Whales, Seals, Fish and Man

Developments in Marine Biology 4 Whales, seals, fish and man Developments in Marine Biology 1. Toxic dinoflagellate...

Author: SCHYTTE

30 downloads 961 Views 36MB 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

Developments in Marine Biology 4

Whales, seals, fish and man

Developments in Marine Biology 1.

Toxic dinoflagellate blooms edited by D.L. Taylor and H.H. Seliger, 1979 (out of print)

2.

Phytoflagellates edited by E.R. Cox, 1980 (out of print)

3.

Toxic phytoplanktonblooms in the sea edited by T.J. Smayda and Y. Shimizu, 1993

Developments in Marine Biology 4

Whales, seals, fish and man Proceedings of the International Symposium on the Biology Marine Mammals in the North East Atlantic Tromsg, Norway, 29 November-1 December 1994

Editors:

Arnoldus Schytte Blix Department of Arctic Biology, University of Tromse, Tromse, Norway Lars Wallge Institute of Physiology, University of Oslo, Blindern, Oslo, Norway 0yvind Ulltang Institute of Marine Research, Nordnes, Bergen, Norway

1995

ELSEVIER

Amsterdam - Lausanne - New York - Oxford - Shannon - Tokyo

(

0 1995 Elsevier Science B.V. All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher, Elsevier Science B.V., Permissions Department. P.O. Box 521, lo00 AM Amsterdam, The Netherlands. No responsibility is assumed by the Publisher for any injury andor damage to persons or property as a matter of products liability, negligence or otherwise, or from use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, the Publisher recommends that independent verification of diagnoses and drug dosages should be made. Special regulations for readers in the USA - This publication has been registered with the Copyright Clearance Center Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, USA. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside the USA, should be referred to the copyright owner, Elsevier Science B.V., unless otherwise specified. Developments in Marine Biology 4 ISBN 0 444 82070 1 This book is printed on acid-free paper. Published by: Elsevier Science B.V. P.O. Box 21 1 lo00 AE Amsterdam The Netherlands

Library of Congress Cataloging in Publication Data: I n t e r n a t i o n a l S y m p o s i u m on t h e B i o l o g y o f M a r i n e M a m m a l s in t h e N o r t h E a s t A t l a n t i c (1994 T r o m s s , N o r w a y ) Whales. seals. fish, and man proceedings o f the International S y m p o s i u m on t h e B i o l o g y o f M a r i n e M a m m a l s i n t h e N o r t h E a s t A t l a n t i c . T r o m s c . N o r n a y . 29 N o v e m b e r - 1 D e c e m b e r 1994 / e d l t o r s . A r n o l d u s S c h y t t e - B I i x . L a r s WalIcre. 0 1 v i n d U l t a n g . p. c m . -- ( D e v e l o p m e n t s i n m a r i n e b i o l o g y 4) I n c l u d e s indexes. I S B N 0-444-82070-1 1. M a r i n e m a m m a l s - - N o r t h A t l a n t l c O c e a n - - C o n g r e s s e s . 2. F i s h e r i e s - - N o r t h A t l a n t i c O c e a n - - C o n g r e s s e s . I. S c h y t t e - B l l x . Arnoldus. 11. Wallse. Lars. 111. Ultang. Oivind. IV. Title. V. S e r i e s . 0~713.25.157 1994 599.5'09163'3--dC20 95-91 10 CIP

.

In order to ensure rapid publication this volume was prepared using a method of electronic text processing known as Optical Character Recognition (OCR). Scientific accuracy and consistency of style were handled by the author. Time did not allow for the usual extensive editing process of the Publisher. Printed in the Netherlands

Preface In February 1988, the Board of the Norwegian Fisheries Research Council agreed to establish an integrated research programme on whales and seals in Norwegian waters. A planning group was appointed to draw up the framework for the programme and recommend how it should be organised. Members of the planning group were Professor Lars WallCe, University of Oslo; (chairman) Professor Amoldus S. Blix, University of Tromsr (senior lecturer) Per Grotnes, Norwegian College of Fishery Science, Tromsr (research officer) Morten Ryg, University of Oslo; (research director) Oyvind Ulltang, Institute of Marine Research, Bergen. A plan was presented in June 1988 and approved by both the Board of the Norwegian Fisheries Research Council, and the two relevant ministries (Environment and Fisheries). The planning group was supplemented by one representative from each of the two ministries, and was appointed as the steering group for the research programme. Dr. Nina H. Markussen later replaced Dr. Morten Ryg, Dr. Ame J. BjCrge and Mr. Helge Lorentzen were representatives of the ministries for most of the programme period, but were temporarily replaced by others. Ms. Sidsel GrCnvik was employed as secretary to the steering group. In 1988, there were two main reasons for the establishment of an extensive research programme on marine mammals. Firstly, in the preceding few years there had been considerable scientific disagreement concerning the information on which Norwegian minke whaling was based. In particular, there had been controversy over the size of the north-east Atlantic minke whale stock, and disagreement as to whether continued harvesting of the stock could be justified. In 1986, the International Whaling Commission decided to classify the northeast Atlantic minke whale stock as a "protection stock", and the Norwegian government subsequently appointed an independent group of scientists to assess the state of our knowledge and review the scientific controversy (Roy M. Anderson, UK; Raymond J.H. Beverton, UK; Ame Semb-Johansson, Norway; and Lars WallCe, Norway). The group concluded that our knowledge of the biology, population structure, and stock sizes of this species was inadequate and recommended a number of research activities designed to provide more data. Secondly, during the winter of 1986-1987 large numbers of harp seals appeared along the Norwegian coast. Altogether, about 60,000 animals were caught in fishing gear. The same happened on a somewhat smaller scale in the winter of 1987-1988. These migrations resulted in heavy economic losses for fishermen. The reasons for the mass migrations were not clear, although many hypotheses were put forward. The steering group gave highest priority to research activities which could provide more information on the questions mentioned above, but funding was also provided for research proposals related to other aspects of the management of marine mammals. The research programme was originally planned to last for five years, but was later extended for another two years until the end of 1994. The total research council

vi spending on the programme added up to more than NOK 100 million (approximately 15 million US$). We believe that the programme has resulted in significant advances in our understanding of marine mammals both in Norwegian waters and elsewhere. Since marine mammal research is a topic of international interest, the steering group decided to arrange an international symposium to mark the conclusion of the programme, to present the results and to provide an opportunity for criticism and discussion. The symposium was organised in Tromsr Norway, at 70~ in the middle of the Arctic winter (November 29 to December 1, 1994), when snow and darkness prevail, but it nevertheless attracted 150 participants from no less than 17 countries. This volume presents the proceedings of the symposium. It outlines the major findings from the Norwegian programme, and in addition contains invited contributions from international authorities on specific relevant topics as well as other communications by non-Norwegian participants. We hope that this volume will be of interest to scientists and managers as well as to the environmentally aware members of the general public.

Arnoldus Schytte Blix Lars WallOe Oyvind Ulltang

vii

Acknowledgements The editorial committee gratefully acknowledges the work of the organising committee for the symposium: Ame J. BjCrge (convener), Oyvind Ulltang, Erling S. NordCy, Sidsel Grcnvik (secretary), and the many helpers from the Department of Arctic Biology, the Norwegian College for Fishery Science, and the Norwegian Institute of Fisheries and Aquaculture, as well as the financial contributions from the Research Council of Norway and the many sponsors, all of whom contributed to a successful symposium. We would also like to express our appreciation of the secretarial assistance provided by Ms. Elin Giaever during the production of this volume.

viii

Sponsors The Research Council of Norway, Oslo The University of Tromsr The Roam Amundsen Centre for Arctic Research, University of Tromsr The Norwegian Institute of Fisheries and Aquaculture, Tromsr TromsO Kommune Norges Fiskarlag, Trondheim Norges R&fisklag, Tromsr Sildemelfabrikkenes Landsforening, Oslo Nord-Norges Sildolje- og Sildemelfabrikkers Forening, Vadsr G.C. Rieber & Co., A/S, Bergen L. Macks Olbryggeri og Mineralvandsfabrik A/S, Tromsr Odd Berg Gruppen, Tromsr

ix

Contents Preface Acknowledgements Sponsors

V

vii

...

Vlll

Stock assessment Estimating the abundance of marine mammals: a North Atlantic perspective P.S. Hammond Effective search width in shipboard surveys of minke whales in the northeastern Atlantic: concepts and methods T. Schweder and G. Hagen Point clustering of minke whales in the northeastern Atlantic G. Hagen and T.Schweder Use of mark-recapture experiments to monitor seal populations subject to catching N. @en and T. @-itsland Estimation of grey seal Halichoerus grypus pup production from one or more censuses S.-H. Lorentsen and Bakke Increased accuracy in the estimation of harp seal (Phoca groenlandica) abundance in whelping patches V.I. Chernook, V.A. Potelov and N. V. Kuznetsov Haul-out behaviour of the Norwegian harbour seal during summer R. Roen and A. Bj@rge Influences on spatial patterns of Gulf of Maine harbor porpoises D. Palka Aerial surveys of harp and hooded seal pups in the Greenland Sea pack-ice T.@itsland and N. @en

a.

Stock identity and social organization Genetic markers and whale stocks in the North Atlantic ocean: a review A. Arnason Genetic variation in northeastern Atlantic minke whales (Balaenoptera acutorostrata) A. K. Danielsddttir, S. D. Hallddrsson, S. Gudlaugsddttir and A. Arnason Preliminary results of a DNA-microsatellite study of the population and social structure of the harbour porpoise L. W. Andersen, L.-E. Holm, B. Clausen and C.C. Kinze Photo-identification of the minke whale Balaenoptera acutorostrata off the Isle of Mull, Scotland A. Gill and R.S. Fairbairns

3

13 27

35 47

53 61 69 77

91 105

119 129

X

Parasites as indicators of social structure and stock identity of marine mammals J.A. Balbuena, F.J. Aznar, M. Ferndndez and J.A. Raga Marine mammalian fatty acids: a source of information 0. Grahl-Nielsen and 0. Mjaavatten Fatty acid composition in blubber, heart and brain from phocid seals B. Fredheim, S. Holen, K.I. Ugland and 0. Grahl-Nielsen Studies of the social ecology of Norwegian killer whales (Orcinus orca) A. Bisther and D. Vongraven Possible effects of previous catch on the present population of Norwegian killer whales (Orcinus orca) D. Vongraven and A. Bisther

Distribution, diet and feeding ecology New approaches to studying the foraging ecology of small cetaceans A.J. Read Distribution and diving behaviour of hooded seals L.P. Folkow and A.S. Blix Distribution and abundance of walruses (Odobenus rosmarus) in Svalbard I. Gjertz and @. Wiig Habitat use and diving behaviour of harbour seals in a coastal archipelago in Norway A. Bjorge, D. Thompson, P. Hammond, M. Fedak, E. Bryant, H. Aarefjord, R. Roen and M. Olsen Spatial and temporal variations in northeast Atlantic minke whale Balaenoptera acutorostrata feeding habits T. Haug, H. Gj@sa?ter,U. LindstrGm, K.T. Nilssen and I. Rottingen Seasonal distribution, condition and feeding habits of Barents Sea harp seals (Phoca groenlandica) K.T. Nilssen Food consumption of the Northeast Atlantic stock of harp seals E.S. Nordoy, P-.E. Mirtensson, A.R. Lager, L.P. Folkow and A.S. Blix Historic variation in the diet of harp seals (Phoca groenlandica) in the northwest Atlantic J. W. Lawson and G.B. Stenson Seasonal and regional variations in the diet of harbour seal in Norwegian waters M. Olsen and A. Bjorge Feeding ecology of harp and hooded seals in the Davis Strait - Baffin Bay region F. 0. Kapel

133 141 153 169

177

183 193 203

21 1

225

24 1 255

26 1

27 1

287

xi

Energetics and other physiological aspects Food requirements of Northeast Atlantic minke whales E.S. Nordgy, L.P. Folkow, P.-E. MGrtensson and A.S. Blix Energetics of pregnancy, lactation and neonatal development in ringed seals (Phoca hispida) C. Lyde rsen Harp and hooded seals - a case study in the determinants of mating systems in pinnipeds K.M. Kovacs Consumption of cod by the Northwest Atlantic grey seal in Eastern Canada M.O. Hammill, M.S. Ryg and B. Mohn Digestive physiology of minke whales S.D. Mathiesen, T.H. Aagnes, W. Sgrmo, E.S. Nordgy, A.S. Blix and M.A. Olsen Body condition of fin whales during summer off Iceland G.A. Vikingsson Seal adaptations for long dives: recent studies of ischemia and oxygen radicals R. Elsner, S. Qyaseter, O.D. Saugstad and A.S. Blix Pineal functions in newborn seals K.-A. Stokkan Changes in metabolic rate and body composition during starvation and semistarvation in harbour seals N.H. Markussen Variation in the metabolic rates of captive harbour seals D. Rosen and D. Renouf Population dynamics Population dynamics: species traits and environmental influence C.W. Fowler Interpretation of growth layers in the periosteal zone of tympanic bulla from minke whales Balaenoptera acutorostrata I. Christensen On the life history and autecology of North Atlantic rorquals J. Sigurjdnsson Aspects of the biology of the harbour porpoise, Phocoena phocoena, from British waters C. Lockyer Aspects of reproduction and seasonality in the harbour porpoise from Dutch waters M.J. Addink, T.B. Sgrensen and M. Garcia Hartmann Migration strategy of southern minke whales to maintain high reproductive rate H. Kato

307 319 329 337

35 1 36 1 37 1 377 383 393

403 413 425 443 459 465

xii Overview of cetacean life histories: an essay in their evolution T. Kasuya Modelling the school structure of pilot whales in the Faroe Islands, 18321994 D. Bloch and L. Lustein Harp seals as indicators of the Bqrents Sea ecosystem Y.K. Timoshenko

Interactions with fisheries Interactions between marine mammals and fisheries: an unresolved problem for fisheries research T.D. Smith Strategies to reduce the incidental capture of marine mammals and other species in fisheries M.A. Hall Ecological implications of harp seal Phoca groenlandica invasions in northern Norway T. Haug and K.T. Nilssen Aspects of the sealworm Pseudoterranova decipiens life-cycle and sealfisheries interactions along the Norwegian coast K. Andersen, S. des Clers and T. Jensen Grey seal (Halichoerus grypus Fabr.), population biology, food and feeding habits, and importance as a final host for the life-cycle of sealworm (Pseudoterranova decipiens Krabbe) in Icelandic Waters E. Hauksson and D. Olafsddttir Pollutants, toxicology and epizootics Toxicological and epidemiological significance of pollutants in marine mammals P.J.H. Reijnders and E.M. de Ruiter-Dijkman Organochlorine contaminants in marine mammals from the Norwegian Arctic J.U. Skaare Seasonal variation of organochlorine concentrations in harp seal (Phoca groenlandica) L. Kleivane, 0. Espeland, K.I. Ugland and J. U.Skaare Biomarkers in blood to assess effects of polychlorinated biphenyls in freeliving grey seal pups B.M. Jenssen, J. U.Skaare, S. Woldstad, A.T. Nastad, 0. Haugen, B. Kloven and E.G. Sormo Toxic, essential and non-essential metals in harbour porpoises of the Polish Baltic Sea P. Szefer, M. Malinga, W. Czarnowski and K. Skdra

48 1 499 509

527 537 545 557

565

575

5 89 599

607 617

...

Xlll

Assessment of the vulnerability of grey seals to oil contamination at Froan, Norway M.Ekker, D. Vongraven, B.M. Jenssen and M. Silverstone Cytochrome P450 in marine mammals: isozyme forms, catalytic functions, and physiological regulations A. GoksQyr Serological investigation of morbillivirus infections in minke whales (Balaenoptera acutorostrata) S. Stuen and P. Have

Management and cultural, social and economic aspects of exploitation The International Whaling Commission’s Revised Management Procedure as an example of a new approach to fishery management J.G. Cooke Multispecies modelling and management with reference to the Institute of Marine Research’s multispecies model for the Barents Sea 0.Ulltang The management of Irish waters as a whale and dolphin sanctuary E. Rogan and S. D. Berrow The scientific background for the management of monodontids in West Greenland M.P. Heide-JQrgensen Marine mammals in the culture of Norwegian coastal communities A. Kalland Management of whaling in coastal communities R. Gambell Impacts of modern seal invasions S. Eikeland

623 629 64 1

647 659 671 683 689 699 709

Index of authors

715

Keyword index

717

This Page Intentionally Left Blank

Stock assessment


9 1995 ElsevierScienceB.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 13. Ulltang,editors

Estimating the abundance of marine mammals: a North Atlantic perspective P.S. H a m m o n d Sea Mammal Research Unit, Natural Environment Research Council, Cambridge, UK Abstract. This paper investigates how studies to estimate the abundance of marine mammals have used the characteristics and behaviour of the target population to help determine the most appropriate method for any specific study. It also addresses how these studies have attempted to overcome the problems associated with ensuring that important assumptions of the method are not violated. Three basic methods used for estimating marine mammal abundance are described: extrapolating counts; mark-recapture analyses of photo-identification data; and sightings surveys. Recent studies in the North Atlantic to estimate abundance include: harbour seals and bottlenose dolphins in the Moray Firth, Scotland; grey seals in the North Sea and the breeding population in Britain; North Atlantic humpback whales; and the NASS and SCANS sightings surveys for whales and small cetaceans in the North Atlantic and North Sea, respectively. The paper demonstrates the wide range of applicability of the available methods, the way that the example studies addressed and (in some cases) overcame potential problems, and the modification or extension of "standard" methodology to fit particular circumstances. Key words: population size, extrapolating counts, sightings surveys, line transect, photo-identification, mark-recapture, model assumptions, North Atlantic

Introduction There are a number of ways to estimate animal abundance that are applicable to marine mammals. But any application of these methods must take into account the inconvenient fact that pinnipeds and cetaceans spend most (or all) of their time at sea and a large proportion of that time underwater. However, studies to estimate marine mammal abundance can take advantage of the variation in the physical and behavioural characteristics of species to arrive at the most appropriate method for a particular case. In this context, pinnipeds spend a proportion of their time on land (or ice) and some of these haulout periods are predictable, particularly during the pupping season. Being out of the water avoids the complication of animals diving out of sight. This means that they can actually be counted. Methods for estimating seal numbers make use of haulout behaviour thus avoiding the problems of sampling at sea. In contrast, cetaceans spend their entire lives at sea. Direct counts are not possible and methods for estimating their abundance must rely on sampling. This has received increasing attention during the last 15 years, particularly through the forum of the IWC Scientific Committee [1,2]. This paper investigates how studies to estimate the abundance of marine mammals have used the characteristics and behaviour of the target population to help deAddress for correspondence: Sea Mammal Research Unit, Natural Environment Research Council, High Cross, Madingley Road, Cambridge CB3 0ET, UK.

termine the most appropriate method for any specific study. But even the most appropriate method will have its drawbacks, particularly the difficulties in ensuring that important assumptions of the estimation model are not violated. This paper also addresses, therefore, how these studies have attempted to overcome these difficulties. The examples chosen are all drawn from the North Atlantic, reflecting the focus of this Symposium and the experience of the author. This paper is not a review but, in describing these example studies, it is also intended to give the reader an idea of the range of work undertaken in the North Atlantic in recent years.

Methods for estimating abundance A fundamental assumption of any method of estimating abundance is that the data must be representative of the study population or area. This applies to all the methods described below.

Extrapolation of counts Rarely, if ever, can it be possible to enumerate an entire population of animals. This is certainly true for marine mammals which are effectively invisible for a large proportion of their lives. However, seals give birth and wean their pups on land or ice, and also haul out during their annual moult and on other occasions, and it is possible during these periods to count the number of animals present in an area. But all the animals in the population will not be hauled out at the same time so it is necessary to extrapolate the number of seals counted to obtain an estimate of the entire population. There are a number of ways of doing this. One is to choose a class of animal which can be counted (pups of the year, for example) and to use those numbers as input to an age- or stage-structured population model which calculates total abundance. An example of this method is given below. For a reliable estimate the population model must be realistic with parameter values estimated from relevant data. Another method of extrapolation is to count all the animals hauled within a given period and to divide this number by an estimate of the proportion of the population which is hauled out during this time. One way that this proportion can be estimated is by using telemetry data. An example of this method is given below. Assuming accurate counts can be made, these methods are most sensitive to the quality of data used to extrapolate counts to total abundance.

Mark-recapture and photo-identification Mark-recapture methods use data on the number of animals marked and the proportion of marked animals in samples of recaptured animals to estimate population size. They makes several basic assumptions about the data including:

marks are unique, cannot be lost and are always reported on recovery; and marking does not affect the subsequent catchability of an animal. In the past, mark-recapture analyses have been used to estimate the population size of marine mammals using data from artificially applied marks. More recently, studies have concentrated on populations of animals in which unique individual natural markings can be photographed and identified. Records of re-identifications from a series of photographic samples (capture histories) can then be used as data for mark-recapture analyses to estimate population size. Three examples of this are given below. Photo-identification has some advantages over the use of artificial marks [1] but obviously cannot be used for those populations whose individuals do not possess recognisable natural markings. In mark-recapture analyses, the assumption most likely to be violated is that each animal must have the same probability of being captured within a sampling occasion. It is very likely that the behaviour of individual animals will lead to heterogeneity of capture probabilities resulting in population estimates which are biased downwards. In the extreme, some animals may never be available to be sampled and will not be included in the population estimate. There are ways to account for this analytically if closed population models can be used but it is clearly better to minimise the problem during data collection. This means designing a study which gives every animal a chance of being captured, and capturing as many animals as possible. This can be achieved more easily if the distribution of the study population is concentrated in a limited area for a period during the year.

Sightings surveys Line transect methods [2,3] were first developed for terrestrial animals but are now widely used to estimate the abundance of cetacean populations via shipboard or aerial surveys. In these sightings surveys, the study area is sampled by the survey platform searching along predetermined transects, placed so that the whole area is representatively sampled. When sightings are made, data are collected which allow the calculation of perpendicular distance from the sighting to the transect line. These data are used to estimate the effective width of the strip searched so that sample density can be estimated and extrapolated to give an estimate of abundance in the whole study area. Sighting surveys thus provide an estimate of the number of animals in an area at a given time, not an estimate of the size of a biological population unless the whole of that population was in the study area during the survey period. The most important assumption made by the line transect sampling method is that every animal (or group of animals if groups are the sighting target) on the transect line itself is seen. That is, the probability of detecting an animal at zero perpendicular distance is unity. Clearly, this can never be the case for cetaceans, which are underwater most of the time. In order to arrive at an unbiased estimate of abundance, therefore, it is necessary to estimate this probability (known as g(0) in the literature). The best way to do this is through the analysis of duplicate sightings data collected

from independent sightings platforms. However, sightings surveys do not always collect data for estimating g(0) and in these cases, abundance will be negatively biased by an unknown amount, notwithstanding other potential biases. An alternative method of data collection and analysis for sightings surveys is that of "cue counting" [4]. This method involves estimating the density of cues (whale blows, for example) per unit area searched per unit time and dividing by an estimate of the rate at which animals of the study species provide cues. This approach can only be used for species which exhibit clear cues. But it has the advantage that the equivalent of g(0) is the probability that all cues at zero (radial) distance from the survey platform will be seen, which is far more likely to be satisfied. Cue counting works best from aerial platforms.

Studies to estimate abundance in the North Atlantic

Harbour seals in the Moray Firth The Moray Firth in northeast Scotland is home to a population of harbour seals which regularly haul out on inter-tidal sandbanks, particularly during the June/July pupping season. This has made it possible to count seals at all known sites in the area during this period of the year with the aim of estimating population size (personal communication). Sex-specific estimates of the proportion of seals hauled out during this time were estimated from VHF telemetry data collected from 26 seals of a range of sizes. Pups were not included in the estimate because they were being born throughout the counting period. The estimate of the average number of seals hauled out of 1007 was extrapolated to an estimate of population size (excluding pups) of 1651 (CV = 0.104). The important factors which had to be taken into account in this study were" the timing of the counts with respect to time of year, time of day and stage of tide; the need to ensure that double counting was not occurring; and the representativeness of the telemetry data. Peak haulout counts were made during the pupping season and 2 h either side of low tide. There was no diurnal pattern. This was in contrast to a similar study in Orkney where peak haulout counts were made during the moult, there was no tidal relationship but time of day had to be taken into account [5]. Double counting was judged minimal based on the movement patterns of individuals. The telemetry data were collected over a number of years and the data for females was biased towards pregnant females so there is a possibility that they were not representative and that the population estimate may be biased. Obtaining representative telemetry data for any study is often difficult and is, perhaps, the major drawback of this method of estimating abundance.

The British grey seal breeding population The number of grey seals breeding around the coasts of Britain is estimated each

year based on counts of pups born during autumn [6]. High resolution aerial photography is used to record all the pups on land at all known major breeding sites, several times during the pupping season. Pup production is estimated by fitting a model, with parameters defining birth, death and time of leaving the site, to the pup counts for each site. The number of females in the population is estimated by feeding the annual pup production figures into an age-structured population model. Total population size is obtained via a simple sex ratio calculation. The most important factors affecting this study are: the quality of the aerial photographs; the adequacy of the model used to estimate pup production from the pup counts (including the number of counts made each year at each site); and the adequacy of the data in the population model. The development of a purpose-built camera system including image-motion compensation [7] has resulted in excellent quality aerial photographs. The model to estimate pup production is sensitive to the timing and length of the pupping season. These vary from site to site, and so pup production is estimated separately for each site. Estimates of population size from the model are sensitive to the fecundity rate data and the sex ratio. The fecundity data were collected from one area in 1981 and it is not known how representative they are to other areas and the present day. There are few data on the sex ratio, which is assumed to be 1:1. Grey seals in the North Sea

The method described above estimates the size of the grey seal population breeding around Britain. But the distribution of seals outside the pupping season, when they are building up energy reserves for the next season, may be different. And it is the number of seals foraging in an area which is of most relevance to concerns about interactions with fisheries. To address this, a study using photo-identification methods to obtain data for mark-recapture analyses has been conducted to estimate the number of seals associated with haulouts along the central North Sea coast of Britain [8]. The pelage patterns of grey seals (Fig. l a) were photographed at 2-week intervals during the summer months at three major haulout areas in 1991-1993. Normally, only adult female grey seals are well enough marked to be recognisable in a subsequent photograph. This is an extreme form of heterogeneity of capture probabilities and to account for it the proportion of well-marked animals present was recorded during each sampling occasion. The pelage pattern was extracted from a standard area of the seal's head and neck by specifically developed image processing software which compensates for differences in the viewpoint of the photograph and the posture of the seal [9]. These standard images served as "marks". All pairs of photographs were compared by another computer program which generated a set of similarity measures, the highest of which were compared by eye to determine genuine matches. In this way, capture histories were established for all individually recognisable seals. A mark-recapture model was developed specifically for these data, structured to

Fig. 1. Examples of natural markings used for photo-identification studies in the North Atlantic: (a) grey seal (photo: Lex Hiby, Sea Mammal Research Unit); (b) bottlenose dolphin (photo: Ben Wilson, University of Aberdeen); (c) humpback whale (photo: Tony Martin, Sea Mammal Research Unit).

allow for movement between the three sampling areas and to account for animals identified from the left or right side [8]. The data showed that mixing of animals in the population occurred rapidly so that it was possible to estimate population size within each year using a model ignoring births and deaths. The estimates were corrected by an estimate of the proportion of well-marked seals in the population. This assumes that males and immature seals spend the same proportion of their time at haulout sites as do adult females. Limited telemetry data on grey seals indicate that this is the case but if it is not, the estimate will be biased. Heterogeneity of capture probabilities within the well-marked part of the population was probably not a problem because sampling covered all major haulout sites in the area and over 50% of this class of animals were identified. In addition, data collected outside the study area indicated little exchange of animals.

Bottlenose dolphins in the Moray Firth In the Moray Firth, one of the aims of a study of a small and isolated population of bottlenose dolphins is to estimate population size. The dolphins can be identified by nicks on the dorsal fin and/or by pigmentation patterns on the skin (Fig. l b) [10]. This is thus an ideal situation for mark-recapture analysis of photo-identification data. As in the case of grey seals described above, left and right sides of an animal are different and were sampled separately. Not all animals have natural markings so data were collected to estimate the proportion which do. The population was sampled twice each month during the summers of 1990-1992 and matches were determined by eye. The first population estimates have recently been calculated (personal communication). Analysis was via a multi-sample closed population model applied to the data for each summer. This allowed the use of identifying marks which were known to last longer than a summer but not necessarily more than a year, thus increasing sample sizes, and allowed heterogeneity of capture probabilities to be taken into account using the program, CAPTURE [11 ]. The model allowing heterogeneity gave consistently higher estimates indicating that this did need to be accounted for. The estimate in 1992, when sampling was extended to cover a larger proportion of the known range, was higher indicating that there were some site preferences within the study area and that results from this year should be used. The estimate for the right side (lower CV than left side) for 1992 using the heterogeneity model was selected as the best estimate. This study is a good example of the use of an extensive and detailed data set to address and take account of some of the potential problems of applying mark-recapture methods of photo-identification data.

North Atlantic humpback whales At the other extreme of scale is an ongoing multi-national study of humpback whales in the North Atlantic. One aim of project YoNAH (Years of the North Atlantic Humpback) is to estimate the size of the whole population via mark-recapture

10 analyses of data from photo-identification of fluke patterns (Fig. l c). Humpbacks breed in winter in the West Indies, migrating to summer feeding areas in high latitudes. There is known to be a high degree of fidelity of an animal to a particular feeding area [ 12]. Project YoNAH sampled whales in the breeding areas and again in all known major feeding areas (Gulf of Maine, Canada, West Greenland, Iceland and Norway) in 1992 and 1993. Although humpback whales occur across the whole North Atlantic, the chosen abundance estimation method is the most appropriate because all animals are well-marked and available to be sampled in relatively discrete and accessible areas. The photographs are currently being processed by eye. In 1992, 819 whales were identified in the West Indies and 855 in feeding areas. In 1993, there were 773 identifications from the West Indies and over 1000 are expected from the feeding areas. Next, the most appropriate mark-recapture model to analyze this particular data set will be developed. The options are limited by the number of sampling occasions but there is a large amount of detailed data which it is hoped can be used to test whether some of the standard assumptions have been violated and to take account of this, if necessary. For example, sampling in the feeding areas was unbalanced; how important was this? In addition, it is expected that analyses of data in the breeding areas may be affected by different behavioural characteristics of animals in different social groups, a form of heterogeneity of capture probabilities. North Atlantic sightings surveys

In 1987 and 1989, several countries in the North Atlantic collaborated in the NASS (North Atlantic Sightings Survey) projects, and a third is planned for 1995. The area covered by the 1987 and 1989 surveys is shown in Fig. 2. Ships were the main survey platform but some areas were covered by aerial survey to estimate the abundance of the target species: fin, minke, sei and pilot whales. The ship surveys used line transect sampling and the aerial surveys used cue-counting methods. On the ships, no data were collected to allow the estimation of g(0) in 1987, but in 1989 duplicate sightings data were collected on some vessels. A new "maximum simulated likelihood" method of analysis was developed to estimate g(0) for minke whales by integrating experimental and survey data [13]. NASS is a good example of one way to enable a large area of ocean to be surveyed in order to estimate cetacean abundance. Estimates of the target species could only be obtained from sightings surveys and a collaborative effort was important in order to cover as much of the North Atlantic as possible. Estimates of abundance for all the target species have been accepted by the IWC Scientific Committee although there is currently a debate about the estimate of g(0) for minke whales. For the other species, g(0) was not estimated and was assumed to be equal to one. Small cetacean abundance in the North Sea

Another, recently conducted, multi-national sightings survey in the North Atlantic

11

% % .

@

O'o/. zr \

6~'~

20.t,/

I 20"~ o* Fig. 2. Areas covered by the 1987 and 1989 NASS sightings surveys in the central and eastern North Atlantic and the 1994 SCANS sightings survey in the North Sea and adjacent waters. was SCANS (Small Cetacean Abundance in the North Sea). SCANS targeted harbour porpoises and other small cetaceans (and including minke whales) in the North Sea and adjacent waters in summer 1994. The area surveyed (shown in Fig. 2) was sampled by nine ships and two aircraft. Harbour porpoises are difficult to detect and it was expected that the probability of detecting them on the transect line would be small, so it was important to obtain good estimates of g(0). Duplicate sightings data were, therefore, collected on all ships. The two searching platforms operated all the time; one searched in "normal" mode, the other tracked sightings detected further ahead. The two aircraft flew in tandem (one directly behind the other) for much of the time, allowing duplicate sightings data to be collected during the aerial survey. New methodology was developed to analyze the duplicate sightings data from the ships and the aircraft. At the time of writing, the data are still being analyzed. It is hoped that the intensive coverage of the survey area and the extensive duplicate sightings data collected will allow the calculation of abundance estimates of higher than usual precision and accuracy.

Concluding remarks

The studies described above demonstrate the range of and variation within the methods available to estimate the abundance of marine mammal populations. In each example, the method was chosen which was most appropriate to the physical and

12

behavioural characteristics of the study population. In all cases, those conducting the research were aware of the particular difficulties of the chosen method and in some cases were able to investigate potential violations of the assumptions. And in several examples, "standard" methods were modified to adapt existing techniques to fit the particular situation. The studies described also demonstrate that there has been much work in recent years to increase our knowledge of marine mammal abundance in the North Atlantic. And the examples given, although important ones, are not exhaustive. Future attempts to estimate pinniped or cetacean population size should take note of the successes, and problems, detailed in the work described above.

References 1. Hammond PS. Estimating the size of naturally marked whale populations using capture-recapture techniques. Rep Int Whal Commn 1986;(Special Issue 8):253-282. 2. Hiby AR, Hammond PS. Survey techniques for estimating abundance of cetaceans. Rep Int Whal Commn 1989;(Special Issue 11):47-80. 3. Buckland ST, Anderson DR, Burnham KP, Laake JL. Distance Sampling: Estimating Abundance of Animal Populations. London: Chapman and Hall, 1993. 4. Hiby AR. An approach to estimating population densities of great whales from sightings surveys. IMA J Math Appl Med Biol 1985;2:201-220. 5. Thompson PM, Harwood J. Methods for estimating the population size of common seals (Phoca vitulina). J Appl Ecol 1990;217:281-294. 6. Ward AJ, Thompson D, Hiby AR. Census techniques for grey seal populations. Symp Zool Soc London 1987;58:181-191. 7. Hiby AR, Thompson D, Ward AJ. Census of grey seals by aerial photography. Photogram Rec 1988;12:589-594. 8. Hiby AR. Abundance estimates for grey seals in summer based on photo-identification data. In: Hammond PS (ed.) Grey Seals in the North Sea and their Interactions with Fisheries. Report to UK Ministry of Agriculture, Fisheries and Food under contract MF 0503, 1994. 9. Hiby AR, Lovell, P. Computer-aided matching of natural markings: a prototype system for grey seals. Rep Int Whal Commn 1990;(Special Issue 12):57-61. 10. Thompson PM, Hammond PS. The use of photography to monitor dermal disease in wild bottlenose dolphins (Tursiops truncatus). Ambio 1992;21:135-137. 11. White GC, Anderson DR, Burnham KP, Otis DL. Capture-recapture and removal methods for sampling closed populations. Los Alamos, NM: Los Alamos National Laboratory, 1982. 12. Katona SK, Beard JA. Population size, migrations and feeding aggregations of the humpback whale (Megaptera novaeangliae) in the western North Atlantic Ocean. Rep Int Whal Commn 1990;(Special Issue 12):295-305. 13. Schweder T, HCst G. Integrating experimental data and survey data to estimate g(0): a first approach. Rep Int Whal Commn 1992;42:575-582.

9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 0. Ulltang,editors

13

Effective search width in shipboard surveys of minke whales in the northeastern Atlantic: concepts and methods Tore Schweder ~ and Gro Hagen 2

_

1University of Oslo, Blindern, Oslo, Norway; and 2Norwegian Computing Center, Blindern, Oslo,

Norway

A b s t r a c t . Concepts and methods for estimating the effective search width from survey data of the line

transect type, independent observer data, dive time data and other data sources are reviewed and further developed. It is argued that effective search width is the primary parameter, which should not be estimated by separate estimates of g(0) and f(0) in cases where the hazard of sighting has been estimated. To estimate the hazard probability of sighting in complex models, systematic simulation is used. Occurrence exposure data for estimating the hazard probability of sighting are obtained from independent observer experiments by an automatic identification rule based on measurement error data. To remove bias introduced by imperfections in the chosen rule, the independent observer experiments are simulated for spatially distributed whales with density and clustering estimated from the survey data. Measurement errors are included in the model when estimating the hazard probability of sighting. The effective search width is calculated as the sighting rate in the fitted simulation model. Emphasis is on methodology for future surveys. Key words: baleen whales, abundance estimation, sighting, line transect survey

Introduction

Shipborne surveys of whales are cases of line transect experiments (see [1,2]). In standard line transect theory, the assumptions are that (i) objects on the line are detected with certainty; (ii) objects are detected at their initial location, and (iii) measurements of time and positions are exact. An additional assumption is: (iv) cues are, with certainty, identified with individual objects. When the objects are minke whales in the north Atlantic, and when the survey is carried out by ship, none of the four assumptions are likely to be completely satisfied. A whale is only available for sighting in the brief moments it breaks the surface for breathing. Observations are made from the barrel, some 16 m above sea surface, with two trained observers using the naked eye. In the Norwegian surveys, the vessels move at a speed of 10 knots, when the sighting condition is satisfactory (ca. 25% of the time). The northern minke whale does not show itself by a visible blow, and it is not easy to see the black back of a surfacing whale. Whales dive, and it is quite possible to pass a whale which is submerged during the period of possible sighting. With g(y) being the probability of detecting a whale at perpendicular distance y from the transect, assumption (i) is that g(0) = 1. When this assumption fails, g(0) must be estimated. Data, concepts and methods for estimating g(0) are discussed in some detail. We argue, however, that the effective search width, 2w, is the parameter of

Address for correspondence: T. Schweder, University of Oslo, Box 1095 Blindern, 0317 Oslo, Norway.

14 primary interest, and this parameter should rather be estimated directly than through separate estimates of g(0) and the probability density at zero of observed perpendicular distances, f(0) = g(O)/w. The basic concept in our model is the hazard probability of sighting. This is the conditional probability of successfully sighting a surfacing whale, given that the whale was not previously sighted. Having estimated this probability as a function of surfacing position, and possibly other observables, the effective search width is estimated by counting the successes in a simulation experiment. Here, whales scattered over the area are surfacing according to observed dive time data, and they are sighted according to the estimated hazard probability of sighting. Whales are moving about. Assumption (ii) is essentially that their movements are random, with no specific directional drift, or without being affected by the observer. If whales are attracted by the approaching vessel, or if they try to avoid it, assumption (ii) is broken. This problem is not further discussed. When a whale has been sighted, the observer presses a button connected to a computer to record the time of the sighting. He then measures the radial distance to the surfacing by eye, and he measures the angle between the sighting line and the transect line with the help of an angle board fixed at the barrel rim. None of these measurements are taken without error. Separate experiments are needed to estimate the error distribution, with sighting of radar buoys or other objects similar to a surfacing minke whale at known positions, and under conditions similar to the survey. In the experiment reported in [3], the vessel moved towards radar buoys at a speed of 10 knots, and the observer was allowed to observe the buoys at discrete points in time. Such experiments serve the additional purpose of training the observers. We use data from Ref. [3] for various purposes, but we do not discuss measurement error experiments any further. We do, however, stress the importance of consistency: if measurement errors are present, substantial errors can occur if it is switched inconsistently between measured and true coordinates for relative whale positions throughout the analysis. In our case, the cues are visible surfacings of minke whales. Cues might be misinterpreted. The problem is not that other species of whales are mistaken for minke whales, but rather that series of minke whale cues erroneously might be linked together as tracks for individual whales. The identification of tracks of cues is done online, and it is based on whale behaviour, direction of movement, size and other visual aspects of the cue. This identification might be affected by two types of errors: the series of sighted cues belonging to one whale is broken into several tracks; and the series of more than one whale are linked together in one track. It is possible to develop a cue counting method which will help to remove bias introduced by erroneous track identification. Space does not, however, allow this to be done in the present paper. The primary purpose of this paper is to present concepts and methods for the shipborn survey of minke whales in Norwegian water, planned for the summer of 1995, and for similar surveys. The plan is to cover the complete northeastern Atlantic management area. We go a bit further than in Ref. [4], where the independent ob-

15 server experiments in 1989 and 1990 were analysed, and an estimate of abundance was presented. In the 1995 survey, the plan is to have two barrels in place on each vessel. This will allow the independent observer experiments to be concurrent with the survey. Heterogeneity in observational conditions, and in sighting efficiency between vessels and teams, will thus be less of a concern than assumed in Ref. [ 11 ]. Due to the criticism raised in Ref. [6], the interpretation of data from independent observer experiments is discussed in extra detail. We present an automated procedure for so-called duplicate identification. This procedure is used on the observed data from the 1990 independent observer experiment (see [5]). The classification found by using the automated rule agrees well with the classification actually used in Ref. [4]. To remove bias due to subjectivity in the classification rule, simulation based estimation is required. This approach is explained, but not carried out. The method to use is that of maximum simulated likelihood [7], which was used in Ref. [4] to estimate g(0).

The search width problem The detection curve, g(y), represents the conditional probabilities of detecting a whale, given that it is at perpendicular distance y _> 0 from the line. The effective search width, 2w, is that width for which the strip of area a = 2wL contains the same expected number of objects as is observed during a transect of length L. When the transect line is randomly placed in the area, and D is the average whale density per unit area, the expected number of sightings is

DL2~g(y) dy The effective half-width is therefore W--

~g(y) dy

If n whales are observed during a transect of length L through an area of size A, and an estimate ff is available for w, the abundance of whales in the area is estimated by ^

An

NA = 2~L

(1)

Perpendicular distances between observed whales and the line has probability density, f(y), which is proportional to the detection curve, f ( y ) = g(y)/w. The effective half-width therefore satisfies g(0) w = ~ f(0)

(2)

16 This relation has been used to estimate w, by first estimating f(0) by fitting a parametric probability density to the observed perpendicular distances in the survey, and then separately to estimate g(0) from data gathered in an independent observer experiment [1,2,4]. The effective search width is, however, a more fundamental concept than g(0). If a direct estimate of w is available, as is the case when the spatial hazard probability of sighting has been estimated, this estimate should be used in density or abundance estimation. There are two main reasons for not estimating f(0) and g(0) separately, and then computing w: (i) it is notoriously difficult to estimate the density at the extreme end, 0, of the range. Erroneous model specification for the observed perpendicular distances can actually introduce quite severe bias in the estimate of f(0). The second reason is (ii) that extra variability is introduced in the abundance estimate. If the hazard probability has to be estimated, both w, f(0) and g(0) are implicitly determined. To estimate f(0) separately from the positions of initial observations, which already have been used when estimating the hazard probability, amounts to adding variability, and it will make the estimation of uncertainty more difficult. In Ref. [4], we were in partial breach with this advice (see section 7). The abundance estimate is sensitive to the estimation of g(0). For the previous surveys of minke whales in the northeastern Atlantic, g(0) was estimated to 0.36 [4]. The abundance estimate not corrected for g(0) (setting g(0)= 1, as assumed in the standard theory) was therefore increased by a factor of 1/0.36 = 2.78.

Hazard

probability

and simulation

Minke whales are available for sighting only at the discrete points in time when they are up breathing. The concept of hazard probability is therefore appropriate when modelling the observational process. Let Q(p) be the conditional probability that a whale surfacing at position p relative to the observer platform is sighted, given that the whale was not previously sighted from the platform. If Pi is the relative position of the whale at its ith surfacing, I I ( 1 - Q(Pi)) is the probability that the whale is passed without being sighted at any of its surfacings. If the whale stays at a fixed absolute position, the relative positions a r e Pi=(Xo-vTi,y), where the first coordinate is its position along the transect direction and the second coordinate is the perpendicular position. Here, Xo is the forward distance at time 0, Ti is the time of the ith surfacing and v is the speed of the observer platform. If the surfacings follow a Poisson process in time, with surfacing rate a, the probability generating functional of the Poisson point process yields (E is the expectation operator)

g(y) =

1 - Eli

(1 -

Q(x o -

vT/, y)) = 1 - e x p ( - c t v

IoQ(X,y) dx)

(3)

It is here assumed that no whale is sighted behind the observer platform, Q (x,y) = 0, x1 year) leaves the normal cycle, and its members are therefore not available for the next year's moulting season. The assignment of a pup to either of these components takes place at birth, for example determined by

44 location in a breeding patch, where those at the outskirts might be more prone to be taken away by prevailing weather and ice conditions. It can then be shown [8] that: (i) the return rate of temporarily emigrated animals will determine how fast the mark-recapture estimates approach the true value over years; and (ii) the proportional distribution of tags on emigrants and non-emigrants during tagging determines whether the true estimate will be approached from below or above. This motivates the following procedures for alternative estimations of pup production: (1) given that the first year after tagging generally is the more sensitive to the assumptions of uniform mixing, estimates might be calculated by deleting data from the first year after tagging; (2) by the assumption that all emigrants will return to the population when sexually mature, estimates might be calculated by deleting data from all the 5 years following a tagging. Although with a scanty data series, such tentative estimates were presented by Oien and Oritsland [8] and indicate that the return rate is rather small the first year, but increasing with the years, giving some merit to the suggested method (2). Unfortunately, several years will pass until the cohorts arrive at the truncation point, and thereafter several years are needed to obtain reasonable sample sizes for the estimates. This procedure might of course be of limited value for management questions in the short term, but in a longer time perspective might prove to be the only way of monitoring on a regular basis such remote seal populations as the harp seals breeding in the Greenland Sea.

Acknowledgements The tagging of harp seals during the Sea Mammal Research Programme has been funded by the Norwegian Council of Fisheries Research (NFFR), project number 4001-701.304.

References 1. RasmussenB, Oritsland T. Norwegian tagging of harp and hooded seals in North Atlantic waters. Fiskeridir Skr 1964;13(7):43-55. 2. BowenWD, Sergeant DE, Oritsland T. Validation of age estimation in the harp seal, Phoca groenlandica, using dentinal annuli. Can J Fish Aquat Sci 1983;40:1430-1441. 3. Bowen WD, Sergeant DE. Mark-recapture estimates of harp seal pup (Phoca groenlandica) production in the Northwest Atlantic. Can J Fish Aquat Sci 1983;40:728-742. 4. Stenson GB, Myers RA, Hammill MO, Ni I-H, Warren WG, Kingsley MCS. Pup production of harp seals, Phoca groenlandica, in the northwest Atlantic. Can J Fish Aquat Sci 1993;50:24292439. 5. Oritsland T, Oien N. Aerial surveys of harp and hooded seal pups in the Greenland Sea pack-ice. In: Blix AS, WallOe L, Ulltang 0 (eds) Whales, Seals, Fish and Man. Amsterdam: Elsevier, 1995;77-83. 6. Oien N. Update of mark-recapture estimates of harp seal pup production in the Greenland Sea. Working paper to the Joint ICES/NAFO Working group on harp and hooded seals, WP SEA-47, Copenhagen, September, 1993.

45 7. Oien N, Oritsland T. Recaptures of harp seals (Phoca groenlandica) tagged as pups in the Greenland Sea; pup production and dispersion patterns. Working paper to the Joint ICES/NAFO Working group on harp and hooded seals, WP SEA-33, Copenhagen, October, 1991. 8. Oien N, Oritsland T. Using mark-recapture methods to estimate pup production of harp seals (Phoca groenlandica) in the Greenland Sea. ICES CM 1992/N:10- Ref.D. 9. Meisfjord J, Fyllingen I, N~evdal, G. A study of genetic variation in Northeast Atlantic harp seals (Pagophilus groenlandicus). ICES CM 1991/N:5. 10. Terhune JM. Geographical variation of harp seal underwater vocalizations. Can J Zool 1994;72:892-897. 11. Bergstad OA, Jr T, Dragesund O. Life history and ecology of the Gadoid resources of the Barents Sea. Fish Res 1987;5:119-161. 12. Seber GAF. The Estimation of Animal Abundance and Related Parameters. London: Charles Griffin, 1982. 13. Oien N. Age compositions in 1982 to 1984 samples from breeding and moulting harp seals in the West Ice, with an evaluation of the age determinations. Working paper to the ICES Working group on harp and hooded seals in the Greenland Sea, SGS WP 12, Copenhagen, October, 1987.


9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCe and ~. Ulltang, editors

47

E s t i m a t i o n of grey seal Halichoerus grypus p u p p r o d u c t i o n f r o m o n e or m o r e c e n s u s e s Svein-HS.kon Lorentsen and O y v i n d B a k k e Norwegian Institute for Nature Research, Tungasletta, Trondheim, Norway Halichoerus grypus~ the spread of pupping dates exceeds the length of the stay ashore of individual pups. Thus, no single census will include all pups born. To cover the whole pupping season is time consuming and expensive. We therefore developed a method, based on maximum likelihood estimation, to estimate annual pup production from one or more censuses in which the pups are classified according to age. The results of using this method on single censuses in the 1993 breeding season in the Froan Nature Reserve, Central Norway, are presented. The total pup production estimates are within 20% of the estimate based on all 8 censuses, except for those based on single censuses, which deviated more. A b s t r a c t . In the grey seal

K e y w o r d s : maximum likelihood, age group duration, stage structure

Introduction

Proper management of animal populations requires rational methods for monitoring population development. The methods used should give a confident estimate of the population size, or of the fraction that is being monitored, and should be easy to carry out, cheap, and not too time consuming. For seals the only fraction of the population that can be accurately counted is the number of pups born. Seals use the same pupping grounds annually and the number of pups born each year can be accurately counted, and compared with previous counts. However, in the grey seal Halichoerus grypus, the spread of pupping dates exceeds the length of the stay ashore of individual pups, and thus, no single census will include all pups born. To cover the whole pupping season is time consuming and expensive. We therefore developed a method for estimating annual pup production from a limited number of censuses in which the pups are classified according to age. After birth grey seal pups go through five age groups defined by external morphology [ 1]. They typically migrate to sea at 3-4 weeks old. A procedure for estimating grey seal pup production from a single census was given by Radford et al. [2] who assumed deterministic age group durations. We wanted to develop a procedure where the number of censuses may be greater than one, as the breeding season seems to be too long at Froan to obtain sensible estimates from one census. We also wanted to take into account the variation in age group durations and take advantage of today's computing power, which makes maximum likelihood estimation possible.

Address for correspondence: S.-H. Lorentsen, Norwegian Institute for Nature Research, Tungasletta 2, N-7005 Trondheim, Norway. Tel: (+47) 73 580500. Fax: (+47) 73 915433. Emaih [email protected].

48 Ward et al. [3] gave a similar estimation procedure, but without going into detail about how to utilize knowledge of age group durations or applying the estimation procedure to data.

Materials and Methods Data were collected in Froan Nature Reserve (64~ 9~ Central Norway, during the breeding seasons of 1990--1993. Froan is one of the main grey seal pupping grounds in Norway and each year about 300 females gather to give birth to their single pup [4]. The whole area used by breeding grey seals was searched approximately every 5 days. All pups encountered were identified by a Rototag mark in their hind flipper and were aged according to Kovacs and Lavigne [1]. In addition, the number of days since birth was determined for pups that were less than 3 days old. We used a procedure for estimating the total grey seal pup production during a breeding season based on one or more censuses in each of which counts of pups of each defined age group are obtained [5]. The procedure resembles the Bellow and B irley model described by Manly [6]. Estimates were obtained for the peak time for pup production, the spread of the breeding season, the observability (invisibility due to stay at sea or death), and the total number of pups produced. We needed to know the distribution and parameters for the duration of the first four age groups. Parameter estimates were obtained by a method described by Bakke and Lorentsen [5]. It is based on repeated counts of marked pups throughout the breeding season, and thus requires a great deal of field work. It seems reasonable to assume that the duration of stay in each age group varies less from season to season than the four parameters mentioned in the preceding paragraph, so that these estimates may also be used in other years. The age group duration parameters of the fifth age group are not used because it is difficult to obtain estimates for the duration of this age group since pups have migrated to sea and are not easily observable when this age group ends. However, the first observation of a pup in the fifth age group is used when estimating the parameters of the fourth age group. We have assumed that the time of birth To is normally distributed with expected value # and standard deviation or. Thus # may be thought of as the "time of peak pup production" and cr as the "spread of the breeding season". Further, we have assumed that the probability of a pup being observable t days after birth is qt, where q is a parameter that may be less than 1 because the pup is at sea or dead. The motivation for choosing this probability is that the probability of being observed t days after birth would be qt if all live pups are observed and they have a constant probability q of surviving from one day to the next, and it is also reasonable to believe that invisibility due to pups entering the sea increases as they get older. The age group durations Tj, 1 < j < 4 were assumed to be log-normally distributed with parameters as estimated in the Results using the method of Bakke and Lorentsen [5].

49 The uncertainty of the estimates may be assessed by bootstrap analysis [7]. For each census, a new census is then simulated by drawing the same number of pups as in the original census. Each pup is assigned a specific age group with probability equal to the ratio of pups in this age group to the total number of pups in the original census. Then the parameters are estimated again on the basis of this simulated data set. This procedure is repeated a number of times, so that a number of bootstrap estimates of the parameters are obtained. The sample standard deviations of those parameters are estimates of the standard deviations of the original ones.

Results

Pup production The data from the 1993 breeding season were organized into 8 census data sets (Table 1). Information on single pups obtained by marking was thus lost. Several subsets of the censuses were used to estimate pup production parameters (Table 2). When all 8 censuses were used to obtain estimates, a total production of 306 pups was estimated. We simulated birth, aging and observability according to the model (Table 1) to see if deviation from the model in the original data had any effect on the estimates (Table 2).

Age group durations The age group durations were assumed to be log-normally distributed. The expected values and standard deviations were estimated to 4.4 _+ 0.8, 2.4 _+ 1.4, 4.5 _+ 4.0 and 6.5 _ 2.5 days for age groups 1, 2, 3 and 4, respectively, using data from 465 pups.

Table 1. Results of eight censuses made during 45 days (with the last count on day 0) at Froan in 1993 Census no.

Day no.

No. of pups in age group 1

2

1

-45

8 (20)

2 3 4 5 6 7 8

-41 -37 -31 -25 -14 -8 0

16 (22) 39 (24) 48 (41) 21 (37) 9 (12) 4 (3) 3 (1)

Numbers in parentheses were simulated by the model.

14 (2)

10 (10) 15 (13) 19 (11) 31 (13) 14 (11) 4 (2) 2 (1)

3

4 0 (4)

5 (7) 15 (7) 18 (10) 19 (16) 18 (9) 9 (10) 4(2)

0 (1) 0(3) 4 (5) 5(8) 21 (13) 18(17) 22 (9) 8 (5)

50 Table 2. Estimates of expected time of birth (/z), standard deviation of this (a), observability parameter (q) and total pup production (N) for various subsets of counts made at Froan in 1993 (Table 1)

No. of censuses

Censuses included

Parameterestimates /z

a

q

N

All real All simulated 2, 4, 6, 8 1, 4, 7 3, 6

-32.2 -33.0 -30.9 -31.8 _ 0.4 -33.5

10.6 11.3 10.5 9.9 _+0.4 9.9

0.93 0.90 0.91 0.93 _ 0.01 0.92

1

3

-41.3

3.4

1.00

1 1 1

4 5 6

-35.2 -32.3 -22.7

3.4 3.3 3.6

1.00 1.00 1.00

306 312 348 328 _ 39 366 84 105 95 61

8 8 4 3 2

An estimate based on the simulated counts is also included. Bootstrapping was used to estimate the standard deviation for the estimate based on 3 censuses.

Discussion

In 1993, a total of 226 pups were found in the Froan Nature Reserve. However, due to a long period of bad weather during the peak of the pup production, it is highly likely that we missed several pups which probably were washed into the sea and drifted out of the study area. Also taking into consideration the probability of failing to see a pup during its stay ashore, the estimated total production of 306 pups using all 8 censuses seems reasonable. All estimates based on more than one census are within 20% of the estimate based on all 8 censuses. In contrast, the mean for the single census estimates (86 pups) is only one-fourth of the total estimate and no single census estimate is more than onethird of the total estimate. This may be due to both the extended spread of pupping dates at Froan, and/or inaccuracies in our model. Nevertheless, these results clearly indicate that total pup production estimates based on a single census should be avoided. To obtain more accurate estimates, it is necessary to know more about the distribution of births and the observability of pups, although the closeness of the estimates derived from the simulated censuses to those from the original data indicates that the estimates are quite robust. No assessment of the age group duration parameters has been made, and more accurate information on the distribution of the age group durations should be acquired before using the pup production estimates for management purposes. The procedure used for age group parameter estimation is extremely computer intensive. The duration of age groups could, thus, be better determined by following individual pups, e.g. by radiotelemetry. The observability q is influenced by mortality, invisibility when the pups are at sea and when being overlooked by the observer. Clearly it would have been desir-

51 able to have separated the first factor from the last two. Again, more information could be obtained by studying individual pups.

Acknowledgements This study was financed by the Norwegian Research Council through grant no. 4001.713.030. We want to thank Steinar Engen for helpful discussions. We are thankful to T. Anker-Nilssen who made helpful comments on the manuscript and to R. Binns who improved the English language. Thanks are also due to the field personnel G. Dahl, H. Hoel, J.M. Meland, M. Olsen, T. Opdahl, T. Rodahl, P.T. Smiseth, O. Vie, and I.J. Oyen. Part of the data was collected by M. Ekker, B.M. Jenssen and D. Vongraven from SINTEF/UNIMED and the University of Trondheim.

References 1. Kovacs KM, Lavigne DM. Growth of Grey seal Halichoerus grypus neonates: differential maternal investment. Can J Zool 1986;64:1937-1943. 2. Radford PJ, Summers CF, Young KM. A statistical procedure for estimating grey seal pup production from a single census. Mammal Rev 1978;8:35--42. 3. Ward AJ, Thompson D, Hiby AR. Census techniques for grey seal populations. Symp Zool Soc London 1987;58:181-191. 4. Wiig 0, Ekker M, Ekker T, RCv N. Trends in the pup production of grey seals Halichoerus grypus in Froan, Norway, from 1974 to 1987. Holarctic Ecol 1990;13:173-175. 5. Bakke 0, Lorentsen S-H. Estimation of offspring production from a limited number of stage structured censuses. Manuscript. 6. Manly BFJ. Stage-structured Populations: Sampling, Analysis and Simulation. London: Chapman and Hall, 1990. 7. Efron B, Tibshirani RJ. An Introduction to the Bootstrap. New York: Chapman and Hall, 1993.


9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand O. Ulltang,editors

53

Increased accuracy in the estimation of harp seal (Phoca groenlandica) abundance in whelping patches V.I. Chernook, V.A. Potelov and N.V. Kuznetsov Knipovich Polar Research Institute of Marine Fishery and Oceanography (PINRO), Murmansk, Russian Federation A b s t r a c t . The course and results of experimental aerial surveys of harp seal whelping patches con-

ducted in March 1993-1994 in the White Sea u s i n g an MI-8 helicopter are presented. Recommendations concerning improvements of the on-board equipment, the method of aerial survey and possible automation of the aerial counting work are given. Key words: seal, white coat pup, air survey, infrared, transect

Introduction

Aerial surveys of harp seal whelping patches are part of the complex research which also includes the distribution and migration analysis (geographical aspects) and biological aspects, and which leads to a rational use of the stocks of this animal. Aerial surveys have been used to estimate the number of harp seals in the White Sea since the middle 1920s. Since the early 1960s scientists have surveyed only suckling females, while estimations of the number of pups themselves, the main indication of the seals' stock condition, have not been made. While suckling their pups the females need to feed themselves and must occasionally leave their pups. According to observations made by Popov [1] in early March (when the patches are formed and the suckling begins) 70-80% of females are in the water in the early morning hours on bright sunny days. During the second part of the day no more than 30--40% stay in the water. After 1 or 2 weeks of suckling, the time the females spend in the water increases greatly. This may cause aerial surveys when only females are counted to err by 50% or more. The number of females on the ice also decreases during snow-storms, or when the ice breaks up [1], but increases during ice compressions. Taking into account that the number of suckling females varies greatly both during a 24-h daily cycle and during the whelping period, the harp seal researchers in the White Sea have introduced a correction factor [2]. Thus it is assumed that 80% of females are on the ice and 20% are in the water during aerial surveys conducted in early March. The validity of the correction factor has not yet been confirmed. Estimates based on aerial surveys of harp seal whelping patches during the first days of suckling have some advantages compared to estimates made at other periods of their life (e.g. when the suckling period is over or on the moulting grounds). Then

Address for correspondence: V.I. Chernook, Knipovich st. 6, Murmansk 183763, Russian Federation.

54 the density of the seals' patch is at its highest; furthermore, the age and sex structures of the patch are stable for a long time (females + pups + males). After approximately 3 weeks the whelping patches break up leaving only the pups on the ice. Because of the ice drift, however, the density of pups in the patch greatly decreases, thus resulting in increases in the cost of flights designed to survey them. For that reason we set out to improve the method of aerial surveys of seals in the suckling period. The main point of the work was to improve the accuracy in the methods used to estimate harp seal pup numbers. The means was a survey where the seals were synchronously recorded visually and in the IR part of the spectrum. The stages of the work were: data collection; determination of optimal survey conditions by considering the technical specifications of the on-board equipment; improving the methods for a synchronous survey using the visual and IR part of the spectrum; simultaneous processing of the shots and images obtained from the two parts of spectrum.

Methods

and

Equipment

In March 1993 and 1994 experimental aerial surveys of harp seal whelping patches were carded out using an MI-8 helicopter in the entrance of the White Sea. Technical facilities used in 1993 included (see Table 1): "Vulkan" thermovision set; aerial photo camera AFA-TES- 10; video set JVC GF-500. It is necessary to emphasize that, unlike photo and video cameras, the thermovision set surveys the space by linear scanning with a narrow momentary sighting angle perpendicular to the flight direction. Therefore the given combined survey may be considered as a "quasisynchronous" one.

Table 1.

Specifications of the equipment used in the 1993 survey

Sensor

Thermovisiona Video Photo

Sight angle

Track width

Ground resolution (m)b

(~

(km)

Across

Along

80 40 84

0.72 1.34

1.29

0.40 0.33 0.15

0.4 0.4 0.2

aThermovision-scanner with scanning frequency of 180 lines/s, sensitivity 0.3~ bGround resolutions are given for a flight altitude of 200 m.

55

Table 2. Specifications of the equipment used in the 1994 survey. Sensor

Thermovisiona Video Photo

Sight angle

Track width

Ground resolution (m)b

(o)

(km)

Across

Along

20 40 44

0.35 0.72 0.80

0.26 0.33 0.08

0.3 0.4 0.1

aThermovision sensitivity is 0.2~ bGround resolutions are given for a flight altitude of 200 m.

Technical facilities used in 1994 included (see Table 2): "Insight 80" series thermal imager; video camera JVC TK-880; videocassette recorders JVC; aerial photo camera PA-39; lap-top computer Toshiba; GPS positioning system Raytheon R-900. To determine the number of whitecoats reliably, the survey should be conducted when the majority of the females have whelped. The experiments were conducted when the number of whelped females exceeded 95%. The experiments both in 1993 and in 1994 were carried out using the following design: 1st flight: searching for the whelping patch and choosing the optimum flight parameters (height and speed; synchronous surveys were carried out successively from 400, 200 and 100 m height, speeds ranging between 100 and 180 km/h). 2ndflight: surveying the seal patch along a net of transects at a predetermined height and speed. In 1993 the flights were carried out on 5 and 7 March, and in 1994 on 8 and 9 March. With the equipment used in 1993, the optimum height and speed was 100 m and 180 km/h. The corresponding values in 1994 were 200 m and 130 km/h. The essence of the method is synchronous surveying of the harp seal whelping patches in the visual and IR parts of the spectrum followed by joint processing of the images, when the number of whitecoat pups lying on the ice is determined. Interpretation of the visual images was unproblematic because this information is the same as obtained through the human eye. Thermal images were formed by three main objects: snow-and-ice cover, water and seals. Snow-and-ice cover and water were hardly distinguishable on IR images and served as background for the surveyed seals. These objects had a temperature close to that of the air which at the beginning of March is usually - 3 to -5~ It is known that the temperature of adult seals and whitecoat pups substantially exceeds 0~ Such thermal contrast is quite enough to image them distinctly by means of thermovision techniques (Fig. 1). -

-

-

-

56

a).

b).

Fig. 1. Computer image processing of an IR frame (there are 3 harp seals on the frame) (a) incoming IR frame; (b) IR frame (a) after image processing. 1, Water; 2, ice; 3, harp seals.

-

-

-

-

The succession of analyses of the photographs obtained was as follows: general analysis of images, sorting of the films according to tracks and heights; selection of image areas and ice-floes surveyed in both the visual and IR parts of the spectrum; identification and calculation of seals on the photos in the visual part of the spectrum; joint analysis and calculation of seals in both the visual and IR parts of the spectrum (Fig. 2);

Fig. 2. (1) Video image; (2) processed IR image; bright (warm) spots on IR image are seals; clark (cold) background is ice and water.

57 studies of spatial distribution and determination of errors in seal calculations; In the 1993 experiment, the number of seals was determined by counting the animals on the same ice floes on visual and IR images. The seal count was made in the following order: counts of adult seals on the photos; counts of total number of seals on the IR images; counts of adult seals and whitecoat pups while analysing both photo and IR images. To ensure the reliability of the survey results, video films were also used. The 1994 results have not yet been analysed, so that the following only presents the 1993 results. -

Results

In Table 3 the results of aerial survey of seals on three tracks from a height of 100 m are presented. On these tracks some ice-floes with clearly visible edges were selected and on these ice-floes seals were counted on photos and IR images. A total 272 females and 60 males (332 in total) were counted on the three tracks. According to presently used methods of estimation of seal numbers in aerial surveys, these data correspond to 272 whitecoat pups, 272 females and 60 males (totally 604 seals). Use of a correction factor of 20% [2] yields a corrected number of seals of 340 females, 60 males and 340 pups; in total 740 animals. On IR images 522 seals were counted. Thus the results obtained by one method alone may include large errors. Combination of the two methods, however, may help to increase the accuracy. As can be seen from Table 4, the number of adults remained the same, 332, but the distribution among males and females was altered when the photos and IR images were analysed in combination. In particular, two new groups of seals appeared: single ones (without pups), which were classified as "females with pups" according to Table 3, and dead ones (no spots on IR image). The number of females with pups decreased from 272 to 188 (by 31%), the number of males increased from 60 to 72 (by 20%). In the combined analysis of the images, 69 pups lying on ice-floes without Number of harp seals detected during photo and IR surveys on whelping patches in the White Sea, 6.03.93, 100 m height (corrected numbers include a surplus of 20% [2]) Table 3.

No. of track 10

Visual survey Females 96

11

109

12 Total Corrected

67 272 340

IR survey total no. of seals

Males

Total

9 27 24 60 60

105

194

136

194

91 332 (604) 400 (740)

134 522

58 Table 4. Results of harp seal numbers estimated on whelping patches in the White Sea obtained by combined analysis of the photo and IR images obtained 6.03.93, 100 m height No. of track

10 11 12 Total

Adults

Pups

Total number

Females with pups

Single seals

Males

Dead or ice-floes

With mothers

Single

88 62 38 188

6 47 17 70

9 27 36 72

2 2

88 62 38 188

28 28 13 69

221 226 142 589

mothers were detected. The total number of pups established by the visual and IR method of survey on the three tracks equalled 257 which was 6.5% less than in Table 3 (or 32% less if the correction factor is taken into consideration). The data of Table 4 also let one determine the number of females in the water, which is equal to the number of single pups. On track 10 they amount to 24.1%, on track 11, 31.1% and on track 12, 25.5%.

Discussion When conducting such surveys, the spatial resolution of the survey equipment is of utmost importance. The ground resolution depends on the angle of view, the equipment's angular resolution and the altitude of the flight. Due to some technical problems the resolution of the thermovision equipment obtained during the present survey was less than that of the video camera and included 200-250 lines in the frame. The technical characteristics of the infrared imager are determined when conducting the survey (video + IR). An infrared imager temperature sensitivity of 0.1 o, a spatial resolution of 300-400 elements in the line, and an angle of view of 15-20 ~ are desirable. The following errors appear in processing of the infrared images: a seal scared away from the ice in the patch can be seen as a luminous spot (heated ice), which may be erroneously interpreted as a second seal; a seal only observed in the water has a very weak infrared contrast and may be not detected in the infrared image, but may be observed in the video image; pup and female lying together may be seen as a large spot; pups lying under standing ice-floes are not observed in the infrared still pictures. The main shortcoming of the infrared equipment (infrared imager "Vulkan") used in 1993 was that the film recording the infrared information was not suitable for operational analysis and correction of the survey results in situ. Furthermore, its sensitivity was insufficient. The infrared equipment used in 1994 had better sensitivity, with an output signal of television standard, making it possible to observe the results

59 of the survey during the flight at once. Additionally, the use of a satellite navigational system made it possible to have a precise spatial attachment of all the data from the survey and to show the route of the flight on the monitor screen. The 1994 results will be published at a later stage. Synchronous video and IR surveys of harp seals whelping patches allow the determination of the number of seals on the ice. A combined image analysis in the visible and IR parts of spectrum increases the accuracy of the number of adult seals determined. Using this method, the pupless females (sterile and non-pregnant), separate males, females moving inside the patch, dead animals (or spots resembling animals) and pups separated from their mothers may be detected on the photos.

Acknowledgements The authors thank the on-board operator V.Yu. Bogomolov for his participation in the experimental survey of the seals. We also thank A.B. Gvozd for his aid in data processing and forming the present report.

References 1. Popov LA. On biological substantiation of aerial photography of harp seals' whelping patches in the White Sea (in Russian). In: IV All-Union Conference on Studying of Sea Animals, Kaliningrad, Moscow, 1969;102-105. 2. Timoshenko Yu K. Observations on harp seal females' behaviour on whelping patches. In: Coil "Sea mammals". Moscow: USSR Academy of Science, 1978;323-324 (in Russian).


9 1995 ElsevierScience B.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. Wallce and O. Ulltang,editors

61

Haul-out behaviour of the Norwegian harbour seal during summer Randi R o e n and Arne BjCrge Norwegian Institute for Nature Research, Blindern, Oslo, Norway Abstract. Background: the haul-out behaviour of the Norwegian harbour seals had previously not been studied thoroughly. Methods: the intent was to examine the haul-out behaviour in relation to three factors: (1) the diel light cycle, (2) the tidal cycle and (3) the interaction between these two cycles. Observations were therefore made on days with low tide around noon, and later repeated on days with high tide around noon. The number of hauled-out seals was counted every hour, day and night. Three places were selected on the basis of their differing characteristics in diel light and tidal variation during summer. Results: the diel light cycle, the tidal cycle, and the interaction between the two cycles, all showed a significant relation with the haul-out behaviour both at Froan and in Kongsfjord. At Hvaler, the haulout pattern was very inconsistent, with large fluctuations from day to day. Conclusions: the higher numbers of hauled-out seals in the daytime, especially in Froan but to a certain degree also in Kongsfjord, may indicate that most seals feed at night. Similarly, the higher numbers hauled out at low tide, especially in Kongsfjord but also in Froan, may indicate that most seals feed around high tides. At Hvaler, the seals may be less at ease by the great deal of commercial and recreational boat traffic. K e y w o r d s : seals, diurnal activity, time series, haul-out pattern

Introduction

Tide is often thought to be a highly important factor influencing the haul-out behaviour of harbour seals [1-3]. Others, however, have found time of day to be the most significant factor [4,5]. Prior to this study, only occasional observations of the haul-out behaviour of harbour seals have been made in Norway. These observations suggest that harbour seals in the Oslofjord haul out in the very early morning, while the seals in West- and North-Norway haul out at low tide [6].

Materials and Methods Three harbour seal colonies at three different locations were selected for the study. At all three locations, haul-out sites are accessible throughout the tidal cycle. Hvaler (59~ 10~ has the smallest tidal amplitude, on average 0.05). To insure an assumption of GAMs, the dependent variable, harbor porpoise density, was log transformed and the gaussian link was used because log(density) was more closely normally distributed than were the Gamma or Poisson normalized (Anscombe) residuals of density. Fish density indices were transformed to log(fish density index + 1) for the same reason. Adding one to the fish density index allowed the areas of zero fish density (which were infrequent) to be used. Using the F-test and AIC for each separate year, the following GAM model was chosen which used the lowest possible order smoothed functions while still capturing the shape of the relationship: log(density) = s(temperature,2) + s(depth, 2) + s(log(fish + 1),2) +/o(latitude,longitude, 1/3) where s(factor, df) indicates a smoothed function of the factor with df degrees of freedom, and/o(factorl,factor2,span) indicates a lowest smoothed fit of factorl and factor2, where the span is the fraction of data used to compute the local regression. Separate smoothed functions of latitude and longitude were investigated but did not fit as well as the bivariate relationship, indicating an interaction between latitude and longitude. The pseudo-R 2 for 1991 and 1992 were 0.25 and 0.32, respectively.

72

Fig. 1. Plots of CAM smooth fits of the environmental factors: temperature (temp), depth and log transformed fish density index (logcalcallnum) for 1991 (top line) and 1992 (bottom line). Solid line is best fit, dashed lines are 2 standard error confidence limits, hash lines on x-axis are the observed values of that factor.

73 GAM models suggested that high densities of harbor porpoises were in waters that were between 10 ~ and 13.5~ especially between 10 ~ and 12 ~ in 1991 (Fig. 1). Spatial contour plots of temperature overlaid with harbor porpoise density (Fig. 2) illustrated that 1992 harbor porpoise density was higher than that in 1991, as was the percentage of the planar area that was covered with water temperatures of 10-13.5 ~ (41% in 1991 versus 59% in 1992). GAMs demonstrated that harbor porpoises were most often found in waters that were between 30 and 70 fathoms deep. Also, more animals were found in shallower waters (down to 10 fathoms) and in deeper waters (out to 100 fathoms) in 1992 than in 1991 (Fig. 1). The relationship between harbor porpoise and fish density is less clear. The GAM models indicated that, for both years, high densities of harbor porpoises were associated with intermediate levels of the fish density indices (1.5-11.0 fish/min; Fig. 1), even though harbor porpoises were distributed in waters with a larger range of fish density indices in 1992 than in 1991 (0-53 fish/min in 1991 versus 0-244 in 1992). In fact, legs with very high fish density indices were in hot waters, generally in temperatures higher than the "preferred" surface temperatures described above. The planar area of the study region that was covered with waters containing the "preferred" fish index values was greater in 1992 than in 1991 (10% in 1991 versus 16% in 1992; Fig. 2). The bivariate relationship of latitude and longitude in the GAM indicated that highest densities during both years were found around Grand Manan Island and offshore of Penobscot and Frenchmans Bays of central Maine (Fig. 1). The Maine area had lower harbor porpoise densities in 1991 than in 1992.

Discussion

The GAM models of harbor porpoise density did predict small scale patterns; however, the magnitude was underestimated. For example, when using environmental data from 1991 in the 1992 model, the density in 1991 was correctly predicted to be lower than that predicted for 1992. However, the predicted 1991 densities were also lower than the actual 1991 densities. This may be due to several possibilities, such as: the smoothing procedure was too smooth; zero density legs should have been incorporated; data should have been collected on a finer spatial scale; there were additional factors influencing density and/or sightability of harbor porpoises that were not incorporated; or density in the summer depends not only on concurrent small scale environmental factors, but also on events which occurred in regions where harbor porpoises spent the winter. Some of these factors can be investigated in the future. Even though the GAM models provided very smooth surfaces, these data indicated that on a large spatial scale, harbor porpoise aggregations were in the same general vicinity during both years. On a smaller scale, the exact location and magnitude of those aggregations were correlated to small scale distributions of water tem-

74 Fig. 2. Contour plots of spatial distribution of temperature and log transformed fish density index for 1991 and 1992. Dark areas of contour plots are warmer ternperatures. Black dots and + marks are the locations of the beginning of legs of effort from the harbor porpoise line transect survey. + indicates no observed harbor porpoises. Size of black dot indicates magnitude of harbor porpoise density; the larger the dot size the greater the density. Total black areas are either land or water that was not surveyed.

75

perature and fish density indices. This can be seen by a close inspection of the two highest density areas. The Maine aggregation extended farther offshore and south in 1992 than in 1991. However, in both years, the aggregation coincided with a finger of water that was 13~ and contained intermediate fish density indices of 2 - 1 0 or more fish per minute. The same trends can be seen in the Grand Manan Island area, although not as clearly. South of the island (along 44.3~ from 67.5~ to 66.5~ temperature, harbor porpoise and fish densities were higher in 1991 than in 1992. Inter-annual changes in the surface temperature and fish density patterns may be a reason for the inter-annual changes in harbor porpoise abundance. Although it is difficult to demonstrate a cause-and-effect relationship, it can be noted that the abundance of the region for 1992 was estimated to be 1.8 times higher than that for 1991 [2], which coincides with a similar magnitude of increase in the planar area covered by "preferred" water temperatures (1.5 x ) and fish density indices (1.6 x ).

References 1. Palka D. Abundance estimate of the Gulf of Maine harbor porpoise. Rep Int Whal Commn 1994;(Special Issue 16) (in press). 2. Smith T, Palka D, Bisack K. Biological significant of bycatch of harbor porpoise in the Gulf of Maine demersal groundfish fishery. NOAA/NMFS/NEFSC Ref Doc 1993;23:1-15. 3. Gaskin DE, Watson AP. The harbor porpoise, Phocoena phocoena, in Fish Harbour, New Brunswick, Canada: occupance, distribution and movements. Fish Bull 1985;83:427-442. 4. Watts P, Gaskin DE. Habitat index analysis of the harbor porpoise (Phocoena phocoena) in the southern coastal Bay of Fundy, Canada. J Mammal 1985;66:733-744. 5. Keckler D. SURFER for Windows User Guide. Golden, CO, USA: Golden Software Inc, 1994. 6. Buckland S, Anderson DR, Burnham KP, Laake JL. Distance Sampling. Estimating Abundance of Biological Populations. New York: Chapman and Hall, 1993. 7. Hastie TJ, Tibshirani RJ. Generalized Additive Models. New York: Chapman and Hall, 1991. 8. SPLUS. Seattle, WA, USA: Statistical Sciences, 1991. 9. Swartzman G, Huang C. Spatial analysis of Bering Sea groundfish survey data using generalized additive models. Can J Fish Aquat Sci 1992;49:1366--1378.


9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand O. Ulltang, editors

77

Aerial surveys of harp and hooded seal pups in the Greenland Sea pack-ice Torger Oritsland and Nils Oien Institute of Marine Research, Nordnes, Bergen, Norway A b s t r a c t . After experimental surveys by ship-borne helicopter and fixed-wing aircraft in 1990, harp

seal breeding patches were surveyed in the Greenland Sea pack-ice (the West Ice) in the spring of 1991. Combined estimates based on data from visual helicopter transects and aerial photographic transect surveys indicate a minimum total production in excess of 55,000 harp seal pups in the West Ice in 1991. Extreme weather conditions and an exceptional westerly distribution of the pack-ice impeded attempts to survey West Ice hooded seal pups in 1994. Classified counts suggest late births of harp seal pups in 1990 when compared to 1991. Less precise data suggest comparable timing of peak pupping but a wider temporal distribution of hooded seal births in 1994 than in 1991. Published results from analyses of still-camera and video material confirm that image-analysis may be used to obtain useful information on environmental conditions and the spatial distribution and sizes of seals in pack-ice breeding patches. Key words: visual, video and photographic transects, pup development, production

Introduction Harp and hooded seals, Phoca groenlandica and Cystophora cristata, were selected as key species for studies under the Norwegian Marine Mammal Research Programme 1989-1993, and surveys of pups in the West Ice, the breeding grounds in the Greenland Sea pack-ice, were among the chosen priority projects planned for the programme. Experience in the operation of a ship-based helicopter was obtained on an expedition to tag harp seal pups in the West Ice during the breeding season of 1989 [1,2]. The project was continued on an expedition to pursue the tagging programme and to test alternative techniques and select methods for aerial transect surveys of pups in the West Ice breeding lairs in 1990 [3]. Harp seal pups were then surveyed and tagged during the breeding season in 1991 [4,5]. Further efforts were delayed by a temporary scarcity of funds, and a postponed expedition dedicated to surveys of hooded seal pups was organized by the Institute of Marine Research in prolongation of the Marine Mammal Research Programme in 1994. Activities and results related to the surveys in 1990, 1991 and 1994 are summarized in this report. Results from the tagging programme [6] are reported separately to this symposium [7].

Logistics The mutually trustful relationship established between the Institute of Marine ReAddress for correspondence: T. Dritsland, Institute of Marine Research, P.O. Box 1870, Nordnes, N5024 Bergen, Norway.

78 search and the companies and crews during the tagging expedition in 1989, formed the basis for further cooperation with Rieber Shipping A/S, Bergen, and Helikopterteneste a.s., Kinsarvik, for the use of the same ship and helicopter on the expeditions in 1990, 1991 and 1994. The "Polarsyssel" is a 50-m long vessel of 499 gross tonnes with reinforced hull for navigation in ice and fitted with a 2,500 hp main engine. She is fully classified as an icebreaker-sealer (DnV + 1AJ). Large flush hatches over the aft shelterdeck are particularly useful because they permit sheltered stowage of the helicopter during crossings in rough seas to and from the ice, as well as sheltered instalment of a container lab. The Ecureuil AS 350 B 1 helicopter proved to be well suited for our purpose, both with regard to ease of handling and stowage on board the ship and because of capacity and range. Both the ship and the helicopter were fitted with satellite navigation systems. The global positioning system (GPS) in the helicopter proved to be useful both for positioning of observations during search flights and for navigation along preplanned transect lines during visual and video surveys. A fixed-wing twin-engine Partenavia P68TC Observer aircraft, operated by Fotonor A.S., Oslo, was based on Jan Mayen Island to be used for search flights and photographic transect surveys of seal patches in all the 3 years covered by this report. For navigation, the Partenavia was fitted with Loran C in 1990 and 1991 and with GPS in 1994, and carded a mounted Wild RC 20 camera with a 15 cm lens for vertical photography. In 1994 an additional fixed-wing twin-engined aircraft operated by the same company, a Piper PA31 fitted with GPS and a vertically mounted Zeiss LMK camera with a 15 cm lens, was based at Longyearbyen airport, Svalbard, to improve the efficiency of search and photographic surveys during the expected brief spells of acceptable weather conditions.

Field work 1990

The primary purpose of the expedition in 1990 was to test alternative techniques for selection of methods for transect surveys of harp and hooded seal pups in the West Ice breeding lairs. A secondary objective was to continue the tagging program for harp seal pups which was started in 1989. "Polarsyssel", with the helicopter on board, spent 20 days (20 March to 9 April) in the ice in 1990. Continued northerly winds prior to and during the breeding season dispersed the pack-ice and also caused a rapid southerly drift of the floes. This drift continued throughout the season. Except for three early days with strong winds and poor visibility, the weather conditions were tolerable or good during the period in the ice. The Partenavia aircraft was available for search and survey flights for 13 days, but

79 the operations were restricted by local weather conditions at Jan Mayen, and flying was limited to a total of 29.5 h. However, the pack-ice between 71~ and 75~ was searched on three flights before the end of March, and another search with intervallic photography from 1,000 ft (305 m) was carded out from 70~ to 75~ during the first few days of April (Fig. 1). Short intervallic and overlapping photo-transects from 500 ft (152 m) and 1,000 ft were also obtained over two small harp seal breeding lairs. Additional information on ice conditions was supplied from two search flights by a Soviet aircraft which covered the area between 72~ and 74~ on 21 March and 23 March. Breeding seals were not discovered during the Soviet flights. "Polarsyssel", in the meantime, worked its way from about 71~ up to 75~ using available time on helicopter search and tagging of harp seal pups. Scattered breeding harp seals were recorded throughout the area from 70~ to 74~ with two small patches of breeding harps which were located around 73~ 12~ in late March. Widely dispersed families of hoods were found all over the open pack-ice from 70~ to 75~ with only one minor concentration of some 300 families at 73~ 13~ on 25 March. In addition to search flights, the helicopter was also used to deploy personnel for tagging and for experimental transect surveys. Seven series of video transects were carried out over the two patches of harps at 100 ft (30 m) altitude intervals between 200 ft and 600 ft (61 and 183 m). A special Dage MTI camera, selected for maximum resolution, proved to be hopelessly inadequate for the purpose, but good images were obtained by an off-the-shelf Sony 8 mm handycam, provisionally mounted for vertical recording through a front-bottom window in the helicopter cockpit [8]. In 1990 the helicopter was also used for visual transects with two observers at 200 ft (61 m) and a series of age (developmental stage) determinations of harp seal pups from altitudes around 50 ft (15 m) in breeding patches between 31 March and 8 April. 1991

In 1991 the Partenavia aircraft was available on Jan Mayen Island for 26 days from 15 March. Operations were impeded by unsuitable or unstable weather for 16 days, but a total of seven search flights and three photosurveys were completed over the pack-ice between 71~ 18~ and 74~ 03~ "Polarsyssel", with the helicopter onboard, was engaged in search, taggings and surveys in the ice between 71 ~ 18'W and 75~ 03'W for 28 days from 16 March. Again the weather was highly variable and less than ideal with winds at gale force or stronger for 11 days, and moderate or poor visibility for 16 days through the period "Polarsyssel" was working in the pack-ice (Fig. 2). Young ice covered extensive areas between strips and patches of consolidated older ice. Currents and changing winds led to frequent rearrangements of the ice with an unexpected easterly drift in late March and a resultant fairly fast drift towards SSW in early April. Scattered harp seals with pups were recorded throughout the area between 71 ~ and 75~ a distance of about 350 nautical miles (650 km), with the largest concen-

80

~~176 9

eeeoeee0

0328

7" *

/

73"

/

/

r

/ 7 ~

/

/

/

/

/

i

i

/

i !

i....

~ ~

'......

....,,. ...9

i..0325

/

i

i

i

i

/

L~

/

/

i

J

o,b

03.23

/

,,/

!

P~;oee,o

/ i i! I ,

72 o

9

oe',t ~176 03.25..;;.

d~ q'o ,.~,~

ce~

.

. ,7~1~a28

I

I I

I

/

!

1

04.03

9 .%.

04.03i~/

,=

%

/

"..

/

9

/

/

/I/I/ 71"

. ///i/---.-'J i "

1 3 4

02

70 ' 4

i

Fig.1.

--

|

I11111

2 """ ---

oo.o

1

"--

6 7

eooo

8

---

10

"-i

Ice edges and the distribution of breeding harp and hooded seals recorded by spotting aircraft in the West Ice 23 March to 3 April, 1990: (1) harp seal breeding patch; (2) hooded seal breeding patch; (3) scattered breeding hooded seals; (4) ice edge; (5-10) search flights 23, 25, 28, 31 March, 2 and 3 April.

81

76"

'•[

.......

'I

,,.I,,

75"

/~

.ii.l~,

GI l~% N.

/

73 9

-I

.

f

.

Ol.lllt .

G|V'

. ~ ~..~' / ~ , ,

//,"//.

;,,,/ol,,lo

/o,,",'~_ ,//,/~ "is.a,

,,-.:.

72' ,--.., ._..,

3 4

71"

I/

/j,,,,

-=-

,,,,,6w

t

I

t

I0"

5"

0"

5

Fig. 2. Ice edges and the distribution of breeding harp and hooded seals, recorded by ship-borne helicopter and spotting aircraft in the West Ice 16 March to 12 April, 1991" (1) hooded seal breeding patch; (2) harp seal breeding patch; (3) scattered breeding hooded seals; (4) drift; (5) ice edge.

Table 1. Estimates of harp seal pup production in four separate breeding patches in the Greenland Sea pack-ice in 1991, based on visual transect counts from helicopter and photographic transect surveys by fixed-wing aircraft, uncorrected for bias

Patch No. 01 02 03 04 Total over all patches 95% confidence interval

Visual 7,100 (0.075) 3,800 52,500 (0.159)

Photographic 2,021 (0.128) 5,905 31,917 (0.271)

Combined 2,991 (0.078) 3,800 5,905 (0.195) 42,574 (0.141) 55,270 (0.141) 44,500-68,500

The estimate for patch No 2 was provided from visual shipboard transects by the Soviet research vessel "Varzuga" without information on uncertainties. Available visual and photographic data were combined for each patch by weighting of variances. Coefficients of variation are given in parentheses.

82 trations towards the north. Four separate harp seal breeding patches were found and three of them were covered by aerial still photo transects and two of these three patches also by visual and video helicopter transects (Table 1). The fourth and smallest patch (No. 02) was covered only by visual shipboard transects carried out by the Soviet research vessel "Varzuga". Age or developmental stages were determined for a total of 4,711 harp seal pups in two of the breeding patches. An additional 20 dead pups (0.4% of the total sum) were recorded during these low-altitude helicopter flights. 1994

As mentioned under logistics, the expedition in 1994 included an extra fixed-wing aircraft stationed at Longyearbyen airport, Svalbard. "Polarsyssel", with the helicopter on board, had 19 days in the ice from 16 March, and both aircraft were available for 18 days from the same date. However, boisterous weather was even more restrictive this year than during any of the previous expeditions. Northerly winds prevailed with strengths from strong breeze (Beaufort 6) to violent storm (B.11) with occasional hour-long gusts of hurricane force (B.12). Visibility was poor to moderate for 16 days through the time spent in the ice. Also the ice conditions were extreme with the ice edge far to the west of the expected position and with a rapid drift of the pack-ice towards SSW-SW (Fig. 3). An Argos satellite buoy deployed on a large floe on 26 March drifted at an average speed of 1.3 knots until it was resighted 116 nautical miles further to the SSW 4 days later, and continued roughly in the same direction along the East Greenland coast until the last signals were received in early June. Because of the ice conditions, the expedition in 1994 operated under special permit from Greenlandic authorities. "Polarsysser' operated south of 72~ towards 68~ throughout the period in the ice and was unable to proceed further north. The helicopter was used on 12 search flights between 68~ and 73~ and two attempts at visual transect surveys of hooded seals with pups on 26 March and 30 March. An attempt to use the helicopter for tagging of harp seal pups at 69~ 20~ on 3 April was discontinued because of icing and poor visibility. The difficult weather conditions and the westerly location of the pack-ice also impeded the operation of the two fixed-wing aircraft. They needed about 1 and up to 3 h to reach the ice from their bases on Jan Mayen and at Longyearbyen, respectively, and therefore required exceptionally stable weather and reliable forecasts. However, three search flights from Jan Mayen covered the ice from 68~ 23~ to 74~ 10~ and the area from 74~ 15~ to 76~ 09~ was searched on four flights from Longyearbyen. Only small numbers of scattered breeding harps and hoods were found north of 74~ A total of 395 photographs were taken at intervals from 800 ft (245 m) during these seven search flights. The Partenavia from Jan Mayen also obtained 180 images from the same altitude in four transects over scattered weaned hooded seal pups around 68~ 23~ on 26 March. On the same date scattered female hoods

83 76 ~

_

74 o

72 ~

'~176 Lf -

6B o -/--

. . . . . .

03,26~ /

Y

=

3

ooo

/-,,

::x

5

--'03.25-03.26

2 "0 o

-

'

1 0''

'

Fig. 3. Ice edges and the distribution of breeding harp and hooded seals recorded by spotting aircraft

and ship-borne helicopter in the West Ice 20 March to 4 April, 1994: (1) hooded seal breeding patch; (2) harp seal breeding patch; (3) scattered weaned hooded seal pups; (4) scattered weaned harp seal pups; (5) ice edge. with relatively young pups were recorded around 74~ 14~ Another hooded seal breeding patch with newly born pups unevenly distributed over an area of about 5 x 10 nautical miles (170 km 2) was discovered the next day at 71~ 18~ A few pictures were obtained but the available flight-time did not permit coverage by transects.

Results and Discussion

Still-camera and video images obtained during the exploratory survey in 1990 have been used in image analyses to assess ice-conditions and measure sizes of seals [8].

84

100~

80

1 g8

,oo, 9

_

,,ol

I

,

.

--

1267

~;, I

,5 2 6

,,o

,

I

I

_

~

@

60

Z uJ

2

3

40-

r

20,--.

1 ~-l-L"~

1

03.25 397 578 { ----~' ......

100

=12

'~.

80

' ~.... l

0 4.0

0 4.0 9

:373 :

497 I -

181

44,5 ~

9

431

~

~--

5

./'"--"

l |

i 1 !

---

z

I i~

l

.

u

~ 9

252 81g ~ 426..~ _ ___.~ -.-

---

60

1-~

0 ~.30

1

"1

1

!

,

~J

:3

o.

! |

2,

e ~"~" e 1 "

0 -

9

|e

~ , ...

,

~

i

03.25

~1

i~11 I1' I I ~

[ "-- ' ~'a"m

03.;30

]~

'

["

'

0 4.0 9

Jl

,

.... '

1

'

-- '

;

04,0g

DATE

Fig, 4. The relative distribution of harp seal pups by developmental stage and date (] = newborn; 5 = moulted beater), recorded by classified counts in the West Ice 28 March to 8 April, 1990 (upper graph) and 22 March to 12 April, 1991 (lower graph). Sample sizes for each count are given along the upper horizontal axis of each graph. The vertical stippled line in the lower graph divides between counts in two separate breeding patches.

85 108 100

~-

80

"--

60

--

140

18 I

148

I

,l

.

.~ .

17 .

x

15 ,

)-,

,

~

- -

4

z

t4J (,J t~ a.,

,

..-

40

,==

20

---

0 03.25

03.30

04,04

04.0~

278 79

16~7

148

306

6O

U ~ e,.

o

0 40

1-3

2O

0

03.2

0

03.25

00,30

DATE

Fig. 5. The relative distribution of hooded seal pups by developmental stage and date (0 = unborn/ parturient females; 5 = weaned), recorded by classified counts in the West Ice 22 March to 7 April, 1991 (upper graph) and 16 March to 31 March, 1994 (lower graph). Sample sizes for each count are given along the upper horizontal axis of each graph.

86 Results indicate that seals prefer medium-sized floes, ranging from 13 to 38 rn in diameter, compared to the size range 0.2--49 rn for all ice floes. Floe shape was also characteristic of occupied floes: seals selected more rounded floes rather than elongated floes. The presence of seals was also related to the ice conditions over a larger area. In the transects that were analysed seals were always found in areas with greater than 60% ice coverage. Data from videos resulted in values of 1.32-2.40 m for lengths of adult and subadult seals, and 1.02 m for pup length. No attempt was made to distinguish between species, but these measurements fall within the combined range of lengths for harp and hooded seals obtained from previous studies [8]. Abundance estimates of harp seal pups calculated from the basic transect data obtained in 1991 are listed in Table 1, with a provisional best estimate of total production exceeding 55,000 pups. For each patch the available estimates have been combined by weighting of variances. The variance associated with the Soviet count of patch No. 02 is not known, and has therefore not been accounted for in the confidence interval for the combined total. Both the visual and the photographic survey estimates have an inherent negative bias caused by the fact that no correction has been made for scattered pups between the patches, nor for the temporal distribution of births. A specific problem with the photographic surveys is caused by errors in the interpretation and reading of the photographic material. Experimental reading of subsamples by alternative techniques indicates that the adopted procedure may underestimate the true counts of pups by about 9% [5]. The results from low altitude developmental stage determinations of harp seal pups in 1991 [4] are compared to corresponding data collected in 1990 in Fig. 4. These graphs suggest that breeding may have occurred earlier by 5-7 days in 1991 than in 1990. To accommodate the sealers' lore, it may be mentioned that the moon occurred in its first quarter on 23 March in 1991 and 10 days later, on 2 April, in 1990. Developmental stages of hooded seal pups [4] were not determined on dedicated flights in 1994, but classified counts recorded on helicopter search flights suggest an exceptionally wide temporal spacing of births; centered roughly around the same dates as in 1991 when stage determinations were recorded on three dedicated low altitude helicopter flights and from the ship in passing through the ice (Fig. 5). Because they covered only limited areas with unevenly distributed hooded seals, representing a minor but unknown proportion of the total, the visual helicopter transects of hooded seal pups in 1994 were methodological exercises rather than surveys which could be applied in assessments. Data from the expedition in 1994 are indicative of possible distributions of breeding hooded seals under extreme conditions in the Greenland Sea. They are, however, totally inadequate for any assessment of production.

Acknowledgements We acknowledge with gratitude the assistance of the crews on the "Polarsyssel" and

87 the crews manning the helicopter and fixed-wing aircraft, and appreciate the support of Rieber Shipping A/S, Helikopterteneste a.s. and Fotonor AS. We also extend our thanks to BjCrn Bergflcdt, Kjell Arne Fagerheim and Karl Tellnes of the Institute of Marine Research, and to Tore Haug, Kjell T. Nilssen and Nils-Erik Skavberg of the Norwegian Institute of Fisheries and Aquaculture in Tromsr who all contributed by enthusiastic participation in the field work. Kjell Arne Fagerheim also handled all visual analyses of the photographic material and Siri Hartvedt provided assistance with the punching and analysis of data. Visitors who joined individual expeditions to pursue independent studies and contributed to activities in the field included Jonny Beyer, University of Bergen, Robert Eisner, University of Alaska, Lars Folkow and Per-Erik MArtensson, University of Tromsr V.A. Potelov and V.F. Prishchemikhin, sevPINRO, Arkhangelsk and Jack Terhune, University of New Brunswick. The project was funded by the Norwegian Fisheries Research Council (NFFR No. 4001701.304) and in 1994 by the Norwegian Research Council (NFR No. 104.500/110).

References 1. Oritsland T, Folkow L. Selmerking i Vesterisen. Rapport Fiskeridirektoratets havforskningsinstitutt, SPS 8904. Bergen: 1989; 1-4. 2. Folkow LP, Blix AS. Tracking harp and hooded seals in the Greenland and Barents Seas. Working Paper ICES Working Group on Harp and Hooded Seals, 16-18 October. SEA- 17. Bergen: 1989; 16. 3. ~ritsland T, Haug T, Oien N, BergflCdt B. Telling, merking og undersCkelser av sel i Vesterisen. Rapport Havforskningsinstituttet, SPS 9004. Bergen: 1990;1-11. 4. IDritsland T, Fagerheim KA, Oien N. West Ice seal survey and tagging in 1991. Working Paper Joint ICES/NAFO Working Group on Harp and Hooded Seals, 14-18 October, SEA-23. Copenhagen: 1991;1-13. 5. Oien N, Oritsland T. Aerial and visual surveys to estimate harp seal pup production in the Greenland Sea. ICES Coun Meet 1993; N9:1-9. 6. ~ien N, 13ritsland T. Recaptures of harp seals (Phoca groenlandica) tagged as pups in the Greenland S e a - pup production and dispersion patterns. Working Paper Joint ICES/NAFO Working Group on Harp and Hooded Seals, 14-18 October, SEA-33. Copenhagen: 1991 ;1-21. 7. Oien N. Use of mark-recapture experiments to monitor seal populations subject to catching. In: Blix AS, WallCe L, Ulltang 13 (eds) Whales, Seals, Fish and Man. Amsterdam: Elsevier, 1995;3545. 8. Estep KW, Maclntyre F, Noji TT, Stensholt B, Oritsland T. Seal sizes and habitat conditions assessed from aerial photography and video analysis. ICES J Mar Sci 1994;51:253-261.


Stock identity and social organization


© 1995 ElsevierScienceB.V. All rightsreserved Whales, seals, fish and man A.S. Blix, L. Wallceand 0. Ulltang,editors

Genetic markers

91

and whale stocks in the North Atlantic ocean:

a review

Alfrec3 Arnason lmmunogenetics Unit, Department of Pathology, University Hospital, Reykjavfk, Iceland Abstract. The discovery of protein electrophoresis and later of applying electrophoresis to DNA fragments is a powerful tool in comparing animals from different areas. We discuss our experience of using these techniques in comparing stocks of whales. The emphasis is laid on fin (Balaenoptera physalus), sei (B. borealis) and minke whales (B. acutorostrata). Of 31 enzyme systems encoded by 40 loci in fin whale liver samples from Iceland and Spain, 11 loci were found to be polymorphic. The average heterozygosity for these two stocks of fin whales was 0.074 and 0.083, respectively. There were between years differences observed in the fin whales suggesting substructures of the populations. The Nei's genetic distance between Icelandic and Spanish fins was 0.016. In a limited study of fin whale samples from Norway, Iceland and Canada there were statistical differences between enzyme loci tested. No between years differences were found in seis that came only from Iceland. The isozyme study of minke whales revealed genetic differentiation between Norway, Iceland and West Greenland. DNA fingerprinting showed similar differentiation. Earlier morphometric comparisons and tagging experiments pointed in the same direction. Immunogenetical studies of the MHC region of fin and sei show lack of polymorphism in these species. The C4 is more similar to terrestrial animals regarding polymorphism. The C4 typing revealed hybridisation between blue and fin whales; three hybrids were found, one fertile female in her second pregnancy and two males. This demonstrates how closely related these two species are. The conclusion from our study is that fin whales from Norway, Iceland, Canada and Spain are separate stocks. The same applies to minke whales from Norway, Iceland and West Greenland; they represent different populations. Key words: fin, sei, minke, isozymes, DNA, MHC, hybrids

Introduction In this short review we concentrate mainly on our investigation over the past 18 years or so. This is bound to reflect the technical advances during the period. The advent of electrophoresis made possible the separation of different proteins and later of DNA fragments. The patterns acquired give the picture of different genes or their products, the proteins; these are the so-called genetic markers. We have applied a variety of these techniques to study several species of whales from different areas of the North-Atlantic Ocean. R/Srvik and Jonsgaard [1] divide fin whales in the North Atlantic Ocean into six stocks: 1. North Norway and Arctic Eastern North-Atlantic; 2. East Greenland and Iceland; 3. West-Norway and the Faroes; 4. British Isles, Spain and Portugal; Address for correspondence: Immunogenetics Unit, Department of Pathology, University Hospital, Reykjav~, Iceland.

92 5. 6.

West Greenland; Nova Scotia, Newfoundland and Labrador. The IWC divides the fin whales of the North Atlantic into seven management stocks, the divisions the same as above, except stock 6 is further subdivided into two stocks: Newfoundland-Labrador and Nova Scotia stocks [2]. Other ideas have been put forward, e.g. that the fin whales in the North Atlantic form a patchy continuum [3]. The sei whales are divided into three management stocks by the IWC [2]: 1. Iceland-Denmark Strait; 2. Eastern stock (Norway, Scottish Isles, Spain and Portugal); 3. Nova Scotia. The IWC division of minke whales into four management stocks is as follows [2]: 1. the Canadian East Coast stock; 2. West Greenland stock; 3. Central stock (Iceland); 4. Northeastern stock (Norway). The IWC stock divisions are mainly based on catch statistics, and some biological characters such as morphometry and tag returns. The aim of the present study was to use genetic marker systems in order to clarify, if possible, the stock identity of the above-mentioned whale species in the North Atlantic Ocean. This review also deals with some comparative immunogenetical markers in the major histocompatibility complex (MHC region) as well as complement factor 4 (C4).We also discuss species hybridization between the blue and the fin whale and how this reflects the close genetic relationship between these species.

Material and Methods

Electrophoresis was the separation method used, adjusted to each marker system. The proteins investigated were mostly isozymes, each locus recognized by specific substrates. The DNA fragments were recognized by specific probes and restriction enzymes and in some cases by PCR magnification. Otherwise we refer to the original papers cited in the text. The samples came from the following areas: fin whales (Balaenoptera physalus) mainly from Icelandic and Spanish waters, but smaller numbers for a pilot study from Canadian and Norwegian waters; sei whales (Balaenoptera borealis) only from Icelandic waters; minke whales (Balaenoptera acutorostrata) from Norway, Iceland and West Greenland. Only a few individuals from various other whale species were studied for the purpose of species comparison.

Results

The results presented here are a compilation from our earlier publications from dif-

93 ferent times thus being a review of what we know today about the whale stocks in the North Atlantic The fin whale

Both protein and DNA markers were used in the study. Protein markers

We started investigating esterases of fin whale blood in 1971, concentrating on carbonic anhydrases of the erythrocytes [4]. The total of 1,099 fin whales caught off Iceland were studied in the years 1971, 1981-1989. This system was in HardyWeinberg equilibrium all years except 1985 and 1986. Young males seemed to be responsible for this deviation from the Hardy-Weinberg equilibrium [4,5]. Liver esterases. The liver esterases gave a complex pattern and only the fast moving esterases were of use in this respect. No significant differences were observed between years or locations [4,6]. Liver esterase patterns were on the other hand useful for species comparison. Cardiac esterases. The electrophoretic esterase patterns of cardiac muscle homogenates of fin whales are highly polymorphic. This is discussed in detail in our earlier reports [4,7]. There is a highly significant difference between Iceland and Spain in frequencies of some fractions [4]. If we use the most frequent phenotypes as markers shown in Fig. 1, differences between Iceland and Spain seem obvious and are statistically significant. The differences between years could probably reflect herd structure. It should be borne in mind that we do not know the genetics behind the cardiac esterase patterns, so we are only using these as markers. Genetic variation at 40 enzyme loci in fin whales. Fin whale samples taken in 1985 and under special permit in 1986-1989 were used for an extensive study of isozymes and other markers [4,8]. A total of 31 out of the 38 enzyme systems examined gave adequate staining and resolution. The zymograms have been described

Phenotypes iI

0,8 0,7

II III

0,6 % 0,5 0,4

0,3

i~ !iiiii

0,2 0,1

!.

Fig.

1981 n=169

1982 1983 n=72 n=71 Iceland

1984 n=129

I

I

1983 n=29

1984 n=12 Spain

I

1. A bar chart demonstrating the frequencies of cardiac esterase phenotypes in fin whales caught off Iceland in the years 1981-1984 and in fin whales caught off Spain 1983-1984 [4].

94 Table 1. Allele frequencies of 11 polymorphic enzymes in fin whales caught off Iceland and Spain Locus

Subunit number

Alleles

Allele frequencies Spain 1985

Sample size Ada

(n) 1

Ak-I

(n) 1

Ca

(n) 1

Gpd

(n) 2

Ldh-A

(n) 4

Mdh-S

(n) 2

Mpi

(n) 1

Pep-A

(n) 1

Pgm-1

(n) 1

Pgi

(n) 2

Sod-A

(n) 2

1 2 3 4 1 2 F S 1 2 A A' 1 2 1 2 3 1 2 3 1 2 1 2 3 1 2

46 0.098 0.293 0.152 0.457 33 0.636 0.364** 24 0.187 0.813 46 0.533 0.467 46 0.674 0.326** 40 0.562 0.438 0.000 45 0.611 0.389 0.000 45 0.100 0.900 46 0.858 0.109 0.033 46 0.315 0.685

Iceland 1985

Iceland 1986

65 49 0.085 0.069 0.231 0.139 0.330 0.271 0.354*** 0.521"** 64 67 0.164 0.172 0.836*** 0.828*** 145 0.276 0.724* 65 0.138 0.862 65 72 0.531 0.729 0.469 0.271 65 72 0.669 0.299 0.331 0.701"* 60 62 0.133 0.113 0.867 0.887 0.000 0.000 20 71 0.725 0.676 0.275 0.317 0.000 0.007*** 65 59 0.108 0.153 0.892 0.847 63 70 0.937 0.915 0.063 0.064 0.000 0.021 65 72 0.208 0.312 0.792 0.688

Iceland 1987

78 0.032 0.346 0.109 0.513"** 73 0.192 0.808*** 78 0.814 0.186 78 0.147 0.853 78 0.038 0.949 0.013 73 0.610 0.370 0.020*** 74 0.101 0.899* 78 0.911 0.051 0.038 78 0.391 0.609

Iceland 1988

68 0.081 0.279 0.051 0.588** 61 0.492 0.508*** 68 0.801 0.198 68 0.044 0.956 68 0.140 0.838 0.022 58 0.715 0.241 0.430** 65 0.262 0.738 68 0.919 0.037 0.044 68 0.140 0.860

Danfelsd6ttir et al. [8]. Significant deviations in genotypic frequency from Hardy-Weinberg I (*) and the sample size (n). *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; -, not screened.

p r e v i o u s l y . I n fin w h a l e s a m p l e s c a u g h t o f f I c e l a n d a n d S p a i n , t h e 31 e n z y m e s r e p r e s e n t 4 0 l o c i o f w h i c h 11 ( 2 7 . 5 % ) w e r e p o l y m o r p h i c b y t h e 0 . 9 5 c r i t e r i o n f o r p o l y m o r p h i s m , i.e. t h e r a r e a l l e l e h a d a f r e q u e n c y o f at l e a s t 0.05. A l l e l e f r e q u e n c i e s f o r t h e p o l y m o r p h i c loci in fin w h a l e s a m p l e s are l i s t e d in T a b l e 1. In f i n w h a l e s , t h e

95 loci suitable for the chi-square goodness-of-fit test for the Hardy-Weinberg equilibrium were: Ada, Ak-1, Ca, Gpd, Ldh-A, Mdh-S, Mpi-1, Pep-A, Pgm-1, Pgi and Sod-A. Where the expected number of a given genotypic frequency was > >

> < > < < <
> > < < >

> < > < < >

Hooded

Harbor

N = 15

N = 17

5.1 0.7 9.1 9.3 1.5 23.6 3.6 0.5 1.7 0.9 2.4 15.4 0.3 5.2 7.3 1.0 2.4 10.1 16 61 23

_ 0.6 _ 0.2 • 1.7 • 1.7 _ 0.3 _-4-2.2 _ 0.6 _ 0.1 _ 0.2 _ 0.2 • 0.5 _ 2.4 • 0.1 • 1.3 _ 1.5 • 0.3 • 0.5 • 1.8

> < >

< > < > >
>

< > <

60

?5

0 .25

r (1.)

3o0

Male

,

t

240 0

180 120

>.

?5

60 0 !

11.25

11.35

11.45

11.55

T i m e of day Fig. 3. Two continuous respiration profiles sampled simultaneously through visual observations of focal whales (readily identifiable in the field), during their cooperative feeding on herring (n = 40 + 45 dive durations).

Discussion The social system and behavioural ecology of Norwegian killer whales are gradually pieced together with photographic, bioacoustic and behavioural data. As for most studies on free-ranging whales, this project also has a long-term perspective. Considering the preliminary insights that have been obtained so far, the killer whales off the Norwegian coast seem to have social characteristics in common with their Canadian conspecifics. The cooperative nature of Norwegian killer whales is stated both in their coordinated feeding activities [9], and in the communal care of young where small individuals are hydrodynamically carried by adult group members. The genealogy of the whales is unknown, but the photographic data still indicate stability in group com-

174 position. If Norwegian killer whales are socially organized in the same way as Canadian killer whales, the groups would be matrilineal kin units where even the adult males refrain from dispersing from their natal "pod" [3]. However, intraspecific variation might be found among the strategies of the males. Some of the Norwegian killer whale males seem to have a nomadic lifestyle where they repeatedly occur together with different groups of whales. The presence of other males in the groups does not prevent these males entering the group, participating in cooperative feeding or leaving the group together with a female. These nomads are few, compared to males that are consistently identified together with the same females, and they might represent one of two strategies in accordance with the mating system of the brown hyena (Hyaena brunnea) [10]. Among these animals, the males either become nomadic and mate with unrelated females, or stay in their natal clans and gain reproductive success through alloparenting the cubs of their female relatives. Another possibility, if the killer whale males do have a nomadic strategy, might be the importance of age, size and social rank upon dispersal from the natal group. Because of the previous catch of Norwegian killer whales, it is also possible that the nomadic males are not representing any strategy but are merely a consequence of disrupted groups. Male mammals usually compete for access to receptive females [ 11 ]. Among bottlenose dolphins (Tursiops truncatus), male competition takes the form of physical fighting resulting in heavily scarred bodies of the oldest animals [ 12]. Male aggression has not been observed from the surface behaviour of Norwegian killer whales, but the males frequently occur in pairs, detached from other whales in their vicinity. At these instances, the males might spend a considerable amount of time in a ritualized activity descriptively termed "body contact-belly up-beak genital". Due to the quantified reciprocity and symmetry of roles in the display, the behaviour resembles the sexual gestures used as ritualized greetings between adult male baboons (Papio cynocephalus anubis) [13]. These male primates form stable coalitions where they take the oestrus females of a troop away from other males. In that respect, male baboons have a similar strategy to male bottlenose dolphins in Shark Bay, western Australia [ 14]. The behavioural studies of Norwegian killer whales are too preliminary to be conclusive about the formation of alliances or coalitions among males. The sexual activities observed to occur between two males and a female might however indicate some form of cooperative relationship between adult males. The size differences between male and female killer whales were discussed by Bain [15]. He quotes the work of Peters [16], and since the larger sized males are able to swim more efficiently at higher speeds than females, they would also have the ability to feed at greater depths. Bain suggests that the philopatric males in the Canadian killer whale population might "disperse" ecologically rather than geographically. Adult males would thereby reduce the cost of sharing food resources with their maternal groups. In a Norwegian version of the idea of ecological dispersal, respiration profiles were measured for pairs of killer whales while they were cooperatively feeding on

175 herring with their groups. The diving behaviour of males and female-sized whales during feeding was however highly synchronous. No differences were found that could attribute a specific role for the larger males when measuring the respiration patterns. However, these data were sampled in an area where food is superabundant, and only in a final stage of the feeding when the herring has been herded towards the surface. The males might still compensate for their larger size and higher energetic needs. Solitary feeding on fish has only been observed to include males, which is also described to occur among Canadian killer whales [ 17]. The acoustic communication of killer whales consists of high frequency whistles and pulsed signals that can be either variable or discrete [18]. Discrete "calls" are produced by the whales in a stereotyped way, and each Canadian killer whale pod was found to have a limited but unique and stable repertoire of discrete calls [19]. The pod-specific vocalization was termed "dialects", and is suggested to have evolved among Canadian killer whales as a result of the social structure of the population [4]. The primary function of dialects might be to promote group cohesiveness and to coordinate behavioural activities. The acoustic properties of Norwegian killer whale vocalization resemble dialects. Different groups can be acoustically distinguished, repertoire sizes of discrete calls are comparable, there is a variable degree of similarity between groups and the relative use of calls is about the same [7]. The knowledge of the social organization of Norwegian killer whales is however incomplete. It remains to be demonstrated that all groups are stable units with no long-term exchange of individuals or sub-groups, which is the prerequisite of dialects. Many years of field work on Norwegian killer whales are still required before the results can be adequately compared with the insights revealed from the Canadian killer whale population. Knowledge of the Canadian whales is meanwhile being used as a "model" in the Norwegian research, to search for similarities or variations. Sometimes both occur, as in a small group of Norwegian killer whales that seem to have developed specialized feeding on marine mammals. In that respect, they are "transient" as the carnivorous killer whales in the Canadian population are termed [20]. The Norwegian "transients" still seem to diverge from the Canadian definition, since they are not socially isolated from fish-feeding groups of whales. Further aspects on comparative social ecology will clarify differences between characteristic traits for the killer whale and intraspecific variations due to various ecological factors.

Acknowledgements For the enthusiastic assistance on board M/S Bella, we acknowledge Marianne Olsen, Bente Brekke, Elle Lettevall, Tyri Askl6f, Leffe Strandh, brother and sister Bisther. The project was financed in the period 1990-1993 by The Norwegian Fisheries Research Council as part of the national research program on marine mammals.

176

References 1. Christensen I. Distribution, movements and abundance of killer whales (Orcinus orca) in Norwegian coastal waters, 1982-1987, based on questionnaire survey. Rit Fiskideildar 1988;11:7988. 2. ~ien N. The distribution of killer whales (Orcinus orca) in the North Atlantic based on Norwegian catches, 1938-1981, and incidental sightings, 1967-1987. Rit Fiskideildar 1988;11:65-78. 3. Bigg, MA, Olesiuk PF, Ellis GM, Ford JKB, Balcomb KC. Social organization and genealogy of resident killer whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Rep Int Whal Commn 1990;(Special Issue 12):383-405. 4. Ford JKB. Vocal traditions among killer whales (Orcinus orca) in coastal waters of British Columbia. Can J Zool 1991 ;69:1454-1483. 5. RCttingen I. A review of variability in the distribution and abundance of Norwegian spring spawning herring and Barents Sea capelin. Polar Res 1990;8:33-42. 6. Bigg MA, Ellis G, Balcomb KC. The photographic identification of individual cetaceans. Whalewatcher (J Am Cetacean Soc) 1986;20(2):10-12. 7. Bisther A. The acoustic communication of social groups of photographically identified killer whales (Orcinus orca) at the coast off Norway. Master thesis, University of Gtiteborg, Sweden, 1991. 8. Lang TG. Hydronamic analysis of cetacean performance. In: Norris KE (ed) Whales, Dolphins and Porpoises. Berkley, CA: University of California Press, 1966;410--432. 9. Simil~i T, Ugarte F. Surface and underwater observations of cooperatively feeding killer whales, Orcinus orca, in northern Norway. Can J Zool 1993;71:1494-1499. 10. Mills MGL. The comparative behavioral ecology of hyenas: the importance of diet and food dispersion. In: Gittleman JL (ed) Carnivore Behavior, Ecology and Evolution. London: Chapman and Hall, 1989;125-143. 11. Poole T. Social Behaviour in Mammals. New York: Chapman and Hall, 1985. 12. Scott MD, Irvine AB, Wells RS. A long-term study of bottlenose dolphins on the west coast of Florida. In: Leatherwood S, Reeves RR (eds) The Bottlenose Dolphin. San Diego, CA: Academic Press, 1990;235-243. 13. Smuts BB, Watanabe JM. Social relationships and ritualized greetings in adult male baboons (Papio cynocephalus anubis). Int J Primatol 1990; 11 (2): 147-172. 14. Connor RC, Smolker RA, Richards AF. Two levels of alliance formation among male bottlenose dolphins (Tursiops sp.). Proc Natl Acad Sci USA 1992;89:987-990. 15. Bain DE. An evaluation of evolutionary processes: studies of natural selection, dispersal and cultural evolution in killer whales (Orcinus orca). Ph.D. dissertation, University of California, Santa Cruz, 1989. 16. Peters RH. The Ecological Implications of Body Size. New York: Cambridge University Press, 1983. 17. Hoelzel RA. Foraging behaviour and social group dynamics in Puget Sound killer whales. Anim Behav 1993;45:581-591. 18. Ford JKB. A catalogue of underwater calls produced by killer whales (Orcinus orca) in British Columbia. Can Data Rep Fish Aquat Sci No 633, 1987. 19. Ford JKB, Fisher HD. Group-specific dialects of killer whales (Orcinus orca) in British Colombia. In: Payne R (ed) Communication and Behavior of Whales. AAAS Sel Symp Series 76, Boulder, CO: Westview Press, 1983;129-161. 20. Baird RW. Foraging behaviour and ecology of transient killer whales (Orcinus orca). PhD thesis, Simon Fraser University, Canada, 1994.

9 1995Elsevier Science B.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand O. Ulltang, editors

177

Possible effects of previous catch on the present population of Norwegian killer whales (Orcinus orca) D a g V o n g r a v e n 1 and A n n a Bisther 2 1Department of Zoology, University of Trondheim-A VH, Dragvoll, Norway; and 2Kristineberg Marine Research Station, University of Gothenburg, 45 034 Fiskebiickskil, Sweden A b s t r a c t . Intensive coastal killer whale catches undertaken in the two decades after 1960 (especially in

1969, 1970 and 1979), might have had effects on reproduction and social behaviour in the present Norwegian killer whale community. The catch was both sex- and age-biased, and this might have triggered compensatory mechanisms. Our approach when studying the social ecology of Norwegian killer whales must take into account the possible presence of such mechanisms. Further modelling studies are in progress. K e y w o r d s : killer whales, catch data, reproduction, social ecology

Introduction

The killer whale population off the Norwegian coast has previously been subjected to a hunting pressure with unknown effects upon the present population. Results and insights from an on-going photo-ID study indicate that this population is stationary and has evolved complex social behaviour. Other species of long-lived, slow-reproducing mammals that live in socially structured societies have been shown to respond to hunting with changes in social strategies, either as an effect of a reduced population density (i.e. simakobu monkey Nasalis concolor [1 ]) or as an effect of selective takes of large and old individuals (i.e. elephants Loxodonta africana [2]). Could peaks in the catch of killer whales one generation ago have had any other effects than a reduced size of the present population?

Materials and Results

There are official catch records of 2435 killer whales caught in the North Atlantic in the period from 1938 to 1981. Of these, 64% were caught in the coastal waters off Norway. Twenty-nine percent of the total catch in the whole period occurred in the three seasons 1969 (231), 1970 (246) and 1979 (221), and 91% of these were caught in the coastal fishery zones. Temporal and spatial intensity is especially characteristic of the catch during these years, during which the majority of the whales were caught

Address for correspondence: D. Vongraven, Department of Zoology, University of Trondheim-AVH, 7055 Dragvoll, Norway.

178 a

1969/1970

b

1979 Sex ratio in total catch (F : M)

Sex ratio in total catch (F : M)

1:2.1 20

1.2:1 20

IMMATURE WHALES

~ ~o

"~

IMMATURE WHALES

10

s 0

MALES

FEMALES BOTHSEXES

0

Fig. 1. Characteristics of the Norwegian killer whale catch, depicting the sex- and age-biases in the peak seasons (a) 1969 and 1970; and (b) 1979.

in the space of 1-2 months within a single fishery zone every season. The 1969/1970 catch was concentrated in the MOre region, while the 1979 catch was concentrated in Lofoten. Females are sexually mature at an average length of 15 ft, and males at 18 ft [3]. Seventy-one percent of the total catch was recorded after 1960, and the fraction of immature whales in the same period was 11.9%. The corresponding fraction for the two high intensity years 1969 and 1970 was 6.9%, and for 1979 15.9%. Of all the whales caught after 1960, a fraction of only 4.2% were smaller than 15 ft (Fig. 1). The overall sex ratio for sexually mature whales is close to 1:1.3 (females/males) for all seasons. However, there are large variations between years. The sex ratio for the two seasons 1969/1970 was 1:2.1, whereas the sex ratio for the 1979 season was 1.2:1 (Fig. 1). There was a switch from a female-biased to a male-biased sex ratio at lengths above 19 ft.

Discussion There are two factors that indicate that the catch in the peak seasons could be regarded as high compared to the total size of the killer whale community. First, a total number of identified whales in the range of 5-700 individuals and fractions of resightings of 65% and 90% in the two on-going photo-ID studies off the Norwegian coast indicate that a population size much bigger than this is improbable (Simil~i personal communication; Bisther and Vongraven, unpublished data). Second, the female bias in the catch from 1979 could be considered as a direct effect of the malebiased catch from previous years, and especially the 1969/1970 catch, if population size was in the size range previously suggested.

179 There is an obvious size bias in the catch from these peak seasons. Given that young whales rely on nursing and care-taking from adult whales for survival, the relative absence of calves and juveniles in the latest catch records could have led to an increase in their future mortality as many parents and potential care-giving individuals were removed. In the Pacific Northwest, when comparing two populations of killer whales with different exploitation histories, Bain [4] found a neonate mortality of 63% in the cropped population and 41% in the uncropped population, whereas adult survivorship was similar. A tendency towards a higher juvenile mortality (up to an age of 15 years) in cropped than in uncropped pods has also been shown by Olesiuk et al. [5]. Live-capture fishery for killer whales in the North-American Pacific Northwest removed approx. 25% of the initial population in the period 1964-1975 [6]. If the population size of Norwegian killer whales in the 1960s was in the order of magnitude previously suggested, then the "coastal" fraction of the 477 whales caught in 1969 and 1970 would at least represent a similar fraction of the population at that time. Findings like these point to the fact that killer whale reproduction depends on social as well as on density dependent determinants. It is also easier to comprehend compensatory mechanisms being induced on a group level rather than on an overall reduced density in the whole area inhabited by the population [7]. Destruction of social structures caused by biased removal of individuals from the population could account for some of the loose group structure suggested by our data. Further modelling studies with basis in the catch data will be carried out. 13y means of different scenarios for population status and catch regime, and previously published vital rates, we will try to investigate what possible effects the catch might have had on the social structure of the population. Finally, we wish to place the emphasis on the potential importance of the catch, and on the influence this ought to have on our approach when studying the social system of Norwegian killer whales. References 1. Watanabe K. Variations in group composition and population density of two sympatric mentawaian leaf-monkeys. Primates 1981 ;22:145-160. 2. Eltringham SK. Elephants. Blandford Books, 1982;52-56. 3. Christensen I. 1982. Killer whales in Norwegian coastal waters. Rep Int Whal Commn 1982;32:633--641. 4. Bain DE. An evaluation of evolutionary processes. Ph.D. Thesis, University of California, Santa Cruz, 1988. 5. Olesiuk PF, Bigg MA, Ellis GM. Life history and population dynamics of resident killer whales O r c i n u s o r c a in the coastal waters of British Columbia and Washington State. Rep Int Whal Commn 1990;(Special Issue 12):209-243. 6. Bigg MA, MacAskie IB, Ellis G. Abundance and movements of killer whales off eastern and southern Vancouver Island with comments on management. Preliminary Report, Arctic Biological Station, Ste. Anne de Bellevue, Quebec, 1976. 7. Fowler CW. Density dependence in cetacean populations. Rep Int Whal Commn 1984;(Special Issue 6):373-379.


Distribution, diet and feeding ecology


9 1995 ElsevierScience B.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand t3. Ulltang,editors

183

New approaches to studying the foraging ecology of small cetaceans A n d r e w J. R e a d Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USA Abstract. Dolphins and porpoises spend the majority of their lives underwater, out of the view of human observers. Consequently, scientists have relied on indirect means to study the foraging ecology of these animals. These indirect methods, such as examining the stomach contents of carcasses, provide an incomplete and often biased view of their feeding ecology. Promising new technological developments with data loggers and satellite telemetry are beginning to offer alternative methods to studying the foraging ecology of small cetaceans. Data loggers have long been used with other marine vertebrates but have not been suitable for dolphins or porpoises because the devices must be retrieved to recover stored data. Recent advances in their recovery have overcome some of these difficulties and loggers have now been successfully deployed on several species of small cetaceans. A new generation of small satellitelinked transmitters has allowed researchers to follow the movements and diving behaviour of animals from their offices. Continuing developments in these fields will afford new opportunities to study the lives of dolphins and porpoises and better understand how they find and obtain food at sea.

Key words: foraging, ecology, small cetaceans

Introduction Dolphins and porpoises spend most of their lives underwater, out of the view of human observers. Many species also live in remote habitats where direct observation of their behaviour is difficult and often impossible. Consequently, for most of these animals, our knowledge of their foraging ecology is limited to inferences made from the examination of carcasses obtained as strandings, incidental catches in commercial fisheries, or from directed catches. Even when we are able to observe dolphins and porpoises catch their prey, we are usually limited to events at or near the surface that are unlikely to be representative of the true range of predator-prey interactions. The traditional approach to studying the foraging ecology of small cetaceans has been to examine stomach contents and attempt to reconstruct the diet from samples of hard parts [21,22]. With this dietary information in hand, it is then possible to make some inferences regarding the foraging ecology and behaviour of the marine mammal. A preponderance of benthic and demersal prey items in the diet would indicate that an animal was feeding at or near the sea floor, for example. This is a valuable exercise that provides important information on the diet of small cetaceans. It is also fraught with potential pitfalls and biases that must be taken into account when reconstructing the diet and making inferences about behaviour. Most of these poten-

Address for correspondence: Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA.

184 tial biases are well known (see the review of Pierce and Boyle [21] for a full description). A few examples will illustrate some of the problems associated with such dietary analysis. First, the sample may not be representative of the diet of the population because of the manner in which specimens were obtained. For example, observations of the stomach contents of stranded animals may not reflect their true diet, particularly if the animals were sick for some time before their death. The stomach contents of animals killed incidentally in fishing operations may be biased towards the target species of the fishery and not represent the diet of the population when it is not interacting with fishing gear. A second class of biases arises through the differential ingestion, digestion, and retention of hard parts. Some animals may ingest only the bodies of prey, without consuming the heads (and otoliths). The rates at which hard parts digest in the stomachs of small cetaceans are not known, but it is clear that these rates differ among prey taxa. Squid beaks are resistant to digestion, for example, and are believed to be retained in the stomach much longer than otoliths [5]. It is particularly difficult, therefore, to reconstruct the diet of a marine mammal that feeds on both squid and fish with any accuracy. Finally, the presence of a particular item in the diet tells us little of how, when or where the marine mammal came to locate and capture the prey. A mid-water fish that undertakes pronounced vertical migration could be captured at depth during the day, or near the surface between dusk and dawn; its presence in the stomach tells us little about the foraging behaviour of the predator. A considerable amount of research has helped to overcome some of these biases. Indirect means of examining the diet of marine mammals have been developed, such as studying the fatty acid composition [ 11 ] or stable isotope ratios [ 1] of tissues and comparing them to potential prey. The digestion rates of hard parts of various sizes and shapes have been measured in the laboratory to help understand the biases associated with differential digestion times [22]. And biochemical techniques have been developed to identify the presence of different prey species without using hard parts [21 ]. Taken together, these methods offer researchers a powerful array of tools with which to study the diet of small cetaceans. More promising, however, is the development of methods that allow researchers to follow the activities of individual animals at sea and thus make more direct observations of foraging behaviour [7]. Early attempts to use conventional radio telemetry to study the diving behaviour of common dolphins (Delphinus delphis) were promising [9], but the tags available at that time were too cumbersome for widespread application. Recent advances in the miniaturization, attachment and recovery of data Table 1. Deployments of time-depth recorders (TDRs) on small cetaceans

Species

N

Maximum depth (m)

Reference

Harbour porpoises Spotted dolphins Orcas

7 5 7

226 203 173

[27] [24] [3]

185 loggers and satellite-linked tags have allowed researchers to begin to use these tools to study how marine mammals find and obtain their food at sea [6,7]. The basic technology for both these approaches has been available for some time, but only recently have they been suitable for use with dolphins and porpoises. These applications will revolutionize the way that we study the foraging ecology of small cetaceans in the next decade.

Data loggers Data loggers are microprocessor-controlled devices that record information obtained from external sensors. The most common and best known application of this system is the time-depth recorder (TDR), which has been used on a variety of marine vertebrates and developed extensively with pinnipeds [13]. The earliest TDRs mechanically recorded depth profiles on film [6]. The current generation of TDRs sample ambient pressure at pre-set intervals and store this information as hexadecimal code on a microchip [6,7]. The data are downloaded to a computer after the TDR is retrieved and the code converted to depth information. The need to recover the TDR has been one of the major obstacles to the use of these devices on dolphins and porpoises. TDRs can be attached to and recovered from pinnipeds when they haul out, often in a predictable location. Unfortunately, dolphins and porpoises are not inclined to haul out and allow researchers to attach devices to them. Three innovative approaches have overcome this obstacle, however, and TDRs have now been successfully deployed on harbour porpoises (Phocoena phocoena), spotted dolphins (Stenella attenuata) and orcas (Orcinus orca) (Table 1). The first successful deployment of a TDR on a cetacean was made by Westgate et al. [27], who studied the diving behaviour of harbour porpoises in the Bay of Fundy, Canada. These researchers developed a tag that incorporated a TDR and a small VHF radio tag in a buoyant epoxy package. The tags were attached to porpoises released from herring weirs with the help of commercial fishermen. The attachment mechanism incorporated two small magnesium nuts that corroded rapidly in salt water, allowing the package to detach from the porpoise and float to the surface, where it could be located by conventional radio telemetry. Seven of eight tags were recovered, with one tag retrieved 17 days after it was deployed. Scott et al. [24] tagged and tracked spotted dolphins associated with yellowfin tuna (Thunnus albacares) in the Eastern Tropical Pacific. These dolphins were captured with a chartered commercial purse seiner, tagged with a saddle-mounted TDR and VHF radio assembly and followed from a research vessel. After a period of several days the animals were recaptured with the purse seiner and the saddles removed. In addition, these researchers investigated the bond between dolphins and yellowfin tuna by attaching acoustic tags to the tuna and simultaneously tracking the movements of fish and mammals. Another unique approach was taken by Baird and Goodyear [3,4], who attached a buoyant TDR-VHF package to orcas in the waters of Vancouver Island, British Co-

186 lumbia. The tags were attached to the whales with suction cups, using a crossbow or long pole. A magnesium disc was incorporated into the wall of the suction cup; when the magnesium corroded and the suction was broken, the package detached and floated to the surface. These tags were recovered after a deployment period of up to 8.4 h [3]. In each of these three cases, TDRs provided a wealth of new information on the diving behaviour of the species in question. For example, in the study of Westgate et al. [27], one harbour porpoise dove to the deepest part of the Bay of Fundy (226 m) and most of the tagged porpoises routinely made dives to over 100 m. Most dives were flat-bottomed, with rapid descent and ascent rates and prolonged periods of time at a relatively constant depth during the mid-portion of the dive. The authors interpreted this behaviour as foraging, probably at or near the sea floor, although it was not possible to obtain synoptic data on water depth. This supports the observations of Recchia and Read [23], who found a preponderance of Atlantic herring (Clupea harengus) and silver hake (Merluccius bilinearis) in the stomach contents of harbour porpoises killed in bottom-set gill nets. Herring and silver hake are found predominantly at depth during the day, although both are believed to migrate upwards in the water column at night. Surprisingly, however, the tagged porpoises made longer and deeper dives at night, when their prey are thought to be closer to the surface. This diel variation in diving behaviour is not currently understood, but will undoubtedly spur further study of the foraging behaviour of this species. These three studies demonstrate some of the potential for using data loggers, and particularly TDRs, to study the foraging ecology of small cetaceans. We are beginning to gain an appreciation for the diversity of diving behaviour in dolphins, porpoises, and small whales and to consider some of the factors responsible for this variation. The three studies also demonstrate some of the limitations of this approach as it is currently employed. The use of TDRs with small cetaceans requires researchers to both attach and recover the tags. The non-invasive attachment of Baird and Goodyear [3,4] obviates the requirement of capturing an animal to attach a TDR, but their approach does not allow for a long-term deployment that will provide information on diel patterns in behaviour. In addition, we are still limited in our understanding of the foraging ecology of these animals because TDRs give us only an indirect view into their feeding behaviour. We assume that flat-bottomed dives represent foraging excursions, but have no way of testing this assumption. I shall return to this point below.

Satellite telemetry The use of conventional radio tracking techniques, such as VHF telemetry, has greatly increased our ability to track the movements and behaviour of small cetaceans at sea. Limiting this methodology, however, is the need to stay in radio contact with tagged animals. During the last two decades, a new technique has revolutionized the ability of biologists to track the movements of animals in remote

187 areas. This technique, known as satellite telemetry, utilizes radio transmitters that send signals to receivers aboard the US National Oceanographic and Atmospheric Administration (NOAA) Tiros-N weather satellites. The receivers are operated by Service ARGOS, of Toulouse, France, which makes the system available to commercial and scientific users. The system has proven extremely effective in tracking large terrestrial mammals [12] and used with great success to telemeter information about the foraging ecology of pinnipeds [7,20]. The ARGOS system uses the Doppler shift to estimate the position of the transmitter or Platform Transmitting Terminal (PTI'). All PTI's transmit on a stable frequency of 401.650 mHz. The accuracy of the estimated position depends on the number and quality of transmissions received during a satellite pass. The transmission rate is limited both by the ARGOS system and by the surfacing behaviour of the animal. PTFs used with marine mammals typically employ a salt water switch to restrict transmissions to periods when an animal is at the surface and the signal can be received by the satellite. Each transmitter is identified by a unique code that is sent at the beginning of every message. In addition to the ID code, the transmitter can send up to 256 bits of data, which typically includes information from environmental sensors aboard the tag. The most frequent type of recorded data is information on depth and temperature, although velocity sensors are also employed on pinniped PTrs [20]. Until very recently, PTI's were too large for use with small cetaceans. In 1987, Mate [ 17] tagged a rehabilitated stranded pilot whale (Globicephala melaena) with a P T r and followed its movements and diving behaviour for 95 days. The whale made long dives at night, presumably to forage on the squid that is believed to comprise its primary prey. During the daylight hours, the pilot whale made shallow dives, perhaps resting or feeding on near-surface prey. Since this early success, advances in miniaturization have allowed PTrs to be used to study the movements and activities of bottlenose dolphins (Tursiops truncatus), narwhals (Monodon monoceros), belugas (Delphinapterus leucas), white-sided dolphins (Lagenorhynchus acutus), common dolphins (Delphinus sp.), and harbour porpoises (Table 2). The most successful programme has been the study of high Arctic belugas by Martin and colleagues [15,16]. Eighteen belugas were tagged with PTTs between 1988 and 1992, yielding a tremendous amount of information on movement patterns Table 2. Deployments of satellite-linked radio tags (PTTs) on small cetaceans

Species

N

Maximum duration (days)

Reference

Bottlenose dolphins Pilot whales Narwhals Belugas White-sided dolphins Bottlenose dolphins Common dolphins Harbour porpoises

14 1 3 18 1 1 2 3

35 95 19 75 6 25 ? 21

[25] [17] [14] [15] [19] [18] P. Thorson, personal communication A. Read, unpublished data

188 and diving behaviour. These belugas made dives of up to 18 min in duration and to depths of 440 m. As was the case with harbour porpoises, most of the long dives made by belugas were fiat-bottomed, with several minutes spent at a constant depth near the sea bed. These dives were interpreted as foraging excursions, supporting previous observations from stomach contents that belugas feed on benthic and demersal prey. The diving capabilities of these animals were unexpected, however, and the results demonstrate that belugas have access to much more of the Arctic sea floor than previously believed. Once again, this new technology has radically changed the way that we view the foraging ecology of these animals. The use of satellite telemetry has overcome the need to recover tags to obtain data, one of the main problems with the use of data loggers. But satellite tags have some of the same other limitations possessed by data loggers, particularly the need to capture animals to attach tags. And, as with data loggers, satellite telemetry provides only an indirect view of foraging behaviour. In addition, satellite-linked tags have their own limitations imposed by the ARGOS system. The maximum limit of 256 bits of data per transmission is a serious impediment to the collection of fine-scale environmental data from more than a single sensor. Even with data compression algorithms and duty cycling of different sensors, it is difficult or impossible to collect the kind of information that can be obtained with data loggers. The number of successful uplinks is also limited by the difficulty of coordinating transmissions with satellite passes; this is a particular problem in equatorial regions where the number of passes per day is very limited. The ARGOS system is currently limited to the two Tiros satellites which are both in polar orbits. Finally, the recent success of long-term tag attachments used with PTI's means that we are approaching the maximum transmission life of the current generation of batteries. It may be possible to address some of these limitations with new technological developments, as discussed below.

Future directions

Recent developments with both data loggers and satellite telemetry have produced impressive advances in our understanding of the foraging ecology of small cetaceans. Nevertheless, many important questions remain. For example, how important is echolocation in finding food? The prey of many dolphins and porpoises are known to produce sound, and it is likely that passive acoustics and vision are also important senses in the initial detection process. How do deep-diving animals find their prey several hundred metres below them near the sea floor? How important are cooperative feeding strategies in group-living species? These and many other questions will require more direct means of observing the behaviour of dolphins and porpoises. Below I list some avenues of technological research and development that could yield the answers to such questions. As with the development of data loggers and satellite telemetry, it is likely that many of these developments will first be tested on pinnipeds, before the constraints of deployment on small cetaceans are overcome. One area of future work that holds great promise is the development of new sen-

189 sors for use with data loggers and satellite-linked tags. Researchers are already experimenting with sensors that record stomach temperature [28] and heart rate [ 10]. It should soon be possible to determine when prey are ingested by monitoring reductions in stomach temperature and then link this information to diving patterns to test the hypothesis that long, fiat-bottomed dives are foraging excursions. Video may also make a dramatic impact on our ability to document foraging behaviour, at least for shallow water species [8]. Finally, loggers are now available that allow researchers to record the occurrence of echolocation clicks and thus relate acoustic behaviour to feeding events. Preliminary work has been conducted on several species in captivity [2,26] and it should soon be possible to adapt some of these techniques to the study of free-ranging animals. At present, it seems likely that the simultaneous use of several sensors will be best achieved with data loggers rather than PTTs. Data loggers are not as power-hungry as PTI's because they do not have to transmit data, thus the loggers can be considerably smaller than PTI's. The fact that loggers record rather than transmit data allows them to store information on a much finer scale. Such detailed records will be critical when we attempt to integrate information recorded simultaneously from several sensors (depth, stomach temperature and echolocation trains, for example). The possibility of obtaining such rich data records is tremendously exciting and will leave us struggling for analytical tools to cope with a wealth of information. The second type of technical developments that will further our study of foraging ecology lies in the development, attachment and recovery of the tags themselves. As noted above, for the most part we still need to capture animals to attach tags and, for many species, this is not possible. More work on non-invasive, remote attachments, like those of Baird and Goodyear [3,4], is required. We also need to incorporate active release mechanisms into the attachments used with data loggers, so that we can retrieve them when it is convenient to do so. Although this is a relatively straightforward task, at present we are limited by the need either to recapture a tagged animal or to wait for a passive mechanism to trigger before we can recover the data logger. Finally, we will need considerable advances in our ability to transmit large quantities of data rapidly to receiving stations, either aboard satellites or other platforms. Clearly, in all these cases a productive collaboration between engineers and biologists will be essential to progress. The last aspect of development in the study of foraging ecology of small cetaceans rests not in our ability to understand what dolphins and porpoises are doing, but in the fine-scale distribution and behaviour of their prey. As we increase our ability to monitor the activities of marine mammals, we rapidly outstrip the ecological information base required to interpret this behaviour. Those of us who have searched vainly through databases on commercial fisheries and stock assessments know too well that the scale of this information is much too coarse to be of value. Even when small cetaceans feed on commercially valuable prey species, we seldom have distributional data on a scale that is fine enough to interpret movement or diving patterns. In other cases, such as the pelagic dolphins that feed on mid-water squid and fishes of the deep scattering layer, even the taxonomy of prey species may be unresolved.

190 To test the hypotheses that we will formulate about the feeding strategies of these animals, it will be necessary to collect synoptic information about the distribution and movements of their prey. This may involve sophisticated means of detecting the presence of prey aggregations, such as the use of research sonar systems, and tracking the prey themselves to monitor movements and activity patterns.

Conclusions To those of us involved in studying the foraging ecology of small cetaceans, this is a tremendously exciting time to be conducting field research. Each data logger and satellite uplink holds the promise of new insight into the ecology and behaviour of these animals. As our toolbox of research techniques grows, we will be able to piece together the puzzle that describes how dolphins and porpoises make a living at sea. Our experience to date has demonstrated that with increased knowledge comes a myriad of unexpected questions and apparent paradoxes. Our present challenge is to develop new and benign tools to resolve these questions and better understand the role of small cetaceans in their ecosystems.

Acknowledgement I would like to thank my colleagues who work on harbour porpoises in the Bay of Fundy and on bottlenose dolphins in Sarasota, Florida for their dedication and friendship in the field. In particular, Andrew Westgate, Michael Scott, Randy Wells and Forrest Townsend have been instrumental in developing safe and effective tagging systems. Peter Tyack, Kurt Fristrup and Bill Watkins of the Woods Hole Oceanographic Institution have generously shared their insight and experiences. I also thank Tony Martin and Bruce Mate for their encouragement and advice. This paper was substantially improved by comments from Andrew Westgate. My research has been supported by the US National Marine Fisheries Service, US Office of Naval Research and World Wildlife Canada. Finally, thanks to Arne BjCrge for inviting me to Tromsr and to prepare this paper.

References 1. Abend A, Finn J, Smith TD. Diet prediction of the long-finned pilot whale (Globicephala melas) using carbon and nitrogen stable isotope tracers. Abstracts, Tenth Biennial Conference on the Biology of Marine Mammals, Galveston, TX, 1993; 19. 2. Akamatsu T, Hatakeyama Y, Kojima T, Soeda H. Echolocation rates of two harbor porpoises (Phocoena phocoena). Mar Mammal Sci 1994;10:401-411. 3. Baird RW. Foraging behaviour and ecology of transient killer whales (Orcinus orca). PhD Dissertation, Simon Fraser University, Vancouver, Canada, 1994. 4. Baird RW, Goodyear JD. An examination of killer whale diving behaviour using a recoverable,

191

5.

6. 7. 8.

9. 10.

11. 12. 13. 14. 15. 16. 17. 18.

19.

20. 21. 22. 23. 24.

25. 26. 27. 28.

suction-cup attached TDR/VHF tag. Abstracts, Tenth Biennial Conference on the Biology of Marine Mammals, Galveston, TX, 1993;25. Bigg MA, Fawcett I. Two biases in diet determination of northern fur seals (Callorhinus ursus). In Beddington JR, Beverton RJH, Lavigne DM (eds) Marine Mammals and Fisheries. London: George Allen and Unwin, 1985;284--299. Costa DP. Methods for studying the energetics of freely diving animals. Can J Zool 1988;66:4552. Costa DP. The secret life of marine mammals. Oceanography 1993;6:t20-128. Davis RW, LeBoeuf BJ, Marshall G, Crocker D, Williams J. Observing the underwater behavior of elephant seals at sea by attaching a small video camera to their backs. Abstracts, Tenth Biennial Conference on the Biology of Marine Mammals, Galveston, TX, 1993;40. Evans WE. Orientation behavior of delphinids: radio telemetric studies. Ann NY Acad Sci 1971;188:142-160. Goodyear JD, Andrews R. The first whale ECG with a self-contained tag and radio and satellitelinked depth and heart-monitoring tags for whales. Abstracts, Tenth Biennial Conference on the Biology of Marine Mammals, Galveston, TX, 1993;54. Iverson S. Milk secretion in marine mammals in relation to foraging: can milk fatty acids predict diet? Symp Zool Soc London 1993;66:263-291. Keating KA, Brewster WG, Key CH. Satellite telemetry: performance of animal tracking systems. J Wildlife Manage 1991;55:160-171. Kooyman G. Diverse Divers: Physiology and Behaviour. Berlin: Springer-Verlag, 1989. Martin AR, Kingsley MCS, Ramsay MA. Diving behaviour of narwhals (Monodon monoceros) on their summer grounds. Can J Zool 1994;66:446--458. Martin AR, Smith TG. Deep diving in wild, free-ranging beluga whales, Delphinapterus leucas. Can J Fish Aquat Sci 1992;49:462-466. Martin AR, Smith TG, Cox OP. Studying the behaviour and movements of high Arctic belugas with satellite telemetry. Symp Zool Soc London 1993;66:195-210. Mate BR. Watching habits and habitats from Earth satellites. Oceanus 1989;32:14-18. Mate BR, Rossbach KA, Nieukirk SL, Wells RS, Irvine AB, Scott MD, Read AJ. Satellitemonitored movements and dive-behavior of a bottlenose dolphin (Tursiops truncatus) in Tampa Bay, Florida. Mar Mammal Sci (in press). Mate BR, Stafford KM, Nawojchik R, Dunn JL. Movements and dive behavior of a satellitemonitored Atlantic white-sided dolphin (Lagenorhynchus acutus) in the Gulf of Maine. Mar Mammal Sci 1994;10:116-121. McConnell B J, Chambers C, Fedak MA. Foraging ecology of southern elephant seals in relation to the bathymetry and productivity of the Southern Ocean. Antarctic Sci 1992;4:393-398. Pierce GJ, Boyle PR. A review of methods for diet analysis in piscivirous marine mammals. Oceanogr Mar Biol Annu Rev 1991;29:409-486. Pierce GJ, Boyle PR, Watt J, Grisley M. Recent advances in diet analysis of marine mammals. Symp Zool Soc London 1993;66:241-261. Recchia CA, Read AJ. Stomach contents of the harbour porpoise, Phocoena phocoena (L.), from the Bay of Fundy, Canada. Can J Zool 1989;67:2140-2146. Scott MD, Chivers SJ, Olson RJ, Lindsay RJ. Radiotracking of spotted dolphins associated with tuna in the Eastern Tropical Pacific. Abstracts, Tenth Biennial Conference on the Biology of Marine Mammals, Galveston, TX, 1993;97. Tanaka S. Satellite radio tracking of bottlenose dolphins Tursiops truncatus. Bull Jap Soc Sci Fish 1987 ;53:1327-1338. Tyack P. A data logger to identify vocalizing dolphins. J Acoust Soc Arm 1991;90:1668-1671. Westgate AJ, Read AJ, Berggren P, Koopman HN, Gaskin DE. Diving behaviour of harbour porpoises, Phocoena phocoena. Can J Fish Aquat Sci (in press). Wilson RP, Cooper J, Plotz J. Can we determine when marine endotherms feed? A case study with seabirds. J Exp Biol 1992;167:267-275.


9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand ~. Ulltang,editors

193

Distribution and diving behaviour of hooded seals L.P. Folkow and A.S. Blix Department of Arctic Biology and Institute of Medical Biology, University of Tromsr Tromsr Norway Abstract. Hooded seals, Cystophora cristata, are abundant in the North Atlantic. This paper reviews current knowledge on the distribution and dive behaviour of these seals. The stock which breeds in sea ice near Jan Mayen may count about 250,000 animals, but little is known about where they stay and what they eat outside the pupping season (March/April) and the moult (July). We used satellite tags to monitor movements and/or dive depths and durations of 19 seals, and we obtained data on -12,000 locations and -120,000 dives, between July 1992 and July 1993. After the moult, most of the seals dispersed to travel, once or repeatedly, between the ice off Greenland and the distant waters off the Faeroe Islands, south of Bear Island, or the Irminger Sea. After breeding, all seals again returned to sea to travel to the waters off northern Ireland, the Faeroes or the Norwegian coast. Hooded seals may dive repeatedly to >1,000 m and stay submerged for >52 min, but usually dive to 100-600 m depth. We suggest that the dietary preferences, and even the fish consumption of hooded seals in different areas may be assessed by comparing their dive depths with the distribution of potential prey. Key words: arctic, feeding ecology, fisheries resources, North Atlantic

Introduction

Hooded seals (Cystophora cristata) are abundant in the ice-filled waters of the North Atlantic. Average lengths and weights of adult males and females are about 2.5 and 2.0 m, and 300 and 160 kg, respectively [1 ]. Previous studies of hooded seals have almost exclusively been conducted in connection with breeding, which takes place in the second half of March/early April, and moulting, which occurs between late June and early August. During these events, hooded seals aggregate in the drift ice and are accessible in large numbers for both studies and harvesting. During the remainder of the year, however, the seals disperse and are seldom found in large concentrations. For this reason, studies of the distribution and feeding ecology of the species during these periods have been difficult to carry out. However, modem satellite tracking techniques have recently made it possible to remotely monitor movements of radiotagged seals, via satellite. Moreover, such satellite tags, which are attached to the seal by gluing it to the fur, may also include different types of sensors allowing collection and subsequent transmission of data on for example dive depth and duration, heart rate, etc. This review of the distribution and dive behaviour of hooded seals integrates previous knowledge based on traditional techniques (field studies on breeding and moulting animals, incidental observations, and interviews with sealers and hunters) with recent data obtained by use of satellite telemetry and tracking techniques.

Address for correspondence: L.P. Folkow, Department of Arctic Biology, University of Tromsr N9037 Tromsr Norway.

194 Distribution and Migration Hooded seals aggregate to give birth and mate in relatively heavy pack ice near Jan Mayen (the West Ice), off Newfoundland (the Front) and in the Gulf of St. Lawrence (the Gulf) (Fig. 1). In addition, whelping also takes place in some areas in the Davis Strait [2]. The sizes of the different breeding stocks are not known, but based on current estimates of pup production rates it appears that the Northwest Atlantic stocks count about 400,000 animals [3], while the West Ice stock may count in the order of 250,000 animals. The segregation into different breeding areas has caused speculation as to whether different breeding stocks represent separate populations.

i' "=============================== F :'Q'''''''''''''"

.. ,,,,.. . . . . 9

.:...

GREENLAND

9

.

v...vv.v..:.v.3.-.-..

,,.

.......

,.,io 9

.--..

.L':,

9 9 9

.. . .'.i.332:..2.2"2-.'.""

". 9

o.

OQO 9

~l.'."

NORTH ATLANTIC OCEAN

~, ,

-~.-~- ~.

..~

,.3"

.

.'-."

".-.: . -

9.

"~'s

9

'

o,.."-. ,, 9 9 9 9 9 . . , , . . . o . , . . o o -o-

'. 9 9 - . " 9 9~ . . . . . , , . . o . . . . , , . . .

9

e

.r

QS

,

qp

""..

..~-~

...

Fig 1. Current knowledge on the distribution of hooded seals in the North Atlantic. Cross-hatched areas: breeding grounds; densely dotted areas: moulting grounds; lightly dotted area: main distribution area. Arrows indicate migration patterns from breeding grounds to moulting areas. The map was drawn based on published data [4,7] and results from a satellite tracking study by Folk 9 M=irtensson and Blix (submitted to Polar Biology).

195 Based on migration patterns, Rasmussen [4] concluded that there is only one stock of hooded seals in the North Atlantic, and this conclusion is supported by Wiig and Lie [5], who in comparing morphometric data from hooded seal skulls from the different breeding stocks, found no significant differences. Still, the two main breeding stocks appear to inhabit different geographical areas during most of the year. From an ecological and managemental point of view, the two stocks should therefore still be looked upon as separate, despite the fact that there is some exchange of genetic material between them. After breeding, hooded seals have been reported to start migrating towards their moulting areas. The main moulting area is located in the Denmark Strait, where animals from both the Northwest Atlantic and the West Ice stocks have been presumed to moult [4,6,7] (Fig. 1). That migration of hooded seals from Newfoundland to the Denmark Strait takes place has been confirmed by mark-recapture data [2,8], as well as in recent satellite tracking studies [9]. All stocks of hooded seals were for long considered to moult in the Denmark Strait. However, Nansen [6] described another lair which was located further north, between 72 and 76~ Folkow, Mhrtensson and Blix (submitted to Polar Biology) observed a lair somewhat further southeast to this (northwest of Jan Mayen), but concluded that this probably corresponded to that described by Nansen [6], and suggested that a large fraction of the West Ice stock is moulting there. Previous reviews of hooded seal migration patterns seem to imply that the seals start their migration towards the moulting lairs as soon as breeding is completed, and that they all travel in one general direction while feeding on their way [4]. However, data obtained by Folkow, Mhrtensson and Blix (submitted to Polar Biology) from eight satellite tagged hooded seals in the period between breeding and moult show that West Ice hooded seals leave the drift ice and distribute widely soon after breeding, to perform long excursions to distant waters (e.g. off the Faeroe Islands, off northern Ireland and in the Norwegian Sea) (Fig. 1), before heading towards the moulting lair northwest of Jan Mayen in the summer. Also in the Northwest Atlantic, satellite tracking studies show that excursions at high sea take place before the moult migration starts [9]. Data on the distribution of hooded seals after the moult off the east coast of Greenland are few, primarily because such studies have remained impossible until the development of satellite tracking techniques. The general opinion has been that the seals then distribute widely, the seals from the Northwest Atlantic either returning to waters off Newfoundland, or following the Greenland coast around Kap Farvel, and north as far as to the Thule district [10], while West Ice hooded seals have been assumed to mainly range north, in the ice between Greenland, Svalbard and Bear Island [4]. However, some seals have been documented to migrate over long distances to other areas, and records exist of observations or catches of hooded seals, from Alaska in the west [11], to the Barents Sea in the east [7], along the coast of Norway [7,12], south of Iceland and off the Faeroe Islands [13]. In fact, stranded specimens have been found on remote locations as far south as the Portuguese coast in the east [14] and the coast of Florida in the west [15], and a stranded female

196 hooded seal has even been found in the Pacific, as far south as the Califomian coast [ 16]. Such seals have usually been considered as "stragglers" and "strays". However, recent satellite tracking data have shown that long travels in the central and eastern North Atlantic are roles rather than exceptions for these mammals. Folkow, MArtensson and Blix (submitted to Polar Biology) have collected data from 15 subadult and adult hooded seals of both sexes which were equipped with satellite transmitters after the moult in lairs at about 71~ 12~ and which were tracked for about 200 days, on average. They found that all but one of the tagged seals on one or several occasions made an excursion which lasted for about 3-7 weeks to distant areas, and then returned to the drift ice edge, somewhere between 65 and 77~ along the east coast of Greenland. The most frequently visited area off the ice edge were the waters off the Faeroe Islands, where altogether eight of the seals stayed at some time during the tracking period. Other areas of importance were the Irminger Sea (southwest of Iceland), the continental shelf break between the Norwegian mainland and Bear Island, and areas in the Norwegian sea and north/northeast of Iceland (Fig. 1). One individual, which was a male, travelled to waters off the west and north coast of Svalbard (up to 81 ~ and remained there throughout the remainder of the tracking period. Apart from this, the study by Folkow, MArtensson and Blix (submitted to Polar Biology) did not reveal any differences in distribution pattems of males and females. Long excursions at high sea were conducted at all times of the year, but the seals always returned to the ice edge and were never recorded to haul out on any coast. In this sense, the hooded seal appears to be a truly pelagic species. At the time of breeding (second half of March), 7 of the 15 tags which were deployed in July were still active. Judging by the size of the animals, which was compared with length-age data presented by Rasmussen [4], one female may have been sexually mature, and this animal retumed to the breeding area southwest of Jan Mayen and spent 6 days hauled-out on the ice before returning to sea. Given the short nursing period of hooded seals [ 17], this animal may well have given birth and nursed a pup in this period.

Diving behaviour It is easier to summarize previous data on the dive behaviour of hooded seals than on their distribution. In 1890, Nansen [6] stated that hooded seals are deep divers, and Oynes [12] reported that hooded seals occasionally were caught in nets off the coast of Norway at depths of several hundred meters. The first actual recording of dive depths of hooded seals was made by Scholander [18] who found that captive young seals could dive down to about 75 m. Data on dive depths of free-swimming hooded seals did not exist until such information could be collected by satellite telemetry. The first confirmation of Nansen's [6] assumption was made by Folkow and Blix [19], who found that their satellite-tagged hooded seals could dive to depths beyond the depth range of the pressure sensor employed, which was 0-1,000 m. Dives as

197

JAN MAYEN DECEMBER 45

1268 dives (n=l)

40 35 30 25

g 2015 g 10

12,1

9,8 5,4

5

0,5

0 0-52

52-100

100-300

300-6(X)

600-968

>968

depth (m) JAN MAYEN DECEMBER 60

1131 dives (11=1)

50

52,9

29,9

qm

lO o

I

6,7

I, ,=

~,4 16,9

8,1

lO

2,1

1,8

600-968

>968

5 o

,= g 15 ~

13,5

10

I

6,8

2-5

I

5-15

15-25

I

4,6 25.52

I

0,06

duration (rain) Fig. 3. Overall dive depths and durations of 16 hooded seals, between July 1992 and July 1993 (from

manuscript by L.P. Folkow and A.S. Blix, submitted to Polar Biology). autumn (August-November), while there was a shift towards deeper dives (300600 m) in winter and spring (December-May). Areas in which particularly deep dives were performed were the waters off Jan Mayen (Fig. 2), and in the Irminger Sea, but even in these areas dives to between 300 and 600 m usually constituted the largest fractions. Comparison of dive depths with water depths in different areas showed that most dives must have been pelagic. Folkow and Blix (submitted to Polar Biology) also found that dive durations generally were well correlated with dive depths and a few dives of more than 52 min duration were recorded in areas where deep dives (>1,000 m) were performed. The average dive duration of hooded seals appears to be between 5 and 15 min (Fig. 3), and Folkow and Blix (submitted to Polar Biology) estimated the aerobic dive limit of

199 hooded seals to be about 15-17 min, which implies that as much as 36% of all dives by hooded seals are in part anaerobic. Dive depth data have also been collected, by use of the same technique, from Northwest Atlantic hooded seals during the period between breeding and moult. These studies showed that the depths of dives were less in this area (

B Gadidae

....

C.harengus ..

..

... ..

0%

Pelagic trawl

Bottom trawl

Seal Intestines

Fig. 3. Mean relative numerical frequency of various organisms in catches (standard 30 min pelagic and bottom trawl hauls) and diet (intestinal contents analysis) of harp seals caught in the area northeast of Cape Kanin in February 1993. From [9].

1994 [20]. Thus, increasing quantities of young herring (0-group and recruits up to 4 years old) are recorded in the southern Barents Sea [21 ], and immature herring are, at present, probably the major winter prey for harp seal in this area [9]. The decline in condition in spring and early summer (Fig. 2) suggests that food intake is decreased in this period. However, some feeding does occur among lactating female harp seals. In the White Sea breeding grounds in early March 1989 and 1993, the main food items were found to be crustaceans [8]. Comparable observations were made of lactating harp seals in the White Sea in 1962 [22]. Nocturnal

247

feeding, and subsequent rapid digestion [23], may account for the low frequency of food remains found in stomachs of lactating harp seal females shot during the day [24]. The localisation of breeding grounds is determined by the drift and distribution of ice, and this may vary both seasonally and from year to year. This may result in considerable variations in food availability for the female seals which spend a lot of time on the ice nursing their pups [15,25]. In his 1962 investigations in the White Sea, Timoshenko [22] found that if food was available, lactating harp seals fed quite intensively. In contrast to the situation in early March in the White Sea, seals examined in Varangerfjord in late March to early April 1992 had stomachs well filled with undigested prey, almost exclusively capelin (Fig. 4). These seals were mostly adult females, and some still had milk in their mammary glands. These females were probably on a westward feeding migration following the lactation period. Capelin (Mallotus villosus) has also been found to dominate the diet of harp seals in this area during late winter also in 1991 [8]. The increased dominance of capelin in the seal

100

NUMERICAL

BIOMASS

I

J

I I I I

v LU

r

zLU n." n'D

~

I

o o 0

0 >o z w D 0

J

J

PRAWNS

CAPELIN

OTHERS

I I

50

I I I

G: W >

I

_J W ,....o...... o o...

J I

VARANGERFJORD (N=35)

EAST ICE (N=6)

,:.:.:.:.:.:.:.:.:.:.

::::::::::::::::::::: VARANGERFJORD (N=35)

EAST ICE (N=6)

Fig. 4. Diet composition from stomach contents analyses of harp seals caught in Varangerfjord (MarchApril) and East Ice (April) 1992. Dietary composition is given, in terms of relative frequency of occurrence of each prey item as numerical frequency and as per cent total biomass. From [8].

248 diet during late winter in eastern Finnmark in recent years differs from observations made during the extensive harp seal invasions in 1986-1988. At that time the seal diet consisted mainly of prawns (Pandalus borealis) and codfish (Gadidae) [4,11]. Earlier, in 1978-1981, harp seals taken as by-catches in winter gill-net fisheries in Finnmark were reported to have eaten mainly capelin [26]. The recovery of the Barents Sea capelin stock in 1990, following the severe collapse in 1985/1986 [20], probably accounts for the increase in importance of capelin as prey for harp seals in Finnmark during the winters of 1991 and 1992. The Barents Sea capelin spawning grounds stretch from the White Sea to the North Norwegian coast (particularly Finnmark) and spawning takes place (March-April) when harp seals are in the Varangerfjord area [27-30]. There was a new collapse in the capelin stock during 1993 [20] and the stock is again at a low level. This, combined with the substantial increase of immature Norwegian spring spawning herring in the southern Barents Sea, has probably resulted in herring becoming the most important harp seal prey in this area during the winter. Restricted feeding by harp seals was observed during the moult. In the East Ice in April 1992 the seals had eaten mainly prawns and capelin (Fig. 4), but codfish, sculpins (Cottidae), snailfish (Liparidae) and flatfish (Pleuronectidae) also occurred in the diet [8]. In April 1993 in this area, harp seal stomachs well-filled with herring were observed (K.A. Fagerheim, Institute of Marine Research, Bergen, Norway, pers. comm.). During moult in the White Sea, the harp seal diet comprised amphipods (Parathemisto spp.), prawns, capelin, herring, eelpout (Zoarces viviparus), sandeels (Ammodytes spp.) and stickleback (Gasterosteus aculeatus) [8]. Results of the Barents and White Seas studies agree with previous observations of harp seals in the western North Atlantic. In late winter and early spring, food intake tends to be reduced [15,18]. The exception seems to be in the brief period between breeding and moult when increased condition of adult seals (Fig. 2) and the occurrence of well-filled stomachs in adult females indicates that feeding is more intense. This observation is supported by previous White Sea data which showed a slight increase in adult female blubber thickness in the time period between whelping and moult [ 17].

Early summer to autumn in the northern Barents Sea The Barents Sea harp seals are found in large numbers along the drifting pack-ice and in open waters in early summer [13,16,31,32]. In the course of a research survey conducted along the ice fringe from Novaja Zemlja to Hopen in June 1991 (Table 1), almost all harp seals were observed to be confined to the pack-ice southeast of Hopen [ 10]. The stomachs of seals captured from ice floes in this area (Fig. 1, area 4) were usually empty, with only a few fragments of prawns and otoliths of long rough dab (Hippoglossoides platessoides), sculpin (Triglops pingelii), capelin and polar cod being found. There was also little sign of harp seal faeces on the ice floes where the seals had hauled out. The seals were in poor condition (Fig. 2), suggesting that little feeding had taken place during the early summer period [ 10].

249 Pelagic and bottom trawling carried out concomitantly with the capture of the seals revealed little or no potential prey in the water column at shallow depths. Prawns, capelin, polar cod and other fish species were, however, abundant in bottom waters at depths of approximately 330 m [10]. These prey items were probably unavailable to the harp seals, due to the great depth of occurrence. The fact that adult harp seals improve in condition from June to September (Fig. 2) indicates that the summer and autumn are periods of intensive feeding. However, no data are available about harp seal feeding habits in the Barent Sea during the summer period (July-August). In autumn the harp seals disperse along almost the entire ice edge, from the east of Spitsbergen to the northeastern parts of the Kara Sea [13,16]. In September 1990 and 1991 harp seals were captured whilst in water close to the pack-ice edge (Fig. 1, area

100 i

.-o~176176

-'-'.'-','""-'.'-:-i .................. " .".'.". '.". .' ". .' " " " : ' i . . . . . . .

O-~

-

#~ vj

.

:::::::::::::

~:

:. : : : :. : : : :.: : : :.: : : :.: : : :.: .

0

-

--

>"

0

Z uJ (~

50

i

'-__

I!

w >

.

.

"-~ ..... ,..,,....,..,. .. .. .~. . . . . . . . .-...................

....................., ..................... - ..9.-......--. ,. .. ',.-'.. .-.. -' ..'-. .' .- . . . . . . . . . .. 2.. : ..: . :.. : ..: . : . : . : . : . ........,...,......, . .. .. .. .. .. .. .. .. .. .. . . . . . . . . . . -..,-.-...'.-.',','.', ,,.,....o.......,..,. .9. . . . 9, . , . , . , . . . . . ..........-....~ "-'-'-'.'-'.'-'-'-'-" :::::::::::::::::::::

.

.

iliiiii;ii!i)i;i!iiiii

. ~.

9

.

.

.

.

.

.

::::::::::::::::::::: :.:.:.i.:':-:.:.:.i.:

:::::::::::

''" ....

iiiiiiiiiiiiii!!i!iiii

:::::::::::::::::::::: :::::::::::::::::::::: - :........... .:-:.:.:.:-:-:.:.:.:

9 OoOe, 000 ~

......

:::::::::::::::::::::

~~J

_

::::::::::::::::::::::: .9, , . . . . , , . . , . , . . . , . . ~ ........... .,.......,..,.,...,.t :-:.:-.-.-.-:-:.:.:.:~ ,....................., '...-.'.'.'.'.'.'.'.-.' . . . . .................. .

:::::::::::::

;!:i:i:i:i:i:i:i:i:i:i i!ii?iiiiiiiii!ii!iiii!i ii!!!!iiiiiiii!!iiiiii! ,......-..,,

.

[

~

.,.,....

.

:::::::::::::::::::::

'

::::::::::::::::::::: iii?!i!iii!!ii!!ii!!! .,........,....,..-..

September 1990 fi!ii!i!ili!iiiii;ili!!

:i:i:i:i:!:!:i:i:i:i: i:!:!:!:!:i:i:i:i:i:i . .. .. .. .. .. .. .. .. .. .. i:i:i:i:i:i:!:i:i:i:i:l ::::::::::::::

9

~

w

:::::::::::::::::::::

.":" . . .:". . :"- . :" . . -:". .:'. . :". . .:". .:". :" ". .. .' .. '. .. '.,.'...' .. .' .. '...'...'...". .

9

~ .........

'~

.

'

-:-:-:-:-:-:-:.:.:.:.: ,...-.-.-.-...-.-.-.-, ,...-.....................%....,....

~

.

:::::::::::::::::::::: ::::::::::::::::::::::

i:i:!:i:i:!:!:i:i:!:i: _

.

'~176176176 boeoooee

:::::::::::::::::::::: . . .. .. .. .. ,. .. . . . . . . . . . . . , . ..%

w

I"1"

.

.: .: .: .: .: .: .: .: .: .: .: : : : : : : : : : :

.. " ' " ' " '9." o' ".'o" '.". ". ". . . . . . . ,. -. .. .' .. -. .. .' .. '.,.'...' .. .-...' .. .' .. '. .

Z

.

.: .: .: .: .: -: :. :. :. :. :. :, :. :. :. :, :. :- :. :-:.:.

ldd-""

.--I

.

o.-.-...,.....-...-.', "'""""'"'"'"'" -.....-.,,.....,-.-,-. "- .' .".' -".' -".' -"."-'."' .' "' .'' . ' . ' . ...................

.:.:.:.:.:.:.:.:.:.:.: : :.:.:.:.:.:.:.:.:.: : : : : : : : : : : : . 9.............o . . . . . ...., . . . . . ...,.....-.-......,... ...o..... 9... 9 ::::::::::::::::::::::

-

(3 z

Fig. 4. Variation in the relative importance of the species of fish found in the faeces of harbour seals during the period of sampling at Hvaler in 1991. were significantly different between the two months, when considering both frequency of occurrence and number of each species (chi-square test, P < 0.05, df = 6). Significant positive correlation was found between order of ranking of species in the trawl and in faeces in September 1990 (Spearman's rank test, P < 0.05). Ranking was made on the basis of number of individuals. The four species, Norway pout, blue whiting (Micromesistius poutassou), haddock and whiting, made up 96% of all the individuals in the trawl in 1990 and they represented 93% of the individuals found in faeces the same year (Fig. 5). In 1991 the composition of species varied between the months (Fig. 4). Herring was the only species occurring in all the months of sampling. The results indicate variation in the relative importance of the species, even though Norway pout and herring were dominating throughout the sampling period. Significant positive correlation was not found between order of ranking of species in the trawl and in faeces (Spearman's rank test, P < 0.05). However, Norway pout was the dominating species both in faeces and in the trawl. The five most numerous species found in the trawl (Norway pout, whiting, haddock, hake Merluccius merluccius and poor cod) made up 65% of all fish caught in total in 1990 and they represented 56% of the number of fish in the faeces (Fig. 5). In faeces there was a large representation of especially sandeel, but also herring and sprat (Sprattus sprattus). None of these species were found in the trawl. All the species found in faeces at Hvaler in 1990 were also found in 1991 (Table 3). Of the five most important species occurring in both 1990 and 1991 at Hvaler (Norway pout, saithe, sandeel, herring and cod), significant differences were found in the distribution from one year to another (chi-square test, P < 0.05, df = 4). The

280 Faeccs 1990

Trawl 1990 " 0

240

1,0~, 5,9 2,7 2,5 2,3 1,7

7,0

iiii!iiiiiiiiiiiiii!iiiiiii!iiiiiiiiiiiii!ii!iiiiiiii!i!iiiiii ! ii!iiii

[~

norway pout whiting

8

poor cod

D

haddock

k~ hake Facces 1991

Trawl 1991

8,4

8,0 2

8,0

1

~]

cod

~]

plaice

~1

other species

I~

'11,83

sandeel herring

,::iiiiiiiiiii!iiiiFi!iiiiiii!ii :i,o.o, i:::i:, 5,6 5,6

kg;'d ['~

sprat

[]

blue whiting

saithe

24,0 52,7

11,0

Fig. 5. Relative importance of species of fish found in the faeces of harbour seals and in the trawl at Hvaler in 1990 and in 1991.

same 7-8 species were dominating in the diet both years (Fig. 3). However, the ranking within these species varied between the years. In May 1990 there was a relative dominance of Gadidae, Clupeidae and Ammodytidae, and in September a relative decrease in the number of Clupeidae and Ammodytidae, relative to the Gadidae. In contrast to 1990 the occurrence of Clupeidae did not seem to decrease in the autumn of 1991. However, Ammodytidae seemed to follow the same pattern both years; decreasing in September of 1990 and not occurring after August of 1991. Significant difference was found when comparing the relative importance of species at Hvaler and at Froan in 1991 (chi-square test, P < 0.05, df = 4). Comparison was based on the number of individuals of the species Norway pout, saithe, dab (Limanda limanda), herring and poor cod, common in the seals' diet at both sites (Fig. 3). Differences were found both when comparing all the samples of 1991 and when comparing the samples of July 1991 only.

281 Discussion and Conclusions

Using analysis of faeces is based on the assumption that the samples found are representative for the population of seals studied. It should be noted that in months like September and October of 1991, when the number of samples were only two and three, respectively, the total number of species occurring was lower than in months with larger sample size. On the basis of the observed rate at which number of species increases when sample size increases, two and three samples may not give a representative description of the diet. Several studies have shown that small fragile otoliths, especially of clupeids, are more vulnerable to digestion than otoliths from other species, and may therefore be under-represented in the faeces of seals [4,28-31]. Even though an underrepresentation of clupeids cannot be neglected in this study, especially herring were still considered one of the most important species in the diet of harbour seals, indicating that the bias may not be great enough to influence relative importance of a species, not considering mass. Variation in the diet between months

Analyses of faeces from Hvaler in both 1990 and 1991 indicated significant seasonal changes in the diet of harbour seals. This is also shown in other studies [1,6]. H~irkiSnen [4] also found variation within the year at Koster, Sweden, which is an area close to and similar to Hvaler. The variations in relative importance of the prey species is probably due to the seals' feeding strategy. Abundance of prey may change through the year and cause variation in the diet. Olesiuk [ 10] found a change of dominating species of fish in the diet of harbour seals, correlated to the species spawning migration. H~irktinen [4] found that fluctuations in the diet often were correlated to migrations or changed behaviour during periods of spawning. Also some species were found to occur at high frequency when preferred prey were unavailable [36]. The trawl catches gave an indication that the seals feed among the most available species, but also that there are species that the seals avoid or that avoid being a prey. However, species not found in faeces certain periods of the year were without exception found in the trawl at the same time of the year. For example haddock were not found in faeces at Hvaler later than July in 1991, but did occur in the trawl in November-December. Trawling as a method for mapping the species distribution has its limitations, being a selective way of sampling. The depth and area chosen for trawling may eliminate some, especially pelagic, species. This is shown by the fact that neither herring, sprat nor sandeel were represented in the trawl at any time in this study. In August of 1991 there was a marked change in the diet of the seals at Hvaler. The seals behave differently due to age, sex and time of year. During August the harbour seals at Hvaler terminate their pupping and suckling time, which implies that the pups must start feeding, first together with their mother and then on their own. This could explain the change in the diet of the population seen in August of 1991.

282 The mother and pup may feed in other habitats than single grown up seals. The presence of hake, flounder, plaice, dab and goby may indicate feeding in more shallow waters than earlier in the year, reflecting the newly weaned pups contribution to the samples of faeces. In addition to feeding in more shallow water, the pups may start feeding on more easily accessible species and on a wider range of species than older seals, since they probably have not yet fully developed their feeding strategy. In general the harbour seals were found to feed on fish smaller than 30 cm. This may be energetically beneficial. There is, however, a chance that erosion of otoliths during passage through the stomach and intestines [28,32,35] and/or seals decapitating larger fish before swallowing them [34,35], may lead to underestimation of the size of the fish found in faeces.

Difference between years Variation in sample size and sampling period may be an explanation of the variation seen in both number of species and the species' relative importance in faeces when comparing 1990 and 1991. Differences in availability of species between the two years were difficult to investigate on the basis of the trawl catches because trawling was not carded out at the same time of the year. However, all the most important species in the diet were abundant in the trawl of both years, even though some of them were found in faeces only one of the years. Recruitment of fish larvae may show great variation between years. Due to this, certain year classes, and therefore possibly preferred sizes of a particular species, may be less available one year than another. At Koster, Harktinen [4] found little variation from one year to another, but reported a change from the period 1977-1979 to 1989 [37]. Also Rae [1] reported a change in the diet of harbour seals with time.

Regional variation in the diet Few species were found in the samples from Froan compared to the samples from Hvaler. This is probably due to a lower sample size, and may explain some of the variation in diet found between the two sites. Regional variation in the diet has been found in other studies of the diet of harbour seals [2], especially where the seals feed in different habitats [3,5]. The results of this study give no reason to state that Froan and Hvaler represent two different habitats, but the occurrence of greater argentine (Argentina silus) in the samples from Froan may indicate that the seals at Froan feed deeper than the seals at Hvaler. Greater argentine live mainly in deep waters at 200600 m [26]. Since trawling was not carded out at Froan it was not possible to investigate the abundance of species in the area, or use the abundance as an explanation for why some species widely distributed and of relative importance at Hvaler, such as cod, blue whiting and sandeel, were not present in the faeces at Froan. Another factor not studied but possible as an explanation for minor diet variations is competition with other piscivores. At Froan there is a colony of grey seals (Halichoerus grypus), not found at Hvaler. Their diet has not been investigated and possible competition between the two species is unknown.

283 Conclusion

This study indicates that the harbour seals mainly prey on benthic schooling species of fish. This is consistent with other studies [3,6,7,12,]. Few species per sample and clear dominance of one or two species throughout the year supports this view. Feeding on schooling fish, close to the bottom to limit the escape possibilities of the prey and in relatively shallow waters to minimise the energetic costs of diving, could be an energetically beneficial strategy of feeding. A relatively small range of species common through the year also supports the theory of harbour seals avoiding some species, as suggested by H~irktinen [4]. Being partly selective, however, does not exclude having an opportunistic strategy of feeding. The variation seen between months in the same year, from one year to another and between sites, does indicate an opportunistic strategy of feeding dependent on the abundance of prey. This was to a certain degree supported by the trawl catches, where the ranking of species in 1990 was correlated to the ranking in faeces the same year, and the most important species in faeces were the most important species in the trawl catches, especially in 1990 but also in 1991. In conclusion, the harbour seals seem to feed opportunistically on some families but not on all species of fish, and the relative importance of the preferred prey species seems to be determined by the abundance at the time, and possibly also to a certain degree by the state of the seals at the time.

Acknowledgements This study was supported by the Norwegian Research Council. Special thanks to John Prime and Randi Roen for helping out with the field work and to John Prime for support on identification of otoliths. Also thanks to Karin Andersen, University of Oslo, for sharing the trawl data.

References 1. Rae BB. Further observations on the food of seals. J Zool 1973;169:287-297. 2. Cubbage J, Calambokidis J, Carter S. Fish otoliths recovered from scat of harbour seals in the inland waters of Washington State. Third Biennal Conf of the Biol of Mar Mammal, 1979. 3. Brown RF, Mate BR. Abundance, movements and feeding habits of harbor seals, Phoca vitulina, at Netarts and Tillamook Bays, Oregon. Fish Bull 1983;81(2):291-301. 4. H~irk/Snen T. Seasonal and regional variations in the feeding habits of the harbour seal, Phoca vitulina, in the Skagerrak and the Kattegat. J Zool 1987;213:535-543. 5. Payne PM, Seizer LA. The distribution, abundance and selected prey of the harbour seal, Phoca vitulina concolor, in southern New England. Mar Mammal Sci 1989;5(2):173-192. 6. Pierce GJ, Boyle PR, Thompson PM. Diet selection by seals. In: Barnes M, Gibson RN (eds) Trophic Relationships in the Marine Environment. Proc 24th Eur Mar Biol Symp, 1990. 7. Pitcher KW. Food of the harbour seal, Phoca vitulina, in the Gulf of Alaska. Fish Bull 1980;78(2) :544-549.

284 8. Behrends G. Analysis of stomach and colon contents of 185 common seals from the Waddensea of Schleswig-Holstein. ICES CM 1982/N:11 Mar Mammal Comm. 9. Boulva J, McLaren IA. Biology of the harbor seal, Phoca vitulina, in Eastern Canada. Bull Fish Res Bd Can 1979;200. 10. Olesiuk PF, Bigg MA, Ellis GM, Crockford SJ, Wigen RJ. An assessment of the feeding habits of harbour seals (Phoca vitulina) in the Strait of Georgia, British Columbia, based on scat analysis. Can Tech Rep Fish Aq Sci 1990; 1730. 11. Frost KJ, Lowry LF. Sizes of walleye pollock, Theragra chalcogramma, consumed by marine mammals in the Bering sea. Fish Bull 1986;84:192-197. 12. Gortsev VN. Feeding of the harbour seal. Ekologiya 1971;2:62-70. 13. Havinga B. Der Seehund (Phoca vitulina L.) in den Hollaendischen Gewassern. Tijdschr Ned Diersk Vereen 1933;3:79-111. 14. Spalding DJ. Comparative feeding habits of the fur seal, sea lion and harbour seal on the British Columbia coast. Bull Fish Res Bd Can 1964;146. 15. Everitt RD, Gearin PJ, Skidmore JS, Delong RL. Prey items of harbor seals and California sea lions in Puget Sound, Washington. The Murrelet 1981 ;Winter:83-86. 16. Bailey KM, Ainley DG. The dynamics of california sea lion predation on pacific hake. Fish Res 1982;1:163-176. 17. Prime JH, Hammond PS. Quantitative assessment of grey seal diet from faecal analysis. In: Huntley AC, Costa DP, Worthy GAJ, Castellini MA (eds) Approaches to Marine Mammal Energetics 1987;165-181. 18. Prime JH, Hammond PS. The diet of grey seals from the south-western north sea assessed from analysis of hard parts found in faeces. J Appl Ecol 1990;27:435-447. 19. Hammond PS, Prime JH. The diet of British grey seals, Halicoerus grypus. In Bowen WD (ed) Population Biology of Sealworm (Pseudoterranova decipiens) in Relation to its Intermediate and Seal Hosts. Can Bull Fish Aquat Sci 1990;222. 20. Daneri GA, Coria NR. The diet of Antarctic fur seal, Arctocephalus gazella, during the summerautumn period at Mossman Peninsula, Laurie Island (South Orkneys). Polar Biol 1992;11:565566. 21. BjCrge A. Status of the harbour seal Phoca vitulina L. in Norway. Biol. Conserv 1991;58:229238. 22. BjCrge A, Fagerheim KA, M~rkved B. Telling av steinkobbe ved Hvaler i 1983. Fisken Hav 1983;3:1-4. 23. Markussen NH. Apparent decline in the harbour seal Phoca vitulina population near Hvaler, Norway, following an epizootic. Ecography 1992;15:111-113. 24. BjCrge A. Status of marine mammal habitat protection in Norway. ICES CM 1986/N:4 Mar Mammal Comm. 25. H~irk6nenT. Guide to the otoliths of the bony fishes of the northeast Atlantic. Hellerup, Denmark: Danbiu Aps, 1986. 26. Pethon P. Aschehougs store fiskebok, Oslo: Aschehoug, 1989. 27. Zar JH. Biostatistical Analysis. Engelwood Cliffs, NJ: Prentice Hall, 1974. 28. DaSilva J, Neilson JD. Limitations of using otoliths recovered in scats to estimate prey consumption in seals. Can J Fish Aquat Sci 1985;42:1439-1442. 29. Murie DJ, Lavigne DM. Digestion and retention of Atlantic herring otoliths in the stomachs of grey seals. In: Beddington JR, Beverton RJH, Lavigne PM (eds) Marine Mammals and Fisheries. London: George Allen and Unwin, 1985;292-299. 30. Jobling M, Breiby A. The use and abuse of fish otoliths in studies of feeding habits of marine piscivores. Sarsia 1986;71:265-274. 31. Jobling M. Marine mammal faeces samples as indicators of prey importance - a source of error in bioenergetic studies. Sarsia 1987;72:255-260.

285 32. Dellinger T, Trillmich F. Estimating diet composition from scat analysis in otariid seals (Otariidae): is it reliable? Can J Zool 1988;66:1865-1870. 33. Harvey JT. Assessment of errors associated with harbour seals (Phoca vitulina) fecal sampling. J Zool 1989;219(1):101-112. 34. Boulva J. The biology of the harbour seal in eastern Canada. Ph.D. Thesis, Dalhousie University, Halifax, 1973;153p. 35. Pitcher KW. Stomach contents and faeces as indicators of harbour seal, Phoca vitulina, foods in the Gulf of Alaska. Fish Bull 1980;78(3):797-798. 36. H~irk/Snen T. Food-habitat relationship of harbour seals and black cormorants in Skagerrak and Kattegat. J Zool (London) 1988;214:673-681. 37. H~irk6nen T, Heide-JCrgensen MP. The harbour seal Phoca vitulina as a predator in the Skagerrak. Ophelia 1991 ;34(3):191-207.


9 1995 ElsevierScience B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 13. UUtang,editors

287

Feeding ecology of harp and hooded seals in the Davis StraitBaffin Bay region Finn O. Kapel Greenland Fisheries Research Institute, Copenhagen, Denmark A b s t r a c t . Results of stomach contents analyses of material collected in West Greenland waters in the

period 1986--1993 are reviewed, and compared with published data and circumstantial information from local hunters. The diet of harp seals feeding in this region is variable but consists mainly of pelagic crustaceans (euphausids and amphipods) and small fish species such as capelin, sandeel, polar cod and Arctic cod. Species of importance for commercial fisheries in Greenland, such as Northern prawn, Atlantic cod, and Greenland halibut play a minor role in the diet of harp seals in this area. Variation in the diet of hooded seals is less well documented, but in addition to the species also taken by harp seals, larger demersal fishes such as Greenland halibut, redfish, cod, and wolffish are apparently important prey items. Published data on harp seal feeding in the Canadian Arctic are briefly reviewed; they indicate a variation in food composition between seasons and areas similar to what was found in Greenland waters. Detailed information on hooded seal diet in the Canadian Arctic was not found. Information on population sizes, distribution and abundance of harp and hooded seals is reviewed. On this basis it is suggested that it is possible to develop density indices by area and season for harp and hooded seal in coastal waters of Greenland, and to use such indices combined with feeding data to estimate food consumption in this region. Similar indices may be developed for the Canadian Arctic, and for offshore areas, to arrive at total consumption by the seal stocks during their migrations and stay in the Davis Strait- Baffin Bay region. Key words: Phoca groenlandica, Cystophora cristata, Greenland, Arctic Canada, food composition, distribution, density indices

Introduction The purpose of this presentation is to review available information on the feeding habits of harp and hooded seals in the Northwest Atlantic during their migrations and stay in northern regions, i.e. outside the whelping and moulting seasons. Information on population sizes, distribution and abundance is reviewed in order to investigate ways of combining data on variation in diet and in abundance with estimates of total food consumption by the seals in these northern regions.

Food Composition Harp seal feeding in Greenland Information from hunters on the food of harp seals in Greenland was collected in the 1970s and presented in two meeting documents [ 1,2]. Address for correspondence: Greenland Fisheries Research Institute, Tagensvej 135 l, DK-2200 Copenhagen N, Denmark.

288

Table I .

Harp seal food composition (W%) in West Greenland

Area and month

N

MALL

MOAC

ARB0

FISH

PRAW

EUPH

PARA

CRUS

CEPH

Southwest (s) M Southwest (s) J Southwest (n) M Southwest (n) J Southwest (n) E Southwest (n) W Central W.(w) J Central W.(w) E Central W.(w) W Central W.(e) A Central W.(e) S Uummannaq A Upernavik (s) J Upernavik (s) S Upernavik (n) A Upv.(n) and Thu S Notes: Food items: MALL, capelin; MOAC, Gadus sp.; ARBO, AtctogaduslBoreogadus; FISH, other fish species; PRAW, prawns; EUPH, euphausids; PARA, Parathemisto; CRUS, other crustaceans; CEPH,cephalopods. Areas: (s) southern, (n) northern, (w) western, (e) eastern part. m =month: M, May; J, June(-July); A, August; S, September-October; E, autumn; W, winter.

289 Between 1985 and 1993, harp seal stomachs were collected, and analyses of the contents of 1,172 stomachs were reported [3-5]. Considerable geographical and seasonal variation was demonstrated, but some general patterns in feeding habits emerged. These are illustrated in Table 1 and Fig. 1, and can be summarized as follows. In the coastal waters of Southwest Greenland two prey items dominate the diet of harp seals: capelin (Mallotus villosus) and "krill" (euphausids). In the early summer, May, krill appears to be the predominant food; later, in June and in the autumn months, capelin constitutes the major part of the food. Codfish (in this region mostly Gadus morhua and G. ogac), other fish species, and prawns (particularly Pandalus borealis) are also taken but altogether they account for less than 20% of the diet. In the winter months, however, these "secondary prey species" appear to contribute to more than one-third of the food biomass, yet still surpassed by capelin and krill. In the western part of Central West Greenland, i.e. the region around the southem entrance to Disko Bay, the pattem resembles that found in Southwest Greenland: in June-July pelagic crustaceans dominate the diet: euphausids as well as amphipods, Parathemisto sp., followed by capelin, and prawns. In the autumn, however, capelin is the completely predominant prey item, whereas the winter diet is composed of a number of other fish species, pelagic crustaceans, and squid (mainly Gonatus fabri-

cii). Few samples were obtained from offshore waters in Southwest and Central West Greenland. They show variation between areas and month, with sandeel (Ammodytes sp.) as a very important food item, supplemented by Parathemisto, Pandalus, redfish (Sebastes sp.), and squid [5]. In the northeastern part of Disko Bay ("Vaigat"), polar cod (Boreogadus saida) was the predominant prey in August, followed by capelin and krill. In SeptemberNovember, capelin took over the "leading role" followed by polar cod, some other fish species and squid, but few crustaceans. In most of the samples from Northwest Greenland, polar cod, and to some extent Arctic cod (Arctogadus glacialis), was a significant or dominant part of the food of harp seals. In the Uummannaq district, however, krill was the predominant prey in August, and in the southern part of Upemavik district, capelin constituted more than half of the calculated biomass of the food in September.

Harp seal feeding in Arctic Canada Published information on the feeding of harp seals during their stay in Arctic Canada, i.e. away from the whelping and moulting regions, is relatively sparse. Foy et al. [6] found that immature and adult harp seal feeding in bays on the north coast of Labrador in late May-June were taking mainly capelin, whereas the diet of juveniles feeding near the offshore archipelago consisted mainly of euphausids. Immatures and adults feeding further offshore had a more varied diet including both fish, euphausids, and bottom-living decapods. In late autumn to early winter (November-January) harp seals feeding in the bays were feeding on a variety of

290

!ii~i i:~:i

iiii: ! ,,~!i~: ~~ii~i ~!i! !iiii: ~

I

CEPH

iii!!:: :i!i!i!: iilii::.......___ .......,_:. .........,. .......i',i',i........;_, ...........i! ::: ::;

::::: :

: : : ::

: : : : : :i

:: ::: :

: : : :.

i:!:!:!

70

:i:!:;

"--,~

""%"i

: : : :::

,~.~.1

i:!:! ....... E - E ~ ........ k"--." ........ .:~,}~:~ ....... !:i:;: ........ - - _ ' ~ ........

~.~.."~'1 ,~-.-'~ ::~:!..'!i~;:i~.s~:: : : : : : :

i!iiii:::.i ::':":'"

::::::::':

:::::':'

= . . . . . . .

~:'~':~:

50

................ ~,-, i!ii~ ~ i. ~i ! ....... i ~:-_--~ l

40

- 9: : : : : : ........ ,'--~ ........ : i : i : ...... g-~-"=, ....... , ' - - " ~ . . . . . . . . . ~

30

........ i._--~ .......

.......i-_'---~ ........

I~::~--~I.......~_~-~ ....... :_-:::.~ E-~

20

F::ra........ g--.~ E-T.-_-q E-7.'~

SWsM

........

SWsJ

:.--::~ FS-: ~-T.-~ Er

SWnM

100-

SWnd ..

SWnE

MOAC

MALL

...... 9 !"--'~ ........

........

:_-.=E-: a.

~7~

~i.~i~,o~:~i ~4~

"--'~

~-

PRAW

FISH

[-"1 ,.___:~.~:.._.-'~:~' .......

!--~: ........~-f_--~

~_:-'J!E:E:~E:I E:E- E

10

::

1

~..~..~

..i~::ii~. ...... ~i::i~; ....... !~::~:: .......~:-_:z~ ....... k_-~ ........[:::!:i~.::ii~;;i!....... !i ::~i~i~ ...............

60

EUPH

........ ~',~

~-~-~ ....... if:-~. ........ ~,:~

E::-::,~-E~ -::_~ii~Nil IFS.~i :f-5-~......... i~l E-'-.-~ E-:-:~, :-:~a ~ z ~

~.:5~

SWnW

CWwJ

CWwE

CWwW

,

I CEPH

90-

EUPH

80-

PRAW 70FISH 60ARBO

NN

50-

40 .

.

.

.

E-:-~ .............. ".-2-_'-~ ............. E--_a.

30-

=-=

MALL

.............I."~"~..............

.

---

i!ii!ii!i',ii::

tm. i m l

::::_-:~ .............

1 0 - .................................. iE--"-3 . . . . . . . . . . . . . . . . .

I 0

, CWeA

~ZZ-'-Z--~ CWeS

, UMQA

UPRJ

UPRS

NUSA

NUTS

Fig. 1. Food composition (weight%) of harp seals in Southwest (upper part) and Northwest Greenland (lower part). Food items: M A L L , capelin; M O A C , Atlantic cod or Greenland cod; ARBO, polar cod or Arctic cod; FISH, other fish species; P R A W , prawns; EUPH, euphausids or amphipods; CEPH, squid. Area: SW, Southwest Greenland; CW, Central West Greenland; s, n, w, e, southern, northern, western and eastern part, respectively; UMQ, U u m m a n n a q ; UPR, Upernavik (southern part); NUS, northern part of Upernavik; NUT, same (NUS) and Thule. Month: M, May; J, June(-July); A, August; S, September(-October); E, autumn; W, winter.

291 small Gadidae, polar cod, and capelin, whereas invertebrates were of secondary importance. Smith et al. [7] working in southeastern Baffin Island (63-64~ in JulySeptember found that the diet of four young-of-the-year harp seals consisted of 64% pelagic crustaceans (mysids and the amphipod Parathemisto), and 36% fish (mainly polar cod). One adult female harp seal was feeding on the pelagic shrimp Sergestes

arcticus. In the western Hudson Strait, Bech et al. [8] found that the stomachs of 14 harp seals caught in September-October near the south coast were dominated by capelin, with polar cod, sculpin, Greenland cod, and flatfish as secondary elements. One seal feeding offshore, near Salisbury Island, had taken Parathemisto and polar cod in almost equal amounts Sergeant [9] presented stomach contents data on 16 harp seals caught in the Canadian Arctic (at four different localities). Polar cod (Boreogadus saida) occurred in 6 of these, mysids in 5, Parathemisto in 4, and euphausids in 3 stomachs. In the High Arctic, Finley et al. [10] studied the feeding ecology of harp seals at Pond Inlet (73~ and Grise Fiord (76~ between mid-August and early October. Detailed analysis of 63 stomachs revealed that polar cod (Boreogadus saida) occurred in all of them, and accounted for 84% of all food items found in the stomachs (% frequency). The related species Arctic cod (Arctogadus glacialis) was found in 63% of the stomachs, but the number was much lower (5% frequency). In terms of biomass, however, polar cod and Arctic cod contributed 66% and 33% of the weight, respectively (as the individuals of the latter species were considerably larger). Other fish species or invertebrates were also found, but accounted for only a minor part of the food (altogether 16% frequency, and about 1% by weight).

Hooded seal feeding in Greenland In Greenland much less information is available on the feeding of the hooded seal than for the harp seal. Accounts in the literature and information gathered from hunters were reviewed in a previous paper by the present author [ 11 ]. When sampling harp seal stomachs, a few hooded seal stomachs were also obtained, and preliminary results of examination of the contents are presented here (Table 2, Fig. 2). From spring hunting in South Greenland, hunters provided information on the contents of 1,236 hooded seals stomachs, 386 (31.2%) of which were empty. Of the 850 stomachs with contents, 828 (97.4%) contained fish, whereas crustaceans and squid were only reported in 16 and 6 stomachs, less than 2 and 1%, respectively. For many stomachs, the contents were only given as "fish", but the fish species reported most frequently were redfish (Sebastes sp.), cod or Greenland cod (Gadus morhua, G. ogac), and capelin (Mallotus villosus). Six stomachs of hooded seals caught in Southwest Greenland have been examined in the laboratory. One was empty, and in the remaining five capelin was the predominant food. Four of the seals were taken in the spring, one in June, and one in

292

Table 2. Hooded seal food com~ositionin Greenland: information from hunters (8occurrence)

Species

South Greenland No.

-

Northwest Greenland % Food

NO.

% A11

1.6 0.2

4

0.2

55

1.3

1.9 0.7

% All

Southeast Greenland % Food

No.

59

0.6 9.0 9.6

0.9 11.9 12.8

-

1 46 1 153 614

0.2 75.1 24.9 100

0.2 100

30

% All

% Food

-

Capelin Codfishes Grl. halibut Redfish Wolffish Other fish Unspecified fish Fish total Decapods Other crustaceans Crustaceans totaI Cephalopods Stomachs with food Empty stomachs Total

14 2 16

6 850 386 1236

1.1

0.5 68.8 31.2 100

100

aThe percentages of the various fish species are adjusted for "unspecified fish".

206 236

12.7 87.3 100

-

100

Table 2 (continued).Hooded seal food composition in Greenland: ureliminary analyses of stomach contents (~01%)

Southwest Greenland (n = 6)

Capelin Polar/Arctic cod Grl. halibut Redfish Wolffish Other fish Pandalus Other crustaceans Cephalopods Unidentifiable Stomachs with food Empty stomachs

% All

% Food

82.7 0.2 0.3 0.2 -

99.2

-

83.3 16.7

-

0.2 0.4 -

0.2

-

100

-

CW Grl. SAQ(1)

Northwest Greenland UMQ(3)

UPV(5) % All

57 23 18 1 1

-

100

64.3 22.7 11.3 0.7 -

1.O

Southeast Greenland ( n = 29)

24.1 52.0 3.4

-

-

0.4

-

-

100 -

80.0 20.0

%Food 30.2 65.1 -

4.3 0.5 100

%All

-

1.9 2.4 1.7 15.5 2.1 42.0 3.5 69.0 31.0

% Food

2.8 3.5 2.5 22.4 3.0 60.9 5.0 100

293

294

I--]

EMPT

90

CEPH 80 CRUS

UNID UNSP 50

FISP

40 30 20

SGRL

NWGR

$EGR

SWGR

UPV

SEGR

INVE

: ~ ....................................

90

~:::::::,x~.~

FISH

80 ._L.t )"-K'~

i !

i

............................= - ' ~ ....... ~' r =--'.,.~ r r

....... . ~ - ~ - ~ ) ........ i

~

] ............................

....... r

~ .......

r

.......

....... ~/2,~ .................................................. ~'C/,~

50

o

........::

I:

.............................. ,',,,5"~'N~....... ....

10 ........................"='~" "='-'=~ 9 .........~'~X ~

o

="-~

"---= ---' "-

"-"-:

SGRL

:-,:==,:.~,

. . . . . . .

~xx>

=,,-_--,,,.-

.-,.--.,.,'-~

-*=-'=~ ".=-.~

....

"-'-~

....

---'-'

- - -

~,.-,,.~

"-"

.. . ... . . .. . . . . . .

NWGR

F-:-"

SEGR

SWGR

SEBA

P,:'/,~ r

NNt REIN

r

r

r

~.x,

, ~, ,~'c"~

,

GADO

MALL

~O ,:' ~", (.1 I:: (1)

:1"

O" G) L. I.I.

I

""

I

9149

K

(n

I

[

i

b

L_ (.1

_=

I II L_

I

-

~ L.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

II_

K Population size

Fig. 2. Rate of increase in relation to population size as shown for species with differing life-history strategy and levels of population variability. (a) The rate of increase for K-selected species in relation to population size resulting from little population variability. (b) The rate of increase for r-selected species in relation to population size resulting from extensive population variability.

less limits to their rates of increase. The solid curved line in Fig. 2b shows how natural selection allows for greater density-dependent elevation in the rate of increase than is possible for species of the type shown in Fig. 2a. Thus, evolution results in a pattern relating the shape of the density-dependence curve (and the equivalent position of the inflection point in population growth curves) to life-history strategy. The effects of natural selection as described above occur on a time scale linked to generation time. The resulting pattern is one in which the curvature of the lines of Figs. 2a,b is correlated with the rate of increase per generation [7,8]. This curvature is an evolved species-level trait. Figure 3, then, is a representation of the evolutionary pattern achieved by species in evolving various lifehistory traits that are adaptations to their respective environments. Species at the lower right of the relationship in Fig. 3 are r-selected species with high population variability and density-dependent rates of increase shown in Fig. 2b [7]. Species at the upper left are K-selected species with lower population variability and densitydependent rates of increase shown in Fig. 2a. Since the life-histories themselves have

408 1.0

Elephant ~ e

9Sheep

0.8

9Fin whale

Stenella

Fur seal

Eschen'chia~~

.......

Haramec/um

0.6

9

............... Grey Whale "

9Grizzly

bear Bighorn sheep " ~ 9Blue whale Deer

~

~ - ~

9

~ Mouse

_ 9",,,,~Striped bass

0.4

Dros

"op~

0.2

0 !

-2

,

!

-1

Fig.3.

,

!

0

i

I

1

,

I

2

In (rT)

The relationship between the position of the inflection point in population growth curves as a function of the rate of increase per generation, from [7]. The position of the inflection point is equivalent to the peak of population productivity curves expressed as a fraction (R) of equilibrium population levels (or mean naturally occurring population levels). some genetic basis, the pattern is largely determined by natural selection. Variability within the pattern (i.e. deviation from the line) represents more short-term effects of environmental factors, incomplete progress in achieving the evolutionarily stable strategy, and measurement error.

T h e Effects of H a r v e s t i n g

Human predation, like other forms of predation, presents harvested species with selective pressures to which they respond through genetic change. Documentation of the genetic effects of harvesting has been accumulating since at least 1957 [15] and is now of concern in fisheries management [16]. Of considerable importance in the present context are the effects of harvesting on life-history strategies [17]. Components of life history that have responded to harvesting are age at first reproduction, reproductive rate, and growth rate [ 15,18-22]. Another example of life-history alteration involves the harvest of female northern fur seals between 1956 and 1968 on the Pribilof Islands, Alaska. During that harvest nearly 300,000 females were killed, predominantly from the younger age groups [23]. Early maturing young females may have been more available for harvest than

409 late maturing females of the same age. For example, the early maturing females would be more likely to return to the island to copulate than the immature females of the same age. This harvest may have selectively removed females which matured early [24]. One possible result of this removal is a genetically based increased age at first reproduction [24]. As shown in Ref. [24], there is a negative relationship between age at first reproduction and juvenile survival (estimated for juvenile males). Within this correlation, age at first reproduction declines with increasing survival as would be expected in a density-dependent change [25]. However, this relationship changed following the harvest (Fig. 4), a change in which the slope of the correlation remained the same but the intercept rose. Thus, for a given level of survival, the age at first reproduction increased by about 0.6 years during the years of the harvest. This change was opposite that expected in a density-dependent reaction and inconsistent with other observed density-dependent reactions for northern fur seals during the population decline caused by the harvest [25]. Furthermore, whatever caused this deviation had to be extensive enough to overcome opposing density-dependent change. Thus, to the extent that age of maturity is genetically determined, the increase in age at maturity may reflect an effect of the harvest on the genetic makeup of the population. The direct effects of the harvest may account for part of the observed

6.45

o u

1956-66 y e a r c l a s s e s e.,

6.20

ee

"0 0 L_

9

D.. 5.95 =._

t..._ OR

r

9

-

----

5.20

_

! 0.08

! 0.16

! 0.24

! 0.32

! 0.40

I 0.48

Juvenile survival Fig. 4. Age at first reproduction for female northern fur seals, from the Pribilof Islands, Alaska, and their correlation with juvenile male survival, grouped according to the 1956-1966 year classes (during the female harvest) and the 1952-1955 year classes (before the harvest, modified from [24]).

410 decline between 1956 and 1968 [23], if for no other reason than its contribution to reduced productivity. The production of pups by young females is a significant contribution to population growth (or in the case of a harvested population, the production of harvestable individuals). Thus, genetic changes may explain part, or all, of what cannot be accounted for by the harvest [23]. Changes in the genetic composition (change in gene frequency related to age at first reproduction) of this population may also help explain the lack of recovery following the termination of the harvest in 1984. Of importance here is the fact that the components of life history (e.g. age at maturation) that may be modified by harvesting also influence rates of increase and generation time. Changes in the components of population dynamics thereby alter the rate of increase per generation which, in turn, would change the shapes of productivity and growth curves of the type shown in Fig. 2 in accord with the relationship shown in Fig. 3. With current knowledge, the direction or magnitude of these changes remain unpredictable. In accounting for the uncertainty of these changes the effects of genetic change must be considered, especially in management applications.

Discussion Because of the links between life-history strategies, population dynamics, and harvesting, conventional resource management models may be misleading and using them may lead to mistakes. A population model fit to data at one point in time is likely to become obsolete owing to the effects of harvesting through their impact on population genetics. In some cases the genetic alteration of the population characteristics of a species may result in conservative management advice. In contrast, change in the genetic composition of a population can lead to overharvesting when initial models recommended harvest levels that populations with altered life-history strategies cannot support. Changes in genetic composition may affect more than just the harvested species. Altering target species by harvesting results in an altered biotic environment for other species in the ecosystem. These changes may result in secondary genetic changes among these other species as well. The complexity and nature of ecosystem interactions make the management of renewable resources difficult and are the source of further uncertainty to be accounted for in management. This conclusion is not meant to imply that fitting population models to empirical data is an invalid means of characterizing populations. The point is that the measurements upon which models are based are for properties of species that are vulnerable to environmental influences and particularly those produced by harvesting.

Acknowledgement I wish to thank Jason Baker, Howard Braham, Gary Duker, Jeff Hard, Tom

411

Loughlin, Rolf Ream, and Anne York for reviews of earlier drafts of this paper and the improvements made as a result of their efforts.

References 1. Smith SJ, Hunt JJ, Rivard D (eds). Risk evaluation and biological reference points for fisheries management. Can Spec Publ Fish Aquat Sci 1993;120. 2. Beverton RHH, Holt SJ. On the dynamics of exploited fish populations. Fisheries Investigation Series 2, Number 19. London, UK: United Kingdom Ministry of Agriculture and Fisheries, 1957. 3. Ricker WE. Handbook of computations for biological statistics of fish populations. Bull Fish Res Bd Can 1958;191. 4. Stubbs M. Density dependence in the life-cycles of animals and its importance in K- and rstrategies. J Anim Ecol 1977;46:677-688. 5. Gilpin ME, Case TJ, Ayala FJ. 0-Selection. Math Biosci 1976;32:131-139. 6. Fowler, CW. A review of density dependence in populations of large mammals. In: Genoways H (ed) Current Mammalogy. New York: Plenum, 1987;401-441. 7. Fowler CW. Population dynamics as related to rate of increase per generation. Evol Ecol 1988 ;2:197-204. 8. Charnov EL. Life History Invariants: Some Explorations of Symmetry in Evolutionary Ecology. Oxford, UK: Oxford University Press, 1993. 9. Butterworth DS, Best PB. Implications of the recovery rate of the South African right whale population for baleen whale population dynamics. Rep Int Whal Commn 1990;40:433-447. 10. Fowler CW, Siniff DB. Determining population status and the use of biological indices for the management of marine mammals. In: McCullough DR, Reginald RH (eds) Wildlife 2001: Populations. London, UK: Elsevier, 1992; 1051-1061. 11. Rosenberg AA, Fogarty MJ, Sissenwine MP, Beddington JR, Shephard JG. Achieving sustainable use of renewable resources. Science 1993;262:828-829. 12. Smith TD. Scaling in Fisheries: The Science of Measuring the Effects of Fishing, 1855-1955. Cambridge, UK: Cambridge University Press, 1994. 13. Fowler CW. Non-linearity in population dynamics with special reference to large mammals. In: Fowler CW, Bunderson WT, Cherry MB, Ryel RJ, Steel BB (eds) Comparative Population Dynamics of Large Mammals: A Search for Management Criteria. Report to U.S. Marine Mammal Commission. Contract #MM7AC013. NTIS #PB80-178627. National Technical Information Service, 1980; 175-220. 14. Fowler CW. Density dependence as related to life history strategy. Ecology 1981 ;62:602-6 10. 15. Policansky D. Fishing as a cause of evolution in fishes. In: Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an International Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993;2-18. 16. Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an International Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993. 17. Grey DR. Evolutionarily stable optimal harvesting strategies. In: Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an International Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993; 176-186. 18. Horwood J. Growth and fecundity changes in flatfish. In: Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an International Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993;37-43. 19. Kirkpatrick M. The evolution of size and growth in harvested natural populations. In: Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an Inter-

412

20.

21.

22.

23. 24. 25.

national Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993; 145-154. Reznick DN. Norms of reaction in fishes. In: Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an International Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993;72-90. Rijnsdorp AD. Selection differentials in male and female North Sea plaice and changes in maturation and fecundity. In: Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an International Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993;19-36. Rowell CA. The effects of fishing on the timing of maturity in North Sea cod (Gadus morhua). In: Law R, McGlade JM, Stokes TK (eds) The Exploitation of Evolving Resources: Proceedings of an International Conference, Julich, Germany, 1991 (Lecture Notes in Biomathematics, 99). Berlin: Springer-Verlag, 1993 ;44-61. York AE, Hartley JR. Pup production following harvest of female northern fur seals. Can J Fish Aquat Sci 1981;38:84-90. York AE. Average age at first reproduction of the northern fur seal (Callorhinus ursinus). Can J Fish Aquat Sci 1983;40:121-127. Fowler CW. Density dependence in northern fur seals. (Callorhinus ursinus). Mar Mammal Sci 1990;6:171-195.

9 1995 Elsevier Science B.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 0. Ulltang, editors

413

Interpretation of growth layers in the periosteal zone of tympanic bulla from minke whales Balaenoptera acutorostrata Ivar Christensen Institute of Marine Research, Bergen, Norway A b s t r a c t . Growth layers in the periosteal zone of tympanic bullae from minke whales, Balaenoptera acutorostrata, have been analysed. Different methods for preparation of bulla sections were used to

show the growth layers. A comparison of the number of growth layers identified in thick etched sections and in thin translucent sections using light microscopy and scanning electron microscopy show similar readability in the different preparation techniques. Acid-etched surfaces of thick sections from 297 bullae and thin translucent sections from 399 bullae were then examined by reflected and transmitted light in a binocular light microscope. All the etched and 390 of the thin sections were examined by two readers, and 157 of the thin sections were also read by a third reader. Comparisons of the readability of growth layers in etched and thin sections were tested for each individual reader and between readers. There is some problem in counting growth layers, but there is substantial evidence suggesting that about 80% of the minke whales were correctly aged within _1 year. The deposition of periosteal growth layers is discussed in relation to body length of the animals, accumulation of corpora in the ovaries of females and other biological data. Key words: age, bone, baleen whale

Introduction Different methods for determining the age of baleen whales have been suggested during the last five to six decades, but so far no method has proved to be satisfactory. Jonsg~rd [1] tried to utilize Ruud's [2] aging method based on growth ridges in the baleen plates for fin whales, but concluded that the rapid wear of minke whale baleen made it difficult to decide more than 2-3 age groups in this species. Ear plugs [3] useful for aging other baleen whales are difficult to use for the minke whale. Sergeant [4] reported that of 11 minke whale ear plugs collected in good condition off Canada, only one could be read clearly, and Sigurjonsson [5] reported that only 42% of the plugs sampled off Iceland were readable. Laws [6] described the layered structure in tympanic bullae from elephant seals. Later, Laws [7] suggested that the laminae in bones could be used for age determination of baleen whales in which the ear plug had proved unsatisfactory, as for example the minke whale. By comparing the number of layers observed in the periosteal zone of the tympanic bullae from minke whale with layers in the ear plug, body length and the num-

Address for correspondence: Sea Mammal Section, Institute of Marine Research, Postbox 1870 Nordnes, N-5024 Bergen, Norway.

414 ber of corpora in the ovary, Christensen [8] concluded that the periosteal growth layers are formed annually. Sukhovskaya et al. [9] reported that in stained thin sections of bullae, the growth layers showed good correlations with age. However, for etched sections, the inner layers are lost and the age is therefore underestimated.

Materials and Methods Material and data were collected by observers from the Institute of Marine Research on board small whale catchers hunting minke whales in the Barents S e a - Svalbard area in the period 1977-1992. All whales examined were measured to the nearest centimetre; tympanic bullae, one from each animal, and sexual organs were collected. Bullae were sampled from a total of 516 minke whales. All of them seemed to be relatively young animals. In addition to these, a 380 cm long minke whale calf, found drifting dead near Bergen, was included in the material. In the laboratory, a set of parallel successive thick (4.5-5.0 mm) and thin (200/zm) sections was cut from the medial part of each bulla (Fig. 1). The surfaces of a set of thick sections were first polished and then etched with 10% formic acid for 60 min, dried and stored dry.

B DORSAL

L ORAL VENTRAL

A LATERAL MEDIAL

CUT

L

.........

ABORAL

Fig. 1. Schematic drawing of a tympanic bullae from minke whale: (A) longitudinal section, and (B) the transversal section in the medial part of the bone (dotted line in A). Growth layer area marked black.

415

Methods of Analysis

Light microscopy (LM) The 517 bullae were examined using a binocular microscope with zoom lens and magnification of x 6 to x50. Thick etched sections were examined by reflected light, and thin sections by transmitted light.

Scanning electron microscopy (SEM) Thick polished (9), thick etched (5), thin untreated (6), and thin polished (10) sections were mounted on aluminium stubs using colloidal carbon paint. The section surfaces were made electrically conductive by deposition of a 20 nm thick carbon film using a vacuum evaporator. Micrographs were taken both in the secondary electron imaging (SEI) and in the backscattered electron imaging (BEI) mode. The SEI mode was used to study specimen topography, but a few major fractures were seen in this mode [ 10]. The content of Ca and P in the sections was studied by X-ray micro analysis using the LINK analyzer. This is an energy dispersive X-ray analyzer (EDXA) permitting simultaneous detection of all elements in the Periodic Table from 11 (Na) to 92 (U).

Structure of Tympanic Bullae

Light microscopy The tympanic bullae consist of reticular dense bone with a narrow periosteal zone of varying thickness. The periosteal zone consists of layers of bony plates, separated by parallel adhesion layers appearing as lines in thin transverse sections (Fig. 2). The boundary between the periosteal and mesosteal or reticular bone is sometimes very diffuse but the postnatal periosteal zone is often easily distinguished from the prenatal reticular bone because the former has a more uniform structure and a different (whitish) colour. In an etched transverse section of a bullae, a periodical arrangement of ridges and furrows can be seen in the periosteal zone. One ridge and one furrow, i.e. the distance between the beginning of one ridge and the beginning of the subsequent ridge is defined as one growth layer (Fig. 3). The first growth layer is much wider (thicker) than the following growth layers. There is no neonatal line or resting line between prenatal and postnatal bone. This was also confirmed by the bullae from the very young calf found drifting near Bergen (Fig. 4). The resting lines or furrows are formed during periods of reduced growth, while the broad layers of the periosteal bone, the ridges, are formed in periods of rapid growth of the periosteal zone [8]. The first growth layer ends at the first, innermost resting line; subsequent growth layers are of variable thickness.

416

AD

M Fig. 2. A 200 r

thick section from the lateral area of a tympanic bullae (L in Fig. 1) showing (P) the periosteal zone, (M) mesosteal zone, and (AD) adhesion layers ( x 40).

In a 200 btm untreated thick section, translucent laminae appear clear or light in transmitted light and dark in reflected light. Opaque laminae appear as dark zones in transmitted light and as light zones in reflected light. Opaque laminae correspond

AD

M

Fig. 3. Etched segment of a minke whale bullae. P, M, and AD as Fig. 2. Twelve ridges and furrows show up in the periosteal ( x 50).

417

).,,

,

,.

,,

,

" ,

,

.

.

.

: %~.i .

9

:. o

Fig.

4. A 200/zm thick section from bullae of a calf. Same area as in Fig. 2 ( x 12).

probably with the ridge and the translucent laminae with the furrows in the etched sections. Scanning electron microscopy showed alternating dark and bright bands in the periosteal zone. One dark and one bright band together were interpreted as one periosteal growth layer. Figure 5 shows growth layers from thick and thin polished sections. As in LM, the growth layers show a varied appearance, from discrete and well defined to broad and poorly defined layers. Splitting and merging of growth layers were occasionally observed.

X-Ray micro analysis Analysis performed on thick polished sections in spot mode showed that the bullae consist of approximately 47% Ca and 22% P giving a total of 69% inorganic material. Significant variations in the composition were not found either between the mesosteal and the periosteal zones, or between the dark and light constituents of the growth layers.

Growth of Bullae

As mentioned by Christensen [8], the resting lines or furrows are formed in periods with low food consumption or slow growth. All bullae in the sample were collected in the Barents Sea early in the feeding season, in May-June. We therefore mostly found fully developed growth layers in the bullae. Two observers with experience using LM in counting growth layers in bones and teeth studied segments from 250 bullae. The results of their readings are given in Table 1.

418

o.

~

.

.

.

.

.

.

.

.

.

Fig. 5. The surface of a thick (a) and thin (b) polished sections imaged in SEM, BEI/COMPO mode. Periosteal growth layers are clearly resolved (arrowheads) and can be seen split an emerge.

Table 1 shows that in the etched section 39% of two successive readings by observer A are identical, while 36% show a difference of +_1 growth layer. The two observers' readings of growth layers in thin sections give about the same percentage scores for two successive readings, 42% (A), 43% (B) for identical readings, and 32% (A) and 38% (B) for difference of _1 growth layer. When comparing A and B readings of the same section, only 20% were identical in numbers of growth layers, while 39% showed a difference of _+1 growth layers (53 readings). Closer examination of these 53 readings (Table 1, _+1), illustrated that B gave 1 more growth layer than A in 39 cases, and 1 less in 14 cases. For the _+2 layer

419 Table 1. Difference in counts of growth layer group in the periosteal zone in middle part of tympanic bullae by two observers A and B and in two successive readings by the same observer

Section

Difference

Etched A (no.) Etched % Thin A (no.) Thin (%) Thin B (no.) Thin (%) Etched-thin, A (no.) Etched-Thin, % Thin A-B (no) Thin A-B (%)

Total

0

_+ 1

_2

__.3

Other

55 39 81 42 51 43 42 27 27 20

51 36 61 32 45 38 55 35 53 39

28 20 31 16 12 10 29 19 36 26

6 4 10 5 6 5 12 8 9 7

3 2 12 6 5 4 18 12 12 9

143 193 119 156 137

Etched section, reflected light (one observer, A). Thin section (200/~m) in transmitted light (two observers).

difference, B gave 2 layers more than A in 31 cases and 2 layers less in 5 cases. In conclusion, B overestimated the number of growth layers relative to A. The readability of growth layers in thin polished and thick polished segments examined in SEM (22%, 44%, 33% for identical _1, __.2layers) is about the same as in thin and etched sections in LM (Table 1). Thin sections in LM and SEM show similar readability (47%, 33%, 7% for identical -+1, _+2 layers) as for two successive readings of thin sections of LM. This comparison shows that the number of growth layers identified using different preparing techniques; etching and thin sections, polished and unpolished segments, reflected and transmitted light, LM and SEM, give similar result for growth layer count in minke whale bullae.

Readability and interpretation Based on these assumptions, three readers (termed 1, 2, 3) with experience in counting growth layers in bones and teeth using LM, studied the sections from the 517 tympanic bullae. Table 2. Average coefficient of variation for age determinations by reader and preparation method

Reader

Males (N)

Etched sections 1 0.0682 (105) 2 0.1074 (9) _

Thin sections 1 0.0919 (125) 2 0.1005 (121) 3 0.0615 (39)

Females (N)

Sexes combined (N)

0.0743 (192) 0.1231 (23)

0.0721 (297) 0.1187 (32)

_

0.0947 (274) 0.1214 (262) 0.0799 (80)

0.0938 (399) 0.1148 (383) 0.0739 (119)

420 Table 3. Percent agreement in age determinations within indicated differences (readers 1, 2, 3" methods: E, etched section; T, thin section) Difference within (years)

Comparison (reader/method) 1 E - IT

2 E - 2T

1 E - 2E

1 T - 2T

I T - 3T

2 T - 3T

0 _1 _+2 _+3

28.6 64.8 85.7 94.5

21.6 56.1 76.6 88.3

33.4 77.5 93.2 97.3

33.1 78.7 94.6 99.0

33.1 77.1 96.2 97.5

34.4 77.5 90.7 97.4

To compare the precision of age determinations between readings for each reader and between the readers, an average coefficient of variation (cv) was calculated for each reader by preparation method for the sections. These cvs are based on sections for which two or more readings are available. As seen from Table 2, it is not possible to select one of the methods as the best with regard to reproducibility of age readings. The table also indicates that for all readers and both methods, age determinations of samples from males seem to be more reproducible than those from females. Agreements in age assigned with selected deviations are given as a percentage in Table 3. These agreements have been calculated using the means of age determinations from sections, rounded to the nearest integer. The percent agreements are comparable within methods, but decrease between methods. The distributions of differences in age determinations between readers and methods are also illustrated in Figs. 6-11. These distributions may indicate problems relating to interpretation if they are asymmetric around 0. Two readers determined age from both etched and thin sections but these results are not conclusive; both distributions are asymmetric. However, while in the case of reader 1 ages determined from etched sections tend to be higher than those from thin sections (Fig. 6), the situation is just the opposite for reader 2 (Fig. 7). For the two readers of the etched sections, reader 1 tended to estimate higher ages than reader 2 (Fig. 8). Thin sections were read by three readers. The highest ages were given by reader 2, then reader 3, with reader 1 giving the lowest age determinations (Figs. 9-11). The

i

-8

I

-6

I

-4

I

-2

I

I

I

0

2

4

1

6

Fig. 6. Comparison of age determination by reader 1 between etched sections and thin sections (E-T, N = 182).

421

I

i

-6

i

-4

-2

i

I

I

0

2

4

Fig. 7. Comparison of age determination by reader 2 between etched sections and thin sections (E-T, N = 171).

comparison between reader 1 and reader 2 indicates that there may be a problem with the interpretation of the thin sections (Fig. 9).

Conclusion Growth layer groups are identified in etched thick sections using reflected light, thin sections using transmitted light in light microscope, thick and thin polished and unpolished sections using scanning electron microscope. The growth layers in etched sections are a periodical arrangement of ridge and furrows observed in the periosteal zone. In thin sections using transmitted light, the growth layers are observed as dark and light lamina. These correspond probably to the ridge and furrow in etched sections.

8

9

'

-6

-4

-2

i

i

i

i

0

2

4

6

i

-4

t

-2

|

0

-

2

4

Fig. 8. Comparison of age determinations from etched sections by reader 1 and reader 2 (El-E2, N = 293).

Fig. 9. Comparison of age determinations from thin sections by reader 1 and reader 2 (T1-T2, N = 390).

422

10

I

-4

11

n

I

-2

I

I

I

0

2

4

~

I

6

-4

I

-2

'

I

I

I

I

0

2

4

6

Fig. 10. Comparison of age determinations from thin sections by reader 1 and reader 3 (T1-T3, N = 157). Fig. 11. Comparison of age determinations from thin sections by reader 2 and reader 3 (T2-T3, N = 151).

In scanning electron microscopy, alternating dark and bright bands are observed in the periosteal zone which also probably correspond to the ridges and furrows seen in etched sections. No variation in the content of Ca (47%) and P (22%) was found in the bullae, either between mesosteal and periosteal, or between dark and light lamina. The number of growth layers identified using different techniques gave similar resuits for the growth layer count in minke whale bullae. The study did not give conclusive evidence of which of the methods, etched sections in reflected light or thin sections in transmitted light were the most reproducible. However, agreement in layers assigned within _1 layer in 77-79% occasions suggests that the method of counting growth layers in tympanic bullae is a useful approach for studying age-related biological data. The etching methods in this study were the same as used by Christensen [8]. He also correlated available biological data with the growth layer, and concluded that one growth layer group is laid down per year. There are still some problems in counting growth layers, but there is substantial evidence to suggest that a minke whale is correctly aged within _+1 year in about 80% of cases.

Acknowledgement This project has been funded by the Norwegian Fisheries Research Council (NFFR), project number 4001-701.247.

References 1. Jonsg~d A. Studies on the little piked whale or minke whale (Balaenoptera acutorostrata Lac6p~de). Report on Norwegian investigations carried out in the years 1943-1950. Norsk Hvalfangst-tid 1951 ;40:209-232.

423 2. Ruud JT. Further studies on the structure of the baleen plates and their application to age determination. Hvalr~d Skr 1945 ;29:1-69. 3. Purves PE. The wax plug in the external auditory meatus of the Mysticeti. Discovery Rep 1955;27:293-302. 4. Sergeant DE. Minke whales, Balaenoptera acutorostrata of the western north Atlantic. J Fish Res Bd Can 1963;20:1489-1504. 5. Sigurjonsson J. A preliminary note on ear plugs from Icelandic minke whales. Rep Int Whal Commn 1980;30:193-194. 6. Laws RM. The elephant seal (Mirounga leonina Linn.) I. Growth and age. Sci Rep Falkland Isl Depend Surv 1953;8:1-62. 7. Laws RM. Laminated structure of bones from some marine mammals. Nature London 1960;187:338-339. 8. Christensen I. Age determination of minke whales, Balaenoptera acutorostrata, from laminated structures in the tympanic bullae. Rep Int Whal Commn 1981;31:245-253. 9. Sukhovskaya LJ, Klevezal GA, Borisov VJ, Lagerev SJ. Use of bone layers to determine age in minke whale. Acta Theriol 1985;30:275-286. 10. Christensen I, Krekling T, Salbu B. Growth layers in tympanic bullae from minke whales (Balaenoptera acutorostrata), determined by light and electron microscopy. Int Whal Commn Sci Comm Pap, 42 (NHMi3) 1990 R~sum~ Section. In: Rep Int Whal Commn 1991;41.


9 1995 ElsevierScience B.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand t3. Ulltang,editors

425

On the life history and autecology of North Atlantic rorquals J6hann Sigurj6nsson Marine Research Institute, Reykjavfk, Iceland Abstract. Background and methods: this paper briefly reviews in qualitative terms available information on exploitation, stock development, life history and autecology of North Atlantic rorquals (mainly blue, fin, sei, minke and humpback whales) with special reference to interspecific variations in abundance. Historical catches and information on growth, reproductive rate and food preferences are examined in relation to present stock sizes. Results and conclusions: recent sighting surveys show that the North Atlantic rorquals cannot be regarded as threatened or endangered by exploitation, although some species are depleted locally in some areas. The minke and fin whales are both euryphagic seasonal feeders that prior to exploitation and still today number in the 50,000-100,000 and 100,000-200,000 range, respectively. Their opportunistic feeding habits and choice for vast open ocean breeding areas seem to give rise to relatively large stocks and to their apparent sustenance for high catches. The near stenophagic crustacean feeding blue and sei whales seem on the other hand to have had smaller historical stock levels of well within 20,000 animals. While the sei whale is probably at present in the 10,00015,000 range, the blue whale is still at a low level, although increasing in some areas, such as off Iceland. The humpback whale seems to be the species that historically occurs in smallest numbers, although the northwestern stock seems to be in a healthy state (increasing by ~10% per year). The strong recovery of the species in recent decades may be linked with its euryphagous lifestyle. However, the relatively small population size may also be correlated with its dependence on the rather limited coastal zones for breeding. Key words: balaenopterids, stock size, exploitation, status, feeding, competition

Introduction

The past century has been a dramatic epoch with respect to the development and exploitation of North Atlantic balaenopterid whales or rorquals, since with the introduction of modem whaling in the late 1860s they became the main target of the newly developing industry [ 1,2]. The rorquals are part of the baleen whale group and are all, except the minke whale (Balaenoptera acutorostrata), true large whales. The other rorquals occurring in the North Atlantic comprise the largest animal on earth, the blue whale (B. musculus), in addition to fin (B. physalus), sei (B. borealis), Bryde's (B. edeni) and humpback whales (Megaptera novaeangliae). The minke whale was, however, not subject to large scale exploitation until during the first decades of this century with the introduction of small-type whaling activities in several areas of the North Atlantic Ocean [3-5]. The intense whaling pressure caused significant reduction in many of the rorqual stocks even before the turn of the century. This paper briefly reviews the history of

Address for correspondence: Marine Research Institute, P.O. Box 1390, 121 Reykjav~, Iceland.

426 exploitation and discusses a few qualitative aspects of development of stock sizes and the present situation in light of several life history and ecological characteristics of the different species and stocks.

Main Factors Affecting the Survival of the Stocks There are a number of factors that need to be considered in this context (see e.g. Ref. [6]), some of which one may explore, while others are very difficult to approach. We have mentioned the exploitation that evidently is important here. Biological limits are important, the rate of reproduction, the growth rate, etc., i.e. the life history characteristics of each species and stock; also, the selection of habitat, food preference and feeding strategy that the species have adjusted to, and the availability of preferred habitats and food in time and space. Finally, behavioural aspects such as territorialism that we know so little about, may play a central role in this game of survival and competition. One can say that the optimization of all these factors forms each of the species' competitive strength.

'3 ~ " ~-~

... ....

I

:.~..:t

1 4

X .:

k

\

7

Y'.sr~.o~,,AA,-, -ICELAND I

-LABRADORI\

I 7 I

I !

.~

:i !

I

....

I

3..J ~

~ I

"-~

-.

r__;o o.

...

.~I GREENLAND~I

~..1~

'NORWAY

...."

I

~. ..'a ** o~i-.~,,."}

NORTH

".~.~

..'::~.~!

:-...

(

.~~

L~.. ~""

J~

I

r~.~

Io

. o v a sco,,~ ~,9 w

I

,po~

..

.

.

.

Fig 1. Traditional IWC stock divisions of N Atlantic fin whales.

OR ~Y FAROE,-

r

I ~'-J~".

.,'+,srA

:

-'.

.t~

~

".~"

~:. i[ ~. .'L, _,-~...-.-" ~J ~: ~'.J

-SPAIN & PORTUGAL..

" 29ow_ I .

~ /&

'

-~

,?.

...i .... 90

-~';-~ . ~ . ~:

427

Migration, Distribution and Stock Units The typical migratory cycle of a large baleen whale is a summer feeding migration to high latitudes and a return migration to wintering/breeding grounds in autumn and winter at low latitudes [2,19]. Although this seems to be the general pattern with respect to the rorquals at high latitudes, there are clear exceptions, e.g. the humpback whales that annually occur during winter at the capelin (Mallotus villosus) grounds north off Iceland and at the turn of the century they occurred during January-March at the northwest coast of Finnmarken, N Norway [20,21]. Detailed accounts on the distribution and movements of the rorquals in the N Atlantic are reviewed, e.g. by Kellogg [22] and Jonsghrd [23] and more specifically for blue and fin whales in Jonsg~d's comprehensive papers [24,25], for sei and minke whales in Horwood's two monographs [3,26] and for humpback whales in Winn and Reichley [27]. The question of stock discreteness of each of the species of rorquals has been dealt with in detail by many authors in conjunction with management of whaling [714]. However, scientists in the mid-1970s made postulations regarding division of the N Atlantic baleen whales into stocks that gave rise to the traditional IWC management units [13]. These were mainly based on the distribution of catch grounds and recent sightings, catch history and whale markings. Accumulating evidence favours the idea of several genetically discrete or near isolated entities for minke, fin and humpback whales [10-12]. Fidelity to site, demonstrated well by traditional whale marking/recoveries in fin, sei and minke whales [9,14] and by repeated sightings of photoidentified blue, fin, humpback and minke whales [15-18], is also an important element in this discussion on reasonable units to manage whales, that accommodates some restricted intermingling, even between distant feeding populations. Figure 1 shows the seven traditional stock areas or management units for fin whales in the N Atlantic. Similarly, some four stock areas have been suggested for minke whales, i.e. two western units, one central and one northeastern stock unit. Three sei whale stock units (Nova Scotia, Iceland-Denmark Strait, and Eastern stock) have been suggested [13]. All these rorquals, except the sei whale, have summer feeding grounds ranging north to the ice edge and as far south as the Gulf of Mexico, the Mediterranean Sea and the coast of NW Africa, and more oceanic offshore winter distribution at lower latitudes. The sei whale on the other hand (Fig. 2a), does not normally reach as far north in summer, is more confined to temperate/boreal waters, and stays also offshore but has more southern distribution in winter than the other species, possibly as far south as the equator [26]. Figure 2b shows the feeding and breeding areas of humpback whales at high and low latitudes, respectively. Due to the humpback's coastal affinity and its often clearly identifiable marks on the ventral fluke that can be photographed, this species has been the subject of intensive investigations since the 1970s. The feeding aggregations seem to form quite discrete matrilineal units, apparently with a panmictic genetic pool [ 12,17]. The whales from the western Atlantic including the grounds off

428 Sei whale

75*

70*

65*

60* 55* 50* 45* 40" 35* 30* 25* 20* 15" 10" 5*

70*

50*

30*

10"

10"

30 ~

10 ~

30"

Humpback whale

75*

70*

8

65*

60* 55* 50* 45* 40* 35* 30* 25* 20* 15" 10" 5*

70 ~

50 ~

30 ~

10 ~

Fig. 2. Summer feeding grounds and winter breeding grounds of N Atlantic sei (a) and humpback (b) whales.

429

I!

3000

Fin

Sei

2500

Blue

2000 O E

15o0

Z 1000 500 0

~-v

1868

1878

1888

~

.,,.,,~,,,~,,~=~,~.w.._..,~.

1898

1908

~,~--~r.

1919

-

1929

.

1939

.

.

.

.

.

.

1949

.

.

.

.

.

.

1959

.

.

.

1969

1979

Year Fig. 3. Catches of blue, fin, sci and humpback whales in the N Atlantic 1868-1985 (cf. text).

Iceland, migrate in winter to the breeding grounds in the Caribbean Sea. There are large feeding aggregations off Iceland and Newfoundland, and smaller ones at W Greenland, in the Gulf of St Lawrence and Gulf of Maine. The humpbacks off N Norway may have a relationship with the breeding grounds in the southwest, although we have traditionally assumed that this feeding aggregation migrates to the eastern breeding area off Cape Verde Islands, which at present is evidently sparsely populated [27,59].

Exploitation The catches of the four large rorqual species from the start of modem whaling last century (based on Refs. [28,29]) are shown in Fig. 3. We can see a period of learning and expansion of whaling activities into new areas that gave improved catch results up to the mid-1920s when over 3,000 animals (where unspecified catches have been incorporated) were taken annually. The largest catches were taken off Great Britain and Spain, while whaling first commenced off the coast of N Norway in the 1860s and in the 1880s at Iceland, but both these areas played a major role in the history of modern whaling in the N Atlantic. Note that in the early years, strike and loss rate was quite high [1 ], so here we are only talking about landed animals that may have been well over 100,000 in the last century or so. Examination of the species composition indicates that some 79,000 fin, 12,000 blue whales, less than 10,000 humpbacks and 16,000 sei whales were landed during commercial whaling since the late 1860s. The blue and humpback whales were mainly taken during the first period of modem whaling, while catches of fin whales

430 increased until the mid-1920s and then decreased with the closures of the fisheries in the British Isles and later the commercial operations in the Faroes in the 1960s, in Canada and N Norway after 1972, and Spain and Iceland in the mid-1980s. In contrast to the large whaling industry, minke whaling with cold-harpoons and motor vessels did not commence until well into this century. It was confined to small fishing boats in nearshore waters of Norway [4] and Iceland [5,31 ] at the beginning of this century. After the World War II, Norwegian minke whaling expanded to the west [30] and local whaling commenced on the Canadian [32,33] and Greenland coasts [34,35]. The bulk of the minke whale catches were taken by the Norwegian fishery, which gradually expanded towards the open sea on board well equipped and larger vessels that were able to catch and process the whales at sea. At most the annual take by the Norwegian fishery was over 4,000 whales [28,29], but the minke whaling activities were less widely distributed than whaling for other rorquals.

Present Status

Intensive hunting in the last century, involving mainly the five species of rorquals, has had major impact on the status of the stocks. This applies particularly to blue and humpback whales, which probably were reduced to very low levels throughout the ocean just after the turn of the century [17,24,38]. Now recent sighting surveys and other investigations have given us information on the current stock levels (see Table 1). Although apparently the humpback whale never was in great numbers in the N Atlantic [59], after decades of depleted status the stock has more or less fully recovTable 1. Present status of stocks of North Atlantic rorquals

Species

Blue

Stock size

1,000-2,000

Fin

50,000+

Sei

13,500+

Minke

Humpback

100,000+

5,500+

Status

Sources

Low level, mainly off Iceland and Gulf of St Lawrence, far less in other past whaling grounds, increasing by 5% per year off Iceland Still in good numbers, although depleted off W Norway and UK; lack of recent sightings estimate in NW Atlantic, so tagging estimate from 1970s and CETAP results from the US coast used Recent survey estimate in the central N Atlantic available, NW Atlantic marking estimate (1,800) used, depleted in some earlier whaling grounds, such as off N Norway Stock size reduced but still abundant in NE Atlantic, the smaller NW stock may number several thousands NW Atlantic near pre-exploitation level, increasing by --10% per year, eastern stock depleted

15,24,33,37-39,59

8,33,39,40,46

26,33,38,42,46

7,46

17,36,37,59

431 ered in the western and central distribution area and may still be increasing [ 17,36,37,59]. However, the whaling grounds at the eastern side of the N Atlantic are almost vacant for this species. It seems that, prior to exploitation, both sei and blue whales were in somewhat greater abundance or in the 10,000-15,000 range [38]. The present level of sei whales may be of similar order of size in some areas as prior to exploitation, while its absence from the earlier grounds off N Norway has been noted [26]. On the other hand, the blue whale is evidently still at a low level, while showing some significant signs of recovery (5% per year) off Iceland [37]. Moving to the far more abundant fin whale stock, which evidently was also subject to heavy taxation at the end of the last century and the first half of this century [25,28], the situation is different. Nevertheless, some major past whaling grounds, such as off W Norway and the British Isles, seem still to be sparsely populated [8]. Finally, we have the stock of minke whales, which may originally have numbered in excess of 150,000 animals and still is probably well above 100,000 in the entire N Atlantic Ocean [7].

Life History Parameters One may like to speculate why these rather closely related species exhibit such different historical stock levels and what are the likely factors affecting recent developments of the stocks. Such speculations may be of help in understanding stock development and can be useful in generating models to study the stocks in the future. Let us first take a look at the life history characteristics. As we see from Table 2 the ranges in life history parameters are not very different in the large species, the age at sexual maturity ranging from 6 to 12 years of age in blue, fin and sei whales, while it seems that both humpback and minke have a slightly faster growth potential, with age at maturity being somewhat lower. The pregnancy rates are of the same order of magnitude, i.e. normally a 2-year reproductive cycle in the larger species with possibilities of up to 1.5 calf per 2-year period.

Table 2. Life history parameters for balaenopterid whales

Blue a Length of newborn (m) Length (m) at sex. maturity Males Females Age (years) at sexual maturity Males Females Pregnancy rates Length of gestation (months)

--7 20-21 21-23 ~10 ~10 10-11

Fin b

Sei c

6.4

4.5

17.7 18.3

12.0-12.8 13.1-13.4

8-12 6-10 0.5-0.73 11.2

7-11.7 5.6-11.7 0.36--0.47 10.7

Sources: a[44]" b[44,45,55]" c[26]; d[3,7]" e[27,57].

Minke d

Humpback e

2.4-2.8

4-5

6.8-7.0 7.3-7.4

11.6 12.1

3--6 5-7 0.86-0.99 10

2-5 2-5 0.3-0.43 11-11.5

432

Fin

12

~vhales

: Age

'al Iransilion

phase

+1

IO cq

9

i ....

,

55 1960

.

.

.

.

i

'

65 19'70 75 19'80 '85 19'90 Yearat transition Fig. 4. Changes in age at maturation in fin whales off Iceland.

However, the typical minke whale cyclus is annual, which presumably gives it a plus on the competition record. The ranges of ages at maturity shown in Table 2 are to a large degree a reflection of growth rate in the populations where we have available information of this kind. Evidence for dramatic changes in the growth rate of fin whales off Iceland demonstrate this well, where after years of decline in age at sexual maturity [41 ], a reversed trend in the most recent year classes has followed [43,45] as shown in Fig. 4. This has been related to the available food resources [6,19,46], where great fluctuations in fecundity have also been linked with changing availability of food. In conclusion, changes in growth rate and the corresponding changes in age at maturity are important elements in the survival of these stocks and here all the balaenopterid species may have a similar chance. Likewise, it seems as if minke whales may have greater potential in utilizing favourable conditions, when these are available, than the larger species, if only reproduction would decide who is to win.

Table 3. Prey group preference in North Atlantic fin and sei whales

Species

Area

1st prey

2nd prey

Fina

N Norway W Norway Faroes Icelandc Nova Scotia Newfoundland N Norway W Norway Icelandc Nova Scotia

Krill Krill krill krill krill Capelin Copepods Copepods Krill Copepods

Capelin/herring

Seib

Sources: a[25,47]" b[26,47]; Cunpublished.

Herring Blue whiting, herring, capelin Sandlance/mackerel Sand1ancellantern Krill Krill Copepods Krill

433

Table 4. Main fish species found in N Atlantic minke whales Area

Herring a

Capelin a

Barents Sea Norwegian coast Great Britain Iceland E Greenland W Greenland Newfoundland Eastern US

x x x x

x

Ammod. sp. a

Gadoids b

Mackerel b

x x x x

x x

X

X

Sources: [3,47-50,53,54]. aGreatest in quantity; bless in quantity.

Food and Feeding Table 3 lists the favourite/common food (lst prey/2nd prey) of fin and sei whales. Usually fin whales go for krill as the first choice, but they are also taking fish in considerable amounts in certain areas, seasons or time periods [25,47]. The fish species are then most often capelin (Mallotus villosus), herring (Clupea harengus), sandlance or Ammodyte spp. The humpbacks feed even more on fish, while also taking krill. The sei whales eat nearly exclusively crustaceans [26], most often the smallest of these, the copepods. The blue whales are also solely crustacean feeders [59], and are

10 Blue ~

Fin - -o -

Humpback--

I::1.-. S e i I

[] ..........

.~

_E

o

~

~9 s r

6 Z

0 , .J. -. .- .-. . . . . .

1-15 JUN

_n. . . . . . . .

...n, ..........

1-15 JUL

cr

w

. . 5 '

..........

I

I

1-15 AUG

1-15 $ E P

Fig. 5. Sightings of blue, fin, sei and humpback whales on board vessels west and southwest of Iceland, June-September 1979-1985.

434 even more specialised, nearly always preying on krill. The minke is perhaps the most extreme opportunistic feeder of the N Atlantic balaenopterid whales, taking mainly fish, but also krill, this all varying greatly between seasons and areas. The main species of fish found in the stomach of minke whales in different areas of the N Atlantic are listed in Table 4. In conclusion, it can be stated that there are generalists/opportunists as Mitchell categorized it [47], the minke, humpback and fin whales forming that group; there are the specialists, the blue and sei whales. No doubt the position in the food web and the degree of specialization counts very much in the whale's ability to cope with variations in the availability of different prey. It is important for us to monitor the status of the principal species of fish stocks as well as crustacean production in order to capture a picture of the situation the whales are in at any given time. Many of the pelagic fish species undergo major natural fluctuations that can greatly influence the livelihood of the whale population in the area, such as the capelin off Newfoundland [51 ] and off Iceland [46], or the herring off Norway and Iceland. Off W Iceland herring comprised up to 30% of the food of fin whales at the turn of the century (unpublished data), while in recent years there is

'

G

~#

foJ.2:.% t-~k',.*,"

~"

;'""

..."

-20"

o

-i

."9

I/

*

9

.

o~ 0

0

0,J ,.~

_@

%

9.•

:" ~

":, :9

.

. . ~,- . . 9. . . "-. :-...... - - , 9

~

.o,

"

"

9

o" .....

,.," -,

,o,'

9

. .-..-

..../

. ....

"

\

;.

-~"

.

....... . . " :

.

~

...o" ~ ~176 ~ .~. '2, ,~176149 9 9 :"

.:" ..--p

0

*,......'-'-"

.-9"""~.: :..'

?...~,

0

:9 ........ ',.[ ..... 9.... ,-'"'"""

9. .

. oo 0

0

o~

~

9

# 0

I

J

I

,

,

I

.... I

9

Sei w h a l e s Fin w h a l e s

I

I,

i

Distribution of fin and sei whales during Icelandic sightings surveys July-August 1989 (500m and 2000m depth contours shown).

Fig. 6.

435

Blue w h a l e

a

,

~.

, :.:,

-

.: .

,.~ i

..-""O

"o

.

,

..-.-

.

.

.

~:~.~:~

.

i;

+++`~::/~.

,---o'~-~o ooe~eo @. ---~

O

-,~:

,

9e o . e . e o . o

9 9

,'

-O

(

.

9

.,~

~+:: +, :++~:!! ",

.

~0,~

eo. ,-

o~O

./I

o"

:.'.

9

/"

+

~,

~

o~;::i

olHxte.,.-o--

9

}l j

._._9,,.,_---*-

9 9 9 oO-

,'.

9

/+" ,dlk v

,

9 db, v

"

"l +o

9 9

9 e O e

~

I

9

--

.....

i,ooo'

--

I 2000

---

I

D e p t h (m)

Minke whale

-: i.

'.

9 '.:....-:'.

....'"" ~,

,~

Q....+.... ..../

':i

,.." ", . . . . . - -

1.'-"

9

9

9I

.'

.

..

/

."..

"'

9 I

i ..

9

/.

.....

~

i

l.;

~

~

~

M

n

~

~

+

~.~

ilr

.o,_,+/oo~

~+" "

O

-"

":"-... .....---."...... r:~;. ~ "--:--I, - 4I !J ; + o I 9-, ~.-.--"._'. .-" j ~....

]

9

9

t

9/ ,

~/'J )lj

I/ ~

, ,.,

"- ~

9 ,,i:!

I ",

.

-

,., ._'

', -

.

,,. :... ~

.

~

~

~

./.

9, .

'I

,

g ~

9 ;O

~,

I

_

.

!!'

9o o e o

"'x./'J

."

.

9 9- . .

('-J

b '.:.

./,

9 9o.q~O gO'o'-" go 9 r 9o O O O o ~ - ~ f 9 9O O O i O + O ~ 9,..~o,.:...,,,

,,_

, ,.,

9

,../O;OOOe0or,

II~'

,

~i: ~+

J

I

"'

),0

..... . . . . j /

,,"

/~

/"

I"

/

,, ,'-

.

--,..__l

,,"

:_.j

..1500

9 ~

t

.....

I

t '~

--

I 2000 -I Depth(m) ,

I

,

..~

7. Sightings of blue (a) and minke (b) whales west and southwest of Iceland June-September 1979-1985.

Fig.

436 1951-1955

a ,~I

U

"9

.."

, u I

l

!

/ :"

I

"

.'

-"

;/

..."~

,"

"'~

..-"

,..,"

i

.

.

I I

;; ;:

.--: ..... .." . . . . "

" .........

." .

I

.

I

..-'~ ""

.

i -.. ~,

/

I

"'%

',

" ...........

......" ...........

~.

.

.

~,,

"',

r""

.~

i

,,

i h"

.

----

,,..,~'

~

""

0

% "'-->, % ~ ": % %

.o-"

:

:

9

,

-

,"

s

.

o.O

-"

9 .o"

9

I

/ ~

%.0

;.-" o-"

"

I

i

,,""

~

~

/

//

~

r

_

I

~

o'

"

-"

;

"-..

9

,

",.".~.

! _ ',.

'--,'~..e

(

el /~O

~.1

r

..~,._,,

4""S/a'-_ ""~,,-. 9

~ 55 cm

35

~

Neonates

30

~

porpoises< 74.3 cm

25

o '(9

20

E = z

15 10

F

,unJ M

u

A

M

!

J

i

J

i

A

S

O

,

N

!

D

Month

Fig. 1. Near-term foetuses and neonates of Phocoena phocoena from Dutch waters, by month of birth. Although there is a clear peak in July, the birth season extends into August.

nificant activity was found in the male from N o v e m b e r (this specimen requires a more thorough study). Seven female porpoises studied for reproduction were aged as b e t w e e n 3 and 5 years old. Table 1 shows that five of these were already sexually mature. The ani-

Fig. 2. Sperm taken from the epididymides of the male stranded on 26 March 1992, photographed during the postmortem (photograph J.C. den Hartog).

462 Table 1. Female Dutch harbour porpoises aged between 3 and 5 years old

Maturity

Date

Place

89.03.15 89.09.01 90.12.05 91.01.02 92.01.09 92.01.18 92.11.22

Bloemendaal Mature Z a n d v o o r t Mature Oosterschelde Immature Texel Mature Texel Immature Camperduin Mature Castricum Mature

Length

Weight (kg)

Age (year)

Pregnant

145 154 132 169 149 147.5 150

59 39 29 51 33.5 48 41

5-5.75 3-4 4.25 5 5 5.25 5

Yes No No No No

(cm)

Five out of seven porpoises are already sexually mature (defined as at least one CL or CA present). mal stranded on 15 March 1989 was pregnant with a 53 cm foetus. It had only one CL of pregnancy and no other scars, so must have become pregnant for the first time when it was between 4 and 5 years old. Although P h o c o e n a is known for its ovarian asymmetry, within a series of 15 sexually mature females with at least one CL or CA, four animals showed functional development of both ovaries (26.6%). One of these, a pregnant animal, had the foetus in the right uterus horn and the CL of pregnancy in the right ovary.

Discussion Although there is a pronounced peak in July, the birth period in Dutch waters appears to be extended (see Fig. 1). SCrensen and Kinze [12] as well as Read [13] report a narrow birth peak, from the middle of June to early July for Denmark and from May to June for Canada. In central Californian waters, however, there appears to be a more extended birth period again, from June into August/early September [ 14]. There are sightings of calves as early as April in British waters (P.G.H. Evans, personal communication) and there is one calf in April among a limited set of sightings in Dutch coastal waters [15]. The causes of these differences between porpoises from different geographic areas are not clear; for a discussion of the topic, see SCrensen and Kinze [ 12]. The studies from Denmark and Canada [12,13] also find a strongly synchronized female reproductive cycle, and this appears to be true for males as well. Because most Dutch porpoises are collected in late autumn/winter, there are almost no females available for studying maturing follicles. So far, based on gross pathological examination, only ten males assumed to be mature have been collected during this study. The two males with aseasonal (as defined by SCrensen and Kinze [12]) sperm activity would be in agreement with an extended female mating period. A previous study found an age of sexual maturity of 6 years for Dutch female porpoises [ 11 ]. In our material, the ASM is lower for five of the seven animals studied (see Table 1). It is suspected that some of the ages published by Van Utrecht [ 11 ] may be too high. Some of the teeth studied by Van Utrecht were cut in a way that

463 could lead to an erroneous age determination (C. Lockyer, personal communication). The animals in question will be aged again. If Van Utrecht's age determinations prove correct, our findings could be evidence of a population under pressure, in which the ASM would have decreased by 1 year or more. A similar reaction was observed in the harbour porpoise population in the Bay of Fundy, Canada [ 16]. Phocoena is known for its ovarian asymmetry, with only the left ovary developing during puberty and all activity confined to this ovary [12]. It is thus very interesting that 4 out of 15 females from Dutch waters have both ovaries functionally developed. Because histological examinations are not completed, we will refrain from speculating about the possible causes of this phenomenon.

Acknowledgements We thank P.J.H. van Bree for giving access to historical reports and archives, T. Olesen for processing the two samples for histology, H. Kremer for processing the teeth, J.C. den Hartog for taking photographs during a postmortem, and C. Smeenk and P. de Wilde for critically reading this manuscript.

References 1. Smeenk C. The harbour porpoise Phocoena phocoena (L., 1758) in The Netherlands: stranding records and decline. Lutra 1987;30:77-90. 2. Verwey J. The cetaceans Phocoena phocoena and Tursiops truncatus in the Marsdiep area (Dutch Waddensea) in the years 1931-1973. Nederlands Instituut voor Onderzoek der Zee, Publikaties en Verslagen, 1975;17a,b:l-153. 3. Evans PGH, Harding S, Tyler G, Hall S. Analysis of cetacean sightings in the British Isles, 19581985. Peterborough, UK: Nature Conservancy Council, 1986. 4. Evans PGH, Scanlan GM. Historical review of cetaceans in British and Irish waters. UK Cetacean Group, Zoology Department, University of Oxford, 1989. 5. Camphuysen C, Leopold MF. The harbour porpoise Phocoena phocoena in the southern North Sea, particularly the Dutch sector. Lutra 1993;36:1-24. 6. Kuiken T, Garcfa Hartmann M (eds). Proceedings of the first European Cetacean Society workshop on cetacean pathology: dissection techniques and tissue sampling, Leiden, The Netherlands, 1991. ECS Newslett 1993;(Special Issue 17):1-39. 7. Perrin WF, Myrick Jr AC (eds). Age determination of toothed whales and sirenians. Cambridge: Rep Int Whal Commn 1980;(Special Issue 3). 8. Deinse AB van. De Fossiele en Recente Cetacea van Nederland. Amsterdam: HJ Paris, 1931. 9. Deinse AB van. De recente Cetacea van Nederland van 1931 tot en met 1944. Zool Med 1946;26:139-210. 10. Slijper EJ. Die Cetaceen Vergleichend-anatomisch und Systematisch. Den Haag: Martinus Nijhoff, 1936. 11. Utrecht WL van. Age and growth in Phocoena phocoena Linnaeus, 1758 (Cetacea, Odontoceti) from the North Sea. Bijdr Dierkunde 1978;48:16-28. 12. SCrensen TB, Kinze CC. Reproduction and reproductive seasonality in Danish harbour porpoises, Phocoena phocoena. Ophelia 1994 ;39:159-176.

464 13. Read AJ. Reproductive seasonality in harbour porpoises, Phocoena phocoena, from the Bay of Fundy. Can J Zool 1990;68:284-288. 14. Hohn AA, Brownell RL. Harbor porpoise in central Californian waters: life history and incidental catches. Doe SCI421SM!47. Int Whal Commn 1990;1-11. 15. Camphuysen C. The harbour porpoise Phocoena phocoena in the southern North Sea. II: a comeback in Dutch coastal waters? Lutra 1994;37:54--61. 16. Read AJ, Gaskin DE. Changes in growth and reproduction of harbour porpoises, Phocoena phocoena, from the Bay of Fundy. Can J Fish Aquat Sci 1990;47:2158-2163.

9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man

A.S. Blix, L. WallCeand t3. Ulltang, editors

465

Migration strategy of southern minke whales to maintain high reproductive rate Hidehiro Kato National Research Institute of Far Seas Fisheries, Orido, Shimizu, Shizuoka, Japan The timing of the southward migration of southern minke whales from their breeding grounds is examined by analyzing length data from 11,953 foetuses recorded in Areas II and IV from 1971/1972 to 1986/1987 by the Japanese whaling expeditions. The few records of lactating females or cows with calves in high latitudes and review of published information suggests that the majority of females conceive while still lactating and that lactating females remain segregated and do not migrate to the Antarctic until after weaning. Although back-calculation of foetal length gives a conception peak in early September, females that conceived in the earlier part of the mating season tended to arrive earlier in the Antarctic than those that conceived later. This appears to be a strategy to maintain a high reproductive rate and may be a density dependent change due to a possible recent increase in carrying capacity. Abstract.

Key words: foetal growth, pregnancy while lactating, conception peak, calving interval, migration pattern

Background While it is reasonable to consider for most balaenopterids that the reproductive cycle is 2 years comprising about a year gestation, a half year lactation and a half year resting period [ 1], minke whales (Balaenoptera acutorostrata) are believed to have a shorter reproductive cycles in both the northern and southern hemispheres [2]. Best [3] considered the mean annual pregnancy rate to be 0.78 or 1.29 years of mean calving interval for southern minke whales; if this is true some specific strategy will be necessary to accommodate a 1 year migration cycle to maintain such a high reproductive rate. Kato and Miyashita [4] examined this feature using data obtained from past Antarctic whaling expeditions; the present study further considers and explores the specific feature of the strategy.

Materials The present study used a total of 11,953 foetal lengths recorded in the Antarctic Areas III and IV by the Japanese commercial whaling expeditions in 1971/1972 to 1986/1987 which took place at high latitudes. Data from both areas were combined

Address for correspondence: National Research Institute of Far Seas Fisheries, 5 - 7 - 1 0 r i d o , Shimizu, Shizuoka 424, Japan

466 without further consideration of the stock boundary. In addition to the body length data, total body weights from Areas III, IV and V in 1978/1979 to 1981/1982 were used to calculate the regression formula for foetal body weight on length. In addition to the standard records collected every year by the technicians, the present study used the condition of the mammary gland of each female which was examined on the deck of the factory ship (Nisshin-maru no. 3) by the author or professional colleagues during the entire period of the operation in 1985/1986. The other data source used was the biological master tape for whales caught between 1971/1972 and 1986/1987 held by the National Research Institute of Far Seas Fisheries, Shimizu. Foetal length was measured to the nearest 1 cm along a straight line from the top of the upper jaw to the notch of the tail flukes (or to the tip of the tail for foetuses smaller than about 12 cm in length). Body weight was measured to the nearest 1 g, 10 g, 100 g and 1 kg for size classes 50 cm, respectively.

Results

Foetal growth curve Model In order to obtain a reasonable model indicating foetal growth of southern minke whales, the formula of Hugget and Widdas [5] was used: Wlt3 = a ( t - to)

(1)

where W is the body weight in grams, a is a growth velocity constant, t is the time in days since conception and to is the intercept where the linear part of the plot, if extrapolated backwards, cuts the time axis. Because of the limited months for fishing and the range of foetal sizes in the minke whale samples, it was not possible to estimate a and to from monthly changes in foetal size composition as noted by Ivashin and Mikhalev [14]. For this study, we have estimated a for southern minke whales by estimating the interspecific relationship among balaenopterids between a and neonatal length. We have accepted the value of to = 74 days proposed for most balaenopterids by Lockyer [1 ] without further consideration. Growth velocity Since balaenopterids have similar gestation periods of 10-11 months, positive correlation can be assumed between neonatal length (Lb) and growth velocity (a). We use this feature here for estimating a. Lockyer [1,6] examined the published foetal growth formula for mysticetes [5,710], and after reviewing previous estimates of a and to by those authors, obtained

467 revised estimates for the southern blue (B. musculus), fin (B. physalus) and sei (B. borealis) whales. Figure 1 shows the correlation between a and ~ (in cm) among the three species (each parameter value is given in Table 1). The fitted regression of a on Lb is expressed as follows, with a high correlation coefficient (r 2 = 0.998):

(2)

a = 0.048 + 0.00067 Lb

Kato [ 11 ] considered 290 cm to be the most appropriate value for the mean length at birth of southern minke whales from reviewing previous studies [3,12-16] and examining the relationship between female maximum lengths and mean neonatal length among several balaenopterids. From eq. (2), this leads to an estimate of a of 0.24.

Weight-length key Using the 867 foetal weight data (Fig. 2) ranging from 1.6 g (4.0 cm in length) to 250 kg (304 cm), the regression between body weight (W, in g) and length (L, in cm) was calculated as W = 0.059L 2-676

(3)

The correlation coefficient between log W and log L is high (r 2 = 0.992).

,•ue

0.6

t--

t~

0.5

o O

0.4

~ O O ,==..

Sei

0.3

Jr"

O t._

I Minke 0.2

O !

I

2

Foetal

1!

I

4

body

I

l

6

length

I

I

I

8

(L b} in c m

Fig. 1. Relationship between the foetal growth velocity (a) and the neonatal body length (Lb) among balaenopterids in the Southern Hemisphere.

468

Table 1. Length distribution of minke whale foetuses in each 10 days, collected from Areas 111 and IV in 1971J1972 to 1986/1987 of Japanese Antarctic operations Length class (cm)

Ten days of month foetus collected; E=early, M=middle, L=late 11E

11M

1lL

12E

12M

12L

1E

1M

1L

2E

2M

2L

3E

3M

3L

0

0

~

0

~

0

~

0

0

0

~

0

0

0

~

0

0

~

0

0

0

0

~

0

0

0

0

0

0

~

0

0

0

0

~

0

0

0

0

0

0

0

0

0

0

0

0

~

0

0

0

0

0

0

0

0

0

0

0

0

0

0

~

0

0

0

~

0

~

0

0

0

0

0

0

0

~

0

~

0

0

0

~

0

0

0

0

~

0

0

0

0

0

0

~

~

0

0

0

469

470

r

v~ o

w = o.os9 L ~ o

0

3

../

"/

f

",~, -.#

o "0 0

eo~,

rn

it

I

!

10

100

Body

length(L)

200 in cm

Fig. 2. Relationship between foetal body length (L) and body weight (W) of southern minke whales in full-logarithmic scale.

Foetal growth formula Using the value of a of 0.24 from eq. (2) and to = 74 days, the foetal growth formula for southern minke whales is W 1/3 = 0 . 2 4 ( t - 74)

(4)

or

t = W1/3/0.24 + 74

(5)

substituting eq. (3) for W, then t = 1.622L ~

+ 74

(6)

However, eq. (4) is not applicable to foetuses in the slow growth stage of early gestation. Although the length (or weight) of a foetus at which the growth rate

471 changes is not known for any cetacean, Laws [17] suggested that change may take place when the placenta is formed in the uterus. From my observation, this phase appears to occur at around 15 cm in body length in southern minke whales. Therefore, we tentatively consider that the growth rate changes at 15 cm (or 82.8 g), at which t is 92 days from eq. (6). Assuming a linear growth rate in body weight between conception (t, W; 0,0) and the change (92, 82.8), the growth for foetuses less than 15 cm in length can be expressed as t = 1.11W

(7)

or t -- 0 . 0 6 5 5 L 2"676

(8)

The overall foetal growth curve is summarized in Fig. 3.

Timing of conception To estimate the date of conception we converted the foetal lengths into ages using eqs. (6) and (8). In the analyses, foetal length data were grouped by 10-cm intervals for foetuses over 19 cm in length (median values were used for the calculations), whereas actual lengths recorded to the nearest 1 cm were used for smaller foetuses.

---~/

300

250 A

E

o

_.e >, '1o o

4) 0 i.!.

200

150

100

50

I

t = 0.06 Lt2"6~6~

I

___......-.---.--,

50

,

,

,

I

100

,

,

i

i

I

150

i

i

i

~

J

200

*

i

i

i.

I

250

.

.

.

.

l

300

9

.

=

I

350

Days since conception (t) Fig. 3. The overall foetal ~rowth curve of southern minke whales established by the present study.

472

Table 2. Incidence of conception by 10 days estimated from foetus length in relation to the month the foetus was sampled: corrected values (Cor.) are from the correction by monthly changes in CPUE, sex ratio, sexual maturity and pregnancy rate

Ten days

Incidence of conception November Crude

December

Cor.

%

January

Crude

Cor.

%

Crude

February %

Crude

Total

March Cor.

%

Crude Cor.

%

Crude

Cor.

%

In previous season 12M 1 0.2

0.1

0

0.0

0.0

0

0.0

0

0.0

0.0

0

0.0

0.0

1

0.2

0.01

In early season 1E 2 2 1L 1 2E I 2M 1 2L 3E 0 3M 0 3L 0 4M 1 4L 0 5E 0 5M 0 5L 0 6E 0 1 6M 6L 14

0.2 0.2 0.1 0.1 0.1 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.1 1.4

4 3 5 7

1.9 1.5 2.4 3.4 0.5 2.4 0.0 0.0 0.0 0.0 0.0 1.0 1.0 2.4 5.8 11.2

0.1 0.1 0.1 0.2 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.3 0.5

0 0 0 1 1 1 1 2 1 1 0 0 1

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.2 0.3 0.8

0 0 0 0 0 I 0 0 2 2 2 2 5 11 21 22

0.0 0.0 0.0 0.0 0.0 0.7 0.0 0.0 1.4 1.4 1.4 1.4 3.4 7.5 14.2 14.9

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.2 0.3 0.7 0.7

0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4 0.0

6 5 6 9 3

2.3 1.0 2.6 4.6 1.7 4.1 1.0 2.0 2.6 2.4 1.4 2.4 5.4 16.9 31.8 54.0

0.03 0.03 0.04 0.06 0.03 0.06 0.02 0.03 0.04 0.03 0.02 0.03 0.07 0.22 0.41 0.69

0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.2 2.9

1

5 0 0 0 0 0 2 2 5 12 23

7

11 25

7

1 2 4 3 2 4 8 23 46 84

7E 7M 7L 8E 8M 8L 9E 9M 9L 1OE 1OM 1OL 11E 11M 11L 12E 12M 12L

12 28 70 120 212 134 14 39 26 46 28 47 51 73 82 0 0 0

In late season 1E 0 0 1M 1L 0 2E 0 2M 0 2L 0 3E 0 Total

1,006

473

474

20

9'..o ~ /',,_ i/ ~ ! I. . i ii ~

15

=,,..A. o.-o o--e ~--zx

November samples December samples January samples Februarysamples

o c o

I

!

., :

10

:.

I

:.o/

".

/i~.1, I

,,o~ '

.~1

9

I

,,"

J,/I

a"

i

1

,A,I

I\

\/

i/

_ _..,ii~"2B

I

\

%-o-t,-o-,, o../\ Xo.~O-~O "--o-a= f O

~

^. - a ,

!

M

J

J

A

S

0

N

D

J

F

Month

Changes in pattern of conception date by month at which foetuses were collected. Incidence of the conception is expressed in 10-day periods obtained from foetal lengths, after correction for monthly changes in CPUE, sex ratio, maturity rate and pregnancy rate in the catch.

Fig. 4.

Table 1 indicates the length distribution of the 872 foetuses by 10-day periods. Conception dates were pooled by 10-day periods. In order to estimate the peak conception date for the population, using foetus samples collected from different months, we obtained the following correlation factors by multiplying the monthly catch per unit effort indices (CSW [ 19]), and the sex ratio, percent mature and pregnancy rate from Kate [16] for Areas III and IV: November,-0.207; December,-0.486; January,-1.000; February,-0.678; March, 0.609. Figure 4 and Table 2 show the incidence of conception by month of collection. Overall there is long-tailed distribution with apparent peaks from late August to early September. However, the apparent conception peaks appear correlated with sample month; the peak of the November samples is in mid-August, of the December samples in late August, and of the January and February samples in mid-September.

Yearly trends in the timing of conception Table 3 examines yearly changes in peak conception dates. The peak of conception in November is fairly constant among years. No trends are apparent for December,

475

Table 3. The peaks of conception expressed by the 10 days of the month calculated from the foetus length in each season of Antarctic operation by the month the foetus was obtained Season

1971/1972 1972/1973 1973/1974 1974/1975 1975/1976 1976/1977 1977/1978 1978/1979 1979/1980 1980/1981 1981/1982 1982/1983 1983/1984 1984/1985 1985/1986 1986/1987

Month foetus obtained a November

December

January

February

8M 8M 8M 8M 8M 8M 8L 8M 8M -

9E 9E 8L 8L 8L-9E 9M 9M 9M 9E-M

9M 9M 9L- 10E 9L 9M-10M 9E-M 9E 10E 9L 9L -

10L 10L 9M 9M

9L 10E

aE, lst-10th; M, 11-20th; L, 21st-30th/31st. January and February, although some fluctuations exist. Insufficient data were available to examine trends for March.

Occurrence of lactating females at the Antarctic The Japanese Antarctic operation took 27,149 female minke whales in seasons from 1971/1972 to 1986/1987, of which only 61 were recorded as lactating (Table 4). Table 4 reveals that the recorded proportions of lactating females to sexually mature females in 1971/1972 (1.42%) and 1985/1986 (0.94%) were relatively high; in these seasons scientists examined all of the whales caught. This suggested that the n u m b e r of lactating females in the other seasons a m o n g mature females in the Antarctic is about 1-1.5%. In 1985/1986, the author and colleagues examined the m a m m a r y glands, uteri and ovaries of all females caught (1,063 individuals) and found 10 lactating females. O f these, five had normal milk and the other five had "late or terminating milk" indicating that they were in the final stage of lactation. Five lactating females were simultaneously pregnant with foetuses of 5 - 1 7 c m in length (Table 5). According to the information from the catcher boats, one had been accompanied by a calf. In addition eight lactating females were taken under the Japanese research catch in 1988/1989, 1989/1990, 1991/1992 and 1992/1993; of those six were simultaneously lactating and pregnant, one was without foetuses and one unknown; none were a c c o m p a n i e d by a calf. Thus 11 simultaneously pregnant and lactating females were confirmed by scientists, and the mean lengths of their foetuses were 11.22 c m (CV 0.46) and 38.5 cm for lactating females with normal milk, respectively.

476

Table 4. Number of females by reproductive status and proportion of lactating females in 1971/1972 to 1986/1987 season of Japanese Antarctic operations Season

1971/1972 1972/1973 1973/1974 1974/1975 1975/1976 1976/1977 1977/1978 1978/1979 1979/1980 1980/1981 1981/1982 1982/1983 1983/1984 1984/1985 1985/1986 1986/1987 Total

Number of females Total

Mature

Pregnant

Unknown

1,942 975 2,597 2,251 1,553 2,276 1,388 1,635 1,327 1,647 1,999 2,140 1,868 1,061 1,063 1,457 27,149

1,643 615 2,031 1,779 1,102 1,834 1,130 1,348 1,090 1,350 1,678 1,845 1,578 991 967 1,362 22,343

1,526 541 1,751 1,620 972 1,602 969 1,181 1,010 1,236 1,563 1,681 1,471 923 907 1,289 20,242

4 11 15 10 18 20 46 21 30 57 64 74 31 5 17 39 461

Lactating 27 5 5 3 4 1 10 3 61

L. %a 1.42 0.52 0.20 0.14 0.25 0.05 0.94 0.21 0.28

aL.%; % of lactating to mature.

Table 5. Biological information of lactating females of the southern minke whale taken from the Antarctic region Date

Body length (m)

Reproductive status a

Foetus size (cm)

Mammary gland status b

T.B. c (cm)

8 Dec. 1985 8 Dec. 1985 9 Dec. 1985 19 Dec. 1985 23 Dec. 1985 24 Dec. 1985 28 Dec. 1985 10 Jan. 1986 10 Jan. 1986 27 Feb. 1986 14 Jan. 1989 11 Dec. 1989 18 Dec. 1989 3 Feb. 1990 1 Jan. 1992 17 Feb. 1992 23 Feb. 1992 3 Jan. 1993

8.9 8.9 8.9 9.3 8.9 8.7 8.8 9.3 8.5 8.9 8.7 9.2 8.8 9.0 9.2 8.6 9.1 8.4

Ovulating Unknown d Pregnant Ovulating Ovulating Pregnant Pregnant Pregnant Resting Pregnant Pregnant Pregnant Pregnant Pregnant Ovulating Pregnant Pregnant Unknown d

?

3.6 3.5 6.7 3.5 3.2 3.9 4.4 4.9 5.0 5.5 9.1 6.0 3.7 5.0 5.0 6.2 5.2 4.4

2.8 3.3 3.1 3.3 3.5 3.2 3.7 3.5 4.2 4.3 3.4 2.8 4.3 3.6 3.4 2.5 3.1 3.2

5 39 15 16 16 17 10 38 14,13 2.8 5.7 ?

L.milk L.milk N.milk L.milk L.milk L.milk N.milk N.milk N.milk N.milk N.milk N.milk L.milk N.milk N.milk N.milk N.milk N.milk

aOvulating means ovulating while lactating and pregnant is pregnant while lactating. bThickness of the mammary gland and types of milk as late (L) and normal (N) milk. CThickness of blubber at the lateral side of the body below the dorsal fin. dUterine horn was damaged by the harpoon.

477 The maximum depth of the mammary glands (Table 5) of the females with normal milk was 4.9-9.1 cm (mean 5.67, CV 0.23). This is within the range (but towards the lower end) of the mammary glands of active lactating females observed off Durban (4-16 cm [3]). The mean depth for females with late milk was 3.2-3.9 cm (mean 3.5, CV 0.24).

Discussion

The present analysis used the Huggett and Widdas [5] growth formula for foetuses over 14 cm in length but in the absence of empirical data we had to assume a simple linear growth rate for the smaller foetus. This is likely to underestimate the age of the smaller foetuses. However, even allowing for this bias, the overall patterns of conception are still acceptable because the magnitude of any bias will be slight. In the absence of empirical data, the lactation of southern minke whales is thought to last from 4 [18] to 6 months [3]. In either case, starting the overall peak of conception in early September revealed by the present analyses, lactation should last

Month ,

/,

M3

j

A ~b

~ b3

gestation(1)~

9-- lactation (1) 9----

?-- gestation(2) "--'-- .~=/~..~,.. 4-'-'~ conception(2)

WINTERGROUND

c•oncep_tion (1)

,

parturition

(1)

)

SUMMERGROUND

A

O~

~~/U

--Q

ICEEDGE

weaning

-Q-

Fig. 5. Schematic illustration of the possible migration pattern of the southern minke whales in relation to reproductive cycle. The double circles mean females and C1 means pregnant females and calf, respectively.

478 from early January to mid-March for most females. However, the percentage of lactating to sexually mature females in the Antarctic catch is only 1-1.5%. This apparent absence of lactating females in the high latitudes is further supported by the low numbers of cow and calf pairs seen during the Antarctic IDCR cruises [20]. It seems, therefore, that females wean their calves before arriving in Antarctic waters, probably because the prevailing oceanographic conditions in the Antarctic are not advantageous for weaned calves. The present study also suggests that most mature females ovulate and conceive while still lactating. This strategy of simultaneous pregnancy and lactation is the most likely way of allowing a short calving interval of about 1 year without breaking the annual migration cycle. It is also common in the Dali's porpoise (Phocoenoides dalli) which has a short calving interval close to 1 year (Gosho and Jones, personal communication; Kasuya, personal communication). In summary, mature females conceive while still lactating in low latitudes, cows and their calves segregate from the other whales, probably in low and middle latitudes, and the females do not move down to high latitudes for feeding until after weaning (Fig. 5). However, this does not mean that females continue to bear calves every year indefinitely. Accepting the annual pregnancy rate of 0.78 by Best [3] would mean that they would miss a pregnancy every 4 years on average. Another feature of their migration pattern revealed by the present analysis is that the peak of conception for November and December foetuses was earlier than that for January and later foetuses. A similar pattern was found in the eastern stock of grey whales [21]. It is unclear, however, whether the females that conceive earlier also leave the feeding grounds earlier. The foetal length distribution of Table 1 provides information on this. Estimates of the length of gestation obtained by regression of mean or modal lengths of foetuses (except those conceived in the previous season) against progress of the feeding season, are always greater than 1 year (up to 18 months) as shown by Masaki [15]. This suggests that females that conceived earlier also leave the feeding ground earlier than those that conceived later. At least two interpretations of the above strategies can be put forward. One is that the minke whale has evolved a 1-year reproductive cycle, particularly given that an annual cycle is seen in both hemispheres, whereas other balaenopterids have 2-year cycles or longer. Another is that the 1-year cycle is a response to changes in carrying capacity. The calving interval is one of several density dependent parameters seen in both terrestrial and marine mammals, and food resources are a major factor in density dependence [22]. As shown by Lockyer [1,6], the energy cost of lactation is generally greater than that of pregnancy and thus simultaneous lactation and pregnancy is energetically expensive and requires sufficient food availability for minke whale populations in the Antarctic, which have increased due to depletion of stocks of the larger baleen whales [23,24]. We believe that the strategy revealed by the present study is probably a result of density dependent changes attributable to an increased carrying capacity.

479

Conclusions 0

2. 3. 0

.

0

A conception peak of southern minke whales exists in early September. The majority of sexually mature females conceive while still lactating. The lactating females segregate and do not migrate to the Antarctic until after weaning. Females that conceived in the earlier parts of the mating season tended to arrive earlier in the Antarctic. Items 2 and 4 above must be specific strategies to maintain short calving interval or high pregnancy rate. This may be a density dependent change due to a recent increase in carrying capacity.

References 1. Lockyer C. Review of baleen whale reproduction and implications for management. Rep Int Whal Commn 1984;(Special Issue 6):27-50. 2. Horwood JW. Biology and Exploitation of the Minke Whales. Boca Raton, FL: CRC Press, 1990; 238 pp. 3. Best PB. Seasonal abundance, feeding, reproduction age and growth in minke whales off Durban (with incidental observations from the Antarctic). Rep Int Whal Commn 1981 ;32:759-786. 4. Kato H, Miyahita T. Migration strategy of southern minke whales in relation to reproductive cycle estimated from foetal length. Rep Int Whal Commn 1991 ;41:363-369. 5. Huggett AStG, Widdas WF. The relationship between mammalian foetal weight and conception age. J Physiol 1951;114:306-317. 6. Lockyer C. Estimation of the energy costs of growth, maintenance and reproduction in the female minke whale (Balaenoptera acutorostrata), from the Southern Hemisphere. Rep Int Whal Commn 1981;31:337-343. 7. Laws RM. Foetal growth rates of whales with special reference to the fin whale, Balaenoptera physalus (L.). Discovery Rep 1959;29:281-308. 8. Frazer JFD, Huggett AStG. Specific foetal growth rates of cetaceans. J Zool London 1973;169:111-126. 9. Frazer JFD, Huggett AStG. Species variations in the foetal growth rates of eutherian mammal. J Zool London 1974;174:481-509. 10. Rice DW. Gestation period and fetal growth of the gray whale. Rep Int Whal Commn 1983;33:539-544. 11. Kato H. Life history of baleen whales, with particular reference to southern minke whales. In: Miyazaki N, Kasuya T (eds), Biology of Marine Mammals. Tokyo: Scientist Inc, 1990;128-150 (in Japanese). 12. Davies JL, Guiler ER. A newborn piked whale in Tasmania. J Mammal 1958;39:593-594. 13. Ohsumi, S. Allomorphosis between length at sexual maturity and body length at birth in Cetacea. J Mammal Soc Jpn 1969;3:3-7. 14. Ivashin MV, Mikhalev YA. To the problem of the prenatal growth of minke whales, Balaenoptera acutorostrata of the Southern Hemisphere of the biology of their reproduction. Rep Int Whal Commn 1978;28:201-205. 15. Masaki Y. Yearly changes of the biological parameters for the Antarctic minke whale. Rep Int Whal Commn 1979;29:375-396.

480 16. Kato H. Year to year changes in biological parameters and population dynamics of southern minke whales. Doctoral thesis, Hokkaido University, 1986;145 pp. (in Japanese). 17. Laws RM. Southern fin whales. Discovery Rep 1966;31:327-486. 18. Williamson GR. Minke whales off Brazil. Sci Rep Whales Res Inst 1975;27:37-59. 19. Free CA. Southern hemisphere minke whales multiplicative regression analysis of Japanese CSW data. Rep Int Whal Commn 1983;33:111-113. 20. Kasamatsu F, Hembree D, Joyce G, Tsunoda L, Rowlett R, Nakano T. Distribution of cetacean sightings in the Antarctic: results obtained from the IWC/IDCR minke whale assessment cruise, 1978/79 to 1983/84. Rep Int Whal Commn 1988;38:449-487. 21. Rice DW, Wolman AA. The Life History and Ecology of the Gray Whale. Special publ 3, The American Society of Mammalogists, 1971; 142 pp. 22. Fowler CW. A review of density dependence in populations of large mammals. In: Genoways H (ed) Current Mammalogy, vol 1. New York: Plenum, 1987;401-441. 23. Kato H. Density dependent changes in growth parameters of the southern minke whale. Sci Rep Whales Res Inst 1987;38:47-73. 24. Kato H, Sakuramoto K. Age at sexual maturity of southern minke whales: a review and some additional analyses. Paper SC/42/SHMill submitted to 42nd annual meeting of IWC/SC, 1990;16 PP.

9 1995 Elsevier Science B.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 0. Ulltang, editors

481

Overview of cetacean life histories: an essay in their evolution Toshio Kasuya National Institute of Research on Far Seas Fisheries, Shimizu, Shizuoka, Japan Abstract. Extant cetacean species exhibit wide variety in body size, habitat choice, reproductive pattern and social structure. Augmentation occurred in several taxa supported by abundant food supply in association with polygyny, seasonal starvation or needs for precocious calves. Fetal growth rate, gestation time and breeding seasonality appear to have adapted for better life time reproductive success under a given environment, although full interpretation is often difficult. Mother and calf is the only stable individual association known to many baleen whales and some toothed whales such as phocoenids, which has further evolved as seen among several delphinids towards association of individuals by age and reproductive status including that of lactating females and also to extended maternal care. The last trait has evolved to a matrilineal school structure of killer whales and probably of long-finned pilot whales for lifetime cooperation of kin of both sexes, while males of sperm whales and perhaps short-finned pilot whales have chosen another mating strategy to fully utilize the female association by moving between nursing schools. The apparent polyandry of Baird's beaked whales could also arise from the same original social structure by assuming paternal investment in calf rearing and kin selec-

tion. Key words: growth, parental investment, polyandry, polygyny, reproduction

Introduction

Life history characteristics of mammals reflect both genetic and environmental differences. The major environmental factors of cetaceans include oceanography and food availability, and could have already been variously modified by human activities including whaling and other fisheries. However, our knowledge on their life history and social structure covers only a short historical period, and is limited to a small number of species. We know almost nothing about the social structure of many species of baleen whales, Ziphiidae, Monodontidae, and Kogia. The present study attempts, using such fragmental information on cetacean life histories, to review their variability and consider the evolutionary significance. This will help to understand the need of management to match characteristics of each species, although this review does not look at cetacean responses to short term environmental changes such as density dependence which is also an important factor of the management of whale stocks.

Address for correspondence: National Institute of Research on Far Seas Fisheries, 5 - 7 - 1 0 r i d o , Shimizu, Shizuoka, 424 Japan.

482 Sources of Information Used

Fetal growth and reproductive seasonality

o

11

o

.

.

.

.

.

Common porpoise Phocoena phocoena: fetal growth is given in Fraser and Huggest [ 1,2], and reproductive seasonality in Mohl-Hansen [3]. Finless porpoise Neophocaena phocaenoides: gestation time is available in Furuta et al. [4] with some uncertainty on conception date. This and assumption that to = 0.135tg [5], give the fetal growth rate. Breeding seasonality and neonatal length are in Shirakihara et al. [6]. Dali's porpoise Phocoenoides dalli: fetal growth is given in Kasuya [7], and reproductive seasonality in Kasuya [7] and Newby [8]. Commerson's dolphin Cephalorhynchus commersonii: average gestation of 345 _+20 days (n = 8) in captivity [9], average neonatal length of 100 cm [10] and to = 0.135tg [5] give fetal growth rate. Spotted dolphin Stenella attenuta and spinner dolphin S. longirostris: no reliable estimate of fetal growth available. Barlow [11] gives reproductive seasonality for each of the populations identified by Perrin et al. [ 12,13]. Bottlenose dolphin Tursiops truncatus: average gestation time is 3 7 0 _ 7 days (n = 77, population not stated) in captivity [9]. This mean neonatal length of 128 cm (WN Pacific) [14] and 117 cm (WN Atlantic) [15], and to = 0.135tg [5] give a range of the fetal growth rate. The fetal growth rate and dates of 96 fetuses (>26 cm) off Japan give the parturition season [ 14]. Short-finned pilot whale Globicephala macrorhynchus: information is available for two populations off Japan [ 16,17]. Long-finned pilot whale G. melas: Martin and Rothery [18] fitted conception curves to the fetal length to obtain likely fetal growth rate. The one used here is based on a single breeding season hypothesis, which gave a slightly slower growth rate than another bimodal model. A single breeding peak is apparent at least in Faroese waters in the occurrence of small fetuses, being the most direct indicator of the breeding season. Killer whale Orcinus orca: gestation of 515 +_7 days (n = 7) in captivity [9], mean of the four neonates from the N Pacific (231-241 cm) and three from the N Atlantic (206-238 cm) [19], and an assumption of to = 0.09tg provide the fetal growth rate. The breeding seasonality is given in Olesiuk et al.

[201. 10.

White whale Delphinapterus leucas: if a linear fetal growth is applied to the fetal lengths in Burns and Seaman [21], the extended fetal growth will cut the axis of time on 1-31 July and reach the mean neonatal length on 1-31 August, which gives a range of t~-to of 242-304 days and fetal growth rate 0.51-0.64 cm/day. If to = 0.135t~ is assumed, 9-12 months is calculated as the time from the start of conception to the mean neonatal length of 155 cm. This period will be followed by a period of body weight increase of variable duration [21 ].

483 11.

12.

13. 14.

15.

Sperm whale Physeter catodon: neonatal length and fetal growth rate are from Fraser and Huggett [1,2], which are within the range of Best et al. [22]. Breeding seasonality is given in Ohsumi [23]. Blue and fin whales Balaenoptera musculus and B. physalus: fetal growth rate is given in Fraser and Huggett [1,2], and reproductive seasonality in Mackintosh and Wheeler [24]. Sei whale B. borealis: the fetal growth rate is given in Fraser and Huggett [ 1,2] and breeding seasonality in Matthews [25]. Bryde's whale B. edeni: fetal growth rate is not available. Reproduction is aseasonal in the Indian Ocean [26] and off South Africa [27] in latitudes 1035~ or weakly seasonal in the South Pacific (10-30~ and North Pacific (20--43~ [26,28]. Minke whale B. acutorostrata: fetal growth rate is given in Fraser and Huggett [1,2]. Breeding seasonality of two populations in the WN Pacific is given in [29-31 ], and that of Antarctic minke whales in [32].

Age composition and age dependent change in reproduction

o

,

,

,

6.

Dall's porpoise: from the Aleutian Islands and Bering Sea stocks in JulyAugust, 1978 and 1980 [8], or during calving and early conception seasons. Thus pregnant females with near term fetuses and females in lactation, which lasts less than 1 year, are combined to obtain the annual pregnancy rate. Striped dolphin: from eight drives off Japan in November-December, 19711977 [33]. Short-finned pilot whale: from the 21 drives of southern form stock off Japan in 1975-1984 [ 17]. Sperm whale: female reproductive status by age is from South Africa in 196567 [22], and age compositions from the Japanese coastal whaling in 1960-1965 nearly randomly sampled by biologists [34]. Baird's beaked whale: from the catch of whaling off Japan in 1975-1987 [35]. Fin whale: data are from Japanese Antarctic whaling in two seasons of 1968/1969 and 1969/1970 in possession of the National Institute of Research on Far Seas Fisheries, Japan.

Seasonality of Calving Mating season and gestation time of mammals have probably evolved to place parturition in such a season that may maximize the survival of the offspring [36]. However, timing of breeding also changes according to the annual change in food availability and climate, and mammal stocks transplanted across the equator can adjust their reproduction to local climate (Fig. 1). The breeding season may also respond to density changes [ 11 ]. Thus the difference in breeding season cannot always

484

Table 1. Fetal growth rate at linear pact of the growth (mmlday) and reproductive seasonality

Species and stocks

P. p h o c o e ~ N. phocaenoides P. dalli C. commersonii S. attenuata S. longirostris S. I. orientalis T. truncarus G. macrorhynchus

G. melas 0. orca D. leucas Physeter catodon 3. m. musculus B. m. brevicauda 3. physalus B. borealis B. edeni

B. acutorostrata M. novaeangliae

Gestation

Baltic Sea Inland S./Pacific Off W. Kyushu OkhotsklPacific Bering/Aleutian S . America N. offshore S. offshore N . whitebelly Inshore Offshore W.NA and W.NP Southern Japan Northern Japan Faroe Is. NP and NA W . Greenland Whole Arctic N . Pacific Antarctic Indian Ocean Antarctic Antarctic Ind. 0.1s. Africa N. and S. Pacific NP/Okhotsk YSlSJlOkhot~k Antarctic Antarctic

Months

Method

10.8 c11.0 11.4

Regression Direct

-

11.3 -

-

-

12.2 14.9

-

11.8 16.9 10.8 9-1 2 15.5 9.5? 11.2 L2.0 -

12.0 -

12.0

-

Regression Direct -

82 Direct Regression -

Model fit Direct Regression Regression Regression Regression Regression Regression -

Regression Regression

Neonatal length (cm) 75 78 -

100 c100 82 77 77 77 117-128 140 c185 177 219-235 155 155 395 700 640

450 270 456

Fetal growth rate 2.81 2.70 3.33 -

3.35 3-5 (1 1-12) -

-

3.74.0 3.40 6.65 4.7-5.0 5.1-6.4 9.01 27.1

-

19.9 17.0 -

10.0 -

17.0

Birth month peak and range

Conception peak and range

617 (6-7) 4 (3-8) 11/12 (8-4) 819 (8-9) 718 (6-10)

718 (7-8) 5 (6-9) l2/ l (9-5) 9 (7-9) 8 (6-9)

Vaguely bimodal 24,8-9 5-8 (2-10) 2-4 (1-12) 6 718 1-2 4-6 (3-10) 1Crl2 (6-5) 3-5 6/7 ( 4 7 + ) 819 (3-2) 5 (4-6) c6, 12 5 (4-9) 7 (1-11) Aseasonal Vaguely seasonal 12 (9-2) 516 (3-8) 718 (1-12) 7-10

-

-

(1-10) 6 (1-10) (2-1) 5 (11-10) 10-1 1 5-7 (4-1 1)

-

5 2-4 4 (10-9) 617 ( 6 9 ) 617 (5-10) 7 (1-1 1) 3/4(14) 10 (7-1) 819 (6-1) 8-1 1

485 I

I

I

I

' I

l

I

I

I

I

I

i

I

-- -?

Wh i t e wha I e

I

I

Killer

whale

Harbor p o r p o i s e

I I,,

. . . . . . .

II

Dali's

p.,

Sanriku

I

Dall's

p.,

Aleutian

I

Long-fin.

I~~l~----~

~ ....

--t

_

--

Short-fin.

p. whale,

N.

' Short-fin.

p. whale,

S.

!

I

I

I I

.... _

I

L

I

I

L

I

-

I

F

p.,

WK dolphin

I

E. s p i n n e r

d.,

Inshore

I E. s p i n n e r

d.,

Offshore

I

t

Whitebelly

I

!

A

I

M

,

J

d. N.

~-- O f f s h o r e

sptd.

d. S.

Sperm whale Minke whale,

_1

I

J

I

A

- Minke whale, I

S

OS

sptd.

I

M

spinner,

Offshore

-

1

I

- Finless

--

I

J

IS/Pac.

-

t

[

p.,

Bottlenose

I I

Finless

I

I

91 I

p. whale

I

0

I.

N

Yell.

Sea

W.N. Pac.

J

D

Fig. 1. Parturition season (line) and its peak months (box) of cetaceans in equatorial and northern waters (roughly in the order of latitude). Determination of the peak season is arbitrary and may not be strictly comparable between stocks.

be evidence of genetic differentiation, although it may suggest at least some degree of genetic isolation. Baleen whales

With the exception of a few species, e.g. Bryde's and bowhead whales, many baleen whales make long distance seasonal migration. They accumulate nutrition in their body during the feeding season in higher latitudes and consume it in the breeding season in lower latitudes [37]. Baleen whales wean the calves during the feeding season following parturition even though some calves may accompany their mothers for nearly 12 months [37,38]. Thus parturition followed by the feeding season offers a good opportunity for calves to switch their nutrition from milk to solid food. Under

486 such an annual cycle, selection could have favored larger body size because it was beneficial for storage of nutrition for reproduction during starvation [39]. The breeding season of normal blue whales that migrate to 60-75~ lasts only 34 months. Using fetal length frequencies in February-March, bimodal breeding has been suggested for pygmy blue whales (B. m. brevicauda) migrating to 40-55~ [40]. The smaller mode (,

o

" "~-

0.01) in 1830-1920, which possibly was caused by the low number of observations in this period.

Discussion and Conclusion Hard exploited whale populations show an increased fecundity by attainment of sexual maturity at an earlier age for the females as observed for spotted and striped dolphins of Japan, sperm whales in the North Pacific and fin whales off Iceland [ 1618]. Further, the male sperm whales in the North Pacific have increased their maximal length caused by better feeding possibilities after depletion of stocks [ 17]. The Faroese pilot whale examination showed that the long-finned pilot whales in the Northeastern Atlantic differed significantly from those in the Northwestern At-

505 lantic [19]. The two populations were proposed to live each in the two gyres of the North Atlantic and separated by the front made by the Mid-Atlantic Ridge in the Irminger Sea. If this is so, the main population for Faroese pilot whaling is distributed in the northeastern part of the North Atlantic and estimated at a size of 778,000 whales (CV =0.295) [20]. This means that the population is exploited with

5000

FARCE ISLANDS

4000

-

3000

-

2000

-

1000

z

4 0 O 0

-

3000

-

2000

-

1000

-

400

ICELAND I

!

NORWAY

-~

300

-

"6

200

-

100

-

500

-

400

-

300

-

I GREENLAND

200100

8 0 0 0

-

6000

-

4000

-

2000

-

40003000

,L ,li_i. I

-

10000

NEWFOUNDLAND

EAST COAST OF USA -

!

20001000 0

i

I 1800

i 1820

1840

1860

1880

1900

1920

1940

1960

1980

2000

Year

Fig. 4. The number of long-finned pilot whales stranded and taken in fisheries in the North Atlantic north of Great Britain in the period long-finned pilot whales [ 2 5 ] .

1800-1994.

The material is sampled by the ICES study group

of

506 a percentage ranging from 0 to 0.6%, and an annual mean take of 0.1% under the condition that the population size on the whole is unchanged. The exploitation of the long-finned pilot whale in the Northeast Atlantic must therefore be considered as negligible for the population and the oscillations observed in the Faroese schools by (a) annual take of whales in the Faroe area, (b) annual AWS, (c) largest whales in the schools, and estimated for (d) the sex ratio over time, and finally, (e) the maturity distribution of the two sexes over time, must be caused by environmental conditions other than the hunting pressure. The peak periods in the number of pilot whales taken in the Faroes (Fig. 1a) occur at the same time as the peaks in abundance of the prey species [3,21 ]. In these peak periods the whales are observed to have a smaller AWS (Fig. l b), especially in the southern part of the Faroes (Fig. 3), a higher male percentage (Fig. lc) [5], and smaller size of the largest whales in the schools (Fig. 2). Moreover, in peak periods there are estimated a larger percentage of immature whales in the schools (Fig. 1d,e). The pilot whales feed mainly on mid-water species of gregarious and luminous squid and feed most commonly at depths between 100 and 500 m and the diet is different according to age, sex and reproductive classes [22]. A difference in composition and occurrence of the main prey species over time could therefore be one possible explanation of the small maximum pilot whale size in peak periods and in the fewer, but larger males in valley periods, which must be considered in future examinations. The smaller AWS observed in the southern part of the Faroes in peak periods (Fig. 3) may be explained by a greater part of immatures in the schools (Fig. l d,e). Differences in school composition can perhaps be explained by variations in the food availability and/or in the migrating pattern according to the current system in the area. Low AWS values are correlated with small school sizes, and this could also indicate the establishment of many smaller schools in peak periods by splitting of the larger schools observed in valley periods. A whale measuring >__14 skinn is a male [5,14], but it should be remarked that for some of the smaller schools in this material the largest whale, measured at 11 skinn, has been a female. This points perhaps to a way of establishing a new school out of an old one and opens the way for further studies of pilot whale statistics [23,24]. The number of catches as well as stranded schools from the whole North Atlantic north of Great Britain show the same occurrence of maximum and minimum numbers as the landed schools in the Faroes, despite the fact that some of the numbers are small (Fig. 4, compiled from Butterworth [25]). A future comparison with oceanographic and biological data may indicate if the long-time oscillations found in the abundance of pilot whales in the Faroese waters as well as the biological parameters in the schools are general for the North Atlantic as a whole.

Acknowledgement Thanks to P.H. Enckell for commenting on this paper.

507

References 1. Bloch D, Desportes G, HCydal K, Jean P. Pilot Whaling in the Faroe Islands. July 1986-July 1988. N Atlantic Stud 1990;2:36-44. 2. Debes HJ. FCroya SCga. T6rshavn: Nort)urlond og FCroyar, 1990. 3. Hr K, Lastein L. Analysis of Faroese catches of pilot whales (1709-1992), in relation to environmental variations. Rep Int Whal Commn 1993;(Special Issue 14):89-106. 4. MUller HC. Whale fishing in the Faroe Isles. Fish and Fisheries. Prize Essays of the Intern. Fisheries Exhibition, Edinburgh, 1882; 1-I 6. 5. Bloch D. Studies on the long-finned pilot whale in the Faroe Islands, 1976-86. Fr6c3skaparrit 1992 (1990);38-39:35-61. 6. Bloch D. Intermale competition in schools of long-finned pilot whales as indicated by abundance of fighting marks. In: Bloch D, Pilot Whales in the North Atlantic. Age, Growth and Social Structure in Faroese Grinds of the Long-Finned Pilot Whale, Globicephala melas. Ph D thesis, the University of Lund, Sweden. 1994;VII:1-16. 7. May RM, Beddington JR. The effect of adult sex ratio and density on the fecundity of sperm whales. Rep Int Whal Commn 1990;(Special Issue 2):213-217. 8. Bloch D, Lastein L. Modeling the school structure of the long-finned pilot whales on the Faroe Islands by use of long-time catch series. In: Bloch D, Pilot Whales in the North Atlantic. Age, Growth and Social Structure in Faroese Grinds of the Long-Finned Pilot Whale, Globicephala melas. Ph D thesis, the University of Lund, Sweden. 1994;VIII:1-21. 9. Bloch D, Desportes G, Mouritsen R, Skaaning S, Stefansson E. An introduction to studies of the ecology and status of the long-finned pilot whale (Globicephala melas) off the Faroe Islands, 1986-1988. Rep Int Whal Commn 1993;(Special Issue 14):1-32. 10. Donovan GP, Lockyer CH, Martin AR (eds). Biology of northern hemisphere pilot whales. Rep Int Whal Commn 1993;(Special Issue 14):1-479. 11. Desportes G, Bloch D, Andersen LW, Mouritsen R. The international research programme on the ecology and status of the long-finned pilot whale off the Faroe Islands: presentation, results and reference. Fr6c3skaparrit 1994 (1992); 40: 9-29. 12. Desportes G, Saboureau M, Lacroix A. Reproductive maturity and seasonality of male long-finned pilot whales, off the Faroe Islands. Rep. int. Whal. Commn 1993;(Special Issue 14):233-262. 13. Martin AR, Rothery P. Reproductive parameters of female long-finned pilot whales (Globicephala melas) around the Faroe Islands. Rep Int Whal Commn 1993;(Special Issue 14):263-304. 14. Bloch D, Zachariassen M. The "skinn" values of pilot whales in the Faroe Islands. An evaluation and a corrective proposal. N Atlantic Stud 1989;1:39-56. 15. Bloch D. Pilot whales in the North Atlantic. Age, growth and social structure in Faroese grinds of the long-finned pilot whale, Globicephala melas. Ph.D. thesis, the University of Lund, Sweden, 1994;1-203. 16. Kasuya T. Effect of exploitation on reproductive parameters of the spotted and striped dolphins off the Pacific coast of Japan. Sci Rep Whales Res Inst, Tokyo. 1985;36:107-138. 17. Kasuya T. Density dependent growth in North Pacific sperm whales. Mar Mammal Sci 1991 ;7:230-257. 18. Lockyer C, Sigurj6nsson J. Rep Int Whal Commn Special meeting on North Atlantic Fin Whales, Reykjavfk 25 February - 1 March 1991. Doc. SC/F9 l/F8, 1991; 1-36 (unpublished). 19. Bloch D, Lastein L. Morphometric segregation of long-finned pilot whales in eastern and western North Atlantic. Ophelia. 1993;38: 55-68. 20. Buckland ST, Bloch D, Cattanach KL, Gunnlaugsson T, HCydal K, Lens S, Sigurj6nsson J. Distribution and abundance of long-finned pilot whales in the North Atlantic, estimated from NASS1987 and NASS-89 data. Rep Int Whal Commn 1993;(Special Issue 14):33-49. 21. Zachariassen P. Pilot whale catches in the Faroe Islands, 1709-1992. Rep Int Whal Commn 1993;(Special Issue 14):69-88.

508 22. Desportes G, Mouritsen R. Preliminary results on the diet of long-finned pilot whales off the Faroe Islands. Rep Int Whal Commn 1993;(Special Issue 14):305-324. 23. Brault S, Smith T. Implications of cohesive pod structures for population studies of cetaceans. Presented to the ICES Study Group Meeting, Copenhagen, August-September 1993. WP-9 1993; 1-4 (unpublished). 24. Smith TD, Read A, Caswell H, Brault S, Barlow J. Puffing pigs and podheads in demographic cyberspace: measuring uncertainty in estimates of rate of increase. Presented to the ICES Study Group Meeting, Copenhagen, August-September 1993. WP- 10 1993; 1-5 (unpublished). 25. Butterworth D (ed). Study Group on Long-Finned Pilot Whales. Report of Meeting, Copenhagen, 30 August-3 September 1993. Int. Counc. Exp. Sea. C.M.1993/N:5. Ref.:A. 1-31.

9 1995 ElsevierScienceB.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 0. Ulltang,editors

509

Harp seals as indicators of the Barents Sea ecosystem Yu. K. Timoshenko Northern Branch of Polar Research Institute of Marine Fisheries and Oceanography (Sev PINRO), Arkhangelsk, Russian Federation Abstract. In 1960-1994, harp seal (Pagophilus groenlandica) population ecology was studied in the White Sea. In the 1980s, substantial changes were revealed in the population ecology, affecting the distribution, migrations, age composition and breeding of the seals. In 1982-1989, unusual harp seal invasions were observed in the estuaries of the rivers flowing into the White Sea (Severnaja Dvina, Koida), movements of seals to the Murmansk coast in summer, and the presence of harp seals in summer and autumn in the White Sea. A sharp reduction in the abundance of younger animals on moulting grounds was also noted, which may indicate high mortality of animals during their first year of life. This trend also characterized the age composition of breeding females on breeding grounds. In the 1980s, harp seal females attained sexual maturity later than normal, and the rates of follicle maturation in ovaries slowed down. Pup weight also decreased. The changes in the ecology of the White Sea harp seal population occurred at a time of profound changes in the Barents Sea ecosystem in the 1980s under the influence of human economic activities. A sharp reduction in fish stocks, particularly polar cod (Boreogadus saida) and capelin (Mallotus villosus), resulted in deterioration of feeding conditions and subsequent changes in the population ecology. The results of the investigations point out the potential of using harp seals as specific indicators of the state of the ecosystem.

Key words: harp seal, distribution, age composition, sexual maturation

Introduction

Marine mammals are an important element of the White and Barents Seas ecosystem. Their significance is conditioned by their role and place in the ecosystem trophic chains. With this in mind, and also the commercial importance of marine mammals, investigations of their biology are being carried out. Great attention is paid to the White Sea population of harp seals ( P a g o p h i l u s g r o e n l a n d i c a ) - an important object of human hunting. More than 30 years of observations of the distribution, migrations, breeding, age-sex composition and other aspects of the harp seal biology have been conducted in the White Sea during the winter-spring periods. The investigations have yielded possibilities to detect changes in the population ecology in recent years. The results of these studies have rarely been reported and have previously been limited to publications on the distribution and migrations of harp seals in the White Sea in 1987 [ 1] and hunting influence on the age composition of breeding females [2].

Address for correspondence: Northern Branch of Polar Research Institute of Marine Fisheries and Oceanography (Sev PINRO), 17 Uritsky Street, 163002 Arkhangelsk, Russian Federation.

510 The harp seal habitat has also undergone substantial changes as a result of human economic activities. Stocks of many fish species (including species comprising the harp seal diet [3-5]) in the Barents Sea have sharply reduced. This paper, therefore, documents the scale of the White Sea harp seal ecology changes and shows how marine mammals can be potentially used as indicators of the ecosystem state.

Materials and Methods

The paper is based on material collected in the White and Barents Seas, mainly in March-May in 1960-1994 during harp seal hunting and but also during other seasons of the year. The harp seal distribution and migrations were studied using airplanes, helicopters, ships, visits to the coast, and direct questioning of local people. Flights were carried out using IL-14, AN-26, AN-30, L-410 planes and MI-8 helicopters. Observations were made by 2-3 observers on the left and right sides of the aircraft at an altitude of 150-300 m at a speed of 200-300 km/h. Age samples were usually collected during the hunt on breeding grounds in the first 10 days of March using helicopters. On moulting grounds, age samples were collected again during the hunt in the third 10 days of April, first 10 days of May from hunting and research ships, and in 1989-1991 (during hunting) from helicopters. Samples characterizing the age composition of breeding females on breeding grounds were collected mainly in the south of the White Sea Basin. On the moulting grounds, samples were collected in the White Sea Voronka, Gorlo and Basin. Neither the hunting nor the collection of material was selective. Therefore, there are reasons to believe that the samples represent the White Sea population. Age was determined by counting layers of dental tissue in the lower jaw canines [6]. Teeth sections 0.11-0.16 mm thick were made and examined under a transmission microscope. Colouring of the harp seal pelages was also taken into account according to an existing classification, as colouring is approximately related to the age of the animals [7]. In total, the ages of 7,106 females obtained on breeding grounds in 1969-1993 and 26,789 males and females obtained on moulting grounds in 1964-1994 are available. Furthermore, 23,819 animals obtained on moulting grounds were grouped according to colouring in 1960-1994. Reproductive organs of females (ovaries) were collected in the third 10 days of April to the first 10 days of May. They were fixed in 4% formalin solution. In the laboratory, ovaries were cut into thin slices and examined visually in order to discover the corpora lutea and follicles. The diameter of large follicles was measured with a ruler. Normally, the largest follicle of the harp seal ovulates, and this allows us to consider its size as a criterion of maturity. Follicles with diameters 10 mm or more were classified as mature based on the fact that corpora lutea ovulationes at early stages are usually of the same size.

511 Females reaching sexual maturity in the current season and breeding for the first time have corpus luteum ovulationis in their ovaries in April-May. Besides corpora lutea ovulationes, corpora lutea lactationes and corpora albicantia were also found at that time in ovaries of females breeding repeatedly. These formations are classified according to colour, thickness, composition and size. Corpora lutea ovulationes are a pale-flesh colour tinged with pink. Corpora lutea lactationes are dark yellow, thicker, and have connective tissue formations. Corpora albicantia are connective tissue infiltrations with luteal tissue remains. Females which have not reached sex maturity do not have these types of corpora lutea in their ovaries. This enables us to group the females into immature, first time breeding and multi-breeding females. The weight of pups was recorded mainly in the second 10 days of March after weaning when they were moulting.

Distribution and Migrations In the 1980s, some unusual peculiarities were revealed in the Barents Sea harp seal distribution and migrations (Fig. 1). During their stay in the White Sea, harp seals appeared in regions in which they usually do not occur. In December 1982, harp seals were observed in the estuary of the river Koida (the Mezen Bay). Their invasion in the river coincided with the mass approach of the gadoid fish navaga (Eleginus navaga). In late December 1982, harp seals were registered in the south of Dvinsky Bay and in the estuary of the river Severnaja Dvina where they penetrated through a channel in the ice made by ships. The animals lay on the ice in groups of 3-5 individuals. Totally, about 50 animals were counted there. Particularly significant deviations from the normal in the harp seal distribution and migrations were observed in the White Sea in 1987. In summer-autumn 1987, individual harp seals or groups of them were repeatedly registered in the coastal regions of Dvinsky Bay, Onega Bay, Kandalakhsha Bay, White Sea Gorlo and Voronka. The animals were very exhausted and behaved unusually in some cases; they did not show their usual caution and they sometimes came out on land. Moreover, the formation of breeding grounds in the White Sea in 1987 was delayed and the density of animals there was low. More information about this is given by Timoshenko [1]. Single animals were also registered in the White Sea Gorlo and Voronka in summerautumn 1988. On June 22, 1985, groups of harp seals consisting of 4-11 animals were discovered in the north-west of the White Sea Voronka and in the Lumbovsky Bay region and near the Cape Svjatoi Nos. Some peculiarities of the harp seal distribution and migrations were also noted in the Barents Sea. On June 22, 1985, groups of harp seals were registered near the Murmansk coast, not far from the estuaries of the rivers Voronja and Varzina. A total of 35 individuals was counted there. On June 24, 1985, 8 animals were observed to the north-west of the Cape Kanin Nos (south of 69~ and 10 animals were observed to the north of the Cape Svjatoi Nos. In March-April 1987, harp seals were registered in the Kolsky Bay, mainly in its northern part.

512 52*

34 ~

36* !

38* ,

40* ,

42*

44*

w

i

BARENTS SEA 69 ~

Q

68*

KOLA

67 ~

66 ~

WWITE S EA 65"

64 ~ I

I

I

I

l

I

.

I

Fig. 1. Localities in the White Sea and in the south of the Barents Sea where harp seals were registered in summer-autumn 1985, 1987 and 1988 ( 9 and winter 1982 (O).

In the 1980s, harp seals were also registered in the coastal regions of the southeastern part of the Barents Sea. For example, in January 1981, 3 adult males were caught (by net) near the entrance to Kolokolkova Bay at the fast ice edge. One of them had 34 navagas of length 18-26 cm in its stomach. The other had navaga otoliths. Single animals were also caught by net in subsequent years.

Age Composition A characteristic peculiarity distinguishing harp seals from other seal species is their change of colouring with respect to age. This can be used in an approximate differentiation of age categories of animals. Thus, results from studies of harp seal composition according to colouring on the White Sea moulting grounds are of some

513 Table 1. Composition of harp seals on moulting grounds in the White Sea according to colouring

Years

Total number of animals

Pelage types Middlings Number

1960 1961 1962 1963 1964 1970 1983 1984 1986 1987 1988 1989 1990 1991 1992 1994

869 2,058 2,307 847 1,508 644 2,526 2,582 763 2,389 1,618 352 1,343 1,473 1,098 1,442

453 1,142 1,143 341 531 419 689 955 178 160 262 17 106 101 99 311

Saddlers % 52.1 55.5 49.5 40.3 35.2 65.1 27.3 37.0 23.3 6.7 16.2 4.8 7.9 6.9 9.0 21.6

Number

%

416 916 1,164 506 977 225 1,837 1,627 585 2,229 1,356 335 1,237 1,372 999 1,131

47.9 44.5 50.5 59.7 64.8 34.9 72.7 63.0 76.7 93.3 83.8 95.2 92.1 93.1 91.0 78.4

Males and females were pooled and the classification was performed in late April to early May, 19601994 in certain years in the period. Middlings are younger animals, saddlers older mature animals.

interest (Table 1). The data show that considerable changes have occurred during the last three decades. In the 1980s and early 1990s, the abundance of middlings (younger animals with "grey-spotted" colouring) on the grounds has reduced gradually, while that of saddlers (older and mature animals with wing-shaped black spots) has increased. Results from studies of the absolute age composition of harp seals caught on the moulting grounds are of special interest, since they probably give the most representative picture of the age composition in the population. Analyses of age samples obtained in the period 1964-1994 demonstrate essential changes in the composition of animals on the moulting grounds (Table 2). First, a pronounced downward trend in abundance of younger animals should be noted. This concerns particularly animals aged from 1 to 4 years old. The number of harp seals from these particular age groups decreased significantly in the 1980s. Simultaneously, the number of older animals, especially those aged 20 years and older, increased. Changes in the age composition of harp seal females congregating in the breeding period in late February to early March have also been revealed. Analyses of age data from the breeding grounds show a reduction in abundance of animals aged 4-9 years from the middle 1970s. Simultaneously some increase in abundance of older harp seals has occurred (Table 3). The data for 1984 do not, however, fit the trend. A factor contributing to this could be that in 1984, material was collected in the third 10 days of March, when the breeding lairs were about to vanish, compared to the first 10 days in other seasons. Moreover, in 1984, females were caught in the north-

514

Table 2. Dynamics of harp seal age composition on moulting grounds in the White Sea Years

Age, years old 1-4

1964 1970 1971 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1994

5-9

10-19

20 and older

No.

%

No.

%

No.

%

No.

%

671 240 140 477 389 601 286 144 111 84 69 81 197 75 92 48 25 40 127 253

28.7 49.2 32.7 48.7 39.3 59.2 20.0 15.1 5.1 8.7 19.4 14.5 14.6 3.4 6.3 2.5 1.5 2.2 7.6 12.7

882 121 146 312 331 218 573 385 687 336 63 223 309 330 492 397 404 496 248 277

37.8 24.8 34.1 31.9 33.4 21.5 40.0 40.5 31.5 35.0 17.7 40.0 22.9 14.9 33.6 20.8 24.1 27.0 14.9 13.9

727 121 124 149 215 151 406 359 977 461 152 197 638 1,254 672 1,056 973 1,028 929 1,055

31.1 24.8 29.0 15.2 21.7 14.9 28.3 37.7 44.9 48.0 42.8 35.3 47.3 56.6 45.9 55.3 58.0 56.0 55.9 52.7

56 6 18 40 55 46 168 63 403 80 71 57 206 557 207 409 277 272 359 415

2.4 1.2 4.2 4.1 5.5 4.5 11.8 6.6 18.5 8.3 20.0 10.2 15.3 25.1 14.1 21.4 16.5 14.8 21.6 20.7


2,336 488 428 978 990 1,016 1,433 951 2,178 961 355 558 1,350 2,216 1,463 1,910 1,679 1,836 1,663 2,000

Males and females were pooled, and samples were taken in late April-early May in certain years during the period 1964-1994.

ern V o r o n k a area, not in the Gorlo or Basin as in the other years. H e n c e , the s a m p l e for 1984 is not directly c o m p a r a b l e with samples for the other years. Thus, the results of investigations for the recent 30 years s h o w that the age c o m position o f harp seals on both m o u l t i n g and b r e e d i n g g r o u n d s in the W h i t e Sea has c h a n g e d substantially. T h e p e r c e n t a g e of y o u n g e r animals has d e c r e a s e d to a minim u m while that of seals of older age has increased, especially on the m o u l t i n g grounds.

R a t e s of F e m a l e Sexual M a t u r a t i o n E x a m i n a t i o n s of ovaries show essential differences in the rates of harp seal f e m a l e sexual maturation f r o m 1 9 6 2 - 1 9 6 4 to 1988 (Table 4). In 1 9 6 2 - 1 9 6 4 , the m i n i m u m age at w h i c h a corpus luteum ovulationis was d i s c o v e r e d in f e m a l e o v a r i e s thus indicating their sexual maturity, was 4 years. In this earliest period, practically all f e m a l e s at age 7 bred. In the samples f r o m 1988, corpus luteum ovulationis w e r e f o u n d in ovaries of females aged 5 years and older. H o w e v e r , the n u m b e r of such y o u n g m a t u r e individuals was low, and very few f e m a l e s y o u n g e r than 7 - 8 bred.

515 Table 3. Dynamics of age composition of breeding harp seal females on breeding grounds in the White Sea in certain years during the period 1969-1993

Years


Age, years old 4-9

1969 1970 1971 1972 1973 1974 1975 1976 1980 1984 1986 1987 1988 1989 1993

10-19

20 and older

No.

%

No.

%

No.

%

144 170 134 119 97 107 156 74 76 333 25 35 53 60 91

30.6 35.9 28.7 25.9 20.3 22.7 32.4 17.1 16.0 35.8 19.5 9.0 11.1 12.7 18.4

280 278 275 274 323 276 246 230 277 462 69 254 329 318 332

59.6 58.6 58.9 59.6 67.4 58.5 51.1 53.0 58.3 49.6 53.9 65.3 68.7 67.4 67.1

46 26 58 67 59 89 79 130 122 136 34 100 97 94 72

9.8 5.5 12.4 14.6 12.3 18.9 16.4 30.0 25.7 14.6 26.6 25.7 20.3 19.9 14.5

T h i s i n d i c a t e s an o b v i o u s t r e n d o f d e l a y in s e x u a l m a t u r a t i o n

470 474 467 460 479 472 481 434 475 931 128 389 479 472 495

of females

in t h e

1980s.

Follicle Maturation in Ovaries O v a r i e s o f l a c t a t i n g f e m a l e s , w h i c h w e r e o b t a i n e d o n b r e e d i n g g r o u n d s in t h e W h i t e S e a in t h e f i r s t 10 d a y s o f M a r c h 1 9 6 2 - 1 9 8 9 , w e r e s t u d i e d . N u m b e r s o f o v a r i e s w i t h

Table 4. Rates of harp seal female sexual maturation based on the results from analysis of ovaries

Age

No. of females 1962-1964 Total

3 4 5 6 7 8 >9

44 62 89 101 95 73 192

1988

Immature

Mature

No.

%

No.

%

44 51 32 9 1 1 5

100.0 82.3 36.0 8.9 1.1 1.4 2.6

-

17.7 64.0 91.1 98.9 98.6 97.4

11 57 92 94 72 187

Total

11 35 52 65 50 29 334

Immature

Mature

No.

%

11 35 51 56 23 8 19

100.0 100.0 98.1 86.2 46.0 27.6 5.7

Material collected in the White Sea, late A p r i l - early May, in 1962-1964 and in 1988.

No. 1 9 27 21 315

%

D

1.9 13.8 54.0 72.4 94.3

516 Table 5. Characteristics of the White Sea harp seal female ovaries based on the availibility of mature follicles (sampling was performed in early March, in several years during the period 1962-1989)

Years

1962 1976 1977 1984 1987 1988 1989

Total number of ovaries (pairs) 312 423 135 121 366 459 442

Ovaries with mature follicles No.

%

190 262 73 59 140 157 364

60.9 61.9 54.1 48.8 38.3 34.2 82.4

mature follicles (>10 m m diameter) were determined. The data (Table 5) indicate that the relative abundance of females with mature follicles in their ovaries in the first 10 days of March fluctuates from year to year. T h e lowest n u m b e r of females with mature follicles in ovaries was registered in 1987-1988.

Weight of Pups Harp seal pups were weighed on the breeding grounds in the White Sea in the second half of March. At that time, the pups have accumulated much h y p o d e r m a l fat and begin to moult. The investigations show that the mean pup weight is not constant from year to year (Table 6). In 1987-1989, the mean weight of the pups was significantly lower than in the other years.

Discussion The distribution area of the White Sea harp seal population covers the W h i t e and Barents Seas. Harp seals do not occur constantly in the White Sea, but the most Table 6. Dynamics of mean body weight (M, in kg) of moulting White Sea harp seal pups by year in the period 1976-1992

Year

Date

No. of animals

M

Weight range

1976 1980 1981 1987 1988 1989 1990 1991 1992

15-20.03 16-21.03 25.03-03.04 19-26.03 19-28.03 23-24.03 17-26.03 17-23.03 12-17.03

78 236 20 151 252 186 731 593 372

36.4 36.3 35.2 32.1 30.6 32.5 34.5 35.6 37.4

20-49 21-52 25-54 18-50 20--44 21-45 20-54 21-58 26-51

517 important stages of their annual life history such as breeding and moulting take place there. The Barents Sea is a feeding area of seals where they feed intensively in autumn and winter. Harp seals spend much time in the open sea and on drifting ice. Another peculiarity of harp seals is their ability to undertake seasonal migrations. They are in the White Sea in December-April, but leave after breeding and moulting in late April to the first half of May. As a rule, they do not occur in the White Sea after the second half of May. Progeny staying in the White Sea after this time in years with anomalous ice and meteorological conditions (extensive ice cover, prevalence of northern winds) is an exception [8]. Seals may appear in the White Sea long before pupping; in December, seal grounds were repeatedly surveyed from airplanes along the edge of the drifting ice in the White Sea Voronka near the Kanin coast. In some years, harp seals may appear in the White Sea even earlier. The invasion of harp seals in the rivers Koida and Severnaja Dvina in winter 1982, and their stay in the White Sea in the summers and autumns of 1985, 1987 and 1988 are probably anomalous phenomena emphasizing substantial disturbance in the migration character of the population. This is confirmed both by the results of our investigations and previous observations of fishermen and hunters living on the coast. It is noteworthy that the movement of harp seals to the river Koida coincided in time with a mass approach of navaga to Koida. This may indicate that harp seals came to the river in search for food. Following navaga, individual harp seals also approached the edge of the fast ice in the south-eastern part of the Barents Sea in the 1980s. Navagas were discovered in the stomachs of some seals obtained there. Why the harp seals appeared in the White Sea in the summer and autumn of 1985, 1987 and 1988 is not completely understood. It has been suggested, that due to unfavourable feeding conditions some seals were not energetically prepared for the long migrations from the White Sea to the northern part of the Barents Sea. This is supported by the registration of individuals in poor condition (skin and fat weight only 31-33% of total weight) on the moulting grounds in late April 1987. Our observations of harp seals near the Murmansk coast in June 1985 agree with other data. Several observations of marine mammals have been conducted near the Murmansk coast, in particular on the Aynov Isles in the period 1963-1989. Harp seals, sick or dead individuals, were found in July 1978 and in May 1985 and 1987 [9]. Similar observations were carried out near the Murmansk coast and the Kharlov Island where harp seals were found in June in 1979 and 1987 (Yury Krasnov, the Kandalakhsha Reserve, personal communication). In other years, harp seals were not observed there. Particularly noteworthy are the deviations in the harp seal migrations in the Barents Sea manifested in their mass invasions to the northern Norwegian coast in 1979-1989 accompanied by mass mortality of animals in fishing gear [1012]. These facts show that there were significant deviations in the harp seal distribution and migration routes in the 1980s in the White and Barents Seas. The results of the harp seal age composition studies on the breeding and moulting grounds are of special interest, with respect to the status and dynamics of the population. Harp seals start to moult in the White Sea in April. In this period seals form

518 dense concentrations on the ice, on the so-called moulting grounds. The age-sex composition of seals on the grounds varies with time. Mature males start to form the moulting grounds. Subsequently, the animal composition on the grounds changes gradually following the appearance of immature animals of both sexes and mature females. In the third 10 days of April and the first 10 days of May males and females of all ages (excluding their young) moult. The overwhelming majority of seals stay on the ice in the White Sea in this period. It is reasonable, therefore, to assume that samples from these grounds obtained in the third 10 days of April to early May may give a representative view of the population composition and its changes. Judging from the investigation results from the moulting and breeding grounds, it seems reasonable to suggest that the White Sea harp seal population age composition underwent considerable changes in the 1980s. However, the data on age composition need correction, taking into account the complicated nature of the moulting ground formation process and possible peculiarities in the distribution of young animals in this period. Thus the data should not be considered as absolute. However, the trend revealed in the population composition should be beyond doubt since it is registered over many years. The sharp reduction in abundance of young animals on the moulting grounds should be considered as a negative symptom, indicating a high mortality of animals during their first year of life. This raises the question as to what caused this high mortality of young animals. Sealing based on the taking of pups during the recent 30 years may have had a certain effect on the state and composition of the population. The intensity of harp seal hunting was high in the 1980s, with a mean annual catch of 68,400 (range 42,900-82,100) in 1980-1989, while in 1970-1979, it was estimated at 40,800 (range 36,300--48,500). However, the present harvest of animals can hardly be considered as the main and only reason for the sharp reduction in the number of young animals in the population. This is supported by several facts. Aerial visual observations, conducted over the White Sea after hunting, have shown the presence of young over a large area. For instance, during a flight on 30 March, 1991, pups were found over practically the whole area of the White Sea Gorlo. Although visual observations can hardly produce an accurate estimate of pup numbers after hunting, it can be considered to give a significant picture of the area of their distribution. Since 1989, the harvest of harp seals in the White and Barents Seas has twice been reduced and now amounts to 40,000 animals per season. But this did not influence the age composition of the population and did not result in any appreciable increase in abundance of young animals on the grounds. The mass tagging of young being carried out in the White Sea since 1989 has revealed long-distance migrations of young animals. The tagged animals were registered in several far-distant regions: on the northern coast of Norway, Spitsbergen and the south-west of Greenland. It is difficult to assess the consequences of such distribution due to the lack of data on tagging from previous years. However, we believe that long migrations contribute to an exchange of animals between the White Sea and the Jan-Mayen populations which may be one of the reasons for the low representation of young animals in some years.

519 Considering the changes in the age composition of females on the breeding grounds, it is necessary to take into account the delay in sexual maturation of females in the 1980s. Along with other reasons, this caused a noticeable reduction in the abundance of females aged 4--9 years in the 1980s compared to the previous 10 years. The changes in rates of follicle maturation in harp seal female ovaries are also worthy of attention. Harp seals copulate in the White Sea in mid-late March [ 13]. By that time, mature follicles able to ovulate (graaphian follicles) should appear in the ovaries. This is an indispensable condition predetermining successful breeding. The reduction in abundance of females with mature follicles in their ovaries by early March in the 1980s should be regarded as a result of the delay in the rate of follicle maturation. The negative consequences of this process are not difficult to foresee; ovulation will occur too late. The study of ovaries collected in the White Sea in late March 1984 support this trend: Most of the females had not ovulated by the end of March in 1984 [14]. Such females have less chance of successful breeding, as it can result in the lack of simultaneous mating and ovulation. As a result, the population's reproductive abilities will be reduced. Animal weight is an important morphophysiological indicator corresponding to animal life conditions and development [15]. Therefore, the decrease in the total body weight of harp seal pups in the 1980s should be regarded as a negative symptom. The changes in the White Sea harp seal population ecology occurred during a period of profound changes in the Barents Sea ecosystem including large reductions in stocks of many fish species on which harp seals feed intensively [3-5]. Particularly, this refers to polar cod (Boreogadus saida) and capelin (Mallotus villosus) which play important roles in harp seal feeding ecology [ 16-19]. The polar cod catch in the Barents Sea in 1966 was estimated at less than 1,000 tonnes but already in 1971 it reached a maximum of 348,000 tonnes [20]. However, the catch was subsequently reduced, and by the late 1970s, polar cod had lost its commercial value. The special role of the polar cod is conditioned by the fact that this frigostable fish is distributed near the ice edge along with harp seals. Moreover, it is widely spread in the Barents Sea where it may form significantly dense concentrations [21,22]. It is therefore not surprising that the harp seal migrations are in many respects correlated with polar cod migrations. The dates and regions of the harp seal autumn-winter migrations to the White Sea coincide in many cases with the pre-spawning migrations of polar cod along the western coast of Novaja Zemlja to the south-east of the Barents Sea [23,24]. Chapsky [25] pointed out that both the migrations and dates of harp seal appearance on the breeding grounds in the White Sea are predetermined by the peculiarities of the polar cod distribution. The fact that polar cod forms pre-spawning concentrations and spawns in the south-eastern part of the Barents Sea in autumnwinter may have made this region particularly important for harp seals. Investigations carried out in the south-eastern part of the Barents Sea in December-February in 1966, 1967 and 1971 from the hunting ship "Chistopol" showed mass occurrence of harp seals in regions near the south-western coast of Novaja Zemlja, the Kolgujev

520

30

45 ~

47 ~

!

49*

i

51"

-1

530

i

SEA

BARENTS

70 ~

0 o

o

o

~

o

o

KOLGUJEV

o

Io

71 ~

69 ~

0

9

i

/-

o o

68*

67 ~

_

I

I_

I

,

I

I

66 ~

Fig. 2. Localities in the south-east of the Barents Sea where harp seals were registered in DecemberFebruary from ships (1971) and airplanes (1985, 1987, 1989 and 1990). 9 large grounds; O groups consisting of tens and hundreds of individuals.

Island and the Kanin peninsula [26,27; unpublished data]. These are the very areas where the most dense pre-spawning and spawning concentrations of polar cod occur [24]. Harp seals were also registered in the south-east of the Barents Sea in subsequent years. Flights carded out in December 1985, 1987, 1989 and 1990 revealed harp seals on the ice in the Cheshskaja Bay, near the northern coast of the Kanin peninsula, and along the southern coast of Kolgujev Island (Fig. 2). Stocks of capelin, another basic fish in the trophic chains of the Barents Sea, have been catastrophically reduced. Large catches of capelin in the 1970s and early 1980s reduced its stock so much that fishing was provisionally stopped in 1987. Record quantities of capelin, 2.5 million tonnes and 2.9 million tonnes, were taken in 1976 and 1977, respectively [28]. Of course, such wide-scale disturbances in the Barents Sea ecosystem could not occur without consequences for the harp seal population. This is indirectly proved by the fact that harp seal invasions to the coast of Norway have been registered since 1978 [11] when the polar cod stocks were significantly reduced. It is therefore possible that the changes in the White Sea harp seal population

521 32 ~

34 ~

36 ~

40 ~

38 ~ I

,42 ~

.44 ~

!

!

!

69*

68 ~ vl

KOLA 67* %

66* I! .....

iv

65*

64* t

i

|

&

I

i

i

!

I

Fig 3. Geographical names of the White Sea regions. I. Kandalaksha Bay; II. Basin; III. Onega Bay" IV. Dvinsky Bay; V. Gorlo; VI. Voronka.

ecology (migrations, age composition and breeding) in the 1980s occurred as a result of the changes in the Barents Sea ecosystem, primarily the changes in the feeding conditions of the species. This may contribute to an explanation of the deviations in the migration routes, the high mortality of young animals, the progeny weight decrease, the delayed sexual maturation of females and the delay in rates of follicle maturation in ovaries. The deterioration of the habitat conditions showed that young animals were the most vulnerable to unfavourable influences of the environment. This study points out the potential of using harp seals as indicators of the Barents Sea ecosystem.

Acknowledgements We are thankful to Yu.I. Nazarenko, V.V. Andrianov, Yu.M. Silinsky, V.V.

522

Sobolev, V.N. Mankov, G.N. Ognotov for their help in the collection and treatment of the material.

References 1. Timoshenko YuK. Osobennosti rasprostranenija I migratsii grenlandskogo tjulenja v Belom more v 1987 g. Ekologia 1992;1:26-33. 2. Yablokov AV, Nazarenko JuI. Poddezhanije optimalnoi vozrastno-polovoi struktury - osnova ratsionalnoi ekspluatatsii populjatsii grenlandskogo tjulenja. Ekologia 1986;2:51-56. 3. Luka GI, Matishov GG, Nizovtsev GP, Orlova EL. K voprosu ob ekologicheskoi osnove postrojenija ratsionalnogo rybolovstva v Barentsevom more. Vsesojuznaja konferentsija po ratsionalnomu ispolzovaniju biologicheskikh resursov okrainnykh I vnutrennikh morei SSSR/Sbalansirovannoje rybolovstvo/. M: 1989;67-73. 4. Glukhov AA. Ekologicheskije problemy Barentseva izuchenija I ratsionalnogo ispolzovanija biologicheskikh resursov okrainnykh I vnutrennikh morei SNG. Materialy vtoroi mezhgosudarstvennoi konferensii. Rostov-na-Donu: 1992;33-35. 5. Drobysheva SS, Matishov GG. Posledstvija antropogennogo narushenija bioticheskogo balansa v Barentsevom more. Problemy izuchenija I ratsionalnogo ispolzovanija biologicheskikh resursov okrainnykh I vnutrennikh morei SNG. Materialy vtoroi mezhgosudarstvennoi Konferentsii. Rostov-na-Donu: 1992;56-57. 6. Yakovenko MI. Opredelenije vozrasta I srokov nastunlenija polovoi zrelosti u Belomorskogo lysuna. Trudy PINRO, Murmansk, 12, 1960;117-118. 7. Chapsky KK. O perekhodnykh tipakh okraski volosjanogo pokrova samtsov grenlandskogo tjulenja. Issledovanija morskikh mlekopitajushchikh. Trudy PINRO, Murmansk, 1967;21:60-79. 8. Timoshenko YuK. Vlijanije ledovykh I meteorologicheskikh uslovii na nekotorykh predstavitelei nastojashchikh tjulenei. Ecologia 1986;3:72-78. 9. Tatarinkova IP, Chemjakin RG. O vstrechakh morskikh mlekopitajushchikh v raione Ainovykh ostrovov (Zapadnyi Murman). Morskije mlekopitajushchije. Tezisy dokladov X vsesojuznogo soveshchanija po izucheniju, okhrane I ratsionalnomu ispolzovaniju morskikh mlekopitajushchikh. M 1990;291-292. 10. Wiig 0. Selinvasjoner til norskekysten. Fiskets Gang 1988;6/7:18-19. 11. Oritsland T. Seals in the northeast Atlantic and Interactions with Fisheries. Comite Arctique International (CAI): Commentary, 2 Feb 1990; 10-13. 12. Haug T, Kroyer AB, Nilssen KT, Ugland KI, Aspholm PE. Harp seal (Phoca groenlandica) invasions in Norwegian coastal waters: age composition and Feeding habits. ICES J mar Sci 1991;48:363-371. 13. Popov LA. O periode sparivanija grenlandskogo tjulenja. Tezisy dokladov pjatogo vsesojuznogo soveshchanija po izucheniju morskikh mlekopitajushchikh. Ch 1, Makhachkala 1972;80-82. 14. Timoshenko YuK. O srokakh ovuljatsii u grenlandskogo tjulenja belomorskoi populatsii. Morskije mlekopitajushchije. Tezisy dokladov IX Vsesojuznogo soveshchanija po izucheniju, okhrane I ratsionalnomu ispolzovaniju morskikh mlekopitajushchikh. Arkhangelsk 1986;381-382. 15. Shvarz SS, Smirnov VC, Dobrinsky LN. Metod morfofiziologicheskikh indikatorov v ekologii nazemnykh pozvonochnykh. Trudy instituta ekologii rastenii I zhivotnykh, AN SSSR, Uralsky Filial, Sverdlovsk 1968;58:387. 16. Smirnov NA. O morskom zverinom promysle na russkikh sudakh. Eksp. dlja nauch.-prom, issl. u beregov Murmana. SPB 1903;157. 17. Sivertsen E. On the biology of the harp seal Phoca groenlandica. Erxl. Investigations carried out in the White Sea 1925-1937. Hvalradets Skr 1941;26:1-166. 18. Lydersen C, Angantyr LA, Wiig O, Oritsland T. Feeding habits of Northeast of Northeast Atlantic

523

19. 20.

21. 22. 23. 24. 25.

26. 27.

28.

harp seals (Phoca grornlandica) along the summer ice edge of the Barents Sea. Can J Fish Aquat Sci 1991;48:2180-2183. Kapel FO, Angantyr LA. Feeding patterns of harp seals (Phoca groenlandica) in coastal waters of West Greenland, with a note on offshore feeding. ICES CM 1989;N:6:1 l+tables. Borkin IV, Ponomarenko VP, Trtjak VL, Shleinik VN. Saika Boreagadus saida (Lepeshin) - ryba poljarnykh morei (zapasy i ispolzovanije). Biologicheskije resursy Arktiki I Antarktiki. M 1987;183-207. Manteifel BP. Saika I jejo promysel. Arkhangelsk 1943;32. Ponomarenko VP. Saika Boreagadus saida V kn.: Promyslovyje biologicheskije resursy severnoi Atlantiki (prilegajushchikh morei Severnogo Ledovitogo okeana. Ch 1, M 1977;309-313. Ponomarenko VP. Osenne-zimneje raspredelenije prednerestovykh I nerestovykh skoplenii saiki (Boreagadus saida) v Barentsevom more. Trudy PINRO, Murmansk, 1963;15:177-197. Ponomarenko VP. Migratsii saiki v Sovetskom sektore Arktiki. Trudy PINRO, Murmansk, 1968;23:500-512. Chapsky KK. Nekotoryje ekologicheskije obosnovanija sezonnoi dinamichnosti areala belomorskoi populjatsii grenlandskogo tjulena. Trudy soveshchanija ikhtiologicheskoi komissii AN SSSR, M, 1961;12:150-162. Beloborodov AG. K osenne-zimney migratsii grenlandskogo tjulenja v Beloje more. Materialy rybokhozjaistvennykh issledovanii Severnogo basseina, vyp XVIII. Murmansk 1971 ;101-106. Timoshenko YuK, Lukin LR. K poznaniju ekologii grenlandskogo tjulenja I morzha v zimny period. Tezisy dokladov konferentsii molodykh uchonykh PINRO po resultatam issledovanii 1971. Murmansk 1972;40-41. Zilanov VK, Luka GI, Ushakov NG. O ratsionalnoi ekspluatatsii zapasov moivy Barentseva morja. Rybnoje khozjaistvo 1984;7:34-37.


Interactions with fisheries


9 1995 ElsevierScienceB.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand O. Ulltang,editors

527

Interactions between marine m a m m a l s and fisheries: an unresolved problem for fisheries research Tim D. Smith Northeast Fisheries Science Center, Woods Hole, Massachusetts, USA Abstract. Food web interactions between marine mammals and commercial fisheries have been subject to relatively little attention in fishery research. Three characteristics of marine mammals, commercial fisheries, and the conduct of fisheries research have contributed to this. These include differences between marine mammals and fishermen in prey size capability and ability to adapt to rapid changes, major undocumented ecological changes which have occurred historically making it difficult to interpret present ecosystem states, and a short term focus which was adopted by fishery research during its development towards the direct effects of harvesting. These characteristics need to be considered in designing and implementing research programs to investigate interactions between marine mammals and fisheries. Key words: ecology, predation, fishery biology

An Early Morning Introduction It is 4:30 in the morning, and still no revelation" why have fishery biologists not addressed the question of the food web interactions between marine mammals and fisheries? From around my upraised mug of morning tea I glimpse the cobweb in the ceiling comer, vowing again to sweep it down. But something moves; emerging cautiously from the shadow is Lepsima saccharina, the silverfish. It is encouraging at this hour that it is not a spider; at least it has the same name as my quandary" "fish". Awake at this hour out of desperation for an idea I call to it encouragingly, hoping for insight: "Bonjour M. petit poisson d'argent, comment s'va?" But unfortunately it speaks only Norwegian, where it is known as sr "poor silver creature." So it is not a fish after all; my hopes for insight fade. The sr moves diagonally down the wall towards my book shelf, irritatingly bold for so poor a creature. I become the predator, poised. It jumps, and lands on a dark blue book, dusty and old; I strike; it disappears into the book's spine; the book falls to my desk. Animal Ecology is the title, written by Charles Elton in 1927 [ 1]. I understand that sr prefer the glue used in older books, and this may be so because I cannot shake him out of the spine. Instead, I start paging through Elton's version of ecology. He has a lot to say about food web interactions, discussing the size of prey and the requirements of predators. The best size prey, he thinks, is one which a predator can capture and ingest efficiently and safely while being able to catch sufficient numbers

Address for correspondence: Northeast Fisheries Science Center, Woods Hole, MA 02543, USA. Tel: +1 508 548 5124; Fax: +1 508 548 5124; E-mail: [email protected]

528 to meet physiological requirements. I continue to read; it is at least a starting point in this morning's dim light.

Sizes of Predators and Prey The killer whale and the blue whale, Elton notes, represent extremes for large predators; one is specialized to the unusual habit of eating prey larger than itself and the other to eating prey far, far smaller than itself. As Tiu Simil~i and Anna Bisther showed in their presentations to this Symposium, however, we have since learned that killer whales along the coast of Norway also feed on much smaller prey, regularly herding and catching Atlantic herring. But killer whales do not appear as efficient at catching small prey as are the more specialized baleen whales, and must resort to stunning the tightly schooled herring and eating them one at a time. One may wonder if killer whales learned to feed on herring after the whale fisheries in the North Atlantic depleted the numbers of baleen whales, and also, if baleen whales were more abundant today, would killer whales chose to compete with or prey on them. Like killer whales and in contrast to most animals, man as a predator is not limited to certain sizes of prey. We harvest the largest animal, the blue whale, its predator, the killer whale, and now even the blue whale's prey, krill. In this respect, Elton argues, man is unique, although this has been so only "in the later part of his history." Were animals other than man not so constrained, Elton muses, there would be much less "variety and specialisation," and food chains as we see them would not exist. But food chains do exist, linking together, as Elton noted, into food webs. Despite our ability to harvest these chains and webs at any and seemingly all levels, man appears to have ignored changes in food chain dynamics in studying the effects of fishing [2]. Fishermen are animals, and as such are subject to the same evolutionary processes as their prey. But fisheries change as the economics, technology, and levels of prey abundance change, and these changes are much more rapid and radical than those due to biological evolution. For example, between 1880 and 1900 the English sailpowered beam trawler was almost completely replaced by the much more efficient and larger steam-powered otter trawler. The increase in efficiency was substantial, as calculated by Garstang in 1900 [3], perhaps of the order of 4--8 times. The fishing fleet changed markedly, and with fewer but more efficient boats, it roughly doubled its predatory power during these 20 years. The fish and their ecological communities also began to change, although at lesser rates. Abundance of target fish species had fallen by 1990 to half of its 1880 level under the greater predation by fishermen, while the abundance of some nontarget species had increased. The rate of change of fisheries has continuously increased, and although some "species" of fisherman have become extinct in the process, the diversity of types of fishing today is enormous. The English language has adapted by generalizing the word fishing to include the harvest of nearly all aquatic and marine organisms. The

529 change that the word has encompassed is illustrated by the variety of harvesters in the Gulf of Maine who have been and are referred to as fishermen: this ranges from the whalemen of a century ago, to those harvesting pelagic species such as Atlantic mackerel and herring, to the Atlantic halibut fishermen of the turn of the century, to the groundfish trawlers of this century, and now even to those who now have learned to harvest sea cucumbers, sea urchins, and seaweed in their quest for survival. Metamorphoses of fisheries over ranges of species such as this have been common around the world, frequently driven by successive overexploitation of one species after another. Such changes as these are facilitated by the increasing diversity of methods of catching fish used on vessels ranging from the size of a man to the size of a playing field, the latter frequently able to reduce prey handling time by on-board processing catches. At least as significant, the sensory capabilities of fishermen, formerly limited to blind groping with nets and hooks, now greatly exceed the capabilities of most of their prey. Fishermen and fisheries are constantly changing to meet continuously changing economic conditions by modifying vessels and equipment, and adopting technological advances. If the abundance of a target species decreases, due to too much fishing or to environmental change, fishermen evolve their methods to become ever more efficient [4]. If changes occur too rapidly, fishermen must radically alter their harvesting strategy and methods to catch different prey, that is they must metamorphose rather than evolve. These changes generally improve overall predatory effectiveness, but because change is the normal course of fisheries, the fishermen must also maintain or increase their capabilities for yet further changes.

The Present State of Ecosystems While Elton did not follow up on the nature of fisheries, he did describe some of their effects. The "doom of the whales" in the Arctic was sealed, he suggested, with the development of Dutch and English whaling in the 16th century. He shows a photograph of the skulls of Atlantic walrus on Moffen Island, just north of Spitsbergen, slaughtered for their ivory. The fate of that fishery, and now more importantly, the prospects for recovery of that population, is reviewed in a poster by Oystein Wiig and Ian Gjertz at this Symposium. Elton also described the rapid recovery of the Alaska fur seal from its fishing-induced near extinction in 1914. He lamented the senselessness that "the capital of animal numbers is destroyed to make the fortune of a few men," and the unfairness "that all possible benefits for any one coming later are lost." But despite his lament, he was more interested in using these and other examples to illustrate the point that: Living as we do in a world which has largely been denuded of all the large and interesting wild animals, we are usually denied the chance of seeing very big animals in very big numbers. Therefore, he continues:

530 the conditions under which the present fauna has evolved are ... rather different from what one might expect from seeing the world in its present state. Elton did not think that all harvesting had grossly distorted the ecosystem. For example, in the summer around Spitsbergen, he noted, Norwegians were then pursuing a simultaneous hunt for the bearded seal, Erignathus barbatus or storkobbe, and its predator, the polar bear. He observed that despite the harvest, the seals were "more abundant than ever", and suggested that this was due to harvesting the predator along with its prey. Although he recognized that such an outcome depended on the numbers of both species harvested, he thought a "balance happens to have been struck" at Spitsbergen. Nonetheless, such a balance entails a change in the ecosystem; in this case a lower abundance of polar bears, and presumably a higher abundance of Norwegians. More generally, the theory of harvesting outlined by Hjort in 1933 [5], drawing on the work of the mathematical ecologists in the 1920s, emphasized that the major effect of harvesting would be a decrease in the abundance of the target species to some lower stable level as the major effect of harvesting. While the theory allowed for sustainable catches, it was apparent to Hjort that a balance had not been struck for the North Atlantic whale fisheries. These fisheries had decreased the abundance of whales to the point that the fleets had to continually move to new grounds. Since the turn of the century, Foyn's fishing vessels had been excluded first from the Barents Sea grounds, then from Icelandic waters, and by 1933 were returning large profits from the Antarctic grounds. In the face of the large and rapid declines in the species targeted by many fisheries, the changes wrought on the rest of the ecosystem by harvesting were not explicitly addressed when biologists began to develop management advice. For example, Hjort focused primarily on developing methods of measuring the biological significance of the repeated pattern of whaling. Similarly, in applying Hjort's approach, Graham in 1935, Schaefer in 1954 [7], and Chapman in 1964 [2] also focused on the direct effects. Thus while harvesting indirectly changes food web relationships, almost always only the often overwhelming direct effect of reducing the target species has been studied; indirect effects have been ignored.

Fishery Biologists Didn't! Elton was not alone in the 1920s in his interest in the nature of the interactions among predators and prey in fisheries. However, this interest was not shared within the field of fishery biology as it developed in the latter 19th century and the first half of the 20th century. Lankester had argued forcefully in 1883 that the effect of fishing can only be understood by considering the effect of the increased predation by fishermen on the equilibrium between natural predators and their prey: the thousands of apparently superfluous young produced by fishes are not really superfluous, but have a perfectly definite place in the complex interactions of the living beings within their area.

531 Lankester's perspective was not taken up in fishery biology, and there was little study of these effects, even though they have been described for centuries. For example, in response to concerns in France about the effect of dolphins on fisheries, a Papal Decree was issued in 1587 "anathematizing this vermin" [2]. Predator control bounties have arisen as a part of fishery management periodically for centuries. An example close to my laboratory was when the town of Wellfleet on Cape Cod in Massachusetts in 1740 began paying a bounty for "porpoise tails" [8]. Between 1740 and 1742 one person was paid the bounty 500 times. In France the government attempted to implement the earlier Papal Decree in the 1880s in the form of bounties for dolphins [9], but apparently with little success. The calls for renewed control of predators were regular by the turn of the century, but with modest scientific bases. For example, in 1887 the German Fishery Association argued for a "premium on seals" based on the observation that their new estimates of the productivity of fishing grounds, which included seals, was less than half that of the protected fish ponds [10]. A similar conclusion was reached in 1889 by the US Fish Commissioner, Spencer Baird, for porpoises and elasmobranchs. He argued from common knowledge, noting that "the extent of destruction to fish caused by the porpoises, skates, and dogfish is well known" and endorsed a plan for a factory in the village of Woods Hole to make agricultural fertilizer from predatory fish [ 11 ]. This, he argued, would have "a marked influence upon the supply of edible fishes." Whether it did so was not measured, although it did have a direct effect on the economy, as well as the smell, of the village of Woods Hole. Despite the wider interest in the effects of predation, most fishery research has focused on the direct and most apparent effects of fishing. For example, the analyses conducted by scientists of the International Council for the Exploration of the Seas on the higher catch rates in the North Sea after the First World War were confined to such single species effects [ 12]. The ICES analyses were criticized [13,14] by some fishery biologists [13,14] for their narrow scope, and additionally the effects of the war years were seen very differently in other fisheries. For example, in the Adriatic Sea, D'Anconna collected landings data which suggested a higher proportion of predatory fish were in the markets when the overall catches were lower during the war years than before or after the war [ 15]. Based on D'Anconna's observation, Volterra, his father-in-law, developed his well-known predator-prey models in 1926 [15]. While this work fueled a minor controversy in the developing field of mathematical ecology [16], it was generally overlooked by fishery biologists despite a translation of Volterra's paper having been published in one of the prominent fishery science journals of the day. One reason predation and other ecological roles have historically not been subject to widespread study in fishery research is, I feel, the commensal relationship between the biologist and the fishermen. Fishery biology is a discipline which continues to evolve to fill a specialized niche defined by the needs of fishermen, and to coevolve with another commensal, the fisheries manager. The needs that the biologists are called on to meet are ever changing as the "bionomic ecosystem" of Gordon [ 17] evolves under the continued and seemingly ever mounting pressure of fishing.

532 Historically, the fishery biologist tended to see the world from the continually changing point of view of the fishermen. This was literally true when fisherydependent data were the primary source of information, with the fishery biologist observing from the bow, or perhaps stem, of the fishing boats. As such, direct observations of predation by marine mammals are more likely to be made during fishing operations than at other times. These observations may not be representative but are often the source of information underlying complaints by fishermen [ 18,19]. The constraints on the biologist's perspective also change over time, with redefinition of what is important occurring as fisheries change their target species and spatial distribution. For example, in New England the demise of the Atlantic halibut fishery around the turn of the century was followed by an expansion of a fishery on haddock, one of its prey. The questions shifted from the dynamics of predators near the top of a food chain to the dynamics of predators which themselves support several predators. That Georges Bank had once been replete with Atlantic halibut has seemingly been forgotten by both fishermen and biologists. Similarly, we are now asking the question if minke whales or other predators have moved into the niche of the Antarctic blue whales. But, with the ending of the blue whale fishery, the emphasis of fishermen and biologists shifted away from the largest of animals to the next most valuable whale species, the fin whale. The data needed to describe the blue whale's niche was not collected, and now may not be collectable. From his niche within the bionomic ecosystem defined by human predation, the fishery biologist has tended to be myopic. One form this short-sightedness takes is that the factors governing the evolution and metamorphoses of the fishermen in response to changing prey availability and economics frequently have not been observed. A second form is that study has often been limited to the immediate area and season of harvesting, an especially constraining practice when considering the food web interactions with marine mammals, with their frequently large annual movements. Thirdly, there has been a tendency to focus on those components of the bionomic system which most affect the short term success of the fishermen, namely the abundance of their prey. Some of this myopia was firmly embedded in fishery research programs by 1955. In his paper to the United Nations' Rome Conference that year Schaefer [20] outlined how he felt fishery research programs should be structured. There he disallowed a place for the economist, and relegated predation to the most remote comer of the third (meaning least needful of attention) level of his diagrammatic prescription. In the 40 years since the Rome conference, the scope of fishery biology has been broadened substantially, frequently at the level of the fishery biologist and at times in the context of fishery management. Fishery biologists have become less dependent on the bow of the fishing vessel as they have developed fishery independent resource surveys, which have been used to monitor target species abundance, and in some places to monitor both abundance and prey abundance in multispecies groundfish assemblages. Long term programs have been developed to monitor other components of the ecosystem, especially zoo-

533 plankton. Fishery biologists have developed and explored multispecies models to help interpret this information, and some perspectives from it have at time affected the fishery managers. In my own region, for example, international management at one time adopted a "two-tiered quota" system in recognition of interactions among the several harvested groundfish species. The historic concerns about the predatory relationships of marine mammals and fishery resources, however, have reemerged in recent years. Questions about the effect of these predators on fishery resources that were not addressed historically still cannot be answered, even with expansion of the scope of the fishery biologist to include multispecies interactions. An important and I feel unresolved problem is how to develop research programs, including both data collection and methods of analysis, which will allow resource management to address these issues on a scientific basis.

"A New but Undocumented Fishery" In this Symposium many of the papers presented have focused on prey consumption, including estimates of rates of consumption for several pinnipeds and for minke whales. David Lavigne argued in his contribution that further improving our understanding of consumption rates is now far less important for management than improving our understanding of the dynamics of food web interactions. However, despite interest among managers and prompting from other disciplines, fishery biology has not addressed this aspect of the effects of fishing, and has not developed needed research methods. Harwood [21 ] suggested that progress might be made "if we were to consider the competing marine mammal as if it were a new, but undocumented, fishery, and to review what information about its operation is required to evaluate the threat it poses to existing fisheries". Under this approach the tools of the fishery biologist might come into play more naturally as fishery biologists become commensal with the marine mammal rather than, or better in addition to, the fishermen. Harwood's relegation of the marine mammal to the role of the "new fishery" reflects the lack of emphasis given to such studies previously, but is jarring given that the niches of marine mammal predators have been defined over much longer time frames than those of the fisheries. How best to organize fisheries research has been debated since at least the 1890s [2]; many forms have been tried and in most places the model adopted has been the government supported fisheries research laboratory. The question of how best to organize research relating to marine mammals, both as a target of and as by-catch in fisheries has also been much discussed. In the United States, the Marine Mammal Protection Act in its several revisions has resulted in an expanded responsibility for scientists in the fisheries laboratories of the National Marine Fisheries Service. While research by scientists outside the government has also been supported, the nature of

534 most research related to marine mammals has been defined by biologists within the government. Norway has taken a different approach to address its interests, and we are seeing the results of their approach in this Symposium culminating the Norwegian Marine Mammal Research Program, begun in 1989. As Lars WallCe indicated, the Norwegian strategy has been to utilize the expertise and methodology of fishery biology, through the Norwegian Marine Research Institute, and to utilize other disciplines through various universities and research agencies. However, in contrast to the US approach, selection of specific projects for funding was determined external to the government fisheries laboratories. The two approaches have resulted in a different mix of research activities. Some of the differences reflect the different objectives and interests within Norway and the US. Others of these differences are due to the disciplinary filter used in the US in deciding the research to be funded. For example, national priorities have focused our research on measuring the direct effects of the by-catch of marine mammals, and have generally given lower emphasis to measuring the effects of rapidly expanding pinniped populations on their fished prey populations. In contrast, in Norway a broader range of research has been supported. The two approaches have different strengths and weaknesses. The US approach may be too narrowly focused, resulting in key gaps in information. The Norway approach may suffer from too limited integration among the broader scope of research projects. This latter was reflected, for example, in the need for additional simultaneous measurement of prey species distribution and availability (i.e. the specialty of the fishery biologist) to complement the studies on the prey consumption of whales and seals (i.e. the specialty of the academic biologist.) Similarly, some Symposium descriptions of now completed research programs included specific reference to the next step being the evaluation of the management significance of the results, again the province of the fishery biologists. Such evaluation would be more effective, however, as an integral part of the overall research program.

Conclusions

The wisdom and utility of the approaches taken in the US and Norway towards understanding the food web interactions between marine mammals and fisheries will ultimately be evaluated by history; this Symposium has established some of the facts for such judgements to be made. In the meantime, some guidelines emerge. The first is that the full complexity of food webs cannot be directly approached; rather, following the lead of the Benguela Ecology Program [ 18], we should strive to determine the minimal complexity required in models of food webs to provide sound, scientifically justifiable management advice. Secondly, improved data collection methods must be developed for measuring predation (for example fatty acid and stable isotope profiles of predators and prey.) Lastly, the nature and timing of shifts in food web relationships need to be determined. In this Symposium, for example,

535 T o r e H a u g suggested that minke whales m a y shift a m o n g prey on an annual basis, whereas J o h a n n Sigurjonsson noted that fin whales may u n d e r g o decadal c h a n g e s , possibly driven by environmental variability. T h e s e time frames m u s t be c o n s i d e r e d in designing and i m p l e m e n t i n g research p r o g r a m s to obtain long t e r m observations necessary to understand the ecological roles of marine m a m m a l s . In i m p l e m e n t i n g these guidelines, robust institutional structures for research are needed, structures far different f r o m those which are now in place. G i v e n rapidly developing c o m m u n i c a t i o n s abilities, these structures m a y usefully be m o r e than just physical buildings and scientists in specific locations. Further, their overall operation will have to account for both regional and country-wide priorities. Strong effective research which will yield increased understanding of these c o m p l e x interactions is dependent on d e v e l o p m e n t of suitable research institutions, and their structure should be discussed as a matter of urgency if the goal of understanding the interactions between m a r i n e m a m m a l s and fisheries is to be met.

References 1. Elton CS. Animal Ecology. New York: Macmillian, 1927. 2. Smith TD. Scaling fisheries: the science of measuring the effects of fishing, 1855-1955. Cambridge: Cambridge University Press, 1994. 3. Garstang W. The impoverishment of the s e a - a critical summary of the experimental and statistical evidence bearing upon the alleged depletion of the trawling grounds. J Mar Biol Assoc 1900;6:1-69. 4. Graham M. The Fish Gate. London: Farber, 1943. 5. Hjort J. Whales and whaling. Hvalradets Skr 1933;7:7-29. 6. Graham M. Modern theory of exploiting a fishery, and application to North Sea trawling. J Conseil 1937;10:264-274. 7. Schaefer MB. Some aspects of the dynamics of populations important to the management of the commercial marine fisheries. Inter-Am Tropical Tuna Commn Bull 1953;1:27-56. 8. Kittredge HC. Cape Cod: Its People and Their History. Boston: Houghton Mifflin, 1930. 9. Anon. Notes. Nature 1889;40:401-402. 10. Anon. Notes. Nature 1887;35:374-377. 11. Baird SF. The sea fisheries of eastern North America. Report of the United States Fish Commission. 1889; 14(Appendix): 1-224. 12. Borely JO, Russell ES, Graham MB, Wallace W, Thursby-Pellam DE. The plaice fishery and the war: preliminary report on investigations. Ministry of Agriculture, Fisheries and Food (UK), Fisheries Investigations (Series 2), 1923;5. 13. Peterson CGJ. On the stock of plaice and the plaice fisheries in different waters. Report of the Danish Biological Station, 1922;29. 14. Garstang W. Plaice in the North Sea - changes in size of catch. The Times (London) 1926;21 April: 15, and 26 April:20. 15. Volterra V. Variations and fluctuations of the number of individuals in animal species living together. (In Italian, translator Wells ME) J Conseil 1926;3:1-51. 16. Kingsland SE. Modeling nature: episodes in the history of population ecology. Chicago: University of Chicago Press, 1985. 17. Gordon HS. An economic approach to the optimum utilization of fishery resources. J Fish Res Bd Can 1953;10:442-457.

536 18. Anon. Report of the Benguela Ecology Program workshop on seal-fishery interactions. Reports Benguela Ecology Program, South Africa, 1991 ;22:65. 19. Anon. Marine Mammal/Fishery Interactions: analysis of cull proposals. Report of the Meeting of the Scientific Advisory Committee of the Marine Mammals Action Plan, Liege 27. United Nations Environment Programme, P.O. Box 30552, Nairobi, Kenya, 1992. 20. Schaefer MB. The scientific basis for a conservation program. In: Papers Presented at the International Technical Conference on the Conservation of the Living Resources of the Sea. Rome: United Nations Publication 1956.II.B. 1, 1956; 194-221. 21. Harwood J. Assessing the competitive effects of marine mammal predation on commercial fisheries. S Afr J Mar Sci 1992;12:689-693.

9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCe and 13. Ulltang, editors

537

Strategies to reduce the incidental capture of marine mammals and other species in fisheries Martin A. Hall Inter-American Tropical Tuna Commission, La Jolla, California, USA Abstract. A brief analysis is made of the strategies that can be used to reduce the bycatches of marine mammals and other species in fisheries. They all fall under two basic types: reduction of the level of effort and reduction of the average bycatch per unit of effort (BPUE). The former frequently results in lower catches of the target species. Reduction in the BPUE, on the other hand, may offer a way to mitigate the problems with fewer negative impacts on the fisheries. Identifying the environmental, biological and technological reasons why bycatches happen is the key point of those strategies that attempt to deal with the problems while at the same time maintaining the use of the resources involved. Five "lines of defense" are identified to try to mitigate or solve bycatch problems. The Tuna-Dolphin Program of the Inter-American Tropical Tuna Commission is used as a case study to illustrate different issues. Finally, some of the conditions that have helped solve this problem are presented. Even though it is clear that each fishery will have to develop its own set of solutions, there are some common traits that may help in the search for solutions.

Key words: bycatch, dolphin, tuna

Introduction Marine mammal bycatches in fisheries have become a very important, if not the dominant, factor in the management of some fisheries. These bycatches range from very rare events, to very large mortalities; from serious conservation threats to negligible impacts on populations [1-3]. In some cases, bycatches are a problem because they affect an endangered species; in other cases, the level of the bycatch is not sustainable. In all cases, the popular perception of the marine mammals in some cultures gives visibility to these situations. Although the emphasis in this paper is on marine mammals, the basic concepts and strategies discussed in this document apply also to the bycatches of other species. The main focus of this paper is on a general characterization of the strategies that can be used to mitigate the problem of bycatches. The different approaches will be illustrated, where appropriate, with examples from the eastern Pacific tuna-dolphin problem, where an international bycatch reduction program has been in place for decades [4,5]. In this fishery, dolphins of several species (Stenella attenuata, S. longirostris, and Delphinus delphis) are entangled in the purse-seine nets used to catch yellowfin tuna, Thunnus albacares.

Address for correspondence." Inter-American Tropical Tuna Commission, 8604 La Jolla Shores Dr., La Jolla, CA 92037, USA

538 In recent years, that program has generated an international agreement that has succeeded in accelerating the efforts to mitigate the problem. In the last 7 years, dolphin mortality has been reduced by 97% [6], while at the same time the fishery has continued operating quite successfully, showing that the two goals were not incompatible.

Some Basic Strategies to Mitigate Bycatch Problems The basic formula to estimate the total bycatch of a given species, caused by a given gear is a good starting point to visualize the strategies that can be used to reduce it: Total bycatch = total effort x bycatch per unit of effort To reduce the total bycatch there are two options: (1) trying to reduce the total effort, or (2) trying to reduce the bycatch per unit of effort (BPUE). Of course, both options can be pursued simultaneously.

Reducing total effort Banning effort or limiting the level of effort This can be done directly, as a regulation promulgated by one or more govemments (e.g. the recent ban on the use of drift gillnets on the high seas) or, indirectly, through the use of economic forces (demand, prices, etc.). Embargoes, consumer campaigns, boycotts, and tariffs can be applied towards achieving this goal (e.g. recent US embargoes on tunas [5] and shrimp related to bycatch problems). These options are open to governments, but also to industries, advocate groups, etc. As a result of these actions, demand for a product may drop, markets may close, prices may decrease, etc., and fishing effort should be affected by those forces. The effectiveness of these measures will depend on the viability of the enforcement and control mechanisms, public response, etc.

Setting limits on the bycatch levels allowed If a bycatch limit is imposed at a level lower than the current mortality, and the fishermen cannot find a way to improve the technology or procedures used, then they will be forced to cut the effort to meet that limit.

Developing alternative ways of fishing One of the ways of reducing a particular bycatch is by switching gear or fishing technique to others that are more selective with respect to that species. But the alternative gear technique may have other problems, and they have to be compared with the original ones. A good example of the dangers of "ecological solutions" that are not based on scientific data and analyses is the "dolphin-safe" policy. In the case of the eastern Pacific tuna fishery, an attempt to eliminate all fishing on dolphins to

539 reduce their mortality was recently promoted by some environmental groups by means of a "dolphin-safe" campaign. This campaign had the objective of closing all the markets to the tuna caught in association with dolphins, and therefore, of promoting alternative ways of fishing. Although alternative ways of fishing available in the area rarely involved dolphins, they did have other consequences that were not considered. If the fishermen had switched to other ways: (i) the production of tunas would have been severely reduced [7] because of the increasing catches of very small and sexually immature tunas; (ii) the discards of tunas would have increased significantly, and (iii) the bycatches of other species would have experienced large increases [5,8]. Overall, this "solution" results in an ecologically unsound use of the target species, and trades a low bycatch of dolphins for a large bycatch of other species with an effect on the ecosystem of unknown magnitude.

Reducing the bycatch per unit of effort (BPUE) Depending on the characteristics of the fishery, different options will be available to achieve this goal.

Technological change In many cases, bycatch problems can be eliminated, or at least reduced, by technological improvements in the fishing gear, mode of operation, materials, etc. The use of turtle-excluder devices (TEDs) in shrimp trawls, the backdown maneuver and the Medina panel while purse-seining for tunas associated with dolphins, pingers in gillnets [9], square mesh or grids in some areas of the net, etc., are examples of this.

Regulations aiming at reducing BPUE (a)

(b)

(c)

(d)

Gear or operational restrictions: examples of this could be restrictions in mesh size, duration of trawl hauls, etc. They may lead to lower B PUEs, either by reducing the probability of encounter with a bycatch species, or by improving the chances of that species surviving the encounter. Individual limits or "acceptable" ratios: another approach is the setting of individual bycatch limits, or "acceptable" ratios of bycatch to total catch. In either case, if the fishermen have any control on the bycatch level, they will change their behavior, area of deployment, or other variables to stay within the limits. In the eastern Pacific, tuna purse-seiners have an annual limit to the number of dolphin mortalities they cause, and if a vessel reaches its limit it must stop fishing for tunas associated with dolphins for the remainder of the year. Partial closures: if some areal or temporal strata have much higher bycatch rates than others, closures of those strata should result in lower average B PUEs. If effort can be re-distributed to other strata, the gains made may not be accompanied by losses in effort or in catches. Incentives: not all fishermen are equally skilled at handling their gear and

540 boats, or at making decisions, and not all are equally motivated. Individual limits or "acceptable" ratios can be considered as incentives, but there are other possibilities. The incentive system should reward the best fishermen (from the point of view of the bycatch), and also promote the development of new techniques, by conferring an economic advantage to those that can find better ways of fishing. The individual vessel mortality limit is an example of a "selective" mechanism that rewards the better operators, but there are many other possibilities, including extended seasons, higher catch limits, access to desirable areas, etc.

Training When there are maneuvers or procedures, or some devices that can reduce bycatches, it is possible to train captains and crews of fishing boats to use them effectively. The IATTC staff conducts seminars at frequent intervals to pass information on techniques for minimizing dolphin mortality to the less skilled or experienced vessel captains.

The Lines of Defense Against Bycatch Problems Reduce incidental captures First line: decisions by fishermen or regulations concerning gear, areas, and seasons Before deploying the net or other type of gear, many decisions are made by the fishermen that may affect the bycatch. They may choose to avoid some areas or seasons with high bycatch rates; they may modify or change the type of fishing gear used to reduce the incidental captures of non-target species. Alternatively, regulations may be passed making some of those choices mandatory, or banning some gears, areas, etc. In the eastern Pacific tuna fishery, for example, the fishermen: (a) avoid areas along the edges of the fishery where dolphin herds are larger and dolphin behavior is less adapted to the fishing operations [10]; (b) avoid setting on species that have higher mortality rates; (c) use modified purse-seines and other auxiliary equipment that have been adapted to reduce the entanglement of dolphins and to facilitate their release [11]; (d) train their fishing captains and crews on the dolphin rescue techniques, a set of special procedures developed over the years; and (e) test their gear to verify that it is performing as intended. Second line: decisions by fishermen or regulations concerning deployment conditions When the gear is being deployed, another set of choices (or regulations) can come into play. The time of day, the duration of the deployment, the fishing depth, the position with respect to currents or other oceanographic or topographic features, are

541 all factors that may affect bycatches. Gillnets, longlines or trawls can be fished at different depths and for different periods to minimize bycatches For the example of the eastern Pacific fishery, the mortality rates of dolphins are greater when the set is completed after dark, so "sundown" sets are prohibited. Also, the fishermen tend to avoid setting in areas with strong subsurface currents. When subsurface currents are present, but not too strong, the orientation of the deployment is modified.

Increase release of bycatch Third line: release from the net (procedures and equipment) After marine mammals are captured, different procedures and equipment can help to facilitate their release. In the eastern Pacific, a procedure called backdown [12] is used to get the dolphins out of the net. After a group is encircled, the seiner gets its engine in reverse, and pulls the net in such a way that the corkline sinks and allows the escape of the dolphins. Also the net has been modified by addition of a small mesh section to keep the dolphins from getting their snouts in the mesh, a raft is used inside the net for hand rescue. In the western North Atlantic, techniques have been developed to release whales caught in gillnets [ 13].

Fourth line: release from the deck (procedures and equipment) When a marine mammal is brought on board a vessel, it is returned to the sea using different methods depending on the species in question. It may be possible to change some of the conditions prevailing on deck (shade, temperature, running water) to reduce the negative effects of the capture on survival (if there were any), or to develop equipment to facilitate the handling of the animals, reducing injuries or traumas. This approach is more likely to prove useful for bycatches of fish or invertebrates.

Change from bycatch to catch Fifth line: utilization Once the marine mammal is dead, it can be returned to the sea or utilized. From the ecological point of view, in some cases it may be wiser to utilize it. Given that the ecological costs of fishing have already been incurred (fuel consumption, pollution, bycatches, damage to the habitat, etc.), the protein or any other product extracted from the bycatch may replace other alternative sources of the same product, and reduce the ecological impact of the other exploitation. On the other hand, an animal returned to the sea dead may be "recycled" faster than one brought ashore, so there may also be some ecological arguments in favor of the opposite action. Only knowledge of the extent of the ecological impacts caused by the fishing operations, and of the characteristics of the ecosystem where the impacts occur can provide the answer, which may be different in different cases.

542 In the eastern Pacific, during the recent peak of dolphin mortality, in 1986, approximately 12,000 mt of dolphins were returned to the sea (133,000 individuals x 90 kg each). Currently, the figure is close to 324 mt (3,600 individuals), still a considerable amount, but one that must be compared with, for instance, the estimated figures of direct harvest of more than 2,000 dolphins for human consumption in Peru in 1987 [14]. The dolphins accidentally killed could have replaced the directed harvest.

Conclusions

The strategies to mitigate bycatch problems are determined by the statistically simple nature of those problems. With only two "levers" available, the solutions will have to be sought in one of them. Fortunately, the options available are quite diverse, and further technological and scientific developments will add more. Scientist must work to identify the factors that cause high bycatches, such as environmental conditions (currents, turbidity, etc.), gear characteristics and "behavior", and behavior and ecology of the species involved. This knowledge must be transferred to the fishermen to improve their decision-making processes. The lines of defense identified above provide a wide range of possibilities for mitigating bycatch problems. In each fishery, some of those options will be available (because of the nature of the fishery and of the problem), while others may not be. All possible lines of defense should be explored for their potential to produce results, with the idea that the solution may be the sum of many improvements, large and small, rather than a "silver bullet" that eliminates all the problems at once. In the eastern Pacific, the average mortality of dolphins per set has gone from about 60 animals per set in the 1960s [15], to about 0.5 in 1993 [6]. But it took many innovations, most of them generated by the fishermen, and 30 years to reach the current point. The education of the fishermen has been another key element of the process. The programs to train the fishermen in the equipment and procedures to release dolphins have been going on for years, and they will have to continue. They produce a steady flow of ideas between fishermen and scientists and generate a constructive communication channel that sparks new initiatives and motivates the participants. Bycatches result from a combination of environmental, biological, ecological, and gear factors. It is vital to identify them, and to assess their relative importance if measures needed to mitigate the problems are to be undertaken. Research programs and management actions should be based on well-established scientific facts. Observer programs that are designed to assist in the search for solutions can provide the data required. Given the large number and complexity of the factors that can be involved, extensive databases are required. To illustrate this complexity, in the eastern Pacific tuna fishery the following factors have some effect on incidental mortality rates: species of dolphin, area, size of dolphin herd, size of tuna school caught, time of day, presence of strong currents, malfunctions on the equipment, use of a rescue

543 raft, condition of equipment (repair, alignment, etc.), and, of course, the skill and motivation of the captain and crew of the vessel. The experience of the eastern Pacific tuna fishery shows that bycatch problems can be tackled successfully, but that some conditions have to be met to reach a solution. Some of these are: to have nations and industries that accept the existence of the problem, and tackle it by instituting programs with clear objectives in which all participate, based on a solid scientific foundation; to work gradually toward these objectives, setting realistic short-term goals, that encourage the fishermen to achieve them, with a system of incentives and disincentives for the individual fishermen, to reward or penalize their skill, motivation, and creativity; to have: (a) an extensive monitoring system, that provides valid estimates and helps diagnose the causes of problems; (b) a sensible scheme of regulations, with fair, but meaningful, sanctions for infractions; and (c) an adequate and transparent compliance program, with the participation of the nations involved; to have an extensive experimental program that generates and tests innovations to address the problems identified; to have a continued and constructive interaction among fishermen, scientists, managers, environmentalists, and industrialists.

Acknowledgements

The author would like to acknowledge Drs. W. Bayliff, J. Joseph and M. Scott for their comments and review of the manuscript.

References

1. Northridge SP. World review of interactions between marine mammals and fisheries. Rome: FAO Fish Tech Pap 1984;251:190 pp. 2. Northridge SP. An updated world review of interactions between marine mammals and fisheries. Rome: FAO Fish Tech Pap 1991;251(Suppl 1);58 pp. 3. Jefferson TA, Curry BE. A global review of porpoise (Cetacea: Phocoenidae) mortality in gillnets. Biol Conserv 1994;67:167-183. 4. Francis RC, Awbrey FT, Goudey CL, Hall MA, King DM, Medina H, Norris KS, Orbach MK, Payne R, Pikitch E. Dolphins and the Tuna Industry. Washington, DC: National Academy Press, 1992;xii, 176 pp. 5. Joseph J. The tuna-dolphin controversy in the eastern Pacific Ocean: biological, economic and political impacts. Ocean Dev Int Law 1994;25:1-30. 6. Lennert C, Hall MA. 1995. Estimates of incidental mortality of dolphins in the eastern Pacific Ocean tuna fishery in 1993. Rep Int Whal Commn 1995;45:(submitted). 7. Punsly RG, Tomlinson PK, Mullen AJ. Potential tuna catches in the eastern Pacific Ocean from schools not associated with dolphins. Fish Bull US 1994;92:132-143. 8. Hall MA. An ecological view of the tuna-dolphin problem. (unpublished).

544 9. Lien J, Todd S, Guigne J. Inferences about perception in large cetaceans, especially humpback whales, from incidental catches in fixed fishing gear, enhancement of nets by "alarm" devices, and the acoustics of fishing gear. In: Thomas JA, Kastelein RA (eds) Sensory Abilities of Cetaceans: Laboratory and Field Evidence. New York: Plenum Press, 1990;347-362. 10. Hall MA, Boyer SD. 1986. Incidental mortality of dolphins in the eastern tropical Pacific tuna fishery: description of a new method and estimation of 1984 mortality. Rep Int Whal Commn 1986;36:375-381. 11. Coe JM, Holts DB, Butler RW. The "tuna-porpoise" problem: NMFS dolphin mortality reduction research, 1970-81. Mar Fish Rev 1984;46:18-33. 12. Coe JM, Sousa G. Removing porpoises from a tuna purse seine. Mar Fish Rev 1972;34(1112):15-19. 13. Lien J. Entrapments of larger cetaceans in passive inshore fishing gear in Newfoundland and Labrador (1979-1990). In: Perrin WF, Donovan G (eds) Int Whal Commn Special Issue 1994;(in press). 14. Van Waebereek K, Reyes JC. Catch of small cetaceans at Pucusana Port, Central Peru, during 1987. Biol Conserv 1990;51:15-22. 15. Lo NCH, Smith TD. Incidental mortality of dolphins in the eastern tropical Pacific, 1959-1972. Fish Bull US 1986;84:27-34. 16. Hall, M.A. and Lennert, C. Incidental mortality of dolphins in the eastern Pacific Ocean tuna fishery in 1992. Rep Int Whal Commn 1994;44:349-352.

9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 12l.Ulltang, editors

545

Ecological implications of harp seal Phoca groenlandica invasions in northern Norway Tore Haug and Kjell Tormod Nilssen Norwegian Institute of Fisheries and Aquaculture, Tromsr Norway. Abstract. In the years since 1978, Barents Sea harp seals Phoca groenlandica have appeared in large numbers in Finnmark, North Norway, in February-May. The size of the "seal invasions" increased dramatically in 1987 and 1988 when large seal herds were observed along the coast of North Norway in January-August. The seal invasions gave rise to seal-fisheries conflicts. In addition to consuming fish (capelin, cod, saithe and haddock), the seals caused substantial damage to gill-nets and gill-net catches. The presence of seals may also have resulted in the emigration of commercial species from traditional fishing grounds to deeper strata or areas unsuitable for fishing. Reduced recruitment to the seal population seems to have prevailed during most of the seal invasion period with particularly dramatic effects in 1986-1988, when first-year-mortality may have been almost total. Food shortage, particularly the two important prey species capelin and herring, is discussed as a possible factor contributing to the seal invasions.

Key words: seal/prey, seal/fisheries, recruitment, food shortage

Introduction

Two populations of harp seals Phoca groenlandica inhabit the northeast Atlantic Ocean. These populations whelp off the east coast of Greenland (the Greenland Sea population) and in the White Sea (the Barents Sea population), respectively [1 ]. The annual migration pattern of the Barents Sea population of harp seals is usually characterized by a north-bound feeding migration in spring and early summer (MayJune) and a south-bound breeding migration during winter [2]. In summer and autumn the seals are found in open waters and along the pack-ice in the northern parts of the Barents Sea, and they move southwards in November. In winter and early spring (December-May), the seals are usually concentrated at the southern edge of the range, primarily in the southeastern parts of the Barents Sea and in the White Sea where breeding and moult occur [2,3]. From 1978 onwards, harp seals started to appear in large numbers along the coast of Finnmark, North Norway, in winter and spring. Although recaptures of tagged seals indicated that some of the immature seals were from the Greenland Sea population [4], it seems reasonable to assume that the majority of the seals arose from the Barents Sea population. In the current paper we attempt to discuss the following questions" What were the consequences of the changes in the migratory patterns of

Address for correspondence: T. Haug, Norwegian Institute of Fisheries and Aquaculture, P.O. Box 2511, N-9002 Tromsr

Norway.

546 the seals - for the prey species of the seals, for fisheries, and for the seal population itself?

The Seal Invasions A new, and apparently aberrant, migratory pattern of the Barents Sea harp seal population persisted throughout the 1980s, and there were dramatic increases in numbers of animals observed along the Norwegian coast in 1987 and 1988 [5]. The losses imposed by the invading seals on coastal fisheries in northern Norway [6] led the Norwegian authorities to introduce compensatory bounty payments for seals taken as by-catch in gill nets. The numbers of harp seals caught increased from 500 to 2,000 animals during the first half of the 1980s to more than 56,000 in 1987 (Fig. 1). Numbers caught were lower in 1988, and from 1989 onwards the numbers seem to have returned to the level of the early 1980s. The numbers of seals returned for compensation purposes are, however, lower than the total numbers drowned, which may have been at least 10,000 per year throughout the early 1980s [4,7] and perhaps as many as 100,000 in 1987 [8]. It should also be stated that the number of seals recorded for compensation purposes cannot be used uncritically as an index to estimate the size of the invasion because of the large geographic and seasonal 60 000

60%)

1,580,000 (10%) 24,600,000 (7%)

Table 2. Sealwoxm infection levels in fish from Vega and Hvaler between 1990 and 1993

Area

Species names

n

Prevalence (%)

Abundance (wormslfish)

Max

Wormskg

Vega Common sculpin Butterfish Sea scorpion Dragonet Long rough dab Cusk Cod

Myoxocephalus scorpius Pholis gunnellus Taurulus bubalis Callionymus lyra Hyppoglossoides platessoides Brosme brosme Gadus morhua

248 1 2 3 10 301 414

76.2 (112) (213) 50.0 60.1 40.6

23.2 1.4 6.0 4.8

287 1 3 2 3 205 150

175.5 (110) (50) (39) 14.0 7.2 3.6

Hvaler (shallows) Common sculpin Hooknose Sea scorpion Eelpout Cod 5-Bearded rockling

Myoxocephalus scorpius Agonus olaphractus Taurulus bubalis Zoarces viviparus Gadm morhua Ciliata mustela

172 2 8 3 128 10

18.8 (212) (518) (113) 62.5 80.0

36.1 -

41 1 2 3 1 315

5

209.1 (80) (46) (30) 26.1 25.9

Hvaler (trawls) 4-Bearded rockling Long rough dab Cod Common dab Plaice

Enchelyopus cimbrius Hyppoglossoides platessoides Gadus morhua Limanda limanda Pleuronectes platessa

3 2 12 1 1

1.2 0.5 0.3 0.1 0.1

195 662 1,173 33 71

14.4 3.2 8.6 (3133) (1171)

-

10.0 2.5

0.16 0.03 0.13 -

559

560 species in the immediate vicinity of haul-out sites (Table 2). We also found that, when comparable habitats could be sampled at both sites, infected fish species were the same in the two areas. Non-commercially exploited fish species have so far been sampled very rarely, which explains our findings of five new host records, four species in shallow waters (dragonet Callyonymus lyra, hooknose Agonus olaphractus, eelpout Zoarces viviparus, five-bearded rockling Ciliata mustela) and one species in trawls (four-bearded rockling Enchelyopus cimbrius). In the species where enough fish could be caught, infection levels were found to increase with the length and age of the host, and the juveniles of larger species (cod Gadus morhua and cusk Brosme brosme) less than 3 0 c m had little or no infection [8]. The importance of small species is further illustrated by the very large number of worms per kilogram in common sculpins (Myoxocephalus scorpius) in both study areas. This species, which is known to be eaten by common seals [9,10] and by grey seals [5,11,12] had an average of 175 and 209 worms/kg at Vega and Hvaler, respectively. The significantly higher infection level in sculpins from the skerries at Hvaler has been attributed to the more sedentary behaviour of common seals, compared to grey seals at Vega [8]. It can also be linked to the more concentrated area of shallow waters around the fewer islands at Hvaler, when compared to the extensive shallow grounds at Vega. The main features of the sealworm transmission route in Norwegian waters revealed by our study are summarised in Fig. 1. Final seal hosts, common seals at

~-

.-

_i

~. 7' ~':' :':,!~4~,;::,:.::" ..:~:7%~ ,~,~--~--.:._.~27::._2_,,,.....r 9.--~- ~ .

m ~.(

~

-

.y

~

*-

e

L3 I

I, Ii , II m a i n

steps

in the lifecycle unimportant

t "

~;

.

"" .

,.

~,~t

\ "" ~ - ' ~ . ~ /

Fig. 1. Major sealworm (Pseudoterranova decipiens) transmission routes identified in N o r w e g i a n waters.

561 Hvaler and mostly grey seals at Vega, become infected by foraging on small benthic fish species close to haul-out sites. Larger fish, including commercially exploited cod and cusk are not important for the completion of the life-cycle. The key role played by non-commercially exploited but long-lived species, such as sculpins, could explain why sealworm infection levels in cod have remained remarkably stable [ 13,14], as dramatic decreases in the abundance of cod stocks would affect sealworm transmission only marginally. This was illustrated in Hvaler, where the seal distemper virus epizootic which killed most of the seals using the haul-out sites in the summer 1988 had only a temporary and relatively little effect on sealworm transmission to cod caught by trawls [15], suggesting the presence of a reservoir fish host other than the locally over-exploited cod.

Seal-fishery interactions Three types of interactions between seals and fisheries are relevant to the two study areas. The competition for a common fish resource between seals and fishermen and the seals' interactions with fishing gear are only briefly reviewed. Infection of commercially exploited fish species by the sealworm parasite is exposed in more detail. First, concerning predation, a diet study for common seals in Hvaler [16; Prime, unpublished] confirmed that common seals foraged on locally and seasonally abundant prey [17,18], pelagic, demersal and benthic in the vicinity of the haul-out sites, and that the preferred size of fish prey was mostly 20 cm. Published studies have also shown that the bigger grey seals generally eat more and eat bigger fish prey than common seals [19]. Second, seals are known to interact mostly with fixed gear such as gill-net or long-lines where they can easily scavenge, and although not quantified, this was observed in both our study areas. Most importantly, our studies in Hvaler and Vega have revealed an important difference between actual and perceived level of interaction. The perception of interactions by fishermen appears to be mainly driven by the type of fishing activity prevailing in the area at the time (Table 3). The behaviour of the seal species, with bigger and less shy grey seals than common seals is also important at Vega. Furthermore, the much greater spatial and temporal overlap between seals and fishing activities at Vega explains a much higher perception of interaction than at TorbjCmskjaer. Similarly for the third type of interaction, although effective sealworm transmission may be locally higher from common seals at Hvaler, the area of shallow waters where infection levels are high is much more extended at Vega (Table 1). This explains that the dominant fishery at Vega targets benthic and demersal species with fixed gear, in shallow waters in the immediate vicinity of seal haul-out sites. At Vega, between 1990 and 1992, it was estimated that sealworm infected fish represented nearly 50% of the weight of landed fish over the year, and more than 60% of its total value (Table 1). This is also exacerbated by the relatively light fishing pressure at Vega, which leads to the presence of seven age groups (2-8 years old) of cod in the fishery, while there are mainly two age classes caught in the dominant trawl fishery (1 and

562 Table 3. Criteria of minimum and maximum levels of perceived interaction between grey and common seals and fishing activities, with observations for the fisheries at Hvaler and Vega between 1990 and 1993

Seal species Fishery

Fishing gear Fish species

Fish growth Fish age in catch

Grey Common Inshore Offshore Close to haul-outs Far from haul-outs Lines, gill-nets trawl Pelagic fish demersal fish benthic fish invertebrate Fast Slow Young Old

Min

Hvaler

Vega

Max

X

X

x

x

X

x

x

x

x

x x

x

X

X

X

X

X

X

X

x x

x x

x

x

X

X

X

X

X

X

2 years old) at Hvaler [8]. At Hvaler, the main fishery targets invertebrate species (Pandalus borealis and Nephrops norvegicus) with trawlers in deeper waters. These invertebrates are not infected with the sealworm, and infection levels in fish caught there are low, particularly in young fish. It was estimated that, between 1990 and 1992, not more than 10% of the total weight of fish and shellfish landed in the Hvaler area concerned known sealworm host species (Table 1). This represented just 7% of an otherwise high-valued catch, a very different situation from Vega.

Discussion

Although the major sealworm transmission routes in Norwegian waters have been identified in our study areas, important aspects of the parasite life-cycle dynamics need to be further investigated before we can fully understand the differences between Vega and Hvaler. The feeding behaviour and habitat use by common seals, for example, appear to change when grey seals are present. Thus more information is needed to compare the seals' diet in Vega and in Hvaler, and the use of haul-out sites by the two species, in isolation or co-habiting during the year. Historical changes in the fishery's main target species in Hvaler, from cod to prawn, are similar to changes in many Scottish coastal fisheries, were the inshore fleet have switched from cod to Nephrops. Hence the competition with seals for a main target species has been replaced by a marginal interaction for a now relatively low valued species caught incidentally [14]. Changes in fishing activities are a key element to analyse historical changes in the perception of seal-fisheries interactions. Although sealworm infection levels in cod may have remained very similar over the last three decades [ 13] in many coastal areas, the dominance of small young fish cur-

563 rently caught by trawlers has led to an apparent decline of sealworm infection levels and consequently, to a decrease in the perceived level of interaction. Finally, the presence of a large and long-lived reservoir of sealworms in small benthic fish has important management implications. First, there is little hope for a reduction in transmission through commercial fisheries. However, an apparent reduction in infection levels in cod caught by long-lines in Vega could be achieved though a higher fishing pressure by decreasing the average age and length of cod caught in the fishery. The economics of a trade-off between smaller fish with fewer worms and larger fish with higher worm burdens and therefore higher processing costs, would be worth studying in the future. Second, given that the parasite spends very little time of its life cycle in seals, there is no reason to believe that a seal cull other than a total cull could control worm numbers in demersal fish. This was clearly illustrated by our monitoring of worm burdens in cod trawled at Hvaler after the epizootic which killed two-thirds of the seal colony in 1988 [ 15].

Acknowledgements This research was part of a 4-year set of projects funded by the Marine Mammal Programme of the Norwegian Research Council for Fisheries (NFFR) whose support and encouragement are gratefully acknowledged. Our collaboration was initiated by Andrew Rosenberg, who made helpful suggestions. Professional fishermen Odd Stirensen, Kjell Arne Hovland and Leif H. Lien provided invaluable support. The collaboration of Arne BjCrge, John Prime, Stein Tveite, Anne SchCnhaug, Giari Langholm, Einar Stromnes and Glenn Boyle made the study possible over the years.

References 1. Bowen WD (ed). Population biology of the sealworm (Pseudoterranova decipiens) in relation to its intermediate and seal hosts. Can Bull Fish Aquat Sci 1990;222. 2. BjCrge AJ. The relationship between seal abundance and cod worm (Phocanema decipiens) infection in cod in Norwegian coastal waters. International Council for the Exploration of the Marine Sea, Copenhagen. Mammals Commn Rep 1985;C.M. 1985/N:4. 3. Jensen T, Idhs K. Infection with (Pseudoterranova decipiens) (Krabbe, 1878) larvae in cod (Gadus morhua) relative to proximity of seal colonies. Sarsia 1992;76:227-230. 4. Templeman W. Historical background to the sealworm problem in Eastern Canadian waters. In: Bowen WD (ed) Can Bull Fish Aquat Sci 1990;222:1-16. 5. H~iuksson, E. Investigations on the Sealworm problem in Icelandic waters: recent findings and future research. In: Mtiller H (ed) Nematode Problems in North Atlantic Fish. International Council for the Exploration of the Sea 1989;Copenhagen. C.M./F:6:30-31. 6. Wootten R, Waddell IF. Studies on the biology of larval nematodes from the musculature of cod and whiting in Scottish waters. J Cons Perm Int Explor Mer 1977;37:266-273. 7. Markussen NH. Apparent decline in the harbour seal Phoca vitulina population near Hvaler, Norway, following an epizootic. Ecography 1992;15:111-113. 8. Jensen T, Andersen K, des Clers S. Sealworm (Pseudoterranova decipiens) infections in demersal fish from two areas in Norway. Can J Zool 1994;72:598-608.

564 9. Behrends G. Zur Nahrungswahl von Seehunden (Phoca vitulina) im Wattenmeer SchleswigHolstein. Z Jagdwiss 1985;31:3-14. 10. Pierce GJ, Thompson PM, Miller A, Diack JSW, Miller D, Boyle PR. Seasonal variation in the diet of common seals (Phoca vitulina) in the Moray Firth area of Scotland 1991. J Zool London 223:641-652. 11. Benoit D, Bowen WD. Seasonal and geographic variation in the diet of grey seals (Halichoerus grypus) in eastern Canada. In: Bowen WD (ed) Can Bull Fish Aquat Sci 1990;222:215-226. 12. Hammonds PS, Prime JH. The diet of British grey seals, Halichoerus grypus. In: Bowen WD (ed) Can Bull Fish Aquat Sci 1990;222:243-254. 13. des Clers S. Functional relationship between sealworm (Pseudoterranova decipiens, Nematoda, Ascaridoidea) burden and host size in Atlantic cod (Gadus morhua). Proc R Soc London B 1991 ;245:85-89. 14. des Clers S, Prime J. Seals and fisheries interactions: observations and models in the Firth of Clyde (Scotland). In: Greenstreet SPR, Tasker ML (eds) Aquatic Predators and their Prey. Oxford: Blackwell Scientific, 1995;(in press). 15. des Clers S, Andersen K. Sealworm (Pseudoterranova decipiens) transmission to fish trawled from Hvaler, Oslofjord, Norway. J Fish Biol 1994; 46:8-17. 16. Olsen M. N~eringsvalg og n~eringsstrategi hos steinkobber (Phoca vitulina). Cand Sci Thesis. University of Oslo, 1993. 17. Thompson PM. Summer foraging activity and movements of radio-tagged common seals (Phoca vitulina L.) in the Moray Firth, Scotland. J Appl Ecol 1990;27:492-501. 18. Thompson PM, Pierce GM, Hislop JRG, Miller D, Diack JSW. Winter foraging by common seals (Phoca vitulina) in relation to food availability in the inner Moray Firth, N.E. Scotland. J Appl Ecol 1991 ;60:283-294. 19. Prime JH, Hammond PS. The diet of grey seals from the south-western North Sea assessed from the analyses of hard parts found in faeces. J Appl Ecol 1990;27:435-447.

9 1995 ElsevierScience B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand ~. Ulltang,editors

565

Grey seal (Halichoerus grypus Fabr.), population biology, food and feeding habits, and importance as a final host for the lifecycle of sealworm (Pseudoterranova decipiens Krabbe) in Icelandic Waters Erlingur Hauksson 1 and Droplaug Olafsd6ttir2 lIcelandic Fisheries Laboratories, Reykjavfk, Iceland; and 2Marine Research Institute, Reykjavfk, Iceland Abstract. A research-project on the sealworm problem of commercial fish in Icelandic Waters, which incorporates studies on population sizes of seals, seal diet, nematode infections of seals and fish species, is currently being implemented. The distribution of the grey seal population in Iceland is divided into two main areas, the West and Northwest coasts and the Southeast coast. The population size was stable at 10,000-12,000 animals, in the period 1982-1990, but may be declining in recent years. Grey seals show regional and seasonal differences in feeding habits and diet. Grey seals from the West and Northwest coasts feed mainly on cod (Gadus morhua) and lumpsucker (Cyclopterus lumpus) in the period from February to August, the active feeding time, while sea-scorpion (Myoxocephalus scorpius) becomes the most dominating food-species, during September to January, the breeding time of grey seals. Seals from the South coast feed largely on sandeels in all seasons. Sealworms are most abundant in grey seals from the West and Northwest coasts, where the seals' ingestion of highly infected seascorpions in the autumn causes a maximum abundance of sealworm in their stomachs. Grey seals are the main final host for sealworm in Icelandic waters. Sea-scorpion seems to be the most important second intermediate host. K e y words: grey seal's diet, nematode infection, intermediate hosts

Introduction

The research project on the sealworm problem in commercial fish in Icelandic waters was started in the year 1980. The project emphasizes studies on the population dynamics of seals, food and feeding habits of seals, and nematode infections of seals and fish. The aim of the project is to elucidate the main pathways of the life-cycle of the sealworm in Icelandic waters.

Materials and Methods Aerial census o f grey seal pups

The population status of the grey seal in Iceland has been estimated by aerial counting of pups. Pups were counted in the breeding areas, in October to November,

Address for correspondence: E. Hauksson, Icelandic Fisheries Laboratories, P.O. Box 1405, 121 Reykjavfk, Iceland, Tel. +354 620240; Fax 354 620740.

566 in the years 1982, 1986, 1990 and 1992. Population size of the grey seal is estimated by multiplying the total number of pups born each year by a factor of 4. Minimum and maximum size limits for the population are estimated by multiplying by the factors 3.5 and 4.5 [ 1].

Food studies Contents of stomachs from grey seals collected in the period 1990-1993 were used for studying diet and nematode infections. Age and stage of maturity of the seals were determined by studying their canine teeth and sex organs. Food remains from the stomachs were identified to species or species-groups by recognizing otoliths, bones and other indigestible parts. The percentages of stomachs with each "foodspecies" are presented in this paper.

Nematode studies The length and weight of each fish were measured. Their visceral organs, fillets and flaps were investigated for nematodes on a candling table. Subsamples of 200--400 worms were taken from the stomachs of grey seals. All nematodes were preserved in a solution of 70% iso-propanol, 5% glycerol and 25% water, and were cleared in glycerol or lactic acid, before identification to species, larval stages and sex under a light microscope [2]. The terms prevalence (percentage of seals infected) and abundance (mean number of parasites per seal, including uninfected seals) used here follow standard usage [3]. Density represents the number of worms per 100 g of whole fish.

Results

Population biology of grey seals The distribution of the grey seal population in Iceland is mainly restricted to the West and Northwest coasts and the Southeast coast (Fig. 1). About 80% of the population breed in the West and Northwest, and much of the rest breeds in the Southeast. Results of aerial censuses of pups on breeding places of grey seals indicate a population size at 10,000-12,000 animals in the period 1982-1990. The pup count in the year 1992, however, was the lowest count recorded, so the population may be starting to decline (Table 1).

Food and feeding habits of grey seals Grey seals show regional and seasonal differences in feeding habits and diet (Table 2). Cod (Gadus morhua) and lumpsuckers (Cyclopterus lumpus) are important in the food of grey seals from the West, Northwest, Northeast and East coasts during the

567

W-NW 6 7~

641-- k,.

\

..-~

24 ~

14~

Fig. 1. Breeding distribution of grey seals (Halichoerus grypus) in Iceland and division of the coast into three areas; West-Northwest, Northeast-East and South. Breeding places are marked with dots: large dots, major breeding sites; small dots, minor breeding sites. Sampling stations for fish are indicated with numbers: 1, the Island of Hvalseyjar; 2, Sn~efellsnes; 3, 61afsv~; 4, Breida-fjord; 5, Lfitrabjarg; 6, West-fjords; 7, Strandir; 8, Htinafl6i; 9, Langanes; 10, LoOmundar-fjord; 11, Horna-fjord; 12, M3)rabugur; 13, Selvogsgrunn (long rough dab); 14, Selvogsgrunn (witch (Glyptocephalus cynoglossus)). Topographical lines show the 200 and 400 m depth lines. actual feeding time in February to August. A striking increase in p e r c e n t a g e s of s t o m a c h s with sea-scorpions (Myoxocephalus scorpius), occurs in grey seals f r o m the W e s t and N o r t h w e s t coasts during breeding time. The diet of grey seals f r o m the South coast consists mainly of sandeels (Ammodytes sp.) in all seasons.

Nematodes in the stomach of grey seals G r e y seals are m u c h m o r e infected with s e a l w o r m s than c o m m o n seals (Phoca vitulina L.). The difference is often tenfold or more, in the s a m e season and in the s a m e coastal area [2]. The prevalence of s e a l w o r m in grey seals was found to be

Table 1. Pup-production of grey seals (Halichoerus grypus) in Icelandic Waters, estimated population size, as 4 times total pup-production each year of counting, and +_12.5% of the population size

Pup-production Stock-size _

1982

1986

1990

1992

2,689 10,756 1,345

2,965 11,860 1,483

3,034 12,136 1,517

2,133 8,532 1,066

568

Table 2. Diet of grey seals (Halichoerus grypus) by seasons and areas in Icelandic Waters (Fig. 1), in the years 1992-1993, presented in percentage of stomachs containing each food-species; February to August; feeding time and September to January; breeding time Areas and food species

February-August (%)

September-January (%)

West and Northwest coasts Cod Lumpsucker Sandeel Catfish Sea-scorpion Hyas sp.

n = 422 32.0 31.5 24.6 23.7 4.0 16.6

n = 150 18.7 2.0 20.0 2.7 38.0 30.0

Northeast and East coasts Cod Lumpsucker Sandeel Catfish Sea-scorpion Hyas sp.

n = 50 30.0 34.0 8.0 20.0 4.0 16.0

n=6 66.7 0.0 0.0 16.7 0.0 0.0

South coast Cod Lumpsucker Sandeel Sea-scorpion Catfish Hyas sp.

n = 21 0.05 0.0 76.2 4.8 4.8 4.8

n = 22 0.0 0.0 22.7 0.0 0.0 4.5

100%

in all s e a s o n s a n d a r e a s , b u t t h e a b u n d a n c e

numerous

varies.

Sealworms

are most

in g r e y s e a l s f r o m t h e W e s t a n d N o r t h w e s t c o a s t s a n d t h e a b u n d a n c e

in

this a r e a i n c r e a s e s f r o m s p r i n g to a u t u m n . A d e c r e a s e in t h e n u m b e r o f s e a l w o r m s in t h e s t o m a c h s o f g r e y s e a l s f r o m t h e S o u t h c o a s t s e e m s , in c o n t r a s t , to o c c u r f r o m t h e s u m m e r m o n t h s a n d t h r o u g h t h e a u t u m n ( T a b l e 3).

Table 3. Sealworm (Pseudoterranova decipiens) infection of grey seals (Halichoerus grypus) 1+ year of age, by coastal areas in Icelandic Waters, surveyed in the period 1990-1993 Coastal area

Month

n

Abundance

Standard error

West-Northwest

May June July October May August April August September October

15 8 37 24 5 5 3 3 25 15

361.9 567.9 639.5 3,972.1 119.8 1,563.0 265.3 726.0 733.0 159.8

48.62 286.44 119.69 974.01 35.35 439.61 37.55 477.42 197.9 57.26

Northeast-East South

Table 4. Prevalence, abundance and density of sealworm (Pseudotermnova decipiens) larvae in some fish species from the Icelandic coastal waters (based mostly on PI)

Fish species -

Length (cm)

n

Location

Year

Prevalence (%)

Abundance Range

Density (worms/100 g fish)

-

Cod Saithe Saithe Whiting Sea-scorpion Sea-scorpion Plaice Plaice Dab Halibut Witch Herring Sand-eel Lumpsucker

Breidafjord West coast Snzfellsnes and West-fjords West coast Hvalseyjar West coast Mfrabugur SE Coast Snzfellsnes, West-fjord and Strandir West coast Hvalseyjar West coast Snzfellsnes and West-fjords West coast Hvalseyjar West coast Snzfellsnes and West-fjords West coast Snzfellsnes and West-fjords West coast Selvogsgrunn South coast Snzfellsnes West coast 61afsvik West coast Hvalseyjar 569

aSamples from 1982 do not include infections from stomachs.

VI Q\ \O

570

Table 5. Density (worms per 100 g of whole fish) of sealworm (Pseudoterranovadecipiens) larvae in long rough dab (Hippoglossoidesplatessoideslimandoides),from various coastal areas of Icelandic Waters Collection site, collection time and coastal areas

Fish length < 25 cm

25.0-34.5 cm >35 cm

L~itrabjarg, March 1991, West coast n

0.47 24

0.71 56

1.76 27

Hfnafl6i, October 1989 and February 1990, Northwest coast n

2.32 87

1.60 46

1.43 8

Langanes, March 1991, Northeast coast n

1.26 21

0.77 32

0.26 18

1.50 8

1.19 65

1.50 25

6.41 38

1.55 39

0.51 21

Horna-fjord, March 1991, Southeast coast n Selvogsgrunn, March 1991, South coast n

Sealworm infections of food species of grey seals Sea-scorpions are extremely infested with sealworm larvae off the West coast. The density is much higher than in any other fish species important in the food of grey seals (Table 4). However, being much lower, the infection in the long rough dab (Hippoglossoides platessoides limandoides) comes second (Table 5). Abundance of sealworms, in sea-scorpion, saithe (Pollachius virens) and plaice (Pleuronectes platessa), is much higher around Hvalseyjar Island, than in the coastal waters off Sn~efellsnes. This supports the idea of regional differences of sealworm infections of fish. The former location is close to a large breeding colony of grey seals, but the latter is not. The infections in long rough dab also differ between areas (Table 5).

Discussion

Status of the grey seal population The observed pup counts do not show a significant trend. A longer time series of the number of pups born annually is needed before any conclusion can be drawn about the status of the grey seal population in Iceland. Is it stable at 10,000-12,000 animals, or did it increase from the years 1982 to 1990, and decline from 1990 to 1992 (Table 1) ?

571

Final and intermediate host of the sealworm and environmental effects on the sealworm's life-cycle Grey seals seem to be the main final host for sealworm in Icelandic Waters. It is far more infected with sealworms than the common seal. Sea-scorpion seems to be the most important second intermediate host, as its sealworm density, as far as is known, exceeds the density in any other fish, and because of its relative importance in the food of grey seals. Sea-scorpions have also proved to be an important intermediate host in the western coastal waters of Norway [4]. Shallow rocky bottoms and large numbers of all necessary hosts in the life-cycle, living in close contact, seem to create perfect conditions for the sealworm in the West and Northwest coastal waters. The deeper sandy bottoms of the South coast seem to be less favourable to the dispersal of the sealworm. In this area, grey seals lose sealworms from their stomachs during breeding and do not accumulate new infections from local food species.

Future research More aerial censuses of grey seal pups will be carried out to investigate possible changes in the Icelandic grey seal population. The next census is planned in the autumn of 1995. Investigations of sealworm infections in fish species important in the diet of grey seals will continue. This includes further studies on the sea-scorpion in other areas of the coast. Nematode infections of catfish (Anarhichas lupus) will also be studied, as this species is important in the food of grey seals, but very little is known about its infections.

Acknowledgements Valur Bogason assisted in working out the food samples from the grey seals. This research is jointly sponsored by the fish sales organizations and fishing companies of Iceland.

References 1. Hauksson E. Aerial census of grey seal (Halichoerus grypus Fabricius) pups in Iceland in 1982. Ntittfrufra~Singurinn 1985;55:83-93 (in Icelandic with an English summary). 2. 61afsd6ttir D. Hringormar f meltingarvegi landsela og titsela vi8 strendur Islands (Nematodes in the digestive track of common seals and grey seals in Icelandic coastal waters). In: Hersteinsson P, Sigurbjarnarson G (eds) Villt Islensk Spender. Reykjav~: Hi8 Islenska Ntitttirufra~Sif61ag Landvernd, 1993;227-239 (in Icelandic). 3. Margolis L, Esch GW, Holmes JC, Kuris AM, Shad GA. The use of ecological terms in parasitol-

572 ogy (Report of an ad hoc committee of the American Society of Parasitologists). J Parasitol 1982;68:131-133. 4. Jensen T, Andersen K. The importance of sculpin (Myoxocephalus scorpius) as intermediate host and transmitter of the sealworm Pseudoterranova decipiens. Int J Parasitol 1992;22:665-668. 5. Hauksson E. Larval Anisakine Nematodes in Various Fish Species, from the coast of Iceland. Hafranns6knir 1992;43:107-122.

Pollutants, toxicology and epizootics


9 1995 Elsevier Science B.V. All fights reserved Whales, seals, fish and man A.S. Blix, L. Walllaeand ~. Ulltang, editors

575

Toxicological and epidemiological significance of pollutants in marine mammals Peter J.H. Reijnders 1 and Elze M. de R u i t e r - D i j k m a n 2 I IBN DLO, Den Burg, The Netherlands; and 2IBN DLO, Wageningen, The Netherlands A b s t r a c t . There is accumulating evidence from epidemiological and experimental research that the

"resilience" of marine mammals can be affected by contaminants. The expected prolonged existence of persistent pollutants already present in oceans and seas warrants intensified research on the impact of pollution on marine mammals. The establishment of a monitoring scheme based on multiple response assessment, investigations on synergistic effects of contaminants and toxic significance of new contaminants of concern are important elements in this research. It should be integrated in an assessment of general habitat degradation of marine mammal species, in which marine pollution is an important contributing factor. K e y words: cetaceans, pinnipeds, pollutants, trends, impact assessment

Introduction The principal objective of studying the interactions between pollution and marine mammals should be to understand the long term biological effects of contaminants on marine mammal behaviour and physiology. This implies more than studying anomalies and residue levels in animal tissue. The focus should rather be on assessing the impact of contaminants on the potential of marine mammals to recuperate from environmental fluctuations, and the way this translates into changes in populations. Comprehensive reviews on biomagnification rates, accumulation and concentrations of contaminants such as heavy metals and organochlorines, have been published [1-7]. Likewise, reviews on toxicokinetics, pathology and toxicology of different contaminants and several marine mammal species are available [8-14]. This paper is therefore not a comprehensive review of processes involved between exposure and excretion. It aims to evaluate available data on occurrence of contaminants and associated effects, from the perspective of the influence pollution has on the resilience of marine mammals. The basic questions to be answered are: what are the global spatial and temporal trends of levels of pollutants observed in marine mammal tissues, what is known about effects in marine mammals and how might this be extrapolated to other marine mammal species, and what are the major lacunae that remain in knowledge required to fully assess the impact of pollution on marine mammals?

Address for correspondence: P.J.H. Reijnders, Institute for Forestry and Nature Research (IBN DLO), Department of Aquatic Ecology, P.O. Box 167, 1790 AD Den Burg, The Netherlands.

576 Given the paucity of data on occurrence, let alone effects, of most types of marine pollution, only trace elements and organochlorines are discussed in more detail in this paper (see e.g. [4,8] for overviews on other types of pollution). This should not be interpreted as complete ignorance about potential effects of the other types of pollution, as will be elucidated further on.

Global Spatial Trends Trace elements Since the uptake of trace elements is predominantly dietary, it can be expected that regional differences in concentrations in prey species would be reflected in marine mammals. However, observed concentration levels in different species vary due to several factors, e.g. species-specific accumulation rates, compound-specific accumulation rates, differences in analytical techniques, individual-specific accumulation rates related to age and sex. Therefore, geographical comparison of pollution burdens in marine mammals should only be carried out in the same tissues, from the same species and from animals of known age and sex. The general pattern arising from such comparisons is that trace-element concentrations in marine mammals appear to be related to their feeding habits and area of exposure [1,5,15,16]. Cadmium, copper and zinc levels are higher in species that feed more on squid, compared to species that feed primarily on fish. This is largely due to the relative higher levels of these elements in squid, which is a general phenomenon in the marine environment [ 17]. Usually levels of trace elements are lower in baleen whales compared to toothed whales and considered to be primarily due to their feeding lower in the food web [5]. Intra-species variation is generally attributed to differences in exposure caused by variation in foraging strategies and locations [ 18,19]. Within the group of trace elements, only for mercury does a large data set of concentrations in marine mammals exist. Considering the geographical distribution of those data, there seems to be no clear latitudinal gradient in mercury levels in marine mammals. This is partly explained by the fact that mercury enters the environment via anthropogenic as well as geological sources. This leads to a rather patchy distribution of certain regions where levels are some orders of magnitude higher than in others. Examples include the relatively high levels in the Arctic, Irish Sea, Mediterranean, South Pacific, Oslofjord and Wadden Sea [ 1,15,16,20-24].

Organochlorines The global distribution of organochlorines (OCs) in marine mammals is demonstrated using data from literature and our own data on levels of PCBs as an example. The data are average concentrations for some species and are used in a rather qualitative way. In a natural situation the observed amoeba (Fig. 1) would have been one point, indicating PCB concentration 0. To compare spatial trends on a more

577

ARCTIC Dall's porpoise

Beluga 9 Grey seal

Whitesided dolphi Str iped dolphi Sea l

PACIFIC

i / "

o n / ~ ~ ~ ? i ~ i ? i ~ i i ~ , ~

\

( _N!i|174

Harbour seal

I

i N

_ .~ . Dusky dolphin

.x IX

...... 11. ........

o

"~Striped dolphin

ANTARCTIC Fig. 1. Global average PCB concentrations in marine mammals in the late 1970s, early 1980s. Concen-

trations are expressed relative to 50/~g 9g-1 lipid weight. Center, 0; inner circle, 50/~g 9g-l; outer circle, 7006tg" g-1 [25-29,53].

quantitative basis, it is necessary to take into account other factors that influence concentration levels of OCs. Like in the trace elements, the more prominent factors in that respect are age, sex and dietary concentration [7,30-32]. Also OC concentrations are generally higher in fish-eating species compared to squid-eating species because fish are more fatty and therefore carry higher OC burdens. If these sources of variation are taken into account, the levels of OCs are usually lower in baleen whales compared to toothed whales [5], largely due to their feeding on less contaminated prey items low in the food web. An additional factor is the lower per capita energetic intake which leads to dilution of lipophilic contaminants in these larger animals [33,34]. Tanabe et al. [35] published a comprehensive global survey on latitudinal distribution of OCs in the atmosphere, surface waters and tissues of marine mammals in the west Pacific. Their data on PCBs were compiled with other literature, including their own data, on concentrations in marine mammals from the east Atlantic (Fig. 2). It is concluded from these data that both in the west Pacific and east Atlantic, marine mammals distributed in the northern hemisphere are exposed to higher PCB concentrations than those in the southern hemisphere. This is confirmed by observations made for baleen whales [5]. The highest levels are usually found in the mid-latitudes, apparently related to the extensive use and production in the industrialized countries in these parts of the world.

578

East Atlantic 80N

West Pacific

'l I ......

60

i

I

80N

........ [ ]

9 O F1E] E~I~]~]]]O[~'Oq

--

[]

[JZ]

[~] 0

m OOD

-

6O

--

40

O

O~I

O O0

40

O

20

--

20

0

20 9

-

40

~o

20

-

40

-

60

9 t

r-]

--

seals

9 cetaceans I

80S

t_

80

,

, I

60

I

40

.....

l

20

80 S

.......

0

20

40

PCB (ug.g wet blubber weight) Fig. 2. Latitudinal ~PCB concentrations (~g 9g-1 wet weight) in blubber tissue of various marine mammals from the West Pacific and the East Atlantic [16,30,34,36-43].

Temporal Trends Trace elements

Glacial records have shown that heavy metals have been in the marine environment from prehistoric times [39-40]. As a result of anthropogenic emissions since the beginning of the present century, increases of for example mercury, lead and cadmium are particularly noticeable in the environment, including marine mammals and humans [ 18]. It is difficult to indicate a general temporal trend from which future trends could possibly be projected. Some major handicaps are lack of longer time series and poor comparability of data, due to differences in analytical as well as sampling procedures, rendering the picture rather incomplete. Despite these complications, it can be observed that in some areas, due to for example closure of plants or improvement of industrial and agricultural processes and sewage treatment,

579 declines of substances such as mercury, lead and cadmium have occurred. In particular, input levels for mercury have been reported to decline in certain areas, leading to a local reduction in concentrations in water and fish [18,46]. On a more regional scale, however, the apparent reduction in inputs has not as yet led to decreasing concentrations in marine biota in places such as Minamata [47] and the North Sea [48]. Concentrations in marine mammals follow this trend, as can be concluded from studies on cetaceans and seals [21,49,50].

Organochlorines Temporal trends for different OCs in marine biota have been established for several oceans and seas [7,51,52]. From a global point of view, Tanabe [53] concluded that PCB levels in marine biota are unlikely to decline in the near future, due to the fact that only 30% of all PCBs produced have so far dispersed into the environment [53]. Global budgets on an organic basis were calculated by Marquenie and Reijnders [28], using the compartmentation from Tanabe [53], but also including recent data on continued production in Europe and estimates for production in the Comecon States. From the more than 20 million tons of PCBs produced, more than 30% is still in use and that implies under control. It is essential that a stringent policy is developed to collect and adequately destroy those amounts. From the rest, only 1% has reached the oceans and 30% has accumulated in dump sites and sediments of lakes, coastal

PCBs (kTonnes) 2 0 - -

- unknown 15-

sea & o c e a n / w a t e r _ land & sea sed ime nt destroyed

10-: ............................... _

_

-"-"

. . . . . .

.

.....

. . . . . . .

.

5_.

IR u s e

O-

Fig. 3. Global budget of produced PCBs (kilotonnes) after Marquenie and Reijnders [28].

580 zones and estuaries (Fig. 3). Their future dispersal into oceans further strengthens Tanabe's conclusion [53] about the unlikeliness of future decline. It should be mentioned that, in areas close to the source of pollutants, due to restricted use and regulated disposal, PCB levels have declined [54,55]. However, from studies on biota from the Arctic [56], the North Sea [57,58] the Baltic Sea [52] and the Pacific [7], it became apparent that the decline levelled off between 1980 and 1985. Tateya et al. [59] presented a comprehensive report on temporal trends in PCB contamination, which started in the early 1930s. They predicted possible future trends in marine mammals, based on data from studies on striped dolphins, which would reach a peak between the years 2000 and 2030. If their data are combined with the estimate of Bletchly [60] that disposal of PCBs will peak at the end of the 1990s, it is postulated that at least until the turn of next century, no apparent reduction in the potential toxic impact of PCBs on marine mammals on a global scale can be expected. The reported global change in distribution of organochlorine residue levels is of special interest. Continued use of insecticides in developing countries and the slower deposition rate in tropical waters led to a prolonged exposure of tropical marine biota. Particularly Arctic waters and adjacent seas and oceans presumably become the major sink for OCs [7].

Epidemiological and Experimental Findings, and Risk Assessment Epidemiological findings There is a suite of epidemiological findings associated with contaminants (Table 1). It is beyond the context of this paper to review the numerous publications on this subject, therefore reference is made to recent reviews [4,6,7,11-13]. The general conclusion from the available literature is that both heavy metals and OCs have been associated with reproductive and immunological disorders in marine mammals, in particular in seals, belugas and small cetaceans. In most studies usually a mixture of compounds have been involved which rendered it impossible to assign an observed

Table 1. Epidemiological findings associated with contaminants Immune dysfunctions Reproductive failure Premature pupping Stillbirths Stenosis Occlusions Osteoporosis Exotosis Epizootics Testosterone reduction Liver disease

Beluga, harbour, grey and ringed seal Harbour and ringed seal Californian sea lion Saima seal Ringed seal Ringed seal Harbour and grey seal Harbour seal Bottlenose, striped and common dolphin, harbour and Baikal seal Dall' s porpoise Bottlenose dolphin

581 Table 2. Experimental findings in harbour seals related to contaminants

Implantation failure Reduced T-cell function Reduced NK activity

Reduced vitamin A level Reduced thyroid hormone level Reduced cortisol level

effect uniquely to a single compound. It is apparent that OCs are the predominant compounds involved, but this may be partly influenced by the fact that they are currently the focus of most ecotoxicological research on marine mammals.

Experimental findings Although there have been many studies made on concentrations of contaminants in marine mammals, the assessment of their physiological effects is often restricted to working hypotheses, given the other, often confounding, factors. Only experimental research under controlled conditions could provide the necessary evidence to establish a cause and effect relationship. Experimental findings obtained from studies [61-63] in which harbour seals were exposed to contaminants in prey they would normally encounter in areas where these fish came from, are given in Table 2. Again here, the problems are of a reproductive or immunological nature. Caution is needed in the interpretation of the different results. With respect to implantation failure, the impact is significant. However, even the reduced hormone levels and the immunerelated parameters, are one step removed from measuring lowered resilience in the wild.

Risk assessment It is also emphasized that the indicators found in seal research cannot be directly extrapolated to other seal species or cetaceans. Apart from the differences in feeding strategy and therefore exposure, most species differ in their response to given compounds and pathogens. Indeed, both in cetaceans and pinnipeds, the dominant system in metabolizing OCs is the P450 enzyme system. That system can be induced by PCBs, mediated by the arylhydrocarbon (Ah) receptor, which has been found in many mammals and birds. However, there is a clear difference in metabolic capacity between cetaceans and seals, between species of both groups, and even between individuals [64]. Some cetaceans lack or possess a lower potential of certain liver microsomal enzymes (Pb-type) than some seal species. Ringed seals and harbour porpoises seem to have metabolic capacities intermediate to those of other cetaceans and seals [10,11]. Species-specific response was also found in in vitro studies with harbour porpoise and harbour seal [65]. Instead of measuring enzyme activity to detect biotransformation capacity, it is also possible to compare the concentration of a given PCB-congener relative to the concentration of a reference (persistent) congener in prey, with the same ratio in a predator [6,66]. PCB-153 is usually chosen as reference congener because of its persistence. Since it is also the dominant

582

Rrel C B 52

[]

B

O~k

~2

9

Rrel C B 149 2.

[] u~k o

oi Rrel C B 101 2

o

0

o

Harbour porpoise

o

Harbour seal 9 Ringedseal Whitebeakeddolphin 9 Striped dolphin

z~ D a l i s porpoise

m Bottlenose dolphin

9 Spinner dolphin

Fig. 4. Relative ratio (Rrel) of PCB congeners 52, 101 and 149 in blubber tissue of different marine mammals. Basic data are derived from [6,23,39,40,64,66--69]. *Indicates only one sample in that specific data point. Rrel = [(CB-X)/(CB-153)]predator/[(CB-X)/(CB-153)]prey.

congener, the ratio CB-X/CB-153 is always lower than 1. A relative ratio (Rr~n) between predator and prey can be obtained by dividing the ratios of both. An Rre 1 distinctly smaller than 1 indicates that metabolization occurs. If that ratio is about equal to or higher than 1, no conclusion on either metabolization or the lack thereof can be drawn. This approach has been applied to the available literature and our own

583 data. The relative ratios of three PCB congeners in different marine mammal species are shown in Fig. 4. The results indicate for each congener that differences between species and even within species from different areas exist. It is furthermore concluded that the two porpoise species differ in metabolization capacity, as was already postulated in Reijnders [6]. The results also demonstrate that besides the harbour porpoises, other cetaceans also possess microsomal Pb-type enzymes. The relative ratios should not be interpreted as absolute values to quantify biotransformation capacity. This is a tentative attempt with diverse data, which could for some data incorporate a certain problem of accuracy comparability. Irrespectively, the latter quantification would anyhow require information on the actual species composition of the consumed prey. This is necessary to check the important prerequisite for this approach that the pattern for persistent PCBs in prey and predator is not significantly different [70]. The conclusion from this chapter is that differences in biotransformation capacity exist between marine mammal species and even between individuals within species. However, as has previously been pointed out [7], this does not per se allow the development of a toxico-vulnerability index for marine mammal species according to their metabolization capacity. That finally depends on the balance between the "toxic costs" of continued induction by persistent congeners and reactive intermediates, and the "revenues" of decreased dioxine type of toxicity. To that end, further research is needed to construct a matrix in which are incorporated, for different species, the relative toxic impact for different types of toxicity, their relative frequency of occurrence and most of all their ultimate effect on physiological processes, in particular the endocrine control system.

Concluding Summary There are strong indications that pollutants can have a significant impact on marine mammals. However, the exact threats are difficult to quantify. This is mainly due to lacunae in knowledge of actual physiological effects of known contaminants, perception of possible threats of unknown, or hitherto scarcely studied, contaminants and knowledge of future trends in global pollution and contaminant disposal. Some of these questions can be addressed by designing a conceptual framework for prediction of pollution impact on marine mammals, based on multiple response assessment as elaborated in Reijnders [6]. Next to this, it is evident that, except for mercury, trace elements are under-represented in ecotoxicological marine mammal research. Monitoring of global trends in OC levels is considered of importance since it provides information on persistence of longer-term threats of these compounds and the expected shift in distribution. In particular research on global distribution and toxicity of less routinely determined compounds in marine mammals is required. These include DDE- and PCB-methylsulphones, PAHs, chlordanes, toxaphenes, naphthalenes, tris(4-chlorophenyl) methanol and tris(4-chlorophenyl) methane, which have already been determined in marine mammal tissue from different parts of

584 the w o r l d a n d s o m e o f w h i c h are k n o w n for t h e i r t o x i c o l o g i c a l a n d p a t h o l o g i c a l impact [ 13,23,71-75].

Acknowledgements W e a p p r e c i a t e the c o l l a b o r a t i o n w i t h A r n e BjCrge f r o m t h e N o r w e g i a n I n s t i t u t e for N a t u r e R e s e a r c h ( N I N A ) . T i s s u e s a m p l e s o f h a r b o u r p o r p o i s e s o b t a i n e d via t h e N o r w e g i a n M a r i n e M a m m a l P r o g r a m m e h a v e b e e n a n a l y s e d at o u r i n s t i t u t e a n d w e w e r e k i n d l y a l l o w e d to use s o m e o f the d a t a for this p a p e r . W e a c k n o w l e d g e t h e c o m m e n t s o f M a r k S i m m o n d s , A l e x A g u i l a r a n d S o p h i e B r a s s e u r on an e a r l i e r draft o f this p a p e r .

References 1. Wagemann R, Muir DCG. Concentrations of heavy metals and organochlorines in marine mammals of northern waters: overview and evaluation. Can Techn Rep Fish Aquat Sci 1984;1297. 2. Law RJ, Fileman CF, Hopkins AD, Baker JR, Harwood J, Jackson DB, Kennedy S, Martin AR, Morris RJ. Concentrations of trace metals in the livers of marine mammals from the Welsh coast and the Irish sea. Mar Poll Bull 1991;22:183-191. 3. Bowles D. An overview of the concentrations and effects of heavy metals in cetacean species. IWC/SC/46/O20, 1994. 4. Moscrop A, Simmonds MP. The significance of pollution for marine cetaceans. IWC/SC/46/O14, 1994. 5. O'Shea TJ, Brownell RL. Organochlorine and metal contaminants in baleen whales: a review and evaluation of conservation implications. Sci Total Environ 1994; 154:179-200. 6. Reijnders PJH. Toxicokinetics of chlorobiphenyls and associated physiological responses in marine mammals, with particular reference to their potential for ecotoxicological risk assessment. Sci Total Environ 1994; 154:229-236. 7. Tanabe S, Iwata H, Tatsukawa R. Global contamination by persistent organochlorines and their ecotoxicological impact on marine mammals. Sci Total Environ 1994;154:163-177. 8. Geraci J, St. Aubin DJ. Sea Mammals and Oil: Confronting the Risks. San Diego, CA: Academic Press, 1990. 9. Tanabe S, Tatsukawa R. Persistent organochlorines in marine mammals. In: Jones KC (ed) Organic Contaminants in the Environment: Environmental Pathways and Effects. London: Elsevier, 1991 ;275-289. 10. Boon JP, Van Arnhem E, Jansen S, Kannan N, Petrick G, Duinker JC, Reijnders PJH, Goks0yr A. The toxicokinetics of PCBs in marine mammals with special reference to possible interactions of individual congeners with the cytochrome P450-dependent monooxygenase system: an overview. In: Walker CH, Livingstone DR (eds) Persistent Pollutants in Marine Ecosystems. Oxford: Pergamon Press, 1992;119-159. 11. Norstrom RJ, Muir DCG. Chlorinated hydrocarbon contaminants in arctic marine mammals. Sci Total Environ 1994;154:107-128. 12. Martineau D, Deguise S, Girard C, Lagac6 A, B61and P. Pathology and toxicology of beluga whales from the St. Lawrence estuary, Qu6bec, Canada. Sci Total Environ 1994;154:201-215. 13. Olsson M, Karlsson B, Ahnland E. Diseases and environmental contaminants in seals from the Baltic and the Swedish west coast. Sci Total Environ 1994;154:217-227. 14. Reijnders PJH. Contaminants and cetaceans: reasons for concern. IWC/SC/46/O8, 1994.

585 15. Hansen CT, Nielsen CO, Dietz R, Hansen MM. Zinc, Cadmium, Mercury and Selenium in Minke Whales, Belugas and Narwhals from West Greenland. Polar Biol 1990;10:529-539. 16. Skaare, JU, Markussen NH, Norheim G, Haugen, Holt G. Levels of polychlorinated biphenyls, organochlorine pesticides, mercury, cadmium, copper, selenium, arsenic and zinc in the harbour seal, Phoca vitulina, in Norwegian waters. Environ Pollut 1990;66:309-324. 17. Ftirstner U, Wittmann GTW. Metal Pollution in the Aquatic Environment. Berlin: Springer, 1979. 18. Wagemann R, Stewart REA, B61and P, Desjardins C. Heavy metals and selenium in tissues of beluga whales, Delphinapterus leucas, from the Canadian Arctic and the St. Lawrence estuary. Can Bull Fish Aquat Sci 1990;224:191-206. 19. Caurant F, Amiard JC, Amiard-Triquet C, Sauriau PG. Ecological and biological factors controlling the concentrations of trace elements (AS, Cd, Cu, Hg, Se, Zn) in delphinids, Globicephala melas, from the north Atlantic ocean. Mar Ecol Prog Ser 1994; 103:207-219. 20. Reijnders PJH. Organochlorine and heavy metal residues in harbour seals from the Wadden Sea and their possible effects on reproduction. Neth J Sea Res 1980;14:30--65. 21. Law RJ, Whinnett JA. Polycyclic aromatic hydrocarbons in muscle tissue of harbour porpoises, Phocoena phocoena, from UK waters. Mar Poll Bull 1992;24:550-553. 22. Leonzio C, Focardi S, Fossi C. Heavy metals and selenium in stranded dolphins of the Northern Tyrrhenian (NW Mediterranean). Sci Total Environ 1992;119:77-84. 23. Muir DCG, Wagemann R, Hargrave BT, Thomas DJ, Peakall DB, Norstrom RJ. Arctic marine ecosystem contamination. Sci Total Environ 1992; 122:75-134. 24. Marcovecchio JE, Gerpe MS, Bastida RO, Rodriguez DH, Mor6n SG. Environmental contamination and marine mammals in coastal waters from Argentina: an overview. Sci Total Environ 1994;154:141-151. 25. DeLong RL, Gilmartin WG, Simpson JG. Premature births in Californian sealions: association with high organochlorine pollutant residue levels. Science 1973; 181:1168-1170. 26. Helle E, Olsson M, Jensen S. PCB levels correlated with pathological changes in seal uteri. Ambio 1976;5:261-263. 27. Martineau D, B61and P, Desjardins C, Lagac6 A. Levels of organochlorine chemicals in tissue of beluga whales, Delphinapterus leucas, from the St Lawrence estuary, Qu6bec, Canada. Arch Env Contam Toxicol 1987;16:137-147. 28. Marquenie JM, Reijnders PJH. PCBs, an increasing concern for the marine environment. CM, 1989, N: 12. Copenhagen: ICES, 1989. 29. Broekhuizen S. First data on contamination of otters in The Netherlands. Bull IUCN Otter Spec Group 1987;2:27-32. 30. Aguilar A, Borrell A. Age- and sex-related changes in organochlorine compound levels in fin whales, Balaenoptera physalus, from the Eastern North Atlantic. Mar Environ Res 1988;25:195211. 31. Reijnders PJH. Accumulation and body distribution of xenobiotics in marine mammals. In: Salomons W, Bayne BL, Duursma EK, FiSrstner U (eds) Pollution of the North Sea: an Assessment. Heidelberg: Springer, 1988;596-603. 32. Subramanian A, Tanabe S, Tatsukawa R. Use of organochlorines as chemical tracers in determining some reproductive parameters in Dalli-type Dali's porpoises Phocoenoides dalli. Mar Environ Res 1988;25:161-174. 33. Aguilar A. Compartmentation and reliability of sampling procedures in organochlorine pollution surveys of cetaceans. Residue Rev 1985;14:349-352. 34. Borrell A. PCB and DDTs in blubber of cetaceans from the northeastern North Atlantic. Mar Poll Bull 1993;26:146-151. 35. Tanabe S, Mori T, Tatsukawa R, Miyazaki N. Global pollution of marine mammals by PCBs, DDTs and HCHs (BHCs). Chemosphere 1983;12:1269-1275. 36. Baumann Ofstad E, Martinsen K. Persistent organochlorine compounds in seals from Norwegian coastal waters. Ambio 1983;12:262-264.

586 37. Morris RJ, Law RJ, Allchin CR, Kelly CA, Fileman CF. Metals and organochlorines in dolphins and porpoises of Cardigan Bay, West Wales. Mar Poll Bull 1989;20:512-523. 38. Luckas B, Vetter W, Fischer P, Heidemann G, P16tz J. Characteristic chlorinated hydrocarbon patterns in the blubber of seals from different marine regions. Chemosphere 1990;21:13-19. 39. Granby K, Kinze CC. Organochlorines in Danish and West Greenland harbour porpoises. Mar Poll Bull 1991 ;22:458-462. 40. Wells DE, Echarri I. Determination of individual chlorobiphenyls (CBs), including non-ortho, and mono-ortho chloro substituted CBs in marine mammals from Scottish waters. Int J Environ Anal Chem 1992;47:75-97. 41. Simmonds MP, Johnston PA, French MC. Organochlorine and mercury contamination in United Kingdom seals. Vet Rec 1993;132:291-295. 42. Borrell A, Aguilar A, Grau E, Gonzalez LM, Lopez-Jurado LM, San Felix M, Lopez-Jurado LF, Hernandez M. DDT and PCB levels in Mediterranean monk seals, Monachus monachus. Proceedings 21st Annual Symposium for Aquatic Mammals, Madrid, 1993. 43. De Kock AC, Best PB, Cockcroft V, Bosma C. Persistent organochlorine residues in small cetaceans from the east and west coasts of southern Africa. Sci Total Environ 1994;154:153-162. 44. Murozumi M, Chow TJ, Petterson C. Chemical concentrations of pollutant lead aerosols, terrestrial dusts and sea salts in Greenland and Antartic snow strata. Geochim Cosmochim Acta 1969;33:1247-1294. 45. Weiss HV, Koide M, Goldberg ED. Mercury in a Greenland icesheet: evidence of recent input by man. Science 1971;174:692-694. 46. Franklin A. The concentration of metals, organochlorine pesticide and PCB residues in marine fish and shellfish: results from MAFF fish and shellfish monitoring programmes, 1977-1984. Aquatic Environment Monitoring Report No 16. MAFF, 1987. 47. Tsubaki T, Irukayama K. Minimata Disease. Tokyo/Amsterdam: Kodansha/Elsevier, 1980. 48. NSTF. Quality Status Report of the North Sea. The Hague: North Sea Task Force, 1993. 49. Reijnders PJH. Organochlorine and heavy metal residues in harbour seals from the Wadden Sea and their possible effects on reproduction. Neth J Sea Res 1980; 14:30-65. 50. Lang D. Untersuchungsergebnisse zur Schadstoffbelastung von Seehunden, Phoca vitulina L., aus dem Wattenmeer, Schleswig-Holstein. S~iugetierk Mitt 1985;32:281-292. 51. Loganathan BG, Kannan K. Time perspectives of organochlorine contamination in the global environment. Mar Poll Bull 1991 ;22:582-584. 52. Bingert A, G6thberg A, Jensen S, Litz6n K, Odsj6 T, Olsson M, Reutergardh L. The need for adequate biological sampling in ecotoxicological investigations: a retrospective study of twenty years pollution monitoring. Sci Total Environ 1993;128:121-139. 53. Tanabe S. PCB problems in the future: foresight from current knowledge. Environ Poll 1988;50:528. 54. OECD. Chemical trends in wildlife: an international co-operative study, No. 700 DH 97 80 06 1. OECD, 1980. 55. Olsson M, Reutergardh L. DDT and PCB pollution trends in the Swedish aquatic environment. Ambio 1986;15:103-109. 56. Norstrom RJ, Simon M, Muir DCG, Schweinsburg RE. Organochlorine contaminants in Arctic marine food chains: identification geographical distribution and temporal trends in polar bears. Environ Sci Technol 1988;22:1063-1071. 57. De Boer J. Trends in chlorobiphenyl contents in livers of Atlantic cod (Gadus morhua) from the North Sea, 1979-1987. Chemosphere 1988;17:1811-1819. 58. De Boer J. Spatial differences and temporal trends of bioaccumulating halogenated hydrocarbons in livers of Atlantic cod (Gadus morhua) from the North Sea, 1977-1992. Proceedings Quality Status Report-Scientific Symposium, Ebeltoft, Denmark, April 1994. 59. Tateya S, Tanabe S, Tatsukawa R. PCBs on the globe: possible trend of future levels in the open ocean environment. In: Schmidtke NW (ed) Toxic Contaminants in Large Lakes. Proceedings

587

60.

61. 62.

63.

64. 65.

66.

67.

68. 69.

70.

71.

72. 73. 74.

75.

World Conference on Large Lakes, May 1986, Mackinac Island, Michigan, US. Lewis Publishers Inc, 1988;237-281. Bletchly JD. Polychlorinated biphenyls: production, current use and possible rate of future disposal in OECD member countries. In: Barros MC, KSnemann H, Visser R (eds) Proceedings of PCB Seminar. The Hague: Ministry of Housing, Physical planning and Environment, 1984;343372. Reijnders PJH. Reproductive failure in common seals feeding on fish from polluted coastal waters. Nature 1986;324:456--457. Brouwer A, Reijnders PJH, Koeman JH. Polychlorinated biphenyl (PCB)-contaminated fish induces vitamin A and thyroid hormone deficiency in the common seal, Phoca vitulina. Aquat Toxicol 1989;15:99-106. De Swart RL, Ross PS, Vedder EJ, Timmerman HH, Heisterkamp SH, Van Loveren H, Vos JG, Reijnders PJH, Osterhaus ADME. Impairment of immunological functions in harbour seals, Phoca vitulina, feeding on fish from polluted coastal waters. Ambio 1993;23:155-159. Tanabe S, Watanabe S, Kan H, Tatsukawa R. Capacity and mode of PCB metabolism in small cetaceans. Mar Mammal Sci 1988;4:103-124. Murk A, Morse D, Boon J, Brouwer A. In vitro metabolism of 3,3', 4,4'-tetrachlorobiphenyls in relation to ethoxyresorufin-O-deethylase activity in liver microsomes of some wildlife species and rat. Eur J Pharmacol Environ Toxicol Pharmacol Sec 1994;270:253-261. Boon JP, Oostingh I, Van der Meer J, Hillebrand MTJ. A model for the bioaccumulation in chlorobiphenyl congeners in marine mammals. Eur J Pharmacol Environ Toxicol Pharmacol Sec 1994;270:237-251. Duinker JC, Zeinstra T, Hillebrand MTJ, Boon JP. Individual chlorinated biphenyls and pesticides in tissues of some cetacean species from the North Sea and the Atlantic Ocean: tissue distribution and biotransformation. Aquat Mamm 1989; 15:95-124. Morris RJ, Law RJ, Allchin CR, Kelly CA, Fileman CF. Metals and organochlorines in dolphins and porpoises of Cardigan Bay, West Wales. Mar Poll Bull 1989;20:512-523. Norstrom RJ, Muir DCG, Ford CA, Simon M, Macdonald CR, B61and P. Indications of P450 monooxygenase activities in beluga, Delphinapterus leucas, and narwhal, Monodon monoceros, from patterns of PCB, PCDD and PCDF accumulation. Mar Environ Res 1992;34:267-272. Boon JP, Eijgenraam F, Everaarts JM, Duinker JC. A structure-activity relationship (SAR) approach towards metabolism of PCBs in marine animals from different trophic levels. Mar Environ Res 1989;27:159-176. Walker W, Riseborough RW, Jarman WM, Lappe BW, Tefft JA, DeLong RL. Identification of tris (chlorophenyl)methanol in blubber of harbour seals from Puget Sound. Chemosphere 1989; 18:1799-1804. Hellou J, Stenson G, Ni I-H, Payne JF. Polycyclic aromatic hydrocarbons in muscle tissue of marine mammals from the northwest Atlantic. Mar Poll Bull 1990;21:469-473. Law RJ, Whinnett JA. Polycyclic aromatic hydrocarbons in muscle tissue of harbour porpoises (Phocoena phocoena) from UK waters. Mar Poll Bull 1992;24:550-553. Zook DR, Buser HR, Bergqvist PA, Rappe C, Olsson M. Detection of tris (chlorophenyl)methane and tris (4-chlorophenyl)methanol in ringed seal (Phoca hispida) from the Baltic Sea. Ambio 1992;21:557-560. De Boer J, Wester PG. Determination of toxaphene in human milk from Nicaragua and in fish and marine mammals from the northeastern Atlantic and the North Sea. Chemosphere 1993;27:18791890.


9 1995 Elsevier Science B.V. All rights reserved Whales, seals, fish and man A.S. Blix, L. WallCeand 13. Ulltang, editors

589

Organochlorine contaminants in marine mammals from the Norwegian Arctic Janneche Utne Skaare Norwegian College of Veterinary Medicine~The Norwegian State Veterinary Laboratories (NCVM /NSVL), Oslo, Norway Abstract. Selected organochlorine contaminants (OCs) have been determined in a large number of different marine mammals in the Norwegian Marine Mammal Programme (NMMP). Alarmingly high PCB levels were found in polar bear at Svalbard (mean 25 ppm in fat) and in harbour porpoise along the Norwegian coast (mean 20 ppm in blubber). Low OC levels were found in ringed and harp seal (3 ppm PCB in blubber), while somewhat higher levels were found in grey seal, particularly in Varanger near the Russian border (6 ppm PCB in blubber) and in hooded seal from the West Ice (5 ppm PCB in blubber). However, the intraspecies variations were large. The average blubber level in minke whale was about 3 ppm PCB, but geographical differences in OC pattern were observed. Geographical differences in OC levels were registered with decreasing contamination from south to north in harbour seal and porpoise along the Norwegian coast, and increasing levels for harp seal from the West Ice to the East Ice.

Key words: PCB, DDT, marine mammals, Norwegian waters, Norwegian Arctic

Introduction The lipophilicity and persistence of organochlorine contaminants (OCs) contribute to their potential for bioaccumulation and biomagnification in nature, resulting in high OC levels particularly in top predator species of the marine food webs, such as odontocetes, phocides, polar bear and polar fox [1-6]. Between 1988 and 1994, mostly through the Norwegian Marine Mammal Programme (NMMP), a substantial number of marine mammals caught along the coast of Norway and in the Norwegian Arctic have been sampled for determination of OC contamination. The data produced, elucidating differences in OC levels and patterns with regard to species, age, sex, nutritional and health status, season of sampling, feeding preferences, geography etc., is or will be internationally published. The objective of this presentation is to provide some of the main results. Materials and Methods Materials Species and number of individuals examined for OC contamination are listed in Table 1. In Fig. 1, the different sampling locations are indicated on a map. Address for correspondence: Norwegian College of Veterinary Medicine/The Norwegian State Veterinary Laboratories (NCVM/NSVL), P.O. Box 8156, dep., 0033 Oslo, Norway

590 Table 1. List of different marine mammals analysed for OC contaminants between 1988 and 1994

Species

Location

Year

N

Harbour seal

Oslo fjord Southern Norway Western Norway Jarfjord, Finnmark Vester~len Jarfjord, Finnmark Jarfjord, Finnmark Vikna, TrCndelag Froan, TrCndelag Froan, TrCndelag Froan, TrCndelag Froan, TrCndelag Skjhnes, Finnmark Jarfjord The West Ice area Barents Sea, north Barents Sea, north The East Ice area Jarfjord Barents Sea, north Svalbard

1988 1988 1988 1988/1989 1990 Oct 1989 Jan/Mar 1990 Mar 1989 1991 1992 1992 1993 Feb 1988/1989 Jan/Mar 1990 Mar 1990 Sep 1990 June/July 1992 Apr/May 1993 Jan/Mar 1990 Sep 1990 1992

35 26 18 9 8 14 10 6 17 7 61 42 13 38 20 22 10 42 12 5 13

The West Ice area Svalbard Svalbard Kattegatt Western Norway Tufjord, Finnmark Barents/Finnmark Barents/Finnmark Barents/Finnmark Svalbard Svalbard Svalbard

Mar. 1990 1991 1992 1988/1989 1988/1989 1988/1989 1988/1989 1992 1993 1978-1989 1992 1993

27 16 53 12 15 7 37 72 64 24 33 40

Norwegian coast

1989-1993

17

Phoca vitulina

Grey seal Halichoerus grypus

Harp seal Phoca groenlandica

Ringed seal Phoca hispida

Sum

96

157

145 30

Hooded seal Cystophora cristata

Walrus Odobenus r. rosmarus

Harbour porpoise Phocoena phocoena

Minke whale Balaenoptera acutorostrata

Polar bear Ursus maritimus

Whales Different species Total number of animals

27 69 34 173 97 17 845

Location of catch, year and/or time of year, number of animals are given. Most of the individuals have been characterized with respect to sex, age, health and nutritional status.

M o s t animals were killed for scientific purposes, except for h a r b o u r seals in 1988 (found d e a d during an epizootic), grey seals f r o m F r o a n 1 9 9 1 - 1 9 9 3 (blood samples and b l u b b e r biopsies taken f r o m live animals), walrus (blubber biopsies taken f r o m live animals), polar bears 1 9 9 2 - 1 9 9 3 (blood samples and b l u b b e r biopsies taken coast.

from

live animals)

and

different whale

species

found

dead

along

the

591 40*W

......

'. i:..:GRE~NL

30*W

20*W

100W

0~

..... . . . .

10~

200E

..........!!i" i:ii

~Jan

~

30*E

~t

1

50*E

1 !

J Sk'~es J x,~ t_

9

40*E

75 ~

70oN THE

"

I

Fig. 1. Map showing the different sampling locations for marine mammals from the Norwegian coast (mid-Norway and northward), Norwegian Arctic areas, and the West Ice listed in Table 1.

Analytical procedure The methods used are described in Bernhoft and Skaare [7], where details on analytical quality assurance are also given. The following OCs were determined" 22 PCB-congeners (ZPCB), IUPAC nos: -28, -52, -74, -99, -101, -105, -110, -114, - 1 1 8 , - 1 2 8 , - 1 3 8 , - 1 5 3 , - 1 5 6 , - 1 5 7 , - 1 7 0 , - 1 8 0 , - 1 8 7 , - 1 9 4 , - 2 0 6 and-209; 5 DDT components and metabolites (ZDDT): p,p'-DDT, p,p'-DDE, p,p'-DDD, o,p'-DDT and o,p'-DDD; chlordanes (ZCHL): heptachlor, heptachlor epoxide, oxy-chlordane and trans-nonachlor; hexa-chlorocyclohexanes (ZHCH): a-HCH, fl-HCH and yHCH; and hexachlorobenzene (HCB).

Results and Discussion

Interspecies differences in OC levels Many factors such as age, sex, season (nutritional and reproductive status, feeding habits) are known to influence the levels and pattern of lipophilic and persistent contaminants. Comparison of OC levels between species and groups of individuals should always be done with this knowledge in mind. Figure 2 shows the distribution of ~:PCBs, ZDDTs in blubber/adipose tissue in marine mammals from the Norwegian Arctic compared to some terrestrial mammals and Norwegian women

592

24

POI.AR BEAR - SVAI.BARD, 1992/93 (8)

FN ARCTICFOX-SVALBAP.~ (6) ItARBOUR PORI~ISE- FINNMARK 1988/89 (1)

22 ,,,.,

o ~

~

20

GREY SEAL- VARANGER, 1989/90 (10) ~ 1 HARP SEAL- VARANGER, JAN/FEB 1990 (9) itARP SEAL - THE WEST ICE, I:F~ 1990 (Espeland & Skaare, in prep.)

18

HOODED SEAL - THE WEST ICE, FEB 1990 (Espeland & Skaare, in prep.)

16

MINKE WHALE- BARENTS SF.ak,1992 (10) [ - 7 NORWEGIAN WOMEN- 1992 (11)

.--= 14 ~

12

~

10

e

8

HARE- NORWAY (12)

E

2-PCB

E-DDT

Fig. 2. Mean levels of EPCB and EDDT in blubber/adipose tissue from different marine mammals from the Norwegian Arctic compared to corresponding levels in Norwegian women and hare. All samples but the samples of harbour porpoise have been analysed using basically the same method of EPCB and ZDDT identification and quantification at the Norwegian College of Veterinary Medicine/National Veterinary Institute. [1,6,8-12]. The interspecies differences found in levels and patterns reflect differences in metabolizing capacity and exposure. Thus, not unexpectedly, the marine mammals generally contained much higher levels of OCs compared to the terrestrial hare (Lepus timidus) [12] and humans (Norwegian women) [11]. Alarmingly high PCB levels were found in polar bear, which is a top predator in the marine food web, eating almost exclusively blubber of different seal species. The arctic fox (Alopex lagopus) from Svalbard is partly dependent on marine food sources and contains higher amounts of EPCBs than fox feeding on terrestrial food only [6]. The higher levels of ZPCBs and ZDDTs found in harbour porpoise compared to the much lower corresponding levels in grey seals probably indicate differences in toxicokinetics since both species feed on coastal fish [ 1]. Ranges of ZPCBs and ZDDTs levels (mg/kg wet weight) in blubber from five different seal species caught on the coast of Finnmark and in the West Ice are given in Fig. 3 [9,10]. The mean levels in all groups are lower compared to the corresponding levels found in seal species from the Baltic [13] and the Wadden Sea [3], and are generally below 10 mg/kg. The difference in OC levels found in adult female harp

593

Ice 1990, n=19 Harp seal, the West Ice 1990, n=10

,. ,..

,, ,, ,, ,. ,,..

,, ,, . . . . .

Harp seal, Jarfiord 1990, n=38

Harp seal, SkjAnes 1988/89, n=|3

. . , ,

. . . .

.

.

.

.

.

.

.

.

,

.

.

R )-PCB }DDT I Mean

,,, ,, ,, ,,,

.

.

.

.

~,~_ :~_~

_

~ .>~~,~.. ~ , ! . ~.....~..... . . . . ~ : . , .,.~ ~ . .................... [ 6 "/.?; 6;.'.';';4 ~ / / / / / / / / / / / / / / 9 J 9

.

~

~

.

@

.

.

.

~

~;>'":'~":"%':" v

t,-.

':

._,, --,-2/,-?-,'-)"

,,, "L'~: 2'

Whales, Seals, Fish and Man

Whales, Seals, Fish and Man

Seals

Seals

Seals

Seals

Seals and Sealing

Great Whales

Whales and Dolphins (Eye Wonder)

Seals (Endangered!)

Whales Whaling and Ocean Ecosystems

Whales, Whaling, and Ocean Ecosystems

Hydrocarbon Seals

Hydrocarbon Seals

Whales, Whaling, and Ocean Ecosystems

Fish

Fish

CIA Flaps And Seals Manual

Fish

Multilingual Dictionary of Fish and Fish Products

Behaviour and Physiology of Fish (Fish Physiology)

Fish is Fish

Save the Whales Please

US Navy SEALs

Dangerous Games (Tempting SEALs)

Hidden Agendas (Tempting SEALs)

The Seals Surprise Baby

The Wind Whales of Ishmael

Fish and amphibians

Nutrition and Fish Health

Fish Cognition and Behavior (Fish and Aquatic Resources)

Whales, Seals, Fish and Man