., and Smithers, R. J. (2005). ‘‘Long term ecological change in British woodland (1971– 2001).’’ English Nature Research Report 653. English Nature, London, UK. Kirschbaum, M. U. F. (2004). Soil respiration under prolonged soil warming: Are rate reductions caused by acclimation or substrate loss? Global Change Biol. 10, 1870–1877. Kirschbaum, M. U. F. (2006). The temperature dependence of organic-matter decomposition—still a topic of debate. Soil. Biol. Biochem. 38, 2510–2518. Knorr, W., Prentice, I. C., House, J. I., and Holland, E. A. (2005). Long-term sensitivity of soil carbon turnover to warming. Nature 433, 298–301. Kuzyakov, Y. (2006). Sources of CO2 efflux from soil and review of partitioning methods. Soil Biol. Biochem. 38, 425–448. Kuzyakov, Y., and Larionova, A. A. (2005). Root and rhizomicrobial respiration: A review of approaches to estimate respiration by autotrophic and heterotrophic organisms in soil. J. Plant Nutr. Soil Sci. 168, 503–520. Lal, R. (1999). Soil management and restoration for C sequestration to mitigate the accelerated greenhouse effect. Prog. Environ. Sci. 1, 307–326. Lal, R. (2004). Soil carbon sequestration impacts on global climate change and food security. Science 304, 1623–1627. Lal, R., Kimble, J. M., Follet, R. F., and Cole, C. V. (1998). ‘‘The Potential of US Cropland to Sequester Carbon and Mitigate the Greenhouse Effect.’’ Ann Arbor Press, Chelsea, MI. Larsen, K. S., Ibrom, A., Beier, C., Jonasson, S., and Michelsen, A. (2007). Ecosystem respiration depends strongly on photosynthesis in a temperate heath. Biogeochemistry 85, 201–213. Lavigne, M. B., Ryan, M. G., Anderson, D. E., Baldocchi, D. D., Crill, P., Fitzjarrald, D. R., Goulden, M. L., Gower, S. T., Massheder, J. M., McCaughey, J. H., Rayment, M., and Striegl, R. G. (1997). Comparing nocturnal eddy covariance measurements to estimates of ecosystem respiration made by scaling chamber measurements. J. Geophys. Res. 102(D24), 28977–28986. Leifeld, J., and Fuhrer, J. (2005). The temperature response of CO2 production from bulk soils and soil fractions is related to soil organic matter quality. Biogeochemistry 75, 433–453. Lenton, T. M. (2000). Land and ocean carbon cycle feedback effects on global warming in a simple Earth system model. Tellus 52B, 1159–1188. Lenton, T. M., and Huntingford, C. (2003). Global terrestrial carbon storage and uncertainties in its temperature sensitivity examined with a simple model. Global Change Biol. 9, 1333–1352. Liao, Q. L., Zhang, X. H., Li, Z. P., Pan, G. X., Smith, P., and Wu, X. M. (2008). Increases in cropland soil organic carbon stocks over the last two decades in China’s Jiangsu Province. Global Change Biol. (in press). Lipson, D. A. (2007). Relationships between temperature responses and bacterial community structure along seasonal and altitudinal gradients. FEMS Microbiol. Ecol. 59, 418–427. Livingston, G. P., and Hutchinson, G. L. (1995). Enclosure-based measurement of trace gas exchange: Applications and sources of error. In ‘‘Biogenic Trace Gases: Measuring Emissions From Soil and Water’’ (P. A. Matson and R. C. Harriss, Eds.), pp. 14–51. Blackwell Scientific Publications, Oxford. Lohila, A., Aurela, M., Tuovinen, J. P., and Laurila, T. (2004). Annual CO2 exchange of a peat field growing spring barley or perennial forage grass. J. Geophys. Res 109, D18116. doi:10.1029/2004JD004715.


39

Luo, Y., Wan, S., Hui, D., and Wallace, L. L. (2001). Acclimatization of soil respiration to warming in a tall grass prairie. Nature 413, 622–625. Malhi, Y., Meir, P., and Brown, S. (2002). Forests, carbon and global climate. Philos. Trans. R. Soc. Lond. A 360, 1567–1591. Maljanen, M., Martikainen, P. J., Walden, J., and Silvola, J. (2001). CO2 exchange in an organic field growing barley or grass in eastern Finland. Global Change Biol. 7, 679–692. Maljanen, M., Komulainen, V. M., Hytonen, J., Martikainen, P., and Laine, J. (2004). Carbon dioxide, nitrous oxide and methane dynamics in boreal organic agricultural soils with different soil characteristics. Soil Biol. Biochem. 36, 1801–1808. Martens, D. A. (2000). Plant residue biochemistry regulates soil carbon cycling and carbon sequestration. Soil Biol. Biochem. 32, 361–369. McCulley, R. L., Boutton, T. W., and Archer, S. R. (2007). Soil respiration in a subtropical savanna parkland: Response to water additions. Soil Sci. Soc. Am. J. 71, 820–828. McGill, W. B. (1996). Review and classification of ten soil organic matter (SOM) models. In ‘‘Evaluation of Soil Organic Matter Models Using Existing, Long-Term Datasets’’ (D. S. Powlson, P. Smith, and J. U. Smith, Eds.), NATO ASI Series I. Vol. 38, pp. 111–133. Springer-Verlag, Heidelberg, Germany. McGill, W. B., Hunt, H. W., Woodmansee, R. G., and Reuss, J. O. (1981). PHOENIX, a model of the dynamics of carbon and nitrogen in grassland soil. In ‘‘Terrestrial Nitrogen Cycles. Processes, Ecosystem Strategies and Management Impacts’’ (F. E. Clark and T. Roswall, Eds.), Vol. 33, pp. 49–115. Ecological Bulletins. Meharg, A. (1994). A critical review of labeling techniques used to quantify rhizosphere carbon-flow. Plant Soil 166, 55–62. Melillo, J. M., Steudler, P. A., Aber, J. D., Newkirk, K., Lux, H., Bowles, F. P., Catricala, C., Magill, A., Ahrens, T., and Morrisseau, S. (2002). Soil warming and carbon-cycle feedbacks to the climate system. Science 298, 2173–2176. Mitchell, T. D., Carter, T. R., Jones, P. D., Hulme, M., and New, M. (2004). ‘‘A comprehensive set of high-resolution grids of monthly climate for Europe and the globe: The observed record (1901–2000) and 16 scenarios (2001–2100).’’ Working Paper 55, Tyndall Centre for Climate Change Research, University of East Anglia, Norwich. Molina, J. A. E., and Smith, P. (1998). Modeling carbon and nitrogen processes in soil. Adv. Agron. 62, 253–298. Moncrieff, J. B., Massheder, J. M., Verhoef, A., Elbers, J., Heutsunkveld, B. H., Scott, S., de Bruin, H., Kabat, P., Soegaard, H., and Jarvis, P. G. (1997). A system to measure surface fluxes of energy, momentum and carbon dioxide. J. Hydrol. 188–189, 589–611. Monson, R. K., Lipson, D. L., Burns, S. P., Turnipseed, A. A., Delany, A. C., Williams, M. W., and Schmidt, S. K. (2006). Winter forest soil respiration controlled by climate and microbial community composition. Nature 439, 711–714. Monteith, J. L., Szeicz, G., and Yabuki, K. (1964). Crop photosynthesis and the flux of carbon dioxide below the canopy. J. Appl. Ecol. 1, 321–337. Moore, T. R., Roulet, N. T., and Waddington, J. M. (1998). Uncertainty in predicting the effect of climatic change on the carbon cycling of Canadian peatlands. Clim. Change 40, 229–245. Muller, C., Eickhout, B., Zaehle, S., Bondeau, A., Cramer, W., and Lucht, W. (2007). Effects of changes in CO2, climate, and land use on the carbon balance of the land biosphere during the 21st century. J. Geophys. Res. Biogeosci 112(G2), Art. No. G02032 JUN 26 2007. Nakayama, F. S. (1990). Soil respiration. Remote Sens. Rev. 5, 311–321. Neller, J. R. (1922). The influence of growing plants upon oxidation processes in the soil. Soil Sci. 13, 139–158.

40

Pete Smith et al.

Ngao, J., Longdoz, B., Granier, A., and Epron, D. (2007). Estimation of autotrophic and heterotrophic components of soil respiration by trenching is sensitive to corrections for root decomposition and changes in soil water content. Plant Soil. doi: 10.1007/s11104007-9425-z. Nicolardot, B., Fauvet, G., and Cheneby, D. (1994). Carbon and nitrogen cycling through soil microbial biomass at various temperatures. Soil Biol. Biochem. 26, 253–261. Niklin´ska, M., and Klimek, B. (2007). Effect of temperature on the respiration rate of forest soil organic layer along an elevation gradient in the Polish Carpathians. Biol. Fertil. Soils 43, 511–518. Norman, J. M., Garcia, R., and Verma, S. B. (1992). Soil surface CO2 fluxes and the carbon budget of a grassland. J. Geophys. Res. 97(D17), 18845–18853. Nyka¨nen, H., Alm, J., Lang, K., Silvola, J., and Martikainen, P. J. (1995). Emissions of CH4, N2O and CO2 from a virgin fen and a fen drained for grassland in Finland. J. Biogeogr. 22, 351–357. Oberbauer, S. F., Gillespie, C. T., Cheng, W., Gebauer, R., Sala Serra, A., and Tenhunen, J. D. (1992). Environmental effect on CO2 efflux from riparian tundra in the northern foothills of the Brooks Range, Alaska, USA. Arct. Alp. Res. 28, 328–338. Ostle, N., Ineson, P., Benham, D., and Sleep, D. (2000). Carbon assimilation and turnover in grassland vegetation using an in situ 13CO2 pulse labelling system. Rapid Commun. Mass Spectrom. 14, 1345–1350. Pare, D., Boutin, R., Larocque, G. R., and Raulier, F. (2006). Effect of temperature on soil organic matter decomposition in three forest biomes of eastern Canada. Can. J. Soil Sci. 86, 247–256. Parkinson, K. J. (1981). An improved method for measuring soil respiration in the field. J. Appl. Ecol. 18, 221–228. Paterson, E., Hall, J. M., Rattray, E. A. S., Griffiths, B. S., Ritz, K., and Killham, K. (1997). Effect of elevated CO2 on rhizosphere carbon flow and soil microbial processes. Global Change Biol. 3, 363–377. Paustian, K. (1994). Modelling soil biology and biochemical processes for sustainable agricultural research. In ‘‘Soil Biota. Management in Sustainable Farming Systems’’ (C. E. Pankhurst, B. M. Doube, V. V. S. R. Gupta, and P. R. Grace, Eds.), pp. 182–193. CSIRO Information Services, Melbourne, Australia. Post, W. M., and Kwon, K. C. (2000). Soil carbon sequestration and land-use change: Processes and potential. Global Change Biol. 6, 317–327. Post, W. M., Emanuel, W. R., Zinke, P. J., and Stangenberger, A. G. (1982). Soil carbon pools and world life zones. Nature 298, 156–159. Raich, J. W., and Nadelhoffer, K. J. (1989). Belowground carbon allocation in forest ecosystems: Global trends. Ecology 70, 1346–1354. Raich, J. W., and Schlesinger, W. H. (1992). The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus 44B, 81–99. Raich, J. W., Bowden, R. D., and Steudler, P. A. (1990). Comparison of 2 static chamber techniques for determining carbon-dioxide efflux from forest soils. Soil Sci. Soc. Am. J. 54, 1754–1757. Raich, J. W., Potter, C. S., and Bhagawati, D. (2002). Interannual variability in global soil respiration, 1980–1994. Global Change Biol. 8, 800–812. Rasmussen, C., Southard, R. J., and Horwath, W. R. (2006). Mineral control of organic carbon mineralization in a range of temperate conifer forest soils. Global Change Biol. 12, 834–847. Rayment, M. (2000). Closed chamber systems underestimate soil CO2 efflux. Eur. J. Soil Sci. 51, 107–110.


41

Reichstein, M., Subke, J.-A., Angeli, A. C., and Tenhunen, J. D. (2005). Does the temperature sensitivity of decomposition of soil organic matter depend upon water content, soil horizon, or incubation time? Global Change Biol. 11, 1754–1767. Rey, A., and Jarvis, P. (2006). Modelling the effect of temperature on carbon mineralization rates across a network of European forest sites (FORCAST). Global Change Biol. 12, 1894–1908. Rey, A., Petsikos, C., Jarvis, P. G., and Grace, J. (2005). Effect of temperature andmoisture on rates of carbon mineralization in a Mediterranean oak forest soil under controlled and field conditions. Eur. J. Soil Sci. 56, 589–599. Rochette, P., Flanagan, L. B., and Gregorich, E. G. (1999). Separating soil respiration into plant and soil components using analyses of the natural abundance of carbon-13. Soil Sci. Soc. Am. J. 63, 1207–1213. Romell, L. G. (1932). Mull and duff as biotic equilibria. Soil. Sci. 34, 161–168. Rustad, L. E., Campbell, J. L., Marion, G. M., Norby, R. J., Mitchell, M. J., Hartley, A. E., Cornelissen, J. H., and Gurevitch, J. (2001). A meta-analysis of the response of soil respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming. Oecologia 126, 543–562. Ryan, M. G., and Law, B. E. (2005). Interpreting, measuring, and modeling soil respiration. Biogeochemistry 73, 3–27. Schimel, D. S. (1995). Terrestrial ecosystems and the carbon-cycle. Global Change Biol. 1, 77–91. Schimel, D. S., House, J. I., Hibbard, K. A., Bousquet, P., Ciais, P., Peylin, P., Braswell, B. H., Apps, M. J., Baker, D., Bondeau, A., Canadell, J., Churkina, G., et al. (2001). Recent patterns and mechanisms of carbon exchange by terrestrial ecosystems. Nature 414, 169–172. Sierra, J., and Renault, P. (1996). Respiratory activity and oxygen distribution in natural aggregates in relation to anaerobiosis. Soil Sci. Soc. Am. J. 60, 1428–1438. Singh, J. S., and Gupta, S. R. (1977). Plant decomposition and soil respiration in terrestrial ecosystems. Bot. Rev. 43, 449–528. Six, J., Conant, R. T., Paul, E. A., and Paustian, K. (2002). Stabilization mechanisms of soil organic matter: Implications for C-saturation of soils. Plant Soil 241, 155–176. Six, J., Bossuyt, H., Degryze, S., and Denef, K. (2004). A history of research on the link between (micro)aggregates, soil biota, and soil organic matter dynamics. Soil Tillage Res. 79, 7–31. Sjursen, H., Michelsen, A., and Jonasson, S. (2005). Effects of long-term soil warming and fertilization on microarthropod abundances in three sub-arctic ecosystems. Appl. Soil Ecol. 30, 148–161. Sleutel, S., De Neve, S., and Hofman, G. (2003). Estimates of carbon stock changes in Belgian cropland. Soil Use Manage. 19, 166–171. Smith, P. (2001). Soil organic matter modeling. In ‘‘Encyclopedia of Soil Science.’’ (R. Lal, Ed.). Marcel Dekker Inc, New York. Smith, P. (2002). Soil and the environment: Role of soil in models for climate change. In ‘‘Encyclopedia of Soils in the Environment’’ (D. Hillel, C. Rosenzweig, D. Powlson, K. Scow, M. Singer, and D. Sparks, Eds.), Academic Press, London. Smith, P. (2004). Soils as carbon sinks—the global context. Soil Use Manage. 20, 212–218. Smith, P. (2008). Land use change and soil organic carbon dynamics. Nutr. Cycl. Agroecosyst (in press). doi: 10.1007/s10705-007-9138-y. Smith, P., and Powlson, D. S. (2003). Sustainability of soil management practices—a global perspective. In ‘‘Soil Biological Fertility—A Key To Sustainable Land Use In Agriculture.’’ (L. K. Abbott, and D. V. Murphy, Eds.), pp. 241–254. Kluwer Academic Publishers, Dordrecht, Netherlands.

42

Pete Smith et al.

Smith, P., and Trines, E. (2007). Agricultural measures for mitigating climate change: Will the barriers prevent any benefits to developing countries? Int. J. Agric. Sustain. 4, 173–175. Smith, P., Powlson, D. S., and Glendining, M. J. (1996). Establishing a European soil organic matter network (SOMNET). In ‘‘Evaluation of Soil Organic Matter Models using Existing, Long-Term Datasets’’ (D. S. Powlson, P. Smith, and J. U. Smith, Eds.), NATO ASI Series I, Vol. 38, pp. 81–98. Springer-Verlag, Berlin. Smith, P., Smith, J. U., Powlson, D. S., McGill, W. B., Arah, J. R. M., Chertov, O. G., Coleman, K., Franko, U., Frolking, S., Jenkinson, D. S., Jensen, L. S., Kelly, R. H., et al. (1997). A comparison of the performance of nine soil organic matter models using seven long-term experimental datasets. Geoderma 81, 153–225. Smith, P., Andreń, O., Brussaard, L., Dangerfield, M., Ekschmitt, K., Lavelle, P., and Tate, K. (1998). Soil biota and global change at the ecosystem level: Describing soil biota in mathematical models. Global Change Biol. 4, 773–784. Smith, P., Falloon, P., Smith, J. U., and Powlson, D. S. (Eds.) (2001). ‘‘Soil Organic Matter Network (SOMNET): 2001 Model and Experimental Metadata.’’ GCTE Report 7 (2nd ed.). GCTE Focus 3 Office Wallingford, Oxon, 224pp. Smith, P., Falloon, P. D., Ko¨rschens, M., Shevtsova, L. K., Franko, U., Romanenkov, V., Coleman, K., Rodionova, V., Smith, J. U., and Schramm, G. (2002). EuroSOMNET— a European database of long-term experiments on soil organic matter: The WWW metadatabase. J. Agric. Sci., Camb. 138, 123–134. Smith, J. U., Smith, P., Wattenbach, M., Zaehle, S., Hiederer, R., Jones, R. J. A., Montanarella, L., Rounsevell, M. D. A., Reginster, I., and Ewert, F. (2005). Projected changes in mineral soil carbon of European croplands and grasslands, 1990–2080. Global Change Biol. 11, 2141–2152. Smith, P., Smith, J. U., Wattenbach, M., Meyer, J., Lindner, M., Zaehle, S., Hiederer, R., Jones, R., Montanarella, L., Rounsevell, M., Reginster, I., and Kankaanpaä¨, S. (2006). Projected changes in mineral soil carbon of European forests, 1990–2100. Can. J. Soil Sci. 86, 159–169. Smith, P., Martino, D., Cai, Z., Gwary, D., Janzen, H. H., Kumar, P., McCarl, B., Ogle, S., O’Mara, F., Rice, C., Scholes, R. J., Sirotenko, O., et al. (2007a). Greenhouse gas mitigation in agriculture. Philos. Trans. R. Soc. Lond. B 363. doi: 10.1098/rstb.2007.2184. Smith, P., Martino, D., Cai, Z., Gwary, D., Janzen, H. H., Kumar, P., McCarl, B., Ogle, S., O’Mara, F., Rice, C., Scholes, R. J., Sirotenko, O., et al. (2007b). Policy and technological constraints to implementation of greenhouse gas mitigation options in agriculture. Agric., Ecosyst. Environ. 118, 6–28. Smith, P., Chapman, S. J., Scott, W. A., Black, H. I. J., Wattenbach, M., Milne, R., Campbell, C. D., Lilly, A., Ostle, N., Levy, P., Lumsdon, D. G., Millard, P., et al. (2007c). Climate change cannot be entirely responsible for soil carbon loss observed in England and Wales, 1978–2003. Global Change Biol. 13, 2605–2609. Smith, P., Smith, J. U., Franko, U., Kuka, K., Romanenkov, V. A., Shevtsova, L. K., Wattenbach, M., Gottschalk, P., Sirotenko, O. D., Rukhovich, D. I., Koroleva, P. V., Romanenko, I. A., et al. (2007d). Changes in soil organic carbon stocks in the croplands of European Russia and the Ukraine, 1990–2070; comparison of three models and implications for climate mitigation. Reg. Environ. Change 7, 105–119. Smith, J. U., Smith, P., Wattenbach, M., Gottschalk, P., Romanenkov, V. A., Shevtsova, L. K., Sirotenko, O. D., and Rukhovich, D. I.Koroleva, P. V., Romanenko, I. A., and Lisovoi, N. V. (2007e). Projected changes in cropland soil organic carbon stocks in European Russia and the Ukraine, 1990–2070. Global Change Biol. 13, 342–356. Stoy, P. C., Palmroth, S., Oishi, A. C., Siqueira, M. B. S., Juan, J.-Y., Novick, K. A., Ward, E. J., Katul, G. G., and Oren, R. (2007). Are ecosystem carbon inputs and outputs coupled at short time scales? A case study from adjacent pine and hardwood forests using impulse-response analysis. Plant Cell Environ. 30, 700–710.


43

Subke, J. A., Hahn, V., Battipaglia, G., Linder, S., Buchmann, N., and Cotrufo, M. F. (2004). Feedback interactions between needle litter decomposition and rhizosphere activity. Oecologia 139, 551–559. Subke, J. A., Inglima, I., and Cotrufo, M. F. (2006). Trends and methodological impacts in soil CO2 efflux partitioning: A metaanalytical review. Global Change Biol. 12, 921–943. Trasar-Cepeda, C., Gil-Sotres, F., and Leiro´s, M. C. (2007). Thermodynamic parameters of enzymes in grassland soils from Galicia, NW Spain. Soil. Biol. Biochem. 39, 311–319. Trumbore, S. (1997). Potential response of soil organic carbon to global environmental change. Proc. Natl. Acad. Sci. USA 94, 8284–8291. Trumbore, S. (2006). Carbon respired by terrestrial ecosystems—recent progress and challenges. Global Change Biol. 12, 141–153. Van Groenigen, K.-J., Six, J., Hungate, B. A., de Graaff, M.-A., van Breemen, N., and van Kessel, C. (2006). Element interactions limit soil carbon storage. Proc. Natl. Acad. Sci. USA 103, 6571–6574. Welles, J. M., Demetriades-Shah, T. H., and McDermitt, D. K. (2001). Considerations for measuring ground CO2 fluxes with chambers. Chem. Geol. 177, 3–13. White, A., Cannell, M. G. R., and Friend, A. D. (1999). Climate change impacts on ecosystems and the terrestrial carbon sink: A new assessment. Glob. Environ. Change 9, S21–S30. Widen, B., and Lindroth, A. (2003). A calibration system for soil carbon dioxide-efflux measurement: Description and application. Soil Sci. Soc. Am. J. 67, 327–334. Xu, L., Furtaw, M. D., Madsen, R. A., Garcia, R. L., Anderson, D. J., and McDermitt, D. K. (2006). On maintaining pressure equilibrium between a soil CO2 flux chamber and the ambient air. J. Geophys. Res. Atmos 111(D8), Art. No. D08S10. Yim, M. H., Joo, S. J., and Nakane, K. (2002). Comparison of field methods for measuring soil respiration: A static alkali absorption method and two dynamic closed chamber methods. Forest Ecol. Manage. 170, 189–197. Zimmermann, M., Leifeld, J., Schmidt, M. W. I., Smith, P., and Fuhrer, J. (2007). Measured soil organic matter fractions can be related to pools in the RothC model. Eur. J. Soil Sci. 58, 658–667. Zimov, S. A., Schuur, E. A. G., and Chapin, F. S. (2006). Permafrost and the global carbon budget. Science 312, 1612–1613.

C H A P T E R

T W O

Some Prospective Strategies for Improving Crop Salt Tolerance M. Ashraf,* H. R. Athar,† P. J. C. Harris,‡ and T. R. Kwon§ Contents 1. Introduction 2. Presowing Seed Treatment: A Shotgun Approach to Enhance Crop Salt Tolerance 2.1. Osmopriming 2.2. Halopriming 2.3. Hydropriming 2.4. Thermopriming 2.5. Hormone-priming 3. Exogenous Application of Osmolytes, Osmoprotectants, and PGRs 3.1. Glycine betaine 3.2. Proline 3.3. Plant growth regulators 3.4. Supply of nutrients through the rooting medium 3.5. Foliar application of macronutrients 4. Breeding for Salt Tolerance 4.1. Mutation breeding 4.2. Identification of traits using advanced molecular techniques 4.3. Molecular breeding for salt tolerance 5. Conclusion and Future Prospects References

46 47 48 48 49 49 50 51 52 53 57 62 64 65 66 68 69 90 92

Soil salinity is a major environmental constraint to crop productivity worldwide. The ‘‘biological’’ approach to this problem focuses on the management, exploitation, or development of plants able to thrive on salt-affected soils. This chapter reviews strategies by which plants can be enabled to grow on saline soils. The first strategy is to prime seeds before planting by treating them with inorganic or organic chemicals and/or with high or low temperatures.

* { { }

Department of Botany, University of Agriculture, Faisalabad 30840, Pakistan Institute of Pure and Applied Biology, Bahauddin Zakariya University, Multan, Pakistan Faculty of Business, Environment and Society, Coventry University, Coventry, United Kingdom The National Institute of Agricultural Biotechnology, Suwon 441-707, Korea


#


45

46

M. Ashraf et al.

The second strategy involves exogenous application of organic chemicals, such as glycine betaine, proline, or plant growth regulators, or inorganic chemicals to plants under salinity stress. Considerable improvements in growth and yield have been reported in a number of crops using these approaches. The third strategy is to employ selection and breeding. Major efforts have been made to develop salt-tolerant lines or cultivars of crops using conventional plant breeding. However, the complexity of the tolerance mechanisms, lack of selection criteria, and variation in responses of plants at different developmental stages have resulted in only limited success. The emphases for developing salttolerant lines/cultivars are now on marker-assisted breeding and genetic transformation. The development of salt-tolerant transgenic plants is still at an early stage but may become increasingly more effective as better knowledge of the complex mechanisms involved in plant salt tolerance is acquired. Furthermore, the rapid expansion in knowledge on genomics and proteomics will undoubtedly accelerate the transgenic and molecular breeding approaches However, to date, there are few conclusive reports indicating successful performance of transgenic cultivars under natural stressful environments.

1. Introduction The world human population is expected to reach 8.0 billion by 2025 and 8.9 billion by 2050 (FAO, 2006). About 80 million people are being added to the population total each year, and 97% of the predicted population growth will take place in the developing countries. It is projected that there is a need to double world food production in order to feed 8.0 billion people by 2025. This will certainly place more pressure on the environment. The developing countries, in particular, are confronted with severe food-security challenges. With a severe limit to the amount of unused land available to bring into cultivation, improving crop yields in both normal soils and less productive lands, including salt-affected lands, is an absolute utter requirement to satisfy future world food needs. Environmental stresses represent a major constraint to meeting the world food demand. There are relatively few ‘‘stress free’’ areas where crops may approach their potential yields. Abiotic environmental factors (abiotic stresses) are considered to be the main source (71%) of yield reductions (Boyer, 1982). The estimation of potential yield losses by individual abiotic stresses are estimated at 17% by drought, 20% by salinity, 40% by high temperature, 15% by low temperature, and 8% by other factors (Ashraf and Harris, 2005). According to one estimate, about 380 million ha, almost one-third of the area under cultivation, is affected by salinity coupled with waterlogging and alkalinity (Ghassemi et al., 1995). Sixty million ha are exposed to overirrigation, where a raised water table transports underground salt, mainly NaCl, to the soil surface. It has been estimated that the resulting

Some Prospective Strategies for Improving Crop Salt Tolerance

47

agricultural salinization can degrade as much land as is newly irrigated each year. Unless checked, this process of salinization will, in due course, result in a net loss of productive land as the area suitable for irrigation becomes exhausted. Both farmers and scientists are familiar with the two major strategies for the utilization of salt-affected lands: first, the use of reclamative and preventive measures to make the salt-affected soils fit for agriculture; second, utilization of salt-affected soils by growing halophytes or salt-tolerant crops/cultivars. The latter approach has been referred to as the ‘‘biological approach’’ (Ashraf, 1994a; Epstein, 1977; Kingsbury and Epstein, 1984) and has considerable potential to mitigate the problem of soil salinity around the world. This chapter focuses on the range of biological strategies by which salinity tolerance of potential crops can be increased.

2. Presowing Seed Treatment: A Shotgun Approach to Enhance Crop Salt Tolerance Germination of a seed primarily depends on the availability of water, but a seed subjected to salt stress faces a shortage of water because of low water potential of the saline soil. As a result, both the rate of germination and total germination of commercial crops may be drastically reduced in salt-affected soils. Restricted water uptake may result in various structural, physiological, and biochemical changes within a germinating seed, and these have been described elsewhere (Ashraf and Foolad, 2005). These processes may reduce seed germination under saline conditions (PoljakoffMayber et al., 1994). Seed germination and early-seedling establishment are generally considered phases of growth, most sensitive to salt stress as plant salt tolerance usually improves with plant development (Foolad, 2004). However, rapid and uniform seed germination and early-seedling growth are of vital importance for crop production in saline soils. Thus, if the adverse effects of salt can be alleviated at the initial growth stages, the chances of establishing a crop in saline soils will be much improved (Ashraf and Foolad, 2005; Ashraf et al., 2003; Foolad, 2000; Sallam, 1999). Many studies have shown that salt tolerance of plants can be improved by treating seed with water or solutions of inorganic or organic salts before sowing (Babaeva et al., 1999; Chang-Zheng et al., 2002; Pill et al., 1991; Rehman et al., 1998; Strogonov, 1964). During presowing seed treatment, a process generally referred to as ‘‘priming,’’ seeds are immersed in a solution either with an external water potential that is low enough to prevent germination but allows some pregerminative physiological and biochemical phenomena to occur or for a duration that is insufficient for germination to pass a critical point (Bradford, 1986). Primed seeds usually germinate more

48

M. Ashraf et al.

rapidly than unprimed seeds once placed in an appropriate germination environment. Over the last two decades, the seed priming approach has been used extensively to enhance the rate and uniformity of field germination and emergence in many important crop plants grown under both saline and nonsaline conditions. Seed priming as a technique is adopted to improve the rate and synchrony of seed germination and can be subdivided into osmopriming, halopriming, hydropriming, thermopriming, and hormone-priming.

2.1. Osmopriming This is also known as osmoconditioning or osmotic conditioning. In this technique, seeds are soaked for a certain time period in solutions of sugars, polyethylene glycol (PEG), glycerol, sorbitol, or mannitol followed by airdrying before sowing. The low water potential of the solutions of these organic compounds allows partial seed hydration so that pregermination metabolic processes are activated but germination does not proceed (Bennett et al., 1992; McDonald, 2000; Pill and Necker, 2001). When the primed seeds are planted in the field, they usually show fast and even germination. For example, seeds of tomato (Lycopersicon esculentum) and asparagus (Asparagus officinalis) osmoconditioned with 0.8 MPa PEG8000 (Pill et al., 1991), that of Bermuda grass (Cynodon dactylon) also with PEG (Al-Humaid, 2002), and that of cucumber (Cucumis sativus) with 0.7 M mannitol (Passam and Kakouriotis, 1994) showed enhanced germination under saline conditions. Osmopriming not only improves seed germination but also enhances general crop performance under nonsaline or saline conditions, for example, osmoconditioning of Italian ryegrass (Lolium multiflorum) and sorghum (Sorghum bicolor) seeds with 20% PEG-8000 for 2 days at 10 C enhanced germination rate, percent germination, seedling growth, and dry matter production under water-stressed, waterlogged, cold-stress, or saline conditions (Hur, 1991).

2.2. Halopriming Halopriming refers to soaking seed in solutions of inorganic salts. A number of studies have shown a significant improvement in seed germination, seedling emergence and establishment, and final crop yield in salt-affected soils as a result of seed halopriming. For example, rice seed treated with a mixed salt solution germinated more rapidly than unprimed seed under saltstress conditions (Chang-Zheng et al., 2002). In another study, presoaking Echinacea purpurea seed in either 0.1% MnSO4 or 0.05% ZnSO4 solution increased germination percentage by 36% or 38%, respectively, and field emergence by 27–41% (Babaeva et al., 1999). Halopriming not only promotes seed germination but may also stimulate subsequent growth, thereby


49

enhancing final crop yield (Eleiwa, 1989; Sallam, 1999). For example, presoaking seed in solutions of inorganic salts improved growth in broad bean (Vicia faba) (Sallam, 1999), grain yield in wheat (Triticum aestivum) (Mehta et al., 1979), and soybean (Glycine max) (Eleiwa, 1989), and both growth and seed yield in Pennisetum americanum and S. bicolor (Kadiri and Hussaini, 1999) under saline conditions.

2.3. Hydropriming Hydropriming involves soaking the seeds in water before sowing (Pill and Necker, 2001) and may or may not be followed by air-drying of the seeds. The process in which seeds are subjected to one or more cycles of wetting and drying prior to germination has also been referred to as ‘‘hardening’’ (Bewley and Black, 1994; Hafeez and Hudson, 1967). Harris (1992) proposed this approach as a low-cost technique, described as ‘‘on-farm seed priming,’’ for several crops in developing countries. Hydropriming may enhance seed germination and seedling emergence under both saline and nonsaline conditions. For example, Roy and Srivastava (1999) found that soaking wheat kernels in water improved their germination rate under saline conditions. Improvement in salt tolerance of maize (Zea mays) (Ashraf and Rauf, 2001), pigeon pea (Cajanus cajan) ( Jyotsna and Srivastava, 1998), and Acacia seeds (Rehman et al., 1998) was also observed following hydropriming. The precise mechanisms by which application of this simple technique can achieve sometimes quite dramatic improvements in plant growth and seed yield in saline or nonsaline conditions remain unclear. Some researchers have considered hydropriming a ‘‘key technology’’ that is simple and cost effective, the impact of which is very high in terms of enhanced yield (Ashraf and Foolad, 2005; Harris et al., 1999).

2.4. Thermopriming Thermopriming refers to treating seed with low or high temperatures to improve germination and seedling emergence under adverse environmental conditions. Low-temperature treatment of seed in some species is a common practice in agriculture either to protect seed from precocious germination in unsuitable environments or to positively improve germination (Ashraf and Foolad, 2005; Bewley and Black, 1994). For example, Sharma and Kumar (1999) found that chilling treatment of Brassica juncea seed for 5, 10, or 15 days resulted in enhanced germination under salt stress. In another study, chilling pearl millet (Pennisetum glaucum) seed for 2 days at 5 C increased the final germination percentage, but not the germination rate under saline conditions (Ashraf et al., 2003). Similar mitigation of adverse effects of salt stress by chilling treatment of seed has also been observed in

50

M. Ashraf et al.

Indian grass (Sorghastrum nutans) (Watkinson and Pill, 1998) and parsnip (Pastinaca sativa) (Finch-Savage and Cox, 1982). Although presowing seed treatment with specific temperatures is effective in improving seed germination and seedling emergence, it can also improve later plant growth and development. However, there is little information available in the literature reflecting the role of temperature treatment of seed in improving growth and seed yield of different crops under saline conditions.

2.5. Hormone-priming Pretreating seed with optimal concentrations of plant growth regulators (PGRs) can effectively improve germination as well as growth and yield of crops under both normal and stress conditions (Darra et al., 1973; Hurly et al., 1991; Lee et al., 1998). Several growth regulators are commonly used for seed priming, including auxins (IAA, IBA, NAA), gibberellins (GA), gibberellin antagonists, kinetin, abscisic acid, polyamines (PAs), ethylene, brasinolide, salicylic acid (SA), triacontanol, and ascorbic acid. Treating seed with growth hormones has been shown to improve crop germination under salt stress. For example, treating seed of Sudan grass (Sorghum sudanense) with chlormequat chloride (CCC) alleviated the adverse effect of salt stress on rate and percentage of germination (Ismaeil et al., 1993). In pigeon pea seed, kinetin and ascorbic acid were effective in mitigating the adverse effects of salt stress on germination ( Jyotsna and Srivastava, 1998). Similarly, seed germination of wheat was improved under salt stress by soaking seed with IAA, NAA, or GA (Balki and Padole, 1982). Likewise, in another study with wheat, seed pretreated with different concentrations of GA showed improved germination under saline conditions, with presoaking in 50 mg liter1 GA having the greatest effect (Parashar and Varma, 1988). Gulnaz et al. (1999a) observed similar findings with wheat seeds presoaked seed in IAA, IBA, or GA and exposed to saline conditions. Similarly, GA was found a very effective priming agent in alleviating the effects of salt stress on tomato (Kang et al., 1996) and okra (Abelmoschus esculentus) (Vijayaraghavan, 1999). Similarly, priming seeds with a moderate concentration of kinetin (150 mg liter1) was effective in improving growth and grain yield of two wheat cultivars, and this appeared to be related to the beneficial effects of kinetin on photosynthetic capacity and water use efficiency under saline conditions (Iqbal and Ashraf, 2005a). While examining the effect of PAs as priming agents, Iqbal and Ashraf (2005b) found that although all three PAs, spermine (Spm), spermidine (Spd), and putrescine (Put) were effective in alleviating the adverse effect of salt stress on wheat plants, their effects in altering the concentration of different ions and growth were cultivar specific. Other plant growth hormones that have proven to be effective


51

priming agents for improving growth or seed yield under saline conditions and not just germination of the pretreated seed include CCC in Sudan grass (Ismaeil et al., 1993), 2,4-D, IAA, NAA, GA, ascorbic acid, thiamin, and sodium salicylate in wheat (Al-Hakimi and Hamada, 2001; Balki and Padole, 1982; Gulnaz et al., 1999b; Parashar and Varma, 1988), ascorbic acid, thiamin, and pyridoxine in sunflower (Helianthus annuus) and maize (Ahmed-Hamad and Monsaly, 1998), and 28-homobrassinolide in mung bean (Vigna radiata) (Fariduddin et al., 2003). Although the above-mentioned reports show considerable benefits of different growth substances as potential seed priming agents, some of the growth substances are very expensive and their use by farmers may not be economic. Thus, there is a need to identify growth regulators that are relatively cheap, easily available, and yet effective (Ashraf and Foolad, 2005). In addition to the above-mentioned seed-priming techniques, there are many other techniques such as matric priming, biopriming, drum priming, seed pelleting, film coating, and fluid drilling that are being practised to improve germination and later crop growth and seed yield under nonstress conditions. However, there is no clear evidence in the literature that they provide a means for improving growth under saline conditions. The seed priming methods discussed above each have advantages and disadvantages and they may not be equally effective in all crops or in protecting plants from saline stress at different developmental stages. To determine the optimum conditions for each priming technique, factors such as concentration and volume of priming agent, seed soaking time, whether and how the seeds are dried after priming, and subsequent seed storability must be investigated (Ashraf and Foolad, 2005). Furthermore, optimum treatments determined in laboratory experiments on seed priming require scaling up to treat large quantities of seed for use under field conditions and this itself may influence the methods used.

3. Exogenous Application of Osmolytes, Osmoprotectants, and PGRs Overproduction of different types of compatible organic solutes is one of the most common responses in plants grown under stress conditions (Serraj and Sinclair, 2002). ‘‘Compatible solutes’’ is a term given to highly soluble, low molecular weight compounds that can accumulate at high levels within cells without impairing cellular function. They protect plants from stress by contributing to cellular osmotic adjustment, detoxifying reactive oxygen species (ROS), protecting membrane structure, and stabilizing proteins (Ashraf and Foolad, 2007; Bohnert and Jensen, 1996; Yancey et al., 1982). Some of these solutes can effectively protect cellular

52

M. Ashraf et al.

components from dehydration injury, so they are generally referred to as osmoprotectants. These solutes include proline, sucrose, polyols, trehalose, and quaternary ammonium compounds such as GB, alaninebetaine, prolinebetaine, choline O-sulfate, hydroxyprolinebetaine, and pipecolatebetaine (Ashraf and Harris, 2004; Rhodes and Hanson, 1993). Although much effort has been devoted to genetically engineering plants for overproduction of osmoprotectants, there has been only limited success in achieving levels of these osmolytes in plants that could help sustain crop growth under saline conditions. It has been reported that transgenic plants overexpressing biosynthetic enzymes for osmoprotectants, such as mannitol, GB, D-ononitol, or sorbitol, accumulate these substances in amounts too low to provide protective effects by osmotic mass action alone (Huang et al., 2000; Sakamoto et al., 1998; Sheveleva et al., 1997). Nonetheless, increased resistance to various abiotic stresses has been achieved through exogenous application of organic solutes. However, this approach that may effectively contribute to enhanced crop production under saline conditions has not received much attention (Ashraf and Foolad, 2007).

3.1. Glycine betaine Glycine betaine, a quaternary ammonium compound, is known to accumulate in response to salt stress in many crops (Fallon and Phillips, 1989; McCue and Hanson, 1995; Weimberg et al., 1984; Yang et al., 2003). However, in some crop plants such as rice (Oryza sativa), Brassica spp., Arabidopsis, and tobacco (Nicotiana tabacum), the naturally produced GB is not sufficient to alleviate the adverse effects of salt stress (Subbarao et al., 2001; Wyn Jones and Storey, 1981; Yancey, 1994). However, exogenous application of GB to nonaccumulating or low-accumulating plants may significantly reduce the adverse effects of salt stress (Agboma et al., 1997a,b,c; Ma˚kela et al., 1998a). It has been observed that exogenously applied GB can rapidly penetrate through leaf tissue and rapidly translocate to other plant organs, where it contributes to enhanced stress tolerance (Ma˚kela et al., 1998a). In addition, because naturally produced GB does not normally break down in plants (Bray et al., 2000), it can easily be extracted as a low-cost by-product from many flowering plants, including sugar beet (Beta vulgaris) (Rhodes and Hanson, 1993). This makes exogenous application of GB a potentially economic technique for offsetting the adverse effects of environmental stresses on crops (Ashraf and Foolad, 2007). Exogenous application of GB has resulted in a significant improvement in salt tolerance of a number of plants (Harinasut et al., 1996; Lutts, 2000). For example, exogenous application of GB to tomato plants subjected to saline conditions or high temperatures caused approximately 40% increase in fruit yield (Table 1) compared with control plants with no GB treatment (Ma˚kela et al., 1998a,b).


53

While evaluating the effects of exogenous application of GB on growth inhibition and ultrastructural damage induced by salt stress in rice seedlings, Rahman et al. (2002) found that GB treatment assuaged salt-induced inhibition of shoot growth but not of root growth. Furthermore, significantly lower Naþ and higher Kþ concentrations accumulated in the shoots of GB-treated plants than of untreated control plants. The ameliorative effects of GB in the shoot were suggested to be due to GB-induced production of additional vacuoles in root cells that resulted in a greater accumulation of Naþ in the root and a low transportation to the shoot. Although the mechanisms by which externally applied GB exerts its effects at the cellular level are not fully elucidated, a study using two GBnonaccumulating cruciferae species, Brassica napus (canola) and Arabidopsis thaliana, suggested that exogenous GB has a destabilizing effect on photorespiration via competitive effects with glycine at the mitochondrial step of the glycolate pathway (Sulpice et al., 2002). However, prior to recommending the application of GB to farmers for improving crop stress tolerance, better elucidation of its mechanism of action and possible side effects is vital. Furthermore, Ashraf and Foolad (2007) reported that effective and efficient doses of GB may vary with plant species and plant developmental stage. The time of application and environmental conditions under which plants are grown may also influence the effectiveness of GB in enhancing stress tolerance. Thus, optimization of the dose–response relationship in different plant species and at different developmental stages before commercial use of GB as a stress-tolerant stimulator is a prerequisite.

3.2. Proline Like GB, proline is also known to accumulate in high concentrations in response to various abiotic stresses (Ali et al., 1999; Ashraf and Foolad, 2007; Hsu et al., 2003; Ozturk and Demir, 2002; Rhodes et al., 1999). Despite the active participation of proline in osmotic adjustment as an osmolyte, it also plays an important role in a number of physiological and biochemical phenomena such as stabilizing subcellular structures, scavenging free radicals, and buffering cellular redox potential under stress conditions (Hare and Cress, 1997; Srinivas and Balasubramanian, 1995). Proline accumulation normally occurs in the cytosol where it contributes considerably to the cytoplasmic osmotic adjustment in plants subjected to drought or salinity stress (Binzel et al., 1987; Ketchum et al., 1991; Leigh et al., 1981). Accumulation of proline under salt stress in many plant species has been correlated with stress tolerance, and its concentration has been shown to be generally higher in salt-tolerant than in salt-sensitive plants (Fouge`re et al., 1991; Gangopadhyay et al., 1997; Madan et al., 1995; Petrusa and Winicov, 1997).

54 Table 1

Different seed priming techniques and their effectiveness in improving growth of various crops under saline conditions

Presowing technique

Nature of priming agent

Crop

Attribute improved

References

Osmopriming

Polyethylene glycol

Tomato and asparagus Bermuda grass Italian ryegrass and sorghum Cucumber

Germination

Pill et al., 1991

Germination Germination and later growth Germination

Al-Humaid, 2002 Hur, 1991

Broad bean

Growth

Passam and Kakouriotis, 1994 Sallam, 1999

Wheat Soybean Pennisetum americanum and Sorghum bicolor Wheat

Grain yield Grain yield Both growth and seed yield

Mehta et al., 1979 Eleiwa, 1989 Kadiri and Hussaini, 1999

Germination rate

Maize Pigeon pea

Germination Germination

Brassica juncea

Germination

Roy and Srivastava, 1999 Ashraf and Rauf, 2001 Jyotsna and Srivastava, 1998 Sharma and Kumar, 1999

Mannitol Halopriming

Hydropriming

Solutions of inorganic salts

Water

Thermopriming Chilling treatment

Pearl millet

Hormonepriming

Indian grass

Improved germination percentage but not germination rate Germination

Parsnip

Germination

Chlormequat chloride Kinetin and ascorbic acid

Sudan grass Pigeon pea

Germination Germination

IAA, NAA, or Gibberellins IAA, IBA, or GA3 GA3

Wheat Wheat Wheat

Germination Germination Germination

Gibberellins

Okra and Pearl millet

28-Homobrassinolide

Mung bean

Ascorbic acid, thiamin, and pyridoxine

Sunflower and maize

Growth and Grain yield Growth and grain yield Growth and grain yield

Ashraf et al., 2003

Watkinson and Pill, 1998 Finch-Savage and Cox, 1982 Ismaeil et al., 1993 Jyotsna and Srivastava, 1998 Balki and Padole, 1982 Gulnaz et al., 1999b Parashar and Varma, 1988 Vijayaraghavan, 1999 Fariduddin et al., 2003 Ahmed-Hamad and Monsaly, 1998

55

56

M. Ashraf et al.

Like GB, exogenously applied proline plays a significant role in enhancing crop tolerance to various abiotic stresses. According to some reports, exogenously applied proline provides osmoprotection and facilitates growth of plants subjected to salt stress (Csonka and Hanson, 1991; Yancey, 1994). It also enhances the activities of antioxidants that in turn protect membranes from salt-induced oxidative stress (Yan et al., 2000). For example, exogenous application of proline to soybean cell cultures under salt stress increased the activities of superoxide dismutase (SOD) and peroxidase (Table 1), which normally contribute to increased salt tolerance (Hua and Guo, 2002; Yan et al., 2000). In another study, exogenous application of proline to stressed plants of the halophyte Allenrolfea occidentalis neutralized the enhanced production of ethylene induced by the salt or drought stress (Chrominski et al., 1989). Accumulation of proline in the cytoplasm causes a reduction in the concentration of less-compatible solutes and an increase in the cytosolic water volume (Cayley et al., 1992). For example, exogenous application of proline to barley (Hordeum vulgare) embryo cultures under saline conditions caused a significant decrease in shoot Naþ and Cl concentrations, thereby causing an increase in growth (Lone et al., 1987). The authors suggested that the ameliorative effect of proline in barley was due to membrane stabilization. This was further supported by Mansour (1998) while assessing the ameliorative effect of exogenous application on salt-stressed onion (Allium cepa). In rice, exogenous application of 30 mM of proline offsets the adverse effects of salinity on early-seedling growth and enhances the Kþ/Naþ ratio (Roy et al., 1993). Despite the above reports, in some cases, exogenous application of proline neither changed the cellular levels of toxic elements nor mitigated the adverse effects of salt stress. For example, foliar application of proline to rice plants subjected to saline conditions did not alter the Naþ or Cl content of the leaves. These results led to the conclusion that exogenous application of proline did not offset salt damage (Krishnamurthy and Bhagwat, 1993). Moreover, Hare et al. (2002) demonstrated that exogenous application of proline caused damage to the ultrastructure of chloroplasts and mitochondria in 21-day-old Arabidopsis plants. It may be that the concentration of proline used for exogenous application to these Arabidopsis plants was too high, resulting in harmful rather than beneficial effects. Considering that high concentrations of proline may be injurious to plant growth or cellular metabolisms, it is essential to determine the optimal concentration for advantageous effects (Ashraf and Foolad, 2007). It is very likely that the optimal concentration of proline is species, and possibly also variety, dependent and needs to be determined before commercial application of exogenous proline to crops to improve their stress tolerance can be recommended. Moreover, because both stress tolerance and response


57

to elevated proline levels are likely to vary from stage to stage throughout the plant development, it is crucial to identify the growth stage at which exogenous application of proline is both safe and effective.

3.3. Plant growth regulators Plant growth regulators (PGRs) influence plant development in many ways and the response of plants to abiotic stress involves complex physiological and biochemical responses, including changes in the concentration and ratios of endogenous PGRs (Wang et al., 2005). Although most of the known PGRs are upregulators of growth, maximum plant growth may depend on the ratio rather than the absolute levels of these substances in plants. These ratios may change dramatically in response to salinity stress, thereby leading to physiological disturbances associated with a general reduction in growth (Lerner and Amzallag, 1994). The synthesis of PGRs in plants generally decreases under saline environments and they may also undergo degradation (Kuiper et al., 1988). Thus, application of PGRs such as auxins, GA, and cytokinins to the seeds at the time of sowing or directly to the growing plants may make up for their deficiency. A number of natural or synthetic PGRs have been applied exogenously to crops to improve their salt tolerance. For example, a foliar spray of IAA increased the grain yield and yield components in a wheat cultivar, Shahkar-92 (Hegazi et al., 1995), and in rice (Kaur and Singh, 1987) grown under saline conditions (Table 2). Gibberellins can also alleviate the adverse effects of salt stress on plants. Ashraf et al. (2002) found that GA treatment stimulated the vegetative growth of two wheat cultivars under salt stress but caused a slight reduction in their grain yield. GA treatment enhanced the accumulation of Naþ and Cl in both shoots and roots of wheat plants under salt stress. It also caused a significant increase in photosynthetic capacity in both lines at the vegetative stage in both saline and nonsaline media. Cytokinins are known to play a significant role in almost all aspects of plant growth and development. Of the many synthetic cytokinins, kinetin (6-furfurylaminopurine) has been extensively researched because of its active role in promoting cell division and enlargement. While assessing the role of kinetin on growth and grain yield of wheat grown under salt stress, Gadallah (1999) found that kinetin application ameliorated the deleterious effects of salinity by reducing the uptake of the toxic ions Naþ and Cl and promoting that of Kþ. Another important class of plant hormones, the PAs such as Spd and Spm and their obligate precursor Put, are polybasic amines that play a considerable role in a number of physiological processes in plants (Galston and Kaur-Sawhney, 1995; Kumar et al., 1997; Tiburcio et al., 1993).

58 Table 2 Exogenous application of different organic osmolytes or plant growth regulators on growth of various crops grown under saline substrate Organic compound

Crop

Character improved

Reference

Glycine betaine

Tomato

Fruit yield

Rice

Improved shoot growth but not root growth; caused low accumulation of Naþ and high of Kþ Destabilizing effects on photorespiration

Ma˚kela et al., 1998a,b Rahman et al., 2002

Proline

Canola and Arabidopsis Soybean Allenrolfea occidentalis Barley Rice

IAA

Wheat Rice

Sulpice et al., 2002

Increased the activities of superoxide dismutase and peroxidase in cell lines Neutralized the enhanced production of ethylene

Yan et al., 2000; Hua and Guo, 2002 Chrominski et al., 1989

Improved growth by causing low accumulation of Naþ and Cl Improved early-seedling growth, thereby enhancing Kþ/Naþ ratio Increased grain yield Increased grain yield

Lone et al., 1987 Roy et al., 1993 Hegazi et al., 1995 Kaur and Singh, 1987

GA3

Wheat

Kinetin

Wheat

Polyamines Salicylic acid

Barley Barley Tomato Barley Sunflower

Jasmonic acid Thiamin 5-Aminolevulinic acid

Cotton

Stimulated the vegetative growth but reduced grain yield Improved growth and caused reduction in Naþ and Cl uptake Improved growth by stabilizing root tonoplast integrity Improved growth Improved growth and photosynthetic capacity Improved growth and photosynthetic capacity Improved uptake of Kþ, increased leaf relative water content, chlorophyll content, and dry mass production Increased seedling growth

Ashraf et al., 2002 Gadallah, 1999 Zhao and Qin, 2004 El-Tayeb, 2005 Stevens et al., 2006 Tsonev et al., 1998 Sayed and Gadallah, 2002 Tanaka and Kuramochi, 2001

59

60

M. Ashraf et al.

Because of their polycationic nature at a physiological pH, PAs occur in plant cells not only as free forms but also as bound forms. Their ionic interactions are considered to increase the stabilization of the subcellular compounds and membranes under stress environments (Roberts et al., 1986). There is a growing interest in uncovering the role of PAs in the defense reactions of plants to various stresses (Bouchereau et al., 1999; Flores, 1990; Galston, 2001; He et al., 2002; Kumar et al., 1997). There are many reports that show that stress tolerance of plants is correlated with their capacity to enhance the synthesis of PAs under stress conditions (Bouchereau et al., 1999; Chattopadhyay et al., 1997; Kasinathan and Wingler, 2004; Kasukabe et al., 2004; Mo and Pua, 2002). Which of the three PAs plays vital roles in stress responses of plants may depend on the plant species and the type of stress (Kasukabe et al., 2004). In most cases, Spd is more closely associated with stress tolerance of plants than either Put or Spm (He et al., 2002; Li and Chen, 2000; Martıńez-Te´llez et al., 2002; Shen et al., 2000). Even if PAs accumulate, this does not mean that they are involved in stress protection, particularly as the role of PAs may depend on their cellular localization and whether they are free, bound to proteins, or conjugated to phenolic acids (Bouchereau et al., 1999). However, a correlation between stress tolerance and PA levels has been reported in a number of studies using a variety of plant species, although the physiological basis for stress-induced PA accumulation remains unknown. The function of PAs is believed to be protective, with a major role in scavenging free radicals (Mansour, 2000). All these reports show that the individual PAs may have different roles during the response of plants to salt stress. While examining the role of exogenously applied PAs in salt-stressed barley, Zhao and Qin (2004) found that the ameliorative effect was associated with their role in stabilizing root tonoplast integrity and function. The NaCl-induced reduction in the content of phospholipids and PAs in tonoplast vesicles isolated from barley seedling roots as well as the activities of Hþ-ATPase, HþPPase, and vacuolar Naþ/Hþ antiport were all partially restored by the application of 0.5 mM Put and 0.5 mM Spd; the effect of the former being more pronounced. In rice, Ndayiragije and Lutts (2006) observed that exogenous application of Put to salt-treated callus promoted its growth by reducing the accumulation of Naþ and Cl in the growing tissue. In another in vitro study, Tang and Newton (2005) investigated the effect of PAs on salt-induced oxidative stress in callus cultures and plantlets in Virginia pine (Pinus virginiana). The PAs reduced salt-induced oxidative damage by increasing the activities of antioxidant enzymes and decreasing lipid peroxidation. Among the PAs used in this study, Put was found to be more effective than Spd or Spm in increasing the activities of ascorbate peroxidase, glutathione reductase (GR), and SOD, reducing the activities of acid phosphatase and V-type Hþ-ATPase, and decreasing lipid peroxidation.


61

Salicylic acid, a phenolic compound, plays a vital role in the plant response to adverse environmental conditions such as salt and osmotic stresses (Borsani et al., 2001; Senaratna et al., 2000). Recently, it has been observed that SA applied exogenously to barley plants is effective in ameliorating the adverse effects of salt stress (El-Tayeb, 2005). In maize, Khodary (2004) found that the application of 0.1 mM SA to plants under saline conditions enhanced their growth and development. Similarly, exogenous application of SA as a foliar spray counteracted the deleterious effects of NaCl on wheat (Sakhabutdinova et al., 2003). In the same crop, application of 100 mg liter1 SA proved to be effective in alleviating the adverse effects of salt stress on wheat seedlings (Hamada and Al-Hakimi, 2001). In tomato, the application of 0.1 mM SA to tomato plants as a root drench provided protection against salt, improving survival, relative growth rate, and photosynthetic capacity ( Jason et al., 2006). Another important plant hormone class is jasmonic acid and its methyl esters, which are ubiquitous in plants and involved in regulation of growth and development. Jasmonic acid applied exogenously to salt-stressed barley seedlings improved their salt tolerance and enhanced their photosynthetic rate (Tsonev et al., 1998). However, because there are, as yet, very few reports in the literature on the effect of exogenous application of jasmonic acid on crops exposed to salt stress, it is not possible to draw general conclusions on a possible role of this hormone in mitigating the deleterious effects of salt stress on crop growth, or on its potential practical application. In addition to the PGRs discussed above, a range of other organic compounds have also been applied exogenously to test their effect on the response of plant to saline conditions. For example, Sayed and Gadallah (2002) applied thiamin (vitamin B1, an antioxidant member of the B complex) as a foliar spray or through the rooting medium to salt-stressed sunflower plants (Table 2). They reported that salt stress significantly reduced the rate of growth and dry biomass accumulation, lowered leaf relative water content (RWC) and leaf and root water potential, and decreased chlorophyll content, soluble sugars, and the Kþ/Naþ ratio in plant tissue. However, application of thiamin improved uptake of Kþ, and increased leaf RWC, chlorophyll content, soluble sugars, and dry mass production. Another organic compound that is receiving attention as an ameliorating agent is 5-aminolevulinic acid (5-ALA), which is a key precursor in the biosynthesis of porphyrins such as chlorophyll and heme. It has been reported that this compound elicits useful responses in crops at low concentrations, ranging from 30 to 100 mg liter1, such as improvement in dry biomass, promotion of photosynthetic activity, and inhibition of respiration (Yoshida et al., 1996a,b, 2003). Tanaka and Kuramochi (2001) found that exogenous application of 100 mg liter1 5-ALA increased salt tolerance of young cotton (Gossypium hirsutum) seedlings (Table 2). Nishihara et al. (2003) applied 5-ALA to the leaves of spinach (Spinacia oleracea) plants

62

M. Ashraf et al.

grown under saline conditions of 50 or 100 mM NaCl. They found that 0.60 and 1.80 mM 5-ALA significantly increased the photosynthetic rate of plants grown in 50 and 100 mM NaCl, while 0.18 mM 5-ALA had no effect at 50 mM NaCl and caused a slight decrease in photosynthetic rate at 100 mM NaCl. These reports suggest that the response of plants to 5-ALA is likely to be both dose- and species-dependent. The above discussion clearly shows that application of PGRs to many crops can at least partially overcome the adverse effects of a saline growing medium. In many cases, the mechanisms involved remain unclear and are not necessarily directly related to the normally perceived physiological role of the PGR. It is unclear whether exogenous application of PGRs operates to reverse the inbalance of PGRs that undoubtedly develops during salinity stress (Debez et al., 2001; Gadallah, 1999; Khan et al., 2000), triggers specific plant stress defense mechanisms, or simply improves plant vigor. Nonetheless, it is now evident that application of PGRs offers a potential practical means to allow crops to grow in more saline environments than is currently the case and to increase the yield of crops in existing environments. PGRs can be applied exogenously by presowing seed treatment, through the growth medium, or as a foliar spray. The application of PGRs through the rooting medium (soil) on a field scale is unlikely to be feasible because the PGRs added to the soil may undergo degradation by soil microbes (Arshad and Frankenberger, 2002). Furthermore, the addition of a PGR to soil at a sufficient application rate to ensure adequate uptake is likely to be prohibitively expensive. Foliar application of PGRs seems to be a more realistic approach to attain enhanced crop growth and yield under saline conditions because it requires a lower application rate than with soil-applied PGRs. In theory, presowing seed treatment would be the most costeffective treatment, but there may be problems in ensuring a sufficient and continuing endogenous concentration to be effective at critical stages during the crop growth and development.

3.4. Supply of nutrients through the rooting medium Of a number of factors responsible for salt-induced inhibition in plant growth, nutrient deficiency is the most crucial factor that reduces plant growth and crop productivity because both macro- and micronutrients are important constituents of enzymes, hormones, and cellular structures. However, nutrient uptake by the plants from soil is influenced by the activity of membrane transporters that mediate their intra- and inter-cellular distribution (Epstein and Bloom, 2005; Marschner, 1995; Tester and Davenport, 2003), for example, Ca2þ-ATPases (Ca2þ-pumps) for Ca2þ (Geisler et al., 2000), inward- and outward-rectifying Kþ channels for Naþ and Kþ (Maathuis and Amtmann, 1999), HVSTI for SO42 (Smith et al., 1997), and different kinds of transporters for nitrate and ammonium (Epstein and Bloom, 2005).


63

In view of a number of reports, it is suggested that salt-induced nutrient deficiency may have been due to Naþ-induced blockage or reduced activity of these transporters (Maathuis and Amtmann, 1999; Qi and Spalding, 2004; Rus et al., 2004; Tester and Davenport, 2003). However, salt-induced nutritional disorders can be alleviated by the addition of mineral nutrients in the growth medium. For instance, addition of N improved the growth and/or yield of maize (Ravikovitch, 1973), tomato (Papadopoulos and Rendig, 1983), grapes (Taylor et al., 1987), and apple (El-Siddig and Ludders, 1993). Long ago, Champagnol (1979) reported in a comprehensive review that 34 of 37 different studies show that addition of P to saline soil increases crop growth and yield. While working with tomato, Awad et al. (1990) demonstrated that addition of P increased crop salt tolerance over a wide range of salinity (10– 100 mM NaCl). Like P, calcium is known to ameliorate the adverse effects of salinity in plants, presumably by facilitating higher Kþ/Naþ selectivity (Hasegawa et al., 2000). For example, Ebert et al. (2002) found that rootapplied 10-mM Ca2þ as Ca (NO3)2 counteracted salt-induced growth inhibition. This ameliorative effect of Ca2þ on growth was suggested to be due to increased leaf Kþ and Ca2þ coupled with decrease in leaf Naþ. In another pot culture experiment, Arshi et al. (2006) found that supplemental 10 mM Ca2þ increased the photosynthetic capacity via stomatal regulation and hence growth in Cassia angustifolia. While working with mung bean, Misra et al. (2001) found that supplemental Ca2þ applied as CaCl2 improved the chlorophyll fluorescence that indicates that plant can protect its photosynthetic apparatus from salt-induced damaging effects in the presence of Ca2þ. Likewise, Navarro et al. (2000) found that the growth promoting effect of supplemental Ca2þ on salt-grown tomato was due to the ameliorative effect of Ca2þ on root hydraulic conductivity as well as water uptake. Recently, Tuna et al. (2007) found that addition of CaSO4 to a saline nutrient medium helped in maintaining membrane permeability and thus increasing concentrations of Ca2þ, N, and Kþ and reducing concentration of Naþ, which could offer an economical and simple solution to tomato crop production under saline conditions. Furthermore, Shabala et al. (2006) demonstrated using a noninvasive ion flux measuring and patch-clamp techniques that supplemental Ca2þ could ameliorate Naþ toxicity in plants by reducing Naþ influx through nonselective cation channels as well as by inhibiting Naþinduced Kþ efflux through outwardly directed Kþ-permeable channels. Furthermore, they found that NaCl-induced Kþ efflux was partially inhibited by 1 mM Ca2þ and fully prevented by 10 mM Ca2þ. Similarly, deleterious effects of salt stress associated with reduced uptake of Kþ can be alleviated by the addition of Kþ to the growth medium in many crops, for example, tomato, maize, sunflower, beans (Grattan and Grieve, 1999). In view of these reports, it is suggested that high level of macronutrients can be supplied exogenously so as to alleviate the adverse effects of salt stress on plant growth.

64

M. Ashraf et al.

3.5. Foliar application of macronutrients Generally, mineral nutrients such as N, P, and K are taken up through the roots by plants (Marschner, 1995). However, some nutrients are absorbed in the leaves via stomata (Taiz and Zeiger, 2002). Uptake of mineral nutrients through leaves is restricted due to the presence of a cuticle on the epidermis. By eliminating this barrier, nutrients can be supplied through leaves. Thus, the effectiveness of foliar application of mineral nutrients depends on rate at which foliarly applied nutrients are absorbed with a subsequent translocation within the leaves (Chamel, 1985). Generally, foliar fertilization is a useful technique to supplement nutrients that are deficient in plants when soil conditions restrict nutrient availability to roots (Mengel, 2002). Under saline conditions, uptake of Kþ, Ca2þ, and N through the roots and its supply to the growing regions of shoots is considerably impaired (Ashraf, 1994b, 2004; Grattan and Grieve, 1999; Marschner, 1995; Munns, 2005) and thus optimal concentration of essential nutrients by the plant is decreased (Marschner, 1995). In this case, foliar fertilization can be used promote mineral nutrition, suppress physiological disorders, and promote plant growth and yield (Alexander, 1985; Howard et al., 1998; Pervez et al., 2004). The efficacy of foliarly applied nutrients depends on their mobility within the plant (Mengel, 2002). Of all nutrients, N, K, and Mg are more phloem-mobile, whereas Ca and Fe are least phloem-mobile. Thus, foliar application of the later two nutrients is not effective (Mengel, 2002). Leaves easily absorb and transport the phloem-mobile nutrients. Potassium is known to play a vital role in various physiological processes such as meristematic growth, enzyme activation, cation/anion balance, stomatal movement, osmoregulation, photosynthesis, and protein synthesis (Epstein and Bloom, 2005). Thus, foliar application of Kþ can be used to ameliorate the adverse effects of salt stress on plant growth and yield. For example, foliar spray of KH2PO4 corrected the deficiencies of both P and Kþ in salt-stressed strawberry (Fragaria ananassa Duch cvv. Oso Grande) (Kaya et al., 2001a) and tomato (Satti and Al-Yahyai, 1995). Kaya et al. (2001b) found that foliar spray of KH2PO4 alleviated the inhibitory effects of salinity on growth of spinach under saline conditions by correcting Nainduced Kþ deficiency as well as improving Kþ/Naþ ratio from 1.61 to 2.72 in plants. Recently, Ahmad and Jabeen (2005) have found that 250 mg liter1 solution of KNO3 not only inhibited toxic effects of salt on fruit formation in Lagenaria siceraria grown under saline conditions but also increased its production by 76.9% with respect to weight per plant. While working with sunflower, Akram (2006) found that foliar application of varying levels of Kþ resulted in increased tissue Kþ contents of salt-stressed plants, which in turn promoted plant growth and yield. However, effect of exogenous application of different potassium sources on leaf or root Kþ was significantly variable.


65

It is also necessary to understand how stress affects foliar nutrient levels and how foliar nutrient levels affect stress resistance mechanisms. The role of plant nutrients on stress resistance mechanisms has gained a significant recognition and opened new avenues for use of foliar applied nutrients to counteract the deleterious effects of stresses. Antagonism is an important feature of ion interaction during uptake whereby reduction of cation uptake by other cation takes place for maintaining charge balance within the cells (Marschner, 1995). Akram (2006) found that exogenously applied Kþ as a foliar spray did not improve the accumulation of macronutrients under saline conditions. However, he found that exogenous application of Kþ caused more partitioning of Naþ in roots, indicating that foliar application of Kþ restricts the Naþ translocation from root to shoot and thus protects photosynthetic tissue. This view was supported by the fact that Kþ applied as a foliar spray caused an increase in chlorophyll a and PSII efficiency (improved Fv/Fm values), and membrane permeability under saline conditions (Akram, 2006). It was suggested that Kþ can be applied foliarly to correct salt-induced Kþ deficiency and thus may be useful either in promotion of acclimation and resistance of plants to salinity or in enhancing repair mechanisms once damage has occurred (Akram, 2006). It is generally believed that leaf-applied Kþ is readily absorbed; however, this variability in accumulation of Kþ in the leaves may have been due to differential chemical properties of different Kþ sources because absorption rate of mineral nutrients by leaves strongly depends on chemical properties and valency of a cation (Mengel, 2002), diameter of cation (Franke, 1967), nature and chemical properties of accompanying anion (Mengel, 2002). Furthermore, foliar uptake of a nutrient depends on the type of plant species and its epidermal structure and composition (amount of wax deposited or type of cuticle), growth conditions (leaf age), type of potassium source used, concentration of Kþ applied at the leaf surface, pH of applied Kþ solution, and absorption of Kþ in leaves with subsequent utilization by the plant (Mengel, 2002; Wo´jcik, 2004). Although N and Mg among macronutrients are also phloem-mobile and their foliar application promotes the growth under restricted nutrient supply (Mengel, 2002), there is no report available in the literature on foliar application of N-containing inorganic compounds alleviating the adverse effects of salt stress.

4. Breeding for Salt Tolerance Although resistance to salinity stress is a complex phenomenon at both the whole plant level and the cellular level, involving the interaction of stress with molecular, biochemical, and physiological processes at different

66

M. Ashraf et al.

stages of plant development, development of crop plants tolerant to this stress is vital to meet the growing food demand through sustainable agriculture. There has been some progress in achieving this goal through conventional breeding. Ashraf (1994a) listed a few salt-tolerant lines/cultivars of different crops that had been developed through conventional breeding. The main challenge confronting conventional plant breeders is the magnitude of genetic variation, which is very low in the gene pool of most crop species. As a result, major improvements in salt tolerance cannot be expected. Utilizing the wild relatives of crop plants as a source of genes conferring salt tolerance can broaden the range of variation that can be used in crop improvement programs. However, incorporating salt-tolerant genes from wild relatives into domesticated crops is difficult because of reproductive barriers and there are very few examples of effectiveness of this approach in the literature. In fact, conventional breeding has had only limited success in improving crop salt tolerance for a number of reasons (Chinnusamy et al., 2005), including: (1) it is time-consuming and laborintensive, (2) undesirable genes are often transferred along with desirable traits, and (3) reproductive barriers restrict transfer of favorable alleles from interspecific and intergeneric sources.

4.1. Mutation breeding In addition to conventional methods, modern techniques including somaclonal variation, protoplasmic fusion, and mutation breeding may also contribute to the development of salt-tolerant varieties if the traits for salt tolerance do not exist or if the genetic variability for specific traits is absent (Ashraf, 1994a). The prime strategy in mutation-based breeding is to induce or alter one or two major traits through mutagens, which may be chemical or radiation. Mutations cause enhancement or reduction in traits that may result in well-adapted plant varieties. Induced mutations play a considerable role in creating crop varieties with traits such as heavy metal stress tolerance and resistance to drought and salinity as a major component of environmentally sustainable agriculture (Ahloowalia et al., 2004). Induced mutations have been used to improve major crops such as wheat, rice, barley, cotton, peanuts, and beans, which are seed propagated. For inducing variation in major traits that enhance the quality as well as quantity of crop yield, the use of ionizing radiation, such as X-rays, g-rays, and neutrons and chemical mutagens, is well established. For example, a research group from the Nuclear Institute for Agriculture and Biology (NIAB), Faisalabad, Pakistan developed rice mutants using g-radiation. Of these, two mutants NIAB Rice-1 and PSR 1–84 have shown greater yield than their respective salt-tolerant check cultivars, Pokkali and Johna 349, under saline conditions (Anonymous, 1986). Likewise, a research group from the Scottish Crop Research Institute, working on the effects of semidwarfing genes on salt


67

tolerance in barley in 1993 found that Golden Promise, a g-ray-induced semidwarf mutant of the cultivar ‘‘Maythorpe,’’ showed considerable tolerance to salinity stress. It is assumed that both the parent and mutant cultivars are isogenic. However, responses of both parent and mutant cultivars were different to salt stress. The positive and pleiotropic effects of the mutant gene, commonly known as GPert, were found to be effective in a number of genetic backgrounds (Forster, 2001). It has been shown that the GPert mutation was allelic to the ari-e mutants in barley. The ari-e mutants were salt tested and found to show the same positive responses to salt stress as Golden Promise, which indicates that the GPert symbol was changed to ari-e.GP. However, the semidwarf mutant sdw1 (also known as denso) and the erectoides semidwarf mutant, ert-k32, were not salt tolerant. These findings led the scientists to conclude that salt tolerance was therefore not a general phenomenon of semidwarf stature but specific to mutations at the ari-e locus in these lines (Forster, 2001). While working with rice, Saleem et al. (2005) found that growth and number of adapted mutagenized callus of indica rice (O. sativa) cv. Basmati 370 (irradiated with 50 Gy of g-rays of 60Co for creating genetic variability against salinity) was more than that of nonmutagenized callus under saline conditions. Furthermore, irradiated callus showed 2.0–4.75% regeneration frequency while nonmutagenized salt-adapted callus did not show any regeneration. They also found two putative lines (M2 generation) with moderate salt tolerance from g-ray mutagenized cultures at the seedling stage. The use of chemical mutagens in developing salt-tolerant rice mutants has also been shown to be successful. For example, Ashraf (1984) induced variability for salt tolerance in the salt-sensitive rice cultivar Taichung 65 by treating fertilized egg cells with varying doses of N-methyl-N-nitrosourea. In M3, two salt-tolerant mutants were detected that had 83% and 90% survival at the seedling stage in 0.5% NaCl. Mutants of Arabidopsis developed through insertional mutagenesis are already being used to analyze genes that determine response to auxins, cytokinins, GA, abscisic acid, and ethylene in plant growth, floral development and senescence, fruit formation, and ripening. These mutants are facilitating the isolation, identification, and cloning of the genes (Ahloowalia and Maluszynski, 2001). Insertion mutagenesis is very labor and cost intensive, whereas radiation and chemical mutagenesis are more efficient but the mutations are anonymous. From all these reports, it has been suggested that induced mutations through radiations such as g-rays may be used as a versatile approach to enhance salt tolerance in crop plants. Genetic markers (RAPD, AFLP, and SSR) that have been routinely used in fingerprinting, genetic mapping, and quantitative trait loci (QTL) analysis may help in identifying useful mutations (Lema-Rumin´ska et al., 2004; Mlcˇochova´ et al., 2004) induced by chemical mutagens or radiations and may help in examining its physiological bases, which will bring a new

68

M. Ashraf et al.

dimension in gene technology. These in turn may lead to a rapid enhancement of crop yield in salt-affected areas.

4.2. Identification of traits using advanced molecular techniques Over the last two decades, advances in molecular marker technology have led to the development of detailed molecular linkage maps for many plant species. DNA-based markers speed up the advance in the improvement of crops for stress tolerance (Vinh and Paterson, 2005). Undoubtedly, DNA markers offer an important means to examine the degree to which adaptations to stress are related to genetic differences in crop productivity under stress. Furthermore, DNA markers can be used as a diagnostic tool to identify genotypes in large populations that bring together superior genetic potential for productivity under stress, with the traits that are necessary to develop a potential commercial cultivar (Vinh and Paterson, 2005). Because of the complexity of abiotic stress tolerance and the difficulty in phenotypic selection, the approach of QTL mapping has become crucial to the use of DNA markers in the improvement of stress tolerance. QTL mapping is a way to estimate the locations, numbers, extent of phenotypic effects, and modes of gene action, and of individual determinants that substantially contribute to the inheritance of continuously variable traits (Vinh and Paterson, 2005). The principal objective is to isolate genetic signals emerging from an individual locus, the background collective effects of nongenetic factors, as well as measurement errors in evaluation of continuous traits. QTL mapping is thus an efficient means for identifying specific components that allow direct assessment of stress tolerance. Several different types of DNA markers are currently in use to examine the inheritance of abiotic stress tolerance including RFLPs, RAPDs, CAPS, PCRindels, AFLPs, microsatellites (SSRs), SNPs, and DNA sequences. Each type of marker has advantages and disadvantages. ‘‘Transposon display’’ is the latest development in DNA markers for use in crop improvement for stress tolerance (Casa et al., 2000; Van den Broek et al., 1998). This enables the researcher to select a sample that is likely to represent near-complete coverage of the genome, or possibly even to target specific chromosomes if warranted (Casa et al., 2000). This method has a special promise because the priming sequence itself may be an effective agent of genetic change, and in some cases is preferentially associated with genes (Zhang et al., 2000). Successful plant breeding undoubtedly depends on a broad understanding of the genetic architecture of relevant traits. Genes with major effects and genes with minor effects play a significant role in controlling abiotic stress tolerance. Although major gene and polygenic variation has been analyzed using conventional means, the use of DNA markers and the techniques of QTL analysis now permit a more integrated approach in


69

dissecting complex traits and assessing gene effects (Humphreys and Humphreys, 2005). QTL analysis of valuable traits has been carried out in a number of crops including rice, wheat, maize, sorghum, barley, forage and turf grasses, Brassica, soybean, pea, alfalfa, beans, tomato, and beet (Duncan and Carrow, 1999; McCouch and Doerge, 1995; Mohan et al., 1997). As an alternative to conventional breeding, engineering can relatively easily be employed to achieve gene transfer across reproductive barriers. Genetic engineering offers a way to create new genetic variation when natural allelic variation may be limited or inappropriate (Humphreys and Humphreys, 2005). Thus, two fundamental genetic approaches are currently being used to improve crop stress tolerance (Yamaguchi and Blumwald, 2005): (1) utilization of natural genetic variation, either through direct selection in stressful conditions or through the mapping of QTLs and subsequent marker-assisted selection, and (2) production of transgenic plants by introducting the novel genes or by modifying the expression of the existing genes to alter the degree of stress tolerance. Dissection of complex salt tolerance trait by means of QTL mapping and identifying chromosomal regions associated with DNA markers can be a useful means for large-scale screening of genotypes.

4.3. Molecular breeding for salt tolerance 4.3.1. QTLs and marker-assisted breeding QTLs and marker-assisted selection offer several advantages over direct phenotypic screening, inasmuch as the PCR-based techniques used to identify the markers reduce the time needed to screen genotypes as well as reduce the environmental impact on the trait under study. Convincing evidence exists that salt tolerance is a complex trait and that both additive and dominance effects are important in the inheritance of many of the traits associated with salt tolerance (Ashraf, 1994a; Foolad, 2004; Yamaguchi and Blumwald, 2005). For example, 38 traits related to tissue or whole-plant Naþ or Cl accumulation were analyzed in salinized and nonsalinized BC1 progeny clones from a Poncirus Citrus cross (Tozlu et al., 1999). QTLs were mapped for traits associated with salinity tolerance and 17 regions of interest of the citrus genome were identified of which 8 contained QTLs of great effect. In rice, QTLs associated with various traits conferring salt tolerance in seedlings were mapped on the molecular map (Lee et al., 2006; Prasad et al., 2000a). Seven QTL were identified for seedling traits under salt stress, comprising two for seed germination, one for seedling root length, three for seedling dry matter, and one for seedling vigor. Of the seven QTLs, four were located on chromosome 6 (Prasad et al., 2000a). Recombinant lines between cv. ‘‘Milyang 23’’ and cv. ‘‘Gilhobyeo’’ produced two QTLs, qST1 and qST3, related to salt tolerance (Lee et al., 2006).

70

M. Ashraf et al.

QST1, detected on chromosome 1 and flanked by Est12RZ569A, is major QTL explaining 27.8% of the total phenotypic variation. It is widely accepted that degree of salt tolerance varies during the plant life cycle (Ashraf and Harris, 2004; Flowers, 2004; Foolad, 2004). For example, in tomato, barley, corn, rice, and wheat, the degree of salt tolerance tends to increase with plant development (Foolad, 2004). QTLs associated with salt tolerance at the germination stage were different from those associated with salt tolerance at the early growth stage in barley (Mano and Takeda, 1997), tomato (Foolad, 1999), and Arabidopsis (Quesada et al., 2002). In the same crops, the plants selected on the basis of their germinability at high salinity did not show similar degree of salt tolerance at the vegetative stage. Advances in molecular biology have led to the development of DNA markers that can be used to identify QTLs. In the last two decades, advances in molecular marker technology have allowed the development of detailed molecular linkage maps and considerable progress has been made in markerassisted selection procedures that in future will permit pyramiding desirable traits to achieve considerable improvements in crop salt tolerance. 4.3.2. Improvement of crop salt tolerance using transgenic approaches Transgenic crops are bioengineered crops that possess a gene or genes that have been inserted by humans into their genome using modern biotechnology. The inserted gene sequence, known as the transgene, may belong to an unrelated plant, or even to a bacterium or animal. Crops that contain these transgenes are also described as genetically modified and are often referred to as genetically modified organisms. This approach is now being intensively used by plant scientists to improve qualitative and quantitative traits including tolerance to biotic and abiotic stresses in a number of crops. As described above, plant responses to salt stress are complex and involve many genes. Under these circumstances, the possibility of improving crop salt tolerance by genetic modification is rather difficult and slow. With respect to the behavior of plants in saline environments, two classes of plants are generally known, halophytes (salt tolerant) and glycophytes (salt sensitive) (Ashraf, 2004; Flowers et al., 1977; Maas and Nieman, 1978). Halophytes are able to maintain high turgor pressure by accumulation of salt (Flowers et al., 1977). These plants sequester and accumulate salt into the cell vacuoles, controlling the salt concentrations in the cytosol and maintaining a high cytosolic Kþ/Naþ ratio in their cells (Glenn et al., 1999). In glycophytes in which salt exclusion is the major mechanism of salt tolerance, the accumulation of organic solutes or inorganic ions is increased to achieve osmotic balance (Ashraf, 1994b; Greenway and Munns, 1980; Wyn Jones, 1981). Advances in molecular biology have led the identification of a large number of genes that are induced as a result of drought or salinity stress


71

(Lea et al., 2004). However, the major focus of researchers is on the genes that encode: (1) compatible osmotica (e.g., GB, proline, sugars), (2) transcription factors, (3) enzymatic and nonenzymatic antioxidants, (4) ion transport proteins, and (5) heat-shock and late embryogenesis abundant (LEA) proteins. A large number of studies are reported in the literature in which transgenic lines with resistance to salt stress have been produced (Table 3). As discussed above, overproduction of compatible organic solutes is one of the most common responses of plants grown under stress conditions (Serraj and Sinclair, 2002). The genes responsible for the synthesis of such organic compounds can be engineered to overproduce these compounds in transgenic plants. For example, the plastid-expressed betaine aldehyde dehydrogenase (badh) gene was reported to confer improved salt tolerance in carrot plants (Kumar et al., 2004). In another study, transgenic tomato plants showed significantly higher levels of mRNA and BADH enzyme activity than wild-type plants and exhibited tolerance to salt stress up to 120 mM NaCl ( Jia et al., 2002). While inducing stress tolerance by engineering plants for compatible solutes, Chen and Murata (2002) found that the level of GB in transformed plants was not more than 5 mM in leaves, suggesting that GB has effects other than simply as an osmolyte. In rice, transgenic plants overexpressing peroxisomal BADH exhibited enhanced ion selectivity under saline conditions by accumulating high amounts of Kþ but low amounts of Naþ and Cl (Kishitani et al., 2000). However, it is not yet clear how GB overproduction increases the Kþ/Naþ ratio, which is regulated by ion transporters and channels. The first direct evidence for regulation of ion fluxes across the plasma membrane by physiologically low concentrations of compatible solutes was provided by Cuin and Shabala (2005), when they exogenously supplied low (0.5–5 mM) concentrations of proline or betaine that significantly reduced NaCl-induced Kþ leakage from barley roots in a dose–response manner. Exogenously supplied betaine enhanced NaCl-induced Hþ efflux thereby creating a pH gradient, which is beneficial to plants under saline conditions. Furthermore, maintenance of cytosolic Kþ homeostasis may have been because of the GB-induced enhanced activity of Hþ-ATPase (Shabala, 2000; Shabala and Newman, 2000), controlling voltage-dependent outward-rectifying Kþ channels and generating the electrochemical gradient necessary for secondary ion transport processes. Salt stress has been shown to stimulate the activity of vacuolar and plasma membrane H-ATPase activity (Ayala et al., 1996, 1997; Nakamura et al., 1992). The resulting acidification of the external solution is proposed to be important in providing the driving force for a plasma membrane Naþ/Hþ exchanger to transport Naþ from the cytoplasm into the apoplast (Ayala et al., 1996). Thus, elevated levels of compatible solutes such as proline and betaine in

72

Table 3

Salt-tolerant genes transformed in different crop species

Group

Gene and its origin

Crop

Attribute improved

References

(1) Compatible osmotica (i) Glycine betaine

bet A from Escherichia coli BADH from Escherichia coli

Tobacco

Improved biomass

Rice

Enhanced Kþ/Naþ ratio

Carrot

Improved growth

Tobacco

Enhanced growth and better photosynthetic activity

BADH from Atriplex

Tomato

Cod A from Arthrobacter globiformis

Arabidopsis thaliana Rice Brassica juncea

Reduced ion leakage and enhanced growth Increased growth and photosynthetic capacity Increased growth and photosynthetic capacity Improved growth

Lilius et al., 1996 Kishitani et al., 2000 Kumar et al., 2004 Liang et al., 1997; Holmstrom et al., 2000 Jia et al., 2002

Tobacco

Enhanced growth

Rice

Increase in growth

(ii) Proline

P5CS (pyrroline-5carboxylate synthase) from Vigna aconitifolia

Hayashi et al., 1997 Sakamoto et al., 1998 Prasad et al., 2000b Kishor et al., 1995 Zhu et al., 1998

Reduced oxidative stress and better germination Improvement in growth and yield

Hong et al., 2000 Hmida-Sayari et al., 2005

Arabidopsis thaliana

High turgor maintenance

Nanjo et al., 1999

Arabidopsis thaliana, Tobacco, and Brassica napus Tobacco

Improved growth

Huang et al., 2000

Moghaieb et al., 2006

Arabidopsis

Improved root growth, enhanced transpiration, improved photosynthesis, improved N, and increased Na and K Germination and growth

Tobacco

Germination and growth

Tobacco P5CS (pyrroline-5carboxylate synthase) from Arabidopsis thaliana ProDH (proline dehydrogenase) antisensense cDNA from Arabidopsis thaliana COX from Arthrobacter pascens

(iii) Ectoin

Hmect. ABC

(iv) Mannitol

mt1D (mannitol-1phosphate dehydrogenase) from Escherichia coli

Potato

Tarczynski et al., 1993 Thomas et al., 1995 (continued)

73

74

Table 3

(continued)

Group

(v) Myo-inositol (vi) Tehalsoe

(2) Transcription factors

Gene and its origin

Crop

Attribute improved

References

Egg plant

Better growth

Wheat

Improved growth

Populus tomentosa Carr.

Improved growth, enhanced stomatal conductance, transpiration rate, and photosynthesis; reduced cellular membrane permeability; stunted growth under nonsaline conditions Improved photosynthetic activity

Prabhavathi et al., 2002 Abebe et al., 2003 Hu et al., 2005

IMT1 (myo-ionositol O-methyltransferase) OstA and OstB

Tobacco

TPS and TPP from Escherichia coli OsDREB1A from rice

Rice

Tsil (EREBP/AP2 transcription factor) DREB1A (AP2 transcription factor)

Tobacco

Improved tolerance to salt and pathogens

Arabidopsis

Improved growth and increased tolerance to abiotic stress

Rice

Arabidopsis

Improved growth and reduced photooxidative damage Increased growth and photosynthetic capacity Increased tolerance to high-salt and freezing stresses

Sheveleva et al., 1997 Garg et al., 2002 Jang et al., 2003 Dubouzet et al., 2003; Ito et al., 2006 Park et al., 2001 Kasuga et al., 1999, 2004

Rice

(3) Enzymatic and nonenzymatic antioxidants

Increased tolerance to drought and high salinity Better recovery from stress and improved growth Improved growth and higher level of stress-tolerant transcripts Germination

Mtzpt2–1 from Medicago trunculata CAP2 from Cicer arietinum CpMYB10 from Craterostigma plantagineum Alfin1

Medicago trunculata Tobacco

Alfalfa

Improved growth

STO


Improved growth

FeSOD Glutathione S-transferase/ glutathione peroxidase

Tobacco

Improved antioxidant capacity

Ascorbate peroxidase

Tobacco Tobacco

Reduced oxidative stress Improved photosynthetic capacity

Glutamine synthase

Rice

Reduced photorespiration


Oh et al., 2005 Merchan et al., 2003 Shukla et al., 2006 Villalobos et al., 2004 Winicov and Bastola, 1999 Nagaoka and Takano, 2003 Van Camp et al., 1996; Roxas et al., 1997, 2000 Badawi et al., 2004a Hoshida et al., 2000 (continued)

75

76

Table 3

(continued)

Group

(4) Ion transport proteins

Gene and its origin

Crop

Attribute improved

References

Glyoxylase

Tobacco

Mn-SOD

Rice

Better protection of photosynthetic pigments Reduced oxygen radicles

MSR4 (methionine sulfoxide reductase) AtLDH3 (Aldehyde dehydrogenase) GmALDH7

Arabidopsis

Increased resistance to oxidative damage

Arabodopsis

Reduced lipid peroxidation

Tobacco

Increased germination

Arabidopsis

Increased germination

AVP1

Arabidopsis

Increased accumulation of salt in vacuole

HAL1

Melon

High K/Na ratio

Tomato

High K/Na ratio

Arabidopsis

High K/Na ratio

Arabidopsis

High K/Na ratio

Veena et al., 1999 Tanaka et al., 1999 Romero et al., 2004 Sunkar et al., 2003 Rodrigues et al., 2006 Rodrigues et al., 2006 Gaxiola et al., 2001 Bordas et al., 1997 Gisbert et al., 2000; Rus et al., 2001a Yang et al., 2001 Albert et al., 2000

AtHal3

AtNHX1

Arabidopsis

Improved growth

Tomato

Improved growth and yield

Brassica napus

Improved growth

Wheat

Improved growth and yield

Cotton

Tall fescue

Improved growth and yield, increased fiber with better quality, photosynthesis, higher nitrogen assimilation Improved growth

AgNHX1

Rice

Improved growth

GhNHX1 SsVP

Tobacco Arabidopsis thaliana

SOD2 (Na/H antiporter)

Rice

Improved growth Improved growth, enhanced V-ATPase and V-PPase hydrolytic activity, increased uptake of Naþ (possibly vacuolar compartmentation of Naþ) Increased uptake of K, Ca, and Mg, and reduce Na; enhanced photosynthesis and enhanced P-ATPase hydrolytic activity; reduced reactive oxygen species generation

Apse et al., 1999 Zhang and Blumwald, 2001 Zhang et al., 2001 Xue et al., 2004 He et al., 2005

Tian et al., 2006 Ohta et al., 2002 Wu et al., 2004 Guo et al., 2006

Zhao et al., 2006a

77

(continued)

Table 3 (continued) Group

(5) Heat-shock and late abundant embryogenesis proteins

Gene and its origin

Crop

Attribute improved

References

SsNHX1 and AVP1

Rice

Zhao et al., 2006b

SsNHX1

Rice

AtHsP17.6A DnaK1 from Aphanothece halophytica LEA

Arabidopsis Tobacco

Increased V-PPase hydrolytic activity, higher Kþ/Naþ ratio, higher net CO2 assimilation rate, reduced H2O2 Increased uptake of K, Ca, and Mg; higher Kþ/Naþ ratio, enhanced Fv/ Fm values, higher net CO2 assimilation rate, and soluble sugars; enhanced leaf and root V-ATPase and root P-ATPase hydrolytic activity; reduced ROS generation Improved biomass Reduced Na and better CO2 fixation

Rice

Improved growth

Zhao et al., 2006c

Sun et al., 2001 Sugino et al., 1999 Xu et al., 1996


79

salinized plants may have a protective effect on plasma membrane stability and cation transporters, particularly those directly responsible or associated with Kþ transport across the cell membrane. In general, proline accumulation in response to abiotic stresses including salt stress is found to be correlated with stress tolerance of many plants (Ashraf and Harris, 2004). Genetic engineering has been used to improve plant stress tolerance by overproducing proline. Rice plants engineered to overproduce proline under stress by expressing moth bean (Vigna aconitifolia) [D1]-pyrroline-5-carboxylate synthetase cDNA under the control of ABA responsive element from the barley Hva22 gene showed faster recovery after a short period of salt stress (Zhu et al., 1998). Similarly, tobacco transgenic plants overexpressing the mutated moth bean (Vigna acontifolia) [D1]pyrroline-5-carboxylate synthetase gene accumulated twofold more proline than the transgenics overexpressing normal [D1]-pyrroline-5-carboxylate synthetase gene. This higher accumulation of proline significantly reduced the levels of free radicals, thereby improving tolerance of the transgenic plants to salt stress (Hong et al., 2000). In another analogous study, a pyrroline-5-carboxylate synthetase (P5CS) cDNA from A. thaliana was inserted into potato plants, which then showed a greater stress-induced increase in proline production compared with nontransgenics (HmidaSayari et al., 2005). The transgenic potato plants also show a greater tolerance to salinity in terms of tuber yield and weight than the nontransgenic plants. Arabidopsis plants engineered with an antisense proline dehydrogenase cDNA showed enhanced accumulation of proline (Nanjo et al., 1999). These antisense transgenics exhibited constitutive freezing tolerance (7 C) and were highly tolerant to 600 mM NaCl. The stress tolerance of these transgenic plants was also compared with rd29A:: DREB1A transgenic plants (Kasuga et al., 1999). The rd29A promoterdriven DREB1A transgenic plants showed enhanced tolerance to drought, salinity, and freezing. Although genetic engineering plants to overproduce proline or GB seems to be an effective means of increasing salinity tolerance, some failures of this approach are also reported in the literature (Ashraf and Foolad, 2007). It is also apparent that salt tolerance in plants depends on both the level of accumulation of compatible solutes and their subsequent transportation to target compartments (Chinnusamy et al., 2005). For instance, transgenic plants expressing choline oxidase targeted to chloroplasts exhibited higher resistance to photoinhibition under salt and cold stresses than did transgenic plants with choline oxidase targeted to the cytosol (Sakamoto et al., 1998). Considering the various strategies available to induce salt tolerance in plants, it is suggested that all factors associated with gene regulation at transcriptional and translational levels should be examined while engineering plants for these compatible solutes.

80

M. Ashraf et al.

Trehalose, a nonreducing disaccharide, has recently attracted considerable interest as an organic osmoticum, which plays a vital role in stress tolerance of plants. The development of transgenic rice expressing the trehalose gene (Garg et al., 2002) is a remarkable achievement in that this line showed improved tolerance not only to salinity stress but also to drought and cold stress. Similarly, transgenic rice plants expressing chimeric gene Ubi1::TPSP accumulated a high level of trehalose that resulted in increased tolerance of drought, salt, and cold, as shown by chlorophyll fluorescence and growth inhibition analyses ( Jang et al., 2003). In another study, transgenic tobacco and melon (Cucumis melo) plants expressing yeast trehalose 6-phosphate synthase (TPS1) showed improved growth when subjected to salt stress (Serrano et al., 1999). However, several pleiotropic effects observed in these transgenics led to the suggestion that trehalose affects other plant developmental processes as well. Polyols are sugar alcohols, such as glycerol, mannitol, sorbitol, and D-ononitol, which serve as potential osmoprotectants in many halophytes (Yancey et al., 1982). Transgenic tobacco plants expressing the mt1D gene, responsible for the biosynthesis of mannitol, were found to be tolerant to salt stress (Tarczynski et al., 1993). Likewise, expression of the mt1D gene for the biosynthesis of mannitol in transgenic wheat was found to result in enhanced tolerance of both drought and salinity (Abebe et al., 2003). In ice plant (Mesembryanthemum crystallinum), myo-inositol acts as a substrate for the synthesis of D-ononitol and D-pinitol under saline conditions. D-ononitol is synthesized from myo-inositol by myo-inositol O-methyl transferase. Tobacco plants engineered to overexpress myo-inositol O-methyl transferase showed accumulation of up to 600 mM D-ononitol in the cytosol, and an enhanced rate of photosynthesis under drought and salt stress (Sheveleva et al., 1997). In another study, germination of seeds of transgenic egg plants (Solanum melongena) expressing the mt1D gene was higher under saline conditions (Prabhavathi et al., 2002). In addition, shoot and root lengths and leaf lamina of transgenic lines of egg plants were also higher than those of nontransformed control plants. Ectoine is another important compatible solute and is generally accumulated in halophytic bacteria. In the halophytic bacterium, Halomonas elongata, ectoine is biosynthesized from aspartate b-semialdehyde in three steps catalyzed by three enzymes 2,4-diaminobutyrate aminotransferase, acetyltransferase, and ectoine synthase (Ono et al., 1999). Like other compatible solutes, its introduction into plants can also alleviate adverse effects of salt stress. For example, Moghaieb et al. (2006) transformed tobacco with three genes Hmect. ABC (ect. A, ect. B, and ect. C) isolated from Halomonas elongate, which are responsible for biosynthesis of ectoine. Transgenic tobacco lines accumulated more ectoine than nontransformed control plants when treated with NaCl and showed greater growth than untransformed lines. The greater growth was attributed to an improved water status of the


81

plants, greater N uptake with subsequent translocation from root to shoot, and increased photosynthetic capacity. Because the salinity-induced reduction in photosynthesis was more due to Na-induced metabolic limitations than stomatal limitations, the authors proposed that high accumulation of ectoin protected the photosynthetic apparatus from high accumulation of Naþ in the leaves. ROS such as superoxide radicals (O2), hydrogen peroxide (H2O2), hydroxyl radical (OH), and singlet oxygen (1O2) are produced in aerobic cellular processes when electrons from the mitochondrial and chloroplast electron transport chains leak and react with O2 in the absence of other acceptors (Halliwell and Gutteridge, 1985; Thompson et al., 1987). The ROS can cause considerable oxidative damage to membrane lipids, proteins, and nucleic acids. However, plants can produce various antioxidants and detoxifying enzymes to efficiently scavenge ROS. The various antioxidants used by plants are ascorbate, glutathione, a-tocopherol, and carotenoids, whereas detoxifying enzymes include SOD, catalase, peroxidase, and enzymes of ascorbate-glutathione cycle. However, it is crucial to target enzymes at the site where the stress-induced ROS production takes place for detoxification and hence improved stress tolerance. This indicates that proteins that are damaged by oxidative stress have a significant adverse effect on plant tolerance to environmental stresses. For example, glutathione S-transferase/glutathione peroxidase, one of the major enzymes of ascorbate-glutathione cycle, was overexpressed in tobacco. These transgenic plants were found to be tolerant to both chilling and salt stresses (Roxas et al., 1997, 2000). Similarly, Badawi et al. (2004a,b) found that tobacco transgenic lines expressing Chl-APX5 showed higher APX activity with subsequent enhanced tolerance to the oxidative stress in chloroplasts. The first generation of this line showed enhanced tolerance to salt, PEG, and water stresses, as determined by net photosynthesis. In contrast, transgenic tobacco plants overproducing FeSOD or MnSOD exhibited elevated levels of antioxidant activities and better protection of photosystem II against oxidative stress (Van Camp et al., 1996) Furthermore, FeSOD provides better protection against oxidative stress than MnSOD in transgenic plants. However, FeSOD-overproducing plants were no more tolerant to salt stress than the MnSOD-overproducing plants, as indicated by biomass accumulation (Van Camp et al., 1996). Methionine (Met) and cysteine, both of which contain a sulfur atom in their side chains, are the most easily oxidized amino acids (Vogt, 1995). Most oxidative stress is triggered by Met oxidation resulting in disruption of protein structure (Berlett and Stadman, 1997; Hoshi and Heinemann, 2001). Oxidized Met can be reduced back to Met by the activity the enzyme methionine sulfoxide reductase (MSR) (Moskovitz et al., 1995; Sadanandom et al., 1996). Romero et al. (2004) demonstrated that plant lines overexpressing MSR4 in chloroplasts have increased resistance to

82

M. Ashraf et al.

oxidative damage and are expected to be salt tolerant because higher antioxidant capacity of a plant is associated with salt tolerance (Gossett et al., 1994, 1996; Sehmer et al., 1995). Cell lines of cotton transformed with GR, SOD, and APX were found to be salt tolerant. Growth studies showed that the transgenic cell lines grew better at both 0 and 150 mM NaCl than did either an NaCl-sensitive or an NaCl-tolerant cell line. At 250 mM NaCl, the transgenic cell lines were more tolerant than the NaCl-sensitive cell line, which died before the end of 28 days; however, the NaCl-tolerant cell line exhibited significantly more NaCl tolerance than any of the transgenics. An analysis of the antioxidant enzymes revealed that while each of the transgenics overexpressed its respective antioxidant enzyme, the activities of the other enzymes remained relatively low, and in some cases, the activity was actually lower than the activity expressed by the NaCl-sensitive cell lines. On the other hand, the NaCl-tolerant cell line expressed significantly higher activities of all five antioxidant enzymes when grown on high NaCl level than did the NaCl-sensitive cell line. These data suggest that in order to achieve a significant level of salt tolerance, the activities of several of the antioxidant enzymes must be upregulated simultaneously (Cotton Incorporated, 2007). In contrast, Oh et al. (2005) observed that cellular activity of MSR was attenuated when Arabidopsis transgenic lines of MSR4 were subjected to salt stress. However, these transgenics accumulated MSR4 transcripts under saline conditions. It is possible that the activity of the MSR enzyme is inhibited by cellular components not known yet that are generated in response to high salt concentrations. In addition to these antioxidant enzymes, some aldehyde dehydrogenases (ALDHs) are involved in detoxifying ROS (Sunkar et al., 2003). Some salt stress-induced ALDHs are known to have a role in osmoregulation by catalyzing the synthesis of osmoprotectancts (Kirch et al., 2004). In view of these findings, it can be predicted that introduction of these ALDHs can prevent Na-induced oxidative stress. Overexpression of aldehyde dehydrogenase (AtALDH3 ) and aldehyde reductase (MsALR) in Arabidopsis and tobacco reduced lipid peroxidation and induced osmotic stress tolerance. Recently, Rodrigues et al. (2006) showed that both tobacco and Arabidospsis expressing aldehyde dehydrogenase (GmALDH7 ) showed greater germination and reduced reactive aldehydes, which are generated by lipid peroxidation, under saline conditions. These findings indicate that GmALDH7 is one of the most effective genes for producing salt stress-tolerant plants. Although most salt-affected soils contain a mixture of salts, NaCl is the dominant one, which, if taken up by the plant, causes cytoplasmic toxicity (Rehman et al., 2005). Most plants, when subjected to saline conditions, accumulate a small amount of absorbed Naþ in their roots and exclude it from their shoots. Such plants are generally categorized as Naþ excluders. Regulation of Naþ uptake by cells and long-distance Naþ transport is


83

considered as one of the significant adaptations of plants to salt stress (Munns et al., 1983). Conversely, some species can efficiently accumulate high amounts of Naþ in their shoots. Such plants are referred to as Naþ includers. It appears that Naþ moves passively through a general cation channel from the saline medium into the cytoplasm of plant cells (Blumwald, 2000; Jacoby, 1999; Mansour et al., 2003), but there is also strong evidence that active transport of Naþ takes place through Naþ/Hþ antiports (Niu et al., 1993; Ratner and Jacoby, 1976; Shi et al., 2003). The low uptake and accumulation of Naþ is mediated through the regulation of influx and/or by active efflux from the cytoplasm to the vacuoles or back to the growth medium (Blumwald, 2000; Grattan and Grieve, 1994; Jacoby, 1999). This control mechanism is dependent on the regulation of proton pumps and antiporters operating at both plasma membrane and tonoplast (Ashraf, 2004). While engineering plants for overexpression of genes for different types of antiporter, Aharon et al. (2003) found that overexpression of the vacuolar Naþ/Hþ antiporter that sequesters Naþ in vacuoles (NHX1) improved the salinity tolerance in Arabidopsis, tomato, and Brassicas. Similarly, overexpression of a plasma membrane Naþ/Hþ antiporter gene (SOS1 ) enhanced salt tolerance in A. thaliana (Shi et al., 2003). Ohta et al. (2002) and Fukuda et al. (2004) found that the salt tolerance of transgenic rice overexpressing halophyte (Atriplex gmelini) gene (AgNHX1 ) and rice gene (OsNHX1 ) was improved compared with the wild types. However, the increase in leaf Naþ was similar in both the transgenic and wild-type plants. These results indicated that the Naþ/Hþ antiporter gene could sequester part of the Naþ into the vacuoles and prevent the toxic effects of excessive Naþ ions on the cells. Similarly, Wu et al. (2005) observed that salt tolerance of perennial ryegrass was improved by overexpression of the rice vacuolar Naþ/Hþ antiporter gene, OsNHX1. The leaves of transgenic plants accumulated higher concentrations of Naþ, Kþ, and proline than those of the control plants. While working with wheat, Xue et al. (2004) demonstrated the constitutive overexpression of AtNHX1 in wheat grown in moderately saline soil (10 dS m1 ¼ 100 mM NaCl). The grain yield of the best T3 homozygous transgenic wheat line was reduced by only 50%, whereas in the untransformed control plants it reduced by 65%. Under saline conditions, accumulation of Naþ in the root remained unchanged, while in the shoots it decreased. In the same way, when tobacco plants were transformed to overexpress a vacuolar antiporter Naþ/Hþ isolated from cotton (GhNHX1 ), they exhibited an increase in salt tolerance (Wu et al., 2004). The HAL1 gene from yeast controls Kþ and Naþ transport and tolerance to salt stress in yeast cells. Expression of this gene in tomato increased fruit yield and enhanced Kþ/Naþ selectivity in leaves (Gisbert et al., 2000; Rus et al., 2001a,b). Likewise, overexpression of HAL1-induced salt tolerance in melon (Bordas et al., 1997) and Arabidopsis (Yang et al., 2001) has

84

M. Ashraf et al.

been reported. However, the exact function of this gene in higher plants is not yet fully known. The Arabidopsis STO gene complements the salt-sensitive phenotype of a yeast mutant (Lippuner et al., 1996) that suggests that STO, like SOS3, is involved in regulating the internal Naþ/Kþ ratio. STO is similar to the Arabidopsis CONSTNS protein in regions that may be zinc fingers (Lippuner et al., 1996). While overexpressing STO in Arabidopsis plants, Nagaoka and Takano (2003) showed that root growth was increased by 33–70% in the transgenic plants compared with wild-type plants under salt stress. Salt stress increased the expression of the HPPBF-1 gene, an H-protein promoter binding factor-1 that binds to STO protein. The protein of HPPBF-1 has been found in the nucleus. These findings suggest that STO and HPPBF-1 function as transcription factors that are involved in salt-stress responses in Arabidopsis. It was found that the intracellular Naþ/Kþ ratio was maintained at a lower level in STO transgenic plants than in wild-type plants under NaCl treatment (Nagaoka and Takano, 2003). Active transport of Naþ across plant cell membranes is usually coupled to the proton (Hþ) electrochemical potential established by Hþ-translocating pumps (Blumwald et al., 2000; Hasegawa et al., 2000; Gaxiola et al., 2001; Shi et al., 2000). The pH gradient across cell membranes increases following imposition of salt stress and this has been attributed to both an increase in activity and an increase in transcripts of H-ATPase (Hasegawa et al., 2000). In two spring wheat cultivars, increase in activities of both plasma membrane and tonoplast Hþ-ATPase has been observed when plants were subjected to salt stress (Ayala et al., 1997). Similarly, overexpression of the vacuolar Naþ/ Hþ antiporter and Hþ-pyrophosphatase pump (Hþ-PPase) has resulted in enhanced plant tolerance to salinity (Apse et al., 1999; Gaxiola et al., 2001) and drought stress (Gaxiola et al., 2001). From these findings, it is suggested that overexpression of cation transporters in combination with Hþ-translocating pumps can increase crop salt tolerance. Recently, Fukuda et al. (2004) observed that expression of the vacuolar Hþ-PPase and Naþ/Hþ antiporter appears to be coordinated in barley subjected to salinity stress. In contrast, expression analysis of salt-stressed wheat plants showed that TNHX1 (orthologues of the Arabidopsis genes AtNHX1 ) was substantially upregulated in roots and to some extent in leaves, while TVP1 (orthologues of the Arabidopsis genes AVP1 ) was only upregulated in roots. However, such type of coordinated expression is lacking in leaves (Brini et al., 2005). All the above reports suggest that coordinated elevated expression of vacuolar Hþ-PPase and Naþ/ Hþ antiporter in roots may improve crop salt tolerance. Recently, Zhao et al. (2006a) reported that coexpression of Suaeda salsala SsNHX1 and Arabidopsis AVP1 in transgenic rice caused greater improvement in salt tolerance than transformation with the single gene, SsNHX1. Vacuolar PPase activity was higher in membrane vesicles of salt-stressed SsNHX1 þ AVP1 transgenic plants than SsNHX1 transgenics that were closely related to the development


85

period of the SA-transgenic seedlings and markedly higher in 3-week-old seedlings than in 5-week-old seedlings. Salt-stressed SsNHX1 þ AVP1 transgenic plants had higher ion accumulation with higher Kþ/Naþ ratio in their shoots than did the SsNHX1-transformed lines. Furthermore, salt-stressed SsNHX1 þ AVP1 transgenic plants had higher levels of photosynthesis and root proton exportation capacity. However, generation of H2O2 was reduced in salt-stressed SsNHX1 þ AVP1 transgenic plants. All the data presented in the study supported the view that simultaneous expression of the SsNHX1 and AVP1 conferred greater salt tolerance to the transgenic plants than did the single SsNHX1. The same research group further demonstrated that expressing the plasma membrane Naþ/Hþ antiporter SOD2 from yeast in transgenic rice increased salt tolerance (Zhao et al., 2006b). Transgenic rice plants accumulated more ions but showed reduced accumulation of Naþ with a subsequent higher leaf Kþ/Naþ ratio under saline conditions. Furthermore, P-ATPase hydrolytic activity was higher in isolated plasma membrane vesicles derived from the SOD2 transgenic rice plant roots. Salt stress reduced the PSII activity as reflected from Fv/Fm values. However, this reduction in PSII activity was lower in SOD2 transgenic rice plants. In addition, SOD2 transgenic rice plants had reduced ROS generation. This suggested that SOD2 transgenic rice plants reduced the ROS generation by reduced uptake of Naþ or enhanced extrusion of Naþ from cytosol, which in turn better protected the photosynthetic apparatus from Naþ toxicity, thereby resulting in improved photosynthetic activity. In general, gene regulation is a complex phenomenon with a number of factors involved in upregulation or downregulation of a gene, and it becomes even more complex under stress environments. During the last two decades, many abiotic stress inducible genes have been cloned and characterized from different plant species, including Osem, Rab16A-D and SalT from rice (Claes et al., 1990; Hattori et al., 1995; Mundy and Chua, 1988), LEA and Dehydrin from cotton and barley (Baker et al., 1988; Close et al., 1989), and Rab17 from maize (Villardell et al., 1990). However, for the expression of these genes, there is a need to identify suitable promoters. Most of the stress-specific promoters are cis-acting elements that are recognized by the appropriate transcription factors. Efforts have been made to identify and characterize stress-induced promoters, particularly those induced by anaerobic conditions, low or high temperatures or salt stress (Busk and Pages, 1998; Grover et al., 2001). For example, a gene DREB1A was expressed in A. thaliana (Kasuga et al., 2004), which was operated by either the constitutive CaMV 35S or the stress-inducible rd29A promoter (Kasuga et al., 1999). It was observed that plants expressing DREB1A constitutively exhibited morphological abnormalities under nonstress conditions but, on the other hand, plants expressing the DREB1A under the

86

M. Ashraf et al.

control of rd29A were vigorous and highly tolerant to abiotic stress. Kasuga et al. (2004) are of the view that a combination of the rd29A promoter and DREB1A is useful for improvement of different types of transgenic plants tolerant to stressful environments. However, Ito et al. (2006) observed that transgenic rice plants overexpressing the OsDREB1 or DREB1 genes not only showed growth retardation under normal growth conditions but also improved tolerance to drought, high-salt, and low-temperature stresses. These transgenic rice plants accumulated elevated levels of osmoprotectants such as free proline and various soluble sugars that have a role in stress tolerance. Dubouzet et al. (2003) isolated rice homologues for DREB1/ CBF and DREB2, four OsDREB1s, and one OsDREB2 from rice genomic sequences and found that they induced strong expression of stressresponsive genes in transgenic Arabidopsis plants, resulting in increased tolerance to high-salt and freezing stresses. Overexpression of OsDREB1A in Arabidopsis revealed that this gene has a similar function in inducing stress tolerance. From this, it can be generalized that similar transcription factors function in dicotyledonous and monocotyledonous plants. Recently, Oh et al. (2005) developed transgenic rice plants that constitutively expressed Arabidopsis DREB1A. The transgenic rice overexpressing DREB1A is tolerant to drought and high salinity, but has a low level of tolerance to freezing stress. However, transgenic rice plants that constitutively expressed DREB1A or OsDREB1A genes (Ito et al., unpublished data) show elevated tolerance to drought, salinity, and cold stress (reviewed by Nakashima and Shinozaki, 2006). In contrast to the DREB1/CBF genes, overexpression of DREB2 in transgenic plants under an inducible promoter does not improve stress tolerance, suggesting that DREB2 proteins require posttranslational activation (Liu et al., 1998). This is further supported by the fact that transgenic Arabidopsis plants overexpressing this gene (DREB2A) under a constitutive promoter showed growth retardation and improved tolerance to drought. However, it is important to note that several genes were upregulated exclusively in DREB2A-CA overexpressing Arabidopsis plants and were not upregulated in the transgenic Arabidopsis plants overexpressing DREB1A (Nakashima and Yamaguchi-Shinozaki, 2006). From these reports, it can be generalized that production of transgenic plants with DREB genes is useful for improvement of tolerance of environmental stresses in a number of species. However, constitutive expression of these genes retards plant growth. Development of transgenic plants with stressinducible promoters along with DREB genes or regulation of expression of DREB genes by stress-inducible promoters can induce stress tolerance and minimize the adverse effects of stress on growth. For the effective application of molecular approaches to producing stress-tolerant plants, it would be desirable to restrict transgene expression to particular tissues by the use of tissue-specific promoters (Gittins et al., 2001) and this approach is being explored. Overexpression of the Tsi1 gene


87

encoding an EREBP/AP2-type transcription factor in tobacco induced expression of several pathogenesis-related genes under normal conditions, resulting in improved tolerance to salt and pathogens. These results suggest that Tsi1 might be involved as a positive trans-acting factor in two separate signal transduction pathways under abiotic and biotic stresses (Park et al., 2001). Furthermore, the transgenic overexpression of another transcription factor Alfin1 (zinc finger) enhanced expression of the endogenous MsPRP2 gene in alfalfa and improved salinity tolerance of the plants (Winicov and Bastola, 1999). Overexpression of another transcription factor CpMYB10, isolated from Craterostigma plantagineum, in Arabidopsis led to desiccation and salt tolerance of transgenics lines. Interestingly, it was found that plants overexpressing CpMYB10 exhibited glucose-insensitive and ABA-hypersensitive phenotypes and this suggested that CpMYB10 in Arabidopsis mediates stress tolerance and alters ABA and glucose-signaling responses (Villalobos et al., 2004). A Kruppel-like transcription factor, Mtzpt2–1, is required for the formation of the nitrogen-fixing region in plants. At high level of salt stress (200 mM), the process of nodule formation was limited to the upper part of the roots indicating that initiation of nodule formation is sensitive to salt stress. In addition, recovery of stressed plants is irreversible when plants were exposed to 150 mM NaCl for 7 days, whereas recovery was 75% when plants were salinized for 4 days. Transgenic Medicago truncatula plants expressing Mtzpt2–1 in antisense configuration are more able to ‘‘recover’’ from salt stress than the wild-type plants. These results suggest that Mtzpt2–1, as a molecular marker, has a role at the transcriptional level and is potentially linked to stress tolerance in M. truncatula (Merchan et al., 2003). Recently, Shukla et al. (2006) isolated and characterized a gene CAP2 from Cicer arietinum (chickpea) that encodes AP2-family transcription factor, a family of transcription factors that play important roles in plant responses to stressful environments. Transcript level of CAP2 increased because of water and salt stresses but not because of low-temperature stress. Furthermore, tobacco transgenic plants with CAP2 expressed higher steady-state transcript levels of abiotic stress response genes NtERD10B and NtERD10C and were more tolerant to salt stress and dehydration stress than wild type. In view of all the above reports, it seems that the ultimate goal of regulon biotechnology is the control of signal transduction networks, a manipulation that in turn is expected to improve stress tolerance in plants. LEA proteins are members of a large group of hydrophilic, glycine-rich proteins occurring in plants, algae, fungi, and bacteria (Soulages et al., 2003). These proteins are preferentially expressed in response to desiccation or hyperosmotic stress. Group 2 LEA (dehydrins or responsive to abscisic acid) proteins are known to stabilize macromolecules against damage by stresses such as freezing, dehydration, ionic, or osmotic stress (Soulages et al., 2003).

88

M. Ashraf et al.

The authors have analyzed the structural properties of a recombinant form of a soybean group 2 LEA (rGmDHN1 ), using a range of advanced techniques. It was found that the protein remained in a highly extended conformation at low temperatures that could account for the functional role of GmDHN1 in preventing macromolecular structures by freezing, desiccation, ionic or osmotic stresses. In another study, two transgenic rice plants expressing a wheat LEA group 2 protein (PMA80 ) gene or the wheat LEA group 1 protein (PMA1959 ) gene were developed (Cheng et al., 2002). The second-generation transgenic plants subjected to dehydration or salt stress exhibited high accumulation of either MA80 or PMA1959, which correlated with enhanced tolerance of transgenic rice plants to these abiotic stresses. Yu et al. (2005) investigated the functions of different members of the group 3 LEA genes. They were able to isolate and characterize two new group 3 LEA genes, TaLEA2 and TaLEA3, from wheat and introduced TaLEA2 and TaLEA3 into Saccharmyces cerevisiae to examine the effect of these genes on yeast cell tolerance to osmotic, salt, and cold stresses. However, overexpression of TaLEA2 and TaLEA3 improved stress tolerance in transgenic yeast cells when cultured in medium containing sorbitol, salt, and freezing treatments, respectively. The yeast transformants with TaLEA2 were found to be more tolerant to hyperosmotic and freezing stress than transformants with TaLEA3. Several groups of LEA protein genes have been demonstrated to confer water deficit and salt-stress tolerance. Expression of HVA1, a group 3 LEA protein from barley, conferred tolerance to soil water deficit and salt stress in transgenic rice plants (Babu et al., 2004; Rohila et al., 2002; Xu et al., 1996). Likewise, transgenic Chinese cabbage (Brassica parachinensis) with group 3 LEA protein gene isolated from Brassica spp. showed accumulation of the rape LEA protein in the vegetative tissues and conferred increased tolerance to water deficit and salt stress (Park et al., 2005). The increased tolerance was reflected by delayed development of damage symptoms caused by stress. The increased tolerance also showed improved recovery upon the removal of stress condition. These results suggest that the genetic modification of Chinese cabbage by LEA protein gene holds considerable potentiality for crop improvement toward environment-stress tolerance (Park et al., 2005). Similarly, Oraby et al. (2005) demonstrated that third generation of transgenic oat (Avena sativa) expressing barley HVA1 exhibited greater growth and showed an increased tolerance to salt stress (200 mM NaCl) for yield and a number of yield components. Transgenic tobacco plants expressing a heat-shock protein DnaK/Hsp70 from a halotolerant cyanobacterium Aphanothece halophytica in the cytosol had greater CO2 assimilation rate than control plants after 3 days of treatment with 0.6 M NaCl (Sugino et al., 1999). However, leaf Naþ was significantly lower than in nontransformed control plants. Salt-induced reduction in total protein contents and ribulose 1,5-bisphosphate carboxygenase oxygenase


89

levels were lower in the transgenic tobacco, indicating that DnaK has a role in protecting photosynthetic machinery with a subsequent increase in photosynthesis by extruding Naþ out of photosynthetic tissue. Xyloglucan endotransglucosylase/hydrolase (XTH) has been recognized as a cell wall-modifying enzyme, participating in a number of diverse physiological phenomena. Transgenic Arabidopsis plants that constitutively expressed the CaXTH3 gene, isolated from water-stressed hot pepper plants, under the control of the CaMV 35S promoter exhibited abnormal leaf morphology; the transgenic leaves showed variable degrees of twisting and bending along the edges, resulting in a severely wrinkled leaf shape. In addition, the 35S-CaXTH3 transgenic plants displayed markedly improved tolerance to severe water deficit, and to lesser extent to high salinity compared with the wild-type plants. These results indicate that CaXTH3 is functional in heterologous Arabidopsis cells, thereby effectively altering cell growth and also the response to abiotic stresses. Although the physiological function of CaXTHs is not yet clear, there are several possibilities for their involvement in a subset of physiological responses to offset dehydration and high-salinity stresses in transgenic Arabidopsis plants (Cho et al., 2006). Although transgenic approaches for enhanced abiotic stress tolerance are gaining ground among both scientists and public, the achievements made so far are not astounding. This has been due to the fact that scientists have been producing transgenics of various crops in the past using single-gene transfer that undoubtedly resulted in transgenics with limited stress tolerance. However, there is a growing trend now to use the multigene approach by which several genes responsible for overall stress tolerance are simultaneously transferred to the transgenics (Bajaj et al., 1999; Cherian et al., 2006). Furthermore, other protocols such as RNAi and transposon insertional knockouts for the candidate stress-tolerant genes and signaling pathways show a great promise to produce highly stress-tolerant crop plants. 4.3.3. Roles of crop physiology as a link between molecular breeding and phenotyping Physiological study of salt tolerance could provide a link between genomics and phenomics (Edmeades et al., 2004; Yin et al., 2004). Molecular biologists prefer to use survival of transgenic plants in the presence of short-term salt stress to show clear evidence of gene function in the laboratory. Many target genes for salt tolerance are identified and characterized under high salinity in vitro environments using clear criteria such as salt-induced damage or mortality. These genomic works are hardly consistent with responses of transgenic plants under long-term weak or moderate salt stress, especially, in the field environments. Plant breeders are mainly concerned with favorable phenotyes such as growth or yield under long-term salt stress in the field.

90

M. Ashraf et al.

So, this gap has to be narrowed to use results of genomic works and salttolerant transgenic plants in plant improvement programs. Crop physiology has traditionally been applied to explain crop performance (phenotype) in the presence of salt stress. It is generally known that crop plants usually tackle salt stress using complicated and multiple physiological traits related to water and ion relations. So, physiological traits can be applied to test the function of target genes and transgenic plants for salt tolerance. Arabidopsis CBF-overexpressing transgenic potato maintained much greater growth when exposed to 50 mM NaCl for 2 weeks than wild-type potato did (TRK, unpublished data). Physiological work with the CBF-transgenic potato plants and wild-type plants also revealed that the gene has a function in osmotic adjustment. The transgenic plants adjusted osmotically twice as much as the wild-type plants did in the same salt stress. The salt-induced osmotic adjustment could be a component physiological trait providing increased salt tolerance of the CBF-transgenic potato.

5. Conclusion and Future Prospects Although all the different strategies discussed earlier show promise in improving crop stress tolerance, they have practical advantages and disadvantages. For example, although a number of seed priming techniques have been developed and employed in different crops species, they may not all be equally effective in all crops. In addition, the efficacy of each priming technique may vary not only from crop to crop but also from cultivar to cultivar as well as at different stages of plant growth and development (Ashraf and Harris, 2005). To determine the suitability and effectiveness of different priming techniques, it would be sensible to determine the factors such as concentration and dose of priming agent, time period for soaking of seed in priming agent, and seed storability conditions such as temperature and humidity. However, such assessments must be based on large-scale experiments if seed priming is to be effectively and economically used for extensive field planting. It will also be necessary to effectively disseminate this information to the end users. From a farmer’s viewpoint, cost–benefit ratio of such techniques will certainly be critical in determining their adoption of these techniques. Exogenous applications of organic compounds, such as GB and proline, and PGRs to plants under saline conditions have been shown to enhance the intrinsic levels of these compounds, thereby improving plant growth and final yield under stress conditions. However, different crops differ in their response to exogenous application of these compounds because not all plant species are equally responsive to such exogenous applications. The application of organic compounds through the soil is impractical because


91

compounds added to the soil may undergo degradation by soil microbes (Arshad and Frankenberger, 2002) and such large quantities may be required that the techniques would be prohibitively expensive for practical economic use. Foliar application of such compounds seems to be a plausible approach so as to attain enhanced crop growth and yield under saline conditions because it can be achieved with smaller volumes and doses using appropriate mechanical spraying techniques. There are a number of factors that contribute to overall response of a crop to these compounds. For example, as with seed priming, the effective concentration of these growth promoting compounds may vary from species to species or even from cultivar to cultivar, and high doses may suppress plant growth. In addition, the response of plants to externally applied compounds will almost certainly vary significantly depending on developmental stage. Thus, for effective commercial utilization of these compounds, it would be essential to determine the optimal concentration of these compounds and the appropriate plant growth stage at which the exogenous application of these compounds should be made in order to enhance crop stress tolerance. Here again the main focus should be selection and recommendation of that compound that has a low cost–benefit ratio while using it on a large scale. The issue of food safety of crops treated with ‘‘growth regulators’’ whether naturally occurring or synthetic would also have to be addressed and some countries have strict requirements for testing of novel agricultural products before approval is given, linked with clear recommendations on use and safe harvest intervals. Recent progress in marker technology and genetic transformation seems to offer considerable promise for the development of stress-tolerant plants in the near future. It is imperative to note that the transformation technology for improving plant stress tolerance is still in its infancy because there is no report to date in the literature, describing the testing of the performance of bioengineered plants under natural stressful conditions. Almost all studies reported so far show the performance of transgenic lines under controlled conditions. It is possible that the performance of laboratory or glasshousetested transgenic lines may differ under natural field conditions, owing to the perplexity and interaction of a number of edaphic and climatic factors and their simultaneous influence on plant growth and the main stress factors. Thus, the usefulness of the research carried out during the last two or three decades using transgenic plants to alter salt tolerance has yet to be established in the field. Our limited understanding of stress-related metabolic phenomena is a major gap in achieving a considerable success in developing highly salttolerant plants. Thus, a comprehensive profiling of stress-responsible metabolites and pathways is necessary to the successful molecular breeding of stress-tolerant crop plants. Exploration of additional stress-associated gene resources, from both crop plants and highly salt-tolerant model plants, will

92

M. Ashraf et al.

allow future molecular analysis of salt tolerance mechanisms in potential crop plants (Vinocur and Altman, 2005). Certainly extensive work is needed to elucidate well the genetics, biochemical, and physiological basis of plant stress tolerance. Future knowledge of components of stress tolerance and the identification and cloning of target genes may allow the transfer of multiple genes to produce highly stress-tolerant transgenic plants. With the current advances in genetic transformation technology, it seems possible to transfer multiple genes that may act in combination to improve plant stress tolerance. Furthermore, the knowledge gained from the genomic and postgenomic studies as well as from the proteomics research assures the applicability of the transgenic research for developing highly stress-tolerant plants in future.

REFERENCES Abebe, T., Guenzi, A. C., Martin, B., and Cushman, J. C. (2003). Tolerance of mannitolaccumulating transgenic wheat to water stress and salinity. Plant Physiol. 131, 1748–1755. Agboma, M., Jones, M. G. K., Peltonen-Sainio, P., Rita, H., and Pehu, E. (1997a). Exogenous glycine betaine enhances grain yield of maize, sorghum and wheat grown under two supplementary watering regimes. J. Agron. Crop Sci. 178, 29–37. Agboma, P., Peltonen-Sainio, P., Hinkkanen, R., and Pehu, E. (1997b). Effect of foliar application of glycine betaine on yield components of drought stressed tobacco plants. Exp. Agric. 33, 345–352. Agboma, P., Sinclair, T., Jokinen, K., Peltonen-Sainio, P., and Pehu, E. (1997c). An evaluation of the effect of exogenous glycine betaine on the growth and yield of soybean. Field Crops Res. 54, 51–64. Aharon, G. S., Apse, M. P., Duan, S., Hua, X., and Blumwald, E. (2003). Characterization of a family of vacuolar Naþ/Hþ antiporters in Arabidopsis thaliana. Plant Soil 253, 245–256. Ahloowalia, B. S., and Maluszynski, M. (2001). Induced mutations—a new paradigm in plant breeding. Euphytica 118, 167–173. Ahloowalia, B. S., Maluszynski, M., and Nichterlein, K. (2004). Global impact of mutationderived varieties. Euphytica 135, 187–204. Ahmad, R., and Jabeen, R. (2005). Foliar spray of mineral elements antagonistic to sodium—a technique to induce salt tolerance in plants growing under saline conditions. Pak. J. Bot. 37, 913–920. Ahmed-Hamad, A. M., and Monsaly, H. M. (1998). Seed soaking presowing in vitamins versus the adverse effects of NaCl salinity on photosynthesis and some related activities of maize and sunflower plants. In ‘‘Proceedings of the XIth International Congress on Photosynthesis,’’ Vol. IV, pp. 2617–2620. Kluwer Academic, Budapest, Hungary. August, 17–22. Akram, M. S. (2006). Influence of Exogenously Applied K from Different Sources on Sunflower (Helianthus annuus L.) Under Salt Stress. University of Agriculture, Faisalabad, Pakistan. Al-Hakimi, A. M. A., and Hamada, A. M. (2001). Counteraction of salinity stress on wheat plants by grain soaking in ascorbic acid, thiamin or sodium salicylate. Biol. Plant. 44, 253–261.


93

Al-Humaid, A. I. (2002). Effects of osmotic priming on seed germination and seedling growth of bermudagrass (Cynodon dactylon L.) under saline conditions. Bullet. Faculty Agric. Cairo Univ. 53, 265–274. Albert, A., Martinez-Ripoll, M., Espinosa-Ruiz, A., Yenush, L., Culianez-Macia, F. A., and Serrano, R. (2000). The X-ray structure of the FMN-binding protein atHal3 provides the structural basis for the activity of a regulatory subunit involved in signal transduction. Structure 8, 961–969. Alexander, A. (1985). Optimum timing of foliar nutrient sprays. In ‘‘Foliar Fertilization: Proceedings of the First International Symposium on Foliar Fertilization’’ (A. Alexander, Ed.), Vol. 22, pp. 44–60. Kluwer Academic Publishers, Berlin. Ali, G., Srivastava, P. S., and Iqbal, M. (1999). Proline accumulation, protein pattern and photosynthesis in regenerants grown under NaCl stress. Biol. Plant. 42, 89–95. Anonymous (1986). ‘‘Fifteen Years of NIAB.’’ Nuclear Institute for Agriculture and Biology, Faisalabad, Pakistan. Apse, M. P., Aharon, G. S., Snedden, W. S., and Blumwald, E. (1999). Salt tolerance conferred by overexpression of a vacuolar Naþ/Hþ antiport in Arabidopsis. Science 285, 1256–1258. Arshad, M., and Frankenberger, W. T., Jr. (2002). ‘‘Ethylene, Agricultural Sources and Applications.’’ Kluwer Academic/Plenum Publishers, New York. Arshi, A., Abdin, M. Z., and Iqbal, M. (2006). Sennoside content and yield attributes of Cassia angustifolia Vahl. as affected by NaCl and CaCl2. Scientia Hort. 111, 84–90. Ashraf, M. (1984). ‘‘Induced Variability for Salinity and Aluminium Tolerance by N-Methyl N-Nitrosourea Treatment of Fertilized Egg Cell of Rice.’’ University of Philippines, Los Banos, Philippines. Ashraf, M. (1994a). Breeding for salinity tolerance in plants. Crit. Rev. Plant Sci. 13, 17–42. Ashraf, M. (1994b). Organic substances responsible for salt tolerance in Eruca sativa. Biol. Plant. 36, 255–259. Ashraf, M. (2004). Some important physiological selection criteria fort salt tolerance in plants. Flora 199, 361–376. Ashraf, M., and Foolad, M. R. (2005). Pre-sowing seed treatment—a shotgun approach to improve germination, plant growth, and crop yield under saline and non-saline conditions. Adv Agron. 88, 223–271. Ashraf, M., and Foolad, M. R. (2007). Roles of glycine betaine and proline in improving plant abiotic-stress resistance. Environ. Exper. Bot. 59(2), 206–216. Ashraf, M., and Harris, P. J. C. (2004). Potential biochemical indicators of salinity tolerance in plants. Plant Sci. 166, 3–16. Ashraf, M., and Harris, P. J. C. (2005). ‘‘Abiotic Stresses: Plant Resistance Through Breeding and Molecular Approaches.’’ Haworth Press, New York. Ashraf, M., and Rauf, H. (2001). Inducing salt tolerance in maize (Zea mays L.) through seed priming with chloride salts: Growth and ion transport at early growth stages. Acta Physiol. Plant. 23, 407–414. Ashraf, M., Karim, F., and Rasul, E. (2002). Interactive effects of gibberellic acid (GA3) and salt stress on growth, ion accumulation and photosynthetic capacity of two spring wheat (Triticum aestivum L.) cultivars differing in salt tolerance. Plant Growth Regul. 36, 49–59. Ashraf, M., Kausar, A., and Ashraf, M. Y. (2003). Alleviation of salt stress in pearl millet (Pennisetum glaucum (L.) R. Br.) through seed treatments. Agronomie 23, 227–234. Awad, A. S., Edwards, D. G., and Campbell, L. C. (1990). Phosphorus enhancement of salt tolerance of tomato. Crop Sci. 30, 123–128. Ayala, F., O’Leary, J. W., and Schmaker, K. S. (1996). Increased vacuolar and plasma membrane Hþ-ATPase activities in Salicornia bigelovii Torr. J. Exp. Bot. 47, 25–32. Ayala, F., Ashraf, M., and O’Leary, W. (1997). Plasma membrane Hþ-ATPase activity in salt-tolerant and salt-sensitive lines of spring wheat (Triticum aestivum L.). Acta Bot. Neerl. 46, 315–324.

94

M. Ashraf et al.

Babaeva, E. Y., Volobueva, V. F., Yagodin, B. A., and Klimakhin, G. I. (1999). Sowing quality and productivity of Echinacea purpurea in relation to soaking the seed in manganese and zinc solutions. Izvestiya Timiryazevskoi Sel’skokhozyaistvennoi Akademii 4, 73–80. Babu, R. C., Zhang, J., Blum, A., Ho, D. H. T., Wu, D. R., and Nguyen, H. T. (2004). HVA1, a LEA gene from barley confers dehydration tolerance in transgenic rice (Oryza sativa L.) via cell membrane protection. Plant Sci. 166, 855–862. Badawi, G. H., Kawano, N., Yamauchi, Y., Shimada, E., Sasaki, R., Kubo, A., and Tanaka, K. (2004a). Over-expression of ascorbate peroxidase in tobacco chloroplasts enhances the tolerance to salt stress and water deficit. Physiol. Plant. 121, 231–238. Badawi, G. H., Yamauchi, Y., Shimada, E., Sasaki, R., Kawano, N., and Tanaka, K. (2004b). Enhanced tolerance to salt stress and water deficit by overexpressing superoxide dismutase in tobacco (Nicotiana tabacum) chloroplasts. Plant Sci. 166, 919–928. Bajaj, S., Targolli, J., Liu, L. F., Ho, T. H. D., and Wu, R. (1999). Transgenic approaches to increase dehydration-stress tolerance in plants. Mol. Breed. 5, 493–503. Baker, J., Steele, C., and Dure, III, L. (1988). Sequence and characterization of 5 Lea proteins and their genes from cotton. Plant Mol. Biol. 11, 277–291. Balki, A. S., and Padole, V. R. (1982). Effect of pre-soaking seed treatments with plant hormones on wheat under conditions of soil salinity. J. Indian Soc. Soil Sci. 30, 361–365. Bennett, M., Fritz, V. A., and Callan, N. W. (1992). Impact of seed treatments on crop stand establishment. Hort. Technol. 2, 345–349. Berlett, B. S., and Stadman, E. R. (1997). Protein oxidation in aging, disease, and oxidative stress. J. Biol. Chem. 272, 203–213. Bewley, J. D., and Black, M. (1994). ‘‘Seeds: Physiology of Development and Germination.’’ Plenum Press, New York. Binzel, M. L., Hasegawa, P. M., Rhodes, D., Handa, S., Handa, A. K., and Bressan, R. A. (1987). Solute accumulation in tobacco cells adapted to NaCl. Plant Physiol. 84, 1408–1415. Blumwald, E. (2000). Sodium transport and salt tolerance in plants. Curr. Opin. Cell Biol. 12, 431–434. Blumwald, E., Aharon, G. S., and Apse, M. P. (2000). Sodium transport in plant cells. Biochim. Biophys. Acta 1465, 140–151. Bohnert, H. J., and Jensen, R. G. (1996). Strategies for engineering water stress tolerance in plants. Trends Biotechnol. 14, 89–97. Bordas, M., Montesinos, C., Dabauza, M., Salvador, A., Roig, L. A., Serrano, R., and Moreno, V. (1997). Transfer of the yeast salt tolerance gene HAL1 to Cucumis melo L. cultivars and in vitro evaluation of salt tolerance. Transgenic Res. 6, 41–50. Borsani, O., Valpuesta, V., and Botella, M. A. (2001). Evidence for a role of salicylic acid in the oxidative damage generated by NaCl and osmotic stress in Arabidopsis seedlings. Plant Physiol. 126, 1024–1030. Bouchereau, A., Aziz, A., Larher, F., and Martin-Tanguy, J. (1999). Polyamines and environmental challenges: Recent development. Plant Sci. 140, 103–125. Boyer, J. S. (1982). Plant productivity and environment potential for increasing crop plant productivity, genotypic selection. Science 218, 443–448. Bradford, K. J. (1986). Manipulation of seed water relations via osmotic priming to improve germination under stress conditions. HortScience 21, 1105–1112. Bray, E. A., Bailey-Serres, J., and Weretilnyk, E. (2000). Responses to abiotic stress. In ‘‘Biochemistry and Molecular Biology of Plants’’ (B. Buchanan, W. Gruissem, and R. Jones, Eds.), pp. 1158–1203. American Society of Plant Biologists, Rockville, MD. Brini, F., Gaxiola, R. A., Berkowitz, G. A., and Masmoudi, K. (2005). Cloning and characterization of a wheat vacuolar cation/proton antiporter and pyrophosphatase proton pump. Plant Physiol. Biochem. 43, 347–354.


95

Busk, P. K., and Pages, M. (1998). Regulation of abscisic acid-induced transcription. Plant Mol. Biol. 37, 425–435. Casa, A. M., Brouwer, C., Nagel, A., Wang, L., Zhang, Q., Kresovich, S., and Wessler, S. R. (2000). The MITE family heartbreaker (Hbr): Molecular markers in maize. Proc. Natl. Acad. Sci. USA 97, 10083–10089. Cayley, S., Lewis, B. A., and Record, J. M. T. (1992). Origins of the osmoprotective properties of betaine and proline in Escherichia coli K-12. J. Bacteriol. 174, 1586–1595. Chamel, A. (1985). Survey of different approaches to determine the behaviour of chemical directly applied to arial parts of plants. In ‘‘Proceedings of the First International Symposium on Foliar Fertilization.’’ pp. 66–86. Berlin (FRG). Champagnol, F. (1979). Relationships between PO4 nutrition of plants and salt toxicity. Phosphorous Agric. 76, 35–43. Chang-Zheng, H., Jin, H., Zhi-Yu, Z., Song-Lin, R., and Wen-Jian, S. (2002). Effect of seed priming with mixed salt solution on germination and physiological characteristics of seedling in rice (Oryza sativa L.) under stress conditions. J. Zhejiang Univ. (Agric. Life Sci.) 28, 175–178. Chattopadhyay, A., Mukherjee, S., Rukmini, R., Rawat, S. S., and Sudha, S. (1997). Ionization, partitioning, and dynamics of tryptophan octyl ester: Implications for membrane-bound tryptophan residues. Biophys. J. 73, 839–849. Chen, T. H. H., and Murata, N. (2002). Enhancement of tolerance of abiotic stress by metabolic engineering of betaines and other compatible solutes. Curr. Opin. Plant Biol. 5, 250–257. Cheng, Z.-Q., Targolli, J., Huang, X.-Q., and Wu, R. (2002). Wheat LEA genes, PMA80 and PMA1959, enhance dehydration tolerance of transgenic rice (Oryza sativa L.). Mol. Breed. 10, 71–82. Cherian, S., Reddy, M. P., and Ferreira, R. B. (2006). Transgenic plants with improved dehydration-stress tolerance: Progress and future prospects. Biol. Plantarum 50, 481–495. Chinnusamy, V., Jagendorf, A., and Zhu, J.-K. (2005). Understanding and improving salt tolerance in plants. Crop Sci. 45, 437–448. Cho, S. K., Kim, J. E., Park, J. A., Eom, T. J., and Kim, W. T. (2006). Constitutive expression of abiotic stress-inducible hot pepper CaXTH3, which encodes a xyloglucan endotransglucosylase/hydrolase homolog, improves drought and salt tolerance in transgenic Arabidopsis plants. FEBS Lett. 580, 3136–3144. Chrominski, A., Weber, D. J., Smith, B. N., and Hegerhorst, D. F. (1989). Is dimethylsulfonium propionate an osmoprotectant of terrestrial glycophytes? Die Naturwissenschaften 76, 473–475. Claes, B., Dekeyser, R., Villarroel, R., Bulcke van den, M., Bauw, G., Van Montagu, M., and Caplan, A. (1990). Characterization of a rice gene showing organ-specific expression in response to salt stress and drought. Plant Cell 2, 19–27. Close, T. J., Kortt, A. A., and Chandler, P. M. (1989). A cDNA-based comparison of dehydration induced proteins (dehydrins) in barley and corn. Plant Mol. Biol. 13, 95–108. Cotton Incorporated (2007). Core Program Crop Improvement. [online] available from [20 October 2007]. Csonka, L. N., and Hanson, A. D. (1991). Prokaryotic osmoregulation: Genetics and physiology. Annu. Rev. Microbiol. 45, 569–606. Cuin, T. A., and Shabala, S. (2005). Exogenously supplied compatible solutes rapidly ameliorate NaCl-induced potassium efflux from barley roots. Plant Cell Physiol. 46, 1924–1933. Darra, B. L., Seth, S. P., Singh, H., and Mendiratta, R. S. (1973). Effect of hormone directed presoaking on emergence and growth of osmotically stressed wheat (Triticum aestivum L.) seeds. Agron. J. 65, 292–295.

96

M. Ashraf et al.

Debez, A., Chaibi, W., and Bouzid, S. (2001). Effect du NaCl et de regulatoeurs de croissance sur la germination d’Atriplex halimus L. Cahiers Agricultures. 10, 135–138. Dubouzet, J. G., Sakuma, Y., Ito, Y., Kasuga, M., Dubouzet, E. G., Miura, S., Seki, M., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2003). OsDREB genes in rice, Oryza sativa L., encode transcription activators that function in drought-, high salt and coldresponsive gene expression. Plant J. 33, 751–763. Duncan, R. R., and Carrow, R. N. (1999). Turfgrass molecular genetic improvement for abiotic/edaphic stress resistant. Adv. Agron. 67, 233–305. Ebert, G., Eberle, J., Ali-Dinar, H., and Ludders, P. (2002). Ameliorating effects of Ca(NO3)2 on growth, mineral uptake and photosynthesis of NaCl-stressed guava seedlings (Psidium guajava L.). Scientia Hort. 93, 125–135. Edmeades, G. O., McMaster, G. S., White, J. W., and Campos, H. (2004). Genomics and the physiologist: Bridging the gap between genes and crop responses. Field Crops Res. 90, 5–18. El-Siddig, K., and Ludders, P. (1993). Interactive effects of salinity and nitrogen nutrition on biomass production and distribution in apple trees. J. Appl. Bot. 67, 68–171. El-Tayeb, M. A. (2005). Response of barley grains to the interactive effect of salinity and salicylic acid. Plant Growth Regul. 45, 215–224. Eleiwa, M. E. (1989). Effect of prolonged seed soaking on the organic and mineral components of immature pods of soybeans. Egypt. J. Bot. 32, 149–160. Epstein, E. (1977). Genetic potentials for solving problems of soil mineral stress: Adaptation of crops to salinity. In ‘‘Plant Adaptation to Mineral Stress in Problem Soils’’ (M. J. Wright, Ed.), pp. 73–123. Cornell University, Ithaca, New York. Epstein, E., and Bloom, A. J. (2005). ‘‘Mineral Nutrition of Plants: Principles and Perspectives.’’ 2nd ed. Sinauer Associates, Sunderland, MA. Fallon, K. M., and Phillips, R. (1989). Responses to water stress in adapted carrot cell suspension cultures. J. Exp. Bot. 40, 681–687. FAO. (2006). World agriculture: Towards 2030/2050. Prospects for food, nutrition, agriculture and major commodity groups. Global Perspective Studies Unit Food and Agriculture Organization of the United Nations, Rome, Italy. Fariduddin, Q., Ahmad, A., and Hayat, S. (2003). Photosynthetic response of Vigna radiata to pre-sowing seed treatment with 28 homobrassinolide. Photosynthetica 41, 307–310. Finch-Savage, W. E., and Cox, C. J. (1982). A cold treatment technique to improve the germination of vegetable seeds prior to fluid drilling. Scientia Hort. 16, 301–311. Flores, H. E. (1990). Polyamines and plant stress. In ‘‘Stress Responses in Plants: Adaptation and Acclimation Mechanisms.’’ (R. G. Alscher and J. R. Cumming, Eds.), pp. 217–239. Wiley-Liss, New York. Flowers, T. J. (2004). Improving crop salt tolerance. J. Exp. Bot. 55, 307–319. Flowers, T. J., Troke, P. F., and Yeo, A. R. (1977). The mechanism of salt tolerance in halophytes. Annu. Rev. Plant Physiol. 28, 89–121. Foolad, M. R. (1999). Comparison of salt tolerance during seed germination and vegetative growth in tomato by QTL mapping. Genome 42, 727–734. Foolad, M. R. (2000). Genetic basis of salt tolerance and cold tolerance in tomato. Curr. Top. Plant Biol. 2, 35–49. Foolad, M. R. (2004). Recent advances in genetics of salt tolerance in tomato. Plant Cell Tiss. Org. Cult. 76, 101–119. Forster, B. P. (2001). Mutation genetics of salt tolerance in barley: An assessment of golden promise and other semi-dwarf mutants. Euphytica 120, 317–328. Fouge`re, F., Le Rudulier, D., and Streeter, J. G. (1991). Effects of salts stress on amino acid, organic acid, and carbohydrate composition of roots, bacteroids, and cytosol of alfalfa (Medicago sativa L.). Plant Physiol. 96, 1228–1236.


97

Franke, W. (1967). Mechanisms of foliar penetration of solutions. Annu. Rev. Plant Physiol. 18, 281–300. Fukuda, A., Nakamura, A., Tagiri, A., Tanaka, H., Miyao, A., Hirochika, H., and Tanaka, Y. (2004). Function, intracellular localization and the importance in salt tolerance of a vacuolar Naþ/Hþ antiporter from rice. Plant Cell Physiol. 45, 146–159. Gadallah, M. A. A. (1999). Effects of kinetin on growth, grain yield and some mineral elements in wheat plants growing under excess salinity and oxygen deficiency. Plant Growth Regul. 27, 63–74. Galston, A. W. (2001). Plant biology: Retrospect and prospect. Curr. Sci. 80, 143–152. Galston, A. W., and Kaur-Sawhney, R. (1995). Polyamines as endogenous growth regulators. In ‘‘Plant Hormones: Physiology, Biochemistry and Molecular Biology’’ (P. J. Davies, Ed.), pp. 158–178. Kluwer Academic Publishers, Dordrecht. Gangopadhyay, G., Basu, S., Mukherjee, B. B., and Gupta, S. (1997). Effects of salt and osmotic shocks on unadapted and adapted callus lines of tobacco. Plant Cell Tissue Organ Cult. 49, 42–52. Garg, A. K., Kim, J. K., Owens, T. G., Ranwala, A. P., Do Choi, Y., Kochian, L. V., and Wu, R. J. (2002). Trehalose accumulation in rice plants confers high tolerance levels to different abiotic stresses. Proc. Natl. Acad. Sci. USA 99, 15898–15903. Gaxiola, R. A., Li, J., Undurraga, S., Dang, L. M., Allen, G. J., Alper, S. L., and Fink, G. R. (2001). Drought- and salt-tolerant plants result from overexpression of the AVP1 Hþpump. Proc. Natl. Acad. Sci. USA 98, 11444–11449. Geisler, M., Frangne, N., Gomes, E., Martinoia, E., and Palmgren, M. G. (2000). The ACA4 gene of Arabidopsis encodes a vacuolar membrane calcium pump that improves salt tolerance in yeast. Plant Physiol. 124, 1814–1827. Ghassemi, F., Jackman, A. J., and Nix, H. A. (1995). ‘‘Salinization of Land and Water Resources.’’ CAB International, Wallingford, Oxon, UK. Gisbert, C., Rus, A. M., Bolarıń, M. C., Lo´pez-Coronado, J. M., Arrillaga, I., Montesinos, C., Caro, M., Serrano, R., and Moreno, V. (2000). The Yeast HAL1 gene improves salt tolerance of transgenic tomato. Plant Physiol. 123, 393–402. Gittins, J. R., Hiles, E. R., Pellny, T. K., Biricolti, S., and James, D. J. (2001). The Brassica napus extA promoter: A novel alternative promoter to CaMV 35S for directing transgene expression to young stem tissues and load bearing regions of transgenic apple trees (Malus pumila Mill.). Mol. Breed. 7, 51–62. Glenn, E. P., Brown, J. J., and Blumwald, E. (1999). Salt tolerance and crop potential of halophytes. Crit. Rev. Plant Sci. 18, 227–255. Gossett, D. R., Millhollon, E. P., and Lucas, M. C. (1994). Antioxidant response to NaCl stress in salt-tolerant and salt-sensitive cultivars of cotton. Crop Sci. 34, 706–714. Gossett, D. R., Banks, S. W., Millhollon, E. P., and Lucas, M. C. (1996). Antioxidant response to NaCl stress in a control and a NaCl-tolerant cotton line grown in the presence of paraquat, buthionine sulfoxime, and exogenous glutathione. Plant Physiol. 112, 803–809. Grattan, S. R., and Grieve, C. M. (1994). Mineral nutrient aquisition and response of plants grown in saline environments. In ‘‘Handbook of Plant and Crop Stress.’’ (M. Pessarakli, Ed.), pp. 203–229. Marcel Dekker Press, Inc., New York. Grattan, S. R., and Grieve, C. M. (1999). Salinity–mineral nutrient relations in horticultural crops. Scientia Hort. 78, 127–157. Greenway, H., and Munns, R. (1980). Mechanism of salt tolerance in non-halophytes. Annu. Rev. Plant Physiol. 31, 149–190. Grover, A., Kapoor, A., Lakshmi, O. S., Agarwal, S., Sahi, C., Agarwal, S. K., Agarwal, M., and Dubey, H. (2001). Understanding molecular alphabets of the plant abiotic stress responses. Curr. Sci. 80, 206–216.

98

M. Ashraf et al.

Gulnaz, A., Iqbal, J., and Azam, F. (1999a). Seed treatment with growth regulators and crop productivity. II. Response of critical growth stages of wheat (Triticum aestivum L.) under salinity stress. Cereal Res. Commun. 27, 419–426. Gulnaz, A., Iqbal, J., Farooq, S., and Azam, F. (1999b). Seed treatment with growth regulators and crop productivity. I. 2, 4-D as an inducer of salinity tolerance in wheat (Triticum aestivum L.). Plant Soil 210, 209–217. Guo, S., Yin, H., Zhang, X., Zhao, F., Li, P., Chen, S., Zhao, Y., and Zhang, H. (2006). Molecular cloning and characterization of a vacuolar H[þ]-pyrophos-phatase gene, SsVP, from the halophyte Suaeda salsa and its overexpression increases salt and drought tolerance of Arabidopsis. Plant Mol. Biol. 60, 41–50. Hafeez, A. T. A., and Hudson, J. P. (1967). Effect of ‘‘hardening’’ radish seeds. Nature 216, 688. Halliwell, B., and Gutteridge, J. M. C. (1985). ‘‘Free Radicals in Biology and Medicine.’’ Clarendon Press, Oxford. Hamada, A. M., and Al-Hakimi, A. M. A. (2001). Salicylic acid versus salinity-drought induced stress on wheat seedlings. Rostlinna Vyroba 47, 444–450. Hare, P. D., and Cress, W. A. (1997). Metabolic implications of stress-induced proline accumulation in plant. Plant Growth Regul. 21, 79–102. Hare, P. D., Cress, W. A., and Staden, J. V. (2002). Disruptive effects of exogenous proline on chloroplast and mitochondrial ultrastructure in Arabidopsis leaves. South African J. Bot. 68, 393–396. Harinasut, P., Tsutsui, K., Takabe, T., Nomura, M., and Kishitani, S. (1996). Exogenous glycine betaine accumulation and increased salt tolerance in rice seedlings. Biosci. Biotechnol. Biochem. 60, 366–368. Harris, D. (1992). Staying in control of rainfed crops. In ‘‘Proceedings of the First Annual Scientific Conference of the SADCC/ODA Land and Water Management Programme,’’ pp. 257–262. Gaborone, Botswana. Harris, D., Joshi, A., Khan, P. A., Gothkar, P., and Sodhi, P. S. (1999). On-farm seed priming in semi-arid agriculture: Development and evaluation in maize, rice and chickpea in India using participatory methods. Exp. Agr. 35, 5–29. Hasegawa, P. M., Bressan, R. A., Zhu, J. K., and Bohnert, H. J. (2000). Plant cellular and molecular responses to high salinity. Annu. Rev. Plant Phys. 51, 463–499. Hattori, T., Terada, T., and Hamasuna, S. (1995). Regulation of Osem gene and the transcriptional activator VP1: Analysis of Cis-acting promoter elements required for regulation by abscisic acid and VP1. Plant J. 7, 913–925. Hayashi, H., Alia, L., Mustardy, P., Deshnium, I., and Murata, N. (1997). Transformation of Arabidopsis thaliana with the codA gene for choline oxidase; accumulation of glycinebetaine and enhanced tolerance to salt and cold stress. Plant J. 12, 133–142. He, C., Yan, J., Shen, G., Fu, L., Holaday, S., Auld, D., Blumwald, E., and Zhang, H. (2005). Expression of an Arabidopsis vacuolar sodium/proton antiporter gene in cotton improves photosynthetic performance under salt conditions and increase fiber yield in the field. Plant Cell Physiol. 46, 1848–1854. He, L., Nada, K., and Tachibana, S. (2002). Effects of spermidine pretreatment through the roots on growth and photosynthesis of chilled cucumber plants (Cucumis sativus L.). J. Jpn. Soc. Hort. Sci. 71, 490–498. Hegazi, A. M., El-Gaaly, F. M., and El-Din, N. M. N. (1995). Effect of some growth regulators on yield and yield components of wheat grown under saline conditions. Ann. Agric. Sci. 33, 709–717. Hmida-Sayari, A., Gargouri-Bouzid, R., Bidani, A., Jaoua, L., Savoure´, A., and Jaoua, S. (2005). Overexpression of D1-pyrroline-5-carboxylate synthetase increases proline production and confers salt tolerance in transgenic potato plants. Plant Sci. 169, 746–752.


99

Holmstrom, K. O., Somersalo, S., Mandal, A., Palva, T. E., and Welin, B. (2000). Improved tolerance to salinity and low temperature in transgenic tobacco producing glycinebetaine. J. Exp. Bot. 51, 177–185. Hong, Z., Lakkineni, K., Zhang, Z., and Verma, D. P. S. (2000). Removal of feedback inhibition of D1-pyrroline-5-carboxylate synthetase results in increased proline accumulation and protection of plants from osmotic stress. Plant Physiol. 122, 1129–1136. Hoshi, T., and Heinemann, S. H. (2001). Regulation of cell function by methionine oxidation reduction. J. Physiol. 531, 1–11. Hoshida, H., Tanaka, Y., Hibino, T., Hayashi, Y., Tanaka, A., Takabe, T., and Takabe, T. (2000). Enhanced tolerance to salt stress in transgenic rice that overexpresses chloroplast glutamine synthetase. Plant Mol. Biol. 43, 103–111. Howard, D. D., Gwathmey, C. O., and Sams, C. E. (1998). Foliar feeding of cotton: Potassium sources, potassium solution buffering, and boron. J. Agron. 90, 740–746. Hsu, Z., Hsu, Y. T., and Kao, C. H. (2003). The effect of polyethylene glycol on proline accumulation in rice leaves. Biol. Plantarum 46, 73–78. Hu, W. W., Gong, H., and Pua, E. C. (2005). The pivotal roles of the plant S-adenosylmethionine decarboxylase 5# untranslated leader sequence in regulation of gene expression at the transcriptional and posttranscriptional levels. Plant Physiol. 38, 276–286. Hua, B., and Guo, W. Y. (2002). Effect of exogenous proline on SOD and POD activity of soyabean callus under salt stress. Acta Agric. Boreali-Sinica 17, 37–40. Huang, J., Hirji, R., Adam, L., Rozwadowski, K. L., Hammerlindl, J. K., Keller, W. A., and Selvaraj, G. (2000). Genetic engineering of glycinebetaine production toward enhancing stress tolerance in plants: Metabolic limitations. Plant Physiol. 122, 747–756. Humphreys, M. O., and Humphreys, M. W. (2005). Breeding for stress resistance: General principles. In ‘‘Abiotic Stresses: Plant Resistance Through Breeding and Molecular Approaches’’ (M. Ashraf and P. J. C. Harris, Eds.), pp. 19–46. Haworth Press, New York. Hur, S. N. (1991). Effect of osmoconditioning on the productivity of Italian ryegrass and sorghum under suboptimal conditions. Korean J. Animal Sci. 33, 101–105. Hurly, R. F., Van Staden, J., and Smith, M. T. (1991). Improved germination in seeds of guayule (Parthenium argentatum Gray) following polyethylene glycol and gibberellic acid pretreatments. Ann. Appl. Biol. 118, 175–184. Iqbal, M., and Ashraf, M. (2005a). Presowing seed treatment with cytokinins and its effect on growth, photosynthetic rate, ionic levels and yield of two wheat cultivars differing in salt tolerance. J. Integr. Plant Biol. 47, 1315–1325. Iqbal, M., and Ashraf, M. (2005b). Changes in growth, photosynthetic capacity and ionic relations in spring wheat (Triticum aestivum L.) due to pre-sowing seed treatment with polyamines. Plant Growth Regul. 46, 19–30. Ismaeil, S. M., Khafagi, O. M. A., Kishk, E. T., and Sohsah, S. M. (1993). Effect of some seed hardening treatments on germination, growth and yield of sudangrass grown under saline conditions. Desert Inst. Bull. Egypt 43, 221–242. Ito, Y., Katsura, K., Maruyama, K., Taji, T., Kobayashi, M., Seki, M., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2006). Functional analysis of rice DREB1/CBF-type transcription factors involved in cold-responsive gene expression in transgenic rice. Plant Cell Physiol. 47, 141–153. Jacoby, B. (1999). Mechanisms involved in salt tolerance of plants. In ‘‘Handbook of Plant and Crop Stress’’ (M. Pessarakli, Ed.), pp. 97–123. Marcel Dekker, New York. Jang, I. C., Oh, S. J., Seo, J. S., Choi, W. B., Song, S. I., Kim, C. H., Kim, Y. S., Seo, H. S., Do Choi, Y., Nahm, B. H., and Kim, J. K. (2003). Expression of a bifunctional fusion of the Escherichia coli genes for trehalose-6-phosphate synthase and trehalose-6-phosphate phosphatase in transgenic rice plants increases trehalose accumulation and abiotic stress tolerance without stunting growth. Plant Physiol. 131, 516–524.

100

M. Ashraf et al.

Jason, S., Senaratna, T., and Sivasithamparam, K. (2006). Salicylic acid induces salinity tolerance in tomato (Lycopersicon esculentum cv. Roma): Associated changes in gas exchange, water relations and membrane stabilisation. Plant Growth Regul. 49, 77–83. Jia, G. X., Zhu, Z. Q., Chang, F. Q., and Li, Y. X. (2002). Transformation of tomato with the BADH gene from Atriplex improves salt tolerance. Plant Cell Rep. 21, 141–146. Jyotsna, V., and Srivastava, A. K. (1998). Physiological basis of salt stress resistance in pigeonpea (Cajanus cajan L.)—II. Pre-sowing seed soaking treatment in regulating early seedling metabolism during seed germination. Plant Physiol. Biochem. (New Delhi) 25, 89–94. Kadiri, M., and Hussaini, M. A. (1999). Effect of hardening pretreatments on vegetative growth, enzyme activities and yield of Pennisetum americanum and Sorghum bicolor. Global J. Pure Appl. Sci. 5, 179–183. Kang, J. S., Cho, J. L., and Jeong, Y. J. (1996). Effect of seed priming on the germinability of tomato (Lycopersicon esculentum Mill.) seeds under water and saline stress. J. Korean Soc. Hort. Sci. 37, 516–521. Kasinathan, V., and Wingler, A. (2004). Effect of reduced arginine decarboxylase activity on salt tolerance and on polyamine formation during salt stress in Arabidopsis thaliana. Physiol. Plantarum 121, 101–107. Kasuga, M., Liu, Q., Miura, S., Yamaguchi-Shinozaki, K., and Shinozaki, K. (1999). Improving plant drought, salt, and freezing tolerance by gene transfer of a single stressinducible transcription factor. Nat. Biotechnol. 17, 287–291. Kasuga, M., Miura, S., Shinozaki, K., and Shinozaki, K. Y. (2004). A combination of the Arabidopsis DREB1A gene and stressinducible rd29A promoter improved drought- and low temperature stress tolerance in tobacco by gene transfer. Plant Cell Physiol. 45, 346–350. Kasukabe, Y., He, L., Nada, K., Misawa, S., Ihara, I., and Tachibana, S. (2004). Overexpression of spermidine synthase enhances tolerance to multiple environmental stresses and upregulates the expression of various stress-regulated genes in transgenic Arabidopsis thaliana. Plant Cell Physiol. 45, 712–722. Kaur, J., and Singh, G. (1987). Hormonal regulation of grain filling in relation to peduncle anatomy in rice cultivars. Indian J. Exp. Biol. 25, 63–65. Kaya, C., Kirnak, H., and Higgs, D. (2001a). An experiment to investigate the ameliorative effects of foliar potassium phosphate sprays on salt stressed strawberry plants. Aus. J. Agri. Res. 52, 995–1000. Kaya, C., Higgs, D., and Kirnak, H. (2001b). The effects of higher salinity (NaCl) and supplementary phosphorous and potassium on physiology and nutrition development of spinach. Bulg. J. Plant Physiol. 27, 47–59. Ketchum, R. E. B., Warren, R. C., Klima, L. J., Lopez-Gutierrez, F., and Nabors, M. W. (1991). The mechanism and regulation of proline accumulation in suspension cultures of the halophytic grass Distichlis spicata L. J. Plant Physiol. 137, 368–374. Khan, M. A., Gul, B., and Weber, D. J. (2000). Germination responses of Salicornia rubra to temperature and salinity. J. Arid Environ. 45, 207–214. Khodary, S. E. A. (2004). Effect of salicylic acid on the growth, photosynthesis and carbohydrate metabolism in salt-stressed maize plants. Int. J. Agri. Biol. 6, 5–8. Kingsbury, R. W., and Epstein, E. (1984). Selection for salt resistant spring wheat. Crop Sci. 24, 310–315. Kirch, H.-H., Bartels, D., Wei, Y., Schnable, P. S., and Wood, A. J. (2004). The ALDH gene superfamily of Arabidopsis. Trends Plant Sci. 9, 371–377. Kishitani, S., Takanami, T., Suzuki, M., Oikawa, M., Yokoi, S., Ishitani, M., AlvarezNakase, A. M., Takabe, T., and Takabe, T. (2000). Compatibility of glycinebetaine in rice plants: Evaluation using transgenic rice plants with a gene for peroxisomal betaine aldehyde dehydrogenase from barley. Plant Cell Environ. 23, 107–114.


101

Kishor, P. B. K., Hong, Z., Miao, G. H., Hu, C. A. A., and Verma, D. P. S. (1995). Over 826 expression of [delta]-pyrroline-5-carboxylate synthetase increases proline 827 production and confers osmotolerance in transgenic plants. Plant Physiol. 108, 1387–1394. Krishnamurthy, R., and Bhagwat, K. A. (1993). Effect of foliar application of pro- 833 line on the salt stressed rice seedlings. Acta Agron. Hung. 42, 267–272. Kuiper, P. J. C., Kuiper, D., and Schuit, J. (1988). Root functioning under stress conditions: An introduction. Plant Soil 111, 249–253. Kumar, A., Altabella, T., Taylor, M. A., and Tiburcio, A. F. (1997). Recent advances in polyamine research. Trends Plant Sci. 2, 124–130. Kumar, S., Dhingra, A., and Daniell, H. (2004). Plastid-expressed betaine aldehyde dehydrogenase gene in carrot cultured cells, roots, and leaves confers enhanced salt tolerance. Plant Physiol. 136, 2843–2854. Lea, P. J., Parry, M. A. J., and Medrano, H. (2004). Improving resistance to drought and salinity in plants. Ann. Appl. Biol. 144, 249–250. Lee, S. S., Kim, J. H., Hong, S. B., Yuu, S. H., and Park, E. H. (1998). Priming effect of rice seeds on seedling establishment under adverse soil conditions. Korean J. Crop Sci. 43, 194–198. Lee, S. Y., Ahn, J. H., Cha, Y. S., Yun, D. W., Lee, M. C., Ko, J. C., Lee, K. S., and Eun, M. Y. (2006). Mapping of quantitative trait loci for salt tolerance at seedling stage in rice. Mol. Cells 21, 192–196. Leigh, R. A., Ahmad, N., and Wyn Jones, R. G. (1981). Assessment of glycine betaine and proline compartmentation by analysis isolated beet vacuoles. Planta 153, 34–41. Lema-Rumin´ska, J., Zalewska, M., and Sadoch, Z. (2004). Radiomutants of chrysanthemum (Dendranthema grandiflora Tzvelev) of the Lady group: RAPD analysis of the genetic diversity. Plant Breed. 123, 290–293. Lerner, H. R., and Amzallag, G. N. (1994). The response of plants to salinity; a working hypothesis. In ‘‘Biochemical and Cellular Mechamsms of Stress Tolerance in Plants’’ ( J. H. Cherry, Ed.), pp. 463–476. Springer Verlag, Berlin. Li, Z.-Y., and Chen, S. Y. (2000). Isolation and characterization of a salt-and drought-inducible gene for S-adenosylmethionine decarboxylase from wheat (Triticum aestivum L.). J. Plant Physiol. 156, 386–393. Liang, Z., Ma, D., Tang, L., Hong, Y., Luo, A., and Dai, X. (1997). Expression of the spinach aldehyde dehydrogenase (BADH) gene in transgenic tobacco plants. Chin. J. Biotechnol. 13, 236–240. Lilius, G., Holmberg, N., and Bu¨low, L. (1996). Enhanced NaCl stress tolerance in transgenic tobacco expressing bacterial choline dehydrogenase. Biotechnology 14, 177–180. Lippuner, V., Cyert, M. S., and Gasser, C. S. (1996). Two classes of plant cDNA clones differentially complement yeast calcineurin mutants and increase salt tolerance of wildtype yeast. J. Biol. Chem. 271, 12859–12866. Liu, Q., Sakuma, Y., Abe, H., Kasuga, M., Miura, S., Yamaguchi-Shinozaki, K., and Shinozaki, K. (1998). Two transcription factors, DREB1 and DREB2, with an EREBP/AP2 DNA binding domain, separate two cellular signal transduction pathways in drought- and low temperature-responsive gene expression, respectively, in Arabidopsis. Plant Cell 10, 1391–1406. Lone, M. I., Kueh, J. S. H., Wyn Jones, R. G., and Bright, S. W. J. (1987). Influence of proline and glycine betaine on salt tolerance of cultured barley embryos. J. Exp. Bot. 38, 479–490. Lutts, S. (2000). Exogenous glycine betaine reduces sodium accumulation in salt-stressed rice plants. Intl. Rice Res. Notes 25, 39–40. Maas, E. V., and Nieman, R. H. (1978). Physiology of plant tolerance to salinity. In ‘‘Crop Tolerance to Sub-Optimal Land Conditions’’ (G. A. Jung, Ed.), pp. 277–299. American Society of Agronomy Special, USA.

102

M. Ashraf et al.

Maathuis, F. J. M., and Amtmann, A. (1999). Kþ nutrition and Naþ toxicity: The basis of cellular Kþ/Naþ ratios. Ann. Bot. 84, 123–133. Madan, S., Nainawatee, H. S., Jain, R. K., and Chowdhury, J. B. (1995). Proline and proline metabolizing enzymes in in vitro selected NaCl-tolerant Brassica juncea L. under salt stress. Ann. Bot. 76, 51–57. Ma˚kela, P., Jokinen, K., Kontturi, M., Peltonen-Sainio, P., Pehu, E., and Somersalo, S. (1998a). Foliar application of glycine betaine—a novel product from sugarbeet—as an approach to increase tomato yield. Ind. Crops Prod. 7, 139–148. Ma˚kela, P., Peltonen-Sainio, P., Jokinen, K., Pehu, E., Setala, H., Hinkkanen, R., and Somersalo, S. (1998b). Effect of foliar application of glycine betaine on stomatal conductance, abscisic acid and solute concentrations in leaves of salt- and drought-stressed tomato. Aust. J. Plant Physiol. 25, 655–663. Mano, Y., and Takeda, K. (1997). Mapping quantitative trait loci for salt tolerance at germination and the seedling stage in barley (Hordeum vulgare L.). Euphytica 94, 263–272. Mansour, M. M. F. (1998). Protection of plasma membrane of onion epidermal cells by glycinebetaine and proline against NaCl stress. Plant Physiol. Biochem. 36, 767–772. Mansour, M. M. F. (2000). Nitrogen containing compounds and adaptation of plants to salinity stress. Biol. Plant. 43, 491–500. Mansour, M. M. F., Salama, K. H. A., and Al-Mutawa, M. M. (2003). Transport proteins and salt tolerance in plants. Plant Sci. 164, 891–900. Marschner, H. (1995). ‘‘Mineral Nutrition of Higher Plants.’’ Academic Press, London. Martıńez-Te´llez, M. A., Ramos-Clamont, M. G., Gardea, A. A., and Vargas-Arispuro, I. (2002). Effect of infiltrated polyamines on polygalacturonase activity and chilling injury responses in zucchini squash (Cucurbita pepo L.). Biochem. Bioph. Res. Co. 295, 98–101. McCouch, S. R., and Doerge, R. W. (1995). QTL mapping in rice. Trends Genet. 11, 482–487. McCue, R. F., and Hanson, A. D. (1995). Drought and salt tolerance: Towards understanding and application. TIBTECH 8, 358–362. McDonald, M. B. (2000). Seed priming. In ‘‘Seed Technology and Its Biological Basis’’ (M. Black and J. D. Bewley, Eds.), pp. 287–325. Sheffield Academic Press, Sheffield, UK. Mehta, P. C., Puntamkar, S. S., and Seth, S. P. (1979). Effect of pre-soaking of seeds in different salts with varying concentration on the germination and yield of wheat grown on salinized soil. New Agric. 6, 73–76. Mengel, K. (2002). Alternative or complementary role of foliar supply in mineral nutrition. Acta Hortic. 594, 3348. Merchan, F., Breda, C., Hormaeche, J. P., Sousa, C., Kondorosi, A., Aguilar, O. M., Megias, M., and Crespi, M. (2003). A Kruppel-like transcription factor gene is involved in salt stress responses in Medicago spp. Plant Soil 257, 1–9. Misra, A. N., Srivastava, A., and Strasser, R. J. (2001). Utilization of fast chlorophyll a fluoresence technique in assessing the salt/ion sensitivity of mung bean and Brassica seedlings. J. Plant Physiol. 158, 1173–1181. Mlcˇochova´, L., Chloupek, O., Uptmoor, R., Ordon, F., and Friedt, W. (2004). Molecular analysis of the barley cv. ‘Valticky’ and its X-ray-derived semidwarf-mutant ‘Diamant’. Plant Breed. 123, 421–427. Mo, H., and Pua, E. C. (2002). Up-regulation of arginine decarboxylase gene expression and accumulation of polyamines in mustard (Brassica juncea) in response to stress. Physiol. Plant 114, 439–449. Moghaieb, R. E. A., Tanaka, N., Saneoka, H., Murooka, Y., Ono, H., Morikawa, H., Nakamura, A., Nguyen, N. T., Suwa, R., and Fujita, K. (2006). Characterization of salt tolerance in ectoine-transformed tobacco plants (Nicotiana tabaccum): Photosynthesis, osmotic adjustment, and nitrogen partitioning. Plant Cell Environ. 29, 173–182.


103

Mohan, M., Nair, S., Bhagwat, A., Krishna, T. G., Yano, M., Bhatia, C. R., and Sasaki, T. (1997). Genome mapping, molecular markers and marker-assisted selection in crop plants. Mol. Breed. 3, 87–103. Moskovitz, J., Rahman, M. A., Strassman, J., Yancey, S. O., Kushner, S., Brot, R., and Weissbach, N. H. (1995). Eschericia coli peptide methionine sulfoxide reductase gene: Regulation of expression and role in protecting against oxidative damage. J. Bacteriol. 177, 502–507. Mundy, J., and Chua, N.-H. (1988). Abscisic acid and water-stress induce the expression of a novel rice gene. EMBO J. 7, 2279–2286. Munns, R. (2005). Genes and salt tolerance: Bringing them together. New Phytol. 167, 645–663. Munns, R., Greenway, H., and Kirst, G. O. (1983). Halotolerant eukaryotes. In ‘‘Physiological Plant Ecology. III. Responses to the Chemical and Biological Environment, Encyclopedia of Plant Physiology, New Series’’ (O. L. Lange, P. S. Nobel, C. B. Osmond, and H. Zeigler, Eds.), Vol. 12C, pp. 59–135. Springer-Verlag, Berlin. Nagaoka, S., and Takano, T. (2003). Salt tolerance-related protein STO binds to a Myb transcription factor homologue and confers salt tolerance in Arabidopsis. J. Exp. Bot. 54, 2231–2237. Nakamura, Y., Kasamo, K., Shimosato, N., Sakata, M., and Ohta, E. (1992). Stimulation of the extrusion of protons and Hþ-ATPase activities with the decline in pyrophosphate activity of the tonoplast in intact mung bean roots under high salt stress and its relation to external levels of Ca2þ ions. Plant Cell Physiol. 33, 139–149. Nakashima, K., and Yamaguchi-Shinozaki, K. (2006). Regulons involved in osmotic stressresponsive and cold stress-responsive gene expression in plants. Physiol. Plantarum 126, 62–71. Nanjo, T., Kobayashi, T. M., Yoshida, Y., Kakubari, Y., Yamaguchi-Shinozaki, K., and Shinozaki, K. (1999). Antisense suppression of proline degradation improves tolerance to freezing and salinity in Arabidopsis thaliana. FEBS Lett. 461, 205–210. Navarro, J. M., Martinez, V., and Carvajal, M. (2000). Ammonium, bicarbonate and calcium effects on tomato plants grown under saline conditions. Plant Sci. 157, 89–96. Ndayiragije, A., and Lutts, S. (2006). Exogenous putrescine reduces sodium and chloride accumulation in NaCl-treated calli of the salt-sensitive rice cultivar I Kong Pao. Plant Growth Regul. 48, 51–63. Nishihara, E., Kondo, K., Parvez, M. M., Takahashi, K., Watanabe, K., and Tanaka, K. (2003). Role of 5-amino-levulinic acid (ALA) on active oxygen-scavenging system in NaCl-treated spinach (Spinacia oleracea). J. Plant Physiol. 160, 1085–1091. Niu, X., Bressan, R. A., and Hasegawa, P. M. (1993). Halophytes up-regulate plasma membrane Hþ-ATPase gene more rapidly than glycophytes in response to salt stress. Plant Physiol. 102(Suppl.), 130. Oh, S. J., Song, S. I., Kim, Y. S., Jang, H. J., Kim, S. Y., Kim, M., Kim, Y. K., Nahm, B. H., and Kim, J. K. (2005). Arabidopsis CBF3/DREB1A and ABF3 in transgenic rice increased tolerance to abiotic stress without stunting growth. Plant Physiol. 138, 341–351. Ohta, M., Hayashi, Y., Nakashima, A., Hamada, A., Tanaka, A., Nakamura, T., and Hayakawa, T. (2002). Introduction of a Naþ/Hþ antiport gene from Atriplex gmelini confers salt tolerance to rice. FEBS Lett. 532, 279–282. Ono, H., Sawada, K., Khunajakr, N., Tao, T., Yamamoto, M., Hiramoto, M., Shinmyo, A., Takano, M., and Murooka, Y. (1999). Characterization of biosynthetic enzymes for ectoine as a compatible solute in a moderately halophilic eubacteria, Halomonas elongata. J. Bacteriol. 181, 91–99. Oraby, H. F., Ransom, C. B., Kravchenko, A. N., and Sticklen, M. B. (2005). Barley HVA1 gene confers salt tolerance in R3 transgenic oat. Crop Sci. 45, 2218–2227.

104

M. Ashraf et al.

Ozturk, L., and Demir, Y. (2002). In vivo and in vitro protective role of proline. Plant Growth Regul. 38, 259–264. Papadopoulos, I., and Rendig, V. V. (1983). Interactive effects of salinity and nitrogen on growth and yield of tomato plants. Plant Soil 73, 47–57. Parashar, A., and Varma, S. K. (1988). Effect of presowing seed soaking in gibberellic acid, duration of soaking, different temperatures and their interaction on seed germination and early seedling growth of wheat under saline conditions. Plant Physiol. Biochem. (New Delhi) 15, 189–197. Park, B.-J., Liu, Z., Kanno, A., and Kameya, T. (2005). Genetic improvement of Chinese cabbage for salt and drought tolerance by constitutive expression of a B. napus LEA gene. Plant Sci. 169, 553–558. Park, J. M., Park, C. J., Lee, S. B., Ham, B. K., Shin, R., and Paek, K. H. (2001). Overexpression of the tobacco Tsi1 gene encoding an EREBP/AP2–type transcription factor enhances resistance against pathogen attack and osmotic stress in tobacco. Plant Cell 13, 1035–1046. Passam, H. C., and Kakouriotis, D. (1994). The effects of osmoconditioning on the germination, emergence and early plant growth of cucumber under saline conditions. Scient. Hort. 57, 233–240. Pervez, H., Ashraf, M., and Makhdum, M. I. (2004). Influence of potassium rates and sources on seed cotton yield and yield components of some elite cotton cultivars. J. Plant Nutr. 27, 1295–1317. Petrusa, L. M., and Winicov, I. (1997). Proline status in salt tolerant and salt sensitive alfalfa cell lines and plants in response to NaCl. Plant Physiol. Biochem. 35, 303–310. Pill, W. G., and Necker, A. D. (2001). The effects of seed treatments on germination and establishment of Kentucky bluegrass (Poa pratense L.). Seed Sci. Technol. 29, 65–72. Pill, W. G., Frett, J. J., and Morneau, D. C. (1991). Germination and seedling emergence of primed tomato and asparagus seeds under adverse conditions. HortScience 26, 1160–1162. Poljakoff-Mayber, A., Somers, G. F., Werker, E., and Gallagher, J. L. (1994). Seeds of Kosteletzkya virginica (Malvaceae): Their structure, germination and salt tolerance. Am. J. Bot. 81, 54–59. Prabhavathi, V., Yadav, J. S., Kumar, P. A., and Rajam, M. V. (2002). Abiotic stress tolerance in transgenic eggplant (Solanum melongena L.) by introduction of bacterial mannitol phosphodehydrogenase gene. Mol. Breed. 9, 137–147. Prasad, S. R., Bagali, P. G., Hittalmani, S., and Shashidhar, H. E. (2000a). Molecular mapping of quantitative trait loci associated with seedling tolerance to salt stress in rice (Oryza sativa L). Curr. Sci. 78, 162–164. Prasad, K. V. S. K., Sharmila, P., Kumar, P. A., and Saradhi, P. P. (2000b). Transformation of Brassica juncea (L.) Czern. with bacterial codA gene enhances its tolerance to salt stress. Mol. Breeding 6, 489–499. Qi, Z., and Spalding, E. P. (2004). Protection of plasma membrane Kþ transport by the salt overly sensitive Naþ/Hþ antiporter during salinity stress. Plant Physiol. 136, 2548–2555. Quesada, V., Garcia-Martinez, S., Piqueras, P., Ponce, M. R., and Micol, J. L. (2002). Genetic architecture of NaCl tolerance in Arabidopsis. Plant Physiol. 130, 951–963. Rahman, M. S., Miyake, H., and Takeoka, Y. (2002). Effects of exogenous glycine betaine on growth and ultrastructure of salt-stressed rice seedlings (Oryza sativa L.). Plant Prod. Sci. 5, 33–44. Ratner, A., and Jacoby, B. (1976). Effect of Kþ, its counter anion, and pH on sodium efflux from barley root tips. J. Exp. Bot. 27, 843–850. Ravikovitch, S. (1973). Effects of brackish irrigation water and fertilizers on millet and corn. Expl. Agric. 9, 181–188. Rehman, S., Harris, P. J. C., and Bourne, W. F. (1998). The effect of hardening on the salinity tolerance of Acacia seeds. Seed Sci. & Technol. 26, 743–754.


105

Rehman, S., Harris, P. J. C., and Ashraf, M. (2005). Stress environments and their impact on crop production. In ‘‘Abiotic Stresses: Plant Resistance Through Breeding and Molecular Approaches’’ (M. Ashraf and P. J. C. Harris, Eds.), pp. 3–18. Haworth Press, New York. Rhodes, D., Verslues, P. E., and Sharp, R. E. (1999). Role of amino acids in abiotic stress resistance. In ‘‘Plant Amino Acids: Biochemistry and Biotechnology’’ (B. K. Singh, Ed.), pp. 319–356. Marcel Dekker, New York. Rhodes, R., and Hanson, A. D. (1993). Quaternary ammonium and tertiary sulfonium compounds in higher plants. Annu. Rev. Plant Physiol. Plant Mol. Biol. 44, 357–384. Roberts, D. R., Dumdroff, E. B., and Thompson, J. E. (1986). Exogenous polyamines alter membrane fluidity in bean leaves: A basis for potential misinterpretation of their true physiological role. Planta 167, 395–401. Rodrigues, S. M., Andrade, M. O., Gomes, A. P. S., Da Matta, F. M., BaracatPereira, M. C., and Fontes, E. P. B. (2006). Arabidopsis and tobacco plants ectopically expressing the soybean antiquitin-like ALDH7 gene display enhanced tolerance to drought, salinity, and oxidative stress. J. Exp. Bot. 57, 1909–1918. Rohila, J. S., Jain, R. K., and Wu, R. (2002). Genetic improvement of Basmati rice for salt and drought tolerance by regulated expression of a barley Hva1 cDNA. Plant Sci. 163, 525–532. Romero, H. M., Berlett, B. S., Jensen, P. J., Pell, E. J., and Tien, M. (2004). Investigations into the role of the plastidal peptide methionine sulfoxide reductase in response to oxidative stress in Arabidopsis. Plant Phys. 136, 3784–3794. Roxas, V. P., Lodhi, S. A., Garrett, D. K., Mahan, J. R., and Allen, R. D. (2000). Stress tolerance in transgenic tobacco seedlings that overexpress glutathione S-transferase/ glutathione peroxidase. Plant Cell Physiol. 41, 1229–1234. Roxas, V. P. J., Smith, R. K., Allen, E. R., and Allen, R. D. (1997). Overexpression of glutathione S-transferase/glutathione peroxidase enhances the growth of transgenic tobacco seedlings during stress. Nat. Biotechnol. 15, 988–991. Roy, D., Basu, N., Bhunia, A., and Banerjee, S. K. (1993). Counteraction of exogenous L-proline with NaCl in salt-sensitive cultivar of rice. Biol. Plant. 35, 69–72. Roy, N. K., and Srivastava, A. K. (1999). Effect of presoaking seed treatment on germination and amylase activity of wheat (Triticum aestivum L.) under salt stress conditions. Rachis 18, 46–51. Rus, A., Yokoi, S., Sharkhuu, A., Reddy, M., Lee, B.-H., Matsumoto, T. K., Koiwa, H., Zhu, J. K., Bressan, R. A., and Hasegawa, P. M. (2001a). AtHKT1 is a salt tolerance determinant that controls Naþ entry into plant roots. Proc. Natl. Acad. Sci. USA 98, 14150–14155. Rus, A. M., Estan˜, M. T., Gisbert, C., Garcia-Sogo, B., Serrano, R., Caro, M., Moreno, V., and Bolarıń, M. C. (2001b). Expressing the yeast HAL1 gene in tomato increases fruit yield and enhances Kþ/Naþ selectivity under salt stress. Plant Cell Environ. 24, 875–880. Rus, A. M., Lee, B., Munõz-Mayor, A., Sharkhuu, A., Miura, K., Zhu, J.-K., Bressan, R. A., and Hasegawa, P. M. (2004). AtHKT1 facilitates Naþ homeostasis and Kþ nutrition in plant. Plant Physiol. 136, 2500–2511. Sadanandom, A., Piffanelli, P., Knott, R., Robinson, C., Sharpe, A., Lydiate, D., Murphy, D. J., and Fairbairn, D. J. (1996). Identification of a peptide methionine sulfoxide reductase gene in an oleosin promoter from Brassica napus. Plant J. 10, 235–242. Sakamoto, A., Murata, A., and Murata, N. (1998). Metabolic engineering of rice leading to biosynthesis of glycinebetaine and tolerance to salt and cold. Plant Mol. Biol. 38, 1011–1019. Sakhabutdinova, A. R., Fatkhutdinova, D. R., Bezrukova, M. V., and Shakirova, F. M. (2003). Salicylic acid prevents the damaging action of stress factors on wheat plants. Bulg. J. Plant Physiol. 21, 114–319.

106

M. Ashraf et al.

Saleem, M. Y., Mukhtar, Z., Cheema, A. A., and Atta, B. M. (2005). Induced mutation and in vitro techniques as a method to induce salt tolerance in Basmati rice (Oryza sativa L.). Int. J. Environ. Sci. Technol. 2, 141–145. Sallam, H. A. (1999). Effect of some seed-soaking treatments on growth and chemical components on faba bean plants under saline conditions. Ann. Agric. Sci. (Cairo) 44, 159–171. Satti, S. M., and Al-Yahyai, R. A. (1995). Salinity tolerance in tomato: Implications of potassium, calcium and phosphorus. Comm. Soil. Sci. Plant Anal. 26, 2749–2760. Sayed, S. A., and Gadallah, M. A. A. (2002). Effects of shoot and root application of thiamin on salt-stressed sunflower plants. Plant Growth Regul. 36, 71–80. Sehmer, L., Alaouisosse, B., and Dizengremel, P. (1995). Effect of salt stress on growth and on the detoxifying pathway of pedunculate oak seedlings (Quercus robur L.). J. Plant Physiol. 147, 144–151. Senaratna, T., Touchell, D., Bumm, E., and Dixon, K. (2000). Acetylsalicylic (Aspirin) and salicylic acid induce multiple stress tolerance in bean and tomato plants. Plant Growth Regul. 30, 157–161. Serraj, R., and Sinclair, T. R. (2002). Osmolyte accumulation: Can it really help increase crop yield under drought conditions? Plant Cell Environ. 25, 333–341. Serrano, R., Mulet, J. M., Rios, G., Marquez, J. A., de Larrinoa, I. F., Leube, M. P., Mendizabal, I., Pascual-Ahuir, A., Proft, M., Ros, R., and Montesinos, C. (1999). A glimpse of the mechanisms of ion homeostasis during salt stress. J. Exp. Bot. 50, 1023–1036. Shabala, S. (2000). Ionic and osmotic components of salt stress specifically modulate net ion fluxes from bean leaf mesophyll. Plant Cell Environ. 23, 825–837. Shabala, S., and Newman, I. A. (2000). Salinity effects on the activity of plasma membrane Hþ and Ca2þ transporters in bean leaf mesophyll: Masking role of the cell wall. Ann. Bot. 85, 681–686. Shabala, S., Demidchik, V., Shabala, L., Cuin, T. A., Smith, S. J., Miller, A. J., Davies, J. M., and Newman, I. A. (2006). Extracellular Ca2þ ameliorates NaCl-induced Kþ loss from Arabidopsis root and leaf cells by controlling plasma membrane Kþ-permeable channels. Plant Physiol. 141, 1653–1665. Sharma, P. C., and Kumar, P. (1999). Alleviation of salinity stress during germination in Brassica juncea by pre-sowing chilling treatments to seeds. Biol. Plant. 42, 451–455. Shen, W., Nada, K., and Tachibana, S. (2000). Involvement of polyamines in the chilling tolerance of cucumber cultivars. Plant Physiol. 124, 431–440. Sheveleva, E., Chmara, W., Bohnert, H. J., and Jensen, R. G. (1997). Increased salt and drought tolerance by D-ononitol production in transgenic Nicotiana tabacum L. Plant Physiol. 115, 1211–1219. Shi, H., Ishitani, M., Kim, C., and Zhu, J.-K. (2000). The Arabidopsis thaliana salt tolerance gene SOS1 encodes a putative Naþ/Hþ antiporter. Proc. Natl. Acad. Sci. USA 97, 6896–6901. Shi, H., Lee, B. H., Wu, S. J., and Zhu, J. K. (2003). Overexpression of a plasma membrane Naþ/Hþ antiporter gene improves salt tolerance in Arabidopsis thaliana. Nat. Biotechnol. 21, 81–85. Shukla, R. K., Raha, S., Tripathi, V., and Chattopadhyay, D. (2006). Expression of CAP2, an AP2-family transcription factor from chickpea enhances growth and tolerance to dehydration and salt stress in transgenic tobacco. Plant Physiol. 142, 113–123. Smith, F. W., Hawkesford, M. J., Ealing, P. M., Clarkson, D. T., VandenBerg, P. J., Belcher, A. R., and Warrilow, A. G. S. (1997). Regulation of expression of a cDNA from barley roots encoding a high-affinity sulphate transporter. Plant J. 11, 83–92. Soulages, J. L., Kangmin, K., Arrese, E. L., Walters, C., and Cushman, J. C. (2003). Conformation of a group 2 late embryogenesis abundant protein from soybean. Evidence of poly (L-proline)-type II structure. Plant Physiol. 131, 963–975.


107

Srinivas, V., and Balasubramanian, D. (1995). Proline is a protein-compatible hydrotrope. Langmuir 11, 2830–2833. Stevens, J., Senaratna, T., and Sivasithamparam, K. (2006). Salicylic acid induces salinity tolerance in tomato (Lycopersicon esculentum cv. Roma): Associated changes in gas exchange, water relations and membrane stabilization. Plant Growth Regul. 49, 73–83. Strogonov, B. P. (1964). Practical means for increasing salt tolerance of plants as related to type of salinity in the soil. In ‘‘Physiological Basis of Salt Tolerance of Plants’’ (V. Poljako, A. Mayber, and A. A. Meyer, Eds.), pp. 218–244. Israel Program for Scientific Translations Ltd, Jerusalem. Subbarao, G. V., Wheeler, R. M., Levine, L. H., and Stutte, G. W. (2001). Glycine betaine accumulation, ionic and water relations of red-beet at contrasting levels of sodium supply. J. Plant Physiol. 158, 767–776. Sugino, M., Hibino, T., Tanaka, Y., Nii, N., and Takabe, T. (1999). Overexpression of DnaK from a halotolerant yanobacterium Aphanothece halophytice acquires resistance to salt stress in transgenic tobacco plants. Plant Sci. 146, 81–88. Sulpice, R., Gibon, Y., Cornic, G., and Larher, F. R. (2002). Interaction between exogenous glycine betaine and the photorespiratory pathway in canola leaf discs. Physiol. Plant 116, 460–467. Sun, W., Bernard, C., van de Cotte, B., Van Montagu, M., and Verbruggen, N. (2001). At-HSP17.6A, encoding a small heat-shock protein in Arabidopsis, can enhance osmotolerance upon overexpression. Plant J. 27, 407–415. Sunkar, R., Bartels, D., and Kirch, H. H. (2003). Overexpression of a stress-inducible dehydrogenase gene from Arabidopsis thaliana in transgenic plants improves stress tolerance. Plant J. 35, 452–464. Taiz, L., and Zeiger, E. (2002). ‘‘Plant Physiology’’, 3rd Ed. Sinauer Associates, Inc. Publication, Sunderland, MA. Tanaka, T., and Kuramochi, H. (2001). 5-Aminolevulinic acid improves salt tolerance. Regul. Plant Growth Dev. 36, 190–197. Tanaka, Y., Hibino, T., Hayashi, Y., Tanaka, A., Kishitani, S., Takabe, T., Yokota, S., and Takabe, T. (1999). Salt tolerance of transgenic rice overexpresing yeast mitochondrial Mn-SOD in chloroplasts. Plant Sci. 148, 131–138. Tang, W., and Newton, R. J. (2005). Polyamines reduce salt-induced oxidative damage by increasing the activities of antioxidant enzymes and decreasing lipid peroxidation in Virginia pine. Plant Growth Regul. 46, 31–43. Tarczynski, M. C., Jensen, R. G., and Bohnert, H. J. (1993). Stress protection of transgenic tobacco by production of the osmolyte mannitol. Science 259, 508–510. Taylor, R. M., Fenn, L. B., and Pety, C. A. (1987). [15]Nitrogen uptake by grapes with divided roots growing in differentially salinized soils. HortScience 22, 664. Tester, M., and Davenport, R. (2003). Naþ tolerance and Naþ transport in higher plants. Ann. Bot. 91, 503–550. Thomas, J. C., Sepahi, M., Arendall, B., and Bohnert, H. J. (1995). Enhancement of seed germination in high salinity by engineering mannitol expression in Arabidopsis thaliana. Plant Cell Environ. 18, 801–806. Thompson, J. E., Ledge, R. L., and Barber, R. F. (1987). The role of free radicals in senescence and wounding. New Phytol. 105, 317–344. Tian, L., Huang, C., Yu, R., Liang, R., Li, Z., Zhang, L., Wang, Y., Zhang, X., and Wu, Z. (2006). Overexpression AtNHX1 confers salt-tolerance of transgenic tall fescue. Afr. J. Biotechnol. 5, 1041–1044. Tiburcio, A. F., Kaur-Sawhney, R., and Galston, A. W. (1993). Spermidine biosynthesis as affected by osmotic stress in oat leaves. Plant Growth Regul. 13, 103–109.

108

M. Ashraf et al.

Tozlu, I., Guy, C. L., and Moore, G. A. (1999). QTL analysis of Na and Cl accumulation related traits in an intergeneric BC1 progeny of Citrus and Poncirus under saline and non saline environments. Genome 42, 692–705. Tsonev, T. D., Lazova, G. N., Stoinova, Z. G., and Popova, L. P. (1998). A possible role for jasmonic acid in adaptation of barley seedlings to salinity stress. Plant Growth Regul. 17, 153–159. Tuna, A. L., Kaya, C., Ashraf, M., Altunlu, H., Yokas, I., and Yagmu, B. (2007). The effects of calcium sulphate on growth, membrane stability and nutrient uptake of tomato plants grown under salt stress. Env. Exp. Bot. 59, 173–178. Van Camp, W., Capiau, K., Van Montagu, M., Inze, D., and Slooten, L. (1996). Enhancement of oxidative stress tolerance in transgenic tobacco plants overproducing Fe-superoxide dismutase in chloroplasts. Plant Physiol. 112, 1703–1714. Van den Broek, D., Maes, T., Sauer, M., Zethof, J., de Keukeleire, P., D’Hauw, M., Van Motagu, M., and Gerets, T. (1998). Transposon display identifies individual transposable elements in high copy number lines. Plant J. 13, 121–129. Veena, J., Reddy, V. S., and Sopory, S. K. (1999). Glyoxylase I from Brassica juncea: Molecular cloning, regulation and its overexpression confer tolerance in transgenic tobacco under stress. Plant J. 17, 385–395. Vijayaraghavan, H. (1999). Effect of seed treatment with plant growth regulators on bhendi (Abelmoschus esculentus L.) grown under sodic soil conditions. Madras Agric. J. 86, 247–249. Villalobos, M. A., Bartels, D., and Iturriga, G. (2004). Stress tolerance and glucose insensitive phenotypes in Arahidopsis overexpressing the CpMYBIO transcription factor gene. Plant Physiol. 135, 309–324. Villardell, J., Goday, A., Freire, M. A., Torrent, M., Martinez, M. C., Torne, J. M., and Pages, M. (1990). Gene sequence, developmental expression, and protein phosphorylation of Rab 17 in Maize. Plant Mol. Biol. 14, 423–432. Vinh, N. T., and Paterson, A. H. (2005). Genome mapping and its implications for improving stress resistance in plants. In ‘‘Abiotic Stresses: Plant Resistance Through Breeding and Molecular Approaches’’ (M. Ashraf and P. J. C. Harris, Eds.), pp. 109–124. Haworth Press, New York. Vinocur, B., and Altman, A. (2005). Recent advances in engineering plant tolerance to abiotic stress: Achievements and limitations. Curr. Opin. Biotechnol. 16, 123–132. Vogt, W. (1995). Oxidation of methionyl residues in proteins: Tools, targets, and reversal. Free Radic. Biol. Med. 18, 93–105. Wang, L.-J., Jiang, W.-B., Liu, H., Liu, W.-Q., Kang, L., and Hou, X.-L. (2005). Promotion by 5-aminolevulinic acid of germination of pakchoi (Brassica campestris spp. chinensis var. communis Tsen et Lee) seeds under salt stress. J. Integr. Plant Biol. 47, 1084–1091. Watkinson, J. I., and Pill, W. G. (1998). Gibberellic acid and presowing chilling increase seed germination of Indiangrass [Sorghastrum nutans (L.) Nash.]. HortScience 33, 849–851. Weimberg, R., Lerner, H. R., and Poljakoff-Mayber, A. (1984). Changes in growth and water soluble solute concentrations in Sorghum bicolor stressed with sodium and potassium. Physiol. Plant 62, 472–480. Winicov, I., and Bastola, D. R. (1999). Transgenic overexpression of the transcription factor Alfin1 enhances expression of the endogenous, MSPRP2 gene in alfalfa and improves salinity tolerance of the plant. Plant Physiol. 120, 473–480. Wo´jcik, P. (2004). Uptake of mineral nutrients from foliar fertilization. J. Fruit Ornam. Plant Res. 12(Special edition), 201–218. Wu, C. A., Yang, G. D., Meng, Q. W., and Zheng, C. C. (2004). The cotton GhNHX1 gene encoding a novel putative tonoplast Naþ/Hþ antiporter plays an important role in salt stress. Plant Cell Physiol. 45, 600–607.


109

Wu, Y.-Y., Chen, Q.-J., Chen, M., Chen, J., and Wang, X.-C. (2005). Salt-tolerant transgenic perennial ryegrass (Lolium perenne L.) obtained by Agrobacterium tumefaciensmediated transformation of the vacuolar Naþ/Hþ antiporter gene. Plant Sci. 169, 65–73. Wyn Jones, R. G. (1981). Salt tolerance. In ‘‘Physiological Processes Limiting Plant Productivity’’ (C. B. Johnson, Ed.), pp. 271–292. Butterworths, London. Wyn Jones, R. G., and Storey, R. (1981). Betaines. In ‘‘The Physiology and Biochemistry of Drought Resistance in Plants’’ (L. G. Paleg and A. Aspinall, Eds.), pp. 171–204. Academic Press, Sydney. Xu, D., Duan, X., Wang, B., Hong, B., Ho, T. D., and Wu, R. (1996). Expression of a late embryogenesis abundant protein gene, HVA1, from barley confers tolerance to water deficit and salt stress in transgenic rice. Plant Physiol. 110, 249–257. Xue, Z.-Y., Zhi, D.-Y., Xue, G.-P., Zhang, H., Zhao, Y.-X., and Xia, G.-M. (2004). Enhanced salt tolerance of transgenic wheat (Triticum aestivum L.) expressing a vacuolar Naþ/Hþ antiporter gene with improved grain yields in saline soils in the field and a reduced level of leaf Naþ. Plant Sci. 167, 849–859. Yamaguchi, T., and Blumwald, E. (2005). Developing salt-tolerant crop plants: Challenges and opportunities. Trends Plant Sci. 10, 615–620. Yan, H., Gang, L. Z., Zhao, C. Y., and Guo, W. Y. (2000). Effects of exogenous proline on the physiology of soyabean plantlets regenerated from embryos in vitro and on the ultrastructure of their mitochondria under NaCl stress. Soybean Sci. 19, 314–319. Yancey, P. H. (1994). Compatible and counteracting solutes. In ‘‘Cellular and Molecular Physiology of Cell Volume Regulation’’ (K. Strange, Ed.), pp. 81–109. CRC Press, Boca Raton, Florida. Yancey, P. H., Clark, M. E., Hand, S. C., Bowles, R. D., and Somero, G. N. (1982). Living with water stress: Evolution of osmolyte system. Science 217, 1214–1222. Yang, S. X., Zhao, Y. X., Zhang, Q., He, Y. K., Zhang, H., and Luo, H. (2001). HAL1 mediates salt adaptation in Arabidopsis thaliana. Cell Res. 11, 142–148. Yang, W.-J., Rich, P. J., Axtell, J. D., Wood, K. V., Bonham, C. C., Ejeta, G., Mickelbart, M. V., and Rhodes, D. (2003). Genotypic variation for glycine betaine in sorghum. Crop Sci. 43, 162–169. Yin, X. Y., Struik, P. C., and Kropff, M. J. (2004). Role of crop physiology in predicting gene-to-phenotype relationship. Trends Plant Sci. 9, 426–432. Yoshida, R., Hotta, Y., Tanaka, T., Tkeuchi, Y., and Konnai, M. (1996a). Promotive effects of 5-aminolevulinic acid on rice plants. Crop Res. Asia: Achiev. Perspective (ACSA) 524–525. Yoshida, R., Tanaka, T., and Hotta, Y. (1996b). Regulation of fructan accumulation in Rakkyo (Allium bakeri) and Shallot (Allium ascalonicum) by 5-aminolevulinic acid. Proc. Plant Growth Regul. Soc. Am. 23, 177–182. Yoshida, R., Watanabe, S., Fukuta, Y., Kusaka, Y., Iwai, K., and Tanaka, T. (2003). Effect of 5-aminolevulinic acid on growth and nutrient uptake of leaf vegetables in alkaline soil. In ‘‘Proceedings of Plant Growth Regulation Society of America’’ (E. A. Curry, Ed.), pp. 142–143. Vancover, British Colombia, Canada. Yu, J.-N., Zhang, J.-S., Shan, L., and Chen, S.-Y. (2005). Two new group 3 LEA genes of wheat and their functional analysis in yeast. J. Integr. Plant Biol. 47, 1372–1381. Zhang, H. X., and Blumwald, E. (2001). Transgenic salt-tolerant tomato plants accumulate salt in foliage but not in fruit. Nat. Biotechnol. 19, 765–768. Zhang, H. X., Hodson, J., Williams, J. P., and Blumwald, E. (2001). Engineering salttolerant Brassica plants: Characterization of yield and seed oil quality in transgenic plants with increased vacuolar sodium accumulation. Proc. Natl. Acad. Sci. USA 98, 12832–12836.

110

M. Ashraf et al.

Zhang, J., Klueva, N. Y., Wang, Z., Wu, R., Ho, T.-H. D., and Nguyen, H. T. (2000). Genetic engineering for abiotic stress resistance in crop plants. In Vitro Cell. Dev. Biol. Plant 36, 108–114. Zhao, F., Guo, S., Zhang, H., and Zhao, Y. (2006a). Expression of yeast SOD2 in transgenic rice results in increased salt tolerance. Plant Sci. 170, 216–224. Zhao, F., Wang, Z., Zhang, Q., Zhao, Y., and Zhang, H. (2006b). Analysis of the physiological mechanism of salt-tolerant transgenic rice carrying a vacuolar Naþ/Hþ antiporter gene from Suaeda salsa. J. Plant Res. 119, 95–104. Zhao, F., Zhang, X., Li, P., Zhao, Y., and Zhang, H. (2006c). Co-expression of the Suaeda salsa SsNHX1 and Arabidopsis AVP1 confer greater salt tolerance to transgenic rice than the single SsNHX1. Mol. Breed. 17, 341–353. Zhao, F. G., and Qin, P. (2004). Protective effect of exogenous polyamines on root tonoplast function against salt stress in barley seedlings. Plant Growth Regul. 42, 97–103. Zhu, B., Su, J., Chang, M. C., Verma, D. P. S., Fan, Y. L., and Wu, R. (1998). Overexpression of a pyrroline-5-carboxylate synthetase gene and analysis of tolerance to water and salt stress in transgenic rice. Plant Sci. 139, 41–48.

C H A P T E R

T H R E E

Succession of Arbuscular Mycorrhizal Fungi: Patterns, Causes, and Considerations for Organic Agriculture Jeff S. Piotrowski* and Matthias C. Rillig† Contents 1. Introduction 2. Are There Predictable Patterns of AMF Succession? 2.1. Succession in natural systems 2.2. AMF succession in organically managed agricultural systems 3. Potential Environmental Drivers of AMF Succession in Organic Management 3.1. Long-term abiotic changes of organic management that can affect AMF 3.2. Long-term biotic changes affecting AMF 3.3. Inoculum immigration 4. Consequences of AMF Succession to Production 5. Can We Manage AMF Succession in Organic Agriculture? 5.1. Managing succession of AMF abundance 5.2. Maintaining AMF species composition 6. Future Research Needs to Improve AMF Application 6.1. Measures of persistence 6.2. Physiological studies and environmental match of AM species and ecotypes 6.3. Characterization of native communities 7. Conclusions References

112 114 114 116 118 119 121 122 122 124 124 124 125 125 125 126 126 127

Arbuscular mycorrhizal fungi (AMF) have been promoted as a biofertilizer for sustainable agriculture, and production of inoculum is a widespread and growing industry. The last decade of AMF research has revealed far more selectivity

* {

Division of Biological Sciences, University of Montana, 507 Health Sciences, Missoula, Montana 59812 Freie Universita¨t Berlin, Institut fu¨r Biologie, D-14195 Berlin, Germany

Advances in Agronomy, Volume 97 ISSN 0065-2113, DOI: 10.1016/S0065-2113(07)00003-X

#


111

112

Jeff S. Piotrowski and Matthias C. Rillig

of AMF community association with their hosts and a greater dependence of resulting functions on particular host/fungus combinations than previously imagined, with the effects of specific host/AMF combinations ranging from beneficial to parasitic; hence, the next step in AMF application will entail an effort toward employing beneficial combinations. AMF communities and abundances may fluctuate throughout seasons and years, for example, because of changes in the abiotic and biotic environments. To date, we have little information on the persistence of applied AMF in systems and how changes in the AMF community through time may affect plant growth. We must consider how to manage the soil environment to direct succession of AMF species and the displacement of applied or beneficial AMF. This chapter attempts to merge our current understanding of AMF succession from natural ecosystems with that of AMF application in agriculture. We discuss the patterns and causes of change in AMF abundance and species compositions through time, considering how common organic farming techniques may affect these fungi. We propose that management techniques could be employed to direct AMF succession and maintain specific, beneficial species or species groups, with the potential to increase the sustainability and benefits derived from AMF in organic agriculture.

1. Introduction Sustainable agriculture and organic agriculture are broad terms that describe numerous crop and land management techniques that are designed to reduce anthropogenic inputs like chemical fertilizers, pesticides, and water while preserving the integrity of the soil for future farming (Altieri, 1995). While there are unique crop specific techniques, commonly employed practices include crop rotation, no-till farming, integrated pest management, green manures, polyculture, and the use of so-called ‘‘biofertilizers’’ (Altieri, 1995). Biofertilizers are beneficial, often mutualistic soil organism that can promote plant growth and reduce inputs; these include nitrogen-fixing bacteria in leguminous crops, ectomycorrhizal fungi like Pisolithus in timber plantations, Azospirillum in rice culture, and arbuscular mycorrhizal fungi (AMF) in many crop plants (Rai, 2006). The last two decades have seen an exponentially increasing interest in the use of AMF as biofertilizers in sustainable agriculture, site amelioration, and renaturalization (Cuenca et al., 1997; Gianinazzi et al., 1995; Hart and Trevors, 2006; Jeffries et al., 2003). Despite a lack of consistent and predictable benefits (Gosling et al., 2006; Ryan and Graham, 2002), the commercial production of these symbiotic fungi has developed into a global industry with dozens of companies producing inoculum (Gianinazzi and Vosatka, 2004). The use of AMF in sustainable agriculture is particularly attractive in that through the symbiotic association with these fungi, plants are able to increase nutrient uptake and reduce inputs of fertilizers, water,

Succession of Arbuscular Mycorrhizal Fungi

113

and pesticides. Even with their popularity, widespread, prophylactic application, and the potential hazards of careless introductions of nonnative mycorrhizal species, there is still a paucity of information regarding the persistence of applied AMF (Schwartz et al., 2006). AMF are susceptible to successional pressures resulting from changes in the abiotic and biotic environments; yet, little definitive information is available on the mechanisms behind AMF community changes. Displacement of applied AMF by indigenous AMF or other competitive soil microbes may occur very shortly after application that has the potential to greatly attenuate the purported benefits of these fungi, resulting potentially in a considerable waste of resources. Over time, soils under certain organic practices may change in certain biotic and physiochemical properties that can affect AMF. Many sustainable practices are designed to increase soil nutrients and organic matter (low till, no-till, green manures), and in many cases, these can stimulate AMF inoculum potential (Gosling et al., 2006; Oehl et al., 2004). However, in order for the benefits of AMF to be maximized or to be even simply positive, agriculturalists must consider that abundance alone will not always translate to benefit, rather the specific fungi selected must be identified (Douds et al., 2005). Moreover, AMF vary widely in their tolerances and requirements, and soil parameters that change under long-term organic management techniques may select less beneficial AMF and displace applied species of known benefit. Owing to the functional diversity of soil microbial groups, the importance of microbial species diversity to sustainable agriculture, not just abundance, has been realized (Kennedy and Smith, 1995). Studies of AMF have highlighted that it is not merely their presence but rather the community composition and specific plant/AMF combinations that can determine the benefit of these fungi. AMF can vary tremendously with respect to promoting host plant growth (Klironomos, 2003), phosphorus uptake ( Jakobsen et al., 2002), conferring drought tolerance (Ruiz-Lozano et al., 1999), soil aggregation (Piotrowski et al., 2004), pathogen protection (Maherali and Klironomos, 2007), and seedling establishment (van der Heijden, 2004). Hence, managing for abundance alone may not yield the greatest benefit that these symbionts can offer, but promoting AMF diversity or even the proliferation of particular species or families of AMF could be a more effective strategy. Furthermore, if specific AMF are applied, then understanding the physiological tolerances and requirements of the applied species will be necessary to promoting their persistence and reduce the need and costs of multiple applications. If specific AMF/host combinations are to be employed to maximize a desired function, then displacement of inoculum by immigrant or indigenous fungi could have significant economic and production consequences. To protect the investment in AMF inoculum by

114


stabilizing to the desired abundance and community structure, we must understand the drivers and patterns of AMF succession. Therefore, the purpose of this chapter is to merge our knowledge of patterns and causes of AMF succession from natural systems with frequently observed soil changes that occur during the course of long-term organic agricultural techniques; the goal is to predict how these changes will affect AMF communities in agroecosystems and what techniques would allow management of specific AMF. We seek to highlight areas of collaboration between both AMF biologists and agriculturalists that desire to use the fungi in order to create more efficient application strategies. We will primarily focus on annually cropped systems, realizing that AMF management in woody production systems may be considerably different. Herein we describe some predictable patterns of AMF succession, identify driving causes of these changes that may occur under long-term organic management, and discuss if it is feasible to direct AMF species succession. Finally, we outline future research needed to properly address AMF succession to ensure more consistently effective application of these fungi.

2. Are There Predictable Patterns of AMF Succession? 2.1. Succession in natural systems Major studies reporting on patterns of successional changes in AMF abundance across decades are presented in Table 1. From these, a consistent pattern has emerged in temperate and boreal systems. Following disturbance, the abundance of AMF increases rapidly with increases in AMFhosting species (0–20 years post disturbance); however, the period of AMF dominance is brief as ectomycorrhizal fungi (ECMF) and ECMFhosting plants replace AMF plants (Barni and Siniscalo, 2000; Boerner et al., 1996; Johnson et al., 1991; Piotrowski et al., in press; Treseder et al., 2004; Trowbridge and Jumpponen, 2004; van der Heijden and Vosatka, 1999). Lowland tropical systems display a continual increase in AMF abundance following disturbance as ECMF-hosting species are not generally as common in these systems ( Janos, 1980). Additionally, assessment of AMF abundance from some dune ecosystems has shown steady increases in AMF abundance during succession, with the greatest abundances at the oldest site (Allen and Allen, 1980; Gemma and Koske, 1990; Greipsson and El-Mayas, 2000; Koske and Gemma, 1997). Regarding changes in AMF species composition through many years, Johnson et al. (1991) provided the first and most complete documentation of AMF succession during old field development. This study suggests that ‘‘early succession’’ versus ‘‘late succession’’ species of AMF species could exist.

115


Table 1 Studies that directly measure long-term changes in AMF abundance during soil development following disturbance using chronosequences

Disturbance

Tillage Barni and Siniscalo, 2000 Johnson et al., 1991

Chronosequence age (years)

AMF measure

Summary of AMF pattern

0 ! 60

MIP

0 ! 60

Spores, MIP

MIP peaks at 10 years then declines Spores increase across most of sequence but are very low at oldest sites. MIP peaks at 19 years then declines Decrease through time

Boerner et al., 5–30 1996 Dune formation Greipsson and 0–245 El-Mayas, 2000

Koske and Gemma, 1997 Volcano Balser et al., 2005

Fire Treseder et al., 2004 Flood Piotrowski et al., in press

MIP

Spore number

Spores increase across entire sequence first sampling, peak in abundance at penultimate site following year Rapid steady increase

0 ! 5 (oldest not determined)

Spores, MIP

300–40,100,000

Hyphal length, PLFA

Peak at middle site, lowest abundance at the oldest site

3–80

Hyphal lengths

Peak and decline

0–70

Hyphal length, MIP

Peak at 10–13 years then decline in both measures

MIP, Mycorrhizal inoculum potential; PLFA, phospholipid fatty acid analysis.

116


Johnson et al. (1991) did not find an increase in AMF richness through time, but increasing species evenness through time as Glomus aggregatum spores became less abundant. Nevertheless, some species were much more abundant in late succession sites (Acaulospora elegans) and some in early sites (Scutellospora persica). Koske and Gemma (1997) presented a similar pattern based on spore data from a dune system of the eastern United States. This study of a 5-year chronosequence documented an increase in species richness across the artificially planted system. Like Johnson et al. (1991) they identified certain species characteristic of certain successional stages. For instance, one species of Acaulospora was only found in mid-to-late successional soils, and Glomus 7243 was only present in the oldest sites. Thus far, no consistent patterns of early versus late AMF species have been defined, and some studies have documented no change in AMF during succession ( Johnson and Wedin, 1997), or a decline in species richness through time (Beauchamp et al., 2007). Nevertheless, similar patterns have been documented in agroecosystems.

2.2. AMF succession in organically managed agricultural systems A large body of literature supports the fact that conventional agriculture practices of tillage and high inputs of phosphorus fertilizers can drastically reduce AMF inoculum in agroecosystems over time (Douds and Millner, 1999; Jansa et al., 2002; Kabir, 2005). The scenario can be different under certain long-term organic management regimes. Many studies have demonstrated that after decades of organic management, AMF abundance, as measured by hyphal density or spore numbers, is significantly increased compared to conventionally managed plots (Kabir et al., 1998; Ma¨der et al., 2002; Oehl et al., 2003). Long-term sustainable practices may also increase AMF diversity compared to conventional agriculture ( Jansa et al., 2002; Oehl et al., 2003, 2004). These studies have also revealed a change in species composition similar to succession in natural systems. Notably, over long-term organic management (BiodynamicTM and no-till), species of Acaulospora increase in abundance, as well as of other slow-growing species like Scutellospora and Entrophospora ( Jansa et al., 2002; Oehl et al., 2004). These data suggest that significant changes in AMF abundance and species composition occur over the years and decades since conversion to organic management, similar to natural systems ( Johnson et al., 1991; Koske and Gemma, 1997). What are the changes that are driving the succession of AMF? Over time, abiotic changes similar to those in natural soil development (e.g., decreasing pH, increasing organic matter and litter leachates, and increasing saprophytic fungi) could affect AMF. A list of potential effects of long-term organic management on AMF is presented in Table 2. Sustainable management practices in long-term cropping may have to consider

117


Table 2 Comparison of common sustainable agriculture practices and their long-term effects on the soil environment and potential effects on AMF

Agricultural practice

Some known long-term effects on abiotic and biotic environments

Hypothesized effects on AMF

No-till

Increased soil organic

Mixed effects on abundance,

matter Alteration of soil pH

Increased soil moisture Increased soil phenolics Increase saprophytic

fungi/hyphal grazers Crop rotation

Alteration of soil

nutrient status Alteration of soil pH Alteration of AMF host

interactions Soil organic matter Green manures/ ley crops Alteration of soil

nutrient status Increase in soil moisture

potential selection for non-Gigasporaceae depending on organic matter chemistry Mixed effects on abundance, potential selection of Acaulospora species with pH decrease Potential decrease in abundance at high moisture levels, unknown selection Decrease in abundance, potential selection for nonGigasporaceae Potential decrease in abundance, unknown selection Mixed effects on abundance, potential selection toward more parasitic Glomus species at high nutrient levels Mixed effects on abundance, potential selection of Acaulospora Mixed effects of abundance and species selection depending on the specific AMF/host combination Mixed effects on abundance, potential selection for nonGigasporaceae depending on organic matter chemistry Mixed effects on abundance, potential selection toward more parasitic Glomus species at high nutrient levels Potential decrease in abundance at high moisture levels, unknown selection (continued)

118


Table 2

(continued)

Agricultural practice

Some known long-term effects on abiotic and biotic environments

Hypothesized effects on AMF

Increased soil phenolics

Decrease in abundance,

Increase saprophytic

fungi Alteration of AMF host

interactions

potential selection for non-Gigasporaceae Potential decrease in abundance, unknown selection Mixed effects of abundance and species selection depending on the specific AMF/host combination

these pressures, as well as AMF host diversity, in order to maintain a beneficial AMF abundance and effective species composition.

3. Potential Environmental Drivers of AMF Succession in Organic Management In natural systems, many studies have correlated changes in AMF abundance with changes in abiotic properties in an attempt to define a specific cause of changes in AMF during succession. During soil development, components of the soil physiochemical environment known to influence AMF growth (e.g., pH, soil moisture, and phosphorus) can change dramatically over relatively short periods. AMF abundance and species composition can also change through time, even without changes in the plant community, indicating that exogenous forces are affecting proliferation of these fungi (Husband et al., 2002; Koske and Gemma, 1997; Wacker, 1988). Organic agricultural practices are designed to reduce seasonal inputs and tillage, and therefore soils in these systems may develop more similarly to natural soil systems following disturbance, compared to conventionally tilled and managed fields. During development of nonmanaged soils, predictable changes can occur in soil texture, pH, organic matter, and nutrient content (Walker and Moral, 2003); similar changes may occur in some organically managed systems. For instance, pH may be reduced (Pekrun et al., 2003), organic matter increases (Marriott and Wander, 2006), soil nutrient status changes (Gosling and Shepherd, 2005), and phenolic compounds


119

accumulate (Blum et al., 1991). All these variables have the potential to affect AMF abundance and species composition. While these changes are correlated with increases in AMF abundance and diversity (Ma¨der et al., 2002; Oehl et al., 2004), selection for less beneficial AMF might also occur depending on the host/AMF interaction. This section discusses what we know about the controls of AMF abundance and species composition from natural systems and how we may use these to predict changes in agroecosystems.

3.1. Long-term abiotic changes of organic management that can affect AMF 3.1.1. Soil pH Decreases in soil pH have been correlated with reduction of AMF in soils (reviewed in Entry et al., 2002). The mechanism may be a function of the pH tolerance of AMF, increasing metal toxicity, or alteration of phosphate availability. There is considerable variability in how pH responds to longterm organic management practices. In general, organic management such as reduced tillage increases soil buffering capacity and reduces large shifts in pH; however, there are instances where long-term no-till can significantly alter soil pH (Pekrun et al., 2003). In a summary of studies, Pekrun et al. (2003) found that half of the soils under no-till management for more than 5 years had significantly reduced soil pH, while the other half had no significant change. Other practices like green manure, compost additions, and crop rotations can also affect soil pH (Astier et al., 2006; Godsey et al., 2007). No research to date has documented that such shifts in pH are correlated with reduced AMF abundance; however, these changes may affect species composition in a predictable manner. A few studies suggest that species of Acaulospora are more tolerant of low soil pH ( Johnson et al., 1991; Porter et al., 1987). This could explain proliferation of these AMF in late succession or after long-term no-till management ( Jansa et al., 2002; Johnson et al., 1991; Oehl et al., 2004). Interestingly, this species is also more adept at phosphorus acquisition than Glomus or Gigaspora species ( Jakobsen et al., 1992). Perhaps the Acaulospora is a preferred associate in late succession soils where low pH limits phosphorus availability. 3.1.2. Soil nutrient status Changes in soil nutrient status may have a strong effect on AMF colonization and abundance. The soil nitrogen to phosphorus ratio has been shown repeatedly to affect AMF colonization ( Johnson et al., 2003; Liu et al., 2000), but the exact mechanism is still unclear. Changes in available (mineralized) soil phosphorus and nitrogen are the basis of Read’s hypothesis regarding the distribution of AMF across ecosystems (Read, 1991), and since then, this idea has been applied to changes in AMF across successional

120


time in temperate and boreal ecosystems (Treseder et al., 2004; Piotrowski et al., in press). While excess phosphorus inputs can reduce AMF abundance in agricultural settings, the role of phosphorus inhibition of AMF during soil development in natural ecosystems is not conclusive because concomitant changes in other variables make it difficult to isolate a single cause. The effect of organic management on soil nutrient status again depends on the particular practice. As with conventional fertilizer amendments, longterm use of organic fertilizers and green manures can affect soil phosphorus and nitrogen levels (Edmeades, 2003), potentially affecting AMF. To date, no studies have described nutrient inhibition of AMF by organic fertilizers that is similar to long-term reduction of AMF through conventional fertilizer regimes, and this may not be a strong selective force under organic regimes. Nevertheless, alteration of the nutrient status over time may result in similar shifts toward less beneficial AMF as described by Johnson (1993). 3.1.3. Soil organic matter and crop residues Following disturbance, soil organic matter can increase rapidly in natural systems as well as under many organic regimes (e.g., no-till, green manure, and compost additions). Read (1991) proposed that as soil organic matter increases, the preferred mycorrhizal associate will be one that can access organic nutrients; hence, AMF will be replaced by ECMF that have extracellular enzyme systems capable of accessing organic nutrients. While this phenomenon is frequently observed, it is unknown if organic matter is directly involved in the displacement of AMF. Experimental tests of organic matter additions on AMF are mixed. Experiments have shown that organic matter additions stimulate AMF colonization and soil abundance (Cavender et al., 2003; Nan et al., 2006). Other studies, however, have shown that organic matter inputs in the form of litter and litter leachates may strongly inhibit AMF colonization (Piotrowski et al., 2007; Yun and Choi, 2002). As with other microorganisms, the effect of organic matter inputs on AMF depends on the litter chemistry. Piotrowski et al. (2007) document that increases in soluble phenolic compounds over time may result in inhibition of AMF colonization. The mechanism may be a product of phenolic toxicity or competitive exclusion by organisms that are capable of detoxifying or tolerating these compounds. Of course, this may only occur in young soils with low humic substance content, as humic materials may bind phenolic substances (Cecchi et al., 2004). Similarly, some no-till systems have significantly higher soil phenolic concentrations than conventional till (Blum et al., 1991). The effect of these substances from crop residues on AMF may be most significant in crops that have a high foliar phenolic concentration and less noticeable in other crops. Phenolics might even contribute to succession of AMF species. A study by Wacker et al. (1990) described inhibition of germ tube elongation of AMF spores by the common phenolic compound ferulic acid. This suggests that accumulation of soil phenolics could lead to selection of AMF species


121

that can infect through other means besides spores alone. In this instance, accumulation of phenolics could potentially reduce root colonization by Gigasporaceae species in a soil (which colonize primarily from spores), compared with members of the Acaulosporaceae and Glomaceae that have more infective soil hyphae (Hart and Reader, 2002).

3.2. Long-term biotic changes affecting AMF In addition to changes in the abiotic soil properties discussed above, many biotic parameters that can affect AMF abundance and/or diversity can change through soil development. Some sustainable agriculture practices allow for greater changes in the biotic environment than conventionally managed fields. Green manures, polyculture, and crop rotation introduce greater plant diversity to a cropped system, and these practices have the potential to determine AMF abundance as well as species composition. Host identity can affect sporulation and abundance of particular AMF species (e.g., Bever et al., 1996; Vandenkoornhuyse et al., 2002). Here the choice of the cover crop or cocultivated crop can strongly affect AMF. It is established that using non-AMF host species as cover crops may reduce soil inoculum (Arihara and Karasawa, 2000; Miller, 2000), but the AMF communities associated with the cover crop may not be optimal for conferring a benefit to the crop. If so, the cover crop has the potential to amplify a less beneficial AMF community and affect the performance of the production crop, similar to the negative feedback phenomena described in the work of Bever (2002). The same negative feedback phenomena could also hold for crop rotations. Rotating one crop species and its associated AMF community with another crop species that does not benefit from the previous AMF community could affect production. Increases in organic matter inputs during soil development can stimulate populations of saprobic organisms. These could compete with AMF for resources within the soil. Also, saprobic organisms parasitic to AMF may be stimulated with increased organic matter (Lozupone and Klein, 1994). Hyphal grazers like collembolans can significantly increase under reduced tillage (Titi, 2003). While collembolans are thought not to prefer AMF over other fungi (Klironomos and Kendrick, 1996), in high abundance they could still reduce AMF hyphal lengths enough to give other proliferating fungi a competitive advantage, or even selectively graze certain AMF species. During succession in natural systems, there typically is a large turnover in plant species, presumably also resulting in larger changes in the AMF community. In most conventional agricultural settings, there are no large changes in the plant community outside of the crop; however, in less intensively managed systems, neighboring plants and weeds with their associated AMF could lead to changes in inoculum identity within a field. For instance, Mummey and Rillig (2006) describe the decrease in diversity of an AMF community of a natural

122


grassland after invasion by Centaurea maculosa; a similar phenomenon may occur as weeds invade cropped fields. Over seasons and in the postharvest periods, inoculum of any applied fungi could quickly be displaced if they have lower sporulation with the established weeds.

3.3. Inoculum immigration The initial AMF community is determined by the resident inoculum, which will be a function of site history. However, over time, immigration of AMF inoculum from off-site has the potential to alter the community composition. AMF inoculum is available as spores, hyphal fragments, or colonized root fragments. While the amount of inoculum available will depend on disturbance intensity (e.g., tillage intensity) and time since disturbance (e.g., hyphal fragments and colonized root lengths do not persist as long as spores), the invading species composition of the inoculum will be almost entirely a product of vegetation surrounding the disturbed site. Unlike most fungi, wind is not the primary dispersal agent of AMF. While aeolian deposition of AMF spores is possible (Allen et al., 1989), their spores are much larger than other fungi and produced entirely underground. AMF spores are transported into disturbed and denuded environments via a variety of vectors, including rodents (Allen et al., 1984; Mangan and Adler, 2002), earthworms and microarthropods (Doube et al., 1994; Klironomos and Moutoglis, 1999), and anthropogenic dispersal (Schwartz et al., 2006). The extent of dispersal may depend on the specific AMF family and spore size (Klironomos and Moutoglis, 1999). Members of the Glomeraceae produce copious, small-volume spores compared with the Gigasporaceae that produce fewer spores of a significantly greater size (Hart and Reader, 2002). In early successional sites, members of Glomeraceae spores will likely be the primary immigrants; however, over longer timescales, species that have larger spores, produce fewer spores, or both will immigrate as well.

4. Consequences of AMF Succession to Production The reduction of AMF abundance and colonization alone has the potential to reduce plant performance in highly mycorrhizal-responsive plants, but what about changing AMF community composition? While AMF diversity can affect plant diversity in natural plant communities (Hartnett and Wilson, 1999; van der Heijden et al., 1998), the composition of the AMF community might not always be beneficial in low-diversity agricultural systems. Increasing AMF diversity can lead to a greater likelihood of a beneficial host/plant combination and functionally


123

complementary combinations of AMF; however, although not yet documented, the opposite might also be possible. During succession of agricultural fields employing AMF, a shift in community composition could potentially affect the benefits that the desired applied community confers. Our knowledge of how changes in AMF communities change in function through time is almost nonexistent, but we can make a few predictions. During soil development, the nutrient status, water-holding capacity, and soil pathogen load can change dramatically. Early successional soils are often exposed and dry, whereas older soils may host a greater density of root pathogens. We hypothesize that as a whole, early successional AMF communities in arid and semiarid environments have a greater capacity for drought tolerance whereas AMF in older soils as a whole may be more adept at conferring pathogen resistance to their hosts. If this predicted change in AMF function through time holds, then new crop rotation strategies may be designed. Plants that greatly benefit from AMF-assisted pathogen protection could be used late in rotation or following crops that increase populations of AMF that confer greater pathogen tolerance. It is difficult to predict how the phosphorus acquisition abilities of an AMF community would change over time. In many natural systems, as sites age, soil pH decreases markedly; hence, phosphorus becomes more limiting as it is bound in iron and aluminum complexes. Late successional species of AMF may be more adept at phosphorus scavenging, and this may not be a function of plant allocating carbon to the most beneficial associate. For instance, some species of Acaulospora are able to tolerate low-pH soils and have a greater capacity to uptake phosphorus ( Jakobsen et al., 1992; Porter et al., 1987). These species are characteristic of late successional soils ( Johnson et al., 1991). Thus, as phosphorus becomes less available, AMF species that can tolerate low-pH soil persist and maintain phosphorus scavenging for hosts. To the organic farmer, this suggests that plants that benefit from Acaulospora associations can be rotated into soils of decreased pH to enhance the occurrence of the association. One known phenomenon that affects production is the ‘‘organic transition.’’ The organic transition is a recognized phenomenon that occurs as a conventionally managed agroecosystems is converted to a lower input, organically managed system (Delatea and Cambardella, 2004; Liebhardt et al., 1989). This period of approximately 3 years entails significantly lower yields before returning to higher production rates (Delatea and Cambardella, 2004). This transition has been attributed in part to the microbial community of the soil (Tu et al., 2006). If traditional agriculture selects for less beneficial, more parasitic AMF ( Johnson, 1993), then the lower yields of the organic transition could result from the succession of less beneficial AMF to ones of greater host specificity, as host controls on populations begin to outweigh edaphic factors. Through a better understanding of controls and succession of AMF, perhaps better management

124


practices acting on the AMF community could be employed to shorten the organic transition period.

5. Can We Manage AMF Succession in Organic Agriculture? 5.1. Managing succession of AMF abundance A number of studies have explored innovative practices that can stimulate the abundance of AMF in agricultural fields (e.g., reduced tillage, liming, organic matter additions, and green manures) (reviewed in Gosling et al., 2006). Moreover, additions of living top soils, organic matter, and commercially produced inoculum are all able to temporarily increase the mycorrhizal inoculum potential of a soil. But do the indigenous AMF communities offer a net benefit to the crop species? An exploration of this interaction prior to planting on the local scale may help maximize the benefits of native AMF, determine management strategies to promote proliferation of beneficial AMF groups, or highlight the need to apply specific fungi.

5.2. Maintaining AMF species composition Our current understanding of patterns of AMF species succession relies on only a few studies, and the mechanisms behind the observed shifts between species or groups are even more obscure; however, some controls on AMF species are apparent and may serve as the basis for testing direct manipulation of AMF in field settings. For instance, members of the Acaulosporaceae have been found to be more abundant in mid-to-late succession ( Johnson et al., 1991; Koske and Gemma, 1997). While this could be a product of the family’s slow-growing life history strategy (Hart and Reader, 2002), genera within this family have demonstrated a tolerance of low soil pH (Porter et al., 1987). During soil development or across some long-term organic management regimes, soil pH may decrease, leading to a greater abundance of Acaulospora species. If these species are beneficial to the crop, artificially reducing soil pH may help select for them in cases where the decrease in pH is not detrimental to the crop species. The opposite is possible as well, if Acaulospora associations are not as beneficial to the specific crop, then their abundance may be suppressed through maintaining a neutral to alkaline pH. Another example of a successional change in the abiotic environment that could be managed is soil moisture content. During soil development in both natural and long-term organic systems, the water holding capacity of a soil can increase with increases in organic matter and texture changes. Changes in soil moisture have been shown to alter the colonizing ability of AMF compared with other root colonizing fungi (Lodge, 1989).


125

Also, certain AMF species are able to tolerate a drying environment and confer greater drought tolerance (Ruiz-Luzano et al., 1999). If the goal of AMF management is to confer greater drought tolerance, then cover crops could be water starved to stimulate the abundance of drought-tolerant AMF species like Glomus deserticola, or one could choose a cover crop that promotes significantly more sporulation of G. deserticola than another. We know that AMF species sporulation can be host dependent (Bever et al., 1996). Thus, a better understanding of crop/AMF interactions with respect to sporulation as well as indigenous noncrop AMF-hosting species is necessary. The abundance of certain beneficial species may be stimulated by cover crops and crops that maximize sporulation of a desired AMF species. Crop rotation may be developed to capitalize on positive feedbacks delivered by AMF communities beneficial to multiple plants. Finally, controlling the immigration of undesirable AMF inoculum that could displace the applied or managed community will be critical to maintaining beneficial AMF communities in agroecosystems and reducing repeat inoculations. Preventing the establishment of weeds that are nonmycorrhizal or promote a change to a less beneficial AMF community will help slow succession toward potentially less optimal crop/AMF interactions.

6. Future Research Needs to Improve AMF Application 6.1. Measures of persistence For AMF to be most useful as biofertilizers, they should persist across seasons. If specific fungi are used to maximize benefits, they must remain when confronted by other edaphic or biotic changes. Our understanding of this is limited and presents a huge gap in the knowledge of application of these fungi. Molecular techniques have made identification of AMF species from roots and soil easier, and recent studies are beginning to apply these to monitor the persistence of applied AMF (Farmer et al., 2007).

6.2. Physiological studies and environmental match of AM species and ecotypes There are approximately MD > EF > RE. We summarized the performance of the simulations as poor (), fair (þ/), good (þ), and very good (þþ) (Table 2).

2.2. Models characteristics 2.2.1. N processes simulated The N processes most often simulated (Table 3) were mineralization (86% of studies), leaching and uptake (79%), nitrification (74%), denitrification (55%), volatilization (45%), and symbiotic N fixation (17%). The latter was generally not addressed when simulating N balance, the main models that did take it into account being ANIMO, CROPSYST, DNDC, and EPIC. Only 28% of the models found had calculated all these N processes except for fixation. This means that most models either leave out some N processes

144 Table 2

P. Cannavo et al.

Grid of model performance evaluation

Criteria

þ/

þ

þþ

MD/RMSE/RE (%) r2/EF

>60% Day

Day

Day Crop cycle

Day Day

Day Crop cycle/ year Year Day

Day Day

Day Day

Day Day Day Day

Day Day Day Day

Day Day Day Day Day Day

Day Day Day Day Day Day

Day

Day

147

148

P. Cannavo et al.

because these are able to easily integrate the variability of pollution over time and space according to pedoclimatic conditions, crop succession, and crop management. Indeed, indicators generally give qualitative information at the end of a crop rotation by making a simplified N balance. This makes it possible to aggregate N balances from several fields to address spatial scales larger than the field. The most common time step for making calculations was the one-day step, although indicators commonly use the crop cycle as their calculation time step. The time step of the output data is generally the same as the calculation time step. There were some exceptions with decision support system tools. For example, FARMN presents the N balance at the end of the crop rotation, without a dynamic representation. NGAUGE uses a oneday time step for its calculations but presents monthly results (Table 3). 2.2.3. Ability to simulate different crop species under various conditions Few models are able to simulate the N cycle in the soil–plant system for more than one crop species. Figure 1 shows the proportionate coverage of different crop families by the models in the studies examined. The most Forestry: 5% Others: 4% Arboriculture: 1%

(sugarcane, cotton, tropical products)

Market gardening: 16% Beet: 39% Cabbage: 13% Potato: 13% Tomato: 9%

Cereals: 48% Corn: 28% Wheat: 27% Barley: 18% Ryegrass: 14%

Legumes: 13% Soybean Lucerne Clover

Pasture: 9% Oilseed plant: 4% Rapeseed: 71% Sunflower: 14%

Figure 1

Crop species simulated by models.

Modeling N Dynamics to Assess Environmental Impacts of Cropped Soils

149

widely simulated crops family was cereals (48% of the papers reviewed). Within this family, corn (28%) and wheat (27%) were the most represented species. The other crop families were, in decreasing order: vegetables (16%), market gardening (13%), pasture (9%), forestry (5%), oilseed crops (4%), and arboriculture (1%). Four percent of the publications dealt with tropical crops like sugarcane and cotton. Land area sown to cereals is the largest in the world (FAO, 2001), which may explain why it is the crop family most often concerned by N balance studies and research into the potential pollution risks for groundwater and atmosphere. Looking more closely at the crop families covered by the models, 56 of the 62 models studied simulated N balance under cereal crops. Of these, 20 simulated only one cereal species, 13 simulated 2 species (generally wheat and corn), only two, CERES and CROPSYST, were able to simulate 6 cereal crops. There were 21 models that simulated another crop family (most commonly market gardening or legumes) in addition to the cereal crop family. Finally, 16, 12, and 2 models were capable of simulating 2, 3, and 4 additional crop families, respectively. The EPIC model is one of the few models that simulate cereals and 4 additional crop families. These results highlight the fact that of the many N models available, most deal with specific objectives in specific agricultural and/or pedoclimatic contexts, while a few aim to be as generic as possible and therefore valid for a range of crops and situations. The use of environmental factors (see Section 3.7) generally makes it possible to extrapolate laboratory measurements to field conditions. This can be successful for particular soil and climate characteristics, but it may be difficult to generalize for other contexts. For example, there is no general agreement as to whether a single water correction factor or temperature factor can be used to describe situations in both temperate and tropical climates.

2.3. Conclusions This survey showed that several categories of model exist, and are able to simulate either all the main N processes or some specific ones. N fixation is generally not addressed when simulating N balance. There are probably two main reasons for this. First, in temperate areas with high-input farming systems or fertile soils, the nitrogen balance does not depend highly on N fixation. However, tropical systems are characterized by low N content and require greater attention to management strategies, including the use of leguminous crops (Ledgard and Giller, 1995). Second, it is difficult to simulate N fixation correctly, as it would require a mechanistic approach simulating water and nitrogen stresses on plant and crop development. Our study reveals that nitrate leaching is the N loss process most often calculated by these models. This reflects the level of world concern about groundwater pollution during the 1980–2000 period and the economic and

150

P. Cannavo et al.

environmental concerns that have become prominent since then. By contrast, gas emissions from denitrification and volatilization and their environmental impacts are a more recent preoccupation and so far less attention has been paid to calculating them with models. These two processes are obviously quite difficult to simulate (see Chapter by Bouma et al., this volume); the denitrification process is not clearly understood and not readily predictable ( Jones et al., 2006), and correct estimation of volatilization requires a large number of input variables that can be difficult to access. Finally, our analysis suggests that when building models to be used in different pedoclimatic contexts, it is necessary to test the sensitivity of the model over a wide range of diverse databases, and to elaborate correction functions taking into account the main explanatory factors of a process. The ALFAM project, for example (Sogaard et al., 2002), used a European database to incorporate the soil, climate, and field management factors affecting volatilization following organic waste spreading.

3. Description of Equations Used for Nitrogen Processes Calculations For the various N processes simulated in the models studied, we found that a wide range of concepts were used, varying according to the process considered. For each process, we classified the parameterization by increasing order of complexity (i.e., by increasing order of difficulty of access to the parameter values) (Table 4). This revealed wide differences in the numbers of input parameters and in their accessibility. The processes that presented the highest diversity in equations were N uptake > nitrification > denitrification > mineralization > volatilization > symbiotic N2 fixation ¼ leaching.

3.1. Nitrogen uptake Two concepts are used to calculate N uptake, that is with or without the use of a crop growth model. The calculation of uptake can be simple, or, with models that use a crop growth module, very complex. We will not detail here the different equations used to model crop growth, but it is important to point out that a large number of input parameters are required and that some are far more difficult to establish than others. Increasing complexity is found where the model uses a mechanistic approach to crop growth. Including root dynamics in the model adds further complexity. For example, the STICS and DAISY models simulate root dynamics and require such parameters as root affinity for soil nitrogen, root radius density, and water flux towards the roots (Table 4).

Table 4 Equations used Process

Equations

Models

Mineralization

Net mineralization: NM ¼ ½Norg 0 0:1 for winter manure NM ¼ ½Norg 0 0:2 for autumn manure Soil organic matter mineralization: NM ¼ Ns as Residue mineralization: NM ¼ NR0 kr FT FW Net mineralization: CM;i ðtÞ ¼ CO;i ð1 expðKi tÞÞ and NM;i ðtÞ ¼ CM;i ðN=CÞi

MANNER (Chambers et al., 1999)

Residue mineralization: NM ¼ NR0 kr FT FW FC=N Residue mineralization: NM ¼ NR0 kr FT FW FText qffiffiffiffiffiffiffiffiffiffiffi Residue mineralization: NM ¼ NR0 0:05 FCNP yycc FT Residue mineralization: NM ¼ NRO Fa Fg Fs Net mineralization: CM;i ¼ CO;i Ki FT FW FX and NM;i ¼ CM;i ðN=CÞi Residue mineralization: NM ¼ NR0 kmin WS b Soil organic matter mineralization: NM ¼ kSOM NSOM FT FW ð0:58=10Þ Residue mineralization: CM ¼ kr fr RES FT FW FC=N 1 and NM ¼ C ðC=NÞ 0:0333 r h i Net mineralization: NM ðtÞ ¼

WS060 ðtÞ Vp 10

6:865

2785 TS060 ðtÞþ273

WSFC;060

DEAC ( Jolivel, 2003) MINERVA (Richter et al., 1998) CREAMS (Stockle and Campbell, 1989), AGRIFLUX (Banton et al., 1993) CERES (Ma and Shaffer, 2001) CENTURY (Ma and Shaffer, 2001) EPIC, GLEAMS (Ma and Shaffer, 2001) NGAUGE (Brown et al., 2005) FARMN (Petersen, 2003) AZODYN ( Jeuffroy and Recous, 1999) GPFARM (Shaffer et al., 2004), NLEAP (Shaffer et al., 2001) AGROSIMWinterWheat (Schultz and Mirschel, 1995), TRITSIM (Mirschel et al., 1991)

Net mineralization: NM;i ¼ NPM;i ð1 expðKi tÞÞ

Chowdary et al., 2004

Net mineralization: CM ¼ CRO ðac bc ekr:t gc el:t Þ

CANDY (Franko, 1996), STICS (Nicolardot et al., 2001)

NM ¼ NRO ðaN bN ekr:t gN el:t Þ PNz P Net mineralization: NM ¼ Vp Dt j¼1 FT;j i¼1 FW ;ij P DCi Net mineralization: NM ¼ ðC=N Þ i fe Cim fh Cm i Net mineralization: NM ¼ Cim N Ci fh Nm r0

LIXIM (Mary et al., 1999) DAISY (McGechan and Wu, 2001) SOILN (McGechan and Wu, 2001)

(continued)

Table 4 (continued) 152

Process

Equations

Models

Nitrification

NN ¼ NH4 ð1 expðKn tÞÞ

CREAMS (Stockle and Campbell, 1989), Chowdary et al. (2004) GPFARM (Shaffer et al., 2004), NLEAP (Shaffer et al., 2001) AGRIFLUX (Banton et al., 1993)

NN ¼ Kn NH4 FT FW NO3 NN ¼ Kn NH4 RNO =NH 3

4

NN ¼ WNH3 ½1 expðFT FW FpH Þ dNH4 dt

40 NH4 ¼ A NH SNH4 4 þ90

NN ¼ 0:01 Ae Kn y NH4 ND Kn

NH

NN ¼ K1=2 þ NH4 4 dNH4 dt

1=2 ¼ K1=2 ðNHH4 Þþ O2 exp kEb Ta s 3 ¼ mnit NH4 NO rmax

dNH4 dt d½NO3 dt

4 ¼ R11 mbn Km½NH þ½NH4

NN ¼ NH4 ð0:005 BIOÞ pH

R dD Where BIO ¼ dG dt dt BIO FT FW Fw dG DOC dt ¼ 0:0166 1þDOC þ 1þFw dD 1 dt ¼ 0:008 BIO ð1þDOCÞ=ð1þFw Þ

EPIC (Ma and Shaffer, 2001) CERES (Ma and Shaffer, 2001) ANIMO (McGechan and Wu, 2001) DAISY, CANDY (McGechan and Wu, 2001) NTRM (Ma and Shaffer, 2001) LEACHM (Ma and Shaffer, 2001) STAL (Morvan and Leterme, 2001) DNDC (Li, 2000)

and NO ¼ 0:0025 NN FT and N2 O ¼ 0:0024 NN Volatilization

NV ¼ NH4 Fc

NV ¼ KV NH4 FT Cumulated NV ¼ ½NVmax ½T =ðT þ Km Þ NV ¼ NH4amend expðKv tÞ

Pervanchon et al. (2005), SUNDIAL (McGechan and Wu, 2001), NGAUGE (Brown et al., 2005), FARMN (Hutchings et al., 2001), INDIGO (Bockstaller and Girardin, 2003) GPFARM (Shaffer et al., 2004), NLEAP (Shaffer et al., 2001) MANNER (Chambers et al., 1999) GLEAMS (Ma and Shaffer, 2001)

NV ¼ NH4 ð1 expðKv tÞÞ bi C NV ¼ ai ðkv u ÞrNH3 0:3X z0 2 TTS0 ea ð1 clayÞ NV ¼ NHK4 OH A Denitrification

Chowdary et al. (2004) Geńermont and Cellier (1997) DNDC (Li, 2000)

ND ¼ Namend fdf soil management factor

Pervanchon et al. (2005)

ND ¼ kden ½FNO3 ½FT ½FW

SOILN (Vallejo et al., 2004), NEMIS (Heńault and Germon, 2000) CANDY (McGechan and Wu, 2001) CERES (Ma and Shaffer, 2001)

ND ¼ kden FT FW NO3 CSOM ND ¼ 6:105 FC FW FT L NO3 dND dt

¼

Kden NO3 NO3 þKm

MINERVA (Kersebaum, 1995)

Fw :FT

AGRIFLUX (Banton et al., 1993)

NO2

ND ¼ Kdc NO2 þK3 3

1=2

ND ¼ Kdc NR0 FT ½twet þ FW ðt twet Þ ND ¼ NO3 ½1 expðKdc FW tÞ ND ¼ NO3 ½1 expð1:4 FT CÞ ND ¼

NO3 expð0:0693Tmoy Þtwet FW 100000 zr ymoy

NLEAP, RZWQM (Ma and Shaffer, 2001), GPFARM (Shaffer et al., 2004) CREAMS (Stockle and Campbell, 1989), CROPSYST (Marchetti et al., 1997) EPIC, GLEAMS (Ma and Shaffer, 2001) CREAMS (Saleh et al., 1994)

ND ¼ fdf NO3 W ðg5gf gd Þ

SUNDIAL (McGechan and Wu, 2001)

ND ¼ NO3 ð1 expðKdc tÞÞ

Chowdary et al., 2004

In the absence of overfertilization: ND ¼ Namend 0:0125 Fsoil Fsoilmanag Famendmanag Firr In case of overfertilization:

INDIGO (Bockstaller and Girardin, 2003)

f

153

(continued)

154

Table 4 (continued) Process

Equations

Denitrification (continued)

ND ¼ ðNamend 0:0125 þ Nre 0:0175Þ Fsoil Fsoil

manag

Famend

manag

Firr

For a legume: ND ¼ 3 kg N2 O ha1 dNO3 dt

Kden NO3 ¼ NO 3 þK1=2

ND ¼

Leaching

Models

LEACHM (Ma and Shaffer, 2001)

0:541Dom ð1Ae Þ Ltd

ANIMO (McGechan and Wu, 2001)

ND ¼ minðKa ¼ FW ad xCO2 ; Kt ¼ bd NO3 Þ ND ¼ NR0 min Fd ðNO3 Þ; Fd ðCO2 Þ FW

DAYCENT (Del Grosso et al., 2000)

NL ¼ Dr NO3 NL ¼ NO3 ½leachable fraction Dr

Chowdary et al. (2004) Pervanchon et al. (2005)

NL ¼ NO3 ½½ð0:86 Pe 60:9Þ=WS 0:4

MANNER (Chambers et al., 1999)

NL1 ¼ ðNO3 Þ1 ½1 expð1:2 Dr1 =POR1 Þ,

NLEAP (Shaffer et al., 2001)

DAISY (McGechan and Wu, 2001)

NL2 ¼ ðNO3 Þ2 ½1 expð1:2 Dr2 =POR2 Þ where ðNO3 Þ2 ¼ ðNO3 Þ2 þ NL1 , and NLfin ¼ ðNO3 Þ3 ½1 expð1:2 Drfin =POR3 Þ where NO3 ¼ ðNO3 Þ3 þ NL2 NL ¼ NP þ NH with NP ¼ SNLi and NLi ¼ Namend i Fl f ðPeETMð2=3Þgf Þti where Fl ¼ ðPeETMð2=3Þg Þti þðyCC =10Þ z

INDIGO (Bockstaller and Girardin, 2003)

f

NH ¼ ðNrr þ Nre þ Nmr þ Nmh þ NXm þ NXo NA Þ Flh ðWBW Þ where Flh ¼ ðWBWup ÞþðyupCC =10Þ z Uptake

N ¼ a þ bliscf X þ As þ Bi þ Cf þ Dc þ ej

GNL (Borgesen et al., 2001)

NA ¼ Tr NH4 NA ¼ Ktrans Tr NA ¼ f ðyieldÞ or NA ¼ f ðnitrogen demandÞ

Chowdary et al. (2004) CANDY (Franko, 1996) DEAC ( Jolivel, 2003)

NA ¼ min½ðNdmd ¼ YG TNU f NU Þ; ðNH4 þ ðNO3 Þ1 þ ðNO3 Þ2 þ ðNO3 Þ3 Þ NA ¼ NUPT1 þ NUPT2 where NUPT NUPT1 ¼ min 1þUPT2 =UPT1 ðNO3 Þ pot =ðNO3 Þ 2

1 WS1 =WS2

; Ndisp;1

NLEAP (Shaffer et al., 2001)

N-DICEA (Habets and Oomen, 1994)

and NUPT2 ¼ minðNUPTpot NUPT1 ; Ndisp;2 Þ f R =FC NA ¼ FC NO3 ðWW f R =FC Þþu0 1 NA ¼ ½ð1 þ Fc Þ expð0:075GÞ 1 þ 60 1 expð0:5GÞ P1 þ expðP2 dÞ

dNA dt

¼

NA ¼

Tr c De NA ðNmax NA Þ Pn i¼1 ½Si ðTr;i NO3 þ

CABALA (Battaglia et al., 2004) SUNDIAL (McGechan and Wu, 2001) WHNSIM (Huwe and Totsche, 1995)

NH4 Þ fifre Kei

where i ¼ soil layer Nr N Nr þK NA ¼ NUPTopt rr L Dt FT 1R RNm nupt 3 NA ðNO3 Þ ¼ Fpr Frac BN NONO 3 þNH4

ANIMO (McGechan and Wu, 2001) CREAMS (Stockle and Campbell, 1989) AGRIFLUX (Banton et al., 1993)

4 NA ðNH4 Þ ¼ Fpr Frac BN NONH 3 þNH4

with Fpr ¼ Fð jÞ Fð j 1Þ 1 and F( j ) ¼ 1þexpð1:5976zð1þ0:04417jz 2 ÞÞ m with z ¼ jm s

and mm ¼ matemer and s ¼ matemer 2 6 h2 2 i1 bc ln bc NA ¼ 4pDðC Cr Þ b2 1 1 if m ¼ 0 c

NA ¼ qw NA ¼

Cr ln b2c ðb2c 1ÞC ðb2c 1Þ ln b2c

qw if m ¼ 2 with m ¼ 2pD and bc ¼ ðrr prr Þ1=2

ðbc2m 1ÞCr ðb2c 1Þð1m=2ÞC ðb2c 1Þð1m=2Þðbc2m 1Þ

if m 6¼ 0 and 2

NA ¼ 10 a1 Vc if DMa < 1 t ha

1

DAISY (McGechan and Wu, 2001) STICS (Brisson et al., 1998)

155

(continued)

Table 4 (continued) Process

Equations

Models

1 1 NA ¼ 10 a1 Vc ð1 b1 Þ DMb a if DMa 1 t ha dKcn ðtÞ 3 NA ¼ Kcn ðtÞ dNO dt þ dt NO3

CREAMS (Seppelt, 1999)

where Kcn ðtÞ ¼ K1 if t < tDC else Kcn(t) ¼ ðKmax K0 ÞeðttDC gÞ=t2 ðKmax K1 ÞeðttDC gÞ=t1 þ K0 N2 fixation

NF ¼ Tr 0:234 under tropical conditions

DNDC (Kiese et al., 2005)

DMa NF ¼ rf DM a þKf

CREAMS (Seppelt, 1999)

NF ¼ FXP MinðFXW ; FXN Þ

EPIC (Bouniols et al., 1991)

NF ¼ DMa N % Pfix ð1 þ Prootþstubble þ Ptrans soil þ Ptrans animal þ Pimmobile Þ

FARMN (Hgh-Jensen et al., 2003)

NF ¼ minðFXD ; FXN ; FXP ; FXW Þ

AFISOL (Vocanson, 2006)


157

There are several simple models that do not require a crop growth module to calculate N uptake. They generally use easily accessible input parameters such as crop yield (NLEAP, SUNDIAL) and potential nitrogen uptake (N-DICEA, AGRIFLUX). In this case, the number of input parameters is reduced but depends on the crop species included in the model. Some models consider N uptake as either the crop N demand value or the available soil N value, whichever is the smaller (e.g., NLEAP, STICS, N-DICEA). Most models do not make the distinction between mineral nitrogen species (i.e., preferential uptake for NO3 or NH4þ). Exceptions are Chowdary’s model and EPIC, which consider NH4þ and NO3 uptake, respectively.

3.2. Nitrification Not all models simulate nitrification. Those that do not calculate nitrification assume that when N is mineralized, it instantly takes nitrate form. This is the case for the CENTURY model, for example, and for the majority of indicators. However, nitrification can generate N2O emissions (Prosser, 1986), contributing to the greenhouse effect. These emissions are less than those produced by denitrification, but seem to be nonnegligible (Henault et al., 2005). Moreover, volatilization fluxes also depend on nitrification, as it is a major process of NH4þ consumption in soils and determines the balance between NH4þ transformation into NH3 and into NO3. Thus, it appears more important to simulate nitrification processes in the soil when gas emissions are a matter of concern. Equations used here can be zero-order kinetics (NLEAP, RZWQM, NCSOIL), first-order kinetics (SOILN, SUNDIAL, EPIC, LEACHN), empirical equation (ANIMO), or Michaelis–Menten kinetics (DAISY, CANDY, CERES). Mechanistic models, such as LEACHN or DNDC, include microbial parameters to take nitrifier populations into account. This leads to a large number of parameters that are often difficult to access without laboratory measurements (Table 4). Few models account for soil pH effects, although pH largely controls nitrification (NTRM, DNDC).

3.3. Denitrification Most models use an empirical approach to describe denitrification. There are some exceptions; for example, the DNDC model uses a mechanistic approach, calculating denitrification on the basis of the description of the denitrifying population. The relative growth rate of denitrifiers is calculated using half-saturation values of solute carbon and N oxides. Apart from this exception, models generally use a ‘‘biological parameter’’ to simulate denitrification. This parameter is often a potential denitrification rate obtained in the laboratory under anoxic conditions with excess nitrate and

158

P. Cannavo et al.

organic carbon input (SOILN, NEMIS, CANDY, CERES, MINERVA, LEACHN, CROPSYST, AGRIFLUX, NLEAP). It may also be soil microbial respiration (SUNDIAL, DAISY, DAYCENT) or organic matter content (EPIC and ANIMO) (Table 4). However, a study by Marchetti et al. (1997) showed the EPIC model to be less efficient than CROPSYST and NLEAP. As expected, this highlights the fact that soil organic matter content is a poor biological indicator for predicting denitrification. Finding an appropriate indicator for denitrifier activity in relation to soil and climate properties remains a challenge. Denitrification is highly dependent on soil water content, soil temperature, and soil nitrate content. These parameters are taken into account with correction factors (see more details in Section 3.7). Anaerobic conditions are estimated using soil water content correction functions that are based on the water-filled pore space (WFPS). Some mechanistic models (ANIMO) prefer to use pore-size distribution as obtained from retention curves. Soil water content is generally the most important parameter to calculate accurately because the moisture correction factor activates denitrification when a given moisture threshold is reached.

3.4. Mineralization Two main approaches exist. The first approach consists in calculating net N mineralization (e.g., NLEAP, Table 4). The second approach consists in coupling the C and N cycles and then calculating separately gross N mineralization and gross N immobilization associated with C cycling. Net mineralization is then deduced from the difference between the two fluxes. This approach is mainly used in mechanistic models, which use several organic matter pools and kinetic factors (Benbi and Richter, 2002). They can incorporate a variable number of pools, but usually three pools are considered: the exogenous organic pool (input, e.g., from crop residues, organic waste, and rhizodeposits), the microbial biomass pool, and the humified organic matter pool. In the model, each pool is basically considered as a compartment defined by a quantity of C or N and a specific mineralization rate, which may itself be calculated from organic matter characteristics (e.g., SOILN, CANDY, and STICS). Some models split the exogenous pool into subpools to take into account its heterogeneous chemical composition. In that case, the subpools are either described explicitly in terms of the proportions of particular biochemical fractions (e.g., lignin-like, hemicellulose, cellulose fractions determined by the Van Soest method) (CERES, NCSOIL, CANTIS, Verberne model), or more simply are defined as fractions with fast or slow decomposition rates (CENTURY, CNSIM, WHNSIM). Despite these differences, all models use first-order kinetics. The exogenous and soil organic matter pools are often modeled separately, but in medium- and long-term models, undecomposed exogenous organic


159

matter is often attributed to the soil organic matter, which modifies its composition and the mineralization rate. One challenge is therefore the compatibility between the mineralization timescales of the exogenous organic matter and soil organic matter pools. Moreover, correction factors are required to take into account the effects of temperature and moisture because the standard coefficients used to parameterize the equations are obtained under constant conditions (see Section 3.7).

3.5. Volatilization When urea is applied to soil (in chemical fertilizer or animal manure), volatilization of ammonia occurs only after the urea has hydrolyzed, but this process is rarely taken into account in models. It is usually assumed that urea is quickly transformed into NH4þ (i.e., there is no simulation of the hydrolysis stage). CERES and NTRM estimate volatilization as a proportion of applied urea-N using standard values, whereas Chowdary’s model uses a first-order kinetic based on a urea hydrolysis constant, and thereafter calculates volatilization using a volatilization rate. A mechanistic approach for describing volatilization is found in models such as DNDC, VOLT’AIR, and STAL. Calculations involve numerous input parameters that are difficult to access. However, these models have the advantage of integrating key parameters affecting volatilization such as soil pH, clay, and organic matter content. VOLT’AIR is one of the few models able to incorporate both organic and mineral forms of fertilizer. Many models take an empirical approach based on experiments and/or the literature. Models commonly include a volatilization factor (SUNDIAL, NGAUGE, FARMN, INDIGO), or a volatilization rate (GPFARM, NLEAP, MANNER, GLEAMS), calibrated or measured in the pedoclimatic context in which the model is then used (Table 4).

3.6. Fixation Symbiotic N2 fixation is the process least often calculated in models. It is calculated in various ways. DNDC uses a basic empirical approach based on crop transpiration. Others models try to incorporate the main factors affecting fixation. One of them is crop biomass production. The CREAMS model regards this crop factor as a Michaelis–Menten kinetic. EPIC and AFISOL models regard fixation as combining a minimum of three and four stress factors, respectively. These factors take into account crop development (based on degree days since emergence), soil water stress (relative to the soil water reserve), soil nitrogen content (linear regression before and after flowering), and, for AFISOL, the crop’s nitrogen needs (according to the maximum N accumulation rate in the crop). FARMN divides fixation between several components of the ecosystem—soil, legumes, and grasses.

160

P. Cannavo et al.

This model operates with the ratio of fixed N in belowground leguminous plant tissue, the ratio of fixed N transferred belowground to the grass, the ratio of fixed N transferred to the grass from the grazing animals, and the ratio of fixed N immobilized in the soil organic pool by rhizodeposition.

3.7. Leaching Unlike the previous processes, leaching is not calculated by widely different methods. The nitrate concentration in drained water is usually calculated from leached soil NO3 and water drainage. However, the water drainage calculation can differ between models. Two types of water balance model exist, using either Darcy’s law (e.g., ANIMO, CANDY, CERES, DAISY, LEACHN, PASTIS, RZWQM, SOILN, and WHNSIM) or soil reservoir cascade theory (e.g., CENTURY, EPIC, NCSOIL, SUNDIAL, NLEAP). Some models like MANNER and INDIGO incorporate the water balance in their leaching equations, requiring some accessible inputs such as rainfall, evapotraspiration, and soil water content. An original conception comes from the NLEAP model, which uses an exponential equation for drainage and introduces soil porosity (Table 4). Water balance calculations, using a reservoir cascade scheme, reach their limits when soil water content is above field capacity. Water drains between two reservoirs when the soil water stock of one is above field capacity. The excess is drained to the next reservoir. So it is a priori impossible to simulate soil water content between field capacity and saturation. This is a severe limitation, especially for calculating denitrification. Some models like STICS have tried to solve this problem by taking soil macroporosity into account. The idea is that rain or irrigation water first fills the micropores of the topsoil layer and then the macropores, creating anoxic conditions. The water stored in the macropores will then fill the micropores of the underlying soil layer, and so on.

3.8. Correction factors applied to N transformation rates Correction factors are used to extrapolate the rate of a process that has been parameterized in optimal (laboratory) conditions or given field conditions, to other soil or climatic conditions. Correction factors taking into account the effects of soil water content and temperature on N processes are very commonly used. Their formalisms are usually simple but can vary widely between models for a given N process. Rodrigo et al. (1997) studied nine models and showed that the difference between the lowest and highest mineralization rates among them was about 325%, resulting solely from differences in the temperature factors. This underlines the importance of choosing appropriate correction functions that are valid for the conditions of the situation modeled.


161

Thus, it is crucial to correctly model the water balance in the soil–crop system and water content in each soil layer. We observed two formalisms for a soil water correction factor: (1) an Arrhenius function describing the effect of water potential on N processes (e.g., CENTURY) and (2) the ratio between the actual matric potential and the optimum matric potential for the process considered (e.g., ANIMO, DAISY, PASTIS, SUNDIAL), or the ratio of actual to optimum water content used in a similar way (e.g., CANDY, CERES, EPIC, GLEAMS, LEACHN, SOILN, STICS). Soil temperature also affects all biological and physical processes, including mineralization, nitrification, denitrification, volatilization and, indirectly, N uptake and fixation. Thus, temperature correction factors are applied to the potential rates of N processes (Table 4). Several equations are used: (1) an Arrhenius function (ANIMO, EPIC, GLEAMS, NLEAP, WHNSIM), (2) the Q10 law (CANDY, LEACHN, SOILN), and (3) linear or exponential relations (CENTURY, CERES, DAISY, NCSOIL, PASTIS). The soil temperature and a reference temperature are the main parameters required in all three equation types. The effect of temperature can be appreciated during freeze–thaw cycles in models such as SOILN (Blomba¨ck et al., 2003) and ANIMO, in addition to changes in hydraulic properties. Other correction factors are also used. For example, mineralization may require correction factors to account for the effects of clay content (e.g., CENTURY, FARMN), the C-to-N ratio of crop residues (CERES, EPIC), or biomass dependence (CANTIS). The RZWQM model uses a correction factor for oxygen and hydrogen contents. Nitrification and volatilization require a correction factor for soil pH (SOILN, CERES, WHNSIM). Soil mineral N correction factors are used to calculate denitrification (NEMIS, MINERVA, AGRIFLUX).

4. Critical Analysis of the Models 4.1. Equations used in models This synthesis has shown that many formalisms are available for modeling N processes. This is partly because different N models have different goals and therefore different priorities in describing the processes concerned. That is why some modelers have preferred to maximize their models’ simulation quality for particular processes. The models’ equations vary especially widely where biological processes such as N uptake, nitrification, and mineralization are concerned because biological processes associated with plant growth, soil microbial dynamics, and enzyme activities are still not well understood. They also generally involve many parameters that are difficult to access. Input parameters for a given process therefore vary widely from model to model, but some models’ conceptualizations are simplified in

162

P. Cannavo et al.

accordance with the model’s objectives. One example is the microbial pool in mineralization: ANIMO, CERES, and LEACHN ignore this pool. Another example is the omission of the nitrification process in models such as CENTURY, SUCROS, NCYCLE, and CREAMS, which assume that NH4þ is instantly transformed into NO3. This assumption is probably acceptable for most agricultural soils with high pH and frequent fertilization, but certainly has its limits for acidic, hydromorphic, or tropical conditions. One major challenge is to correctly simulate mineralization for all type of soils and for a wide range of organic inputs (crop residues, roots and root exudates, waste, compost), which are subject to wide variations in their chemical and physical characteristics and the timing of their return to the soil. Some models are able to simulate mineralization of organic waste, slurry, and crop residues (e.g., ANIMO, DAISY, MANNER, SOILN, and STICS). However, it is difficult for such models to incorporate the longterm effects of repeated inputs of organic matter after the current crop cycle, although this can considerably affect nitrogen cycling (ShengMao et al., 2007). The processes presenting the least diversity among equations were, in decreasing order, volatilization, leaching, and fixation. Leaching is rather easy to simulate and the formalisms proposed found a good consensus in the literature. By contrast, fixation concerns a limited number of crops and requires a mechanistic approach based on the factors that affect this process (i.e., crop growth, soil water, and nitrogen content). For these reasons, it seems that this process is often left out of account or that an empirical approach is used. Improvements are needed to assess the impact of these stress factors on the fixation process. Some investigations have already been undertaken for pea (Vocanson, 2006) and should be reproduced for other species. Finally, being able to predict volatilization is important because of its environmental impact. It appears that the physical and chemical mechanisms involved have been well characterized, as has their dependence on agricultural practices (e.g., forms of N applied, N concentrations, role of spreading, or incorporation). By contrast, NH3 emissions remain difficult to predict (see Section 4.3). The main reason for this seems to be the large number of parameters involved, which are difficult to incorporate and not always easy to access. Thus, as with N fixation, the difficulty in parameterizing volatilization modules often leads scientists to prefer a very empirical approach.

4.2. Model performance CERES, CENTURY, EPIC, APSIM, and SOILN are the models most often published, with more than 45 published papers based on their use (Table 5) (numbers of publications indicated are those obtained when entering the model name as topic in the CAB Abstracts database).

Table 5 Models’ capacities to simulate experimental data for each N process and crop growth Model

CERES CENTURY/ DAYCENT EPIC APSIM/ I-WHEAT SOILN/SOILSOILN GLEAMS CROPSYST DAISY NCSOIL RZWQM DNDC NLEAP STICS AFRCWHEAT CREAMS LEACHN SUNDIAL NCSWAP NTRM ANIMO PASTIS Sinclair SOILNDB MANNER

Publications number

Mineralization

92 83

þ þ

63 48

þ

45

þ

35 35 27 26 24 24 23 21 20 14 13 13 10 10 10 10 8 5 5

Leaching

Denitrification

Absorption

þ þ

þ/

þ/ þ

þ/

þ þ

þ þ

þ

þ

þ

þ

þ

þ

Volatilization

Nitrification

þ/ þ/ þ þ/ þ

þ/ þ/ þ/ þ

þ þ

þ/ þ/ þ

þ/ þ/

þ þ þ/ þ/ þ þ

þ þ/ þ/ þ/

þ/ þ/

þ/

N2 Fixation

Crop growth

þ þ þ/

þ þ þ

þ þ þ

þ þ þþ þ

þ þ

þ þ

þ þ/ þ þ/ þ

þ þ/ þ

(continued)

Table 5 (continued) Model

MINERVA Chowdary model WHNSIM SOMM AGRIFLUX CANDY AZODYN VOLT’AIR LIXIM NGAUGE TRITSIM N-DICEA AGROSIM CABALA CN-SIM GPFARM NOE/NEMIS GNL STAL

Publications number

5 4 4 4 4 3 3 3 3 3 3 2 2 2 2 2 2 1 1

Mineralization

Leaching

Volatilization

þ/

þ þ þ þ/ þ þ/ þ

Nitrification

Denitrification

Absorption

N2 Fixation

þ/ þ

þ

Crop growth

þ

þ þ

þ þ

þ þ/

þ þ

þ

þþ

þ

þ þ

þ/ þ þ/

þ þþ

þ

þ þ

þ/

þþ

Symbols refer to Table 2. Blanks signify either that we were not able to evaluate model performance or that the model does not simulate this process.

þ/ þ


165

Few papers show comparisons between model simulations and field measurements for N processes. Most papers are purely simulations and scenario tests. Some models (e.g., CERES) are more focused on crop yield simulation and therefore produce few references for N process performances. This is particularly true for volatilization. This underlines the scarcity of field measurements designed to adequately evaluate model performance. Some models have only been used in the pedoclimatic area where they were calibrated, suggesting that their ability to perform in other pedoclimatic conditions has not been tested. Generally speaking, it is the most recent models that have been least often used in published studies. These recent models are mainly deterministic, incorporate modules from other published models, and are intended as decision support tools. For example, GPFARM (Great Plains Framework For Agricultural Resource Management) aims to determinate the economic and environmental sustainability of farm and crop management systems. It uses NLEAP as its C and N cycles module. The decision support tool NGAUGE uses the NCYCLE model for the nitrogen balance. The deterministic model TNT2 (topography-based nitrogen transfer and transformation) uses a modified version of STICS. AGROSIM is a crop growth model coupled with TRITSIM for wheat growth (Schultz and Mirschel, 1995) and with CANDY for beet growth (Franko and Mirschel, 2001). It was difficult to analyze model performance as most of the published works do not present statistical data. Many of the papers only compare simulated and experimental data visually, the experimental data sometimes having no standard error bar. In such cases, the most relevant performance criterion is the absolute relative error estimation. The ‘‘mean’’ performance of each model was deduced from the analysis of all papers concerned (Table 5). However, the ability of a model to simulate observed data also depends on crop species and pedoclimatic conditions. For example, nitrification in SUNDIAL presented relative errors of more than 90% for a beet crop and a maize–barley rotation (Gabrielle et al., 2000) and an RMSE of 14% for winter wheat (Gabrielle et al., 2003). With regards to mineralization, N-DICEA has shown fair performance with a beet crop (Habets and Oomen, 1994) but has performed better than the previous crop (RMSE of 1.2% and EF of 0.25) with a market gardening rotation (Koopmans and Bokhorst, 2002). When assessing model performance process by process, we found an overall good performance for processes such as leaching, mineralization, nitrification, and uptake, although for the latter two processes, the models varied widely in their conceptualization. Crop growth was also usually well simulated by models. By contrast, performance for denitrification and volatilization was poor. At present, it is obvious that the microbial processes involved in denitrification and their interactions with soil physical and chemical properties are not well known and therefore not well described

166

P. Cannavo et al.

mathematically. The absolute relative error frequently reached 1000%; this level of error is probably due to the spatial heterogeneity of the process (Marchetti et al., 1997). DNDC was the most successful model for simulating denitrification in various pedoclimatic conditions, with relative errors of about 50%. Few papers have examined volatilization, although there are 28 models that can calculate volatilization, only 6 papers made the comparison between simulated and measured data. To accurately model a specific process, one must be able to accurately simulate the underlying parameters. For example, we cannot expect a model to give a precise and satisfactory estimate of N2O emissions by denitrification if soil NO3 concentrations are not accurately estimated. Whichever the N process simulated, we did not observe a lower performance for empirical models compared to the more mechanistic ones. This shows that empirical approaches, with fewer and more accessible data, can be as precise as mechanistic models. However, many of the empirical models were developed to apply to a specific context and performed well in that context. They have also been covered in fewer published papers so it is difficult to assess their adaptability to other contexts. Many mechanistic models have proven their ability to satisfactorily simulate a range of situations. Nonetheless, at present there is a lack of model testing and validation.

5. Conclusions: Current Limits and Challenges When de Willigen (1991) compared 14 simulation models for nitrogen behavior in the vadose zone, he concluded that such models simulated N behavior more or less satisfactorily. His three main conclusions were (1) heterogeneity in the number of N processes taken into account and in their degree of complexity; (2) failure to integrate biological processes, particularly C–N coupling; and (3) lack of model testing at the field scale. Since then, many models have been elaborated and performed in order to couple physical and biological processes. This was the main trend that emerged from the present study. However, simulation quality generally decreased when an effective model of organic matter cycling was associated with other modules such as crop growth (Smith et al., 1997). This is due to the uncertainties generated by the models after coupling with soil organic matter decomposition (Gabrielle et al., 2002). The main trend in the shift from mechanistic models to functional models has been a simplification of the equations involved and the use of correction functions. This simplification often consists in using potential values, measured in specific contexts, and then adjusted by correction factors (e.g., soil temperature, water content, and mineral N content).


167

These functional models are often used to estimate volatilization, denitrification, and nitrogen uptake. However, the equations used and the parameterization of the correction functions vary widely between models and this leads to large differences in simulated rates, irrespective of the model formalisms themselves (Rodrigo et al., 1997). Another trend is the frequent adoption of models in which the core element is a mechanistic module. Researchers often modify these modules to adapt them to their own contexts (see Section 2.2.3). Thus, it is difficult to access the details of these new models without consulting the authors of the study because we do not know how much the initial model has been modified. Clearly, there is a need for modular systems with modules describing various processes with various level of complexity from which models can be built ‘‘a` la carte,’’ depending on the requirements—as was done for the APSIM model (McCown et al., 1996). With respect to indicators, the first difficulty is to find methods to evaluate their validity. They generally give qualitative information and are intended for decision aid purposes. They characterize particular situations and it is difficult to select a variable to validate them. The second difficulty is to evaluate the environmental impacts of practices when these impacts generally cannot be directly perceived at field level. Some indicators such as INDIGO and DEAC evaluate environmental impacts of practices at field or farm level using environmental indicators correlated to the contribution of the field or the farm to the global nuisance. Decision support tools follow the same trend ( Jorgensen et al., 2005); however, they still do not incorporate many crop species, which is necessary for scenario simulations. Few of them have been validated on crop rotations. One model that does is ROTOR (Bachinger and Zander, 2007) that is able to generate and evaluate site-specific and agronomically sustainable crop rotations for organic farming systems in Central Europe. A set of annual crop production activities is assembled semiautomatically from a single site and crop-specific field operations using a relational database. The rotation scale is the relevant one for assessing the impacts of changes in cropping systems. Predicting N losses accurately at rotation scale needs calculations of N harvest index, N partitioning between shoots and roots and root characteristics (e.g., density, rooting depth). Progress in the modeling of medium- and long-term changes in the C and N content of the soil should, in the near future, provide new tools for managing organic matter at the rotation level (Meynard et al., 2002). On this timescale, the impact of the parameters of a single plant on the model outputs is smoothed. The work of Beaudoin et al. (2005) suggests that the accuracy of simulations obtained by running a model continuously over a rotation would be more sensitive to the pedoclimatic variables that drive soil–crop functioning: labile organic N pool, maximum rooting depth, and N fertilizer use efficiency. Finally, very few models aim to rank agricultural practices according to economic and/or

168

P. Cannavo et al.

environmental objectives before evaluating them experimentally (Tixier et al., 2007). Therefore, integrating multicriteria analysis into models to evaluate crop successions remains a major challenge for the future. At present, then, despite considerable in-depth scientific knowledge of N loss, no model correctly simulates all N losses. This chapter shows that the use of dynamic crop models for fertilizer management can only progress as the number of input variables—inherently large for a majority of models—is reduced, without affecting their predictive value. Future models will need to develop generic equations capable of simulating crop growth in a larger range of crop species. This will mean finding formalisms simple enough to limit the number of parameters, and easy to access for numerous crops. Finally, to evaluate environmental impacts, the uncertainty of the input data describing soil and climate will have to be incorporated, as is more widely done, for example, in hydrology (Thorsen et al., 2001). Modeling approaches should significantly facilitate risk assessment—but these approaches will not be satisfactory unless the climate is accurately predicted. We suggest two modeling methods, depending on the users and objectives concerned. The first is suitable for scientists building integrated frameworks for computer models. This approach facilitates interconnecting modules selected according to need. For example, the SEAMLESS framework (www.seamless-ip.org) adopts this strategy and can be used to assess how future alternative agricultural and environmental policies would affect sustainable development in Europe. The second is suitable for agricultural advisers and environmental agency managers who need tools to produce benchmarks that are easy to evaluate and communicate, for agricultural development purposes. These users often question the scientific reliability of the simple tools they are currently using. This second approach may involve users in the design of new tools (Dore´ et al., 2008; Meynard et al., 2002). This implies the development of multidisciplinary models combining agricultural science (e.g., indicators) and social science (e.g., models describing the knowledge and activities of the users).

ACKNOWLEDGMENTS The authors wish to thank Dr. S. Pellerin (INRA) for his fruitful comments. The authors also wish to thank Ms. Harriet Coleman for reviewing the English version of the manuscript.

REFERENCES Acutis, M., Ducco, G., and Grignani, C. (2000). Stochastic use of the LEACHN model to forecast nitrate leaching in different maize cropping systems. Eur. J. Soil Sci. 13, 191–206. Ahuja, L. R., Rojas, K. W., Hanson, J. D., Shaffer, M. J., and Ma, L. (2000). ‘‘Root Zone Water Quality Model: Modeling management effects on water quality and crop production.’’ p. 372. Water Resources Publications, LLC, Highland Ranch, CO.


169

Arnold, J. G., Srinivasan, R., Muttiah, R. S., and Williams, J. R. (1998). Large area hydrologic modeling and assessment. Part I: Model development. J. Am. Water Resour. Assoc. 34, 73–89. Bachinger, J., and Zander, J. (2007). ROTOR, a tool for generating and evaluating crop rotations for organic farming systems. Eur. J. Agron. 26, 130–143. Banton, O., Larocque, M., Surateau, F., and Villeneuve, J.-P. (1993). Agriflux: Logiciel d’e´valuation des pertes environnementales de contaminants agricoles. Manuel d’usager. Scientific Report No. R-380. Ste-Foy, QC, Canada, INRS-eau, 146pp. Battaglia, M., Sands, P., White, D., and Mummery, D. (2004). CABALA: A linked carbon, water and nitrogen model of forest growth for silvicultural decision support. Forest Ecol. Manage. 193, 251–282. Beaudoin, N., Saad, J. K., Laethem, C. V., Machet, J. M., Maucorps, J., and Mary, B. (2005). Nitrate leaching in intensive agriculture in Northern France: Effect of farming practices, soils and crop rotations. Agric. Ecosyst. Environ. 111, 292–310. Beaujouan, V., Durand, P., Ruiz, L., Aurousseau, P., and Cotteret, G. (2002). A hydrological model dedicated to topography-based simulation of nitrogen transfer and transformation: Rationale and application to the geomorphology-denitrification relationship. Hydrol. Process. 16, 493–507. Benbi, D. K., and Richter, J. (2002). A critical review of some approaches to modelling nitrogen mineralization. Biol. Fertil. Soils 35, 168–183. Blomba¨ck, K., Eckersten, H., KLewan, E., and Aronsson, H. (2003). Simulations of soil carbon and nitrogen dynamics during seven years in a catch crop experiment. Agric. Syst. 76, 95–114. Bochu, J.-L. (2002). PLANETE: Me´thode pour l’analyse eńerge´tique de l’exploitation agricole et l’e´valuation des e´missions de gaz a` effet de serre. Quels diagnostics pour quelles actions agroenvironnementales? pp. 68–80. Les Actes, Toulouse, France, Solagro. Bockstaller, C., and Girardin, P. (2003). Mode de calcul des indicateurs agri-environnementaux de la me´thode INDIGO (version 1.51 du logiciel). p. 115. Colmar, France, INRA. Boote, K. J., Jones, J. W., and Pickering, N. B. (1996). Potential uses and limitations of crop models. Agron. J. 88, 704–716. Borgesen, C. D., Djurhuus, J., and Kyllingsbaek, A. (2001). Estimating the effect of legislation on nitrogen leaching by upscaling field simulations. Ecol. Modell. 136, 31–48. Bouniols, A., Cabelguenne, M., Jones, C. A., Chalamet, A., and Charpenteau, J. L. (1991). Simulation of soybean nitrogen nutrition for a silty clay soil in southern France. Field Crops Res. 26, 19–34. Bouwman, A. F., Boumans, L. J. M., and Batjes, N. H. (2002). Estimation of global NH3 volatilization loss from synthetic fertilizers and animal manure applied to arable lands and grasslands. Global Biogeochem. Cycles 16, 1–15. Bradbury, N. J., Whitemore, A. P., Hart, P. B. S., and Jekinson, D. S. (1993). Modelling the fate of nitrogen in crop and soil in the years following application of 15N-labelled fertilizer to winter wheat. J. Agric. Sci. 121, 363–379. Brisson, N., Mary, B., Ripoche, D., Jeuffroy, M. H., Ruget, F., Nicoullaud, B., Gate, P., Devinet-Barret, F., Antonioletti, R., Durr, C., Richard, G., Beaudoin, N., et al. (1998). STICS: A generic crop model applied to wheat and corn. Agronomie 18, 311–346. Brown, L., Scholefield, D., Jewkes, E. C., Lockyer, D. R., and del Prado, A. (2005). NGAUGE: A decision support system to optimise N fertilisation of Bristish grassland for economic and environmental goals. Agric. Ecosyst. Manage. 109, 20–39. Chambers, B. J., Lord, E. I., Nicholson, F. A., and Smith, K. A. (1999). Prediciting nitrogen availability and losses following application of organic manures to arable land: MANNER. Soil Use Manage. 15, 137–143.

170

P. Cannavo et al.

Chowdary, V. M., Rao, N. H., and Sarma, P. B. S. (2004). A coupled soil water and nitrogen balance model for flooded rice fields in India. Agric. Ecosyst. Manage. 103, 425–441. Christensen, L., Riley, W. J., and Ortiz-Monasterio, I. (2006). Nitrogen cycling in an irrigated wheat system in Sonora, Mexico: Measurements and modeling. Nutr. Cycl. Agroecosyst. 75, 175–186. Cole, C. V. (1996). Agricultural options for mitigation of GHG emissions. In ‘‘Climate Change 1995, Second Assessment Report.’’Chapter 23, pp. 745–771. Cambridge University Press Publication, UK. Dai, Q., Flectcher, J. J., and Lee, J. G. (1993). Incorporating stochastic variables in crop response models: Implications for fertilization decisions. Amer. J. Agric. Econ. 75, 377–386. De Cara, S., Houze´, M., and Jayet, P.-A. (2005). Methane and nitrous oxide emissions from agriculture in the EU: A spatial assessment of sources and abatement costs. Environ. Res. Econ. 32, 551–583. de Willigen, P. (1991). Nitrogen turnover in the soil-crop system. Comparison of fourteen simulation models. Fertil. Res. 27, 141–149. Del Grosso, S. J., Parton, W. J., Mosier, A. R., Ojima, D. S., Kulmala, A. E., and Phongpan, S. (2000). General model for N2O and N2 gas emissions from soils due to denitrification. Global Biogeochem. Cycles 14, 1045–1060. Del Grosso, S. J., Parton, W. J., Mosier, A. R., Hartman, M. D., Brenner, J., Ojima, D. S., and Schimel, D. S. (2001). Simulated interaction of carbon dynamics and nitrogen trace gas fluxes using the DAYCENT model. In ‘‘Modeling Carbon and Nitrogen Dynamics for Soil Management,’’ pp. 83, 303–332. CRC Press, Boca Raton, FL. Dore´, T., Clermont-Dauphin, C., Crozat, Y., David, C., Jeuffroy, M.-H., Loyce, C., Makowski, D., Male´zieux, E., Meynard, J.-M., and Valantin-Morison, M. (2008). Methological progress in on-farm regional agrnomic diagnosis. A review. Agron. Sustain. Dev. 28 (in press). Dubrulle, P., Machet, J.-M., and Damay, N. (2003). Azofert: A new decision support tool for fertiliser N recommendations. 12th Nitrogen Workshop, Exeter, Devon, UK, 21–24 September, Wageningen Academic Publisher, pp. 500–501. FAO (2001). In ‘‘Production Yearbook.’’ FAO, Rome, Italy. Franko, U. (1996). Modelling approaches of soil organic matter turnover within the CANDY system. In ‘‘Evaluation of Soil Organic Matter Models,’’ pp. 247–254. Springer-Verlag, Berlin Heidelberg, Berlin. Franko, U., and Mirschel, W. (2001). Integration of a crop growth model with a model of soil dynamics. Agron. J. 93, 666–670. Freibauer, A., and Kalschmitt, M. (2000). Emission rates and emissions factors of greenhouse gas fluxes in arable and animal agriculture. European summary report of the EU Concerted Action ‘‘Biogenic Emissions of Greenhouse Gases cause by Arable and Animal Agriculture’’ (FAIR3-CT96–1877) financed by EU DGVI Project Report Task 1. Stuttgart, Institut fu¨r Energiewirtschaft und Rationelle Energianwendung (Universita¨t Stuttgart). http://opus-ho.uni-stuttgart.de/hop/volltexte/2002/22/pdf/Freibauer_diss. pdf, 196pp. Gabrielle, B., Agostini, F., and Donatelli, M. (2000). Limits to the accuracy of the water component of a decision (support-oriented agronomic model). Ital. J. Agron. 3, 87–99. Gabrielle, B., Roche, R., Angas, P., Cantero-Martinez, C., Cosentino, L., Mantineo, M., Langensiepen, M., Heńault, C., Laville, P., Nicoullaud, B., and Gosse, G. (2002). A priori parametrisation of the CERES soil-crop models and tests against several European data sets. Agronomie 22, 119–132.


171

Gabrielle, B., Mary, B., Roche, R., Smith, P., and Gosse, G. (2003). Simulation of carbon and nitrogen dynamics in arable soils: A comparison of approaches. Eur. J. Agron. 18, 107–120. Garnier, P., Neél, C., Mary, B., and Lafolie, F. (2001). Evaluation of a nitrogen transport and transformation model in a bare soil. Eur. J. Soil Sci. 52, 253–268. Geńermont, S., and Cellier, P. (1997). A mechanistic model for estimating ammonia volatilization from slurry applied to bare soil. Agric. Forest Meteorol. 88, 145–167. Habets, A. S. J., and Oomen, G. J. M. (1994). Modeling nitrogen dynamics in crop rotations in ecological agriculture. Nitrogen mineralization in agricultural soils. pp. 255–268. Institute for Soil Fertility Research, Haren, NL, AB-DLO Thema’s, AB-DLO, Haren. Hanks, J., and Ritchie, J. T. (1991). ‘‘Modelling Plant and Soil Systems.’’ Madison, WI, 545pp. Soil Science Society of America. Hansen, S., Jensen, H. E., Nielsen, N. E., and Svendsen, H. (1991). Simulation of nitrogen dynamics and biomass production in winter wheat using the Danish simulation model DAISY. Fertil. Res. 27, 245–259. Heńault, C., and Germon, J. C. (2000). NEMIS, a predictive model of denitrification on the field scale. Eur. J. Soil Sci. 51, 257–270. Henault, C., Bizouard, F., Laville, P., Gabrielle, B., Nicoullaud, B., Germon, J. C., and Cellier, P. (2005). Predicting in situ soil N2O emission using NOE algorithm and soil database. Global Change Biol. 11, 115–127. Hgh-Jensen, H., Loges, R., Jensen, E. S., Jrgensen, F. V., and Vinther, F. P. (2003). Empirical model for quantification of symbiotic nitrogen fixation in leguminous crops. Agric. Syst., unpublished. Available from: http://orgprints.org/316/, 31pp. Houot, S., Cadot, L., and Molina, J. A. E. (1996). Simulation by NCSWAP of the nitrogen dynamics under crop amended with sewage sludge in two soils. In Progress in Nitrogen Cycling Studies: Proceedings of the 8th Nitrogen Workshop held at the University of Ghent, 5–8 September 1994. Kluwer Academic Publishers, Dordrecht, Netherlands, 1996, 18 ref., pp. 413–418. Hutchings, N. J., Sommer, S. G., Andersen, J. M., and Asman, W. A. H. (2001). A detailed ammonia emission inventory for Denmark. Atmos. Environ. 35, 1959–1968. Huwe, B., and Totsche, K. U. (1995). Deterministic and stochastic modelling of water, heat and nitrogen dynamics on different scales with WHNSIM. J. Cont. Hydrol. 20, 265–284. Jeuffroy, M.-H., and Recous, S. (1999). Azodyn: A simple model simulating the date of nitrogen deficiency for decision support in wheat fertilization. Eur. J. Agron. 10, 129–144. Johnsson, H., Bergstrom, L., Jansson, P.-E., and Paustian, K. (1987). Simulated nitrogen dynamics and losses in a layered agricultural soil. Agric. Ecosyst. Environ. 18, 333–356. Johnsson, H., Larsson, M. H., Martensson, K., and Hoffman, M. (2002). SOILNDB: A decision support tool for assessing nitrogen leaching losses from arable land. Environ. Modell. Softw. 17, 505–517. Jolivel, C. (2003). AQUALEA: Guide de reálisation du diagnostic. La Chapelle St-Sauveur, ARVALIS—Institut du Ve´ge´tal, 22pp. Jones, C. G., Gutierrez, J. L., Groffman, P. M., and Shachak, M. (2006). Linking ecosystem engineers to soil processes: A framework using the Jenny State Factor Equation. In ICSZ—Soil animals and ecosystem services. Proceedings of the XIVth International Colloquium on Soil Zoology, Rouen, France, 30 August–3 September 2004. Eur. J. Soil Biol. 42, S39–S53. Jorgensen, M. S., Detlefsen, N. K., and Hutchings, J. (2005). FarmN: A decision support tool for managing nitrogen flow at the farm level. EFITA Conference, Vila Rea, Portugal, pp. 1–7. http://130.226.173.223/farmn. Kersebaum, K. C. (1995). Application of a simple management model to simulate water and nitrogen dynamics. Ecol. Modell. 81, 145–156.

172

P. Cannavo et al.

Kiese, R., Li, C. S., Hilbert, D. W., Papen, H., and Butterbach-Bahl, K. (2005). Regional application of PnET-N-DNDC for estimating the N2O source strength of tropical rainforests in the Wet Tropics of Australia. Global Change Biol. 11, 128–144. Knisel, W. G. (1993). ‘‘GLEAMS, groundwater loading effects of agricultural management systems. Version 2.10.’’ UGA-CPES-BAED 5, 259pp. Koopmans, C. J., and Bokhorst, J. (2002). Nitrogen mineralisation in organic farming systems: A test of the NDICEA model. Agronomie 22, 855–862. Lafolie, F., Bruckler, L., Cockborne, A. M. D., and Laboucarie, C. (1996). Modeling the water transport and nitrogen dynamics in irrigated salad crops. Irrig. Sci. 17, 95–104. Lanquetuit, D., and Sebillotte, M. (1997). ‘‘Protection de l’eau: le guide FERTI-MIEUX pour e´valuer les modifications de pratiques des agriculteurs.’’ ANDA Eds., Paris, 179pp. Le Cadre, E. (2004). Mode´lisation de la volatilisation de l’ammoniac en interaction avec les processus chimiques et biologiques du sol: le mode`le Volt’Air. Ecole doctorale ABIES. Paris, France, INA-PG, 211pp. Ledgard, S. F., and Giller, K. E. (1995). Atmospheric N2 fixation as an alternative N source. In ‘‘Nitrogen Fertilization in the Environment’’ (P. E. Bacon, Ed.), pp. 129–169. Marcel Dekker, New York. Li, C. (1996). The DNDC model. In ‘‘Evaluation of Soil Organic Matter Models.’’ pp. 263–267. Springer-Verlag, Berlin Heidelberg, Berlin. Li, C. (2000). Modeling trace gas emissions from agricultural ecosystems. Nutr. Cycl. Ecosyst. 58, 259–276. Lin, Z., and Routray, J. K. (2003). Operational indicators for measuring agricultural sustainability in developing countries. Environ. Manage. 31, 34–46. Ma, L., and Shaffer, M. J. (2001). A review of carbon and nitrogen processes in nine U.S. soil nitrogen dynamics models. In ‘‘Modeling Carbon and Nitrogen Dynamics for Soil Management.’’ pp. 55–102. Lewis Publishers, Boca Raton, FL, USA. Marchetti, R., Donatelli, M., and Spallacci, P. (1997). Testing denitrification functions of dynamic crop models. J. Environ. Qual. 26, 394–401. Mary, B., Beaudoin, N., Justes, E., and Machet, J.-M. (1999). Calculation of nitrogen mineralization and leaching in fallow soil using a simple dynamic model. Eur. J. Soil Sci. 50, 549–566. McCown, R. L., Hammer, G. L., Hargreaves, J. N. G., Holzworth, D. P., and Freebairn, D. M. (1996). APSIM: A novel software system for model development, model testing and simulation in agricultural systems research. Agric. Syst. 50, 255–271. McGechan, M. B., and Wu, L. (2001). A review of carbon and nitrogen processes in European soil nitrogen dynamics models. In ‘‘Modeling Carbon and Nitrogen Dynamics for Soil Management,’’ pp. 103–171. Lewis Publishers, Boca Raton, FL, USA. Meynard, J. M., Cerf, M., Guichard, L., Jeuffroy, M. H., and Makowski, D. (2002). Which decision support tools for the environmental management of nitrogen? Agronomie 22, 817–829. Mirschel, W., Kretschmer, H., Matthaus, E., and Koitzsch, R. (1991). Simulation of the effects of nitrogen supply on yield formation processes in winter wheat with the model TRITSIM. Fertil. Res. 27, 293–304. Molina, J. A. E., Clapp, C. E., Shaffer, M. J., Chichester, F. W., and Larson, W. E. (1983). NCSOIL, a model of nitrogen and carbon transformations in soil: Description, calibration, and behavior. Soil Sci. Soc. Am. J. 47, 85–91. Morvan, T., and Leterme, P. (2001). Vers une pre´vision ope´rationnelle des flux d’azote re´sultant de l’e´pandage de lisier: Parame´trage d’un mode`le dynamique de simulation des transformations de l’azote des lisiers (STAL). Ingeńieries 26, 17–26. Mosier, A. R., Halvorson, A. D., Peterson, G. A., Robertson, G. P., and Sherrod, L. (2005). Measurement of net global warming potential in three agroecosystems. Nutr. Cycl. Agroecosyst. 72, 67–76.


173

Nicolardot, B., Recous, S., and Mary, B. (2001). Simulation of C and N mineralisation during crop residue decomposition: A simple dynamic model based on the C:N ratio of the residues. Plant Soil 228, 83–103. Nolot, J. M. (2001). Dispositif ‘‘Syste`mes de Grande Culture Inte´gre´s,’’ les outils d’aide a` la dećision associe´s. Compte rendu d’essai SGCI, INRA Domaine d’Auzeville, France, 20pp. Parton, W. J., Ojima, D. S., Cole, C. V., and Schimel, D. S. (1994). A general model for soil organic matter dynamics: Sensitivity to litter chemistry, texture and management. In ‘‘Quantitative Modeling of Soil Forming Processes,’’ Vol. 39, pp. 147–167. SSSA Spec. Publ., Madison, WI. Pervanchon, F., Bockstaller, C., Amiaud, B., Peigne´, J., Bernard, P.-Y., Verte`s, F., Fiorelli, J.-L., and Plantureux, S. (2005). A novel indicator of environmental risk due to nitrogen management on grasslands. Agric. Ecosyst. Manage. 105, 1–16. Petersen, B. M. (2003). C-TOOL version 1.1—A tool for simulation of soil carbon turnover. Description and users guide, http://www.agrsci.dk/c-tool/ctool.pdf. Petersen, B. M., Jensen, L. S., Hansen, S., Pedersen, A., Henriksen, T. M., Sorensen, P., Trinsoutrot-Gattin, I., and Berntsen, J. (2005b). CN-SIM: A model for the turnover of soil organic matter. II. Short-term carbon and nitrogen development. Soil Biol. Biochem. 37, 375–393. Porter, J. R. (1993). AFRCWHEAT2: A model of the growth and development of wheat incorporating responses to water and nitrogen. Eur. J. Agron. 2, 69–82. Prosser, J. I. (1986). ‘‘Nitrification.’’ IRL Press, Oxford. Richter, G. M., Beblik, A. J., Schmalstieg, K., and Richter, O. (1998). N-dynamics and nitrate leaching under rotational and continuous set-aside—a case study at the field and catchment scale. Agric. Ecosyst. Manage. 68, 125–138. Ritjema, P. E., and Kroes, J. G. (1991). Some results of nitrogen simulations with the model ANIMO. Fertil. Res. 27, 189–198. Rodrigo, A., Recous, S., Neél, C., and Mary, B. (1997). Modelling temperature and moisture effects on C-N transformations in soils: Comparison of nine models. Ecol. Modell. 102, 325–339. Saleh, A. R., Bengtson, R. L., and Fouss, J. L. (1994). Performance of the DRAINMODCREAMS model with an incorporated nutrient submodel. Trans. ASAE 37, 1109–1114. Scholefield, D., Lockyer, D. R., Whitehead, D. C., and Tyson, K. T. (1991). A model to predict transfromations and losses of nitrogen in UK pastures grazed by beef cattle. Plant Soil 132, 165–177. Schultz, A., and Mirschel, W. (1995). Simulating soil water balance, nitrogen behaviour and biomass components, using the agroecosystem model AGROSIM-WinterWheat and data from the north german Krummbach catchment. Ecol. Modell. 81, 133–144. Seppelt, R. (1999). Applications of optimum control theory to agroecosystem modelling. Ecol. Modell. 121, 161–183. Shaffer, M. J., Bartling, P. N. S., and McMaster, G. S. (2004). GPFARM modeling of corn yield and residual soil nitrate-N. Comp. Elect. Agr. 43, 87–107. Shaffer, M. J., and Larson, W. E. (1987). In ‘‘NTRM, A Soil-Crop Simulation Model for Nitrogen Tillage, and Crop-Residue Management.’’ US Department of Agriculture, Conservation Research Report, 34–1, 103pp. Shaffer, M. J., Lasnik, K., Ou, X., and Flynn, R. (2001). NLEAP Internet tools for estimating NO3-N leaching and N2O emissions. In ‘‘Modeling Carbon and Nitrogen Dynamics for Soil Management,’’ pp. 403–426. Lewis Publishers, Boca Raton, FL, USA. ShengMao, Y., Malhi, S. S., FengMin, L., DongRang, S., MingGang, X., Ping, W., GuoJu, X., Yu, J., TianWen, G., and JianGuo, W. (2007). Long-term effects of manure and fertilization on soil organic matter and quality parameters of a calcareous soil in NW China. J. Plant Nutr. Soil Sci. 170, 234–243.

174

P. Cannavo et al.

Sieling, K., and Kage, H. (2006). N balance as an indicator of N leaching in an oilseed rape— winter wheat—winter barley rotation. Agric. Ecosyst. Environ. 115, 261–269. Sinclair, T. R., and Amir, J. (1992). A model to assess nitrogen limitations on the growth and yield of spring wheat. Field Crops Res. 30, 63–78. Six, J., Feller, C., Denef, K., Ogle, S. M., Sa, J. C.d. M., and Albrecht, A. (2001). Soil organic matter, biota and aggregation in temperate and tropical soils—effects of notillage. In 11th Nitrogen Workshop, Agronomie, Vol. 22, pp. 755–775. Reims. Smil, V. (1997). Global population and the nutrient cycle. Sci. Am. 227, 58–63. Smith, P., Smith, J. U., Powlson, D. S., McGill, W. B., Arah, J. R. M., Chertov, O. G., Coleman, K., Franko, U., Frolking, S., Jenkinson, D. S., Jensen, L. S., Kelly, R. H., et al. (1997). A comparison of the performance of nine soil organic matter models using datasets from seven long-term experiments. Geoderma 81, 153–225. Sogaard, H. T., Sommer, S. G., Hutchings, N. J., Huijsmansc, J. F. M., Bussinkd, D. W., and Nicholson, F. (2002). Ammonia volatilization from field-applied animal slurry—the ALFAM model. Atmos. Environ. 36, 3309–3319. Stockle, C. O., and Campbell, G. S. (1989). Simulation of crop response to water and nitrogen: An exemple using spring wheat. Trans. ASAE 32, 66–74. Stockle, C. O., Martin, S. A., and Campbell, G. S. (1994). CropSyst, a cropping systems simulation model: Water/nitrogen budgets and crop yield. Agric. Syst. 46, 335–359. Strong, W. M. (1995). Nitrogen fertilization of upland crops. In ‘‘Nitrogen Fertilization in the Environment’’ (P. E. Bacon, Ed.), pp. 129–169. Marcel Dekker, Inc., New York. Supit, I., and Goot, V. D. (2003). In ‘‘Updated System Description of the WOFOST Crop Growth Simulation Model as Implemented in the Crop Monitoring System Applied by the European Commission.’’ 120pp, http://www.treemail.nl/download/treebook7. Thorsen, M., Refsgaard, J. C., Hansen, S., Pebesma, E., Jensen, J. B., and Kleeschulte, S. (2001). Assessment of uncertainty in simulation of nitrate leaching to aquifers at catchment scale. J. Hydrol. 242, 210–227. Tixier, P., Malezieux, E., Dorel, M., Bockstaller, C., and Girardin, P. (2007). Rpest—an indicator linked to a crop model to assess the dynamics of the risk of pesticide water pollution: Application to banana-based cropping systems. Eur. J. Agron. 26, 71–81. Vallejo, A., Diez, J. A., Lopez-Valdivia, L. M., Cartagena, M. C., Tarquis, A., and Hernaiz, P. (2004). Denitrification from an irrigated soil fertilized with pig slurry under Mediterranean conditions. Bio. Fertil. Soils 40, 93–100. van Ittersum, M. K., Leffelaar, P. A., van Keulen, H., Kropff, M. J., Bastiaans, L., and Goudriaan, J. (2003). On approaches and applications of the Wageningen crop models. Eur. J. Agron. 18, 201–234. Verberne, E. L. J., Hassink, J., Willigen, P. D., Groot, J. J. R., and Veen, J. A. V. (1990). Modelling organic matter dynamics in different soils. Netherlands J. Agric. Sci. 38, 221–238. Vocanson, A. (2006). In Evaluation ex ante d’innovations varie´tales en pois d’hiver (Pisum sativum L.): Approche par mode´lisation au niveau de la parcelle et de l’exploitation agricole. Ecole Doctorale ABIES. Paris, INA-PG, 294pp. Willians, J. R. (1995). The EPIC model. In ‘‘Computer Models of Watershed Hydrology,’’ pp. 909–1000. Water Resources Publications, Highlands Ranch, CO.

C H A P T E R

F I V E

The Role of Scientists in Multiscale Land Use Analysis: Lessons Learned from Dutch Communities of Practice J. Bouma,* J. A. de Vos,* M. P. W. Sonneveld,* G. B. M. Heuvelink,* and J. J. Stoorvogel* Contents 1. Introduction 2. The Available Toolkit for Up- and Downscaling 2.1. Introduction and rationale 2.2. Tools for upscaling 2.3. Tools for downscaling 3. Four Considerations When Dealing with Interactive Research 3.1. Introduction 3.2. The role of scientists as knowledge facilitators 3.3. Different levels of knowledge in a given discipline 3.4. Long-term engagement when following the policy cycle 3.5. Avoid getting lost 4. Introduction to the Case Studies 4.1. Selecting relevant spatial levels to be considered in a demand analysis 4.2. Format for analyzing the case studies 5. Case Studies at Farm Level 5.1. Introduction 5.2. Case study Van Bergeijk: Precision agriculture 5.3. Case study Spruit: Production of ‘‘new’’ manure 6. Case Study at Regional Level 6.1. Introduction 6.2. Case study NFW: Agriculture in a national landscape 7. General Conclusions and Recommendations Acknowledgments References

*

177 180 180 181 187 188 188 188 188 190 190 191 191 192 193 193 193 201 211 211 215 231 233 234

Wageningen University and Research Centre, Wageningen, The Netherlands


#


175

176

J. Bouma et al.

Many research and scientific organizations emphasize the importance of science for society in their strategic plans. This is certainly true for land use studies being discussed in this chapter as new environmental policies are introduced at European and national level. Such policies reflect concerns of society so that a structural link between science and policymaking would appear to be logical and desirable. Rather than following traditional top-down and disciplinary research approaches, emphasis is increasingly being placed on interactive, interdisciplinary work in Communities of Practice (CoPs) in which scientists work together with various stakeholders and policymakers in a joint learning mode. But this requires new research approaches including long-term engagement during the entire policy cycle asking for a new attitude of scientists. Few experiences have been reported so far. Three Dutch case studies are therefore discussed to illustrate the functioning of CoPs by focusing on up- and downscaling (called multiscaling hereafter), a key element of land use research. Five types of multiscaling were used in the three case studies. Three were technical: (1) use of model- or design-based (geo)statistical techniques, (2) extrapolation of data obtained from experimental plots to larger areas, and (3) use of quasi-3D process models to upscale grid data in a Geographic Information System (GIS) to regions. Two were policy oriented: (1) nutrient balances for farms to allow upscaling from fields to farm and (2) a research framework for regions, based on the DPSIR approach, which sequentially covers aspects mentioned in environmental laws as being important for sustainable development. Quite diverse and unrelated questions about land use issues by different members of the CoP cannot be a fruitful basis for research programs. Scientists have therefore an important role to play within a CoP in orchestrating a demand analysis that puts questions into context and defines existing knowledge as well as knowledge gaps. Defining research on the basis of a demand analysis in a CoP creates innovative ideas, creates commitment of participants, and allows definition of needed research that is functional. This includes cutting edge research publishable in literature and requires for land use studies updating of valuable existing soil survey information to a level that can be used in modern modeling techniques including functional characterization of soil series, development of pedotransfer functions, and definition of phenoforms. Particular attention is needed for introducing modern monitoring techniques for soil and water because the high cost of traditional methods implies that little monitoring is done now with detrimental effects for the calibration and validation of simulation models that increasingly secure a live of their own. The scientific community needs to take a fresh look at its paradigms. Next to the establishment of CoPs, we therefore advocate development of Communities of Scientific Practice (CSP) within the research community that define different functions for members of the scientific community in terms of (1) communication within CoPs by shaping the demand analysis and to the outside world and (2) defining research needs and its execution, using knowledge chains including basic research.

The Role of Scientists in Multiscale Land Use Analysis

177

1. Introduction Increasingly, research on land use is made in the context of CoPs where scientists work together with stakeholders, policymakers, and other involved citizens to analyze and solve problems in a participatory, jointlearning mode (e.g., Bouma, 2005; Lave and Wenger, 1991; Wenger et al., 2002). This form of interaction is inspired by an increasing number of environmental laws and regulations at different administrative levels. The relevance of introducing the CoP concept to the scientific community becomes evident when studying strategic plans of many research organizations that emphasize the importance of science for society. The International Council for Science (www.icsu.org) speaks about ‘‘. . . where science is used for the benefit of all. . .and where scientific knowledge is effectively linked to policy making.’’ Their goal is to ‘‘. . .strengthen international science for the benefit of society.’’ The Royal Netherlands Academy of Arts and Sciences includes in its mission statement the phrase: ‘‘. . . the Academy commits itself to make an optimal contribution of Dutch science to the cultural, social and economic development of society.’’ And one of the three main lines of action in the strategic plan of The National Science Foundation in the Netherlands concerns ‘‘Science for Society.’’ The CoP concept follows from the realization that it is very difficult for scientists, working in disciplinary isolation, to satisfy demands of society. Issues are complex and require an open, learning attitude from all participants involved, including scientists. This certainly applies to people-centered land use studies where soil scientists are generally asked to work together with economists, agronomists, farmers, and policymakers. Although attractive as a concept, CoP definitions are diverse and little practical experiences have been reported in literature. Moreover, the issue of scale, highly relevant for land use studies, is hardly being addressed in CoP literature so far. Attention will therefore be paid in this chapter to the establishment and functioning of CoPs in three case studies to be discussed, paying due attention to experiences obtained elsewhere. However, sets of nonrelated primary questions raised by diverse participants in a CoP are usually only the start of formulating broader and more basic questions that reflect the common interests of the entire group and the intentions of environmental legislation that, ideally, reflects interests of society. We will therefore move beyond simply trying to answer an often incoherent set of primary questions by stakeholders toward a comprehensive demand analysis within the CoP in which researchers can play an important, if not crucial and central, role. This process is time consuming but turns out be to worthwhile in the end. From a policy cycle perspective, this would be a part of the signaling function and would contribute to the design function (e.g., Bouma et al., 2007).

178

J. Bouma et al.

The call for establishing effective CoPs as a means for science to be embedded in society is, however, disputed by some researchers. Fears such as ‘‘moving away from the core business,’’ ‘‘loosing integrity,’’ or ‘‘loosing time for real in-depth research’’ make the research approach through CoPs controversial. Some scientists feel that research, including data collection, analysis, and publication of results, is their primary and consuming responsibility, whereas communication and interaction with stakeholders, although important, is another matter. In fact, in earlier times, this approach was institutionalized by the Transfer-of-Technology model (Chambers et al., 1989) where extension services functioned as de facto communicators between scientists and stakeholders. However, because of privatization, most extension services have substantially been reduced in capacity and this model is hardly operational anymore. But, irrespective of the decreased capacity of extension services, our basic premise in this chapter is the essential role of a CoP in effective land use research involving structural interaction of scientists with stakeholders, policymakers, and other involved citizens. Cases to be discussed are intended to support this premise and show that this type of interaction still leaves room for basic research and, in fact, demonstrates the need for it. Participants in land use studies operate on a wide variety of scales. Farmers generally deal with fields and farms while policymakers deal not only with farms but also with administrative levels ranging from villages and regions to national levels and beyond. For example, environmental guidelines of the European Union (EU) in the areas of water, habitat, and soil are based on the regional or catchment level while regulations at the national level are also based on the farm level. Upscaling of process descriptions and data from samples, plots, fields, and farms to regions and beyond is one of the key issues in land use studies. How to make sure that by necessity limited number of data points, field samples, or plot experiments can accurately represent properties and features of larger areas of land? In turn, how can such representations be ‘‘downscaled’’ to lower scale levels, for instance, when articulating the implications of regional land use phenomena for individual land users? Aside from considering up- and downscaling in terms of space, it is also necessary to consider it in terms of time: computing averages over time or singling out values at points in time from timeintegrated values. Stakeholders want rapid solutions while researchers are bound to time-consuming procedures when setting up and executing research projects. Current environmental laws and regulations often define indicators and thresholds but are painfully short on defining operational procedures in both space and time. Numerous studies have been made on upscaling in space and time and at least two types of studies can be distinguished: process-driven studies that use mechanistic models and data from experimental plots to be extrapolated to larger areas and those that use various mathematical, (geo)statistical, and


179

remote sensing techniques for upscaling point data. Examples of the first category are fertilizer trials on small plots, used to determine fertilizer response rates for regions. Small erosion plots are used to predict erosion rates of watersheds. Also, soil survey interpretations implicitly assume that behavior of ‘‘representative’’ pedons for a given mapping unit represents the behavior of at least 80% of the corresponding large area on the soil map. Many studies on kriging and application of remote sensing represent the second type of upscaling, reviewed in Chapter by Ashraf et al., this volume. Most emphasis in studies on up- and downscaling been has so far given to development of either a ‘‘(geo)statistical toolkit’’ or operational procedures. Even though most of these studies present case studies to illustrate the results of applying any given technique, less attention is usually paid to the generic question as to if, when, and how such techniques can be best applied, if at all. Often, there appears to be a rather strong technology push from the supply side by researchers when defining up- and downscaling research with less attention to operational aspects in land use studies. Therefore, this chapter will, in contrast, focus on the demand side when researching practical questions on land use being raised by a wide range of stakeholders and policymakers. What are characteristic questions at different spatial levels and what is their time frame? What is the relation with environmental rules and regulations? How can diverse and isolated questions on land use be framed into a coherent format by a demand analysis? How, in turn, can available up- and downscaling techniques be used to answer questions arising from the demand analysis and are available techniques adequate or do we need new techniques and procedures? Aside from the technical issues related to up- and downscaling, as discussed, policy issues also differ at different scales both in terms of space and time. For example, a governmental official at community level has a different focus than the one at provincial, national, or EU level. As stated, this is reflected in the type of questions being raised that partly result from different environmental laws and regulations at successive administrative levels. A particular problem to be discussed in one of the case studies arises when policy questions at EU or national level are ‘‘translated’’ to the farm level without consideration of the operational and institutional questions that are being raised at that level. Finally, to avoid continued reference to upand downscaling in the text, the comprehensive term ‘‘multiscaling’’ will frequently be used hereafter. In summary, this chapter has the following objectives: 1. To discuss multiscaling of land use data and characteristics based on an analysis of three Dutch case studies on land use. What questions are raised at what spatial level by land users, stakeholders, and policymakers, working together with scientists in CoPs and what is the relation with environmental laws and regulations?

180

J. Bouma et al.

2. To clarify and integrate such often unrelated primary questions, raised by members of a CoP, into a coherent demand analysis that satisfies the signaling function and contributes to the design function of the policy process. 3. To illustrate the formation and functioning of CoPs in a number of case studies on multiscaling with particular emphasis on the role of scientists. 4. To define which techniques for multiscaling were most successfully used in the various case studies, considering the requirements of the demand analysis. What research, if any, is still needed to fill the experienced gaps in knowledge? 5. To analyze whether our current scientific practices can adequately address issues being raised.

2. The Available Toolkit for Up- and Downscaling 2.1. Introduction and rationale As explained in Section 1, up- and downscaling are important steps in land use studies. This section reviews technical solutions for both cases, with a focus on upscaling. Before doing so, it is useful to precisely define what is understood by up- and downscaling because the literature is not very clear and consistent about what these terms mean. Scale issues play an important role in many disciplines, including those that are important for land use, such as soil science, meteorology, hydrology, and ecology (Bierkens et al., 2000; Bloschl and Sivapalan, 1995; Bouma et al., 1998). Upscaling is taken as synonymous to spatial aggregation, where the interest is in obtaining an integrated value of a spatial variable over an area of a given size. The integrated value will usually be the spatial average, but it may also be some other aggregate such as the median or the proportion of the area for which the variable has a value above a given threshold (Heuvelink and Pebesma, 1999). The upscaling must usually be based on point values of the variable of interest, collected at multiple locations within the area. However, in many cases, the upscaling can also benefit from auxiliary information, such as maps of correlated variables and process knowledge, available in the form of process-oriented models. For instance, the average nitrate concentration in the Northern Frisian Woodlands (NFW) area (see Section 6.2) can be estimated using only observations of nitrate concentrations in groundwater and surface water at points within the region, but the accuracy of the upscaling may likely be improved by also including information about soil type, land use, land and water management, and predictions of groundwater quality models.


181

Downscaling is the opposite of upscaling. Here, one knows the spatially aggregated value over a given area and would like to infer from it the value of the variable at points within or subareas of the entire area. For instance, groundwater quality may be measured at the inlet of a pumping station. This provides an average value of the water quality of the aquifer below the station, but the interest may be in the groundwater quality of individual farms that feed the aquifer. In such a case, the average groundwater quality must be scaled down to that of the individual farms. Note that up- and downscaling need not be restricted to the spatial domain but may also be defined over time. For instance, estimating the annual greenhouse gas emission from a sample of half-hour measurements taken at various time points within the year requires temporal upscaling. Estimating the annual greenhouse gas emission for an entire region or country from a large sample of half-hour chamber measurements with a spatial support of 0.5 m2 at multiple points within the region requires spatiotemporal upscaling. In this section, we present methodologies for up- and downscaling in space, but in fact these methods can be easily transferred to the temporal and spatiotemporal domain. The importance of up- and downscaling to CoPs is clear: stakeholders such as farmers, decision makers, and land managers need to evaluate the effectiveness of their measures on aggregated values, whereas measurements are often only available at point level. Likewise, stakeholders may need to disaggregate national statistics to regions, communities, or individual farms to be able to adjust policies and practices.

2.2. Tools for upscaling 2.2.1. Design-based methods Design-based methods (De Gruijter and Ter Braak, 1990; De Gruijter et al., 2006) rely on classical sampling theory (Cochran, 1977). These methods have various important advantages over model-based approaches discussed in Section 2.2.2. The most important advantage is that design-based methods do not make any assumptions about the spatial variability structure of the variable of interest. Thus, they also do not suffer from making wrong assumptions, which increases their validity and would for instance make a better case in court. The downside is that design-based upscaling puts high demands on the number and way in which the point data are collected. Let z be the variable of interest, measured on a continuous numerical scale. Because z varies with location, we denote it by z(x), where x is a twodimensional geographic coordinate (extension to three dimensions is straightforward). If the goal of the upscaling is to compute the average of z over a spatial domain D, then our target variable zD is defined as:

182

J. Bouma et al.

1 zD ¼ jDj

Z zðxÞdx

ð1Þ

x2D

where |D| is the size of D. In order to compute zD without error, we would need to know z(x) without error for all x in D. In practice, such information is not available (otherwise the problem would be trivial). Instead, we might have to make do with a sample of observations taken within D. Now suppose that the sample is collected using probability sampling. In the simplest case, this would be a simple random sample {z(x1), z(x2),. . .,z(xn)}, whereby the locations xi are randomly and independently selected and where each location in D has equal probability of being selected. In such a case, the best estimate of zD is simply the unweighted mean of the observations:

z^D ¼

n 1X zðxi Þ n i¼1

ð2Þ

It is well-known that this yields an unbiased estimate (i.e., repeating the procedure many times, each time drawing a new set of n locations, yields a distribution of estimates that is centered on the true value zD) and that the standard deviation of the estimation error is given by:

s SDðzD z^D Þ ¼ pffiffiffi n

ð3Þ

where s is the square root of the spatial variance s2, defined as:

1 s ¼ jDj

Z

ðzðxÞ zD Þ2 dx

2

ð4Þ

x2D

In practice, the spatial variance s2 is unknown and estimated with the sample variance. This slightly increases uncertainty but the differences are small when n is sufficiently large (say n > 30). Note also that for sufficiently large n, the error zD z^D will be normally distributed, so that confidence intervals and exceedance probabilities can easily be computed. Now suppose that the observations z(xi) are not error-free. If the observation error can be characterized with a probability distribution that has zero mean (i.e., it has no systematic error or bias) and constant variance (i.e., all observations have the same accuracy), then Eqs. (2)–(4) are still valid, now of course with the true values z(xi) replaced with their


183

measured value. The extra uncertainty caused by the measurement errors is automatically incorporated in the increased spatial variance s2. If different observations have different measurement error variances, then the optimal estimate of zD becomes a weighed combination of the observations. Observations with smaller error variance have more credibility and get larger weights. The case when the measurement errors are mutually correlated also has a straightforward solution, provided the correlations are known. A widespread and persistent misconception states that design-based methods cannot be used for spatially correlated variables because the method requires that observations are independent. However, independence is satisfied because the stochasticity in design-based methods is in the coordinates of the sampling locations x (which are sampled independently), and not in the spatial variable z. This is explained in great detail in De Gruijter and Ter Braak (1990). Design-based methods also apply to the case where the objective is not the spatial mean but some other spatial aggregate. In all these cases, the sample statistic is used to estimate the population statistic. For instance, to estimate the proportion of an area with a nitrate concentration above 50 g liter1, one may simply use the proportion of the random sample taken in the area for which the nitrate concentration is above the threshold. The uncertainty in the estimate can also be quantified, although the equations can become more involved than Eq. (3). Equation (3) shows that the accuracy of the upscaled variable is inversely proportional to the square root of the number of observations. This means that accuracy slowly progresses with sample size. To double the accuracy, one must quadruple the sample size. In practice, very large sample sizes may be needed to reach an accepted accuracy level. For variables that are expensive to measure, this may be prohibitive. However, there are also other ways to improve the accuracy of design-based upscaling. For instance, simple random sampling may be replaced by alternative probability sampling schemes that avoid redundancies. Stratified sampling is an important example of a more efficient sampling scheme. In this case, the area of interest is subdivided in multiple strata that are sampled and analyzed separately, after which the results are merged. An important advantage of stratified sampling over simple random sampling is that the between-stratum variance is removed, in effect yielding a smaller spatial variance s2, hence reducing upscaling uncertainty [(i.e., compare Eq. (3))]. Uncertainty can be further reduced by optimizing the stratum sample sizes. This usually means that the sample size is greater for large strata and strata with a large within-stratum variance. Stratified sampling is but one example of more elaborate sampling designs. Examples of other sampling methods that improve estimation accuracy or reduce sampling cost are poststratification regression estimators, two-phase random sampling, and cluster sampling. It is beyond the scope of this section to discuss these methods. Instead, we refer to De Gruijter et al. (2006).

184

J. Bouma et al.

Notwithstanding the great flexibility of design-based methods for upscaling, there will always be cases where these methods may not be applicable or not provide sufficiently accurate results. This will happen when the number of observations in the domain of interest is small or even zero, or when the observations are not obtained with probability sampling. In practice, one may not be in a situation where a sampling plan can be designed and executed but where one must do with existing legacy data. These data will often have been collected using preferential or convenience sampling. In such cases, and when the number of observations is small, it may be better to turn to model-based upscaling approaches. 2.2.2. Model-based methods Model-based methods take an entirely different approach to upscaling. In this case, a statistical model is proposed that characterizes the spatial variation of z, the variable of interest. A statistical model is used because the true spatial distribution of z is unknown (otherwise upscaling would be a trivial exercise, see previous section). The general statistical model satisfies:

ZðxÞ ¼ mðxÞ þ eðxÞ

ð5Þ

where m(x) is a trend function (explanatory part) and e(x) is a stochastic residual. The trend m(x) may be a linear regression but can also be the result of a mechanistic process model. Note that here we have introduced upper case Z, to distinguish between the random variable Z(x) (the statistical model) and the true (but unknown) value z(x). In fact, we assume that the true reality z is a realization of the statistical model Z. The rationale behind this approach is that because of uncertainty, we are not able to characterize z itself, but we do feel sufficiently confident to characterize a probability distribution of it. For instance, we might not be able to tell what the exact ammonia votalization from manure for a given parcel of the Spruit farm is (see Section 5.3), but perhaps we may know or have sufficient confidence to state that it is in between the bounds of an interval with a certain probability. If we can do this for all intervals then we have effectively produced a probability distribution. The center of the probability distribution of Z(x) may depend on known factors contained in m(x), whereas the spread of the distribution reflects our uncertainty about the ammonia votalization and is represented by e(x). An important characteristic of the model Eq. (5) is that the residual e may be spatially autocorrelated. This implies that measurements z(xi), i ¼ 1, . . . , n may be used to narrow down the probability distribution of Z(x0), where x0 is an unobserved location with distances from the observation points that are within the spatial correlation length. This is the essence of kriging. In particular, in model-based upscaling, block kriging is used to scale up the point observations z(xi) to estimate the spatial average zD.


185

2.2.3. Ordinary block kriging Ordinary kriging is a well-known technique that we only briefly summarize here. In ordinary kriging, it is assumed that the trend m is an unknown constant. Further, it is assumed that e is second-order stationary, which means that the covariance between e(x) and e(x þ h) only depends on the separation distance h. Next, the value of z(x0) is estimated as:

z^ðx0 Þ ¼

n 1X li zðxi Þ n i ¼1

ð6Þ

where the so-called kriging weights li are chosen such that the squared expected estimation error is minimized, under the condition of unbiasedness (i.e., the sum of the weights must equal one). The kriging weights are obtained by solving a linear system of n þ 1 equations (Goovaerts, 1997). The estimation error variance can easily be computed from the weights li and the covariance function of e. In ordinary block kriging, the region of interest is first discretized, ordinary kriging estimations are made at all discretization nodes and the results are averaged. Averaging linear combinations of observations remains a linear operation on the observations, so that we can write: n 1X z^D ¼ zðxi Þ n i¼1 i

ð7Þ

Comparison of the model-based estimate Eq. (7) with the design-based estimate Eq. (2) shows that weights have been introduced. This is because in the model-based approach we take the model Eq. (5) as a starting point, the residual e of which may be spatially correlated. This means that the information content of observations depends on their spatial configuration. Points that are on average closer to the points in the target domain are likely to be assigned larger weights, unless they are close to other observation points that causes redundancy between the observations. Note also that in the model-based approach, there is no impediment against using observations that are outside the area of interest. This is because points that are outside the area may still be correlated with points in the area, and can thus be used to improve estimation. This implies that model-based upscaling can also be used in cases where there are no observations in the target domain D. The uncertainty about the model-based estimate Eq. (7) is characterized by the standard deviation of the estimation error. As the estimate z^D is linear in the z(xi) and zD is linear in the z(x), it can be easily computed because it depends only on the covariance function of e and the spatial configuration of the observation points and domain D.

186

J. Bouma et al.

2.2.4. Regression block kriging There are many extensions to ordinary kriging. Here we will only discuss regression kriging, which is the most valuable extension for the purpose of upscaling. Unlike ordinary kriging, regression kriging assumes that the trend m is a linear combination of known, spatially exhaustive covariates:

mðxÞ ¼

k X

bi fi ðxÞ

ð8Þ

i¼1

Here, the fi(x) are the covariates and the bi are unknown regression coefficients. For instance, if the variable of interest is the ammonia emission in the Spruit farm, then candidate covariates might be manure application and water table level. Clearly, the covariates fi must have predictive power about the dependent variable z, otherwise there is no need to include them. The computation of the regression kriging estimate and associated kriging variance is along the same lines as for ordinary kriging (Hengl et al., 2004). The equations are somewhat more involved because the regression coefficients bi also need to be estimated and the uncertainty about them must also be taken into account. Upscaling to the region D is again done by discretizing the region and visiting the discretization nodes one by one. Regression kriging may be interpreted as a merger between linear regression and kriging because it contains a regression term and estimates the regression coefficients bi, while it also includes a spatially correlated residual that can be kriged. Many studies ignore the spatial autocorrelation structure of the residual and use only regression to estimate the dependent variable at all points in the target area D (e.g. Stehfest and Bouwman, 2006). However, in such an approach, it will be difficult to quantify the uncertainty about the upscaled values because the uncertainty about the aggregated e in Eq. (5) will completely cancel out if it is spatially uncorrelated. This does not appear to be very realistic. Regression kriging is increasingly popular because recently many accurate and detailed covariates have become available in environmental research, such as derived from remote sensing imagery and high-resolution digital elevation models. Indeed there is nothing against using the outcome of a dynamic process model as a covariate or simply as the trend itself. In case of the latter, the residual e represents model error. For instance, if a deterministic surface- and groundwater quality model were available for the NFW case and would predict the nitrate concentration of the groundwater, then it would be foolish not to use it in the upscaling procedure. One could run the model at all point locations within the area, average the results, and


187

add to it the spatially averaged kriged residual, obtained by kriging the deviations between the observations and deterministic model predictions.

2.3. Tools for downscaling There are far fewer tools available for downscaling than for upscaling. This is partly because practical problems are more often about upscaling, but also because downscaling is more difficult than upscaling and not all upscaling methods described in Section 2.2 have a natural analogue for downscaling. For instance, design-based approaches seem not applicable for downscaling because there is no means to infer characteristics about the objects of a population (i.e., the values at points in an area) from only information about the population as a whole (i.e., the average of the variable over the entire area). Model-based approaches can be used but they must typically make bold assumptions to obtain results. The goal of downscaling is to compute z(x) for all x in the spatial domain D, from knowledge about zD. The obvious requirement is that equation (1) must be satisfied. However, this requirement does not yield a unique solution because there typically are an infinite number of spatial distributions of z that will average to zD. Additional assumptions or requirements are needed, and in one way or another, these all boil down to assuming that the spatial pattern of z is known up to a (multiplicative) constant, which is derived from equation (1). Bierkens et al. (2000) distinguish three cases for downscaling, depending on whether the average zD is known exactly and whether one seeks a single deterministic function z(x) or an infinite number of possible functions z(x), each with a certain probability of occurrence and thus characterized by a random function Z(x). In all three cases, one requires additional assumptions about the spatial pattern of z(x). One frequently used method is multiple linear regression, in which it is assumed that z(x) is a linear function of spatially exhaustive covariates. Another solution is to let the spatial distribution of z(x) be characterized by a (mechanistic) process model. In the case of the latter, one should calibrate the process model such that the average of the process-model outcomes z(x) equals the areal average zD. Downscaling has recently received much attention in climate change research, where the outcomes of Global Circulation Models (GCMs) need to be scaled down for local studies. Many of these studies use statistical methods such as multiple linear regression or regional climate models to scale down the course predictions of GCMs to the finer scale (e.g., Schmidli et al., 2007; Spak et al., 2007).

188

J. Bouma et al.

3. Four Considerations When Dealing with Interactive Research 3.1. Introduction The CoP is the starting point of the discussions in this chapter. Following the definition of Wenger et al. (2002), a CoP is ‘‘a group of people informally bound together by shared expertise and passion for a joint enterprise.’’ Land use questions asked by various land users, policymakers, scientists, and other stakeholders are discussed and analyzed in a CoP providing the platform for the demand analysis, as discussed above. There is no set of standard rules defining what is needed to achieve the objectives of a CoP, nor will there ever be one. Every project, its stakeholders, and its policy context are different. Still, sharing experiences is necessary to improve our still rather limited ability as researchers to participate in and, if feasible, guide interactive activities within CoPs. Later, in this chapter, we will pay particular attention to the establishment and functioning of CoPs in the various case studies to be discussed. But we do not start from scratch and four important aspects related to interactive work come to mind as experienced during our work in The Netherlands so far:

3.2. The role of scientists as knowledge facilitators Land use research is most effective when researchers focus their contributions on the interaction process between stakeholders and policymakers, or, more broadly, between citizens and their government (see also Bouma, 2005). Exclusive interaction with stakeholders runs the risk of yielding results that may be quite interesting and publishable but irrelevant for the public policy process as policymakers are usually not amused when presented with solutions they see for the first time. But exclusive attention for policymakers, often resulting from restrictive funding mechanisms, is not advisable either because noninvolvement of stakeholders in the research process is a major and usually fatal barrier toward their acceptance of results obtained. The role of researchers is therefore to facilitate the interaction process between government and their citizens by using their knowledge and expertise as a tool. This is not simple as it resembles a balancing act that can only be effective in the end when the ultimate goals of any particular project are constantly kept in mind, realizing that different roads lead to ‘‘Rome.’’ But to Rome they should lead.

3.3. Different levels of knowledge in a given discipline Be aware that different levels of knowledge are being tapped when communicating with citizens, governmental officials, and colleague scientists. The scheme, originally proposed by Hoosbeek and Bryant (1992) to discuss

189


soil research, is helpful to distinguish a number of categories on the basis of two distinguishing scales: one ranging from empirical to mechanistic and the other from qualitative to quantitative. Bouma (1997) extended this concept to five knowledge levels, as follows (see Fig. 1). Stakeholders and land users can contribute important K1 and K2 knowledge to the debate and many have reached K3 level by means of education or by consulting internet. Research in subdisciplines of soil science to be published in international refereed journals is often of the K5 type. Soil survey reports, on the contrary, are more K2 or perhaps K3. This may at least partly explain lack of communication between these two activities within soil science research. Also, contacts between different disciplines depends on their knowledge levels. Social science studies seem to operate on K3 level, even though K4 and K5 studies are also made. Econometric studies, on the contrary, are often of the K5 type that makes communication with other disciplines difficult. The general suggestion here is that creating awareness of different levels of knowledge in different disciplines helps to improve communications, also within a given discipline when dealing with several subdisciplines. Knowledge chains represent both the sequence from K1 to K5, indicating by a cost–benefit analysis what is being gained by increasing the detail of scientific information and the sequence from K5 to K1 that is important for communicating results of basic research to stakeholders.

Empirical

K3

K1

Qualitative

K4

K2

Quantitative

K5

Mechanistic

Figure 1 The five knowledge levels. K1 ¼ practical expertise, sometimes referred to as tacit knowledge; K2 ¼ expert knowledge, with a better understanding of mechanisms involved but often qualitative and descriptive in character, even though sometimes measurements are made with standard techniques; K3 ¼ domain of the applied scientist: empirical but expressing relationships quantitatively, for example, by regression analysis and by making specific relatively simple measurements. K4 ¼ domain of the basic scientist, using relatively simple, partly empirical models to describe systems and using existing and new measurement techniques; K5 ¼ same, but use of more complex deterministic simulation models and cutting-edge measuring methodology.

190

J. Bouma et al.

3.4. Long-term engagement when following the policy cycle The policy cycle distinguishes five phases: signaling, designing, deciding, implementing, and evaluating (e.g., Bouma, 2005, Bouma et al., 2007). It pays in the end to spend considerable time in the beginning of any project on signaling. The demand analysis, to be discussed in this chapter, is an important element of the signaling phase and also contributes to the later design. Signaling raises questions such as: How is the problem perceived by those involved? What is their background and frame of reference? What are their images as to what should and could be accomplished? It is important here to take all ideas seriously, even the ones that appear to be weird on first sight. It would not be the first time when what appear to be weird ideas now turn out to be visionary and innovative in the end. Once the problem to be studied has been explored this way, the design of the research follows. In land use studies, this often means that a scenario analysis is made in which all options for future land use are studied and expressed in terms of their economic, social, and environmental implications if sustainability is the ultimate aim, which it usually is. Certain extreme options will smoothly disappear without a trace when the tradeoffs involved lead to unacceptable consequences. The design activity is the logical core activity for the research process in land use studies. Often, researchers are only involved with this activity. Scientists can advise decision makers, if so invited, but their role is not to make decisions nor should they imply they should do so. Continued involvement of researchers during implementation of the decisions that have been made is, however, advisable. Many questions are bound to arise and the ultimate success of a project is better guaranteed when those questions are answered by researchers that were involved in signaling and designing. Finally, the entire process needs to be evaluated as the ultimate activity of the CoP. Only then can joint learning result and can experiences from one project to be used in the next. This may all imply that the participation of scientists within a CoP is much more time consuming and costlier than when only a design is made or an experiment is done. For research to be effective for society, we believe, however, that long-term involvement of scientists during the entire policy cycle is important.

3.5. Avoid getting lost Realizing interdisciplinary research is very difficult and sometimes impossible. Four suggestions to avoid and overcome at least some of the problems arise from our own experience: (1) Try to understand the overall scope of the particular problem being considered. Who is involved and why? Are there hidden agendas or skeletons in the closet? Are all relevant partners involved or are some missing? If that is the case, what is the reason? Jumping too


191

quickly into any project can be rather risky. One can never be sure about all conditions surrounding a given project and having some vision about its setting helps. Of course, the joint learning concept implies that once the project is started, a joint development path is created and what might appear to be adverse conditions at first sight may turn into unexpected opportunities later. (2) Analyze what the particular strength and contribution can be of your own profession: what are our ‘‘niches.’’ Complaints by soil scientists of being sidelined in land use projects are a universal and rather worrying international phenomenon. Rather than complain, soil scientists should focus in a proactive manner on specific opportunities provided and extend this analysis to the five policy phases discussed above. Once such opportunities are identified, a focused effort should follow to make sure that other members of the team agree and that participation is assured. Realize that other members of the team often speak a different scientific language and come from a different scientific culture. Try to understand this and use such insights to better promote your own input. (3) Be pragmatic in terms of selecting the most promising scientific partners in any CoP. It may be intellectually exciting to work, for example, with psychologists on interaction processes, but this in the end may not result in significant input from your work in the overall project. Logical partners for soil scientists are hydrologists and agronomists because together they can define water regimes in given areas, which is an important element of land use planning (see also Sections 5 and 6); (4) Define emergency exits that allow you to leave any given project without too many problems if the project appears doomed. This happens even in projects that appeared attractive at first sight.

4. Introduction to the Case Studies 4.1. Selecting relevant spatial levels to be considered in a demand analysis Land use questions are most often raised at the enterprise level (most often a farm), the regional and national level and at the international level where a distinction can be made between groups of countries unified in some type of institutional structure, such as the EU, and the global level. We will discuss three case studies, two at the farm level (Spruit and Van Bergeijk) and one at the regional level the Northern Frisian Woodlands (NFW) (Fig. 2). In these case studies up- and downscaling procedures will be analyzed that have been used by scientists in the three quite different CoPs. The farm level represents the basic management unit while the regional level is increasingly relevant for both national and European legislation. Finally, confronting results of

192

J. Bouma et al.

Figure 2

Locations of the three case study areas.

the different demand analyses in the case studies with selected scaling procedures used to signal and design, we will discuss whether new multiscaling techniques are needed.

4.2. Format for analyzing the case studies Following the objectives of this chapter, four elements will systematically be addressed in each of the three case studies: 1. What questions are raised about a given land use problem by various stakeholders and policymakers? 2. How was the demand analysis performed in the CoP and how did this contribute to the signaling, design, and implementation process? How are decisions made and can the entire process be evaluated? 3. How was the CoP established and how did it perform? In particular, what was the role of the soil scientists and what research was done? 4. What techniques for multiscaling were used and what areas require new research?


193

5. Case Studies at Farm Level 5.1. Introduction Moving upscale from samples, soils, and fields, the first spatial level where stakeholders ask questions is the farm level, or broader, the level where various enterprises operate that use land in a given area. In this chapter, the focus will be on farms. Current questions by farmers have a strong economic focus: prices of agricultural produce decrease while costs increase. Moreover, environmental regulations become more restrictive and many questions relate to the implementation of such laws and how consequences can be minimized. Questions of politicians focus on environmental regulations as well but from their point of view covering not only environmental quality but also enforceability. Two cases studies will be analyzed, presenting different conditions.

5.2. Case study Van Bergeijk: Precision agriculture 5.2.1. Questions raised Some of the prime agricultural soils in The Netherlands can be found on the former island of Voorne-Putten in the central-western part of The Netherlands. The commercial farm of the Van Bergeijk family was studied. The farm is located in Voorne-Putten (51 480 N and 4 160 E) and measures approximately 100 ha. The soils consist of calcareous marine deposits with a fine loam to heavy clay loam texture. The soils are classified as Typic Fluvaquents (Soil Survey Staff, 2006) with an average summer temperature of 17.3 C and an average annual rainfall of 760 mm. The farm uses a crop rotation of potato, sugar beet, and winter wheat. Here, we will focus on winter wheat. With a European Nitrates Directive (EC, 1991) in place that defines very specific goals in terms of groundwater quality (i.e., nitrate levels below 50 mg liter1 in the upper groundwater), The Netherlands formulated specific legislation to reach these European goals. At the time of this research (1999), Dutch legislation included a system that required farmers to maintain an accounting system on nitrogen fertilization and nutrient uptake (Oenema et al., 1997). In addition, they were provided with generic fertilizer recommendations that were established for the region. Fertilizer recommendations for winter wheat are aimed at production levels of 12 t ha1 assuming 25 kg of mineral N is required for each ton of wheat. Recommendations are provided for individual fields and incorporate average soil mineral N levels measured in the root zone at the start of the growing season. For example, with an initial soil mineral N level of 60 kg N ha1, the total recommended fertilizer rate is 280 kg N ha1 (including a

194

J. Bouma et al.

topdressing of 40 kg N ha1 just before flowering). Mineral fertilizer is applied using a split application strategy. Typically, four applications are scheduled: a base application in February or March, two applications during April and May (using development stage and coloring as triggers), and a topdressing before flowering in June. Despite European and Dutch legislation and general fertilizer recommendations, farmers are dealing with the very specific conditions at their farm with soil variation between and even within their fields. Discussions with the farmer were initially focused on the issue as to how fertilization practices could be improved as the farmer felt that official recommendation guidelines were too generic for the different soil types on his farm. Can fertilizer management be more tailored to the specific soil and crop conditions while complying with Dutch legislation and considering the generic recommendations? 5.2.2. Demand analysis The farmers (two brothers) were pioneers in the region while managing their own farm but also managing an equally large business on contract farm management. Also through one of their sons, a PhD student in farm management, they were well connected to the research community. As a result, they had a specific interest as to how new developments in agricultural research could support their farm management. During the last two decades, agricultural research on farm management paid much attention to precision agriculture (Bongiovanni and Lowenberg-DeBoer, 2004; Mulla, 2005). Although new machinery was equipped with GPS and yield mapping equipment, no operational methods were available to make use of these technological innovations. In addition, farm operations were increasingly constrained through national and European legislation. Discussions with the farmers resulted in a decision to explore the potential of precision agriculture on the farm to reduce the use of fertilizer and biocides as a proactive measure to be able to deal with further constraints on agricultural inputs. In this proactive approach, supply of nutrients and chemicals to the plant is matched with plant demand in space and time: ‘‘just in time at the right place.’’ As a result, it avoids generation of excess fertilizer and biocides that could pollute surface- and groundwater. This is superior in economic and environmental terms than using generic approaches for fertilization and for applying agrochemicals, which are not fine-tuned to local conditions. European and Dutch legislation are the boundary conditions for the research. The demand analysis resulted in broadening the initial rather elementary questions as to interpreting current fertilizer application management schemes into a farm-wide systems analysis on agrochemical dynamics based on principles of precision agriculture. In this case, we are dealing with a proper multiscale problem. At the European level, a threshold is defined for the nitrate concentration in groundwater. At the national level, policies are developed to reach this


195

threshold through a nitrogen accounting system. Fertilizer recommendations for the region are based on this accounting system but they include the nitrogen stock in a farmers field at the beginning of the growing season. In the study that is being reported here, we refined these recommendations to adapt fertilizer applications to the variation within a field. 5.2.3. Functioning of the CoP and research performed In this case, the CoP was simple and consisted of the Van Bergeijk family discussing research with a group of researchers and students. One wonders whether the term CoP should apply here as other land users in the area were not directly involved, nor were extension workers or local representatives of the Ministry of Agriculture. Still, contacts were strongly but indirectly influenced by these persons so the term CoP is still appropriate, even though the Community is somewhat diffuse. Research focused on the development of a system for precision agriculture that allowed for a properly timed, site-specific fertilizer recommendation. This was done through (1) a description of the growing conditions on the farm, (2) the definition of management units, and (3) real-time simulations to determine the nitrogen requirements. 5.2.3.1. Farm survey Following the standard procedures for soil survey as laid down in, for example, the Soil Survey Manual (Soil Survey Staff, 1993), a detailed 1:5000 soil survey was conducted in the study area with approximately six soil auger observations per hectare. Results were stored in a soil database containing soil physical and chemical properties for individual soil layers. Texture and soil organic matter content were estimated directly in the field and tested against a limited number of laboratory measurements to ensure accurate characterization. On the basis of these properties, soil layers were grouped into relatively homogeneous classes as defined by the Staringreeks (Wo¨sten et al., 1994). This classification distinguishes between topsoil and subsoil layers that are further differentiated by textural composition and SOM content. Sixteen classes were identified and sampled in the field. Average bulk density and saturated moisture content were determined for each class using at least four replicate samples. Soil hydraulic characteristics were derived through a continuous pedotransfer function (Wo¨sten et al., 1998). Based on the layer information, soil profiles were classified according to the standards of the Dutch 1:50,000 national soil map (De Bakker and Schelling, 1966). 5.2.3.2. Determination of management units In the past, precision agriculture has often focused on managing microvariability within fields. Although technically feasible, this variability is often not achievable for practical farm management and too costly. In a discussion with the farmer,

196

J. Bouma et al.

we therefore decided to establish management units at a higher scale level (Van Alphen and Stoorvogel, 2000a). The farmers considered the larger management units to be easier to manage. For the researchers, it became thus possible to run real-time crop growth simulation models for a limited number of locations. The 1:5000 soil survey provided us with large soil units in terms of the Dutch national soil map, but the relevance of the classification for agricultural and environmental purposes was not clear. We, therefore, decided to carry out a functional soil classification on the basis of four functional properties, each relevant from a production and environmental quality point of view 1. 2. 3. 4.

water stress in a dry year; N-stress in a wet year; N-leaching from the root zone in a wet year; residual N-content at harvest in a wet year.

The first two properties describe the sensitivity of a soil to the effects of major growth-limiting factors and, as such, they are linked to crop production and are relevant from an economic perspective. The other two properties are environmental factors and directly relate to N-leaching. The four factors are determined by applying the mechanistic-deterministic simulation model WAVE (Water and Agrochemicals in Soil and Vadose Environment) (Vanclooster et al., 1994) for each of the observation points of the original soil survey. Subsequently, fuzzy clustering was applied resulting in four typical management units that differed in terms of water stress, N-leaching, and residual N. Membership values were assigned to each observation point indicating the similarity of the points to the management unit. Each of the four membership values were next interpolated with ordinary kriging (Section 2.2). The four maps were combined into a single confusion index (Burrough et al., 1997) that formed the basis for a boundary detection for creating the map with four management units (Fig. 3). These management units were therefore not distinguished on the basis of measured yield patterns in different years, as is widely done elsewhere, but on a functional characterization of soils. Table 1 shows the results for the farm. During the simulations, no nitrogen stress was observed. As to be expected, the four management units describe very well water stress, nitrogen leaching, and residual nitrogen at the end of the growing season. Soil texture, an important classification parameter, is equally well described in the taxonomic and functional map. Soil organic matter is much better described by the functional map as large variation occurs within the soil types. This makes the soil types of limited use for fertilizer recommendations. 5.2.3.3. Real-time simulation of crop growth and nutrient depletion and resulting fertilization regime A simulation model and real-time weather were used to monitor soil mineral N levels in the management units during the

197


Management units 1(21.7 ha) 2(8.3 ha) 3(45.8 ha) 4(9.2 ha)

200 m

Figure 3 Four management units for the Van Bergeijk farm as determined by functional characterization.

Table 1 Soil functional and soil physical properties for the four management units in the experimental field

Unit

1 2 3 4 R2 (%) Func. Tax.

Water stressa N-stressa

N-leaching (kg N ha1)

N-residual (kg N ha1)

Clay, % (0– 100 cm)

SOM, % (0– 100 cm)

1 29 2 4

0 0 0 0

29.8 37.1 36.8 56.3

90.1 106.3 86.9 154.5

41.8 46.3 30.7 33.7

0.9 1.8 1.2 3.1

86 13

100 100

71 41

70 32

56 53

70 17

a Stress is defined as the percentage yield loss as a result of water and or nitrogen stress. R2 indicates the percentage of spatial variation explained by functional (Func.) and traditional ‘‘taxonomic’’ classes (Tax.) [adapted from Van Alphen and Stoorvogel (2000a)].

198

J. Bouma et al.

entire growing season (Van Alphen and Stoorvogel, 2000b). The farmers reported on a weekly basis the weather conditions at the farm after which an early warning was provided when soil mineral N concentrations dropped below a critical threshold (defined by twice the expected N-uptake in the coming week). N fertilizer was only applied when the N-stock was low. Figure 4 presents the results for the management units 1 (left) and 3 (right). After an initial application of 80 kg N ha1 on day 52 (application 1), on the first management unit, the threshold level for N depletion was first reached on day 110 (application 2). Soil available N levels had dropped to 26 kg N ha-1 and crop uptake had reached 6 kg N ha-1 week1. If decision rules had been applied strictly, the threshold level would have been set at 12 kg N ha-1 (twice the uptake rate), indicating no action was required. However, uptake had been extremely low in that particular week, caused by unfavorable weather conditions (low radiation and temperature). Two weeks earlier, uptake rates had been up to 14 kg N ha-1 week1, indicating action was required (threshold at 28 kg N ha-1) under normal conditions. Taking this into consideration, it was decided to apply a second fertilizer dose on day 113. The fact that this coincided with standard fertilization was purely coincidental and was caused by weather conditions. The standard application had been planned 2 weeks earlier, but heavy rainfall had kept machinery from entering the field. The fertilizer rate for the second application in the ‘‘precise’’ system was established at 60 kg N ha-1. This quantity was derived through an exploratory or ‘‘forward-looking’’ simulation. Starting with the situation on day 110, the simulation covered a period of 4 weeks, corresponding to the estimated interval between applications two and three (the latter was scheduled for the second half of May). The calculated rate proved accurate, as threshold levels were reached for the second time on day 138. By then, soil available N concentrations had

kg NO3 ha−1 (0–90 cm)

80

A1 A2

60

A3

60

40

40

20

20

0 30

A3

80

90

150

0 30 210 Day number

A4

A1 A2

90

150

210

Figure 4 Total N in the rootzone and N-uptake by the plants during the growing season as a function of the precision fertilization regime for the two main management units 1 (left) and 3 (right).


199

dropped to 24 kg N ha-1 and crop uptake had reached 17 kg N ha-1 week1. The third application occurred soon after on day 140. Again traditional and ‘‘precise’’ fertilization coincided, only this time both were applied according to schedule. A second exploratory simulation was conducted, resulting in a recommended fertilizer rate of 45 kg N ha-1 for application three. The scheduled fourth and final application is traditionally applied just before anthesis. This was expected to occur on day 161. When flowering was witnessed on day 156, a fungus infestation was found throughout the field and additional fertilizer was expected to only deteriorate crop condition. Therefore, the fourth application was canceled. A comparable sequence can be described for management unit no.3, except that a fourth application was feasible here. Figure 4 illustrates that the dynamics and timing for the different management units are different because the soils function differently. Precision management resulted in a total fertilizer dosage of 185 kg N ha-1; a reduction of 55 kg N ha-1 or 23% compared to traditional management. Simulated nitrogen concentrations were validated against a limited number of field observations giving satisfactory results (Van Alphen, 2002; Van Alphen and Stoorvogel, 2000b). Also, some exploratory measurements of groundwater quality were made but no systematic monitoring program was performed. The quantitative mechanistic simulation model (K5, Fig. 1) was overruled twice by qualitative expert knowledge (K1 in Fig. 1). First, a fertilizer application was temporarily suspended because field conditions were not considered appropriate to enter the field with the fertilizer spreader. Later, an application was cancelled due to the presence of a fungus infestation. This illustrates the importance of K 1 expertise for farm management. The application of this high resolution time and space nutrient management procedure resulted in a reduced fertilizer input by 23% and by a significant decrease of leaching of nitrates by an estimated 55 kg N ha1. It should be noted that under Dutch conditions, residual mineral N at the end of the growing season will leach during the humid fall and winter to the groundwater. As a result, there is little residual effect of excess fertilizer on next year crop. The key threshold of the EU for nitrogen content in groundwater is 50 mg NO3 l1. A simple calculation supports the importance of precision agriculture. Under Dutch circumstances, one kilogram of N leached to the environment roughly corresponds to an increase of 0.04 mg N l1 in runoff and groundwater (assuming a precipitation excess of 250 mm water year1). As a result, a reduction of 55 kg fertilizer N ha1 leads to a significant reduction in ground and surface water concentrations corresponding to an average of 10 mg NO3 l1. In other words, using current application rates suggested by extension specialists, farmers are much more likely to exceed the threshold values. Applying precision agriculture, on the contrary, it is likely that they will stay under the threshold.

200

J. Bouma et al.

As crop yields under precision agriculture management were as high as under conventional management, environmental quality goals are reached without adverse effects for productivity. 5.2.4. Multiscaling techniques used and needed new research Fuzzy clustering and ordinary Kriging were used as an upscaling technique on the farm to obtain management units from point data. These are modelbased techniques (Section 2.2.2) that make specific assumptions about the spatial variability of the relevant soil properties. Data obtained by functional characterization of soil pedons were used rather than primary soil data. Upscaling to other farms with possibly identical soils was not possible because no link was made with soil survey classification, allowing possible extrapolation of data obtained for the dominant soil type Mn25A to other areas where this soil type occurs. Questions by farmers relate to their entire operation and still have a dominantly economic focus even though environmental concerns become much more important not in the least because of stringent environmental rules and regulations. How data on samples, soil horizons, and pedons are upscaled to the farm level is a concern of scientists, not of farmers. The demand analysis, as presented in Section 5.2.2 has resulted in a research that allowed reduction of use of chemical fertilizer by 23%. Clearly, the often proclaimed opinion that ‘‘what is good for the environment is bad for business’’ does not apply here. In contrast, producing in an environmentally friendly manner can save considerable money! A demand analysis followed by research also shows where lack of knowledge has impaired progress of the work. In other words, it shows where white spots occur in our understanding. The methodology that has been developed now faces an interesting future: How to implement a high-tech system for precision agriculture on a large scale. Tools for real-time simulations are still very much research tools and they can only be applied by research teams. This means that the CoP is equally important for the implementation and support of the tool. A good example where simulation models are used at a large scale to support farm management is the Agricultural Production Systems Simulator (APSIM; McCrown et al., 1996) that is used by Australian farmers. The Van Bergeijk case study indicated the following research needs: 1. Linking up with Soil Survey. The soil survey made for the farm was based on sampling in a grid and was not associated with available soil surveys that were too general in character to be useful. Still, the challenge exists to link up with the soil survey system in terms of their soil series definitions to build a database for particular soil series occurring in the area that can be used in future in areas where the same soil series occurs. The alternative is to use a grid survey time and time again for


201

different farms, relating basic soil data to model parameters, without creating any opportunity for learning and extrapolation. Expanding soil survey interpretations beyond the current K2 level to the K5 level, which is necessary for modern applications (as demonstrated on the Van Bergeijk farm), is therefore necessary. This includes functional characterization as made in this study and implies an entirely new field of research. 2. Translation of existing soil data into data and parameters that can be used for functional characterization, which often implies application of modeling and simulation techniques. The term ‘‘pedotransfer functions’’ has been coined for this process (Bouma, 1989) and many studies have been made already (e.g., Pachepsky and Rawls, 2004). This is still an attractive area of research considering the large soil databases in many countries and the increasing use of modeling techniques. 3. Distinction of soil management effects in soil survey. Traditionally, management effects on soil classifications have been avoided by requiring the upper 15 cm of soils to be mixed before classifying. The purpose of this was to avoid changes of classification following tillage. However, we find considerable differences in soil properties within a given soil series as a result of different forms of management. Droogers and Bouma (1997) proposed the term phenoform to express such differences within a given soil series, the latter representing the genoform. The phenoform concept was worked out in detail for a loamy, mixed, mesic Typic Fluvaquent (Soil Survey Staff, 2006) by Pulleman et al., 2000, and by Sonneveld et al., 2002, for a coarse loamy, siliceous, mesic Plagganthreptic Alorthod (Soil Survey Staff, 1998). Bouma and Droogers (2007) analyzed these two examples. Distinction of phenoforms is an important tool for linking soil survey expertise and data with soil management. After this study, conditions changed for the farmers. Excess chicken and pig manure from other areas in the country presented such severe local problems that farmers elsewhere were paid to use this organic manure. The Van Bergeijk farmers accepted this opportunity for obvious economic reasons and replaced their precision system based on mineral fertilizers with a system using organic fertilizers. As the nutrient content of organic manure and mineralization rates are rather variable, organic manure is much more difficult to fit into a precision system as described.

5.3. Case study Spruit: Production of ‘‘new’’ manure 5.3.1. Questions raised Research was done in 2004 and 2005 at the Spruit farm in the Central Peat district of The Netherlands (Fig. 2). The farm covers some 37 ha of grassland on peat soils that are classified as Terric Histosols (Soil Survey Staff, 2006).

202

J. Bouma et al.

For the years 2004 and 2005, annual rainfall was 926 and 835 mm, respectively. Average annual temperature for both years were 10.6 and 10.8 C, respectively. The research followed a public outcry following several court convictions of dairy farmer Spruit who refused to apply farm manure to his grassland in the manner prescribed by law. To reduce ammonia volatilization from manure, farmers in The Netherlands are since 1995 required by law to inject their manure into the soil or apply it in small bands onto the very soil surface (Huijsmans, 2003). Traditional aboveground surface spreading of manure on grassland is now forbidden by law as plot studies have shown that surface spreading results in significantly higher ammonia volatilization as compared with injection. However, a number of farmers refused to follow the new regulations because of assumed detrimental side effects of injecting manure to soil fauna and meadowbird populations in addition to soil compaction by the heavy, special equipment. This equipment is expensive and can often only be bought by contractors serving many farmers. This strongly reduces management flexibility of individual farmers who are now frequently confronted with applications under nonideal circumstances as contractors determine de facto the moment of application. In contrast, traditional surface spreading is done by the farmers themselves allowing them to select the proper moment. This is important because weather conditions affect ammonia volatilization (Klarenbeek and Bruin, 1990) whereby relatively low emissions are found during dark, rainy weather and higher rates under sunny and windy conditions. One way farmers have tried to deal with this problem was to produce and apply ‘‘new’’ manure resulting from feeding less protein to the cows that, in turn, produces manure with a relatively low total N and ammonia-N content (Kebreab et al., 2001) and, as a result, less possible ammonia volatilization. Ironically, one farmer even received an environmental innovation award because of his ‘‘new’’ manure, but was nevertheless fined when he applied his ‘‘new’’ manure in the traditional manner by surface spreading. The farmer claimed that traditional surface spreading of this ‘‘new’’ manure, to be done by himself under dark, rainy weather conditions, was at least as effective in reducing ammonia volatilization as compared with injecting conventional manure with more total N and a relatively higher proportion of ammonium-N. Being a well-known and respected farmer in the so-called ‘‘Green Hart’’ of The Netherlands (one of the Dutch National Landscapes), his convictions inspired a number of prominent citizens to write a letter of protest to the Minister of Agriculture. Writers included members of Parliament, the chairman of the prominent Socioeconomic Council of The Netherlands, a famous writer, a famous actor, and several professors of Wageningen University. As a result, the Secretary of Agriculture agreed with a proposal to initiate a monitoring program to investigate the claims being made as there was as yet no scientific proof for this farm that


203

‘‘new’’ manure indeed reduced ammonia volatilization nor of the adverse effects of injecting manure. There is more at stake here than a simple question as to the effects of different treatments of manure. Existing rules and regulations on ammonia volatilization in 1995 were derived top-down by policymakers and researchers with little involvement of farmers. If farmers do not follow the rules, fines are imposed, defining lack of compliance and distrust as guiding principles. The alternative would be to set environmental guidelines for, in this case, air quality in terms of allowable ammonia volatilization and allow farmers to reach these goals using their favorite management practice, whatever it may be. Of course, field measurements would have to be made to see whether quality standards were met. In this approach, management abilities of farmers would receive most emphasis, defining trust as the guiding principle. In a time when relations between citizens and their government are widely considered to be problematic, such a shift in emphasis would be most welcome. Allowing this program to proceed implied that Spruit was exempted by the authorities from the existing regulations for a period of 2 years. Initial questions included: (1) how much ammonia volatilization occurs from ‘‘new’’ manure? and (2) how does injection affect soil fauna and structure? 5.3.2. Demand analysis Questions raised covered only a few aspects of environmental quality and to arrive at meaningful and applicable research results at farm level, these questions had to be placed into a broader farm management context, also considering EU regulations, such as the Nitrate Directive (EC, 1991) and the Water Framework Directive (EC, 2000). The former aims for a maximum nitrate concentration in the upper groundwater of 50 mg per liter, to be reached by a maximum application of 250 kg N from manure per hectare for farms that comply with a number of criteria, such as having more than 70% of the area consisting of grassland. The latter outlines requirements for surface water quality. Here, the aim was to document the environmental effects of producing and applying ‘‘new’’ manure, not only focusing on ammonia volatilization but also on water quality. As is, regulations for ammonia volatilization, nitrate pollution of groundwater and nitrogen, and phosphorous pollution of surface waters are all separate legal entities. This is difficult to work with in practice and impossible to explain to farmers because processes in air, water, and soil are highly interrelated. The water-quality aspect will not further be discussed here (for results see Sonneveld et al., 2008) as attention will be focused on ammonia volatilization. In addition, the decision was made to work in an operational whole-farm context where the farmer was followed in his actual management rather than guided by the researchers. From a scientific point of view, it is, of course, also possible to study ammonia volatilization in small plot experiments where

204

J. Bouma et al.

each plot receives manure with a certain composition. However, this does not reflect important real-world management effects that play a dominant role in determining emissions at farm level. What, for example, is the relation between the composition of manure and the feeding regime being followed at any particular farm? When and how frequent is manure applied to the field and why at such particular times? What are soil and environmental conditions at such times of application? As mentioned earlier, the latter is important as it strongly affects volatilization. To allow a comprehensive analysis of N-fluxes on the farm, a farm budget model was used (Schro¨der, 2000). This model represents a policy-oriented method for upscaling processes occurring in the various fields to the farm level (Fig. 5). In summary, the primary outcome of the demand analysis was the decision to investigate whether environmental threshold values for air and water quality were exceeded when using different forms of management, with particular emphasis on the effects of producing ‘‘new’’ manure. To do so, a budget model was applied. This broadened the initial isolated questions on ammonia volatilization and effects of soil traffic to a comprehensive farmwide systems analysis. Taking the time to do a demand analysis in close interaction with the farmer turned out to be rewarding in the end because it allowed meaningful expansion of research beyond answering partial, isolated questions. Continuous involvement of the farmer in the discussions assured his cooperation and commitment. This procedure is time consuming and asks communication skills of scientists involved that are not part of their University education.

Imported feed

Milk and meat

Herd

Forage

Indoor excr. manure

Outdoor excr. manure

Exported manure Mineral fertilizer

Clover

Deposition and mineralization

Soil

Other losses

Ammonia loss

Figure 5 Schematic diagram for a farm summarizing N-fluxes that functions as a policy-oriented upscaling tool from field to farm (after Schro¨der, 2000).


205

5.3.3. Functioning of the CoP and research performed This CoP was an unusual one as it included, at least initially, several prominent Dutch citizens. However, later interaction was also not confined to scientists and the farmer but included representatives of the Ministry of Agriculture and of the Environment that followed the experiments closely, as did representatives of the farmers union and local action groups striving for preservation of the ‘‘Green Hart.’’ Based on the demand analysis, an experimental research program was designed and performed with the following steps: 5.3.3.1. Setup of a monitoring system for ammonia volatilization From a farming system perspective, ammonia is not only lost after application of manure to the field (which is considered in the law) but lost from manure inputs during grazing, from stables, and from manure storage locations. The farm consists of approximately 37 ha of grassland, divided into 14 fields, and holds some 80 dairy cows, along with young calves and bulls. The integrated horizontal flux method was used to measure the emission from one field (4 ha), which was considered to be representative for farm operations. This method is based on the integration of the product of ammonia concentration and wind speed over a pole-height of 12 m perpendicular to the wind direction. The ammonia concentrations were determined in passive samplers located on the pole (Sonneveld et al., 2008). The total ammonia emission was determined over a period of 36 h in different phases: a first phase during spreading, a second phase until the morning of the next day, and a third phase until the end of the next day. When application of the manure took place, both the amount, composition, time span, and fertilized area were determined, as were meteorological conditions including wind direction, wind speed, temperature, relative humidity, and precipitation. These complex measurements of ammonia emission using the integrated horizontal flux method were determined three times in 2004 and once in 2005 (Tables 2 and 3). Note that this work at the Spruit farm in 2004 represents the first field measurements of ammonia volatilization under real management conditions because the environmental law on ammonia volatilization was introduced in 1995. The number of field measurements was constrained by budget limitations. To obtain additional data, also the micrometeorological mass balance method was used to measure the emission from small plots of about 0.2 ha (e.g., Huijsmans, 2003). This allows comparison with previous ammonia studies where only this method was used. Two experiments were carried out with the micrometeorological mass balance method. In 2004, the manure from the Spruit farm was applied on two grassland fields using broadcast spreading. In 2005, the manure from the Spruit farm was applied

Table 2

Slurry application during the experiments Manure Composition

Experiment

Area (ha)

Application (m3 ha1)

Method of application

N-total (g kg1)

N-mineral (g kg1)

Dry matter (g kg1)

1 2

– 3.85

– 15.6

– 3.9

– 1.4

– 108

3

3.85

15.6

3.4

1.3

88

4

3.70

11.4

– Water sludge Water rain Water

3.7

1.5

95

207


Table 3 Ammonia emissions after slurry application measured with the integrated horizontal flux method

Experiment

Emission (kg NH3-N ha1)

Emission (% applied NH4-N)

Emission (% applied total-N)

2 (two phases) 3 (all phases) 4 (all phases)

3.7 3.5 10.9

18.5 17.8 68.4

6.5 6.9 27.3

using broadcast spreading on one field, whereas the manure from the nearby Zegveld experimental farm was applied using band application as prescribed by law. Measurements on gaseous emission of ammonia from the naturally ventilated stables were made using the internal tracer ratio method (Sonneveld et al., 2008). This was done for two periods in 2004 and during two periods in 2005. Ammonia emissions from the open manure storage facility outside the barn were measured using the open-path tunable diode laser (Boreal Laser, Inc., Gasfinder 2.0Ò ) and a photoacoustic monitor (Innova 1312Ò ). These measurements were performed for 1 day in 2004 before and for 2 days in 2005. Further details of the monitoring methods used are presented by Sonneveld et al. (2008). Manure samples were also taken for chemical analysis at three different dates and analyzed for total-N, mineral-N, pH, dry matter, and ash content. In summary, cutting-edge technology (K5 level of Fig. 1) had to be and was mobilized to answer questions raised by the demand analysis. To really answer those questions, simpler and cheaper methods would not have provided adequate answers. Obviously, this approach has major financial and operational consequences, but we found that funding agencies are positively inclined to fund well-reasoned proposals. Field experiments on the effects of manure injection on soil biology have shown that earthworm populations are strongly affected by weather conditions in the previous winter and studies on the impacts of application method on earthworms are inconclusive (Van Vliet and De Goede, 2006; De Goede et al., 2003). Results do not support local opinions about the detrimental effects of injection on earthworm populations. Research on soil compaction is still in progress. 5.3.3.2. Data acquisition Following the management practice on the Spruit farm implied that scientists had to be ready to monitor whenever the farmer was planning to spread his manure, and such plans were usually communicated on short notice. This created a logistic challenge that worked out well, except in experiment 4. Experiment 1 failed (Table 3)

208

J. Bouma et al.

because of changing wind directions during the measurement period. The other three experiments were successful. Experiments 2 and 3 showed that relative ammonia emissions were substantially lower (18%) than the commonly assumed 68% for surface spreading, a percentage that is based on experimental studies (Van der Hoek, 2002). This value of 68% was measured at the fourth experiment but this experiment was made at the request of the researchers at a time with sunny and windy weather and two weeks after the farmer intended to spread the manure. Measurements with the micrometeorological mass balance method also resulted in lower emission values (30–36%) compared to what was reported by Van der Hoek (2002). Farm data on N inputs and outputs were collected from regular farm visits and stored in a spreadsheet format. As the N-budget model also required data on soil N-mineralization; grass cages were installed to estimate N-uptake under nonfertilized conditions. Grass was mown regularly after consultation with the farmer to fit into his farm management and were analyzed for dry matter and N-content. Results of collected data are presented in Table 4 that shows the N-contents of the elements of the budget model in Fig. 4. The overall farm efficiency of N turnover is estimated to be 28%. 5.3.3.3. Evaluation of management measures Three components can be distinguished that contribute to the observed low ammonia emissions from this farm, as discussed above: (1) the low-protein feeding strategy, (2) surface spreading of slurry under cloudy, rainy weather conditions, and (3) the independent ability of the farmer to apply manure at the best time, also considering the growing stage of the grass. The study showed that innovative management leading to the production of ‘‘new’’ manure can satisfy environmental requirements as well as when injecting traditional manure. Results were reported to the Minister of Agriculture who demanded to see additional results before a change of law could be considered. So to test whether results of this study can be reproduced by other farms, with other soil-landscape conditions, a follow-up study was started with other farmers in the NFW region, to be discussed in Section 6.2.

5.3.4. Multiscaling techniques used and needed new research The upscaling issue at stake here was the calculation of ammonia emissions at the entire farm based on measurements at individual fields. As a method, this resembles soil fertility or erosion and runoff experiments on relatively small plots, to be extrapolated to larger areas. Each measurement reflected the effects of one type of management including the type of manure and the application method and is considered to be representative for this particular farm. But only two measurements (numbers 2 and 3) reflected the genuine practices of the farmer. Number 4 did not reflected because he was forced to

209


Table 4 Estimated N surplus (kg ha1 year1) and efficiencies (kg kg1) for 2003, ¨der 2004, and 2005 at the spruit dairy farm according to the budget model of Schro (Sonneveld et al., 2008). N

Input

Output

Concentrates replacers By-products, brewery’s spent grains, maize pellets Straw Solid manure Fertilizer N2-fixation (clover) Mineralizationa Deposition Total Milk Meat Manure Total

Surplus Efficiency a

2003

2004

2005

89 89

69 83

57 75

19 0 0 40 142 29 408 89 19 0 109 297 0.27

14 6 0 40 142 29 383 83 23 6 112 269 0.29

10 6 0 40 142 29 358 78 18 6 102 254 0.28

Estimated, partly on the basis of N-yields from in situ trials.

apply his manure at a convenient time for the measurement equipment but not for him, while measurement 1 did not succeed because of changing wind directions during the measurement. Examples obtained were convincing to policymakers even though two replicates of the Spruit management (numbers 2 and 3) and no replicates of the measurement under poor weather conditions (number 4) would hardly be convincing from a statistical point of view. Effects are demonstrated but no scientific proof is provided because of a limited number of replicates. The current work on ammonia volatilization in the NFW (see Section 6.2) involves two groups of 29 farmers, one group producing ‘‘new’’ manure and the other following traditional management. This is expected to provide, in time, answers that are statistically significant as shown, using a design-based analysis (Section 2.2.1). In terms of the policy cycle (Section 3.4), the signaling, designing and implementing functions were clearly performed. However, decisions to change the environmental laws will depend on results obtained in the NFW, a study that could only materialize because of the work at the Spruit farm. Experiences obtained led to clear indications for needed new research. This is an attractive procedure to define research priorities as compared with the usual procedure where priorities follow from disciplinary or

210

J. Bouma et al.

instrumental considerations only that do not necessarily reflect environmental regulations or broader concerns of society. 1. The measurement method with the integrated horizontal flux method was cumbersome and inflexible. Wind direction strongly determined the measurements. Also, equipment needed to be operated manually so a system had to be devised to inform researchers on very short notice about a planned manure application. Such a system cannot produce statistically significant results. For large-scale and long-term monitoring, there is a need for a large number of (preferably) automated measurement points that register air quality continuously and that can be accessed from distance. Such systems are being used elsewhere but are still expensive. Further development of these techniques is urgent. Currently, emission factors for the various application techniques of manure that have been derived from experiments at plot level are being used for emission calculations at a national scale level (e.g., RIVM, 2006). However, an ammonia gap has been observed between the few measurements at national level and model calculations (e.g., Erisman et al., 2001; Erisman and Monteny, 1998). It has been suggested that one of the reasons for the poor performance of simulation models is the inability to incorporate variability in the emission factors due to different forms of management. More onsite measurements, as being made now in the NFW, could solve this problem. 2. The approach taken can be described as a farming systems approach (Goss et al., 1995) where a dairy farm was taken as a basic unit of analysis and where actual management practices by the farmer formed the basis for research. This contrasts with common approaches where management–effect relationships are determined for single isolated parameters, such as ammonia volatilization from manure, at small experimental plots. More research is needed to streamline the attractive farming systems approach that yields data fitting directly into the management schemes of farmers. 3. Separate treatment of air, soil, and water quality, as prevails in current laws, needs to be replaced by an integrated analysis because there are many interactions between the various compartments. A comprehensive farm N-budget model was therefore used in this study to calculate environmental N-fluxes to soil, water, and air as a function of management. The budget model acts as an integrative device combining the different measurements on ammonia fluxes and soil N-mineralization and allows upscaling by translating field data to farm level for yearly periods. More research is needed to better characterize the components of the budget model, for instance, in terms of greenhouse-gas emissions and flow of nutrients from the soil to surface water.


211

6. Case Study at Regional Level 6.1. Introduction 6.1.1. New laws presenting conflicting demands at regional level Several new laws and regulations covering land use on a regional level are being proposed and implemented on national and international level, the latter in the EU. A new Dutch law on spatial planning will take effect in 2007. It emphasizes the need to develop spatial plans for communities and provinces, while the impact of planning at the national level decreases. A new Dutch soil policy, broadening its focus from cleaning chemically polluted soils in the former law to sustainable land use in the new one, was introduced in December 2003 (Second Chamber of Parliament, 2003). In 2004, the Ministry of Spatial Planning proposed and initiated 20 so-called National Landscapes that have characteristic and unique features that deserve to be preserved (VROM et al., 2004). However, no funds to do so were provided. They also defined the so-called ‘‘three-layer’’ concept to be applied in spatial planning. The first layer describes the dynamic soil and water system and the associated ecology; the second, the transportation network such as roads and railways; and the third, human settlements [see also Figure 5 in Bouma (2005)]. A practical distinction can be made between layer 1a, defining soil and water dynamics and layer 1b defining ecological conditions. Any new land use plan needs to consider the sequence from layer one to three when developing the plan. This, obviously, offers a unique opportunity for soil and water research. At the level of the EU, legally binding guidelines have appeared for: (1) ecological Habitat features (2000) and (2) water quality (2000) and soil quality (2007). The latter defines seven technical soil functions such as production of food and other biomass, storing, filtering and transformation of compounds, providing a gene pool and habitat for organisms, acting as a carbon pool and source of raw materials. But the soil quality guideline also considers social functions such as providing the physical and cultural environment for humankind and preserving an archive of geological and archeological heritage (see also Fig. 7). All these guidelines have to be implemented at national level, where the Dutch government delegates its responsibilities to communities and provinces so the regional level becomes increasingly important. No doubt born out of good intentions, all these laws and regulations are often not mutually compatible and all of them focus on analyses and goals. Unfortunately, they are painfully short on suggestions and criteria for implementation. As a result, immense communication and implementation problems arise. An attempt was made by working groups in the EU to define procedures and necessary research to implement the soils

212

J. Bouma et al.

guideline (Van-Camp et al., 2004). However, they ultimately defined at least 200 areas for priority research, reflecting disciplinary approaches that essentially resulted in the end in adding up a series of different disciplinary shopping lists (e.g., Bouma and Droogers, 2007). The case study on the NFW to be discussed in Section 6.2 reflects all these developments and problems broadly sketched above. However, before this case study is discussed, a general demand analysis is needed for the regional level because many diverse and conflicting questions that are raised in the regional context require first a systematic analysis to focus the necessary discussion within CoPs toward a demand analysis including a description of needed research. 6.1.2. General demand analysis for the regional level The demand analysis for the farm level, presented in Sections 5.2.2 and 5.3.2, is quite simple compared with the one to be made here for the regional level. The wide range of questions arising from the introduction of a series of new laws and regulations, as mentioned above, cannot be the basis of a systematic approach without a logical framework in which different activities have its particular place and function in contributing toward the goals envisaged by the lawmakers. This framework is lacking at this time. Long discussions with stakeholders, policymakers, and various stakeholder groups resulted in the following proposal for a framework that attempts to combine various elements of the different pieces of legislation offering a platform to consider the many concerns and questions mentioned in Section 6.1.1 (see Figs. 6 and 7). Past DPSIR -system: D = Drivers P = Pressures I = Impact R = Response S = State

Future(s)

Present 1

Spa

D+P=I+R=

Spr

Sf1,2• • •,n

2 n Scenarios: R=D+P=I=

Cost−benefit analysis Decision making

Figure 6 Illustration of the DPSIR procedure, illustrating how the past (Spa), Current (Spr) and future states of a landscape (Sf 1, 2, . . ., n) can be derived from considering D, P, R, I, and R.


213

Landscape

3-layer model

7 soil functions

State (S)

1a. Land and water dynamics 1b. Ecology 2. Roads, railways, canals 3. Settlements

1. Production food and biomass 2. Storing, filtering, transformation 3. Habitat, genepool 4. Physical and cultural environment 5. Source of raw materials 6. Carbon pool 7. Archive geological and archeological heritage

Figure 7 Flowchart indicating how the state (S) of a given landscape can be characterized by systematically considering the three-layer model and the seven soil functions.

First, we decided to follow the DPSIR framework as proposed by the EU Soil Protection Strategy (see also Van-Camp et al., 2004). Here, S represents the state of the land; D represents drivers of land use change, P are the resulting pressures on the land, I is the impact, and R, finally, indicates development of strategies and operational procedures for the mitigation of the threats. The flowchart in Fig. 6 shows the past, present, and future state S of the land. Drivers and pressures in the past have lead to impacts and, most likely, certain responses. This all results in a present state S. Of particular interest, of course, are future developments that are considered in terms of different scenarios defined by scientists, stakeholders, and policymakers. Each one of these is based on a given response to problems encountered in the present state S. Each response is associated with characteristic drivers, pressures, and impacts. Scenarios may represent different visions on sustainability, effects of climate change, etc. Each one of the scenarios results in a new, future state S that is, of course, an estimate only. All the scenarios together are input into the policy process. A selection has to be made from all options being presented and this selection is made by politicians and citizens, not by scientists. The described scenario approach defining a series of states with all its attributes is therefore more realistic than presenting only one, ‘‘ideal’’ option as defined by a group of researchers. When considering sustainable development, environmental, social, and economic considerations and approaches have to be mutually balanced to achieve some type of compromise that is acceptable to environmentalists, social scientists, and economists alike even though it probably does not

214

J. Bouma et al.

represent the ideal solution as seen from their purely disciplinary point of view. Usually, economic considerations largely determine the outcome of this type of interdisciplinary analysis. The scheme in Fig. 6 suggests an alternative approach where environmental and social aspects—the latter expressed by environmental legislation—are considered first and economic considerations come later in terms of a cost–benefit analysis for each of the Sf scenarios. Recently, the DPSIR scheme has been criticized by Svarstad et al. (2008). They believe that the system is not neutral and reproduces discursive positions that applicants bring into it. Also, they observe a lack of effort to find a way of dealing with multiple attitudes and definitions of issues by stakeholders and the general public. In our opinion, the scenario approach, as illustrated above, can avoid the potential shortcomings as noted by Svarstad et al. (2008). When those scenarios are prepared by the participants in the CoP, they should reflect attitudes and issues of stakeholders and the general public as well as those of the participating scientists. The observations of Svarstad et al. (2008) are valuable as they implicitly illustrate the importance of open discussions within the CoP. Next, we need to define how S is characterized. What is the quality of a given state S? (see Fig. 6). We first decide to consider the landscape scale, as is promoted by all the new laws. Next, we combine the three-layer model with the seven soil functions, as defined by the EU Soil Protection Strategy: 1. 2. 3. 4. 5. 6. 7.

production of food and other biomass storing, filtering, and transformation of compounds providing habitat and gene pool providing the physical and cultural environment for humankind source of raw materials acting as a carbon pool archive of geological and archeological heritage

At first sight, there appears to be a slight overlap between the three-layer model and the seven soil functions, particularly for layer 1a and function 2 and for layer 1b and function 3. This is not the case as layer 1a represents the hydraulic regime in the area not considering solutes and their transformations that are considered in function 2. Layer 1b considers ecological conditions in a broad sense in terms of flora and fauna, while layer 3 interprets these conditions in terms of habitat. Consideration of soil as a gene pool represents a separate item of much greater detail. For a given area, the hydrological regimes are characterized first using hydrological data for the geological substrate and soil physical data for the unsaturated soil. Ecological conditions are defined in terms of vegetation patterns and animal occurrences. This completes layer 1, where a distinction between layer 1a for the abiotic aspects and layer 1b for the biotic ones is useful. Next, transportation routes (layer 2) and settlements (layer 3)


215

are distinguished. Then, the seven soil functions are systematically and sequentially investigated, as indicated in Fig. 7. This requires, ideally, application of many monitoring and observational techniques to the entire area. This may not be feasible and many data may not be available. Different types of knowledge may therefore have to be used (Section 3.3, Fig. 1). It is better to have K2 data than no data at all. Any given region will have a characteristic set of functions that are most important and deserve most emphasis. Still, consideration of all functions is desirable in a demand analysis for any given area, even though some functions may be minor at this time. However, they may become important in future. Ideally, each soil function in the area should be defined in terms of a certain and reproducible quality measure, which is, as yet, not available. The state S is defined when the given area has been characterized in terms of the seven functions. This, in turn, makes it possible to determine which of the scenarios is most attractive in policy terms after running cost– benefit analyses to indicate economic implications of each scenario (Fig. 6). The demand analysis thus results in a systematic and comprehensive framework for approaching the problems and questions being raised by a wide variety of CoP members. The framework functions as a constant reminder of the overall scope of land use studies as established in legally binding rules and regulations. Following the framework is attractive because it helps to avoid disciplinary bias and tunnel views. At the same time, the scheme implies that all regional studies must consider the soil functions when defining future land use scenarios. This is an excellent way to bring soil science back to where the action is. In summary, the general demand analysis for regional land use studies, as presented here, puts a large number of quite different questions on diverse soil functions by stakeholders and policymakers into context. The usual assumption that researchers should directly respond to any of such questions is likely to lead to chaos. Scientists can and should play a crucial proactive role in structuring the questions into some type of logical framework as presented here. How the discussions within the CoP proceeded in the regional NFW case study being presented here and how the demand analysis was translated into a research program will be discussed in the following sections.

6.2. Case study NFW: Agriculture in a national landscape 6.2.1. Introduction The NFW covers 60,000 ha and occurs in the northern part of The Netherlands (Fig. 2). The area is 1 of 20 so-called National Landscapes with landscape features that are to be preserved (VROM, 2004). Here, small elongated fields separated by hedgerows are characteristic. The area also contains geological phenomena from the last ice age, such as pingo remains

216

J. Bouma et al.

that are also considered to represent unique features to be preserved for future generations. Unique for The Netherlands, about 800 farmers in the area have united to form an environmental cooperative, striving for the development of sustainable forms of agriculture, using innovative management systems. Members of this cooperative are important actors in the CoP to be discussed later. 6.2.2. Questions raised Questions at regional level are particularly raised at governmental level: how to realize sustainable development with its economic, social, and environmental dimensions, the latter being partly reflected within EU regulations? Questions are particularly acute in areas where National Landscapes are to be established in The Netherlands, such as the NFW. But there are also major concerns at the lower levels of individual enterprises, among which farmers play a dominant role since approximately 60% of the Dutch land area is used by agriculture. In the NFW area, the percentage is almost 80%, mainly grassland. It has been suggested that strict compliance with the EU Water Framework Directive (EC, 2000) will be the end for Dutch agriculture. Questions are raised by farmers and politicians alike whether this is realistic and how water quality can be monitored and interpreted. EU agricultural policies will change in the coming decade. Export subsidies and import duties are likely to be removed in the context of free-trade agreements being negotiated. Also, restrictions on the volume of milk that individual farmers can produce (the EU quota system) will most likely disappear. Future land use scenarios will have to reflect such developments. At this time, farmers are particularly concerned about the water quality guidelines. In summary, according to the EU nitrate guideline, groundwater should not contain more than 50 mg NO3 liter1 and for surface water preliminary Dutch threshold values are 2.2 mg N liter1 and 0.15 mg P liter1. Because measurements of water quality are quite expensive, proxy values were defined to reach water quality goals. The EU Nitrate Groundwater Guideline of 1991 (EC, 1991) specified that not more than 170 kg N ha1 from manure should be applied on grassland per year. A request by the Dutch government was, however, honored to increase this standard to 250 kg N ha1 for the period 2006–2009 for dairy farms with more than 70% grassland. This application rate of manure roughly corresponds with 2.2 cows per ha (assuming an excretion rate of 114 kg N per cow per year). Some farmers feel that this rule infringes on their independence and want to be judged by water quality itself rather than some proxy value. With proper management, good water quality with respect to nutrients can be observed even at higher cattle densities than 40 cows ha1 (Schroder et al., 2007). Nitrogen excretion rates can also be lowered, for example, through lowering the protein content of cattle feed, producing manure with less mineral N (‘‘new’’ manure). This is also thought to be


217

important to reduce ammonia volatilization from manure as discussed in Section 5.3. Yet another concern has to be considered. The way in which manure regulations in future will be implemented needs to be reviewed. The farmers cooperative would like to take responsibility for reaching the environmental goals in the NFW region. Environmental goals are not questioned as much as the manner in which measures are formulated by the government to reach such goals. The Dutch government has traditionally focused on top-down, uniform, and enforceable policies for the entire country. It is, however, now much more open to decentralization of responsibilities and differentiation. Regional differentiation in manure policy is strongly supported by groups of farmers, such as in the NFW. All the concerns mentioned above were the major reasons for farmers in this area to form an environmental cooperative, thereby coordinating their questions and activities ensuring that farmer concerns would be heard in the debate. 6.2.3. Demand analysis The various environmental laws and guidelines, as analyzed in Section 6.1.1, present many different requirements leading to many and quite diverse questions of land users and governmental officials alike. Just reacting to any of these questions in isolation is not meaningful. In the general demand analysis, presented in Section 6.1.2, a scheme was therefore developed that allows a characterization of the state of the region in which the requirements of the various guidelines are systematically considered, following the DPSIR approach (as in Fig. 6). In fact, the scheme represents a form of policy-oriented upscaling from the farm to the regional level as well as downscaling from region to farm as criteria at regional level have to be acceptable at local, farm level. This demand analysis, following the schemes of Figs. 6 and 7, will be followed for the NFW area, focusing on questions related to water quality only because of space limitations. The overall study considers more aspects of the demand analysis as discussed in Section 6.1.2. 6.2.4. Functioning of the CoP and research performed 6.2.4.1. Discussions within the CoP Based on initiatives from the NFW region itself, a vibrant CoP is already operational for at least four years. The CoP is chaired by the Deputy Governor of the Province of Friesland and it has representatives of major stakeholder groups and governmental entities that operate in the area, including the NFW Farmers Cooperative, the Water Board ‘‘Wetterskip Fryslaˆn,’’ the Nature Preservation Society, the national Ministries of Agriculture and Environment; the major Farmers Organization; representatives of all municipalities within the area; and Wageningen University and Research Centre.

218

J. Bouma et al.

Four subgroups meet regularly covering four themes: (1) Agriculture, Environment and Water; (2) Landscape and Nature; (3) Meadow birds; and (4) Regional Economy. These hands-on groups constitute the core of the CoP for NFW. For the water quality work, researchers interacted most with representatives of the NFW farmers cooperative, the Ministries of Agriculture and Environment; the Province of Friesland; and the Water Board. Contacts with the latter group were most productive, resulting in funding for joint research on monitoring. It is important to start any particular subgroup CoP immediately when a certain activity is being planned, to ensure that all relevant parties are included and to define roles and responsibilities. Researchers operating within this CoP have a delicate role in reacting to questions and demands by the various stakeholders while at the same time working towards a coherent program that offers the promise to contribute towards sustainable development of the NFW area in future. At the same time, researchers have to guard their scientific integrity making sure that research performed in the end meets scientific quality standards. This delicate process was started by initiating a demand analysis that requires consideration of much tacit K2 knowledge of the various CoP members (see Section 3.3) while designing research and providing effective communication to make sure that the CoP is aware of the role of research and does not lose touch with research plans being developed. The position of researchers within a CoP, such as the one for NFW, is not necessarily assured from the start and is difficult to establish. In the egalitarian society of The Netherlands, one has to earn its position: it is not granted without being challenged. Land users are particularly sensitive to researchers collecting data after which they disappear to write their articles. A long-term commitment is necessary in tune with the policy cycle (Section 3.4), together with an ability to listen and communicate research plans and results in a manner that the land user can understand. Scientists are not trained to do this as they tend to fit in the traditional scheme of doing either fundamental, basic, or applied research or extension. New abilities have to be learned. Bouma (2005) suggested this to occur within CSP within which the scientific community prepares itself for the demands of modern society. The Royal Netherlands Academy of Arts and Sciences (KNAW, 2005) and Spaapen et al. (2007) have provided some indicators and guidelines for such CSPs These need to be tested and implemented in common practice. Ammonia volatilization and water quality were the starting points for the discussion in the CoP. A large monitoring program was initiated for ammonia volatilization from manure as a follow-up to the work reported in Section 5.3. Because of space limitations, attention will be focused here on water quality using the schemes presented in Section 6.1.2, covering the demand analysis.


219

The outcome of the demand analysis yielded three major questions: (1) How does the regional water system function; (2) What is the current quality of ground- and surface water in the Northern Friesian Woodland compared to these standards; and (3) is it possible to develop an environmental monitoring scheme to evaluate whether the region complies with water quality standards? Discussions on the first question were started by presenting a water systems analysis characterizing the dynamics of layer 1a and soil function 2 (storing, filtering, and transformation of compounds). The STONE model (Wolf et al., 2003) was used that is the standard national nutrient emission model used in the evaluation of Dutch manure policy. STONE calculations for the year 2000, using downscaled national data, were considered to represent a base level, reflecting national conditions. It was decided to also perform STONE calculations for the year 2004 using regional farm data of the NFW region rather than such downscaled national data. Farmers expected results to be rather different. The choice of the years 2000 and 2004 was dictated by the availability of data in existing national and regional databases. In addition, future exploratory scenarios (Sf in Fig. 6) reflecting water quality impacts following the application of applying ‘‘new’’ manure with less mineral nitrogen would be of great interest to both farmers and policymakers but have not yet been produced. Also future scenarios reflecting climate change and effects of various rural development schemes are foreseen for future work. Discussions about these topics showed that interaction with members of the CoP resulted in a better understanding of local practices making scientists more aware of problems when downscaling a national nutrient emission model to the regional scale without taking local conditions into account. Spatial modeling, however, should always be combined with monitoring of real conditions in the field, in this case in terms of water quality. Such monitoring data define present conditions (Spr in Fig. 6) and are the only objective way to check whether environmental goals have been reached. But monitoring data are also essential to calibrate and validate simulation models that are needed to explore the effects of possible future management practices. Please note that the design of a monitoring program to verify whether threshold values of environmental laws are met (‘‘compliance monitoring’’) is quite different from a monitoring program to calibrate a process-oriented model (see De Gruijter et al., 2006). Obviously, future conditions cannot be measured and modeling provides the only way to explore future effects. Discussions resulted in the widely shared conclusion that a scientifically sound monitoring system needed to be developed and the Water Board agreed to fund it. Such a conclusion would most probably not have been reached when scientists would have proposed such a monitoring system by themselves without discussions within the CoP. A number of key questions

220

J. Bouma et al.

needed to be answered when defining a robust monitoring system: (1) How many samples should be taken, where, and when?; (2) What methodologies should be used to measure environmental parameters?; (3) What are the environmental threshold values with respect to N and P?; (4) What are the sources of N and P in surface water and groundwater?; (5) Who is responsible for the monitoring?; and (6) What are acceptable costs as defined by procedures being used, sampling density, and intensity? These questions were discussed in several workshops and research was initiated to answer them. 6.2.4.2. Research following the demand analysis 1. Basic data: GIS data on soil types, hydrology, and land use were available in national Dutch databases and were allocated to each of the 398 grid cells in the area with a resolution of 250 250 m (Fig. 8). Farmers provided data on actual nutrient use on the farms, which was compared with data from nation-wide databases. The local Water Board ‘‘Wetterskip Fryslaˆn’’ possessed maps of the surface water system (Fig. 9) and indications on the quantities of inlet and outlet water into and out of the region. Also, there is upward seepage of water from greater depth in lower parts of the area and this is also likely to affect water quality flowing into either surface water or groundwater. All these sources are significant for N and P enrichment of surface water and groundwater. Use of the STONE model resulted in a refined impression of the functioning of the water system in the NFW region and the nutrient loads on surface water and groundwater. These values were compared with limited historical monitoring data.

N

N

Sand Clay Peat No data

Figure 8 Soil types and land use in the NFW area in the year 2000.

No data Grass Maize Arable Woods Other

221


Drained Controlled with weirs Free draining

Unknown Seepage No seepage of infiltration Infiltration

N

N

0

5

10

15

20 Kilometers 0

5

10

15

20 Kilometers

Figure 9 The local surface water system in the NFW area (left) and infiltration and seepage zones used in the STONE calculations for the year 2000 (right).

Groundwater and surface water quality had been monitored in the past by the Province of Friesland and by the Water Board, respectively. For six locations, the quality of the surface water had been monitored for almost 10 years on a 3 months basis. However, this monitoring setup by the Water Board was not meant to be used for compliance monitoring in the NFW area. Groundwater quality data of the monitoring network from the Province unfortunately turned out to be inaccessible due to major problems with information systems being used. 2. Water system analysis: The STONE model (Wolf et al., 2003) is used to calculate nutrient loads from the soil system to the surface water and groundwater. STONE calculates nutrient fluxes for different forms of land use, like grassland, arable land, and nature; and concentrations of N and P of soil water entering the surface water. In this study, STONE was applied at the regional scale of the NFW. The region was divided in 398, 250 250 m grid cells (Figs. 8–13), for which specific hydrological, soil physical, and soil chemical characteristics could be derived by downscaling existing soil survey maps. All cells are hydrologically separated, but are connected to various drainage levels, like ditches, canals, and subsoil. This implies that STONE behaves spatially like a semi three-dimensional model. All cells have a specific soil type and corresponding soil hydraulic characteristics for the unsaturated zone. The P status and mineralization capacity of the soil are also input for the model. In the STONE model, the water balance model SWAP (Belmans et al., 1983; Van Dam et al., 1997) is used to simulate water

222

J. Bouma et al.

Nitrogen (mg/Iiter) 5

5

National STONE plots

Regional NFW plots

Figure 10 Mean N concentrations in the soil water entering the surface water, calculated with the ‘‘downscaled 2000 data’’ using the national STONE plots (left) and the ‘‘downscaled regional 2004’’ using the regional NFW plots (right).

Phosphorus (mg/Iiter) 0.6

0.25 mm), which were most important for a stable soil structure (see Baver, 1940). Various weighted averages of aggregate size have also been suggested (e.g., Van Bavel, 1950). Organic matter and iron oxides were the accepted ‘‘cementing’’ agents holding soil particles together, that is stabilizing structure. The mechanisms for action of organic matter were vaguely known; it was considered that decomposition produced sticky substances that acted as glue. Clay was another important component required to achieve aggregation. There was considerable interest in the properties of clays affecting soil structure. Russell (1938) reviewed these studies. Baver’s second edition in 1948 added more material on types of clay minerals and aggregation. Coherence of clay particles was a key, and it depended on whether the clay was flocculated and on the influence of drying. Cultivation of loam soils at an intermediate water content, but not in wet or dry soils, could bring smaller units together to form aggregates. Bradfield (1936) pointed out that ‘‘granulation is flocculation plus.’’ There was speculation on the bonding forces, but little work had been done on mechanisms involved. It was realized that the internal makeup of aggregates needed to be specified to distinguish porous, granular surface aggregates from dense, angular aggregates. Baver (1940) stated that the

252

Benno P. Warkentin

macrostructure we see depends ultimately on the arrangement of particles of a size not visible to the naked eye. He devoted nearly 20% of his book to a chapter on soil structure. Russell (1938) drew attention to the role of microaggregates or units from which larger units are built up—a concept developed later as the hierarchy of soil structure. This concept of a hierarchal buildup of soil structure was not fleshed out until 50 years later (e.g., Oades and Waters, 1988).

4.2. Aggregation Warington (1900) had discussed the formation of compound particles that arise spontaneously under certain natural conditions, for example volume change from wetting and drying or frost. Russell (1933) advanced the hypothesis that clay particles were held together by oriented molecules of water around cations, through dipole–cation–dipole links. A modern concept. The water held the cations away from the clay surface. Crumbs formed when the water was removed on soil drying. He observed in the laboratory that kaolin did not form crumbs. He related this to low surface charge on kaolin. The cation exchange capacity had to exceed 30 meq/100 g to get crumbs in this way. Heńin (1935) in France carried out a series of studies on effects of clay properties on soil structure. He discussed orientation of clay particles affecting secondary particle formation. These studies are reviewed in his textbook (Heńin, 1976). Recently, researchers in France have explored this idea more completely to understand the structure of clay soils (e.g., Tessier, 1991). The considerable body of field studies carried out by Russian soil scientists in the early 1930s on formation of aggregates and the effects of different plants on structure was evaluated by Russell (1938) and summarized by Baver (1940). The Russian work stemmed from the interests and leadership of Williams (1935) who taught that structure was the key to soil productivity. He was a strong proponent of the desirable effects of a grassland system of agriculture on soil structure. Bradfield (1950) summarized the knowledge about soil structure at the end of this period. The dominant influence of organic matter on improving aggregation was aided by factors such as frost action and wetting and drying. He pointed out the increasing awareness by soil scientists of the role of soil structure in crop production, even though there was no common definition of the term. He saw conclusive evidence that the management systems then used resulted in deterioration of structure of arable soils in many parts of the world. ‘‘Available evidence makes it quite clear, I think, that we cannot make up for the structural deficiencies of a soil by liberal fertilization nor can we make up for fertility deficiencies completely by providing good structure.’’ He expected research to provide solutions for soil management.

253

Soil Structure

‘‘In conclusion I would like to say that it seems to me that investigations in soil structure offer an exceedingly promising field for future investigations.’’ This did not occur for another quarter of a century until two major ideas, structure as an hierarchal buildup of aggregates and structure as providing habitat for soil functions, were explored (Section 6.1). The interest in soil structure, to judge from the number of research papers published, peaked about 1950 (Table 5). This is consistent with my professional experience after 1950. While the importance of structure for crop growth was acknowledged, it was not a productive area for research. Aggregate-size distribution and its stability in water had been measured on many soils, and correlated with the known aggregating agents—clay, organic matter, iron, and metal oxides. The wet sieving method did correlate with field observations of good structure in soils where organic matter was the main aggregating agent, and under climatic conditions where the cultivated soils had originally been under forest. It was a not a useful method for soils in more arid areas with smaller aggregates, or for soils with clay contents above about 25%. After 1930, all textbooks on soil physics, beginning with Baver’s first edition in 1940, contained a chapter on soil structure. The second (1948) and third (1956) editions had only minor new additions to the soil structure chapter, indicating the sparse knowledge accumulated. The 1948 book had added extra pages on clay and on microbial activity. The chapters on soil structure discussed classification of structure, how structure is formed, stability and methods of evaluation. A blip in interest occurred in the 1950s with the introduction of synthetic polyacrylic molecules for aggregate stabilization. It was known that the aggregating action of organic matter resided in the decomposable fractions, for example polysaccharides. But these materials decomposed rapidly in soils. The idea then was to produce a molecule with similar action, but stable in soils (Quastel, 1952). In 1951, such a material, Krilium# was announced by the Monsanto Co. Numerous studies in the ensuing Table 5

Published papers on soil structure

Soil Science 1916–1928 1941–1953 1954–1972 Soil Science Society of America Journal 1957–1966 1972–1976 1977–1981

Structure

Porosity

7 32 1

0 2 10

17 2 4

10 4 5

254

Benno P. Warkentin

‘‘feeding frenzy,’’ prodded by availability of research money, showed it stabilized aggregates, although it did not form aggregates from dispersed soils (e.g., Hedrick and Mowry, 1952). However, it also became evident that nature produced aggregating materials at a small fraction of the price charged by the chemists. The products were not used for treating soils on large acreages. The chain of history leads from manure to synthetics and back to manure. However, the chemical products later found application for special soil stabilization projects and in stabilizing arid soils under furrow irrigation (Sojka and Lentz, 1994). The period from 1930 to 1950 was a time of limited resources, people and money, for soils research. Many studies were by individual soil scientists interested in soil structure as a (the) fundamental soil property as well as a property determining long-term (sustainable) crop response. Yoder (1937) discussing the significance of soil structure in relation to tilth, wrote ‘‘. . . soil tillage as practiced today . . . is an art rather than a science’’ and again ‘‘. . . soil science has not kept pace with practice in the field of tillage.’’ He asked rhetorically whether farmers would manage resources to sustain soil structure, or whether they would substitute fertilizer to maintain yields (Yoder, 1937). The period from 1950 to 1980 answered this question. Research money flowed after 1950 and more tractable problems were researched.

4.3. Field description of soil structure The field descriptions of soil structure adopted for soil classification and soil survey interpretations concentrate on structure of subsoil units, not tilth of surface soils. The field descriptions are based on categories suggested by Nikiforoff (1941) from a scheme for classification described by Zakharov (1927). Shape, arrangement, size, distinctness, and durability are the categories. These descriptions allow predictive estimates to be made of properties such as drainage, rooting, and slope stability. But the descriptions do not, and are not intended to, give information on processes of surface soil aggregate formation. A scheme based on classification of cavities and pores (Pigulevsky, 1932) was later developed for microstructure features (Brewer, 1964; Kubiena, 1938). Descriptions of structure are not now diagnostic criteria in the international soil classification system. Is this because structure is reasonably uniform across soil bodies, or because our understanding and description of structure is too general to be useful, or that structure is not an important characteristic of soils, or because structure is not quantifiable in terms useful for classification? Qualitative descriptions of soil structure have been developed and used in describing soils in the field based on three features: shape and arrangement, size, and distinctness or hardness. Shape can be platy, prismatic,

Soil Structure

255

columnar, blocky (angular or subangular), granular (with nonporous peds), and crumb (porous peds that occur in surface layers of soil) (USDA Handbook No. 18, 1951). The most common shapes are blocky, where surfaces of peds are adjacent, or granular where there is little accommodation between polyhedron units. The former occur largely in subsoils, the granular peds in surface soils. The term crumb, for very porous granules, is rarely used now. The other types occur in soils influenced by ice or by salt. Nearly equidimensional shapes are expected in natural materials. This is a minimum energy condition. There are no linear granules (Thompson, 1961). Terms such as ‘‘interaggregate porosity,’’ and ‘‘macrostructure’’ are used without definition. They are qualitative descriptions, using terms in common use and understanding. USDA Handbook No. 18 (1951) has a separate section for describing structure, later publications omit this section, assuming a common knowledge. Grain-size distribution is an important diagnostic criterion for classifying soils, but structure or porosity is not. Structure is not listed among analytical procedures in the international classification system, as shown in the World Reference Base for Soil Resources [WRB; FAO (2006), Rome, World Soil Resources Report 103]. ‘‘Structural development’’ and ‘‘platy structure,’’ as well as macro- and microstructure, are used as descriptive terms but are not defined. While structure is not a component of diagnostic criteria, it occurs under descriptive criteria for soil horizons (USDA, 1999). Structure is not a component of the higher orders in Soil Taxonomy or in the WRB, except for vertisols, where alternate swelling and shrinking results in wedge-shaped structural elements in the subsurface soil. In the WRB, the diagnostic criteria for the veronic (but not the mollic) horizon include a ‘‘granular or fine subangular blocky structure’’ in the surface horizon. The stability and type of structure described in the field for classification of soils is useful as qualitative information on drainage, bulk density, and rooting characteristics (root room). Microstructure studies in soil micromorphology have much more emphasis on pores. The results are used as confirmatory information in soil genesis and classification studies. Characteristics such as slickensides, which give information on genesis of the soil, are important in field descriptions (Brewer, 1964). A scheme for a more detailed description of soil structure at various scales is given by Babel et al. (1995). The first level specifies the scale of observation of aggregates from mega to micro scales. The second level specifies shape, the third packing of structural units, the fourth describes the surfaces, and the fifth other properties of distinct structure units that are useful for description. For specific studies, such a system would provide useful information for soil classification; for general use in routine mapping, it is likely too detailed.

256

Benno P. Warkentin

5. From 1950 to 1980 5.1. Aggregation and tilth And still the issues of tilth: how to measure it, how to predict resulting tilth under different management systems, and how to go from knowledge about structure to tilth. All the methods in common use, for example size distribution of stable aggregates, are indirect measurements of soil structure, which then have to be related to soil tilth or to the effects of cultivation. This eludes us to this day. Frese (1960) asked whether we have an outline or draft of understanding (Konzept) for a science of soil cultivation or tillage. His answer was in the negative. A USDA Joint Committee on Tilth reported in 1943 that empirical experimentation would not tell us which type of tillage was superior because we did not know what physical state was desired for a grain-cropping system (Shaw, 1952). They recommended a systematic approach to measurements of physical properties. The book ‘‘Soil Physical Conditions and Plant Growth’’ edited by B.T. Shaw (1952) was a review of the state of knowledge. The call for systematic, intensive measurements of a range of soil properties has been repeated a number of times (e.g., Hartge and Stewart, 1995). The results have not been encouraging. The alternative approach has involved modeling the processes leading to soil structure. The complex nature of structure, and lack of understanding of many of the processes, hinders the usefulness of modeling. We have not yet achieved our aim. Tillage during this period was still generally considered necessary to achieve satisfactory structure. While soil scientists were trying to define and predict tilth, farmers were substituting increasingly inexpensive fertilizers and new varieties to compensate for the gradual decline in tilth from intensive cropping systems. Some writers had pointed to undesirable effects of cultivation. In one of the early observations, Lee (1849) stated that the soil became exhausted by excessive plowing and hoeing. Two-thirds of the tillage done in the southern states impaired the natural fertility of soils because of oxidation of organic matter and leaching of soluble mineral elements.

5.2. Soil structure and soil erosion Some of the impetus for soil structure studies during this period came from the need to understand and predict erosion by water and wind. Soil erosion has an obvious connection to soil structure, but structure turns out to be a small part of a large issue. One aspect of soil conservation is prevention of soil erosion, especially from cultivated soils. Stability of soil structure is again only one factor in preventing erosion (Hudson, 1995). Where aggregate

Soil Structure

257

structure is not stable, loose grains of sand, silt, or even microassemblages of clay can be easily separated and transported. Stability and size of aggregates were obviously important. Measurements were made of particle sizes stable against mild mechanical forces, rather than of ultimate grain size by complete dispersion. The great importance of stable microaggregates was shown. Even though there were early research studies on soil erosion in the soil science literature (e.g., Wollny, 1878–1898), the later impetus was soil conservation in the United States developed as a social program for support of farmers’ income and for improvement of rural life. It did not arise as an application of soil science to a natural phenomenon, except possibly in studies of surface geomorphology done by geographers, not soil scientists. However, the methods developed during this period in the United States became the model for soil erosion control in the western world (Showers, 2006). This section concentrates on the stability of structure or the erodibility of soil, after a short statement about the larger aspects of soil conservation. Soil erosion is multifaceted in concept, in understanding by people, and in methods used to conserve soil. The concept includes optimum use of land, sustainable use, prevention of erosion for economic reasons, and preserving soil because of its inherent value as part of our biosphere. It can include degradation of soils by compaction or loss of fertility. Further, loss of soil is recognized by people with a range of interests from scientists studying natural resources to citizens with concern for future livability on the planet. Statements such as ‘‘a nation that destroys its soils destroys itself,’’ or ‘‘we hold the soil as a future inheritance for our children’’ resonate especially with the wider audience. Given this importance of erosion losses and the public support for solving the problem, engineers and scientists have tried to sort out the physical factors leading to erosion. Erosion is a complex result of interaction between rainfall characteristics and temporal soil diversity. There is only a general relationship with soils categorized in various ways, for example from soil mapping. The large diversity in soils and rainfall characteristics makes erosion prediction and control a formidable technical task. There is a large literature in the technology, as well as in economics and sociology of erosion control. Erosion control methods include biological or vegetation means and physical means such as structures to prevent concentrated water flow over the land surface. The strongest control over soil erosion that soil management has access to is keeping vegetation on the surface. This technology can be applied even where the diversity of land and water makes modeling and scientific understanding of the process possible only on a macroscale. It is the socioeconomic factors and landscape diversity, rather than predicting the stability of soil structure, that appear to be the less tractable problems. Social and economic factors drive protection against erosion. Recently, community-based efforts of various kinds have had some

258

Benno P. Warkentin

success, for example the ‘‘Landcare’’ movement in Australia (Ewing, 1998). To appropriate a statement: ‘‘It takes a community to save the soil.’’ The shear forces generated by moving water and wind break up aggregates; erosion removes soil particles from one place on the landscape and deposits them at another. Description of the influence of rainfall requires specification of intensity and duration of rainfall, slope of the land and length of slopes, the cover on the soil that intercepts entering water, and a soil resistance factor—erodibility or detachability. This erodibility is determined by bonding forces operating at the different aggregate sizes (Table 6). Most of the eroded soil carried by water is of silt and fine sand size, less than 250 mm, either single grains or aggregates of this size bonded by clay. The bonding forces in these aggregates are physical–chemical surface forces between grains in association with degraded organic molecules. Our understanding of these forces is insufficient to allow direct measurement for use in erosion modeling. Erodibility of soils has been estimated in several ways. Aggregate stability against moving water can be measured by wet sieving methods, measuring decrease in size and paying special attention to breakdown of larger aggregates to silt size. The first of these was the Middleton 1930 Dispersion Ratio, the proportion of silt and clay fractions easily dispersed by shaking in water. Many studies have developed equations for regression of erodibility on soil properties such as % clay, % organic matter, Ca content, and % silt. Higher clay, organic matter, and Ca generally provide higher stability, that is lower erodibility; high silt content increases erodibility. Ro¨mkens (1985) has a review of these equations. Numbers for many of these properties are available in databases from classification of soils, for example the WRB or the Soil Taxonomy. The descriptions of soil structure are adequate only for very general and inferred information on erodibility. The standard method to measure erosion is a field measurement, with a defined, standard applied rainfall, and standard slope and cover, with receptacles for measuring soil loss. These values are then used in the Universal Soil Loss Equation (USLE) to calculate average annual erosion losses for a region or a field. Erosion, in common with tilth, has not been modeled satisfactorily. Models of the erosion process are used for either prediction or for Table 6 Hierarchy of structure units Compound particles

Particle size

Pores

Domains Microaggregates Aggregates Clods

1 mm 10 mm 1 mm 10 cm

Micropores Mesopores Macropores

Soil Structure

259

explanation (Morgan, 2005). The very complex process of erosion requires that predictive models have components for the different processes, and that simplifications are necessary to provide predictions useful for land managers. Many of these erosion estimates are made for rill or sheet erosion, but soil erosion and water quality problems often occur from the concentrated water flow arising from major rainfall events (e.g., Cox, 2007). A one-in-ten year storm may exceed the damage in 10 ‘‘normal’’ years. These extreme events may become more common as global climate changes produce more severe events. Our fixation with the concept of averages makes consideration of these erosion events more difficult because we ignore them as being unusual. This does not make them less real. Does this call for redundancy in conservation planning? This ‘‘planning for extremes’’ requires consideration of individual events rather than averages, and makes the role of soil structure an even smaller part of the immediate answer. Our concern in this paper is with the soil resistance to erosion, the factor termed soil erodibility, or K value in the USLE. With knowledge of the other factors in the USLE, the remaining factor is soil erosivity. For prediction purposes, this can be estimated from measured soil loss on various soils, holding the other factors in the equation constant or putting in the values calculated from the physical processes for the other factors. Attempts to predict K from soil properties such as aggregate stability, organic content, permeability over a wide range have been only moderately successful. While resistance to erosion of soil by water and wind is certainly related to structure, our understanding of the components makes this only a general relationship that is of limited value in predicting erodibility from measured soil properties. In this it is similar to tilth, the diversity at all scales from micron to meter makes detailed specification difficult. This leaves soil knowledge as a secondary factor in soil conservation. The physical factors, rainfall, slope, and soil cover are more important. In the final picture, the stability of arrangement of particles in structure is delicate; exposed surface soil structure is unequal to the forces of nature. Soil must be protected from raindrops and wind. ‘‘Soils are meant to be covered’’ is the motto of Steve Groff (2006), a conservation farmer in Pennsylvania who uses a permanent cover-cropping system of soil management. This is the main lesson. Soil science, particularly soil structure studies in soil physics, has not made a major contribution to the practice of soil conservation. Agronomy, with emphasis on covering the soil surface with plants, has been much more important.

5.3. Soil mechanics studies of soil structure The stress/strain relationships of soil materials, studied in soil mechanics, depend on arrangement of solids and voids in the soil architecture. Strain is the deformation or movement of a soil mass that results from application

260

Benno P. Warkentin

of a load or a stress. Standard methods exist to measure this soil characteristic as a function of nature of the applied stress—one-, two-, or threedimensional and of soil properties such as water content. Pore water pressure, positive or negative, has a large influence on strength, that is resistance to applied strain. These measurements are applied dominantly to subsurface soils. The experimental results can be used to predict stability of a soil mass, but do not identify how it arises. In engineering practice, it is usually desirable to decrease the volume of voids to increase strength and resistance to shear; in agricultural practice, an increase in porosity is desired for better tilth and drainage. Studies on soil structure reported in the soil mechanics literature influenced soil science ideas in two areas, the influence of particle orientation on strength and stability and the nature of soil compaction. Many studies of the microarrangement of solid particles in natural or modified soils were reported in the soil mechanics literature in the 1950–1970 period. They provided conceptual models of arrangements, explanations for measured stress/strain curves, and contributed to ideas about bonds between particles. The work is summarized by Mitchell (1976), Smart (1975), and Yong and Warkentin (1966). The expectation that such knowledge would be useful in interpreting stress/strain relationships for engineering practice was only partly realized, and the interest declined. The mechanics concepts were applied after 1970 to aggregated surface soils in soil science studies to understand compaction and tilth. Clay and organic matter are dominant factors in determining size and stability of aggregates under mechanical and friction forces. Both inter- and intraaggregate stability are affected by cultivation and other soil management activities, although cultivation or even compaction scarcely changes structure at the micropore level (Table 6). Contacts between aggregates are formed, stabilized, and broken during seasonal weather changes and from soil management activities. The application of soil mechanics concepts and measurements to aggregates and aggregated soils meets the additional complication of heterogeneity. Soil mechanics concepts usually assume homogenous soils, an assumption more valid for subsoils. In soils with more than about 30% clay-size grains, the clay particle arrangement becomes a dominant factor in tilth. Stability at the microlevel can be assigned to physicochemical bonding between clay particles and is subject to manipulation such as flocculation/deflocculation. At clay contents above about 10%, aggregates are formed. The results of soil science studies on stress/strain relationships have been useful in designing cultivation tools and management options that maintain points of contact and void space of aggregated soils at the macroaggregate (larger than 0.5 mm) level (e.g., Larson et al., 1989). Understanding and predicting transmission of stresses is useful in designing equipment that minimizes compaction. These relationships are often expressed as

Soil Structure

261

experimentally measured compaction related to measured soil properties such as water content, or % clay.

5.4. Soil compaction Soil compaction results from forcing soil aggregates into closer proximity, usually by applying stress but also from applying shear forces. The result is a rearrangement of voids, with an increase in bulk density and a decrease in porosity. Size and volume of the larger pores is decreased with some volume increase in smaller size pores. The pore-size level at which these changes occur in the hierarchy of soil structure depends on soil conditions and on the force applied. The amount of these changes depends on soil characteristics and water content. Slurrying a soil could break down most of the aggregates, a dry soil would show the least change on compaction. There is a large soil engineering literature on methods of achieving compaction to increase strength of soil. Soil engineering practice and its literature is concerned with consolidation curves, the relationship between stress applied and resulting bulk density, and how this varies with water content (Yong and Warkentin, 1975). There is an optimum water content, between dry and wet, at which maximum bulk density can be achieved at a given level of stress. These results on soil compaction gave information on processes of compaction, although the direction for surface soils is usually to avoid or decrease compaction. There is also a large agricultural soils literature on methods to avoid or to reverse compaction. Methods for increasing porosity include use of cultivation implements, which rearrange structure at the higher levels, for example creating large interaggregate pores by breaking aggregates. These methods would increase drainage, aeration, and root growth. Interceding at lower levels would be by interparticle swelling, possibly by freezing and by actions of biota. This section on compaction will be restricted to discussions of the changes in soil architecture and the arrangement of solids and voids on compaction of tilled surface soils. Tillage and compaction are the two dominant mechanical ways in which structure of soils is altered rapidly through mechanical forces. These alterations are at the macroaggregate (5 mm) and interaggregate level. Biological and chemical forces altering structure of the micro (5 mm) level work more slowly through processes of organic material addition and degradation, and through addition of chemicals such as fertilizers or stabilizing agents such as calcium compounds. The mechanical forces are described in static or dynamic stress relationships and the strain results as changes in bulk physical properties of soil, for example bulk density, root room, seedbed quality, tilth, or hydraulic conductivity. Measurements of changes in pore water pressure during compaction may give indications of changes in structure at the microlevel.

262

Benno P. Warkentin

Information on changes at the microlevel has been obtained from measuring the changes in compaction after modifying surface forces, for example surface charge manipulation by altering soil pH or by adsorption of strongly bonded inorganic or organic ions (Shanmuganathan and Oades, 1982). The application of these concepts to single-grained soils (the assumption often made) is more straightforward than for aggregated or structured agricultural soils. A simple model could predict that compaction decreases interaggregate (larger) voids, while tillage rearranges and increases the volume of these larger voids, but is wholly inadequate to explain what happens with interparticle voids. Most of the emphasis on compaction of agricultural soils has been on soil and equipment management to prevent compaction, and on breaking up compaction. Recent studies in tillage have identified and described changes in microstructure. Models to summarize, generalize, and predict these changes have been very difficult to formulate because of the multitude of soil, environmental, management, and loading type variables. A summary of such studies was assembled in Larson et al. (1989).

5.5. The tilth story summary Soil structure studies from 1950 to 1980 were in the doldrums. This was a period of inactivity during the third quarter of the 20th century. Original ideas were few, especially those that could point the way to research studies that would further our understanding. Many measurements had been made of stability of aggregates against gentle water forces. Many regression relationships were drawn between aggregate sizes remaining after immersing in water with gentle shaking and percent clay or silt, or sand, and the total organic matter content. Water content of samples before testing was a factor, slowly prewetting aggregates resulted in higher stability. This stability was related to crop growth response in a corroborative general way and only mildly to root growth, but did not yield soil factors that could be used to predict response. Aggregate stability measurements seemed useful for originally forested soils where organic matter was the main aggregant. The lack of progress in describing and defining soil structure in the 1960s led to several large scale studies, with many soil scientists concentrating their specific expertise on the issue of soil structure. Examples of large-scale studies are the work on tilth organized by a USDA Committee (Shaw, 1952), the extensive study of agricultural soils in Belgium by De Leenherr and cooperators (De Leenheer, 1977), the 6-year study in Germany reported by Hartge and Stewart (1995), which contains a detailed key for soil aggregates. Soil structure is a concept similar to soil fertility, with no single measurement sufficient for description. Soil structure studies did not show the advances in theoretical background and in applications that soil chemistry, for example, had made.

Soil Structure

263

During this period, it was increasingly realized that deterioration of soil structure led gradually to lower yield, often masked by increasing application of fertilizer. This was the answer to Yoder’s question (Section 4.1). Compaction of surface soils and gradual loss of organic matter from cropped soils had resulted in deterioration of soil structure, which manifested itself in increased erosion, decreased readiness of drainage, and hence decreased traction. These changes were embedded in the cropping systems used, which could not easily be changed. Increasing interest in soil structure at the end of the 20th century is indicated by the papers in the journal, Soil and Tillage Research. Our approach to diversity has been to take more samples to obtain a statistically valid mean value. This may not be the tactic that gives the most information. Is there a method to describe and use this diversity for predictive purposes? The diversity may be more valuable for management, for example erosion control, than the average of measurements. What scale of information is needed—field or plot? Wringing of hands there was aplenty but no directions pointing to solutions or new thinking about how to approach the problem. A number of attempts at modeling soil structure, and predicting changes in soil structure did not overcome the basic issue of diversity at all scales of soil structure. The breakthroughs came in thinking about soil diversity and about structure as a habitat for soil functions.

6. Architecture for Soil Functions After 1980 6.1. Structure or architecture and ecosystems During the last third of the 20th century, it became generally recognized in soil science that soils performed necessary and unique functions in ecosystems. Ecologists had developed these ideas earlier for unmodified (natural) ecosystems, but now they were also being applied to agro ecosystems. Decreased research funding for agricultural crop production and increased funding for environmental quality research was an important driver for soil scientists. Concepts of soils viewed within a context of ecology began to appear in research studies and textbooks. Soil architecture was an important component of these concepts, the soil geobiochemical functions were carried out in the habitat created by the arrangement of voids and solids, that is the soil architecture (structure). Concepts of soils in an ecological context appear sporadically in the earlier soils literature. King (1897) in the introductory chapter to ‘‘The Soil’’ begins with ecological concepts of cycles, growth, life in the soil, energy, movement, water, and sunshine. Further discussion in the book is

264

Benno P. Warkentin

based on processes, for example soil ventilation rather than status of pore space, tilth rather than stability of structure, noncapillary spaces rather than pores. But the contents of his books were technological descriptions of the soil, for agronomic applications. Pedology in Switzerland developed with a bias toward soil functions in ecosystems (Sticher, 2001). Other statements that could be interpreted as ecology of soils were scattered in the soils literature. But these are unusual examples. The overwhelming impetus for considering soils in an ecological context came after 1970. The focus on soil structure now changed from aggregate size and stability to the arrangement of pores. Pore structure was based on a size hierarchy of solid structure units, from clay grains through microaggregates to aggregates visible in the field (e.g., Oades, 1987). Arrangement of the smaller units formed the next larger ones, with pores of a range of size and continuity (Table 6). Biotic organism size determined where in the poresize distribution they could exist. The size and continuity of pores would determine flux of gases, and hence whether the habitat was aerated or anoxic. Roots could develop in larger channels or would need the ability to enter small pores, and expand them as root diameter increased. At the same time, the relationship between hydraulic conductivity and nature of pore space has become an active area of research (hydropedology) on the vadose zone by earth scientists. Tuller and Or (2002) have a review. The term ‘‘soil architecture’’ seemed apt to describe different pore or void space sizes, connected by doorways of varying sizes, with walls available for attachment of water, chemicals, and microorganisms. From grand ballroom spaces to closets under the stairs, they provided different habitat for different organisms and different functions. And all of this depended on a stable structure or arrangement of solids. The emphasis in soil architecture is, however, on open spaces, or pores. ‘‘The Reality of the Building does not consist in roof and Walls but in the space within to be lived in’’ (Lao Tse, 6th century, BPE). A fitting motto for soil science now. The concept of tilth now becomes a concept of pores, not solids. This has likely always been the essence of ‘‘tilth.’’ The concern is transmission of water, chemicals, and fine solids in unsaturated soil. The interactions of these three in colloidal transport need to be unraveled (e.g., McCarthy and McKay, 2004). The model of soil structure becomes one of a hierarchy of aggregate sizes, from clay associations at the nanometer level to compound particles at the 5-mm scale. The association at different sizes led to the concept of aggregate hierarchy in soils, and its importance in relating structure to other soil properties and soil management (e.g., Dexter, 1988; Edwards and Bremner, 1967; Oades and Waters, 1988; Warkentin, 1991). Our understanding of the mechanisms for bonding at different sizes is at an early stage. Inter-clay particle forces at the 0.01–0.1 mm level have been described, based on our knowledge of physical–chemical forces of water

Soil Structure

265

sorption and interparticle attraction and repulsion (e.g., Quirk, 1994). Again at the macroaggregate level, generally taken as >250 mm, the effects of organic matter bonding, inorganic oxides, and roots have been described, even if only qualitatively. Our understanding of bonding is least for the middle range from 1 to 100 mm, the size in the hierarchy that is widely considered to be the most important for stability of soil structure (e.g., Edwards and Bremner, 1967). Kay (1990) defines form, stability, and resiliency as the characteristics of structure. Resilience is the ability to recover from adverse changes. Dexter (1988) defines structure as ‘‘the spatial heterogeneity of the different components or properties of soil.’’ This definition accommodates the different aspects of the many size scales. ‘‘. . . spatial heterogeneity ¼ spatial variability ¼ structure.’’ Bypass flow in large macropores (cracks), a consequence of heterogeneity, is important in soil management (e.g., Edwards et al., 1988). The SSSA 1997 Revision of the Glossary of Terms defines soil structure in an older concept as ‘‘the combination or arrangement of primary soil particles into secondary units or peds, ‘‘. . . characterized on the basis of size, shape, and grade. . .’’ Nieder (2004) has defined soil on the basis of voids and functions, ‘‘Soil is . . . (the) life sustaining, biologically active, porous and structured medium at the Earth’s surface formed by . . . (inorganic and organic) particles, air and living organisms.’’ A stable soil structure, resulting in a diversity of habitats, is necessary to allow important soil functions to proceed. The ecological approach to soil science has led us to think in terms of the functions that soil performs, both in unmodified (natural) ecosystems and in agro ecosystems. This approach to soil structure can be appreciated from a paper submitted to a workshop, ‘‘Soil Structure: Key to Soil Functions’’ (Dexter, 2001). These functions, biochemical and geochemical, include recycling of carbon and nutrients, partitioning of water at the soil surface, physical and chemical buffering, storage and release of water and nutrients, energy partitioning at the surface, and a physical base for plants, animals, and man-made structures. All depend on a stable structure or soil architecture within which they can proceed. The different sizes, shapes, and connections of pores allow these functions.

6.2. The char story A recently described management material, char, mimics the desirable features of an aggregate with good structure, stability, high porosity, and active surfaces for water and nutrient retention. It is produced from organic materials burned under exposure to limited oxygen to produce charred material. This is then mixed into soil. Experimental results have

266

Benno P. Warkentin

shown large increases in yield, especially in infertile tropical soils (e.g., Glaser et al., 2002; Lehmann and Rondon, 2006). The ‘‘char’’ story goes back to the Amazon region, where well-defined small areas of dark soil, one or more feet thick were identified, in which plants grew more luxuriantly than in surrounding soil. These areas, terra preta, became objects of curiosity for soil scientists and naturalists studying soils in South America. What was their origin? General agreement formed around the idea that these were the middens of settlements and fields in which the inhabitants had spread charred organic materials beginning about seven millennia ago.

6.3. Tillage after 1980 The understanding of tillage has changed from tillage required to create tilth (structure) to a concern that tillage destroys structure. Warkentin (2001) has argued that the net beneficial effects of tillage were lost when machine power became generally available and led to excessive tillage and compaction. With human or animal power, cultivation probably improved the seedbed. Destruction of structure has been described as a major cause of land degradation, but has not been taken seriously until recently. Decreased infiltration rate is the most visible result. The interest in decreased tillage (no-till or direct drilling) in cropping systems has kept alive the attempts to predict changes in soil structure from different tillage practices. The desire to understand what happens to soils in the change from intensive cultivation to no-till has led to more studies on soil structure. Success in understanding and predicting is still eluding us. ‘‘It is not possible to predict the resulting soil condition from any given tillage operation’’ (Dexter, 1988). ‘‘None of the models for predicting the changes in soil structure of the tilled layer at present available can be used to forecast the effects of a cropping system on these changes over space and time, especially when the soil is ploughed’’ (Roger-Estrada et al., 2000). Two main approaches to predicting structure resulting from tillage have been to model the processes or to measure a sufficiently large number of physical properties of soils to gain an understanding of the changes and hence to be able to predict them; both approaches have their adherents. Numerous papers in the last 35 years have measured changes under different tillage systems. The comprehensive studies such as the 15-year study by De Leenheer (1977) and colleagues on structure changes by compaction on high-silt soils in Belgium found that significant differences between soil types and treatments could not generally be established because the variability was too high. Part of this was due to year-to-year variability. The study in Germany, also discussed in Section 5.5 (Hartge and Stewart, 1995),

Soil Structure

267

measuring various properties in different research laboratories, again concluded that we do not have a sufficiently complete understanding of aggregate stability or of the processes of aggregation to use as a guide in measurements for evaluating tillage. Soil compaction, mostly from heavy machinery used in intensive cropping systems, is now one of the major global causes of soil degradation on productive agricultural land under intensive management (Horn et al., 2000). While the compaction problem has been recognized from the early preRenaissance writings, work on understanding and combating compaction has been a major line of research on soil structure only in the past 30 years. Compaction destroys one level of structure, the transmission pores, and increases a group of smaller size pores (Dexter, 1988). Recent work, applying concepts of soil mechanics, has contributed to understanding the forces between aggregates at different size levels (e.g., Horn et al., 2000).

6.4. The developing story of architecture (or structure) as habitat The story of how arrangement of soil solids, resulting in arrangement of the voids that give soil its unique properties of habitat for soil functions, is a story stretching over millennia from the ‘‘first soil scientists’’ who observed soil behavior to today’s scientists with modern measuring equipment in field and laboratory. They have all tried to understand and predict soil behavior, and draw conclusions about soil management. The concepts have changed as our soils knowledge increased. Structure as habitat for biota has revitalized the study of soil structure. These studies are led largely by soil biologists and landscape ecologists. The review paper by Six et al. (2004) has a good discussion of the background of this approach. The focus for tilth is now on natural processes, leading to association (combination) of soil particles rather than on fragmentation processes resulting from tillage. Soil biology has replaced soil chemistry at the frontier of soil research, and in non-soils journal articles. We now see a geobiochemical approach, with special knowledge coming from all parts of soil science. This chapter ends where the new soil scientists, the geobiochemical earth scientists, the ‘‘geodermists,’’ with new concepts of soil architecture as biological habitat and as the zone of essential soil functions, and with new tools and new specializations, face current issues of soil management. Tilth has a new framework. This story will be played out in the next decades, with active research on concepts and additions to knowledge in papers published in a range of earth science journals.

268

Benno P. Warkentin

7. Summary Knowledge about soils was developed, in its early history, by practitioners who worked with soils. This knowledge was related to the social conditions of the times and to society’s needs. The more practical Roman contributions succeeded the Greek concepts, based more on philosophy. The knowledge increased, largely on a technical base, in the last 200 years with the participation of soil scientists. Soil structure of surface horizons was perceived for millennia as tilth of the seedbed and plowing to achieve it. This was the concept until about 1850 while mostly practitioners wrote about soils; then the soil scientists stepped in and defined it as aggregation. The early social and technical history of tilth was largely the history of the plow—how to plow well. With the advent of soil science laboratories, the dominant concern soon became measurement of static properties expected to be related to tilth and stability of structure: aggregate-size distribution, bulk density to calculate porosity, grain-size distribution, shape and size of aggregates, and stability of aggregates. It was a concept based on the solid particles. The importance of organic matter and clay content in soil structure was generally recognized. All these properties were related to changes in soil from cultivation, and cultivation to tilth and yield, but the relationships were not close enough to allow prediction of tilth from measured properties. By 1950, soil structure research was in the doldrums. Two major changes occurred after 1970 to reinvigorate it. The concept of hierarchical arrangement of different aggregate sizes, and the bonds responsible for stability, drew attention to different void sizes and to the soil functions of each. And then the recognition of the unique role of soil in ecosystems led to considering soil structure as defining the habitat for soil biophysicochemical functions such as decomposition, water routing, etc. Structure now became a concept centered on voids, and the term soil architecture became more appropriate; the spaces and surfaces of the spaces were more important that the solids of walls and roof. This is the venue of studies begun in the last decades. Today’s soil structure research is again productive in concepts and applications.

Glossary Aggregates Hierarchal associations of groups of solids and the enclosed voids of different sizes.

Soil Structure

269

Soil architecture The arrangement of voids and solids in soils. Used as equivalent to soil structure. Preferred because of its emphasis on spaces for soil functions. Tilth The ‘‘condition’’ of a soil in relation to the habitat provided for seed germination and plant growth. A soil can be in ‘‘good tilth.’’ Soil compaction Decrease in void volume of a soil. Pores The spaces between solid soil particles. Used as equivalent to voids. Soil structure Arrangement of solids and pores in soils. Usually includes size, shape, and arrangement of solids and size, shape, arrangement, and connections of voids. Voids The spaces between soil particles. Used as equivalent to pores. The term voids is preferred because it has no connotation of shape. Bulk density Mass of dry soil per unit volume.

ACKNOWLEDGMENTS Some of the early papers were consulted in the USDA Agricultural Library, in the British Museum, and in the Rothamsted Library where the librarian kindly pulled early papers and books out of storage. Many ideas came from discussions with colleagues. My thanks also to Tracy Mitzel, Department of Crop and Soil Science, Oregon State University for preparing the chapter. Some of the concepts were abstracted from an earlier draft of this chapter and published as the second section of chapter 13 on the changing understanding of physical properties of soils, in Footprints in the Soil, B. P. Warkentin (ed.) Elsevier, Amsterdam, 2006.

REFERENCES Amundson, R. (2006). Philosophical developments in pedology in the United States. In ‘‘Footprints in the Soil: People and Ideas in Soil History’’ (B. P. Warkentin, Ed.), pp. 149–165. Elsevier, Amsterdam. Atterberg, A. (1911). Die plastizita¨t der Tone. Int. Mitt. fu¨r Bodenkunde 1, 7–9. Babel, U., Benecke, P., Hartge, K. H., Horn, R., and Wiechmann, H. (1995). Determination of soil structure at various scales. In ‘‘Soil Structure—Its Development and Function’’ (K. H. Hartge and B. A. Stewart, Eds.), CRC Press, Inc., Boca Raton, FL. Baver, L. D. (1940). ‘‘Soil Physics.’’ Wiley, New York. [The 2nd (1948) and 3rd (1956) editions contained only a few additional paragraphs on tilth and soil structure.]. Bradfield, R. (1936). The value and limitations of calcium in soil structure. Am. Soil Surv. Assoc. Bull. VII, 31–32. Bradfield, R. (1950). Soil structure. Trans. Int. Cong. Soil Sci. Amsterdam II, 1–11. Brewer, R. (1964). ‘‘Fabric and Mineral Analysis of Soils.’’ Wiley, NY. Burkett, H. (1907). ‘‘Soils.’’ Orange Judd, NY.

270

Benno P. Warkentin

Cleme`nt-Mullet, J.-J. (1864). Le Livre de l’Agriculture. Translation into French of Ibn-AlAwam manuscript. Libraire A. Franck 2 volumes, Paris. Cox, C. (2007). Improving conservation planning for extreme events. J. Soil Water Conserv. 62, 75A–77A. Crome, G. E. W. (1812). ‘‘Der Boden und seine Verha¨ltnisz zu den Gewa¨chsen.’’ Gebru¨der Hahn, Hannover. Davy, H. (1813). ‘‘Elements of Agricultural Chemistry.’’ Longman et al., London. De Leenheer, L. (1977). Structure et Fertilite´ de Sols Limoneux sur Fermes Mecanisees. Mededelingen van de Faculteit der Landbouwwetenschappen. Rijksuniversiteit Gent. 2 volumes. Dexter, A. R. (1988). Advances in characterization of soil structure. Soil Till. Res. 11, 199–238. Dexter, A. R. (2001). Soil structure: Key to soil functions. Paper title for workshop on soil structure. Edwards, A. P., and Bremner, J. M. (1967). Microaggregates in soils. J. Soil Sci. 18, 64–73. Edwards, W. M., Norton, L. D., and Redmond, C. E. (1988). Characterizing macropores that affect infiltration in untilled soil. Soil Sci. Soc. Am. J. 52, 483–487. Ehrenberg, M. (1989). ‘‘Women in Prehistory.’’ British Museum Press, London, UK. Emerson, P. (1930). ‘‘Soil Technology.’’ Macmillan, New York. Ewing, S. (1998). Australia’s community ‘‘landcare’’ movement: A tale of great expectations. Adv. Geoecol. 31, 953–956. FAO (2006). World Reference Base for Soil Resources, Report 103, Rome. Faulkner, E. H. (1943). ‘‘Plowman’s Folly.’’ University of Oklahoma Press. Fitzherbert, E. H. (1534). ‘‘The Boke of Husbandry.’’ Thomas Berthelet, London, 90 sheets. Frese, H. (1960). Haben wir ein Konzept fu¨r eine Wissenschaft von der Bodenbearbeitung? ‘‘7th International Congress Soil Science,’’ Vol. 6, pp. 54–66. Madison, WI. Glaser, B., Lehmann, J., and Zech, W. (2002). Ameliorating physical and chemical properties of highly weathered soils in the tropics with charcoal—A review. Biol. Fertil. Soils 35, 219–230. Groff, S. (2006). Communication on a World Congress of Soil Science field trip in Pennsylvania, July 2006. Hall, A. D. (1931). ‘‘The Soil.’’ Dutton, NY (the 1st ed. was published in 1912). Hartge, K. H., and Stewart, B. A., (Eds.), (1995). In ‘‘Soil Structure: Its Development and Function’’ Adv. in Soil Sci. CRC Press. Hedrick, R. M., and Mowry, D. T. (1952). Effect of synthetic polyelectrolytes on aggregation, aeration, and water relationships of soil. Soil Sci. 73, 427–441. Heńin, S. (1935). Soil Structure. Ann. Agron 5, 44–50 (seen in Russell, E. W. 1938). Heńin, S. (1976). Cours de Physique du Sol. Vol. 1.—Texture, Structure-Aeration, Orstom and Editext, Brussels. Hilgard, E. W. (1879). On the flocculation of particles and its physical and chemical bearing. Am. J. Sci. 17, 205–214 (seen in Baver, 1940). Horn, R., van den Akker, J. J. M., and Arvidsson, J. (Eds.), (2000). In ‘‘Subsoil Compaction—Distribution, Processes and Consequences’’ Adv. Geoecology No. 32. Catena Verlag, Reiskirchen, Germany. Hudson, W. (1995). ‘‘Soil Conservation.’’, 3rd ed. Iowa State Univ. Press, Ames, IA. Karlen, D. L., Erbach, D. C., Kaspar, T. C., Colvin, T. S., Berry, E. C., and Timmons, D. R. (1990). Soil tilth: A review of past perceptions and future needs. Soil Sci. Soc. Am. J. 54, 153–161. Kay, B. D. (1990). Rates of change of soil structure under different cropping systems. Adv. Soil Sci. 12, 1–52. Keen, B. A. (1931). ‘‘The Physical Properties of the Soil.’’ Longmans Green, London. King, F. H. (1897). ‘‘The Soil.’’ Macmillan, New York. King, F. H. (1907). ‘‘A Textbook on the Physics of Agriculture.’’ 4th ed., published by the author. [King published several editions of this book, but they were printings without changes rather than editions in the modern sense.]

Soil Structure

271

Kubiena, W. L. (1938). ‘‘Micropedology.’’ Collegiate Press, Ames, IA. Lamond, E. (1890). ‘‘Walter of Henley’s Husbandry.’’ Translation of ‘‘Le Dite de Hosebondrie.’’ Longmans Green, London. Laotse (Li, Erh, 6th century BPE). Seen on a Plaque at Taliesin West, Phoenix, AZ. Larson, W.E., et al., (Eds.), (1989). In ‘‘Mechanics and Related Processes in Structured Agricultural Soils.’’ Kluwer Academic Publishers, Dordrecht. Lee, D. (1849). The philosophy of tillage. Trans. N.Y. Agric. Soc. 8, 342–358 (seen in Sewell, 1919). Lehmann, J., and Rondon, M. (2006). Bio-char soil management on highly weathered soils in the humid tropics. In ‘‘Biological Approaches to Sustainable Soil Systems’’ (N. Uphoff, Ed.), pp. 517–530. CRC Press, Boca Raton, FL. McCarthy, J., and McKay, L. (2004). Colloid transport in the subsurface. Vadose Zone J. 3, 326. Mitchell, J. K. (1976). ‘‘Fundamentals of Soil Behavior.’’ Wiley, New York. Morgan, R. P. C. (2005). ‘‘Soil Erosion and Conservation.’’ 3rd ed., Blackwell Science, Oxford. Mortimer, J. (1707). ‘‘The Whole Art of Husbandry.’’ Mortlock at the Phoenix, London. ( J. M. Esq. FRS, is listed as the author on the title page.). Nieder, R. (2004). Future of soil science. In ‘‘The Future of Soil Science’’ (A. E. Hartemink, Ed.), pp. 97–100. International Union of Soil Sciences, Wageningen, Netherlands. Nikiforoff, C. C. (1941). Morphological classification of soil structure. Soil Sci. 52, 193–212. Nye, M. J. (1996). ‘‘Before Big Science.’’ Twayne, New York. Oades, J. M. (1987). Association of colloidal materials in soils. Trans. 13th ISSS Congress, Hamburg (1986), Vol. 6, pp. 660–674. Oades, J. M., and Waters, A. G. (1988). Aggregate hierarchy in soils. Aust. J. Soil Res. 29, 815–828. Olson, L. (1943). Columella and soil science. Agric. Hist. 17, 65–72. Olson, L. (1944a). Vergil and conservation. Agric. Hist. 18, 153–155. Olson, L. (1944b). Cato’s views on the farmer’s obligation to the land. Agric. Hist. 19, 129–132. Olson, L., and Eddy, H. L. (1943). Ibn-Al-Awam—A soil scientist of Moorish Spain. Geogr. Rev. 33, 100–110. Pigulevsky, M. C. (1932). Soil as an object for work in agriculture. Trans. 2nd Int. Cong. Soil Sci. 1, 82–93, Moscow (seen in Russell, E. W. 1938). Quastel, J. H. (1952). Influence of organic matter on aeration and structure of soil. Soil Sci. 73, 419–426. Quirk, J. P. (1994). Interparticle forces: A basis for the interpretation of soil physical behavior. Adv. Agron. 53, 121–183. Roger-Estrada, J., Richard, G., Boizard, H., Boiffin, J., Caneill, J., and Manichon, H. (2000). Modeling structural changes in tilled topsoil over time as a function of cropping systems. Eur. J. Soil Sci. 51, 455–474. Ro¨mkens, M. J. M. (1985). The soil erodibility factor: A perspective. In ‘‘Soil Erosion and Conservation’’ (S. A. El-Swaify, W. C. Moldenhauer, and A. Lo Eds.), pp. 445–461. Soil Conservation Society of America, Ankeny, IA. Chapter 42. Russell, E. J. (1912). ‘‘Soil Conditions and Plant Growth.’’ Longmans Green, London. Seven editions to 1932. Russell, E. J., Sir (1957). ‘‘The World of the Soil.’’ Fontana Library, London. Russell, E. W. (1933). The binding forces between clay particles in a soil crumb. Trans. 3rd Int. Congr. Soil Sci. 1, 26–29. Russell, E. W. (1938). Soil structure. Imp. Bur. Soil Sci., Tech. Comm. 37, 40. Schu¨bler, G. (1838). ‘‘Grundsa¨tze der Agrikultur Chemie.’’ 2nd ed., Baumga¨rtner, Leipzig. Sewell, M. C. (1919). Tillage: A review of the literature. J. Am. Soc. Agron. 11, 269–290. Shanmuganathan, R. T., and Oades, J. M. (1982). Modification of soil physical properties by manipulating the net surface charge on colloids through addition of Fe (III) polycations. J. Soil Sci. 33, 451–465.

272

Benno P. Warkentin

Shaw, B.T., (Ed.), (1952). In ‘‘Soil Physical Conditions and Plant Growth’’ Agronomy Monograph No. 2. Academic Press, New York. Showers, K. B. (2006). Soil erosion and conservation: An international history and a cautionary tale. In ‘‘Footprints in the Soil: People and Ideas in Soil History’’ (B. P. Warkentin, Ed.), pp. 369–406. Elsevier, Amsterdam. Six, J., Bossuyt, H., Degryze, S., and Dener, K. (2004). A history of research on the link between (micro) aggregates, soil biota, and soil organic matter dynamics. Soil Till. Res. 79, 7–31. Smart, P. (1975). Soil microstructure. Soil Sci. 119, 385–393. Sojka, R. E., and Lentz, R. D. (1994). Time for yet another look at soil conditioners. Soil Sci. 158, 233–234. Sticher, H. (2001). Der Alpenraum, eine Wiege der Bodenokologie 1926–2001. In ‘‘Deutsche Bodenkundliche Gesellschaft 1926–2001’’ (H. P. Blume, Ed.), Vol. 97, pp. 349–366. Chapter 14. Tessier, D. (1991). Behaviour and microstructure of clay minerals. In ‘‘Soil Colloids and their Associations in Micro-aggregates’’ (M. F. DeBoodt, M. H. B. Hayes, and A. Herbillon, Eds.), Plenum Press, New York. Thompson, D. W. (1961). ‘‘On Shape and Form.’’ Cambridge University Press, New York. Tiulin, A. F. (1928). Seen in Baver, 1948. ‘‘Soil Physics’’, 2nd ed. p. 173. Wiley. Tull, J. (1731). ‘‘The New Horse-Houghing Husbandry.’’ Published by the Author, London. Tull, J. (1751). ‘‘Horse-hoeing Husbandry.’’ 3rd ed., Miller, London. Tuller, M., and Or, D. (2002). Unsaturated hydraulic conductivity of structured porous media. A review. Vadose Zone J. 1, 14–37. USDA. (1951). Soil Survey Manual, Handbook 18. USDA. (1999). Soil Taxonomy, Handbook 436, 2nd ed. Van Bavel, C. H. M. (1950). Mean weight diameter of soil aggregates as a statistical index of aggregation. Soil Sci. Soc. Am. Proc. 14, 20–23. von Humboldt, A. (1850). ‘‘Kosmos.’’ Harper, New York (Translated from German). von Leibig, J. (1841). ‘‘Organic Chemistry and its Application to Agriculture and Physiology.’’ Taylor and Walton, London. Warington, R. (1900). ‘‘Lectures on Some of the Physical Properties of Soil.’’ Oxford, Claredon. Warkentin, B. P. (1991). Clay soil structure related to soil management. Trop. Agric. 59, 82–91. Warkentin, B. P. (1999). The return of the other soil scientists. Can. J. Soil Sci. 79, 1–4. Warkentin, B. P. (2000). Tillage for soil fertility before fertilizers. Can. J. Soil Sci. 80, 391–393. Warkentin, B. P. (2001). The tillage effect in sustaining soil functions. J. Plant Nutr. Soil Sci. 164, 1–6. Williams, V. R. (1935). Seen in Baver (1948). Winiwarter, V. (2006). Soil scientists in ancient Rome. In ‘‘Footprints in the Soil: People and Ideas in Soil History’’ (B. P. Warkentin, Ed.), pp. 3–16. Elsevier, Amsterdam. Wollny, E. (1878–1898). Forsch. Gebiet Agr. Phys. Journal with research articles, abstracts and reviews. Yoder, R. E. (1937). The significance of soil structure in relation to the tilth problem. Soil Sci. Soc. Am. Proc. 2, 1–23. Yong, R. N., and Warkentin, B. P. (1966). ‘‘Introduction to Soil Behavior.’’ Macmillan, New York. Yong, R. N., and Warkentin, B. P. (1975). ‘‘Soil Properties and Behaviour.’’ Elsevier, Amsterdam. Zakharov, S. A. (1927). Seen in Russell (1938). Zwerman, P. J., and Blake, C. R. (1958). An index to Forsch. Gebiet Agr. Phys. Soil Sci. 86, 350–354.

C H A P T E R

S E V E N

Long-Term Cereal-Based Rotation Trials in the Mediterranean Region: Implications for Cropping Sustainability John Ryan,* Murari Singh,* and Mustafa Pala* Contents 274 276 276 277 279 279 282 284 285 286 288 290 290 291 294 294 298 298 300 303 305 306 307 308

1. Introduction 2. Perspective on Global Long-Term Cropping System Trials 2.1. The concept of sustainable cropping 2.2. Relevance to current conditions 3. The Mediterranean Region 3.1. Climate and environmental conditions 3.2. Soil and water resources 3.3. Cropping systems and rotations 4. Overview of Long-Term Trials in the Mediterranean Region 4.1. North Africa 4.2. West Asia 5. Cropping System Trials at ICARDA in Syria 5.1. Rationale for long-term experimentation 5.2. Multiyear rotation and tillage trials 6. Synthesis of Long-Term Cropping System Trials at ICARDA 6.1. Crop yield trends: Cereal, food, and forage legumes 6.2. Quality components of grain and straw 6.3. Soil mineral nitrogen and nitrogen cycling 6.4. Potential benefits for soil quality 6.5. Crop water use and water use efficiency 6.6. Phosphorus dynamics in arable soils and rangeland 6.7. Economic assessment 7. Statistical Trends in Crop Rotations 7.1. Estimation of time trends

* International Center for Agricultural Research in the Dry Areas (ICARDA), Aleppo, Syria Advances in Agronomy, Volume 97 ISSN 0065-2113, DOI: 10.1016/S0065-2113(07)00007-7

#


273

274

John Ryan et al.

8. Lessons Learned 9. Future Strategies Acknowledgments References

309 311 312 312

With increasing global populations particularly in developing countries, and a limited or even shrinking supply of arable land, the challenge to agriculture is to meet the world’s food and fiber needs without reducing the capacity of the resource base (soil and water) to enable guaranteed production for posterity and also to accommodate society’s environmental and energy concerns. The issue of production sustainability is all the more acute in semi-arid and arid regions of the world where drought and related biophysical factors create a fragile and uncertain environment for production. In the West, mainly in temperate regions, long-term agronomic trials have been invaluable in identifying new technologies and crop management systems that have contributed to enhanced crop output that is sustainable from the biological, environmental, and economical standpoints. Many of these trials continue to guide cropping trends into the foreseeable future. The Mediterranean region has served climatic constraints to its agriculture and despite being cultivated for millennia, it is largely food deficient. Yet long-term cropping experiments that could direct agricultural production in a sustainable manner are relatively rare, and even most of such trials are of recent vintage. This review offers a background perspective on factors related to crop, production, and subsequently examines the various multiyear cropping system/ tillage trials in countries of North Africa and West Asia that border the Mediterranean. Special emphasis is given to the wide range of trials conducted in Syria by the International Center for Agriculture Research in the Dry Areas across a range of rainfall zones that are typical of the region as a whole. The goal of many trials was to identify cropping systems as a substitute for fallow and continuous cereal cropping with implications for improved water-use efficiency (WUE), crop quality, soil quality, and fertilizer use. Lessons learned from the trials are highlighted as well as future directions for cropping systems research.

1. Introduction Since the dawn of man’s existence, the quest to acquire enough food and clothing has been a constant challenge in order to guarantee survival of the human species. The turning point for mankind took place about 10–12 thousand years ago when hunting/gathering slowly gave way to nomadic animal herding and settled agriculture with cultivated crops. Thus began the concept of land use (Harlan, 1992). With permanent settlements, which developed in various parts of the world, notably in the Mediterranean region, that is, the ‘‘Fertile Crescent of the Near East,’’ the beginning of human civilization was set in motion. A more stable food supply from cultivated crops in well-watered river valley ensured better nutrition and consequently high survival rates and rapid population growth. World

Long-Term Cereal-Based Rotation Trials in the Mediterranean Region

275

populations grew exponentially to such an extent that concerns were raised that mankind would outgrow the very resources that sustained them. Despite continued population growth in the 19th and 20th centuries, the gloomy predictions of Malthus were downplayed as advances in agricultural technology, combined with exploitation of new lands for cultivation, produced food and fiber in abundance, indeed surpluses, especially in developed countries (Borlaug, 2003). However, in the modern era, actual and predicted growth rates in developing countries leave little grounds for complacency as far as balancing the food demand–supply equation is concerned. As we face a world of over 6 billion people, current global food stocks are at historically low levels, with dim prospects for Africa, and some countries of Asia and Latin America, getting out of the food-deficit hole. As society grapples with this monumental concern, with its implication for the environment and maintenance of global peace, the major question is ‘‘Can the world meets the food and fiber needs of its burgeoning populations without damaging its resource base?’’ In short, production sustainability that is compatible with soil and environmental presentation is a priority for mankind (Lal, 2001). Of major agro-ecosystems of the world, particular concern in relation to food production system and its impact on the land-resource base is focused on semi-arid areas of the world where mostly rainfed cropping is practiced within the prevailing climatic constraints (Steiner et al., 1988). Much of the world’s drylands are in developing countries and are characterized by low crop productivity, limited irrigation potential, weak administrative and agricultural research infrastructures, and unrelenting pressure on natural resources due to escalating growth in human and animal populations (Ryan, 2002a,b). Despite the attention that has been given to irrigated and the successes of the Green Revolution, dry areas of the world supply over 60% of the world’s food (Stewart et al., 2006). As water scarcity has now been fully realized, along with urban industrial and recreational demands for water, the expansion in irrigational well inevitably stall (Rosegrant et al., 2002) and with it comes renewed crop production demands on semi-arid lands. Despite the challenges inherent in dryland agriculture and the concerns about system sustainability (Stewart and Robinson, 1997), there are grounds for optimism that, with available soil and crop management technologies and improvements in water conservation, crop production in dry areas can be improved and sustainably managed (Lal, 1987; Peterson et al., 2006; Ryan, 2002a,b; Steiner et al., 1988). Indeed, given the global extent of dry areas, dryland agriculture may potentially contribute to carbon sequestration and thus mitigate the effects of climate change (Lal, 2002). As agriculture, whether involving crops or animals, influences its environment (FAO, 2002), future agricultural practices will have to be concerned about the ways in which it can do so either positively or negatively.

276

John Ryan et al.

Nowhere are the concerns about the practice of agriculture in rainfalllimited areas more acute than in the lands bordering the Mediterranean, especially to the south and east. Though the region is where settled agriculture began, especially in the cradle of civilization in the Fertile Crescent (Dregne, 2006) and the center of origin of many of the worlds food crops, notably cereals and pulses (Harlan, 1992), and the place where many of our farm animals (cattle, sheep, and goats) were domesticated, the region is largely a food-deficit one. Despite its antiquity, agriculture in the region is rapidly changing with more intensive land use, driven by population pressure (Ryan, 2002a,b). Given the drought-stressed environment of the region (Cooper et al., 1987; Smith and Harris, 1981) and its fragile soils (Matar et al., 1992), the implications of these changes on soil, water, and cropping systems call for a broad assessment as to their sustainability as the only way to reliably monitor long-term cropping practices, particularly with the introduction of new technologies and crop varieties, is through the medium of long-term cropping system trials. Thus, in this review, we examine the various long-term trials with emphasis on crop rotations, a practice that dates back to Roman times in the Middle East (Karlen et al., 1994). Emphasis is given to food and forage legumes as alternatives to fallow, which is disappearing, and continuous cereal cropping, which is deemed sustainable. Consideration is also given to crop quality, nutrient cycling, soil quality, and water-use efficiency (WUE). As the fundamental issue is long-term viability of agricultural practices, a brief discussion follows on sustainability and lessons that have emerged from long-term cropping system trials elsewhere.

2. Perspective on Global Long-Term Cropping System Trials Agriculture by definition is a long-term process, and one that has evolved and changed with time, and is likely to continue to do so in ways we cannot predict for the future. Sustainability of agricultural practices is a relatively recent concern from the historical perspective. It only became an issue when people were obliged to farm the same land indefinitely and maintain its productivity. The fundamental rationale for long-term agricultural experiments is to assess where particular practices are likely to endure in any agro-ecosystem. Prior to teasing out the essence of various global longterm trials, it is pertinent to consider sustainability and how it is measured.

2.1. The concept of sustainable cropping Sustainability is a concept of recent origin and one that is applicable at different levels and timescales, and indeed to different people. As anxieties developed about man’s impact on the biosphere, sustainability enshrined the


277

goal of conservation of habitats and biodiversity rather than exploitation, degradation, and pollution ( Jones, 1993). Those concerns extend to the maintenance of food security, the necessary inputs to agriculture, and human equity. Agricultural sustainability must accommodate the array of interactions, whether positive or negative, between agricultural activities and the environment and society as a whole. Jones (1993) expanded on the different perspectives of sustainability depending on the stage of economic development. In developed countries, sustainability involves crop diversification and less dependence on nutrients and pesticides, while in developing countries the food production is the priority with lesser concerns for preserving the natural resource base. In both circumstances, the notion of sustainability extended to both internal and external resources. In short, the capacity of agriculture to maintain or even increase production over time without any damage to that potential or the environment is of concern for all mankind. As the sustainability impinges on all sectors of society, the inevitable question raised is how to objectively measure the complex changes involved with sustainability. Detailed data of statistical validity are vital is such an endeavor (Barnett et al., 1994). Long-term agro-ecosystem experiments have been seen as the logical approach to evaluating long-term change in cropping systems. They provide an invaluable resource for assessing biological, biogeochemical, and environmental dimensions of agricultural sustainability, as well as serving as a basic for predicting the impact of climate change (Rasmussen et al., 1998). Notwithstanding the complexities of long-term trial evaluation, the most convincing evidence of the sustainability of any agricultural system comes from a long-term experiment with positive results (Steiner, 1995). Though the timeframe for agricultural sustainability assessment needs millennia (Sander and Eash, 1991), most global long-term experiments, with a few notable exceptions, fall within the human lifespan. These agroecosystem experiments constitute the largest temporal and spatial database presently available for determining the impacts of ecosystem change. Some brief comments on these trials as a background to considering agronomic systems trial in the Mediterranean region are given below.

2.2. Relevance to current conditions Most long-term agronomic trials have been established in stable developed countries, with few in tropical climates or developing countries (Rasmussen et al., 1998), and none approaching the duration of conventional ‘‘longterm’’ trials. The standards by which such trials are measured are those established in Rothamsted ( Jenkinson, 1991; Johnston, 1997, Johnston and Powlson, 1994). It is of interest to note that these trials were not initiated as long-term trials that they survive today and are of interest to many scientific disciplines even though the problems these trials set out to solve, mainly

278

John Ryan et al.

crop nutrient requirements, have long been solved. Because the trials were well designed and carefully managed, the Rothamsted trials survived despite the considerable financial and labor costs involved, while many others failed. Of relevance to our consideration of Middle Eastern trials is the fact that cereal monoculture was sustainable in the trials provided organic or inorganic nutrients were used, but they also provided information on issues not even thought of at the beginning, that is, nitrogen (N) and sulfur (S) cycling and atmospheric contributions of nutrients, cadmium in P fertilizers, and soil organic matter (SOM) dynamics, all issues related to current environmental concerns. The message that long-term trials may yield new information not previously considered as time progresses is an important consideration related to any well-planned and well-designed long-term trial. As we reflected on the Middle Eastern trials, it is worth recalling the broad objectives of long-term experimentation as outlined by Johnston and Powlson (1994). These are: (1) test the sustainability of a particular cropping system over a long time span and determine what changes are needed to enhance productivity and maintain sustainability, (2) generate data of value to farmers to improve cropping systems, (3) provide a means for further scientific research on soil and plant processes which control soil fertility and crop production, (4) allow a realistic assessment of nonagricultural anthropogenic activities on soil fertility and crop quality, and (5) produce datasets for development of mathematical models to predict the likely effects of management practices and climate change on soils and their productive capacity. An overview of the historic long-term trials in the United States, for example, ‘‘Morrow Plots’’ in Illinois (1876), ‘‘Sanborn Field’’ in Missouri (1888), ‘‘Magrudar Plots’’ in Oklahoma (1892), and Alabama’s ‘‘Old Rotation’’ (1896) have messages of relevance to our study (Mitchell et al., 1991). Most important among these conclusions is that crop rotations and attention to established fertility practices, which may or may not include legumes and manuring, are essential to maintaining high and sustained production. While the most obvious concerns of applied research and farmers is maintenance of crop yields and providing cropping options and yield responses to fertilizers, particularly N, most of the world’s long-term trials provide invaluable information on soil quality (Reeves, 1997). The long-term information on SOM has direct implications for maintenance of soil structure and is fundamental to conservation of agriculture and the shift in cultivation practices from conventional tillage to minimum tillage and no-tillage. These trials also provided a means of assessing the effect of soil management on carbon sequestration, an issue of increasing importance in the growing public debate on climate change and how it is impacted by agriculture.


279

3. The Mediterranean Region Areas of the world with a typical Mediterranean climate occur principally in the lands surrounding the Mediterranean Sea and in limited areas of other continents, for example, California in the North America, Chile in South America, Southwestern Australia, and at the southern fringes of South Africa. Because of its comparatively lesser developed economic state than the northern shores of the Mediterranean, we focus our attention in this review on the southern and eastern Mediterranean, that is, West Asia and North Africa (WANA) (Fig. 1). Therefore, it is pertinent to outline climatic and soil conditions that influence cropping in the WANA region.

3.1. Climate and environmental conditions The Mediterranean region has a unique climate, mainly a cool moist season followed by a hot rainless one, and has been described by various authors (Cooper et al., 1987; Kassam, 1981). Typically, the cropping season commences in the fall (October–November) with the onset of the first rains and a decrease in temperature and evapotranspiration (ET). Rainfall generally peaks in the January–February period which coincides with minimum temperatures. Rainfall falls off in April–May which coincides with maximum crop growth prior to May–June harvest. Typical weather features, that is, temperature, rainfall, and ET, are illustrated in Fig. 2. Terminal drought invariably occurs at the end of the season in April–May. The key to coolseason winter cropping in Mediterranean environments is an excess of soil moisture from rainfall over evaporation demand at a time when temperature conditions permit crop growth. Later as evaporative demand exceeds rainfall, growth depends on residual soil moisture. Rainfed or dryland crop production, which dominates agriculture in the Mediterranean region, takes place against a background of limited rainfall that is variable in both time and space. Rainfall is generally highest near the sea and in high-elevation areas. The rainfall gradient decreases inland and eventually rainfed cropping gives way to vast expanses of rangeland or steppe and deserts in some cases (Kassam, 1981). The discrepancies were clearly illustrated by Harris (1995) for lowland and highland areas in North Africa and West Asia in terms of rainfall, and temperature patterns, and consequently growing conditions. In some highland areas, for example, Anatolia in Turkey and Zagros Mountains in Iran, snow is common in winter, while crop-damaging frosts can also occur at high elevations and inland where the climate is more continental. Dryland or rainfed cropping is generally practiced in the range of 200–600 mm year1. Rainfall range and such farming occupies about

Kazakhstan Uzbekistan Turkmeniztan

Turkey Tunisia Morocco Algeria

Mauritania

Libya

Cyprus Syria Iran Lebanon Iraq Jordan Kuwait Bahrain Egypt Qatar UAE Saudi Arabia Oman Eritrea

Kyrgyzstan Tajikistan

Afghanistan Pakistan

Yemen

Sudan Ethiopia Somalia

Figure 1 Map of Mediterranean region, indicating countries under the mandate of the International Center for Agricultural Research in the Dry Areas (ICARDA).

281

80

40 Rain Max-T

60

30 20

40 Min-T

10

20

Temperature (⬚C)

Rain (mm);A-Pan E (mm/day)


A-Pan E 0 0 Sept Oct Nov Dec Jan Feb Mar Apr MayJune July Aug

Figure 2 Long-term average for seasonal rainfall, A-Pan evaporation and maximum and minimum temperatures at Tel Hadya, Aleppo, Syria, approximating weather conditions in the Mediterranean region.

Indicative percentage land use

100 Highland

Deserts Native pastures and steppeland

Barley/livestok systems

Wheat-based systems

Rainfall(mm/yr) climatic zone Farming system

Horticultural crops

Irrigated agriculture

0 0

100

200

Arid Livestock farming

300

400

500

Semi-arid

600 1000 Semi-humid

Cereal farming Highland: Mixed farming

Figure 3 Schematic representation of agricultural production systems of the West Asia-North Africa (WANA) region.

125 mha in the WANA region (Kassam, 1981). There is a rainfall association between seasonal rainfall and cropping systems as illustrated in Fig. 3. Where rainfall is less than 200 mm, cropping is unreliable without irrigation. Without irrigation, such low-rainfall areas are occupied by rangeland or true deserts. As rainfall increases, barley, which is relatively drought tolerant, commonly grows in the 200–300 mm zone and is associated with livestock production, mainly sheep. This then gives way, as rainfall increases, to wheat-based system (300–500 mm) and above 500 mm where a range of field and horticultural crops/trees are grown. While rainfall is

282

John Ryan et al.

variable in any given year, the variability increases as the mean annual rainfall decreases, particularly with irregular distribution. Rainfall variability and its distribution pattern have implications for soil moisture dynamics and crop yields and underlines the need to analyze long-term weather data in terms of probability for accessing the implications of improved production practices. As the length of the growing season and rainfall totals are related (Kassam, 1981), based on degree-day requirement of crops and soil moisture availability, Smith and Harris (1981) indicated that the growing season would range from 90 to 180 days in the Mediterranean environment (Fig. 4). The well-established relationship between crop yields and rainfall in dryland agriculture has been borne out in numerous field studies at The International Center for Agricultural Research in the Dry Areas (ICARDA) and elsewhere in the Mediterranean zone. As seasonal rainfall increased at three experiment stations on a rainfall transect, cereal yields increased (Keatinge et al., 1985). Similarly, seasonal rainfall had the greatest influence on yields in on-farm trials both for barley ( Jones and Wahbi, 1992) and wheat (Pala et al., 1996). The above studies suggest that in crop rotation trials conducted over several years, rainfall would be a dominant influence on growth, which would be likely modified by the particular cereal rotation and its influence on residual soil moisture.

3.2. Soil and water resources Agriculture in the Mediterranean region is inextricably linked to soil quality and water supply, whether from rainfall or irrigation. The soils of the region show considerable variation, reflecting the cumulative influence of climate, 3.5

500

2.5

450

2.0

400

1.5

350

1.0

300

0.5

250

Rainfall (mm)

Grain yield (t ha−1)

3.0

550 Grain yield Rainfall

200 0.0 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000

Figure 4 Relationship between average yearly rainfall and overall wheat grain yield from the Cropping Systems Productivity trial at Tel Hadya, Aleppo, Syria.


283

topography, and the influence of man. The major soil orders are Inceptisols, Lithosols or shallow soils, Entisols, and Aridisols, with Vertisols being common in some areas (Kassam, 1981). Calcareous soils are dominant in the region, being derived from limestone residuum and are highly variable in terms of texture, depth, slope, and stoniness (Matar et al. 1992). While all soil properties impinge in one way or another on crop management and WUE, heavy clay soils pose a limitation on tillage operations while shallow soils have limited water-holding capacity and sloping soils promote erosion. The dominant characteristics of the soils are the presence of free calcium carbonate (CaCO3) and the low level of SOM, as both have a controlling influence on soil nutrients and their availability to plants. Levels of SOM are invariably low, being generally about 1–2% of the soil volume, with many soils in low rainfall areas being considerably less than 1% (Ryan, 1998). As a source of soil N, SOM can serve as a slow release nutrient source. However, the low equilibrium levels of SOM in semi-arid climates are conditioned by the temperature–moisture regime. Cultivation, especially with conventional tillage, exacerbates the decline of SOM in Mediterranean agriculture through CO2 emission to atmosphere (Ryan and Pala, 2007). Consequently, given the large demand of crops for N, deficiencies of N are a constant feature of Mediterranean agriculture (Harmsen, 1984) and N fertilization is invariably required on an annual basis for nonlegume crops, especially cereals (Ryan, 2004). While CaCO3 is normally a sizeable fraction of the soils in the region, its influence on soil chemical equilibria and solubility is significant, especially for P (Matar et al., 1992) and micronutrients such as iron and zinc (Rashid and Ryan, 2008). Prior to the advent of extensive chemical fertilization, P deficiency was widespread and endemic (Matar et al., 1992). However, with routine fertilizer use, available soil P levels have built up in the soil and, as a result, severe deficiencies are rare (Ryan, 2004). While Zn deficiency is less directly related to CaCO3 than P, deficiency of Zn is important in some areas, for example, Turkey (Cakmak, 1998) and Syria (Materon and Ryan 1995). Although boron (B) deficiency is influenced by both SOM and CaCO3, the incidence of deficiency can range from minimal in some countries (Khan et al., 1979) to severe in others (Rashid et al., 1994). While excessive levels of B can exist in Mediterranean soils, leading to toxicity to crops (Yau et al., 1995), the problem is related to geology rather than specific soil properties. While the above brief comments reflect the possible influence of soil properties as they relate to crop production, these influences must be seen in the context of available water to sustain crop growth, whether it comes from the environment as rain or is provided from irrigation based on groundwater or surface sources. Despite the fact that supplemental irrigation has increased in the past decade (Oweis et al., 1998), both sources are limited and are unlikely to increase substantially; groundwater levels are dropping at

284

John Ryan et al.

a level of almost 1 m year1, with insignificant re-charge, while urban areas are competing with agriculture for surface water source. Only urban waste water is likely to increase as a source for irrigation with the additional benefits of nutrients for crops (Ryan et al., 2006). With this dismal scenario for expansion of irrigated agriculture, there is little option but to work within the confines of existing rainfall conditions and endeavor to exploit dryland cropping systems to achieve overall increases in output and maximize WUE.

3.3. Cropping systems and rotations Notwithstanding the changes that have taken place in Mediterranean agriculture, descriptions of its cropping systems by Gibbon (1981) and Cooper et al. (1987) are still relevant as has been the case for millennia, the principal rainfed crops are cereals, mainly wheat and barley with small areas of rye, oats, and triticale of a local importance. While bread wheat and durum wheat dominate the more favorable rainfall zones (þ350 mm), barley tends to be grown mainly in the lower rainfall zone. Bread wheat dominates in Turkey, while durum wheat mainly grows in Arab countries for flat bread, or burghul, and couscous. Barley is grown as a livestock feed (grain and straw) and is used also as human food in Morocco and Tunisia. As demand for livestock has increased, so too has the area devoted to barley, exceeding that of wheat in Syria and Morocco. Food legumes mainly chickpea and lentils in rainfed areas, and faba bean in wetter areas or irrigated areas, are important crops of relevance to human nutrition in the Mediterranean region, and occupy about 5–10% of the area devoted to cereals. Local demand for these crops is generally met, and some countries even export them. Other crops are diverse and are mainly grown in summer on stored moisture in favorable rainfall zones; these include watermelon, cantaloupe, sesame, cotton, sunflower, sorghum, and maize. Given the importance of livestock, efforts were made to include forage legumes as an alternative or supplement to grazed cereal stubble. Despite much effort to introduce and promote the use of a self-regenerating grazed forage, medicago or medic, the attempt has not met with success (Christiansen et al., 2000) but research on vetch is more promising ( Jones and Singh, 2000a,b). Though crop rotation, involving a cereal crop after an alternate crop or fallow, has its origins in Mediterranean agriculture (Karlen et al., 1994), the practice is of enduring relevance in dryland agriculture. Crop rotation serves as a break in disease buildup and promotes the conservation of stored moisture for the subsequent cereal crop; in the case of fallow, it leads to increased available N when a legume is used. In the past, the most common rotation was cereal/fallow, a moisture-conserving strategy to ensure an acceptable yield in the cereal year. Two types of fallow are practiced in


285

the Middle East, that is, cultivated or clean fallow, which is common in West Asia, and ‘‘weedy’’ fallow which is common in North Africa (Cooper et al., 1987). In the latter case, the uncultivated fallow produces palatable weeds and volunteer cereals as a cheap livestock feed during the winter and spring months. With increasing land-use pressure, the incidence of fallow has progressively decreased to be replaced by continuous cereal cropping. In the higher rainfall areas, food legumes and summer crops are often grown in place of fallow, especially on deep soils in a 2-year rotation with cereals and, in some cases, a 3-year rotation with cereal, legumes, and summer crops. While the benefits of rotations on cereal yields were recognized by farmers, we now know that these are attributed to a savings in soil moisture, though fallow efficiency is low and added fertility in the form of N for the fallowing cereal (Karlen et al., 1994). Additional benefits in terms of SOM are also important for soil aggregation and water relations, and enhanced biological activity (Rasmussen and Collins, 1991). While the beneficial effect of crop rotations can be masked by fertilizers, Karlen et al. (1994) concluded that ‘‘no amount of chemical fertilizer or pesticide can be fully compensated for crop rotation effects.’’ The complexity of crop rotations in the Mediterranean region is accentuated by the grazing animal, especially in relation to animal droppings and the extent of grazed stubble.

4. Overview of Long-Term Trials in the Mediterranean Region Despite the antiquity of agriculture in the Middle East, being the center of origin of some of the world’s food crops, for example, cereals, pulses, and nuts, and the major site of settled agriculture (Damania et al., 1998), allied to the concerns about land use intensification in this fragile ecosystem, few long-term cropping system trials exist in the region (Steiner and Herdt, 1993). In fact only one trial in Egypt dating back almost a century (1916) meet the criteria of Rasmussen et al. (1998) of at least 20 years to qualify to be considered as a long-term trial. A number of administrative and socioeconomic factors could have contributed to the paucity of such cropping system trials. Primary among these is the general absence in most Middle Eastern countries of strong, well-advanced agricultural research and extension systems since such trials require considerable commitment of human and biophysical resources for many years. An additional factor is related to limited land availability for such extensive experiments at experimental station, which are usually relatively small (Ryan et al., 1990), or in farmers fields, where holdings are small (often

Advances in Agronomy, Volume 97

Advances in Agronomy, Volume 60

Advances in Agronomy, Volume 50

Advances in Agronomy, Volume 68

Advances in Agronomy, Volume 54

Advances in Agronomy, Volume 11

Advances in Agronomy, Volume 31

Advances in Agronomy, Volume 33

Advances in Agronomy, Volume 65

Advances in Agronomy, Volume 21

Advances in Agronomy, Volume 64

Advances in Agronomy, Volume 105

Advances in Agronomy, Volume 20

Advances in Agronomy, Volume 100

Advances in Agronomy, Volume 108

Advances in Agronomy, Volume 69

Advances in Agronomy, Volume 83

Advances in Agronomy, Volume 91

Advances in Agronomy, Volume 46

Advances in Agronomy, Volume 111

Advances in Agronomy, Volume 62

Advances in Agronomy, Volume 56

Advances in Agronomy, Volume 28

Advances in Agronomy, Volume 53

Advances in Agronomy, Volume 14

Advances in Agronomy, Volume 26

Advances in Agronomy, Volume 103

Advances in Agronomy, Volume 81

Advances in Agronomy, Volume 40

Advances in Agronomy, Volume 45

Advances in Agronomy, Volume 101

Advances in Agronomy, Volume 97

Advances in Agronomy, Volume 60

Advances in Agronomy, Volume 50

Advances in Agronomy, Volume 68

Advances in Agronomy, Volume 54

Advances in Agronomy, Volume 11

Advances in Agronomy, Volume 31

Advances in Agronomy, Volume 33

Advances in Agronomy, Volume 65

Advances in Agronomy, Volume 21

Advances in Agronomy, Volume 64

Advances in Agronomy, Volume 105

Advances in Agronomy, Volume 20

Advances in Agronomy, Volume 100

Advances in Agronomy, Volume 108

Advances in Agronomy, Volume 69

Advances in Agronomy, Volume 83

Advances in Agronomy, Volume 91

Advances in Agronomy, Volume 46

Advances in Agronomy, Volume 111

Advances in Agronomy, Volume 62

Advances in Agronomy, Volume 56

Advances in Agronomy, Volume 28

Advances in Agronomy, Volume 53

Advances in Agronomy, Volume 14

Advances in Agronomy, Volume 26

Advances in Agronomy, Volume 103

Advances in Agronomy, Volume 81

Advances in Agronomy, Volume 40

Advances in Agronomy, Volume 45

Advances in Agronomy, Volume 101

Recommend Documents