Mycorrhizal Symbiosis, Second Edition

Preface Thirteen years have elapsed since the publication of the first edition and the statement made in 1983 that it ...

Author: Sally E. Smith (Author) | David J. Read (Author)

226 downloads 1715 Views 47MB 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

Preface

Thirteen years have elapsed since the publication of the first edition and the statement made in 1983 that it was 'almost impossible for one person or even two to keep up with all the experimental work and speculation' on the subject of mycorrhizas is even more true today There has been such an enormous amount of new work that the book has been essentially rewritten and detailed reference to much early research has had to be severely reduced, although we have attempted to retain a feeling for the way the subject has developed. Again, it has been impossible to review all topics in detail and the new edition once more reflects the interests of the authors and gives our personal views of the subject. There have been at least three important changes of emphasis. Where the first edition explicitly avoided evolutionary discussions, on the grounds that insufficient information was available to make this profitable, we now believe that information from the fossil record and other new work based on molecular phylogeny of the fungi has filled in many gaps and that techniques are now available which will make this a rapidly developing area that can be addressed analytically. A second important change is the increased emphasis on the extraradical mycelium, which is of key significance in all types of mycorrhizas. Thirdly, whereas in the earlier stages of development of the subject the emphasis was on unifying ideas, there is now an increasing awareness of the diversity of structure and function, superimposed on a few basic patterns. Exposing the extent to which this diversity is important will be one of the challenges for the next phase of research. Throughout the book we have emphasized the experimental approaches that have proved useful and that may need to be applied in the future. We have retained the same general structure of the book, with four sections providing general accounts of the main types of mycorrhiza, including information on the identity of the symbionts, structure and development of the mycorrhizas formed by them, as well as their function and ecological significance. The fifth section is devoted to general themes, in which we have integrated ideas and information essential for clear understanding. These chapters build on the material presented earlier, but so that they can be read separately we have deliberately been repetitive and have also tried to improve the cross-referencing in the text. The general themes have replaced the essays in the first edition. There are several reasons for this: some material from the essays is now of central importance and is incorporated into the main chapters; some of the essays have largely served their purpose and have therefore been eliminated, so that readers interested in these topics will need to use the first edition; and not all readers (or reviewers) appreciated the essays, believing them to be too speculative. Throughout the text we

viii

Preface

have tried to indicate where future experimental research might be directed, although we realize that our personal views may have introduced some bias, which will be recognized by those who know us. Many friends and colleagues have helped us in countless ways, especially by discussions and by allowing us to use unpublished results of their work. Many have unstintingly made available original photographs, and in this context Hugues Massicotte and Diane and Jack Robertson deserve special mention. Drafts of some of the chapters have been read by Susan Barker, Gaby Delp, Lindsay Harley, Jonathan Leake, David Lewis, Francis Martin, Larry Peterson, Andrew Smith, Andy Taylor and Judy Tisdall. We are grateful for their critical and helpful comments, many of which have been incorporated into the text. However, we must take responsibility for all the ideas and views expressed whether good or bad; the errors are all ours. We have also received invaluable help in preparation of the manuscript. Kathie Stove laboured valiantly to try to attain consistency in style and presentation and corrected the worst of our grammatical errors. Jayne Young typed many drafts of chapters and produced the final manuscript, and Jan Ditchfield provided essential secretarial assistance. We also wish to thank Jennie Groom, Glyn Woods and David HoUingworth for photographic assistance and Marcus Brownlow for producing some figures. We have used much previously published material in both tables and figures. The sources are acknowledged in each case, but we wish to give special thanks to all the authors and publishers who generously allowed us to use their copyright material. We also thank The Cooperative Research Centre for Soil and Land Management, Adelaide, for providing financial support towards production of the manuscript. Last, but by no means least, our thanks and love go to the 'long-sufferers', Andrew Smith and Chris Williams, who provided invaluable moral (and immoral) support which helped to see us through the many months of largely self-centred devotion to writing. The book is dedicated to the memory of Jack Harley, whose influence upon the development of our subject was so enormous. Amongst his many contributions, two stand out as being of pre-eminent importance. Frustrated in his earlier career as an ecologist by much woolly thinking about the biology of mycorrhiza. Jack determined to subject the topic to rigorous physiological analysis. Over several decades and using as his research material the ectomycorrhizal roots of beech, he and his collaborators evaluated the basic processes whereby nutrients are exchanged between partners in the symbiosis. The second and arguably more important contribution arose out of his skill as a communicator. The Biology of Mycorrhiza, published in 1959, was the first attempt to synthesize the many and sometimes disparate strands of thought which had developed in over 100 years of research. With characteristic incisiveness he cut through much, often pedantic, debate to focus upon those questions which were in need of resolution. The work was of inestimable importance to many who, like ourselves, were struggling to take their first steps in research on this subject. These combined contributions provided impetus to a further expansion of research to the extent that by 1983 a new and even more substantial volume, the first edition of Mycorrhizal Symbiosis, was required. One of us had the privilege to collaborate in that enterprise. The influence of Jack Harley goes on. Although sadly, he has not been here to

Preface

assist us in updating the book, we have both been conscious of his legacy. Without him it is doubtful whether the subject would be in the pre-eminent position it enjoys today, increasingly recognized by physiologists and ecologists alike as being of central importance in plant and fungal biology. It is with some trepidation that we seek to emulate Jack's achievements in the production of this second edition.

INTRODUCTION

The interest in symbioses noted in the Introduction to the first edition has continued unabated in recent years. On the one hand, the importance of symbioses between different prokaryotes in the evolution of eukaryotic cells is now firmly established. On the other hand, there is increasing recognition that symbiosis at the level of more complex organisms is the rule, rather than the exception. Compared with those at the cellular level, the mutually beneficial associations between identifiably different organisms, such as those between fungus and alga in lichens, plant and fungus in most mycorrhizas, alga and coelenterate in corals, and bacteria and plant in N-fixing symbioses are of relatively recent origin, but they play exceedingly influential roles in natural ecosystems and may also be important in man-made ecosystems. The term 'symbiotismus' (symbiosis) was probably first used by Frank (1877) as a neutral term that did not imply parasitism, but was based simply on the regular coexistence of dissimilar organisms, such as is observed in lichens. De Bary (1887), who is often credited with the introduction of the term, certainly used it to signify the common life of parasite and host as well as of associations in which the organisms apparently helped each other. In time the meaning of the terms symbiosis and parasite changed. Symbiosis was used more and more for mutually beneficial associations between dissimilar organisms, and parasite and parasitism came to be almost synonymous with pathogen and pathogenesis. De Bary had pointed out that there was every conceivable gradation between the parasite that quickly destroys its victim and those that 'further and support' their partners, and in recent years biologists have come back to this view. In this book we use 'symbiosis' in the broad sense originally developed by de Bary. Mutualistic symbioses are those in which both partners can benefit from the association, although there is unresolved discussion as to what actually constitutes 'benefit', increased 'fitness' or 'aptness' for particular environmental circumstances. Most mycorrhizal symbioses are now clearly recognized, together with lichens, corals and N-fixing systems, as being common and significant representatives of the mutualistic end of the symbiotic spectrum, whereas biotrophic pathogens such as rusts and mildews are parasitic symbioses. However, as will become clear, not all of those associations that are described as mycorrhizas have been shown to be mutualistic by experimental analysis of nutritional interactions or determination of fitness. If a symbiosis is mutualistic and based on bidirectional exchange of nutrients, then the description of one partner as the 'host' seems inappropriate. In general we have adopted the terms 'plant' and 'fungus' to describe the partners in mycorrhizal symbioses, although we have retained terms such as 'host range' to describe the diversity of

2

Mycorrhizal symbiosis

plant species with which a single fungus associates. Similarly, and responding to forceful pleas by some, we have where possible substituted the (politically correct) term 'colonization' for 'infection', to avoid the suggestion of parasitic attack by the fungus on the plant. Again, where 'infection' has been used as part of an established term, as in 'infection unit', it has been retained. De Bary (1887) believed that there was some degree of common life, i.e. of symbiosis, in all or almost all examples of association between plants and fungi. The associations are often classified as biotrophic or necrotrophic, depending on whether both symbionts remain alive or the death of one was necessary before substances could be absorbed by the other. There is, however, a great range of behaviour between these two apparently clear-cut extremes, not only between different types of association but also at different times or under different environmental conditions in the same association. These variations in function, sometimes also apparent in changes in structural relationships between the symbionts, are seen in mycorrhizal associations as much as in other symbioses. One of the important features of recent research on mycorrhizas is the recognition of the considerable diversity of structure, development and function that exists even within a single mycorrhizal type. The changes that occur during the normal development of a symbiosis may be great, but until recently they have not been given much attention. Furthermore, the identity of the symbionts may have both major and minor influences on the structure (and presumably also function) of mycorrhizas formed between them. Examples include the colonization of roots by Wilcoxina, which forms ectendomycorrhizas with Pinus but ectomycorrhizas with Picea. Mycorrhizas formed by glomalean fungi can also be quite variable, the same fungus forming extensive intracellular coils and rather few arbuscules in some species of plant, and intercellular hyphae and many arbuscules in others. Many of these variations were appreciated by the early workers such as Janse (1897) and Gallaud (1905), but were forgotten as the attention of researchers was drawn to find unifying hypotheses that emphasized the structural and functional similarities of different mycorrhizal types. It is even clearer now than it was in 1983 that there is an immense diversity in what we call mycorrhizas. The description, in structural, developmental and physiological terms, of this diversity and the understanding of its importance in ecosystems are likely to be significant areas of research in the next decades. We take this opportunity to re-emphasize a point made forcibly in the Introduction to the first edition: 'It is clear that since these kinds of variation occur, much of the discussion based on classification and nomenclature is pedantic unless it is helpful in formulating clear questions which will lead to experiments from which answers will be obtained'. Although the term 'mycorrhiza' implies the association of fungi with roots, relationships called mycorrhizal associations, which are involved in the absorption of nutrients from soil, are found between hyphal fungi and the underground organs of the gametophytes of many bryophytes and pteridophytes, as well as the roots of seed plants and the sporophytes of most pteridophytes. Indeed, mycorrhizas, not roots, are the chief organs of nutrient uptake by land plants and recent work has amply confirmed the earlier observations (see Nicolson, 1975) that the earliest land plants, which had no true roots, were colonized by hyphal fungi that formed vesicles and arbuscules strikingly similar to modern vesicular-arbuscular mycorrhizas. The fossil record thus confirms previous speculation that the colonization of

Introduction

3

the land was achieved by symbiotic organisms. The location of the fungal symbiont in the root and its hyphal connections with the soil ensure that it can influence the absorption of soil-derived nutrients, while in many cases obtaining organic C as recent photosynthate from the plant. This bidirectional transfer of nutrients is the basis of mutualism in most mycorrhizal symbioses. One of the important advances in the last decade of research is the increased emphasis on the structure, organization and function of the external mycelium and the role it plays in exploitation of, and nutrient mobilization in, the soil. Mycorrhizal fungi are specialized members of the vast population of microorganisms that colonize the rhizosphere. With a few exceptions mentioned below, they are completely dependent on the plant for organic C. Being independent of the scarce and patchily distributed organic C resources in soil, they are likely to be in a good position to compete with saprotrophs in the mobilization of N, PO^ (phosphate) and other nutrients. Another vital difference between mycorrhizal associations and the general association of organisms with the root surface or rhizosphere lies in the closeness of the relationship and the recognizable and consistent structures formed. In mycorrhizas there is always some penetration of the tissues of the root by the fungus, or a recognizable structure conforming to one of the common types, or both. The difference between the mycorrhizal symbiosis and those symbioses caused by parasites which lead to disease is that the mycorrhizal condition is the normal state for most plants under most ecological conditions. With the increasing awareness of the ecological importance of mycorrhizas and their diversity, research must be directed to experiments and surveys that will elucidate quantitative aspects of the distribution of different types and their contribution to the function of ecosystems, as opposed to simple records of their occurrence or casual speculation. This is one area of research in which tools of molecular biology are already being used to good effect in identification of the fungi present in single mycorrhizal roots by DNA fingerprints. These methods have, for example, enabled confirmation of the way in which plants are linked together by their fungal associates, the same fungus being found in the roots of separate individuals of the same or different species. This approach is proving particularly illuminating for associations involving achlorophyllous plants. These mycoheterotrophic associations are mycorrhizal in the sense that they are normal (indeed essential) for the plant and are important for nutrient acquisition. They differ from most mycorrhizas, in that C movement occurs in the same direction as mineral nutrients, so that strictly the plant is primarily parasitic on its fungal associate and secondarily dependent on another (photosynthetic) mycorrhizal partner. Molecular biology is also contributing to our knowledge of the taxonomy and phylogeny of mycorrhiza-forming fungi and of ways in which changes in gene expression in both symbionts are involved in establishment and function of the symbioses at different stages of development The Introduction to the first edition raised an important point with respect to the taxonomy of the fungi and the variability to be found within a single fungal species. The potential biological significance of genetic and biochemical variations within single fungal taxa has now been appreciated and investigations are in progress which will help to unravel the genetic bases of mycorrhizal development. However, taxonomy and nomenclature constitute continuing problems in the description of mycorrhizas and in understanding their developmental and functional differences.

4

Mycorrhizal symbiosis

For example, the well known mycorrhizal basidiomycete Pisolithus tinctorius should now be called (on grounds of precedence) P. arhizus. The species is notoriously variable in many ways, including its mycorrhiza-forming abilities, and is, we understand, soon to be split into a number of species so that P. arhizus as a name for this complex array of fungi will be short-lived. Consequently, we intend to continue (incorrectly) to use the name P. tinctorius to avoid unnecessary confusion at a later date. There have been many other changes in names of mycorrhizal fungi for reasons of nomenclature, as well as changes resulting from improved understanding of the taxonomy and phylogeny of particular groups. Where the new names are clearly established we have tried to use these, in preference to those in the original papers. In some cases we have indicated both the original names and those in current use. In any event, we are aware of the potential for confusion and apologize for it.

Classification of Mycorrhizas The classification of mycorrhizas adopted here has not changed significantly since the publication of the first edition and is shown in Table 1, which is taken from that edition with minor modifications which reflect advances in knowledge made in the last decade or so. As before, the classification aims to be descriptive and to emphasize problems in need of solutions, rather than to gloss over difficulties. The kinds of mycorrhiza are divided as before, on the basis of their fungal associates, into those involving largely aseptate endophytes in the zygomycetous order Glomales (formerly the Endogonales) and those formed by septate fungi in the Ascomycetes and Basidiomycetes. The plant symbionts in mycorrhizas are so numerous and so taxonomically diverse that a primary classification on that basis would be wholly impractical. We have retained the term 'vesicular-arbuscular mycorrhiza' (VA mycorrhiza) to describe those associations formed by members of the Glomales. The unifying characters of this order are currently defined as the ability to form a mutualistic symbiosis and to produce arbuscules within the cells of compatible plants. Since vesicles are not formed by all members of this order there is a trend, advocated strongly by some, towards adoption of the term 'arbuscular mycorrhiza', to replace the well established VA mycorrhiza (see Berch, 1987; Walker, 1995). The first argument against changing the name was expressed by Walker in 1992. He doubted the wisdom of defining this group of fungi, as Morton and Benny (1990) had done, on the basis of the symbiotic ability of its members. To this we add our own doubts about the assumptions made with respect to the functions of the arbuscule by Morton and Benny (1990). Furthermore, there are problems in extrapolating from structure to function in what now appear to be structurally and possibly also functionally diverse symbioses. The second argument is more pragmatic. The explosion of experimental work in the last decade has led to the appreciation of the potential importance of mycorrhizal symbioses by scientists in a wide range of disciplines and by foresters, horticulturalists and agriculturalists in many parts of the world. They have become used to the term 'VA mycorrhiza', or VAM, and a change of name is likely to cause confusion outside the restricted circle of mycorrhizologists. We therefore retain the general term, which covers a diversity of

O u ,

>^ E

c

s O ,

u

.2

E

IE

5

•D

a o.

8 O c o Z

N

E

c o LU

E Q

^

+

+

I

+

I

I + I

I

I

I

I

I + + + I

o

+ + +

o

+ + +

I

•X+

0

!§ m

a OQ

I

c> d

( N ON GO ON ( N r^ ro ro ON GO hs 1 ^

O

-^ -^

r^

>o 0 0 NO 00 lO ro f>l ON ^ - 60

^

50

N

40

c O

S

30

20

10

10

15

20

25

30

35 Days

40

45

50

55

60

F i g u r e 2.13 The effects of irradiance and additions of P on VA mycorrhizal colonization of roots of Allium cepa in low-P soil. High light, l o w P, • ; high light, added P, • ; l o w light, l o w P, • ; l o w light, added P, Q. N o t e the combined effect of added P and l o w light in reducing the percentage of the r o o t length colonized. D r a w n f r o m data of C . L Son and S.E. Smith.

Colonization of roots

77

Bruce et al (1994), using the modelling approach of Smith and Walker (1981), have shown that P applied to mycorrhizal Cucumis plants increased rates of initiation and extension of lateral roots, but these effects did not occur early enough to provide a complete explanation for reduced percentage colonization which was first observed at 10 days. Growth of infection units was reduced by added P from as early as day 8 and although the length of the lag phase of colonization was unaffected, the rate of production of entry points was very much lower in the presence of added P from day 20 onwards. This was probably the result of reduced secondary colonization. Effects of P on arbuscule development are somewhat variable. Sanders and Tinker (1973) and many subsequent observers (e.g. Hayman, 1974; Graham et ah, 1982a; Smith and Gianinazzi-Pearson, 1990; Amijee et al, 1989a; Bruce et ah, 1994) reported reductions in the densities of arbuscules in roots, but although Abbott and Robson (1979) observed lower percentage colonization in Trifolium suhterraneum, Erodium botrys and Lolium rigidum they saw no effects on arbuscule development. Given that high P concentrations in the medium reduce the growth of germ tubes, it might be expected that P in the soil might have an effect on hyphal growth from propagules and the initiation of colonization. Alternatively, high P within the roots might influence the rate of spread of the fungus and the growth of the extraradical mycelium. As shown above, the results are variable. Sanders (1975) investigated this problem by injecting P solution into the hollow leaves of onion plants. P was translocated from shoot to root and at the final harvest the percentage colonization was reduced, probably as a result of slower growth of the hyphae along the cortex of roots of high P content, a point which has been confirmed by Bruce et al (1994). The amount of external mycelium produced per centimetre of infected root was reduced from 3.5 to 2 mg, and there were also changes in the anatomy of the infection units. Sanders (1975) therefore concluded that the effects of soil P in reducing colonization were mediated via the root and need not involve any direct effects upon fungal growth in the soil. This early demonstration has been confirmed with other plant species using transplanting experiments and split-root techniques. However, the importance of direct effects via fungal growth in soil are supported by the results of Amijee et al (1993) who showed that the 'odds' on colonization were strongly reduced as soil P was increased. Menge et al. (1978) showed a reduction in numbers of chlamydospores produced by Glomus fasciculatus on both halves of a split root system of sudan grass, even though only one half received high levels (750 mg kg~^ soil) of P. Additional experiments indicated that the reduction in numbers of arbuscules and external hyphae, as well as chlamydospores, was more closely related to the P concentration in the roots than to that in the soil. Jasper et al. (1979) observed that a lower percentage of the root volume was colonized and fewer entry points per unit length of root were produced when Trifolium subterraneum was transplanted from sterilized soil low in P to a soil high in P containing spores of Glomus monosporus. Such an effect was not observed when plants from sterilized high-P soil were transplanted to infective low-P soil. In this experiment high P concentrations were associated with lower soluble carbohydrate concentrations in the roots and the authors suggested that carbohydrate availability to the endophytes might be important in determining fungal establishment, a point which will be elaborated later. Following early work which indicated reductions in colonization at low light and

78

Vesicular-arbuscular mycorrhizas

which provided speculation on the relationships with carbohydrate supply (Peuss, 1958; Schrader, 1958; BouUard, 1959), the effects of light and defoliation have continued to be investigated and it is clear that percentage colonization is almost always reduced in situations where supply of photosynthates might also be expected to be lower. Daft and El Giahmi (1978) found that both defoliation and either shading or short day length reduced percentage colonization and numbers of secondary spores produced by Glomus macrocarpus var. geosporus or G. mosseae in a variety of host plants. Hayman (1974) found that colonization was higher at higher light intensities and that this was correlated to sugar concentrations in the roots. The reduction in colonization at low irradiance is more marked if P supply is high (e.g. Graham et al, 1982a; Son and Smith, 1988; Smith and Gianinazzi-Pearson, 1990; Thomson et al, 1990a) and there have now been a number of investigations of the way light and P supply interact with the pools of soluble carbohydrate in the root and the amount and composition of root exudates, with the aim of correlating this with mycorrhizal colonization. Ratnayake et al. (1978) formed the hypothesis, based on experiments with Citrus, that exudation of substances from the roots of plants growing in low P conditions is increased as a result of the decrease in phospholipids and an increase of permeability of the cell membranes. They found that there was a much greater leakage of amino acids and sugars from the roots and they suggested that these might stimulate the growth of the fungus and the development of mycorrhizal colonization. They followed this up (Graham et al, 1981,1982a; Johnson et al, 1982; Schwab et al, 1982, 1984; and see Schwab et al, 1991) by studying other conditions where increased exudation results in an increase of mycorrhizal colonization, including an example of a non-host plant {Chenopodium quinona) in which extremely slight colonization was induced by the application of the herbicide simazine. In other investigations, the correlation between percentage colonization and concentrations of soluble carbohydrates in the roots systems was closer than that between percentage colonization and concentrations of exudates (Jasper et al, 1979; Thomson et al, 1990b, 1991). The picture is not clear, because negative correlations between P and soluble carbohydrates have also been found (Amijee et al, 1993; Pearson and Schweiger, 1993) and these more recent investigations have highlighted problems in interpretation of the earlier work. Deduction of causal relationships between the size of the soluble carbohydrate pool and extent of colonization is risky, because the pool size is determined by both input and output, i.e. by production and use. Data have been interpreted in two main ways: (1) the higher the pool size the more carbohydrate is available for the fungus (this is the argument used when a positive correlation between colonization and carbohydrate concentration is found) and (2) the higher the pool size the less carbohydrate the fungus has consumed (this is the argument used to explain negative correlations). The possible effects of the fungus itself in inducing changes in carbohydrate status or effects of carbohydrate status on root growth, and hence percentage colonization, have been more or less ignored. Amijee et al, (1993) measured the rate of growth of the fungus within the roots (rather than percentage colonization) and failed to find a close correlation between that and soluble carbohydrate concentration. They emphasized that in Allium the carbohydrate pools were similar in mycorrhizal and non-mycorrhizal plants that contain the same shoot P concentrations. They did observe some effects of mycorrhizal

Colonization of roots

79

colonization on the relative concentrations of monosaccharides (glucose and fructose). Light and P do not necessarily operate in the same way in reducing percentage colonization: Tester et al. (1986) found no effect of reduced irradiance on the rate of growth of infection units in Trifolium, whereas this parameter was influenced by P supply in Allium and Cucumis (Amijee et al., 1989a; Bruce et al., 1994). In summary, while it is clear that both low irradiance and high P reduce colonization, the mechanism(s) by which they do so are unclear and will not be elucidated until we have a better idea of the mechanisms by which the fungus obtains carbohydrate from the roots.

Root Growth The effect of mycorrhizal colonization on root growth has been investigated in Allium and although some work indicates that there are no changes in rates of apical extension or initiation of branches (Buwalda et al., 1984) it is now well established that changes do take place even if the colonization of the root by mycorrhizal fungi is relatively slight. Berta et al. (1990, 1991) have demonstrated that the rate of growth of root apices slows down after colonization by a Glomus species and that this is associated with a decrease in the mitotic index, because of extensions of G l , S and metaphase and marked reductions in the duration of G2. At the same time an increase in initiation of lateral roots is observed, presumably stimulated by the loss of activity of the apices of the main adventitious roots. The mechanisms underlying these changes are obscure and clearly operate at a distance from the actual sites of colonization, but as Koske and Gemma (1992) have pointed out, the increased branching of roots as well as that of hyphae would increase the chances of encounters between roots and infective hyphae. This is not the place to discuss environmental effects on root growth and development, but factors such as nutrient availability, temperature and soil compaction clearly exert marked effects on both apical extension and branching, and these will affect the interpretation of measurements of percentage colonization.

Conclusions The propagules that can initiate VA mycorrhizal colonization have been identified as spores, root fragments and hyphae. The latter form a complex network which links plants of the same and different species and when these are growing the infectivity of the mycelium is very high. The networks appear to be able to survive in both dry and cold conditions, which is probably very important in initiating colonization early in the following season. The processes leading to the colonization of roots and the way in which the fungi develop in the root systems is well understood in a few plant species. The outcome of the colonization process is that the fungus comes to occupy two probably different apoplastic compartments in the root, the intercellular spaces and a more specialized intracellular, arbuscular apoplast. There are thus two interfaces between the symbionts which show different specializations and may well have different functions. Outside the root an extensive mycelium develops and appears to undergo differentiation, so that different types of hyphae perform different functions in colonization, nutrient

80


acquisition and, possibly, survival. Although variations in colonization patterns were described early in this century, the comparative study of Arum- and Paris-type mycorrhizas has been neglected. Structural, developmental and physiological investigations are required to understand these diverse associations in field situations, as well as for managing mycorrhizas in primary production. The complexities of field situations, where many different species of fungi and plants coexist with other soil organisms, are appreciated but have scarcely been investigated and will provide very significant challenges.

Plate I. (a) Spores arid subtending hyphae oi Gigaspora margaritOy approximate diameter 400-450 |im. (b) Spore of Acaulospora laevis (arrowed), attached to the neck of the sporiferous saccule (s). Spore diameter approximately 190-210 |im. (c) Spores of Scutellospora n/gro, approximate diameter 300|Lim. Photographs courtesy of V. Gianinazzi-Pearson.

Genetic, cellular and molecular Interactions in the establishment of VA mycorrhizas

Introduction The development of vesicular-arbuscular (VA) mycorrhizas involves a well coordinated sequence of events, during which morphogenetic changes to both fungus and plant take place, supporting the maintenance of a compatible, biotrophic symbiosis (see Chapter 2). This chapter will describe the progress of the interactions between plant and fungus as they are influenced by cultivars, mutants and nonhost species, with the aim of highlighting stages in the colonization processes which may act as control points in the different interactions. Research on these topics is relatively recent and although much of the information is preliminary, rapid advances are to be expected in our knowledge of the molecular-genetic regulation of the symbiosis, as new approaches and methods are applied. These methods include specific staining or affinity labelling of different molecules in the walls or interfaces, in situ mRNA hybridization with nucleic acid probes to detect expression of genes likely to be involved in mycorrhizal development and function, and methods specifically designed to determine changes in gene expression during colonization. These last methods include construction and screening of cDNA libraries, differential display and analysis of changes in polypeptide profiles. When linked to detailed studies of mycorrhiza development in host plants with well characterized genomes and in mutants deficient for mycorrhizal colonization, these will be powerful techniques for investigating mycorrhizal symbiosis. In particular, they will help to overcome the difficulties arising from the unculturability of the fungi and they will allow direct investigation of the molecular composition and gene expression in intact mycorrhizal plants. It is important to describe precisely the stages of mycorrhizal colonization, in order to pinpoint differences in the interactions that occur in different plant-fungus combinations. Currently, descriptions are based almost entirely on morphological changes in the development of plant and fungus because there is little supplementary physiological or molecular information (Gianinazzi, 1991; Bonfante-Fasolo and

82


Perotto, 1992). As outlined in Chapter 2, the main features of the sequence are spore germination, growth of the external mycelium, formation of appressoria and colonization of the root cortex by an intraradical mycelium, development of arbuscules and growth of the extraradical mycelium (see Giovannetti et al., 1994). However, some details are not included in this list and Table 3.1 provides an expanded list of stages at which there may be important developmental changes, together with abbreviations which will be used later in the chapter (Smith, 1995). Schemes of this type are useful if they provide a means of accurate description and comparison of different interactions and consequently help to predict steps which may be under genetic control. However, they must remain flexible and be updated as new information becomes available. The morphological changes that occur in each organism during colonization indicate that a number of developmental switches occur during the establishment of the biotrophic and compatible interaction, and these require the exchange of signals leading to changes in gene expression. In the root it seems very likely that a number of different genes will be expressed in the different cell types as these are progressively colonized by the fungus. We already know that in cells colonized by arbuscules these probably include H^-ATPases, which show increased activity (Gianinazzi-Pearson et al, 1991a; Murphy, 1995; Murphy et ah, 1996) and, in Medicago, phenyl alanine annmonia lyase (PAL) and chalcone synthase (CHS) (Harrison and Dixon, 1994). Conversely, changes in the fungus (including branching pattern and wall characteristics) are probably induced by preinfection signals produced by the plant and by contact with, or penetration of, the different cell types. Unfortunately, there are only a few clues about which genes are involved in either organism or how the spatial and temporal coordination of expression is organized. In VA mycorrhizas the work has centred on the plant responses and genes, rather than on the fungus, for a number of reasons. First, the developmental sequence in Arum-type mycorrhizas is well known and comparisons of different species and cultivars indicate that the plant genotype can affect the extent of colonization and the response (see below and Chapter 4). Second, the plant 'hosts' can be grown with or without mycorrhizal inoculum, so that their development and gene expression can be studied in both mycorrhizal and non-mycorrhizal states. Furthermore, it is now realized that the choice of plant species for molecular genetic work should be based on a number of criteria which will significantly aid the analysis. These are listed in Table 3.2, together with their applicability to four genera currently in use for mycorrhizal research in this area, namely Medicago, Lycopersicon, Pisum and Hordeum. As shown in Chapter 1, there is no evidence in VA mycorrhizas for narrow cultivar-race specificity and the effects of mycorrhizal colonization on plant defence responses are minor. A plant species can probably accept most, if not all of the known glomalean fungi as mycorrhizal partners, while the fungi are similarly undemanding with respect to the identity of plant species that they can colonize. Only a few species of fungi (around 150) can enter into this kind of symbiosis and we know almost nothing of their genetics, except that they may be asexual and may have relatively large genomes compared with other fungi. However, research is becoming increasingly practicable, with the application of methods based on the PCR which permits amplification of specific DNA sequences from very small amounts of target DNA. This is an area of research in which rapid

u

u

c

o u ,

^ c

E _o

u

(U

O c

|2

c O

c CL

O c

"D

O

U

U

u (U

o Z

0)

U

O

u o > C

u Si

E ^

NO

00 ^ .

a;

^

£ 0 0

a.

J3 D

E

c

u O

txo

*= -9 « c S O

r

O

ON

ON

4>

3S? ._ c> »

T3

0 c

U 3 C

0)

E c (U u 0 V4-

0-

< D u c

00 00 ON

o

_0

c

o CO

c 0

Q.

0

0

0

£

^ "r5 ^ c rt

4-1

Q-

8 c 0 L. 0)

U

O

o U

o

0) XI

0)

0 Q.

(^ ^

tXO -a C 3 M -

Rr

0)

c

s

•D >v

3

x: "a c

^

JD

0) O-

c O

E

I 0 !5

L. Q-

(/)

tci

c

c

|-B

^

^r 4. s

U

c

Q.

(J

0)

O

W.

£5

^

«/>

0)

-I -^ 1-

«^

^

cd C

9-

_C

3

= .y 1^ g

C (d i-

*l_ U

«4-

Q(U

_Isd? c O CO Co

0 ) 'i_ 4_l 4-* Q . C/) rJ

^ u

c

cd

^

c 0

Q.

Q. •J C 03

t«

E

^ u £ ^ o s £

•

_c c

(d 0) i-

3 U

u c

U) 0) t>0

> Ueac C

c

U

p

(U

>

lap I

O

o £

"D

•^ .E

(D

3

u

C

a U txO ^

^ "O c c

^

D

Is O

O

Q. W) I

O

I

|2

o

8. 8

'5

U

U) (/)

:2^

£

3

c IS

0)

E ^"^

W)

(/)

3

—

. ro

—

ON

Q) ON Q_ ON Q. —

to

cd . _

Q-

O

U

Cd

— . (u

£

t^

r c

C

< z

r-

g3 I_.

f^

£ £ cd cd

c/t

1 1 u X X

to

3

£ £ E E 3 c/>

cC d:

£l

£ & E

£ fd

T u

>^ £ E CO

E i-

Q)

O

0

U

•D

0)

z

c

(A

+J

£ o u

•a

c

c o

^ , Q-1

N

o u ,

£ £ O c

c O u O

Z

|2

o

c O

CO

"5. J5 O

00

c o ro

Q-

c^

V o (d (2 ta.

to

92


identified phenolics produced by plants at physiologically relevant concentrations has been confirmed in only a few cases. These include exudates from roots of carrot seedlings containing the flavonols quercetin and kaempferol, which stimulate hyphal growth of Gigaspora margarita (Poulin et ah, 1993), and from seeds of Medicago containing quercetin-3-O-galactoside, which stimulates hyphal growth from spores of Glomus species (Tsai and Phillips, 1991). It is a reasonable assumption that the stimulatory effects of root exudates from other host plants are based on the production of similar compounds (Gianinazzi-Pearson et aL, 1989; Giovannetti et aL, 1993a, 1994). However, a note of caution is needed - Becard et al. (1995) have failed to detect any flavonoids in exudates of Ri T-DNA transformed roots of Daucus carota. As the roots were colonized by Gi. margarita, flavonoids are not absolutely essential for the process. The non-host plants Lupinus albus, Dianthus caryophyllus and Spinacea oleracea as well as Brassica species do not appear to produce compounds with morphogenetic effects because neither the presence of roots nor exudates from them increase growth or stimulate branching of hyphae growing from spores or sporocarps (Gianinazzi-Pearson et aL, 1989; Giovannetti et aL, 1993a,b; 1994). The same is true for plant species which form other types of mycorrhizas - Abies alba, Pinus nigra (ectomycorrhiza only). Arbutus unedo (arbutoid mycorrhiza) and Vaccinium myrtillus (ericoid mycorrhiza) (Giovannetti et aL, 1994) - but the dual host Alnus glutinosa (ecto- and VA mycorrhiza) does stimulate hyphal morphogenesis and branching of VA mycorrhizal fungi, as does the myc~^ mutant of Pisum (Giovannetti et aL, 1993a, 1994).

Appressorium Formation Hyphal contact with roots is followed by formation and adhesion of appressoria (Apr). The morphogenetic stimulus has not been identified, but it does not seem to be thigmotropic even though appressoria are often seen in grooves between cells on the root epidermis. No mechanisms of adhesion, such as the fibrils seen in ericoid systems, or the various mechanisms of adhesion found in fungal plant pathogens, have been detected. Both contact and appressorium formation are reduced or absent on roots of Brassica napus and Lupinus albus (Tommerup, 1984c; Glenn et aL, 1985, 1988; Gianinazzi-Pearson and Gianinazzi, 1992). There is evidence that a shoot factor is involved in the inhibition in Lupinus, because when shoots of Lupinus were used as scions grafted to rootstocks of Glycine max or Pisum sativum, appressorium formation was reduced, just as it was in intact Lupinus plants (Gianinazzi-Pearson and Gianinazzi, 1992; Tommerup, 1984c). When fungi are grown with the non-host plants Brassica, Dianthus caryophyllus, Eruca sativa. Nasturtium officinale and Spinacea oleracea, hyphae attach to the surface and develop swellings at their tips, which are morphologically quite different from normal appressoria on Basilicum (Giovannetti et aL, 1994). The swellings are more frequent on plants which act as hosts for other types of mycorrhizas than on species which are generally non-mycorrhizal. In no case did hyphal contact with non-host roots induce fluorescence in the host roots, nor was there deposition of lignin or callose, which would indicate initiation of a defence response. Francis and Read (1994) also investigated interactions between mycorrhizal fungi and the nonhost species Arenaria serpyllifolia (Caryophyllaceae) and Arabis hirsuta (Cruciferae).

Genetic, cellular and molecular interactions

93

They observed hyphal contact with roots, but there was no penetration and no evidence of lesions or any visually apparent defence responses. These results are quite different from the observations of Allen et at. (1989), who observed penetration of roots of Salsola kali (Crucifereae), associated with marked fluorescence and eventual rejection of the fungi. The myc~^ Pisum mutant (Pen~) does not affect appressorium formation (Apr"^), which is normal both with respect to frequency and morphology (Fig. 3.1a; GoUotte et aL, 1993; Giovannetti et al., 1993b). However, a number of plant genes seem likely to affect this stage because overproduction of abnormally shaped appressoria has been observed in mutants of both Medicago sativa and Lycopersicon esculentum (Fig. 3.1c; Bradbury et al, 1991). The mutants in Pisum myc~^ Medicago and Lycopersicon are all resistant to penetration (Pen~) and it has been suggested that overproduction of appressoria may be a fungal response to failure of tissue colonization (Bradbury et al., 1993a). Alternatively, the swollen structures may not be true appressoria, but may be equivalent to the much-branched hyphae produced by mycorrhizal fungi when actual contact with the root is prevented by artificial membranes, as in the studies by Giovannetti and co-workers. In this case the implication is that the myc~ genotypes produce compounds that stimulate hyphal growth and branching (Pif^ and Pab"^) but do not provide the appropriate stimulus for the formation of appressoria and are therefore actually Apr". This is an area where more research could be very productive, in particular with respect to the stimuli required for appressorium formation and mechanisms of adhesion. Normal appressoria are easily identified and in the Pisum myc~^ mutants they occur in the almost complete absence of other developmental stages. Consequently, fungal genes involved could be isolated by differential screening of cDNA libraries prepared from roots at the early stages of colonization of mutant or non-mutant peas, or a similar screen using a fungal genomic library, as has been done for Magnaporthe grisea (Lee and Dean, 1993). If the abnormal, branched structures on Medicago are true appressoria, increased expression of 'appressorium genes' might be expected in the interaction between mycorrhizal fungi and these plants.

Penetration Penetration (Pen) of the outer tangential walls of the epidermal cells follows appressorium formation and is marked by narrowing of the fungal hypha to form an infection peg. Localized production of wall-degrading hydrolytic enzymes by the fungus, as well as hydrostatic pressure exerted by the hyphal tip are both probably involved at this stage (see Chapter 2). In the plant, thickening of the epidermal cell wall occurs in some plant-fungus combinations (Fig. 3.2a) and indicates recognition of fungal contact. This step is clearly affected in the myc~^ mutant, in which well defined wall thickening, with increased deposition of phenolics and P(l,3)glucans (possibly callose), occurs beneath the appressoria (GoUotte et al, 1993), although the appressoria themselves remain alive and apparently normal (Fig. 3.2b). In contrast, in the hypertrophied 'appressoria' on Medicago nod~myc~ genotypes, electron-dense deposits are soon observed, indicating that both plant and fungal responses are involved in the failure of colonization in these interactions.

94


Intraradical Colonization As penetration and colonization of the root tissues proceed, the plant responds in a number of ways, which probably vary in different plant-fungus combinations, although this has not been systematically investigated. In particular, nuclear size increases, possibly indicating genome reduplication or increased rates of gene transcription. In the myc~^ Pisum mutant this occurs to a lesser extent than in wild-type peas, but information for other mutants and for non-host plants is not available (Berta et al, 1991; Sgorbati et al, 1993). Internal colonization of the root involves the formation of intercellular hyphae (Ih) coils (Cof and Cip) and arbuscules (Arb). The plant-fungus interactions have been studied by following changes in the deposition of molecules in the intercellular and intracellular interfaces, including fungal and plant walls (BonfanteFasolo, 1988; Lemoine et ah, 1995) and by the striking changes in fungal morphology during arbuscule formation. In arbuscule-containing cells, the plant nuclei migrate from the periphery of the cortical cells to their centre. They increase in size and the chromatin decondenses, as shown by antibody labelling (Balestrini et al., 1992). These changes, together with changes in the spatial localization of enzymes such as H^-ATPases and other phosphatases, clearly indicate a remarkable degree of ongoing coordinated development (see Chapter 2). Changes in wall chemistry of the fungus could possibly be important in determining the response of different types of plant cell to colonization. For example, there is no evidence that either intercellular hyphae or intracellular coils (both of which have relatively thick and well developed walls) elicit any defence responses in the host cells. In contrast, development of arbuscules, which have very thin and highly modified walls, does result in localized defence responses. A highly speculative explanation might be that stimulatory or elicitor molecules become exposed in the walls of arbuscules (see below). Formation of arbuscules is blocked (Arb~) or reduced (Arb^^^) both in myc~^ Pisum mutants (Gianinazzi-Pearson et al., 1991b) and in non-host plants such as Brassica, Salsola and Atriplex, where some tissue penetration has been reported. However, in none of these examples is the nature of the block understood. In the fungal interaction with myc~^ mutants, the increase of H^-ATPase activity normally observed on the PAM does not occur (Pat~), and a reasonable prediction would be that P transfer from fungus to plant would be considerably reduced, although this has not yet been investigated. Distribution of activity of plasma membrane H"ÂTPase in the fungus is normal: intercellular hyphae have high activity and the reduced arbuscules have none (Gianinazzi-Pearson et al., 1995; GoUotte et al., 1996; see also Gianinazzi-Pearson et al., 1991a). The vigorous growth of the intercellular

Figure 3.2 Cellular reactions during epidermal penetration of roots by mycorrhizal fungi, (a) Acriflavine-positive wall thickenings in the epidermal cells of Allium porrum (arrowed), in response to normal colonization by Glomus versiforme. Bar, 10 jiim. From Garriock et al. (1989), with permission, (b) Callose-containing wall thickenings in the epidermal cells of myc~' mutants of Pisum sativum cv. Frisson, in response to appressorium formation (ap) and colonization. Bar, 10 |im. From Gollotte et al. (1993) Planta, 191, Fig. 2, p. I I 5 , with permission.


F i g u r e 3.2 (Caption opposite)

95

96


hyphae in these mutant plants and the distribution of activity of ATPases indicate that arbuscules are not required for the fungus to obtain C from the plant, confirming the suggestion of Gianinazzi-Pearson et al. (1991a; see Smith and Smith, 1995, 1996b). Parallel investigations of nodulation and mycorrhizal interactions at the level of molecules present in the symbiotic interfaces are proceeding with the aim of identifying features which will provide clues to the nature of the common controlling genes in the two symbioses. The expression of several nodulins is increased in mycorrhizal plants, including Nod 26, (Gianinazzi-Pearson and Gianinazzi, 1988; Wyss et al., 1990) which is a membrane protein with sequence homology to a number of genes in plants and animals. The high degree of sequence conservation of Nod 26 in distantly related organisms indicates that it may have arisen in a common and very ancient ancestor and it has been suggested that it is a channel protein in plasma membranes (see Verma et al., 1992). If this is so, then its increased expression in both nodulated and mycorrhizal plants may be a consequence of the increase in membrane production during the development of the PBM and PAM. The PAM shows both similarities to, and differences from the PBM, as indicated by the use of antibodies raised to cell surface components in nodules of Pisum (Gianinazzi-Pearson et al, 1990; Perotto et al, 1994; Gollotte et al, 1995). Similarities between the PBM and PAM are not surprising as they are both at least partly derived from the plasma membrane of the cell and both have important transport functions associated with the symbioses. Apart from this, the results do not yet provide a coherent picture of common functional aspects of the membranes and interfacial compartments of the rhizobial and mycorrhizal symbioses.

Changes in Gene Transcription during VA Mycorrhizal Colonization The hypertrophy of plant nuclei, increased staining with DAPI (4',6-diamidino-2phenylindole) and susceptibility to degradation by DNAase have been accepted as evidence that colonization induces higher rates of gene transcription in mycorrhizal roots, leading to observed increases in accumulation of RNA on a fresh-weight basis (Schellenbaum et al, 1992; Franken and Gnadinger, 1994; Murphy et al, 1996). The relative contributions of transcription of specific symbiosis-related plant genes and up-regulation of genes which are also expressed in uncolonized roots is not yet known, but it seems likely that both are involved. Analysis of soluble protein composition (e.g. Pacovsky, 1989; Schellenbaum et al, 1992; Arines et al, 1993; Dumas-Gaudot et al, 1994b) has shown the production of new polypeptides in mycorrhizal roots; some of these are certainly fungal, but others may be plant proteins related to particular stages of colonization, as shown by their presence in Pisum myc~^ mutants with appressorial colonization by Glomus mosseae. So far, no clues on the functions of the new polypeptides (called mycorrhizins) have been obtained. A second method of detecting changes in gene expression in response to mycorrhizal colonization is the preparation of cDNA libraries, which provide a 'snapshot' of the mRNA species being produced at any one time. Differential screening of such libraries identifies cDNA clones representing mRNAs of particular origin, and can therefore be used to help identify genes expressed by plant or fungus at different


97

stages of colonisation. Differential display PCR (Liang and Pardee, 1992; Bauer et al., 1993) also of use in detecting genes with low copy number, including fungal genes. There are several preliminary reports of these approaches being used to investigate VA mycorrhizal associations (Rosewarne, 1993; Barker ei al., 1994; van Buuren and Harrison, 1994; Murphy, 1995; Martin-Laurent et al, 1995; Murphy et al, 1996; Ridgeway et al, 1996). Murphy et al (1996) isolated polyA"" mRNA from barley roots in the early stages of colonization by Glomus intraradices and used it to construct a cDNA library. The resulting clones were screened with ^^P-labelled cDNA, obtained by reverse transcription of RNA from non-colonized and mycorrhizal roots, in order to distinguish clones representing mRNAs which were differentially accumulated in the mycorrhizal interaction. Colonization resulted in both up- and down-regulation of genes. Of those clones selected for further analysis, BM 78 shows close sequence homology with H'^-ATPases from Arabidopsis and Lycopersicon and does not hybridize with DNA from fungal spores. Up-regulation of this gene is particularly interesting because it may be associated with the increased activity of ATPase on the PAM and be involved in symbiotic transport. When used as a probe to investigate expression of the gene in roots treated in different ways, accumulation of the mRNA hybridizing to BM 78 increased as mycorrhizal colonization increased, but was unaffected by P nutrition, infection by the pathogen Gaeumannomyces gratninis or water stress. BM 78 therefore appears to be clearly related to mycorrhizal colonization. Using barley addition lines in the wheat variety Chinese Spring, the gene has been mapped to the short arm of chromosome 2 of barley (Murphy et al, 1995; and see Fig. 3.3). The existence of mapping populations of Lycopersicon will facilitate isolation of genes from this genus (Giovannoni et al, 1991). Other up- and down-regulated mRNA species have also been found and mapped to particular chromosomes, but no information that would give clues to the functions of the corresponding genes is yet available. By screening a cDNA library from mycorrhizal Medicago tuncatula, Harrison (1996) has identified a clone with sequence homology to a membrane-bound sugar transporter, showing a four- to five-fold increase in accumulation of transcripts in mycorrhizal roots, but no change in either non-colonized roots or roots of the myc~ M. sativa genotype. Again, the link with an aspect of symbiotic transport is exciting, but in all cases much more work is required. Some of the up-regulated genes may be associated with resistance responses to pathogens (see next section) because small increases in both activity of the enzymes and accumulation of mRNA transcripts of genes involved in resistance have been found. It is for these that a third approach, using cloned probes for genes with known functions, has been applied with some success. This approach could usefully be extended to genes coding for proteins involved in other activities, such as membrane transport (see above), which can be predicted from physiological or biochemical evidence to be important in symbiotic development. The progress in application of molecular methods is promising and, when coupled with immunolocalization of the proteins a n d / o r in situ mRNA hybridization, has the potential to expand enormously our knowledge of the details of symbiotic interactions. So far, gene expression in the fungi has not been investigated to any great extent and the application of DNA technology has been applied only to investigations of relationships between taxa and to the development of taxon-specific probes for ecological investigations. However, a partial genomic library from Scutellospora

98


CSB123

4 5 6

7

Figure 3.3 Use of barley addition lines of Triticum aesHvum cv. Chinese Spring to determine the location of the up-regulated gene represented by BM 78 in the genome of Hordeum vulgare. DNA from 7! aestivum Chinese Spring (lane CS), H. vulgare cv. Betses (lane B) and Chinese Spring containing chromosomes 1-7 of barley (lanes 1-7) probed with BM 78. The gene is located on chromosome 2, as shown by the major bands (arrowed) common to barley and to the addition line containing barley chromosome 2. From Murphy et al. (1995), with permission.


99

castanea has now been produced and there is some evidence that three of the clones in this library may represent repeat sequences present at high copy number. As they hybridize to DNA from spores and mycorrhizal roots but not to DNA from noncolonized roots, these could be useful as probes for the occurrence of mycorrhizas, as tools for later mapping (Zeze et al., 1994) and, by comparison of base sequences, for analysis of genome evolution. Fungal genes with potential regulatory functions have recently been identified (Delp et ah, 1996) and from a functional standpoint the differential expression of a fungal PO^ transporter in extra- and intra-radical fungal structures is very important (Harrison and van Buuren, 1996) and will be discussed in Chapter 14.

Effects of VA Mycorrhizal Colonization on Resistance Responses Comparisons of plant defence responses to parasitic and mycorrhizal fungi are beginning to yield coherent information. However, it is important to recognize that although compatible interactions in these two types of symbiosis are superficially similar, there are also important differences. In mycorrhizas the only sign of loss of biotrophic status is the regular degeneration of arbuscules (see Chapter 2); it is the fungus, not the plant cell, which collapses and dies. In contrast, in biotrophic symbioses involving fungal parasites the plant cells die, either rapidly in hypersensitive, resistant responses or more slowly in compatible interactions. VA mycorrhizal associations are consequently analogous to compatible interactions, but direct comparisons between them must be made with care. One of the questions asked frequently is how does a mycorrhizal fungus avoid triggering a plant defence response? As described below, some defence responses are certainly mobilized for a short period but are later suppressed. VA mycorrhizas may represent a state of basic compatibility between plant and fungus which probably underlies all the more taxon-specific resistance mechanisms. Having evolved so long ago, the mycorrhizal recognition and signalling systems may predate evolution of specific mechanisms of resistance to plant disease. It seems likely that there would have been selective advantages in independent operation of systems of defence against pathogens and establishment of mycorrhizal interactions. All the work on deployment of defence responses against pathogens during VA mycorrhizal colonization indicates that they are weak and transitory. In typical mycorrhizal interactions there are no major changes in synthesis of lignin or callose in the plant cells. Metabolism of secondary metabolites, including flavonoids, is altered, but to a much lesser extent than in response to pathogen attack. Enzymes of the phenylpropanoid pathway have been investigated at levels of transcription and activity, particularly in legumes. During mycorrhizal development, early stimulation of both transcription and activity of PAL, CHS and isoflavone reductase (IFR) occurs in Medicago and Phaseolus, at least to a small extent. In Medicago an increase in transcription of chalcone isomerase (CHI) has also been reported (Harrison and Dixon, 1993; Lambais and Mehdy, 1993; Volpin et al, 1994). The increase in activity of PAL and CHI is transitory in M. sativa (Volpin et al, 1994), but in M. truncatula accumulation of PAL and CHS transcripts is maintained at 1.75- and 2.25-fold, respectively, above uninoculated controls for some weeks, while IFR transcripts decline. The increases are small compared with plant-pathogen interactions, but

100


they are apparently consistent. It must be remembered that the values are averaged over the different cell types in the roots, not all of which show increases in levels of transcripts (see below). The pattern of transcription of IFR is correlated with changes in accumulation of the phytoalexin medicarpin, and it appears to indicate the suppression of a defence response in mycorrhizal plants, which is not observed in the nod~myc~ M. sativa genotype (Harrison and Dixon, 1993). In Glycine, small increases in glyceoUin production have been observed following inoculation with Glomus species, but these are either slow to develop or not significantly greater than the uninoculated controls (Morandi et ah, 1984; Wyss et ah, 1991). In M. truncatula, in situ mRNA hybridization has shown that the increased levels of PAL and CHS mRNAs are confined to cells containing arbuscules, possibly indicating a localized stress response or a mechanism controlling intracellular invasion (Harrison and Dixon, 1994). This may help to explain the short life span of arbuscules. At present there are no data on responses in cells containing intracellular coils (e.g. the hypodermal passage cells), so that it is impossible to determine whether or not they differ from arbuscule-containing cells. No increases associated with intercellular hyphae were observed. Localization of HPRG in the arbuscular interface also suggests mobilization of small and localized defence responses (Bonfante-Fasolo et al, 1992), but although Franken and Gnadinger (1994) observed consistent increases in accumulation of transcripts of a gene coding for HRGP during mycorrhizal colonization of Fetroselenium, preliminary in situ hybridization analysis showed that the tissues involved were the stele and root apex, but not the cells colonized by arbuscules. In this non-legume, which has been used extensively to study plant-pathogen interactions, there were no major changes in the accumulation of phenolic compounds, of transcripts from PAL, CHS or 4coumaratexoA ligase (4CL) genes in mycorrhizal plants, nor of genes encoding other enzymes which respond to treatment with elicitors from Phytophthora megasperma f. sp. glycinea, including peroxidase. Chitinases and P(l,3)glucanases are also implicated in responses of plants to parasites, while peroxidase is important in the final stages of lignin deposition, again as a defence response. Chitinase and peroxidase sometimes show transitory increases in activity in plants infected by mycorrhizal fungi (Spanu and BonfanteFasolo, 1988; Spanu et al, 1989; Lambais and Mehdy, 1993; Dumas-Gaudot et al, 1994a; Vierheilig et al, 1994; Volpin et al, 1994), whereas activities of p(l,3)glucanases are either unaffected or suppressed, compared with uninoculated controls (Lambais and Mehdy, 1993; Vierheilig et al, 1994). Activities of these enzymes do not appear to be related to the control of colonization, because the progress of colonization is normal in transgenic Nicotiana plants expressing different forms of chitinase. Spanu et al (1989), using gold-labelled antibodies to bean chitinase, showed that in Allium porrum these enzymes are localized in plant vacuoles and intercellular spaces of both mycorrhizal and non-mycorrhizal plants and they are never found bound to the walls of G. versiforme. Indeed, the lack of reaction may be because the chitin in the fungal walls is rendered inaccessible to the enzymes by the presence of other wall components. In the non-hosts Brassica, Spinacea and Lupinus, increases in chitinase and P(l,3)glucanase activity, as well as ethylene production, have been observed in response to the presence of glomalean fungi, but the effects were all weak or transitory and did not markedly differ from responses of host plants (Vierheilig et al, 1994). In the myc~^ mutant of Pisum the defence responses appear


10!

to be stronger, accompanied by phenolics and callose, and exclude the fungus (GoUotte et al, 1993). The occurrence of a pathogenesis related protein (Pbrl) in the wall thickenings of myc~^ mutants indicates that other features of host defence against pathogens are mobilized (GoUotte et aL, 1994) and it will be interesting to discover whether or not the plant cells in this and the myc~^ mutant exhibit other characteristics of a hypersensitive response. In conclusion, it appears that rather than inducing major deployment of defence responses, mycorrhizal colonization may provoke a minor, transitory response which is followed by general suppression in both host and non-host plants. This is consistent with the persistent state of compatibility which mycorrhizas represent and there is no evidence from the species so far tested that failure of colonization by glomalean fungi in non-mycorrhizal plants involves mechanisms similar to those which are mobilized in plants resistant to attack by pathogens.

Possible Control Steps In VA Mycorrhizal Colonization Analysis of the interactions of glomalean fungi with normal host plants, with mutants having abnormal mycorrhizal phenotypes and with non-hosts has shed some light on the mechanisms of control operating in the symbiosis and allows some speculative predictions to be made. Figure 3.4 shows how the interactions vary and provides evidence that the mechanisms conferring non-host status are different in different taxonomic groups. The development of the fungus within the root is clearly influenced by the cell type colonized, so that hyphal coils invariably develop within hypodermal passage cells and sometimes within cortical parenchyma, particularly in P(irzs-type mycorrhizas (see Smith and Smith, 1996a). Intercellular hyphae only develop extensively in roots characterized by extensive intercellular spaces {Arum-type mycorrhizas), while arbuscules are characteristic of the cortical parenchyma and are sometimes localized in the innermost cell layer, in a ring surrounding the endodermis. Coordinated changes in development of both fungus and plant are striking, but we have few ideas on either the identity of the signals or the genes controlling the cellular modifications. It is clear that many genes must be involved in both organisms. The general absence of non-host variants in mycorrhizal species suggests that the significant genes could well be present in multiple copies or be so important in root function that their loss is lethal. Study of mutants has the potential to yield important information about the genes involved in normal mycorrhizal development and on any key control steps. Based on their investigations of the Pisum myc~^ mutant, GoUotte et aL, (1993) have proposed a scheme to explain the way in which the defence responses are suppressed in normal host plants. The hypothesis would explain a number of observations. It requires that a mycorrhizal (host) plant carries a dominant gene which codes for a receptor. This receptor would recognize a signal molecule from an appropriate mycorrhizal fungus and, if this occurs, the host defence responses would not be mobilized and colonization would proceed normally. Different receptors could be involved for each mycorrhizal type, and species with more than one receptor would detect signals from more than one type of mycorrhizal fungus. Mutants in which penetration does not occur could have evolved by loss of


102

Non host

Host/Mutants

Germ(*) Brassica spp.

Germ

Pit

Pif(-) Brassica spp.

^ PLANT ROOT EXUDATES

Pab-

Pab+

NOAPPRESSORIUM FORMATION

Apr" or AprH

Pen

Pen-1

UNSUCCESSFUL

myc""" mutants Apr-*Pen-

myc- 2 mutants

C

a= 'y

myc~2 mutants Apr-^ ih+ arb~

F i g u r e 3.4 Diagrammatic representation of the sequence of events during colonization of hosts, non-hosts and mutants by glomalean fungi. Modified from Giovannetti et al. (1994).

the receptor, explaining their recessive status as well as the mobilization of some defence responses. This picture contrasts with models for gene-for-gene resistance to plant pathogens, in which the plant-fungus interaction is always highly specific and resistance is dominant. In these interactions susceptibility (recessive) has evolved by the loss of gene function. At present we must emphasize that no signal compounds or receptors have been identified in mycorrhizal plants or fungi. However, molecules of this type are known for the interaction between Alternaria alternata and a range of host species (see Smith and Smith, 1996b). In the plants, resistance is recessive, as in mycorrhizal mutants, and it is controlled by a single gene which is thought to code for a receptor. Lack of the receptor prevents colonization of the resistant plants carrying


103

the recessive allele. The fungus produces a signal molecule (toxin) which binds to the receptor and confers compatibility on the association. In the Alternaria system the signal molecules produced are specific for particular host varieties (host-specific toxins) and confer specificity as well as compatibility. However, this feature would not be required in mycorrhizas and could very well have been selected against, because having a wide host range would confer considerable advantages for a mutualistic, root-colonizing fungus. There are some problems with the hypothesis. The defence responses potentially mobilized against mycorrhizas and pathogens must be different, otherwise mycorrhizal plants, having switched off their defences, would be much more susceptible to pathogens; in fact, the reverse is normally observed. The mechanisms which exclude mycorrhizal fungi from non-host species are quite varied, and involve toxin production by brassicas and apparent failure to produce morphogenetic signals in a number of other taxa. It is necessary to propose that the different mechanisms are the result of the action of different genes. The simplest in concept would be those which actually block colonization at an early stage; these might be expected to be dominant. Others, representing the loss of one or more key signals or receptors, would be recessive. There is as yet no systematic evidence for variation in susceptibility within non-host species, and no mutant screens have yet been undertaken to identify mycorrhizal phenotypes. Such mutants could give important leads in understanding the genes and mechanisms by which different taxonomic groups of plants achieve non-host status.

Independent Fungal Growth Perhaps the most striking result of the normal plant-fungus interaction is the change in growth rate and reproductive capacity of the fungus following colonization of the root. Possible blocks in fungal metabolism in the absence of the plant have been sought (e.g. availability of particular nutrients, absence of key enzymes or metabolic processes and nuclear division; see Chapter 2), but none has been identified. However, there are indications that membrane transport processes may not be fully operational in germ-tubes growing from spores, and that activity of the plasma membrane H"^-ATPase on the fungus, which is important in nutrient uptake and transfer, depends on colonization (Thomson et al., 1990a; Lei et ah, 1991). There are inconsistencies in the results, possibly because of differences in the age of the experimental material (F.A. Smith, personal communication), so that more experiments are needed. It has also been suggested that the fungus lacks some essential genetic information (see Hepper, 1984a). In symbiosis this might mean that the fungus is dependent on the plant at the metabolic level, with key and as yet unidentified processes coded for by plant genes. Alternatively, genetic material might actually be transferred from plant to fungus, before continuing fungal growth can occur. This suggestion is highly speculative and has not been investigated. Conclusions Much of the material presented in this chapter is preliminary and it is most important to appreciate that many of the ideas are still speculative and unconfirmed by

104


experiment or observation. The aim has been to provide an overview of variations in mycorrhizal colonization at the cellular level which may help to indicate the directions of research which may be fruitful. We know that the development of mycorrhizas must be under the control of plant and fungal genes, which act in a coordinated manner to produce the characteristic, biotrophic and compatible interaction in mycorrhizal host species. However, our understanding of the details of the process at the genetic level is rudimentary. Several research groups are now addressing the problems and have identified plant-fungus combinations and methodologies which will facilitate molecular genetic investigations. It is possible to describe the colonization process in precise terms, so that key regulatory steps can be identified and consequently investigated. At this stage the mechanisms which prevent mycorrhizal colonization in non-host species are still not completely clear, but there are a number of useful leads and a clear recognition that different taxa may exclude the fungi by different mechanisms. As the mycorrhizal condition is probably primitive in land plants, non-host species are likely to have retained many of the genes controlling colonization in host species. When these have been identified we will have powerful tools with which to investigate non-host plants and also to compare the compatible biotrophic symbioses involving mycorrhizal fungi and fungal plant pathogens. Understanding the genetic control of mycorrhizal symbiosis will provide insights into the evolution of the symbiosis and hence the present-day interactions in natural ecosystems. It will also be important in the biology of crop plants, particularly in the development of cultivars which can be designed to fit particular production systems. For example, in low-input agriculture where high efficiency of nutrient uptake is required, a cultivar that is both highly susceptible and highly responsive to mycorrhizas would be appropriate. Conversely, in highly fertilized situations a non-susceptible and non-responsive cultivar might be more suitable, always assuming that mycorrhizal effects on interactions with pathogens or soil structural stability had been adequately considered (see Chapter 16).

Growth and carbon economy of VA mycorrhizal plants

Introduction This is the first of two chapters on the physiological interactions between the symbionts in vesicular-arbuscular (VA) mycorrhizal plants. It will cover the basis of the symbiotic relationship in terms of transfer of organic C and soil-derived nutrients between the symbionts and it will describe the C economy of the symbioses and the growth responses of the plants to colonization by the fungi. The importance of mycelial links between plants will be considered, in so far as they affect C allocation among members of a group of plants. Interactions between glomalean fungi and non-photosynthetic (heterotrophic) plants will be discussed briefly, because of their unusual method of organic C nutrition. The symbiosis between VA mycorrhizal fungi and autotrophic plants is generally regarded as mutualistic, with the basis of mutualism assumed to be the bidirectional transfer of nutrients. With the exception of a few achlorophyllous species, VA mycorrhizal plants are autotrophic and, although normally colonized by mycorrhizal fungi in the field, they are usually capable of satisfactory growth in the absence of colonization, provided that mineral nutrient supplies are adequate. In contrast, the fungi are ecologically obligate symbionts. There is no good evidence that any glomalean fungi have significant saprophytic ability and the limited capacity of their propagules to produce mycelium in the absence of colonization of a plant is based on mobilization of reserves (see Chapter 2). In consequence, the fungus depends on recent photosynthate supplied by the autotroph and, as will be discussed later, utilizes a considerable proportion of the assimilated C. Conversely, the development of the external mycelium of the fungal symbiont in the soil increases the capacity of the plant to absorb nutrients because both roots and hyphae are involved in uptake. The mycelial habit ensures that the fungal symbiont has access to soil-derived nutrients, some of which are passed to the host plant. This simple view of mutualism takes no account of changes in the quantitative balance of transfer between symbionts at different times as the plants develop, nor of the fact that nutritional interactions within an uneven-aged plant community, composed of different species, may be very complex indeed. Furthermore, it

106


ignores non-nutritional features of the symbiosis. For example, the more stable physicochemical conditions in the root apoplast (e.g. water potential, ion concentrations, pH), compared with the soil environment, may be advantageous for the fungus. For the plant there is increasing evidence that mycorrhizal colonization may increase resistance to pathogens and insect herbivores, as well as tolerance of water deficit. Consequently, it may be difficult to demonstrate that the symbiosis is mutualistic, especially in natural ecosystems, and the term mutualism should be used cautiously. The questions that will be addressed in this chapter concern particularly the way in which mycorrhizal plants allocate photosynthate to growth of roots and shoots and to the development of the symbiotic fungus, and how this is affected by nutrient supply. Understanding these physiological relationships, including movement of organic C and growth responses, requires some understanding of the role played by the fungal symbiont in mineral nutrition, as well as its requirements for C. This will be covered briefly here and the details discussed in Chapter 5. Fungal translocation to the roots of nutrients absorbed from soil and the cellular bases for nutrient transfer from one partner to the other are included in Chapter 14.

Plant Growth The association between the development of VA mycorrhizas and increased growth of the host was made by Asai (1944) in his studies of mycorrhizal colonization and nodulation in a large number of legumes. He concluded that colonization was important both in plant growth and in the development of nodules. Subsequently, many investigators have carried out a large number of experiments which in general demonstrate that colonization is followed by considerable stimulation of growth. This early work has been extensively reviewed, with particular emphasis on the importance of mycorrhizas in P nutrition (e.g. Gerdemann, 1968, 1975; Mosse, 1973; Tinker, 1975a,b; Smith, 1980; Gianinazzi-Pearson and Gianinazzi, 1983; Harley and Smith, 1983; Hayman, 1983; Smith and Gianinazzi-Pearson, 1988; Koide, 1991a). Pioneering work on the potential significance of mycorrhizas in plant nutrition was carried out by Mosse (1957) on apples, Baylis (1959, 1967, 1972a) on Griselinia and other New Zealand plants, and Gerdemann (1968) on Liquidambar and maize. Subsequently, Daft and Nicolson (1966, 1969a, b, 1972), and Hayman and Mosse (1971, 1972; see also Mosse and Hayman, 1971) independently investigated the basis for this growth response in a number of plant species, in particular with respect to soil conditions and inoculum density. They demonstrated that development of mycorrhizal roots and their effect on plant growth is greater in soils of low or imbalanced nutrient status, in particular if P is in short supply, and they made valuable advances in interpretation of the mechanisms of these effects, which are discussed in Chapter 5. Increased growth has been demonstrated for a very wide variety of plant species including many crop plants and trees (Plate 2); it is manifest as increased growth of roots as well as shoots, reduced root:shoot ratio and increased tissue P concentrations (see Table 4.1 and Fig. 4.1). In a few plant species increased flower production and yield have also been demonstrated. Nodulation and N fixation in mycorrhizal legumes and dually colonized actinorrhizal plants are also increased and have been

s 3 0)

c o fc a; 3

E

VJ

.3

^ |S "o ^ o o T3 C

O

9 c O c u c o u C

txO

2 c O c O C

-a c (d

c O

N

c _o o u

o u

E

•a

^ _ |2 on

Q)

(/) C O O-

g 'M

0 0

x>c

— to

•J

±

1 a:

5

o o

fi

O 0

"D

E c 0

c 0)

u c o u

Q.

Q.

c E

t>0

_0

o U

£ c

•go -D "O

m

^

avOO. O^ C> 00 tN — LO 00 O — — f S —* ro —* r4

r o ' ^ i O f M O O h v r o irj ro hs — / '-y^î/^

I

yit

Membrane (0-45 fim) Nets (30//m)

/

- 37 /fm mesh 8 cm / • '

„^/ y

f\

/ /

\

JC^\^'

/ V v ^ X ^\. /VyrV" v^^s. ^

/ \ / /Hyphae/ /^x^ r "Pdish

\ .

1. -A.

J

J~ > v l lOO/tm ) | mesh ."^

•

^

-

Bulk Bulk soil Roots soil 3 i^2i--2i--.2.^ 3 — Hyphae Hyphae

V ^ 165 mm

Figure 5.3 Compartnnented pots used for investigation of the effects of extraradical myceliunn on nutrient uptake by plants. The pots are divided into compartments using mesh, so that roots but not hyphae are excluded, (a) System devised by Jakobsen et al. (1992b), in which roots of the plant (in this case Trifolium subterraneum) are excluded from the hyphal compartment by 37 jim mesh. Reprinted with permission, (b) System devised by Li et al. (1991b), in which the plant compartment is separated from the hyphal compartment by 30 |im mesh. The outermost compartments (representing bulk soil) are further separated by 0.45 jim membranes which exclude both roots and hyphae. Reprinted with permission.

to be equivalent to a cylinder of root hairs 1 mm long - a surprisingly small effect. The data in Figure 5.4a demonstrate the effect of hyphal colonization on the depletion of P from soil with two levels of P applied in the hyphal compartment. The percentage colonization of the roots and the hyphal length density in the hyphal compartment differed little between the P treatments (76-79% and around 5.0 m cm~^ soil, respectively; see Fig. 5.4b) and, assuming all the hyphae were alive, values of P inflow to the hyphae of 3.3 and 4.3 X 10"^^ mol P m"^ s"^ for the two P treatments were calculated and are very close to the value of 2.25 X 10~^^ mol P m~^ s"^ that can be derived from the data of Sanders and Tinker (1973) making the same assumptions (Table 5.2). Hyphal uptake extended at least 12 cm into the hyphal compartment (the maximum possible in the experimental system) and had a significant effect on the mycorrhizal plants, which had both higher tissue P concentrations and higher dry weights than the non-colonized controls. Furthermore, the proportional contribution of the hyphae to total uptake was slightly higher with the higher P supply, emphasizing that localized patches of high P in soil (in this case the soil in the hyphal compartment) can be effectively exploited by the mycorrhizal mycelium. This is an important point, because

136


2

4

6

8

Distance from the roots (cm)

4

6

8

10

Distance from the roots (cm)

Figure 5.4 Phosphorus depletion and hyphal length density in soil in compartmented pots planted with either mycorrhizal or non-mycorrhizal Thfolium repens. (a) Depletion profiles of NaHCOa-extractable (Olsen method) P in the hyphal compartments of mycorrhizal and non-mycorrhizal 7! repens, grown at two P levels. In both cases the mycorrhizal treatments ( I or # ) have depleted the soil to a greater extent than the non-mycorrhizal treatments ( D or O)- (b) Hyphal length density at different distances from the boundary of the root compartment, at the same two P levels used in (a). Note that there were small effects of P on hyphal development, so that the different depletion profiles cannot be attributed to different hyphal densities. From Li et al. (1991a), by permission of Kluwer Academic Publishers.

although the effects of mycorrhizal colonization on inflow are frequently greatest when soil P supply is low, significant effects may be apparent even when levels are adequate for near-maximal plant growth, particularly during later stages of plant development when depletion zones would be maximal and much of the P initially available in the pots had been absorbed (e.g. Smith, 1982; Smith et al, 1986a; Son and Smith, 1988; Dunne and Fitter, 1989). Use of the compartmented pot system to study the effects of an established mycelium on P uptake from temporally discontinuous P supplies might provide further insights into the roles of the external mycelium, particularly in habitats (such as wet tropical forests) where competition between plants and microorganisms for nutrients mineralized from large flushes of organic matter are important in 'tight cycling' (Janos, 1987; Lodge et al, 1994). The importance of hyphal translocation in soil was demonstrated directly as early as 1973 (Hattingh et al, 1973; Hattingh, 1975; Rhodes and Gerdemann, 1975, 1978a,b; and see Rhodes and Gerdemann, 1980). Using compartmented perspex chambers, ^^P, ^^S and ^^Ca were injected at different distances from mycorrhizal onion plants and the appearance of the tracer followed. ^^P was translocated by hyphae of Glomus mosseae and G. Jasciculatus' up to 7 cm through soil to roots of onion. When hyphae between the source of "^^P and the root were cut, no translocation to the root was observed. Extension of this approach using larger volumes of soil and measurements of hyphal length densities in the hyphal compartments (HC) has provided quantitative data for several different fungi (Jakobsen et al, 1992b; see Figs 5.3 and 5.5). Hyphal uptake and translocation of ^^P from a source localized in the HC at different distances (0, 1.0, 2.5, 4.0 and

Mineral nutrition, heavy metal accumulation and water relations

137

5 I

^

10

0-1

1-2

2-3

3-4

4-5

S-7

Distance from root compartment (cm)

Figure 5.5 a-d The time course of appearance of radioactivity in young leaflets of Trifolium subterraneum in association with: • , Acaulospora laevis; • , Glomus sp.; • , Scutellospora calospora; or x, non-mycorrhizal. Distances (cnn) between the ^^P-labelled soil and the root compartment were: (a) 0; (b) 1.0; (c) 2.5; and (d) 4.5. Bars are standard errors of means, (e) Length of hyphae in soil sections from the hyphal compartments of the experimental units with T. subterraneum in association with: • A laevis; M Glomus sp.; or H S. calospora. Values were corrected for background hyphae. From Jakobsen et al. (1992b), with permission. 7.0 cm) from the root plus hyphal compartment (RHC) into Trifolium subterraneum was followed for u p to 37 days. Hyphae of a Glomus sp. (WUM 10(1)) were relatively dense close to the roots and this fungus transferred most ^^P to the plants when the P source was similarly close. In contrast, Acaulospora laevis had a higher hyphal length density between 2 and 5 cm from the root than closei; to it and absorbed ^^P effectively from more distant placements. Scutellospora calospora

138


T a b l e 5.2 Inflow of P t o hyphae f r o m soil Plant/fungus

Nutrient/system

Inflow (mol m~' s " ' )

Reference

Allium cepa/Qomus

P/soil

2.25 X 10"'^

Sanders and Tinker (1973); see also Tinker (1975)

Trifolium repens/ G. mosseae

P/soil compartmented

3.3-4.3X10"'^

Li et o/. (1991a)

(WUM 12(2)) had relatively low hyphal length densities and did not transfer much '^^P to the plants from any distance. However, hyphae of this fungus contained around four times more ^^P than the others and the conclusion was that although the S. calospora was able to absorb and translocate ^^P, transfer to the plants occurred at a low rate and consequently P accumulated in the hyphae. Only a relatively small total amount of P need be absorbed and retained by the hyphae for this effect to be apparent. The reasons for the low rate of transfer have not been determined, but might relate to low density of active arbuscules (Smith and Dickson, 1991) or to inherently low transfer fluxes across the symbiotic interface(s). The relative effectiveness of the same Glomus and S. calospora has been confirmed in dual labelling experiments in which ^^P was supplied to roots plus hyphae in an RHC and ^^P to hyphae alone in an HC (Pearson and Jakobsen, 1993b). However, hyphal uptake from HC of a third fungus. Glomus caledonium, was as great as uptake from RHC, leading to the suggestion that root uptake was completely inhibited by the presence of the fungus, via an unknown mechanism (Pearson and Jakobsen, 1993b; Jakobsen, 1995). Subsequent re-evaluation of the data indicates that G. caledonium was extremely effective in exploiting the soil and that combined uptake by roots plus hyphae was likely to have depleted the P in the RHC completely. Consequently the apparently low value for root uptake was probably a result of this, rather than more subtle interactions between the symbionts (Jakobsen, personal communication). The rates of translocation of ^^P (as well as ^^Zn and ^^S) in hyphae of G. mosseae to Trifolium repens and Allium cepa plants have been measured in split agar plates by Pearson and Tinker (1975) and Cooper and Tinker (1978, 1981). In these investigations the fluxes were one or two orders of magnitude lower than the theoretical flux calculated for entry point hyphae mentioned above (1.02 X 10~^ to 2.0 X 10"^ mol m~^ s~^; Table 5.3). These measurements involved counting and measuring the hyphae traversing a diffusion barrier in a split agar plate. The discrepancy between the measured fluxes and those calculated from P uptake has been attributed to three factors (Cooper and Tinker, 1978): (a) the strong likelihood of higher fluxes in entry-point hyphae than in those of the general extraradical mycelium; (b) lower growth rates and P uptake of plants growing in agar culture, compared with plants growing in soil and hence lower demand and transfer across the interface; and (c) some of the hyphae used to calculate crosssectional area for flux determinations may have been dead. The second point may be of particular importance, since it has been shown that a mass flow component of the translocation process is increased by rapid transpiration rates in associated host

139


T a b l e 5.3 Translocation of nutrients in external hyphae, calculated per unit cross sectional area of hyphae Plant/fungus

System

Translocation flux (molm"^s"')

Reference

Allium cepa/Glomus sp.

Soil/P entry points

3.8 X 10""*

Sanders and Tinker (1973)

Agar/^^P

3-10 X 10"^

Pearson and Tinker (1975)

Agar/^^P

2.0-20 X 10~^

Cooper and Tinker (1978, 1981)

Soil/'^N

7.42 X 10"^

Ames eta/. (1983)

Agar/"Zn

2.1 X I0~®

Cooper and Tinker (1978)

Agar/^^S

16.5 X 10"

Cooper and Tinker (1978)

Thfolium repens/ G. mosseae Thfolium repens/ G. mosseae Apium graveolens/ Glomus sp. Thfolium repens/G. mosseae Thfolium repens/G. mosseae

plants (Cooper and Tinker, 1981). Plants grown in agar may well have lower transpiration rates than plants in pots. Suggested mechanisms of translocation in fungi involve processes based on both mass flow and cytoplasmic streaming (see Chapter 14). In VA mycorrhizal fungi any mechanism must take into account the potential for bidirectional movement of different nutrients (e.g. P and organic C) and the fact that translocation is temperature sensitive and inhibited by cytochalasin b, which also inhibits cytoplasmic streaming. It seems likely that the system of pleiomorphic, motile, vacuolar tubules recently demonstrated in members of the Eumycota, including members of the Glomales (Basidiomycotina, Ascomycotina and Zygomycotina; see Shepherd et al, 1993a,b; Rees et al, 1994; S. Dickson and A.E. Ashford, unpublished data), is involved in fungal translocation and investigations of the factors influencing their behaviour and translocating potential can be expected to shed light on translocation in mycorrhizal systems. Considerable transfer of P from fungus to plant occurs in most of the plantfungus combinations studied. At this point it is essential to emphasize that transfer of P from plant to fungus must involve membrane transport steps at an interface which includes the membranes of both symbionts and an apoplastic region between them (see Chapters 2 and 14). Furthermore, it must be reiterated that arbuscules do not contain sufficient P for their 'digestion' to be a credible mechanism for P transfer (Cox and Tinker, 1976). It is generally assumed that the site of transfer is the arbuscular interface, and although the distribution of membranebound H"^-ATPases does appear to support this (Smith and Smith, 1990; GianinazziPearson et al., 1991a; and see below) there is no definitive evidence that would exclude intercellular hyphae or hyphal coils from involvement in this process. There is no doubt that the interface between the symbionts in cells colonized by arbuscules would provide, as Cox and Tinker (1976) emphasized, a relatively large surface area across which P and other nutrients could also be transferred. They


140

Table 5.4 Net transfer of P to the plant across the symbiotic interface, assuming the arbuscular interface alone is responsible for transfer Plant/fungus

Nutrient/system

Transfer flux (mol m~^s~')

Reference

Allium cepa/Glomus mosseae

P/soil P/soil P/soil

13 X 10"' 4-29 X 10"' 2.0-3.2 X 10"'

Cox and Tinker (1976) Sukarno et al. (1996) Smith eto/. (1994)

P/soil

5.0-12.8 X 10"'

Smith eto/. (1994a)

A. cepa/Glomus s p . ( W U M 16) A. porrum/G. mosseoe A. porruml Glomus sp.

(WUM 16)

Hyphal contribution to inflow calculated from total P uptake in mycorrhizal and non-mycorrhizal plants. Area of interface deternnined from numbers of arbuscules and invagination of the plant plasma membrane (see Cox and Tinker, 1976).

measured the area of interface in such cells and calculated the flux required to support transfer of P from fungus to plant. P flux via arbuscules of Glomus mosseae to Allium cepa was 13 X 10"^ mol m~^ s~^. Similar calculations have now been made for Glomus sp. (WUM 16) and G. mosseae colonizing Allium porrum (leek) and found to be of the same order (Smith et al, 1994b; and see Table 5.4). The magnitude of these transfer fluxes is of the same order as uptake of P by free living plant and fungal cells, and very much larger than measured rates of efflux (Beever and Bums, 1980; Elliott et al, 1984). As net transfer across a symbiotic interface involves both efflux and uptake operating in series, the two processes must occur at equal rates. The inevitable conclusion is that P transfer from fungus to plant involves special modifications to increase the efflux from the fungus and probably also to suppress reabsorption (uptake) by the fungus, as this would negate the efflux (Smith et al., 1994b,c; and see Chapter 14). The mechanisms underlying the increased efflux are unknown at present. In conclusion, the involvement of hyphae in effectively and economically extending the root system of colonized plants seems very clear and provides a rationale for the negative relationship between development of root hairs (which play a similar role) and the responsiveness of species to mycorrhizal colonization. The idea that mycorrhizas might replace root hairs was first mooted in the nineteenth century and more recently Baylis has discussed the evolution of root systems and their associated mycorrhizas (Baylis, 1972a, 1975). His general thesis has been that plants with thick unbranched roots and few root hairs (e.g. Allium, Coprosma, Citrus) are apparently more responsive to mycorrhizal colonization when growing in low P soils than plants with finely branched roots and long or numerous root hairs, although all may be susceptible to colonization (St John, 1980; see Chapter 1). Competition Between Hyphae of VA Mycorrhizal Fungi and Soil Microorganisms The possibility that extraradical hyphae of mycorrhizal fungi might effectively increase the 'competitive ability' of the mycorrhizal root system, vis a vis soil microorganisms, in acquiring P from the soil solution and thus circumvent the problems of immobilization of P in the soil biomass has been canvassed from time to time (see Linderman, 1992). The dependence of mycorrhizal fungi on recent photosynthate from the plant means that their activity would not be affected by the


141

C: P ratio of organic matter in soil, nor by the availability of the C substrates in soil, giving them considerable advantages over saprophytic microorganisms. Barea et al., (1975) were among the first to investigate interactions between P solubilizing bacteria and mycorrhizal inoculation in the mobilization of P from rock phosphate (RP). They observed positive, synergistic effects in growth and P uptake by Zea mays and Lavendula spica, that were significant in some soil-plant combinations. Similarly, the potential of a P-solubilizing fungus Penicillium balaji to increase the availability of RP to Triticum aestivum and Phaseolus vulgaris depended on mycorrhizal activity (Kucey, 1987; Kucey et ah, 1989). Whereas RP alone had no effect on plant growth, the separate positive effects of inoculation with a mycorrhizal fungus and P. halaji were additive. Furthermore, in sterile soil mycorrhizal inoculation was an absolute requirement for the RP-P. halaji system to provide additional P to the plants. Jayachandran et al. (1989) used three soils and showed that if strong iron-chelating agents were added, the P released was available to mycorrhizal but not to non-mycorrhizal plants. They found no evidence for production of chelating agents by the mycorrhizal fungi themselves and concluded that the outcome was the result of effective exploitation of the soil and competition with resident microflora. P mineralized from organic P sources appears to be more readily available to mycorrhizal than to non-mycorrhizal plants, but there is no good evidence that mycorrhizal fungi are actually involved in the mineralization process (Jayachandran et ah, 1992; Joner and Jakobsen, 1994). Joner and Jakobsen (1994) concluded that Glomus sp. (WUM 10) and Glomus caledonium were both capable of intercepting Pi released during mineralization of Po by microorganisms and preventing immobilization in the biomass or sorption on clay minerals. The external hyphae have frequently been seen to proliferate preferentially in organic matter in soil (St John et ah, 1983a,b; Warner, 1984), which would be an appropriate situation for the operation of this competitive effect. Furthermore, the location of living hyphae of VA mycorrhizal fungi within dying roots appears to be important in the redistribution of ^^P to neighbouring plants which are linked into the external mycelium (Ritz and Newman, 1985; Eason and Newman, 1990; Eason et al., 1991; and see Newman, 1988). More work is needed to understand the interactions between the activity of the soil microflora in both mineralizing and immobilizing Pi and the potential capacity of mycorrhizal fungi to short-circuit this aspect of nutrient cycling. The interactions between mycorrhizal colonization and populations of particular functional groups of organisms both in the rhizosphere and further from the root where hyphae may proliferate are complex and as yet their significance is rather poorly understood (Azcon-Aguilar and Barea, 1992; Linderman, 1992; Fitter and Garbaye, 1994). The findings might be significant in studies of the fate of nutrient flushes and of fertilizer P, especially in situations where leaching contributes to losses from the soil and possible accumulation in water supplies. In this context, seasonal development of mycorrhizas in the fine lateral and cluster roots of some dryland species in the Restionaceae and Cyperaceae has been observed to coincide with the first period of winter rains, when nutrient mobilization might be maximal (Meney et ah, 1993). The potential for mycorrhizal interception of nutrient flow through soil deserves investigation.

142


Differences in Kinetics of Uptake Between Hyphae and Roots Mosse et al. (1973) suggested that mycorrhizal colonization might alter the threshold concentration from which plants were able to absorb P and consequently increase the effectiveness of absorption from the soil solution. This possibility was investigated using tomato {Lycopersicon esculentum) and cassava (Manihot sp.; Cress et al, 1979; Howeler et al, 1979). P uptake from solution by tomato roots was studied at concentrations between 1 and 100 |LIM KH2PO4. The lower part of this range corresponds realistically with the concentration in soil solutions, an important point if we are to use the results to help interpret mycorrhizal effects on plant growth in soil. Both non-sterile and axenically grown non-mycorrhizal roots were used for comparison with mycorrhizal roots and the initial internal P concentra-x tions of the roots used were fairly similar. This was achieved by growing mycorrhizal plants on Ca3(P04)2 and non-mycorrhizal plants on NaH2P04. Over the solution concentration range 1-20 |XM, the Vmax of the uptake system (based on uptake per unit fresh weight of roots) was not very different in mycorrhizal and non-mycorrhizal plants, but the lower X^ (1.6 as opposed to 4.0 |IM) indicated that the affinity of the uptake sites for P was higher in the mycorrhizal roots. Thomson et al. (1990a) determined the kinetics of uptake in germ tubes of Gigaspora margarita grown with different P supplies and concluded that, like roots and other fungi, there were two transport systems which could operate simultaneously. The high affinity System I had a Xm of 1.8-3.1 |LIM and a Vâx of 3.13.6 nmol mg protein"^ h~^. AH that can be concluded from the comparison of the two sets of data is that the affinity of System I in hyphae is similar to that of the colonized roots and it does not seem likely that the hyphae would influence the threshold concentration for absorption by mycorrhizal roots. It is possible to argue that if P uptake is limited by the rate of diffusion of ions through a depleted zone around a root or hypha, then uptake characteristics of the root, mycorrhiza or hypha are irrelevant to considerations of P uptake from the soil, except when they grow into microsites with localized high solution concentrations and before any depletion zones develop. The work with cassava (Yost and Fox, 1979) illustrates two points. This species appears to have a very high P requirement, coupled with a very inefficient P uptake system in the absence of mycorrhizal colonization. Despite this, cassava is well known for its growth on soils of low fertility and its efficiency of uptake is markedly increased when roots are colonized by mycorrhizal fungi. Thus the importance of mycorrhizal colonization to a particular species or variety of host plant (its responsiveness or dependency; see Chapter 4) may depend on the concentration of P in the soil, the relative affinities of the root and fungal systems for P, the characteristics of the root system in extending into the soil and also upon the P requirement of the host (Koide, 1991a).

Use of Sources of P Unavailable to Roots The suggestion has frequently been made that mycorrhizal roots might be able to exploit sources of P in soil not normally available to plants. These sources include relatively insoluble forms of Pi, such as rock phosphate (RP) and Fe and Al

143


phosphates, as well as sources of Po such as phytate. It has certainly been shown that growth of mycorrhizal plants does respond to the application of RP or tricalcium phosphate, whereas these fertilizers had little or no effect on the growth of non-mycorrhizal plants at the rates of application used (e.g. Murdoch et ah, 1967; Wibawa et ah, 1995). Similar results have been obtained following application of insoluble phosphate fertilizers for a variety of host plants, usually in soils of low pH. In most investigations, comparisons were made at one or two rates of fertilizer application and Pairunan et al. (1980) have suggested that the conclusion that mycorrhizal plants had increased access to the P supply was invalid. They showed that it is essential to compare growth over complete P response curves for both soluble and insoluble fertilizers. The curves for superphosphate and RP are of similar form (Fig. 5.6), but there are important quantitative differences between them. First, the amount of P as RP which has to be added to soil to achieve maximum growth of Trifolium subterraneum was about 40 times greater than the amount of P from superphosphate that was required both for mycorrhizal and nonmycorrhizal plants. Second, the maximum growth with RP was less than that with superphosphate, irrespective of mycorrhizal colonization. These differences apart, and assuming that the RP contained no traces of soluble P, nor indeed toxic substances, the results do suggest that there is no absolute difference in the availability of RP to mycorrhizal and non-mycorrhizal plants. Nevertheless, at moderate and realistic levels of application, and at any level of P equivalent to the superphosphate range (0-0.8 g P kg soiP^; see Fig. 5.6), mycorrhizal plants were more effective at extracting P from the fertilizer. The mechanisms underlying the increased uptake might depend upon hyphal exploitation of the soil volume. In addition, both synergistic action between mycorrhizas and P-solubilizing microorganisms (see above), and the possible excretion of H"^ or hydroxyacids by hyphae, would increase the availability of RP to mycorrhizal plants (Johnston, 1956; Johnston and Miller, 1959; Smith, 1980). As indicated above, there is no experimental support for production of chelating agents by

1-2 2-4 0 43-2 Phosphorus applied (g per pot)

86-4

Figure 5.6 The effect of phosphorus fertilization and inoculation with Glomus mosseae on dry weight of shoots of Trifolium subterraneum after 7 weeks' growth, (a) Superphosphate fertilizer; (b) C-grade rock phosphate: O, uninoculated control plants; A , plants inoculated with G. mosseae. From Pairunan et al. (1980), with permission.

144


hyphae of VA mycorrhizal fungi. However, reductions in, soil pH (up to 1.0 unit) in hyphal compartments has been shown concurrently with P depletion in a cambisol, when P was supplied as Ca(H2P04)2 and N as (NH4)2S04 (Li et al, 1991b; Fig. 5.7a,b). The mechanism by which the pH was reduced is likely to have been via the extrusion of H^, following use of ammonium by the hyphae (Raven and Smith, 1976; Smith, 1980; Smith and Smith, 1984), but other mechanisms such as increased CO2 production may also have contributed (Li et ah, 1991c). In addition to studies specifically on the use of insoluble phosphate fertilizers, there has been considerable interest in the ability of mycorrhizal fungi to access P in the non-labile fraction in soil. This question has been addressed by assuming that only the labile P would exchange with added ^^P and that it was the pool from which non-mycorrhizal plants absorbed P. Although mycorrhizal onions, rye grass and soybeans took up more total P from soil with the labile fraction labelled than did non-mycorrhizal plants, there was no difference in specific activity of P in the two groups of plants. These results suggest that mycorrhizal plants had no access to non-labile P sources, which would have diluted the ^^P and lowered the specific activity (Sanders and Tinker, 1971; Hayman and Mosse, 1972; Mosse et al, 1973; Powell, 1975a; Pichot and Binh, 1976; Gianinazzi-Pearson et al, 1981a). In contrast, mycorrhizal potatoes apparently did take up fixed P from the latosols on which they were grown, the evidence being that the specific activity of "^^P in mycorrhizal (a)

Net (30 /im)

Membrane (0-45 ;/m)

(b)

0

5 10 15 20 25 30 35 40 45 Distance from root plane (mm)

50

Figure 5.7 Depletion profiles of: (a) H20-extractable P; and (b) soil pH in the hyphal and bulk soil compartments (refer to Fig. 5.3b) from non-mycorrhizal ( Q or Q ) and mycorrhizal ( # or • ) Trifolium repenSy grown in a cambisol. Bars represent standard errors of means. From Li et al. (1991b), with permission.


145

plants was higher than in uncolonized plants when they were absorbing ^^P from the fixed P fraction (Swaminathan, 1979). Furthermore, heating soil to provide a range of concentrations of fixed P gave no evidence of significant difference between mycorrhizal and non-mycorrhizal plants in accessing P in the different fractions (Barrow et ah, 1977). This confused picture probably results from the assumptions made about the labelling of the different P fractions in soil (Bolan et al, 1984b; Bolan, 1991) and a careful re-examination of the problem has shown the need for more work in this area. Bolan et al. (1984b) added iron hydroxide to soil, with the expectation that P 'fixation' as iron phosphates would take place. The P was labelled by the addition of ^^P. Subsequent extraction with 10 mM CaCl2, 0.5 M NaHCOa or acid NH4F was used to determine the amount and specific activities of P in different fractions. Addition of iron hydroxide reduced the amount of P available to non-mycorrhizal plants, but made no difference to mycorrhizal plants of Trifolium subterraneum. However, despite the fact that addition of iron hydroxide reduced the amount of P that could be extracted by CaCl2 or NaHCOs (but not by NH4F) there were no differences in the specific activities of any of the extracts nor between mycorrhizal and non-mycorrhizal plants. The conclusion must be that labelling techniques of this sort are of doubtful usefulness and results cannot eliminate the possibility that mycorrhizal plants obtain P that is unavailable to non-mycorrhizal plants. There has not been much work on the ability of enzymes from roots or VA mycorrhizal fungi to hydrolyse Po. Mosse and Phillips (1971) found that phytates were satisfactory sources of P for plant and fungal growth in agar cultures, and that calcium phytate stimulated fungal growth. It has also been shown that both mycorrhizal and non-mycorrhizal soybeans could hydrolyse calcium phytate in soil, but that there was no differential growth response (Gianinazzi-Pearson ef al, 1981a). Such an effect might be mediated by increased acid phosphatase activity in the rhizosphere, as shown by Dodd et al (1987), although its significance is still doubtful and has not been observed in all investigations (Allen et al., 1981a; Krishna and Bagyaraj, 1982). More recently, '^^P labelling techniques have been applied in pots compartmented by mesh (Joner, 1994; Joner and Jakobsen, 1994) and again there was no evidence for direct effects of mycorrhizal hyphae on solubilization of Po; rather, the hyphae increased the competitive ability of the plants for P mobilized by other soil organisms. Phosphate Metabolism in the Fungus External hyphae of VA mycorrhizal fungi must absorb Pi by active transport after the roots have become colonized. They have an active H^-ATPase on the plasma membrane which would be capable of generating the required proton-motive force (Lei et al., 1991) to drive H'^-phosphate co-transport, and P is certainly accumulated to high concentrations. Rapid absorption of P by mycorrhizal fungi is followed by the synthesis of inorganic polyphosphate in the fungal vacuoles. The granules observed following staining with toluidine blue and in the electron microscope (Cox and Tinker, 1976; Callow et al, 1978) are probably artefacts of the preparation methods, but polyphosphate is certainly important in mycorrhizal P metabolism (see Chapter 14). Recent work using freeze-substitution techniques has shown that in the ectomy-

146


corrhizal fungus Pisolithus tinctorius, and probably also in many other fungi, the polyphosphate is present as soluble, short chain 15-mers, stabilized by K"^ ions (Orlovich and Ashford, 1993; Ashford et ah, 1994). Synthesis of polyphosphate prevents excessive accumulation of orthophosphate in fungal cells, when external supply is plentiful and uptake is rapid. Such a storage role is common in microorganisms and may be important in reducing osmotic stress; it is found in ectomycorrhizas, in rusts and other obligate fungal parasites, in lichens, in many unicellular algae and in some free-living fungi and in bacteria (Harold, 1966; Beever and Bums, 1980). In VA mycorrhizas formation of polyphosphates might account for the increased P concentrations in the root (Sanders, 1975; Smith and Daft, 1977; Smith, 1982), which are associated with mycorrhizal colonization. Polyphosphate accumulation might also constitute an energy store but this is not likely to be large (Beever and Bums, 1980). Polyphosphate is probably also important in translocation of P in the hyphae, although not in granule form, as originally envisaged by Cox and Tinker (1976). The system of pleiomorphic, vacuolar tubules observed in many fungi contains polyphosphate and pulsation of the tubules results in transfer of their contents over short distances. Transfer by this mechanism over the long distances which would be required in external mycelium of VA mycorrhizas is still an open question (Ashford and Orlovich, 1994). Calculations of P content based on numbers of (artefactual) granules and amount of P per granule are probably still valid. Combined with rates of cytoplasmic streaming they indicate that movement of polyphosphate within the hyphae might account for measured fluxes into the root (Cox and Tinker, 1976; Cox et ah, 1980). Enzymes of polyphosphate metabolism have been detected in VA mycorrhizas by Capaccio and Callow (1982). The distribution is consistent with synthesis of polyphosphate in external hyphae and breakdown in hyphae within the root. Furthermore, the absence of polyphosphatases in external hyphae is interesting, as it suggests that polyphosphate turnover and unloading from tubules might be low. This might have consequences for long-distance translocation, but much more work on control of polyphosphate metabolism is required before this could be regarded as certain or even likely (see Chapter 14). Other information on P metabolism by the fungi still does not give a very coherent picture. 'Mycorrhiza-specific' alkaline phosphatases have been detected cytochemically in the vacuoles of the mature arbuscules and intercellular hyphae (Gianinazzi et ah, 1979). There is some circumstantial evidence that development of activity is linked to the presence of arbuscules and transfer of P to the plant (Gianinazzi-Pearson and Gianinazzi, 1978), leading to the suggestion that this enzyme might be a useful marker for efficient P metabolism in the fungi. Furthermore, it has been claimed that activity develops only in structures within the host tissues. Activity in germ tubes was restricted to the tip region and there was apparently a lag in development of activity within the root, which did not occur with succinate dehydrogenase activity (Tisserant et al., 1993). However, these observations are at odds with those of Dodd (1994) who suggested that alkaline phosphatase activity might be an effective vital marker for the external mycelium. Furthermore, Larsen et al. (1996) have shown that while the fungicide benlate inhibits P uptake and transfer to plants via mycorrhizal hyphae, it does not affect alkaline phosphatase activity, so that alkaline phosphatase activity was not a


147

suitable physiological marker in this instance. The actual role of the alkaline phosphatase in P metabolism of mycorrhizas has yet to be determined. Membrane-bound H'^-ATPases in the plant-fungus interfaces are likely to be important in transfer processes between the symbionts, because these enzymes are essential for generation of proton-motive force, which drives H^ co-transport of solutes across membranes. TTie electrochemical potential gradient for orthophosphate between the soil solution or the mycorrhizal interface and either hyphae or root cells is such that uptake is always active and the distribution of H'^-ATPase on the plant plasma membrane surrounding the arbuscules indicates that this membrane is likely to be energized and capable of the necessary transport of P (or other actively transported ions) to the plant. There is further discussion of the roles of HÂTPases in different interfaces in Chapter 14.

Nitrogen Nutrition Mycorrhizal Effects on Modulation and M Fixation Increased N concentrations have been reported in VA mycorrhizal plants. Of course, where they are also symbiotic with N-fixing bacteria or actinomycetes this can be attributed to increased rates of N fixation induced secondarily, e.g. by increased P uptake, rather than to direct uptake of N compounds from the soil. The assimilation of N fixation of dinitrogen in rhizobial root nodules is certainly increased when plants, growing in low-phosphate soils are also colonized by mycorrhizal fungi. This effect was probably first observed by Asai (1944) who made detailed observations of growth, nodulation and mycorrhizal status of a large number of legumes. More recently, nodulation and N fixation by mycorrhizal and non-mycorrhizal legumes, as well as plants nodulated by Frankia, have been the subject of many experiments (e.g. Rose, 1980; Rose and Youngberg, 1981; Gardner, 1986). In most cases, improved nodulation and N fixation in mycorrhizal plants appears to be the result of relief from P stress and possibly uptake of some essential micronutrients, which result in both a general improvement in growth and indirect effects upon the N-fixing system. The differences between mycorrhizal and non-mycorrhizal plants usually disappear if the latter are supplied with a readily available P source. More detailed information can be obtained from the many reviews of this area (e.g. Bowen and Smith, 1981; Barea and Azcon-Aguilar, 1983; Bethlenfalvay and Newton, 1991; Azcon-Aguilar and Barea, 1992; Barea et ah, 1992; Bethlenfalvay, 1992b). The fact that the effects are not directly attributable to the mycorrhizal fungi themselves should not be allowed to detract from the interest of these tripartite symbioses, which may be very important both in natural ecosystems and in revegetation programmes, where nutrients are in short supply in soil (Lamont, 1984; Pate, 1994). Uptake of Mineral Nitrogen (NH4 and NO3") The N concentration in plants is about ten times greater than the P concentration, emphasizing a much greater requirement for this nutrient. The availability of

148


mineral N in soils depends to a very great extent on the activities of microorganisms in mineralizing N from organic sources, effecting conversions such as nitrification and denitrification, and immobilizing N in the soil biomass. Major inputs derive from organic matter and from biological N fixation. In contrast to the situation with other mycorrhizal types, particularly ericaceous mycorrhizas (see Chapter 12), there is no unequivocal evidence that the extraradical hyphae of VA mycorrhizas are directly involved in mineralization of organic forms of N. However, there is increasing evidence that mycorrhizal fungi are involved in uptake and transfer of inorganic N, although it is not clear whether this always occurs in amounts that are significant for plant nutrition. The two most important sources of inorganic N for plants, and potentially for VA mycorrhizal fungi, are nitrate (NO^) and NH4 (NH4) ions. In agricultural soils NO^ usually predominates, because of the rapid nitrification of NH4. In contrast, in many undisturbed soils, particularly those which are on the acid side of neutrality, ammonium predominates and nitrate may be almost entirely absent (Rice and Pancholy, 1974; Stewart, 1991; Stewart et al, 1993). N O J is not adsorbed on soil colloids and is mobile, at least in moist soil, so that mass flow of the soil solution to roots absorbing both water and nutrients allows uptake to be maintained at rates dependent on the root absorbing power. Consequently, absorbing power and the surface area available for absorption (dependent on root length and radius) are likely to be the factors limiting uptake. A mycorrhizal effect would not be expected for this ion, except in dry soil when the mobility was reduced. In contrast, NH4 is adsorbed and is relatively non-mobile, so that even in moist soils depletion zones develop readily and, as with P, diffusion rather than root absorbing power, limits the rate of uptake and so mycorrhizal fungi might be important in increasing the rate of uptake. Preliminary results on N uptake and transfer were somewhat inconclusive. As early as 1976, Haines and Best reported that loss of NH4, NO3 and nitrite (NO2) from soil by leaching with water was retarded when plants of Liquidambar styraciflua were mycorrhizal with Glomus mosseae. Unfortunately, the root systems of the mycorrhizal plants were considerably larger than the uncolonized systems, so that the results did not indicate unequivocally that the mycorrhizal fungi themselves were involved. Subsequently, Ames et al. (1983) showed that although ^^N supplied to extraradical hyphae in organic form reached mycorrhizal Apium graveolens (celery), transfer required a considerable period of time. They assumed that mineralization by soil microflora was an essential step in making the organic N available and that this caused the delay. Increased inflow of N to roots of Trifolium subterraneum supplied with inorganic N was also observed (Smith et al., 1986d), but the results of different experiments were not consistent (Smith et ah, 1986c) and uptake of N by the hyphae did not seem to play an important role in supplying N to the plants because increased uptake per plant was not observed. The picture is now clearer and again the use of mesh-compartmented systems has played an important part. Ames et al. (1983), supplied (^^NH4)2S04 to the HC of such a system and observed considerable ^^N transfer to the mycorrhizal plants. The amount was correlated with the percentage colonization of the roots of the celery plants {Apium graveolens) by Glomus mosseae, with the hyphal length density in the HC and with the number of hyphal crossings of the mesh. Using the number of crossings, and again assuming that all these hyphae were alive.

149


Ames et ah (1983) were able to calculate a translocation flux for N of 7.42 X 10~^ mol N m~^ s~^. This is roughly the same as the flux of P through entry-point hyphae calculated by Sanders and Tinker (1973) and, considering the high N requirement of plants, the value seems low (see Table 5.3). A similar approach has now been used with different plant-fungus combinations and the data have confirmed hyphal ^^N transfer to mycorrhizal Cucumis sativus, Trifolium subterraneum and Zea mays when ^^NH4 was supplied in the HC colonized by hyphae of Glomus intraradices (Frey and Schuepp, 1992; Johansen et ah, 1992, 1993a). In two of these investigations the ^^N in the HC was significantly depleted (Fig. 5.8) in the hyphal compartment (Frey and Schuepp, 1992; Johansen et ah, 1992). However, none of these demonstrations of N translocation and transfer have been associated with increased plant N content or growth. Johansen et ah (1992) suggested that the small physical size of the experimental systems might have contributed to this. They proposed that where soil N is present as NH4, or strong competition exists for recently mineralized N, a mycorrhizal mycelium might play a significant part in N acquisition. Differences between species of fungi in accessing ^^N were found to be related to differences in distribution of hyphae in the HC (Frey and Schuepp, 1992). Additional evidence for the competitive effects of mycorrhizas in accessing less available forms of N has been obtained using N dilution techniques (Azcon-Aguilar et ah, 1993; Tobar et ah, 1994b). There has been only one investigation of mycorrhizal effects on ^^N-labelled NO^ (Tobar et ah, 1994a). Again using mesh-compartmented pots, no difference in ^^N enrichment between mycorrhizal {Glomus fasciculatum) and non-mycorrhizal plants of Lactuca sativa supplied with ^^N-labelled NO3 under well-watered conditions was observed. However, in dry soil the enrichment of the mycorrhizal plants was four times higher, probably reflecting the much lower mobility of NO3 in dry soil.

13

5

13

5

1 3 B

13

5

Distance from root compartment (cm) F i g u r e 5.8 Effect of mycorrhizal hyphae of Glomus intraradices on depletion of K C I extractable N H 4 and N O ^ in the hyphal c o m p a r t m e n t of plants of Cucumis sativus fertilized ( H C A ) , o r n o t ( H C B ) , w i t h added N O ^ . G l , mycorrhizal plants; N M , n o n mycorrhizal plants. Bars are standard e r r o r s of the means of f o u r replicates. From Johansen et al. (1992), w i t h permission.

150


Mycorrhizal effects on N nutrition have been studied under field conditions and the potential for increased uptake of N from soil, as well as the P-mediated effects on N fixation demonstrated in Hedysarum coronarium (Barea et al, 1987; Table 5.4). In mixed plantings a twofold increase in ^^N transfer from soybean to maize has been observed in mycorrhizal plots, together with a relative increase in productivity of maize (Hamel and Smith, 1991). This suggests that mycorrhizal fungi may be involved in redistribution of N in a plant community. The type of mycorrhiza formed by plants also appeared to have an influence on their capacity to use NO^ in a Banksia woodland. At recently burned sites N O ^ predominated over NH4 and was stored and used most effectively by non-mycorrhizal and VA mycorrhizal, herbaceous plants. Woody species, including some that could be hosts to both VA and ectomycorrhizal fimgi, appeared to use a wider range of N sources. There is also an indication that 6^^N values may differ in mycorrhizal and non-mycorrhizal plants, due to imknown fractionation processes (Handley et al, 1993), and this might account for the variability in 5^^N values in plants of different mycorrhizal status and life form in the Banksia woodland. It might also have contributed to the problems encoimtered by Hamel et al. (1991) using natural abundance methods to elucidate the mycorrhizal interactions between soybean and maize (see Chapter 15). Both absorption and assimilation of inorganic N by the fungi are prerequisites to translocation and transfer to the plants. As NH4 is frequently present at very low concentrations in soil it is thought that assimilation depends on the activity of glutamine synthetase (GS) and glutamate synthase (GOGAT), rather than glutamate dehydrogenase (GDH), because of the higher affinity of GS for NH4 (Miflin and Lea, 1976). GS activity is increased in mycorrhizal root systems, partly due to a contribution from the fimgi themselves because activity has been detected in fungal tissue separated from VA mycorrhizal roots. Improved P nutrition in the plants resulted in only a small increase in activity, confirming the important contribution of the fungi. In contrast, GDH activity showed no direct relationship with colonization (Smith et al, 1985). This limited evidence suggests that the fungi may have the capacit)^ to assimilate NH4 and in consequence N is likely to be transferred from fungus to plant in organic form. Nitrate reductase activity has been detected, albeit at very low levels, in isolated spores of Glomus mosseae and G. 'macrocarpus' (Ho and Trappe, 1975). Since most fungi that reduce N O ^ have been shown to have an NADP-dependent enzyme, Oliver et al. (1983) assayed NAD- and NADP-dependent activity in mycorrhizal roots of Trifolium subterraneum (80% colonized). They observed that although there was no NADP-dependent activity, NAD-dependent nitrate reductase activity was present. This enzyme, more characteristic of higher plants than of fungi, increased both in colonized plants and in plants receiving additional P. The effect was apparent in shoots as well as roots and the most likely explanation for the mycorrhizal effect is that increases in enzyme activity were the result of improved P nutrition. The results confirm those of Carling et al. (1978) who, however, only assayed nitrate reductase in non-mycorrhizal tissues of colonized and nonmycorrhizal soybeans, i.e. their nodules and shoots. The capacity of the fimgus in mycorrhizal roots to assimilate NO^ must remain in doubt and it seems likely that the main site of nitrate reductase is in the plant. An increase in N uptake in mycorrhizal roots (whether directly mediated by the


151

fungus or not) may be important, in order to compensate for the relative loss of absorbing area in mycorrhizal plants with reduced root:shoot ratios. The involvement of VA mycorrhizas in N transformations obviously needs more attention and the interactions between enzyme activity (in both plant and fungus) and P nutrition will need to be addres^sed in further experiments on the apparent i^N/i^N fractionation. Transfer of N between plants via mycorrhizal hyphae has attracted considerable attention and is discussed separately below.

Uptake of Other Nutrients Cu, Zn and Other Micronutrients After some years during which the importance of mycorrhizas in micronutrient uptake received only limited emphasis, there is now consistent evidence that the efficiency of uptake of both Zn and Cu is increased in VA mycorrhizal plants. Some of the earliest work on physiology of VA mycorrhizal plants showed an increase in concentration of Cu in mycorrhizal apple seedlings (Mosse, 1957); subsequently, similar results have been obtained in such diverse species as Zea mays (Daft ei al, 1975), Avena sativa (Gnekow and Marschner, 1989), Phaseolus vulgaris (Kucey and Janzen, 1987), Allium porrum (Gildon and Tinker, 1983) and Trifolium repens (Li et al, 1991c). Using a compartmented mesh system Li et al. (1991c), demonstrated hyphal uptake and translocation of Cu to T. repens. This not only contributed up to 62% of the total Cu uptake, but was also independent of the effects of P nutrition (see below). Increased Cu uptake in mycorrhizal plants has also been confirmed for a number of plant-fungus combinations (e.g. Manjunath and Habte, 1988; Killham and Firestone, 1983) although transfer from the fungus to the plant is often very small (see Manjunath and Habte, 1988) and the mechanisms controlling transfer have not been investigated. As with P, the mobility of Zn in soils is very low and its uptake by organisms is diffusion limited. Consequently, similar effects of mycorrhizal colonization on uptake were expected but proved difficult to demonstrate because of strong PZn interactions. It was shown at an early stage that VA mycorrhizal colonization increased uptake of ^^Zn by Araucaria roots (Bowen et al., 1974) and that ^^Zn was translocated along hyphae of Glomus mosseae into Trifolium repens growing in an agar plate system (Cooper and Tinker, 1978). The rate of ^^Zn translocation was 2.1 X 10~^ mol m~^ s~^, considerably lower than the rate of P translocation (Table 5.3), but probably adequate given the lower requirement of plants for this micronutrient. The tracer studies gave little quantitative information on the transfer of Zn to the plants and effects on Zn nutrition, but Zn deficiency symptoms in peach disappeared as mycorrhizas developed (Gilmore, 1971). Subsequently, a number of studies have shown unequivocally that Zn uptake via mycorrhizas is important and can alleviate Zn deficiency in several species in both pot and field experiments (e.g. Triticum aestivum, Thompson, 1990; Trifolium subterraneum, Burkert and Robson, 1994; Cajanus cajan, Wellings et al., 1991; Zea mays, Evans and Miller, 1988; Lu and Miller, 1989; Leuceana leucocephala, Manjunath and Habte, 1988; Faber et al, 1990). Vesicular-arbuscular mycorrhizal involvement in Zn nutrition

152


(a)

(b)

. ^ c3 ' ^ •Q.

X

J?'^ s £

si .21

5040-

T

^.-A^

30-

\

20-

4*

10-

Y / ^

0-

T

K X\ t-^ 1—

0

•

t

Ul

'1

UJ

10

20

30

40

65zn placement (mm from roots) Figure 5.9 Effect of placement of ^^Zn at different distances from the roots of nonmycorrhizal and mycorrhizal clover plants on the ^^Zn activity of the whole plants at (a) 21 days and (b) 35 days. # Non-mycorrhizal plants; / \ Acou/osporo laewis; O Qomus sp.; and • Scutellospora calospora. Bars indicate standard errors of means. From Bukert and Robson (1994), with permission.

has been implicated in the negative effects of both tillage (Evans and Miller, 1988; Fairchild and Miller, 1988) and long fallow (Thompson, 1987, 1990; Wellings et ah, 1991) on crop growth. Differences in the abilities of the same three VA mycorrhizal fungi used by Jakobsen et al. (1992a,b) to absorb and transfer ^^Zn to plants, have been demonstrated. Using a mesh-compartmented system and different distances of placement of ^^Zn source in the hyphal compartment, Acaulospora laevis obtained Zn from distances of up to 40 cm, whereas Glomus sp. (WUM 10(1)) and Scutellospora calospora (WUM 12(2)) were less effective. As with P uptake, the distribution and length density of hyphae in soil were important contributors to the differences between the fungi (Burkert and Robson, 1994; Fig. 5.9). Uptake of other micronutrients via VA mycorrhizal hyphae is not well established (Marschner and Dell, 1994) and the uptake of Mn is most commonly reduced when plants are mycorrhizal. This effect was attributable to lower Mn"^ reducing potential in the rhizosphere mycorrhizal plants, probably because the populations of microorganisms responsible were lower (Kothari et al,, 1991). Interactions between phosphate fertilization and deficiencies (or toxicities, see below) of trace elements, are well known in several species of typically mycorrhizal plants (Wallace et aL, 1978). In general, when the availability of P is increased, P uptake and plant growth also increase. Concentrations of Cu and Zn in the tissues fall, sometimes to levels at which deficiency symptoms become apparent. Mycorrhizal colonization has been shown to affect these interactions (Lambert et al., 1979; Timmer and Ley den, 1980), so that at moderate levels of phosphate fertilization


153

deficiency symptoms are alleviated as mycorrhizal fungi increase uptake of the trace elements and tissue concentrations rise. At very high levels of P, mycorrhizal colonization itself may be reduced (see Chapter 2) with consequent reductions in uptake and reappearance of the deficiency symptoms. Interactions such as these may be involved in the apparent alleviation of Zn toxicity in polluted sites (e.g. Dueck et ah, 1986). If the sites are P deficient then mycorrhizal P uptake could result in increased growth and dilution of Zn in the tissues.

Potassium Analyses of potassium (K) concentrations in plant tissues have occasionally ir\.dicated increases in K uptake, which might be expected considering the relative immobility of this ion in soil (Mosse, 1957; Holevas, 1966; Possingham and Groot-Obbink, 1971; Powell, 1975b; Huang et al, 1985). However, in the majority of investigations K was found to be at lower concentrations in the tissues of mycorrhizal plants than in those of non-mycorrhizal plants. Extrapolation from tissue concentrations can be dangerous, as we have seen for other nutrients, because of the simultaneous effects of P nutrition on growth. Smith et al, (1981) observed elevated concentrations of K in shoots (but not roots) of mycorrhizal Trifolium subterraneum when plants were grown on P-deficient soils. If sufficient P was supplied to soil to remove any mycorrhizal growth response, then K concentrations in both groups of plants were very similar. This suggests an indirect effect of mycorrhizas on PO^ uptake in P-deficient plants, an effect which has also been indicated with sulphate (Rhodes and Gerdemann, 1978c). However, K depletion in a hyphal compartment colonized by Glomus mosseae, and increased accumulation in associated mycorrhizal Agropyron repens has been observed (George et ah, 1992) so that direct effects are also possible. Accumulation of K is strongly influenced by the form of N available (NOs" or NH4), as well as by other cations, particularly Na"^. It might also be influenced by the synthesis and storage of polyphosphate (see Chapter 14) and carefully designed experiments to investigate the influence of mycorrhizal colonization on K nutrition need to take all these potentially confounding factors into account.

Toxic Elements Some essential elements are required in very small quantities and when accumulated at high concentrations in plants they may become toxic. Consequently, heavy metal toxicity may derive from excessive uptake of Zn, Cu, Fe and Co as well as from other elements and ions which are normally regarded as toxic (e.g. Pb, Cd, Ni, Ti, Ba). General aspects of the interactions between fungi, including mycorrhizal fungi, and these metals have been reviewed by Gadd (1993). Large effects of mycorrhizas in increasing accumulation of Cu, Ni, Pb and Zn in the grass Ehrhartia calycina have been found, especially at low soil p H (Killham and Firestone, 1983). However, El-Kherbawy et al. (1989) observed increased tolerances, at least at some soil p H values. Results are not always consistent between species: Rogers and Williams (1986) found marked increases in ^^^Cs in Melilotus officinalis, in which

154


no additional ^°Co was accumulated, whereas in Sorghum sudanense the converse was true. The possibility that metals may be sequestered in the hyphae and not transferred to the plant has been examined in mycorrhizal roots of Pteridium aquilinum. Electron energy loss spectroscopy showed greater accumulation of Cd, Ti and Ba in the fungal structures than in the root cells themselves. The plants had been growing on a site previously treated with Cd dust and were regarded as tolerant of the toxicity. It was suggested that sequestration of the metals by polyphosphate in the fungus might have been important in minimizing transfer to the plant (Tumau et ah, 1993) but this requires confirmation. The possibility that heavy metal accumulation might result in damage to the fimgi and consequently have significant negative effects on mycorrhizally mediated P or Zn uptake has also been addressed. It appears that prolonged exposure to Cd can result in the development of tolerance in Glomus sp., but again the mechanism is not known, Weissenhom et al. (1994). Arsenate tolerance in plants is most interesting because it appears to depend on modifications to the P transport system (which also transports arsenate) and hence the root absorbing capacity (Meharg, 1994). Tolerant genotypes of Hocus lanatus are more highly colonized by mycorrhizal fungi in the field than non-tolerant genotypes, leading to the suggestion that the tolerant plants are dependent on the fungi for P uptake and consequent success (Meharg et al., 1994). A number of questions still require clarification. If uptake of P and arsenate are diffusion limited it is not clear how modification of transport (root absorbing power) will actually affect the rates of uptake (see discussion on P uptake above). Nor is the mechanism of fungal tolerance to arsenate apparent. If the fungi also have modified P transporters, how can they take up P and compensate for the effects in roots? It is clear that interactions with P nutrition and other aspects of mycorrhizal physiology must be taken into account in studies of accumulation and tolerance of toxic elements. Tissue dilution of a toxic element can occur as a consequence of improved P nutrition and increased plant growth, even in situations where the uptake of the heavy metal per plant is actually greatly increased (El-Kherbawy et al., 1989). Only measurements of inflow or use of matched plants a n d / o r compartmented systems will determine whether the hyphae of VA mycorrhizal fungi are directly involved in uptake of heavy metals by plants.

Interplant Transfer of Nutrients Hyphal links between plants offer potential pathways for the movement of soilderived nutrients, just as they do for plant-derived C, and could play important roles in interplant and interspecies competition and redistribution of nutrients in ecosystems. Consequently, the technical problems in determining whether observed tracer fluxes of these nutrients between plants are likely to be nutritionally or ecologically significant need to be overcome (Newman, 1988; Miller and Allen, 1992; and see Chapter 4). The survival of mycorrhizal hyphae within senescent roots could lead to rapid transfer and redistribution of P released by autolysis, or by the activity of microorganisms, to plants linked by the mycelial network. Transfer of mineral nutrients between living plants is more problematic. There is certainly evidence for the


155

transfer of ^^P and ^^N from 'donor' to 'receiver' plants (Whittingham and Read, 1982; Ritz and Newman, 1984; van Kessel et al, 1985; Newman and Ritz, 1986; Haystead et al., 1988; Hamel and Smith, 1991), but in most cases the results did not show net movement of the nutrient in question. Francis et al. (1986) did not use tracers, but instead supplied either nutrient solution or water to one half of the root system of the donor plants (either Plantago lanceolata or Festuca ovina). Receivers were grown with the other half of the root system either inoculated with VA mycorrhizal fungi or uninoculated. Mycorrhizal receivers (P. lanceolata or F. ovina) responded positively in both growth and nutrient content when nutrients were supplied to the donors, but the non-host Arabis hirsuta and non-mycorrhizal P. lanceolata and F. ovina did not, apparently indicating net transfer of nutrients from the larger nutrient-sufficient donors to the receivers via the hyphal network. However, Newman (1988) has suggested that these results could be explained not by direct transfer but by changes in the competitiveness of the donor plants as a result of differences in nutrient supply to them, with large nutrientsufficient donors competing less strongly with the receivers. Newman's suggestion is supported by data of Ocampo (1986) indicating that competition could have been the dominant effect in nutrient redistribution between large and small plants of Sorghum vulgare. Eissenstat (1990) re-examined this question using both nutrient applications and ^^P and ^^N in a split pot system. By measuring the ratio of tracer in the receiver to tracer in the donor, he concluded that P transfer between individuals of Plantago lanceolata was increased following P fertilization, while N transfer was unaffected. In quantitative terms the transfer of P was too small to have any effect on plant growth, whereas the transfer of N was about tenfold higher, as would be expected from the relative requirements of plants for these nutrients, and could have had an effect on the receivers in very nutrientdeficient soils. His overall conclusion was that alteration in the competitive balance, rather than direct transfer, was the most important effect. This is supported by the results of Johansen et al. (1992), who found that although external hyphae of Glomus intraradices effectively depleted total N from soil in a hyphal compartment (HC) and transferred it to plants of Cucumis sativus, very little ^ N was transferred via the plant to a second hyphal compartment. The results do not indicate a major pathway for N transfer from plant to fungus across a symbiotic interface, which is what would be expected if the transfer systems in the interfaces are polarized for transfer of different nutrients in different directions (Smith and Smith, 1990; see Chapter 14). All in all, considerable uncertainty still surrounds the idea that competition between plants may be modified by direct net transfer of nutrients (P and N) from plant to plant via a common mycelium. The lack of convincing data underlines the fact that very carefully designed experiments are required in this area. In addition, short-cycling of nutrients from a dying plant to a linked living plant may be considerable and should not be overlooked in ecological studies. W a t e r Relations Mosse and Hayman (1971) observed that mycorrhizal onions did not wilt when transplanted, but that non-mycorrhizal plants did. Subsequently, several similar observations have been made (Busse and Ellis, 1985; Huang et al., 1985) and there is

156


no doubt that mycorrhizal colonization does affect the water relations of plants. As with other aspects of the physiology of mycorrhizal plants, it is relevant to distinguish direct effects of fungal colonization from indirect effects resulting from changes in plant size or P status. The subject is complex and there are many inconsistencies in the literature, not all of which can be easily explained (Fitter, 1988; Koide, 1993; and see Nelsen, 1987). The problem was first investigated systematically with soybean by Safir et ah (1971, 1972), who showed that mycorrhizal plants had lower resistances to water transport than uncolonized plants, and in this instance it appeared that most of the difference was attributable to changes in root resistance, since shoot resistances were small and did not differ in the two groups of plants. Safir et al. (1972) concluded that the effect was probably due to improved nutrition, because the differences could be eliminated if nutrients or fungicide were applied. Transpiration rates of mycorrhizal plants are generally higher than those for nonmycorrhizal plants (Allen et al, 1981b; Allen, 1982; Nelsen and Safir, 1982; Huang et al, 1985; Koide, 1985b; Fitter, 1988). As Koide (1993) shows, the discussions and inconsistencies centre on whether this increase is due to increased stomatal conductance or whether decreased resistance to water transport in the below-ground system is also important, as suggested by the data of Safir et al (1971, 1972). Levy and Krikun (1980) used Citrus jambhiri and a fungus similar to Glomus Jasciculatus', with growth and fertilizer conditions that permitted the comparison of mycorrhizal and non-mycorrhizal plants of similar size and growth rate. The major effect of mycorrhizal colonization was an increase in transpirational flux and stomatal conductance, both during stress and recovery. There were apparently no differences between mycorrhizal and non-mycorrhizal plants in terms of resistance of the root to water movement. Similar conclusions were reached with Trifolium pratense (Fitter, 1988), Helianthus annuus (Koide, 1985b), Leucaena leucocephala (Huang et al, 1985) and Bouteloua gracilis (Allen et al, 1981b; Allen, 1982). Again, transpiration rates were increased in mycorrhizal plants, while stomatal resistances were greatly reduced. Both Koide (1985b) and Fitter (1988) came to the conclusion that the high stomatal resistance in P deficient non-mycorrhizal plants was a nutritional effect and Figure 5.10 shows the good correlation between stomatal conductance and leaf P concentration in T. pratense, regardless of whether the differences were induced by P fertilization or mycorrhizal colonization. However, there have been suggestions that stomatal behaviour is influenced by the hormonal changes in the plant that certainly occur as a result of either changes in P nutrition or mycorrhizal colonization per se (e.g. Allen et al, 1980, 1982; Allen, 1982; Dixon et al, 1987; Baas and Kuiper, 1989; Danneberg et al, 1992; Druge and Schonbeck, 1993). In B. gracilis the changes in stomatal conductance could not be explained in terms of differences in gross anatomy or morphology and there were no changes in size of mesophyll cells, bundle sheath cells or stomata, nor was the stomatal density altered following mycorrhizal colonization. Resistance to water transport was not separated into contributions of root and shoot by Allen et al (1981b), but they did comment that increased branching of the roots in mycorrhizal plants could Pb to substantial increases in root surface area without changes in root biomass, and that this might reduce the root resistance to water uptake. However, Koide (1985b) measured the hydraulic resistance to water transport between soil and leaf, and between soil and stem below the leaf (root plus stem resistance) over a range of


157

transpiration rates in well watered soils. When mycorrhizal and non-mycorrhizal plants of the same size and root length were compared there were no differences in hydraulic properties and the effects of mycorrhizas appeared to be entirely explicable on the basis of decreased stomatal resistance. Increased stomatal conductivity and increased transpiration rates in Leucaena were accompanied by higher xylem pressure potentials and the stomata responded more rapidly to changes in humidity. The mycorrhizal plants lowered the water potential of the soil in the pots, indicating much higher water uptake than into the non-mycorrhizal plants (Huang et ah, 1985). The conclusion in this case was that higher water use in mycorrhizal plants was offset by increased C gain, because the stomata remained open for a greater part of the day. Large root systems and rapid stomatal response were also important in the overall water economy. The results of a number of investigations have, however, suggested that VA mycorrhizal colonization can reduce the hydraulic resistance to water uptake in the roots (e.g. Hardie and Leyton, 1981; Nelsen and Safir, 1982; Graham and Syvertsen, 1984; Bildusas et al., 1986). On a whole-plant basis this could occur by increases in the size or branching of the root system. However, Hardie and Leyton (1981) could attribute only part of the decrease in hydraulic resistance of the root systems of Trifolium pratense colonized by Glomus mosseae to this effect. They therefore concluded that hyphal growth in the soil was important in reducing root resistance to water flow and this point is discussed further below. However, Koide (1993) has emphasized that because root hydraulic conductivity is not linearly related to root system size (see Ficus and Markhart, 1979), it is not valid to compare values obtained from root systems of different sizes, even if the results 121

^ 10 I w E

£

8

®

o c

B § "D C

o o E

6 o

o

CO

Leaf P concentration (mg g~^)

F i g u r e 5.10 Relationship between stomatal conductance and tissue P concentration, compiled by Fitter (1988). Symbols represent Rosa ( A A ) ' HelianthuSy ( 0 # ) and Trifolium ( D B ) . Mycorrhizal plants, solid symbols and non-mycorrhizal plants, open symbols. Reprinted f r o m the Journal of Experimer)tal Botariy by permission of O x f o r d University Press.

158


are expressed per unit root length. Consequently, orUy data from matched mycorrhizal and non-mycorrhizal plants can be used for comparison and in these cases either no effects were observed (Graham and Syvertsen, 1984; Koide, 1985b; Graham et ah, 1987) or there was a decrease in hydraulic conductivity (Levy et ah, 1983). Koide (1985b) further noted that apparent effects of mycorrhizal colonization on the root hydraulic conductivity could have been recorded because this parameter varies with transpirational flux, which is normally higher in mycorrhizal plants. The possible roles of mycorrhizal hyphae in water uptake require discussion, despite the fact that there is little imequivocal evidence implicating them in water uptake and transport processes or in direct effects on water relations of colonized plants. Hardie (1985) studied transpirational flux in non-mycorrhizal and mycorrhizal Allium porrum and Trifolium pratense that were not nutrient limited. Mycorrhizal clover had slightly higher transpirational fluxes and lower stomatal resistances than the non-mycorrhizal plants. Removal of external hyphae did not affect the stomatal resistance but appeared to reduce the transpirational flux, although the differences were not significant. Hardie used the maximum reduction in transpirational flux caused by hyphal removal and transplanting (9.9 X 10~^ 1 m~^ root length s~^) to calculate that damage to 40 hyphal entry points per metre of colonized root would account for this reduction if they had the same water flux as Phycomyces blakesleeanus. This is a bold assumption, given the lack of the significance of the data. However, Faber et al, (1991) produced some evidence for hyphal depletion of soil water in a compartmented system and calculated an apparent water flux through the hyphae of 375 X 10~^ 1 h~^ per hypha crossing the mesh (or, using internal radii of hyphae of 5 |Lim, 1.32 1 m~^ s~^; see Table 5.3). Kothari et al. (1990) also examined the question using Zea mays. The difference in transpiration rates between mycorrhizal and non-mycorrhizal plants was considerable and could largely be attributed to increased leaf area. However, root lengths in mycorrhizal plants were reduced by 31% and in consequence the water uptake per unit length of root was much higher than in non-mycorrhizal plants (A 7.3 X 10~^ 1 m"^ s~^; a value similar to that found by Hardie, 1985). Assuming 60% colonization of the root system and 1000 entry points (of 10 |Lim diameter) per metre of colonized length (Kothari et al., 1990), the cross-sectional area of hyphae through which the water would enter can be estimated as 4.7 X 10~^ m^ m~^ and the flux of water as 0.15 1 m~^ s~^, which is an order of magnitude less than the estimate of Faber et al. (1991). Kothari et al. (1990) calculated that if P was delivered in the hyphae by mass flow of a solution of concentration 16 mmol 1~^ the inflow should have been 10 X 10~^° mol m"^ s~^. This is 1-2 order of magnitude higher than the actual inflow, which was in the range of 6.222.0 X 10~^ mol m~^ s~^. They concluded that the apparent bulk flow of water in hyphae did not seem to make an important contribution to P inflow (see also Sanders and Tinker, 1973; Cooper and Tinker, 1981). The conclusion that mass flow of solution in the hyphae did not occur to any great extent was supported by the fact that consumption of water from the hyphal compartments in their experiments was negligible. Furthermore, George et al. (1992) confirmed these findings and showed no difference in water depletion in a hyphal compartment, regardless of whether the plants were well watered or water stressed, whether or not the hyphae were cut.


159

Mass flow has been proposed as one of the possible mechanisms of translocation in fungi (Jennings, 1987). In a multihyphal strand or in a hypha growing apically and with all resource allocation polarized towards the apex, this idea could have some credibility. However, it is well established that mycorrhizal hyphae translocate different nutrients in opposite directions. Organic C moves from the plant to the external mycelium and supports its apical growth in soil. Conversely, mineral nutrients are absorbed from the soil and translocated towards the plant. Any mechanism of translocation or of mass flow of solution must take this into account. It seems highly unlikely that mass flow of solution through VA mycorrhizal hyphae towards the plants takes place at rates able to account for the apparent extra water uptake by mycorrhizal roots, while at the same time permitting C transfer in the opposite direction. It remains to be seen whether the pleiomorphic vacuolar tubules, which do seem to pulsate in opposite directions, move water as well as the solutes within them at rates adequate to support the measured fluxes. The current consensus is that hyphae do not play an important part in reducing the hydraulic resistance of roots and many workers would feel that there was no need to invoke this in any case, given the complexities of measurement and the dependence of hydraulic conductivity on the (variable) rate of transpiration and size and branching of the root systems. However, there are still those who consider that a non-nutritional effect of mycorrhizal colonization on water relations may be important and research in this area continues (Druge and Schonbeck, 1993; Ebel et flZ., 1994). Mycorrhizal colonization may have effects on drought tolerance that are not directly related to water relations. Nutrients, as we have seen, become less and less available as soil dries because of the increasing tortuosity of the diffusion path. Under these conditions growth of non-mycorrhizal plants is likely to be increasingly limited by nutrient availability, and reduced root growth would limit the accessibility of water. Under these conditions the hyphal contribution to uptake of nutrients would become more and more important.

Conclusions There is excellent evidence to demonstrate that external hyphae of VA mycorrhizal fungi absorb non-mobile nutrients (P, Zn, Cu) from soil and translocate them rapidly to the plants, thus overcoming problems of depletion in the rhizosphere, which arise as a consequence of uptake by roots. Transfer across the symbiotic interface results in increased nutrient acquisition by the plant. These processes act in series and, with efficient fungi, lead to depletion of nutrients in the soil well beyond the rhizosphere. The sources of P (and other nutrients) available to the fungi are less clear. The soil solution, in equilibrium with the so-called labile fraction, must be the primary source. Hyphae are able to penetrate soil pores inaccessible to roots and may also be able to compete effectively with soil-inhabiting microorganisms for recently mineralized nutrients. Rapid removal from solution at sites of release will accelerate the use of that K fraction that exchanges rapidly with the solution.

160


The evidence that mycorrhizal fungi lower the threshold concentration for uptake is slim. There is no good evidence that VA mycorrhizal fungi actually hydrolyse organic P or N sources. The use of non-labile inorganic sources or applied RP is still open to question. Mycorrhizal plants do seem to grow better and take up more P from these sources, but until methods are worked out which allow the P fractions in soil to be distinguished and their accessibility to plants (whether mycorrhizal or not) determined, little progress is to be expected in sorting out what mechanisms actually underlie the effects. At present it appears that localized alterations in pH might play a role in increased P mobilization in microsites. Production of chelating agents that would increase the availability of Fe or Al phosphates has not been demonstrated, despite the apparent differences in accessibility to mycorrhizal and non-mycorrhizal plants. The interactions between mycorrhizal colonization and accumulation of heavy metals and other toxic elements is an area of considerable interest. A number of different mechanisms may be involved, including tissue dilution of the toxic element due to interactions with P nutrition, sequestration of the toxic metal in the fimgus, and development of tolerance by the fungus. Water relations of plants are modified in some ways by the mycorrhizal interactions. The mechanisms are difficult to determine, but most of the effects can be related to changes in nutritional status. There is little evidence either for actual water transport via the fimgal hyphae or for alterations in root or shoot hydraulic properties or water potentials that are independent of increased P uptake or of changes in growth as a result of this.

Structure and development of ectomycorrhizal roots

Introduction The ectomycorrhizal root is characterized by the presence of three structural components: (a) a sheath or mantle of fungal tissue which encloses the root; (b) a labyrinthine inward growth of hyphae between the epidermal and cortical cells called the Hartig net; and (c) an outwardly growing system of hyphal elements which form essential connections both with the soil and with the fruit bodies of the fungi forming the ectomycorrhizas. In the absence, until recently, of fossil ectomycorrhizas, discussion of the origin of this type of symbiosis was based largely on conjecture. Fossil leaves and wood of families of plants, such as the Pinaceae, which are today almost exclusively ectomycorrhizal, appear for the first time in the Cretaceous around 130 million years BP (Axelrod, 1986) and molecular clock evidence (Berbee and Taylor, 1993) indicates an origin of the holobasidiomycete group, within which many of their current fungal associates occur, at around the same time. Trappe (1977) suggests that the disjunct distribution of many hypogeous fungi, which are dependent upon animals for dispersal, can best be explained by their presence in the Laurasian subcontinent before the land migration routes of the vectors between Europe and North America were separated by tectonic events some 50 million years BP. On the assumption, now widely accepted, that epigeous forms were ancestral to those of hypogeous habit, the mushroom-forming ectomycorrhizal fungi, and therefore the potential to form this type of the symbiosis, would have originated before that. Support for this contention, in the form of well-preserved ectomycorrhizas, has now been provided in fossils recovered from the approximately 50 million years BP Princeton cherts of British Columbia of the Eocene period (Le Page ei al., 1996). These provide the first unequivocal evidence of ectomycorrhizas in the fossil record. Small roots of a Pinus sp., each having a diameter of 3-5 mm, bear coralloid clusters of attenuated and thickened, dichotomously branched rootlets 0.1-0.5 mm in diameter (Fig. 6.1). Transverse sections reveal the presence of a Hartig net-like fungal tissue between the cells of the cortex up to the position of the endodermis (Fig. 6.1c). Only traces of mantle tissue remain and extraradical hyphae are scarce.

164

Ectomycorrhizas

'0

' ^^^

Figure 6.1 Fossil ectomycorrhizas of a Pinus sp. recovered from the Princeton cherts of Eocene age in British Columbia, (a) Coralloid clusters of ectomycorrhizal roots. Bar, I mm. (b) Individual rootlets showing the dichotomous branching characteristic of ftnus mycorrhizas. Bar, I mm. (c) Transverse section showing Hartig net-like fungal tissue (arrowed) surrounding root cells and showing evidence of labyrinthine structure. Bar, 20 ^ m . From Le Page et al. (1996), with permission. Copyright National Academy of Sciences, USA.

Those that are present lack clamp connections and have diameters of 1-2 ^lm. Among the small number of fungal associates of extant Pinus spp. that produce such coralloid clusters of rootlets and lack clamp connections are members of the hypogeous genus Rhizopogon. If, as proposed by Le Page et al. (1996), the fossil mycorrhizas of Pinus were formed by a Rhizopogon sp., these findings would confirm the view of Trappe (1987) that hypogeous basidiomycetes originated more than 50 million years BP and provide support for the molecular clock-based datmg of the origin of the ancestral epigeous forms, concurrently with that of the plant genus, around 130 million years BP.


165

Almost all of the plants upon which ectomycorrhizas develop are woody perennials. The anatomical structures of the sheath and the emanating mycelia are stable at least at the level of the fungal genus and are increasingly used to facilitate characterization of the ectomycorrhizas (Agerer, 1987-1993; Ingleby et al., 1990). This type of organ is also clearly distinguishable from all other types of mycorrhiza on the basis of the absence of intracellular penetration by the fungus. In the event of penetration of healthy root cells by a mycorrhiza-forming fungus, whether from the Hartig net or from the hyphae of the sheath, the structure is referred to as an 'ectendomycorrhiza' (see below and Chapter 10). The identity of the plant can influence the outcome. Some fungi, for example the ascomycete Wilcoxina mikolae, routinely produce ectendomycorrhizas on young plants of Finns and Larix in nursery soils while forming ectomycorrhizas on Ahies, Picea and Tsuga (Mikola, 1988; and see Chapter 10). However, most ectomycorrhizal fungi are capable of forming intracellular penetrations in senescent parts of the rdot axis, or when the nutrient balance of the association is disturbed. In these circumstances the fungus appears to be behaving in a weakly pathogenic manner. Associations of this kind were called 'pseudomycorrhiza' by Melin (1917) but the lack of precision of this term is strong justification for abandoning its use (Mikola, 1965; Harley, 1969). Many fungi that produce typical ectomycorrhizas on members of the Pinaceae and Fagaceae form extensive intracellular growths as well as Hartig net and sheath in certain ericaceous hosts (see Chapter 11). This type of colonization is recognized as being of the distinct arbutoid category. Intracellular penetration in the form of haustorium-like structures can sometimes be observed in ectomycorrhizal roots. These are produced by fungi of uncertain status that are associated with, but not involved in the formation of, the mycorrhizal mantle. Thus members of the Gomphidiaceae grow in mycorrhizal sheaths formed by Rhizopogon and Suillus spp., from which they penetrate the cortical cells of Pinus spp. (Agerer, 1990). Similarly, the ascomycete Leucoscypha leucotricha, a ubiquitous resident of sheaths formed by Lactarius subdulcis on Fagus sylvatica, does the same on that species producing 250-310 haustoria in every 2 mm of mycorrhizal root tip (Brand, 1991, 1992; Fig. 6.2). While the presence of the three structural elements signifies an ectomycorrhiza, there may be considerable variation in the extent to which Hartig net, sheath and extraradical mycelium are developed. Indeed, even structures that have only a patchy sheath as in some Asteraceae (Warcup, 1980), or that lack a Hartig net as in the roots of Pisonia grandis (Ashford and AUaway, 1982) and of Pinus spp. colonized by Tricholoma matsutake (Ogawa, 1985), have been referred to as 'ectomycorrhiza'. The danger in broadening the category to this extent is that relationships between structure and function established by study of 'typical' forms may break down. This is well illustrated in the case of the Tricholoma 'ectomycorrhiza' reported by Ogawa (1985) to be parasitic. It may be acceptable to include the association seen in Pisonia grandis as a true ectomycorrhiza because, despite the absence of a Hartig net, wall proliferations in the interface are apparently similar to those seen in some typical ectomycorrhizas. Cairney et al (1994) have shown that a basidiomycete forming 'ectomycorrhizas' on Pisonia formed a sheath and some intercellular penetration on Picea sitchensis, but not on Eucalyptus pilulans. In view of the diversity of niches and of plant-fungus combinations possible in

166

Ectomycorrhizas

4

•%.!•,,:,:: ::^r\;:^/

WW

Figure 6.2 Mycorrhizas of Lactarius subdulds on Fagus sylvaticOy showing a dual colonization with an ascomycete Leucoscypha. (a) Ascomycete hyphae (arrowed) penetrating the fungal mantle, (b) Ascomycete hyphae (arrowed) on the inner surface of a dissected mantle, (c) Haustorium of an ascomycete (h) and intercellular hypha (arrowed). From Brand (1992), with permission.

nature it is not surprising that a range of structural forms is found. However, it is important not to lose sight of the fact that the vast majority of associations defined as ectomycorrhizal conform to the basic pattern in which a sheath, Hartig net and some, if only seasonal, development of extraradical mycelium are all present, while intracellular penetration is lacking. These well-conserved features appear to have been favoured because they confer particular functional attributes which further distinguish them from mycorrhizal types that do not have all of these defining characteristics.


167

That is not to say that there is no diversity of form or function within the ectomycorrhizal category defined in this way. Indeed, schemes of classification based upon differences of mantle structure are now enabling us to characterize ectomycorrhizas, and in many cases to identify the fungal symbionts involved, in the absence of fruiting structures. These advances will greatly enhance the precision of analysis of field-collected material, enabling relationships between structural and functional aspects of biodiversity to be evaluated. Techniques of molecular fingerprinting of the fungi associated with individual mycorrhizal roots tips are also playing a greater and greater role in determining the fungal species composition in natural populations (Gardes et al, 1991a,b; Gardes and Bruns, 1993, 1996).

The Plants: Taxonomy and Geographic Occurrence While a relatively small number, probably around 3% (Meyer, 1973), of phanerogams (seed plants) are ectomycorrhizal, their global importance is greatly increased by their disproportionate occupancy of the terrestrial land surface and their economic value as the main producers of timber. Thus the Pinaceae, members of which form the major component of the vast boreal forests of the northern hemisphere, the Fagaceae, dominants of the northern temperate forests, and their counterparts in the temperate and subtropical regions of the southern hemisphere, the Myrtaceae, are all families of predominantly ectomycorrhizal species. Table 6.1 lists examples of families and genera within which ectomycorrhizal colonization has been reported. Some are exclusively ectomycorrhizal, but others may also form vesicular-arbuscular (VA) mycorrhizas and indeed this may be the typical mycorrhizal type for the taxon. In the tropics of south-east Asia the most important family of moist and monsoon forests, the Dipterocarpaceae, is almost exclusively made up of ectomycorrhizal species (Alexander and Hogberg, 1986; Smits, 1992). This habit is also found in certain leguminous plants of the tropics most notably in the subfamily Caesalpinioideae, members of which are characteristically not nodulated. Here, most of the tribe Amherstieae and some genera such as Afzelia, Intsia and Eperua in the tribe Detareae have this type of association (Alexander, 1989a,b). In drier savannah woodlands of the miombo type the prominent leguminous genera Brachystegia and Jubernaldia are also ectomycorrhizal (Hogberg, 1982; Hogberg and Pearce, 1986). In contrast the subfamilies Mimosoideae and Papilionoideae appear, with a very few exceptions that need to be confirmed, to be made up of species colonized by VA mycorrhizal fungi (Alexander, 1989a). There are occasional reports of 'ectomycorrhizal' colonization in species such as Acacia (Mimosoideae; Warcup, 1985; McGee, 1986) even though neither Hartig net nor sheath development are typical. As most reports suggest that such genera have VA mycorrhizas under natural conditions there seems little to be gained by describing them as being ectomycorrhizal. Some genera of shrubs and a very small number of herbaceous species of angiosperm are routinely found to be ectomycorrhizal. Of these, the shrubs Dryas (Rosaceae) and Helianthemum (Cistaceae) are of particular ecological importance. Amongst the herbaceous species the dicotyledonous herb Polygonum viviparum and the cyperaceous monocot Kobresia myosuroides have typical ectomycorrhizal short

Ectomycorrhizas

168

Table 6.1 Genera reported to contain at least one species on which ectomycorrhiza has been described. Family or subfamily

Genus

Family or subfamily

Genus

Aceraceae

B Acer

Epacridaceae

Astroloma

Betulaceae

B Alnus

Ericaceae

Arbutus Arctostaphylos

Betula

Caesalpinioideae

B Carpinus

ChirDQphila

B Corylus

Gaultheria

B Ostrya

Kalmia

B Ostryopsis

Ledum Leucothoe

B Afzelia

Rhododendror)

Aldina

Vacdnium

Anthonota Berlinia Brachystegia

Euphorbiaceae Fagaceae

Eperua

6 Fagus 6 Lithocarpus

B Julbernardia

6 Nothofagus

Monopetalanthus

B Pasania 6 Quercus

Paramacrolobium Swartzia

6 Trigonobalus

Tetraberlinia B Sambucus

Casuarinaceae

B Casuahna B Allocosuorina

Cistaceae

6 Helianthemum 6 Cistus

Cupressaceae

6 Castanea 6 Castar)opsis

B Gilbertiodendron B Intsia

Caprifoliaceae

Porar)thera

Gnetaceae Goodeniaceae

Gnetum Brur)onia* B Gooder)ia*

Hammamelidaceae Juglandaceae

Parrotia B Carya

B Cupressus

Engelhardtia B Juglar)s

B Juniperus

Pterocarya

Cyperaceae

6 Kobresia*

Mimosoideae

B Acacia

Dipterocarpaceae

6 Anisoptera

Myrtaceae

6 Angophora

Ralanocarpus

B CoWistemon

Cotylelobium

6 Campomones/o

Dipterocarpus

B Eucalyptus

Dryobalanops

B Leptospermum

Hopea

B Me/a/euco

Monotes

B Thstar)ia

6 Shoreo Valica Elaea^naceae

Shepherdia

Nyctaginaceae

B Neeo B Torrubia B Pisor)ia

169


Fannily or subfamily

Papilionoideae

Genus

Brachysema

Family or subfamily

Rhamnaceae

Daviesia

B Rhamnus

Dillwynia

Spyridium

Eutaxia

Trymalium

B Gompholobium B Hardenbergia

Rosaceae

B Crataegus

Kennedya

B Dryas

B Mirbelia

B Malus

B Oxylobium

B Prunus

Platylobium

B Pyrus

Pultenaea

B Rosa

B Viminaria Abies Cathaya

B Sorbus Salicaceae

Larix Picea B Pinus Pseudolarix B Pseudotsuga B Tsuga

Polygonaceae

B Platanus Coccoloba Polygonum'^

B Populus B Salix

B Cedrus Keteleeha

Chaembatia Cirocarpus

Jacksonia

Platanaceae

Cryptandra Pomaderris

Chohzema

Pinaceae

Genus

Sapotaceae Sterculiaceae

Glycoxylon B Lasiopetalum Thomasia

Stylidiaceae

B Stylidium

Thymeliaceae

B P/me//a

Tiliaceae

B T/7/0

Uapacaceae

B Uapaca

Ulmaceae

B U/mt/s

This list cannot pretend to be exhaustive but illustrates the wide range of families and genera of Angiospermae and Gynnnospermae in which ectomycorrhizas have been observed. A record of the presence of ectomycorrhizal individuals in a genus does not mean that all species are or may be ectomycorrhizal, nor does it mean ectomycorrhizal colonization is necessarily consistendy or even normally present in any species of that genus. Those marked * are herbaceous and those marked B may form both ecto- and VA mycorrhizas, with the latter in many cases being the most common mycorrhizal type observed. Modified from Harley and Smith (1983).

roots with sheath and Hartig net. Neither of these plants would normally be colonized by VA mycorrhizal fungi. One further category of woody plants is of interest because it shows the facultative ability to be either VA or ectomycorrhizal. Members of the Salicaceae fall into this group. Species in genera such as Salix and Populus, depending upon local circumstances, can be predominantly colonized by VA or ectomycorrhizal fungi. The nature of the mycorrhizal type first formed and the extent to which either type of colonization persists into the adult condition appears to depend upon local soil conditions: VA mycorrhizas are typical of Salix or Populus species growing on mineral- or nutrient-rich soils, while ectomycorrhizas predominate in organic soils. Several plants, including Eucalyptus (Lapeyrie and Chilvers, 1985; Chilvers

170

Ectomycorrhizas

et al., 1987) and Helianthemum (Read et al., 1977) may be VA mycorrhizal in the seedling stages, while Alnus is reported to be VA mycorrhizal under some circumstances and shows growth responses to inoculation with VA and ectomycorrhizal fungi (Fraga-Beddiar and Le Tacon, 1990; Jha ei al, 1993). The situation in the Pinaceae is less clear and the reports (Cazares and Smith, 1992, 1996) of colonization by vesicles and hyphae, with occasional arbuscules, are of as yet unknown significance.

The Fungi Forming Ectomycorrhizal Associations In contrast to the situation seen in the VA mycorrhizal symbiosis, a large number of fungal species have been recorded as forming ectomycorrhizas. The majority of these are in the basidiomycetes, but there is a significant representation from the ascomycetes including the so-called 'E-strain' fungi, which also form ectendomycorrhizas on Pinus and Larix (Laiho, 1965; Mikola, 1965). A few species of zygomycetous fungi in the genus Endogone are also known to form this type of association. Molina et al. (1992) estimate that between 5000 and 6000 species of fungi form ectomycorrhizas or ectendomycorrhizas. While most of these are of the epigeous types (around 4500) a number, perhaps up to one-quarter, are hypogeous. Members of the ascomycetes are particularly conspicuous in the latter group. Lists (e.g. Table 6.2) provide valuable pointers to the possible extent of mycorrhizal involvement in a given genus, but it must be emphasized that they are based largely upon observed associations between hosts and sporophores in the field, and that in only a small proportion of cases has the mycorrhizal status been confirmed. These observations do not, therefore, serve as substitutes for direct analysis of the nature of the relationships between the fungi and their hosts. These can only be established by a combination of experimental approaches amongst which tracing of connections from fruit body to plant root (Agerer, 1991a) and resynthesis of mycorrhizas under axenic conditions (Duddridge, 1987), together with molecular fingerprinting are recommended. Records of plant-sporocarp associations suggest that the majority of ectomycorrhizal fungi have a broad host range (Trappe, 1962; Molina et al., 1992; and see Table 6.2). It is certainly the case that some species, such as Amanita muscaria, Cenococcum geophilum, Hebeloma crustuliniforme, Laccaria laccata, Pisolithus tinctorius and Thelephora terrestris, have worldwide distribution on a very wide range of plants. This should not disguise the fact that specificity, at least at the level of the plant genus, can be recognized in a diverse range of fungi, hosts and habitats. Studies of the plant genera Eucalyptus (Malajczuk et al., 1982), Nothofagus (Garrido, 1988) and Pseudotsuga (Molina et al, 1992) show that each has an extensive assemblage of specific fungi. In the case of Pseudotsuga, Molina et al. (1992) estimate that there may be around 250 such genus-specific fungi. Once again, observations of this kind, which are almost exclusively based upon plant-sporocarp associations, need to be verified by synthesis of mycorrhizas. Such syntheses must, however, be established and interpreted in a sensitive manner because environmental and culture conditions can have important effects on the outcomes. The use of media containing large quantities of exogenous C, as originally recommended for synthesis of ectomycorrhizas (Molina and Palmer, 1982), can


171

lead to considerable distortion of the apparent host range of a fungus. Species such as Suillus grevillei which are confined to the plant genus Larix in nature, were shown (Duddridge, 1986a,b), in the presence of glucose, to form mycorrhizas on a number of other genera. Under similar circumstances, Molina and Trappe (1982b) observed that Rhizopogon vinicolor, the sporocarps of which, in nature, appear to be exclusively associated with Douglas fir, was induced to colonize species of Picea, Pinus and Tsuga. Clearly, syntheses should be carried out in such a way as to reflect the natural inoculum potential of the fungus, a situation that can best be achieved in the laboratory by enabling it to grow from a true host through natural substrates to the roots of the test plant. This provides a measure of 'ecological specificity' and reflects a situation which is perhaps as close as possible to that prevailing in nature. While ectomycorrhizal fungi can be categorized as having either a narrow or a broad host range, the root system of an individual tree, for example of Eucalyptus, Pinus or Pseudotsuga, will normally be colonized by several members of each category. The factors that may have favoured the selection of this pattern of cooccurrence in forest ecosystems is discussed in chapter 15. Furthermore, successions of fungi - both on individual roots and in communities - have been repeatedly described and are discussed in a later section. The taxonomic status of fungi forming ectomycorrhizas is normally described at the level of the genus or species, but there is increasing recognition of the extent of diversity, both structural and functional, seen within the fungal species. Understanding of the genetic basis of this variability is now being sought, not only because of the intrinsic interest in the factors which determine plasticity, but also because the potential exists for manipulation of the genome and selection of functional attributes which are likely to be favourable to the hosts.

Genetics of Ectomycorrhizal Fungi There has long been an awareness of the extent of interspecific variability between ectomycorrhizal fungi in the structure and function of the mycorrhizas that they form. Recent studies have demonstrated, however, that in some of the most widely occurring fungi such as Pisolithus tinctorius (Lamhamedi ei al, 1990; Lamhamedi and Fortin, 1991; Burgess et al, 1994,1995), Laccaria hicolor (Kropp et al, 1987; Wong et al, 1989, 1990; Wong and Fortin, 1990) and Hebeloma cylindrosporum (Debaud et al, 1986; Marmeisse et al, 1992a) the magnitude of intraspecific variability can be as great as that between species. Most basidiomycetes are heterothallic, with very complex genetic control of mating, there being several thousand mating types determined by 1-4 multi-allelic loci (Aa, A(3, Ba, Bp; Kiies and Casselton, 1992; Casselton and Kiies, 1994; Debaud et al, 1995). A sexually sterile monokaryotic mycelium is produced from germination of a basidiospore during the formation of which meiosis has occurred. In mycorrhizal basidiomycetes, such monokaryons are normally unable to produce fully developed ectomycorrhizas (see below) but the mating of two monokaryotic mycelia enables the resulting dikaryon both to produce fruit bodies and to form mycorrhizas. Two mating systems are recognized. In bipolar forms, which constitute approximately 25% of heterothallic species, compatibility between homokaryons is controlled by multiple alleles at a single locus, A. The remaining 75% are

172

Ectomycorrhizas

T a b l e 6.2 Examples of ectomycorrhizal fungal species w i t h little host restriction (broad host range) by fruiting habit, class, family and genus Habit, class, family

Genus

Species

Epigeous habit Basidiomycotina Amanitaceae

Amanita

Astraeaceae Boletaceae

Astraeus Boletus

aspera, fulva, gemmata, inaurata, muscaria, pantherina, phalloides, rubescens, solitaha, spissa, strobiliformis, vaginata, verna, virosa hygrometricus, pteridus appendiculatus, calopus, edulis, erythropus, luridus, minatJoolivaceus, pulverulentus, regius castaneus, cyanescens rhodoxanthus ravenelii chromapes, felleus, gracilis, porphyrosporus armeniacus, badius, chrysenteron, rubellus, spadiceus, subtomentosus, truncatus dbarius, infundibuliformis, tubiformis aurea, botrytis, flava, formosa, mairei, subbotrytis atrovirens sublutea byssinium, croceum, sulphureum acutus, anomalus, bicolor, bivelus, everneus, hemitrichus, leucophanes, mucosus, multiformis, obtusus, phrygianus, saniosus anthracina, cinr)amomea, malicoria, palustris, phoenicea crustuliniforme, cyliridrosporum, hiemale, lor)gicaudum, mesophaeum, mir)us, pumilum, siriapizans asterospora, bongardii, brur)nea, cincinr)ata, dulcamara, fastigiata, jurana, lacera, lanuginella, petJgir)osa, terrigena, umbrina caperata repar)dum velutinum rufescer)s capreolarius, comarophyllus, chrysodor), discoideus, hypothejus, karstenii, marzulus, pudorinus ir)volutus cristatus decipiens, fuligiriosus, helvus, necator, piperatus, repraesentar)eus, rufus, scrobiculatus, spinosulus, uvidus, vellereus, volemus aerugmea, albonigra, amoeria, anthracina, cyanoxantha, densifolia, emetica, foetens, heterophylla, lutea, nigricans, ochroleuca, odorata, oHvacea, paludosa, palumbina, parazurea, vesca, virescens, xerampelina

Gyroporus Phylloporus Pulveroboletus Tylopilus Xerocomus Cantharellaceae Clavariaceae Corticiaceae

Cortinariaceae

Cantharellus Ramaria Byssocorticium Byssoporia Piloderma Cortinarius

Dermocybe Hebeloma

Inocybe

Hygrophoraceae

Rozites DentJnum Hydnellum Hydnum Hygrophorus

Paxillaceae Polyporaceae Russuiaceae

Paxillus Albatrellus Lactahus

Hydnaceae

Russula


Habit, class, family

Genus

173

Species

Epigeous habit Basidiomycotina (contd) Sclerodermataceae Strobi lomycetaceae Thelephoraceae Tricholomataceae

Scleroderma

bovista, cepa, citrinum, laeve, polyrhizum, verrucosum

Pisolithus

tinaorius

Boletellus

betula, chrysenteroides

Strobilomyces

floccopus

Thelephora

ar)thocephala, atrocitrina, per)icillata, terrestris

Sarcodon

imbricatus, scabrosus

Laccaria

amethystina, bicolor, laccata, montana, proximo

Tricholoma

caligatum, columbetta, flavobrur)neum, flavovirer)s, myomyces, sapor)aceum, sulphureum

Hypogeous habit Ascomycotina Balsamiaceae

Balsamia

magnata, platyspora, vulgaris

Elaphomycetaceae

Cenococcum"^

geophilum

Elaphomyces

ar)thracir)us, grar)ulatus, muricatus, mutabilis,

Ger)abea

cerebriformis

reticulatus, variegatus Geneaceae

Ger)ea

gardneri, harknessii, intermedia

Helvellaceae

Hydriotrya

tulasr)ei

Pezizaceae

Pachyphloeus

citrir)us, ligericus, melanoxanthus

Terfeziaceae

Choiromyces

alveolatus, venosus

Tuberaceae

Tuber

aestivum, borchii, brumale, califorr)icum, excavatum, melanosporum, puberulum, rapaeodorum, rufum

Basidiomycotina Cortinariaceae

Hymenogaster

bulliardii, calosporus, citrinus, decorus, HIacmus, luteus,

Hysterangiaceae

Hysterangium

membrar)aceum

Leucogastraceae

Leucogaster

r)udus

Melanogastraceae

Melanogaster

ambiguus, broomeiarius, euryspermus, intermedius,

Russulaceae

Basmomyces

mattiroHar)us

Zelleromyces

stephensii

olivaceus, populetorum, tener, vulgaris

tuberiformis, variegatus

Sclerodermataceae

Scleroderma

hypogaeum

Strobi lomycetaceae

Gautieria

graveolens, mexicana, otthii

Endogorie

lactiflua

Zygomycotina Endogonaceae

Data modified from Molina et al. (1992). * Taxonomic status uncertain.

tetrapolar species in which there are two unlinked mating-type loci, A and B, again with multiple alleles. In the tetrapolar system, homokaryons are compatible with each other when they have different alleles at both mating-type loci. Mycorrhizal basidiomycetes are much harder to work with than the saprophytes that have been used to reveal these genetic systems. However, mating types of the bipolar kind have been identified in three Suillus species (Fries and Neumann, 1990; Fries and Sun, 1992), while H. cylindrosporum (Debaud et al., 1986) and Pisolithus tinctorius (Lamhamedi et

174

Ectomycorrhizas

Mycorrhiza

2

P. tinctorius Roots

(a)

^

Figure 6.3 (a) Cell wall polypeptides in ectomycorrhizas formed between Eucalyptus globulus and Pisolthus tinaorius 441 and in non-colonized Eucalyptus roots and cultured Pisolithus tinaorius 441. Densitograms of l-D SDS-PAGE from cell wall proteins (CWP). Note the increased accumulation of a fungal band (32-kDa CWP) in mycorrhizas (arrowed). Data of De Carvalho and Martin, unpublished. From Martin and Tagu (1995), with permission, (b) Changes in gene expression during the formation of ectomycorrhizas between Eucalyptus globulus and Pisolithus tinctorius 441. Free-living mycelium ( ^ ) and ectomycorrhizas ( • ) . From Tagu et al. (1993), with permission.

al, 1990) are of the tetrapolar type. Pairings between monokaryons of H. cylindrosporum that were derived from the progeny of six wild dikaryotic strains of disjunct geographical distribution, have demonstrated the occurrence of multiple alleles at the A and B mating-type loci in this species (Debaud et al, 1986). Within the genus Laccaria a complex pattern of mating systems is revealed. The most important ectomycorrhizal species L. amethystea, L. hicolor, L. laccata and L. proxima have a tetrapolar mating system (Fries and Mueller, 1984), but one in which all four mating types are rarely found in the progeny of a dikaryon. Doudrick et al (1990) showed the presence of a large number of alleles in L. laccata var. moellerl By pairing isolates obtained from different regions of North America, they estimated the outbreeding efficiency of this system to be 88%. However, there is also evidence that in L. laccata, genes other than those determining mating types can restrict pairings between homokaryons from morphologically similar strains. Thus Fries (1983) found two incompatible groups of the species in a restricted area of Sweden, and Mueller (1991) detected three such groups in North America that were also incompatible with the isolates of Fries. Laccaria bicolor also shows evidence of incompatibility groups (Kropp and Fortin, 1988; Doudrick and Anderson, 1989). Such analyses highlight the inadequacy of our understanding of the 'species' as a unit, and emphasize the need to characterize and describe the origin of isolates used in any experimental study Intraspecific genetic variation can be expressed at the physiological level in the form of differences in growth, production of enzymes or of auxins (Gay and Debaud, 1987), or at the level of mycorrhizal infectivity or aggressiveness (Wong


• u

175

2

n

r ^ ^ T3

-

2J

o c

u

&

4

n

0) 0)

rin

n

10

3J

"^-1 2J

ML, 100

I

a

200 400 300 lAA synthesizing activity

Figure 6.4 Specific indole acetic acid (lAA) synthesizing activity of Hebeloma. (a) Different Hebeloma species, (b) Different wild strains of H. cylindrosporum. (c) Sib-monokaryons being the progeny of the HCI dikaryotic strain of H. cylindrosporum. (d) The dikaryons synthesized from all possible fusions between these monokaryons. Specific activity expressed as nmol lAA synthesized mg~' protein h~'. From Gay and Debaud (1987), with permission.

and Fortin, 1990; Burgess et al, 1994, 1995). Gay and Debaud (1987) measured production of indole acetic acid (lAA) by different species of Hebeloma and compared the observed rates with those of wild strains, of monokaryons produced by germination of spores from a single fruit body (sib-monokaryons) and of dikaryons synthesized from the monokaryons of H. cylindrosporum (Fig. 6.4). Variation of auxin production was as large in the intraspecific strains as between species. Analyses of differences in growth and of enzymes such as glutamate dehydrogenase (Wagner et a/., 1988), nitrate reductase (Wagner et al, 1989) and acid phosphatase (Meysselle et al, 1991) reveal relatively broad variability (Table 6.3). Similar differences in expression of the last enzyme have been observed in progenies of controlled dikaryons of L. laccata (Kropp, 1990).

o •ex c

£ "53 -Q

X DC C

a N 1-

O u

>^ E o u

c

-2 »2

E

:±

(/)

>» — CO T3

0)

O

c O

o CM

O

T3

-

u

"O ^_^

c o rt

ro - ^ Q. i

^

f03 "tj

o 8

0)

i 5 r5

&0 £5

I 5

o

t

1^

I

0)

•s i = E

ii

°° O^ —

00

I

—

^

>o ^ m (N

ON 00

1*

o

"I

1^

P

c^

"^

—

NO

o

o O

-

r^ I

vO

— I

—

—

LO

°

— I

— o^

Q)

ti

c

U

O

^ > S S

•J

t-j

*^

E

O

(U

^

C

0) a> c Q- eg «/> O 0) Z

»-

3 &0

0) Q.

C

Q . u)

> 4-*

^

O 1_

L.

*>« 2

I

O O

O

vO

O

00

00

_

—

o

S .E

S =

:

O

in

o

oi,

+''

Ectendo

Hyphal

ND

ND

Hartig net

ND

ND

ND

ND

Hartig net

ND

ND

complexes Arbutoid

Hyphal complexes

NA, not applicable; ND, no data available; -^, ATPase activity, demonstrated cytochemically, always present; (+), ATPase activity not always present *, Vanadate-sensitiv€5; **, DES-sensitive.

rhizas are reported to be relatively long-lived but more work on these associations is needed. In the mycorrhizas of orchids the fungus forms physiologically active coils in the cells which later degenerate. The cell of the host outlives the fungus and may be recolonized (Chapter 13). Ecto and monotropoid-mycorrhizal roots appear to have greater longevity than other mycorrhizal types (Chapters 2 and 6). However, not only are there distinct stages during development which exhibit different cellular interactions in the interface, but many root systems are also differentiated into long and short roots which have differing life spans and, in the most extreme examples, the short roots persist for months only This diversity of structure and development (see Table 14.1) offers the potential for differentiation of function, both temporally and spatially in a root system. Any of the stages of development or degeneration may be involved in transfer of nutrients and might contribute to this specialization. The exact location of the symbiotic interfaces within the tissues of the root may also be relevant to the mechanisms and controls of nutrient transfer. Localization of the Fungus in an Apoplastic Exchange Compartment Despite the variations in structure and development, the interfaces at the cellular level have the same basic structure (Fig. 14.6). In all instances the fungus actually colonizes the root apoplast and remains strictly outside the plant protoplast. The fungus may be located in the intercellular spaces of the root, or closely associated with the walls of the epidermis and cortical cells, or it may penetrate intracellularly and create a new apoplastic compartment between the fungal cell wall and the

Uptake, translocation and transfer of nutrients in mycorrhizal symbioses INTERFACIAL '^-jmrri-z^ APOPLAST

FUNGUS

efflux

395

PLANT

influx

mineral nutrients P, N, Zn, S etc

—-•

influx

efflux 'carbohydrate

FPM

FW

PPM

Figure 14.6 A highly simplified diagrann of a symbiotic interface, indicating the key components including the interfacial apoplast bounded by the fungal plasma membrane (FPM) and plant plasma membrane (PPM). The arrows indicate membrane transport processes involved in transfer of carbohydrate and soil-derived nutrients. Dashed lines, efflux; solid lines, influx. Wall of the fungus is indicated by shading, but plant wall and walllike components not shown.

plant plasma membrane (Smith and Smith, 1990). In the mature state, cells of both symbionts are alive and the interface is therefore delimited by two plasma membranes, one from the plant and one from the fungus, separated by an interfacial zone or apoplast. There are, of course, differences in detail between different types of interface, particularly in the wall components and other materials found in the interfacial apoplast, as well as variations in membrane activity deduced from structure and ATPase activity. However, in all cases the fungus actually colonizes the apoplast and the environmental conditions there will be very important for all aspects of fungal development and also for nutrient transfer. Indeed, the need for an apoplastic step is recognized as a means of exercising control at the membrane transport level in phloem unloading and in the transfer of sugars to fruits and seeds (e.g. Humphreys, 1988; and see Patrick, 1989). Transfer of sugars from the plant to the fungus is obviously similar to these processes in many ways and the transfer of mineral nutrients needs similar control steps operating in the opposite direction. The interfacial zone normally includes the wall of the fungal hypha which may or may not show some structural modification and/or reduction, and also material derived from the activity of the plasma membrane of the host. In ectomycorrhizas and in the intercellular interfaces of other kinds of mycorrhizas, the interfacial apoplast may consist of a relatively unmodified host wall and a more or less modified middle lamella; or, it may be extremely modified to form a specialized contact layer referred to as the 'involving layer' (see Chapter 6). In the interfaces formed within cells there is a matrix between the plasma membrane of the host and the fungal wall which under the electron microscope is unlike the plant cell wall. At the base, where the trunk of the arbuscule or

396

General themes

penetrating haustorium or hypha passes through the wall of the host, it is usual to find that a layer of host wall, the collar, encloses the hyphae. This wall layer becomes progressively thinner away from the peripheral cell wall and is very thin or absent over the surface of most of the intracellular fungus. Instead, an interfacial kyer, often containing vesicles and fibres derived from the host plasma membrane, may be present. It appears as though the plant maintains some ability to secrete carbohydrate and polymerize it into fibres, but is unable to organize them into recognizable wall. This ability may be regained as the fungus senesces and becomes encapsulated in a thick fibrous layer. How far the wall and wall-like materials are involved in or influence transfer of nutrients across the interface is not at all clear. In the interfaces where transfer is assumed to occur there is no real evidence as to the permeability of wall-like materials, although the assumption must be that, like primary cellulose walls of plants and chitinous walls of fungi, they are permeable to the solutes that pass from one organism to the other. Impermeable structures, like the 'neckbands' associated with the haustoria of some biotrophic fungal pathogens, have rarely been observed (Smith and Smith, 1990). However, an electron-dense ring occurs in monotropoid mycorrhizas where the membranous sac bursts through the 'peg' (see Fig. 11.24) and at the point of entry of ericoid hyphae into cells of the root. These neckbandlike structures may be important in isolating part of the plant-fungus interface, permitting tighter control of conditions in the interfacial apoplast, but their effectiveness in sealing the fungus into an intracellular exchange compartment has not been tested. It could be significant that both these examples involve situations where the fimgus colonizes the epidermis of the plant and is thus topographically outside the hypodermis (see below). The apoplastic phase of the fungal sheath of some ectomycorrhizas is impermeable, and hence must be a barrier to nutrient movement between soil and root (Ashford et al, 1988,1989). Plate 6 shows the way in which entry of cellufluor to the roots of Eucalyptus pilularis is prevented by the fungal mantle of Pisolithus tinctorius (Plate 6a) unless the rootlets were deliberately damaged to remove the outer, unwettable region of the mantle (Plate 6b). Sections to which cellufluor was applied directly (Plate 6c) showed the characteristic turquoise fluorescence of this dye. The presence of the impermeable layer means that all solutes reach the root cells via the fungal symplast of the sheath, first by translocation in the external mycelium and subsequently by efflux to the interfacial apoplast in the Hartig net region. Conversely, solutes from the root cells effluxing to the apoplast must pass to the fungal symplast and cannot 'leak' to the soil via the apoplast of the fungal sheath. The impermeable layers thus offer an opportimity for control of conditions and solute concentrations in the cortical apoplast and Hartig net region, where transport between the ectomycorrhizal symbionts must occur. As Ashford et al. (1989) point out, maximum efficiency requires that material must not be allowed to escape from the 'exchange compartment' (Fig. 14.7). The initial work on apoplastic impermeability was carried out with Pisonia grandis, in the mycorrhizas of which there is no Hartig net. A similar, but slightly differently organized exchange compartment could exist where the Hartig net penetrates as far as the outer layer cortical cells; in this case the inner boundary would be the endodermis, which again provides a block to apoplastic transfer. In this context it is relevant that a fungal hydrophobin gene is up-regulated during ectomycorrhizal development (Martin

Uptake, translocation and transfer of nutrients in mycorrhizal symbioses

wall

outer apoplastic barrier sheath or hypodermis

397

inner apoplastic barrier hypodemriis or endodermis exchange compartment

cytoplasi

fungus in soil

fungus in root (intracellular or intercellular)

plant cell associated with intracellular fungus or Hartig

net

Figure 14.7 Diagramnnatic representation of the proposed apoplastic exchange compartments in mycorrhizal roots (see text), v, Vacuolar compartments; light-shaded areas, cytoplasm (symplast); dark-shaded areas, apoplast; black areas, impermeable apoplastic barriers. Transport of phosphate (P) and sugar (C) through the apoplast (dashed lines and arrows) and symplast (solid lines and arrows).

and Tagu, 1995; and see Chapter 6). The products of this gene are likely to play a significant role in cementing together the fungal hyphae and forming an impermeable and unwettable layer. The fungal sheath has not been reported to be impermeable in all investigations and the discrepancies may relate to different plant-fungus combinations as well as to different experimental methods. The external mycelium of some species of fungi is also covered with non-wettable material (Unestam, 1991). It is not yet clear whether this is important in long distance translocation within the rhizomorphs or whether it might play a role in preventing desiccation as has been suggested for the aerial hyphae of some saprophytic fungi, or indeed whether both these attributes are important. In mycorrhizas without a fungal mantle, any control of apoplastic conditions in the root cortex must lie with the root cells themselves. Here, the suberized cell walls in a well developed hypodermal layer beneath the epidermis may play the same role as an impermeable sheath in ectomycorrhizas. This hypodermis has been known for at least a century (e.g. Janse, 1897) and its widespread occurrence and significance has been emphasized in recent reviews (Peterson, 1988). The tangential walls of the hypodermis become increasingly impermeable as the root matures, essentially insulating the cortex from the soil. Entry (or exit) of solutes must occur in very young regions, or through the symplast of non-suberized 'passage cells', or via the hyphae of mycorrhizal fungi, which themselves enter via the passage cells (Peterson, 1987; Smith et al, 1989). Thus the fungus does not breach the suberized layers and may very well be exploiting the mechanism by which the non-mycorrhizal roots control the solute concentrations and conditions in the cortical apoplast. Again, the apoplast of the root cortex may function as a shared 'exchange compartment'.

398

General themes

Membrane Transport in the Interfaces The first step in transfer across a symbiotic interface must be efflux from the 'donor' organism into the interfacial apoplast. It is important here to correct the impression given by Harley and Smith (1983) that increased permeability, or 'leakiness', of a non-specific sort could be important in such mycorrhizal transfer systems. It is much more likely that increased rates of loss of individual solutes occurs to support overall rates of transfer between the symbionts, and that this is achieved by separate control of specific carriers or channels (Tester et al., 1992; and see above). No channels have yet been found, but techniques that will make the search easier are now becoming available. Conditions in the.apoplast that would promote efflux must be sought. It seems likely that mechanisms that are involved in phloem unloading in other tissues, could operate in the mycorrhizal exchange compartments, for loss of sucrose from the plant. This would require only a change in the tissues involved in expression of the relevant transport proteins. Mechanisms that might be involved in high efflux of P, Zn or amino acids from the fungus are unknown. Efflux must be followed by uptake from the interfacial apoplast by the 'receiver' organism. Membranes involved in uptake by proton co-transport must be energized by an H^-ATPase to generate the necessary PMF (Michelet and Boutry, 1995). The distribution of H^-ATPases on the membranes in the interfaces may therefore provide information on which membranes are important in uptake from the apoplastic compartments (Table 14.1). The distribution of H'^-ATPases in VA mycorrhizas has been determined cytochemically by Pb deposition (Fig. 14.8). The results of Marx et al. (1982) suggested that active ATPases were present on both plant and fungal membranes in the arbuscular interface and thus provided support for bidirectional transfer of nutrients (Fig. 14.9a). Later work has thrown some doubt on this interpretation. Some of the ATP-hydrolysing enzymes have been identified as H^-ATPases but others are probably non-specific phosphatases (Gianinazzi-Pearson et al., 1991a; see Fig. 14.8). Important points to note are that when non-specific phosphatases are inhibited with Mn^"*^, the periarbuscular membrane of the plant and the fungal plasma membrane of the intercellular hyphae retain consistent high ATPase activity. In both cases the opposed membranes have weak or absent activity (Table 14.1), which suggests that spatial separation of transfer functions may occur (Fig. 14.9b,c). The suggestion is that sugar is transferred from plant to fungus in the intercellular interface. Here, sucrose would efflux from the cortical cells of the root, be hydrolysed to hexose, which would then be actively absorbed by the fungus. In the arbuscular interface it is the plant that has an active H'^-ATPase. Here, efflux of P (or Zn, or organic N, etc.) to the apoplast would be followed by active uptake by the plant. The absence of H"^-ATPase activity on one of the membranes in each interface suggests a mechanism whereby reabsorption of solutes by the 'donor' cell (plant or fungus, depending on which interface) might be reduced, thus polarizing transfer across the interface as a whole. Similar methods have been applied to orchid and ectomycorrhizas, although work with the latter system did not involve inhibitors (Serrigny and Dexheimer, 1985; Lei and Dexheimer, 1988; and see Table 14.1). In these symbioses the interfaces were

399


B

:• \ s i * -

V ^^.^^

- . • /• / ' •

. — ^

y

..

^^,u

•

,

r .

Figure 14.8 Transmission electron micrographs of details of the mycorrhizal interaction between Allium cepa and Glomus intraradices. Sections of roots fixed and stained to demonstrate ATPase activity by the deposition of Pb phosphate which appears as fine electron-dense precipitates. (A), (B) and (C) No inhibitors; (D) in the presence of molybdate, which inhibits non-specific phosphatases but not H'*'-ATPases. (A) Extraradical hypha (h) at the root surface; epidermal cell (ec). (B) Intercellular hypha, in the intercellular space (is) between host cortical cells (he). (C) Arbuscular trunk (at) hypha in a cortical cell (he). (D) Fine arbuscular branches (ah) within a cortical cell (he) surrounded by periarbuscular membrane (RAM). Arrows in all parts of the figure indicate the presence of ATPase activity on fungal (A)-(C) and plant (D) membranes. Bars, O.Sjiim. The plate is reproduced from Smith and Smith (1996b), with permission. Origins of figures: (A)-(C), V. GianlnazziPearson, unpublished; (D), from Gianinazzi-Pearson et al, (1991a).

400

General themes

characterized by ATPase activity in opposed membranes of single interfaces. Data are not available for other mycorrhizal types, as indicated by the gaps in Table 14.1. What must emerge from a discussion of nutrient transfer between the symbionts is a picture of considerable complexity (see also Smith and Smith, 1990). For each solute molecule or ion transferred there must be a separate membrane protein, whether it be a carrier or pump or channel (Bush, 1993). This point is made forcibly in Figure 14.1 which shows a schematic representation of transport processes operating in fungi. Similar processes operate in plants, and in both organisms modifications to promote transfer between the symbionts are probable. Even if the complexities are initially intimidating. Fig. 14.1 at least emphasizes that the view of mycorrhizal transport systems as simple phosphate-sugar exchange must be unrealistic. Melin and Nilsson (1950-1958) found that P compounds, cations including Ca^"^ and N compounds derived from NH4 or glutamate are transferred from the substrate to the host via the fungus, and that ^^C-labelled photosynthates pass from the host to the fungus. Other observers have shown that ^^C-labelled compounds may also pass back across the interface into the host, a process which could have considerable significance in interplant transfer of organic C, if the transfer is shown to occur in nutritionally significant quantities. These general findings have been confirmed using a variety of plant-fungus combinations. In the early studies there were no real clues on the identity of the solutes moving across the interface, but there is now more information, especially for ectomycorrhizas. Organic C probably effluxes from the plant to the interfacial apoplast as sucrose, and data supporting this have been obtained for ectomycorrhizas. It is envisaged that sucrose is then hydrolysed by an acid invertase of plant origin and the hexoses resulting are absorbed by the fimgus. The way in which the transfer might be controlled by the pH of the apoplast and the concentrations of fructose has been discussed (see Chapter 7; and Fig. 7.3) and at this stage the hypotheses must be regarded as speculative, pending the availability of data on apoplastic conditions. NMR spectroscopy has shown that ^^C-glucose fed to VA mycorrhizal roots does pass to the fungus (Shachar-Hill et al., 1995; and see Chapter 4). In intact systems sucrose probably effluxes from the plant and is hydrolysed in the apoplast, so that Figure 14.9 Diagrammatic representation of the possible spatial distribution of H"*"ATPases and associated transport processes in VA mycorrhizal interfaces, based on data for molybdate-sensitive ATPase activity. (A) Bidirectional transfer of sugar and PO^ across the same arbuscular interface. Both plant and fungal membranes are shown with H'^-ATPase activity. (B) Transport of PO^ from fungus to plant across an arbuscular interface. Absence of H"*'-ATPase on the fungal plasma membrane (Fig. I4.8d) could be associated with passive loss from hyphae to interfacial apoplast, and presence with active uptake on the plant plasma membrane. (C) Transport of sugar from plant to fungus across the interface between cortical parenchyma cells and intercellular hypha. Little or no activity of H"ÂTPase on the plant plasma membranes could be associated with passive loss of sugar to the interfacial (intercellular) apoplast and presence on the fungal membranes vîth active uptake into the hyphae (Fig. 14.8c). fpm. Fungal plasma membrane; pam, periarbuscular membrane; rpm, plasma membrane of root cortical cell. From Gianinazzi-Pearson et 0/. (1991a), with permission.


401

the uptake of glucose by the fungus represents only part of the process and data on the fates of fructose and sucrose are required. A mechanism involving sucrose efflux and hydrolysis seems likely because plants possess the genetic information that results in specialized sucrose efflux in some tissues. Phloem unloading, for example, involves high rates of sucrose loss from cells to apoplastic compartments where, as in ectomycorrhizal systems, the operation of invertase helps to maintain a concentration gradient favouring further efflux (e.g. Humphreys, 1988). What is needed in a mycorrhizal system is expression of the relevant genes for sucrose

(A)

Soil

Arbuscular interface fpm pann

Fungus

fpm

- Sucrose or hexose

Hexose "

Fungal sugars H*

lATPasel

t

i

•P:

(B) Soil

t

i

lATPasel

lATPasel

t

i

^P:

- ^ PTranslocation

Arbuscular interface fpm pam

Fungus

fpm

Root

lATPasel

lATPasel

i

4^1

PT-

(C)

Root

•PT

Translocation'

Soil

Intercellular Interface

Fungus fpm

fpm

Hexose -4 i H^

Fungal sugars

i H^ lATPasel

F i g u r e 14.9 (Caption opposite)

"^^^

rpm Sucrose or hexose

402

General themes

transporters in cells associated, either inter- or intra-cellularly, with the fungal symbiont (Patrick, 1989). Uptake of hexoses by the fungal symbionts is presumably active, involving proton co-transport. The presence of H'^-ATPase activity on the fungal hyphae of the Hartig net region of Pinus sylvestris-Laccaria laccata ectomycorrhizas (Lei and Dexheimer, 1988) and in the intercellular hyphal interfaces of VA mycorrhizas formed between Allium and Glomus species, would be consistent with these being the sites of hexose absorption. However, ATPase activity on the fungal membranes in the arbuscules is patchy at best (Gianinazzi-Pearson et al., 1991a; and see Fig. 14.8), so that it is unlikely that they play a continuing role in uptake of sugars (Smith and Smith, 1995). More evidence on the localization of different ATPases and sugar transporters in the interfaces, and of their activity, is urgently required. In symbioses like ectomycorrhizas in which a single interface is presumably involved in bidirectional transfer, mechanisms which essentially maintain polarized transfer of different solutes need to be sought. For sugar transfer there is evidence for the operation of a ^biochemical valve' (Lewis and Harley, 1965c; Smith et al., 1969). The first step is the hydrolysis of sucrose by a plant invertase in the apoplast, which maintains sucrose efflux (see above and Chapter 7). This is followed by absorption of hexoses by the fungus and conversion to forn\s such as mannitol and glycogen, which are unavailable to the plant. Thus not only does the synthesis of fungal metabolites maintain a concentration gradient in favour of continued uptake, but any reversal of transfer is also prevented. If any solute did move back to the apoplast it would not be absorbed by the plant, but would accumulate and rapidly dissipate any concentration gradient and prevent further passive efflux. With respect to movement of soil-derived nutrients from fungus to plant, there are data for P and for N. Harley and Loughman (1963), through short-term labelling experiments with excised roots of Fagus, showed that orthophosphate passed from fungus to host and it is generally assumed (with little or no experimental data to support the contention) that the same applies to all mycorrhizal systems in which FO4 transfer takes place. Again, the electrochemical potential difference between the hyphae (Hartig net, hyphal complexes, arbuscules and so on) and apoplast will favour passive efflux from the fungus. Despite this, efflux of P from cells, including fungi (see above) is usually much lower than the measured rates of POJ transfer from fungus to plant in VA mycorrhizas. Mechanisms which promote efflux, and at the same time reduce retrieval or reabsorption, seem likely and the low expression of the fungal PO^ transporter (GvPT) within roots, compared with external mycelium (Harrison and van Buuren, 1996; and see above), is consistent with this suggestion. Data for rates of transfer across the symbiotic interfaces of other types of mycorrhizas are urgently required. There is little doubt that in ectomycorrhizas assimilation of N (either inorganic or organic) by the fungus is followed by transfer of organic N to the plant. In Fagus mycorrhizas, the important role of glutamine in this process was indicated the incorporation into it of ^^C, following dark fixation of ^^€©2 in the presence of NH4 and rapid transfer to the plant. The enzymic pathways show that there is a glutamine-glutamate shuttle which results in the net transfer across the interface of one N per glutamine, which is then transferred to a-ketoglutarate in the plant with the formation of glutamate (see Chapter 8 and Fig. 8.5a). In other plant-fungus


403

combinations, different distributions of enzymes involved in N assimilation give somewhat different pictures (see Fig. 8.5b,c). In Abies mycorrhizas there is also a glutamine-glutamate shuttle, with the N of glutamine being used in the formation of asparagine from aspartate in the root cells. Alanine is also transferred from fungus to plant, with the C skeletons required for a-ketoglutarate synthesis in the fungus being derived from sugars transferred from the plant. In mycorrhizas involving P. tinctorius a simpler picture emerges, with glutamine again the major organic N molecule transferred. However, as all the enzymes for organic N conversions are present in the fungus, there is in this case no cycling of N across the interface and the C skeletons are provided by sugars from the plant. The transfer of glutamine is important because its use does not perturb intracellular pH (Raven and Smith, 1976; Smith, 1980; Smith and Smith, 1990). In the mycorrhizal context this means that all export of H"^ that would be required to maintain intracellular pH during NH4 assimilation would be from the fungus, which has a large surface area in contact with the soil solution. If large amounts of H"^ were generated in the plant, export to the soil would have to be via the symplast of the fungal sheath. Figure 14.10 illustrates movements of K^ and H"^ that would be expected to be transferred across the interface concurrently with organic N in order to maintain charge balance across each membrane. Of course, charge balance also needs to be maintained during transfer of Pi, Zn or Cu. Given that we now believe that polyphosphate is stabilized by K^, it seems logical to propose that both Pi and K might be transferred to the plant, thus maintaining charge balance in both symbionts. The Rates of Transfer There is surprisingly little information on the flux of solutes across the interface. Calculations based on the uptake of P via the fungus and the area of interface available for transfer have only been made for VA mycorrhizas. Values for this flux are in the range 3-30 nmol m~^ s~^ (see Chapter 5), which suggests that abnormally high efflux from the fungus may be required (Smith et ah, 1994a,b; and see above). Data are required for other mycorrhizal types and other nutrients. Bidirectional Nutrient Transfer: Does it Occur at the Same Interface? Simultaneous bidirectional transfer of molecules or ions between the symbionts can only be envisaged when the interface separates two living cells. In mycorrhizas where only one type of interface exists (see Table 14.1) it is likely that the interface and component membranes are modified to promote this transfer. However, where there are two or more different interfaces they could individually be involved in unidirectional transfer, with the combined outcome of bidirectional transfer at the level of the whole mycorrhizal root (Smith and Smith, 1990, 1995, 1996a,b,c; Gianinazzi-Pearson et ah, 1991a; and see above). This idea of spatial separation of function has some attractions for the membrane transport processes that may operate to achieve polarized transfer. What must be emphasized is that whether transfer is bidirectional or unidirectional it is controlled by the two interfacial membranes. Moreover, transfer of any solute will involve efflux from one symbiont

404

General themes

{a)

(b) Soil

fp

Fungus

fp rp

Root

Soil

fp

Fungus

fp

rp

Root Sugar

Sugar

*
2H*4—t

i-^

2H*

2H*4—•

2H*

i

l-j

i-^

aKG^Gin

t—4

•-

^

.^—\

NH4

Glu" i

N

f _y1

1—4

aKG' 2Gln-

A

2NH/ •

Glu'

K*

->

K*

-¥

Soil

fp

Fungus

2H* 4 -

fp

2H* i

rp

Root

\

^

Soil

fp

1—t

2Glu'

•

J

Sugar

Gin

2K*

NH4.

•> 2K*

1

•-

rp

Root

•

^

-

4—\—•Gin

^

Glu" aKG^

NH4* — t

• - ^

2H* i—\

4-Nt

aKG' Glu"

1

fp

IAI

^

NH4^

Fungus

f

Gin

NH/

Gin

(d)

IA

2K*

2Glu"
Gln-J

t—IN

2Glu-

•-^

Sugar4-

f

Sugar

crKG^ 2H* 4 -

r^

Sugar 4

f

- Sugar

F i g u r e 14.10 Possible mechanisms f o r compartmentation of NH4-assimilating enzymes in mycorrhizas (compare w i t h Fig. 8.6). A , interfacial apoplast; G l u ~ , glutamate; G i n , glutamine; a K G 2 ~ , «-ketoglutarate. (a) Glutamine synthetase (GS) in the fungus and glutamate synthase ( G O G A T ) in the roots, w i t h net gain of glutamate in the roots, (b) GS in the fungus and G O G A T in the r o o t , w i t h net gain of glutamine in the r o o t (c) glutamate dehydrogenase ( G D H ) and GS in the fungus and G O G A T in the r o o t , w i t h net gain of glutamine in the r o o t , (d) G D H and GS in the fungus w i t h net gain of glutamine in the r o o t . N e t fluxes of N H 4 , H"^ and K^ are shown and give charge balance across each membrane. In (c) fluxes of N H 4 and H"*^ associated w i t h different parts of the pathway are ( f o r clarity) s h o w n separately at the plasma membrane of the fungus, but it is n o t implied that m o r e than one transporter f o r each ion is required there. Active and passive fluxes are not distinguished. From Smith and Smith (1990), w i t h permission.

into the interfacial apoplast and uptake by the other symbiont (Fig. 14.6). We need to consider whether these processes are modified in symbiosis and if so, how. The concentrations of the molecules transferred will be important, as will conditions such as membrane PD, p H and so on, which could markedly affect membrane function. Furthermore, mechanisms that reduce retrieval of solutes from the apoplast by the donor symbiont are likely to play an important role in polarizing


405

transfer. One of these is the 'biochemical valve' discussed above; others are the altered distribution of H^-ATPases and the low expression of GvPT, the fungal high affinity P transporter, in fungal structures within the root, compared with extraradical mycelium (Harrison and van Buuren, 1996). In situations when one or other symbiont is senescent or dead, only unidirectional transfer can be envisaged. For example, in the pre-Hartig net zone of ectomycorrhizas the cells of the root cap and epidermis are crushed and destroyed so that the fungus might derive material from the dying cells, but transport in the reverse direction to the host would be very unlikely. Similarly, the endophytes of Calluna colonize the mucigel layer of the root surface before penetrating the tissues and may derive C from this source. Conversely, nutrients may be transferred from fungus to root cells following hyphal or arbuscular collapse in orchid, ericoid and VA mycorrhizas, but the amounts transferred would be restricted to the contents of the collapsed hj^hae and, as already mentioned (Chapter 5) are probably insufficient to account for the measured rates of transport. The spatial separation of transfer functions does provide opportunities for differences in efficiency of the symbioses either at different stages of development or in different plant-fungus combinations, depending on the relative abundance and activity of the intercellular and arbuscular interfaces (Smith and Smith, 1995, 1996a,c). The fact that there are differences in efficiency of P transferred per unit C (see Chapters 5 and 8), indicates that the conceptually attractive model outlined by Woolhouse (1975), in which P and organic C were exchanged on a single transporter at the interface, as they are at the chloroplast envelope, is unlikely to provide the correct picture. This was certainly realized by both Clarkson (1985) and Smith and Smith (1986), who appreciated the sigiuficance of the interfacial apoplast and showed separate membrane transport systems operating for FO4 and sugars. More recently, Clarkson (personal communication) has suggested that if, as outlined above, the root apoplast can be viewed as an exchange zone in which conditions can be controlled and solute concentrations are relatively high, then it may be that FO4 uptake by the plant is via the low-affinity uptake system. If, as now seems possible, the high-affinity system is switched off in the fungus, this would provide a way of polarizing transfer in the direction of the plant. Again, probes for these transporters would provide useful data, especially if they could be used at a high level of resolution to show localization on the different membranes. It must be reiterated that the model of spatial separation of hexose and mineral nutrient transfer on different interfaces is still hypothetical. It is based on the distribution of H^-ATPase activity revealed by cytochemistry and on the observations that VA mycorrhizal fungi grow intercellularly in mutant plants of Pisum, in which arbuscules are reduced or absent, implying that the arbuscular interface is not required for C transfer. The idea should be useful in stimulating investigations of transfer processes but it must not be taken as established fact. Indeed, there are a number of problems with it, not least relating to the diversity of structures formed by glomalean fungi in roots. Intercellular interfaces are not present in all associations (Smith, 1995; Smith and Smith, 1996a; and see chapter 2), so that alternative ways of establishing spatial separation of function may need to be sought, perhaps involving intracellular hyphal coils.

406

General themes

Interplant Transfer The widely held belief that nutrients are transferred between plants linked by mycorrhizal mycelium is based on much uncritical discussion and does not recognize the essential difference between net transfer, which might be nutritionally important to the plants, and apparent transfer, brought about by equilibration of tracers with pools of unlabelled solutes (Newman, 1988). It is clear from the discussion above that a mechanism does exist at the membrane transport level which explains tracer movement of ^^C in both directions across a symbiotic interface. Sugars or organic N move from plant to fungus and organic N moves in the reverse direction. In a system involving a single plant and a fungal mycelium, net C movement must occur in the direction of the fungus, because C skeletons are required for synthesis of the organic N. Of course, if the fungus obtains these C skeletons from another source then there can be net C transfer as amino compoimds from the fungus into the plant (Fig. 14.11). Potential alternative sources of C are either another photosynthetic plant linked into the mycelium, or organic N from the soil. It is very difficult to differentiate experimentally between the situation outlined above (which actually involves net transfer across the symbiotic interfaces) and one in which the two photosynthetic plants contribute different amounts of sugar to support the mycelium (see Chapter 4). However, there is now one report in^vhich double labelling with ^^C and ^^C has been used to investigate transfer between Betula and Abies; the results are consistent with an important role for Donor plant

(active interface)

Receiver plant

Soil

NH4"

(active interface)

Al

Pi" -

Pi"

(senescent interface)

Gin

- • Gin

-• Pi"

- • Pi"

-•

Sugar-

1 ^ ^''

U

r

Fungus

A (senescent root)

F i g u r e 14.11 Possible nnechanisnns f o r transfer of organic C and P O ^ between mycorrhizal plants. Possible shuttles across the Veceiver' Interface are not shown. From Smith and Smith (1990), w i t h permission.

Uptake, translocation and transfer of nutrients in mycorrhizal synnbioses

407

glutamine in net transfer both N and C between plants (Simard, 1995; and see Chapter 15). As discussed in Chapter 5, it is less easy to envisage mechanisms at the membrane transport level which would promote reversal of transfer of PO^, so that it occurred from plant to fungus. We have suggested (see above) that the 'normal' symbiotic situation involves considerable modification of fungal membrane transport, so that PO^ efflux is markedly increased and uptake switched off in order to maintain polarized net transfer to the plant. Until we know how this change is achieved and whether it is readily reversible, it seems simplest to agree with Newman (1988) that the apparent transfer could be the result of effective exploitation by the fungus of P in senescent roots, which would indeed be likely to be generally permeable (leaky) to all solutes including P (Fig. 14.11).

Myco-heterotrophs Myco-heterotrophic plants obtain their supplies of organic C from a fungus, which in turn obtains it from a photosynthetic plant or, in the case of some orchids (see Chapter 13), from organic matter in soil. They also obtain mineral nutrients in this way, so that transfer of all solutes is polarized in their favour. Symbiotic modifications to membrane transport processes would certainly be different in the two 'host' plants, because in one (the autotroph) net C transfer would be in the direction of the fimgus and in the other (the heterotroph) it would be in the direction of the plant. Even in green orchids there is as yet no good evidence for organic C transfer to the fungus (see Chapter 13), and it is possible that in the adult photosynthetic stages the main role of the fungus is to supply mineral nutrients to plants which frequently have very poorly developed root systems. Transfer of mineral nutrients (N, P and so on) to the heterotroph could be by mechanisms similar to those operating in the autotroph. We know almost nothing about what actually happens in these symbioses and it will be a fertile but difficult area for experimentation. Measurement of rates of transfer and investigations of distribution and activity of HÂTPases and other transport proteins would be most interesting.

Conclusions Transport processes are central to the function of all mycorrhizal symbioses because they are nutritionally based. In the most common mycorrhizal types, explanations for mycorrhizal function must be sought which promote polarized transfer of different nutrients in opposite directions. Mechanisms that enhance efflux and maintain influx must coexist, at least in the same plant if not in the same cells, and consequently increased efflux cannot be based on general increases in permeability (leakiness) and loss of membrane integrity. Transport processes, whether at the whole-plant or at the cellular level, have been studied in only a few examples of the major mycorrhizal types. There is a wealth of information yet to be gained both in terms of the efficiency of the symbioses (nutrients gained for C expended) and the mechanisms involved. The increasing realization that within each major mycor-

408

General themes

rhizal type there is considerable diversity in structure and efficiency means that there is also diversity at the levels of uptake, translocation and transfer of nutrients. Myco-heterotrophic associations pose a particular challenge with respect to mechanisms of transport between the symbionts. The problems should become easier to address as the identity of the fungal symbionts and their links with autotrophic plants become clearer. This information will pave the way for realistic experiments on the 'unusual' transport processes that support the growth of nonphotosynthetic plants.

Plate 6. Penetration of cellufluor into the fungal sheath of Pisonia grandis. (a) Intact rootlet showing absence of turquoise fluorescence in the cortex, (b) Rootlet showing penetration of cellufluor where the fungal sheath has been damaged, (c) Section bathed in cellufluor, showing turquoise fluorescence of cellufluor throughout. Bars, 100 |im. From Ashford et al. {Australian Journal of Plant Physiology^ 1989), with permission.

15 The roles of mycorrhizas in ecosystems

Introduction In the century following the first experimental investigations of mycorrhizal function by Frank (1894) a wealth of information has accumulated, much of which is covered in this book. The basic structural and physiological attributes of mycorrhizal fungi, and of colonized roots, have been elucidated in detailed laboratory studies. The growth responses of individual species to colonization, usually in pots and as pure stands, have been subjected to innumerable investigations. There has been a small number of experiments investigating the impact of mycorrhizal fimgi upon interactions between different species of plant, grown in pots, but there remains a conspicuous gap in our understanding of the function of mycorrhizas at the community level. The mycorrhizal symbiosis, in all likelihood, co-evolved with the first land plants (Chapter 1) and invariably in nature the plants have occurred in communities of mixed species. The fungi themselves also coexist in similar communities, so that very diverse and interlinked populations are the norm. The questions which arise, and which will in all probability increasingly occupy the thoughts of those investigating the biology of mycorrhizas in the next century, concern the role of the symbiosis, if any, in the biology of plants growing in natural substrates and mixed communities in the field. Are there impacts of mycorrhizal colonization upon the fitness of individual plants? Do mycorrhizal fungi influence the outcomes of competitive interaction between species of plant? If so, under what circumstances and in what way are the influences mediated? Moreover, we may also ask how the populations of plants influence similar features of the fungal community. These are the sorts of questions which need to be addressed and for which, when we find positive or negative interactions, explanations will need to be sought at the molecular, biochemical and physiological levels. In other words, the detailed investigations directed at understanding mechanisms, which frequently must be done in simplified and ecologically unrealistic experimental systems, must be to an increasing extent focused on answering questions of ecological relevance.

410

General themes

Some researchers have already turned their thoughts in these directions. It is, for example, increasingly a prerequisite of any field-based ecological study to know what kind of mycorrhizal association predominates in a given community or ecosystem. This is frequently supplemented with information on seasonal and spatial distribution of mycorrhizal types and activity. Information of this kind is not difficult to obtain and we now have lists of 'occurrences' of broad mycorrhizal types in many different ecosystems, although details of structural variations are not always available. More difficult and time-consuming to obtain is information on function, so that investigations directed towards understanding the roles of mycorrhizas in community dynamics and ecological processes are in their infancies. What is emerging is a picture suggesting that the functions of the symbioses go far beyond the simple capture of mineral nutrients by individual plants of organic C by the associated fimgi. There is clearly considerable diversification in structure and in function, not only in the major types or even in mycorrhizas formed between different species of symbionts, but also imposed by genotypic variation of both fungus and plant. In any distinctive type of ecosystem, selection may have favoured specialized attributes of the symbiosis and symbionts that are appropriate to that particular set of environmental circumstances. What follows is a survey of the major types of biome, considered in a latitudinal gradient, with a view to determining the extent and distribution of mycorrhizas occurring within each, and to elucidate, as far as is known, the functions that might be important in the circumstances prevailing in each system. Inevitably, at this early stage of development, there are more questions than there are answers, but what is emerging is a picture in which the symbioses are multifunctional. It is to be hoped that identification of the questions can at least be of assistance to the next generation of researchers who will be seeking answers relevant to the real world.

Mycorrhizas in High Arctic and Alpine Biomes The small number of species that occur, usually as individual plants, in the nival zone of the Alps and in the very high Arctic are only intermittently uncovered by snow, grow in mineral soils and appear to be largely uncolonized by mycorrhizal fungi. Read and Haselwandter (1981) found that in this zone of the Austrian Alps, Ranunculus glacialis appeared to be free of all fungal colonization. Vare et ah (1992) reported that none of the six Ranunculus spp. examined in the Arctic at Spitsbergen were colonized by vesicular-arbuscular (VA) fungi. This contrasts with the situation for this genus below the nival zone (Mullen and Schmidt, 1993) and in temperate latitudes (e.g. Harley and Harley, 1987) where colonization by VA mycorrhizal fungi appears to be the norm. Similarly, in the maritime Antarctic, Christie and Nicolson (1983) could find no VA colonization on plants such as Deschampsia antarctica which were mycorrhizal at less extreme sub-Antarctic sites in the Falkland Islands. Haselwandter et al (1983) point out that during the short growing season of R. glacialis in the nival zone the plant is supplied with N and P contained in melt-water, so that climatic rather than nutritional factors are likely to determine success at the highest elevations. Below the nival zone of alpine areas and in those regions of the Arctic which have a consistent snow-free growing season, there is a continuous vegetation cover and

The roles of mycorrhizas in ecosystems

41 I

some accumulation of organic matter, at least at the surface. In their study of Ranunculus adoneus, which was carried out at 3500 metres in the alpine zone of the Colorado Front Range, Mullen and Schmidt (1993) showed that while the plant was lightly colonized by coarse and fine VA endophytes throughout the year, arbuscules were present only during the short growing season. Arbuscule formation was followed by increases of P concentration in both shoots and roots. It was proposed that P acquired in this period was stored for use during growth and flowering the following spring, both of which occur before soils thaw to release nutrients. More studies of this kind, in which the dynamics of colonization and nutrient acquisition are followed through the year in the natural environment of the plant, are much needed. They provide not only ecologically relevant information, but also valuable pointers for physiological investigations of arbuscule and hyphal function. Often, especially in those regions of the Arctic where the water-table lies permanently near the surface, plant communities contain a preponderance of species in families like the Cyperaceae, which are typically non-mycorrhizal. While this may help to explain the observation made both in alpine (Haselwandter and Read, 1980; Read and Haselwandter, 1981) and Arctic (Bledsoe et al, 1990; Vare et al, 1992) environments that VA mycorrhizal colonization is low, it does not provide a full explanation, since again many species which are 'hosts' to VA fungi at lower altitudes, are uncolonized or only lightly so in arctic-alpine situations. Even where VA mycorrhizas are observed, they are often formed by 'fine' rather than 'coarse' endophytes, there being a progressive increase in colonization by Glomus tenuis with altitude (Crush, 1973; Haselwandter and Read, 1980). One striking feature to emerge from studies of arctic-alpine plants is the extensive occurrence on their roots of fungi with dark septate (DS) hyphae (Haselwandter and Read, 1980; Christie and Nicolson, 1983; Kohn and Stasovski, 1990; Vare et al, 1992; Treu et al, 1996). In a study of 179 vascular plant species of Alberta, Currah and Van Dyk (1986) found that roots of 87% of alpine species were colonized by DS fungi, in contrast to only 9% in non-alpine environments. The preponderance of fungi of this general type on plants growing in alpine conditions is reflected in analyses of soil microflora, which suggest that in Antarctic (Heal et al, 1967), Arctic (Vare et al, 1992) and alpine soils (Haselwandter and Read, 1980), fungi with DS hyphae dominate the soil microbial community. Their quantitative importance in these habitats means that attention must be directed towards their possible taxonomic and functional status. Based upon the presence of sclerotium-like bodies or of aggregates of pigmented swollen hyphae with pores, Haselwandter and Read (1982) tentatively identified DS fungi of alpine plants as being of the genera Rhizoctonia and Phialophora, respectively. Fungi of the 'Rhizoctonia' type have been known, since early studies of Peyronel (1924), to be capable of colonizing roots, even those also occupied by VA mycorrhizal fungi, while Phialophora is recognized as a casual occupant or weak parasite of many roots, particularly of grasses, where they appear to be associated with programmed cortical senescence (Deacon, 1987). Stoyke and Currah (1991) observed that some cultures of DS fungi, originally isolated from roots of alpine plants, produced the large fan-shaped conidial appendages typical of Phialocephala fortinii (see Fig. 10.4), but only after prolonged storage at low temperature. They suggest that many of the DS isolates hitherto ascribed to Phialophora are congeneric

412

General themes

with Phialocephala and may even be conspecific with Rfortinii, If so, this contributes to an understanding of the status of these associations. Rfortinii can be, albeit under culture conditions of high C content, a pathogen of Pinus (Wilcox and Wang, 1987b) and is one of a number of dematiaceous fungi, previously included in Mycelium radicis atrovirens, which have been regarded as weakly pathogenic (see Chapter 10). It has also been considered to be a commensal saprotroph (O'Dell et ah, 1993) when growing on roots of Lupinus latifolius in temperate environments. Cluster analysis and ordination based on RFLPs of ribosomal DNA extracted from 117 dematiaceous and 10 sterile, hyaline fungi isolated from the roots of 26 species of subalpine plants indicated that two-thirds of the fimgi were closely related to or conspecific to P. fortinii (Stoyke et al, 1992). The prominence of DS fungi in general and of P. fortinii in particular, is clearly a feature of high elevations and latitudes, but the presence alone of these fungi in and aroimd roots is not sufficient to justify claims of mycorrhizal status. The fact is that we know virtually nothing of the biology of these associations in their natural environments and there is an urgent need for experimental analysis of their functional attributes. Haselwandter and Read (1982) isolated DS fungi from healthy, field-collected roots of the alpine sedges Carex firma and C. sempervirens, and obtained a positive growth response in C. firma when the plant was inoculated and grown in sand with one of the isolates. However, since neither the substrate nor the conditions employed in the experiment reflected those of the alpine habitat, they urged caution in the interpretation of these responses, referring to them as being evidence of an 'association' rather than of a typical mycorrhizal relationship. Of equally uncertain status are the ectomycorrhiza-like structures which are frequently, but not consistently, found on roots of herbaceous species such as Kobresia (Fontana, 1963; Haselwandter and Read, 1980; Kohn and Stasovski, 1990) and Polygonum viviparum (Hesselman, 1900; Read and Haselwandter, 1981; Lesica and Antibus, 1986). Where they occur, these associations too are normally formed by fungi with dark mycelia. Amongst these Cenococcum geophilum appears to be prominent, although the frequent presence of hyphae with clamp cormections indicates that basidiomycetous fungi may also be involved. There is a suggestion (Vare.ef al, 1992) that these species are more frequently colonized in this way when growing with typically ectomycorrhizal plants. Again, there is a need for experimental analysis of the status of these types of colonization. Trappe (1988) observed that when Pinus albicaulis occurred as a pioneer on recently exposed moraines in the Cascades Range (Oregon), it formed ectomycorrhizas predominantly with Cenococcum geophilum. The mechanisms of dispersal of fungal propagules at very high altitudes are not understood, but such occurrences, coupled with the observation that the surfaces of snow pack are frequently encrusted with soil particles, indicate that winds are sufficiently strong in these exposed environments to enable transfer of vegetative propagules of fungi. C. geophilum becomes progressively more important as a mycorrhizal colonist with altitude. In 'krunmiholz' formed by coniferous species above 2700 m in the Cascades, Trappe (1988) found that over 90% of ectomycorrhizas were formed by C. geophilum. This suggests that propagules of this fungus must be abundant in soils not far below mountain summits.

413


Mycorrhizas in Heathland Biomes Heathlands occur as major biomes under two environmental circumstances. In the first, the upland or montane heath is found in both continental and island locations (Fig. 15.1) at a distinct altitudinal position between the alpine zone and the tree line. The second, lowland heath, occupies areas of particularly impoverished acid soil at low elevation. These biomes are characterized by the presence of shrubby, sclerophyllous, evergreen plants of the families Ericaceae, Epacridaceae, Empetraceae, Diapensiaceae and Prionotaceae (Specht, 1979), many of which normally have hair roots colonized by ericoid mycorrhizal fungi (Read, 1983,1996; and see Chapter 12). Analysis of environmental gradients across which plants with ericoid mycorrhizas become increasingly prevalent, have shown that such communities arise primarily in response to nutrient impoverishment (Specht, 1981; Rundel, 1988; Read, 1989). Their occurrence in warm mediterranean climate zones as 'dry heath', or 'sand plain' formations as well as in subalpine environments, serves to emphasize the fact that nutritional rather than climatic factors play the primary role in determining the distribution of these communities. The response to low availability of N and P is to allocate increasing proportions of fixed C to the structural components lignin and cellulose rather than to molecules rich in protein or P, a process which leads directly to sclerophylly (Specht and Rundel, 1990) and to the release of residues of high C:N ratio and considerable recalcitrance. These accumulate at the soil surface to provide the matrix of complex residues in which ericoid mycorrhizal roots proliferate. In northern heaths, the hair roots of dominant plants such as Calluna vulgaris, Erica spp. and Vaccinium spp. are characteristically confined to the top 10 cm, or less, of the soil profile, where they are closely associated with the litter (Reiners, EUROPE 5000

TYROLEAN ALPS

ISLANDS FERNANDO PO

TRISTAN DA CUNHA

AFRICA O PICO

MT ELGON

MALAYA rSOOO

4000

3000

h 3000

2000

h 1000

sea level

sea level

Figure 15.1 Simplified global pattern of distribution of major biomes, highlighting the segregation of predominant mycorrhizal types in association with distinctive types of plant community. Mycorrhizal types: black shading, ericoid; grey shading, ecto; and no shading, VA. Note that in the dlpterocarp forests there will be important canopy and understory plants that are VA mycorrhizal (see text). From Read (1993), with permission.

414

General themes

1965; Gimingham, 1972; Persson, 1980). Interestingly, when herbaceous species such as Deschampsia flexuosa, Molinia caerulea, Eriophorum vaginatum and Carex spp. coexist with ericaceous shrubs, their roots are concentrated at greater depths in the soil profile (Gimingham, 1972), so the two groups of plants are not competing for the same resources. Similar mechanisms promoting coexistence in the sand plain communities of Western Australia have also been described. Indeed, the grasses may be colonized by VA mycorrhizal fungi, again emphasizing separate strategies of resource acquisition. Read (1993) presented a schematic view of the way in which distinctive mutualisms, together with modifications of root distribution and anatomy, might promote species diversity in northern heaths by enabling exploitation of different sources of the critical growth-limiting element N (Fig. 15.2). The coexistence of ericaceous, leguminous and carnivorous species, typically seen in heaths of moderate acidity, was facilitated by their abilities to use sources of N derived, respectively, from soil organic matter, the atmosphere and captured animals. Structural modifications, in particular production of aerenchyma, enables cyperaceous species like E. vaginatum to penetrate water-logged horizons where

(a) Ericoid mycorrhizal shrubs

(b) Insectiverous plants

(c) Leguminous plants

(d) Members of Cyperaceae, Restionaceae and Proteaceae, forming deep roots or cluster roots

F i g u r e 15.2 Schematic representation of compartmentation of resource acquisition in heathland ecosystems, based on the occurrence of distinctive mutualistic associations o r r o o t specializations, (a) Ericoid mycorrhizas occur in dwarf shrubs and play an important role in the mobilization of N in plant litter and microbial protein, (b) Insectivorous plants capture insects, f r o m which they release N . (c) Leguminous plants f o r m nodules which fix N . (d) Members of the Cyperaceae, Restionaceae and Proteaceae either produce deep roots that tap N in l o w soil horizons and / o r proteoid o r cluster roots that are important in capture of nutrients (particularly P, but also possibly N in the surface horizons). From Read (1993), w i t h permission.

415


they exploit N sources, including organic forms (Chapin et al, 1993), untapped by the other groups that are essentially surface rooting. Evidence in favour of a role for mycorrhizas in providing discrimination in nutrient use is increasing. Comparative analysis of 5^^N enrichment of leaf tissues of Picea, Vaccinium and Calamagrostis all growing in the tundra heathboreal forest transition zone of Alaska (Schulze et al, 1994) revealed significant difference of enrichment between these life forms, P. mariana having significantly lower 6^^N value (-6.496) than V. vitis idaea (-3.837) and C. canadensis (+0.585). Distinctive rooting depths may have contributed to these differences but isotope discrimination, facilitated by ecto-, ericoid and VA mycorrhizas respectively, is also a possibility (Schulze et al. 1994). Support for this suggestion has been obtained in a field-based study, in which the extent and type of mycorrhizal colonization was examined in subarctic, fellfield and heathland communities (Michelsen et ah, 1996). 6^^N enrichment of leaf tissue was determined in plants of known mycorrhizal status representative of ericoid, ecto-, VA and non-mycorrhizal categories. In the fellfield the mean 8^^N of the ericoid mycorrhizal species was -53, that of ectomycorrhizal species was -4.1 and of VA or non-mycorrhizal species, 0.0 (Fig. 15.3). In the heath the mean 5^^N values of the same groups were -7.6, -6.4 and -1.8, Leaf 5^^N (%o) -10

-8

-6

-4

-2

0 ERICOID MYCORRHIZAL DICOTS Cassiope tetragona Empetrum hermaphroditum Rhododendron lapponicum Vaccinium uliginosum ECTOMYCORRHIZAL DICOTS Arctostapiiylos alpinus Betula nana Dryas octopetala Salix myrsinites Salix reticulata NON-MYCORRHIZAL HEMIPARASITE Bartsia alpine NON-MYCORRHIZAL MONOCOTS Calamagrostis iapponica Carex vaginata Festuca ovina Tofieldia pusilla NON-MYCORRHIZAL NODULATED LEGUMES Astragalus alpinus Astragalus frigidus

Heath

LICHENS Cetraria nivalis Nephroma articum

ERICOID MYCORRHIZAL DICOTS Cassiope tetragona Empetrum hermaphroditum Vaccinium vitis-ideae ECTOMYCORRHIZAL DICOTS Betula nana Dryas octopetala Polygonum viviparum Salix herbacea\/ar.p '

Fellfield

NON- OR ARBUSCULAR MYCORRHIZAL SPR Calamagrostis Iapponica Carex bigelowii Festuca vivipara Luzula arcuata Lycopodium selago

F i g u r e 15.3 5 ' ^ N (0/00) of the most c o m m o n species of plant and t w o lichens in the treeline heath and fellfield in a subarctic ecosystem (Abisco, Sweden). Values are means of six replicate determinations ± standard e r r o r s . From Michelsen et al. (1996), w i t h permission.

416

General themes

respectively. In all cases values obtained from the ericoid and ectomycorrhizal plants were significantly different from those of the VA or non-mycorrhizal species. Although the differences between ericoid and ectomycorrhizal species were not significant (P = 0.051 in fellfield, and 0.270 in heath) the latter clearly appeared to occupy an intermediate position in the hierarchy. There was no evidence of segregation of roots by depth in these sites, almost all roots in the fellfield being restricted to the humus layer having a thickness of only 2-3 cm and to a similar zone in heathland of 10-15 cm. Ericoid and ectomycorrhizal colonization facilitates access to organic sources of N which are not only the predominant forms of N in the soil but frequently also have relatively low levels of 8 N enrichment. In addition to their abilities to use organic sources of N and P (Chapters 8, 9 and 12), it is increasingly evident that ericoid and some ectomycorrhizal fungi produce extracellular enzymes that break down complex polymers of C, thereby exposing new sources of organic N and P to attack (Table 15.1). A hitherto unrecognized association between the ericoid mycorrhizal fungus Hymenoscyphus ericae and the rhizoids of leafy liverworts in the families Adelanthaceae, Caphaloziellaceae, Cephaloziaceae, Calypogeiaceae and Lepidoziaceae has been described by Duckett and Read (1991, 1995). This relationship may be of considerable ecological importance because, whether in the moist acidic peats and mor-humus soils of high latitudes, or in the epiphytic cloud forest communities of tropical mountains, plants with ericoid mycorrhizas normally grow together with liverworts of these families. Microscopic analysis of the swollen tips of the rhizoids of the liverworts show that they are invariably occupied by dense hyphal complexes (Fig. 15.4). When these were exposed to the fluorescent dye 3,3'-dihexyloxacarbocyanine (DiOC6(3)), which selectively stains ascomycetes, they were seen to fluoresce brightly (Duckett and Read, 1991). Liverworts from these families were subsequently grown (Duckett and Read, 1995) in pure culture and inoculated either with H. ericae or with Oidiodendron spp., an orchid mycorrhizal fungus (Ceratohasidium cornigenim) and several ectomycorrhizal fimgi. H. ericae was capable of forming the typical hyphal complexes and swellings in the rhizoidal apices (Table 15.2) of all species seen to produce them in nature. The ericoid fungus failed to colonize those liverworts in families Jungermanniaceae, Amelliaceae and Aneuraceae which are normally associated with basidiomycetous fungi or those in the Pelliaceae, Fossombroniaceae, Lunulariaceae, Conocephalaceae and Marchantiaceae that have zygomycetous associates. None of the other fungi formed associations with any of the liverworts. The failure of Oidiodendron spp. to colonize the liverworts is of particular interest in view of their apparent ability to form ericoid mycorrhizas. Duckett and Read (1995) suggest that one of the important consequences of this pattern of colonization will be that the mats of leafy liverworts so extensively covering acidic organic surfaces will provide effective sources of inoculum for ericaceous seedlings germinating in them. The further possible ecological and physiological consequences of the susceptibility of these phylogenetically distinct groups of plants to colonization by the same fungal endophyte remain to be investigated. Heathlands in regions with a mediterranean-type climate experience a wet winter season, in which most of the root growth and mycorrhizal activity occurs (Ramsay et «/., 1986; Meney et aZ., 1993; Bell et al, 1994). These, and indeed most.

-O T3 C

z Q U

u (d •g

> o

£ N C o u j2

«

U o

•^

-I -I

^ 1^

o ^ 'J fd ^ ^

C

'

0)

^^ U

td 0) e-

Jd

O

Cd ON

<w

O p OS "^ U^ r~

O

-D

^-^

u Z I Z ^

r:

:D^

e-

•D

fd

—.

M

"53 U

0 _3

0)

U

3.

^

"2 cd

3 "D Q. c C cd D c 0 10 cd cd OJ a —I Q-

]o 0^

.C M

10

ro

ON ON

OJ 0

s.

>s

4-J c cd T3 Q-

^ 'S D ^ •2 ^

"cd

C •+3 U

>^

0 Q-

u

3

-D 0) U

txO

cd

cd

•^j

0

c

u

D

C

]Q-

Tf

0

U) 0) X

cd U-

Q-

i>^

s

p.

D

U

o

•g

ON

r

ti

c 0

2 •J

5b c ^^

-= ^ t "^ o S ^^ 8 v (U

c cd

U

_o _3

v

£

u

"QJ

I

(d 0)

§: «

ON

^

X

cd c Q) (U —1 OQ

3. ^

o c

0)

2 e

M

cd 4 5

cd - 3

cd

t5 1^ :§>

.ti

fc E -5. ^ -o -rf ^

o

TJ -a

L.

U

2

(-

^

u

(U

z z z z

I- I- I- I- I- I-

0) >s

I

r^

g

V

uj ^ u;

(Q n

NTNTNTZ

M M M g

E £ E -, o o o ^

>s

u

z z z z z z

U >>. C

(U

I I I I

V

I I I I I I

4-> ' - 0.5 _c

E 5 20 c

Z 15

S "Vi c

^

v> c

1^b '^ 4) 0.0 o. o

1

2 10

i

Q. O

)b

"5

j2

« e

^-0.5

H

o>

u.

^ :3^

35

5 50 100 Sun Ava ilable to Douglas-fir (%)

o

o i2

5

2

0

"1

Ob*

1

Ii"- 11 5 50 100 Sun Available to Douglas-fir (%)

Figure 15.7 Net isotope transfer between Betula and Pseudotsuga, in full sun (100%) and in 50% and 5% of full sun. (a) 1993; (b) 1994. Means denoted by the same letter do not differ significantly (P = 0.01). See text for explanation. From Simard (1995), with permission.

Net transfer from Betula to Pseudotsuga coincided with net photosynthetic rates of the seedlings, which were 1.5 and 4.3 times greater for B. papyrifera than P. menziesii in full sun and full shade, respectively. Furthermore, foliar N concentrations were 1.2 and 6.7 times higher in B. papyrifera than in P. menziesii. It seems likely that net transfer was determined by the gradient of assimilate and nutrient concentrations between the two species. Since rates of photosynthesis of the Betula in full sun were so much greater than those of Pseudotsuga in shade it is likely that C supply to the colonized roots of Betula would be greater than to those of Pseudotsuga, This, combined with the fact that plants of Betula contained significantly more N than P. menziesii, may have contributed to the apparent sink effect. There is still much to be learned about the biochemical and biophysical pathways associated with transfer of C compounds between the symbionts, but where gradients of C and N cooccur the transfer of C in combination with N, perhaps as glutamine (see Chapters 8 and 14), is a possibility. This study should provide further impetus for investigation of transfer processes at the cellular level. Since gradients of nutrient availibility are the norm in nature, confirmation that mycorrhizal interconnections can facilitate the net transfer of C, which is one of the most patchily distributed of all resources, is important. The transfer demonstrated by Simard (1995) took place between plants of broadly equivalent age and size, but might be of even greater significance where the gradients of irradiance are from large, fully illuminated adult plants to small shaded individuals in the understory a situation typical of natural forest ecosystems. Not only might the gain to the receiver be greater but the relative costs to the donor will also be smaller. Again, there is a need to investigate these processes in communities of naturally regenerating plants.


431

The attributes of ectomycorrhizas revealed under simplified laboratory conditions may be constrained in nature where herbivory and fungivory are among numerous factors with the potential to affect mycorrhizal function. It has been shown (Gehring and Whitham, 1991, Del Vecchio et ah, 1993) that races of Pinus edulis that were genetically susceptible to chronic herbivory, either by a stem-boring moth, or a needle-feeding scale insect, suffered a 30% reduction in ectomycorrhizal colonization relative to non-susceptible races. This reduction was subsequently observed to lead to a 20% loss of shoot biomass (Gehring and Whitham, 1994). Such effects can persist for at least one year after removal of the herbivore and may be more severe in poor than in fertile soils (Gehring and Whitham, 1995). In view of the sensitivity of ectomycorrhizal fungi to reduction in assimilate supply, revealed in experiments involving artificial defoliation (Last et ah, 1979); shading (Lamhamedi et ah, 1994); and separation of mycelium from root (Soderstrom and Read, 1987), the negative impacts of shoot herbivory are predictable, but clearly they must be taken into account when evaluating likely responses to colonization in nature. Analysis of the effects of below-ground grazing upon ectomycorrhizas have concentrated upon fungivory. This is appropriate, because most of the youngest root tissue is enveloped in mycelium and therefore protected from direct attack by herbivores. These analyses may be flawed, however, if carried out under simplified conditions. CoUembolans, for example, show highly selective grazing habits in the laboratory, they are influenced by a variety of factors, even down to the culture medium used to support fungal growth (Leonard, 1984); analysis of field-collected animals, however, show a diverse range of food materials in their guts (Anderson and Healy, 1972). Obviously it is desirable to use intact mycorrhizal plants grown on natural substrates with realistic communities of animals when evaluating these effects. Ek et ah (1994) determined the effects of different densities of the coUembolan Onychiurus armatus upon ectomycorrhizas of P. contorta formed by Paxillus involutus. The impacts of fungivory upon nutrient uptake by the extramatrical mycelium was examined by placing cups, containing ^^NH4^ or phytin, to which the fungus alone had access, through the soil. Low densities induced greater development of the mycorrhizal mycelium and an enhancement of uptake and transfer of N to the plants by 76%. Mycelial growth was impeded only at a high density of O. armatus. Studies did not show a significant increase in the coUembolan population when mycorrhizal systems were compared to non-mycorrhizal systems. Setala (1995) exposed birch and pine plants, both colonized (as would be expected in nature) by a number of fungal symbionts, to a naturally complex or highly simplified microfaunal population. After 57 weeks, despite reduced colonization of both species in the presence of the complex animal assemblage, their shoot growth and N and P concentrations had significantly increased relative to those seen in the microcosms with simplified populations. This suggests that fungivory may not be harmful to the symbiosis provided that the composition of the microbial community is sufficiently complex to ensure efficient mobilization and turnover of soil resources. More experiments involving manipulation of target species in otherwise normal microfaunal populations are clearly needed.

432

General themes

VA Mycorrhizas in Temperate Biomes Plant communities which are dominated by trees or herbaceous plants that are colonized by VA mycorrhizal fungi progressively replace those with ectomycorrhizal roots on a global scale as, with decreasing latitude, mean annual temperatures and evapo-transpiration rates increase (Read, 1991a,b). The climatic changes lead to a reversal of the leaching tendencies prevalent in heath and boreal forest systems and to a consequent increase in base status and p H close to the soil surface. This, in turn, results in acceleration of the rates of turnover of organic matter and nitrification (EUenberg, 1988). As a consequence, the relatively mobile NO^ ion progressively replaces organic N or NH4 as the principal source of N for plants. As availability of the element increases, the proportion of nitrophilous species increases along the gradient and P, the mobility of which is low, can be predicted progressively to replace N as the major growth-limiting nutrient in the ecosystem (Read, 1991a,b). Whereas selection appears to have favoured the prevalence of VA rather than ectomycorrhizal colonization in many ecosystems which are primarily P limited, it has proved surprisingly difficult to demonstrate that plants growing in the field under natural conditions benefit from enhanced access to P (Fitter, 1985, 1990; and see Chapters 5 and 16). This situation contrasts strongly with that seen in pots under controlled conditions where reported values of P inflow to roots colonized by VA mycorrhizal fungi are an order of magnitude greater than those seen in the field (Fitter and Merryweather, 1992). A number of factors may combine to reduce the impacts of VA colonization in the field. Of primary importance is the fact that the rates of growth of many plants in nature are limited by environmental factors, for example water shortage, other than P deficiency. In stress-tolerant species, growth rates may be inherently so low that P requirements can be satisfied by diffusion, without involvement of mycorrhizal hyphae. In the case of VA mycorrhizal plants, as in that of plants with ectomycorrhizas, it is important to consider the extent to which the potential of the symbiosis, identified under controlled conditions, is influenced in nature by other biotic factors. One of the proposed reasons for the apparent 'ineffectiveness' of VA mycorrhiza in the field is grazing by coUembolans. Here, as in studies of their impact upon ectomycorrhizal fungi, much that has been written about the possible impacts of grazing coUembolans is based upon experiments of unrealistic design. It has been shown in pot experiments (Warnock et al., 1982; Finlay, 1985) that grazing by coUembolans can reduce or even eliminate the beneficial effects of mycorrhizal colonization, although at low densities of the animals the growth and P concentration of mycorrhizal plants can be increased (Harris and Boemer, 1990). Application of insecticides such as chlorfenvinphos with the aim of specifically reducing coUembolan populations has been shown to provide increases of growth and P inflow to Holcus lanatus in the field (McGonigle and Fitter, 1988). However, since this compound is known to be toxic to non-target animals such as earthworms (Rabatin and Stinner, 1991) and root feeding insects (Brown and Gange, 1990) the results of the experiment are confounded. Using a compound, chlorpyriphos, shown not to effect non-target organisms, Gange and Brown (1992) were, however, able to demonstrate that benefits of mycorrhizal colonization were significantly greater in the absence than in the presence of coUembolans.


433

While coUembolans may have important negative effects upon the mycorrhizal symbiosis under some circumstances, it is increasingly evident that a large number of animal species can influence the extent and outcome of colonization by VA fungi and that tripartite interactions between plant, fungal symbiont and microfauna are both complex and far-reaching. Only by assessing the impacts of all these interactions, many of which are occurring simultaneously in the ecosystem, can a balanced view of faunal effects on the mycorrhizal symbiosis be obtained. Application of the nematicide carbofuran to undisturbed prairie vegetation (Ingham et al, 1986) led to increases of VA mycorrhizal colonization in roots of the dominant grass Bouteloua gracilis, suggesting that nematodes had adverse effects upon the extraradical mycelium. There were no associated effects upon P concentration in shoots, but without data on plant growth and therefore total P uptake by the plants this cannot (as discussed in Chapter 5) provide unequivocal information on the contribution of mycorrhizas to P nutrition. A closer approach to the natural condition was obtained by Barker (1987) who grew Lolium perenne with or without its shoot (Acremonium loliae) and root {Glomus fasciculatum) endophytes, in order to determine their combined and individual contribution to defence against the Argentine stem weevil Listronatus bonariensis. The foliar endophyte, when present, deterred feeding and oviposition by the weevil but its effects were reduced by the presence of the mycorrhizal fungus. In the absence of the foliar endophyte the VA fungus had no effect upon herbivory. There is clearly the possibility that while the presence of shoot endophytes provides advantages to the plant in terms of direct anti-herbivore effects upon aboveground grazes, rhizophagous insects would be deterred by the presence of the mycorrhizal fungus. Obviously, experiments to investigate this more complete scenario would be worthwhile, especially in plants such as L. perenne, which have fibrous root systems and retain mycorrhizal colonization despite the likelihood that they are able to scavenge effectively for nutrients in the absence of the symbiosis. Resistance to herbivory may be a selective factor, along with that of fungal pathogens reported below, contributing to the retention of mycorrhizal colonization in plants such as grasses in temperate ecosystems. It is increasingly recognized that examining responses of plants to mycorrhizal fungi over short periods of their vegetative growth cycle, whether in pots or in the field, may provide misleading results. The extent of this weakness has been highlighted by studies of the bluebell Hyacinthoides non-scripta carried out throughout the life cycle of the plant in a deciduous woodland (Merryweather and Fitter, 1995a,b, 1996). This work clearly demonstrates, apparently for the first time in a natural population of field-grown plants, that the inflows necessary to maintain a positive P budget can only be achieved in the mycorrhizal condition. H. non-scripta is a perennial, vernal geophyte which characteristically dominates the herb layer of deciduous woodland in parts of north-west Europe. It has a coarse root system made up of thick (0.5-1.0 mm) unbranched elements which are produced annually from the base of the bulb. A new bulb and root system are produced every year. By regularly sampling undisturbed plants of H. non-scripta throughout their annual life cycle in a deciduous woodland, Merryweather and Fitter (1995a) explored the relationship between mycorrhizal colonization and inflows of P. There was a rapid increase in the proportion of root length colonized by VA mycorrhizal fungi over the period from root emergence in September (autumn), to reach a maximum value

434

General themes

in excess of 70% in January and February (Fig. 15.8a) even before the shoots appeared above ground. Thereafter, as confirmed by declining numbers of entry points (Fig. 15.8b), new colonization slowed. From the time of root emergence, P inflow increased rapidly at a rate similar to that of colonization, although until December values were negative, indicating a net loss of P. Maximum inflows were reached during the photosynthetic phase (Fig. 15.8c), but these subsequently declined, again at the same rate as that of colonization. The data for P inflow and percentage root length colonized showed a significant correlation (Fig. 15.8d). The individual plants lose significant amounts of P, particularly at the end of the growing season, in seeds, old leaves and roots as they are shed. Glasshouse-grown plants, lacking mycorrhizal colonization, are imable to absorb sufficient P from the soil to balance their P budget and therefore end the season with a large P deficit, which would not permit survival in the field. Soils supporting otherwise undisturbed colonies of H. non-scripta growing in the field were subjected to a benomyl drench at 2-monthly intervals over two years, which greatly reduced mycorrhizal colonization without having any effect on P availability (Merryweather and Fitter, 1996). This led to a large reduction of the P concentration of all vegetative parts of the plant, relative to that in untreated colonies (Fig. 15.9). In contrast, the flowers and seeds of the benomyl-treated plants had the same P concentration as the controls after the first season, with reduction in their P status being observed only after two years (Fig. 15.9). This suggests that when P uptake is reduced, H.

75

162

162

233

162

233

Days from 6th September

233



20

40

60

80

Mean %rootlength colonized

Figure 15.8 Seasonal pattern of (VA) mycorrhizal colonization and P inflow in Hyacinthoides non-scripta growing in the field, (a) Mean percentage colonization, (b) Mean abundance of entry points, calculated as entry points per intersection. The points on the curves are three harvest running means, (c) Curve of fitted values for P inflow in the field, (d) Correlation between P inflow and mean percentage root length colonized. Linear regression: r^ = 28.8, P = 0.15. From Merryweather and Fitter (1995a), with permission.


435

non-scripta protects its reproductive structures by selectively allocating P to them. Selective allocation of P to reproductive structures by mycorrhizal plants growing in P-deficient soils has also been observed in pot-grown plants of wild oat Avena fatua (Koide et al, 1988b; Bryla and Koide, 1990b). This pattern of allocation in VA mycorrhizal plants may significantly influence fecundity and so played a direct role in determination of fitness in the field. Plants such as bluebell with very coarse root systems can be predicted on theoretical grounds to be responsive to mycorrhizal colonization. However, questions remain as to the role play by the symbiosis in plants such as grasses which, despite the fibrosity of their root systems, retain high levels of colonization in nature. There are many glasshouse experiments with grasses demonstrating that increases in PO^ uptake can lead to increases of yield, but such effects have been difficult to observe in natural communities. Thus Hetrick et al, (1988, 1990), examining the responsiveness of two grass species that dominate the tall grass prairies of the USA, found evidence in the C3 species Bromus inertnis that, despite greater P acquisition in the mycorrhizal condition, there was little or no growth response. In nature, B. inertnis makes most of its growth in the cool seasons of autumn and spring and it is at these times that arbuscule production is at a maximum. Early season growth enables the plant to avoid competition with the other dominant grass species Andropogon gerandii, a C4 plant which grows in the warm season and is extremely responsive to VA colonization. The question arises as to the nature of the mycorrhizal relationship in B. inermis. A subsequent study confirmed that P acquisition by B. inermis was significantly increased by colonization, especially at 18°C compared with 29°C but, again, no growth response followed (Hetrick et al,, 1994). As in the annual grasses described above, benefits of P acquisition may only be expressed late in development in terms of increased fecundity and improved

4 Y 'o> 3 r O)

572 V 1 V

March

April

May

June

Early April

Late April

May

Figure 15.9 Effect of benomyl drench on the concentration of P in the leaves of Hyacinthoides non-schptay measured at intervals over two growing seasons in the field. Shaded columns, control; unshaded columns, benomyl. From Merryweather and Fitter (1996), with permission.

436

General themes

offspring performance. It is possible that luxury accumulation of P early in the growing season increased the competitive ability of the grass by pre-empting availability of the element to other species. The responses of these two grass species to seasonal changes in environmental conditions, as well as to mycorrhizal colonization provides a further example of niche separation, this time a temporal one. An alternative explanation for the apparently beneficial impact of VA mycorrhizal fungal colonization upon the annual grass Vulpia ciliata has recently been proposed (Newsham et al., 1994, 1995). Again using benomyl to control colonization of the roots in the field, no relationship was found between the extent of occurrence of VA fungi and the rates of P uptake, but the biocide also controlled weakly pathogenic fungi such as Fusarium oxysporum, which were known to reduce fecundity of the plant (West et al., 1993a,b). It was therefore suggested by Newsham et al. (1994) that the benefits of VA fungi arose through their ability to protect the plant from pathogens. This possibility was examined further in field-grown populations of V. ciliata exposed to different concentrations of benomyl, so as to control the extent of colonization by VA and pathogenic fungi (Newsham et al, 1995). Fecundity was largely unresponsive to fungicide application, despite the fact that benomyl significantly reduced the abundance of both types of fungi in roots. However, the abundance of root pathogens, especially F. oxysporum, was negatively correlated with fecundity, even though plants displayed no disease symptoms. The poor relationship between fecundity and benomyl application contrasted markedly with the effects of benomyl on VA mycorrhizal and pathogenic fungi, and with the negative effects of root pathogens on fecundity. These effects could be explained if the two groups of fungi interacted, so that when both were greatly reduced by fungicides, the net effect on fecundity was slight. Such a hypothesis was supported by statistical analysis of the data, showing that there was a positive effect of VA mycorrhizal fungi, but only in relation to the negative effect of the pathogens. The interaction between VA mycorrhizal fungi and root pathogens was resolved using a transplant approach. Seedlings of V. ciliata were grown in a growth chamber with a factorial combination of inoculum of F. oxysporum or a Glomus sp., both isolated from V. ciliata at the field site, and then planted into a natural population of the grass in the field. After 62 days of growth, clear evidence was obtained that colonization by Glomus gave a protective effect. Plants inoculated with that fungus performed as well as control plants, even when simultaneously inoculated with F oxysporum, whereas those inoculated with F oxysporum alone grew significantly less well (Table 15.3). The Glomus sp. had a negligible effect on the performance of the plants in the absence of the pathogen. There was no correlation between shoot P concentration and the abundance in roots of either pathogenic or mycorrhizal fungal hyphae. Rather, the differences between treatments seem to have been due to a reduction in the frequency of pathogenic hyphae within roots brought about by VA mycorrhizal colonization. These experiments reveal that the effects of mycorrhizal colonization in the field may involve subtle interactions with other microorganisms that can only be detected by combining sensitive experimental design with careful data analysis. In so far as they deal with single species of plant, however, the studies of Hyacinthoides and Vulpia still provide a very simplified view of the possible impacts of


437

T a b l e 15.3 Effects of a factorial combination of Fusarium oxysporum (F) and a Glomus sp. (G) on shoot biomass and r o o t length of Vulpia ciliata plants g r o w n in the laboratory, transplanted into the field and sampled f r o m the field after 62 days* g r o w t h Variable

Treatment -G -F

Log (In) shoot biomass (mg) 2.4a Root length (cm) 217a

Main effects

Interaction GXF

+G -F

-G +F

+G +F

G

F

2.2a 203a

1.4b I 11 b

2.2a 228a

F = 3.5 F=I7.3*** F = 9.4** F = 9.0**

F = 4.8* F = 9.0**

Means are of 16 replicates; where followed by different letters they differ at P

0

m

T

r

o

Nodulated

d

—I

2

4

8

S '^N (%o) F i g u r e 15.11 Frequency diagram showing ' ^ N abundances of foliage of m i o m b o tree species w i t h different r o o t symbioses at different sites: solid bars, Zambia 1995; hatched bars, Zambia 1991; open bars, Tanzania. For discussion see t e x t . From Hogberg and Alexander (1995), w i t h permission.

zal type and soil conditions, especially in the wet and seasonally wet tropical forests. Whereas trees with VA mycorrhizal colonization predominate over large areas of such systems, there are localized occurrences in Amazonian (Singer and Araujo, 1979, 1986) and west African (Newbery et al, 1988, 1996) forests, of communities dominated by ectomycorrhizal species. The communities are characteristically restricted to the most nutrient-poor soils with a surface accumulation of litter and raw humus in which the colonized roots proliferate. Even where legumes occur in such systems they appear to be largely of the non-nodulated type. Along a transect in seasonally wet forest in Korup (Cameroon) within a flora of 200 tree species, nodulated legumes make up only 1% of the total basal area (Hogberg and Alexander, 1995). Such observations suggest on the one hand that N fixation is not an advantage for trees in rain forest, and on the other that ectomycorrhizal

446

General themes

associations are favoured by extreme nutrient deficiency. By analogy with the situation observed in boreal and temperate forests, it has been predicted that these associations are involved in mobilization of organic N from the litter (Alexander, 1989a; Read, 1991a) thus providing advantages over VA mycorrhizas. However, recent analyses of foliage from Korup Forest indicate that ectomycorrhizal species have higher percentage P and lower N:P ratios than their VA mycorrhizal counterparts, suggesting that ectomycorrhizas are more important in P nutrition in the rain forest (Hogberg and Alexander, 1995; Newberry et al, 1996). Positive correlations have been observed between growth, ectomycorrhizal colonization and foliar P concentration in dipterocarp seedlings (Lee and Lim, 1989; Lee and Alexander, 1994). It may be that, as in temperate systems, the ability of the ectomycorrhizal fungi to produce extensive mycelial networks in soil and to selectively exploit localized pockets of organic matter enriched in P or N, provide the selective advantage for this type of symbiosis, but much experimental research remains to be done. In the moist tropics the majority of species probably form VA mycorrhizal associations, with rather non-specific relationships between plants and fungi. Many of the soils are acid, highly leached and P deficient, often because of sequestration as poorly available Al phosphates. Here, VA mycorrhizas are likely to play an important role in rapid cycling of Pi as it becomes available through mineralization. The fact that in these forests ectomycorrhizal plants often occur together in distinct guilds or 'groves' (Singer and Araujo, 1979; Newbery et al, 1988), where their roots proliferate in acidic himiic materials of low quality, is suggestive of specialized functions for this type of mycorrhiza. A factor favouring the maintenance of such guilds, once they are established, is that ectomycorrhizal fungi of tropical, as of temperate and boreal forests, show a greater level of host specificity than do their VA mycorrhizal counterparts (Alexander, 1989a; Thoen and Ba, 1989; Smits, 1992; and see Chapter 6). The greater availability and vigour of the requisite inoculum in and around the guild, relative to that in the surrounding VAdominated systems, will increase the chance of successful establishment of siblings in the vicinity of the guild and promote its integrity. At the same time, host specificity, by favouring the survival of a relatively small number of compatible species, would be expected to lead to reduced diversity within the guild. There is a limited amount of evidence that diversity is, indeed, lower in ectomycorrhizal than in surrounding VA-dominated communities (Alexander et ah, 1989) but the relative importance of nutritional and biological factors in determining these differences remains to be investigated. The Roles of Mycorrhizas in Primary Successions The sequential development of plant communities following major environmental perturbations such as glaciation (Crocker and Major, 1955) and volcanic activity (Simkin and Fiske, 1983) are well documented. It is acknowledged, also, that scarcity of nutrients in the poorly weathered materials exposed by such events may determine the early stages of the primary succession which is initiated on them (Gorham et al, 1979). Under these circumstances it is tempting to suggest a role for mycorrhizal fungi in facilitating the succession, but as yet there is little direct


447

evidence for such a role in nature. Thus, while normally mycorrhizal plants with light seeds such as Dryas (Crocker and Major, 1955) are known to be among the first to colonize recently deglaciated soil, there have been no studies of the pattern of mycorrhizal development on them in situ or, more importantly, of the role of any such development in facilitating establishment. There is evidence that both wind and animals can act as vectors of mycorrhizal inoculum. Warner et ah (1987) demonstrated that spores of VA fungi could be wind blown for up to 2 km, whereas animals were able to transport VA propagules over several miles of sterile pumice on Mount St Helens after its eruption (Allen, 1988). In all likelihood the small spores of ectomycorrhizal fungi will be transported over even greater distances. Trappe (1988) reports the invasion of recently exposed glacial till in the Oregon Cascades by coniferous trees which are abimdantly colonized by ectomycorrhizal fungi. In this case the glacial valley is surrounded by afforested ridges which can provide an abimdance of propagules of both partners. The proximity of established vegetation probably frequently complicates the successional process by providing local sources of reproductive propagules. Allen (1984) found such undisturbed patches of the original plant communities in the centre of the main region of pyroclastic flow on Mount St Helens following its eruption. Evidence from species lists compiled for isolated islands formed by volcanic eruptions does little to clarify the picture, although non-mycorrhizal or facultatively mycorrhizal species, the latter often grasses, figured more prominently as early colonists on Krakatoa. These were succeeded by species likely to be more responsive to mycorrhizal fungi, including orchids and a Casuarina sp. (Simkin and Fiske, 1983). Critical questions concerning the role of mycorrhizal colonization in facilitating establishment of plants in virgin sites can ultimately only be answered by manipulative experimentation, involving addition of plants or fungi to such sites and monitoring over a chronosequence the relationship between colonization, nutrient capture, growth and survival. Primary succession on sand dune ecosystems is a more predictable process in which relationships between soil quality disturbance and mycorrhizal status are apparent (Read, 1989). In the disturbed and nutrient-enriched conditions of the drift line, non-mycorrhizal species predominate, particularly those of the families Chenopodiaceae and Brassicaceae. Under more stable conditions a succession of communities made up of species that are responsive to VA mycorrhizal colonization are found, ranging from open grassland dominated by the grass Ammophila arenaria on the fore-dunes, to herb-rich closed communities on the more stable dimes. The succession from drift line to stable back-dunes typically covers a gradient of decreasing pH and increasing soil organic matter content (Fig. 15.12) over which communities dominated by plants with ectomycorrhizas or ericoid mycorrhizas become increasingly important, with forest or heathland replacing grassland as the climax vegetation type. The Roles of Mycorrhizas in Secondary Successions While primary successions often commence under conditions of nutrient impoverishment, those processes referred to as secondary succession, which follow disturbance of existing vegetation, are normally initiated in an environment of

General themes

448 Occurrence of and Responsiveness to Mycorrhizal Infection

HIGH

LOW HIGHnr~^_

species DRIFT

LINE

(ANNUAL a BIENNIAL

FACULTATIVELY VAMYCORRHIZAC

MOBILE

FACULTATIVELY VA MYCORRHIZAL

O

z

FORE-DUNES

(PERENNIAL

GRASSES)

EARLY

FIXED

DUNES

mycorrhizal hyphae = non-mycorrhizal roots > control. The contribution of roots and hyphae to aggregate stability increases as the concentration of organic matter in soils increases under pasture. The organic binding agents are relatively transient compared with the more persistent agents that cement the smaller particles. Consequently, the larger aggregates, which are so important in determining the occurrence of free-draining pores, are not only relatively temporary but also fragile and subject to disruption by tillage. Miller and Jastrow (1990) emphasized that because of the transitory nature of the bonds, stabilization by roots and hyphae depends on their continued production. This

Vesicular-arbuscular mycorrhizas in agriculture and horticulture

467

Size of particle (^g) E

••

(b)

16

o o CM A CO i z •7? 6 .yi 00

iQ. B O CO

•

12

•

•

• 8

•

0)

4

1

01

• •

1

6

1

1

1

'

8 10 12 14 Hyphal length (m g"^)

*

•

16

18

Figure 16.3 (a) Length of hyphae connbined with water-stable particles in soil after 14 weeks* growth with ryegrass (R), white clover (W) or unplanted (U). Vertical bars, 2X standard errors of the mean, (b) Relationship between hyphal length in whole soil and percentage water-stable aggregates >2000 |im diameter. From Tisdall and Oades (1979), with permission of Australian Journal of Soil Research.

468

General themes

point is obviously relevant for the management of plant-mycorrhizal populations to increase and maintain the structural stability of soil and again highlights the potentially negative effects of bare fallow.

Conclusions Biological activities in soil are widely recognized as playing a vital part in nutrient cycling and availability to plants and in developing and maintaining soil structure, and contributing to 'soil health'. Sustainable land-use requires that soil degradation ceases and that soil management practices are adopted to conserve and augment soil resources. Mycorrhizal fungi comprise just one of the functional groups of organisms that are important in the soil ecosystem, but their position in forming direct links between roots of plants and the soil fabric means that they play key roles in soil-plant interactions. The dependence of mycorrhizal fungi on the photosynthesizing plant means that they do not deplete reserves of soil organic matter as saprophytic microorganisms do, but contribute to its accumulation directly as hyphae and spores and indirectly via their effects on plant growth. The effects of mycorrhizal associations on agricultural and horticultural systems are almost all potentially beneficial, with only a very few reports of growth depressions in field situations that remain imperfectly explained (Modjo and Hendrix, 1986; Modjo et aL, 1987). There is increasing evidence that mycorrhizal fungi contribute to crop productivity and are important in the way in which crops respond to fertilizer applications. It is possible to make estimates of the potential savings in fertilizer if a given crop is mycorrhizal, but real monetary gains from managing mycorrhizal symbioses in broad-scale agriculture remain elusive. Direct inoculation of VA mycorrhizal fungi is currently limited to relatively small-scale, high value systems or subsistence farming. This picture may change when inoculum of VA mycorrhizal fungi can be readily produced in a form that is convenient for wider application. At present, it is important to view mycorrhizal associations as integral coniponents of a complex soil ecosystem and to manage that system in order to maximize the contributions that mycorrhizas most certainly make to soil processes and growth of plants. Research directed towards understanding the activities of mycorrhizal fungi in agricultural and horticultural systems is valuable both in determining appropriate management strategies and as a background against which inoculation techniques will be developed. Less easily quantified benefits of mycorrhizal activity include their effects on soil structure and soil structural stability. Good soil structure is certainly important, both for plant productivity and for control of erosion and the losses that stem from it. The benefits here are in fertility, erosion control and water management, and are inextricably linked to the growth of plants and their fungal symbionts. One aspect of environmental degradation which is receiving increasing attention is the transfer of nutrients, frequently applied as fertilizer, to aquatic systems. The consequence is loss of nutrients for crops, and reduced water quality and eutrophication. The roles played by plants in reducing these nutrient transfers are being recognized in land management and the potential of mycorrhizal fungi to contribute to effective scavenging of nutrients from soil deserves attention. Integrated management of the soil-plant system is clearly important and mycor-

Vesicular-arbuscular mycorrhizas in agriculture and horticulture

469

rhizal symbioses should be viewed in this context. There is no real argument for breeding crops which are more dependent on mycorrhizal symbiosis than current varieties in order to cash in on the sale of inoculum when production and application become economically viable. However, there is an argument for ensuring that levels of colonization are maintained or increased, because this will be important in all aspects of management of mycorrhizas, including nutrient absorption, production of propagules and maintenance of soil structural stability. Furthermore, selection of efficient fungi will require both an understanding of their ecological requirements and their physiological integration with the plants.

17 Mycorrhizas in managed environments: forest production, interactions with other microorganisms and pollutants

Introduction With the realization of the extent of occurence of ectomycorrhizal fungi on trees in many natural ecosystems came the recognition that the symbiosis might be manipulated to enhance productivity in afforestation programmes. Since many commercial practices, particularly those employed in tree nurseries are inimical to the growth of all but a few ruderal species of mycorrhizal fimgi, special techniques have been developed which enable selected fungi to colonize plants prior to outplanting. The application of these techniques has facilitated superior performance in tree crops in many parts of the world, particularly those that lack natural sources of mycorrhizal inoculimi (White, 1941; Wilde, 1944; Shemakhanova, 1962; Stoeckeler and Slabauch, 1965; Mikola, 1969, 1973; Hacskaylo and Vosso, 1971). Over the same period interest has grown in the possibility of harvesting edible fruit bodies of ectomycorrhizal fungi which have been used as commercial inoculum both to supplement diet and revenue. Experience of the use of inoculated seedlings has indicated that responses to mycorrhizal colonization are often greatest under the most extreme conditions, particularly those involving exposure to drought, metal contamination or pathogens. Such observations have led to analysis of the functional basis of the ameliorative effects of mycorrhizal fungi. Considerable advances have been made towards an understanding of the role of the symbiosis in providing resistance to these stresses which, although they also occur in natural ecosystems, are often locally increased by previous land-use practices or by the afforestation process itself. In so far as the atmosphere and climate of the earth are being changed globally as well as locally by the activities of man it is necessary also to consider the extent to which, and the mechanisms whereby, mycorrhizal colonization may respond to

Mycorrhizas in managed environments

471

these larger scale events. Increased direct inputs to soils of N, S and H ions, largely as wet deposition, may have adverse effects upon growth and nutrition of both partners in the symbiosis. Such effects are thought to be contributory factors in the forest decline syndrome experienced in Europe and north-eastern USA. On an even wider scale, progressive global enrichment of atmospheric CO2 can be predicted to influence the C balance of plants and hence, indirectly, that of their fungal symbionts. As a result, there is an emerging interest in the extent to which mycorrhizal systems may act both as sinks for any additional CO2 assimilated by plants, and as pathways through which this C may pass into the soil C pool. This chapter presents an overview of the advances made in the technology of large-scale ectomycorrhiza inoculation programmes, an analysis of present knowledge of the role of this symbiosis in amelioration of selected environmental stresses, and describes some early indications of the responses of both ecto- and vesiculararbuscular (VA) mycorrhizas to elevation of atmospheric CO2. Ectomycorrhizal Inoculum Production and Inoculation Practice The use of defined inoculum consisting of fungi that were physiologically and ecologically appropriate for the planting site, with a view to improving performance of the crop, was pioneered by Moser (1958) in Austria, Takacs (1967) in Argentina, and ITieodorou and Bowen (1973) in Australia. Prerequisites for the widespread use of ectomycorrhizal inoculation programmes are the selection of appropriate fungal symbionts and the development of methods for the large-scale production of inoculum. The two requirements are interrelated because, in addition to providing enhancement of performance of the inoculated crop, the selected fimgus must be able to withstand the physical, chemical and biological stresses involved in the production of the inoculum, as well as those imposed by the soil, usually of a forest nursery, into which it is to be introduced. To date, the most widely used and most successful inoculation programmes have employed Pisolithus tinctorius. Interest in this fungus was prompted by its wide geographic distribution, broad host range (Marx, 1977) and the knowledge that it became prominent on adverse sites, particularly those subject to drought, high temperature or contamination (Schramm, 1966). Various commercial inoculum formulations and inoculation techniques have been developed for use in seedling production systems (Marx and Bryan, 1975; Marx, 1975, 1980; Marx and Kenney, 1982; Marx and Cordell, 1989; Marx et al, 1991). The most successful have involved the growth of vegetative mycelium in vermiculite-peat mixtures moistened with liquid nutrient medium (Marx and Kenney, 1982). Vermiculite provides a well-aerated laminated substratum, within which the mycelium is protected, and addition of peat in different ratios enables adjustment of pH to the required range, usually 4.8-5.5. The recommended nutrient solution has a C: N ratio of between 50 and 60 and is added in volumes sufficient to ensure that all free C is utilized by the fungus in the course of its development in the medium. The presence of available C at the time of inoculation leads to competitive exclusion of the mycorrhizal fungus by saprophytes. The major challenge in the commercial development of inoculation procedures is the scale of the operation required. In the case of pine, for example, 1.5 billion

472

General themes

seedlings are produced per year in nurseries of the southern USA (Marx, 1985). A modification of the practice employed for production of edible mushroom {Agaricus bisporus) spawn has been used to enable scaling-up of inoculum production. The vermiculite, peat-moss and nutrients are mixed and sterilized in a large-volume rotating blender. Starter mycelial inoculum is added, thoroughly mixed, and 10-1 batches of inoculated substrate are aseptically dispensed into sterile plastic bags. The bags are sealed but have 'breather-strips' installed to enable ventilation of the medium during fungal growth. Bags are incubated at room temperature for 5-7 weeks, after which time they can be shipped to nurseries for use. Inoculum of P. tinctorius together with that of two other fungi Hebeloma crustuliniforme and Laccaria laccata, is currently produced and marketed by Mycorr Tech. Inc., University of Pittsburgh Applied Research Center, Pittsburgh, USA. A tractor-drawn nursery seedbed applicator has been developed (Cordell et al, 1981) which places inoculum directly into seed-beds prior to sowing at doses of 0.33 1 m~^ of nursery soils and at a depth of 4-6 cm. At a cost of approximately $7.50 per litre of inoculum, this operation is calculated to represent about 5% of the total cost of establishment of a pine plantation in the southern USA (Marx, 1991). In order to eliminate weeds, pathogens and other symbiotic fungi which are potential competitors, seed-beds are routinely fumigated, most commonly with a methylbromide-chloropicrin mixture, before inoculation. Even so, re-invasion of fumigated soil by spores of naturally occurring mycorrhizal fungi, particularly Thelephora terrestris, normally occurs within days, and it is a requirement of the inoculant fungus that it has the ability to colonize roots quickly. T. terrestris appears to be the dominant mycorrhizal fungus of nursery soils worldwide (Mikola, 1970; Ivory, 1980; Marx et ah, 1984a) and, whether as a result of re-invasion after fumigation, or natural occurrence, its presence as a potential mycorrhizal colonist of roots must be recognized in all nursery studies. Because of the ubiquitous occurrence of T. terrestris, experiments designed to evaluate the influence of an inoculant fungus are complicated by the fact that most of the uninoculated 'control' plants, as well as some of those in the inoculated treatment, are invariably colonized by the local, naturally occurring species. There may also be other 'casual' colonists, amongst which 'E-strain' fungi (see Chapter 10), and Laccaria species are common. Such trials are therefore comparisons of performance between T. terrestris and the inoculated symbiont. Experience with P. tinctorius as the introduced organism strongly suggests that large numbers of mycorrhizas must be produced consistently on the roots of the seedlings if maximum promotion of growth is to be achieved when they are outplanted to reforestation sites. In these situations, Marx et al. (1976,1988) have shown that if less than half of all mycorrhizas are formed by P. tinctorius, no growth promotion relative to that seen in Thelephora-colomzed plants is observed. Outplanting trials in several coimtries (Trotymow and van den Driessche, 1990) have confirmed that the benefits of inoculation increase progressively with extent of colonization by Pisolithus. It is a striking feature of the programme of inoculum production using P. tinctorius that only one vegetative isolate of the fungus has been used throughout. This socalled 'super-strain' was originally obtained from a sporophore found under Pinus taeda in Georgia, USA. Its aggressive traits have apparently been enhanced by annual re-isolation over 30 years from seedlings growing in inoculation trials. Problems with the use of solid substrates for inoculum production include the


473

large space required for storage, difficulties in maintaining homogeneity of conditions within and between batches, and the inability to control physicochemical conditions in the medium in the absence of water. Because of these difficulties there have been various attempts to use liquids or gels as culture media. The main advantages of submerged, liquid culture are the homogeneity of the medium and the control which can be obtained over physical and chemical conditions. Vessels suitable for large-scale axenic production of fungal inoculum have been developed for other purposes in the chemical and pharmaceutical industries. They are designed to facilitate careful regulation and optimization of culture conditions for particular organisms, reducing the period of culture compared with solid substrates (Le Tacon et al, 1985; Boyle et ah, 1987). Inoculum produced in this way can be applied directly as a slurry (Boyle et ah, 1987; Gagnon et ah, 1988), requiring some form of fragmentation. Unfortunately, this treatment greatly reduces the vigour of many ectomycorrhizal fungi. Attempts have been made to retain viability of fragmented inoculum by incorporation in a protective carrier medium. Sodium alginate has been successfully used, either applied as a gel directly to the bare roots (Deacon and Fox, 1988), granulated (Kropacek et al., 1989) or as beads (Le Tacon et al, 1985; Mauperin et al, 1987). The susceptibility of many fungi to fragmentation damage, even when protected in this way, has led to a search for alternative culture methods. One approach which has considerable promise involves the production of the inoculum inside hydrogel beads, which can be applied directly, circumventing the fragmentation phase (Jeffries and Dodd, 1991; Kuek et al, 1992). Several ectomycorrhizal fungi, including species of Descocolea, Hebeloma, Laccaria and Pisolithus, have been successfully grown as inoculum in this way, and it has been shown that viability can be retained after storage for up to seven months at low temperature (Kuek et al, 1992). Basidiospore inoculum of P. tinctorius has been used on an experimental basis in the USA and elsewhere. This can yield growth responses, but rarely produces as many mycorrhizas per plant as does the 'super-strain' of vegetative inoculum and so is less effective. Spores can be sprayed onto fumigated plots as a suspension in water, to which a wetting agent such as Tween 20 is added. Marx et al (1991) report that spore doses of 0.5-1.0 g m~^ of soil surface are effective in enhancement of performance of southern pines. Alternatively, spores can be placed in a clay encapsulating mixture (Marx et al, 1984b) to be applied as pellets to the soil or as a coating to the seed. A delay in production of mycorrhizas from spores might be expected because, as described in Chapter 6, colonization would not normally take place from monokaryotic mycelia. Only after hyphal fusion and the formation of dikaryons does mycorrhizal colonization occur. Unfortunately, the extensive work on methods of inoculum production has not been matched by a demonstration of efficacy in forest production. As a result, while there are theoretical studies of the economic advantages to be gained from use of the new inoculation technologies (e.g. Kuek, 1994), prospects for their extensive application are not promising. Furthermore, a consequence of the recognition of the advantages of fungal diversity in ecosystems (Chapters 15 and 16) will be an increasing reluctance to introduce into mixed communities, single, potentially dominant species. A highly competitive fungus, suitable for inoculation programmes, might have considerable influence on the fungal diversity and the gene pool of the resident population.

474

General themes

Although the responses to inoculation with P. tinctorius have been good in warmer and more drought-susceptible parts of the world, this fimgus has proved less successful in cooler climates. In the Pacific north-west of the USA, for example, the 'super-strain' of P. tinctorius performed less well that did local isolates of the fungus (Perry et ah, 1987). In this region, the US Forest Service developed a spore inoculation programme based upon the use of mycorrhizal fungi known to be important in local ecosystems, including species of Laccaria, Heheloma, Rhizopogon and Suillus (Castellano and Molina, 1989). Spores have been applied to seed-beds through the nursery irrigation system, or to container-grown plants using mistpropagation units. Poor colonization was obtained with Rhizopogon and Suillus spp. (Perry et ah, 1987). In contrast, several strains of Laccaria produced abundant mycorrhizas in container-grown plants (Molina, 1982). One strain, subsequently referred to as L. bicolor S238, was found to have particular promise as an inoculant. Some Laccaria and Hebeloma strains have been developed as commercial inoculum, producing high levels of colonization on Pseudotsuga menziesii in containers and imder nursery conditions (Hung and Molina, 1986). It appears, however, that despite success in achieving colonization by vigorous strains of fungi, outplanting performance of the seedlings has improved little (Perry et ah, 1987). This has also been the experience in Europe. Le Tacon et al. (1988,1992) describe a number of experiments in France, Spain and Britain in which the performance of nursery inoculated plants of P. menziesii and Pinus sylvestris has been followed for several years after outplanting. The fungi used were mostly strains of Thelephora, Hebeloma and Laccaria, including the vigorous Oregon strain S238 of L. bicolor, originally isolated by Molina. The extent of success in obtaining colonization by the inoculant fungus varied from nursery to nursery, apparently being determined largely by the rate of re-invasion and vigour of indigenous Thelephora strains. Even where high levels of colonization by inoculant fungi were achieved, improvements in performance of the outplanted trees were rarely observed. Inoculation of P. menziesii with L. bicolor S238 provided significant increases of height growth and a doubling of wood volume at one site in central France 6 years after outplanting, but at the remaining sites differences between control plants colonized by T. terresfris and those that were inoculated were small. Jackson et al (1995) report a similar experience with container-grown P. menziesii and Picea sitchensis which were inoculated with a wider range of fungi and transplanted, after colonization, to six nursery sites across the UK. Few significant effects upon growth were observed and none of the kind that would be useful to a forester in a practical situation. Unfortimately, it appears that many of those fungi selected to achieve optimal colonization in the nursery are poor competitors in the field, especially when outplanting sites contain indigenous populations of mycorrhizal fungi. McAfee and Fortin (1986) preinoculated seedlings of Pinus banksiana with L. bicolor, P. tinctorius and R. rubescens before transplanting them to denuded, burned or natural pine stands. After 2 months in the natural stand, colonization by L. laccata and P. tinctorius had declined significantly whereas that of R. rubescens showed a modest increase. P. tinctorius showed an ability to colonize new roots in a denuded site which lacked competition from an indigenous mycorrhizal population. These are the circumstances in which the greatest successes have been achieved in inoculation programmes involving this fungus. The failure of L.


475

bicolor to compete with indigenous fungi is in line with the observation of Bledsoe et al. (1982) that the closely related L. laccata failed to persist on seedlings of P. menziesii when challenged by native fungi on outplanting sites in Washington. There are a number of possible explanations for the failure of inoculation to produce beneficial effects at outplanting sites. Probably amongst the most important of these is the inability of introduced inoculum to persist on the roots of planting stock after transfer from the nursery to the field. In addition to the fact that soil conditions experienced by nursery and container grown plants are very different from those in most outplanting sites, the lifting, storage and transport of seedlings, especially those raised in bare-root nurseries, can be expected to reduce the vigour of fine roots and their fungal associates. These are circumstances under which replacement of introduced fungi by those resident in soil of the replant site are likely to occur most readily. It is noteworthy in this context that the most strongly beneficial effects of inoculation have been observed where plants are transferred to disturbed or treeless sites in which inoculum potential of any indigenous fungi is likely to be low. Here, in contrast to the situation so often reported in soils with a pre-existing vegetation cover, responses to inoculation can be quite dramatic (Table 17.1), involving improvements in survival as well as increases in yield (Marx, 1991), and they appear to be most marked where the soil is contaminated with metal ions (see below). In this context, the ability of ectomycorrhizal Betula spp. to colonize mine spoils spontaneously is widely recognized.

Table 17.1 Percentage increase in survival and volume growth of pine seedlings after 2-4 years with Pisolithus tinctohus ectomycorrhizas over controls with naturally occurring ectomycorrhizas on various adverse sites Pinus species

P. resinosa P. echinata P. virginiana P. taedataeda P. rigida P taeda

P. virginiana P taeda

Site

Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Coal spoil Kaolin spoil Fullers' earth Copper Basin Copper Basin Borrow Pit

Data from Marx (1991) and Marx et al. (1989).

Adversity

pH 3.0 pH4.l pH 3.1 pH 3.8 pH 3.8 pH 3.4 pH4.3 pH3.3 pH 3.4 pH4.l pH3.4 pH4.3 Low fertility Low fertility Eroded Eroded Droughty

% Increase in seedling Survival

Volume

214 5 87 480 0 57 8 20 14 41 96 16 0 0 0 0 17

60 400 444 422 420 215 180 415 750 400 800 380 MOO 47 45 88 412

476

General themes

The Use of Ectomycorrhizal Inoculation Programmes to Produce Edible Fungi While emphasis in applied research on ectomycorrhizas has so far concentrated on improvement of tree production, there is an increasing awareness of the potential to exploit the commercial value of the fruit bodies produced by ectomycorrhizal fimgi. At present, a small number of mycorrhizal species (Table 17.2) are prized for their gastronomic quality and are hence of high value. They are collected mostly from natural stands and constitute only a small fraction of the total global production of edible fungi (Fig. 17.1), most of which are saprophytes grown under controlled conditions. In order to increase supply of mycorrhizal fruit bodies, the current demands for which far outstrip supply, numerous commercial organizations are involved in planting trees which have been pre-colonized by inoculation with appropriate fimgi. Particular emphasis has been placed on truffles {Tuber spp.) because of their extremely high economic value, the most important of these being the black truffle, T. melanosporum. Techniques for the germination of the ascospores of this fungus and for the aseptic production of mycorrhizas by a number of Tuber Table 17.2 Some high-priced mycorrhizal truffle and edible mushrooms Botanical name

Common names

Markets

Tuber melanosporum Perigord black truffle Worldwide Vitt truffe due Perigord (Fr) tartufo nero pregiato (It)1 schwarze Truffel (G)

Tuber magr)atum Pico

Italian white truffle truffe d'Alba tartufo bianco pregiato weisse Truffel Boletus edulis cep, penny bun Bull, ex Fr. cepe de Bordeaux porcino steinpilz Car)tharellus dbarius chanterelle Fr. girolle gallinaccio pfifferlinge Tricholoma matsutake matsutake (S. Ito et Imai) Sing.

Worldwide

Approximate recent prices

Fresh 550 (wholesale, London) Fresh 3250 (retail, London) Fresh IOO--430 (picker, France) Bottled 860-1800 (retail, London) Canned 500 (wholesale, Cahors) Bottled 1000 (retail, London) Fresh 800 (wholesale. Bologna)

Europe North America

Fresh 45 (retail, Hamburg) Fresh 10 (retail, Bologna) Dried 60 (retail, Zurich)

Europe North America

Fresh 10 (retail, Hannover)

Japan

72-720 (wholesale, domestic produce) 75 (wholesale, China) 72 (wholesale. South Korea) 36 (wholesale, from North Korea)


477

Wood ear 465 Shiitake 526

.....i^S^^Ws^.^*^^^

^^\, ,

,,,

^..^mmKimmmMm^m,^^j.Amm>^

EnokltaKO

187

Silver ear 140 Nameko 40 Others 155 Oy^t^'' 917 ^ ^ ^ ^ ^ ^ T - ^ ^ ^ - - - . - - s i j g | H ^ Mycorrhizal >200

Button 1590 Figure 17.1 The contribution of mycorrhizal fungi to the approximate world production of edible mushrooms in 1991. Values are tonnes XI000. From Hall et al. (1994), with permission.

spp. were pioneered in France (Grente et ah, 1972; Chevalier and Desmas, 1975; Chevalier and Grente, 1978) and Italy (Palenzona, 1969; Fontana and BonfanteFasolo, 1971). T. melanosporum has a broad host range and can be successfully grown on calcareous soils with the hardwood genera Corylus, Quercus, Carpinus and Castanea, as well as softwoods such as Pinus. Commercial production of colonized seedlings, particularly of Quercus and Corylus, now takes place in a number of centres in both the northern and southern hemispheres. In France alone about 160000 plants colonized by T. melanosporum are produced annually, mostly by Agri-Truffe of St Maixant; some are exported to the USA (Hall et al, 1994). On a smaller scale, colonized plants are being produced in New Zealand, by the New Zealand Institute for Crop and Food Research, in a programme pioneered by Hall, involving the introduction of both fungus and plant as exotic species. Truffieres have been established on both North and South Islands, usually as mixed plantings of Quercus and Corylus on potentially favourable sites, at some distance from any other ectomycorrhizal communities to reduce competition between pre-existing and introduced fungi. Truffles were collected from the introduced, inoculated plants within 5 years of establishment of a truffiere at Gisboume, New Zealand (Hall et ah, 1994). Since in other parts of the world (e.g. Europe and California) the first truffles are usually produced only after 7-10 years, the prospects for production of truffles in New Zealand and development of an export industry appear bright. Despite advances in science and technology which provide the prospect of largescale production of a number of edible mycorrhizal fungi, the commercial success of any venture is not assured. To a large extent the value of the commodities (especially of truffles) is based upon its limited availability, so that prices will certainly drop if large-scale production is achieved. However, especially in crops that can be used for timber, harvesting of edible fruit bodies could be an additional source of revenue or food, especially in developing countries. The large-scale establishment of eucalypt plantations in China, using planting stock pre-inoculated with edible fungi, has considerable potential to provide an important dietary supplement (Dell and Malajczuk, personal communication).

478

General themes

Improvement of Drought Resistance and W a t e r Balance In Ectomycorrhizal Systems There has long been an interest in the possibility that ectomycorrhizal fungi might improve the water relations of trees. Such effects would be of particular significance in an applied context where seedlings were being transplanted from nursery beds, especially if the site being afforested was subject to drought. There are reports that mycorrhizal colonization can be of some benefit to trees planted into such environments. Cromer (1935) believed that colonization increased the drought resistance of Pinus radiata by protecting the roots from shrinkage and providing increased uptake of water from soil at low water potentials (\|/). Some support for this view was supplied by Pigott (1982), who showed that roots of Tilia cordata colonized by Cenococcum geophilum were able to survive soil \|/ as low as - 5 . 5 MPa. While Cenococcum mycorrhizas could not store sufficient water to sustain transpiration for more than a short period when soil was dry, their protective effect enabled the colonized roots to recover their absorptive functions rapidly after soil was rewetted. In studies of P. ponderosa, Goss (1960) showed that mycorrhizal colonization improved the survival of plants after short periods of exposure to drought but had little effect over longer periods. Under conditions of extreme drought in the arid steppes, Zerova (1955) found that mycorrhizal colonization afforded little protection to oak seedlings, although colonization improved their vigour at intermediate levels of soil moisture. It must be emphasized, however, that in the absence of direct measurements of tissue water balance, such benefits could be attributable to nutritional rather than hydrological effects. Dixon et al. (1983) studied the water balance of bare-root and container-grown seedlings of Quercus velutina which had been inoculated with Pisolithus tinctorius or left uninoculated. They found that container-grown plants with extensive mycorrhiza development had a significantly improved water balance following transplantation, pre-dawn shootwater potential (\|/shoot) values being significantly higher in the mycorrhizal plants during mild drought. Similar responses to colonization by this fungus have been observed in Pinus virginiana (Walker et ah, 1982) and P. taeda (Walker et al., 1989). In the latter case, the plants had been grown on a mine spoil for 7 years with or without colonization by P. tinctorius. Colonized plants had a midday ¥xyiem 0.2 MPa higher than that of the uninoculated controls. These results are at variance with those of some other studies (Sands and Theodorou, 1978; Sands et al, 1982; Lehto, 1989; Coleman et al, 1990) which report either no impact of ectomycorrhizal colonization upon \|/piant or a negative effect. Sands and Theodorou (1978) found greater resistance to liquid flow in the soilplant pathway when seedlings of Pinus radiata were colonized by Rhizopogon roseolus than when they were uncolonized. However, these authors made the important observation that the potential benefits of mycorrhizal colonization were likely to have been reduced in their experiment because the extraradical mycelial system, and in particular its rhizmorphs, had failed to develop. The potential of the mycelial system to provide conduits for the transport of water was indicated by Boyd et al (1986) who showed that if rhizomorphs of Suillus bovinus connecting mycorrhizal seedlings of Pinus sylvestris to moist soil were cut, transpiration declined markedly within minutes. The importance of the extraradical

479


phase for water absorption was further emphasized by Lamhamedi et al. (1992) who examined the ability of a number of genetically distinct dikaryons of Pisolithus tinctorius to influence V|/xyiem of Pinus pinaster seedlings growing under moderate drought. They found significant correlations between \|/piant/ total root system resistance, and both the extension growth and the rhizomorph diameters of the different strains (Fig. 17.2); those genotypes that produced the most extensive system of thick rhizomorphs enabled plants to sustain the highest \|/xyiem ^t low Vsoil.

It is likely that there are also considerable intra- and inter-specific differences in the ability of ectomycorrhizal mycelial networks to resist exposure to drought. Rapid resumption of absorptive activity following a prolonged dry period would clearly be important for the plant. Experiments which determine the resilience of ectomycorrhizal mycelia, perhaps along the lines of those described by Jasper et al. (1989) examining VA systems, would be valuable. These workers demonstrated that, provided that the mycelial networks were intact, they could retain viability for at least 36 days, even when \|/soii was as low as - 2 1 MPa. A complete evaluation of the role of mycorrhizas in the water relations of plants cannot be achieved simply by measurement of tissue \j/ (see also Chapter 5). Colonized plants are likely to be larger than their non-mycorrhizal counterparts and, as a result, in pot studies at least, they may use available water resources more quickly, possibly even developing lower \|/xyiem- Thus measurements of \|/piant provide an indication of the impact of mycorrhizal colonization on

O

3x28

•

37x34

A

34x25

•

2x36

a

8x28

•

11x15

O

34x20

•

9x22

*

27x34

•

17x20

25

extension oU.m.Pâse^

Figure 17.2 Effects of the dianneter of rhizomorphs (|Lim) and of extension rate of the extraradicle mycelial system (cm^) of a range of dikaryotic isolates (designated by number on left of figure) of Pisolithus tinctorius, upon the xylem water potential (vj/xyiem) of P"^^^ pinaster seedlings. From Lamhamedi et al. (1992), with permission.

480

General themes

water relations only when effects of plant size, transpiration rate and tissue nutrient status are taken into account. Despite these complexities there is sufficient evidence to suggest that the extraradical mycelia of ectomycorrhizal roots, by playing a role in absorption and transport of water as \|/soii declines, may be of direct benefit to plants. There is a need for experiments to determine the extent and nature of the involvement of this mycelium, especially its rhizomorphs, in transfer of water, and to evaluate the importance of mycorrhiza for the water relations of plants grown under controlled nutrient regimens.

Improvement of Metal Resistance by Ectomycorrhizal Fungi A number of groups have investigated, in the laboratory, the basis of the apparent alleviation of metal toxicity arising from mycorrhizal colonization. The acidity typical of most substrates of ectomycorrhizal plants, and in particular of many of the man-made substrates such as mine-spoils which they colonize, is conducive to the solubilization of potentially toxic metal ions. Amongst these, Al and Fe are quantitatively the most significant in natural soil, while mine-spoils may be polluted by Ni, Pb, Zn and Cd, either separately or in combination. Colonization of clonal cuttings of Betula pubescens by Paxillus involutus provides significant increase in resistance to Zn toxicity and the growth enhancement associated with mycorrhiza formation is related to reduced transfer to the leaves (Brown and Wilkins, 1985; Denny and Wilkins, 1987). Jones and Hutchinson (1986), using Ni, obtained a similar result in seedlings of B. papyrifera which were colonized by Scleroderma flavidum, which is widely present on spoil heaps. This associate was far more effective in providing resistance than was Lactarius rufus. Ni was preferentially concentrated in the roots of the plants where it appeared to be sequestered with polyphosphate (Jones and Hutchinson, 1988). In contrast to the pattern of distribution of Zn, there was some evidence that Ni was accumulated in senescent leaves so that avoidance of exposure to the metal was largely a feature restricted to stem tissue. Large differences occur between ectomycorrhizal symbionts in their effectiveness in providing resistance to metal toxicity. There is evidence both at the species and strain levels that exposure to metals in the soil can lead to selection for resistance. Colpaert and van Assche (1987b) isolated strains of Suillus luteus from Zn-contaminated soil that were able to grow in the presence of 1000 lîg g~^ Zn. Strains of the same fungus obtained from fruit bodies growing on uncontaminated soil showed little or no growth above 100 jig g~^ Zn. Zn-resistant strains of S. bovinus conferred significantly more Zn tolerance upon plants of Pinus sylvestris than did non-resistant strains, and tolerance was most probably attributable to binding of the metal in the extraradical mycelium of the fungus (Colpaert and van Assche, 1987a,b).

Disease Suppression by Ectomycorrhizal Fungi When crowded together as dense monocultures in nursery beds, seedlings are particularly susceptible to attack by fungal pathogens. As a result, there is much


481

interest in the ability of ectomycorrhizal fungi to act as biological control agents. It appears that ectomycorrhizal fungi can provide their hosts with enhanced resistance to attack by fungal pathogens. Both Pisolithus tinctorius and Thelephora terrestris have been shown to reduce the impacts of the root pathogen Phytophthora cinnamomi on Pinus spp. (Marx, 1969, 1973), while inoculation with Laccaria laccata has been shown to reduce the incidence of disease caused by the pathogen Fusarium oxysporum in Pseudotsuga menziesii (Sylvia and Sinclair, 1983a), Picea abies (Sampangi and Perrin, 1985) and P. sylvestris (Chakravarty and Unestam, 1987a,b). It has been suggested that protection against fungal pathogens is achieved as a result of the physical barriers imposed by the hyphal mantle (Marx, 1973) or by the production of phenolic compounds in the plant tissues in response to the presence of the mycorrhizal fungus (Sylvia and Sinclair, 1983b). While both of these effects may indeed be involved in contributing to defence in the adult plant, there is evidence that ectomycorrhizal fungi exert direct antibiotic effects upon would-be pathogens. Duchesne et al. (1988a,b) observed that inoculation of seedlings of Pinus resinosa with the fungus Paxillus involutus significantly reduced pathogenicity of Fusarium oxysporum before mycorrhizal colonization took place. Increases of seedling survival were associated with a sixfold decrease in sporulation of F. oxysporum in the rhizosphere of the plant (Duchesne et al, 1987). Ethanol extracts of the rhizosphere indicated that fungitoxic effects of P. involutus were present within 3 days of inoculation of seedlings with P. involutus (Duchesne et ah, 1989). Disease suppression at this critical stage of plant development prior to formation of the ectomycorrhizal symbiosis may be of particular significance both in the nursery situation and on the natural regeneration niche. Little is known about the chemical basis of the observed antibiotic activity, although Kope et al. (1991) isolated two antifungal compounds from a liquid medium in which Pisolithus tinctorius had been growing. These were identified as hydroxy forms of benzoylformic and mandelic acid and given the names pisolithin A and B, respectively. It is regrettable that most of the experimeni;s on antibiotic effects of ectomycorrhizal fungi have been carried out under rathl^r unrealistic conditions. Epidemiological studies under natural conditions are necessary to determine if, or at what stage, colonization of roots by ectomycorrhizal fungi can enhance disease resistance. The recent recognition of the stimulatory effects of some classes of bacteria upon the processes of mycorrhizal colonization of roots has heightened awareness of the complexity of microbial interaction in soil, as is shown below.

Mycorrhiza Helper Bacteria Transmission electron microscopy of ectomycorrhizal roots collected from the field has revealed the presence of bacteria in the mantle (Foster and Marks, 1966), and the population of bacteria isolated from mycorrhizal roots of Pinus is distinct from those growing in association with uncolonized roots (Rambelli, 1973). The presence of such bacteria as casual associates df the mycorrhizosphere is to be expected. However, the possibility that they are directly involved in the dynamics of mycorrhiza formation was suggested by studies of Bowen and Theodorou (1979) which showed, in vitro, that the ability of the fungus Rhizopogon luteolus to colonize roots of

482

General themes

P. radiata was enhanced in the presence of some bacterial isolates but inhibited by others. Subsequently, Garbaye and Bowen (1987), using three fungal symbionts, examined the process of mycorrhiza formation in different steam-sterilized soils inoculated with a population of bacteria obtained from one of the soils. The effects of addition of bacteria were different in each of the soils and each of the fungi responded in a distinctive manner to the inoculum. Positive effects upon mycorrhizal colonization outnumbered those that were negative. When bacteria growing in the mantle of surface-sterilized roots of P. radiata mycorrhizal with Rhizopogon luteolus were isolated and characterized, up to 10^ colony forming units were obtained, most of which were fluorescent pseudomonads (Garbaye and Bowen, 1989). Significant positive effects upon mycorrhiza formation were produced by 80% of these isolates on re-inoculation of this plant-fungus combination. Stimulatory effects of the presence of fluorescent pseudomonads of the Pseudomonas putida group have also been reported in the case of VA (Mosse, 1962; Meyer and Linderman, 1986) and orchid (Wilkinson et aL, 1989) mycorrhizas. This suggests that the effects of the bacteria may be rather general in nature. However, a number of recent experiments have indicated that relationships between fungi and bacteria in the ectomycorrhizal symbiosis may be more specific, and on the basis of these, Garbaye (1994) has proposed that a special category of 'mycorrhization helper bacteria' (MHB) should be recognized. The ability of 47 strains, mostly of fluorescent pseudomonas, to enhance mycorrhiza formation in the P. menziesiiLaccaria laccata partnership, was compared under nursery, glasshouse and axenic conditions (Garbaye et aL, 1990; Duponnois and Garbaye, 1991). The most efficient isolates increased the amount of mycorrhizal colonization from 67% in the control to 97% and produced the same result in the soil as under sterile conditions, suggesting that the effects occur independently of the influence of the general soil microflora. Further evidence for specificity in the effects of MHBs was gained in experiments showing that, under a wide range of conditions, isolates obtained from mycorrhizas formed by a strain of L. laccata (S238) on P. menziesii consistently stimulated colonization by this strain, and that the closely related L. bicolor had no effect on L. proxima but had negative impacts upon colonization by fungi of other genera (Garbaye and Duponnois, 1992). To date, there is little information on the possible mechanisms of stimulation apparently induced by the bacterium. Garbaye (1994) proposes five different hypotheses (Fig. 17.3) to explain the phenomenon, each of which is amenable to experimental analysis. The possibility that the MHB might enhance the susceptibility of the plant to colonization, either by production of cell wall softening enzymes (Hypothesis 1) or by enhancing the root-fungus recognition process (Hypothesis 2), is considered. Evidence that MHBs produce five enzymes, endoglucanase, cellobiase, hydrolase, pectate lyase and xylanase (Duponnois, 1992), that are known to have wall softening properties, and that cell-free culture filtrates of Pseudomonas spp. containing these enzymes are capable of enhancing colonization of roots by VA fungi (Mosse, 1962) lends support to Hypothesis 1. It is clearly desirable, however, to establish relationships between virulence of the MHBs and enzyme production. Nutritional enhancement of fungal growth by MHBs (Hypothesis 3) is also a possibility but in view of the apparent specificity of the effect it must be of a highly specialized type. Detoxification of compounds present in soil could, indirectly.

483


t «

\

'^

• 9

C

Mycorrhizal Symbiosis, Second Edition

Mycorrhizal Biotechnology

Insect Symbiosis

Insect Symbiosis

Cyanobacteria in Symbiosis

Molecular Basis of Symbiosis

The Philosophy of Symbiosis

Topological Geometry, Second edition

Mendelism, Second Edition

Building Expertise, Second Edition

Applied Geophysics, Second Edition

Bronchial Asthma, Second Edition

JavaServer Pages, Second Edition

Radiation Biophysics, Second Edition

Pediatric Dialysis, Second Edition

Platelets, Second Edition

Contemporary Japan, Second Edition

Information Marketing, Second Edition

Beginning JavaScript Second Edition

Analytical Electochemistry. Second Edition

Applied Sedimentology, Second Edition

Learning Python, Second Edition

Swing, Second Edition

V.S. Naipaul, Second Edition

Forensic Engineering, Second Edition

Electromagnetic Waves, Second Edition

Swing, Second Edition

Galois Theory (Second Edition)

Ocean Circulation, Second Edition

Nonlinear Systems, Second Edition

Mycorrhizal Symbiosis, Second Edition

Mycorrhizal Biotechnology

Insect Symbiosis

Insect Symbiosis

Cyanobacteria in Symbiosis

Molecular Basis of Symbiosis

The Philosophy of Symbiosis

Topological Geometry, Second edition

Mendelism, Second Edition

Building Expertise, Second Edition

Applied Geophysics, Second Edition

Bronchial Asthma, Second Edition

JavaServer Pages, Second Edition

Radiation Biophysics, Second Edition

Pediatric Dialysis, Second Edition

Platelets, Second Edition

Contemporary Japan, Second Edition

Information Marketing, Second Edition

Beginning JavaScript Second Edition

Analytical Electochemistry. Second Edition

Applied Sedimentology, Second Edition

Learning Python, Second Edition

Swing, Second Edition

V.S. Naipaul, Second Edition

Forensic Engineering, Second Edition

Electromagnetic Waves, Second Edition

Swing, Second Edition

Galois Theory (Second Edition)

Ocean Circulation, Second Edition

Nonlinear Systems, Second Edition

Recommend Documents