Population Dynamics of Whitefish ( Coregonus Suidteri Fatio) in Artificially Oxygenated Lake Hallwil, with Special Emphasis on L

Home , Coregonus fera, Coregonus lavaretus, Hallwil, Lake Hallwil

Diss. ETH No. 13706

Population dynamics of whitefish ( Coregonus suidteri Fatio) in artificially oxygenated Lake Hallwil, with special emphasis on larval mortality and sustainable management

Dissertation submitted to the

SWISS FEDERAL INSTITUTE OF TECHNOLOGY, ZURICH for the degree of Doctor of Natural Sciences

presented by Carole Andrea Enz Dipl. Natw. ETH Zurich bon1 August 3, 1972

Citizen of Sch(1nholzerswilen (TG), Switzerland

accepted on the recommendation of Prof. Dr. J. V. Ward, examiner Prof. Dr. H. Lehtonen, co-examiner Dr. R. Muller, co-examiner

Kastanienbaum, 2000 Meinen Eltern und Max Copyright ~;i 2000 by Carole A. Enz, EA WAG Kastanienbaurn

All rights reserved. No part of this book rnay be reproduced, stored in a retrieval systen1 or transmitted, in any fonn or by any ineans, electronic, rnechanical, pho- tocopying, recording or otherwise, without the prior written permission of the copyright holder.

First Edition 2000 PUBLICATIONS

CHAPTER 3 OF THE THESIS HAS BEEN ACCEPTED FOR PUBLICATION:

ENZ, C. A., SCHAFFER E. & MOLLER R. Growth and survival of Lake Hallwil whitefish (Co reg onus suidteri) larvae reared on dry and live food. - Archiv fUr Hyclrobiologie.

CHAPTERS 4, 5, 6 Ai~D 7 OF THESIS HA VE BEEN SUBMITTED FOR PUBLICATION:

ENZ, C. A., MBWENEMO BIA, M. & MULLER. R. Fish species diversity of Lake Hallwil (Switzerland) in the course of eutrophication, with special reference to whitefish ( Coregonus suidteri). Submitted to Conservation Biology.

ENZ, C. A .. SCHAFFER, E. & MULLER, R. Importance of prey movement, food particle and tank circulation for rearing Lake Ha11wil whitefish (Coregonus suidteri) larvae. Submitted to North Alnerican Journal of Aquaculture.

ENZ, C. A., BtTRGl, H. R., STOSSEL, F. & M(LLER, R. Food preference of whitefish (Coregonus suidteri) in eutrophic Lake Hallwil (Switzerland), and the question of cannibalism. Submitted to Freshwater Biology.

ENZ, C.A., MOLLER, R., MBWENEMO BIA, M. & HEBB, J. A population dynamics 1nodel for evaluating mortality factors in Lake Hallwil whitefish (Coregonus suidteri) larvae. Submitted to Arch. Hydrobiol. Spec. Issues Advanc. Lirnnol.

ENZ, C. A., BORG!, H. R., MBWENEMO BIA, M. & MOLLER, R. Factors controlling growth and maturity of whitefish (Coregonus suidteri) in eutrophic Lake Hallwil (Switzerland). Submitted to Journal of Fish Biology.

CONTENTS

Publications v Contents vii Summary xi

Zusammenfassung xiv

1 Introduction 1 ,.., 1.1 Description of Lake Hallwil ;_, 1.2 Population dynamics of Lake Hall wil whitefish 7 1.3 Objectives 15 1.4 References 17

2 Fish species diversity 35 2.1 Abstract 35 2.2 Introduction 35 2.3 Methods 38 2.4 Results 39 2.5 Discussion 42 2.6 Acknowledgements 49 2.7 References 49 3 Growth and survival of whitefish larvae reared on dry and live food. 53 3.1 Abstract 53 3.2 Introduction 54 3.3 Materials & Methods 55 3.4 Results 59 3.5 Discussion 62

Vll 3.6 Acknowledgements 69 3.7 References 70

4 Importance of prey nioven1ent, food particle size and tank circulation 79 4.1 Abstract 79

4 •"-'') Introduction 80 4.3 Methods 81 4.4 Results 83 4.5 Discussion 87 4.6 Acknowledge1nents 89 4.7 References 90

5 Food preference of adult whitefish, and the question of cannibalism 95 5.1 Abstract 95 5.2 Introduction 96 5.3 Methods 97 5.4 Results 99 5.5 Discussion 105 5.6 Acknowledgements 109 5.7 References 109

6 A model for evaluating mortality factors in whitefish larvae 117 6.1 Abstract 117 6.2 Introduction 117 6.3 Methods 119 6.4 Results 126 6.5 Discussion 131 6.6 Acknowledgements 135 6.7 References 135

Vlll 7 Factors controlling growth and maturity 141 7.1 Abstract 141 7.2 Introduction 141 7.3 Methods 144 7.4 Results 146 7.5 Discussion 150 7.6 Acknowledgements 156 7.7 References 157 8 Synopsis 163 8.l Introduction 163 8.2 Population dynan1ics of L. Hallwil whitefish 163 8.3 Implications for artificial reproduction 165 8.4 Implications for stocking practice 166 8.5 Implications for the fisheries 167 8.6 Effects of the restoration measures 168 8.7 References 169 Dank 173 Curriculum vitae 177

lX x SUMMARY

This work basically describes the fish fauna and the dynamics of a whitefish population in a typical eutrophic lake. Because whitefish are cmnmercially important, and because their natural reproduction in eutrophic lakes is ineffective, the life history, management and conservation of whitefish are investigated, taking the example of eutrophic Lake Hallwil in the Swiss Midland.

The structure of today's fish assemblages in Swiss lakes is the result, first of all, of colonization after the last Ice Age some 15,000 years ago. After the retreat of the glaciers, fish immigrated fron1 the previously ice-free regions into the newly- formed lakes. Lake Hallwil is such a post-glacial lake. Today, it is inhabited by twenty fish species. In 196L a rapid decrease in oxygen concentration due to cultural eutrophication led to a devastating fish kill. In the following years, most of the fish populations successfully recovered on their own (chapter 2). However, in cu trophic lakes. sahnonid fish like whitefish are greatly endangered due to unsuccessful natural propagation and the threat of fish kills. In Lake Hallwil, the so called Lake Hallwil "Ballen" (Coregonus suidteri, Fatio), a :fast-growing autoch- thonous coregonid, was commercially important already during the Middle Ages. In 1898, the first algal bloom heralded the beginning of eutrophication. Stocking whitefish larvae to sustain the decreasing whitefish population was conducted already at the beginning of the 20th Century. Nevertheless, whitefish yield contin- ued to decrease due to the negative effects of progressing eutrophication. Follow- ing the fish kill of 1961, Lake Hallwil whitefish probably disappeared from the lake. Thanks to stocking Lake Sen1pach "Ballen", a conspecific whitefish form in a near-by lake, a whitefish population was successfully restored in Lake HallwiL

Since 1970, an increasing munber of whitefish larvae have been pre-fed for at least six weeks in hatcheries prior to stocking in the lake. These pre-fed larvae are expected to survive ten times better in the lake than newly hatched larvae. Larvae are usually reared on live zooplankton or on dry diets. In order to further improve the rearing of whitefish larvae, smne feeding experhnents were carried out (chapters 3 and 4 ). Thereby it was found that live zooplankton is still the best diet to rear

xi whitefish larvae. However, low zooplankton density in the lake in spring often leads to difficulties in supplying sufficient numbers of zooplankton in the hatcheries. Nauplii of Artemia salina, various dry diets (chapter 3) or frozen zooplankton (chapter 4) may serve as alternative food. The results of the experiments showed that larvae reared on dry diet need an adaptation phase to switch from dry food to live zooplankton (chapter 3). As an important detail when setting up rearing tanks, the water inflow should be placed below the water surface of the tank in order to increase floating time and thus availability of the dry diet (chapter 4 ).

Thanks to stocking pre-fed larvae, whitefish has been the dmninant fish in Lake Hallwil since 1977. Nevertheless, annual yield fluctuates strongly. This is a result of variation in year class strength caused by a number of mortality factors acting primarily during the early life stage of whitefish. Among the inortality factors investigated, intraspecific food cmnpetition due to high stock size and cannibalism were found to be unimportant. Larvae and adults showed different food preferences, leading to niche segregation during the most critical time in spring (chapter 5). However, strong evidence for early life mortality due to eutrophication- related processes was found using computer modelling: Sunny weather in May, together with high nutrient concentration, leads to strongly increased algal bimnass and photosynthesis. This may result in extreme oxygen supersaturation which causes lethal gas bubble syndrome in the whitefish larvae (chapter 6). The large whitefish stock in Lake Hallwil is certainly of great con11nercial interest. It also has to be seen as the positive result of fish species conservation. Growth retardation, however, occurred in parallel with increasing whitefish density (chapter 7).

Trophic state in tenns of total phosphorus concentration of Lake Hallwil decreased since 1977. External and internal restoration measures (chapter 1), in particular sewage diversion and artificial oxygenation since 1986, have had a positive effect on the lake ecosystem. Hypolinmetic oxygenation has enlarged the living space of fish and benthic organisms (chapters 5 and 8). Within the context of these water protection ineasures, this thesis provides new knowledge on the biology and ecology of Lake Hallwil whitefish in a eutrophic lake undergoing restoration. The results presented herein may set the basis for better conservation and rnanagen1ent through improved rearing, stocking and protecting this valuable fish (chapter 8). The new findings may also help to assess, on a more general basis, the

Xll usefulness of external and internal lake restoration measures for conservation and management of lake whitefish.

Xlll ZUSAMMENFASSUNG

Die vorliegende Arbeit befasst sich im wesentlichen nut den Then1enbereich der Fischfauna und der Populationsdynamik von Coregonen in einc1n cutrophen See. Weil Felchen fischereiwirtschaftlich von Bedeutung sind und weil die natiirliche Fortpflanzung dieser Fische in eutrophen Seen nicht mehT funktioniert, standen in dieser Arbeit Lebenszyklus, Fischerei1nanagement und Arterhaltung der Felchen im V ordergrund. Das Untersuchungsgewasser, der Hallwilersee, ist ein Beispiel ftir einen eutrophen Coregonensee im Schweizer Mittelland.

Die heutige Zusammensetzung der Fischgesellschaften in Schweizer Seen ist primar das Ergebnis der Neubesiedlung nacheiszeitlicher Gewasser vor ungefahr 15'000 Jahren. Nach dem Rilckzug der Gletscher wanderten Fische aus eisfreien Gebieten in die neuentstandenen Seen ein. Der Hall wilersee ist ein solcher nacheiszeitlicher See. Heute beherbergt er zwanzig verschiedene Fischarten. 1961 flihrte ein Sauerstoffzusammenbruch als Folge der kulturell bedingten Eutrophie- nmg zu eine1n verheerenden Fischsterben. In den darauffolgenden Jaluen erholten sich die meisten Fischpopulationen von selbst (Kapitel 2). Die lachsartigen Fische jedoch, zu denen auch die Felchen gehoren, sind in eutrophen Seen wegen Fort- pflanzungsmisserfolg und moglichen Fischsterben bedroht. Der sogenannte Hall- wiler «Ballem> (Co reg onus suidteri, Patio), eine schnellwtichsige autochthone Felchenform, war bereits im Mittelalter von grosser wirtschaftlicher Bedeutung. 1898 Hiutete cine erste AlgenblUte die beginnende Eutrophierung ein. Bereits zu Beginn des 20sten J ahrhunderts versuchte man 1nit Besatzmassnahmen, dem RUckgang der Felchen entgegenzuwirken. Trotzdem sanken die Felchenfanger- trage kontinuierlich, dies aufgrund der negativen Auswirkungen der zunehmen- den Eutrophierung. Nach elem Fischsterben von 1961 sind die Felchen vennutlich aus dem See verschwunden. Dank Besatz mit Brtitlingen des Sempacher «Bal- len», einer moglicherweise identischen Felchenfonn ans einen1 benachbarten See, konnte der Felchenbestand im Hallwilersee neu aufgebaut werden.

Seit 1970 wurden in den Brutanstalten zunelunend auch Felchenbri.Hlinge wahrend mindestens sechs Wochen angefi.tttert. Man schatzt, class diese sogenannten

xiv Vorsornmerlinge im See eine zehmnal hohere Oberlebensrate als frischge- schliipfte Briitlinge haben. Die Felchenlarven werden rnit lebendem Zooplankton oder mit Trockenfutter geflittert. U1n die Aufzuchttechnik zu verfeinern, wurden Filtterungsexperirnente durchgefiihrt (Kapitel 3 und 4). Die Ergebnisse zeigten, dass lebendes Zooplankton nach wie vor das beste Aufzuchtfutter fitr Felchenlar- ven darstellt. A1lerdings fithren niedrige Zooplanktondichten im See im FrUl~jahr im1ner wieder zu Nahrungsengpiissen in den Brutanstalten. Als alternative Futter- mittel ki:)nnen Nauplien des Salinenkrebses (Artem.ia salina), verschiedene Trok- kenfuttennittel (Kapitel 3) oder gefrorenes Zooplankton (Kapitel 4) verwendet wcrden. Die V crsuchsergebnisse weisen zudem darauf hin, dass 1nit Trockcnfutter geftitterte Larven bei der U mstellung auf lebendes Zooplankton einc Anpassungs- phase durchmachen (Kapitel 3). Von zentraler Bedeutung bei dcr Felchenaufzucht rnit Trockenfutter ist die Positionierung des Wassereinlaufs im Aufzuchtbecken: Der Einlauf tnuss unter der Wasseroberflache milnden, um die Verweilzeit des Trockenfutters auf der Wasseroberflache zu verlangern und dadurch die Verfitg- barkeit der Nahrung zu verbessern (Kapitel 4).

Dank dem Besatz mit VorsOimnerlingen sind die Felchen seit 1977 die dominie- rende Fischart des Hallwilersees. Trotzdem schwanken die Jahresfangertdige enorm. Dies ist eine Folge der von Jahr zu Jahr schwankenden Jahrgangsstarken, verursacht durch eine ganze Reihe von Mortalitatsfaktoren, die hauptsachlich im frilhen Lebensstadimn der Felchen wirken. Diese Faktoren konnen unter Urnstan- den ganze Felchenjahrgange ausWschen. Zwei der untersuchten potentiellen Mor- talitatsfaktoren, intraspezifische Nahrungskonkurrenz wegen hoher Populations- dichte und Kannibalis1nus, zeigten sich als vernachlassigbar. FelchenbriHlinge und adulte Felchen haben wahrend der kritischen Friihjahrsperiode unterschiedliche Nahrungspraferenzen. Dies ftihrt zu einer ausgepragten Nischenseparation (Kapitel 5). Im Computermodell zeigte sich, dass eutrophierungsbedingte Pro- zesse im wesentlichen fitr die zeitweise sehr hohen Mortalitiitsraten wahrend der Larvenentwicklung verantwortlich sind. Sonniges Wetter im Mai. zusamrnen rrrit hoher Nahrstoffkonzentration, filhrt zu stark erhohter Algenbiomasse und Photo- syntheseakti vitat. Dadurch kann sich eine extreme Sauerstoffilbersattigung auf- bauen, die das tOclliche Gasblasensyndrom bei Felchenlarven bewirkt (Kapitel 6). Der reiche Felchenbestand des Hallwilersees ist fischereiwirtschaftlich von grosser Bedeutung. Dies kann als positives Ergebnis der Bestandeserhaltung durch die Besatzmassnahmen angesehen werden. Fischereiwirtschaftlich negativ zu werten

xv ist hingegen der beobachtete Wachstmnsrlickgang wegen der hohen Felchenbe- standsdichte (Kapitel 7).

Seit 1977 hat sich der trophische Zustand des Hallwilersees signifikant verbessert. Exteme und inteme Sanierungs1nassnahn1en, spezieU die Fernhaltung der Abwas- ser, aber auch die ktinstliche Seebelllftung, haben sich positiv auf das Okosystem des Sees ausgewirkt (Kapitel 1). Die kunstliche Sauerstoffanreichenmg im Hypo- limnion hat den Lebensraum fur Fische und Bodentiere wesentlich vergrossert (Kapitel 5 und 8). Die Ergebnisse dieser Dissertation sind iin Rahmen dieser See- sanierungs1nassnah1nen zu sehen, da sie neue Erkenntnisse zur Biologie und Oko- logie der Hallwiler «Ballen>> in eine1n ki.instlich beli.tfteten, eutrophen See vermit- teln. Weiter liefert diese Arbeit Gnmdlagen for die Arterhaltung und die fischereiliche Bewirtschaftung der einzigartigen Hallwiler «Ballen» in Form von Empfehlungen fur Aufzucht, Besatz und Fischerei (Kapitel 8). Die neuen Erkenntnisse konnen schliessbch auch in eine1n etwas allgemeineren Rah1nen dazu dienen, den Erfolg von externen und internen Sanienmgsmassnahmen im Hinblick auf den Erhalt und die Bewirtschaftung von Felchenbestanden in eutrophen Seen zu beurteilen.

XVI 1 INTRODUCTION

Eutrophication, i.e. the gradual increase in nutrient concentration and primary production, is a natural ageing process oflakes (LAMPERT & SOMMER, 1993). Hyper- eutrophication due to cultural development, however, constitutes one of the n1ost common and severe anthropogenic disturbances of aquatic ecosyste1ns. At the early stages of eutrophication, fish stock size increases due to an increasing food base. With progressing eutrophication, however, devastating fish kills due to oxygen depletion may occur, and natural reproduction of salmonid fish like whitefish (Coregonus sp.) is no longer possible due to the anoxic sediment (Ml)LLER, 1992). Without artificial propagation, whitefish disappear frmn eutrophic lakes (BRUTSCHY & GONTERT, 1923; COLBY et al, 1972; HARTMANN, 1977 & 1979). In view of the commercial ilnportance of whitefish, artificial breeding and stocking is being conducted not only to sustain populations without natural reproduction (STEFFENS, 1995) or to restore locally extinct populations (LUCZYNSKI et al., 1998). It also aims to increase year class strength (YCS) and yield (LESKELA et al., 1995; TURKOWSKI & BONAR, 1995). In spite of great efforts undertaken to enhance whitefish stocks, intensive stocking does not guarantee strong year classes and high sustained yield. Knowledge of the factors affecting whitefish survival from larvae to adult fish is essential for sustainable and effective management of whitefish populations. YCS of fish, and thus stock size is determined during the early life stage (MAY, 1974; SALOJARVI, 199la/b). During this early period, several factors are known to influence YCS: rearing conditions, size at the time of stocking (HOAGMAN, 1974; FLUCHTER, 1980; TAYLOR & FREEBERG, 1984), the nun1ber of fish stocked (CHRISTIE, 1963, MlTLLER, 1990) and environmental conditions (PONTON & MlTLLER, 1989; ECKMANN et al., 1988).

It was the overall objective of the research reported in this thesis to study whitefish population dynamics in eutrophic lakes taking the example of Lake Hallwil. The specific objectives of the research are presented in subchapter 1.3, following the presentation of the study site and the population dynamics of whitefish (subchap- ters 1.1 and 1.2).

1 1 Introduction

1.1 Description of Lake Hallwil

1.1.1 Main characteristics

Lake Hallwil is a post-glacial lake in the Swiss Midland (MARKI & SCHMID, 1983; SCHEIDEGGER et al., 1994). The original basin had included the Lakes Hallwil and Baldcgg, but today the situation is as shown in Figure 1.1 (LIECHTI, 1994). The main characteristics of L. Hallwil are presented in Table 1.1.

Lake Hallwil

100 km

Ballwil 0 Figure 1.1 The study site: L. Hallwil and its catchment area including L.Baldegg.

2 1.1 Description of Lake Hallwil

The shores of L. Hallwil border mainly agricultural land and villages (AKERET, 1993). Most parts of the shores are still hennned by reeds followed by a narrow littoral zone. The shores are quite steep; the littoral descends abruptly to 9 meters depth (GeNTERT, 1921). Fr01n 9 to 12 meters depth, the gradient is flatter, followed by a another steep slope (GONTERT, 1921). STOCKLI & SCHMID (1987) defined the border between epilin111ion and metalimnion at 8 m depth and the border between rnetalimnion and hypolirnnion at 13 m depth. According to SCHEIDEGGER (1992), the epilirnnion comprises 26% of the total lake volume, the inetalimnion 15% and the hypolirnnion 59%.

Table 1.1 L. Hallwil according to LIECHTI (1994).

Lake Hallwil

altitude 449 n1 above sea level

'1 surface area 10 km"' length 8.4 kln width 1.6km

maximal depth 46.5 In 1nean depth 28.4 in volume 0.28 kin3 inean discharge 2.35 m3/s theoretical mean residence time 3.8 years

'1 catclunent area (incl. L. Baldegg) 128 kin"'

1.1.2 History of eutrophication

Lake Hallwil is a naturally oligonlictic lake. Tllis fact is due to the hills along the east and west side of the lake, reaching to 878 meters above sea level (AKERET, 1993) and protecting the lake fron1 circulation-inducing winds (SCHEIDEGGER et al., 1994; RCRGI, 1994). The sheltered location, the considerable depth and the

3 1 Introduction

long residence ti1ne of the water (Table 1.1) are the main reasons why oxygen conditions of L. Hallwil are critical (SCHEIDEGGER et al., 1994) .

...... '§i 300~----·------··------··---·-·--·-·---·------~~~~~~----, 3 § 250 :.;::; CTI I- C 200 8 c 150 0 L) (/) :J 100 '- 0 .c 0.. 50 U) 0 .c 0.. 0 CJ) .,.-- ("') L0 I'- (J') .,.-- (I) L0 I'- m r- ("') L() I'- (J) C') U) I'- (J) cu L0

Figure 1.2 Phosphorus concentration at spring overturn in L. Hallwil between 1960 and 1999 (source of data: Kanton Aargau).

L. Hallwil has been fully eutrophic since about 1960 (MOLLER et al., 1994; Figure 1.2). In 1898, the first bloorn of the cyanobacterimn Planktothrix rubescens, an organis1n characteristic for eutrophic lakes (BORGI, 1994), heralded the beginning of eutrophication (BRUTSCHY & GONTERT, 1923). Waste water fro1n settlements and s1nall industries was identified as the 1nain cause (KELLER, 1945; AMB(llIL, 1960). As a first restoration rneasure in 1961, waste waters from households and industries were kept away frorn L. Hallwil by means of a sewer diversion systen1 (MARKI & SCHMID, 1983; AKERET, 1993; BURG! & JOLIDON, 1998). Since 1964, waste waters have been purified in a sewage treat1nent plant below the lake outlet (MARKI & SCHMID, 1983; AKERET, 1993; BORG! & JOLIDON, 1998). Nevertheless, fertilizers from agriculture still contributed to the eutrophication process (BURCH, 1994). By 1977, total phosphorus had reached concentrations exceeding 250 n1icrogrmns per liter (MiJLLER et al., 1994 ), while the hypohmnion and large parts of the sediment had become anaerobic, at least in sununer (AMBUHl,, 1960; BERNER, 1980). Several other measures were taken to lower the phosphorus input and to iinprove living conditions for fish. New regulations on the use of fertilizer in agriculture (Bl.JRGI, 1994)j a ban on phosphorus

4 1.1 Description of Lake Hallwil

in laundry detergents (AKERET, 1993) and an aeration system to oxygenate the anaerobic hypoli1nnion were introduced in 1986 (STOCKLI & SCHMID, 1987; AKERET, 1993; BtrRGI, 1994). In smnmer, the oxygenation systen1 injects pure oxygen into the hypolin1nion, while in winter compressed air is used to enhance circulation (STOCKLI & SCHMID, 1987; AKERET, 1993; BURGI, 1994; WErmu & WUEST, 1996). Thus, living space for fish (AKU et al., 1997) and for benthic organisms has been enlarged (STOSSEL, 1992). However, the anoxic sediment has remained a great problem for fish whose eggs develop on the sediment, such as coregonids (MfTLLER, 1992; VENTLING-SCHWANK, 1992; 1v10LLER, 1993; Mt'ILLER et al., 1994).

1.1.3 l.ake Hallwil whitefish

According to the lake classification of KLEE (1991), L. Hallwil is a coregonid lake. FATIO (1890) hypothesized that fish have colonized the lake using the large post-glacial rivers as migration pathways. After the retreat of the glaciers, the discharge of those rivers decreased and species like whitefish, which were not adapted to traverse water cascades and rapids, were isolated (FATIO, 1890). Thus, the single whitefish form called L. Hallwil "Ballen" could have evolved. Several village names (e.g. Ballwil) that obviously had been derivations of local fish nan1es reflect the richness of fish and the importance of the fisheries since the Middle Ages. Taxonmny of L. Hallwil "Ballen11 has been the subject of several investigations elating back to the 18th century (LlNNE, 1758; FATIO, 1885; BERG, 1932; WAGLER, 1941 & 1950; STEINMANN. 1950a,b & 1951; DOTTRENS, 1959; HTMBERG & LEHTONEN, 1995). Finally, KcrrTELAT (1997) revised coregonid nomenclature and named L. Hallwil whitefish Coregonus suidteri (Fatio).

As a result of progressing eutrophication, total whitefish yield decreased dramatically fro1n 3,600 kg in 1914 to 900 kg in 1919, despite stocking of newly hatched whitefish lmvae (BRUTSCHY & GUNTERT. 1923). \Vater pollution was recognized quite early as the tnain cause of the whitefish population decline in L. Hallwil (BRUTSCHY & Gf1NTERT, 1923). Natural reproduction of coregonids has becmne unsuccessful due to the anoxic sediment (MOLLER, 1992 & 1993; VENTLING- SCHWANK & MOLLER, 1991; VENTLING-SCHWANK, 1992; NlOLLER et aL, 1994; VENTU~G-SCHWANK & LIVINGSTONE, 1994). The development climaxed in autunm 1961 when low oxygen concentration led to a destn1ctive fish kill, causing

s 1 Introduction a severe reduction or even the elimination of the most sensitive fish species, especially whitefish (HEIMANN, 1962). Thanks to stocking of larvae, mainly fr01n L. Sempach which holds a conspecific species (KOTTELAT, 1997), a whitefish population was restored in L. Hallwil in the following years. Beginning in 1970, prc- fed whitefish larvae reared in round tanks for six weeks were stocked annually in addition to the newly hatched larvae (MtTLLER, 1990 & 1993). Because of the intensive rearing practice and stocking of pre-fed larvae, whitefish has been the dominant fish species of L. Hallwil since 1977 (MOLLER, 1990 & 1993). Although phosphorus concentration has significantly decreased due to the restoration measures (ZIMMERMANN et al., 1991), it is still at 40-50 micrograms per liter today (A. STOCKLI, unpubl. data). Thus, natural reproduction of L. Hallwil "Ballen" is still unsuccessful (R. MtTLLER, unpubl. data).

Despite intensive stocking of newly hatched or pre-fed larvae, yield from the sport and the commercial gillnet-fishery showed high fluctuations (Figure 1.3, black bars). By assuming a constant mortality rate of whitefish larvae over the years, a yield decrease in 1992-94 was not expected to occur (Figure 1.3, gray bars).

-100000 s 80000 -a -- ·-··· modeled yield f------· · · --- - Qi ·;;, 60000 ..c -i---~---~-~-5-_:~~-:~_Xi_~-~._J------(/) ;;::: 40000 (}) :!::: ..c 20000 ~ 0 ~ m ..... M m ~ m ..... M m ~ m ~ M m ~ m w w ~ ~ ~ ~ ~ ro ro ro ro ro m m m m m .....m m .....m m m m m m m m m m m m m m m year

Figure 1.3 Observed (black) and modeled (gray) whitefish yield in L. Hallwil (modelling with constant mortality rate, see chapter 1.1.3).

Thus, son1e weak or nonexistent whitefish year classes were the cause of these yield fluctuations (MlJLLER, 1993; Muller et al., 1994). No yield fluctuation would have occurred if larval survival was equal in every year (Figure 1.3). The principal aim of this thesis research was therefore to investigate whitefish population dynamics and to focus on larval mortality factors that could have affected

6 1.2 Population dynamics of Lake Hallwil whitefish whitefish YCS which, in tun1, would be responsible for stock size and yield fluctuations.

1.2 Population dynamics of Lake Hallwil whitefish

The fundmnental processes governing the dymunics of natural populations of fish are reprodrn.:t1on, mortality and growth (SHUTER, 1990): Reproduction is an addition of new individuals to the population, whereas mortality is the loss of existing population members. Growth is the process that renders a new individual capable of reproduction. RICKER (1977) found for several salmonid fish populations a direct connection between the nmnber of spawning fish and the recruits they produce. In this case, some classic mathematical approaches to stock-recruit1nent relationship were available to explain yield fluctuations (RICKER, 1954 & 1975; BEVERTON & HOLT, 1957: CUSHING, 1973; PAULIK, 1973; SAILA & LORDA, 1980; SHEPHERD, 1982). But in populations without successful natural reproduction as is the case for L. Hallwil whitefish, recruitment entirely depends on the human skill to breed and rear whitefish 1.arvae in the hatcheries. Nevertheless, optimal breeding and rearing conditions do not at all guarantee that the stocked larvae will yield a strong year class. According to MAY (1974) and SALOJARVI (199la/b), year-class strength (YCS) in nature is determined by mortality during the early life stage. GOODYEAR (1980) segregated possible inortality factors deter- nrining YCS of fish into fish density-independent (e.g. weather, food quantity and quality) and fish density-dependent mortality factors (e.g. predation, cannibalism, inter- :md intraspecific competition). According to SALOJARVI (1987), populations fluctuate also due to fishing. Fisheries management has therefore to be adjusted to the population dynamics of the exploited fish species.

In Switzerland, all coregonid populations are intensively 111anaged due to their high conunercial importance. However, proper management of coregonids has always been difficult because whitefish tend to disappear frmn eutrophic lakes (BRUTSCHY & GGNTERT, 1923; COLBY et al, 1972; HARTMANN, 1977 & 1979). In order to allow for sustainable development of whitefish populations, two management principles are important as stated by NIOLLER (l 990) and MtTLLER et al. (1994 ): First, it is crucial that whitefish are able to reproduce at least once before they are harvested. Second, populations can be enhanced by adjusted stocking

7 1 Introduction practices in order to increase stock size and thus yield. In eutrophic L. Hallwil, artificial breeding, rearing and stocking of larvae had to cmnpletely replace natural reproduction (Mt'rLLER, 1992 &1993; MULLER et al., 1994).

1.2.1 Artificial propagation

While artificial breeding, rearing and stocking are conducted to increase YCS (LESKELA et al., 1995) and yield (TURKOWSKI & BONAR, 1995), these methods are also used to sustain populations without natural reproduction (STEFFENS, 1995) or even to restore locally-extinct populations (LUCZYNSKI et al., 1998). Because stocked larvae originate frmn only few parental fish and are reared under non-natural conditions, changes in the gene pool (BRZUZAN et al., 1998; MAMON- TOV & Y AKHNENKO, 1998) and the phenotype, e.g. gillraker number (TODD, 1998), may occur. Additionally, assemblages of different coregonid forms exist in son1e lakes (RtJFLI, 1978) and represent species flocks (DOUGLAS et al., 1999). This requires an adjusted stocking practice to inaintain specific genetic variability (LUCZYNSKI &RITTERBUSCH-NAUWERCK, 1995; DOUGLAS, 1999), to avoid cross breeding (JOKIKOKKO & HUHMARNIEMI, 1998) and to optimize stocking efficiency (MliLLER, 1990; ECKMANN et al., 1998, WANZENBOCK & JAGSCH, 1998).

In L. Hallwil hatcheries, eggs are incubated in so-called "Zug jars". The newly hatched larvae are reared in circular tanks for at least six weeks (pre-fed larvae) or are stocked directly into the lake (Figure 1.4 ).

...... 40~--~----~----~------~- ..r:: (() 6 (/) cu '+= 2: <.O cu _ !fill newly hatched larvae 0 30 4.5 .,.- -0 '.2 ..__. w .~ • pre-fed larvae (J) ..r:: 4- 20 u <.O -'--~------~ 3 cu a 2: ..r::ro .,.- cu >. ...__.IO 1.5 -0 (J) 4- I (() c~ 0 ..µhjllb: 0 '- 0... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ m m m m m m m mro m m m m m m m m .,--

year

Figure 1.4 Number of newly hatched and pre-fed whitefish larvae stocked in Lake Hallwil between 1967 and 1998.

8 1.2 Population dynamics of Lake Hallwil whitefish

The effectiveness of the different stocking practices is not equal. Stocking newly hatched larvae or pre-fed larvae results in different survival rates due to size-selective mortality (HOAGMAN, 1973 & 1974; rLUCHTER, 1980; PONTON & MlTLLER, 1989 & 1990; MtJLLER, 1990). In addition, the number of stocked whitefish larvae also influences YCS (MULLER 1990 & 1993). In Finnish and Polish lakes, SALOJARVI (1988) and WOLOS et al. (1995) demonstrated that there is an upper limit for stocking intensity beyond \Vhich the carrying capacity of the ecosystem was exceeded and the effectiveness decreased. lJnless stocked whitefish found a vacant niche (NILSSON, ] 985), poor results can be expected. SALOJARVI (1986) proposed that the nm11ber of stocked larvae should be adjusted on the basis of the relation between stocking number and observed yield. Stocking of newly hatched whitefish larvae was perfonned in L. Hallwil since the beginning of the 20th Cen- tury (BRUTSCHY & GONTERT, 1923). According to ROTH (1954), KLEIN (1987) and MtrLLER (l 990), less than 0.5% of the stocked nevv'ly hatched larvae survive up to the adult stage. Pre-fed larvae \Vere stocked since 1970 (MULLER, 1990 & 1993; Figure 1.4). The survival rate of pre-fed larvae in the lake is estimated to be about ten times higher than that for newly hatched larvae (MOLLER, 1990).

Several factors are known to influence survival and, thus, success of the corcgonid larvae stocked. Rearing conditions and size at the time of stocking are important when artificial reproduction is practiced (FUJCHTER, 1980; URBAN-JEZIERSKA & BIERNACKI, 1992). The three hatcheries of L. Hallwil use different rearing and stocking methods: In one hatchery, larvae are reared on dry food during the first few weeks, and on natural food when the zooplankton concentration in the lake has risen. In the second hatchery, larvae are fed live zooplankton only, and in the third one all fish are stocked as newly hatched larvae. Several authors tested different rearing conditions and diets (artificial feeds and live organisms) for their efficiency to rear larvae, with varying success (ROTH & GEIGER, 1968; MEDGYESY & WIESER, 1982; 0RTLEPP, 1984; ROSCH, 1986, 1987, 1988, 1989 & 1992; ECKMANN, 1985 & 1987; ROSCH & APPELBAUM, 1985; ECKMA~'N et al., 1986; ROSCH & DABROWSKI, 1986; ROSCH & Ec~~1ANN, 1986; KLEIFELD-KRIE- BITZ & ROSCH, 1987; NlESSLBECK 1987; KOSKELA, l 988; BURKHARDT-HOL.'vl et al., 1989; REY & ECKMANN, 1989; CHAMPIGNEULLE & ROJAS-BELTRAN, 1990; R()SCH & NlARMULLA, 1990: ROSCH & SEGNER, I. 990; SEGNER & ROSCH, 1990; KOHO et al., 1991; KOSKELA & ESKELINEN, 1992; STEINHART & ECKMANN, 1992; SCHMID, 1993). When 50-80% of the larvae survive the first 35-45 days

9 1 Introduction

after hatching, the efficiency of artificial diet is usually considered satisfactory (R()SCH, 1988). ECKMAI\TN (1985) showed that mortality after 26 days is only 10% when the larvae are reared on optimal food iten1s, e. g. live f\rternia nauplii. Due to their small size, nauplii of this salt-water crustacean are optinml food for whitefish larvae (ROSCH, 1989; ECKMANN, 1985). In particular, its fatty acid composition is optimal for inetamorphosis (WATANABE et al., 1983; FLCCHTER & RE~1- BOLD, 1986). Nevertheless, the ability of the larvae to switch frmn dry diet or Artemia nauplii to lake zooplankton is also of importance - an aspect often neglected in stocking artificially reared whitefish lmvae.

1.2.2 Feeding

While the larvae pass through their ontogenetic development, their appearance (LUCZYNSKI et al., 1988) and feeding preferences change. They first select rotifers and copepod nauplii, then gradually switch to adult copepods and cladocerans (HOAGMAN, 1973; GUlSSANI & DE BERNARDJ, 1977; ENDERLEIN, 1981; HAMRIN, 1983; PONTON & MtJLLER, 1989 & 1990; HARTMANN & KLEIN, 1993). As larvae grow, foraging abilities improve due to increasing 1nouth size (HARTMANN & KLEIN, 1993), better developed digestive systc111 (LOEWE & ECKMANN, 1988), better visual and swimming capacity (BLAXTER, 1986; BROWN & TAYLOR, 1992; HARTMANN & KLEIN, 1993; MOOKERJI & RAO, 1993, 1994 &1995).

In late winter and spring, when the larvae hatch and start feeding, different food preference and spatial niche segregation of juveniles/adults and larvae arc known to ease or even to avoid intraspecific food competition in some whitefish populations. In winter and spring, when zooplankton concentration is low, juveniles and adults of some whitefish populations switch from zooplankton to benthic organisms such as annelids, insect larvae and insect pupae (NAGY, 1990; P0?\1EROY, 1991; HARTMANN & PROBST, 1995; MICHEL, 1996; MoOKEIUI et al., 1998). In eutrophic lakes, the density of benthic organis1ns is often low due to anoxic conditions in the hypolimnion and on the sediment (STOSSEL 1992). After the start of artificial oxygenation of L. Hallwil in 1986 (STOCKLI & SCHMID, 1987; SCHEIDEGGER et al., 1994) the density of benthic organis1ns increased considerably (STOSSEL, 1992).

Juvenile and adult whitefish are known to exploit both pelagic and benthic food resources. They are opportunistic feeders. Thus. stomach contents largely reflect

10 l .2 Population dynamics of Lake Hallwil whitefish

seasonal changes in prey availability (MOOKERJI et al., 1998). The period of intensive feeding and growth begins in May or June (LEHTONEN, 1981). During this time, whitefish are found in the upper 20 in of the lake (SKURDAL et al., 1985b) where they prey primaiily on cladocerans. Large and non-evasive species, such as Daphnia sp. and Bythotrephes longimanus, are thereby highly selected (GERST- MEIER, 1985 & 1986; MOOKERJI et al.. 1998). Smaller phyllopods like Bosmina sp., and evasive organis1ns like copepods, are consumed mainly at times when the preferred prey are rare (MOOKERJI et al., 1998). Thus, in smne lakes, intraspecific food competition in so1ne whitefish forms is seasonally reduced to surnn1er and autumn. Nevertheless, other biotic and abiotic factors are known to influence larval mortality in spring.

1.2.3 Larval mortality factors

The effectiveness of stocking whitefish larvae is neither predictable (SALOJARVL 1992) nor convincing (STEFFENS, 1995) and depeuds on larval mortality in the lake (MAY, 1974: SALOJARVI, 199la/b). Environmental conditions (LEHTOl\'EN, 1985; GERDEAUX & DEWAELE, 1986; ECKMANN et al., 1988; PONTON & MCLLER, 1989; LESKELA et al., 1995) or fish density-dependent factors (CHRISTIE, 1963; HAMRIN, 1979; SALOJARVI, 1987; AUVINEN, 1988; Mt'JLLER, 1990; CARANHAC & GERDEAUX, 1998; LEHTONEN & NIEMELA, 1998; SALONEN et al., 1998; VILJANEN, 1986) are of priine i1nportance. Figure 1.5 shows hypothetical inortality factors of L. Hallwil whitefish larvae.

AASS (1972), SALOJARVI (1987), GOODYEAR (1980), VILJANEN (1986), VIL- JANEN (1988), JURVELIUS (1991) and SANDLUND et al. (1991) reported fish density-dependent factors to significantly contribute to population fluctuations: A strong year class develops only when fish density is low or medium. Because in L. Hallwil yield and thus fish stock increased enormously since 1967 (Figure 1.3), food competition between larvae, juveniles and adult whitefish (HA:t'vIRIN & PER- SSON, 1986) or even cannibalism (GRIM, 1951) may occur (Figure 1.5).

Cannibalism on eggs has been reported for whitefish (FABlUCIUS & IJNDROTH, 1954; SKURDAL et al., 1985a). Cannibalism on larvae was found for whitefish under laboratory conditions (GRIM, 1951 ), but has never been reported from nature. However, according to RUDSTAM & MAGNUSON (1985) adult whitefish usually stay in the hypolimnion, whereas the juveniles prefer the warmer cpilirn-

11 1 Introduction

nion (JURVELHJS & HEIKKINEN, 1987; PONTON & MENG, 1990). HAMRIN (1986) found that niche segregation between larvae and adults occurs 24 hours a day. He postulated that fish-size-dependent temperature preference, light, and intraspecific competition are responsible for the different behavior patterns, which change through ontogenetic development. SANDLUND et al. (1991) also found different food preference and niche segregation between larvae and adults from May until July. Those findings show that there are mechanis1ns in whitefish populations to ease intraspecific competition and to avoid cannibalism. Furthennore, not only adult whitefish stock size, but also the nmnber of larvae stocked in L. Hallwil greatly increased after 1967 (Figure 1.4 ). Intraspecific competition between whitefish larvae (Figure 1.5) may occur as a consequence of high larval density (AASS, 1972; SALOJARVI, 1987).

interspecific predation competition ~ intraspecific competition larvae-larvae

intraspecific competition adults-larvae/cannibalism ~

Figure 1.5 Hypothetical factors affecting year-class strength of L. Hallwil whitefish larvae. White: fish density-dependent inortality factors; gray: fish density-independent mortality factors.

The stock size of fish species other than whitefish is often high in eutrophic lakes. Therefore, interspecific competition (Figure 1.5) may also be i1nportant as

12 1.2 Population dynamics of Lake Hallwil whitefish described by BERG & GRIMALDI (1966). Several authors stated that predation by perch may affect whitefish year classes (SvARDSON, 1977; GERDEAUX & DEW- AELE, 1986; SALOJARVI, 1987; MOLLER et al., 1994). Whitefish fingerlings have occasionally been found in the stomach of perch frmn L. Sempach (R. MULLER, unpubl. data).

According to LEHTONEN (1983), HILDEN et al. (1984) and SALOJARVI (1987), populations fluctuate due to fishing (Figure 1.5). But fishing also eases the effect of competition by removing substantial numbers of adult fish, thus lowering fish density that otherwise could lead to a weak year class (AASS, 1972).

Weather (Figure 1.5), as an expression of climate, is a fish density-independent factor and exerts a strong influenc on ecosystems in different ways: First, meteorological conditions in spring directly influence water temperature, which according to LEHTONEN (1985) is the most important factor influencing YCS. Total gas supersaturation due to water warming (WEITKAMP & KATZ, 1980) and oxygen supersaturation due to excessive algal photosynthesis in eutrophic lakes during Jong sunshine periods (MATHIAS & BARICA, 1985) influence YCS indirectly via temperature: Oxygen supersaturation leads to lethal gas bubble syndrome (GBS) in whitefish larvae (BOUCK, 1980; STADELMANN, 1988; VENTLING-SCHWANK, 1992). Second, weather effects could also be caused by wind. Low wind in spring favours thermal stratification, which was found to be of prime importance for YCS (ECKMANN et al., 1988). Third, low spring temperature also affects zooplankton density (PONTON & MULLER, 1989; Figure 1.5). The findings of different authors on this topic are controversial. According to ECKMANN et al. (1988), zooplankton concentration during spring had no significant influence on YCS in eutrophic L. Constance. Low food density (TAYLOR & FREEBERG, 1984; NAESJE et al., 1986; RICE et al., 1987; PONTON & MULLER, 1989; DABROWSKI, 1991) or the lack of suitable food (AMMANN & STEINMANN, 1948; HOAGMANN, 1973; GUISSANI & DE BERNARDI, 1977; HAMRIN, 1983; VILJANEN, 1983; ECKMANN, 1985; ROSCH, 1988; PONTON & MULLER, 1989) may increase larval mortality. But not only quantity could have an effect. Zooplankton quality (Figure 1.5) was presumed to be the cause of up to 90% larval mortality in L. Constance due to smne unknown chemical components of zooplankton like copepods and rotifers (ECKMANN, 1985; ROSCH, 1994). It was thought that the zooplankton got poisoned by ingesting toxic phytoplankton. Phytoplankton could affect the food of

13 1 Introduction whitefish larvae also in another way: SCHULTZ (1992) found a near-elimination of some cladocerans like Daphnia sp. at high densities of the cyanobacterimn Plank- tothrix rubescens (Figure 1.5). In some lakes, daphnids are the most abundant organisms in the diet of whitefish fr0111 May until September (MAYR, 1998; MOOKERIJ et al., 1998). However, NAESJE et al. (1986) found a positive correlation between the density of cladocerans and YCS. A lack of daphnids would result in a delayed growth of whitefish larvae which would attain a "predation-secure" size later than under optimal feeding conditions (PONTON & MULLER, 1989 & 1990). Thus, optimal feeding and growth are also of importance for determining YCS.

1.2.4 Growth

Growth, i.e. the rate of increase in length and weight, is one of the ultimate indicators of health and condition in fish because it integrates all a biotic and biotic factors affecting the organism. Fast growth reflects good overall living conditions, while slow growth is usually associated with unfavorable conditions. Poor growth may also reflect chronic stress (LECREN, 1972; WATERS, 1977; LARKIN, 1978). Growth is a key parameter in studying population dynamics (BOWEN et al., 1991) and stress in fish (GOEDE & BARTON, 1990). Growth rate is controlled by a number of factors. Several authors found fish density to be the most important factor influencing growth: According to SALOJARVI (1992), SCHULTZ (1992) and MULLER et al. (1994 ), growth rate is usually low at high stock density. HENDER- SON & BROWN (1985) and VILJANEN (1986 & 1988) stated that strong year classes grow slower than weak ones. Other authors found that fish density-independent factors like temperature or trophic state may be responsible for different growth rates in whitefish. Growth increase with increasing temperature is reported by AASS (1972), SARVALA et al. (1988), PONTON & MULLER (1989) and GRIFFITH et al. (1992). MILLS & CHALANCHUK (1988), PONTON & MtJLLER (1989), KIRCH- HOFER (1995) and MtJLLER & BIA (1998) found that growth is faster in eutrophic than in oligotrophic lakes. Differences in zooplankton density were thereby identified as the primary cause (SVARVAR & MULLER, 1982; MULLER, 1990; KIRCH- HOFER, 1995).

In L. Hallwil, growth is fastest in the first two years (Figure 1.6). Whitefish reach maturity at the age of two or three years (MULLER et al., 1994). Because it is essen-

14 1.3 Objectives

tial that whitefish reproduce at least once before they are harvested (!vtOLLER et al., 1994), 1ninimtm1 mesh size is individually set for each lake whitefish population on the basis of local whitefish growth rate (M"CLLER ct aL 1994 ). Growth of a specific whitefish population is influenced by both environmental and endoge- nous factors (KIRCHHOFER & LINDT-KIRCHHOFER, 1998). Even within a population, growth varies from year to year: Growth of L. Hallwil whitefish declined between 1983 and 1988 (MULLER et al.. 1994: Figure 1.6). According to ALM (1959), COLBY (1984) and MUNKITTRICK & DIXON (1989) there is a tight connection between growth rate and age at maturity in fish: growth is faster and age at maturity lower at low stock size and vice versa.

500~--.-~~-~--~~--~-~~~~~~~~~~-~~~

E 400 -j...... , ...... , ...... , ...... , ...... , ...... ::'+m _tJar.::a.,,.....,"'· :.;;;'~·"' ...... 1...... ~ ...... ,, ...... , ...... ~ ...... ,, ...... , ...... ! ...... ! ...... ! ...... I __,E -.c 300 ~ ..... ,...... ; ..... ;.. fr ... :l·'""J'°'''"'0""'f'··'~""" .. , .... ,r., ... ~"~/--: .... ,,,.~ ... ,... /;~<. ..;.7~»·7'~0 °'ffi 200 -+ .. .,;.:.: .. !,:::;·.. f...... ,, ...... /; ...... T--:;:;:P°"' ..i""-' .... !f· .. '""';"·/"'·:·"""i ro -Q 10 0 -I·""'";""'"'"/·"'"+ , .. ; .... > --+··'-·+.. >...... >. ... ; .. o. ···/"'· ...f ..... +c.... +.c. ....1 .. • .... ,c ..· .... ,c.c .... / .. C...... c.... ,~ ..c .... / •• >...... c...... 1

year

Figure 1.6 Growth of L. Hallwil whitefish between 1980 and 1999.

1.3 Objectives

1.3.1 A brief overview

The overall aim of this study was to find the cause for the strong whitefish yield fluctuations in L. Hallwil (Figure 1.3). This thesis therefore focuses on L. Hallwil whitefish population dynmnics and particularly on the biotic and abiotic factors that could have affected whitefish YCS and, thus, yield. Furthermore, this work should provide a basis for analyzing the effects of restoration on L. Hallwil ecol-

15 1 Introduction

ogy, in order to 1nake recommendations for conservation and sustainable management of L. Hal1wil whitefish. The following questions have been addTessed:

• Arc there differences in stock size of fish before and after the disastrous fish kill of 1961? Are there differences in L. Halhvil whitefish phenotype before and after 1961? • How do the rearing conditions and the diet types fed in L. Hallwil hatcheries influence growth and survival of the larvae? Do the larvae easily switch from dry diet to live zooplankton? • What is the food composition and preference of adult L. Hallwil whitefish? Is there intraspecific food competition between adults and larvae? To which extent does cannibalism on larvae represent a mortality factor in L. Hallwil whitefish population? • Which abiotic and biotic factors influence survival of the whitefish larvae stocked in L. Hallwi1? • What were the causes for the growth decline in L Hallwil whitefish observed between 1983 and 1988?

1.3.2 Contents of the chapters 2~ 7

In chapter 2, a fish species list was established, and the actual L. Hallwil whitefish phenotype was described to compare the fish populations before and after the disastrous fish kill of 1961 .

The aim of chapters 3 and 4 was to investigate the effects of the rearing conditions and the diet types fed in L. Hallwil hatcheries on growth and survival of the larvae. We further looked at the ability of the larvae to switch frmn dry diet to live zooplankton, an aspect often neglected in stocking artificially reared whitefish fry.

Chapter 5 focuses on food composition and preference of adult L. Hallwil whitefish, with special en1phasis on the situation in late winter and spring, i.e. the ti1ne when whitefish larvae are stocked and are vulnerable to cannibalism and food competition.

In chapter 6, larval n10rtality factors that could have affected L Hallwil \vhitefish YCS and which could be responsible for the yield collapse in 1992-94 were investigated. In order to test a nmnber of environmental and population-related parameters for their effect on larval mortality, a population dynamics n10del was developed using the run-time software Stella.Researchrn 5.0 (HANNON & RUTH, 1997: COSTANZA et al., 1998).

16 1.4 References

The objective of chapter 7 was to identity the causes for growth decline in Lake Hallwil whitefish between 1983 and 1988. The influence of eutrophication, temperature, food density, stock size and YCS was investigated.

1.4 References

AASS, P. (1972): Age determination and year-class fluctuation of cisco, Coregonus albula L., in the Mjosa hydroelectric reservoir. - Report Institute of Freshwater Research (Drottningholm) 52: 5-22.

AKERET, B. (1993): Zur Biologie von Chaoborusflavicans, Leptodora kindtii und Bythotrephes longinwnus unter de1n Einfluss interner Restaurierungs- massnahmen in drei Schweizer Seen. - Dissertation an der Eidgenbssischen Technischen Hochschule Zitrich.

AKU, P. M. K .. RUDSTAM, L. G. & TONN. W. M. (1997): linpact of hypolimnetic oxygenation on the vertical distribution of cisco ( Coregonus artedi) in Amisk Lake, Alberta. - Can. J. Fish. Aquat. Sci.: 54/9: 2182-2195.

ALM, G. (1959): Connection between maturity, size and age in fish. - Report of the Institute of Freshwater Research, Drottingholm 40: 7-145.

AMBlJHL, H. (1960): Die Nahrstoffzufuhr zum Hallwilersee. - Schweiz. Z. Hydrol. 22: 563-597.

AMMANN, E. & STEINMANN, P. (1948): Die Verbesserung der Methoden in der Felchenzucht. Konu11ission flir die Erforschung fischereiwirtschaftlicher Fragen. Sonderbericht Nr. 1. W. KUNZ, Pfaffikon-Ziirich.

AUVINEN, H. (1988): Factors affecting year-class strength of vendace ( Coregonus albula) in Lake Pyhajarvi. - Finnish Fish. Res. 9: 235-243.

BERG, L. S. (1932): lJbersicht der Verbreitung der Siisswasserfische Europas. - Zoogeographica 1: 107-208.

BERG, A. & GRIMALDI, E. (1966): Ecological relationships between planktophagic fish species in the Lago Maggiore. V erhandlungen Internationale Vereinigung fitr theoretische und angewandte Limnologie 16/1-2: 1065-1073.

17 1 Introduction

BERNER, P. (1980): Lin1nologischc Untersuchungen nn Hallwilersee. - Diplomarbeit an der Universitat Bern.

BEVERTON, R. J. H. & HOLT, S. J. (1957): On the dynamics of exploited fish populations. U. K. Min. Agr. And Fish., Fish. Invest. Ser. II 19: 533pp.

BLAXTER, J. H. S. (1986): Development of sense organs and behaviour of teleost larvae with special reference to feeding and predator avoidance. - C. Hubbs (Ed.), Proceedings, Ninth Larval Fish Conference, Texas. - Trans. An1. Fish. Soc. 115: 98-114.

BOUCK, G. R. (1980): Etiology of gas bubble disease. Trans. -Atn. Fish. Soc. 109: 703-707.

BOWEN, S. H., D'ANGELO, D. ]., ARNOLD, S. FL, KENIRY, M. J. & ALBRECHT, R. J. (1991 ): Density-dependent maturation, growth, and female dominance in Lake Superior lake herring (Coregonus artedii). Can. J. Fish. Aquat. Sci. 48/4: 569-576.

BROWN, R. W. & TAYLOR, W.W. (1992): Effects of egg composition and prey density on the larval growth and survival of lake whitefish (Coregonus clupeafonnis Mitchill). - J. Fish Biol. 40: 381-394.

BRUTSCHY, A., GfJNTERT, A. (1923): Gutachten iiber den Ri.lckgang des Fischbestandes iin Hallwilersee. - Arch. Hydrobiol. 14: 523-571.

BRZCZAN, P., YAKHNENKO, V. M., MAMONTOV, A.M., MARKOWSKA, A. & TROFIMOV A, I. N. (1998): Mitochondrial DNA variation in whitefish Coregonus lavaretus from Lake Baikal as revealed by restriction analysis. Arch. Hydrobiol. Spec. Issues Advanc. LirnnoL 50: 357-362.

BfJRGI, H. R. (1994): Seenplankton und Seesanienrng in der Schweiz. Linu1ologische Berichte Donau 2: 71-95.

RORGI, H. R. & JOLIDON, C. (1998): 10 Jahre Seesanierung Hallwilersee. Die Rcaktion des Planktons. - Wasser, Energie, Luft 5/6: 109-116.

BURKHARDT-HOLM. P., ECKMANN. R. & STORCH, v. (1989): Schadigung des Darn1epithels von Coregonenlarven ( Corcgonus fera) durch Artemia- Ftitterung, Eine bakterielle Infektion. - Journal of Applied Ichthyology 1: 2- 11.

18 1.4 References

CARANHAC, F. & GERDEAUX, D. (1998): Analysis of the fluctuations in whitefish ( Coregonus lavaretus) abundance in Lake Geneva. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 197-206.

CHAMPIGNEULLE, A. & ROJAS-BELTRAN, R. (1990): First attempts to optimize the 1nass rearing of whitefish (Coregonus lavaretus L.) larvae from Leman and Bourget lakes (France) in tanks and cages. - Aquat. Living Resour. 3: 217- 228.

CHRISTIE, W. J. (1963): Effects of artificial propagation and the weather on recruitment in the Lake Ontario whitefish fishery. - J. Fish. Res. Board Can. 20: 597-646.

COLBY, P. J. (1984): Appraising the status of fisheries: rehabilitation techniques. In: V. W. Cairns, P. V. Hodson and J. 0. Nriagu (eds). Contaminant effects on fisheries. - Wiley, New York: 233-258.

COLBY, P. J., Spangler, G. R., Hurley, D. A. & McCombie, A. M. (1972): Effects of eutrophication on salmonid communities in oligotrophic lakes. - J. Fish. Res. Bd. Can. 29: 975-983.

COSTANZA, R., DUPLISEA, D. & KAUTSKY, U. (1998): Ecological Modeling and economic systems with STELLA. - Ecological Modeling 110/1: 1-4.

CUSHING, J.M. (1973): The dependence of recruitment on parent stock. - J. Fish. Res. Board Canada 30: 1965-1976.

DABROWSKI, R. (1991): Growth and feeding of Coregonuspeled Gmel. larvae in illuminated cages. - Pol. Arch. Hydrobiol. 38/3-4: 463-474.

DOTTRENS, E. (1959): Systematique des coregones de l'Europe occidentale, basee sur une etude biometrique. - Revue Suisse de Zoologie 66/1: 1-66.

DOUGLAS, M. (1999): Central alpine Coregonus (Teleostei, Coregonidae): Evolution and conservation of a unique asse111blage. Dissertation an der Philosophischen Fakultat der Universitat Zurich.

DOUGLAS, M. R., BRUNNER, P. C. & BERNATCHEZ, L. (1999): Do assemblages of Coregonus (Teleostei: Sal111011iformes) in the Central Alpine region of Europe represent species flocks? - Molec. Ecol. 8:589-603.

19 1 Introduction

ECKMANN, R. (1985): Histopathological alterations in the intestine of whitefish ( Coregonus sp.) larvae reared on zooplankton from Lake Constance (West Gennany). - Diseases of Aquatic Organisms 1/1: 11 18.

ECKMANN, R. (1987): Pathological changes in the midgut epithelju111 of grayling,Thymallus thymallus L., larvae reared on different kinds of food, and their relation to n1ortality and growth. Journal of Fish Diseases 10: 91 99.

ECKMANN, R., CZERKIES, P., HELMS, C. & KLEIBS, K. (1998): Evaluating the effectiveness of stocking vendacc (Coregmms albula L.) eleutheroe111bryos by alizarin marking of otoliths. - Arch. Hydrobiol. Spec. Issues Advanc. Li111nol. 50: 457-463.

ECKMANN, R., GAEDKE, U. & \VETZLAR, H. J. ( 1988): Effects of clitnatic and density-dependent factors on year-class strength of Coregonus lavaretus in Lake Constance. Can. J. Fish. Aquat. Sci. 45/5: 1088-1093.

ECKMANN, R., ROSCH, R., 0RTLEPP, J. & KLEIFELD, G. (1986): Survival and growth of corcgonid larvae from Lake Constance fed on zooplankton of different origin. Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 203- 214.

ENDERLEIN. 0. (1981): When, where, what and how much does the adult cisco, Coregonus albula (L.) eat in the Bothanian Bay during the ice-free season. Report Institute of Freshwater Research (Drottningholm) 59: 21-32.

FABRICll.!S, E. & LINDROTH, A. (1954): Experilnental observations on the spawning of whitefish, Coregonus lavaretus L., in the streain aquarimn of the Holle Laboratory at River Indalsalven. Rep. Inst. Freshw. Res. Drottningholm 35: 105-112.

FA TIO, V. (1885): Les coregoncs de la Suisse (ferns di verses) classification et conditions de frai. Receuil Zool. Suisse 1/2: 649-665.

PATIO, V. ( 1890): Faune des vertebres de la Suisse. Histoire naturelle des poissons. H. Georg (ed.) 5/2.

FLOCHTER, J. (1980): Review of the present kno\vlcdge of rearing whitefish (Coregonidae) larvae. - Aquaculture 19: 191-208.

20 1.4 References

FUJCHTER, J. & REMBOLD, H. (1986): Soluble factor essential for inetamorphosis of coregonid larvae has been partially purified frmnArtemia salina. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 197-202.

GERDEAUX, D. & DEWAELE, P. (1986): Effects of the weather and of artificial propagation on coregonid catches in Lake Geneva. - Arch. Hyclrobiol. Spec. issues Advanc. Limnol. 22: 343-352.

GERSTMEIER, R. (1985): Nahrungsokologische Untersuchungen an Fischen im Nationalpark Berchtesgaden. - Arch. Hydrobiol. Suppl. 72: 237-286.

GERSTMEIER, R. (1986): Investigations on the feeding biology of coregonid fishes (Coregonus species) from the Lake Starnberg (Bavaria, West Germany). - Spixiana 9/3: 225-232.

GIUSSANI, G. & DE BERNARDI, R. (1977): Food selectivity in Coregonus sp. [zooplanktophagous fish] of Lake Maggiore [Italy], an energetical approach. - Me1norie dell'Istituto Italiano di Idrobiologia, Dott. Marco de Marchi Verbania Pallanza (Italy) 34: 121 130.

GOEDE, R. W. & BARTON, B. A. (1990): Organismic indices and an autopsy-based assessn1ent as indicators of health and condition of fish. - Am. Fish. Soc. Sy1np. 8:93-108.

GOODYEAR, C. P. (1980): Con1pensation in Fish Populations. - Biological inonitoring of fish: special session at the 110th Annual meeting of the American Fisheries Society (Louisville) 11: 253-280.

GRIFFITHS, W. B., GALLAWAY, B. J., GAZEY, W. J. & DILLINGER, R. E. (1992): Growth and condition of arctic cisco and broad whitefish as indicators of causeway-induced effects in the Prudhoe Bay region, Alaska. Trans. Am. Fish. Soc. 12115: 557-577.

GRIM, J. (1951): Kannibalisrnus bei Blaufelchen und seine moglichen Folgen. - Osteneichs Fischerei 4: 165-171.

GtrNTERT, A. (1921 ): Der Hallwilersee. Ein verschwindender Glacialsee und seine Tiefenfauna. - Dissertation an der Universitat Basel.

21 1 Introduction

HAMRIN, S. F. (1979): Population dynamics, vertical distribution and food selection of ciscoe ( Coregonus albula) in south Swedish lakes. Dissertation at the University of Lund.

HAMRIN, S. F. (1983): The food preference of vendace (Coregonus albula) in South Swedish forest lakes including the effect on zooplankton populations. - Hydrobiologia 101: 121-128.

HAMRIN, S. F. (1986): Vertical dist1ibution and habitat partitioning between different size classes of vendace, Coregonus albula, in thermally stratified lakes. Can. J. Fish. Aquat. Sci. 43/8: 1617-1625.

HAMRIN S. F. & PERSSON L. (1986): Asynnnetrical competition between age classes as a factor causing population oscillations in an obligate planktivorous fish species. OIKOS 47/2: 223-232.

HANNON, B. & RUTH, M. (1997): Modeling dynamic biological systems. - Springer, New York, Berlin, Heidelberg.

HARTMANN, F. & KLEIN, M. (1993): Nahrungsselektion von Renkenbrut ( Coregonus lavaretus) unter Aufzuchtbedingungen. - Fischerei und Teichwirt 44/8: 279-283.

HARTMANN, J. (1977): Fischereiliche Verandenmgen in kulturbedingt eutrophierenden Seen. - Schweiz. Z. Hydrol.. 39/2: 243-254.

HARTMANN, J. (1979): Unterschiedliche Adaptionsfiihigkeit der Fische an Eutrophienmg. - Schweiz. Z. Hydrol. 41/2: 374-382.

HARTMANN, J. & PROBST, L. (1995): Divergent distributions of prey and predator, a fishing effect? - Aquatic Sciences 57/2: 106-118.

HEIMANN, R. (1962): Schonvorschriften for die Fischerei 1m Aargau. - Schweizerische Fischerei-Zeitung 70/9: 267.

HENDERSON, B. A. & BROWN, E. H. (1985): Effects of abundance and water te1nperature on recruitment and growth of alewife (Alosa pseudoharengus) near South Bay, Lake Huron, 1954-82. - Can. J. Fish. Aquat. Sci. 42: 1608- 1613.

22 1.4 References

HILDEN, M., LEHTONEN, H. & BOHLING, P. (1984): The decline of the Finnish vendace, Coregonus albula, catch and the dynamics of the fishery in the Bothnian Bay. -Aqua Fennica 14/1: 33-48.

HIMBERG, M. K. J., LEHTONEN, H. (1995): Systematics and nmnenclature of coregonid fishes, particularly in Northwest Europe. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 39-47.

HOAGMAN, W. J. (1973): The hatching, distribution, abundance, growth and food of the larval whitefish ( Coregonus clupeafonnis Mitchill) of Central Green Bay, Lake Michigan. - Report Institute of Freshwater Research (Drottningholm) 53: 1-20.

HOAGMAN, W. J. (1974): Feeding by alewives (Alosapseudoharengus) on larval lake whitefish (Coregonus clupeafonnis) in the laboratory. - J. Fish. Res. Board Can. 31/2: 229-230.

JOKIKOKKO, E. & HUHMARNIEMI, A. (1998): Stocking practices of anadromous whitefish, Coregonus lavaretus lavaretus, in Bothnian Bay, Finland: Evidence for gillraker numbers. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 507-515.

JURVELIUS, J. (1991): Distribution and density of pelagic fish stocks, especially vendace (Coregonus albula (L.)), monitored by hydroacoustics in shallow and deep southern boreal lakes. -Finn. Fish. Res. 12: 1-19.

JURVELIUS, J. & HEIKKINEN, T. (1987): The pelagic fish density, bimnass and growth of vendace, Coregonus albula L., 1nonitored by hydroacoustic methods and trawling in a Finnish lake. - Aqua Fennica 17/1: 27-34.

KELLER, R. (1945): Limnologische Untersuchungen im Hallwilersee. - Dissertation an der Eidgenossischen Technischen Hochschule Zurich.

KIRCHHOFER, A. (1995): Growth characteristics of coregonid populations in three lakes with different trophic states and decreasing nutrient concentrations. - Arch. Hydrobiol. Spec. Issues Advanc. Li1m101. 46: 61-70.

KIRCHHOFER, A. & LINDT-KIRCHHOFER, T. (1998): Growth and development during early life stages of Coregonus lavaretus from three lakes in Switzerland. - Arch. Hydrobiol. Spec. Issues Advanc. Li1m101. 50: 50-59.

23 1 Introduction

KLEE, 0. (1991): Angewandte Hydrobiologie. Trinkwasser, Abwasser, Gewasserschutz. 2nd edition. Georg Thieme Verlag, Stuttgart, New York.

KLEIFELD-KRIEBITZ, G. & ROSCH, R. (1987): A simple method of feeding coregonid larvae on frozen zooplankton. Journal of Applied Ichthyology 3: 119-124.

KLEIN, M. (1987): Die Bedeutung von Besatzrnassnahmen zur Stabilisierung und Steigerung von Renkenertragen iln Starnberger See. - Fischer & Teichwirt 38/12.

KOHO, J., KARJALAINEN, J. & VILJANEN, M. (1991): Effects of temperatures, food density and time of hatching on growth, survival and feeding of vendace (Coregonus albula L.) larvae. - Aqua Fennica 21/1: 63-73.

KOSKELA, J. (1988): Rearing of larvae of the whitefish Coregonus muksun Pallas with artificial diets. - Finn. Fish. Res. 9: 333-337.

KOSKELA, J. & ESKELINEN, U. (1992): Growth of larval european whitefish (Coregonus lavaretus) at different te1nperatures. - Pol. Arch. Hydrobiol. 39/3-4: 677-682.

KOTTELAT, M. (1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the former USSR), with an introduction for non-systematists and cmnments on nmnenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

LAMPERT, W. & SOMMER, U. (1993): Lirnnoc)kologie. -Thierne Verlag, Stuttgart.

LARKIN, P. A. (1978): Fisheries managernent: an essay for ecologists. - Annual Review of Ecology and Systernatics 9: 57-73.

LECREN, E. D. (1972): Fish production in freshwaters. - Symposia of the Zoological Society of London 29: 115-133.

LEHTONEN, H. (1981): Biology and stock assess1nent of coregonids by the Baltic coast of Finland. - Finn. Fish. Res. 3: 29-83.

LEHTONEN, H. (1983): Scientific basis for fisheries management of vendace, Coregonus albula, in the Bothnian Bay (Baltic Sea). -Aquilo Ser Zoologica 22: 77-82.

24 1.4 References

LEHTONEN, H. (1985): Changes in commercially itnportant freshwater fish stocks in the Gulf of Finland during recent decades. Finn. Fish. Res. 6: 61-70.

LEHTONEN, H. & NIEMELA, E. (1998): Growth and population structure of whitefish (Coregonus lavaretus L.) in 1nountain lakes of northern Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 81-95.

LESKELA, A., BUDD, R., LEHTONEN, H. & SANDSTROM, 0. (1995): Abiotic factors, whitefish stockings and relative year-class strength of anadron1ous whitefish (Coregonus lavaretus L.) spawning populations in the Gulf of Bothnia. - Arch. Hydrobiol. Spec. Issues Advanc. Li1nnol. 46: 241-248.

LIECHTI, P. (1994): Der Zustand der Seen in der Schweiz. - BUWAL Schrifrcnreihe Umwelt 237.

LINNAEUS, C. (1758): Syste1na naturae. I. edition X.

LOEWE, H. & ECKMANN, R. ( 1988): The ontogeny of the alin1entary tract of coregonid larvae, nonnal development. J. Fish Biol. 33: 841-850.

LUCZYNSKI, M., FALKOWSKI, S. & KOPECK!, T. (1988): Larval development in four coregonid species ( Coregonus albula, C. lavaretus, C. 1nuksun and C. peled). - Finn. Fish. Res. 9: 61-69. LUCZYNSKI, M., KUZMINSKI, H., DOBOSZ, s. & GORYCZKO, K. (1998): Gene pool characteristics of whitefish (Coregonus lavaretus) fingerlings produced in a hatchery for restoration stocking purposes. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 317-321.

LUCZYNSKI, M. & RITTERBUSCH-NAUWERCK, B. (1995): Bioche1nical genetic study of European whitefish ( Coregonus lavaretus) in Lake Mondsee, Upper Austria. - Aquatic Sciences 57/2: 127-135.

MAMONTOV, A. M. & YAKHNENKO, V. M. (1998): lsoenzyn1e polymorphisrn in 1.,ake Baikal omul ( Coregomts autumnalis migratorius GEORGI). - Arch. Hydrobiol. Spec. Issues Advanc. Linmol. 50: 375-381.

MARKI. E. & SCHMID, M. (1983): Der Zustand des Hallwilersecs. - Wasser, Energie, Luft 75/4: 105-112.

MATHIAS, J. A. & BARICA, J. (1985): Gas supersaturation as a cause of early spring 1nortality of stocked trout. - Can. J. fish. Aquat. Sci. 4211: 268-279.

25 1 Introduction

MAY, R. C. (197 4): Larval mortality in marine fishes and the critical period concept: 3-19. The Early Life History of Fish. - Blaxter J. H. S. (Ed.), Springer Verlag, Berlin.

MA YR, C ( 199 8): Zurn Einfluss von Trophie, Fischdichte und Habitatwahl auf die Nahrungs- und Wachstmnsbedingungen von Renken (Co reg onus lavaretus L.) in vier oberbayerischen Seen. - Dissertation an der Fakultat der fUr Biologie der Ludwig-Maxilnilians-UniversiUit MUnchen.

MEDGYESY, N. & WIESER, W. (1982): Rearing whitefish (Coregonus lavaretus) with frozen zooplankton by means of a new feeding apparatus. - Aquaculture 28: 327-337.

MICHEL, M. (1996): Untersuchungen zur Nahrungsokologie von Grossfelchen im Vierwaldstattersee wahrend des Sommerhalbjahres 1996. Diploma Thesis ETH. No. 10366: 73pp.

MILLS, K. H. & CHALANCHUK, S. M. (1988): Population dynamics of unexploited lake whitefish (Coregonus clupeaformis) in one experi1nentally fertilized lake and three exploited lake. - Finn. Fish. Res. 9: 145-153.

MOOKERIJ, N., HELLER, C., MENG, H. J., BlJRCJI, H. R. & MULLER, R. (1998): Diel and seasonal patterns of food intake and prey selection by Coregonus sp. in re-oligotrophicated Lake Lucen1e, Switzerland. - J. Fish Biol. 52/3: 443-457.

MOOKERJI, N. & RAO, T. R. (1993): Patterns of prey selection in rohu (Labeo rohita) and singhi (Heteropneustes fossilis) larvae under light and dark condition. - Aquaculture 118: 85-104.

MOOKERJI, N. & RAO, T. R. (1994): Influence of ontogenetic changes in prey selection on the survival and growth of rohu, Labeo rohita, and singhi, Heteropneustesfossilis, larvae. - J. Fish Biol. 44: 479-490. MOOKERJI, N. & RAO, T. R. (1995): Prey capture success, feeding frequency and daily food intake rates in rohu, Labeo rohita (Ham.), and singhi, Heteropneustes fossilis (Bloch), larvae. - Journal of Applied Ichthyology 11: 37-49.

26 l .4 References

MULLER, R. (1990): Managen1ent practices for lake fisheries in Switzerland. In: van Densen, W. L. T.; Steinmetz, B.: Hughes, R.H., (Eds), Management of freshwater fisheries. Pudoc. Wageningen: 477-492

MULLER, R. (1992): Trophic state and its implications for natural reproduction of saln1onid fish. - Hydrobiologia 243-244: 261-268.

MOLLER, R. (1993): Einige fischereibiologische Aspekte von Seesanierungen. - Fortschr. Fisch.wiss. 11: 43-56.

MULLER, R. & BIA, M. M. (1998): Adaptive 1nanage1nent of whitefish stocks in lakes undergoing re-oligotrophication: The Lake Lucerne example. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 391-399.

MllLLER, R., BIA, M. M. & MENG, H.J. (1994): Die Felchenfischerei in einigen Seen der Zentralschweiz und des Mittellandes. l:3 UW AL Mitteilungcn zur Fischerei 55.

MUNKITTRICK, K. R. & DIXON, D. G. (1989): Use of white sucker (Catostornus commersoni) populations to assess the health of aquatic ecosystems exposed to low-level contaminant stress. - Can. J. Fish. Aquat. Sci. 46: 1455-1462.

NAESJE, T. F., SANDLUND, 0. T. & JONSSON, B. (1986): Habitat use and growth of age-0 whitefish, Coregonus lavaretus, and cisco, Coregonus albula. - Environn1ental Biology of Fishes 15/4: 309-314.

NAGY, S. (1990): The food of the whitefish (Coregonus lavaretus L.) in the Strbske pleso mountain lake (Czechoslovakia). - Zivocisna Vyroba 35110: 915-920.

NIESSLBECK, P. (1987): Aufwuchsbcdingungen fiir nattirlich geschliipftc und kUnstlich erbri.Hete Larven von Coregonus lavaretus im Starnberger See (Oberbayern) . - Forschungsbcricht ans dcr Bayrischen Landesanstalt fUr Fischerei in Stamberg.

NILSSON, N. A. (1985): The niche concept and the introduction of exotics. - Rep. Inst. Fresh w. Res. Drottning ho hn 53: 51-77.

27 l Introduction

ORTLEPP, J. (1984): Beobachtungen zurn Verlauf des Wachstmns und der Sterblichkeit von Coregonenlarven bei Fiittenmg nrit Zooplankton. - Diplomarbeit Biologic, Universitlit Freiburg.

PAULIK, G. J. (1973): Studies of the possible fonn of the stock-recruitment curve. Rapp. P.-v. Reun. Cons. perm. int. Explor. Mer. 164: 302-315.

POMEROY, P. P. (1991): A comparative assessment of temporal variation in diet of powan, Coregonus lavaretus (L.), from Loch Lomond and Loch Eck, Scotland, UK. J. Fish Biol. 38/3: 457-478.

PONTON, D. & MENG, H.J. (1990): Use of dual-bean1 acoustic technique for detecting young whitefish, Coregonus sp., juveniles: first experiments in an enclosure. - J. Fish Biol. 36: 741-750.

PONTON, D. & Mt}LLER, R. (l 989): Alirnentation et facteurs de n1ortalite des larves de coregones (Coregonus sp.): Exemple de deux lacs de niveaux trophiques differents, les lacs de Sa111en et de Hallwil (Suisse Centrale ): - Aquatic Sciences 5111: 67-83.

PONTON, D. & MlrLLER, R. (1990): Size of prey ingested by whitefish, Coregonus sp., larvae, Are Coregonus larvae gape-1in1ited predators? J. Fish Biol. 36/l: 67-72.

REY, P. & ECKMANN, R. (1989): The influence of lake temperature on growth of Coregonus lavaretus L. larvae. - Arch. Hydrobiol. 116/2: 181-190.

RICE, J. A., CROWDER, L. B. & BINKOWSKI, F. P. (1987): Evaluating potential sources of mortality for larval bloater (Coregonus hoyi): Starvation and vulnerability to predation. - Can. J. Fish. Aquat. Sci. 44/2: 467-472.

RICKER, W. E. (1954): Stock and recruit1nent. J. Fish. Res. Board Can. 11: 559- 623.

RICKER, W. E. (1975): Computation and interpretation of biological statistics of fish populations. - Bull. Fish. Res. Board Can. 191: 1-382.

RICKER, W. E. (1977): The Historical Development. In: Gulland J. A. (eel): Fish population dynamics. 2nd edition, 1988, 1-26. John Wiley & Sons, Ltd., New York.

28 1.4 References

ROSCH, R. (1986): Satiation of juveniles of Coregonus lavaretus L. fed on Daphnia pulicaria. - Schweiz. Z. Hydrol. 48/l: 44-52.

ROSCH, R. (1987): Effect of experimental conditions on the stomach evacuation of Coregonus lavaretus L. - J. Fish Biol. 30: 521-531.

ROSCH, R. (1988): Mass rearing of Coregonus lavaretus larvae on a dry diet. Finn. Fish. Res. 9: 345-351.

ROSCH, R. (1989): Beginning of food intake and subsequent growth of larvae of Coregonus lavaretus L. - Pol. Arch. Hydrobiol. 36/4: 475-484.

ROSCH, R. (1992): Food intake and growth of larvae of Coregonus lavaretus L., effect of diet and light. - Pol. Arch. Hydrobiol. 39: 671-676.

Rosen, R. (1994): Problen1e bei der Aufzucht von Felchenlarvaen (Coregonus lavaretus) mit Lebendplankton. - European Association of Fish Pathologists, 88-91.

ROSCH, R. & APPELBAUM S. (1985): Experi1nents on the suitability of dry food for larvae of Coregonus lavaretus L. - Aquaculture 48: 291-302.

ROSCH, R. & DABROWSKI, K. (1986): Tests of artificial food for larvae of Coregonus lavaretus from Lake Constance. Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 273-282.

ROSCH, R. & ECKMANN, R. (1986): Survival and growth of prefed Coregonus lavaretus L. held in illuminated net cages. - Aquaculture 52: 245-252.

ROSCH, R. & MARMULLA G. (1990): Maximun1 daily ration of juvenile fish fed on living natural zooplankton. - J. Fish Biol. 36: 789-801.

ROSCH, R. & SEGNER, H. (1990): Development of dry food for larvae of Coregonus lavaretus L. 1. Growth, food digestion and fat absorption. - Aquaculture 91: 101-115.

ROTH, H. & GEIGER, W. (1968): Aufzucht von Besatzfischen in Trogen: 3-39. - Veroffcntlichungen des Eigenossischen Amtes fur Gewasserschutz und der Eidgenossischen Fischereiinspektion 25: 3-39.

ROTI-I, H. (l 9 54): Massnalnnen zur Hebung der Felchen besttinde in den Seen des Kantons Bern. - Schweiz. Fischerei-Zeitung 62: 94-971125-127.

29 l Introduction

RUDSTAM, L. G. & MAGNUSON, J. J. (1985): Predicting the vertical distiibution of fish populations: Analysis of cisco, Coregonus artedii, and yellow perch, Percaflavescens. - Can. J. Fish. Aquat. Sci. 42/6: 1178-1188.

RUFLI, H. (1978): Die heutigen sympatrischen Felchenpopulationen (Coregonus spp.) des Thuner- und Bielersees und ihre Morphologie. Schweiz. Z. Hydrol. 40: 7-31.

SAILA, S. B., LORDA, E. (1980): A critique and generalization of the Ricker stock- recruitment curve. Graduate School of Oceanography, Narrangansett.

SALOJARVI, K. (1986): Review of whitefish (Coregonus lavaretus L. s.L) fingerling rearing and stocking in Finland. Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 99-114.

SALOJARVI, K. (1987): Why do vendace (Coregonus albula L.) populations fluctuate? - Aqua Fennica 1711: 17-26.

SALOJARVI, K. (1988): Effect of the stocking density of whitefish (Coregonus lavaretus L. s.l.) fingerlings on the fish yield in Lake Peranka, Northern Finland. - Finn. Fish. Res. 9: 407-416.

SALOJARVI, K. (1991 a): Stock recruitment relationships in the vendace (Coregonus albula L.) in Lake Oulujarvi, northern Finland. -Aqua Fennica 2112: 153-161.

SALOJARVI, K. (1991 b ): Recruitment mechanisms of the vcndace (Co reg onus albula L.) in Lake Oulujarvi, north em Finland. - Aqua Fennica 21/2: 163- 173.

SALOJARVI, K. (1992): Compensation in whitefish (Coregonus lavaretus L. s.L) populations in Lake Oulujitrvi, northern Finland. Finn. Fish. Res. 13: 31- 48.

SALONEN, E., AHONEN, !\1. & :MUTENIA, A. (1998): Developrnent of whitefish ( Coregonus lavaretus) population and effects of large-scale compensation stocking in Lake Inari, northern Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 439-448.

30 1.4 References

SANDLUND, 0. T., JONSSON, B., NAESJE, T. F. & AASS, P. (1991): Year-class fluctuations in vendace, Coregonus albula (Linnaeus): Who's got the upper hand in intraspecific cornpetition? - J. Fish Biol. 38/6: 873-886.

SARVALA, J., RAJASILTA, M., HANGELIN, C., HIRVONEN, A., KIISKILA, M. & SAARIKARI, V. (1988): Spring abundance, growth and food of O+ vendace (Coregonus albula L.) and whitefish (C. lavaretus L.) in Lake Pyhajarvi, SW Finland. -Finn. Fish. Res. 9: 221-233.

SCHEIDEGGER, A. (1992): Sauerstoffliaushalt 1m Hallwilersee: Eine Untersuchung des Einflusses der internen Massnalnnen zur Seesanierung.- Dipl01narbeit an der Eidgenossischen Technischen Hochschule.

SCHEIDEGGER, A., STOCKLI, A. & WlJEST, A. (1994 ): Einfluss der inten1en Sanierungsmassnahn1en auf den Sauerstofflrnushalt im Hallwilersee. - Wasser, Energie, Luft 86/5-6: 126-131.

SCHMID, W. (1993): Tagesfressrhyth1nik von Felchenlarven (Coregonus lavaretus L. ), Vergleich von Lebend- zu Trockenfutter. - Diplome work, University of Constance.

SCHULTZ, H. (1992): Stock density, growth rate and zooplankton consun1ption of the vendace ( Coregonus allmla) and other fish species in Lake Arendsee. - Limnologica 22/4: 355-373.

SEGNER, H. & RbSCH R. (1990): Development of dry food for larvae of Coregonus lavaretus L. II. Liver histology. - Aquaculture 91: 117-130.

SHEPHERD, J. G. (1982): A versatile new stock-recruitinent relationship for fisheries, and the construction of sustainable yield curve. - J. Cons. int. Explor. Mer. 40: 67-7 5.

SHUTER, B. J. (1990): Population-level indicators of stress. - Ain. Fish. Soc. Symp. 8:93-108.

SKURDAL, J., BLEKEN, E. & STENSETH, N. C. (1985a): Cannibalis1n in whitefish (Coregonus lavaretus): - Oecologia (Berlin) 67/4: 566-571.

SKURDAL, J., HESSEN, D. 0. & BERGE, D. (1985b): Food selection and vertical distribution of pelagic whitefish Coregonus lavaretus in Lake Tyrit]orden, Norway. - Fauna Norvegica Series A 6: 18-23.

31 1 Introduction

STADELMANN, P. (1988): Zustand des Sempachersees. - Wasser, Energie, Luft 80/3-4: 81-96.

STEFFENS, W. (1995): Yield and stocking of vendace (Coregonus albula) in northeast Germany. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 405-412.

STEINHART, M. & ECKMANN, R. (1992): Evaluating the nutritional condition of individual whitefish (Coregonus spp.) larvae by the RNA/DNA ratio. J. Fish Biol. 40: 791-799.

STEINMANN, P. (1950a): Ein neues Syste1n der mitteleuropaischen Coregonen. - Revue Suisse de Zoologie 57/10-32: 517-525.

STEINMANN, P. (1950b ): Monographie der schweizerischen Coregonen I. Teil. - Schweiz. Z. Hydrol. 12.

STEINMANN, P. (1951): Monographie der schweizerischen Coregonen II. Teil. Schweiz. Z. Hydrol. 13.

STOCKLI, A. & SCHMID, M. (1987): Die Sanierung des Hallwilersees. Erste Erfahrungen mit zwangszirkulation und TiefenwasserbeliiJtung. - Wasser, Energie, Luft 7917-8: 143-149.

STOSSEL, F. (1992): Die Bodenfauna im Hallwilersee dringt vor. EAWAG-news 34D: 23-26.

SVARDSON, G. (1977): Interspecific Population Dominance in Fish Communities of Scandinavian Lakes. - Rep. Inst. Freshw. Res. (Drottningholm) 55: 144- 171.

SVARVAR, P. 0. & MlJLLER, R. (1982): Die Felchen des Alpnachersees. - Schweiz. Z. Hydrol. 44/2: 295-314.

TAYLOR, W.W. & FREEBERG, M. H. (1984): Effect of food abundance on larval lake whitefish, Coregonus clupeafonnis Mitchill, growth and survival. - J. Fish Biol. 25: 733-741.

TODD, T. N. (1998): Environmental modification of gillraker number in coregonine fishes. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 305-315.

32 1.4 References

TURKOWSKI, K. & BONAR, A. (1995): Effects of species composition and stocking on cmmnercial catches of vendace, Coregonus albula (L.) in Ostroda lakes (northern Poland). - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 397-403.

URBAN-JEZIERSKA, E. & BIERNACKI, T. (1992): The effect of natural food and starters on growth and development of Coregonus lavaretus in the 1st inonth of life. -Pol. Arch. Hydrobiol. 39/1: 109-132.

VENTLING-SCHWANK, A. R. (1992): Reproduk6on und larvale Entwicklungs- phase der Felchen ( Coregonus sp.) im eutrophen Sempachersee. - Dissertation an der Philosophischen FakulHH der Universifat Ziirich.

VENTLING-SCHWANK, A. R & LIVINGSTONE, D. M. (1994): Transport and burial as a cause of whitefish ( Coregonzts sp.) egg n10rtality in a eutrophic lake. - Can. J. Fish. Aquat. Sci. 5119: 1908-1919.

VENTLING-SCHWANK, A. R. & MULLER, R. (1991): Survival of coregonid (Coregonus sp.) eggs in Lake Sernpach, Switzerland. - Verh. Intenrnt. Verein. Lirnnol. 24: 2451-2454.

VILJANEN, M. (1983): Food and food selection of cisco (Coregonus albula L.) in a dysoligotrophic lake. - Hydrobiologia 101: 129-138.

VILJANEN, M. (1986): Biology, propagation, exploitation and management of vendace (Coregonus albula L.) in Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 73-97.

VILJANEN, M. (1988): Population dynarnics of vendace (Coregonus albula L.) in Finland. - Joensuun Yliopiston Luonnontieteellisia Julkaisuja 12: 1-19.

WAGLER, E. (1941): Die Lachsartigen (Salmonidae). II. Teil: Coregonen. In: Dernoll, R. and Maier H.N. (eds.). - Handbuch der Binnenfischerei Mitteleuropas III. Schweizerbart. Stuttgart.

WAGLER, E. (1950): Die Coregonen in den Seen des Voralpengebietes. XI. Herkunft und Ein wanderung der Voralpencoregonen. - Veroffentl. Zool. Staatssarmnl. Miinchen, 1/3: 3-62.

33 1 Introduction wANZENBOECK, l & JAGSCH, A. (1998): Comparison of larval whitefish densities in lakes with different schemes of larval stocking and fishing practice. - Arch. Hydrobiol. Spec. Issues Advanc. Litnnol. 50: 497-505.

WATANABE, T., KITAJIMA, C. & FUJITA, S. (1983): Nutritional values of live organisms used in Japan for mass propagation of fish, a review. - Aquaculture 34: 115-143.

WATERS, T. F. (1977): Secondary production in inland waters. - Adv~mces in Ecological Research 10: 91 164.

WEHRLI, B. & WDEST, A. (1996): Zelm Jahre Seenbeliiftung: Erfahnmgen uncl Optionen. - Schriftenreihe EA WAG 9.

WEITKAMP, D. E. & KATZ, M. (1980). A review of dissolved gas supersaturation literature. - Trans. Amer. Fish. Soc. 109: 659-702.

WOLOS, A., FALKOWSKI, S. & ABRAMCZYK, A. (1995): Nlanagement of coregonines in the big State Fish Farm Elk production, stocking practice, and effectiveness. Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 387-396.

ZIMMERMANN, C., KOLL\. & SCHLATTER. J. (1991): Zuf1ussuntersuchung zur Nahrstoffbelastung 1988/90. - \Vasser, Energie, Luft 84

34 1 Abstract

2 FISH SPECIES DIVERSITY

2.1 Abstract

A study of the fish fauna of Lake HallwiL Switzerland. with an emphasis on whitefish, was conducted to investigate the consequences of the devastating fish kill in 1961.This event caused a reduction or even the extinction of 1110st fish species in Lake Hallwil. Oxygen deficiency was found to be the cause of the fish kill. Little was known about the recovery and the actual fish species diversity. Therefore, a species list was established on the basis of fishing yield statistics before and after the fish kill. Twenty fish species were found. Three of them were typical river- living species. While yields of most of the fish species showed 1noderate fluctuations, yields of arctic chaff (Salvelinus alpinus), lake-d\velling brown trout (Salmo trutta lacustris), roach (Rutilus rutilus) and whitefish (Coregonus suidteri) were especially variable. Roach is commercially unimportant, but has s01netimes contributed heavily to the yield. Because arctic charr and lake-dwelling brown trout have always been caught irregularly, stock size of both species was assmned to be s1naJl and to rely on stocking. Thanks to intensive stocking, whitefish has become dominant after 1977. Coregonid larvae stocked after 1961 originated predmni- nantly fr01n Lake Sempach. Therefore, the actual phenotype of Lake Hallwil whitefish is similar to that of Lake Sempach \Vhitefish. Overall, fish recovery since 1961 was successful even for species without stocking. Still, eutrophication is the 111ain proble1n related to restricted or unsuccessful reproduction.

2.2 Introduction

Eutrophication of lakes is a natural aging process. Hypereutrophication or cultural eutrophication, however. constitutes one of the most conunon and devastating anthropogenic disturbances of aquatic ecosystems (LA:\fPERT & SOMMER, 1993). At the early stages of eutrophication, fish stock increases due to an increasing food base. With increasing eutrophication, however, natural reproduction of

35 2 Fish species diversity sahnonid fish, e.g. whitefish (Coregonus sp.) and arctic charr (Salvelinus alpinus), is no longer possible due to the anoxic sediment (MtJLLER, 1992). Furthennore, devastating fish kills due to oxygen depletions inay occur (LAMPERT & SOMMER, 1993).

1 Aegeri 2 Baldegg 3 Hallwil 4 Lucerne 5 Lungern 6 Sempach 7 Zug

Figure 2.1 Swiss lakes historically inhabited by coregonid fonns called "Ballen" that are closely related to Lake Hallwil whitefish.

Lake Hallwil, Switzerland (Figure 2.1), is an example of a lake that has undergone cultural eutrophication. According to criteria in KLEE (1991), Lake Hallwil would naturally be a coregonid lake. PATIO (1890) hypothesized that fish have colonized Swiss lakes using the large post-glacial rivers as pathways and following coloni- sation fanned lake specific fish communities. In Lake Hallwil, the local coregonid form is called "Ballen" (Coregonus suidteri Patio; nomenclature according to KOTTELAT, 1997; Figure 2.2). It is a fast growing coregonid fonn that spawn on gravel at the shore during November and December (STEINMANN, 1950b). Several village naines (e.g. Ballwil) apparently are derivations oflocal fish names, reflect- ing the importance of the fisheries in the Middle Ages. The establishment and expansion of settlements brought an increasing load of pollutants. The first bloom of the cyanobacterimn Planktothrix rubescens in 1898 heralded the beginning of

36 2.2 Introduction

cultural eutrophication. As a consequence, total whitefish yield in Lake Hallwil decreased drastically from 3,600 kg in 1914 to 900 kg in 1919, despite the stocking of newly hatched whitefish larvae (BRUTSCHY & GUNTERT, 1923). Because natural reproduction of coregonids is unsuccessful in eutrophic lakes (VENTLING- SCHW ANK & M-OLLER, 1991) stocking is necessary to 1naintain populations. Although BRUTSCHY & G-ONTERT (1923) recognized that water pollution was the main cause of the whitefish population decrease in Lake Hallwil, forty years passed before the first sewage treatment plant was built. In October 1961, low oxygen concentration as a consequence of severe eutrophication (LAMPERT & SOMMER, 1993), led to a devastating fish kill (HEIMANN, 1962), causing a major reduction or extirpation of 1nost fish species in Lake Hallwil. From 1964 on, waste waters from households and industries were treated and no longer discharged into Lake Hallwil. Nevertheless, eutrophication further increased and reached its peak in 1977. Several other measures were taken to lower the phosphorous input and to iinprove conditions for fish. A ban on phosphorous in detergents and an aeration system to oxygenate the anaerobic hypolimnion were introduced in 1986. As a consequence, phosphorus concentration decreased to 40-SOµg/l by 2000. Never- theless, Lake Hallwil maintains a eutrophic state.

Knowledge on the recovery of fish populations after the kill of 1961 is limited. Fry of the c01mnercially exploited fish species (pike, lake-dwelling brown trout, arctic charr and whitefish) were stocked in the lake, employing a n1easure commonly used to restore s1nall or extinct populations (LUCZYNSKI et al., 1998). The eggs and fry stocked in Lake Hallwil originated almost exclusively frmn Lake Sempach (H. MINDER, pers. com.). Intensive stocking of whitefish since 1970 has led to the recovery of this species, making it the dominant and conunercially most impmiant fish species (MlTLLER et al., 1994).

The aims of this work are, first, to document the fish species cmnposition in Lake Hallwil and, second, to conduct a detailed analysis of the whitefish currently inhabiting the lake. If the original fon11 of whitefish in Lake Hallwil was extfr- pated in 1961, the current whitefish population in Lake Hallwil should closely resemble the Lake Se1npach form, whose progeny was principally used to restore the Lake Hallwil stock. We therefore cmnpared morphometric paran1eters of Lake Hallwil and Sempach whitefish measured by STEINMANN (1950b, 1951), before the fish kill in 1961, with our own measurements in 1997 and 1998. Our aim was

37 2 Fish species diversity

to determine whether or not Lake Hallwil whitefish remained phenotypically similar to the population inhabiting the lake before the fish kill in 1961.

2.3 Methods

2.3.1 Fish species diversity

Species lists were tnade on the basis of yield statistics frmn conunercial and sports fisheries. Mean yields from the periods 1938-57 and 1967-98 were used to estimate population abundance and recovery of the species after the fish kill in 1961. Our list was further cmnpared to the findings of JEAN-RICHARD (1997) who identified 15 fish species by direct observation in the littoral zone of Lake Hallwil.

To determine why whitefish has become the dominant species since 1977, year class strength was estimated for 1967-96 by virtual population analysis (VPA) on the basis of age distribution in the catch and total annual yield. Then, numbers of whitefish stocked as newly hatched or prefed larvae were regressed against year

class strength.'-"

2.3.2 Lake Hallwil whitefish phenotype

}figure 2.2 Measuretnents conducted on whitefish, according to RUFLf (1978) and STEINMANN (1950b and 1951): girth (1), eye diatneter (2), pectoral fin length (3), number of scales on lateral line (4) and gillraker nmnber (5). Scales for age determination were taken frmn area 6.

Just before the spawning season at the end of October in 1997 and 1998, 127 adult

38 2.4 Results

Lake Hallwil whitefish aged 2 to 6 years were caught using gillnets of bar mesh sizes between 20 and 40 mm. Only freshly caught and mature individuals were used for the study. According to RUFU (1978) and STEINMANN (1950b), girth, eye diameter, pectoral fin length (all in percent of standard body length), number of scales and gillrakers were chosen to characterize Lake Hallwil whitefish. Girth and pectoral fin length were measured to the nearest 1 mm, and eye dimneter to the nearest 0.5 irnn. Gillraker numbers of the 127 whitefish were checked for normal distribution by a MACDONALD & PITCHER (1979) mixture analysis.

2.4 Results

2.4.1 Fish species diversity

The fish fauna of Lake Hallwil includes twenty species and subspecies in seven fa1nilies (Table 2.1). Fish were grouped into four categories of estimated population abundance: dominant species; very com1non species, which were caught regularly in large ainounts; common species, which appeared in the catch regularly or irregularly in low ainounts; and rare species, that were either typical river- dwelling species or species not captured. In Table 2.1, regularly caught means that fish species appeared every year in the catch statistics, whereas irregularly caught means that fish species were absent in the yield statistics in single years or even for a longer period.

Whitefish (Figure 2.2) has been the dominant fish species in J_,ake Hallwil (Figure 2.6) since 1977 (MlJLLER et al., 1994). Coregonid yield before the fish kill of 1961 has fluctuated between about 200 and 2000 kg per year (mean 1000 kg, Table 2.1 and Figure 2.5). The mean yield before 1961 was comparable to the mean whitefish yield of about 700 kg per year for 1967-76 (Figure 2.6). In 1977, more than 3,199 kg whitefish were caught (Figure 2.6). Since 1977, this 111ark has not been underscored any more (3,441 kg in 1993). Overall, mean whitefish yield increased after the fish kill of 1961to2000% of the value before 1961(Table2.1). Perch and pike, also important for the fisheries, and smne connnercially less important species like roach, bream and tench were always very con11non in the catch (Table 2.1, Figure 2.5 and Figure 2.6). Pikeperch, carp, burbot and eel appeared regularly in relatively small amounts ( <350 kg per year).

39 2 Fish species diversity

Table 2.1 Fish species list and mean annual yield [kg] for 1939-57 J 1967-98 =riverine species that rnay be accidentally in Lake Hallwil: 0 A= introduced species; • =regularly in catch; =irregularly in catch; ? =no data or not captured).

very family rare common common dominant

c illa (eel) Centrarchidae Lepomis gibbosus11 .) ')ff) (pumkinseed)

I Cyprinidae 11hramis brama (bream) • ?/290

Barhus barbus (barbel) 0 '?/'.1 Cyprinus ca17.?io (carp) • ?/35

') !') Ciobio gobio (gudgeon) ,; . Leuciscus cephalus (chub) ?/? Rutilus mtilus (roach) • 600/3,000

erytlzroplitllalmus 0 ?/?

Tinca tinca (tench) • 20/150 ---- Esocidae Esox lucius (pike) • 2001700 ----·-·- Gadidae Lota Iota (burbot) • 15/25 Percidae •

11 ')I~ - Sander lucioperca (pikeperch) .; ;_.) Salmoniclae Coregonus suidteri (whitefish) •

40 2.4 Results

Arctic charr and lake-dwelling brown trout were in the smne range of abundance but were caught irregularly (Table 2.1, Figure 2.5 and Figure 2.6). Some other fish species (barbel, gudgeon, pumpkinseed, chub and rudd) were recorded by JEAN-RICHARD (1997), but were seldom caught and not clearly declared in the statistics (Table 2.1)

Rainbow trout, river-dwelling brown trout and grayling (Table 2.1) are typical for running water. Therefore, they appear irregularly and in small amounts in the catch. Pikeperch, pumpkinseed and rainbow trout are not native to Lake Hallwil. One of them, pikeperch, is caught regularly (Table 2.1 and Figure 2.6). Ruffe (Gymnocephalus cernuus), a native species that has become established in some Swiss lakes in recent years, has not been reported for Lake Hallwil (HOLTEY, 1996).

2.4.2 Lake Hallwil whitefish phenotype

All five parameters 1neasured showed very close agreement with the findings of STEINMANN (1950b; see Table 2.2) for whitefish collected in Lakes Hallwil and Sempach before the fish kill in 1961. Lake Hallwil whitefish gillraker numbers were normally distributed (Figure 2.3).

30 .c (/) t;:: (1) :t= .c 20 ~ 4- 0 I- (1) ..0 10 E ::l c 0 -- 27 30 33 37 41 gill raker number

Figure 2.3 Gillraker number of 127 whitefish frmn Lake Hallwil, srunpled in 1997/98. The curve indicates a normal distribution.

41 2 Fish species diversity

Most whitefish were five years old (Figure 2.4) All fish analysed were sexually mature. Because all whitefish were fast-growing, there is no evidence for the occunence of a slow growing form in Lake HallwiL This agrees with present knowledge based on much larger smnple size (MOLLER et al., 1994).

• 2+ 20 m 3+ ~ 4+ 15 D 5+ t'.3 6+

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ~ w ~ ro m o ~ N ~ ~ ~ w ~ ro m o N N N N N M M M M M M M M M M ~ length [mm]

Figure 2.4 Length and age distribution of mature Lake Hallwil whitefish (n=l27), sainpled in 1997/98.

2.5 Discussion

2.5.l Fish species diversity

Direct observations by JEAN-RICHARD (1997) revealed that at least fifteen fish species occur in Lake Hallwi1 today: barbel, bremn, burbot, carp, chub, eel, gudgeon, lake-dwelling brown trout, perch, pike, pumpkinseed, roach, rudd, tench, whitefish. If we compm·e our findings shown in Table 2.1, based on catch statistics, with the observations by JEAN-RICHARD (1997), then arctic charr, pikeperch, grayling, river-dwelling brown trout and rainbow trout (none of which were recorded by JEAN-RICHARD, 1997), have to be added to the list. Overall, we conclude that 17 lake fish species permanently live in Lake Hallwil, while 3 river fish species visit the lake periodically.

42 2.5 Discussion

Population estimations based on yield statistics can be adequate for exploited fish species like whitefish, pike and perch (Table 2.1, Figure 2.5 and Figure 2.6).

o other fish species D perch B 3000 [)pike 6 "O ~lake-dwelling trout a:; 11 whitefish >, 2000 cu ::J c c cu 1000

year

Figure 2.5 Annual fishing yield by species in Lake Hallwil between 1938 and 1957, based on commercial harvest.

However, for commercially unimportant and small fish species, this link is less obvious because those species are caught accidentally or not at all. Nonetheless, the larger a population, the higher the probability that non-exploited fish species are also caught in gillnets, unless they live in areas which are not fished at all (e.g. gudgeon in the littoral zone). Based on Table 2.1 we conclude that not only whitefish, pike and perch are abundant, but also the very commonly caught species bream, roach and tench. Unfortunately, information on yield from before the fish kill of October 1961 is less detailed than for the following period (Table 2.1, Figure 2.5, and Figure 2.6). This restricted our study to a few fish species. There- fore, we have no data on the yield of arctic charr, breain, carp and pikeperch for the pre-fish-kill period. Furthermore, because barbel, gudgeon, pumpkinseed, chub and rudd were seldmn caught and not separately recorded, we lack the information to estimate their stock size for both time periods. We treated the111 as rare species. Still, chub could also be a conunon species because JEAN-RICHARD (1997) observed more than 200 of them in the littoral zone. The three typical river fish species (rainbow trout, river-dwelling brown trout and gray ling) do not reproduce in the lake and seem to occur only near inflowing rivers. Yield and thus stock

43 2 Fish species diversity

size of eel, burbot, lake-dwelling brown trout, bream, pike, roach, tench and whitefish was also higher after 1961 than before. This 1nay be explained ba an increased food base utilized by fish species whose natural reproduction was still functional, or by fish species which were stocked.

80000 70000 __ •whitefish 60000 - - D all others 50000 40000 30000 20000 10000 0 ,...... , 14000 ~roach OJ ~ C:J perch ...... 12000 (]pike "O 10000 .~ D bream >. 8000 •tench cu :::; 6000 c c 4000 cu 2000 0 rn pike perch D burbot ------350 [Jlfil carp ~ arctic charr 300 - a lake trout •eel 250 200 150 100 50 0 ~ m ~ ~ ~ ~ m ~ ~ ~ ~ m ~ M ~ ~ ill ill ~ ~ ~ ~ ~ ro ro ro ro ro m m m m m m m m m m m m m m m m m m m m ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

year

Figure 2.6 Annual yield of the 12 fish species appearing in the catch statistics of Lake Hallwil between 1967 and 1998, based on conunercial harvest. Note different scales on the vertical axis.

44 2.5 Discussion

Whitefish yield exhibited the highest difference between pre- and post-fish-kill period. Before 1961, the decline frorn 3,600 kg in 1914 to 900 kg in 1919 (BRUT- SCHY & G(TNTERT, 1923) was followed by a increase between 1939 and 1957 (Figure 2.5). As shown in Figure 2.5, yield during this period was always lower than in 1914 (3,600 kg). Between 1967 and 1976, mean whitefish yield was always less than 2,000 kg, but in 1977, yield started to increase and reached its maxirnum in 1997 (over 70,000 kg; Figure 2.6). After the fish kill, the whitefish population was not only successfully restored by stocking larvae, but it has even become the dorninant fish species (Figure 2.6). The enormous increase of whitefish yield since 1977 could be explained by an intensification of the stocking practice. Since 1968, part of the whitefish larvae were fed in circular tanks and stocked after at least six weeks of rearing. Before 1968, only newly hatched larvae were stocked (BRUTSCHY & GONTERT, 1923; H. MINDER, pers. com.). In addition, after 1968, number of larvae stocked increased enormously (H. MINDER, unpubl. data). As shown in Figure 2. 7, regression of the nurnber of prefect larvae stocked versus year class strength gives a better fit than that with the nmnber of newly hatched larvae stocked. 600000 500000 400000 :c tll t;:: 300000 ..... 0 200000 •.• , ..•... 1.- ...... • , Qi D 100000 E ::J .s 0 .c: 10 20 30 40 50 Di c :!! newly hatched whitefish larvae [Mio] u; tll 600000 tll Ill y = 58902x - 33591 (3 500000 ro = 0,5 - 300000 200000 100000 0 0 2 4 6 prefed whitefish larvae [Mio]

Figure 2.7 Regressions between nmnber of stocked whitefish as newly hatched larvae (above) or prefect larvae (below) and year class strength (1967- 96) for Lake Hallwil.

45 2 Fish species diversity

Nevertheless, coregonid stock size showed enormous fluctuations and even a yield collapse in 1992-94 (Figure 2.6). The causes for this collapse are the subject of further investigations (chapter 6). Yield of other fish species except roach, arctic charr and lake-dwelling brown trout experienced less pronounced fluctuations (Figure 2.5 and Figure 2.6). Roach stock size was estimated to be about five times higher after the fish kill than before, but it showed extre1ne fluctuations (Table 2.1, Figure 2.5 and Figure 2.6). Because roach is generally unimportant for the fishery, no stocking has ever been clone. Roach is much less sensitive to eutrophication-induced changes as compared to whitefish and arctic charr (tv10LLER, 1992). Arctic charr has also been reintroduced by stocking, with limited success (e.g. 1983/84; Figure 2.6). Because arctic charr and lake-dwelling brown trout are potential predators of whitefish juveniles, stocking of these species was less intensive than of whitefish. Generally, yields of the two piscivorous salmonids have been low and irregular (Table 2.L Figure 2.5. and Figure 2.6). Never- theless, lake-dwelling brown trout stock size increased after the fish kill (Table 2.1). Overall, fish population recovery after the kill in 1961 seemed successful even for non-exploited and unstocked species. Still, for smne years to come, eutrophication appears to be the n1ain proble111 for natural reproduction of smne fish species, principally the salmonids (MOLLER, 1992).

2.5.2 Lake Hallwil whitefish phenotype

The lack of well developed reproductive barriers between the coregonid fish makes a delimitation of whitefish species and populations difficult (HIMBERG & LEHTONEN, 1995). Therefore, the question about Lake Hallwil whitefish taxonmny remained uncertain for centuries. Several atten1pts to classify the coregonids of Switzerland were done since the 16th century by several authors as described by STEINMANN (1950b). LINNE (1758) gathered all coregonid forms in northern Europe in the two species Coregonus albula and Coregonus lnvaretus. The latter is commonly cited for all Swiss coregonid forms until today. The next main attempt was done at the end of the 19th century. FATIO (1885) built up a system close to the local names. The "Ballen" of Lake Hallwil was called Coregonus annectus. Because the ;'Ballen" of Lake Sempach, which is considered conspecific with the original Lake Hallwil whitefish (KOTTELAT, 1997), was nan1ed Coregonus suidteri, the classification of FATlO (1885) has to be considered inadequate. In the 20th century, BERG (1932) recognized one single species in central

46 2.5 Discussion

Europe ( C. lavaretus L. ), including several sub- and infrasubspecies. However, his natio concept was never really accepted outside eastern Europe (KOTTELAT, 1997). WAGLER (1941 and 1950) concluded that there were four species spreading all over Europe and occupying different lakes in different cmnbinations. The findings of DOUGLAS (1999), based on microsatellite rnarkers, rejected Wagler's concept. Just ten years before the fish kill in Lake Hallwil, STEINMANN (1950b & 1951) published his "Monography of the Swiss Coregonids". On the basis of girth, head length, eye diameter, tail stem, pectoral fin length, nmnber of scales and gillrakers (Table 2.2), he c01nbined the Lake Hallwil, Baldegg and Sempach (Figure 2.1) whitefish into Coregonus lavaretus L. intermedia prirnigenius (STEINMANN, 1950b). Intermedia is a geographical description for "Ballen" of the Swiss midland. Primigenius is a term for ecotype specification and reflects the silnilarity of those fish to the evolutionary original whitefish form (STEINMANN, 1950a). But Steimnann' s concept of a single species colonizing all lakes and radi- ating into different ecotypes was also rejected by DOUGLAS (1999). In contrast, DOTTRENS (1959) recognized six species, but his findings mainly relied on gillraker numbers and were therefore not satisfactory. Based on age, growth, spawning period, body proportion, scale and gill raker number, SvARDS ON (1970) stated that there are only five coregonid species in Scandinavia and central Europe. Fur- thermore, HIMBERG & LEHTONEN (1995) discussed systematics andnmnenclature of north and west European coregonid fish on the base of 111orpho-ecological and protein characters and reduced them to three indigenous (C. albula, C. autumnalis, and C. lavaretus) and one introduced species (C. peled), smne of the1n with further subdivisions. Thus, all Swiss coregonids were still gathered under the name C. lavaretus (L.). Finally, KOTTELAT (1997) reviewed the systematics and nomenclature of the European freshwater fish using the phylogenetic species concept and postulated, based on taxonmnic principles, that most of the different coregonid forms should be treated as species. Therefore, he combined the "Ballen" of Lakes Lucerne, Zug, Se1npach, Baldegg and Hallwil (Figure 2.1) into Coregonus suidteri, the name FA TIO (1885) reserved for Lake Se1npach whitefish. Kottelat' s phylogenetic species concept is basically in concert with the results of DOUGLAS (1999). The species flock concept of DOUGLAS et al. (1999) argues that coregonid forms within a lake are 1nore closely related to one another genetically than are forms among lakes. Using rnicrosatellite markers to investigate in relationships between the Swiss coregonid forms, DOUGLAS (1999) unfortunately didn't test

47 2 Fish species diversity

1naterial of Lake Hallwil. Data of the Lake Sempach "Ballen" showed extreme differences from fast growing coregonid forms of large Swiss lakes (DOUGLAS, 1999), despite the enormous number introduced into Lake Se1npach (HEER, 1983). DOUGLAS (1999) found close relationship between "Ballen" from Lake Sempach, Lungern and Aegeri, lakes within 60 kn1 in the Swiss Midland (Figure 2.1). Lake Hallwil is sited within this area and the actual local whitefish fonn is thought to be conspecific with Lake Sempach whitefish (KOTTELAT, 1997). DOUGLAS (1999) hypothesized that Lake Se1npach "Ballen" remained indigenous despite stocking practice.

Our data confirm the findings of STEINMANN (1950b) for Lake Hallwil and Sem- pach, indicating one fast-growing coregonid forn1 in each lake. The coregonid fonns from the two lakes were extre1nely similar, even before the fish kill in 1961 (Table 2.2).

Table 2.2 Morphometric parameters of whitefish frmn Lakes Sempach and Hall- wil according to Steinmann (1950b) and data of 1997/98 (girth, eye diameter and pectoral fin length as percent of total fish length).

Steinmann Steinmann 199711998 199711998 (1950b) (1950b) (n=127) (n=127) L. Sempach L. Hallwil L. Hallwil L. Hallwil jparameter min - max nun - max min - max mean± sd girth 47.5 - 60 45.5 - 59.5 45.4 - 64.6 54 ± 3.9 eye diameter 2.75 4 3 - 4.25 3 - 4.1 3.5 ± 0.2 pectoral fin length 11.5 - 14 9 - 16.5 9.7 - 14.8 12.9 ± 0.7 number of scales 78 - 92 76 - 94 76 - 97 88 ± 4 - nmnber of gillrakers 26 - 38 23 - 39 I 27 - 39 33 ± 2

Nevertheless, some differences 1night have been expected for gillraker nun1ber, because TODD (1998) reported less gillrakers for artificially bred and reared coregonids, as cmnpared to naturally hatched ones. This does not appear to be the case here (Table 2.2). Therefore, we conclude that Lake Hallwil "Ballen" has preserved its phenotype, via the very closely related Lake Sempach "Ballen", in spite

48 2.6 Acknowledge1nents

of the severe population reduction or even extinction in 1961. DOUGLAS ( 1999) stated that local whitefish populations should be treated as 1nanagement units to preserve the diversity of central alpine coregonids. Because KOTTELAT (1997) found that Lake Se1npach "Ballen" is to be considered conspecific with Lake Hall- wil "Ballen", re-stocking of Lake Hallwil with Lake Se1npach whitefish after the fish kill of 1961 to restore whitefish population is regarded an adequate 1neasure. Although some original genetic resources have probably been lost in 1961, maintaining a closely related whitefish in Lake Hall wil represents an irnportant conservation act, apart from the economical importance of this whitefish population.

2.6 Acknowledgements

Thanks are due to H. Minder, Aarau, and Dr. C. Friedl, Bern, for detailed fisheries statistics and to my mother for her help on the sarnpling days. Commercial fisheries Delphin and Weber are acknowledged for providing whitefish. The constructive con11nents by Prof. J. V. Ward helped to i1nprove earlier drafts of this paper.

2. 7 References

BERG, L. S. (1932): Ubersicht der Verbreitung der Siisswasserfische Europas. Zoogeographica 1: 107-208.

BRUTSCHY, A. & GUNTERT, A. (1923): Gutachten iiber den RUckgang des Fischbestandes irn Hallwilersee. Arch. Hydrobiol. 14: 523-571.

DOTTRENS, E. (1959): Systematique des coregones de l'Europe occidentale, basee sur une etude biometrique. - Revue Suisse de Zoologie 66/1: 1-66.

DOUGLAS, M. (1999): Central alpine Coregonus (Teleostei, Coregonidae): Evolution and conservation of a unique assemblage. Dissertation an der Philosophischen FakulUit der Universitat Ziirich.

DOUGLAS, M. R., BRUNNER, P. C. & BERNATCHEZ, L. (1999): Do assemblages of Coregonus (Teleostei: Salmoniformcs) in the Central Alpine reg10n of Europe represent species flocks? - Molec. Ecol. 8:589-603.

49 2 Fish species diversity

PATIO, V. (1885): Les coregones de la Suisse (ferns diverses) classification et conditions de frai. - Receuil Zool. Suisse 1/2: 649-665.

FATIC\ V. (1890): Faune des vertebres de la Suisse. Histoire naturelle des poissons. - H. Georg (ed.) 5/2.

HEER, L. (1983): Der Sempachersee: Die Fischerei in friiheren Zeiten, sein heutiger Zustand und die Sanierungsmassnahmen. - Schriftenreihe Fischerei 41: 1-20.

HEIMANN, R. (1962). Schonvorscluiften fiir die Fischerei im Aargau. Schweiz. Fischerei-Zeitung 70/9: 267.

HIMBERG, M. K. J. & LEHTONEN, H. (1995): Systematics and nornenclature of coregonid fishes, particularly in Northwest Europe. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 39-47.

HOLTEY, R. (1996): Verbreitung, Wachstum und Ernaluung des Kaulbarsches ( Gyrnnocephalus cernua L.) in den GewLissern der Schweiz. - Diplomarbeit an der Universitat Konstanz.

JEAN-RICHARD, P. (1997): Krebse im Hallwilersee. - Pro natura Aargau. 19pp.

KLEE, 0. (1991): Angewandte Hydrobiologie. Trinkwasser, Abwasser, Gewasserschutz. 2nd edition. - Georg Thierne Verlag, Stuttgart, New York.

KOTTELAT, M. (1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the forn1er USSR), with an introduction for non-syste1natists and cmnments on nomenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

LAMPERT, W. & SOMMER, U. (1993): Limnookologie. -Thieme Verlag, Stuttgait

LINNAEUS, C. (1758): Systema naturae. I. edition X.

LUCZYNSKI, M., KUZMINSKI, H., DOBOSZ. S. & GORYCZKO, K. (1998). Gene pool characteristics of whitefish (Coregonus lavaretus) fingerlings produced in a hatchery for restoration stocking purposes. - Arch. Hydrobiol. Spec. Issues Advanc. UnmoL 50: 317-321.

50 2.7 References

MACDONALD, P. D. M. & PITCHER, T. J. (1979): Age-groups from size-frequency data: a versatile and efficient inethod of analysing distribution mixture. - J. Fish Res. Bd 36: 987-1001.

MlTLLER, R. (1992): Trophic state and its implications for natural reproduction of salmonid fish. - Hydrobiologia 243/244: 261-268.

MULLER, R., BIA, M. M. & MENG, H.J. (1994): Die Felchenfischerei in einigen Seen der Zentralschweiz und des Mittellandes. - Mitteilungen zur Fischerei 55. Bundesamt fiir Umwelt, Wald und Landschaft (BUW AL).

RUFLI, H. (1978): Die heutigen sy1npatrischen Felchenpopulationen ( Coregonus spp.) des Thuner- und Bielersees und ihre Morphologie. - Schweiz. Z. Hydrol. 40: 7-31.

STEINMANN, P. (1950a): Ein neues System der rnitteleuropaischen Coregonen. Revue Suisse de Zoologie 57/10-32: 517-525.

STEINMANN, P. (1950b): Monographie der schweizerischen Koregonen I. Teil. Schweiz. Z. Hydrol. 12.

STEINMANN, P. (l 951): Monographie der schweizerischen Koregonen II. Teil. - Schweiz. Z. Hydrol. 13.

SVARDSON, G. (1970): Significance of introgression in coregonid evolution. In: Lindsey, C. C. & Woods, C. S. (Eds). Biology of Coregonid Fishes. - University of Manitoba Press, Winnipeg. 33-59.

TODD, T. N. (1998): Environmental modification of gillraker number in corcgo- nine fishes. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 305-315.

VENTLING-SCHWANK, A. R. & MfJLLER, R. (1991): Survival of coregonid (Coregonus sp.) eggs in Lake Sempach, Switzerland. - Verb. Internat. Verein. Limnol. 24: 2451-2454.

WAGLER, E. (1941): Die Lachsartigen (Salmonidae). II. Teil: Coregonen. In: De1noll, R. and Maier H.N. (eds.): Handbuch der Binnenfischerei Mitteleuropas III. - Schweizerbart. Stuttgart.

WAGLER, E. (1950): Die Coregonen in den Seen des Voralpengebietes. XI. Herkunft und Einwanderung der Voralpencoregonen. - Veroffentl. Zool. Staatssam1nl. Miinchen, 113: 3-62.

51 2 Fish species diversity

52 3.1 Abstract

3 GROWTH AND SURVIVAL OF WHITEFISH LARVAE REARED ON DRY AND LIVE FOOD.

3.1 Abstract

Two commercial dry diets and live zooplankton were tested as initial food for larvae of Lake Hallwil whitefish ( Coregonus suidteri, n01nenclature according to KOTTELAT, 1997) in comparison with a reference food type (Arternia nauplii). Dry diet Trouvit had never been tested scientifically before. We further looked at the ability of the larvae to switch from dry diet to live zooplankton, an aspect often neglected in stocking artificially reared whitefish fry. After 3 weeks of feeding, the diets in all experimental tanks were changed to live zooplankton caught in Lake Hallwil and larvae were fed for another 3 weeks. Mean total length and dry weight of the Artenzia-prefed larvae at the end of the experi1nent were 18 nun and 3 ing at 5°C, and 30 nun and 23 mg at 13°C. Dry-diet prefed larvae were 15 mm and 1.5 mg at 5°C, and 25 mm and 12.5 mg at 13°C. Zooplankton-prefed larvae reached 17 mm. and 2.5 mg at 5°C, and 27 nu11 and 19 mg at 13°C. Overall mortality varied between 5 and 45%. Larvae fed live zooplankton suffered less mortality than those fed dry diets. Feeding condition during the first three weeks after hatching had an influence: After the switch to zooplankton diet at the encl of the third week, slower growth of Artemia and dry-diet prefed larvae was recorded. We conclude that the dry diets tested give satisfactory rearing results, but zooplankton is still the best diet for mass rearing of whitefish.

53 3 Growth and survival of whitefish larvae reared on dry and live food.

3.2 Introduction

Artificial breeding, rearing and stocking of whitefish larvae are conducted to increase year class strength (LESKELA et al., 1995) and yield (TURKOWSKI & BONAR, 1995), to sustain populations without natural reproduction (STEFFENS, 1995) or even to restore extinct populations (LUCZYNSKI et al., 1998). But often, the effectiveness of stocking whitefish larvae is neither predictable (SALOJARVI, 1992) nor convincing (STEFFENS, 1995) and depends on abiotic (LESKELA et al., 1995) or fish density dependent factors (LEHTONEN & NIEivfELA, 1998: SALONEN et al., 1998). Because stocked larvae usually originate from only few parental fish and are reared under non-natural conditions, changes in the gene pool (BRZUZAN et al., 1998; "MAMONTOV & YAKHNENKO, 1998) and the phenotype, e.g. gillraker number (TODD, 1998), tnay occur. Additionally, the coexistence of several coregonid forms in some lakes requires an adjusted stocking practice (LUCZYNSKI & RITTERBUSCH-NAUWERCK, 1995) to maintain specific genetic variability and to avoid cross breeding (JOKIKOKKO & HGHMARJ'\IEMI, 1998). Therefore, in lakes with high natural reproduction, fishing regulation is preferable to stocking for maintaining high and stable yield (SALONEN et al.. 1998; WANZENBOCK & JAG~ SCH, 1998). In eutrophic lakes, stocking whitefish larvae could also lead to changes in the food web and, thus, may delay the recovery of the lake or even lead to enhanced eutrophication (BERG et al., 1994).

Lake Hallwil (Switzerland) is a eutrophic lake, in which nearly all whitefish eggs die during en1bryonic development because of the anoxic sediment (VENTLING- SCHW ANK & MULLER, 1991; MOLLER, 1992 & 1993: VENTLING-SCHWA:'\K & LIVINGSTONE, 1994). Thanks to artificial breeding, prefeeding during six weeks and stocking, Lake Hallwil whitefish, a fast growing coregonid fonn, has been the dominant and c01mnercially most in1p01tant fish species in the lake since 1977 (MULLER et al., 1994). The three hatcheries at Lake Hallwil follow different rearing and stocking policies: In one hatchery. larvae are reared on dry food during the first few weeks, and on natural food when zooplankton concentration in the lake rises. In the second hatchery, larvae are fed zooplankton only, and in the third one all fish are stocked as unfed f1y.

In spite of the intensive stocking. the high yields are not stable. In 1992, 1993 and 1994 yield was very low. Year class strength of fish, and thus stock size is deter-

54 3.3 Materials & Methods mined during the early life stage (MAY, 1974; SALOJARVI, 1991). Several factors are known to influence survival and, thus, success of the coregonid fry stocked: rearing conditions, size at the time of stocking (FLDCHTER, 1980), the number of fish stocked (CHRISTIE, 1963, Ml'.hLER, 1990), competition, predation, diseases (SALOJARVI, 1987) and enviromnental conditions (ECKMANN et al., 1988; PONTON & MlJLLER, 1989). The ai1n of the present work was to investigate the effect of the rearing conditions and the diet types fed in Lake Hallwil hatcheries on growth and survival of the larvae. We further looked at the ability of the larvae to switch fron1 dry diet to live zooplankton. an aspect often neglected in stocking artificially reared whitefish fry.

3.3 Materials & Methods

3.3.1 Experimental design

The feeding experi111ents were conducted in a hatchery at Lake Hallwil. The whitefish larvae used originated frmn artificially incubated eggs of different parental Lake Hallwil whitefish. Therefore, they were 1nixed before assigned to the experin1ental groups. Twenty-four plastic tanks, each of 9 litres, were used as experi1nental units (Figure 3.1). Initial fish density was set at 30 newly hatched fry per litre, according to ROSCH & SEGNER (1990).

Water was supplied to the experi111ental units by a pmnp whose intake was in the lake at eight meters depth. For the experilnental group C (=cold) kept at 5°C, water was aerated with an air pmnp in a 30 litre overhead storage tank to avoid gas oversaturation. For the group H (=heated), water was first heated to the experimental te111perature of 13 °C in a separate 30 litre tank and then lead by gravity flow into a second 30 litre overhead storage tank for aerating. Larvae of the warm- water group were accli111ated to the experimental temperature of 13 °C for 2 days.

Fron1 both storage tanks, water flowed by gravity into the experi1nental tanks. Flow rate was regulated by a tube clamp to maintain a flow rate of 0.3 litres per minute per tank (ROSCH & SEGNER, 1990). The deposits were siphoned and the outflow tubes cleaned every second day to prevent toxicity problen1s (TAYLOR & FREEBERG, 1984) and to record losses ainong the larvae. Outflow tubes with

55 3 Growth and survival of whitefish larvae reared on dry and live food. plankton net strainers of 670 µm 1nesh size (phase 1) and of 1000 pm mesh size (phase 2) were used.

Phase 1 Phase 2

-Artemia -- Zooplankton dry diets

0 days

Figure 3.1 Experiinental design: Three replicate tanks per treatment, water volume of 9 litres per tank, 30 newly hatched larvae per litre (Phase 1: A = Artemia diet; = zooplankton diet; T = Trouvit diet; K = Kyowa diet; H =water heated to l3°C; =cold water of 5°C; Phase all groups: zooplankton diet) and daily food ration (below).

Light was provided from 6 white OSRAM 18 watt fluorescent tubes. One tube illuminated 4 experiinental tanks to provide a light intensity of 200 lux on the surface during 12 hours per day (ROSCH, 1992). On days 0, 8, 13~ 20, 27, 34 and 39, ten larvae per tank were removed to measure length (to the nearest 0.5 nun), wet weight (blotted, to the nem·est 0.1 mg) and dry weight (to the nearest 0.1 mg after

56 3.3 Materials & Methods drying to constant weight at 105°C during 24 hours). Additionally, at day 20 larval stage according to LUCZYNSKI et al. (1988) was determined.

Larvae were checked daily for normal behaviour, particularly while feeding. Abnormal behaviour was expected to occur during the change from dry food and Artemia diet to live zooplankton (pers. experience M. KELLER).

3.3.2 Diet types

Food was provided ad libitum beginning on the first day after hatching, designated day 0 of the experi1nent. Although satiation level is lower at low ten1perature than at high temperature (RC)SCH, 1986), the san1e food amounts were given to the 5° and l3°C groups. During the experi1nent, the daily ration was increased by increasing the number of fed portions according to RbSCH (1987).

During the first 3 weeks, designated phase 1 of the experiment, fish were fed different diets (Figure 3.1). Arternia represented the reference treatment of the experiment, because anArtemia diet gives the best survival and growth due to small size (ROSCH, 1989; ECKMANN, 1985) and essential fatty acid composition (WATANABE et al., 1983; FLUCHTER & REMBOLD, 1986). During the second 3 weeks, designated phase 2 of the experiment zooplankton was provided in all treatments (Figure 3 .1 ).

The caloric contents of newly hatched and dried Artemia nauplii and dried zooplankton (drying during 24 hours at 105 °C) were detennined using an oxygen bomb calorimeter. Those values were compared to the dry diet energetic content taken from the nutritional information given by the factories.

Artemia nauplii (A) were provided twice and with increasing daily ration three times a day. According to BROWN & TAYLOR (1992), an ad libitum feeding requires 50 Artemia per larva fed twice a day at 7°C. Initial daily ration was 0.6 g eggs (Figure 3.1) or nearly 60'000 Artemia nauplii per tank. Artemia eggs hatch after 24-48 hours at 25°C in water of 2% salinity. Only newly hatched Artemia nauplii were delivered.

Dry diets Trouvit Perla Marine A/B (T), a product manufactured by Hendrix SpA, Mozzecane (Italy), or FryFeedKyowaA (K), a product manufactured by Kyowa Hakko Kogyo Co., Ltd., Tokyo (Japan), were given in portions of 0.1 g between

57 3 Growth and survival of whitefish larvae reared on dry and live food.

9 a.m. and 7 p.m.. Initial daily ration was 0.6 g per tank (ROSCH & SEGNER, 1990) and gradually increased to 0.9 g by day 20 (Figure 3.1). As soon as the first larvae reached a length of l 6n11n, particle size was gradually changed frmn < 200 ~nn (Trouvit A) to 200-400 ~un (Trouvit B). Particle size of K was< 250 ~un all the time.

Zooplankton was collected daily from the open water of Lake Hal1wil by hori- zontal tows between 1 and 4 meters depth with two plankton nets of 200~Lm mesh size (Roni & GEIGER, 1968) and stored at 5°C in an aerated 200 litre tank. Daily ration, expressed as dry weight in Figure 3.1, was given in two equal portions during phase 1 and in three equal portions during phase 2. Due to cold weather, the lake plankton density decreased dramatically soon after the start of the experiment. Therefore, daily ration was low during the first two weeks and could only be increased from the third week on (Figure 3.1). Daily plankton san1ples were also preserved in 5% formalin. In subsamples of the days 1 and 4, the organis1ns were determined and counted using a counting chamber and a WILD binocular microscope with 25-50x magnification. The results were expressed using the percentage composition by number method (HYSLOP, 1980).

3.3.3 Statistics

Analyses of variance at p<0.05, one-way ANOVA (LOZAN, 1992), were made on a mean length base to investigate differences between the experimental groups. Specific growth rates [%/day] were calculated on a n1ean dry weight base according to the method of WINBERG (1956). Cumulative mortality [%] was computed (Figure 3.3) according to the formula of SCHMID (1993), taking into account sainpling reduction according to ORTLEPP (1984).

For an overall estimation of the rearing methods (diet type and temperature), we rnade a classification on the base of literature data about predation-secure size and rearing mortality: HOAGMAN (1974) found that at 17-22 n11n length, the larvae were mostly secure frmnpredation. According to FLlICHTER (1980), effectiveness of stocking is better when fish are 20-30 nun long at the time of stocking. ROSCH (1988) rated the efficiency of a dry diet as satisfactory when 50-80% of the larvae survived the first 35-45 days after hatching. This allows a rough estimation of the rearing method as done in Figure 3.4.

58 3.4 Results

3.4 Results

3.4.1 Quality of diets

Energetic value of the dry diets T and K were 22.7 kJ/g and 23.5 kJ/g, respectively. Reference diet A (Artemia) and test diet Z (Zooplankton), both live food organisms, had energetic values of 21.4 kJ/g and 21.5 kJ/g, respectively.

In the zooplankton samples of day 1 and 4, about 80% of the organisms were copepods (Crustacea: Copepoda): adults and juveniles at copepodite stage. About 20% of the organisms were juvenile or adult cladocerans (Crustacea: Phyllopoda). No rotifers (Nemathelminthes: Rotatoria) or copepod nauplii (Crustacea: Copepoda) could be found.

3.4.2 Feeding behaviour

Larvae fed live diets pursued their prey actively, while those fed dry diets seemed to move around randomly and eat what happened to drift in front of their 1nouth.

The first days after the switch to zooplankton on day 21, not only the fish previously fed dry diets, but also those reared on newly hatched Artemia nauplii bent their bodies after ingesting the zooplankton. Within few days, this strange behaviour gradually disappeared. Larvae fed zooplankton already during phase 1 showed no special behaviour.

3.4.3 Growth

At 13 °C, the larvae grew significantly faster than the fish held at 5°C (Figure 3.2).

Negative growth rates in the heated water groups could be seen during the first week only, whereas fish in some of the tanks with cold water had a negative growth rate even at the end of the experiment (Table 3.2). In phase 1, statistica1ly significant differences in length were found between the reference group fed Artemia and the other 3 groups (Table 3 .2)

59 3 Growth and survival of whitefish larvae reared on dry and live food.

35 30

15 10 35 30

25 -o-TC1 20

15 10 0 10 20 30 40 0 10 20 30 40 days

Figure 3.2 Growth of larvae as mean value and standard deviation of each experimental group. Abbreviations as in Figure 3.1.

At l3°C. differences occurred between ZH and dry-diet fed fish. And at the end of phase 1, KH larvae were significantly longer than TH larvae. In phase 2, there were no significant differences between AH and ZH groups, while differences between zooplankton fed fish and dry-diet groups occurred for both temperature regimes (Table 3.2). At the end of phase 2. TH and KH groups could catch up with ZH fish, and TH larvae even with AH group.

To have a reliable parameter to estimate fish development speed, larval according to LUCZYNSKI et al. ( 1988) was determined. In accordance with KIRCH- HOFER & LINDT-KIRCHHOFER (1998), AH larvae sampled on day 20 had reached body sizes around 20 mn1, were at larval stage 6 and had the swim bladder filled. Larvae frmn all the other groups sampled on day 20 were inainly at stage 2-4, and no filled swim bladder was found.

60 3.4 Results

Table 3.1 Differences in length between the experi1nental groups (one-way- ANOVA; black square= p<0.05). Phase 1: A: Artemia; Z: zooplankton; T: Trouvit; K: Kyowa; Phase 2: A, Z, T and K: zooplankton.

heated water (13°C) cold water (S°C)

phase 1 phase 2 pairs compared phase 1 phase 2 A vs Z AvsT AvsK ZvsT Zvs K Tvs K

3.4.4 Survival

Results frmn the three replicates are given as inean value and standard deviation for every treat1nent group. Most of the losses occurred during the first 25 days (Figure 3.3). In the AH tanks, the highest losses were recorded earlier than in the other groups.

Overall inortality ainong larvae fed zooplankton was 10% at both temperatures. About 6% mortality occurred in the AC group, but the AH group showed the highest losses of all groups, up to 45%. Larvae fed dry diet T suffered 40% inortality at 13° and only 27% at 5°C. Dry diet Kat both temperature regirnes yielded nearly the same proportion of dead larvae, about 25 %

61 3 Growth and survival of whitefish larvae reared on dry and live food.

-AH -AC 50 •••••ZH -----zc ...... r...... ··············TC KH ··KC (ij t:: 0 E 30 Q)

.::::+-' (ij ::J 20 ...... -...... E ::J l) 10 ......

5 10 15 20 25 30 35 40 days

'Figure 3.3 Cunrnlative mortality as mean value and standard deviation of the three replicates of each experilnental group. Abbreviations as in Figure 3.1.

After the switch to zooplankton as the only diet for all groups, mortality in the groups fed dry diets nearly ceased within half a week.

3.5 Discussion

3.5.1 Quality of diets

The energetic values of all four diets did not differ very much fr01n each other, and our results for zooplankton agree with the findings by GIUSSANI & DE BERNARDI (1977). Therefore, we estimate that pararneters other than the energetic value were responsible for the different growth rates (Table 3.2 and Figure 3.2).

In the case of live diet. species con1position is important. The reference diet

62 3.5 Discussion

(Artemia) was a single species and one size food ite1n (FLtJCHTER, 1980), but zooplankton (Z) included copepods and cladocerans of different sizes. HOAGMAN (1973) and NIESSLBECK (1987) found that fish larvae do not eat all the plankton species, but prefer small and non-evasive prey types (MOOKERJI & RAO, 1993), they select the rnost abundant species, even if they would be able to eat bigger prey types (PONTON & MtJLLER, 1990), and they are gape-limited (MOOKERJI & RAO, 1994). Small larvae prey upon rotifers, copepod nauplii (AMMANN & STEINMANN, 1948; HOAGMAN, 1973; MOOKERJI & RAO, 1993, 1994 & 1995) and small fonns of Daphnia spp.(FLOCHTER, 1980). In our zoop1ankton sarnples, neither rotifers nor copepod nauplii could be found, what n1ight have caused a lack of opti1nal food for the larvae and, thus, slow growth in phase 1 (Table 3.2 and Figure 3.2). Due to the fact that mortality \Vas very low ainong zooplankton fed larvae (Figure 3.3), we conclude that whitefish larvae were able to prey, even at first feeding, on a nlixture of zooplankton caught in nets of 200 µin mesh size. Accord- ing to HARTMANN & KLEIN (1993) and the results in chapter 4, only about 30(Yo of the fed zooplankton mixture are edible for Lake Hallwil whitefish larvae at first feeding. For older larvae, cladocerans seemed to be the preferred diet: NAESJE et al. (1986) found a positive c011elation between the density of cladocerans and year class strength. According to HARTMANN & KLEIN (1993), big speci1nens of Daph- nia spp. would be edible for the larvae when they had reached a body length of about 35111111. If the zooplankton diet is composed of copepods only, larvae might not complete metamorphosis and die (FLtJCHTER, 1980). This may result in a weak year class.

3.5.2 Feeding behaviour

Feeding preferences change as the larvae pass through ontogenetic changes. They gradually switch to adult copepods and cladocerans. Changes in parameters that affect foraging ability make this possible: bigger mouth size, better developed digestive system, better visual and swim1ning capacity (BLAXTER, 1986; MOOK- ERJI & RAO, 1994). In addition to ontogenetic changes in larval feeding prefer·- ences, experience seemed also to be important. The observed bending behaviour of larvae previously fed dry diets and Artemia at first contact with live zooplankton suggests that preying on zooplankton 1nay require an adaptation phase. Slower growth than in zooplankton prefed groups is the consequence of this adaptation phase (Table 3.2 and Table 3.2)

63 3 Growth and survival of whitefish larvae reared on dry and live food.

Table 3.2 Growth rates [% dry weight per day] of larvae for every week in columns 2-7, for phase 1 and 2 in columns 8 and 9, and for the whole experiment in colmru1 10 (Phase 1: A =A rtemia diet; Z = zooplankton diet; T = Trouvit diet; K = Kyowa diet; H = 13°C; C = S°C; Phase 2: all groups: zooplankton diet).

days 0-8 8-13 13-20 20-27 27-34 34-39 0-20 20-39 0-39

AHl 4.3 10.3 9.2 7.0 13.l 11.1 7.5 10.3 8.9 AH2 5.3 9.2 9.3 10.9 5.8 9.6 7.7 8.7 8.2 AH3 2.5 5.3 11.7 6.6 11.0 12.5 6.4 9.8 8.0 ACl -0.9 4.0 5.6 3.3 4.6 5.7 2.6 4.4 3.5 AC2 -0.3 3.7 5.2 2.1 4.6 2.0 2.6 3.0 2.8 ' AC3 -0.1 5.8 2.8 3.2 4.9 4.0 2.4 4.0 3.2 ZHl 0.3 6.8 8.2 10.8 11.6 11.0 4.7 11.1 7.8 ZH2 -3.3 14.2 4.4 16.9 11.4 11.9 3.8 13.6 8.5 ZH3 -0.4 1.6 12.9 10.2 12.0 6.5 4.8 9.9 7.3 ZCl -5.1 1.7 2.7 5.2 3.7 10.8 -0.7 6.1 2.6 ZC2 -4.0 6.0 -1.2 5.4 7.5 6.1 -0.6 6.3 2.8 ZC3 I -4.0 -0.3 4.5 6.8 3.0 4.5 -0.l 4.8 2.3 THI -0.6 0.0 5.4 9.6 17.3 17.8 1.7 14.6 8.0 TH2 -2.7 -1.2 11.2 11.3 6.6 15.8 2.6 10.8 6.6 TH3 -2.2 0.0 9.6 12.9 11.3 13.4 2.5 12.4 7.3 TCl -1. l -1.6 l.5 0.3 4.7 5.8 -0.3 3.4 l.5 TC2 -3.3 2.0 -0.6 0.2 7.1 -0.2 -1.0 2.6 0.8 TC3 I -1.5 -2.8 0.6 6.8 3.8 3.6 -1.l 4.9 1.8 KI-Il I -4.2 10.0 7.8 3.5 14.2 10.l 3.5 9.2 6.3 KH2 -1.9 4.5 9.4 0.5 18.2 9.3 3.7 9.4 6.4 KH3 -2.5 7.1 7.1 5.7 12.7 2.8 3.2 7.5 5.3 KCl -3.3 1.7 -0.8 3.6 4.6 5.9 -1.2 4.6 1.6 KC2 -1.7 -l.6 1.4 6.3 -0.4 3.0 I -0.6 3.0 KC3 -0.6 -4.3 3.2 3.9 3.3 l.1 -0.2 3.0 I I I :~I

64 3 .5 Discussion

3.5.3 Growth

The efficiency of utilisation of the different diets and the resulting gain in body length and weight was not unifonn (Table 3.2,Table 3.2 and Figure 3.2). Accord- ing to CHAMPIGNEULLE & ROJAS-BELTRAN ( 1990), the chosen larval density of 30 larvae per litre is well below the upper limit of 200 larvae per litre for opti1nal rearing. Therefore, larval density should not have influenced growth in our experiment, but te1nperature did (Table 3.2, Figure 3.2). As reported by REY & ECK- MANN (1989), higher ten1perature caused accelerated growth. At low temperature, fish metabolis1n could not allow fast growth even on opti1nal diet. Therefore, sig- njficant differences due to the feeding regime n1ostly were found later at 5°C than at 13°C (Table 3.2). Significant differences already occuned from the beginning on at l3°C (Table 3.2).

According to STEINHART & ECKMANN (1992), who reared larvae under food-limited and non-limited conditions. we conclude that our Artenzia-fed larvae were reared under non-limited conditions. Experiinental group ZH shows a peculiar pattern: Differences between ZH and AH larvae were significant in phase 1, but not in phase 2 (Table 3.2). According to the experin1ents of ROSCH & DABROWSKI ( 1986) that showed opti1nal growth of zooplankton fed fish frmn the beginning on, we conclude that ZH and ZC fish were reared under food-limited condition in phase 1 (Figure 3.1 ). In phase 2, according to HARTMANN & KLEIN (1993) and own zooplankton length measurements (chapter 5), about 60 to 100% of the zooplankton organisn1s were edible. Therefore, optimal growth of the ZH and ZC larvae was possible in phase 2 (Table 3.2 and Table 3.2). Because of the strange behaviour and slow growth, we conclude that Artemia prefed fish needed an adaptation phase after the switch to zooplankton diet in phase 2 (Table 3 .2 and Table 3.2). As a consequence of the adaptation phase, zooplankton prefed larvae could catch up with Artemia prefed fish (Table 3.2).

As reported by KOSKELA (1988), fish fed dry diets showed very slow growth during the first few weeks. Between dry diet groups, statistically significant differences were found only at the end of phase 1at13°C (Table 3.2): KH fish were significantly longer than TH larvae. Statistically significant differences between the Arternia and the dry diet groups (Table 3.2) may be explained by lower food intake of dry-diet fed fish due to a lack of aromatic attraction (FLtJCHTER, 1980)

65 3 Growth and survival of whitefish larvae reared on dry and live food. and absent prey inovement (chapter 4). In phase l, differences between ZH group and dry-diet fed fish were son1etimes significant, smnetimes not (Table 3.2). High variation in standard deviation of the ZH group (Figure 3.2) and the selective 1nor- tality of smaller individuals, that apparently accelerated the growth rate of the remaining dry-diet fed fish (chapter 4), could be the cause. Thus, the not significant differences between AH and TH resp. ZH and TH/KH (Table 3.2) may be explained.

Fast larval growth is vital. According to PURCELL et al. (1987) s1nall fish larvae are Jnore vulnerable to predation than large ones. Inadequate food supply and low food density result in slow growth, thereby increasing the duration of larval expo- sure to predation (HOAGMAN, 1974, TAYLOR & FREEBERG, 1984). Therefore, the rearing method in hatcheries may directly influence larval growth and survival (lJRBAN-JEZIERSKA & BIERNACKI, 1992). Thus, larvae fed zooplankton and Arternia have a better chance to reach a predation-secure size before the time of stocking into the lake than do fish fed dry diets (Figure 3.4).

3.5.4 Survival

While temperatures above l2°C allow faster growth, the risk of diseases also increases (KOSKELA & ESKELINEN, 1992: ENZMANN et al., 1993). This may be the cause for different survival of TH and TC resp. AH and AC larvae (Figure 3.3). According to the results of ROSCH & APPELBAUM (1985) and ECKMANN (1985), high mortality occurred suddenly and within a short period (Figure 3.3). Survival of the ZH and ZC group was up to 90%, similar to the results by BROWN & TAYLOR (1992) and KOHO et al. (1991). ECKMANN (1985) concluded that 10% m011ality after 26 days means that the larvae were reared on optimal food.

With respect to Lake Constance. ECKMANN (1985 & 1987), ECKMANN et al. (1986) and ROSCH ( 1994) hypothesized that there could be chemical con1ponents called «stressors» in plankton organisms like Cyclops spp. or rotifers, or a deficiency in some essential lipoproteins. This could be the cause of up to 1ooq;1 inortality (ROSCH, 1994). No high mortality due to possible «stressors» could be found for Lake Hallwil in 1998 (Figure 3.3). This pheno111enon has also not been rep011ed earlier (pers. connn. E. FISCHER and M. KELLER).

66 3.5 Discussion

The phenomenon of high losses in the reference group AH could be the result of a bacterial intestinal infection due to feeding with nauplii of the Artemia strain Great Salt Lake/Utah (BURKHARDT-HOLM et al., 1989; LOEWE & ECKMANN, 1988). The fact that mortality of the AH larvae started and ceased earlier than in the other experiinental groups (Figure 3.3) agrees with the results of LOEWE & ECKMANN (1988) and sustains the bacterial disease hypothesis. The fact that the AC group did not suffer from the san1e proble1ns could be due to the lower water te1nperature that limited bacterial growth.

The test groups Kand T, both dry diets, also showed high mortality. R(>SCH (1988) rated the efficiency of a dry diet satisfactory when 50-80% of the larvae survived the first 35-45 days after hatching. Rearing mortality with dry diets T and K was in the saine range in our experiinent (Figure 3.3), and was thus considered to be satisfactory. TAYLOR & FREEBERG (1984) found that starving larvae suffer 50% 1nortality at 12°C after 3 weeks. In our experiments, mortality was below 50%. Therefore, dry diets tested have been ingested and digested by the surviving larvae.

Mortality in a hatchery occurs when the larvae are not able to ingest the given diet, or they 1nay not have fed due to illness (ROTH & GEIGER, 1968), or the larvae did not start feeding (ROSCH, 1989 & 1992). The large particle size of diet Tin particular may be the cause of high losses at 13 °C (Figure 3 .3) and the very low growth rates at the beginning of phase 1 (Table 3.2). Diet K had a finer consis- tency and could therefore also be ingested by small larvae (pers. comm. M. M. BIA), thus providing nourishment and reducing mortality. At higher te1nperature, the start of feeding is earlier (ROTH & GEIGER, 1968) and the food amount ingested per fish is higher than at low temperature (STEINHART & ECKMANN, 1992). Therefore, it appears that the dry diet K provides better nourishment for srnall whitefish larvae at 13°C than diet T (Table 3.2 and Figure 3.2). Neverthe- less, TH larvae showed in phase 2 higher growth rates than KH larvae (Table 3.2) and could catch up with AH fish (Table 3.2). Therefore, diet T may have a positive effect for adaptation to natural food.

By providing only zooplankton to all groups fr01n day 21 onward, we could investigate how the larvae reacted to diet change. This change of diet occurs regularly when the larvae are stocked in the lake. Mortality ceased within half a week.

67 3 Growth and survival of whitefish larvae reared on dry and live food.

Therefore, prefeeding condition does not seem to have a direct influence on survival of the larvae in the lake, but an indirect influence via growth according to HOAGMAN (1974) and FU)CHTER (1980) as shown in Figure 3.4.

,...... , 100 :.:?. .__,0 80 m ·::;> 50 '-- :::J !/) 0 30 'E .__,E 22 .c...... CTl c 17 QJ 10

Figure 3.4 Estin1ation of the rearing methods after six weeks: Survival (80- 100% is opti1nal, 50-80% satisfactory and 0-50% very bad) and length (>221m11, larvae have predation-secure size; 17-22m1n, most of the larvae have predation-secure size; <17mm, larvae are not secure from predation). Abbreviations as in Figure 3.1.

3.5.5 Summary & Outlook

ROSCH & DABROWSKI (1986) reported different success of dry diets with different fonns of coregonids. Actually, diet K is reported to be the best dry diet for mass rearing of coregonid larvae (pers. c01n1n. R. ROSCH). Diet Thad not been tested scientifically on coregonids before. Growth and survival of Lake Hallwil coregonids reared on both dry diets tested was found to be satisfactory (Figure 3.4). Therefore, the feasibility of rearing coregonid larvae with dry food and the adapt- ability of dry diet prefed larvae to zooplankton diet were demonstrated. At the beginning, slower growth and higher inmiality than under opti1nal conditions are the consequences. After the switch to zooplankton as it happens when the larvae are stocked into the lake, a short adaptation phase is needed, but no losses due to diet change were recorded. FLtJCHTER (1980) concluded that the lack of aromatic attraction was responsible for the lower acceptance of artificial diet relative to that

68 3.6 Acknowledgements of natural food. A dry diet with the taste of Artemia might further i1nprove rearing results.

Natural food appears to be the optimal food for larvae of Coregonus sp. (Figure 3.4). But due to low zooplankton densities in spring and the collecting method using plankton nets of a mesh size of 200 µ1n, it is hardly possible to provide an ad libitum feeding and optimal growth frmn the beginning on. The collecting method could be opti1nized: Smaller mesh size would make available rotifers and small copepod nauplii. Alternatives could be feeding Arten1ia nauplii, dry diet or frozen zooplankton as reported by KLEIFELD-KRIEBITZ & ROSCH (1987) during the first three weeks after hatching to overcome zooplankton shortage. Another possible alternative is rearing the larvae in their natural environment using illmni- nated net cages (ROSCH & ECKMANN, 1986), successfully done in Lake Sempach, a lake within 15km of Lake Hallwil and of comparable trophic state.

Further research on the topic of diet switch may be of interest: New nrnrking techniques with fluorochrmns (ECKMANN et al., 1998: ROJAS-BELTRAN et al., 1998) allow now the mass marking of eyed eggs or even larvae. Therefore, survival and growth of larvae reared until different stages with different foods may be investigated after stocking in a natural or semi-natural enviromnent.

3.6 Acknowledgements

Thanks are due to Prof. R. Ecklnann, Konstanz, Dr. A. Kirchhofer and Prof. W. Meier, Bern, Dr. N. Mooke1ji, Montreal, and Dr. R. Rl)sch, Langenargen, for giving most valuable advice concerning the rearing inethods. Kyowa Hakko Kogyo Co., Tokyo, Fannix AG, Malters, and Hendrix SpA, Mozzecane, provided free sainples of the dry diets used. Special thanks to U. Fischer, President of Sport- fischerverein Hallwilersee, for authorizing conduct of the experiment at the hatchery facilities, and to M. Keller for his great interest and active help during the experiments. Special thanks also to Dr. E. Meister, ETH Ziirich, for providing an oxygen bomb calorimeter and for his help during the experiments. The help and support by iny parents and many colleagues at EAWAG, especially M. M. Bia, Dr. M. Brunke, Dr. H. R. Biirgi, B. Germann, Dr. M. Gessner, E. Grieder, C. Heller, Dr. H.J. Meng, Dr. A. Peter, A. Steffen, D. Steiner and A. Zwyssig is greatly acknowledged. The constructive c01nn1ents by Prof. J. V. Ward,

69 3 Growth and survival of whitefish larvae reared on dry and live food.

Ditbendorf, Dr. R. Rosch, Langenargen, Dr. I. J. Winfield, Windermere, and an anonymous reviewer helped to improve earlier drafts of this paper.

3. 7 References

AMMANN, E. & STEINMANN, P. (1948): Die Verbesserung der Methoden in der Felchenzucht. Kommission fUr die Erforschung fischereiwirtschaftlicher Fragen. Sonderbericht Nr. 1. - W. Kunz, Pfaffikon-Zilrich.

BERG, S., JEPPESEN, E., SONDERGAARD, M. & MORTENSEN, E. (1994). Environmental effects of introducing whitefish, Coregonus lavaretus in Lake Ring. - Hydrobiologia 275-276/0: 71-79.

BLAXTER, J. H. S. (1986): Developtnent of sense organs and behaviour of teleost larvae with special reference to feeding and predator avoidance. - C. Hubbs (Ed.), Proceedings, Ninth Larval Fish Conference, Texas. - Trans. Am. Fish. Soc. 115: 98-114.

BROWN, R. W. & TAYLOR, W.W. (1992): Effects of egg composition and prey density on the larval growth and survival of lake whitefish ( Coregonus clupeafonnis Mitchill). - J. Fish Biol. 40: 381-394.

BRZUZAN, P., YAKHNENKO, V. M., MAMONTOV, A.M., MARKOWSKA, A. & TROFIMOVA, I. N. (1998). Mitochondrial DNA variation in whitefish Coregonus lavaretus from Lake Baikal as revealed by restriction analysis. Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 357-362. BURKHARDT-HOLM, P., ECKMANN, R. & STORCH, v. (1989): Schadigung des Darmepithels von Coregonenlarven (Coregonus fera)

CHAMPIGNEULLE, A. & ROJAS-BELTRAN, R. (1990): First attempts to optirnize the mass rearing of whitefish ( Coregonus lavaretus L.) larvae frorn Leman and Bourget lakes (France) in tanks and cages. - Aquat. Living Resour. 3: 217-

')')~ ...... 8

70 3. 7 References

CHRISTIE, W. J. (1963): Effects of artificial propagation and the weather on recruitment in the Lake Ontario whitefish fishery. - J. Fish. Res. Board Can. 20: 597-646.

ECKMANN, R. (1985): Histopathological alterations in the intestine of whitefish (Coregonus sp.) larvae reared on zooplankton frorn Lake Constance (West Gerniany). - Diseases of Aquatic Organisn1s 1/1: 11-18.

ECKlYIANN, R. (1987): Pathological changes in the 111idgut epithelium of grayling,Thymallus thyrnallus L., larvae reared on different kinds of food, and their relation to mortality and growth. - Journal of Fish Diseases 10: 91 99.

ECKMANN, R., CZERKIES, P., HELMS, C. & KLEIBS, K. (1998): Evaluating the effectiveness of stocking vendace (Coregonus albula L.) eleutheroembryos by alizarin marking of otoliths. - Arch. Hydrobiol. Spec. Issues Advanc. L.imnol. 50: 457-463.

ECKMANN, R., GAEDKE, U. & WETZLAR, H. J. (1988): Effects of clirnatic and density-dependent factors on year-class strength of Coregonus lavaretus in Lake Constance. - Can. J. Fish. Aquat. Sci. 45/5: 1088-1093.

ECKMANN, R., ROSCH, R., ORTLEPP, J. & KLEIFELD, G. (1986): Survival and growth of coregonid larvae from Lake Constance fed on zooplankton of different origin. - Arch. Hydrobiol. Spec. Issues Advanc. Li1m101. 22: 203- 214.

ENZMANN, P. J., KONRAD, M. & PAREY, K. (1993): VHS [viral haernonhagic septicaemia] in wild living fish and experimental transrnission of the virus. - Fisheries Research (Netherlands) 1711-2: 153-161.

FLDCHTER, J. (1980): Review of the present knowledge of rearing whitefish (Coregonidae) larvae. - Aquaculture 19: 191-208.

FLDCHTER, J. & REMBOLD, H. (1986): Soluble factor essential for 1netamorphosis of coregonid larvae has been partially purified from Arteniia salina. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 197-202.

GIUSSANI, G. & DE BERNARDI, R. (1977): Food selectivity in Coregonus sp. [zooplanktophagous fish] of Lake Maggiore [Italy], an energetical

71 3 Growth and survival of whitefish larvae reared on dry and live food.

approach. - Memorie dell'Istituto Italiano di Idrobiologia, Dott. Marco de Marchi Verbania Pallanza (Italy) 34: 121-130. HARTMANN, F. & KLEIN, M. (1993): Nahrungsselektion von Renkenbrut ( Coregonus lavaretus) unter Aufzuchtbedingungen. Fischerei und Teichwirt 44/8: 279-283.

HOAGMAN, W. J. (1973): The hatching, distribution, abundance, growth and food of the larval whitefish (Coregonus clupeqf(Jnnis Mitchill) of Central Green Bay, Lake Michigan. - Report Institute of Freshwater Research (Drottningholm) 53: 1-20.

HOAGMAN, W. J. (1974): Feeding by alewives (Alosa pseudoharengus) on larval lake whitefish (Coregonus clupeqfonnis) in the laboratory. - J. Fish. Res. Board of Canada 31/2: 229-230.

HYSLOP, E. J. (1980): Stomach contents analysis - a review of methods and their application. - J. Fish. Biol. 17: 411-429.

KLEIFELD-KRIEBITZ, G. & ROSCH, R. (1987): A simple method of feeding coregonid larvae on frozen zooplankton. - Journal of Applied Ichthyology 3: 119-124.

KOHO, J., KARJALAINEN, J. & VILJANEN, M. (1991): Effects often1peratures, food density and tin1e of hatching on growth, survival and feeding of vendace (Coregonus albula L.) larvae. - Aqua Fennica 21/l: 63-73.

KOSKELA, J. (1988): Rearing of larvae of the whitefish Coregonus nzuksun Pallas with artificial diets. - Finn. Fish. Res. 9: 333-337.

72 3. 7 References

KOSKELA, J. & ESKELINEN, U. (1992): Growth of larval european whitefish ( Coregonus lavaretus) at different temperatures. - Pol. Arch. Hydro biol. 39/3-4: 677-682.

KOTTELAT, M. ( 1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the fonner USSR), with an introduction for non-systematists and comrnents on nomenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

LEHTONEN, H. & NIEMELA, E. (1998): Growth and population structure of whitefish (Coregonus lavaretus L.) in mountain lakes of northern Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 81-95.

LESKELA, A., RUDD, R., LEHTONEN, H. & SANDSTROM, 0. (1995): Abiotic factors, whitefish stockings and relative year-class strength of anadromous whitefish (Coregonus lavaretus L.) spawning populations in the Gulf of Bothnia. - Arch. Hydrobiol. Spec. Issues Advanc. Li1m101. 46: 241-248.

LOEWE, H. & ECKMANN, R. (1988): The ontogeny of the alimentary tract of coregonid larvae, normal development. - J. Fish Biol. 33, 841-850.

LOZAN, J. L. (1992): Angewandte Statistik for Naturwissenschaftler. - Paul Parey, Berlin/Hmnburg.

LUCZYNSKI, M., FALKOWSKI, S. & KOPECK!, T. (1988): Larval development in four coregonid species (Coregonus albula, C. lavaretus, C. muksun and C. peled). -Finn. Fish. Res. 9: 61-69.

LUCZYNSKI, M., KUZMINSKI, H., DOBOSZ, S. & GORYCZKO, K. (1998): Gene pool characteristics of whitefish ( Coregonus lavaretus) fingerlings produced in a hatchery for restoration stocking purposes. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 317-321

LUCZYNSKI, M. & RITTERBUSCH-NAUWERCK, B. (1995): Biochemical genetic study of European whitefish (Co reg onus lavaretus) in Lake Mondsee, Upper Austria. - Aquatic Sciences 57/2: 127-135.

MAMONTOV, A. M. & YAKHNENKO, V. M. (1998). Isoenzyme polymorphis1n in Lake Baikal 01nul ( Coregonus auturnnalis rnigratorius GEORGI). - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 375-381.

73 3 Growth and survival of whitefish larvae reared on dry and live food.

MAY, R. C. (1974): Larval mortality in marine fishes and the critical period concept: 3-19. The Early Life History of Fish. - Blaxter J. H. S. (Ed.), Springer Verlag, Berlin.

MOOKERJI, N. & RAO, T. R. (1993): Patterns of prey selection in rohu (Labeo rohita) and singhi (Heteropneustes fossilis) larvae under light and dark condition. - Aquaculture 118: 85-104.

MOOKERJI, N. & RAO, T. R. (1995): Prey capture success, feeding frequency and daily food intake rates in rohu, Labeo rohita (Han1.), and singhi, Heteropneustes fossilis (Bloch), larvae. - Journal of Applied Ichthyology 11: 37-49.

MULLER, R. (1990): Management practices for lake fisheries in Switzerland. 477- 492. Management of freshwater fisheries. - Van Densen, W. L. T.; Steinmetz, B.; Hughes, R.H., (Eds), Pudoc. Wageningen.

MULLER, R. (1992): Trophic state and its implications for natural reproduction of salmonid fish. - Hydrobiologia 243/244: 261-268.

MULLER, R. (1993): Einige fischereibiologische Aspekte von Seesanierungen. - Fortschr. Fisch.wiss. 11: 43-56.

MULLER, R., BIA, M. M. & MENG, H.J. (1994): Die Felchenfischerei in einigen Seen der Zentralschweiz und des Mittellandes. Mitteilungen zur Fischerei Nr. 55. - Bundesarnt fur Umwelt, Wald und Landschaft (BUWAL).

NAESJE, T. F., SANDLUND, 0. T. & JONSSON, B. (1986): Habitat use and growth of age-0 whitefish, Coregonus lavaretus, and cisco, Coregonus albula. - Environmental Biology of Fishes 15/4: 309-314.

NIESSLBECK, P. (1987): Aufwuchsbedingungen fur natiirlich geschliipfte und ki.instlich erbrlitete Larven von Coregonus lavaretus iin Starnberger See (Oberbayern) . - Forschungsbericht aus der Bayrischen Landesanstalt fi.ir Fischerei in Starnberg.

74 3.7 References

ORTLEPP, J. (1984): Beobachtungen zurn Verlauf des Wachstums und der Sterblichkeit von Coregonenlarven bei Fi.itterung mit Zooplankton. - Diplomarbeit Biologic, Universitat Freiburg.

PONTON, D. & MULLER, R. (1989): Aliinentation et facteurs de rnortalite des larves de coregones (Coregonus sp.). Exemple de deux lacs de niveaux trophiques differents, les lacs de Sarnen et de Hallwil (Suisse Centrale). - Aquatic Sciences 51/1: 67-83.

PONTON, D. & MULLER, R. (1990): Size of prey ingested by whitefish, Coregonus sp., larvae, Are Coregonus larvae gape-limited predators? - J. Fish Biol. 36/l: 67-72.

PURCELL, J.E., SIFERD, T. D. & MARLIAVE, J.B. (1987): Vulnerability of larval herring ( Clupea harengus pallasi) to capture by the jellyfish Aequorea victoria. - Marine Biology 94: 157-162.

REY, P. & ECKMANN, R. (1989): The influence of lake ternperature on growth of Coregonus lavaretus L. larvae. - Arch. Hydrobiol. 116/2: 181-190.

ROJAS-BELTRAN, R., RUI-ILE, C., KUGLER, M. & GALAN, S. (1998): Preliminary study a mass-marking method for coregonid eggs. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 465-4 70.

ROSCH, R. (1986): Satiation of juveniles of Coregonus lavaretus L. fed on Daphnia pulicaria. - Schweiz. Z. Hydrol. 48/l: 44-52.

ROSCH, R. ( 1987): Effect of experimental conditions on the stomach evacuation of Coregonus lavaretus L. - J. Fish Biol. 30: 521-531.

ROSCH, R. (1988): Mass rearing of Coregonus lavaretus larvae on a dry diet. - Finn. Fish. Res. 9: 345-351.

RoSCII, R. (1989): Beginning of food intake and subsequent growth of larvae of Coregonus lavaretus L. - Pol. Arch. Hydrobiol. 36/4: 4 7 5-484.

ROSCH, R. (1992): Food intake and growth of larvae of Coregonus lavaretus L., effect of diet and light. - Pol. Arch. Hydrobiol. 39: 671-676.

ROSCH, R. (1994): Probleme bei der Aufzucht von Felchenlarvaen (Coregonus lavaretus) mit Lebendplankton. - European Association of Fish Pathologists, 88-91.

75 3 Growth and survival of whitefish larvae reared on dry and live food.

RC)SCH, R. & APPELBAUM S. (1985): Experiments on the suitability of dry food for larvae of Coregonus lavaretus L. - Aquaculture 48: 291-302.

ROSCH, R. & DABROWSKI, K. (1986): Tests of artificial food for larvae of Coregonus lavaretus from Lake Constance. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 273-282.

R()SCH, R. & ECKMANN, R. (1986): Survival and growth of prefed Coregonus lavaretus L. held in illuminated net cages. - Aquaculture 52: 245-252.

ROSCH, R. & SEGNER, H. (1990): Development of dry food for larvae of Coregonus lavaretus L. I. Growth, food digestion and fat absorption. - Aquaculture 91: l 01-115.

ROTH, H. & GEIGER, W. (1968): Aufzucht von Besatzfischen in Tr(jgen: 3-39. - Veroffentlichungen des Eigenossischen Atntes fi.lr Gewasserschutz und der Eidgenossischen Fischereiinspektion Nr. 25.

SALOJARVt K. (1987). Why do vendacc (Coregonus alhula L.) populations fluctuate? - Aqua Fennica 1711: 17-26.

SALOJARVI, K. Cl 991). Stock recruittnent relationships in the vendace ( Coregonus albula L.) in Lake Oulujarvi, northern Finland. - Aqua Fennica 21/2: 153- 161.

SALOJARVI, K. (1992). C01npensation in \vhitefish ( Coregonus lavnretus L.) populations in Lake Oulujarvi, northern Finland.·-Finn. Fish. Res. 13: 31- 48.

SALONEN, E., AHONEN, M. & MCTENIA, A. (1998). Development of whitefish ( Coregonus lavaretus) population and effects of large-scale compensation stocking in Lake Inari, northern Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 439-448.

SCHMID, W. (1993): Tagesfressrhythmik von Felchenlarven (Coregonus lavaretus L.), Vergleich von Lebend- zu Trockenfutter. - Diplome work, University of Constance.

STEFFENS, W. (1995): Yield and stocking of vendace (Coregonus albula) in northeast Germany. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 405-412.

76 3.7 References

STEINHART, M. & ECKMANN, R. (1992): Evaluating the nutritional condition of individual whitefish (Coregonus spp.) larvae by the RNA/DNA ratio. - J. Fish Biol. 40: 791-799.

TAYLOR, W.W. & FREEBERG, M. H. (1984): Effect of food abundance on larval lake whitefish, Coregonus clupeaforrnis Mitchill, growth and survival. - J. Fish Biol. 25: 733-741.

TODD, T. N. (1998). Enviromnental modification of gillraker nmnber in coregonine fishes. - Arch. Hydrobiol. Spec. Issues Advanc. Li1m101. 50: 305-315.

TURKOWSKI, K. & BONAR, A. (1995). Effects of species composition and stocking on com1nercial catches of vendace, Coregonus albula (L.) in Ostroda lakes (northern Poland). - Arch. Hydrobiol. Spec. Issues Advanc. Linn101. 46: 397-403. lJRBAN-JEZIERSKA, E. & BIERNACKI, T. (1992): The effect of natural food and starters on growth and development of Coregonus lavaretus in the 1st month of life. - Pol. Arch. Hydrobiol. 39/1: 109-132.

VENTLING-SCHWANK, A. R. & LIVINGSTONE, D. M. (1994): Transport and Burial as a Cause of Whitefish (Coregonus sp.) Egg Mortality in a Eutrophic Lake. - Can. J. Fish. Aquat. Sci. 51/9: 1908-1919.

VENTLING-SCHWANK, A. R. & MULLER, R. (1991): Survival of coregonid (Coregonus sp.) eggs in Lake Se1npach, Switzerland. - Verh. Internat. Verein. Li1nnol. 24: 2451-2454.

WANZENBOECK, J. & JAGSCH, A. (1998). Comparison oflarval whitefish densities in lakes with different schemes of larval stocking and fishing practice. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 497-505.

WATANABE, T., KlTAJIMA, C. & FUJITA, S. (1983): Nutritional values of live organisms used in Japan for mass propagation of fish, a review. - Aquaculture 34: 115-143.

WINBERG, G. G. (1956): The rate of 1netabolism and the food requirement of fishes. - Publ. Bieloruss. State Univ. Minsk.

77 3 Growth and survival of whitefish larvae reared on dry and live food.

78 4.1 Abstract

4 IMPORTANCE OF PREY MOVEMENT, FOOD PARTICLE SIZE AND TANK CIRCULATION

4.1 Abstract

A commercial dry diet, live and frozen zooplankton were compared as food for larvae of Lake Hallwil whitefish ( Coregonus suidteri Patio) during the first 3 weeks after hatching. Whitefish larvae fed live zooplankton grew significantly faster than those fed frozen zooplankton, but mortality did not differ. Larvae fed the dry diet reached nearly the same length after three weeks than did the fish fed frozen zooplankton. Howevec inortality of fish fed the dry diet was significantly higher. We conclude that prey rnovement is a key factor for food acceptance by whitefish larvae, accounting for the higher growth rates by larvae fed live zooplankton. High inortality is still the main proble1n when feeding dry diet. Therefore, we tested for the first time the effect of two additional parameters on larval inortality: food particle size and tank circulation. The size of food pellets was reduced frmn 200-400 ~tm to 100-200 ~nn by crushing. The water inflow to the tanks was placed below the water surface to increase the floating tiine of the dry diet. Our results showed that neither particle size nor inflow placernent seemed to affect mortality when fish were fed ad libitum. Nevertheless, floating ti1ne was prolonged significantly when the inflow was placed below the water surface, making it possible to lower the feeding frequency, to reduce the food losses and, thus, to lower the daily ration fed.

79 4 In1portance of prey 1110vement, food particle size and tank circulation

4.2 Introduction

Management of coregonids has always been difficult because whitefish disap- pears frmn eutrophic lakes (BRUTSCHY & GONTERT, 1923; COLBY et al, 1972; HARTMANN, 1977 & 1979). 'Therefore, stocking whitefish larvae could not only be conducted to increase YCS (LESKELA et al., 1995) and yield (TURKOWSKI & BONAR, 1995), but also to sustain populations without natural reproduction (STEF- FENS, 1995) or even to restore extinct populations (LUCZYNSKI et al., 1998). Nev- ertheless, intensive stocking does not guarantee strong year classes and high yields. To allow for sustainable n1anagemcnt and development of whitefish populations, it is therefore iinportant to investigate the factors affecting whitefish survival. Year class strength (YCS) of fish, and thus stock size is determined during the early life stage (MAY, 1974; SALOJARVl, 1991). Several factors are known to influence YCS: rearing conditions, size at the time of stocking (HOAGMAN, 1974; FLlJCHTER, 1980; TAYLOR & FREEBERG. 1984), the nu1nbcr of fish stocked (CHRISTIE, 1963, MOLLER, 1990) and enviromncntal conditions (PONTON & MtJLLER, 1989; Eckmann ct al., 1988).

Lake Hallwil (Switzerland) is a eutrophic lake, in which practically all whitefish (Coregonus suidteri Fatio, non1enclature according to KOTTELAT, 1997) eggs die during e1nbryonic developn1ent because of the anoxic sediment (VENTLTNG- SCHWANK & MOLLER, 1991; MOLLER, 1992 & 1993; VENTLING-SCHWANK & LIVINGSTONE, 1994). Nevertheless, thanks to artificial breeding and stocking, whitefish is the dominant and commercially lnost irnportant fish species in Lake Hallwil since 1977 (MULLER et al., 1994). In spite of the intensive stocking, the yields were not stable. In 1992, 1993 and 1994 yields were very low. The ai1n of the present work was therefore to improve the rearing technique of whitefish larvae with respect to growth and survival, and to determine why there are differences between rearing on dry diets and live zooplankton (chapter 3). The potential importance of prey movement on larval growth and survival were tested using live versus frozen zooplankton. Two additional parameters hypothesized to influence rearing success on dry diets were tested: food particle size and circulation patterns within rearing tanks.

80 4.3 Methods

4.3 Methods

4.3.1 Experimental design

4°C 4°C TC cb

' ' - ZH live 0.9 g ...... ,. ... .,.f...... +...... ;...... 1. 0 -- ZH frozen or ZH live 0.6 g '""''"'"'"['"''""'"'"!•······· ; "'.'_·:·::::~:~r-:~·::·::::::·:. 0' 8 dry diet ~.~~~~'"''''"1'·~-~...... 0.6 ., ...... f ...... + ...... f...... +...... 0.4 : : : : rel ...... l ...... , ...... f...... +.., ...... 1...... 0.2 -0 . . ' . : ! 0 0 5 10 15 20 days

Figure 4.1 Experimental design: Three replicate tanks per treatment, water volmne of 9 litres and 300 newly hatched larvae per tank (top; Z = zooplankton, T =dry diet Trouvit c =crushed, r =raw, a= inflow above water surface, b = inflow below water surface, H = heated to 12 °C, c = cold, 4 °C) and daily food ration per tank (bottom).

The whitefish larvae used originated from artificially incubated eggs of Lake Hall- wil whitefish. Rearing procedures used in the experin1ents are described by ROSCH & SEGJ\TER (1990), ROSCH (1992) and in chapter 3. Experi1nents on prey movement were conducted in 12 experimental tanks (each 9 litres and 300 larvae). Water was heated to 12°C and fish were fed live or frozen zooplankton or dry diet (Figure 4.1) over a 20 day period. In additional 12 tanks (each 9 litres and 300 lar-

Sl 4 Iinportance of prey movement, food particle size and tank circulation vae), inflow tubes were alternatively placed above or below the water surface, and the dry diet was crushed before feeding (particle size 100-200 µin) or fed as manufactured (particle size 200-400 µm). Te111perature was kept at 4 °C, and all larvae were reared on dry diet for 24 days (Figure 4.1). For abbreviations see legend of Figure 4.1. Different test periods were chosen for 12°C and 4°C tanks, clue to delayed mortality expected in cold water (chapter 3).

4.3.2 Diet types

Food was provided ad libitum beginning on the day after hatching, designated day 09 0 of the experiment. Dry diet Trouvit Perla Marine , a product manufactured by Hendrix SpA, Mozzecane (Italy), was fed in six equal portions given at 10, 11 and 12 a.111. and 2, 4 and 6 p.111. every day. According to KOSKELA (1988) a daily ration of 0.6 g provides ad libitum feeding for 300 larvae (Figure 4.1 ).

Zooplankton was collected every day in Lake Hallwil with two plankton nets of 200 µrn mesh size (ROTH & GEIGER, 1968) and stored at 4 °C in an aerated 30 liter tank. Some zooplankton subsamples were frozen in portions of 0.2 g dry weight. Daily rations of 0.6 g (ZH live and ZH frozen) respectively 0.9 g (ZH live) dry weight was given in three equal portions (Figure 4.1). A zooplankton sample was preserved on day 7 for species identification and length nieasurernent (a subsam- ple of 200 individuals was n1easured to the nearest 0.1 nn11). Length of zooplankton was expressed in standard Artemia nauplii length (0.7 mm). We did this because Artemia is the best food for whitefish larvae due to its small size (ROSCH, 1989; ECKMANN, 1985). This allows an overall estimation of the size distribution of the zooplankton fed in the experi1nent. Zooplankton s1naller (values < l) or at Artemia size (values = 1) were optimal, zooplankton larger than Artemia (values > 1) vvere less suitable for newly hatched whitefish larvae.

4.3.3 Statistics

On days 0, 8, 13 and 20 ten larvae were removed from each tank and their length was detennined to the nearest 0.5 mm. Cmnulative mortality was computed according to the formula of Schmid (1993), taking into account sainpling reduction (ORTLEPP, 1984).

82 4.4 Results

An analysis of variance at p<0.05, ANOVA (LOZAN, 1992) was nrnde to compare the mean length on day 20 in all groups, mortality at day 20 in groups reared at l2°C and n1ortality on day 24 in groups reared at 4 °C. Because length and Inortal- ity did not differ significantly within all TC groups (dry diet, 4°C), the live larvae from the 12 4°C tanks were pooled for mean length calculation, and so were the dead larvae. Then, at-test at p<0.05 (LOZAN, 1992) was made to c01npare length of 120 living and 41 dead larvae on day 20.

4.4 Results

4.4.1 Prey movement

In the zooplankton smnple of clay 7, about 95% of the organisms were copepods, 5% were cladocerans. Only about 30% of the zooplankton fed hadArtemia size or was sn1aller, the rest was bigger, especially cladocerans like Daphnia sp. (Figure 4.2).

m:mi:!i[i~ii@:;;itN'il available when larvae 10-1 Smm 11111111111 available in addition when larvae 15-20mm - available in addition when larvae 20-35mm V//b available in addition when larvae >35mm _...... 40 ('(j~ +-'0 .._.0 +-' c 30 '+- 0 0 +-' Oaphnia sp. +-' ~ cc 20 Q) cu () a.. 10 I- 0 Q) 0 0.. N 0 0.6 0.8 1.2 1.6 2.0 2.4 2.8 3.2 zooplankton size as standard Artemia length

Figure 4.2 Size composition of the zooplankton collected from Lake Hallwil on day 7, calculated as standard Artemia size (indications of larval length according to HARTMANN & KLEIN, 1993)

83 4 Importance of prey 1novement, food particle size and tank circulation

Larvae fed live zooplankton pursued their prey actively in the middle and top of the tank, while those fed frozen zooplankton svvitchcd during the experilncnt from swimming in the middle to staying at the botto1n of the tank. Larvae fed dry diet were randomly distributed all the time.

Energetic value of dry diet T and zooplankton was 22.7 kJ/g and 21.5 kJ/g, respectively, based on hmnb calorimetry analyses (chapter 3).

4.4.2 Growth

On day 20, ]arvae of the group ZR live 0.6 g and 0.9 g had reached mean body lengths of 16.1 mm and 16.9 mm, respectively; larvae fed frozen plankton averaged 14.4 mm; and larvae fed dry diet averaged 13.6 mm (Figure 4.3 and Figure 4.4).

r--1 ~ E 15 ...__,.E ? f f ..c +-' ANOV A: p=0.178 O> c Q) f 10 + TC (larvae still alive >- t-test: u • on day 20, n=120) 0 p=0.00001 ..0 + TC (larvae died 5 on day 20, n=41) +-'co 0 D ZH live 0.6g +-' <> ZH live 0.9g 0

Figure 4.3 Length of dead and living larvae of all TC groups (left) and larvae fed 0.6 g and 0.9 g live zooplankton (right, 3 tanks each treat1nent) on day 20. For abbreviations see Figure 4.1.

84 4.4 Results

--ZH live 0.9g

18 ., ...... , • • • • • ZH frozen ,.... ~ ...... ,...... ,. . ., .., ...... E ~®_·-~'~'~_'T_H~,----~ E ~ ...... 16 O'.l c <].) >. "O 0 14 ...Cl ...... (1j 0 ....., 12

.~ \····1·· 10 0 5 10 15 20 time [days]

Figure 4.4 Length of larvae as 1nean value and standard deviation of the experi- 1nental groups ZH live 0.9 g, ZH frozen and TH (3 tanks each treat- 1nent). For abbreviations see Figure 4.1.

Larvae fed a daily ration of 0.9 g live zooplankton were not significantly longer than the larvae of the ZH live 0.6 g group (Figure 4.3). Length differences were significant between ZH live 0.6 g I 0.9 g and ZH frozen (p<0.05) and between ZH live 0.6 g I 0.9 g and TH (p<0.005), whereas no differences between TH and ZH frozen were found (p>0.05).

4.4.3 Mortality

M01tality among larvae fed live zooplankton was 0% and only ca. 3% Inortality occurred in the ZH frozen group. This difference was not significant (p>0.05). The TH group showed the highest losses of all groups (34% ). Differences in lnortality

85 4 Importance of prey movement, food particle size and tank circulation on day 20 between ZH live and TH (p<0.001) and between ZH frozen and TH (p<0.005) were significant. ZH live and ZH frozen suffered therefore less mortality than TH.

In all TC groups, no significant differences in mortality (Fig. 4.5) or length occurred (p>0.05), but dead larvae were significantly s1naller (t-test, p=0.00001) than the living ones (Figure 4.3).

----·TC cb 1 5 ,_ ...... i-TC rb ·· ..... Tc ra ,...... , >!2_ "'""'""""""TC ca ...... 0 >. 10 -(ij t 0 E

0 0 5 1 0 15 20 25

time [days]

Figure 4.5 Cumulative mortality as mean value and standard deviation of the three replicates of the experimental groups TCca, TCcb, TCrb and TCra. For abbreviations see Figure 4.1.

4.4.4 Inflow placement and particle size of dry diet

After feeding, the main part of the dry diet given to the fish of the intlow-below- water-surface tanks was floating until next feeding. Feeding intervals were 1 hour in the morning.._. and 2 hours in the afternoon. In the inflow-above-water-surface tanks, the dry diet had sunk to the bottmn within a quarter of an hour. Mean final

86 4.5 Discussion lengths of the TCcb and TCca groups were 11.8 mm and 11.5 mm, respectively. Mean lengths of the TCrb and TCra groups were 11.6 mm and 11.8 mm, respectively. Neither the placement of the inflow nor diet particle size had a significant effect on mortality (Figure 4.5) or final length (p>0.05).

4.5 Discussion

4.5.1 Prey movement

Small larvae prefer non-evasive prey types (MOOKERJI & RAO, 1993). They select the most abundant species, even if they would be able to eat bigger prey types (PONTON & MtJLLER, 1990), and they are gape-limited (MOOKERJI & RAO, 1994). According to HARTMANN & KLEIN (1993) and our measurements (Figure 4.4 ), whitefish larvae at first feeding were about 101run long. These larvae were thus able to ingest only 30% of the zooplankton mixture (Figure 4.2). At the end of the experiment, the larvae reared on live zooplankton at 12°C were longer than 15 1run and showed optimal growth silnilar to the findings of ROSCH & DABROWSKI (1986). According to HARTMANN & KLEIN (1993), the larvae were probably able to ingest at least 60% of the zooplankton (Figure 4.2 and Figure 4.3). There were no significant differences in larval size between ZH live 0.6g and ZH live 0.9g groups (Figure 4.3). Thus, even a zooplankton ration of 0.6g dry weight provides ad libitmn feeding for 300 larvae.

Prey 1novement occurred only in the two ZH live groups. Frozen zooplankton and dry diet slowly sank to the tank bottmn after feeding. In the ZH frozen group, the larvae stayed close to the tank bottmn from the second week until the end of the experiment. Their stomachs were slightly filled at any time of the day. The behaviour of the ZH frozen-fed larvae may indicate that they preyed not only on sinking zooplankton at feeding tilne, but also on zooplankton at the tank bottmn. TH group larvae also had their stomach filled throughout the day because food was given every hour or every second hour, but they never stayed near the tank bottmn.

The energetic values of all diets were similar. Our results for zooplankton agree with the findings by GIUSSANI & DE BERNARDI (l 977) and URBAN-JEZIERSKA & BIERNACKI (1992). Therefore, parmneters other than the energetic value inust have been responsible for different growth (Figure 4.4). We conclude that prey

87 4 Importance of prey move1nent, food particle size and tank circulation move1nent is a key factor for food acceptance by whitefish larvae and thus influences growth.

4.5.2 Growth

The observed growth differences (Figure 4.4) lead us to conclude that food acceptance was low in ZH frozen and TH groups. KLEIFELD-KRIEBITZ & ROSCH (1987) also reported slower growth of frozen zooplankton-fed whitefish larvae as compared to fish fed live zooplankton. ROSCH & DABROWSKI (1986) and R()SCH (1989) found that zooplankton-fed larvae start ingesting food earlier than dry diet- fed larvae. Therefore, prey inove1nent see111s to be a key factor for food acceptance and, indirectly, for growth (Figure 4.4 ), but not for mortality.

4.5.3 Mortality

Feeding frozen zooplankton had no negative effect on larval survival. Mortality was lowest in ZH live and ZH frozen groups, whereas dry diet-fed groups suffered high 1nortality. Therefore, Lake Hallwil zooplankton was suitable for whitefish larvae and contained no «stressors» that would have caused high mortality as reported for Lake Constance (ECKMANN, 1985; ROSCH, 1994).

Natural live food appears to be the optimal food for larvae of Coregonus sp .. How- ever, zooplankton availability is often limited. In early spring, collecting sufficient zooplankton with nets (ROTH & GEIGER, 1968) is very time-consuming due to the low abundance of zooplankton. In view of its fair to good perfonnance with respect to growth and mortality, the use of frozen zooplankton for rearing whitefish larvae is a good alternative. Zooplankton can be harvested and frozen in large quantities in late spring when abundance is high, and stored until the next year. A simple feeding method using frozen zooplankton was developed by KLEIFELD- KRIEBITZ & Rosen (1987): The inflow water was led through a header tank of 1 litre in which the frozen block could thaw before the water with the thawed zooplankton reached the rearing tank. MEDGYESY & WIESER (1982) had developed a basically similar but more complicated feeding inethod: Their apparatus has a dis- pensing system pennitting either a continuous or an intermittent supply of freshly thawed zooplankton. making it possible to rear whitefish larvae on frozen zooplankton as successfully (low mortality, high growth rate) as with Arteniia nauplii (MEDGYESY & WIESER, 1982).

88 4.6 Acknowledgements

4.5.4 Inflow placement and particle size of dry diet

High mortality in larvae fed dry diet seems to be a proble1n of food acceptance by the smallest larvae (Figure 4.3) because dead larvae were significantly smaller than the living ones (Figure 4.3). However, neither particle size of the dry diet nor inflow place1nent influenced mortality (Figure 4.5) when ad libitun1 feeding was provided. Still, inflow placement is i1nportant when long-ti1ne availabilty of dry diet is needed without having to feed manually every hour.

When using dry diet for rearing whitefish larvae, the water inflow should be placed below and not above the water surface. The inflow-below-water-surface method has several advantages: First, it increases floating time of a dry diet and therefore reduces feeding frequency. S1nall portions of the floating particles keep constantly sinking and are available to the larvae. Second, it reduces cleaning effort due to lower food loss. Third, it reduces the costs for feed, because the dry diet daily ration can be reduced, and it makes the use of costly artificial feeders unnecessary. A first positive long-scale experience with this 1nethod was made at Lake Samen, Switzerland (AMSTALDEN, pers. cmmn.).

4.6 Acknowledgements

Thanks are due to Prof. R. Eckmann, Constance, Dr. A. Kirchhofer and Prof. W. Meyer, Bern, Dr. N. Mookerji, Montreal, and Dr. R. Rosch, Langenargen, for giving 1nost valuable advice concerning the rearing methods. Farmix AG, Malters, and Hendrix SpA, Mozzecane, provided free smnples of the dry diet used. Special thanks to U. Fischer, President of the Sport Fishing Club of Lake Hallwil, for authorizing to use the hatchery facilities, and to M. Keller for his great interest and active help during the experiments. Special thanks also to Dr. E. Meister, ETH ZUrich, for providing an oxygen bomb calorimeter and for his help during the experiments. The help and support by my parents and many colleagues at EAWAG, especially Dr. H. R. Biirgi, B. Germann, C. Heller, Dr. A. Peter, C. Rell- stab and Dr. C. Wedekind is greatly acknowledged. The constructive comments by Prof. J. V. Ward and anony1nous reviewers helped to iinprove earlier drafts of this paper.

89 4 Importance of prey movement, food particle size and tank circulation

4. 7 References

BRUTSCHY, A., GUNTERT, A. (1923): Gutachten Uber den Rlickgang des Fischbestandes i1n Hallwilersee. - Arch. Hydrobiol. 14: 523-571.

CHRISTIE, W. J. (1963): Effects of artificial propagation and the weather on recruit1nent in the Lake Ontario whitefish fishery. - J. Fish. Res. Board Can. 20: 597-646.

COLBY, P. J., Spangler, G. R., Hurley, D. A. & McCombie, A. M. (1972): Effects of eutrophication on salmonid comnmnities in oligotrophic lakes. - J. Fish. Res. Bd. Can. 29: 975-983.

ECKMANN, R. (1985): Histopathological alterations in the intestine of whitefish (Coregonus sp.) larvae reared on zooplankton from Lake Constance (West Germany). - Diseases of Aquatic Organisms 1/1, 11-18.

ECKMANN, R., GAEDKE, U. & WETZLAR, H. J. (1988): Effects of clilnatic and density-dependent factors on year-class strength of Coregonus lavaretus in Lake Constance. - Can. J. Fish. Aquat. Sci. 45/5, 1088-1093.

FLDCHTER, J. (1980): Review of the present knowledge of rearing whitefish (Coregonidae) larvae. - Aquaculture 19, 191-208.

GHJSSANI, G. & DE BERNARDI, R. (1977): Food selectivity in Coregonus sp. [zooplanktophagous fish] of Lake Maggiore [Italy], an energetical approach. - Memorie dell'Istituto Italiano di Idrobiologia, Dott. Marco de Marchi Verbania Pallanza (Italy) 34, 121-130.

HARTMANN, J. (1977): Fischereiliche Veranderungen in kulturbedingt eutrophierenden Seen. - Schweiz. Z. Hydrol.. 39/2: 243-254.

HARTMANN, J. (1979): Unterschiedliche Adaptionsfahigkeit der Fische an Eutrophierung. - Schweiz. Z. Hydrol. 41/2: 374-382.

HARTMANN, F. & KLEIN, M. (1993): Nahrungsselektion von Renkenbrut ( Coregonus lavaretus) unter Aufzuchtbedingungen. - Fischer & Teichwirt 44/8, 279-283.

90 4.7 References

HOAGMAN, ·w. J. (1974 ): Feeding by alewives (Alosa pseudoharengus) on larval lake whitefish (Coregonus clupeaformis) in the laboratory. - J. Fish. Res. Board Can. 31/2, 229-230.

KLEIFELD-KRIEBITZ, G. & ROSCH, R. (1987): A simple method of feeding coregonid larvae on frozen zooplankton. - Journal of Applied Ichthyology 3, 119-124.

KOSKELA, J. (1988): Rearing of larvae of the whitefish Coregonus muksun Pallas with artificial diets. - Finn. Fish. Res. 9, 333-337. KOTTELAT, M. (1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the fmmer USSR), with an introduction for non-systematists and connnents on nomenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

LESKELA, A., HUDD, R., LEHTONEN, H. & SANDSTROM, 0. (1995): Abiotic factors, whitefish stockings and relative year-class strength of anadromous whitefish (Coregonus lavaretus L.) spawning populations in the Gulf of Bothnia. - Arch. Hydrobiol. Spec. Issues Advanc. Li1m101. 46: 241-248.

LOZAN, J. L. (1992): Angewandte Statistik fiir Naturwissenschaftler. Paul Parey, Berlin/Hainburg, 237 pp. LUCZYNSKI, M., KUZMINSKI, H., DOBOSZ, s. & GORYCZKO, K. (1998): Gene pool characteristics of whitefish ( Coregonus lavaretus) fingerlings produced in a hatchery for restoration stocking purposes. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 50: 317-3 21. MAY, R. C. (1974): Larval mortality in marine fishes and the critical period concept. In: Blaxter J. H. S. (Ed.), The Early Life History of Fish. Springer Verlag, Berlin, pp. 3-19.

MEDGYESY, N. & WIESER, W. (1982): Rearing whitefish (Coregonus lavaretus) with frozen zooplankton by means of a new feeding apparatus. - Aquaculture 28: 327-337.

MOOKERJI, N. & RAO, T. R. (1993): Patterns of prey selection in rohu (Labeo rohita) and singhi (Heteropneustes fossilis) larvae under light and dark condition. - Aquaculture 118, 85-104.

91 4 Importance of prey movement, food particle size and tank circulation

MULLER, R. ( 1990): Management practices for lake fisheries in Switzerland. In: van Densen, W. L. T.; Steinmetz, B.; Hughes, R.H., (Eds), Manage1nent of freshwater fisheries. Pudoc. Wageningen, pp. 477-492.

MOLLER, R. (l 992): Trophic state and its iinplications for natural reproduction of salmonid fish. - Hydrobiologia 243/244, 261-268.

MOLLER, R. (1993): Einige fischereibiologische Aspekte von Seesanierungen. - Fortschr. Fisch.wiss. 11, 43-56.

MULLER, R., BIA, M. M. & MENG, H.J. (1994): Die Felchenfischerei in einigen Seen der Zentralschweiz und des Mittellandes. - Mitteilungen zur Fischerei 55. Bundesmnt ftir Umwelt, Wald und Landschaft (BUWAL).

ORTLEPP, J. (1984): Beobachtungen zmn Verlauf des Wachstums und der Sterblichkeit von Coregonenlarven bei FiHterung mit Zooplankton. - Diplomarbeit Biologie, Universitat Freiburg.

PONTON, D. & MtJLLER, R. (1989): Alirnentation et facteurs de mortalite des larves de coregones (Coregonus sp.). Exemple de deux lacs de niveaux trophiques differents, les lacs de Smnen et de Hallwil (Suisse Centrale). - Aquatic Sciences 51/l, 67-83.

PONTON, D. & MOLLER, R. (1990): Size of prey ingested by whitefish, Coregonus sp., larvae, Are Coregonus larvae gape-lin1ited predators? - J. Fish Biol. 36/l, 67-72.

ROSCH, R. (1989): Beginning of food intake and subsequent growth of larvae of Coregonus lavaretzts L. - Pol. Arch. Hydrobiol. 36/4, 475-484.

ROSCH, R. (1992): Food intake and growth of larvae of Coregonus lavaretus L., effect of diet and light. - Pol. Arch. Hydrobiol. 39, 671-676.

ROSCH, R. (1994): Probleme bei der Aufzucht von Felchenlarvaen (Coregonus lavaretus) mit Lebendplankton. - European Association of Fish Pathologists, 88-91.

92 4. 7 References

ROSCH, R. & DABROWSKI, K. (1986): Tests of artificial food for larvae of Coregonus lavaretus fr01n Lake Constance. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22, 273-282.

ROSCH, R. & SEGNER, H. (1990): Development of dry food for larvae of Coregonus lavaretus L. I. Growth, food digestion and fat absorption. - Aquaculture 91, 101-115.

ROTH, H. & GEIGER, W. (1968): Aufzucht von Besatzfischen in Trogen. Veroffentlichungen des Eigenossischen Amtes fi.tr Gewasserschutz und der Eidgenossischen Fischereiinspektion 25: 3-39.

SALOJARVI, K. ( 1991 ): Stock recruitment relationships in the vendace ( Coregonus albula L.) in Lake Oulujarvi, northern Finland. - Aqua Fennica 21/2: 153- 161.

SCHMID, W. (1993): Tagesfressrhythmik von Felchenlarven (Co reg onus lavaretus L.), Vergleich von Lebend- zu Trockenfutter. - Diplomarbeit an der Universitat Konstanz.

STEFFENS, W. (1995): Yield and stocking of vendace (Coregonus albula) in northeast Gennany. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 405-412.

TAYLOR, W.W. & FREEBERG, M. H. (1984): Effect of food abundance on larval lake whitefish, Coregonus clupeafonnis Mitchill, growth and survival. - J. Fish Biol. 25, 733-741.

TURKOWSKI, K. & BONAR, A. (1995): Effects of species composition and stocking on cmnmercial catches of vendace, Coregonus albula (L.) in Ostroda lakes (northern Poland). - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 397-403.

URBAN-JEZIERSKA, E. & BIERNACKI, T. (1992): The effect of natural food and starters on growth and develop1nent of Coregonus lavaretus in the 1st 1nonth of life. - Pol. Arch. Hydrobiol. 39/1, 109-132.

VENTLING-SCHWANK, A. R. & LIVINGSTONE, D. M. (1994): Transport and Burial as a Cause of Whitefish (Co reg onus sp.) Egg Mortality in a Eutrophic Lake. - Can. J. Fish. Aquat. Sci. 51/9, 1908-1919.

93 4 Importance of prey movement, food particle size and tank circulation

VENTLING-SCHWANK, A. R. & MULLER, R. (1991): Survival of coregonid (Coregonus sp.) eggs in Lake Sempach, Switzerland. - Verh. Internat. Verein. Limnol. 24, 2451-2454.

94 5.1 Abstract

5 FOOD PREFERENCE OF ADULT WHITEFISH, AND THE QUESTION OF CANNIBALISM

5.1 Abstract

Whitefish (Coregonus suidteri Patio, nomenclature according to KOTTELAT, 1997) is the dominant and cormnercially most important fish species in eutrophic Lake Hall wil. Yield started to increase in 1977 but has not been stable due to strong fluctuations in year class strength (YCS). Because food is one of several key parameters influencing population dynamics, food preference of adult whitefish was studied. Our results show that in spring, 20-35 percent of the whitefish exclusively preyed on lake bottom fauna when zooplankton density was low due to low water temperature. From 50-65 percent of the stomachs contained zooplankton exclusively during spring. The percentage of whitefish eating both planktonic and benthic organisms was about 15-20 percent over the whole year. Artifi- cial hypolimnetic oxygenation since 1985 has substantially increased living space of benthic organisms. This has resulted in an improved benthic food base for Lake Hallwil whitefish. In spring, as soon as the zooplankton density had risen, whitefish switched from benthic prey to planktonic crustaceans. In sununer and fall, only 1 percent of the stomachs contained exclusively benthic prey while 80 percent of the whitefish preyed exclusively on zooplankton. Phyllopods like Daphnia sp. and Bythotrephes longirnanus were highly selected, whereas copepods like Cyclops sp. were consumed only when phyllopod density was very low. The question of potential food competition between adult whitefish and larvae was studied by comparing our data on feeding habits and feeding depth of adults with the

95 5 Food preference of adult whitefish, and the question of cannibalis1n respective data for larvae from the literature. Because cannibalis1n of adult whitefish on larvae was suspected to influence population dynamics of Lake Hallwil whitefish, 990 whitefish stomachs were checked for fish larvae in spring, but no larvae were found. We conclude that the different feeding depth and food preference of whitefish larvae and adults in spring makes intraspecific cmnpetition and cannibalism unlikely.

5.2 Introduction

Lake whitefish, Coregonus sp., are known to exploit both pelagic and benthic food resources. They are opportunistic feeders. Thus, stmnach contents largely reflect seasonal changes in prey availability (MOOKERJI et al., 1998). During the period of intensive feeding and growth (May - September), the fish are found in the upper 20m of lakes where they prey primarily on cladocerans (Crustacea: Phyl- lopoda). Large and non-evasive species, such as Daphnia spp. and Bythotrephes longimanus, are thereby highly selected (GERSTMEIER, 1986; MOOKERJI et al., 1998). S1naller species like Bosniina sp. (Crustacea: Phyllopoda) and evasive organisms like copepods (Crustacea: Copepoda) are consmned mainly at times when the preferred prey are rare (MOOKERJI et al., 1998). In autmm1, whitefish move to deeper layers, show an increased proportion of empty stomachs and prepare for spawning (SKURDAL et al., 1985b). In winter and spring, when zooplankton concentration is low, some whitefish forms switch frmn zooplankton to benthic organisms, such as annelids, insect larvae and insect pupae (NAGY, 1990; POMEROY, 1991; HARTMANN & PROBST, 1995; MICHEL, 1996; MOOKERJI et al., 1998).

In eutrophic lakes, the density of benthic organisms is often low due to anoxic conditions of the hypoli1m1ion and the sediment. Our study site, Lake Hallwil (Switzerland), is a eutrophic lake in which practically all whitefish eggs die during embryonic development because of the anoxic sedi1nent (VENTLING-SCHW ANK & MOLLER, 1991; MULLER, 1992 & 1993; VENTLING-SCHWANK & LIVINGSTOI\1E, 1994). After the start of artificial oxygenation of L. Hallwil in 1985 (STOCKLI & SCHMID, 1987; SCHEIDEGGER et al., 1994), the density of benthic organis1ns increased considerably (STOSSEL, 1992). Thanks to artificial breeding and stocking, Lake Hallwil whitefish (Coregonus suidteri Patio; nmnenclature according to

96 5.3 Methods

KOTTELAT, 1997), a fast growing type, has been the dominant and commercially inost important fish species in Lake Hallwil since 1977, reaching yields of 50 kg/ha (Ml'JLLER et al., 1994). In spite of intensive stocking, stock size and yield strongly fluctuate. ln 1992, 1993 and 1994 yield was very low. Year class strength of fish, and thus stock size and yield is determined during the early life stage (MAY, 1974; SALOJARVI, 1991 ). The hypothesis of this work was that cannibalis1n or food competition between adults and larvae might decimate a new whitefish year class. Cannibalis1n on eggs has been reported for whitefish (FABRICJUS & LINDROTH, 1954; SKURDAL et al., 1985a). Additionally, cannibalis1n on larvae was found for whitefish under laboratory conditions (GRIM, 1951 ), but never in nature. The aim of this study was therefore to investigate the role of food uptake and preference of adult Lake Hallwil whitefish on year class strength, with special emphasis on the situation in late winter and spring, the time when whitefish larvae hatch and tnay be vulnerable to cannibalism and food con1petition.

5.3 Methods

The adult Lake Hallwil whitefish used in this study were obtained from the c01n- mercial fishery that employed gillnets of 35-38 nnn bar mesh size. Age was deter- 1nined using scales. In 1998 and 1999, 30 stomachs were taken weekly during late winter and spring, i.e. after the beginning of stocking of artificially bred whitefish larvae. In 1999, 30 stmnachs were also taken inonthly during sununer, fall and early winter. The st01nachs were preserved in 70% alcohol, taken to the laboratory and processed according to BOWEN (l 996). E1npty st01nachs were recorded, and the content of stmnachs containing food was preserved in 70% alcohol. Each stomach was checked for larval whitefish, and the prey organis1ns were identified and counted using a counting chamber and a WILD binocular microscope at 25- 50x magnification. Samples consisting of less than l 00 organisms were counted in their entirety, whereas large samples were subsampled and the mean value of three subsmnples of 30-50 organisms detennined. If digestion had made proper counting itnpossible, stomach contents were esti1nated as reported by PILLA Y (1952).

The number of whitefish ston1achs in which a prey category was found was deter- 1nined in relation to the total number of stomachs containing food (DINEEN, 1951),

97 5 Food preference of adult whitefish, and the question of cannibalism using the frequency of occurrence method (f.o.o.-rnethod; HYSLOP, 1980). This results in a qualitative overview of the whitefish food preference. To assess the importance of each prey category for whitefish nutrition, the percentage c01npo- sition by number 1nethod (p.c.n.-1nethod; HYSLOP, 1980) and the Index of Rela- tive In1portance IRI (PINKAS et al., 1971: PRINCE, 1975) was used. The IRI was calculated for each prey category according to DRAGOVITCH (1970), taking into account the stomach content volume (V), the f.o.o. (F) and the p.c.n. (N): IRI = (%N + % V)*%F. Stomach content volume was estilnated volumetrically (JUDE, 1973) to the nearest O.lrnl and numerically (MCCOMISH, 1966) to calculate satu- dty (S), according to GERSTMETER ( 1985) with the formula S=(V/G)*lOO, where Vis the food volmne [ml] and G the total body weight of whitefish [g]. The results of MICHEL (l 996), RORGI & EGLI (1984) and our own measuren1ents on 120 prey organisms were taken into account: Body length, width and height of prey organisms were measured to the nearest 0.051nm, excluding tails and legs (BtJRGI & EGLI, 1984 ). For volume calculations, crustaceans were treated as ellipsoids (RORGI & EGLI, 1984), and insects and annelids as cylinders. The volumes found for the different prey categories were multiplied with the total number of individuals per category and sampling day. Leptodora kindtii, Bosniina sp. and arachnids were found in very low amounts only and were therefore omitted from the volmne calculations (except for Bosmina sp. on 25 October 1999).

Zooplankton was sampled monthly from the middle of Lake Hallwil using vertical hauls fron1 44m to surface with nets of 95µm inesh size (BDRGI, 1983). Te1nper- ature and oxygen saturation was measured siinultaneously using a YSI 6000 UPG inulti-parameter instrument. A inetal dredge of 30 cm width and 280 ~nn mesh size in the bag was used to sample benthic organisn1s. To do this, the dredge was pulled 15 111 along the lake bottmn. Organisms were caught at seven different depths. To investigate food selectivity of Lake Hallwil whitefish, the index a (CHESSON, 1978, 1983) was calculated for plankton organisms. This index is calculated as ai=Cr1/n1)/((r1/n1)+(r2/n2)+(r3/n3)+ ... ), where rand n are the nun1ber of prey type 1,2,3 ... in the stomach [Ind./ stomach] and the lake [Ind./m3], respectively. For index a calculation, the mean organis111 density was used, taking into account the average depth where whitefish stayed. Pie diagrains were used to show whether the fish preyed on both benthic and planktonic organisms.

98 5.4 Results

5.4 Results

In spring 1998, 60% of the whitefish sampled were two years o1d, 30% were three and 10% four to six years old. In 1999, 38% of the fish were two years old, 60%) were three and 2% four to six years old. Commercial fishing was conducted at different depths, but large amounts of whitefish were only caught at the depths indicated in Figure 5.1 (pers. conm1. E. Fischer and H. Weber).

s~10- 'E_20·· d! "O 30

40-

0 oo Oo 0 go go 0 0 0 N N N N N oxygen saturation[%]

Figure 5.1 Depth where whitefish were mainly caught during 1999 (gray area) and monthly oxygen saturation profiles for Lake Hallwil.

1998 1999 100 ;:?. 0 :1 -o~ (j) Ul = .r: ' -.;:: () fllI, I! i - o E(\J

In 1999 (and also in 1998, pers. comm. E. Fischer), whitefish stayed at depths between ten and forty meters from January until the 1niddle of April (Figure 5.1). From May to June, they lived in the epilirnnion between the surface and ten 1ncters

99 5 Food preference of adult whitefish, and the question of cannibalism depths. In July-August they moved down to depths beyond twenty meters; during this period a pronounced oxygen minimmn developed in the metalimnion (Figure 5 .1). In fall, the fish caine back closer to the surface and stayed around ten to fifteen meters depths, while in December, they were caught between fifteen and twenty ineters depth

In 1998, whitefish were already feeding in February when san1pling began, whereas in 1999, all stomachs examined were empty until the iniddle of March (Figure 5.2). At the end of November, with the onset of the spawning season, feeding stopped. Food organisn1s found in the 990 stomachs sainpled were grouped into six categories: Copepods and phyllopods (Crustacea), chironmnids and chaoborids (Insecta), annelids (Figure 5.3, Figure 5.4 and Figure 5.5) and inites (Arachnida). Because mites were found in very low numbers in only one stomach on 19 April 1998, this category does not appear in Figure 5.3, Figure 5.4 and Figure 5.5.

1998 ~ 100 ~ I QJ u c 80 ~ 1 ::J u 60 u 0 0 40 ········· ...... ·--··-· >. u c f ...... QJ 20 l]i ::J rr QJ 111~ ~ ~ f "1 l" u: 0 I m t ll J:l J:l J:l ... '- ...... >. >. >. c. 0. 0. c. u. u. u. ~"' 4'"' 4'"' 4'"' . E !I u E E c ,·' "--- ~ QJ 20 ..... i ::J ' ,t t rr ·~ QJ . I I l .. IE E n ~ill ~1 u: 0 I 1. 1 '- '- ... '- '- >. >. c:: :; c...... c. c. c. ::J u "' "'

Figure 5.3 Frequency of occurrence of the five main food categories found in the stomachs of adult whitefish in 1998 and 1999.

100 5.4 Results

1998 100

40 ...--. :::R.__.0 I... 20 Q) ..0 E :J 0 c i.. ~ i.. i...... a...... a... .c .c .c 0. >i >i >. >. Cl> Cl> Cl> ro ro ro ro 0. c. c. ro ro ro ..0 LL. ~ ~ ~ ~ <( <( <( <( LL. LL. I I I I I I I I ~ ~ ~ I I I I c ("') LO T- I I en co "':f' 0 ...... en <.D ("') 0 ("') N N 0 :.+;:; T- N N """"' N . ii) """"' """"' """"' """"' 0 0... E 1999 0 (.) 100 Q) OJ cu +-' 80 c Q) (.) I... Q) 60 0... 40

0 /./ ' a... a... i.. i.. a...... >i >i ::i c. (.) ro ro c. c. c. ro <1) <( ro ::i= """) ~ I (./) :!:I I ~ I ~ ~ ~ .., 0I I I I lO I en tO tO ("') 0 ,...... ll> LO ("') 0 ...... N 'I'"" N N N """"' """"'

I [lliJ Crustacea: Phyllopod~-----o-1·~-~~cta~ChironomidaJ !Ill Crustacea: Copepoda • Insect.a: Ch~oboridae L___ u Annelida: Ol1gochaeta. ----- Figure 5.4 Percentage composition by number for the f oocl in the stomachs of adult whitefish in 1998 and 1999. Food categories as in Figure 5.3.

101 5 Food preference of adult whitefish, and the question of cannibalism

1998 20000

15000

0::: 10000

5000

0- a.. a.. a.. a.. a.. a.. a.. a.. .0 .0 .0 (13 (13 (13 (13 >. >i >. Q.) 0. 0. 0.. 0.. (13 t'tl C\'l CL> CL> <( <( u. :E :E :E :E I I :;:: I u.I u. I I I ~ ~ :;:: :;:: LO 'I"'" CT> I I CT> CX) ~' 0 ' r--. M 'I"'" r--. u:> M 0 'I""" 'I"'" 'I""" N N ' 'I"'" N N M 'I"'" N 1999 20000

15000

a: 10000

a.. a.. 0. 0.. 0.. CL> CJ) ~ ~ I M 0 N M "' Crustacea: Phyllopoda ...... _ lnsecta: Chironomidae m,,,,,~~"""~Crustacea: Copepoda --lnsecta: Chaoboridae , "" '"''Annelida: Oligochaeta

Figure 5.5 Index of Relative Importance (IRI) calculated for the five main food categories found in the stomachs of adult whitefish in 1998 and 1999.

102 5.4 Results

In the category phyllopods, daphnids were dominant. Very rarely son1c individuals of Bosmina sp. were found. In the sa111ples of 20 May 1999, Bosmina sp. made up 3%, on 15 June 1999 only 1 q, and on 6 July 1999 5<7o of the phyllopods. How- ever, in the samples of 25 October 1999. they were highly selected by one fish. Therefore, they made up nearly 50% by number of the total food (Figure 5.4). On 6 July 1999, Bythotrephes longimanus dominated the ston1ach contents (90% of the phyllopods). On 15 June and 25 October 1999, some rernains of Leptodora kindtii were found. \Vithin the category copepods, cyclopoids were the inain prey items (Figure 5.6), whereas calanoids made up smne 10% of the copepods. Cope- pods were highly selected in spring 1999~ whereas in spring 1998 (like in sum1ner and fall of 1999) daphnids were preferred (Figure 5.6). Zooplankton always pre- vailed in the whitefish diet during sununcr and autumn (Figure 5.4, Figure 5.5 and Figure 5.6).

D only benthic ~ zooplankton and • only zooplankton organisms benthic organisms

1 spring 1998 spring 1999 summer/fall 1999 x 0.8 Q.> "O c 0.6 c 0 ~ 0.4 ID ..!:: 0 0.2

Figure 5.6 Food selectivity of Lake Hallwil whitefish for four zooplankton groups (bars, Chesson index). Chesson values lower than 0.2 indicate no selection, while values above 0.6 indicate strong.... selection. Prefer- ence for benthic or planktonic organis111s (pie diagrams).

In spring, benthic prey (Table 5.1) represent an i111portant pait of the whitefish diet (Figure 5.3, Figure 5.4, Figure 5.5 and Figure 5.6). Chironmnid pupae were found

103 5 Food preference of adult whitefish, and the question of cannibalism in the stmnach samples of 29 April 1998 and 13 May 1998, while chaoborid pupae were present in the samples of 29 April 1998, 30 April 1999, 6 July 1999 and 7 September 1999. In all other samples containing chironomids and chaoborids, these were present as larvae (Figure 5.3 and Figure 5.4). Due to the high degree of digestion, annelids could not be clearly identified. Still, it appeared that the fish ate at least two species of Oligochaeta (Annelida).

Table 5.1 Abundance of benthic organisms, by density classes, in deeper sedi- 1nents of L. Hallwil in 1998.

Depth Oligochaeta Chironomidae Chaoboridae 2 ? 2 I [m] (Annelida) [lnd.!m ] (lnsecta) [Ind.Im--] (lnsecta) [lnd.!m ] ' I

20 200-800 0-50 5-800 I

I I 25 I 200-800 0-5 10-200 30 200-800 0-1 10-200 35 200-800 0-1 10-200 1 40 200-800 0 10-50 45 50-200 0 5-50 I 46 50-200 0 I 5-50 I I I

For calculation of IRI and saturity values, stmnach content volume had to be estimated (Figure 5.5). The values used for food volmne calculation were (expressed as the volmne of 1000 individuals) 0.07 ml for bosminids, 0.1 ml for copepods, 0.5 ml for daphnids, l.5 ml for Bythotrephes longimanus, 4 ml for annelids, 9 ml for insect larvae, and 16 ml for insect pupae. A regression between stomach content volun1es calculated nmnerically and volumetrically accounted for more than 70% of the variation in the mean outcon1e (r-squared value). Therefore, we conclude that both rnethods yield satisfactory results. S aturity in Figure 5. 7 was

104 5 .5 Discussion drawn on the basis of nmneric volume, revealing an unexpected decrease during summer 1999.

1998 - - - Temperature in 7.5m - Saturity Plankton density phyllopocls (light) cope pods ( clark)

Figure 5.7 Lake Hallwil te1nperature at 7.5 111 depth, zooplankton density and stomach saturity of adult whitefish during the smnpling periods 1998 and 1999. Saturity is a1neasure of food intake per unit of body weight (see text).

In spring, annelids had the highest IRI values, whereas in smnmer and fall, this was the case for phyllopods (Figure 5.5). For some short periods, when insect larvae were available in large number, chaoborids becmne important food items (Figure 5.5). Dming the diet change frorn benthic prey to plankton organisms in spring, copepods also contributed to high IRI values (Figure 5.5).

5.5 Discussion

5.5.1 I~ood preference of adult whitefish

Water temperature has a profound effect on the physiological processes in fish. Temperature effects are particularly important in spring when the surface waters, n1ediated by varying meteorological conditions~ warn1 up n1ore or less quickly. GRIFFITH ct al. (1992) found a positive coffelation between growth of subadult

105 5 Food preference of adult whitefish, and the question of cannibalism whitefish and te1nperature. Low water te111perature in February slowed down the developtnent of zooplankton (PONTON & Mt:TLLER, 1989), as can also be seen in Figure 5.7. In the cold spring of 1999, adult \Vhitefish ate substantial numbers of benthic organis1ns two weeks longer than in the warn1er spring of 1998 (Figure 5.4). Furthermore, the start of feeding was delayed in 1999 due to low water te1nperature (Figure 5.2 and Figure 5.7). In spring 1998, decreasing plankton density appeared to be the cause of a short decline in saturity. The phenmne- non of a decrease in saturity in smmner has also been described for fast growing Lake Lucerne whitefish (MICHEL, 1996). After intensive feeding in April and May, food intake of L. Hallwil whitefish was low during June, but increased again in July (Figure 5.7). We hypothesize that these changes resulted frmn the switch from the large-sized benthic organisms to the s1nall-bodied zooplankton. High zooplankton densities in July (Figure 5.7) might have favored intensive feeding on zooplankton and. thus, have lead to a second saturity increase. However, in August-October 1999, whitefish retreated to below the pronounced metalimnetic oxygen minimmn (<0.5 mg 0 2/1at10 in depth; Figure 5.1) where feeding apparently was sub-optin1al. Metalimnetic oxygen ininima of less than 1 111g 0 2/l had regularly occurred before 1996 but not again until 1999. In 1999 it was suspected to be the result of a flood event in Nlay with high nutrient input that led to highly increased p1i1nm·y production and, consequently, to an oxygen depression in the metali1m1ion (H. MINDER, pers comm.).

Artificial oxygenation of L. Hallwil since 1985 (STOCKLI & SCHMID, 1987; SCHEIDEGGER et al., 1994) obviously had a positive effect on whitefish in the sense that it enlarged the living space for the fish. Oxygen cnricluncnt of the hypolinmion also enabled an abundant lake bottom fauna to develop (STOSSEL, 1992). This increased the nmnber of food organis1ns available to adult whitefish (Figure 5.3, Figure 5.4, Figure 5.5 and Figure 5.6), allowing adult whitefish to avoid starvation during spring, when zooplankton density was low (Figure 5.7). In spring, annelids were highly selected, whereas in surn1ner and fall, phyllopods were most abundant in the whitefish diet (Figure 5.4, Figure 5.5 and Figure 5.6). Adult whitefish switched frorn benthic prey to planktonic crustaceans as soon as zooplankton density had risen in late spring. \Ve found that L. Hallwil whitefish feed opportunistically, similar to Lake Lucerne whitefish (Mooket~ji et al., 1998): During periods when insect larvae are present in large numbers, the fish are able to use this food type with great efficiency (Figure 5 .5). Overall, food preference

L06 5 .5 Discussion of L. Hallwil whitefish co111pares well \Vith that of fast growing coregonids from other lakes (NAGY, 1990, POMEROY, 1991, HARTMANN & PROBST, 1995, MICHEL, 1996, MOOKERJI et al., 1998).

5.5.2 Intraspecific food competition

NAESJE et al. (1986) reported that whitefish larvae in Norwegian lakes start feeding in the littoral zone where they stay during the entire first growing season. Adults usually stay in the hypolimnion (RUDSTAM & MAGNUSON, 1985), whereas the juveniles prefer the wanner epilimnion (JURVELIUS & HEIKKINEN, 1987; PONTON & MENG, 1990). HAMRI'-.J (1986) found that niche segregation between larvae and adults occurs 24 hours a day. He postulated that fish size dependent temperature preference, light, and intraspecific interaction were responsible for the different behaviour patten1 which changes through ontogenetic develop1nent SAND LUND et al. (l 991) found different food preference and niche segregation between larvae and adults fron1 spring (May) until July. In L. Hallwil, adult whitefish prefer to stay in deep zones during winter and spring (Figure 5.1) as reported by MAYR (1998). Therefore, spatial niche segregation might help to avoid food competition in spring. Potential competition between larvae and adult inay be limited to surmner and autmnn. HoweveL the high density of prey organisms in this eutrophic lake probably precludes food competition at all.

HOAGMAN (1973) and PONTON & MtJLLER (1989) found that coregonid larvae do not eat all the plankton species but prefer s1nall prey types. Coregonid larvae are gape-li1nited (HARTMANN & KLEIN, 1993) and feed in a way as to increase foraging efficiency by choosing the more abundant prey available (PONTON & MULLER, 1990). According to PONI'ON & MULLER (1989), L. Hallwil whitefish larvae cat n1ainly copepod nauplii and calanoid and cyclopoid copepods from first feeding until the end of April. In May, daphnids begin to dmninate their food. At this time, they arc 15-191n1n long (PONTON & MfTLLER, 1989) and fairly secure from predation (HOAGMAN, 1974) and cannibalism (GRIM, 1951). In the adult whitefish we sainpled, no copepod nauplii were found in the stomach. In comparison with the findings of PONTON & MlJLLER (1989) for larvae, food preference in spring may be assumed to be different for adult and larval whitefish. In late winter and spring, when the larvae were present and zooplankton concentration was low, 20-35 percent of the adult whitefish population exclusively preyed on

107 5 Food preference of adult whitefish, and the question of cannibalism lake bottom fauna (Figure 5.3, Figure 5.4. Figure 5.5 and Figure 5.6). In contrast, in surmner and fall, adult whitefish inainly preyed on phyllopods, especially daphnids (Figure 5.3, Figure 5.4, Figure 5.5 and Figure 5.6), in accordance with findings by MAYR (1998) and MOOKERJI et al. (1998). Only when phyllopod density was low (Figure 5.7), did the fish selected copepods (Figure 5.6). For larval whitefish, PONTON & Ml)LLER (1990) found a significant increase in prey length with increasing fish length for calanoids (Copepoda) and daphnids (Phyllopoda). According to BERG et al. ( 1994 )~ the proportion of copepods in the diet decreases with increasing fish length while daphnids becmne inore ilnportant. Large and vis- ible prey are usually strongly selected by adult whitefish (MA YR, 1998; MOOKERJI et al., 1998). Thus, prey size and visibility is inore i1nportant than prey density (GIUSSANI & DE BERNARDI, 1977). This is well docun1ented for the large can1iv- orous cladoceran Bythotrephes longimanus: During a period when B. longinwnus was not found in the plankton sainples at all, son1e stomachs contained up to 90% B. longirnanus. Obviously, B. longinumus was selected very efficiently, as described by MAYR (1998) and MOOKERJI et al. (1998). In contrast) the large but transparent cladoceran Leptodora kindtii was rarely found in whitefish stomachs. Overall, we conclude that in L. Hallwil whitefish, different food preference and spatial niche segregation of adult and larval whitefish are tnechanisrns which help to ease or even to avoid intraspecific food competition.

5.5.3 Cannibalism

Predation of adult whitefish on their own larvae is a priori an important consider- ation in detennining year class strength. It has been frequently docmnented that strong year classes can only develop when fish density is low or inedium (AASS, 1972; SALOJARVI, 1987; GOODYEAR, 1980; VILJANEN, 1986, 1988; JURVELIUS, 1991; SANDLUND et al., 1991). Basically, cannibalis1n may occur in whitefish if the gain in tenns of fitness is higher than the reduction in survival caused by this behavior (SKURDAL et al., 1985). At very high fish density, mechanisms prevent- ing food competition might break down. Cannibalisn1 would thus help to reduce the population to tolerable density.

Although cannibalism in whitefish has never been proven in nature, CARANHAC & GERDEAUX (1998) hypothesized that stock abundance controls recruitment through cannibalis1n. Based on laboratory studies with larval and adult whitefish,

108 5.6 Acknowledgen1ents

GRIM (1951) concluded that larval cannibalism of adult whitefish was responsible for a four-year cycle of low and high whitefish yield in Lake Constance. He further concluded that larvae with good swinuning ability, after six weeks of rearing in tanks, were not vulnerable to cannibalism. If cannibalis1n on larvae exists in nature, then it should occur just after stocking newly hatched larvae in spring. In our sample of 990 whitefish stomachs collected from L. Hallwil during this period, only one larva-like ren1ain was found in one sample on 20 May 1998. Therefore, we conclude that cannibalisn1 on larvae is not likely to occur in L. Hall- wil whitefish. The spatial niche segregation between larvae and adults mentioned above appears to effectively prevent cannibalisn1.

Other than larval cannibalisn1, egg cannibalism in whitefish is well-documented (FABRICIUS & LINDROTH, 1954; SALOJARVL 1987; VENTLING-SCHWANK, 1992). In one case it was found to account for about lQC}{) egg mortality (SKURDAL et al., 1985). In L. Hallwil, egg cannibalism. is not of concern because natural reproduction of whitefish is not functional (MULLER, 1992, 1993). All whitefish originate frmn artificial reproduction and stocking.

5.6 Acknowledgements

Thanks are due to cmnmercial fisheries Delphin, Meisterschwanden, and Weber, Birrwil, for providing freshly caught whitefish. \Ve especially thank Dr. A. Stackli, Aarau, for lnonitoring zooplankton and abiotic parmneters, the laboratory of the Umweltschutzabteilung for analysis of abiotic parameters and C. J olidon, Dtibendorf, for the counting of L. Hallwil zooplankton san1ples. Kindly thanks to my mother for her active help during stomach sainpling days. Also. D. Hohmann, Kastanienbaum, and 0. Heiri, Bern, are greatly acknowledged for their help during the taxonomic investigations of the stomach contents. The constructive comments by Prof. J. V. Ward helped to i111prove earlier drafts of this paper.

5. 7 References

AASS, P. (1972): Age determination and year-class fluctuation of cisco, Coregonus albula L., in the Mjosa hydroelectric reservoir. - Report Institute of Freshwater Research (Drottningholm) 52: 5-22.

109 5 Food preference of adult whitefish, and the question of cannibalism

BERG, S., JEPPESEN, E., SONDERGAARD, M. & MORTENSEN, E. (1994): Environmental effects of introducing whitefish, Coregonus lavaretus in Lake Ring. - Hydrobiologia 275-276: 71-79.

BOWEN, S. H. (1996): Quantitative Descrition of the Diet. Chapter 17 in Fisheries Techniques, 2nd edition (B.R. Murphy & D.\V. Willis, eds). American Fisheries Society, Bethesda: 513-532.

BtJRGI, H. R. & EGLI, B. (1984): Crustaceen-Plankton des VierwaldsHittersees. Grossenfraktionierung und Abundanzdynarnik in den J ahren 197 6 bis 1979. - Schweiz. Z. Hydrol. 46: 247-268.

BORGI, H. R. (1983): Eine neue Netzgarnitur mit Kipp-Schliessmechanisnrns fdr quantitative Zooplanktonfange in Seen. Schweiz. Z. Hydrol. 45/2, 505-507.

CARANHAC, F. & GERDEAUX, D. (1998): Analysis of the fluctuations in whitefish (Coregonus lavaretus) abundance in Lake Geneva. Arch. Hydrobiol. Spec. Issues Advanc. Linmol. 50: 197-206.

CHESSON, J. (1978): Measuring preference in selective predation. - Ecology 59: 211-215.

CHESSON, J. (1983): The estimation and analysis of preference and its relationship to foraging inodels. - Ecology 64: 1297-1304.

DINEEN, C. F. (1951): A comparative study of the food habits of Cottus bairdii and associated species of Sal1110nidae. - A1n. ~1idl. Nat. 46: 640-645.

DRAGOVITCH, A. (1970): The food of blue fin tuna (Thunnus thymms) in the Westen1 North Atlantic Ocean. - Trans. A1n. Fish Soc. 99: 726-728.

FABRICIUS, E. & LINDROTH, A. (1954): Experimental observations on the spawning of whitefish, Coregonus lavaretus L, in the stream aquarium of the Htille Laboratory at River Indalsalven. Rep. Inst. Freshw. Res. Drottninghohn 35: 105-112.

GERSTMEIER, R. (1985): Nahrungsokologische Untersuchungen an Fischen in1 Nationalpark Berchtesgaden. - Arch. Hydrobiol. Suppl. 72: 237-286.

GERSTMEIER, R. (1986): Investigations on the feeding biology of coregonid fishes (Coregonus species) from the Lake Starnberg (Bavaria, West Germany). - "p1xiana.S ' ' 9/3 : ")'15..,,;., -..,~,.,""'·')'"l')

110 5. 7 References

GIUSSANI, G. & DE BERNARDI, R. (1977): Food selectivity in Coregonus sp. [ zooplanktophagous fish] of Lake Maggiore [Italy], an energetical approach. - Memorie dell'Istituto Italiano di Idrobiologia, Dott. Marco de Marchi Verbania Pallanza (Italy) 34: 121 130.

GOODYEAR, C. P. (1980): Compensation in Fish Populations. - Biological monitoring of fish: special session at the 110th - Annual meeting of the Ainerican Fisheries Society (Louisville) 11: 253-280.

GRIM, J. (1951): Kannibalismus bei Blaufelchen und seine moglichen Folgen. - Osteneichs Fischerei 4: 165-171.

HAMRIN, S. F. (1986): Vertical distribution and habitat partitioning between different size classes of vendace, Coregomts albula, in thennally stratified lakes. - Can. J. Fish. Aquat. Sci. 43/8: 1617-1625.

HARTMANN, F. & KLEIN, M. (1993): Nahrungsselektion von Renkenbrut (Coregonus lavaretus) unter Aufzuchtbedingungen. - Fischerei und Teichwirt 44/8: 279-283.

HARTMANN, J. & PROBST, L. (1995): Divergent distributions of prey and predator, a fishing effect? - Aquatic Sciences 57/2: 106-118.

HOAGMAN, W. J. (1973): The hatching, distribution. abundance, growth and food of the larval whitefish (Coregomts clupeqf'onnis Mitchill) of Central Green Bay, Lake Michigan. - Report Institute of Freshwater Research (Drottningholm) 53: 1-20.

HOAGMAN, W. J. (1974): Feeding by alewives (Alosa pseudoharengus) on larval lake whitefish (Coregonus clupeqf'onnis) in the laboratory. - J. Fish. Res. Board Can. 31/2: 229-230.

HYSLOP, E. J. (1980): Stomach contents analysis - a review of methods and their application. - J. Fish. Biol. 17: 411-429.

111 5 Food preference of adult whitefish, and the question of cannibalisrn

JUDE, D. J. (1973): Food and feeding habits of Gizzard Shad m Pool 19, Mississippi River. - Trans. Amer. Fish. Soc. 102: 378-383.

]URVELIUS, J. & HEIKKINEN, T. (1987): The pelagic fish density, biomass and growth of vendace, Coregonus albula L., n1onitored by hydroacoustic methods and trawling in a Finnish lake. Aqua Fennica 1711: 27-34.

KoTTELAT, M. (1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the former USSR), with an introduction for non-systematists and cmntncnts on nomenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

MAY, R. C. (1974): Larval mortality in rnarine fishes and the critical period concept: 3-19. The Early Life History of Fish. - Blaxter J. H. S. (Ed.), Springer Verlag, Berlin.

MAYR, C. (1998): ZumEinfluss von Trophie, Fischdichte und Habitatwahl auf die Nahrungs- und Wachstumsbcdingungcn von Renken (Co reg onus lavaretus L.) in vier oberbayerischen Seen. - Dissertation an der Fakultat der fiir Biologic der Ludwig-Maximilians-Universitiit Mtinchen. 221 pp.

MCCOMISH, T. S. (1966): Food habits of bigmouth and smallmouth buffalo in Lewis and Clarke Lake and the Missouri River. - Trans. Am. Fish. Soc. 96: 109-128.

MICHEL, M. (1996): Untersuchungen zur Nahrungsokologie von Grossfelchen im Vierwaldstattersee wahrend des Sonnnerhalbjahres 1996. - Diploma Thesis ETH. No. 10366: 73pp.

MOOKERIJ N., HELLER C., MENG H. J., BlTRGI H. R. & MOLLER R. (1998): Diel and seasonal patterns of food intake and prey selection by Coregonus sp. in rc-oligotrophicated Lake Lucen1e, Switzerland. - J. Fish Biol. 52/3: 443- 457.

MDLLER, R. ( 1992): Trophic state and its implications for natural reproduction of salmonid fish. Hydrobiologia 243/244: 261-268.

112 5. 7 References

MDLLER, R. (l 993): Einige fischereibiologische Aspekte von Seesanierungen. - Fortschr. Fisch.wiss. 11: 43-56.

MiJLLER R., BIA M. M. & MENG H.J. (1994): Die Felchenfischerei in cinigen Seen der Zentralschweiz und des Mittellandes. - Mitteilungen zur Fischerei 55. Bundesamt fur Urnwe1t, \Vald und Landschaft (BUWAL).

NAESJE, T. F., SANDLUND, 0. T. & JONSSON, B. (1986): Habitat use and growth of age-0 whitefish, Coregonus lcrvaretus, and cisco, Coregonus albula. - Environmental Biology of Fishes 15/4: 309-314.

NAGY, S. (1990): The food of the whitefish (Coregonus lavaretus L.) in the Strbske pleso n1ountain lake (Czechoslovakia). Zivocisna Vyroba 35/10: 915-920.

PILLA Y, T. V. R. (1952): A critique of the methods of study of food of fishes. - J. Zool.Soc. India 4: 185-200.

PINKi\S, L, OLIPHANT, M. S. & IVERSON, I. L. K. (1971): Food habits of albacore, bluefin tuna and bonito in Californian Waters. - Calif. Fish. Garne. 152: 1- 105.

POMEROY, P. P. (1991): A cmnparative assessment of te111poral variation in diet of powan, Coregonus lavaretus (L.), fro111 Loch Lomond and Loch Eck, Scotland, UK. - J. Fish Biol. 38/3: 457-478.

PONTON, D. & MENG, HJ. (1990): Use of dual-heain acoustic technique for detecting young whitefish, Coregonus sp., juveniles: first experiments in an enclosure. - J. Fish Biol. 36: 741-750.

PONTON, D. & MOLLER, R. (1989): Ali1nentation et facteurs de rnortalite des larves de coregones (Coregonus sp.): Exemple de dcux lacs de niveaux trophiques differents, les lacs de Samen et de Hallwil (Suisse Centrale): - Aquatic Sciences 5111: 67-83.

PONTON, D. & MlJLLER, R. (1990): Size of prey ingested by whitefish, Coregonus sp., larvae, Are Coregonus larvae gape-limited predators? J. Fish Biol. 36/1: 67-72.

PRINCE, E. D. (1975): Pinnixid crabs in the diet of young-of-the-year Copper Rockfish (Sebastes caurinus). - Trans. Arri. Fish. Soc. 104: 539-540.

113 5 Food preference of adult whitefish, and the question of cannibalisn1

RUDSTAM, L. G. & MAGNUSON, J. J. (1985): Predicting the vertical distribution of fish populations: Analysis of cisco, Coregonus artedii, and yellow perch, Perea flavescens. - Can. J. Fish. Aquat. Sci. 42/6: 1178-1188.

SALOJARVI, K. (1987): Why do vendace (Coregonus albula L.) populations fluctuate? - Aqua Fennica 1711: 17-26.

SALOJARVI, K. (1991): Stockrecruit1nentrelationships in the vendace (Coregonus albula L.) in Lake Oulujarvi, northern Finland. - Aqua Fennica 21/2: 153- 161.

SANDLUND, 0. T., JONSSON, B., NAESJE, T. F. & AASS, P. (1991): Year-class fluctuations in vendace, Coregonus albula (Linnaeus): Who's got the upper hand in intraspecific competition? - J. Fish Biol. 38/6: 873-886.

SCHEIDEGGER A., STOCKLI A. & \VOEST A. (1994): Einfluss der inten1en Sanierungsmassnahmen auf den Sauerstofihaushalt in1 Hallwilersee. - Wasser, Energie, Luft 5/6: 126-131.

SKURDAL J., BLEKEN E. & STENSETH N. C. (1985a): Cannibalism in whitefish (Coregonus lavaretus). - Oecologia (Berlin) 67/4: 566-571.

SKURDAL. J., HESSEN, D. 0. & BERGE, D. (1985b): Food selection and vertical distribution of pelagic whitefish Coregonus lavaretus in Lake Tyrifjorden, Norway. - Fauna Norvegica Series A 6: 18-23.

STOCKLI, A., SCHMID, M. (1987): Sanierung des Hallwilersees. Wasser, Energie, Luft 7917,8: 143-149.

STC>SSEL~ F. (1992): Die Bodenfauna im Hallwilersee dringt vor. - EAWAG-news 34D: 23-26.

VENTLING-SCHW ANK A. R. ( 1992): Reproduktion und larvale Entwicklungsphase der Felchen ( Coregonus sp.) im eutrophen Scrnpachersee. - Dissertation an der Philosophischen FakulHit der Universitilt Ziirich.

VENTLlNG-SCHWANK A. R & LIVINGSTONE D. M. (1994): Transport and burial as a cause of whitefish ( Coregonus sp.) egg mortality in a eutrophic lake. Can. J. Fish. Aquat. Sci. 51/9: 1908-1919.

114 5. 7 References

VENTLING-SCHWANK, A. R. & MtLLER R. (1991): Survival of coregonid (Coregonus sp.) eggs in Lake Sen1pach, Switzerland. - Verh. Internat. Verein. Lhnnol. 24: 2451-2454.

VILJANEN, M. (1986): Biology, propagation, exploitation and managen1ent of vendace (Coregonus albula L.) in Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 73-97.

VILJANEN, M. (1988): Population dynamics of vendace (Coregonus albula L.) in Finland. - Joensuun Yliopiston Luonnontieteellisia Julkaisuja 12: 1-19.

115 5 Food preference of adult whitefish, and the question of cannibalism

116 6.1 Abstract

6 A MODEL FOR EVALUATING MORTALITY FACTORS IN WHITEFISH LARVAE

6.1 Abstract

A population dynamics 111odel was used to evaluate mortality factors detennining year-class strength (YCS) and yield of Lake Hallwil whitefish (Coregonus suidteri), a fast growing coregonid form in a eutrophic Swiss lake. We used the number of whitefish larvae stocked annually and the mean age of gillnetted whitefish as input variables, while data on annual whitefish yield and YCS were used to validate the model. The following factors affecting survival of the larvae were tested: intra- and interspecific c0111petition, predation, fishing pressure, weather, zooplankton density and quality, effect of the cyanobacteria Planktothrix rubescens on density of phyllopods, the preferred food of larger whitefish larvae. Our 111odeling results indicated that the weather situation in May negatively influenced YCS and yield. Thus, it sustained our hypothesis that gas supersaturation, resulting from intensive oxygen production by algae dming sunny weather in spring, caused lethal gas bubble syndrome in L. Hallwil whitefish larvae. Our inodeling results further indicated that high phyllopod densities in May were associated with strong year classes.

6.2 Introduction

Lake Hallwil whitefish (Corerzonus~ suidteri Fatio.. n0111enclature according~ to KOTTELAT, 1997), a fast growing coregonid form, represents the economically most important fish species in L. Hallwil today. Due to eutrophication, however, natural reproduction of L. Hallwil whitefish is nonexistent (MOLLER, 1992). This

117 6 A model for evaluating mortality factors in whitefish larvae problem has been overcome by stocking whitefish larvae. Nevertheless, yield from sport and comrnercial gillnet-fishery showed high fluctuations (Figure 6.1) .

....-. Q) 40~~~~~~~~~~~~~~~~~~~~~~~-.- 6 ..c ro f/) > y:: ~ .!:'.! • prefect larvae Q) ..c(.) (()- 20 3 ro >...... c-mo ...-- ro >. '--'10 1.5 -0 3 Q) (]) ~ 0 0 .._ -0. 70 .--. Em yield ..c 5 (/) y:: ~50 • YCS u age at harvest .~ >. ..c -2 30 Q) .. ~ ..c ~ 10

0 ~--+--+-~~"'F"'.-tlll+ml

year

Figure 6.1 Nmnbers of newly hatched larvae (gray) and prefed larvae (black) stocked in Lake Hallwil fro1n 1967 to 1998 (above). Year class strength (YCS) calculated by virtual population analysis (black), whitefish yield (gray) frmn 1967 to 1998 and rnean fish age at harvest (below). YCS data for years after 1996 are not available yet.

Some weak or nonexistent whitefish year classes were the cause of these yield fluctuations (MfTLLER, 1993). We hypothesized that whitefish year-class strength (YCS) is determined by mortality during the early life stage similar to the results of MAY (1974) and SALOJARVI (1991). GOODYEAR (1980) segregated general mortality factors detennining YCS of fish into density-independent (e.g. weather,

118 6.3 Methods food quantity and quality) and density-dependent mortality factors (e.g. predation, cannibalis111, inter- and intraspecific con1petition). On the base of these findings for fish in general and our comprehension of L. Hallwil whitefish ecology, we tried to identify larval n1ortality factors that could have affected whitefish YCS and might be responsible for the yield collapse in 1992-94 (Figure 6.1 ).

To achieve our objectives, we developed a population dynainics inodel to test our hypotheses. Although models can be used as very effective tools to investigate dynamic systems, one has to be aware of their possibilities and limits. A population dynmnics model is liinited because it is a si1nplification of reality and could not explain every mechanisrn. However. model-generated results used in concert with field data help to elucidate the mechanis1ns of a dynamic ecosysten1. Cmn- puter modeling also helps to direct research and avoid unnecessary experlinents that may be costly and disruptive to the ecosystem..

6.3 Methods

Our population dynan1ics model was built using the HPS software Stella.ResearchTM5.0. A detailed introduction to ecological modeling with Stella is given by COSTANZA et al. (1998). On the base of our c01nprehension of biology and management of L. Hallwil whitefish we developed the basic n1odel SirnBal 2.0 (Sin1 for si111ulation and Bal for "Ballen'\ the local nan1e of whitefish). It inodels whitefish life history from the time of stocking, either as newly hatched larvae (NHL) or as prefed larvae (PL) reared in the hatcheries for six weeks, until harvested by gillnets (Figure 6.2).

In the language of Stella, four different n1odel symbols were available to describe the different steps in whitefish 1ife history as shown in Figure 6.2. The time series on nmnbers of stocked larvae were stored in two symbols called conveyor. For exan1ple, the purpose of the conveyor "newly hatched larvae" is to convey its content by an inforn1ation arrow to the input arrow "stocking NHL" (Figure 6.2).

119 6 A model for evaluating mortality factors in whitefish larvae

(}) 0 05 ~ ~ · critical extra food :;,.-ro ro ·"-u 2:

mean age at harvest

fishing

stocking PL

prefed larvae

Stella symbols pool input/output arrow conveyor 0 ~ system border 0 •information arrow Q

Figure 6.2 Structure of the basic 1nodel SimBal 2.0. and the extension to SimBal 4.4 (gray) for n10deling the gas bubble syndrome (GBS) scenario combined with phyllopod density data. Above: s-shaped curves for conveyors "critical sunshine" and "extra food". Below: explanation of Stella symbols.

120 6.3 Methods

Numbers of whitefish larvae stocked, available since 1967 (Figure 6.1), were calculated by assuming that one liter of artificially bred eggs was equivalent to 70'000 larvae (H. MINDER, pers. cornm.). Stocking of larvae was 1nodeled using two input arrows filling the pool "young of the year". A pool accumulates and releases values. In our exmnple (Figure 6.2), the pool "young of the year" accu- 1nulates number of whitefish stocked per year and releases every year class after a delay of one virtual year. Releasing is modeled using the output arrow "growing" which becon1es an input arrow for the pool "juvenile and adult fish" and transfers every whitefish year class from one pool to the other. The equations used for both pools (Figure 6.2) Rte:

young of the year(t) =young of the year(t - dt) + (stocking NHL+ stocking PL - growing)* dt (6.1)

juvenile and adult fish(t) =juvenile and adult fish(t - dt) + (growing fishing) * dt (6.2)

Finally, fishing is inodeled using the output arrow "fishing" and the mean age of whitefish at harvest (Figure 6.1) in a 1nodel elernent called first-order delay (ALFELD & GRAHAM, 1976):

fishing =juvenile and adult fish/(mean age at harvest-1) (6.3)

In our model, the first-order delay could simulate increased {reduced} fishing pressure by reducing {increasing} the 1nean age at harvest. Thus, more {less} virtual fish leave the pool "juvenile and adult fish" through the output arrow "fishing" as it is niathe1natically shown in equation 3. This simplification is based on our experience of L. Hallwil fishery. Furthennore, it was necessary to subtract 1 from the n1ean age at harvest (see equation 3) due to the one-year-delay caused by the pool "young of the year". For modeling the harvest of three years old fish, we have to wait for two virtual years and not three because virtual fish were already one year old when they entered the pool "juvenile and adult fish". Fish age at harvest was determined using scales (BAGENAL, 1974 ). For modeling the various factors, age at harvest was kept constant at 3 years according to our records of whitefish age at harvest since 1986 (100 fish were sampled three tirnes per year).

121 6 A model for evaluating mortality factors in whitefish larvae

To model mortality factors, it is important to know the survival rate of the whitefish larvae in nature until the age at harvest. Mean natural survival frmn 1967 to 1998 was calculated using the fonnula YCS=(SNm",*NHL)+(SpL *PL). S is the survival rate (RICKER, 1975). Mean survival of whitefish larvae was assumed to be ten titnes higher for PL than for NHL (MOLLER, 1990), thus, lO*SNHL=SPL· We used data on fish age in the catch to calculate YCS of L. Hallwil whitefish by virtual population analysis. \Ve then calculated mean survival rate by transfonn- i n g the fonnula YCS=(SNHL *NHL)+( 1 O*SNHL *PL) into SNHL=YCS/(NHL+lO*PL)=0.002. Furthermore, SpL=lO*SNHL=0.02. On the basis of this result for n1ean natural survival rate in nature, we set in our rnodel optimal survival rate to 0.004 for NHL and 0.04 for PL, and minirnmn survival rate to 0 for both. Mortality inducing values of the tested factors were modeled as s-shaped curves (Figure 6.2) which were fitted empirically. The resulting survival rate was then multiplied with the number of stocked larvae in the input arrows "stocking NHL" and "stocking PL". This way, the SimBal 2.0 111odel was extended to test different n1ortality factors (Figure 6.2). In the basic 111odel structure, fish went through virtual life history as number of fish. In the output arrow "111odeled yield" in the extended 1nodel, nun1ber of fish [Mio] was 111ultiplicd with a weight factor for converting to real yield [kg]. Yield statistics of sport and c01n- 1nercial gillnet fishery were available since 1967 frmn mandatory reporting logs. Observed yield was regressed against predicted yield. Also 1nodeled and observed YCS [Mio J was compared by regression. Results of the extended model versions were supposed to be closer to observed YCS and yield data and thus to show higher r-squared values and lower percent deviation than the results of the basic SimBal 2.0 model.

Based on detailed biological information on coregonids fr01n the literature, we also checked if our hypotheses were plausible, how we had to 1nodel a particular effect and which factor combinations wou Id need our special attention. Our first hypothesis concerned intraspecific competition (SimBal 3.1). Because in L Hall- wil yield and thus fish stock increased enormously since 1967 (Figure 6.1), competition between larvae and adult whitefish for food or even cannibalism may occur at high whitefish density. In literature, many authors report that a strong year class develops only when fish density is low or n1edium (AASS. 1972; SALOJARVl, 1987; GOODYEAR, 1980; VILJANEN, 1986, 1988; ]URVEUUS, 1991; SANDLUND et al.. 1991). Furthermore, not only yield but also the number of larvae

122 6.3 Methods stocked in L. Hallwil greatly increased after 1967 (Figure 6.1). Intraspecific competition between whitefish larvae (Si111Bal 3.2) may occur as a consequence of high larval density (see AASS (1972) and SALOJARVI (1987)). Also the stock size of other fish species is often high due to eutrophication. Therefore, interspecific competition (SimBal 3.3) inay also be important (BERG & GRIMALDI, 1966). There are several piscivore fish species in L. Hallwil, such as perch (Percafluvi- atilis), pike (Esox lucius), pike perch (Sander lucioperca), brown trout (Salrno trutta Lacustris). According to fisheries statistics, only perch and pike were abundant. Several authors stated that predation by perch (Simbal 3.4) 1nay affect whitefish year classes (SV ARDSON, 1977; SALOJARVI, 1987; MULLER et al., 1994). Sev- eral whitefish fingerlings were found in the stomach of perch frmn L. Sempach (Ml)LLER, unpubl. data). To model the effect of cmnpetition and predation on whitefish YCS, survival of whitefish larvae decreased with increasing values of the 1nortality inducing factor (SimBal 3.1: SpL =0.04 up to 30 t whitefish yield, then decreasing to 0 up to 60 t whitefish yield (overall SrLllO=SNHIJ; SimBal 3.2: SpL =0.04 up to 35 inillions of stocked larvae, then decreasing to 0 up to 40 mil- lions of stocked larvae; Si1nBal 3.3: SpL=0.04 up to 2 t yield of other fish species, then decreasing to 0 up to 8 t yield of other fish species; Si1nBal 3.4: SpL =0.04 up to 1 t perch yield, then decreasing to 0 up to 3 t perch yield). Yield also played a major role in our fifth hypothesis that should model the effect of fishing pressure on whitefish yield fluctuation (SimBal 3.5). According to SALOJARVI (1987), populations fluctuate due to fishing. But fishing also eases the effect of competition by removing substantial numbers of adult fish, thus lowering fish density that otherwise could lead to a weak year class (AASS, 1972). To n10del fishing pressure, mean age at catch was allowed to fluctuate between 2 and 5 years as it was recorded for L. Hallwil whitefish (Figure 6.1).

While fishing yield is related to the density of the fish stock weather (SimBal 3.6) is a fish density-independent factor and could influence ecosysterns in different ways. First, meteorological conditions in spring directly influence water temperature which may lead to total gas supersaturation due to water wanning (WEIT- KAMP & KATZ, 1980) and to oxygen supersaturation due to excessive algal photosynthesis in eutrophic lakes during long sunshine periods (MATHIAS & BARICA, 1985). Total gas saturation of more than 120% at 0-3m depth1nay cause lethal gas bubble syndrotne (GBS) in whitefish larvae (BOUCK, 1980; STADELMANN, 1988; VENTLING-SCHWANK, 1992). \Vhile GBS inay occur already at a total gas satura-

123 6 A model for evaluating mortality factors in whitefish larvae tion of 120%. higher values of oxygen saturation are required to induce GBS (VENTLING-SCHWANK, 1992). Therefore, only when oxygen saturation was higher than 140%, GBS effect was modeled by decreasing survival of whitefish larvae with increasing minutes of sunshine per month (for details see Figure 6.2). For survival rate modeling, the n1ore reliable sunshine tin1e series ([Min./n1onth], daily records) were used instead of oxygen saturation ([%], monthly records). March, April and May situations were inodeled separately or together. Second, weather effects could also be caused by wind. Low wind in spring favours thennal stratification, which is of prime itnportance for YCS (ECKMANN et al., 1988). Strong wind 1nay ease the detrin1ental effect of gas supersaturation by mixing water layers (A. STOCKLI, pers. cmnm.). L. Hallwil is an oligomictic lake due to hills protecting the lake frmn circulation inducing winds (SCHEIDEGGER et al., 1994). Since fall 1985, lake circulation from October until March is technically induced by an aeration system using con1pressed air (STOCKLI & SCHMID, 1987). Therefore, thennal stratification regularly occurs fr0111 April on. Wind force per day was below 3 Beaufort in 1985-1997. According to ECKMANN et al. (1988), wind induced water tnixing occurs above 6 Beaufort. The hypothesis of supersaturation-easing effects of wind was therefore rejected and was not modeled. Fur- thermore, the artificial aeration of L. Hall wil 1nay be thought to act as another 1n01tality factor, because it may theoretically increase oxygen supersaturation. However, WEHRLI & WUEST (1996) found that artificial aeration does not influence the epilimnion of lakes in spring. It was therefore not taken into account as mortality factor, also because weak or nonexistent year classes were recorded even before spring 1986 when the first year class of whitefish larvae occurred at the presence of artificial aeration. Third, low spring te1nperatures also affect zooplankton density (PONTON & ~{OLLER, 1989). This effect was separately modeled using zooplankton density (SilnBal 3.7a-i). The findings of different authors on this topic are controversial. According to ECKMANN et aL (1988), zooplankton concentration during spring had no significant influence on YCS in eutrophic L. Constance. However, NAESJE et al. (1986) found a positive correlation between the density of cladocerans and YCS. Our hypothesis that low {high} food density may cause increased larval n10rtality {survival} is supported by the findings of several authors for different lakes (TAYLOR & FREEBERG, 1984; NAESJE et al., 1986; RICE et al., 1987; PONTON & MtrLLER, 1989; DABROWSKI, 1991). Lack of suitable food causes increased larval 1nortality (AMMANN & STEINMANN, 1948;

124 6.3 Methods

VILJANEN, 1983; ECKMANN, 1985; ROSCH, 1988; PONTON & M1JLLER, l989). Therefore, we tested the effect of changes in densities of different zooplankton organisms: protozoans in March (a), rotifers in March (b) and April (c), herbivorous or carnivorous copepods in April (d/e), phyllopods in April (f), herbivorous or carnivorous copepods in May (g/h) and phyllopods in May (i). In models SimBal 3.7a-i, larval survival increased frmn zero to niaximmn survival rate with increasing zooplankton density similar as for phyllopods in Figure 6.2. But not only quantity could have an effect. Zooplankton quality was presumed to be the cause of up to 90% larval mortality in L. Constance due to smne unknown chemical con1ponents of zooplankton like Cyclops sp. or rotifers (ECKMANN, 1985; Rosen, 1994), which are ingested by small fish larvae (HOAGMAN, 1973; Gurs- SANI & DE BERNARDI, 1977; l-IAMRIN, 1983). It could be thought that the zooplankton got poisoned ingesting unsuitable phytoplankton (SimBal 3.8: in years the effect was observed in L. Constance SpL =0, else SpL=0.04; survival rate between 1967 and 1974 was SrL=0.02 (mean natural survival) due to a lack of data). Phy- toplankton could affect the food of whitefish larvae also in another way: SCHULTZ (1992) found a near-extinction of s01ne cladocerans like Daphnia sp. at high densities of the cyanobacterium Planktothrix rubescens. In smne lakes, daphnids are the tnost frequent organis1ns in the diet of whitefish from May until Septe1nber (MOOKERIJ et al., 1998). A lack of daphnids would result in a delayed growth of whitefish larvae which would attain a size refuge fron1 predation later than under optimal feeding conditions (PONTON & MOLLER, 1989 & 1990). This factor was modeled directly by reducing larval survival with increasing cyanobacteria density (SimBal 3.9: SpL =0.04 up to 0.6g/m2 cyanobacteria density, then decreasing to 0 up to 1.8g/m2 cyanobacteria density).

Sorne c01nbinations of the factors presented above were tested in SimBal 4.1-4.5. The choice was made according to our ecosystern con1prehension. For example, if intraspecific cmnpetition were a mortality factor, it would be likely to occur not only between larvae and adults but also between larvae and larvae (SimBal 4.1). Because whitefish is the dominant and the only pelagic fish species in L. Hallwil in spring, larvae would rather suffer intraspecific competition at high fish density than interspecific cotnpetition. Therefore, intraspecific competition was tested in combination with weather influence (SimBal 4.2) and plankton density (SimBal 4.3). Because weather effects are linked to plankton density, we combined the tnost successful zooplankton tnodel SimBal 3.7i with our weather n1odel SimBal

125 6 A model for evaluating mortality factors in whitefish larvae

3.6 into Sin1Ba1 4.4. Because phyllopod density is not likely to be critically low in eutrophic lakes as we stated before, we avoided a negative food effect by converting phyllopod density to a factor between l and 1.5. Then this factor (extra food, see Figure 6.2) was nrnltiplied with the survival rate modeled by GBS. Finally, if whitefish larvae were affected by perch predation in L. Ha11wil, they would have to stay in the littoral zone where interspecific competition is also high (Si1nBal 4.5).

6.4 Results

Intraspecific (SimBa1 3.1 and 3.2) and interspecific competition (Sin1Ba1 3.3) led to a yield rnininmm, but strong interspecific competition is inconsistent with the increase in whitefish yield fr01n 1967 to 1991 (Figure 6.3). All those 1nodels over- estimate yield before 1985, whereas con1petition among larvae seriously underes- timated the strong 1992 year class (Figure 6.1 and Figure 6.3).

Modeling of perch predation (Sin1Ba1 3.4) resulted in a disastrous stock collapse, but it occurred years before the observed collapse (Figure 6.3). Regressions between modeled and observed YCS or yield were unsatifactory because the r- squared values were low and often not even higher than those of the basic Sin1Bal 2.0 n1odel (Table 6.1). Also the modeling of fishing pressure (Sinillal 3.5) was unsuccessful (Table 6.1 and Figure 6.3).

High spring ternperatures (SimBal 3.6) causing gas supersaturation gave the best fit between real and modeled YCS or yield (Table 6.1 and Figure 6.3). May data, alone or together with March and/or April data, successfully modeled the yield collapse in 1992-94, while March and/or April data did not. Our model could replicate the recovery of yield since 1995, but it was lower than observed (Figure 6.3) - only 30-50 tons were predicted and 50-70 tons were really harvested. This could be ameliorated in the combination model SimBal 4.4. We included a positive influence of phyllopod density (SimBal 3.7i), that was the only scenario frmn SimBal 3.7a-i with satisfactory results (Table 6.1 and Figure 6.4). In SimBal 4.4, high densities of phyllopods increased YCS in 1987 and 1992 and subsequently total yield in 1995-97 (Figure 6.5).

126 6.4 Results

model version YCS YIELD 2.0 +500

3.1

3.2 "O Q3 ·:.;. "Oc: 3.3 (\I (/) (.) >- "O Q) +500 3.4 2:(!) (/I ..Q 0 E 0 3.5 +500 .::: c: 0 :;:: (\I -100 ·::; Q) 3.6 +500 -0..... c: Q) ...(.) 11 Q) .I ••---··••--~ ... 0.. 3.8 +500 k~_Jtl~• ~ 3.9 +500

....------~"'--"""' .t -100 1967 1997 1967 1997

Figure 6.3 Percent deviation of modeled values from observed YCS (left) and whitefish yield (right) for SimBal 2.0, 3.1-3.6, 3.8 and 3.9 (Gray area indicates yield collapse in 1992-94. If values exceed 500 percent deviation, bars cross the graph border).

127 6 A 1nodel for evaluating mortality factors in whitefish larvae

Table 6.1 R 2 value and equation of the regression between real and 1nodeled values of the different SimBal 1nodels. For model version number see rnethods (y =observed YCS, w =observed whitefish yield, bold= r2 values are higher than the r2 value of the basic 111odel SimBal 2.0).

SimJJal 1 YCS whitefish yield

2 model r2 value predicted YCS = 1 value predicted yield = 2.0 0.45 0.47y + 87637 0.29 0.59w + 23654 3.l 0.39 0.42y + 52428 0.37 0.42w + 15895 3.2 0.02 -0.05y + 73885 0.06 0.16w + 17192 3.3 0.00 0.03y + 50835 0.06 O. l4w + 11959 I 3.4 0.25 0.3 ly + 64252 0.15 0.28w + 18358 I 3.5 0.45 0.47y + 87637 0.19 0.50w + 22796 3.6 0.62 0.57y + 32929 0.53 0.52w + 8375 3.7a 0.00 -0.0ly + 37619 0.01 0.05w + 10955

3.7b 0.00 -0.02y + 33935 0.08 O.l4w + 7685 3.7c 0.01 -0.04y + 49538 0.08 0.19w + 11575 3.7d 0.03 0.14y + 60837 0.24 0.45w + 12548 3.7e 0.02 O.lOy + 52827 0.15 0.32w + 12918 3.7f 0.01 0.08y + 62679 0.15 0.36w + 14092 3.7g 0.00 0.04y + 66690 0.20 0.39w + 13184 3.7h 0.01 0.09y + 61224 0.22 0.43w + 12914 i 3.7i 0.74 0.88y + 14802 0.49 0.60w + 7706 3.8 0.48 0.94y - 18756 0.44 0.79w + 651 3.9 0.00 0.03y + 63734 0.14 0.30w + 13616 4. 1 0.18 0.15y + 57311 0.20 0.23w + 16072 4.2 0.50 0.30y + 49509 0.35 0.34w + 13506 4.3 0.37 0.26y + 56189 0.32 0.32w + 15954 4.4 0.72 0.86y 28713 0.58 0.7lw + 8049 4.5 0.12 0.17y + 57274 0.13 0.21w+15158

128 6.4 Results

model version YCS YIELD +500 3 7

. • l .. lU ... J•... L.llJ J .. ___ ..... -... -100 +500 3.7b

l1ll .. __ .... _.... -100 +500 ~ 3.7c Q...... "Cl (ii .I •• ___ .... ___ .... ·s;. -100 "C +500 c: 3.7d ro Cl) u .~k >- .... _____ . "Cl -100 +500 ~ 3.7e Cl) (/) .0 .. ___ ...... __ -.. 0 -100 E 3.7f +500 .g c: 0 :;:; •. ___ ..... _ ro .... • -100 ·:; 3.7g +500 ~ ...c: Cl) •. ______....(.) .... Cl) a.. 3.7h •. ___ .... ___ .. _

3.7i

.J_.. __ .... __ ··· ·100 1967 1997 1967 1997

i;~igure 6.4 Percent deviation of modeled values fron1 observed YCS (left) and whitefish yield (right) for SirnBal 3.7a-i (Gray area indicates yield collapse in 1992-94. If values exceed 500 percent deviation, bars cross the graph border)

129 6 A model for evaluating mortality factors in whitefish larvae

model version YCS YIELD ...... 4.1 rf?...... "C ·:;.Cl) "C c: 4.2 C'l'.l en u 11.J.11. ____ _ > 1•• .1.. .• •... "C Cl> 4.3 500 £:: Q,) f/l .0 0 E 0 4.4 ....I.. c: 0 +:i ·s:C'l'.l Q,) 4.5 "C c: Q,) -(.) -100 c.a; 1967 1997 1967 1997

Figure 6.5 Percent deviation of n10deled values from observed YCS (left) and whitefish yield (right) for SimBal 4.1-4.5 (Gray area indicates yield collapse in 1992-94. If values exceed 500 percent deviation, bars cross the graph border).

The zooplankton quality proble1n (SimBal 3.8) found to be in1portant in L. Con- stance modeled the major collapse in 1992-94 but led to another, earlier decrease at a tin1e when yield actually started to increase strongly (Figure 6.3). Further- more, we could not model a negative effect of Planktothrix rubescens (SimBal 3.9) on herbivorous phyllopod density (Figure 6.3). The combination models SinlBal 4.1-4.3 and 4.5 modeled the yield decrease in 1992-94, but regressions between the predicted and the observed YCS or yield (Figure 6.5) accounted for only 10 to 50 percent of the variation in the mean outcome (Table 6.1 ). This was less than for Simbal 4.4 (Table 6.1).

130 6.5 Discussion

6.5 Discussion

A basic requirement of a population dynainics model is the reliability of data and a reasonable n1odel structure. The basic model SimBal 2.0 was tested using sensi- tivity and error analysis (HANNON & RUTH, 1997). Model predictions depend neither on variations in the initial values (nmnber of stocked larvae) nor on changes in time interval (dt) used during n1odeling. The data used for SimBal 2.0 are quite reliable because the quantity [liters] of artificially bred eggs was n1easured every year and in all hatcheries since 1967. Unfortunately, data reliability of the mortality inducing factors is not co1nparable. Stock size estimates of the other fish species in L. Hallwil from yield statistics are questionable because some of these fish species are conunercially not interesting (SirnBal 3.3). Still, for the heavily exploited fish species whitefish and perch it is acceptable. Oxygen saturation and plankton densities were monitored 1nonthly since 1985 and minutes of sunshine were recorded daily since 1967. Therefore, in Si111Bal 3.7 and 3.9, which were closely related to weather influence, sunshine data were used to 1nodel the years 1967-1984. Fortunately, in this period no collapse of the whitefish stock occurred. Years with zooplankton quality problems in L. Constance were recorded since 1975 (ROSCH, 1994). Overall, continuity and quality of the data used is estimated as satisfactory or even good.

Furthem10re, SimBal 2.0 modeled a steady yield increase fr01111967 to 1997 without a yield collapse in 1992-94 (Figure 6.1 and Figure 6.3). Therefore, if one of the extended 1nodels (SimBal 3.1-3.9) was responsible for weak YCS and thus a yield collapse in 1992-94 and provide a good fit to observed YCS and yield, this would indicate an influence of the modeled factor on survival of the whitefish larvae in nature. But what if more than one factor would give satisfactory results? To answer this question, we estinmted on the basis of literature and our own comprehension of L. Hallwil ecology, how likely each of the tested mortality factors might have occurred. In the following, the hypotheses with the highest likelihood and good n1odeling results (Table 6.1 , Figure 6.3, Figure 6.4 and Figure 6.5) were used as explanation for the yield collapse in 1992-94 in L. Hallwil.

First, we investigated intraspecific competition. Depressed growth is a sign of high fish density (VILJANEN, 1986 & 1988; SALOJARVI, 1992) and was reported by MULLER et al. (1994) for L. Hallwil whitefish. In other whitefish populations,

131 6 A model for evaluating mortality factors in whitefish larvae niche segregation of larvae and adults occurs from hatching in February/March until May-June/July (SANDLUND et al., 1991 ). Therefore, intraspecific cmnpetition between larvae and adults (SimBal 3.1) is probably reduced to late surmner and autumn and may not be of great importance for whitefish larvae. Egg canni- balis1n is reported (SKURDAL, 1985; SALOJARVI, 1987; VENTLING-SCHWANK, 1992), but in L. Hallwil it is not a mortality factor because there is no natural reproduction in whitefish. Cannibalism on larvae might be a possible factor but was not detected in 1998 and 1999 (chapter 5). Thus, cannibalism is very unlikely to occur. Due to unsatisfactory modeling results (Table 6.1 and Figure 6.3), intraspecific competition between adult whitefish and larvae is not an in1portant n1ortality factor but may lead to depressed growth of juvenile whitefish in summer and fall. lntraspecific competition an1ong larvae (SimBal 3.2) also yielded unsatisfactory results (Table 6.1 and Figure 6.3). Because survival was high in 1992, when the number of stocked larvae was high too (Figure 6.1 ), c01npetition ainong larvae is not expected to be an important mortality factor. Regressions between nmnbers of stocked larvae and YCS resulted in a positive relationship but explained only 1 percent of the variation in the mean outcome for NHL, but 50 percent for PL. Therefore, high stocking led not in every case to high YCS. This indicates that other mechanisms rnust affect YCS. Interspecific cornpetition (Siin- Bal 3.3) or even predation (SimBal 3.4) could be in1portant, because L. Hallwil is a eutrophic lake holding abundant fish stocks. Overall, whitefish has been the dominant fish species since 1977 (MULLER, 1993; MULLER et al., 1994). On the average, there are 12 whitefish per perch, 15 vvhitefish per pike and 19 whitefish per other fish species in the catch. According to SVARDSON (1977), coregonids may case interspecific competition by switching to small prey items. Therefore, interspecific co1npetition rnay depress growth but not survival. Predation is 1nore likely to affect whitefish YCS. GERDEAUX & DEWAELE (1986) found significant conclation between perch and coregonid yield with a time lag of 2 years. Regres- sion of coregonid yield against perch yield, taking into account the time lag, gave a positive correlation for L. Hallwil and explained about 50 percent of the variation in the mean outcome. VENTUNG-SCHWANK (1992) found, however, that small coregonid larvae prefer to stay near the surface in the open water where they are relatively secure frmn predation. Our modeling results suggested that perch predation on whitefish larvae is a less important mortality factor (Table 6.1 and Figure 6.3). Furthermore, the effect of fishing (SimBal 3.5) on stock-recruitn1ent

132 6.5 Discussion relationships was low due to artificial breeding and stocking. Therefore it is obvious that the fishing pressure could not be responsible for the yield collapse in 1992-94. In eutrophic lakes, the occurrence of the gas bubble syndrmne (GBS) appears to have the strongest negative effect on whitefish larvae (ShnBal 3.6). GBS repeatedly occurred between 1986 and 1993 in eutrophic Lakes Sen1pach and Baldegg (STADELMANN, 1988 and unpubl. data). Both lakes are within 15 km of L. Hallwil and of comparable trophic state. In L. Hallwil, larvae suffering GBS were observed in several years (M. KELLER, pers. comm.). VENTLING-SCHWANK (1992) considered oxygen supersaturation clue to algal blo01ns to be the most important mortality factor for L. Se1npach whitefish. Occurrence of GBS in eutrophic lakes (MATHIAS & BARICA, 1985) 1nay blur the positive effect of high spring temperatures on larval n1etabolis1n (ECKMANN et al., 1988), larval reaction ability and hunting success (PONTON & M-CTLLER, 1989 & 1990) and, thus, survival. While the 1nodel sitnplified reality, occurrence of GBS in nature is a c01n- plcx matter. VENTLING-SCHWANK (1992) stated that not every case of oxygen supersaturation leads to GBS: One requirement of GBS is that whitefish larvae stay near the surface. This behaviour is age detennined or induced by high algal biomass (VENTLING-SCHWANK, 1992). Another requiren1ent is that not only oxygen but total gas saturation and/or pH is significantly elevated. All those effects were omitted in the 1nodel, for the sake of sin1plicity. As stated above, every population dynamics 1nodel is a simplification of reality and cannot explain every inechanis1n. Therefore, due to good modeling results and high probability of occurrence, we identified GBS as the main mortality factor for L. Hallwil whitefish larvae.

Due to the eutrophic state of L. Hallwil, zooplankton concentration (Sim.Bal 3.7) in spring is quite unlikely to be a critical n1ortality factor. ECKMANN et al. (1988) reported no significant influence of zooplankton densities on YCS in eutrophic L. Constance. Still, results of SimBal 3.7i compared extremely well to observed YCS and yield (Table 6.1 and Figure 6.4). Therefore, we concluded that there has to be an effect, but because eutrophic lakes are hardly food li1nited, it is more probable that the effect is rather positive than negative. The zooplankton quality problem (SimBal 3.8) reported by ECKMANN (1985) and ROSCH (1994) for L. Constance has never occurred in L. Hallwil hatcheries (E. FISCHER & M. KELLER, pers. c01nm. ). Although modeling results were quite satisfactory (Table 6.1 and Figure 6.3), this is not likely to be a factor in L. Hallwil. And although the cyano-

133 6 A model for evaluating mortality factors in whitefish larvae bacteria model (SimBal 3.9) gave a quite good fit (Figure 6.3), the regression results were bad (Table 6.1 ). A direct regression of the density of cyanobacteria, which were dominated by P. rubescens in L. Hallwil (BDRGI, pers. com1n.), plot- ted against phyllopod concentration explained only 5 percent of the variation in the mean outcmne. Thus, our last hypothesis is quite unlikely to be of i1nportance.

SALOJARVI (1987 & 1988) concluded that if environmental conditions are suitable, density-dependent population mechanisms play a major role in determining yield from stocking. However. environmental conditions in eutrophic L. Hallwil have not been suitable for whitefish eggs and larvae since about 1960 (MlTLLER et al., 1994). Therefore, we estimated fish density-independent mortality factors to be inore important than density-dependent factors. Thus, GBS is likely to be the main mortality factor for L. Hallwil whitefish larvae and high phyllopod densities favoured strong year classes. These two combined scenarios (Simbal 4.4) gave results that fitted best with observed YCS and yield (Table 6.1 and Figure 6.5). For both factors, conditions during May were found to be most i1nportant. How- ever, very strong year classes as in 1992 can only develop when the weather in May is as warn1 as possible to enhance phyllopod development, but at the same time as inoderate as possible to avoid oxygen supersaturation and thus GBS. The probability for this to occur is quite low, and therefore strong year classes as in 1992 are the exception. The basic requirement is of course that high numbers of larvae are stocked. So, the SimBal 4.4 model may help to optilnize the stocking practice, thereby increasing the chance of the stocked larvae to grow up. PL were stocked at the end of April or in the beginning of May. Therefore, two options for improvement of the stocking practice are available: First, stocking of PL could be delayed if Secchi depth is less than 1.5 meters. VENTLING-SCHWANK (1992) found no OBS in L. Sempach in 1989 and 1990 when Secchi depth was inore than 1.5 meters. This parameter allows a simple estimation of algal biomass and therefore risk of GBS. Second, prefed larvae could be released stepwise during May to minimize risk of total loss caused by GBS.

Although the trophic state of L. Hallwil is on a steady decrease, OBS is likely to continue for some years. Therefore, our model could be used to atte1npt forecast- ing YCS and yield, based on the nun1ber of larvae stocked, and by using sunshine minutes, phyllopod density and lake oxygen data. Furthermore, data from other lakes of similar state as L. Hallwil could be tested. Once natural reproduction of

134 6.6 Acknowledgements

L. Hallwil whitefish is again possible. our model will need an update to take into account natural reproduction: stock-recruitment relationships, egg predation and egg cannibalistn would then become important.

6.6 Acknowledgements

Thanks to SMA Meteo Schweiz for providing long-term weather records, to Dr. H. R. Bi.irgi, Dr. H. Bi.ihrer, G. Ribi and G. Meier for infonnation and records on plankton, chen1istry and physics ofL. Hallwil, to Prof. H. U. Fuchs, Dr. M. Simon, Technical College (TWI) Wintertlrnr, and to many colleagues fron1 the system. dynamics courses in Fribourg and Winterthur for interesting discussions. We thank Dr. W. Donni, AquaPlus, for his help to in1prove the tnodel structure. Spe- cial thanks to Fl. Minder and Dr. A. StOckli, Aarau, for providing fisheries records and for helpful discussions on fisheries and chemistry ofl.,. Hallwil. We also thank the Finanz- and Baudeparte1nent Aargau for financial support. The constructive comments by Prof. J. V. Ward, Dr. W. Dl1nni and two anonymous reviewers helped to improve the manuscript.

6. 7 References

AASS, P. (1972): Age determination and year-class fluctuation of cisco, Coregonus albula L., in the Mj0sa hydroelectric reservoir. - Report Institute of Freshwater Research (Drottninghohn) 52: 5-22.

ALFELD, L. E. & GRAHAM, A. K. (1976): Introduction to urban dynan1ics. - Wright-Allen Press, Inc. Cambridge, Massachusetts.

AMMANN, E. & STEINMANN, P. (1948): Die Verbesserung der Methoden in der Felchenzucht. Konunission fitr die Erforsclrnng fischereiwirtschaftlicher Fragen. Sonderbericht Nr. l. - W. KUNZ, Pfriffikon-Zi.lrich.

BAGENAL, T. B. (Ed.) (1974): The ageing offish. Proceedings of an International Symposimn held at the University of Reading, England, 19-20 July 1973. - The Gresham Press. Surrey. England.

135 6 A model for evaluating mortality factors in whitefish larvae

BERG, A. & GRIMALDI, E. (1966): Ecological relationships between planktophagic fish species in the Lago Maggiore. Verh. Int. Verein. Limnol. 16/1-2: 1065-1073.

BOUCK, G. R. (1980): Etiology of gas bubble disease. Trans. - An1. Fish. Soc. 109: 703-707.

COSTANZA, R., DUPLISEA, D. & KAUTSKY, U. (1998): Ecological Modeling and econornic syste.ms with STELLA. - Ecological Modeling 110/l: 1-4.

DABROWSKI, R. (1991): Growth and feeding of Coregonus peled G1nel. larvae in illuminated cages. Pol. Arch. Hydrobiol. 38/3-4: 463-474.

ECKMANN, R. (1985): Histopathological alterations in the intestine of whitefish (Coregonus sp.) larvae reared on zooplankton frmn Lake Constance (West Gennany). - Diseases of Aquatic Organisms 111: 11-18.

ECKMANN, R., GAEDKE, u. & WETZLAR, H. J. (1988): Effects of climatk and density-dependent factors on year-class strength of Coregonus lavaretus in Lake Constance. - Can. J. Fish. Aquat. Sci. 45/5: 1088-1093.

GERDEAUX, D. & DEWAELE, P. (1986): Effects of the weather and of artificial propagation on coregonid catches in Lake Geneva. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22. 343-352.

GllJSSANI, G. & DE BERNARDI, R. (1977): Food selectivity in Coregonus sp. [zooplanktophagous fish] of Lake Maggiore [Italy], an energetical approach. Me1norie dell1Istituto Italiano di ldrobiologia, Dott. Marco de Marchi Verbania Pallanza (Italy) 34. 121-130.

GOODYEAR, C. P. (1980): Compensation in Fish Populations. - Biological monitoring of fish: special session at the 1 lOth Annual meeting of the American Fisheries Society (Louisville) 11: 253-280.

HAMRIN, S. F. (1983): The food preference of vendace (Coregmms albula) in South Swedish forest lakes including the effect on zooplankton populations. - Hydrobiologia 101: 121-128.

HANNON, B. & RUTH, M. (1997): Modeling dynarnic biological systen1s. - Springer, New York, Berlin, Heidelberg.

136 6. 7 References

HOAGMAN, W. J. (1973): The hatching, distribution, abundance, growth and food of the larval whitefish (Coregonus clupeafonnis Mitchill) of Central Green Bay, Lake Michigan. - Report Institute of Freshwater Research (Drottningholm) 53: 1-20.

JURVELTUS, J. (1991): Distribution and density of pelagic fish stocks, especially vendace (Coregonus albula (L.)), 1nonitored by hydroacoustics in shallow and deep southern boreal lakes. Finn. Fish. Res. 12: 1-19.

KOTTELAT, M. (1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the former USSR), with an introduction for non-systematists and comments on nomenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

MAY, R. C. (1974): Larval mortality in marine fishes and the critical period concept: 3-19. The Early Life History of Fish. - BLAXTER J. H. S. (Ed.), Springer Verlag, Berlin.

MATHIAS, J. A. & BARICA, J. ( 1985): Gas supersaturation as a cause of early spring inortality of stocked trout. - Can. J. Fish. Aquat. Sci. 42/1: 268-279.

MOOKERIJ, N., HELLER, C., MENG, H.J., BT)RGI, H. R. & M1JLLER, R. (1998): Diel and seasonal patterns of food intake and prey selection by Coregonus sp. in re-oligotrophicated Lake Lucerne, Switzerland. - J. Fish Biol. 52/3: 443-457.

MtJLLER, R. (1990): Manage1nent practices for lake fisheries in Switzerland. 477- 492. Manage111cnt of freshwater fisheries. VAN DENSEN, vV. L. T., STEINMETZ, B., HUGHES, R.H., (Eds), Pudoc. Wageningen.

M(JLLER, R. (1992): Trophic state and its implications for natural reproduction of salmonid fish. - Hydrobiologia 243-244: 261-268.

MULLER, R. (1993): Einige fischereibiologische Aspekte von Seesanierungen. Fortschr. Fisch.wiss. 11: 43-56.

MULLER, R., MBWENEMO BIA, M. & MENG, H.J. (1994): Die Felchenfischerei in

einigen~ Seen der Zentralschweiz und des Mittellandes. Mitteilungcn~ zur Fischerei Nr. 55. Bundesmnt filr U1nwelt, Wald und Landschaft (BlJWAL):

137 6 A model for evaluating mortality factors in whitefish larvae

NAESJE, T. F., SANDLUND, 0. T. & JONSSON, B. (1986): Habitat use and growth of age-0 whitefish, Coregonus lavaretus, and cisco, Coregonus albula. Enviromnental Biology of Fishes 15/4: 309-314.

PONTON, D. & MOLLER, R. (1989): Alimentation et facteurs de nlortalite des larves de coregones ( Coregonus sp.): Exe1nple de deux lacs de niveaux trophiques differents, les lacs de Samen et de Hallwil (Suisse Centrale): - Aquatic Sciences 5111: 67-83.

PONTON, D. & MOLLER, R. (1990): Size of prey ingested by whitefish, Coregonus sp., larvae, Are Coregonus larvae gape-limited predators? - J. Fish Biol. 36/l: 67 -72.

RICKER. W. E. (1975): Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Board Can. 191: 382 pp.

ROSCH, R. (1988): Mass rearing of Coregmws lavaretus larvae on a dry diet. Finn. Fish. Res. 9: 345-35 l.

ROSCH, R. (1994): Probleme bei der Aufzucht von Felchenlarvaen (Coregonus lavaretus) rnit Lebendplankton. European Association of Fish Pathologists: 88-91.

SALOJARVI, K. (1987): Why do vendace (Coregonus albula L.) populations fluctuate? - Aqua Fennica 17/l: 17-26.

SALOJARVI, K. (1988): Effect of the stocking density of whitefish ( Coregonus lavaretus L. s. l.) fingerlings on the fish yield in Lake Peranka, Northern Finland. - Finn. Fish. Res. 9: 407-416.

SALOJARVT, K. ( 1991): Stock recruitment relationships in the vendace ( Coregonus albula L.) in Lake Oulujarvi, northern Finland. - Aqua Fennica 21/2: 153- 161.

SALOJARVI, K. (1992): Compensation in whitefish (Coregonus lavaretus L. s.l.) populations in Lake Oulujarvi, northern Finland. Finn. Fish. Res. 13: 31- 48.

138 6.7 References

SANDLUND, 0. T., JONSSON, B .. NAESJE, T. F. & AASS, P. (1991): Yem·-class :fluctuations in vendace, Coregonus albula (Linnaeus): Who's got the upper hand in intraspecific competition? - J. Fish Biol. 38/6: 873-886.

SCHEIDEGGER, A., STbCKLI, A. & W-OEST, A. (1994): Eintluss der internen Sanienmgsmassnahrnen auf den Sauerstot1lrnushalt iln Hallwilersee. - Wasser, Energie, Luft 86/5-6: 126-131.

SCHULTZ, H. ( 1992): Stock density, growth rate and zooplankton consmnption of the vendace (Coregonus albula) and other fish species in Lake Arendsee. - Limnologica 22/4: 355-373. SKURDAL, J., BLEKEN, E. & STENSETH, N. c. (1985): Cannibalisrn in whitefish (Coregonus lavaretus): - Oecologia (Berlin) 67/4: 566-571.

STADELMANN, P. (1988): Zustand des Sempachersees. - Wasser, Energie, Luft 80/3-4: 81-96.

STbCKLT, A. & SCHMID, M. (1987): Die Sanienmg des Hallwilersees. Erste Erfahrungen mit zwangszirkulation und Tiefenwasserbeliiftung. - Wasser, Energie, Lun 79/7-8: 143-149.

SVARDSON, G. (1977): Interspecific Population Dmninance in Fish Cmrununities of Scandinavian Lakes. - Rep. Inst. Freshw. Res. (Drottninghohn) 55: 144- 171.

TAYLOR, W.W. & FREEBERG, M. H. (1984): Effect of food abundance on larval lake whitefish, Coregonus clupeafonnis Mitchill, growth and survival. J. Fish Biol. 25: 733-741.

VENTLTNG-SCHW ANK, A. R. (1992): Reproduktion und larvale Entwicklungsphase der Felchen (Coregonus sp.) in1 eutrophen Sempachersee. - Dissertation an der Philosophischen Fakultiit der Universitiit Zurich.

VILJANEN, M. (1983): Food and food selection of cisco (Coregonus albula L.) in a dysoligotrophic lake. Hydrobiologia 101: 129-138.

VTLJANEN, M. (1986): Biology, propagation, exploitation and management of vendace (Coregonus albula L.) in Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 73-97.

139 6 A model for evaluating mortality factors in whitefish larvae

VILJANEN, M. (1988): Population dynamics of vendace (Coregonus albula L.) in Finland. - Joensuun Yliopiston Luonnontieteellisia Julkaisuja 0/12: l-J 9.

WEHRLI, B. & WtJEST, A. (1996): Zchn Jahre Seenbeltiftung: Erfahnmgen und Optionen. - Schriftenreihe EAWAG 9.

WEITKAMP, D. E. & KATZ, M. (1980). A review of dissolved gas supersaturation literature. - Trans. A1ner. Fish. Soc. 109: 659-702.

140 7.1 Abstract

7 FACTORS CONTROLLING GROWTH AND MATURITY

7.1 Abstract

Fish growth and age at maturity are indicators of health because they integrate all biotic and abiotic factors that affect the organis1n. In eutrophic lakes, living conditions for fish 1nay vary frmn very favourable to adverse, depending on the fish species and on the degree of eutrophication. Using whitefish ( Coregonus suidteri) of Lake Hallwil, Switzerland, factors controlling growth and maturity of fish in this eutrophic lake were studied. Whitefish has been the d01ninant and corniner- cially most important fish species in Lake Hallwil since 1977. Between 1980 and 1982, whitefish exhibited high growth rates and attained mean lengths frmn 217 to 237 1nm already at one year of age. Although Lake Hallwil renrnined eutrophic, growth of whitefish decreased significantly between 1983and1988. Young of the year (YOY) whitefish reached a inean length of only 150 nun in 1988. From 1988 to 1996, mean YOY length at the end of the year varied between 150 and 175 mm. Age at maturity of males and females increased frmn two years during 1980-1991 to three years thereafter. Between 1980 and 1998, zooplankton density remained high although n1ean phosphorus concentration decreased fro1n 220 to 50 µg/l. Food shortage is therefore excluded as a major factor affecting growth and maturity of whitefish in this lake. Because the whitefish population size increased greatly after 1982, and because other growth-controlling factors were not found to exert a noticeable effect, we conclude that growth of Lake Hallwil whitefish was pri111arily controlled by population density.

7.2 Introduction

Growth, i.e. the rate of increase in length and weight, is one of the ultimate indicators of health and condition in fish because it integrates all abiotic and biotic fac-

141 7 Factors controlling growth and maturity

tors affecting the organism. Fast growth reflects good overall living conditions, while slow growth is usually associated with unfavorable conditions. Poor growth may also reflect chronic stress (LECREN, 1972; WATERS, 1977; LARKIN, 1978). Growth is a key parameter in studying population dynamics (BOWEN et al., 1991) and stress in fish (GOEDE & BARTON, 1990). Growth rate is controlled by a number of factors. Several authors found fish density to be the most important factor influencing growth: According to SALOJARVI (1992), SCHULTZ (1992) and MtJLLER et al. (1994), growth rate is usually low at high stock density. HENDER- SON & BROWN (1985) and VILJANEN (1986, 1988) stated that strong year classes grow slower than weak ones. Other authors found that fish density-independent factors like ten1perature or trophic state may be responsible for different growth rates in whitefish. Growth increase with increasing temperature is reported by AASS (1972), SARVALA et al. (l 988), PONTON & MULLER (1989) and GRIFFITH et al. (1992). MILLS & CHALANCHUK (l 988), PONTON & M1JLLER (1989), KIRCH- HOFER (1995) and MOLLER & BIA (l 998) found that growth is faster in eutrophic than in oligotrophic lakes. Differences in zooplankton density were thereby identified as the primary cause (SVARVAR & MfJLLER, 1982; MlJLLER, 1990; KIRCH- HOFER, 1995).

Lake Hallwil (Switzerland) is a eutrophic lake with a surface area of l 0 square kilometers and a mean depth of 28 111 (SCHEIDEGGER et al, 1994). Most eggs of whitefish (Coregonus suidteri Patio, nomenclature according to KOTTELAT, 1997) die during embryonic develop1nent in the lake because of the anoxic sediment (VENTLING-SCHWANK, 1992; MtlLLER, 1992, VENTLING-SCHWANK & LIVING- STONE 1994). In the fishery this proble1n has been circumvented by incubating whitefish eggs in hatcheries. Thanks to artificial breeding and stocking, whitefish has becon1e the dominant and c01nmercially most in1portant fish species in Lake Hallwil after 1977 (MfrLLER et al., 1994). Whitefish growth was extre1nely fast during the 1nost eutrophic period of Lake Hallwil between 1980 and 1982, but slowed down considerably between 1983 and 1988. Growth has remained ]ow until 1997 (Figure 7 .1 ). In view of these changes in growth, fishing with nets of constant mesh size during the whole period of 1980-1998 111ay not have been optimal at all times.

142 7 .2 Introduction

600 - growth: total length [mm] -o- phosphorus concentration [µg/I] 500 -- population size [1O3 fish] c:::::J whitefish yield [1 o3 fish)

400 - YCS [1 o3 fish] 3rd year

300

200 -- -- -

0 (j) .,..... ("') l{) I'- (J) .,..... ("') l{) I'-

Figure 7.1 Long-term changes in phosphorus concentration and growth, population size, year-class strength and yield of whitefish in Lake Hallwil.

To identify the factors controlling growth and maturity of whitefish in eutrophic Lake Hallwil, the following alternative hypotheses were tested:

l. Growth is independent of fish density. The main factors controlling growth are variations in lake temperature and phosphorus concentration and, thus, zooplankton density. 2. Growth is density-dependent. Population size and/or year class strength (YCS) of whitefish are the p1imary factors controlling growth.

Because factors influencing growth may also affect fecundity and age at maturity, these parameters were studied as well. SALOJARVI (1987) found that in eutrophic lakes and at ]ow fish density fish reach maturity earlier than in oligotrophic lakes

143 7 Factors controlling growth and 1naturity and at high fish density. ALM (1959), COLBY (I 984) and MUNKITI'RICK & DIXON (1989) fmther stated that growth rate, body condition and fecundity are all tightly linked. Therefore, condition index, fecundity and egg quality (caloric content) were taken into account. Egg quality and fecundity also depend on fish population density and food concentration. At low fish density, females produce 1nore eggs of higher quality than at high density (SALOJARVI. 1987). Egg quality is ecologi- cally significant because it influences larval size at hatching, resistance to starvation and overall health (RICE et al., 1987; BROWN & TAYLOR, 1992) and thus survival of a year class (VIUANEN, 1986).

7.3 Methods

7.3.1 Environmental data

Growth of young of the year (YOY) whitefish was compared by regression to the following parmneters: sunshine hours during the 1nain growing season (sum of May to Septe1nber, [hours per growth period]), phosphon1s concentration (circulation value of January, [µg/l]) and inean phyllopod density during the main growing season (middle of May to September, [g/in2 per growth period]). Because water te1nperature data were not available continuously for the whole study period, sunshine data were used to estimate temperature differences for the years 1981-1996. Phosphorus concentrations had been n1easured since 1959, while data on phy11opod density were available from 1986 on.

7 .3.2 Fish data

Annual records of whitefish length, weight and age at maturity were available for years 1981-1982, 1986-1992 and 1994-1999. Despite the lack of data fron1 smne years) annual growth rates for year classes 1980-1997 could be reconstructed using scales for back-calculating whitefish length. Annuli on scales were measured to the nearest 0.1 mm using a WILD binocular microscope at 20x magnification. Differences in growth were tested by analysis of variance (ANOVA) at p<0.05 (LOZAN, 1992). Growth of YOY whitefish was compared by regression to stock size and year class strength (YCS) of whitefish [number of fish]. Year class strength was calculated by virtual population analysis (VPA) for the years 1968- 1997. Year class 1997 was still inco111plete when this study was terminated and

144 7.3 Methods has not been included in the regressions. Fishing yield [kg/year] was available from 1nandatory reporting logs of the gillnet fishery for the years 1967-1999. Because mean weight of the whitefish caught in 1980-1999 was 0.33 kg~ yield was multiplied by 3 to convert weight into nun1bers of fish catch). Thus, whitefish population size in year t [nun1ber of fish in the lake] was calculated using the formula

Population(t) = Population(t-l) + (YCS(t) - Catch(t)). (7.1)

Initial population size (to) for the year 1967 was calculated frmn the three million newly hatched larvae stocked in 1967. Assun1ing a survival rate frmn larval to adult fish of about 0.004 (Mt'J:LLER et al., 1994; chapter 6)~ the stocking in 1967 produced about 12,000 adult whitefish. Because population size in the previous year was not likely to be zero, the size of the 1967 population was empirically set to 30,000 whitefish.

Fulton's (1904) condition index CI of the whitefish fro1n the years 1981-82, 1986- 92 and 1994-96 was calculated according to the formula

CI= (weight [g] * 100) I (body length [cm])3 . (7.2)

Age at maturity was assessd by checking our records for the maturity stage of 1+ and 2+ whitefish in autumn.

Fecundity and egg quality was determined on 55 fe1nale whitefish caught at the end of October in 1997 and 1998 using gillnets of bar 1nesh sizes between 20 and 40 nllll. In 1997 (n = 36), 6% of the fish were two years old, 3% three years, 3% four years, 85% five years and 3% six years old. In 1998 (n = 19), 26l71c) were two years old, 5CJ;1 three years, 16% five years and 53% six years old. Age was determined using scales. Total weight, and total and body length of the fish were measured to the nearest 1 g and l 1run, respectively. The gonads were re1noved to measure total gonad weight to the nearest 2 g and then frozen for later investigations. Frmn one of the two gonads per female, three subsainples of 100 eggs were taken. One subsainple was taken frmn the oral part, one frmn the iniddle and one from the caudal part of the thawed gonad according to MOOKERJI (pers. comm.). Egg dia1nctcr of twelve eggs per female (four per subsa1nple) was 1neasured using a WILD binocular microscope at 40x magnification. Wet weight of the three sub-

145 7 Factors controlling growth and inaturity samples was then determined to the nearest 0.1 1ng. Absolute fecundity (total nmnber of eggs per fe1nale) was calculated taking into account total gonad wet weight and wet weight of 300 eggs. Relative fecundity was calculated by dividing total egg number by total fe1nale body wet weight.

Egg quality was detennined in 1997 (large population size relative to 1998) and 1998 (low population size relative to 1997). The caloric content of dlied eggs was determined using an oxygen bmnb calori1neter. Eggs fron112 whitefish were thor- oughly mixed and then dried at 1OS°C for 24 hours. The ripe whitefish females in 1997 were five years old, while those in 1998 were two years old. The values of 1997 and 1998 were compared by analysis of variance (ANOVA) at p<0.05 (LOZAN, 1992).

7.4 Results

7 .4.1 Environmental effects

The difference in growth of YOY in 1981-82 (before growth decline) and 1990- 96 (after growth decline) was highly significant (p<0.001, Table 7.1 and Figure 7.1 ). Even the small difference between 1990 and 1996 was highly significant (p<0.001, Table 7.1 and Figure 7.1).

Regression between growth of YOY whitefish and sunshine hours accounted for only 25% of the variation in the mean outcome (Figure 7.1, Figure 7.2 and Figure 7.3). The regression trendline shows an increase of growth with decreasing sunshine hours for 1981-1996. Growth of YOY and phosphorus concentration decreased during the same time period (Figure 7. l ). The regression accounted for 64% of the variation in the mean outcome (Figure 7.3). Regression between growth and phyllopod density was less definitive: Growth shows a s1nall decrease with decreasing phyllopod density (Figure 7.3).

146 7.4 Results

,...... , ,...... , N_§ ' §: •• )K ••• • 1000 CJ) ....__, 200 .._. c -----~-* 0 :t::>- :;:: (/) m •• 800 c .;:. - ::1 ...c: l- 400 o.. 0 (1) ...c: - •sunshine c c a. .c. cu (/) 50 1 {j) Q.) 0 * mean phyllopod density 200 c ::I E -§_ _.,.9.., Phosphorus cone. {j) 1 0+-----+-----+-----+------+0 1980 1985 1990 1995 2000

Figure 7.2 Evolution of total phosphorus concentration, sunshine, and phyllopod density during the main growing season (May-September) in L. Hallwil, years 1986-1996.

Regression between growth of YOY whitefish and sunshine hours accounted for only 25 % of the variation in the mean outcome (Figure 7 .1, Figure 7 .2 and Figure 7.3). The regression trcndline shows an increase of growth with decreasing sunshine hours for 1981-1996. Growth of YOY and phosphorus concentration decreased during the same time period (Figure 7.1). The regression accounted for 64% of the variation in the mean outc01ne (Figure 7.3). Regression between growth and phyllopod density was less satisfactory: Growth shows a small decrease with decreasing phyllopod density (Figure 7.3).

7 .4.2 Fish population effects

Regression between growth and calculated population size accounted for only 14% of the variation in the inean outcome (Figure 7.3). However, the trendline shows a decrease of growh with increasing stock size (Figure 7.3). The regression

147 7 Factors controlling growth and niaturity between growth and YCS accounted for only 0.2% of the variation in the inean outcome and showed no growth variation consistently associated with strong or weak year classes (Figure 7.3). 1.50 800000 y=0.001x+1.2 y =-2295x + 719924 2 R2 = 0.5 R = 0.14 600000 ~ 1.45 Ill s '"O II ()) c N II '(ii § '1.40 400000 c E .Q "O c m 0 3 u '1.35 200000 n. ~II 0 0.. Ill 1.30 Ill 0 1100 • 800000 1000 (/) • ,_ ' .A 600000 ::; _g 900 y -223x + 150112 (1J 2 s c R = 0002 400000 ([) -ffi 800 u c .A >- ::J A (/) • • 200000 700 y =-1.9x + 1245 1,;.. • R2 0.25 .A ~ 600 t"i A 0 300 300 N"' y =0.3x + 93 y = 1.7x 169 E 250 2 ....._ R2 0.004 R = 0.6 0) ; 200 200 ·;;:; ~ 150 "O 0 - c 100 -ga.. 100 8 0 (L 50 ..c>. 0.. 0 0 150 200 150 200 first year growth [mm] first year growth [mm]

Figure 7.3 Regressions of YOY growth versus condition index, and YOY growth versus five parameters that were tested as potential determi- nants of first year growth of whitefish.

148 7.4 Results

Condition index CI was 1.45 when growth was fastest but decreased with decreasing growth. In 1990 and 1991 (when whitefish stock size, YCS and also growth were low) CI reached values above the n1ean value for 1981-1996 (Table 7.1). The regression between growth of YOY and CI accounted for 50% of the variation in growth (Figure 7.3).

Until about 1991, inaturity was attained already at the end of the second year, even in females. After 1992, age at niaturity generally increased by one year in both sexes (Table 7.3).

No differences in fecundity of the same age class were found in 1997 and 199 8 (Table 7.2 and Figure 7.4). Six year old whitefish have significantly higher absolute fecundity than younger fish (Table 7 .2 and Figure 7.4 ). Caloric content (Figure 7.4) was significantly lower (p<0.01) h1 1997 (5+ fish) than in 1998 (2+ fish).

Ii absolute fecundity o relative fecundity ~caloric content 40000 100 ...... O>

(/) 30000 75 ~ ...... Q) >- ."t: 20000 50 -0 c :::; u ~ 10000 - 25 Q) .2:..... cu Q) 0 0 .._ 2+ 5+ 2+ 5+ 6+ 5+ 2+ 1997 1997 1998 1998 1998 1997 1998 age I sampling year

Figure 7.4 Absolute and relative fecundity of fe1nale whitefish in 1997 and 1998, and caloric content of whitefish eggs in 1997 and 1998.

149 7 Factors controlling growth and maturity

7 .5 Discussion

7 .5.1 The role of environmental factors

Phosphorus concentration and weather indirectly affect whitefish growth via food quantity and availability (SVARVAR & MULLER, 1982; PONTON & MtTLLER, 1989; MOLLER, 1990; KIRCHHOFER, 1995). According to NAESJE et al. (1986), strong year classes often correlate with high density of cladocerans (Crustacea: Phyl- lopoda). Cladocerans also dmninate the phyllopod cmrnnunity of Lake Hallwil. Because of their abundance, and because they are highly selected by whitefish (chapter 5), phyllopods were expected to affect whitefish growth. The period of May to September was found to be the 1nain feeding period of Lake Hallwil white- 2 fish (chapter 5). Mean phyllopod density [g/m ] and weather situation [sunshine hours] were therefore taken into account for this period. Phyllopod density showed a s1nall decrease over the period of 1981-1996, while sunshine hours increased slightly over the same period (Table 7.1, Figure 7.1 and Figure 7.2). Because growth decreased with increasing sunshine hours (Figure 7.3) - a result inconsistent with the findings of PONTON & MffLLER (1989) -, we conclude that sunshine is not likely a factor influencing Lake Hallwil whitefish growth in the years 1981-1996.

Quite unexpectedly for a eutrophic lake, growth of whitefish decreased with decreasing phosphorus concentration (Figure 7.3). MOLLER et al. (1994), MOLLER & BIA (1998) and MAYR (1998) found a decline in growth of coregonids associated with decreasing phosphorus concentration, but only in oligotrophic lakes with total phosphorus concentration near or below 10 µg/l. In 1998, Lake Hallwil was still eutrophic at about 40 µg/l total phosphorus (Figure 7.1). Lake Hallwil whitefish is able to switch its diet from zooplankton to benthic organisrns (chapter 5). The density of the latter had increased enormously since 1986 (STOS- SEL, 1992), i.e. since the beginning of artificial aeration. Even in re-oligotrophicated Lake Lucerne (total phosphorus 5-6 µg/l), MICHEL (1996) found no imme- diate growth decline in large-type whitefish, as compared to the preceding mesotrophic period. These large-type whitefish, like the ones in Lake Hallwil, are able to prey on both benthic and zooplankton organisms. \Ve therefore conclude that the decrease in phosphorus concentration observed so far cannot be the cause for the growth decline of Lake Hallwil whitefish.

150 7 .5 Discussion

Table 7.1 Parameters which are suspected to have influenced growth and condition of L. Hallwil whitefish. Values with possibly negative effect on growth are in bold; values indicating low growth and low condition are also in bold. Phyll. =mean biomass (wet weight) of phyllopods frmn May to Septernber, 0 2 sat.= oxygen saturation rneasured in May, YCS = year class strength, YOY young of the year white-· fish. 0 =mean of years 1981-1996.

~ ~ ~ ~ '-l '-l g. ~ ~ ~ t.I'.) ...:::; ,..:;: ~ • g. ""'1 ""1 ~ t.I'.) ~8' ~8' .... s I ~ ..... ~ ...... ~~ ~~ ~~ .... Year ~ ~ ~~~ ~ s ~ ...'-l ~., ... .§ R' :::: I;)' 8' '-l ~"' ..::::: '-l ~ ..... ""'-! ~ ...... ~8' ~ ~ ·~ ..t::) ~ ..t::) ~ .-::::: \"f") ~ c--, ~ ...:.:: O'\ .... """I ..:c! E \"f") E ""'-! ~~~ ~~ ~ -., II ~~II ~ ,..._, II t..i ~ II ~ II ~ E II ::;: II :::,..:;:~ ..::: 0.0~ -S .... ~ U V') -.....:. '--" ~ -...... :; '--" ~~~...... '--" IJ) ..:::::. '--" ~~~ c3~~ a~ 1981 747 no data no data 100,000 60,000 233 1.45 1982 901 no data 121 100,000 40,000 237 1.45 1983 9031 no data 124 300,000 200,000 208 no data 1984 805 no data 152 300,000 60,000 212 no data 1985 1.015 142 151 250,000 0 no data no data 1986 970 176 209 200,000 40,000 184 1.40 1987 790 209 116 300,000 200,000 172 1.40 1988 923 134 146 300,000 90,000 151 1.33 1989 1,025 110 161 300,000 80,000 160 1.34 1990 1,057 122 140 150,000 40,000 165 1.38 1991 1.036 122 140 50,000 20,000 158 1.42 1992 963 94 100 600,000 600,000 172 1.33 1993 749 124 143 600,000 20,000 171 no data 1994 993 196 142 550,000 10J)00 158 1.33 I 1995 877 1371 133 450,000 40,000 I 1711 1.33 ' I 1996 881 ssl 116 400,000 100,000 I 175 1.341 l·---·~--

151 7 Factors controlling growth and maturity

In Lake Hallwil, the c01nbination of eutrophication and sunny weather may still lead to algal bloon1s and thus to oxygen supersaturation in spring (MATHIAS & BARICA, 1985). Furthennore, in late summer and fall, a pronounced n1etahnmetic oxygen rninimmn (AKU et al., 1997) may develop due to high algal production and sedimentation. This may represent a severe stress factor for whitefish and may even be the cause of retarded growth. Prolonged periods of sunny weather are a prerequisite for oxygen supersaturation and metalimnetic oxygen ininin1um to occur in eutrophic lakes. If this process were the cause for slow growth in whitefish, then the pattern of fast and slow growth would not be a steady decrease but rather an up and down of year-to-year growth rates. We therefore exclude eutrophication-induced stress related to dissolved oxygen as a possible cause for general growth retardation.

7.5.2 The role of population-related factors

Published results on the effect of YCS on growth are not consistent: SANDLCND et al. (1991) found that the size of fish from weak and strong year classes did not differ. Contrasting results were published by VILJANEN (1986, 1988). We found that first-year growth of Lake Hallwil whitefish did not vary with YCS (Table 7.1 and Figure 7 .3 ). We thus conclude that YCS does not noticeably affect growth of YOY whitefish in Lake Hallwil. Stock size. other than YCS, was described by various authors to have an effect on growth (AASS, 1972; VILJANEN, 1983). In Lake Hallwil, growth of whitefish tends to decrease with increasing stock size. This is consistent with the findings of SALOJARVI (1992), SCHULTZ (1992) and MliLLER et al. (1994). However, stock size accounted for only 14% of the variation in growth of YOY whitefish (Figure 7 .3 ), but it scored a still better r-squared value than the regression of phyllopod density versus YOY growth ( 4%, Figure 7.3).

Based on an overall assessment of the various factors, we conclude that stock size, or rather population size, is probably the main factor controlling growth of whitefish in this lake. Neve1theless, the r-squared value was low, and growth was slow in 1990 and 1991 when stock size was also low. Other factors are therefore likely to influence growth as well. The combination of slow growth and low stock size in 1990/91 occurred just before the yield collapse in 1992-94 (Figure 7 .1 ), which was found to be the result of gas bubble syndrome-related mortality in whitefish

152 7 .5 Discussion larvae (chapter 6). One possible explanation is that the saine density-independent factor, hypereutrophication, not only caused excessive larval mortality in spring but may also have affected growth of YOY whitefish in 1990 and 1991. Oxygen saturation of 140 percent in May 1990 and May 1991 (Table 7 .1) indicates that stress due to oxygen supersaturation in spring and due to metali1nnetic oxygen mininmn1 in late summer and autmnn was likely to occur in 1990 and 1991. This would be an exception to the earlier statement concerning the effect of eutrophication-induced oxygen stress on growth in general.

Table 7.2 ANOVA results of con1parisons for absolute and relative fecundity and egg diameter of L. Hallwil whitefish of different age fron1 years 1997 and 1998. Significant p values (p < 0.05) are in bold.

Absolute Relative Age groups Egg diameter fecundity fecundity compared p= p= p= 5+/97 vs 6+/98 0.003 0.24 0.91

5+/97 VS 5+/98 0.62 0.48 0.33 5+/97 vs 2+/98 0.1 l 0.56 0.50

5+/97 VS 2+/97 0.09 0.62 0.53

2+/97 VS 6+/98 0.045 0.48 0.54 2+/97 vs 5+/98 0.35 0.46 0.08 2+/97 vs 2+/98 0.16 0.59 0.52 2+/98 vs 6+/98 0.01 0.29 0.45 2+/98 vs 5+/98 ll09 0.30 0.22

5+/98 VS 6+/98 0.25 0.14 0.17

WEATHERLEY & GILL (1987) and JONES (1989) found parallel variations in body condition and growth rate similar to the one sho"vn in Figure 7.3. This is common a111ong healthy fish and appears to be driven by variations in food availability. High fish density by itself could be the cause of variations in food availability due

153 7 Factors controlling growth and maturity to increased intraspecific food competition. A condition index of 1.3 is considered a high value in coregonids (WILKONSKA & ZUROMSKA, 1988). The fact that Lake Hallwil whitefish had a n1ean condition index above 1.3 betvveen 1981and1996 (Table 7.1) indicates that fish were well-fed throughout the whole study period. We therefore conclude that general food shortage has never played a major role. During the time of fastest growth, condition was highest (Table 7 .1 ). In 1990 and 1991 (Table 7.1), when stock size and YCS were relatively low and sunshine hours were abundant, whitefish had a 1nean condition index above the mean condition for 1981-1996, but growth was slow (Table 7.2). This indicates that food was plentiful in 1990 and 1991. Thus, decreasing eutrophication, followed by food shortage, may definitively be ruled out as a cause for the growth decline.

The reason why growth of whitefish in food-rich eutrophic lakes varies with high and low stock density is difficult to explain. Our observations point to a possible effect of schooling behavior on food availability to the individual fish. During daytime, i.e. the main foraging period (BECKER & ECKMANN, 1992), whitefish form schools which can be detected by echosounding (R. Mi.iller, unpubl. data). The schools appear particularly large when the population is very abundant. lt is conceivable that fish inside large schools suffer temporary food shortage although prey density outside the school is high, because fish along the front edge of the slowly n10ving school consume 111ost of the available prey. This hypothesis, however, re111ains to be verified.

VILJANEN (1986) stated that egg quality is a decisive factor for YCS, making it an important element in whitefish management. Our investigations of egg quality and fecundity focused on years 1997 and 1998. In 1998, yield was only half the yield of 1997 (Figure 7.1), which let us assmne that stock size in 1998 was lovver than in 1997. Whitefish eggs in 1998 had a significantly higher caloric content than in 1997 (Figure 7.4 ), although the difference was small. This result inay indicate that in 1998 stress for the fish was lower than in 1997. The condition index was 1.37 in 1997 and 1.35 in 1998, the difference not being significant (p>0.05). In addition, fish age in both years was different. Therefore, our results on egg quality hold no evidence for different stock size or different food conditions in these two years.

154 7 .5 Discussion

Table 7.3 Age at maturity of L. Hallwil whitefish from 1981until1998 (numbers that matured at different ages). l + = autunm of second year, etc. n1 =mature, im1n =imn1ature, - =no data available. Mesh size indicates type of gear used for smnpling fish.

! ,. mesh size Year female 1+ ,~ 2+ male 1+ male 2+ I·' [mm]

m llllll1 111 imm 111 ll1lill Ill iimn j 1981 7 0 1 0 6 4 24 40

1982 48 0 1 0 48 0 r - 24 38 1986 - - 8 0 - - 11 0 42 1987 4 2 - 5 2 20-40 1988 2 0 24 0 12 () 31 1 38 -40 1989 3 0 92 0 - - 4 0 38 -40 1990 3 () 33 () 1 0 12 0 40 1991 - 26 0 3 0 27 0 40

1992 - - ,.,..!(: l 0 - - 27 2 40 1993 1 16 0 30 - - 6 - 30 I 1994 - - 84 () - 11 1 35 I i 1995 - - 6 () -- .• - 35 1996 ------40

') 1997 0 64 3 1() "-"' 61 4 9 20-40 1998 0 19 6 10 1 12 10 5 20-40

Differences in relative fecundity of whitefish of the sa1ne age frorn both years were not significant (Table 7.2). ROJAS-BELTRAN & GILLET (1995) found that larvae of females with high relative fecundity and smaller eggs were less successful in rearing experi1nents. Egg diameter and thus egg size of Lake Hallwil white-

155 7 Factors controlling growth and maturity fish was not significantly different between the years (Table 7 .2). For absolute fecundity, differences were observed between different ages (Table 7.2). Accord- ing to ELPERS (1988), egg nmnber depends on fish size and age. Therefore, differences in absolute fecundity between females of different age are to be expected and do not appear to reflect stress. Fecundity and egg quality pointed to an overall large stock size.

ALM (1959), WEATHERLEY & GILL (1987) and SALOJARVI (1987) found that not only fecundity and egg quality, but also age at first spawning may vary in concert with growth rate and body condition. Age at maturity tends to increase as growth slows down, and vice versa. As shown in Table 7.3, age at first spawning increased from two years in 1980-1991 to three years after 1992. This is in agreernent with the developn1ent of growth. Because Lake Hallwil is still eutrophic, and because condition of whitefish has remained high since 1981 (Table 7.1), we conclude that the main reason for the shift in maturity and for growth deceleration in 1983-1988 is the observed increase in stock size which was particularly pronounced after 1982 (Figure 7.1).

7.6 Acknowledgements

SMA Meteo Schweiz provided long-term sunshine records for 1981-1996. H. MINDER and Dr. A. Stackli, Aarau, made available fishing statistics and limnolog- ical data. The sport fishing club of Lake Hallwil, Meisterschwanden, the c01m11er- cial fisheries Delphin, Meisterschwanden, and Weber, Birrwil, are greatly acknowledged for providing whitefish frmn their regular catch. I kindly thanks my inother for her active help with field sampling. Special thanks also to Dr. E. Meis- ter, ETH Zurich, for providing an oxygen bomb calorimeter and for his help during the experiments, to Dr. N. Mooke1:ji, Montreal, for useful advice concerning fecundity measure1nents. The constructive com1nents by Prof. J. V. Ward helped to i1nprove earlier drafts of this paper.

156 7. 7 References

7.7 References

AKlJ, P. M. K., RUDSTAM, L. G. & TONN, W. M. (1997): hnpact of hypolimnetic oxygenation on the vertical distribution of cisco ( Coregonus artedi) in A1nisk Lake, Alberta. - Can. J. Fish. Aquat. Sci.: 54/9: 2182-2195.

ALM, G. (1959): Connection between maturity, size and age in fish. Report of the Institute of Freshwater Research, Drottingholm 40: 7-145.

BECKER, M. & ECKMANN, R. (1992): Plankton selection by pelagic European whitefish in Lake Constance: Dependency on season and time of day. - Pol. Arch. Hydrobiol. 39: 393-402.

BOWEN, S. H., D'ANGELO, D. J., ARNOLD, S. H., KENIRY, M. J. & ALBRECHT, R. J. (1991): Density-dependent maturation, growth, and female dominance in Lake Superior lake herring (Coregonus artedii). - Can. J. Fish. Aquat. Sci. 48/4: 569-576.

BROWN, R. W. & TAYLOR, W.W. (1992): Effects of egg composition and prey density on the larval growth and survival of lake whitefish (Coregonus clupeaformis Mitchill). - J. Fish Biol. 40: 381-394.

COLBY, P. J. (1984): Appraising the status of fisheries: rehabilitation techniques. In: V. W. Cairns, P. V. Hodson and J. 0. Nriagu (eds). Contamirnmt effects on fisheries. - Wiley, New York: 233-258.

ELPERS, C. (1988): Untersuchungen Zlnn Fortpflanzungserfolg von Coregonen verschiedener Altersklassen des Bodensecs. - Diplomarbeit an der Biologischen Fakultat der Universitiit Freiburg.

FULTON, T. W. (1904): The rate of growth of fish. -·Fisheries Board of Scotland Annual Report 22: 141-241.

GOEDE, R. W. & BARTON, B. A. (1990): Organisrnk indices and an autopsy-based assessment as indicators of health and condition of fish. - Am. Fish. Soc. Symp. 8:93-108.

157 7 Factors controlling growth and maturity

GRIFFITHS, \V. B., GALLAWAY, B. J., GAZEY, W. J. & DILLINGER, R. E. (1992): Growth and condition of arctic cisco and broad whitefish as indicators of causeway-induced effects in the Prudhoe Bay region, Alaska. - Trans. Am. Fish. Soc. 12115: 557-577

HENDERSON, B. A. & BROWN, E. H. (1985): Effects of abundance and water temperature on recruitment and growth of alewife (Alosa pseudoharengus) near South Bay, Lake Huron, 1954-82. - Can. J. Fish. Aquat. Sci. 42: 1608- 1613.

JONES, R. (1989): Towards a general theory of population regulation in marine teleosts. - Jounrnl de Conseil, Conseil International pour l'Exploration de la Mer 45: 176-189.

KIRCHHOFER, A. (1995): Growth characteristics of coregonid populations in three lakes with different trophic states and decreasing nutrient concentrations. Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 61-70.

KOTTELAT, M. (1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the former USSR), with an introduction for non-systematists and comments on nomenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

LARKIN, P. A. (1978): Fisheries management: an essay for ecologists. Annual Review of Ecology and Systematics 9: 57-73.

LECREN, E. D. (1972): Fish production in freshwaters. - Symposia of the Zoo logical Society of London 29: 115-13 3.

LOZAN, J. L. (1992): Angewandte Statistik for Naturwissenschaftler. - Paul Parey, Berlin/Harnburg.

MATHIAS, J. A. & BARICA, J. (1985): Gas supersaturation as a cause of early spring inortality of stocked trout. - Can. J. Fish. Aquat. Sci. 42/1: 268-279.

MA YR, C ( 1998): Zurn Einfluss von Trophic, Fischdichte und Habitatwahl auf die Nahnmgs- und Wachstmnsbedingungen von Renken (Coregonus lavaretus L.) in vier oberbayerischen Seen. - Dissertation an der Fakult~it der ftir Biologic der Ludwig-Maximilians-Universittit Mtinchen.

158 7. 7 References

MICHEL, M. (1996). Untersuchungen zur Nahrungslikologie von Grossfelchen im Vierwaldstattersee wahrend des Somrnerhalbjahres 1996. - Diplo1na Thesis ETH. No. 10366: 73pp.

MILLS, K. H. & CHALANCHUK, S. M. (1988): Population dynamics of unexploited lake whitefish ( Coregonus clupeaf(nmis) in one experimentally fertilized lake and three exploited lake. - Finn. Fish. Res. 9: 145-153.

MOLLER, R. (1990): Management practices for lake fisheries in Switzerland. In: van Densen, W. L. T.; Steinmetz, B.; Hughes, R.H., (Eds), Manage1nent of

freshwater fisheries. Pudoc. Wageningen:(,,..,' '- 477-492.

MOLLER, R. (1992): Trophic state and its implications for natural reproduction of salmonid fish. - Hydrobiologia 243-244: 261-268.

MULLER. R. & BIA, M. M. (1998): Adaptive 111anagement of whitefish stocks in lakes undergoing re-oligotrophication: The Lake Lucerne example. Arch. Hydrobiol. Spec. Issues Advanc. Linmol. 50: 391-399.

MULLER R., BIA M. M., MENG H. J. 1994: Die Felchenfischerei in einigen Seen der Zentralschweiz und des Mittellandes. Aus: Mitteilungen zur Fischcrei Nr. 55. Bundesaint fiir lJmwelt Wald und Landschaft (BUWAL).

MUNKlTTRICK K. R. & DIXON, D. G. (1989): Use of white sucker (Catostom.us commersoni) populations to assess the health of aquatic ecosystems exposed to low-level contaminant stress. Can. J. Fish. Aquat. Sci. 46: 1455-1462.

NAESJE, T. F., SANDLUND, 0. T. & JONSSON, B. (1986): Habitat use and growth of age-0 whitefish, Coregonus lavaretus, and cisco, Coregonus albula. - Environmental Biology of Fishes 15/4: 309-314.

PONTON, D. & MtJLLER, R. (1989): Alirnentation et facteurs de morta1ite des larves de coregones (Coregonus sp.): Exemp1e de deux lacs de niveaux trophiques differents, les lacs de Sarnen et de Hallwil (Suisse Centrale): Aquatic Sciences 51/1: 67-83.

RICE, J. A., CROWDER, L. B. & BINKOWSKI, F. P. (1987): Evaluating potential sources of mortality for larval bloater ( Coregonus hoyi): Starvation and vulnerability to predation. - Can. J. Fish. Aquat. Sci. 44/2: 467-472.

159 7 Factors controlling growth and maturity

ROJAS-BELTRAN R. & GILLET C. (1995): The quality of eggs and larvae of whitefish Coregonus lavaretus L. fr01n Lake Leman: effect of female origin. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 46: 309-314.

SALOJARVI, K. (1987): Why do vcndacc (Coregonus albula L) populations fluctuate? - Aqua Fennica 17/1: 17-26.

SALOJARVI, K. (1992): Compensation in whitefish (Coregonus lavaretus L.) populations in Lake Oulujarvi, northern Finland. - Finn. Fish. Res. 13: 31- 48.

SARVALA, J., RAJASILTA, M., HANGELIN, C., HIRVONEN, A., KIISKILA, M. & SAARIKARI, V. (l 988): Spring abundance, growth and food of O+ vendacc (Coregonus albula L.) and whitefish (C. lavaretus L.) in Lake Pyhajarvi, S\V Finland. -Finn. Fish. Res. 9: 221-233.

SCHEIDEGGER, A., STOCKLI, A. & WDEST, A. (1994): Einfluss der intemen Sanierungsn1assnahmen auf den Sauerstoffhaushalt im Hallwilersee. - Wasser, Energie, Luft 86/5-6: 126-131.

SCHlJLTZ, H. (1992): Stock density, growth rate and zooplankton consumption of the vendacc ( Coregonus albula) and other fish species in Lake Arendsee. Liinnologica 22/4: 355-373.

STOSSEL, F. (1992): Die Bodenfauna im Hallwilersee dtingt vor. - EA\VAG-news 34D: 23-26.

SVARVAR, P. 0., MtJLLER, R. (1982): Die Felchen des Alpnachersces. - Schweiz. Z. Hydrol. 44/2: 295-314.

VENTLING-SCHWANK, A. R. ( 1992): Reproduktion und larvale Entwicklungs- phase der Felchen (Coregonus sp.) im. eutrophen Se1npachersee. Dissertation an der Philosophischen Fakulttit der Universitat Ziirich.

VENTLING-SCHWANK, A. R & LIVINGSTONE, D. M. (1994): Transport and Burial as a Cause of Whitefish (Co reg onus sp.) Egg Nlortality in a Eu trophic Lake. - Can. J. Fish. and Aquat. Sci. 5119: 1908-1919.

160 7. 7 References

VILJANEN, M. (1983): Food and food selection of cisco ( Coregonus albula L.) in a dysoligotrophic lake. - Hydrobiologia 101: 129-138.

VILJANENi M. (1986): Biology, propagation, exploitation and rnanagernent of vendace (Coregonus albula L.) in Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Lirnnol. 22: 73-97.

VILJANEN, M. (l 9 88): Population dynamics of vendace ( Coregonus albula L.) in Finland. - Joensuun Yliopiston Luonnontieteellisia Julkaisuja 12: 1-19.

WATERS, T. F. (1977): Secondary production in inland waters. - Advances in Ecological Research 10: 91-164.

WEATHERLEY, A. H. & GILL, H. S. (1987): The biology of fish growth. - Academic Press, New York.

WILKONSKA H. & ZUROMSKA H. (1988): Effect of enviromnent on Coregonus albula (L.) spawners, and influence of their sexual products on the nun1bers and quality of offspring. - Finn. Fish. Res. 9: 81-88.

161 7 Factors controlling growth and maturity

162 8.1 Introduction

8 SYNOPSIS

8.1 Introduction

On the basis of all the results presented in this thesis, the life cycle of Lake Hallwil whitefish ( Coregonus suidteri, Patio) is portrayed in sub-chapter 8.2. The rearing experin1ents (see chapters 3 and 4) have led to rcc01nmendations for improved rearing practice as stated in the sub-chapter 8.3. The inain question of the thesis, the cause for whitefish yield fluctuation, has been investigated using a system dynamics approach with computer n1odelling (see chapter 6). Implications frmn this part of the work are discussed in sub-chapter 8 .4 with respect to opti1nizing stocking of whitefish larvae, and in sub-chapter 8.5 concerning sustainable whitefish management. The conclusions drawn in chapter 2 on whitefish phenotype and in chapter 7 on growth are included here as well. Finally, in sub-chapter 8.6, the impact of lake restoration ineasures on whitefish is briefly assessed on the basis of the results in chapter 5.

8.2 Population dynan1ics of L. Hallwil whitefish

Today, all L. Hallwil whitefish larvae originate frmn artificially bred eggs, because natural reproduction is unsuccessful due to the still anoxic sediment (VENTLING-SCHWANK & LIVINGSTONE, 1994). After hatching, in February or March, the major part of the reared hu-val population is stocked as unfed fry. Fron1 February until April, different feeding depth in the lake and food preference of whitefish larvae and adults niake intraspecific con1petition and cannibalism unlikely to occur. Adults stay in the metalinu1ion and hypolimnion where they n1ainly prey on benthic organis1ns, particularly chaoborids and annelids. The larvae stay close to the surface (PONTON & MENG, 1990) and eat s1nall zooplankton species, especially rotifers, copepod nauplii and calanoid and cyclopoid copepods. This usually lasts until the end of April (PONTON & MOLLER, 1989). Unfed fry has a survival rate - up to the adult stage in the order of 0.5 percent. The sur-

163 8 Synopsis vival rate of pre-fed larvae is estimated to be about ten times greater in the lake than that of newly hatched fry. It is interesting to note that the number of stocked pre-fed larvae correlates better to year class strength than the nmnber of stocked newly hatched (unfed) fry. This difference is due to the better swimn1ing and foraging ability (HARTMANN & KLEIN, 1993) and the lower predation risk (HOAG- MAN, 197 4) of the bigger pre-fed larvae. Therefore, a small part of the larval population is kept in the hatcheries for at least six weeks. They are reared in circular tanks and fed lake zooplankton or dry diets. The pre-fed larvae are about 15-17 mm long when they are stocked in April. Newly hatched L. Hallwil whitefish fry are about 10-11 mm long.

At the beginning of May, when all unfed fry and pre-fed larvae have been stocked, daphnids begin to dominate their food (PONTON & MULLER, 1989). All larvae are then 15-19 mm long (PONTON & MULLER, 1989) and fairly secure from predation (HOAGMAN, 1974) and cannibalism (GRIM, 1951). Nevertheless, survival rate varies from year to year. According to ECKMANN (1990), mortality is mainly a consequence of larval behavior to stay near the surface where they are extremely vulnerable to changes in environmental conditions. Meteorological conditions in spring directly influence water temperature (LEHTONEN, 1985). Computer inodel- ling revealed that in warm and sunny spring months, especially May, oxygen supersaturation due to excessive photosynthesis by algae may lead to gas bubble syndrome (GBS) in whitefish larvae. GBS was found to be the single most important inortality factor that could severely reduce a whitefish year class or even anni- hilate it. Apart from this negative aspect, warm spring months inay also be bene- ficial. When warm and sunny weather occurs in May but does not lead to GBS, high phyllopod densities are correlated with maximum larval survival. Therefore, very strong year classes can only develop when the weather in May is as warm as possible to enhance phyllopod development, but at the same time as moderate as possible (without abrupt short-tenn changes) to avoid oxygen supersaturation and thus GBS. The probability for this situation to occur is quite low. That's why strong year classes as in 1983, 1987 and 1992 are the exception. Strong year classes seem to occur at intervals of 3-5 years, as found by Aass (1972). However, a logical prerequisite for a strong year class to occur is that high numbers of larvae are stocked.

164 8.3 Implications for artificial reproduction

In May, adult whitefish switch frmn benthic prey to zooplankton. However, during periods when insect larvae are present in large numbers, whitefish are able to use this food type with great efficiency. During the main feeding period from May until September, both adult and young of the year whitefish prey preferen- tially on phyllopods. Because prey size increases with increasing fish length (PONTON & MlJLLER, 1990), juvenile (small) and adult (big) whitefish prey on different size classes of phyllopods. This may ease intraspecific food cmnpetition. If stock size is large, then the young of the year whitefish tend to grow 1nore slowly than at low fish density. At small stock size, the young of the year whitefish could reach a body length of more than 200 mm by winter, and reach maturity the following year (as l + fish). In contrast, at high stock density, they would reach about 150-175 mm by the first winter and not mature until the end of the third year (as 2+ fish). Therefore, a shift in age at maturity could also be a sign of changing fish density (SALOJARVI, 1987). In autumn, 1nature whitefish 1nove to deeper layers, show an increased proportion of empty stomachs and prepare for spawning (SKURDAL et al., 1985). During late autmnn, whitefish are captured for artificial propagation.

8.3 Implications for artificial reproduction

Natural reproduction of L. Hallwil whitefish will probably continue to be unsuccessful for a few more decades. Therefore, artificial propagation of L. Hallwil whitefish is necessary. It is thus important to optimize the method of rearing whitefish larvae. Natural food appears to be the optimal food for whitefish fry. The major problem is that zooplankton density is low in the lake during spring. This makes it nearly impossible to provide ad libitum feeding and optimal growth for large numbers of whitefish larvae in hatcheries.

In order to cope with this proble1n, four alternatives are possible: First, rearing the larvae in their natural environment using illuminated net cages (R(>SCH & ECK- MANN, 1986) has proven to be feasible. It is successfully done at large scale in L. Sempach, a nearby lake of comparable trophic state. Second, feeding live Arternia nauplii could be even better at first feeding than lake zooplankton due to their small size (ROSCH, 1989; ECKMANN, 1985) and essential fatty acid cmnposition (FLtJCHTER & REMBOLD, 1986). Unfortunately, some strains of Artemia nauplii

165 8 Synopsis are contaminated and may cause bacterial intestinal infection (BURKHARDT- HOLM et al., 1989). Such strains, particularly the one fr01n Great Salt Lake/Utah, should be avoided. Third, frozen zooplankton, as reported by KLEIFELD-KRIEBITZ & ROSCH (1987), could be used as larval feed. Fourth, coregonid larvae can now be successfully reared on dry diet. When feeding dry diet, three aspects have to be considered. First, the circular rearing tanks have to be illuminated separately to enhance food uptake. A light intensity of at least 200 lux at the water surface is required. Second, only dry diets that result in more than 70 percent survival at temperatures between 4 and 13°C should be used. Third, the water inflow should be placed below, not above, the water surface. A submerged inflow increases floating time of the dry diet, allowing reductions in feeding frequency and the daily ration. As important side-effects, cleaning effort can be lowered due to lower food loss, and the costs for feed can be cut without negatively affecting the fish.

Larvae reared on diets other than live zooplankton need an adaptation phase during the switch to live zooplankton. As a consequence of the adaptation phase, slower growth during the first week after the change of diet may occur. Therefore, the switch from an alternative diet to live zooplankton should be done stepwise and not abruptly. The larvae should not be stocked into the lake without having gained experience with live zooplankton in the hatchery.

8.4 Implications for stocking practice

The best rearing 1nethod is useless if the reared larvae are stocked at the wrong mmnent. Although the trophic state of L. Hallwil is on a steady decrease, GBS is likely to occur for some years. Timing of stocking the larvae in the lake is therefore critical in order to avoid total loss of year classes. Stocking fingerlings in late smnmer would yield the best result. Fingerling survival was reported to be around 75 percent (SALOJARVI, 1986). However, this is hardly feasible because in May the circular tanks of the L. Hallwil hatcheries are used to rear pike, Esox lucius. Pre-fed larvae should be stocked only during cloudy weather and stocking should be delayed if Secchi depth is less than 1.5 meters. The 1nain requirement for GBS to occur is a period of several days of sunny weather. VENTLING-SCHWANK (1992) found no GBS in L. Sempach in 1989 and 1990 when Secchi depth was more than 1.5 meters. Secchi depth provides a simple esti1nate of algal bi01nass

166 8.5 Implications for the fisheries and therefore the risk of GBS. Third, pre-fed larvae should be released batchwise throughout May to minimize the risk of total loss caused by GBS.

When discussing stocking practice, the issue of stocking whitefish larvae fron1 other lakes is of concern. After the disastrous fish kill in 1961, a whitefish population was restored in L. Hallwil by introducing 1nainly L. Se1npach whitefish larvae. DOUGLAS (1999) stated that local whitefish populations should be treated as management units to preserve the diversity of central alpine coregonids. This view is reflected in today's fisheries legislation which requires that whitefish stocked have to originate from parent fish frmn the smne lake. KOTTELAT (1997) found that L. Sempach whitefish is conspecific to L. Hallwil whitefish. In view of this close relationship, the stocking of L. Sempach whitefish after the presmned erad- ication of L. Hallwil whitefish in 1961 was the best possible procedure for restor- ing a whitefish population in L. Hallwil.

8.5 Implications for the fisheries

It is well known that coregonid populations fluctuate due to fishing (LEHTONEN, 1983). In order to ensure that the whitefish of L. Hallwil will be conserved in the longer term, management of the L. Hallwil whitefish population has to be sustainable. GULLAND ( 1985) stated that maximum sustainable yield could reflect both economic benefit and conservation of natural fish populations. For economic benefit, yield fluctuation is a serious problem. For the conservation aspect it is only a problen1 when fish stock size decreases to a critically low level (GULLAND, 1985). VILJANEN (1988) found that yield oscillations disappear when fishing mortality of the adults is lowered and that fish attain greater ages. In agreernent with these findings, our modelling results showed that with increasing fish age at harvest, stock size fluctuations can be reduced. Strong year classes fill up the yield gap resulting from weak or nonexistent year classes. The model showed, however, that yield fluctuations disappear only at an unrealistic mean fish age at harvest of ten years. According to MULLER et al. (1994 ), fish should be allowed to reproduce at least once before they are harvested. Whitefish age at 1naturity has increased from 1+ to 2+ in L. Hallwil, most likely because of increased fish density. This leads to the conclusion that whitefish should attain at mini1num an age of three years at harvest (3+ fish). GULLAND (1985) found that mean individual fish weight is higher when

167 8 Synopsis fishing is moderate, as compared to fishing at the edge of maxirnmn sustainable yield. Indeed, at the age of three years, size and weight are optimal from a commercial point of view because the most rapid growth occurs during the first two to three years (MULLER et al., 1994). Moderate fishing increases average fish weight, reduces fluctuations in stock size and yield and allows for extensive egg production. It is therefore in the ecological and commercial interest that L. Hallwil whitefish are caught at a minimum age of three years. As in any other lake, monitoring whitefish age in the catch and adjusting minimmn n1esh size of the gillnets are of prime importance for achieving a sustainable fishery in L. Haliwil.

8.6 Effects of the restoration n1easures

The external lake restoration measures, pri1narily wastewater diversion and new regulations on the use of fertilizers in agriculture, have lowered the 1nean phosphorus concentration of the lake from 250 mg/l in 1977 to about 40 n1g/l in 1999. Nevertheless, the lake is still eutrophic. High primary production, and thus oxygen supersaturation in spring and the fonnation of a metalimnetic oxygen mini1num in late summer/autun1n, are the well-known consequences. Luckily, this happens infrequently. On the positive side, high primary production also increases zooplankton density and thus the food base for whitefish.

The internal restoration 1neasures of L. Hallwil, i.e. artificial oxygenation and forced 1nixing (SCHEIDEGGER et al., 1994), obviously have had a positive effect on whitefish. The living space for fish was enlarged, and an abundant benthic fauna could develop (STOSSEL, 1992). This increased the nmnber of food organ-

isms available to adult whitefish,, allowing'- the adult whitefish to avoid starvation and to reduce intraspecific con1petition with the larvae during spring, when zooplankton density is low. In spite of this positive development, the sedi1nent surface is still anoxic, which represents the major obstacle to successful natural reproduction of L. Hallwil whitefish.

168 8.7 References

8. 7 References

AASS, P. (1972): Age determination and year-class fluctuation of cisco, Coregonus albula L., in the Mj¢sa hydroelectric reservoir. - Report Institute of Freshwater Research (Drottninghohn) 52: 5-22.

BURKHARDT-HOLM, P., ECKMANN, R. & STORCH, v. (1989): Schadigung des Darmepithels von Coregonenlarven (Coregonus fera) durch Artemia- Fitttenmg, Eine bakterielle Infektion. - Journal of Applied Ichthyology 1: 2- 11.

DOUGLAS, M. (1999): Central alpine Coregonus (Teleostei, Coregonidae): Evolution and conservation of a unique assemblage. Dissertation an der Philosophischen Fakultat der Universitlit Zurich.

ECKMANN, R. (1985): Histopathological alterations in the intestine of whitefish (Coregonus sp.) larvae reared on zooplankton frmn Lake Constance (West Germany). - Diseases of Aquatic Organisms 1/1: 11-18.

ECKMANN, R. (1990): The distribution of coregonid larvae (Co reg onus lavaretus and Coregonus fera) fr0111 Lake Constance in a vertical temperature gradient. - Pol. Arch. Hydrobiol. 36/4: 485-494.

FLOCHTER, J. & REMBOLD, H. (1986): Soluble factor essential for tnetamorphosis of coregonid larvae has been partially purified from Artemia salina. - Arch. Hydrobiol. Spec. Issues Advanc. Linmol. 22: 197-202.

GRIM, J. (1951): Kannibalismus bei Blaufelchen und seine moglichen Folgen. - Osterreichs Fischerei 4: 165-171.

GULLAND, J. A. (1985): Fish stock assessment. A tnanual of basic rnethods, 1st edition, 1983. - Food and Agriculture Organization of the United Nations (FAO) and John Wiley & Sons, Ltd., New York.

HARTMANN, F. & KLEIN, M. (1993): Nahrungsselektion von Renkenbrut (Coregonus lavaretus) unter Aufzuchtbedingungen. - Fischerei und Teichwirt 44/8: 279-283.

HOAGMAN, W. J. (1974): Feeding by alewives (Alosa pseudoharengus) on larval lake whitefish (Coregonus clupeafonnis) in the laboratory. - J. Fish. Res. Board Can. 31/2: 229-230.

169 8 Synopsis

KLEIFELD-KRIEBITZ, G. & ROSCH, R. (1987): A simple rnethod of feeding coregonid larvae on frozen zooplankton. - Journal of Applied Ichthyology 3: 119-124.

KOTTELAT, M. (1997): European freshwater fishes. An heuristic checklist of the freshwater fishes of Europe (exclusive the fonner USSR), with an introduction for non-systematists and comments on nomenclature and conservation. - Biologia, Sec. Zool. 52 (Suppl. 5): 1-271.

LEHTONEN, H. (1983): Scientific basis for fisheries rnanagement of vendace, Coregonus albula, in the Bothnian Bay (Baltic Sea). -Aquilo Ser Zoologica 22: 77-82.

LEHTONEN, H. (1985): Changes in conunercially important freshwater fish stocks in the Gulf of Finland during recent decades. - Finn. Fish. Res. 6: 61-70.

MOLLER, R., BIA, M. M. & MENG, H.J. (1994): Die Felchenfischerei in einigen Seen der Zentralschweiz und des Mittellandes. - BUWAL Mitteilungen zur Fischerei 55.

PONTON, D. & MENG, H.J. (1990): Use of dual-beam acoustic technique for detecting young whitefish, Coregonus sp., juveniles: first experiments in an enclosure. - J. Fish Biol. 36: 741-750.

PONTON, D. & MlJLLER, R. (1989): Alimentation et facteurs de mortalite des larves de coregones (Coregonus sp.): Exernple de deux lacs de niveaux trophiques differents, les lacs de Sa111en et de Hallwil (Suisse Centrale): - Aquatic Sciences 51/1: 67-83.

PONTON, D. & MULLER, R. (1990): Size of prey ingested by whitefish, Coregonus sp., larvae, Are Coregonus larvae gape-lirnited predators? - J. Fish Biol. 36/l: 67-72.

ROSCH, R. & ECKMANN, R. (1986): Survival and growth of prefed Coregonus lavaretus L. held in ilhnninated net cages. - Aquaculture 52: 245-252.

RoscH, R. (1989): Beginning of food intake and subsequent growth of larvae of Coregonus lavaretus L. - Pol. Arch. Hydrobiol. 36/4: 475-484.

170 8. 7 References

SALOJARVI, K. (1986): Review of whitefish (Coregonus lavaretus L. s.l.) fingerling rearing and stocking in Finland. - Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 22: 99-114.

SALOJARVI, K. (1987): Why do vendace (Coregonus albula L.) populations fluctuate? - Aqua Fennica 17 /1: 17-26.

SCHEIDEGGER, A., ST6CKLI, A. & WUEST, A. (1994): Einfiuss der internen Sanierungs1nassnahmen auf den Sauerstoffhaushalt im Hallwilersee. - Wasser, Energie, Luft 86/5-6: 126-131.

SKURDAL, J., HESSEN, D. 0. & BERGE, D. (1985): Food selection and vertical distribution of pelagic whitefish Coregonus lavaretus in Lake Tyrifjorden, Norway. - Fauna Norvegica Series A 6: 18-23.

STOSSEL, F. (1992): Die Bodenfauna i1n Hallwilersee dringt vor. EAWAG-news 34D: 23-26.

VENTLING-SCHWANK, A. R. (1992): Reproduktion und larvale Entwicklungs- phase der Felchen (Coregonus sp.) im eutrophen Sempachersee. - Dissertation an der Philosophischen FakulUit cler Universitat Ziirich.

VENTLING-SCHWANK, A. R & LIVINGSTONE, D. M. (1994): Transport and burial as a cause of whitefish (Coregonus sp.) egg inortality in a eutrophic lake. - Can. J. Fish. Aquat. Sci. 51/9: 1908-1919.

VILJANEN, M. (1988): Population dynamics of vendace (Coregonus albula L.) in Finland. - Joensuun Yliopiston Luonnontieteellisia Julkaisuja 12: 1-19.

171 8 Synopsis

172 DANK

Mcinc1n Bctrcucr der Dissertation, Dr. Rudolf Muller, gilt mcin erster Dank. Er errn6glichte rnir, die vorliegende Arbeit an der EAWAG auszuflihren und liess rneiner Forschung viel Freiramn. Falls jedoch n1eine Gedankenspirale sich zu stark verdrehte, fand er iinmer cine M(iglichkeit, den Gordischen Knoten zu lOsen. Ausserdem danke ich ihm flir das Zurechtbiegen rneines Englisch-Schreibstils.

Meinem Doktorvater, Prof. Dr. J. V. Ward, danke ich fur die Leitung der Disser- tation und fUr die kritische Durchsicht 1neiner Manuskripte. Sein stilsicheres Eng- lisch und seine Fachkornpetenz halfen, manch einen Sachverhalt kurz uncl treffencl zu beschreiben.

Prof. Dr. Hannu Lehtonen, ein finnischer Koregonenspezialist, hat sich freundli- cherweise bereiterklart, das Korreferat meiner Dissertation zu libern.elnnen. Daflir danke ich ihm herzlich.

Deni Finanzdeparte1nent und elem Baudeparternent des Kantons Aargau danke ich flir die finanzielle Untersttitzung. Spcziell danken rnochte ich Herrn Hans Minder und Dr. Arno Stbckli ftir die ausgczeichnete Zusa1111nenarbeit, die wertvollen Fachgespdiche betreffend Hallwilersee und das unkornplizierte Bereitstellen von Datenmaterial.

Prof. Dr. Rainer Eck1nann, Dr. Arthur Kirchhofer, Prof. Dr. Willi Meyer, Dr. Nan- dita Mooke1:ji, Dr. Roland Rosch und Dr. Claus Wedekind haben einen wesentlichen Beitrag zum Gelingen rneiner A ufzuchtversuche geleistet. Ausserst wertvoll flir das Vorankomrnen rneiner Cmnputerrnodelle war das Fachwissen von Dr. Werner Donni, Prof. Dr. Hans U. Fuchs, Dr. Johannes Reeb und Dr. Martin Simon. Dr. Erich Meister errnoglichte inir, Kalorimetrie-Experimente clurchzu- flihren, und stand n1ir stets init Rat und Tat zur Seite. Oliver Heiri gab den ent- scheidenden Hinweis in Insektensyste1natik uncl crleichterte 1nir dadurch das Bestimmen der Felchennahrung. All diesen Fachleuten sei an dieser Stelle herzlichst gedankt.

173 Mein Dank gilt auch Henn Urs Fischer, clamalige111 Prasiclent des Sportfischerver- eins Hallwilersee, welcher cler Brutanstaltbenutzung fUr 111eine Aufzuchtversuche wohlwollencl gegentiberstand. Herrn Max Keller vom Sportfischerverein m()chte ich meinen herzlichsten Dank aussprechen, weil seine be111erkenswerte Beobach- tungsgabe uncl sein reicher Erfahrungsschatz bezliglich der Aufzucht von Fel- chenlarven bei meinen Experimenten unbezahlbar waren. Herrn Ernst Fischer und HeITn Stefan Laib von cler Fischerei Delphin, Herrn Heinz Weber und seiner Frau Rita von der Fischerei Weber danke ich herzlichst ftir das Interesse an meiner Tatigkeit, die reibungslose Zusaimnenarbeit und das Vertrauen.

Die Finnen Kyowa Hakko Kogyo Co. in Tokyo (Japan), Farmix AG in Malters (Schweiz) und Hendrix SpA in Mozzecane (Italien) stellten mir kostenlos Trok- kenfutter flir meine Aufzuchtversuche zur Verfligung. Die Schweizerische Meteorologische Anstalt SMA Meteo Schweiz lieferten mir unbtirokratisch Lang- zeitwetterdaten, welche die Genauigkeit ineiner Computern10delle positiv beein- tlusst haben.

Allen an der EA WAG inochte ich ganz herzlich flir die angenehrne Zeit, die inter- essanten Gesprache und die haufig spontan geleistete Hilfe danken. Genannt seien hier Dr. Heinrich Btihrer, Dr. Hans-Rudolf Btirgi, Brigitte Germann, Dr. Mark Gessner, Erwin Grieder, Dr. Christine Heller, Doris Hohmann, Sonja Janser, Silvia Jost, Gaby Meier-Blirgisser, Dr. Hans-Jiirg Meng, Dr. Armin Peter, Chri- stian Rellstab, Georg Ribi, Andreas Steffen, Daniel Steiner, Francisco Vazquez, Alois Zwyssig und alle Doktorierenden. Mampasi Mbwenemo-Bia gilt 111ein spe- zieller Dank fur seine Einsatze im Feld und die fachkundigc Einftihrung in die Biometric. Meinem Btirokollegen Erwin Schaffer inochte ich ganz besonders danken, da sein tagelanger, selbstloser Einsatz bei111 Bau einer Versuchsanlage wesentlich zmn Gelingen meiner Aufzuchtexperimente beigetragen hat.

Meinen Eltern 111ochte ich einen lieben Dank aussprechen fur die seelische und finanzielle Unterstiitzung und fiir die spontanen Einsatze im Feld. Freundinnen und Freunden danke ich for die Zerstreuung bei Gesprachcn, Ausfliigen sowie Karten-, Brett- und Videospielen und fiir die dummen SprUchc. Den Mitgliedern des Vereins Aquarimn Ziirich danke ich fUr das rege Interesse. Schllesslich hat auch mein personlicher Apple-Faclunann, Max Schlapfer, einen lieben Dank for Geduld, Versttindnis und wertvolle Hilfe vcrdient.

174 175 I

176 CURRICULUM VITAE

Carole Andrea Enz geboren arn 3. August 1972 in Zurich

Schulbildung

1979-1985 Pri111arschule in Zurich

1985-1991 Kantonsschule (Matura Typus B) in ZUrich

Studium und Promotion

1991-1995 Studium der Biologie an der ETH in Zurich Diplomarbeit an der EAWAG/ETH in DUbendorf

1996 Ausbildung for das Hdhere Lehramt an der ETH in Zurich

1997-2000 Dissertation mn Forschungszentrum fur Lirnnologie der EAWAG in Kastanienbaum

Anschrift Carole Enz EAWAG CH-6047 Kastanienbamn

Carole Enz Hesenlooweg 3 CH-8038 Zitrich

177