The Comparative Roles of Suspension-Feeders in ...

The Comparative Roles of SuspensionFeeders in Ecosystems

NATO Science Series A Series presenting the results of scientific meetings supported under the NATO Science Programme. The Series is published by IOS Press, Amsterdam, and Springer (formerly Kluwer Academic Publishers) in conjunction with the NATO Public Diplomacy Division.

Sub-Series I. II. III. IV.

Life and Behavioural Sciences Mathematics, Physics and Chemistry Computer and Systems Science Earth and Environmental Sciences

IOS Press Springer (formerly Kluwer Academic Publishers) IOS Press Springer (formerly Kluwer Academic Publishers)

The NATO Science Series continues the series of books published formerly as the NATO ASI Series. The NATO Science Programme offers support for collaboration in civil science between scientists of countries of the Euro-Atlantic Partnership Council. The types of scientific meeting generally supported are “Advanced Study Institutes” and “Advanced Research Workshops”, and the NATO Science Series collects together the results of these meetings. The meetings are co-organized by scientists from , NATO countries and scientists from NATO s Partner countries – countries of the CIS and Central and Eastern Europe.

Advanced Study Institutes are high-level tutorial courses offering in-depth study of latest advances in a field. Advanced Research Workshops are expert meetings aimed at critical assessment of a field, and identification of directions for future action. As a consequence of the restructuring of the NATO Science Programme in 1999, the NATO Science Series was re-organized to the four sub-series noted above. Please consult the following web sites for information on previous volumes published in the Series. http://www.nato.int/science http://www.springeronline.com http://www.iospress.nl

Series IV: Earth and Environmental Series – Vol. 47

The Comparative Roles of Suspension-Feeders in Ecosystems edited by

Richard F. Dame Marine Science Department, Coastal Carolina University, Conway, SC, U.S.A. and

Sergej Olenin Coastal Research and Planning Institute, Klaipeda University, Klaipeda, Lithuania

Proceedings of the NATO Advanced Research Workshop on The Comparative Roles of Suspension-Feeders in Ecosystems Nida, Lithuania 4–9 October 2003

A C.I.P. P Catalogue record for this book is available from the Library of Congress.

ISBN-10 1-4020-3029-0 (PB) Springer Dordrecht, Berlin, Heidelberg, New York ISBN-13 978-1-4020-3029-1 (PB) Springer Dordrecht, Berlin, Heidelberg, New York ISBN-10 1-4020-3028-2 (HB) Springer Dordrecht, Berlin, Heidelberg, New York ISBN-10 1-4020-3030-4 (e-book) Springer Dordrecht, Berlin, Heidelberg, New York ISBN-13 978-1-4020-3028-4 (HB) Springer Dordrecht, Berlin, Heidelberg, New York ISBN-13 978-1-4020-3030-7 (e-book) Springer Dordrecht, Berlin, Heidelberg, New York

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands.

Printed on acid-free paper

All Rights Reserved © 2005 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed in the Netherlands.

Table of Contents List of Contributors…………...………………………………………………………...…..…vii Preface…………………………………………………………………………………...…..… xi 1. Modelling particle selection efficiency of bivalve suspension feeders P Zemlys and D Daunys……………………………………………………….……...…..….1 2. Field measurements on the variability in biodeposition and estimates of grazing pressure of suspension-feeding bivalves in the northern Baltic Sea J Kotta, H Orav-Kotta and I Vuorinen…………………………………………………...….11 3. Can bivalve suspension-feeders affect pelagic food web structure? T Prins and V Escaravage……………………………………………………….…...…….31 4. Motile suspension-feeders in estuarine and marine ecosystems D Bushek and D M Allen………………..………………………………………...………..53 5. Impact of suspension-feeding nekton in freshwater ecosystems: patterns and mechanisms H Ojaveer………………………………………………………………..………….……….73 6. Influence of eastern oysters on nitrogen and phosphorus regeneration in Chesapeake Bay, USA R I E Newell, R R Holyoke and J C Cornwell…….…………………….…………………..93 7. How does estimation of environmental carrying capacity for bivalve culture depend upon spatial and temporal scales? P Duarte, A J S Hawkins and A Pereira………..……………………….…………..……..121 8. Impact of increased mineral particle concentration on behavior, suspension feeding and reproduction of Acartia clausii (Copepoda) N Shadrin and L Litvinchuk………………………………………………………...….…137 9. Suspension-feeders as factors influencing water quality in aquatic ecosystems S A Ostroumov…………………………………………………………………...…….….147 10. Neoplasia in estuarine bivalves: effect of feeding behaviour and pollution in the Gulf of Gdansk (Baltic Sea, Poland) M Wolowicz, K Smolarz and A Sokolowski……………………………….…...…….….165 11. Bivalves as biofilters and valuable by-products in land-based aquaculture systems M Shpigel……………………………………………………………………..…………..183 12. Significance of suspension-feeder systems on different spatial scales H Asmus and R M Asmus………………………………………………………...……...199 13. Invaders in suspension-feeding systems: variations along the regional environmental gradient and similarities between large basins S Olenin and D Daunys…………………………………………………………………..221

v

vi

14. Contrasting distribution and impacts of two freshwater exotic suspension feeders, Dreissena polymorpha and Corbicula fluminea A Y Karatayev, L E Burlakova and D K Padilla................................................................239 15. Functional changes in benthic freshwater communities after Dreissena polymorpha (Pallas) invasion and consequences for filtration L E Burlakova, A Y Karatayev and D K Padilla….……………………….………...…..263 16. Does the introduction of the Pacific oyster Crassostrea gigas lead to species shifts in the Wadden Sea? A Smaal, M van Stralen and J Craeymeersch………………………………………...….277 17. One estuary, one invasion, two responses: phytoplankton and benthic community dynamics determine the effect of an estuarine invasive suspension-feeder J K Thompson…………………………………..………………………………………...291 18. Development of human impact on suspension-feeding bivalves in coastal soft-bottom ecosystems W J Wolff………………………………………………………………………………...317 19. Oyster reefs as complex ecological systems R Dame…………………………………………………………………..…………….....331 20. Synthesis/Conclusions………………………………………………………….………..345 Index…………………………………………………………….…………………………….355

List of Contributors (Mailing Addressses) Drs. Harald and Ragnhild Asmus Alfred-Wegener-Institut für Polar- und Meeresforschung Wattenmeerstation Sylt Hafenstraße 43 25992 List/Sylt Germany

Dr. David Bushek Haskin Shellfish Research Lab Rutgers University Port Norris, NJ 08349 United States

Dr. Lyubov Burlakova Department of Biology Stephan F. Austin State University SFA Station Nacogdoches, Texas 75962 United States

Dr. Richard Dame (Co-Director NATO) Coastal Carolina University P.O. Box 261954 Conway, SC 29528 United States

Dr. Darius Daunys Coastal Research and Planning Institute Klaipeda University, Manto 84 LT-5805 Klaipeda Lithuania

Dr. Pedro Duarte Universidade Fernando Pessoa Praça 9 de Abril, 349 4200 Porto Portugal

Dr. Alexander Y. Karatayev Department of Biology vii

viii

Stephan F. Austin State University SFA Station Nacogdoches, Texas 75962 United States

Dr. Jonne Kotta Estonian Marine Institute Marja 4d 10617 Tallinn Estonia

Dr. Roger Newell University of Maryland Center for Environmental Studies Horn Point Laboratory PO Box 775, Cambridge, MD 21613 United States

Dr. Sergej Olenin (Co-Director NATO-Partner) Coastal Research and Planning Institute Klaipeda University, Manto 84 LT-5805 Klaipeda Lithuania

Dr. Henn Ojaveer Estonian Marine Institute Maealuse 10a 12618 Tallinn Estonia

Dr. Sergei A. Ostroumov Department of Hydrobiology, Faculty of Biology, Moscow State University Moscow 119899, Russia

Dr. Theo Prins National Institute for Coastal and Marine Management/RIKZ POBOX 8039 4330 EA Middelburg The Netherlands

ix Dr. Nikolay Shadrin Institute of Biology of the Southern Seas 2, Nakhimov Ave. Sevastopol, 99011 Ukraine.

Dr. Muki Shpigel Israel Oceanographic and Limnological Research; National Centre for Mariculture PO Box 1212 88112 Eilat Israel

Dr. Aad Smaal Shellfish Research Centre, RIVO-DLO Korringaweg 5, P.O. Box 77 4400 AB Yerseke The Netherlands

Dr. Jan Thompson U.S. Geological Survey MS496 345 Middlefield Rd. Menlo Park, CA 94025 United States

Dr. Wim J. Wolff Dept. of Marine Biology Groningen University P.O. Box 14 9750 AA Haren The Netherlands

Dr Maciej Wolowicz Laboratory of Estuarine Ecology, University of Gdansk Al. Pilsudskiego 46 81 378 Gdynia Poland

x Dr. Petras Zemlys Coastal Research and Planning Institute Klaipeda University, Manto 84 LT-5805 Klaipeda Lithuania

PREFACE Animals are a major link between the water column (pelagic) and the bottom (benthic) habitats in most shallow systems. This coupling is dominated by active processes such as suspension-feeding in which the organism actively uses energy to pump water that is then filtered to remove suspended particles that are consumed while undigested remains are deposited on the bottom. As a result of this feeding on and metabolism of particles, the animals excrete dissolved inorganic and organic waste back into the water column, and thus, become major components in the cycling and feedback of essential elements. With relatively high weight specific filtration rates of 1– 10 liters/hour/gram dry tissue and a propensity to form large aggregated populations (beds, reefs, schools and swarms), these organisms can play an important role in regulating water column processes Although estuarine bivalve molluscs such as oysters and mussels dominate the suspension-feeder literature, other groups including plankton and nekton that are found in estuarine as well as other aquatic systems are also potentially important removers of suspended particles. Thus, a significant part of the NATO Advanced Research Workshop focused on suspension-feeders as controllers of plankton abundance, biomass and diversity, system metabolism, nutrient cycling and scale dependency. Systems dominated by suspension-feeders are typically impacted by human activities including recreation, aquaculture, human and industrial pollution, and bilge water from shipping. Suspension-feeders are often impacted by fisheries and over-exploitation. These impacts commonly result in changes in ecosystem structure either through the food chain concentration of harmful substances or diseases, the introduction of alien species of suspension-feeders, or the instability of suspension-feeders systems through species displacement or phase shifts in the dominance between different suspension-feeding components such as nekton or zooplankton. These issues were addressed near the close of the workshop along with conclusions and syntheses developed by the working groups. In the almost 10 years since one of us (RFD) led a NATO ARW in The Netherlands on bivalve filter feeders, interest in suspension-feeders as major influences on aquatic ecosystem processes has grown dramatically. This development is particularly evident in freshwater systems, yet the communications between the freshwater and the estuarine-marine scientific communities are weak (probably because of scientific, societal and funding agency structure and habitat separation). Thus, one of our major goals was to balance process orientated topics with presentations from the three general aquatic environments, freshwater, estuarine and marine. An additional overarching aim was to bridge the geographical distribution of NATO and partner countries. Our workshop proposal is timely and compliments

xi

xii

NATO’s new approach involving partner countries because many partner countries have mainly freshwater and brackish systems while most NATO members also have large estuarine-marine components. In addition to chronicling the current status of suspension-feeder research, we believe that this workshop has and will foster greater communications between the various groups and support the cross-fertilisation strategy has been shown to have a strong positive effect on the generation of new scientific approaches, theories and knowledge. The participants are grateful for the financial and logistics support and guidance provided by the NATO staff. We also thank the Kluwer editorial staff for their timely and constructive support.

MODELLING PARTICLE SELECTION EFFICIENCY OF BIVALVE SUSPENSION FEEDERS

Petras Zemlys, Darius Daunys Coastal Research and Planning Institute, Klaipeda University, H.Manto 84, Klaipeda 5808, Lithuania

Abstract: The choice of an appropriate index to adequately describe the efficiency of preingestive organic material selection is important for modelling the material flux within the suspension feeding process. Recently, a new selection efficiency index was suggested by Zemlys et al. (2003) which simplifies the quantification of the selection activity. A simple equation with interpretable parameters calculates the selection efficiency index using literature values of uptake rate and food quality. This analysis suggests the possibility of developing more general and biologically interpretable models. Keywords: Pre-ingestive selection efficiency; suspension feeding bivalves; modelling food processing

INTRODUCTION Bivalve suspension feeders reject part of the food they filter as pseudofeces depending on seston concentration. An important process accompanying pseudofeces production is the selection of particles that result in an increase in the organic material fraction of the ingested food (Fig. 1). Although the physiological regulation of feeding and selection of particles in particular is rejected by some authors (Jørgensen 1996), a number of in vitro and in situ investigations (Kiørboe and Mølenberg, 1981; Hawkins et al. 1996; Defossez and Hawkins 1997; Ward et al. 1998; Hawkins et al. 1998; Schneider et al. 1998; Baker et al. 2000) confirm the selective feeding by bivalves. The selection efficiency of organic matter can be as high as 60% under certain conditions (Hawkins et al. 1998). Therefore, this efficiency may considerably change the organic and inorganic material ratio of ingested food and biodeposits. Although the pre-ingestive selectivity of particles by bivalves is generally recognised, the factors influencing preferential ingestion remain 1 R.F. Dame and S. Olenin (eds.), The Comparative Roles of Suspension-Feeders in Ecosystems, 1–9. ©2005 Springer. Printed in the Netherlands.

2 uncertain. Various proposed criteria include particle size, shape, motility, density, and chemical cues such as algal ectotrines (for review see Defossez and Hawkins 1997). The organic material selection activity defined as selection efficiency, however, can be quantitatively described by experimental relationships that already have been determined for some species of marine bivalves (Hawkins et al. 1996, 1998). These findings strongly suggest that particle selectivity should be included in models of bivalve feeding.

Food uptake by filtration

Preingestive selection, rejection

Suspended particles

Increasing the fraction of low quality particles

Pseudofaeces

Decreasing the fraction of low quality particles Ingestion

Digestion

Fig. 1. A conceptual model of food processing by suspension feeders exhibiting active particle selection.

When selection activity is absent the quantitative description of the allocation of seston fractions (organic, inorganic material, etc.) in pseudofeces and ingested food is easily expressed by the difference between food uptake rate and pseudofeces production or ingestion rate. In the case of selection activity there are two options. One option is to model the rejection (ingestion) rates of different seston fractions separately (e.g. Hawkins et al. 2002; Scholten and Smaal 1999). In this case, the selection efficiency index serves an auxiliary role and can be used as output variable characterising the selection activity. Alternatively, the selection efficiency index can be used to determine by mass balance the different seston fractions beforehand and to estimate the allocation of seston fractions in pseudofeces and ingested food (e.g. Bendell-Young and Arifin 2004). Problems with the last approach arise when the seston is fractionated into organic and inorganic material (Zemlys et al. 2003). They found that the choice of an appropriate selection efficiency index is critical. The most widely used selection efficiency index is that defined in Bayne et al. (1993) and based on the comparison of organic content

3 in pseudofeces and seston. However, Zemlys et al. (2003) demonstrated that using a different selection efficiency index based on comparison of organic content in seston and ingested food could result in series of advantages for food processing modelling. These advantages can be summarised as follows (see Zemlys et al. (2003) for details): x The traditional selection efficiency estimate is generally applicable to obtain organic content in pseudofeces and ingested food, however in some cases it may lead to analytically unsolvable equations. The introduction of an alternative selection efficiency considerably simplifies the solution and never requires iterative methods; x The analysis of constraints originating from mass balance determinations revealed an advantage of employing the alternative selection efficiency estimate. These constraints are more straightforward and obvious for alternative selection efficiency, for example the traditional selection efficiency is limited by certain value that is less than one while alternative index is limited by number one only (see formula (4) in the text below for more details); x Utilising the response surface approach for estimating the alternative selection efficiency as a function from food uptake rate and seston organic content is expected to produce a monotonously increasing function. This function might also have interpretational and analytical advantages in comparison to the traditional bell-shaped response surface. The aim of the study is to compare the response surfaces of traditional and newly defined selection efficiency index for three bivalve species Mytilus edulis, Cerastoderma edule and Crassostrea gigas and to determine an analytical expression for the alternative selection efficiency index. Data from the experimental studies of Hawkins et al. (1998) are used in this paper.

CALCULATION OF RESPONSE SURFACES Traditionally, the selection efficiency (SE) is defined as (Bayne et al., 1993) SE

(1

/

)

(1)

where FPOM M is organic content of seston (fraction particulate organic material), i.e. food quality; FPOMPF F is organic content of pseudofeces (for complete list of variables and parameters used see Table 1). An alternative

4 Table 1. List of notations with explanations. Notation a ase bse cse dse FPIM FPOM FPOM M0 FPIMING FPOMING FPOMPF IRMAX r SE SE1 SEMAX UPR

Explanation Parameter in selection efficiency equation (7) Parameter in equation (3) Parameter in selection efficiency equation (3) Parameter in selection efficiency equation (3) Parameter in selection efficiency equation (3) Inorganic content of seston Organic content of seston (fraction) The value of seston organic content at which the selection activity starts Inorganic content of ingested material Organic content of ingested material Organic content of pseudofeces Maximal ingestion rate, g day-1 Coefficient of proportionality in equation (6) Traditional selection efficiency Alternative selection efficiency Maximal possible selection efficiency value Seston uptake rate, g day-1

definition of selection efficiency based on organic content of ingested food (Zemlys et al. 2003) was defined as SE1 (

) /(1

)

(2)

where SE1 is alternative selection efficiency; FPOMING G is organic content of ingested food. SE1 is similar to selection index reported in Hawkins et al. (1998) but it is normalized by maximal value of difference FPOMING FPOM instead of FPOMING. The equation (2) can be expressed in form similar to (1) but in terms of inorganic content: SE1 1 FPIM / FPIMING

where FPIM is inorganic content in seston; FPIMING is inorganic content in ingested material. The advantages of the alternative index were demonstrated (Zemlys et al. 2003) however equations for evaluation of selection efficiency as a function of food uptake rate and seston organic content exist only for the traditionally defined selection efficiency (Hawkins et al. 1996; Hawkins et al. 1998). The limitations of these empirical equations due to complex response surfaces were also determined by Zemlys et al. (2003). They hypothesized that the response surface shape of newly defined selection efficiency should be considerably simpler. As the first step the response surfaces for both selection efficiency definitions should be constructed. Traditional selection efficiency depends on M). For three food uptake rate (UPR) and organic content of seston ((FPOM

5 bivalve species M. edulis, C. edule and C. gigas the following regression equation was proposed by Hawkins et al. (1998): SE

ase bse / FPOM

cse UPR dse UPR / FPOM

(3)

where ase, bse, cse and dse are parameters (see Hawkins et al. (1998) for numeric values). Unfortunately this equation is based on the rather narrow range of organic content and does not contain any theoretical information about possible shape of response surface outside this range. The analysis of material balance revealed (Zemlys et al., 2003) that traditionally defined selection efficiency is not increasing with UPR and FPOM M but is limited by certain value (SEMAX) X defined by formula (Zemlys et al., 2003) SEMAX

[(1

)/

]

/(

)

(4)

where UPR is seston uptake rate by bivalve; IRMAX X is maximal ingestion rate. SEMAX X is the value of SE E which corresponds to the saturation condition FPOMING 1 (Zemlys et al., 2003). The response surface for SE was constructed assuming SE min i ^SEMAX , SET ` , where SET T are selection efficiency index calculated by equation (3). The empirical equations for the calculation of UPR and IRMAX X that depend on organic content and seston concentration are taken from Hawkins et al. (1998). Does the saturation take place in real conditions or not the question is still open and needs further experimental investigation. However, the shape of SEMAX X surface gives enough information about possible shape of SE E surface because SE E is always less or equal to SEMAX X (Zemlys et al. 2003). The obtained SE surfaces for all three species are shown in Fig. 2a, 2c, 2e). The response surface for SE1 can be easy obtained from SE E surface using the following relationship between SE1 and SE E (Zemlys et al., 2003): SE1 [

/(( /(1

))]



/

1 SE

(5)

The results of these calculations are presented in Fig. 2b, 2d, 2f. For all three species the response surface of SE E is asymmetric bellshaped surface with the maximum along FPOM M axis which makes it difficult to approximate by simple interpretable analytical expression. In contrast, the SE1 surface is an increasing function with regard to both arguments monotonically approaching the value 1 and therefore might be approximated by traditional saturation functions, like Michaelis-Menten function or similar. Of course, the numerical analysis of three species only support the hypothesis but does not let to conclude that SE1 will be increasing function for other bivalve species, however as it is shown by Zemlys et al. (2003) the sufficient

6

Fig. 2.The SE E and SE1 response surfaces for three bivalve species: a), b) – Mytilus edulis; c), d) – Cerastoderma edule e), f) – Crassostrea gigas.

condition to be SE1 monotonically increasing is the monotonic increasing of relative pseudofeces production. It seems that this condition is realistic for majority of suspension feeding bivalve species at least regard to the uptake rate.

APPROXIMATION FUNCTION The description of SE1 directly as a function of UPR and FPOM is very important in order to use the alternative selection efficiency for modeling of food processing. We will demonstrate here that such a function could be

7 obtained using simple assumptions. The simplest assumption that seems to be acceptable for analyzing SE1 for all three species is the linearity of the relationship between SE1 changing rate regard to FPOM M and SE1 that can be formulated by following equation: wSE1 wF FPOM F

r((

) (1

(6)

)

where r(UPR) is coefficient of proportionality depending on uptake rate. Assuming also linearity for r(UPR), i.e. r (UPR ) a UPR after integration the following expression is obtained SE1 max(0 max(0, 1 exp(

(

0

)

))

(7)

0 , a is a parameter and is assumed that SE1=0 at where SE1(( 0) FPOM FPOM 0 . While SE1 and SE E equals to zero simultaneously the equation for FPOM0 can be obtained by assuming SE 0 in (3) and solving it regard to FPOM, M what results in FPOM0((

)

bse dse UPR ase cse UPR

(8)

The equation (6) has only one unknown parameter a which can be interpreted as multiplier determining the rate of SE1 approach to saturation condition (value 1). Together with (8) the equation (6) was used to approximate the surfaces given in fig. 2a, 2c, and 2g by minimum square method. The estimated values of parameter a and mean square error are given in Table 2.

Table 2. The results of approximation SE1 by equation (6) a

Mean square error

Mytilus edulis

7.89

0.034

Cerastoderma edule

10.08

0.066

Crassostrea gigas

3.59

0.159

Bivalve species

The estimated values of parameter a allows us to compare the ability of the different species to increase ingested food quality by means of selection

8 activity when seston food quality and uptake rate are increased. The highest value a 10.08 for C. edule (Table 2) shows that this species has the highest capability. Less, but still comparable ( a 7.89 ) capability has M. edulis while C. gigas is able to exploit less than half of selectivity potential estimated for other two species (Table 2).

DISCUSSION AND CONCLUSIONS The pre-ingestive food selection by bivalves is an important phenomenon determining the organic content in pseudofeces and ingested food, simultaneously controlling the energy fluxes inside the organism and between the organism and environment. In this paper we parameterize the newly proposed alternative selection efficiency index. We show that the response surface for the alternative selection efficiency index for three bivalve species has a simpler shape that is more proper for approximation by analytical expressions. An analytical expression based on simple assumptions for alternative selection efficiency index is proposed and we demonstrate that it can satisfactorily approximate the values of selection efficiency recalculated from traditionally defined selection efficiency for three bivalve species. The proposed relationship contains a parameter that enables the comparison of selection activity of different bivalve species, i.e. the capability to exploit the increase of food quality and food uptake to improve the ingested food quality. Our results lead us to believe that modeling of the allocation of organic and inorganic material in pseudofeces and in ingested food can be based on mass balance and alternative selection efficiency index that can be expressed by simple functions with interpretable parameters. This approach may make food processing models more general than models consisting of purely site and species specific regression equations. It is important to note that equation (7) could be more simplified. As can be seen from Fig. 2, the selection efficiency index depends much more on seston organic content than uptake rate, thus the dependence on uptake rate can be neglected (at least in some cases). The right side of equation (8) is a constant in this case. The traditional selection efficiency index as a function of seston organic content only was used Bendell-Young and Arifin (2004). In some cases other fractions of seston than organic and inorganic material are necessary. For example, the organic material was divided to phytoplankton and non-phytoplankton organics (Hawkins et al. 2002). The further development of the approach considered above for this case is an important task for future investigations.

9 REFERENCES Baker SM Levinton JS Ward JE 2000 Particle transport in the Zebra Mussel, Dreissena polymorpha (Pallas). Biol Bulll 199: 116-125 Bayne BL Iglesias JIP Hawkins AJS Navarro E Héral M Deslous-Paoli JM 1993 Feeding behavior of the mussel Mytilus edulis L.; responses to variations in both quantity and K 73: 813-829 organic content of seston. J Mar Biol Assoc UK Bendell-Young LI Arifin Z 2004 Application of a kinetic model to demonstrate how selective feeding could alter the amount of cadmium accumulated by the blue mussel (Mytilus ( trossolus). J Exp Mar Biol Ecol 298::21– 33 Defossez JM Hawkins AJS 1997 Selective feeding in shellfish: size-dependent rejection of large particles within pseudofaeces from Mytilus edulis, Ruditapes philippinarum and Tapes decussatus. Mar Bioll 129: 139-147 Hawkins AJS Bayne BL Bougrier S Héral M Iglesias JIP Navarro E 1998 Some general relationships in comparing the feeding physiology of suspension-feeding bivalve molluscs. J Exp Mar Biol Ecoll 219: 87-103 Hawkins AJS Duarte P Fang JG Pascoe PL Zhang JH Zhang XL Zhu MY 2002 A functional model of responsive suspension-feeding growth in bivalve shellfish, configured and validated for the scallop Chlamys farreri during culture in China. J Exp Mar Biol Ecoll 281: 13-40 Hawkins AJS Smith RFM Bayne BL Héral M 1996 Novell observations underlying the fast growth of suspension-feeding shellfish in turbid environments: Mytilus edulis. Mar Ecol Progr Ser 131: 170-190 Jørgensen CB 1996 Bivalve feeding revisited. Mar Ecol Prog Serr 142:287-302 Kiørboe T Mølenberg F 1981 Particle selection in suspension-feeding bivalves. Mar Ecol Prog Ser: 5: 291-296 Schneider DW Madon SP Stoeckel JA Sparks RE 1998 Seston quality controls zebra mussel ((Dreissena polymorpha) energetics in turbid rivers. Oecologia 117: 331-341 Scholten H Smaal AC 1999 The ecophysiological response of mussels (Mytilus edulis) in mesocosms to a range of inorganic nutrient loads: simulation with the model EMMY. Aq Ecoll 33: 83-100 Ward JE Levinton JS Shumway SE Cucci T 1998 Particle sorting in bivalves: in vivo determination of pallial organs of selection. Mar Bioll 131: 283-292 Zemlys P Daunys D Razinkovas A 2003 Revision of pre-ingestive selection efficiency definition for suspension feeding bivalves: facilitating the material fluxes modelling. Ecol Modell 166: 67-74

FIELD MEASUREMENTS ON THE VARIABILITY IN BIODEPOSITION AND ESTIMATES OF GRAZING PRESSURE OF SUSPENSION-FEEDING BIVALVES IN THE NORTHERN BALTIC SEA

Jonne Kotta1, Helen Orav-Kotta1,2 and Ilppo Vuorinen3 1

Estonian Marine Institute, University of Tartu, Mäealuse 10a, 12618 Tallinn, Estonia Institute of Zoology and Hydrobiology, University of Tartu, Vanemuise 46, 51014 Tartu, Estonia 3 Archipelago Research Institute, University of Turku, SF-20500 Turku, Finland 2

Abstract: Functional relationships between environmental variables, the biodeposition and clearance rates of Dreissena polymorpha and Mytilus edulis were estimated in the northern Baltic Sea. The biodeposition and clearance of the bivalves increased with ambient temperature. In more eutrophicated regions biodeposition and clearance rates increased curvilinearly with ambient concentrations of chlorophyll a and leveled off at high food concentrations. In less eutrophicated conditions a linear model gave the best fit suggesting that saturation level was not obtained. Additional variation in the biodeposition and clearance was explained by the interaction of water temperature, current velocity and chlorophyll a. Salinity had a significant effect on the biodeposition and clearance of D. polymorpha. The population of suspension-feeders cleared daily on average from 3 to 2426% of overlaying water in the littoral area constituting an important sink for primary production. Keywords: Baltic Sea, benthic grazing, Dreissena polymorpha, Mytilus edulis

INTRODUCTION Owing to elevated nutrient levels and consequent phytoplankton blooms a dramatic increase of dense populations of benthic suspension feeders has been recorded world-wide (Barnes and Hughes 1988; Kautsky 1995; Dame 1996). At high densities the suspension-feeders are capable to deplete phytoplankton (Cloern, 1982 Fréchette and Bourget 1985) and therefore control the standing stock and production of primary producers and limit via competition the growth of pelagic herbivores and fish (e.g. Officer et al. 1982; Møhlenberg 1995). Consequently, suspension-feeders are considered to play a key role in the stability of coastal ecosystems (Herman and Scholten 1990). 11 R.F. Dame and S. Olenin (eds.), The Comparative Roles of Suspension-Feeders in Ecosystems, 11–29. ©2005 Springer. Printed in the Netherlands.

12 In situ studies quantifying broad-scale effects of suspension-feeder populations are scarce and usually they are based on indirect evidence and modelling approaches (e.g. Cloern 1982; Møhlenberg 1995). It has been demonstrated that laboratory measurements are often difficult to interpret and compare (Riisgård, 2001) and they may overestimate the filtration rate by 1300% (Doering and Oviatt 1986; Cranford and Hill 1999). It suggests that suspension-feeders in nature exploit their full clearance capacity for short periods and more often feed at a much-reduced rate (e.g. Cranford 2001). Water temperature, salinity, the quality and concentration of seston coupled with flow regime have significant impact on the activity of suspension feeders (Bayne et al. 1977; Kiørboe et al. 1980; Widdows 1985; Fréchette at al. 1989; Asmus and Asmus 1993) and may account for the major variability of in situ feeding behaviour. Hence, there is a need for field measurements of the feeding behaviour of suspension-feeders combined with the measurement of those environmental variables in the near-bottom layer. These functional relationships have to be estimated for different areas and different times of the year, to assess the importance of suspension feeder grazing to the coastal ecosystem. Suspension-feeders derive their food by filtering the water column and retaining particulate matter on their gills. Clearance rate refers to an amount of water that is cleared per time unit by animal or biomass. Biodeposition is defined as the production of faeces and pseudofeces. An in situ biodeposition approach has been used to evaluate the variations in the feeding behaviour of mussels (Kautsky and Evans 1987; Hawkins et al. 1996; Cranford et al. 1998; Cranford and Hill 1999). By applying an in situ trap technique, the biodeposition was quantified in terms of carbon and nutrients. However, as phytoplankton is considered to be the prime food for benthic suspension-feeders, we chose, in contrast to these previous studies, chlorophyll a (Chl a) as a tracer. The present study focuses on the grazing impact of suspension-feeders on the pelagic algal community in the northern Baltic Sea. The blue mussel (Mytilus edulis Gould) and the zebra mussel ((Dreissena polymorpha (Pallas)) were selected as experimental species due to their ubiquity and, hence their significant potential contribution to phytoplankton removal. The functional relationships between ambient temperature, salinity, current velocity, phytoplankton biomass and the biodeposition of the suspension-feeders were estimated at five sites differing in their eutrophication level during different times of the year. Based on these functional relationships algal grazing by the mussel populations was estimated in multiple areas taking into account the data on ambient temperature, salinity, Chl a concentration, mussel abundance and size distribution.

13 MATERIALS AND METHODS M. edulis and D. polymorpha are the most conspicuous suspensionfeeders in the northern Baltic Sea. The species are most prevalent on hard bottoms above the halocline where, owing to low predation and high input of nutrients, they often form extensive multilayered mats (Segerstråle 1957; Kautsky 1981; Kautsky 1995; Öst and Kilpi 1997). D. polymorpha dominates at salinities less than 5 psu and M. edulis in more saline environments (Kotta, 2000). The study was carried out on three transects in the littoral zone of the Gulf of Riga (GOR) and two transects in the Gulf of Finland (GOF) during one year period between 1996 and 2002 (Fig. 1). Sampling was performed during ice-free period in spring (T=2–15ºC), summer (T>15ºC) and autumn (T=2–15ºC). Northern GOR was characterised by a wide and sheltered coastal zone with diverse bottom topography and extensive reaches of boulders. Depending on the salinity, a scattered population of M. edulis or D. polymorpha occurred on the boulders. The southern transect had a narrow and exposed coastal zone. Coarse sandy substrate prevailed down to a depth of 4 m being replaced by boulders at greater depths. The boulders harboured a dense population of D. polymorpha. Hard substrate prevailed at the northern GOF site. The coverage of M. edulis was almost 100 % along this transect. The southern GOF was characterised by a mixture of sand, pebbles and boulders above 3-m depth. Deeper down only sandy substrate is found and, hence, the area was practically devoid of suspension-feeding bivalves. As a result of the differences in exposure to deep waters, the frequency of upwelling was higher in GOF than GOR sites. Due to high riverine load and moderate water exchange the nutrient concentrations were on average twice as high in GOR than in the Baltic Proper. Northern GOR sites were moderately eutrophicated and southern GOR site was highly eutrophicated. The southern GOF site was moderately eutrophicated due to municipal pollution load of Tallinn City. The concentration of nutrients in the northern GOF site was similar to the values of the Baltic Proper and, hence, representing the least disturbed environment in terms of eutrophication (Astok et al. 1999; Hänninen et al. 2000). In each season the abundance, biomass and size-frequency distribution of the suspension-feeders were estimated along the five abovementioned and an additional transect in a more exposed part of northern GOF. Samples were collected from the seashore down to 12 m depth at steps of 1 m. Metal frames of 20u20 cm surface area were placed randomly on the bottom by a diver. All suspension-feeders within the frame were collected. Three replicates were taken at each location. The length of the bivalves was measured to the nearest 0.1 mm using vernier callipers. In situ grazing rates of M. edulis and D. polymorpha were estimated by quantifying the defecation of Chl a by the mussels at 1 m in each transect

14 during different seasons. Bivalves of 9–31 mm shell length were collected by a diver in the vicinity of deployment. Three individuals were placed on the net of the funnel allowing biodeposits to sediment to the collecting vial below. The near-bottom temperature and salinity were monitored at the beginning and at the end of the deployment using CTD profiling. Each incubation lasted 4 hours. In each season we performed at least five incubations replicated three times.

Fig. 1. Study area. The transects of M. edulis are indicated by crosses and that of D. polymorpha by open circles.

Except for northern GOF plaster balls were used to estimate the water currents in near-bottom layer. The method is recognised as a simple and inexpensive tool for measuring integrated water motion over a wide range of flow rates. The dissolution rates of plaster balls are mainly a function of water velocity and less influenced by salinity and temperature within a range of our study (Thompson and Glenn 1994).

15 After deployment the shell lengths were recorded, the sedimented material in the vials was sorted under a dissecting microscope; faeces were collected with a pipette and filtered on Whatman CF/F filters within 4 h of retrieval. Filters were extracted in dark in 96% ethanol overnight. Chl a was quantified fluorometrically correcting for phaeopigments (Pha) (Strickland and Parsons 1972). The values of Chl a equivalent or total Chl a (Chl a eq) were calculated as Chl a eq = Chl a + 1.52 u Pha. During deployment water for Chl a measurement was daily sampled by a diver at near-bottom layer along the whole transect at steps of 1 m. Additional samples were taken at a distance of 25 cm from the cages in connection with retrieving biodeposits (i.e. in every 4–12 h). Hence, the average concentration of Chl a sampled at the start and end of an incubation was used as a measure of food concentration during incubation. Filtration and extraction of these samples were carried out within 1 h after sampling. The water samples were filtered onto Whatman GF/F filters. Chl a and Pha were measured as noted above. In order to estimate the loss of Chl a during gut passage separate experiments were carried out aboard ship. The mussels were incubated in 5 l buckets for 4 h. Buckets without experimental animals served as controls. The animals were fed natural sea water. At the end of the incubation the biodeposits were cleaned from the buckets by careful pipetting and water samples for Chl a were taken. The content of Chl a and Pha were estimated in biodeposits and water samples as described above. The loss of Chl a during gut passage was estimated as the ratio of the loss of Chl a in water to biodeposit production taking into account the algal growth and sedimentation in the control bucket. Clearance rate by the mussel population was calculated from the estimates of biodeposition. The functional relations between biodeposition and environmental variables were determined after correction for loss of Chl a during gut passage. The data on ambient temperature, salinity, Chl a concentration, mussel abundance and size distribution were taken into account when estimating population grazing potential in multiple areas. Annual population grazing potential is defined as the average of the calculated clearance rates of each incubation by transect and depth interval. The minimum and maximum values represent the extremes of the calculated clearance rates. Grazing by individuals of different size (Gl) was scaled by shell length, i.e. Gl = G20 u l2/202, where G20 is the grazing rate of 20 mm individuals and l the shell length (Kiørboe and Møhlenberg, 1981). We assumed no significant spatial variation in current velocity and complete vertical mixing along the transect. At low current velocities water exchange was likely not sufficient to supply the local suspension-feeder communities with phytoplankton. Hence, the grazing potential tends to overestimate the impact of mussels on phytoplankton communities when the water exchange is low.

16 The biodeposition and clearance of the mussels were analysed by factorial ANOVA including transect and season as the main effects. We employed linear and second-order polynomial linear regression analyses to describe the relationships between the biodeposition and ambient environmental variables. Polynomial regression results are only reported if significantly better fits were achieved using this method compared with the linear model. Table 1. The mean values r S.E. of water temperature, salinity, Chl a eq (Pg l-1) and current velocity (cm s-1) at the study sites in the Gulf of Riga (GOR) and the Gulf of Finland (GOF) during different seasons. Site Seili (GOF-N)

Kakumäe (GOF-S)

Kõiguste (GOR-N)

Audrurand (GOR-N)

Saulkrasti (GOR-S)

Year 1998

Season summer

Temperature 16.4 r 0.2

Salinity 5.9 r 0.1

Chl a 4.9 r 1.4

1998

autumn

8.2 r 0.2

5.8 r 0.1

5.0 r 1.2

1999

spring

2.4 r 0.2

5.8 r 0.1

1.2 r 1.1

2002

spring

12.6 r 0.4

6.1 r 0.1

9.4 r 2.5

Current not measured not measured not measured 0.1r0.0

2002

summer

20.6 r 0.2

4.9 r 0.1

5.7 r 1.6

29.6r4.3

2002

autumn

4.4 r 0.5

5.2 r 0.2

8.0 r 3.5

43.0r2.5

1996

spring

5.9 r 0.2

5.5 r 0.1

19.2 r 1.0

12.1r0.3

1996

summer

16.2 r 0.1

5.7 r 0.0

4.0 r 0.7

0.1r0.0

2002

spring

13.3 r 0.4

3.2 r 0.1

1.7 r 2.5

1.3r0.8

2002

summer

24.0 r 0.3

3.1 r 0.1

6.5 r 1.9

42.9r7.6

2002

autumn

2.0 r 0.5

5.4 r 0.2

14.9 r 3.5

61.3r3.0

1996

spring

4.6 r 0.1

5.0 r 0.0

65.0 r 0.9

0.1r0.0

1996

summer

16.0 r 0.2

5.1 r 0.1

14.4 r 1.0

30.4r5.6

RESULTS In summer water temperature was higher in southern GOF and northern GOR site of D. polymorpha. Salinity values were slightly higher in GOF than in GOR sites. All means were less than 7 psu. The values of maximum water Chl a eq (i.e. a measure of eutrophication level) were higher in GOR than in GOF sites. Highest Chl a eq values were measured in the

17 southern GOR site during the spring bloom. The summer values in the southern GOR site were in the same magnitude as the spring bloom values in other studied sites. In general, current velocities were lowest in spring, intermediate in summer and highest in autumn. The values varied between sites being lowest in the northern GOR site of M. edulis, intermediate in southern GOF and southern GOR site of D. polymorpha and highest in the northern GOR site of D. polymorpha (Table 1). Table 2. The models of multiple linear regressions describing the biodeposition and clearance rates of Mytilus edulis and Dreissena polymorpha. The abbreviations are as follows: T – temperature, S – salinity, Chl – Chlorophyll a eq, Curr – current velocity, multiple terms indicate their interaction. P values of the regressions are lower than 0.001. Site Seili (GOF-N)

Kakumäe (GOF-S)

Kõiguste (GOR-N)

Audrurand (GOR-N)

Saulkrasti (GOR-S)

All sites

All sites

Species Mytilus

Mytilus

Mytilus

Dreissena

Dreissena

Mytilus

Dreissena

Model Biodeposition

Model terms T, T2, Chl2T2

R2 0.85

Clearance

TChl

0.72

Biodeposition

TChl, TCurr, Chl2T2

0.92

Clearance

TChlCurr

0.92

Biodeposition

T, T2Chl

0.85

Clearance

T2Chl

0.73

Biodeposition

TCurr

0.91

Clearance

Chl, ChlCurr

0.92

Biodeposition

SChl, T2Chl, T2Chl2

0.91

Clearance

T, TChl2, T2Chl, ST2Chl2

0.91

Biodeposition

T, T2, Chl, Chl2, TChl, Chl2T

0.57

Clearance

T, T2, Chl, Chl2, T2Chl, Chl2T2

0.57

Biodeposition

TChl

0.89

18

Fig. 2a. Biodeposition rate (Pg ind-1 h-1) of D. polymorpha as a function of ambient temperature, salinity and Chl a eq.

19

Fig. 2b. Biodeposition rate (Pg ind-1 h-1) of M. edulis as a function of ambient temperature, salinity and Chl a eq.

The clearance rates (l ind-1 h-1) increased curvilinearly with ambient temperature. There was a significant interaction between temperature and Chl a eq. The effect of Chl a eq varied between sites and seasons. The clearance rate of D. polymorpha decreased with increasing salinity. In southern GOF

20 and northern GOR site of D. polymorpha current velocity interacting with temperature and Chl a eq had significant effect on biodeposition (Table 2, Fig. 4a). M. edulis had significantly higher clearance rates than D. polymorpha (Fig. 4b). Similarly to the biodeposition values the clearance values were higher at GOF sites than at GOR sites (ANOVA: F1,0.4 = 77.76, p < 0.001) and increased from spring to autumn (ANOVA: F2,0.2 = 39.03, p < 0.001) (Fig. 5). Biodeposition rate (µg Chl a eq ind-1 h-1) was mainly a function of ambient temperature and Chl a eq. The biodeposition values increased curvilinearly with temperature and ambient Chl a eq. The effect of temperature interacted with Chl a eq. The biodeposition of D. polymorpha decreased with increasing salinity. There were statistically significant differences in the regressions between different basins, sites within a basin and seasons. In southern GOF and northern GOR site of D. polymorpha current velocity interacting with temperature had significant effect on biodeposition (Table 2, Fig. 2a, b). The two studied bivalve species did not differ in their biodeposition rates. In general, the biodeposition values were higher at GOF sites than at GOR sites (ANOVA: F1,5 = 24.75, p < 0.001). Biodeposition was lowest in spring, intermediate in summer and highest in autumn (ANOVA: F2,4 = 21.12, p < 0.001) (Fig. 3).

Fig. 3. Mean biodeposition rate (± 95% CI) of M. edulis (solid line) and D. polymorpha (broken line) in relation to region and season.

21

Fig. 4a. Clearance rate (l ind-1 h-1) of D. polymorpha as a function of ambient temperature, salinity and Chl a eq.

22

Fig. 4b. Clearance rate (l ind-1 h-1) of M. edulis as a function of ambient temperature, salinity and Chl a eq.

23

Fig. 5. Mean clearance rate (± 95% CI) of M. edulis (solid line) and D. polymorpha (broken line) in relation to region and season.

Fig. 6. Annual averages of population grazing potential of M. edulis and D. polymorpha along each transect (% of overlaying water cleared m-2 h-1). The minimum and maximum values represent the extremes of the calculated clearance rates of each incubation.

24 The major variability in population grazing potential (% of overlaying water cleared m-2 h-1) was due to the spatial differences in the density of bivalves. The grazing potential was highest at 2–6 m. The lack of hard substrate indirectly reduced the grazing pressure in the shallower areas of southern GOR and in the deeper areas of northern GOR and southern GOF (Fig. 6). When averaged over the transect the suspension-feeders removed daily on average from 3 to 2426% of phytoplankton stock in the coastal area. Population grazing decreased with increasing Chl a eq i.e. eutrophication level (Fig. 7).

Fig. 7. Relationships between annual averages of water Chl a eq and population grazing potential of bivalves in the littoral zone (0-12 m).

DISCUSSION Results of this study indicate that temperature and phytoplankton biomass were the major causes for temporal and spatial variations in biodeposition and clearance rates of mussels. Salinity was important factor for D. polymorpha. Current velocity affected biodeposition and clearance rates at sites where mussels were confined to the shallower depth of the transect (GOF-S, GOR-N D. polymorpha site). The relationships varied significantly between species, sites and seasons. The models predicted between 57 and 92% of the variability in the biodeposition or clearance rates of the bivalves. Regressions with low coefficient of determination were described in the areas where food conditions were unstable due to upwelling (GOF-N, GOR-N M. edulis site). At low temperatures (< 8 ºC) the biodeposition of the studied suspension-feeders was lower regardless of food conditions. Similar results were found earlier for M. edulis and D. polymorpha in the GOR and in the northern Baltic Sea (Kautsky and Evans, 1987; Kotta and Møhlenberg, 2002). Low pumping rates at low temperatures can be caused by temperature induced

25 changes in ciliary beat frequency and an increased viscosity of the water (Jørgensen et al., 1990; Loo, 1992). An active regulation of suspensionfeeding through food concentration (Newell and Bayne, 1980) is less likely. However, as low temperatures and low food levels coincide in many areas, low temperature can be considered a favourable condition during such periods of food shortage, reducing high costs of suspension-feeding when concentration of food particles is low. At higher temperatures the relationship between suspension-feeding and the water temperature was not constant but depended on the food supply and features of the studied basin. Suspension-feeding increased with rising ambient concentrations of Chl a and levelled off at high food concentrations. The saturation reduction occurred above 5–10 µg Chl a eq L-1. It is known that high algal concentrations may lead to reduced valve gapes and a reduction of the filtration rate (Riisgård and Randløv, 1981; Riisgård, 1991, 2001). Such high Chl a values were prevailing in GOR suggesting that fitted polynomial functions between suspension-feeding and the food supply reflected the saturation at high Chl a concentrations in GOR (Kotta and Møhlenberg, 2002). In the northern GOF site, however, a linear functional response between Chl a eq and clearance showed that such a reduction due to saturation was never reached. D. polymorpha is found in a wide range of salinities but each of its subspecies has narrow and different tolerance to salinity. In the Baltic Sea D. polymorpha can live in salinities up to 6 psu (Karatayev et al., 1998). In accordance to these findings the biodeposition and clearance of D. polymorpha decreased with increasing salinities at the southern GOR site. Although, salinity varied within the same range at the northern GOR site, the effect of salinity was not significant and interactions among other factors had a much greater impact on D. polymorpha. We may assume that differences in functional responses are due to food availability. Owing to high riverine load the events of low salinity and high Chl a eq coincided at the southern GOR site (correlation analysis: r = -0.37, p = 0.006) whereas Chl a eq increased with salinity at the northern GOR site (r = 0.78, p < 0.001). The comparison of the clearance rates of M. edulis showed about three times lower suspension-feeding activity in GOR than in the northern GOF site despite of similar temperatures and Chl a concentrations. Hence, temperature and Chl a concentrations per se can not explain the behavioural variation in filtering activity. Due to the difference in exposure the current velocities are likely higher in the northern GOF than in GOR sites. The vertical mixing of the water column increases the amount of available food and promotes the filtration activity of mussels (Dolmer 2000; Newell et al. 2001). In southern GOF and northern GOR site of D. polymorpha the inclusion of the current velocity interacting with temperature and Chl a eq significantly improved the models of biodeposition and clearance rate of M. edulis and D. polymorpha. Besides, the quality of the seston affects the filtration activity of mussels (Asmus and Asmus 1993). The GOR sites are

26 characterized by higher variability of phytoplankton communities than GOF sites. The share of diatoms is higher in GOR than GOF sites (HELCOM 1996, 2002). At the cleaner GOF site, where Chl a concentrations were more stable, the bivalves seemed to use their full filtration capacity regardless of the ambient food level. Due to the aggregated distribution the grazing rate of the populations of suspension-feeders varied significantly within and between study areas. The grazing rate was orders of magnitude higher in GOF than in GOR. The northern GOR has extensive shallow areas and moderate water exchange. Thus, the daily removal of 14% of Chl a on average in GOR may be sufficient for the benthic control of phytoplankton in the area. This was also reflected by the depletion of Chl a at the near-bottom water of mussel beds. In GOF, owing to a relatively small share of coastal area and significant water exchange, it is difficult to estimate the effect of suspension-feeders grazing on phytoplankton population at the regional scale. Nevertheless, the accumulation of biodeposits through mussel filtration was an important process at GOF sites, especially in the sites, that were more exposed to deep waters and housed higher densities of suspension-feeders. The utilisation rate of phytoplankton by suspensionfeeders varies with benthic boundary-layer flow conditions (Fréchette et al. 1989; Dolmer 2000; Newell et al. 2001). On the population level the biomass of suspension-feeders increases with the current intensity and frequency (Gili and Ballesteros 1992; Lesser et al. 1994). It is likely that in the outer archipelago stronger vertical mixing increased the amount of food available to the suspension-feeders and hence, supported higher biomasses. Lower values of population grazing in the middle archipelago might be attributed to the lower wave energy input to the system. In shallower areas phytoplankton production is low due to high turbulence and dense macrophyte assemblages whereas the growth of suspension-feeders is controlled by ice scouring. In deeper areas the suspension-feeders are food limited due to low current velocities. Consequently, the suspension-feeders have highest biomasses and impact at intermediate depths unless substrate is not limiting their distribution. The inverse relationship between water Chl a concentration and population grazing of bivalves indicated that under eutrophicated conditions the impact of suspension-feeders on pelagic communities is small relative to cleaner environments. The studied bivalves exploit their filtration capacity to about 5 µg Chl a l-1 (Clausen and Riisgård 1996). Above that level a considerable reduction of filtration rate is observed presumably caused by overloading of the alimentary canal (Riisgård 1991). Besides, very high sedimentation and concentration of inorganic particles often observed in eutrophicated areas are detrimental to suspension-feeding bivalves (Kiørboe et al., 1980). In this study we have observed the feeding behaviour of suspensionfeeders under a wide range of environmental conditions. We have demonstrated that an important share of phytoplankton is grazed by benthic

27 suspension-feeders in the northern Baltic Sea. There were significant interactions between temperature, phytoplankton biomass and current velocities when affecting the biodeposition and clearance rates of the bivalves. The values of clearance rates measured in this study were in accordance with earlier field observations (Cranford et al. 1998). This study does not only contribute to the knowledge of the functional relationships between suspension-feeding and environmental settings but it also shows the necessity of further in situ measurements on undisturbed suspension-feeders in order to explain and model large natural spatial variations of suspension-feeding.

ACKNOWLEDGEMENTS This research was supported by the grants of the Estonian Governmental Programme no 0182578s03 and Estonian Science Foundation grant no 5103.

REFERENCES Asmus H Asmus RM 1993 Phytoplankton-mussel bed interactions in intertidal ecosystems. In: Bivalve Filter Feeders in Estuarine and Coastal Ecosystem Processes, RF Dame (Ed), Springer-Verlag, Heidelberg, pp 57-84 Astok V Otsmann M Suursaar Ü 1999 Water exchange as the main physical process in semienclosed marine systems: the Gulf of Riga case. Hydrobiologia 393: 11-18 Barnes RSK Hughes RN 1988 An Introduction to Marine Ecology. Blackwell Scientific Publications, Oxford, 351 p Bayne BL Widdows J Worrall C 1977 Some temperature relationships in the physiology of two ecologically distinct bivalve populations. In: Physiological Responses of Marine Biota to Pollutants, FJ Vernberg A Calabrese FP Thurberg WB Vernberg (Eds). Academic Press, New York. pp 379-400 Clausen I Riisgård HU 1996 Growth, filtration and respiration in the mussel Mytilus edulis: no evidence for physiological regulation of the filter-pump to nutritional needs. Mar Ecol Prog Serr 141: 37-45 Cloern JE 1982 Does the benthos control phytoplankton biomass in South Francisco Bay? Mar Ecol Prog Serr 9: 191-202 Cranford PJ 2001 Evaluating the ‘reliability’ of filtration rate measurements in bivalves. Mar Ecol Prog Serr 215: 303-305 Cranford PJ Emerson CW Hargrave BT Milligan TG 1998 In situ feeding and absorption responses of sea scallops Placopecten magellanicus (Gmelin) to storm-induced changes in the quantity and composition of the seston. J Exp Mar Biol Ecoll 219: 45-70 Cranford PJ Hill PS 1999 Seasonal variation in food utilization by the suspension-feeding bivalve molluscs Mytilus edulis and Placopecten magellanicus. Mar Ecol Prog Ser 190: 223-239 Dame RF 1996 Ecology of Marine Bivalves: An Ecosystem Approach. CRC Press, Boca Raton, FL, 254 p Doering PH Oviatt CA 1986 Application of filtration rate models to field population of bivalves: an assessment using experimental mesocosms. Mar Ecol Prog Ser 31: 265275 Dolmer P 2000 Feeding activity of mussels Mytilus edulis related to near-bed currents and phytoplankton biomass. J Sea Res 44: 221-231

28 Fréchette M Bourget E 1985 Energy flow between the pelagic and benthic zones: factors controlling particulate organic matter available to an intertidal mussel bed. Can J Fish Aquat Sci 42: 1158-1165 Fréchette M Butman CA Geyer WR 1989 The importance of boundary-layer flows in supplying phytoplankton to the benthic suspension feeder, Mytilus edulis L. Limnol Oceanogrr 34: 19-36 Gili JM Ballesteros E 1992 Structure of cnidarian populations in Mediterranean sublittoral benthic communities as a result of adaptation to different environmental conditions. In: Homage to Ramón Margalef or Why There is Such Pleasure in Studying Nature, JD Ros N Prat (Ed), Publications Universitat de Barcelona. pp 243-254 Hawkins AJS Smith RFM Bayne BL Héral M 1996 Novel observations underlying the fast growth of suspension-feeding shellfish in turbid environments. Mar Ecol Prog Serr 131: 179-190 Hänninen J Vuorinen I Helminen H Kirkkala T Lehtilä K 2000 Trends and gradients in nutrient concentrations and loading in the Archipelago Sea, Northern Baltic, in 19701997. Estuar Coast Shelf Scii 50: 153-171 HELCOM 1996 Third Periodic Assessment of the State of the Marine Environment of the Baltic Sea, 1989-1993; Background document. Balt Sea Environ Proc 64B HELCOM 2002 Environment of the Baltic Sea Area 1994-1998. Balt Sea Environ Proc 82B Herman PMJ Scholten H 1990 Can suspension-feeders stabilize estuarine ecosystems? In: Trophic Relationships in the Marine Environment. Proc 24th Eur Mar Biol Symp, M Barnes RN Gibson (Eds). Aberdeen University Press, Aberdeen, pp 104–116 Jørgensen CB Larsen PS Riisgård HU 1990 Effects of temperature on the mussel pump. Mar Ecol Prog Serr 64: 89-97 Karatayev AY Burlakova LE Padilla DK 1998 Physical factors that limit the distribution and abundance of Dreissena polymorpha (Pall.). J Shellfish Res 17: 1219-1235 Kautsky N Evans S 1987 Role of biodeposition by Mytilus edulis in the circulation of matter and nutrients in a Baltic coastal ecosystem. Mar Ecol Prog Serr 38: 201-212 Kautsky N 1981 On the role of blue mussel Mytilus edulis L. in the Baltic ecosystem. Doctoral dissertation. Stockholm University, Sweden, 22 p Kautsky U 1995 Ecosystem processes in coastal areas of the Baltic Sea. Doctoral dissertation. Stockholm University, Sweden, 25 p Kiørboe T Møhlenberg F 1981 Particle selection in suspension-feeding bivalves. Mar Ecol Prog Serr 5: 291-296 Kiørboe T Møhlenberg F Nøhr O 1980 Feeding, particle selection and carbon absorption in Mytilus edulis in different mixtures of algae and resuspended bottom material. Ophelia 19: 193-205 Kotta J 2000 Impact of eutrophication and biological invasions on the structure and functions of benthic macrofauna. Dissertationes Biologicae Universitatis Tartuensis, 63. Tartu University Press, Tartu, 160 p Kotta J Møhlenberg F 2002 Grazing impact of Mytilus edulis L. and Dreissena polymorpha (Pallas) in the Gulf of Riga, Baltic Sea estimated from biodeposition rates of algal pigments. Ann Zool Fenn 39: 151-160 Lesser MP Witman JD Sebens KP 1994 Effects of flow and seston availability on scope for growth of benthic suspension-feeding invertebrates from the Gulf of Maine. Biol Bull 187: 319-335 Loo LO 1992 Filtration, assimilation, respiration and growth of Mytilus edulis L. at low temperatures. Ophelia 35: 123-131 Møhlenberg F 1995 Regulating mechanisms of phytoplankton growth and biomass in a shallow estuary. Ophelia 42: 239-256 Newell CR Wildish DJ MacDonald BA 2001 The effects of velocity and seston concentration on the exhalant siphon area, valve gape and filtration rate of the mussel Mytilus edulis. J Exp Mar Biol Ecoll 262: 91-111

29 Newell RC Bayne BL 1980 Seasonal changes in the physiology, reproductive condition and carbohydrate content of the cockle Cardium (Cerastoderma) edule (Bivalvia: Cardiidae). Mar Bioll 56: 11-19 Officer CB Smayda TJ Mann R 1982 Benthic filter feeding: a natural eutrophication control. Mar Ecol Prog Serr 9: 203-210 Öst M Kilpi M 1997 A recent change in size distribution of blue mussels (Mytilus edulis) in the western part of the Gulf of Finland. Ann Zool Fenn 34: 31-36 Riisgård HU 1991 Filtration rate and growth in the blue mussel, Mytilus edulis Linnaeus, 1758: dependence on algal concentration. J Shellfish Res 10: 29-35 Riisgård HU 2001 On measurement of filtration rates in bivalves – the stony road to reliable data: review and interpretation. Mar Ecol Prog Serr 211: 275-291 Riisgård HU Randløv A 1981 Energy budgets, growth and filtration rates in Mytilus edulis at different algal concentration. Mar Bioll 61:227-234 Segerstråle SG 1957 Baltic Sea. Mem Geol Soc America 67: 751-800 Strickland JDH Parsons TR 1972 A practical handbook of seawater analysis. Bull Fish Res Bd Can 167: 1-310 Thompson TL Glenn EP 1994 Plaster standards to measure water motion. Limnol Oceanogr 39: 1768-1779 Widdows J 1985 The effects of fluctuating and abrupt changes in salinity on the performance of Mytilus edulis. In: Marine Biology of Polar Regions and Effects of Stress on Marine Organisms, JS Gray ME Christiansen (Eds), J Wiley, Chichester. pp 555–556

CAN BIVALVE SUSPENSION-FEEDERS AFFECT PELAGIC FOOD WEB STRUCTURE?

Theo Prins1 and Vincent Escaravage2 1

National Institute for Coastal and Marine Management/RIKZ, PO Box 8039, 4330 EA Middelburg, The Netherlands 2 Netherlands Institute of Ecology, Yerseke, The Netherlands Abstract: Bivalve suspension-feeders are considered to be keystone herbivores in many estuarine ecosystems. However, bivalves can also feed upon organisms that belong to the microzooplankton and on mesozooplankton. Laboratory experiments showed that nauplii of the copepod Temora longicornis were filtered by mussels at the same rate as algae. Adult T. longicornis were also susceptible to filtration by mussels and oysters, but at a lower rate. Mesocosm experiments compared plankton dynamics in systems with and without mussels. Biomass of diatoms, heterotrophic dinoflagellates and copepods was strongly reduced in the presence of mussels. Some components of the microbial food web, like ciliates and Phaeocystis, did not show a significant effect, due to cascading effects of declining copepod abundance. It is suggested that in the presence of mussels, the pelagic food web may be shifted towards a more dominant microbial food web. Key words: Mytilus edulis, microzooplankton, copepods, grazing, trophic interactions

INTRODUCTION Grazing by bivalve suspension-feeders is considered a major process in many shallow coastal ecosystems, with the potential to control phytoplankton biomass development to a large extent. Observations and model calculations for a wide variety of estuaries and coastal systems, and results from mesocosm and enclosure experiments, support the notion that top-down control of phytoplankton biomass by bivalve suspension-feeding is a widespread phenomenon in bivalve dominated systems (see e.g. Dame 1996). In view of the extensive documentation of bivalve grazing control of phytoplankton biomass, it is rather surprising that there is only limited information on the effects of bivalve grazing on other components of the plankton, particularly zooplankton. Bivalve suspension-feeders may have an indirect effect on herbivorous zooplankton through food competition, but may 31 R.F. Dame and S. Olenin (eds.), The Comparative Roles of Suspension-Feeders in Ecosystems, 31–51. ©2005 Springer. Printed in the Netherlands.

32 also directly influence zooplankton populations through filtration (Davenport et al. 2000). Carlson et al. (1984) made in situ observations of decreased zooplankton abundance on ebb tides after passage over an intertidal flat with mussels. However, differences were not tested statistically, and the significance of other factors causing the observed decrease in zooplankton abundance (e.g. settlement, other predators) could not be excluded. In recent years, several studies have addressed the effects of bivalve filtration on zooplankton species. There is experimental evidence of filtration of various microzooplankton species by the oyster Crassostrea gigas (Le Gall et al. 1997, Dupuy et al. 1999, Dupuy et al. 2000). It is suggested that feeding by oysters on microzooplankton like protists may constitute a trophic link between the microbial food web and benthic suspension-feeders (Le Gall et al. 1997, Dupuy et al. 1999). Other evidence of a negative impact of bivalves on microzooplankton comes from 3 enclosure experiments, lasting 2-3 weeks, with a mussel biomass in the enclosures high enough to filter the water volume once a day (Riemann et al., 1988). Significantly lower biomass of ciliates and rotifers was observed in enclosures with mussels Mytilus edulis in contrast to enclosures without mussels. However, no effects of the mussels on the abundance of the copepod Acartia tonsa was observed, presumably due to escape responses of the copepods (Horsted et al., 1988). Field observations of declines in copepod abundance, following the introduction of Potamocorbula amurensis in San Francisco Bay, were ascribed to predation by this clam on copepod nauplii (Kimmerer et al. 1994). Experiments demonstrated that bivalve suspension-feeders could capture and ingest copepod nauplii and adults (Kimmerer et al. 1994, Davenport et al. 2000). Escape responses of copepods show interspecific differences (Titelman 2001, Green et al. 2003). Differences in vulnerability of copepod species to bivalve predation could lead to bivalve control of zooplankton species composition, as suggested by Kimmerer et al. (1994). It can be concluded that bivalve filtration can have direct effects through predation on the abundance of zooplankton species. Depending on the zooplankton species involved, bivalve filtration could have an effect on the microbial food web by removing microzooplankton species, or it could have an impact on the classical food chain by predation on copepods. Interspecific differences in susceptibility of zooplankton organisms to bivalve grazing, combined with trophic interactions between different components of the planktonic food web, may result in cascading effects of bivalve grazing on pelagic food web structure. In this study, we use results from various sources to address the hypothesis that bivalve suspension-feeders change pelagic food web structure through predation on zooplankton. Laboratory experiments, aimed at establishing the effects of bivalve filtration on different life-stages of copepods, were carried out to determine the direct, short-term effects of bivalve filtration on copepod survival. Experimental ecosystems (mesocosms)

33 were used for experiments to study the longer-term effects of bivalve filtration on the development of pelagic populations and the composition of the planktonic community.

MATERIALS AND METHODS

Lab Experiments To establish the direct effects of filtration by the blue mussel M. edulis and the pacific oyster C. gigas on naupliar larvae and adult stages of the copepod T. longicornis experiments were carried out between September and December 2002, under controlled conditions at the RIKZ field station. Mussels and oysters had been collected in September 2002 in the Oosterschelde estuary (SW Netherlands). The individual animals were glued on small sticks, and were stored in raceways with flowing natural seawater between experiments. In experiments with mussels (between 15-59 mm shell length), individual animals were placed in 1.15 l bottles, containing a mixture of seawater, algal cells (Rhodomonas ( sp.) and either nauplii or adults collected from a culture of T. longicornis. Mussels were incubated for 30-60 minutes in the bottles that were placed on a rotating wheel to ensure complete mixing of the water in the bottles. Experiments with oysters (67-112 mm shell length) were carried out by placing the oysters in 2.2 l bottles with a mixture of seawater, algae and adult copepods. The water in the bottles was mixed with air bubbles. Initial algal concentrations were kept between 10 and 20 •106 cells l-1, initial copepod concentrations were between 10 and 40 animals l-1. Algal cell concentrations were below levels where the bivalves reduced filtration rates in response to high algal concentrations. Algal cell concentrations at the start and the end of the experiment were counted with a Coulter Counter. Copepods at the start of the experiment were counted while manually adding the copepods to the experimental vessel. At the end of the experiment, the water from the experimental bottles was filtered through a net with a 55 µm mesh size, and copepod abundances were counted under a stereoscope. Each experiment consisted of a series of incubations, with up to 12 bivalve incubations and 2 or more controls (experimental vessels without bivalves). In total, six experiments were carried with nauplii and various sizes of mussels. With adult copepods, two experiments were done with mussels, and four with oysters. Clearance rates CR on algae and on copepods were calculated from the following formula (Coughlan 1969):

34 CR = V/t * ln(C0/Ct) where V = volume of incubation bottle t = time of incubation Co = plankton concentration at the start of the experiment Ct = plankton concentration at the end of the experiment For each experiment, a t-test was used to determine if there was a significant decrease in algal or copepod concentrations in the control incubations.

Design of the Mesocosms The land-based mesocosms are located at the field station of RIKZ near the mouth of the Oosterschelde estuary. The mesocosms consist of black solid polyethylene tanks (height 3 m, diameter 1.2 m, volume 3000 l). The water column in the mesocosms was completely mixed. Daily cleaning prevented fouling of organisms on the walls of the tanks. Inorganic nutrients were continuously added to each of the mesocosms at a rate of 3.7 mmol N m3 , 0.06 mmol P m-3, and 0.8 mmol Si m-3. The mesocosms were flushed with natural seawater at a rate of 100 l day-1, resulting in a residence time of 30 days. Water from the bottom layer of each mesocosm was circulated through a 16 l benthos chamber containing mussels, at a rate of 45 l h-1 by means of a tubing pump. An extensive description of the mesocosms is given in Prins et al. (1995).

Design of the Mesocosm Experiment The period considered was from 30 March 1998 to 20 May 1998, as part of a more extensive mesocosm experiment (see Escaravage and Prins 2002 for details). In this study, we will compare two mesocosm units with a high biomass of mussels M. edulis (“MUS”) with two systems without a benthos and hence a solely pelagic community (“ZOO”). All other experimental conditions were similar in both treatments. In the MUS treatment, 80 mussels were added to each mesocosm. This was equivalent to a mussel biomass of 2.1 g ADW m-3 and a filtration rate of approximately 300 l day-1, or 10% of the mesocosm volume per day. Mussels had been collected a week before the experiment started, from the low tide level at a site close to the field station. At the start of the experiment, natural seawater from the Oosterschelde estuary was added to the mesocosms. Nutrient additions were the same for all mesocosms, and equivalent to loadings to the Dutch coastal zone (Prins et al. 1999). As earlier experiments had indicated that copepod

35 development in the mesocosms could be limited due to low concentrations at the start of the experiment (Prins et al. 1999), copepod biomass was enhanced by addition of zooplankton collected in the field. At the day of the start of the experiment, 12 m3 of water was pumped from approximately 5 m below the surface in the mouth of the Oosterschelde estuary. This water was immediately filtered through a 55 µm plankton net. The collected net plankton was subdivided into 4 subsamples (equivalent to the biomass in 3 m3 of water), and added to each mesocosm within 3 hours after collection. Extensive descriptions of sampling methods and analytical procedures for particulate and dissolved nutrients, chlorophyll-a, phytoplankton and micro- and mesozooplankton biomass, and primary production and bacterial production are given in Escaravage et al. (1995), Prins et al. (1995, 1999), Escaravage and Prins (2002). Samples were collected once a week. Differences between treatments were tested with a two-way ANOVA, with the experimental treatment (presence/absence of mussels) and time as independent factors, and the replicate mesocosms nested within the treatment factor. There was one observation per cell, and the error term for the treatment effect is the nested factor. This design is equivalent to a repeated measures design. Abundance data were log-transformed to reduce heterogeneity of variances.

RESULTS

Lab Experiments Six experiments were done with mussels and nauplii, and two with mussels and adult copepods. In three of these experiments, small but significant changes in concentrations of algae ((Rhodomonas sp.) were observed in the controls, but changes in control concentrations of the copepod T. longicornis were not observed in any of the experiments. Four experiments were done with oysters and adult copepods. In two of these experiments the controls showed small but significant decreases in algal concentrations, but no changes in concentrations of the copepods occurred. In the experiments with significant changes in control algal concentrations, clearance rates were corrected for algal sedimentation in the controls.

36

4

CRcopepods (l h-1)

3

2

1

0 0

1

2

3

-1

CRalgae (l h )

Figure 1. Clearance rates of mussels, calculated from changes in numbers of T. longicornis nauplii (CRcopepods), plotted against clearance rates calculated from Rhodomonas sp. concentration changes (CRalgae).The line y = x is indicated.

3

-1

CRcopepods (l h )

Mussel Oyster

2

1

0 0

1

2

3

4

-1

CRalgae (l h )

Figure 2. Clearance rates of mussels and oysters, calculated from changes in numbers of T. longicornis adults (CRcopepods), plotted against clearance rates calculated from Rhodomonas sp. concentration changes (CRalgae). The line y = x is indicated

37 Individual mussel clearance rates measured in experiments with T. longicornis nauplii are shown in Figure 1. There was no significant difference between clearance rates estimated from changes in algal concentrations (CRalgae), and clearance rates estimated from the decrease in copepod abundance (CRcopepods) (paired t-test, p>0.05). In the experiments with adult copepods, significant decreases in the concentrations of copepods during incubation were observed. However, in contrast to the results with nauplii, there was a lack of correlation between CRcopepods and CRalgae (Figure 2). The clearance rate of mussels on adult copepods (mean ± SD: 0.19 ± 0.21, n=24) was significantly lower (t-test, p2.5 l h-1) were excluded.

Mesocosm Experiment Water temperatures increased from values around 10 °C at the start of the experiment (day 90) to temperatures above 15 °C at day 140. Initial inorganic nutrient concentrations were high (DIN: 178 µM, DIP 4.0 µM, Si 49 µM), and decreased constantly during the course of the experiment. At the end of the experiments inorganic nutrient concentrations (DIN: 114 µM; DIP 0.16 µM; Si 2.0 µM) were still above levels that are considered to be limiting phytoplankton growth (limiting levels: DIN: 2 µM; DIP 0.1 µM; Si 2 µM; Escaravage et al. 1999). No significant differences in nutrient concentrations between treatments occurred. Average biomass of phyto- and zooplankton, and levels of bacterial and primary production are shown in Table 1. Initial phytoplankton biomass was low, with a chlorophyll-a concentration of 1.1 µg l-1. There was a significant treatment effect on chlorophyll concentrations (ANOVA, p 80% of zebra mussel production (Lvova 1977, Yablonskaya 1985) and birds may eat 20 to 70% of annual D. polymorpha production (Mikulski et al. 1975, Stempniewicz 1974). However

249 there is no evidence of any long-term decline of zebra mussel populations due to the effects of predation (Molloy et al. 1997). Similar animals that feed on zebra mussels are reported to consume C. fluminea, including 14 species of fish, 13 species of ducks, racoons, crayfish, and flatworms (reviewed in Sickel 1986). There is some evidence suggesting that fish predation may be a major cause of reduction in C. fluminea density (Dreier and Tranquilli 1981, Robinson and Wellborn 1988). In Fairfield Reservoir, Texas, fish predation reduced C. fluminea abundance 29 fold (Robinson and Wellborn 1988).

Parasites 34 species of endosymbionts are known to be associated with zebra mussels, including ciliates, trematodes, mites, nematodes, leeches, chironomids, oligochaetes, and bacteria (reviewed in Molloy et al. 1997). At least six species of ciliates (Conchophthirus acuminatus, C. klimentinus, Hypocomagalma dreissenae, Sphenophrya dreissenae, S. naumiana, and Ophryoglena sp.) are known to be species specific. There is also some evidence that the trematodes Bucephalus polymorphus, Phyllodistomum folium, and P. dogieli, are quite specific to Dreissena. All of these parasites are found exclusively in Europe. Only nonspecific symbionts (e.g., nematodes, chironomids, oligochaetes, mites) are found in North American zebra mussels. Only one parasite, the trematode B. polymorphus, has been well documented as being seriously debilitating to zebra mussels (i.e., it destroys gonads) (Molloy et al. 1997). In contrast to the wide variety of endosymbionts found in zebra mussels, only two species are known to be associated with C. fluminea: the oligochaete Chaetogaster limnaei (Sickel and Lyles 1981) and a mite (authors unpublished data). The endosymbiotic fauna of C. fluminea in their native region may be more diverse. C. fluminea could be a second intermediate host of Echinistoma revolutum and may be a vector of echinostomiasis in humans (Anazawa 1929). There are no data on the effect of C. fluminea parasites on their abundance.

ECOSYSTEM IMPACTS Local Effects D. polymorpha attaches by byssus to hard substrates and each other, and can create new 3-D habitat, providing not only food, but shelter for bottom invertebrates. These effects on benthic communities are well documented (reviewed in Karatayev et al. 1997, 2002). Zebra mussels have positive effects on isopods (Wolnomiejski 1970, Karatayev and Lyakhnovich

250 1990, Kuhns and Berg 1999), larval chironomids (Wolnomiejski 1970, Botts et al. 1996, Stewart et al. 1998, Kuhns and Berg 1999), leeches (Wolnomiejski 1970), snails (Karatayev et al. 1983, Stewart et al. 1998, reviewed in Strayer 1999), amphipods (Karatayev et al. 1983, Karatayev and Lyakhnovich 1990, Botts et al. 1996, Stewart et al. 1998, Kuhns and Berg 1999, Riccardi 2003), oligochaetes (Afanasiev 1987, Botts et al. 1996), turbellarians (Botts et al. 1996), and hydrozoans (Botts et al. 1996, Stewart et al. 1998). Negative effects are documented for native unionids (reviewed in Schloesser et al. 1996, Karatayev et al. 1997, Burlakova et al. 2000, Riccardi 2003), chironomid larvae (Sokolova et al. 1980, Karatayev et al. 1983) and sphaeriid bivalves (Strayer et al. 1998, Lauer and McComish 2001, Mills et al. 2003). In contrast C. fluminea is solitary, burrows in sediments and does not change the surface of the sediments. Therefore its role in benthic communities may be much smaller. To date there is no evidence of effects of C. fluminea on benthic macroinvertebrates (Karatayev et al. 2003a) or on meiofauna (Hakenkamp et al. 2001). McMahon (1999) hypothesised that C. fluminea detrital feeding could negatively impact other burrowing detritovores. However, in experiments conducted in Lake Nacogdoches, where clams were placed at different densities in trays with sand and the benthic community was allowed to develop for 30 d, there was no difference in species composition or the density of benthic animals with or without live C. fluminea, independent of clam density (R. Mood, A. Karatayev and L. Burlakova, unpublished data). The shells of zebra mussels may accumulate in large quantities and alter the sediments and change benthic community (Karatayev et al. 2002). Similarly, C. fluminea dead shells may affect the benthos (Prokopovich 1969). In the experiments described above, amphipod densities were significantly higher in trays of sand with dead C. fluminea shells than pure sand without shells (R. Mood et al. unpublished data).

System-wide Effects Zebra mussels and Asiatic clams are functionally different than most benthic invertebrates in freshwater. They filter large volumes of water and transport material removed from the water column to the benthos, providing a direct link between processes in the plankton and those in the benthos (benthic-pelagic coupling). The shift of suspended matter from the water column to the bottom induces changes in all aspects of freshwater ecosystems they invade (reviewed in Morton 1997, Karatayev et al. 1997, 2002, McMahon 1999, Vanderploeg et al. 2002, Mayer et al. 2002, Mills et al. 2003) (Table 3).

251 To a large extent the overall impact of D. polymorpha and C. fluminea as suspension feeders on freshwater ecosystem may be similar, however, information is much more available for zebra mussels than Asiatic clams. The filtering activity of both species causes water transparency to increase and decreases seston concentration, BOD, and phytoplankton density (Table 3). With increased transparency, a greater portion of the lake bottom covered with macrophytes. Increased macrophyte beds may provide additional substrate for the zebra mussel attachment and thus increase D. polymorpha populations. In contrast, increased macrophyte beds may cover previously available substrate for C. fluminea and negatively affect their overall density in a waterbody. Table 3. The impact of Dreissena polymorpha and Corbicula fluminea on freshwater ecosystems. Parameter Water transparency Seston concentration

D. polymorpha Increases (reviewed in Karatayev et al. 1997, 2002, Vanderploeg et al. 2002) Decreases (reviewed in Karatayev et al. 1997, 2002, Vanderploeg et al. 2002)

BOD in the water Nutrients

Decreases (reviewed in Karatayev et al. 1997, 2002) Alters nutrient cycling (Johengen et al. 1995, Arnott and Vanni 1996, Makarewicz et al. 2000) Decreases density and chlorophyll content (reviewed in Karatayev et al. 1997, 2002, Vanderploeg et al. 2002)

Phytoplankton

Macrophyte coverage Periphyton Zooplankton Zoobenthos Fish

Increases (reviewed in Karatayev et al. 1997, 2002, Vanderploeg et al. 2002) Increases (Lowe and Pillsbury 1995). Decreases (reviewed in Karatayev et al. 1997, 2002) Increases (reviewed in Karatayev et al. 1997, 2002) Increases quantity of benthophages (reviewed in Karatayev et al. 1997, 2002)

C. fluminea Increases (Buttner 1986, Phelps 1994) Decreases (Buttner 1986, Leff et al. 1990, McMahon 1999) Decreases (Buttner 1986) Alters nutrient cycling (Beaver et al. 1991, Lauritsen and Mozley 1989) Decreases density and chlorophyll content (Cohen et al. 1984, Beaver et al. 1991) Increases (Phelps 1994, McMahon 1999) No data No data No data Increases (Phelps 1994)

No comparable data are available for the impacts of C. fluminea on periphyton, zooplankton, and benthic animal communities. In contrast, the impacts of D. polymorpha in these communities is well documented (Table 3). Zebra mussel filtering results in periphyton and benthic algal increases in both standing stock and primary productivity. Total zooplankton density and biomass decreases. Introduction of both C. fluminea and D. polymorpha may result in increased fish production (Table 3). Although much more data are available on the impact of zebra mussels on fish, generalizations are far from being

252 clear. Many authors have reported an enhancement of all benthic feeding fishes, even those that do not feed on zebra mussels, because zebra mussel invasion is often associated with an increase in biomass of native benthic invertebrates (e.g., Kharchenko and Protasov 1981, Lyakhnovich et al. 1988, Karatayev and Burlakova 1995a, Stewart and Haynes 1994). In contrast, planktivorous fishes could be negatively affected because of decreased phytoplankton abundance and associated decreases in zooplankton, competition with benthic species, and by increasing fish predation on larvae due to increased water transparency (Francis et al. 1996, Lozano et al. 2001). The decline in abundance and body condition in lake whitefish (Coregonis clupeaformis) in lakes Ontario and Michigan (USA) is believed to be related to a decline of Diporea hoyi, an important item in fish diets, following the appearance and proliferation of dreissenid mussels (Hoyle et al. 1999, Pothoven et al. 2001).

GENERAL FINDINGS AND FUTURE DIRECTIONS Although many generalisations can be made about the impacts of Asiatic clams and zebra mussels and their function in freshwater systems, specific predictable impacts are far from clear. The most important aspects of this problem and needed targets for future study are:

Methodological Problems Because these two species are invaders and can cause environmental and economic damage, many aspects about their biology and ecosystem impacts are simplified or exaggerated to draw attention to the problem of invasive species and their spread. Many of these generalisations and exaggerations are then repeated or assumed proven without scientific rigor. Scientists must be careful, especially when extrapolating from short-term laboratory experiments to large scale and long term effects of invaders. We need more studies that link these two approaches before we can draw accurate predictions or assess real impacts. The methodology used to determine impacts is also critical because different methods often yield different results. For example, filtering rates for both Dreissena and Corbicula when feeding on mixed plankton versus single species, as well as filtered versus unfiltered lake water (with seston concentration higher than the incipient threshold seston concentration) may differ, and measures in small volumes of still water are likely to be different than measures made in larger volumes and flowing water. Filtration rates are

253 also reported based on different units. They may be calculated on shell length, wet total mass (WTM, shell plus soft tissue), dry body mass (DBM, soft tissue only), or per ash-free dry mass (AFDM, soft tissue only), and it is not often clear how to convert among these different units. Former Soviet Union scientists generally calculate the filtration rate of D. polymorpha based on shell length or WTM (Lvova 1977, Karatayev and Burlakova 1995b), as do many other Europeans (Reeders and Bij de Vaate 1990, Wisniewski 1990), although some Europeans calculate filtering rate per DBM (Kryger and Riisgård 1988). The majority of North American scientists also calculate the filtration rate of zebra mussels per DBM (Aldridge et al. 1995) or per AFDM (Fanslow et al. 1995, Lei et al. 1996). Common units are essential for cross-study comparisons. We suggest that for zebra mussels the most appropriate units to use are mL of water filtered per g WTM per hour. WTM is very easy to measure, even in the field, and for individuals is much more highly correlated with filtration rate than other measures such AFDM or DBM, which vary greatly with season and reproductive condition (Karatayev 1983). We also recommend that field estimates of filtering rates for Dreissena and Corbicula be calculated as a function of WTM, not density. Different sized mussels will filter at different rates, and similar densities of mussels with different size frequency distributions will have dramatically different filtering rates and therefore their ecosystem impact may vary widely (Young et al. 1996). In any case, appropriate conversions among measures need to be established.

C. fluminea and D. polymorpha Co-effects To date, there are no data on the co-effects of D. polymorpha and C. fluminea invasion on aquatic communities. Both of these invaders continue to spread throughout North America and Europe, and increasingly they are both found in the same freshwater bodies. The effects of both Dreissena and C. fluminea may be additive, or we may see synergistic effects, where their impacts are much more than would be expected by the impacts of either species alone.

C. fluminea vs. D. polymorpha Distribution Although both of these invaders are frequently lumped together because they are bivalves and invade fresh waters, the ecology of D. polymorpha and C. fluminea are different, and therefore their ecosystem impacts are likely to be different. D. polymorpha are more abundant in lakes and large rivers and do not occur in high densities in streams. In contrast, in addition to lakes and large rivers, C. fluminea may be extremely abundant in small streams. Therefore, we may expect to find both similarities and

254 differences in the ecosystem response to the presence of these two invaders. Although they overlap, each has areas where the other is less abundant and they rarely compete for space. The presence of one will not necessarily eliminate the other, and the relative interactions between the two will depend on characteristics of the system, e.g., streams vs. lakes vs. canals.

C. fluminea vs. D. polymorpha Local Effects The impacts of Asiatic clams and zebra mussels, or any other biological agent, are likely to be most intense close to individuals. The biology and natural history of the zebra mussel and Asiatic clam are different. D. polymorpha can live only on the surface of the sediments where they attach to hard substrates creating structure, and providing food and shelter for benthic species. In contrast C. fluminea lives solitary, burrows in sediments and does not alter the surface of sediments. However the accumulation of dead shells of both species may have a similar effect by altering substrate and thereby affecting the benthic community. Moreover, although both species are suspension feeders, C. fluminea also collects food particles from the sediment. Therefore C. fluminea may compete for food with benthic infauna.

C. fluminea vs. D. polymorpha System-wide Effects Depending on water mixing rates, lake morphology, and turnover rates, the effects of suspension feeders on aquatic ecosystems will vary greatly (Ackerman et al. 2001) and may be very local in deep water lakes (ReedAndersen et al. 2000). Although D. polymorpha and C. fluminea impacts on the environment may be similar, feed backs may be different for different invaders. Invasion of both species may increase macrophyte coverage of waterbodies they invade, however increased macrophyte community will provide additional substrate for zebra mussels attachment and therefore may cause further increase of D. polymorpha population size and their impact on ecosystem. In contrast, increased macrophyte coverage may decrease habitat available for C. fluminea and therefore may cause a decrease of Asiatic clam population size and their impact on ecosystem.

Freshwater vs. Marine Bivalves Marine and estuarine bivalves have long been recognized as ecosystems engineers (reviewed in Dame 1993). Invasive zebra mussels and

255 Asiatic clams function similarly in fresh waters. Contrasting the impact of freshwater invasive bivalves on cosystems they recently colonized vs. estuarine, and marine bivalves in ecosystems were they were naturaly dominant but are recently lost, may help us to undestand the role of these important suspension feeders as ecosystem engineers in various waterbodies. Both Dreissena and Corbicula are important freshwater invaders and, as a consequence, have been the focus of much research. However, we still have to learn a great deal about both their biology as well as their impacts on ecosystems that they invade. It is clear from the direct comparison of these two species there remain many “missing pieces” of the picture. We hope that our review will help to focus future efforts such that we will be able to construct the “whole picture” for these two aggressive invaders and their effects on the freshwater ecosystems. These two invaders will continue to provide important information about the capabilities of suspension feeders as well as the functioning of freshwater ecosystems. In addition, these species may be important models that will help us predict the spread and impacts of future invaders.

ACKNOWLEDGEMENTS We would like to acknowledge the support provided by Stephen F. Austin State University (Faculty Research Grant # 14123 to AYK, LEB and DKP, 2003 - 2004).

REFERENCES Abbott TM 1979 Asiatic clam (Corbicula fluminea) vertical distributions in Dale Hollow Reservoir, Tennessee. In: Proceedings, First International Corbicula Symposium, JC Britton (Ed). Fort Worth, Texas, USA, Oct. 13-15, 1977. pp 111-118 Ackerman JD Loewen MR Hamblin PF 2001 Benthic-pelagic coupling over a zebra mussel reef in Western Lake Erie. Limnol Oceanogrr 46: 892-904 Afanasiev SA 1987 The differences in oligochaete distribution in periphyton on substrate with different structure. In: Water Oligochaeta. Proceedings of the 6th All Union Symposium, Riga, Lithuania. pp 38-41 (in Russian) Aldridge DW Payne BS Miller AC 1995 Oxygen consumption, nitrogenous excretion, and filtration rates of Dreissena polymorpha at acclimation temperatures between 20 and 32oC. Can J Fish Aquat Scii 52: 1761-1767 Alimov AF 1981 Functional Ecology of Freshwater Bivalve Moluscs. Nauka Press, Leningrad. 248 p (in Russian) Anazawa K 1929 On a human case of Echinostoma revolutum and its infection route. J Formosan Med Assoc 288:221 Arnott DL Vanni MJ 1996 Nitrogen and phosphorus recycling by the zebra mussel ((Dreissena polymorpha) in the western basin of Lake Erie. Can J Fish Aquat Sci 53: 646-659

256 Arias J 2004 Population dynamics and growth rate of Corbicula fluminea in two lotic systems of East Texas. Master of Science Thesis, Stephen F. Austin State University, Nacogdoches, Texas, 73 p Baker SM Levinton JS Evan WJ 2000 Particle transport in the zebra mussel, Dreissena polymorpha (Pallas). Biol Bull 199: 116-125 Beaver JR Crisman TL Brock RJ 1991 Grazing effects of an exotic bivalve (Corbicula fluminea) on hypereutrophic lake water. Lakes Reserve: Res Manage 7: 45-51 Belanger SE Farris JL Cherry DS Cairns J 1985 Sediment preference of the fresh-water Asiatic clam, Corbicula fluminea. Nautilus 99: 66-73 Boltovskoy D Izaguirre I Correa N 1995 Feeding selectivity of Corbicula fluminea (Bivalvia) on natural phytoplankton. Hydrobiologia 312: 171-182 Boltovskoy D Correa N Cataldo D Stripeikis J Tudino M 1997 Environmental stress on Corbicula fluminea (Bivalvia) in the Parana River Delta (Argentina): complex pollutionrelated disruption of population structures. Arch Hydrobiol 138: 483-507 Botts PS Patterson BA Schloesser DW 1996 Zebra mussel effects on benthic invertebrates: physical or biotic? J N Am Benthol Soc 15: 179-184 Britton JC Morton B 1982 A dissection guide, field and laboratory manual for the introduced bivalve Corbicula fluminea. Malacol Rev 1-82 Britton JC Murphy CE 1977 New records and ecological notes for Corbicula manilensis in Texas. Nautilus 91: 20-23 Burlakova LE 1998 Ecology of Dristan polymorpha (Pallas) and its role in the structure and function of aquatic ecosystems. Candidate Dissertation, Zoology Institute of the Academy of Science, Republic Belarus, Minsk, Belarus 168 p (in Russian) Burlakova LE Karatayev AY Padilla DK 2000 The impact of Dreissena polymorpha (Pallas) invasion on unionid bivalves. Int Rev Hydrobiol 85: 529-541 Buttner JK 1986 Biology of Corbicula in catfish rearing ponds. In: Proceedings of the Second International Corbicula Symposium, JC Britton (Ed). American Malacological Union, Hattiesburg. Am Malacol Bull 2 (Special ed.) pp 211-218 Cahoon LB Owen DA 1996 Can suspension feeding by bivalves regulate phytoplankton biomass in lake Waccamaw, North Carolina? Hydrobiologia 325: 193-200 Cohen RRH Dresler PV Phillips EJP Cory RL 1984 The effect of the Asiatic clam, Corbicula fluminea, on phytoplankton of the Potomac River, Maryland. Limnol Oceanogrr 29: 170-180 Dame RF 1993 Bivalve Filter Feeders in Estuarine and Coastal Ecosystem Processes. Springer-Verlag, Heidelberg, 579 p Dreier H Tranquilli JA 1981 Reproduction, growth, distribution, and abundance of Corbicula in an Illinois cooling lake. Ill Nat Hist Surv Bull 32: 378-393 Eng LL 1979 Population dynamics of the Asiatic clam, Corbicula fluminea (Müller), in the concrete-lined Delta-Mendota canal of Central California. In: Proceedings, First International Corbicula Symposium, JC Britton (Ed). Fort Worth, Texas, USA, Oct. 1315, 1977. pp 39-78 Evans LPJr. Murphy CE Britton JC Newland LW 1979 Salinity relationships in Corbicula fluminea (Müller 1774). In: Proceedings, First International Corbicula Symposium, JC Britton (Ed). Fort Worth, Texas, USA, Oct. 13-15, 1977. pp 193-214 Hakenkamp CC Palmer MA 1999 Introduced bivalves in freshwater ecosystems: the impact of Corbicula on organic matter dynamics in a sandy stream. Oecologia 119: 445-451 Hakenkamp CC Ribblett SG Palmer MA Swan CM Reid JW Goodison MR 2001 The impact of an introduced bivalve (Corbicula fluminea) on the benthos of a sandy stream. Freshwat Biol 46:491-501 Hamburger KP Dall C Jonasson PM 1990 The role of Dreissena polymorpha Pallas (Mollusca) in the energy budget of Lake Esrom, Denmark. Verh Int Verein Limnol 24: 621-625

257 Hebert PDN Muncaster BW Mackie GL 1989 Ecological and genetic studies on Dreissena polymorpha (Pallas): a new mollusc in the Great Lakes. Can J Fish Aquat Sci 46: 1587-1591 Horgan MJ Mills EL 1997 Clearance rates and filtering activity of zebra mussel (Dreissena ( polymorpha): implications for freshwater lakes. Can J Fish Aquat Sci 54: 249-255 Howlett D Baker R 1999 Corbicula fluminea (Müller): New to UK. J Conchol 36:83 Hoyle JA Schaner T Casselman JM Dermott R 1999 Changes in lake whitefish (Coregonus clupeaformis) stocks in eastern Lake Ontario following Dreissena mussel invasion. Great Lakes Res Rev 4: 1430-1441 Graney RL Cherry DS Rodgers JH Cairns J 1980 The influence of thermal discharges and substrate composition on the population-structure and distribution of the Asiatic clam, Corbicula fluminea, in the New River, Virginia. Nautilus 94: 130-135 Gubanova IF 1968 The influence of waste water on epifauna of flooded wood in Kamskoe Reservoir. In: First Conference on Studying Waterbodies of the Volga Basin, NA Dzubav (Ed). Abstracts. Tolyatti, Russia. pp 231-232 (in Russian) Fanslow DL Nalepa TF Lange GA 1995 Filtration rates of the zebra mussel (Dreissena ( polymorpha) on natural seston from Saginaw Bay, Lake Huron. J Great Lakes Res 21: 489-500 Fast AW 1971 The invasion and distribution of the Asiatic clam (Corbicula manilensis) in a Southern California Reservoir. Bull Southern California Acad Sci 70: 91-98 Francis JT Robillard SR Marsden JE 1996 Yellow perch management in Lake Michigan: a multi-jurisdictional challenge. Fisheries (Bethesda) 21: 18-20 French JRP Schloesser DW 1996 Distribution and winter survival health of Asian clams, Corbicula fluminea, in the St Clair River, Michigan. J Freshwat Ecol 11: 183-192 Fritz LW Lutz RA 1986 Environmental perturbations reflected in internal shell growth patterns of Corbicula fluminea (Mollusca, Bivalvia). Veligerr 28: 401-417 Johengen TH Nalepa TF Fahnenstiel GL Goudy G 1995 Nutrient changes in Saginaw Bay, Lake Huron, after the establishment of the Zebra Mussel ((Dreissena polymorpha). J Great Lakes Res 21: 449-464 Joy JE 1985 A 40-week study on growth of the Asian clam, Corbicula fluminea (Müller), in the Kanawha River, West-Virginia. Nautilus 99: 110-116 Karatayev AY 1983 Ecology of Dreissena polymorpha Pallas and its effects on macrozoobenthos of the thermal power plant's cooling reservoir. Candidate Dissertation, Zoology Institute of Academy of Science Belarusian SSR, Minsk, Belarus 178 p (in Russian) Karatayev AY Burlakova LE 1992 Changes in trophic structure of macrozoobenthos of an eutrophic lake, after invasion of Dreissena polymorpha. Biol Vnutr Vod/Biol Inland Waterr 93: 67-71 (in Russian) Karatayev AY Burlakova LE 1995a The role of Dreissena in lake ecosystems. Russian J Ecol 26: 207-211 Karatayev AY Burlakova LE 1995b Present and further patterns in Dreissena population development in the Narochanskaya lakes system. Vestsi Akademii Navuk Belarusi. Seriya Biyologichnikh Navuk 3: 95-98 (in Belorussian) Karatayev AY Burlakova LE Padilla DK 1997 The effects of Dreissena polymorpha (Pallas) invasion on aquatic communities in Eastern Europe. J Shellfish Res 16: 187-203. Karatayev AY Burlakova LE Padilla DK 1998 Physical factors that limit the distribution and abundance of Dreissena polymorpha (Pall.). J Shellfish Res 17: 1219-1235 Karatayev AY Burlakova LE Padilla DK 2002 Impacts of zebra mussels on aquatic communities and their role as ecosystem engineers. In: Invasive Aquatic Species of Europe: Distributions, Impacts and Management. Monoraphiae Biologicae Series, E Leppäkoski S Gollasch S Olenin (Eds.), Kluwer Scientific Publishers, Dordrecht, pp 433-446 Karatayev AY Burlakova LE Kesterson T Padilla DK 2003a Dominance of the Asiatic clam, Corbicula fluminea (Müller) in the benthic community of a reservoir. J Shellfish Res 22(2): 487-493

258 Karatayev AY Burlakova LE Padilla DK Johnson LE 2003b Patterns of spread of the zebra mussel ((Dreissena polymorpha (Pallas)): the continuing invasion of Belarussian lakes. Biol Invasions 5(3): 213-221 Karatayev AY Lyakhnovich VP 1990 Effect of Dreissena polymorpha Pallas on benthic crustaceans (Gammarus lacustris Sars, Pallasea quadrispinosa Sars and Asellus aquaticus L.) in Lukomskoe Lake. In: Species within their Range: Biology, Ecology, and Productivity of Aquatic Invertebrates, NN Khmeleva et al. (Eds). Navuka i Tekhnika Press, Minsk. pp 123-125 (in Russian) Karatayev AY Tishchikov GM Karatayeva IV 1983 The specific community of benthic animals associated with Dreissena polymorpha Pallas. Biol Vnutr Vod/Biol Inland Waterr 61:18-21 (in Russian) Kat P 1982 Shell dissolution as a significant cause of mortality for Corbicula fluminea (Bivalvia: Corbiculidae) inhabiting acidic waters. Malacological Review 15:129-134 Kharchenko TG Protasov AA 1981 On consortia in water ecosystems. Gidrobiologicheskii Zhurnall 17:15-20 (in Russian) Khusainova NZ 1958 The Biological Distinctive Features of Some Abundant Bottom Invertebrates Used by Fishes in the Aral Sea. Kazakhstan State University Press, AlmaAta 116 p (in Russian) Kryger J Riisgård HU 1988 Filtration rate capacities in 6 species of European freshwater bivalves. Oecologia 77: 34-38 Kuhns L Berg M 1999 Benthic invertebrate community responses to round goby (Neogobius melanostomus) and zebra mussel ((Dreissena polymorpha) invasion in southern Lake Michigan. J Great Lakes Res 25: 910-917 Lauer TE McComish TS 2001 Impact of zebra mussels ((Dreissena polymorpha) on fingernail clams (Sphaeriidae) in extreme Southern Lake Michigan. J Great Lakes Res 27(2): 230-238 Lauritsen DD 1986 Filter-feeding in Corbicula fluminea and its effect on seston removal. J N Am Benthol Soc 5: 165-172 Lauritsen DD Mozley SC 1989 Nutrient excretion by the Asiatic clam Corbicula fluminea. J N Am Benthol Soc 8: 134-139 Leff LG Burch JL McArthur JV 1990 Spatial distribution, seston removal, and potential competitive interactions of the bivalves Corbicula fluminea and Elliptio complanata, in a coastal plain stream. Freshwat Biol 24: 409-416 Lei J Payne BS Wang SY 1996 Filtration dynamics of the zebra mussels, Dreissena polymorpha. Can J Fish Aquat Scii 53: 29-37 Lowe RL Pillsbury RW 1995 Shifts in benthic algal community structure and function following the appearance of zebra mussel ((Dreissena polymorpha) in Saginaw Bay, Lake Huron. J Great Lakes Res 21: 558-566 Lozano SJ Scharold JV Nalepa TF 2001 Recent declines in benthic macroinvertebrate densities in Lake Ontario. Can J Fish Aquat Sci 58(3): 518-529 Luferov VP 1965 Organisms living anabiotically frozen in ice of the littoral zone of Rybinskoe Reservoir. Tr Inst Biol Vnutr Vod Akad Nauk SSSR 8(11): 151-154 (in Russian) Lvova AA 1977 The ecology of Dreissena polymorpha (Pall.) in Uchinskoe Reservoir. Candidate Dissertation, Moscow State University, Moscow, USSR 114 p (in Russian) Lvova AA 1980 Ecology of zebra mussel (Dreissena ( polymorpha polymorpha (Pall.). Trans Hydrobiol Soc RAN 23: 101-119 (in Russian) Lyakhnovich VP Karatayev AY Mitrakhovich PA Guryanova LV Vezhnovets GG 1988 Productivity and prospects for utilizing the ecosystem of Lake Lukoml, thermoelectric station cooling reservoir. Soviet J Ecoll 18: 255-259 Lyakhnovich VP Karatayev AY Andreev NI Andreeva SI Afanasiev SA Dyga AK Zakutskiy VP Zolotareva VI Lvova AA Nekrasova MY Osadchikh VF Pligin YV

259 Protasov AA Tischikov GM 1994 Living conditions. In: Freshwater Zebra Mussel Dreissena polymorpha (Pall.) (Bivalvia, Dreissenidae). Systematics, Ecology, Practical Meaning, JI Starobogatov (Ed). Nauka Press, Moscow. pp 109-119 (in Russian) Lyakhov SM Mikheev VP 1964 Distribution and density of Dreissena in the Kuibyshevskoe Reservoir seven years after its construction. Tr Inst Biol Vnutr Vod Akad Nauk SSSR 7(10): 3-18 (in Russian) Marsh PC 1985 Secondary production of introduced Asiatic clam, Corbicula fluminea, in a Central Arizona Canal. Hydrobiologia 124: 103-110 Makarewicz JC Bertram P Lewis TW 2000 Chemistry of the offshore surface waters of Lake Erie: pre- and post-Dreissena introduction (1983-1993). J Great Lakes Res 26 (1): 8293 MacIsaac HJ Lonnee CJ Leach JH 1995 Suppression of microzooplankton by zebra mussels: Importance of mussel size. Freshwat Biol 34: 379-387 Mattice JS 1979 Interactions of Corbicula sp. with power plants. In: Proceedings, First International Corbicula Symposium, JC Britton (Ed), Fort Worth, Texas, USA, Oct. 1315, 1977, pp 119-138 Mayer CM Keats RA Rudstam LG Mills EL 2002 Scale-dependent effects of zebra mussels on benthic invertebrates in a large eutrophic lake. J N Am Benthol Soc 21(4): 616-633 McMahon RF 1983 Ecology of an invasive pest bivalve, Corbicula. In: The Mollusca. Vol. 6. Ecology, WD Russel-Hunter (Ed), Academic Press, Inc., pp 505-561 McMahon RF 1999 Invasive characteristics of the freshwater bivalve Corbicula fluminea. In: Nonindigenous Freshwater Organisms. Vectors, Biology, and Impacts, R Claudi JH Leach (Eds.), Lewis Publishers, Ann Arbor, Michigan, pp 315-343 McMahon RF Bogan AE 2001 Mollusca: Bivalvia. In: Ecology and Classification of North American Freshwater Invertebrates, 2ndd Edition, JH Thorp AP Covich (Eds.), Academic Press, Inc., pp 331-430 Mikheev VP 1967 The nutrition of zebra mussels ((Dreissena polymorpha Pallas). Summary of the Candidate Dissertation, State Research Institute for Lakes and Rivers Fishery Industry, Leningrad 22 p (in Russian) Mikheev VP 1994 Ration of Dreissena in natural conditions. In: Freshwater Zebra Mussel Dreissena polymorpha (Pall.) (Bivalvia, Dreissenidae). Systematics, Ecology, Practical Meaning, JI Starobogatov (Ed). Nauka Press, Moscow, pp 129-132 (in Russian) Mikulski J Adamczak B Bittel L Bohr R Bronisz D Donderski W Gizinski A Luscinska M Rejewski M Strzelczyk E Wolnomiejski N Zawislak W Zytkowicz R 1975 Basic regularities of productive processes in the Ilawa Lakes and in the Goplo Lake from the point of view of utility values of the water. Pol Arch Hydrobiol 22: 101-122 Mills EL Casselman JM Dermott R Fitzsimons JD Gal G Holeck KT Hoyle JA Johannsson OE Lantry BF Makarewicz JC Millard ES Munawar IF Munawar M O’Gorman R Owens RW Rudstam LG Schaner T Stewart TJ 2003 Lake Ontario: food web dynamics in a changing ecosystem (1970-2000). Can J Fish Aquat Sci 60: 471-490 Minchin D 2000 Dispersal of zebra mussel in Ireland. Verh Internat Verein Limnol 27: 15761579 Molloy DP Karatayev AY Burlakova LE Kurandina DP Laruelle F 1997 Natural enemies of zebra mussels: predators, parasites and ecological competitors. Rev Fish Scii 5: 27-97 Mordukhai-Boltovskoi FD 1960 Caspian Fauna in the Azov and Black Sea Basins. Academia Nauk Press, Moscow (in Russian) Morton B 1979 Corbicula in Asia. In: Proceedings, First International Corbicula Symposium, JC Britton (Ed) Fort Worth, Texas, USA, Oct. 13-15, 1977, pp 1-54 Morton B 1997 The aquatic nuisance species problem: a global perspective and review. In: Zebra Mussels and Aquatic Nuisance Species, FM D'Itri (Ed). Ann Arbor Press, Inc., Chelsea, Michigan Petrie S Knapton R 1999 Rapid increase and subsequent decline of zebra and quagga mussels in Long Point Bay, Lake Erie: possible influence of waterfowl predation. J Great Lakes Res 25: 772-782

260 Phelps HL 1994 The Asiatic clam (Corbicula fluminea) invasion and system-level ecological change in the Potomac River Estuary near Washington, DC. Estuaries 17: 614-621 Pimentel D Lach L Zuniga R Morrison D 2000 Environmental and economic costs of nonindigenous species in the United States. Bioscience 50: 53-65 Pothoven SA Nalepa TF Schneeberger PJ Brandt SB 2001 Changes in diet and body condition of lake whitefish in southern Lake Michigan associated with changes in benthos. N Am J Fish Managg 21: 876-883 Prokopovich NP 1969 Deposition of clastic sediments by clams. J Sediment Petrol 39: 891901 Protasov AA Afanasiev SA Ivanova OO 1983 The distribution and role of Dreissena polymorpha in the periphytone of Chernobyl water cooling reservoir. In: Molluscs: Systematics, Ecology and Patterns of Occurrence. Abstracts of the 7th Meeting on the Molluscs Study. Nauka Press, Leningrad. pp 220-222 (in Russian) Ramcharan CW Padilla DK Dodson SI 1992 Models to predict potential occurrence and density of the zebra mussel, Dreissena polymorpha. Can J Fish Aquat Scii 49: 26112620 Reed-Andersen T Carpenter SR Padilla DK Lathrop RC 2000 Predicted impact of zebra mussel (Dreissena polymorpha) invasion on water clarity in Lake Mendota. Can J Fish Aquat Sci 57: 1617-1626 Reeders HH bij de Vaate A 1990 Zebra mussels (Dreissena ( polymorpha): a new perspective for water quality management. Hydrobiologia 200/201: 437-450 Reeders HH bij de Vaate A Slim FJ 1989 The filtration rate of Dreissena polymorpha (Bivalvia) in three Dutch lakes with reference to biological water quality management. Freshwat Biol 22: 133-141 Reid RGB McMahon RF Foighil DO Finnigan R 1992 Anterior inhalant currents and pedalfeeding in bivalves. Veligerr 35: 93-104 Riccardi A 2003 Predicting the impacts of an introduced species from an invasion history: an empirical approach applied to zebra mussel invasion. Freshwat Bioll 48: 972-981 Robinson JV Wellborn GA 1988 Ecological resistance to the invasion of a freshwater clam, Corbicula fluminea: Fish predation effects. Oecologia (Berl.) 77: 445-452 Roditi HA Caraco NF Cole JJ Strayer DL 1996 Filtration of Hudson River water by the zebra mussel ((Dreissena polymorpha). Estuaries 19: 824-832 Rodgers JHJr. Cherry DS Dickson KL Cairns JJr 1979 Invasion, population dynamics and elemental accumulation of Corbicula fluminea in the new river at Glen Lyn, Virginia. In: Proceedings, First International Corbicula Symposium, JC Britton (Ed). Fort Worth, Texas, USA, Oct. 13-15, 1977. pp 99-110 Schloesser DW Nalepa TF Mackie GL 1996 Zebra mussel infestation of unionid bivalves (Unionidae) in North America. Am Zool 36: 300-310 Schneider DW Stoeckel JA Rehmann CR Blodgett KD Sparks RE Padilla DK 2003 A developmental bottleneck in dispersing larvae: implications for spatial population dynamics. Ecol Lett 6 (4): 352-360 Shkorbatov GL Karpevich AF Antonio PI 1994 Ecological physiology. In: Freshwater Zebra Mussel Dreissena polymorpha (Pall.) (Bivalvia, Dreissenidae). Systematics, Ecology, Practical Meaning, JI Starobogatov (Ed). Nauka Press, Moscow, pp 67-108. (in Russian) Sickel JB 1986 Corbicula population mortalities: factors influencing population control. Am Malacol Bulll 2 (Special Ed.):89-94 Sickel JB Lyles MB 1981 Chaetogaster limnaei (Oligochaeta: Naididae) inhabiting the mantle cavity of the Asiatic clam, Corbicula fluminea, in Barkley Lake, Kentucky. Veligerr 23: 361-362

261 Silverman H Achberger EC Lynn JW Dietz TH 1995 Filtration and utilization of laboratorycultured bacteria by Dreissena polymorpha, Corbicula fluminea, and Carunculina texasensis. Biol Bull 189: 308-319 Silverman H Nichols SJ Cherry JS Achberger E Lynn JW Dietz TH 1997 Clearance of laboratory-cultured bacteria by freshwater bivalves: differences between lentic and lotic unionids. Can J Zooll 75: 1857-1866 Sokolova NY Izvekova EI Lvova AA Sakharova MI 1980 Structure, distribution and seasonal dynamics of benthic densities and biomass. Tr Vses Gidrobiol O-va 23: 7-23 (in Russian) Spiridonov YI 1973 The mineralization by zebra mussels in Volgograd Reservoir. Tr. Kompleksn Eksped Sarat Univ Izuch Volgogr Sarat Vodokhran 3: 131-134 (in Russian) Sprung M Rose U 1988 Influence of food size and food quality on the feeding of the mussel Dreissena polymorpha. Oecologia 77: 526-532 Stanczykowska A Lewandowski K 1993 Thirty years of studies of Dreissena polymorpha ecology in Mazurian Lakes of Northeastern Poland. In: Zebra Mussels: Biology, Impacts, and Control, TF Nalepa DW Schloesser (Eds). Lewis Publishers, Ann Arbor, Michigan, pp 3-33 Starobogatov JI Andreeva SI 1994 Distribution and history. In: Freshwater Zebra Mussel Dreissena polymorpha (Pall.) (Bivalvia, Dreissenidae). Systematics, Ecology, Practical Meaning, JI Starobogatov (Ed.). Nauka Press, Moscow, pp 47-55 (in Russian) Stempniewicz L 1974 The effect of feeding of the coot (Fulica ( atra L.) on the character of the shoals of Dreissena polymorpha Pall. in the Lake Goplo. Acta Univ Nicolai Copernici, Ser Mat Przyrr 34: 83-103 Stewart TW Haynes JM 1994 Benthic macroinvertebrate communities of Southwestern Lake Ontario following invasion of Dreissena. J Great Lakes Res 20: 479-493 Stewart TW Miner JG Lowe RL 1998 Quantifying mechanism for zebra mussel effects on benthic macroinvertebrates: organic matter production and shell-generated habitat. J N Am Benthol Soc 17: 81-94 Stites DL Benke AC Gillespie DM 1995 Population dynamics, growth, and production of the Asiatic clam, Corbicula fluminea, in a blackwater river. Can J Fish Aquat Scii 52: 425437 Strayer DL 1999 Effects of alien species on freshwater mollusks in North America. J N Am Benthol Soc 18: 74-98 Strayer DL Caraco NF Cole JJ Findlay S Pace ML 1999 Transformation of freshwater ecosystems by bivalves. A case study of zebra mussels in the Hudson River. BioScience 49: 19-27 Strayer DL Smith LC Hunter DC 1998 Effects of the zebra mussel (Dreissena polymorpha) invasion on the macrobenthos of the freshwater tidal Hudson River. Can J Zool 76: 419-425 Ten Winkel EN Davids C 1982 Food selection by Dreissena polymorpha Pallas (Mollusca: Bivalvia). Freshwat Bioll 12: 553-558 Vanderploeg HA Nalepa TF Jude DJ Mills EL Holeck KT Liebig JR Grigorovich IA Ojaveer H 2002 Dispersal and emerging ecological impacts of Ponto-Caspian species in the Laurentian Great Lakes. Can J Fish Aquat Scii 59(7): 1209-1228 Way CM Hornbach DJ Millerway CA Payne BS Miller AC 1990 Dynamics of filter feeding in Corbicula fluminea (Bivalvia, Corbiculidae). Can J Zool 68: 115-120 Wisniewski R 1990 Shoals of Dreissena polymorpha as a bio-processor of seston. Hydrobiologia 200/201: 451-458 Wolnomiejski N 1970 The effects of Dreissena polymorpha Pall. aggregation on the differentiation of the benthonic macrofauna. Zecz nauk UMK K 25: 31-39 Wong WH Levinton JS Twining BS Fisher N 2003 Assimilation of micro- and mesozooplankton by zebra mussels: a demonstration of the food web link between zooplankton and benthic suspension feeders. Limnol Oceanogr 48: 308-312 Yablonskaya EA 1985 The Caspian Sea: Fauna and Biological Productivity. Nauka Press, Moscow (in Russian)

262 Young BL Padilla DK Schneider DW Hewett SW 1996 The importance of size-frequency relationships for predicting ecological impact of zebra mussel populations. Hydrobiologia 332 (3): 151-158 Zhadin VI 1946 The traveling shellfish Dreissena. Priroda 5: 29-37 (in Russian)

FUNCTIONAL CHANGES IN BENTHIC FRESHWATER COMMUNITIES AFTER DREISSENA POLYMORPHA (PALLAS) INVASION AND CONSEQUENCES FOR FILTRATION

Lyubov E. Burlakova1, Alexander Y. Karatayev1, and Dianna K. Padilla2 1

Department of Biology, Stephen F. Austin State University, Nacogdoches, TX, USA Department of Ecology and Evolution, Stony Brook University, Stony Brook, NY, USA

2

Abstract: Dreissena is extremely abundant in waters it invades, and dramatically changes benthic invertebrate communities in terms of total biomass, species composition, and the relative abundance of functional groups. We analyzed the relative abundance of feeding functional groups of the benthic community before and after zebra mussel invasion in three Belarussian lakes, four lakes after invasion only, and one lake in the same region that has not been invaded. After invasion, benthic structure was dominated by one trophic group – filterers. This group accounted for greater than 96% of the total biomass of benthic invertebrates. We found that the relative abundance of feeding functional groups in the rest of the benthic community, without including Dreissena biomass, was also different in lakes examined before and after zebra mussel invasion. Before invasion and in the un invaded lake, planktonic invertebrates filtered a volume equivalent to the volume of the lake within few days, and were more than 200 times more effective than benthic filterers, which would take about 4 years to filter an equivalent volume. After Dreissena invaded the lakes, the total average biomass of all benthic invertebrates (including zebra mussels) increased more than 20 times. The filtration efficiency of the benthic community increased greater than 70 times, and the time required to filter the volume of the lake was not significantly different than that for zooplankton. These dramatic changes will alter the relative roles of the plankton and benthos in a variety of ecosystem functions, especially the movement of carbon from the plankton to the benthos. Key words: Zebra mussels, benthic community, trophic structure, feeding functional groups, filtration efficiency.

INTRODUCTION The zebra mussel, Dreissena polymorpha Pallas (1771), continues to spread throughout the freshwaters of Eurasia and North America, and new lakes and rivers are constantly being invaded (Kinzelbach 1992, McMahon and Bogan 2001, Minchin et al. 2002, Karatayev et al. 2003). Species of 263 R.F. Dame and S. Olenin (eds.), The Comparative Roles of Suspension-Feeders in Ecosystems, 263–275. ©2005 Springer. Printed in the Netherlands.

264 Dreissena are the only bivalves in freshwater to attach to hard substrates and possess a dispersing planktonic larval stage. D. polymorpha is extremely abundant in waters it invades, is frequently competitively dominant over native freshwater fauna, and has large impacts on all parts of the ecosystem, especially benthic animals (reviewed in Karatayev et al. 1997, 2002). Characterizing the feeding functional group composition in lakes before and after invasion by zebra mussels can provide insights into how benthic communities respond to invasion. The feeding functional group approach enables a quantitative assessment of the degree of dependence of the invertebrate biota on particular food resources, and the linkages between food sources and morphological and behavioral adaptations (Merritt and Cummings 1996). Although the effect of D. polymorpha invasion on species composition and abundance within the benthic community has been documented for certain lakes, to date there are few studies of resultant changes in the trophic structure of communities (Sokolova et al. 1980a, Karatayev and Burlakova 1992). When zebra mussels invade, they create a large population of effective suspension-feeders that can cause radical changes in the benthic community (Lvova-Kachanova and Izvekova 1978, Sokolova et al. 1980a, Karatayev and Burlakova 1992, Karatayev et al. 1997). Native suspensionfeeders can be out-competed by D. polymorpha, and decrease in abundance, while animals feeding on the sediments can increase in abundance (Sokolova et al. 1980a, Karatayev and Burlakova 1992, Karatayev et al. 1994). In this study, we assessed the impacts of invasion by zebra mussels on the structure of the benthic communities by examining feeding functional groups within the benthic community of 8 Belarussian lakes. By comparing the structure of communities without zebra mussels with those that have been invaded, we can assess the impacts of invasion on trophic processes and some aspects of ecosystem function.

MATERIALS AND METHODS Study Sites and Sampling To study the trophic structure of benthic invertebrates in lakes with and without Dreissena, we used data collected from glacial lakes (Karatayev et al. 2003), as part of a larger survey of Belarussian lakes. The Republic of Belarus is situated between Poland and Russia, and was part of the former Soviet Union. A variety of chemical, geological, physical, and biological data were collected in mid-summer for each of these lakes by the Lakes Research Laboratory of the Belarussian State University. We collected additional data for some lakes in the summer of 1998 and 1999. For three lakes we have data

265 before and after invasion, for four lakes we have data after invasion only, and we have data for one lake, in the same region, that has not been invaded (Table 1). To determine the species composition, density and biomass of benthic invertebrates 7 – 16 samples were collected from each lake, depending on the lake size. Sample sites were selected to maximize coverage of the lake bottom and include all major habitat types. For all benthic samples we used a Petersen grab for hard substrates and an Eckman grab for soft substrates (sample area 0.025 m2). Samples were washed through a 500 Pm mesh. Retained macroinvertebrates were preserved with 10% neutral buffered formalin. All macroinvertebrates were identified to the lowest possible taxon, counted and weighed to the nearest 0.0001 g after being blotted dry on absorbent paper (wet mass). For three lakes, Myadel, Boginskoe, and Svir, we had data both before and after invasion, which allowed us to do paired comparisons (Table 1). We also had data for four additional lakes, Bolshie Shvakshty, Volchin, Dolzha, and Bolduk that have been invaded by zebra mussels. All of these lakes were invaded by Dreissena between 1980 and the mid-1990s, however, unfortunately, we do not know the exact years of invasion. Table 1. Limnological parameters of the studied Belarussian lakes. Lake

Year Studied Surface Volume Maximum Secchi Trophic Before After area (106 m3) depth (m) depth (m) status invasion invasion (km2) 1973 1999 1.2 9.12 16.2 2.4 eutrophic Boginskoe Svir 1980 1998 22.3 104.3 8.7 1.8 eutrophic 1980 1998 16.4 102.0 24.6 4.8 mesotrophic Myadel Dolzha 1998 1.0 5.4 13.7 2.6 eutrophic Bolshiye Shvakshty 1998 9.6 22.3 5.3 3.1 eutrophic Volchin 1998 0.5 7.9 32.9 3.8 mesotrophic Bolduk 1999 0.8 11.9 39.7 4.5 oligomesotrophic Ikazn 1973 not 2.4 7.9 8.4 1.4 eutrophic invaded

Functional Groups We used the classification scheme by Merritt and Cummins (1996) for functional feeding groups. For invertebrates that were identified to species, we used data from Izvekova (1975), Sokolova et al. (1980b) and Monakov (1998, 2003) to assign feeding functional group. For invertebrates identified to genus, we used Merritt and Cummins (1996) and Thorp and Covich (2001). However, some species and genera fit into more than one group. For example, some collectors are known to filter-feed and gather (e.g., Microtendipes cloris, Bithynia tentaculata, B. leachi, Tanytarsus sp.)

266 (Izvekova 1975, Merritt and Cummins 1996, Monakov 1998). As these species were abundant in the lakes sampled, we considered them as “filtering + gathering collectors”, a sub-group within the group “collectors”. To determine if there were changes in the relative abundance of feeding functional groups associated with zebra mussel invasion, we compared the relative proportions of biomass of each functional group before and after zebra mussel invasion. Functional Consequences Shifts in the feeding functional groups in a lake will have functional consequences for both the processing of benthic carbon, and for links between the benthic and the planktonic communities (benthic-pelagic coupling). To estimate the filtration capability of the zooplankton community in each lake we used Kryuchkova's (1989) estimate that zooplankton can filter 120 mL mg wet mass-1d-1 in a cladoceran dominated eutrophic lake. To estimate the filtration capacity of the benthos, we used the literature values for the filtration rates, based on wet total mass, of individual species (Izvekova 1975, Alimov 1981, reviewed in Monakov 1998) weighted by the average biomass (wet total mass (body plus shell), g m-2) of that species in the lake. For species whose filtration rates were not known, we used the rates for the closest related taxon whose rate was known. The total filtration capacity of the benthic community was estimated by multiplying the filtration rates by the average biomass of each taxon determined to be in the filtering collectors functional group. As it is very difficult to determine the relative proportion of time a species filters or gathers (references in Monakov 1998), the impact of members of the filtering + gathering collectors was divided in half. For zebra mussels, we used a filtration rate of 58 mL g-1 h-1 (literature estimates range from 35 - 110 mL g-1 h-1, Karatayev et al. 1997). To estimate the total filtration capacity of zebra mussels, we multiplied the average biomass of zebra mussels (g m-2) by this filtration rate. In this way we were able to compare the filtration rate of the entire community of zooplankton with the amount of filtration for the entire benthic community. Statistical Analyses To compare the relative abundance of trophic groups in lakes with and without zebra mussels, we used either a t-test or Mann-Whitney U test on percentage data for each lake (Zar 1996). To compare the structure of functional trophic groups of benthic community before and after zebra mussel invasion we used Fisher-Freeman-Halton test (a generalization of the Fishers Exact test to r by c contingency table) with a Monte Carlo estimate of the Pvalue to test for homogeneity in contingency tables. Effects were considered

267 statistically significant at P < 0.05. Analyses were performed with StatXact-4 (version 4.0.1, Cytel Software Corp.) and Statistica software (STATISTICA version 6, StatSoft, Inc. 2001). When multiple tests were conducted on the same data, we used a sequential Bonferroni correction (Rice 1989) to adjust the critical alpha considered for statistical significance. Where appropriate, we present the critical alpha (Į) with the results of each statistical test.

RESULTS Feeding Functional Group Composition Before Invasion Collectors (filterers, filterers + gatherers and gatherers) dominated the benthic macroinvertebrate community in lakes uninhabited by zebra mussels, and comprised approximately 70% of the total biomass of the community. Filterers + gatherers were the largest group (50% of total biomass). Predators constituted about 17% of the total biomass, comparable to shredders, scrapers and scavengers combined (Table 2). Table 2. Relative proportions of feeding functional groups of benthic macroinvertebrates (% of total biomass) in Belarussian lakes with and without zebra mussels (average ± SE). Feeding functional group Collectors: Filterers Collectors: Filterers + Gatherers Collectors: Gatherers Shredders Scarpers Scavengers Predators

Lakes without zebra mussels (n = 4) 4.2 ± 0.9 50.4 ± 5.1 14.5 ± 5.2 2.5 ± 2.3 9.6 ± 3.5 2.3 ± 1.1 16.5 ± 1.8

Zebra mussel invaded lakes excluding zebra including zebra mussels (n = 7) mussels (n = 7) 7.6 ± 2.0 96.7 ± 0.8 34.4 ± 8.5 21.9 ± 3.2 0.3 ± 0.3 6.0 ± 1.5 7.4 ± 3.2 22.5 ± 5.2

1.5 ± 0.7 0.7 ± 0.2 < 0.01 0.2 ± 0.1 0.3 ± 0.2 0.7 ± 0.1

After Invasion, excluding Dreissena: Lakes with before and after data For the three lakes where we had before and after invasion data, the changes in functional group structure (excluding Dreissena biomass) were significant (Lake Boginskoe: P = 0.0008, Į = 0.017; Lake Myadel: P = 0.004, adjusted critical Į = 0.025; Lake Svir: P = 0.044, Į = 0.05; Fisher-FreemanHalton test) (Table 3).

268 Table 3. Relative proportions of feeding functional groups of benthic macroinvertebrates (% of total wet biomass) excluding Dreissena biomass in three Belarussian lakes studied before (B) and after (A) zebra mussel invasion. Feeding functional group Collectors: Filterers Collectors: Filterers + Gatherers Collectors: Gatherers Total Collectors Shredders Scarpers Scavengers Predators

Lake Myadel B A 5.7 2.4

Lake Boginskoe B A 5.1 10.6

Lake Svir B A 1.6 12.4

36.0 29.8

36.4 12.7

55.9 7.9

51.3 21.4

50.9 11.1

51.0 13.6

71.5 0.0 2.3 5.4 20.8

51.5 0.0 4.5 3.7 40.3

68.9 0.8 14.8 0.8 14.7

83.3 0.0 4.0 5.4 7.3

63.6 0.0 16.4 1.9 18.1

77 0.0 7.7 2.5 12.8

After Invasion, excluding Dreissena: Pooled data from all lakes The pooled data for all invaded (7 invaded lakes - Myadel, Bolshie Shvakshty, Boginskoe, Svir, Bolduk, Dolzha, and Volchin) and uninvaded (4 uninvaded lakes - Myadel, Boginskoe, Svir and Ikazn) lakes were tested for differences in the relative biomass of different feeding functional groups. For these pooled data, the changes in the functional trophic groups after zebra mussel invasion were not significant (P = 0.068, adjusted critical Į = 0.013, Fisher-Freeman-Halton test). Collectors remained more than 60% of total community biomass (excluding D. polymorpha), and the relative abundance of filterers + gatherers did not change significantly (P = 0.12, adjusted critical Į = 0.017, t-test). The portion of collectors-gatherers and filterers also did not significantly change (P > 0.30, adjusted critical Į = 0.025) (Table 2). The change in the relative abundance of all types of collectors before and after invasion was also not significant (P = 0.065, Fisher-Freeman-Halton test adjusted critical Į = 0.010). The portion of scavengers increased, but this change was not significant (P = 0.039, Mann-Whitney U test, adjusted critical Į = 0.008) and the proportion of predators also did not change (P = 0.36, ttest, critical Į = 0.05). After Invasion, including Dreissena If D. polymorpha is considered with the rest of the benthic community, the trophic structure of the benthic community was characterized by an extremely high dominance of one trophic group – collectors filterers, which accounted for > 96 % of the total biomass of benthic invertebrates in lakes populated by zebra mussels (Table 2). The relative proportions of feeding

269 functional groups (in terms of biomass) in the benthic community before and after zebra mussel invasion were significantly different (P 70 times. The time required for the benthic community to filter the volume Table 4. Mean biomass of zooplankton and benthic communities, and the time required to filter a volume of water equal to the lake volume before and after zebra mussel invasion. Parameter

Lake Bolshie Shvakshty

Before Dreissena invasion Zooplankton*: biomass (g m-3)

Zoobenthos:

days to filter biomass (g m-2) days to filter

0.9 7 9 20. 0 63

Lake Svir

Lake Volchin

1 .55 5 1 0.1 5 9

After Dreissena invasion Zooplankton*: biomass (g m-3) days to filter Zoobenthos excluding Dreissena: biomass (g m-2) days to filter

2

0.5 2 16

.78

22. 7 7

6.2

3 1 3 9

Dreissena:

biomass (g m-2) days to filter

16 3 10

2 38 1 4

Zoobenthos including Dreissena: biomass (g m-2) days to filter

18 5 4

2 54 1 0

*data from Karatayev and Makritskaya (1999)

1 .03 8 1 .2 3 688 1 .37 6 9 .1 3 03 1 92 5 5 2 02 4 7

Lake Dolzha

2 .08 4 9 .8 1 732 1 .25 7 4 .4 5 06 2 57 1 5 2 61 1 4

270 equivalent to that of the lake decreased to 19 ± 10 days, and was not significantly different from that for the planktonic community (8 ± 3 days, P = 0.40, paired t-test). The total biomass of benthic invertebrates excluding Dreissena did not change significantly after zebra mussel invasion (P = 0.41, paired t-test). There was also no significant difference in the biomass of zooplankton before and after zebra mussel invasion (1.4 ± 0.3 g m-3 vs. 1.5 ± 0.5 g m-3, paired t-test, P = 0.88), nor in the time to filter the equivalent of the volume of the lake (6.5 ± 1.1 vs. 7.9 ± 2.8, P = 0.65, t-test). The high filtration rate for the zoobenthos in Lake Bolshie Shvakshty was attributed to an unusually large biomass of chironomids, which can have a very high filtration rate (> 1,700 mL g-1 hr-1, Izvekova 1975).

DISCUSSION Feeding Functional Group Composition We found a dramatic shift in the benthic trophic structure after D. polymorpha invasion. The structure of feeding functional groups in the community including Dreissena was overwhelmingly dominated by collectors-filterers. D. polymorpha was the dominant benthic species in terms of biomass. These results are consistent with findings from other lakes and reservoirs in the former Soviet Union including Uchinskoe Reservoir, Russia (Lvova-Kachanova and Izvekova 1978, Sokolova et al. 1980a, Sokolova et al. 1980b) and Lake Lukomskoe, Belarus (Karatayev and Burlakova 1992, Karatayev et al. 1997). The invasion of D. polymorpha in Uchinskoe Reservoir resulted in the replacement of the dominant species, the chironomid filter-feeder Glyptotendipes paripes, and drastic changes in the relative abundance of different species and trophic groups (Lvova-Kachanova and Izvekova 1978, Sokolova et al. 1980a, 1980c). Following the invasion of D. polymorpha in Lake Lukomskoe the benthic community was characterized by an exceedingly high dominance of filterers, which accounted for 95% of the total benthic animal biomass (Karatayev and Burlakova 1992). As a result, the trophic structure of the littoral zone was impoverished, and the remaining trophic groups contributed relatively little to the total biomass. Similar patterns were found for six other waterbodies across the Former Soviet Union, where Dreissena comprised > 93% of the total biomass of benthic community (Karatayev et al. 1994, 1997). Other studies of benthic communities in zebra mussel beds have shown dramatic differences in the density and biomass of associated taxa compare to substrates without mussels (reviewed in Karatayev et al. 1997,

271 2002). The creation of new habitat is perhaps the most important effect that zebra mussels have on the benthic community. Several experimental studies have shown that the structure created by zebra mussels provides refuge for a variety of species, and this impact is seen even in the absence of the biological activity of filtering mussels (Slepnev et al. 1994, Ricciardi 1997, Stewart et al. 1998, 1999). Another possible mechanism through which zebra mussels may impact benthic community functional structure is through trophic impacts. Some studies have reported increases in the relative abundance of collectors due to enrichment of the benthos with organic substances from feces and pseudofeces created by zebra mussels or decreases in filterers due to competition for food with zebra mussels (Izvekova Lvova-Katchanova 1972, Stewart et al. 1998, Berezina 1999). All of these changes may be obvious within zebra mussel beds and areas with high mussel densities. However, in all lakes with zebra mussels much of the bottom is not covered by mussels, especially in the profundal zone (Stánczykowska and Lewandowski 1993, Burlakova, Karatayev, personal observations). Therefore, the overall effect on the entire benthic community might be much less pronounced than in areas with high zebra mussel densities. When the average structure of benthic trophic groups was pooled across lakes with and without zebra mussels, differences in the relative abundance of the feeding functional groups did not significantly change after zebra mussel invasion. However, we had relatively few lakes in our sample, limiting the power of our statistical tests. With greater sample sizes, the quantitative changes in scavengers could be significant. These results suggest that the major change in the community was the addition of zebra mussels rather than the displacement of other functional groups. However, the data for individual lakes indicated significant changes in several functional groups before and after zebra mussel invasion, but the changes were not in concert, and were often in opposite directions for the same functional group. Thus, these changes were masked when the data were pooled. These results do, however, highlight the importance and power of before and after data. To fully understand the impacts of zebra mussels on communities we need much more community data on lakes before and after invasion, and, unfortunately, these data are rare. Suspension-feeders are an integral component of aquatic ecosystems. They feed upon a very dilute food resource and convert previously dispersed, minute materials to larger animal biomass, i.e., their own bodies (Wallace and Merritt 1980). The overpowering dominance of collectors-filterers after zebra mussel invasion will drive changes in ecosystem function as they greatly enhance the rates of deposition of both organic and inorganic material on the bottom and thus build a direct connection between the planktonic portion of the water body and the benthos (benthic-pelagic coupling) (reviewed in Karatayev et al. 2002). In addition, Dreissena as well as many associated benthic animals are prey for benthivorous fishes (reviewed in Molloy et al.

272 1997, Karatayev et al. 1994). They may also provide an important path for moving energy from the benthic community to higher trophic levels.

Zebra Mussel as a Biofilter The role of bivalves in and impacts on aquatic ecosystems has long been recognised for marine and estuarine ecosystems (reviewed in Dame 1993, 1996). Bivalves can affect nutrient cycling by consuming particulate and dissolved organic matter and excreting inorganic nutrients. They affect community structure (both in the water column and on the benthos) and can influence community stability, diversity and interspecies links (Dame 1996). Zebra mussels create high densities over large areas in lakes and efficiently filter large volumes of water. They deposit substantial amounts of pseudofeces and feces on the bottom. Thus, they play the same ecosystem engineering role as marine bivalves (Karatayev et al. 2002). We found that before invasion by zebra mussels, the planktonic community filtered the equivalent to the volume of each lake within a few days, and were on average > 200 times more effective than the benthic community, which took four years to filter the same volume. We found no significant changes in the total biomass or filtration capacity of the zooplankton community in lakes after zebra mussel invasion. However, the total average biomass of the benthic community, including zebra mussels, increased 22 times, and filtration ability of the benthic community increased > 70 times. Consequently, the time required to filter the volume of each lake for the benthos was no different than that for the zooplankton community. The impacts of increased benthic filtration on the ecosystem will depend on the size of the Dreissena population, lake morphometry and rate of water exchange, and will be more pronounced in littoral zone where zebra mussels are most dense and less in profundal areas of deep lakes where zebra mussels are rare or absent. Our results are consistent with the findings of other studies. During the summer, D. polymorpha has been estimated to filter the volume of water equivalent to that of an entire waterbody from 5 to 90 days (Mikheev 1967, Stanczykowska 1977, Lvova et al. 1980, Protasov et al. 1983, Reeders et al. 1989, Karatayev and Burlakova 1995a, Petrie and Knapton 1999). After D. polymorpha invaded Lake Lukomskoe (Belarus), the filtration capacity of the benthic community increased 320 times, and the time to filter the equivalent of the volume of the lake decreased from 15 years to 17 days. At the same time zooplankton abundance declined, and the time required for the zooplankton community to filter the equivalent of the volume of the lake increased from 5 to 17 days (Karatayev and Burlakova 1992, 1995b).

273 In most freshwater ecosystems, benthic production is driven by the slow rain of suspended organic material to the bottom and to a lesser extent by the filtration activity of bottom suspension-feeders where most species feed on detritus or other benthic organisms. Consequently, the typical benthic freshwater system is considered to be detritus dominated, rather than relying on large amounts of primary productivity or direct links to planktonic processes. Usually, the benthos are not capable of controlling processes or dynamics in the planktonic system. Zebra mussels, filtering vast amount of water in a short period of time, provide a direct link between processes in the plankton and those in the benthos and by their deposition of pseudofeces and feces, provide a direct conduit for primary productivity in the water column to the benthos. Thus, they are able to control pelagic processes by removal of particulate matter, increasing water transparency and hence the volume of the photic zone, impacting phytoplankton standing stock, and, therefore, they can influence planktonic trophic interactions (reviewed in Karatayev et al. 2002). As a result, the role of the benthic community in lakes populated by zebra mussels increases tremendously and the benthos become capable of controlling processes and dynamics in the planktonic system and, therefore, the whole freshwater ecosystem. All of these effects are the direct result of changes in the trophic structure of the benthic community after zebra mussel invasion and the overwhelming dominance of one trophic group – filterers.

ACKNOWLEDGMENTS We would like to acknowledge our colleagues and former students of Belarussian State University Igor Rudakovsky, Galina Vezhnovets, Nataliya Lisovskaya, Yulyana Shilenko, Elena Makritskaya, Andrei Usov, Sergej Mastitsky and Vladimir Volosyuk for help in data collection and species identification. We appreciate the help of Dr. Tatyuna Zhukova and the staff of Narochanskaya Biological Station of Belarussian State University during field research. This study was supported by grants from Belarussian State University (Research Grant # 444/50 to AYK and LEB) and Stephen F. Austin State University (Faculty Research Grant # 14123 to AYK, LEB and DKP).

REFERENCES Alimov AF 1981 Patterns of the Growth of Freshwater Bivalve Molluscs. Nauka Press, Leningrad, 248 p (in Russian) Berezina NA 1999 Peculiarities of development of macrozoobenthos communities under influence of Dreissena polymorpha Pallas in experimental mesocosms. Zhurnal Obshchei Biologiii 60: 189-198 Dame RF 1993 Bivalve Filter Feeders in Estuarine and Coastal Ecosystem Processes. Springer-Verlag, Heidelberg, 579 p

274 Dame RF 1996 Ecology of Marine Bivalves: An Ecosystem Approach. CRC Press, Boca Raton, FL, 254 p Izvekova EI 1975 The nutrition and feeding links of the most abundant species of Chironomid larvae in the Uchinskoe reservoir. Summary of the Candidate Dissertation, Moscow State University, Moscow, USSR (in Russian) Izvekova EI Lvova-Kachanova AA 1972 Sedimentation of suspended matter by Dreissena polymorpha Pallas and its subsequent utilization by chironomid larvae. Pol Arch Hydrobioll 19: 203-210 Karatayev AY Burlakova LE 1992 Changes in trophic structure of macrozoobenthos of a eutrophic lake, after invasion of Dreissena polymorpha. Biologiya Vnutrennikh Vod. Inform. Byull 93: 67-71 Karatayev AY Burlakova LE 1995a Present and further patterns in Dreissena polymorpha (Pallas) population development in the Narochanskaya lakes system. Vestsi Akademii Navuk Belarusi. Seriya biyalagichnikh navuk 3: 95-99 (in Belarussian) Karatayev AY Burlakova LE 1995b The role of Dreissena in lake ecosystems. Russian J Ecol 26: 207-211 Karatayev AY Burlakova LE Padilla DK 1997 The effects of Dreissena polymorpha (Pallas) invasion on aquatic communities in Eastern Europe. J Shellfish Res 16: 187-203 Karatayev AY Burlakova LE Padilla DK 2002 Impacts of Zebra Mussels on aquatic communities and their role as ecosystem engineers. In: Invasive Aquatic Species of Europe - Distribution, Impacts and Management, E Leppäkoski S Gollasch S Olenin (Eds). Kluwer, Dordrecht, pp 433-446 Karatayev AY Burlakova LE Padilla DK Johnson LE 2003 Patterns of spread of the zebra mussel ((Dreissena polymorpha (Pallas)): the continuing invasion of Belarussian lakes. Biological Invasions 5(3): 213-221 Karatayev AY Lyakhnovich VP Afanasiev SA Burlakova LE Zakutsky VP Lyakhov SM Miroshnichenko MP Moroz TG Nekrasova MY Skalskaya IA Kharchenko TG Protasov AA 1994 The place of species in ecosystem. In: Freshwater Zebra Mussel Dreissena polymorpha (Pall.) (Bivalvia, Dreissenidae). Systematics, Ecology, Practical Meaning, JI Starobogatov (Ed). Nauka Press, Moscow, pp 180-195 (in Russian) Karatayev AY Makritskaya EN 1999 Zooplankton of lakes in Naroch Region. In: The Results and Future of Aquatic Ecology Research. Proceedings of the International Conference on Aquatic Ecosystems. Minsk, BSU Press, Belarus, pp 108-114 (in Russian) Kinzelbach R 1992 The main features of the phylogeny and dispersal of the zebra mussel Dreissena polymorpha. In: The Zebra Mussel Dreissena polymorpha: Ecology, Biological Monitoring and First Applications in the Water Quality Management, D Neumann and HA Jenner (Eds.). Gustav Fisher, Stuttgart, pp 5-17 Kryuchkova NM 1989 Trophic Relationships Between Zoo- and Phytoplankton. Nauka Press, Moscow (in Russian), 124 p Lvova AA Izvekova EI Sokolova NY 1980 The role of benthic animals in the conversion of organic matter and in the processes of waterbody self-cleaning. Trudy Vsesoiuznogo Gidrobiologicheskogo Obshchestva 23: 171-177 (in Russian) Lvova-Kachanova AA Izvekova EI 1978 Zebra mussel and chironomids from Uchinskoe reservoir. In: Plant and Animal Life in Moscow and its Environments, TN Dunaeva et al. (Eds). Moscow University Press, Moscow, pp 119-121 (in Russian) McMahon RF Bogan AE 2001 Mollusca: Bivalvia. In: Ecology and Classification of North American Freshwater Invertebrates, 2ndd Edition, JH Thorp and AP Covich (Eds). Academic Press, New York, pp 331-430 Merritt RW Cummins KW 1996 An Introduction to the Aquatic Insects of North America (3rd. ed.). Kendall/Hunt Publishing Co., Dubuque, IA, 862 p

275 Mikheev VP 1967 The nutrition of zebra mussels ((Dreissena polymorpha Pallas). Summary of the Candidate Dissertation. State Research Institute for Lakes and Rivers Fishery Industry, Leningrad, USSR (in Russian) Minchin D Lucy F Sullivan M 2002 Zebra mussel: impacts and spread. In: Invasive Aquatic Species of Europe - Distribution, Impacts and Management, E Leppäkoski S Gollasch and S Olenin (Eds). Kluwer, Dordrecht, pp 135-148 Monakov AV 1998 The Feeding of Freshwater Invertebrates. Russian Academy of Sciences. A. N. Severtsov Institute of Ecological and Evolutionary Problems, Moscow, 318 p (in Russian) Monakov AV 2003 Feeding of Freshwater Invertebrates Belgium, Kenobi Prod, 400 p Molloy DP Karatayev AY Burlakova LE Kurandina DP Laruelle F 1997 Natural enemies of zebra mussels: predators, parasites and ecological competitors. Rev Fish Sci 5(1): 27-97 Petrie S Knapton R 1999 Rapid increase and subsequent decline of zebra and quagga mussels in Long Point Bay, Lake Erie: Possible influence of waterfowl predation. J Great Lakes Res 25: 772-782 Protasov AA Afanasiev SA Ivanova OO 1983 The distribution and role of Dreissena polymorpha in the periphytone of Chernobyl water cooling reservoir. In: Molluscs: Systematics, Ecology and Patterns of Occurrence. Abstracts of the 7th Meeting on the Investigation of Molluscs, Nauka Press, Leningrad, pp 220-222 (in Russian) Reeders HH bij de Vaate A Slim FJ 1989 The filtration rate of Dreissena polymorpha (Bivalvia) in three Dutch lakes with reference to biological water quality management. Freshwater Biologyy 22: 133-141 Ricciardi A Whoriskey FG Rasmussen JB 1997 The role of the zebra mussel (Dreissena ( polymorpha) in structuring macroinvertebrate communities of hard substrata. Can J Fish Aqua Sci 54: 2596-2608 Rice WR 1989 Analysing tables of statistical tests. Evolution 43: 223-225 Slepnev AE Protasov AA Videnina YL 1994 Development of a Dreissena polymorpha population under experimental conditions Hydrobiol. J 30: 26-33 Sokolova NY Izvekova EI Lvova AA Sakharova MI 1980a Structure, distribution and seasonal dynamics of benthic densities and biomass. Trudy Vsesoiuznogo Gidrobiologicheskogo Obshchestva 23: 7-23 (in Russian) Sokolova NY Izvekova EI Lvova AA Sakharova MI 1980b The ecology of mass species of bottom invertebrates. Trudy Vsesoiuznogo Gidrobiologicheskogo Obshchestva 23: 39131 (in Russian) Sokolova NY Izvekova EI Lvova AA Sakharova MI 1980c Features of benthos forming in small waterbodies using Uchinskoe Reservoir as an example. Trudy Vsesoiuznogo Gidrobiologicheskogo Obshchestva 23: 161-170 (in Russian) Stanczykowska A 1977 Ecology of Dreissena polymorpha (Pall.) (Bivalvia) in lakes. Pol Arch Hydrobioll 24: 461-530 Stánczykowska A Lewandowski K 1993 Thirty years of studies of Dreissena polymorpha ecology in Mazurian Lakes of northeastern Poland. TF Nalepa DW Schloesser (Eds). Zebra Mussels: Biology, Impacts, and Control, Lewis Publishers, Boca Raton Stewart TW Gafford JC Miner JG Lowe RL 1999 Dreissena-shell habitat and antipredator behavior: combined effects on survivorship of snails co-occurring with molluscivorous fish. J N Am Benthol Soc 18: 274-283 Stewart TW Miner JG Lowe RL 1998 Quantifying mechanism for zebra mussel effects on benthic macroinvertebrates: organic matter production and shell-generated habitat. J N Am Benthol Soc 17: 81-94 Thorp JH Covich AP 2001 Ecology and Classification of North American Freshwater Invertebrates (2ndd Ed.) Academic Press, New York, 950 p Wallace JB Merritt RW 1980 Filter-feeding ecology of aquatic insects. Ann Rev Entomoll 25: 103-132 Zar HJ 1996 Biostatistical Analysis. (3rdd Ed.) Prentice Hall, New York, 662 p

DOES THE INTRODUCTION OF THE PACIFIC OYSTER CRASSOSTREA GIGAS S LEAD TO SPECIES SHIFTS IN THE WADDEN SEA?

Aad Smaal1, Marnix van Stralen2, Johan Craeymeersch1 1

Netherlands Institute for Fishery research, Centre for Shellfish research, Yerseke, The Netherlands 2 MarinX Consultancy, Elkerzee, The Netherlands Abstract: Over centuries dramatic changes have occurred in the species composition of the Wadden Sea, a shallow coastal sea bordering the North Sea. Natural dynamics as well as direct and indirect anthropogenic influences have resulted in the introduction and the disappearance of important benthic populations. Historic records and extensive surveys show large variability in benthic suspension-feeder stocks. Infaunal species like the cockle (Cerastoderma edule) are extremely variable over time and space, hence show a typical resilient response. Mussel (Mytilus edulis) beds seem to be more stable over time. Once lost, mussel beds need more time to re-establish bed structures. It is hypothesized that infaunal populations have a high resilience, while epifauna species are characterized by resistance to changes as they form structures like reefs or beds. On the basis of this hypothesis the consequences of new introductions can be evaluated. It can be expected that the recent introduction of the resistant reef-building epifaunal Pacific oyster Crassostrea gigas, will lead to shifts in benthic suspensionfeeder populations and eventually will develop a new stable state for the Wadden Sea that potentially offers less food for birds. This situation may deviate considerably from the actual nature conservation objectives that focus on the role of the Wadden Sea as one of Europe’s most important wetlands for migratory bird populations. Keywords: mussels, cockles, oysters, resistance, resilience, fishery.

INTRODUCTION The Wadden Sea (8000 km2) is a shallow estuarine ecosystem that stretches from Denmark to the Netherlands (Fig. 1). About 50 % of the Wadden Sea consists of tidal flats, and these are important habitats for benthic populations and their predators, predominantly wader birds. About 400 macrobenthic species occur in the Wadden Sea (Petersen et al., 1996). Cockles (Cerastoderma edule) and mussels (Mytilus edulis) dominate macrobenthos abundance. These suspension-feeder populations show a high natural variability over time and space, reflecting impacts of storms, 277 R.F. Dame and S. Olenin (eds.), The Comparative Roles of Suspension-Feeders in Ecosystems, 277–289. ©2005 Springer. Printed in the Netherlands.

278

Fig. 1. The Wadden sea area including transport routes of Crassostrea gigas (after Reise, 2004)

severe winters and variable recruitment success. As they are exploited species, there are also fishing or harvesting influences. These species are typically rstrategists showing high fecundity, broadcast spawning and a high potential for colonizing areas. Mussel stocks are less variable than cockle stocks, which might be related to the ability of the mussels to form bed structures. However, when these bed structures disappear due to storms or harvesting the restoration of bed structures may take quite a while (Dare et al 2004). Historic records show both extirpations (Wolff 2000) and introductions (Reise et al. 2002) of species in the Wadden Sea. With regard to benthic suspension-feeders the decline of the flat oyster (Ostrea edulis) in the period 1920 – 1940 is well-described (Dijkema 1997; Hagmeier 1941). Introductions of Mya arenaria (ca 1250), Mytilopsis leucophaeta (1835, brackish water species), Mercenaria mercenaria (1864), Crepidula fornicata (1887), Petricola pholadiformis (1890) and Ensis americanus (1978) have recently been reviewed by Reise et al (2002). According to Wolff (2000), the main causes of extermination are overexploitation and habitat destruction. To date, there is little evidence that introductions have driven native species to extinction (Reise et al. 2002). A relatively new species for the area is C. gigas

279 that was first observed on hard substrates along the dikes in 1983 (Bruins 1983). It has expanded rapidly since and also colonized tidal flats areas. In contrast to the introduced infaunal species it might be expected that the oyster, that builds reef structures on the tidal flats, will cause shifts in suspensionfeeder populations and potentially will bring the system to a new stable state. This potentially new state raises questions for nature management as present objectives focus on dominant bird populations that depend on the existing suspension-feeders such as cockles and mussels and not on those depending on oysters. On the basis of new data on the abundance and distribution of the introduced Pacific oyster in relation to the dynamics of existing benthic suspension feeder populations, we want to test the hypothesis that the rapid proliferation and the dominance of the Pacific oyster will result in shifts of suspension-feeders and a change in the state of the Wadden Sea ecosystem.

MATERIALS AND METHODS Annual Assessment of Shellfish Stocks From 1990 onward, each spring a survey is carried out by the Netherlands Institute for fishery Research (RIVO) to estimate the total standing stock of dominant bivalve species of the tidal flats of the Dutch Wadden Sea. The primary aim of the survey is to establish whether the stock size exceeds a given threshold that is conserved as food for wader birds. If the stock is below the threshold no cockle or mussel seed fishery is allowed. Following a stratified approach samples are taken at 1500 stations in the Wadden Sea (Fig. 2). The sampling grid is based on data from previous surveys. Also, results of explorative cockle surveys by fishermen just prior to our survey are used as well as aerial surveys prior to sampling in order to identify mussel and oyster bed structures. Sampling is done with a cockle-fishing vessel with a draught of 50 cm, allowing sampling of intertidal areas at high water. Samples are taken with an adapted mechanical cockle dredge that allows sampling of 0.5 m2/sample. Stations higher in the intertidal zone are sampled with a handheld sampling device (0.1 m2), operating on a small boat, or at low water with a corer (0.1 m2). Sampling depth is 7 cm, and mesh size is 5 mm. Quantitative sampling is realized for 15 species (Table 1). Samples are sorted, analyzed and wet weight is registered on board and data are directly stored in a database. In addition to sampling, mussel and oyster beds are quantified by satellite positioning (GPS) of the location and size of the individual beds, and by an estimate of composition and density. These data are stored in a database and analyzed with an Arcview geographic information system (GIS).

Ameland TERSCHELLING

Vlieland

Texel

Fig. 2. RIVO survey area of tidal flats in the Dutch Wadden Sea, with stratified sampling stations.

280

SCHIERMONNIKOOG

281 Table 1. Dominant benthic suspension-feeder species in the Dutch Wadden Sea with densities and standing stock in 2002, on the basis of an extensive spring survey on the tidal flats. (*) introduced species, see text Species Cerastoderma edule Crassostrea gigas* Donax vittatus

N/m2

106 kg

18.988 0.003

168.70 0.07

0.047

0.08

Ensis americanus*

2.061

0.76

Macoma balthica

51.910

50.78

Mya arenaria*

11.940

9.70

Mytilus edulis

34.140

50.30

Petricola pholadiformis*

0.007

0.11

Tellina tenuis

0.091

0.02

RESULTS AND DISCUSSION Dominant Populations The top 10 benthic suspension-feeder species for 2002 as they are found in our annual survey of the tidal flats of the Dutch Wadden Sea (Table 1). Densities vary over several orders of magnitude with Macoma as the most abundant species. In terms of biomass dominant species are the cockle (C. edule) and the mussel (M. edulis). Total stocks show large fluctuations over time (Fig. 3) and there is considerable spatial heterogeneity as shown in Fig. 4A and B for cockle and mussel beds. Large-scale fishery of cockles exists in the Dutch Wadden Sea – it was abandoned in the German and Danish Wadden Sea, while mussels are fished and cultured along the whole Wadden Sea (Smaal 2002, Kamermans and Smaal 2003). Among the top 10 list there are 4 introduced species: M. arenaria, P. pholadiformis, E. americanus and C. gigas. M. arenaria is the oldest record of anthropogenic impacts, related to transatlantic expeditions of the Vikings (Petersen et al. 1992, Wehrmann et al, 2000). It is a dominant species in the Wadden Sea and generally not considered as exotic. In contrast to the US East Coast where it originates from, no commercial exploitation occurs of Mya in NW Europe. Ensis has increased its biomass rapidly since first observations in 1978 near Helgoland, where it was introduced in the ballast water of ships. The species expanded through larval drift at about 100 km/year both in western and northern direction (Armonies and Reise 1999, Armonies 2001). It is now also dominant along the Dutch North Sea coast (RIVO data). Biomass is underestimated due to

282 insufficient sampling gear. At present 3 ships exploit the species in this area. The species is an important prey for Eider ducks (Reise 2004). Petricola was introduced more than 100 years ago by shipping and is less dynamic, has a low density and is not commercially exploited. The abundance of the Pacific oyster has increased dramatically on the tidal flats of the Dutch Wadden Sea and spatial distribution of Pacific oysters in the Dutch Wadden Sea is given in Fig 4C. Beds were found near Texel, in the central area, near Ameland and near Schiermonnikoog. The data in Table 1 shows the results of the 2002 survey. From our recent survey of 2004, total stock was calculated as 11.5 million kg wet weight forming at least 400 ha of bed area of > 200 g/m2. New beds were found over the whole Dutch Wadden Sea and also in the Ems estuary.

Proliferation of Crassostrea gigas The proliferation of the Pacific oyster in Europe started after the severe winter of 1962/63 when the species was introduced to restore oyster cultures (flat oyster, Ostrea edulis) that almost became extinct due to the frost. Apart from unsuccessful introductions in the Wadden Sea in 1913 in Lower Saxony (Wehrmann et al. 2000), it was not until the Pacific oyster was introduced in 1964 in the Oosterschelde estuary (SW Netherlands) that they developed rapidly (Drinkwaard 1999). Since 1974 culture trials were done in Lower Saxony, but none of these introductions resulted in an economically feasible population (Wehrmann et al. 2000). In the eastern Wadden Sea a culture started near Sylt in 1986 (Reise 1998). It is obvious that the occurrence of wild Pacific oysters in the Wadden Sea is due to aquaculture introduction, but it is remarkable that proliferation occurred over quite large distances. First records of wild Pacific oysters date from 1983 near Texel, the western part of the Dutch Wadden Sea, and from the size it was estimated that these oysters had been there since 1977. The oysters may have been introduced accidentally by mussel boats that frequently sail between the Oosterschelde and the Wadden Sea, although the only transport of shellfish is from the Wadden Sea to the Oosterschelde and not vice versa. Also larval drift may have bridged the gap, as a distance of 150 km can be reached by larvae within 20 – 25 days on the basis of a residual current of 0.1 m/s, while the life span of pelagic larvae is up to 30 days (Wehrmann et al. 2000, but see also Cadée 2000). Further records date from 1998 in the Ems-Dollard area, mainly on hard substrates on the dikes (Tydeman 1999) and along the Lower Saxony coast, particularly on mussel beds (Wehrmann et al. 2000). Around the island of Sylt first observations of wild Pacific oysters are from 1991 near the culture area, and in 1995 wild beds were observed south of Sylt and later on also in northern areas (Fig. 1). Apparently the local culture

2004

2002

2000

1998

1996

1994

A. Cockles

1992

500 450 400 350 300 250 200 150 100 50 0

1990

mln kg fresh weight

283

mln kg fresh weight

70 60

B. Mussels

50 40 30 20 10

1996

1997

1998

1999

2000

2001

2002

2003

1996

1997

1998

1999

2000

2001

2002

2003

1995

1994

1993

1992

1991

90 80 70 60 50 40 30 20 10 0

1995

1994

1993

1992

1991

C. Macoma

1990

mln kg fresh weight

1990

0

Fig. 3. Variations in stock size of major benthic suspension-feeder populations in the Dutch Wadden Sea (A) cockles, (B) mussels and (C) Macoma

284

A. Cockles

B. Mussels

Fig. 4. Spatial distribution of cockles (2002), mussels (2002) and Pacific oysters (2004). In Figs. 4A and B, the intertidal area is grey. In Fig. 4C, the sampling stations are shown in grey.

285 population was the source for proliferation near Sylt, while the other Wadden Sea areas have been colonized by long distance larval dispersal with the Oosterschelde population as a resource.

Dynamics of Crassostrea gigas The dynamics of the oyster population is clearly related to recruitment success. Observed increases in abundance always followed years with good recruitment, and these correspond with extended periods of high summer temperature (Wehrmann et al. 2000). This was documented for the Oosterschelde in 1976, 1982, 1986, 1987, 1989, 1992 (Drinkwaard 1999) and a series of recent years. From aerial photographs, a reconstruction was made of the expansion of oyster beds in the Oosterschelde from 1980 onwards. It shows a development from 30 ha (1980) to 300 ha (1990) and is now 700 ha (2000) (Kater and Baars 2003). Meanwhile sublittoral stocks have been estimated on the basis of side scan sonar survey as approximately 700 ha (Kater et al. 2002). Climate events like severe winters or warm summers, or biotic processes like predation or diseases do not seem to hamper the increase of the population. Warm summers are considered as the main vector for extension of the population (Reise 2004). In 2002 a good spatfal occurred and this is reflected in the dramatic increase of biomass as recorded in the Wadden Sea, from 0.07 million kg in 2002 to 11.5 million kg in 2004. In 2003 summer was extremely warm and high larval concentrations were measured in a monitoring program in the Dutch Delta area (Kamermans, pers. com.), hence a further biomass increase is to be expected. For the Dutch Wadden Sea, early observations are quite limited and most records are from 1995 onwards (Dankers et al. 2004). It is only since 2002 that oysters have become apparent in our annual spring survey in a quantitative way.

Further Developments It is expected that the Pacific oyster will expand further on the tidal flats in the Wadden Sea. From our survey we are able to quantify the standing stock and the oyster bed area in the Dutch Wadden Sea, and a rapid increase in only a few years is evident. Observations from the Oosterschelde estuary show a continuous expansion on the tidal flats with a doubling of covered area in 10 years time. Proliferation started on hard substrates along the dikes in the seventies but by 1980 about 30 ha of bed area could be quantified from aerial

286 photographs of tidal flats. The colonialization of tidal flats starts with settlement of larvae on shell fragments and alike, and then the oyster shells form substrate for new spat. Hence a positive feedback exists when oysters have settled and start colonizing an area. This mechanism obviously results in the formation of reefs and dense oyster beds. We also observed the occurrence of adult oysters in areas that had no oysters the year before. Apparently, wave driven transport of individual oysters or clumps of oysters occurs and new areas can be colonized in the vicinity of existing beds through this mechanism. As shown, recruitment is the driving force of pacific oyster expansion and this is particularly successful in warm summers. As climate change may result in an increase of summer temperatures in the Wadden Sea area, it is quite likely that the expansion rate will increase. This fits in the observation that recruitment of benthic species changes under the influenced of temperature rise (Beukema 1992). Without control measures it is expected that the Pacific oyster will become a dominant species in the Wadden Sea and will change the landscape from tidal flats with dominant infauna and mussel beds to areas that are at least partly covered with oyster reefs. The oyster has a large filtration and biodeposition capacity, hence will play a dominant functional role as well in the Wadden Sea. As shown by Dame (1996, 2005), oyster reefs are complex systems, characterized by multiple feedbacks and a high level of selforganization. These feedbacks stabilize the reef system and create resistance to changes. Eventually a new stable state of the ecosystem may be established.

Potential Impacts and Measures The distribution of the Pacific oysters on the tidal flats may change the landscape. As mentioned, oysters form reefs and dense beds, and outcompete local infauna. Particularly in the Wadden Sea, oysters also settle on existing mussel beds. This was observed in the Dutch Wadden Sea (Dankers et al. 2004) and in Lower Saxony (Wehrmann et al. 2000) and Sleswich Holstein (Reise 1998). It is possible that mussel beds will gradually be taken over by the oysters. The Pacific oyster has a large filtration capacity and filters on average 5 l/g/h (Bougrier et al. 1995) but values up to 25 l/g/h have been recorded (K. Troost pers. comm, Ren et al 2000). As a consequence, there is a potential competition for food with other suspension-feeders and this may have impact on dominant mussel and cockle populations. Current research is focused on the potential of oysters to filter bivalve larvae. It is hypothesized that the oyster reefs may filter large quantities of larvae from the water column, hence reduce the recruitment success of species like macoma, cockle and mussel.

287 Also the competition for food may have impact on the early life stages of suspension-feeders as they also depend on phytoplankton. If the Pacific oysters have a strong competition potential when compared to other benthic suspension-feeders, a decrease in the population size of the other species can be expected. This may have consequences for the food chain, as the Wadden Sea has large bird populations feeding on bivalves like cockles, macoma and mussels. These birds, like - despite its name - the oystercatcher, are not able to feed on the Pacific oyster. The question is therefore what may limit the Pacific oyster. As it is an exotic species it can be expected that it may take quite a while before a natural regulation process will evolve. As shown, predation is not a serious threat so far for the Pacific oyster in this area, as major bird populations cannot handle the species. In France the predatory gastropod Ocinebrellus inornatus has impact on cultured oyster stocks, but this species does not occur in the Wadden Sea. Harvesting Pacific oysters is at present not allowed due to nature conservation directives. The quality of the oysters for consumers is excellent. Therefore there is a chance to harvest good quality product and meanwhile manage the size and spatial distribution of the stock. However, once the oysters have developed reefs the product quality for the consumers decreases dramatically due to clumping, increase of shell size and decrease of meat content. Combining management and harvesting requires therefore timely decisions.

CONCLUSIONS The Pacific oyster is rapidly becoming a dominant suspension-feeder on the tidal flats of in the Wadden Sea. The oyster reefs form a complex system with multiple feedbacks, a high level of self-organization, and a high resistance. As they colonize both mussel-beds and infaunal habitats, and they have a large filtration and biodeposition capacity, the species has a strong competition potential. Shifts in benthic suspension-feeder population may be expected and as a consequence food availability for dominant bird populations may decrease. In contrast to earlier invasions of resilient infaunal suspensionfeeders like E. americanus, the introduction of C. gigas may result in a new stable state in the Wadden Sea ecosystem with large areas of oyster reefs and lower stocks of other benthic suspension-feeders and related bird populations.

Acknowledgements The authors are grateful to Richard Dame for comments and improvements of the manuscript, and to Josien Steenbergen, Divera Baars and

288 Joke Kesteloo for help with data collection and analysis. Data collection was done under contract of the WOT Programme of the Dutch Ministry of LNV. NATO ARW is acknowledged for financial support of the first author to participate in the Nida workshop.

REFERENCES Armonies W 2001 What an introduced species can tell us about the spatial extension of benthic populations. Mar Ecol Prog Ser 209: 289-294 Armonies W Reise K 1999 On the population development of the introduced razor clam Ensis americanus near the island of Sylt (North Sea). Helgoländer Meeresunters 52: 291-300 Beukema JJ 1992 Expected changes in the Wadden Sea benthos in a warmer world: lessons from periods with mild winters. Neth J Sea Res 30: 73-79 Bougrier S Geairon P Deslous-Paoli JM Bacher C Jonquieres G 1995 Allometric relationships and effects of temperature on clearance and oxygen consumption rates Crassostrea gigas (Thunberg). Aquaculture 54: 143-154 Bruins RBW 1983 Crassostrea gigas op Texel. Correp Blad Ned Malacol Verr 215: 14361438 (in Dutch) Cadée G 2000 Japanse oester (Crassostrea gigas) populaties tussen Oudeschild en Mok, Texel. Het Zeepaardd 60: 260-269 Dame RF 1996 Ecology of Marine Bivalves: An Ecosystem Approach. CRC Press Boca Raton, FL, 254 p Dame RF 2005 Oyster reefs as complex ecological systems. In: The Comparative Roles of Suspension-Feeders in Ecosystems, RF Dame S Olenin (Eds), Kluwer Academic Publishers, Dordrecht, pp 331-344 Dankers NMJA Dijkman EM Jong ML de Kort G de Meijboom A 2004 De verspreiding en uitbreiding van de Japanse oester in de Nederlandse Waddenzee. Alterra rapport 909, 51 p(in Dutch) Dare PJ Bell MC Walker P Bannister RCA 2004 Historical and current status of cockle and mussel stocks in the Wash, CEFAS, Lowestoft, 85 p Dijkema R 1997 Molluscan fisheries and culture in The Netherlands. NOAA Technical Report NMFS 129: 115-136 Drinkwaard AC 1999 Introductions and development of oysters in the North Sea: a review. Helgoländer Meeresunters 52: 301-308 Hagmeier A 1941 Die intensive Nutzung des nordfriesischen Wattenmeeres durch Austernund Muschelkultur. Z Fischerei 39: 105-165 Kater B Baars D 2003 Reconstructie van oppervlakten van litorale japanse oesterbanken in de Oosterschelde in het verleden en een schatting van het huidige oppervlak. RIVO rapp C017/03 Kater B Baars D Perdon J van Riet M 2002 Het inventariseren van sublitorale oesterbestanden in de Oosterschelde mbv side scan sonar. RIVO rapp C058/02, 26 p Kamermans P Smaal AC 2003 Mussel culture and cockle fishery in the Netherlands: finding a balance between economy and ecology. J Shellfish Res 21: 509-517 Petersen GH 1996 Red list of macrofaunal benthic invertebrates in the Wadden Sea. Helgoländer Meeresunters 50(suppl): 69–76 Petersen KS Rasmusen KL Heinemeier J Rud N 1992 Clams before Columbus? Nature 359: 679 Reise K 1982 Long-term changes in the macrobenthic invertebrate fauna of the Wadden Sea: are the polychaetes about to take over? Neth J Sea Res 16: 29-36

289 Reise K 1998 Pacific oysters invade mussel beds in the European Wadden Sea. Senckenbergiana maritima 28: 167-175 Reise K 2004 Contribution to Wadden Sea quality status report, in press Reise K Gollasch S Wolff WJ 2002 Introduced marine species of the North Sea coasts. In: In: Invasive Aquatic Species of Europe: Distribution, Impacts and Management, E Leppäkoski S Gollasch S Olenin (Eds), Kluwer Academic Publishers, Dordrecht, pp 260-266 Ren JS Ross AH Schiel DR 2000 Functional descriptions of feeding and energetics of the Pacific oysterr Crassostrea gigas in New Zealand. Mar Ecol Prog Serr 208 : 119-130 Smaal AC 2002 European mussel cultivation along the Atlantic coast: production status, problems and perspectives. Hydrobiologia 484: 89-98 Tydeman P 1999 Japanse oesters in de Eemshaven. Het Zeepaard 59(2): 58-64 Wehrmann A Herlyn M Bungenstock F Hertweck G Millat G 2000 The distribution gap is closed – first records of naturally settled Pacific oysters Crassostrea gigas in the East Frisian Wadden Sea, North Sea. Senckenbergiana maritima 30: 153-160 Wolff WJ 2000 Causes of extirpations in the Wadden Sea, an estuarine area in The Netherlands. Cons Bioll 14(3): 876-885 Wolff WJ Reise K 2002 Oyster imports as a vector for the introduction of alien species into northern and western European coastal waters. In: In: Invasive Aquatic Species of Europe: Distribution, Impacts and Management, E Leppäkoski S Gollasch S Olenin (Eds), Kluwer Academic Publishers, Dordrecht, pp 193-205

ONE ESTUARY, ONE INVASION, TWO RESPONSES: PHYTOPLANKTON AND BENTHIC COMMUNITY DYNAMICS DETERMINE THE EFFECT OF AN ESTUARINE INVASIVE SUSPENSION-FEEDER

Janet K. Thompson United States Geological Survey, Menlo Park, CA, US Abstract: Invasive suspension-feeding bivalves have reduced phytoplankton biomass in many aquatic systems, which has resulted in loss of trophic complexity in some systems. Using an example of one invasive bivalve in San Francisco Bay, Potamocorbula amurensis, the causes of differing system level responses are explored. San Francisco Bay, similar to of other shallow, turbid, non-nutrient limited, but low productivity systems, is likely to be most stressed by the loss of primary producers. While the northern bay has lost primary production following the invasion of P. amurensis, the southern bay (SB) has not and these differences are shown to be due to the different mechanisms responsible for the seasonal turbidity in the systems. Because the period of lowest turbidity in SB is coincident with the period of lowest bivalve grazing, the southern bay has not seen a reduction in its high magnitude but short spring bloom. A method for predicting if a shallow, turbid and nutrient replete estuary might lose phytoplankton production with a sudden increase in suspension-feeders is explored. Keywords: suspension-feeder, bivalve, exotic, grazing, San Francisco Bay, phytoplankton

INTRODUCTION Estuaries are recipients of a high rate of human-mediated invasions by non-indigenous species. Ruiz et al. (2000) report a minimum of 374 successful invasions of algae and invertebrates, more than 100 successful invasions of fish, and over 200 successful invasions of vascular plants in North American coastal marine communities. Anthropogenic transport of species is recent in North America, so it is expected that these numbers are not large, relative to the historical invasions of Europe (Leppakoski et al. 2002). Because human mediated invasions are believed to be accelerating in many geographic areas on all continents (Ruiz and Hewitt 2002), it is important to understand how 291 R.F. Dame and S. Olenin (eds.), The Comparative Roles of Suspension-Feeders in Ecosystems, 291–316. ©2005 Springer. Printed in the Netherlands.

292 they affect aquatic ecosystems. One group of invasive organisms, those that suspension-feed, are of particular interest because they have changed estuaries at the system scale by reducing phytoplankton biomass (Alpine and Cloern 1992, Cohen et al. 1984) resulting in declines at higher trophic levels (Kimmerer 2002, Ojaveer et al. 2002, Shiganova 1998), by changing substrate character and thereby changing the suitability of the bottom for commercially important species (Chauvaud et al. 1998), and by altering nutrient regeneration rates (Chauvaud et al. 2000, Ojaveer et al. 2002) thereby potentially affecting phytoplankton community composition and growth rates. However, there have also been many estuaries with a rich history of invasive suspensionfeeder species that have not responded with these types of system level effects. Because of these disparate responses, invasions of suspension-feeders in aquatic systems can be viewed as system-scale experiments that allow us to explore what characteristics allow systems to be vulnerable to suspensionfeeder control of the food web. We can predict, based on an elegant and simple relationship relating average annual system clearance time by bivalves (CT) to water mass residence time (RT) developed by Smaal and Prins (1993), if the phytoplankton have the potentiall of being controlled by bivalve suspensionfeeders in most systems. The question that we are interested in here is slightly different: Is it likely that an increase in benthic suspension-feeding will control the phytoplankton biomass of a system? To address this question, we should include a combination of factors that are required for phytoplankton growth and accumulation: sufficient nutrients, sufficient light, and grazing loss rates that are low, relative to phytoplankton growth rates. We will limit our need to account for all of these factors by focusing on that class of estuaries whose phytoplankton and trophic webs are most vulnerable to overgrazing - estuaries that have low primary production despite non-limiting nutrients, due to a combination of light limitation and grazing pressure. As shown by Monbet (1992) and Cloern (2001) these types of systems are broadly distributed and common. We begin by exploring the vulnerabilities of estuarine food webs to invasive suspension-feeders by examining the specifics of an invasion of one bivalve suspension-feeder in the San Francisco Bay. We will see that the physical characteristics of the different parts of this system resulted in different phytoplankton responses to the invader. Using this example as a foundation, I will then develop a relationship to test when phytoplankton in turbid, low production estuaries are most vulnerable to significant reduction in biomass with invasive benthic suspension-feeders. I will conclude with a discussion on what we have learned about estuarine ecosystems that may or may not have changing, alternate ecosystem states.

293 The Estuary as an Environment for Benthic Suspension-feeders and for Phytoplankton Estuarine ecosystems are vulnerable to change by invasive suspension-feeders due to the characteristics of a successful invader, the inherent vulnerability of primary producers to suspension-feeders in shallow systems, and the low primary production of many estuaries (Cloern 2001, Heip et al. 1995). All benthic species must survive a range of environmental stresses, such as rapidly changing salinity and turbidity, in order to be successful in the estuarine environment. However, successful invasive species may be particularly tolerant and resilient to the wide temporal and spatial ranges of physical stressors having survived the selection process during transit to, and inoculation into a new system. Once established, those invasive benthic suspension-feeders that can expand into the full range of physical environments and can tolerate the largest range of physical stresses induced by tidal action can benefit from these same physical forces. Food supply for the benthos is increased in tidal regions due to: (1) increased vertical mixing rates, (2) advective transport of food sources and (3) in some cases increased water column residence time resulting in higher primary producer growth and accumulation.

Figure 1 Map of San Francisco Bay with place-names used in text.

The physical environment can limit primary production in estuaries relative to coastal systems and most freshwater systems (Cloern 2001).

294 Although phytoplankton in some estuaries are nutrient limited, many estuaries receive high concentrations of nutrients with little effect on primary production (Heip et al. 1995) reflecting the importance of other physical factors in estuarine phytoplankton growth. Light limitation is so common in estuaries that many estuaries have photic depths that are limiting to phytoplankton growth in the deepest portions of the system and during at least some portions of the year in the shallow water due to a combination of tidal mixing, wind resuspension of sediments, and freshwater inflow. As we have begun to understand more about the origin and variability of phytoplankton blooms (Lucas et al. 1999a, b and Heip et al. 1995) and estuarine turbidity (Cloern 1999, Monbet 1992), we see that the shallow water regions and the exchange between shallow and deep water in each system may determine the temporal and spatial variability of primary producers in estuarine systems. Thus any natural or anthropogenic interruption of this exchange may further limit primary production.

San Francisco Bay: Physical Setting and Exotic Species History San Francisco Bay (SFB) is a temperate, middle-latitude estuary located in the center of a large urban center (>8 million people), and is the natural terminus for 70% of the state’s annual freshwater runoff. However, 80% of the freshwater needs occur in the southern portion of the state and thus about 50% of the freshwater that historically flowed into SFB is now being retained for urban and rural consumption and hydroelectric power (Peterson et al. 1985). The bay is very shallow (median depth is 2.8 m at MSL), mesotidal, and has unequal, semi-diurnal tides. The northern bay can be divided into two sections (Figure 1), North Bay (NB: Honker, Suisun, and Grizzly Bays), through which the majority of freshwater enters, and San Pablo Bay (SPB), the next down-bay system. The most southerly bay, South Bay (SB), is the most saline system, except for the bay at the ocean interface (Central Bay), due to the diversion of most of the freshwater from the southern watershed. The two major systems, NB and SB, are therefore hydrologically and hydrodynamically very different. Freshwater flow peaks in NB in early winter and again in spring with the release of snowmelt water from reservoirs and varies in magnitude between years and seasons. High freshwater flow results in short residence times (§1 day) and high suspended loads in winter and spring. High turbidities also occur in summer due to the resuspension of the sediment deposited during freshwater runoff, by semi-diurnal tides and summer diurnal winds. Large interannual variability in precipitation (Figure 2) is responsible for ill-defined “average” conditions in NB. Salinity ranges

295 from 0-25 psu and the salinity gradient is substantial throughout NB, which is classified as a river-dominated partially mixed estuary. In contrast to the NB, the majority of the freshwater flowing into SB is from sewage effluent, and during very wet years, the southerly advection of the snowmelt runoff. Residence times are longer than in NB (in excess of 14 days during summer and fall, Gross 1997, Walters et al. 1985) and the turbidity, although less than in NB, is still sufficient to restrict phytoplankton growth in the narrow channels (Cloern et al. 1985). The SB is classified as a lagoonal estuary (salinity range of 15-30 psu). Persistent stratification is rare in either system for periods longer than about two weeks except in limited areas of NB where channels have sharp bathymetric gradients (Monismith et al. 2000). Neither system is very productive (20 g C m-2 yr-1 in NB and 150-200 g C m-2 yr-1 in SB, Alpine and Cloern 1992 and Cloern 1987) or nutrient limited. Primary production is predominantly light limited with significant top down control by benthic grazers in both systems (Cloern 2001). The major organic carbon source throughout the system is phytoplankton (Jassby et al. 1993). San Francisco Bay may be unique due to the dominance and number of invasive species in the system: there are a minimum of 234 known invasive species, and an additional 125+ cryptogenic species in this system (Cohen and Carlton 1998) with an average of one new species arriving every 24 weeks since 1970 (Cohen and Carlton 1995). The discovery of gold in the foothills and mountains east of San Francisco Bay in 1849 resulted in an influx of 48,000 people into the area by ship in just the first two years of the gold rush. The introduction of organisms living on the ships and in their dry ballast is likely responsible for many of the early invasions, although the largest pulse of invasions prior to 1940 occurred with the near continuous import of oyster spat from the western Atlantic (1869-1910) and the eastern Pacific (19321939) for culturing (Bonnott 1935). The rapid increase in invasive species that began in the 1940’s, following the conversion of dry ballast to wet ballast in most ships, is convincingly argued to be due to ballast water inoculations by Carlton (1985) and Cohen and Carlton (1998). The majority of the identified invasive species are invertebrates (147, Cohen and Carlton 1995) and the majority of those are suspension-feeders. Some of the invasive bivalve suspension-feeders became sufficiently successful that they were commercially harvested until a combination of poor meat quality and contamination made them unpalatable. Given this history of species invasions, it was not surprising to find a new bivalve, Potamocorbula amurensis, a suspension-feeding, euryhaline corbulid, in the NB in 1986. It was however, quite surprising to watch the speed with which it spread and its apparent tolerance of a wide range of environmental conditions (Carlton et al. 1990). Today, we find P. amurensis in the full sediment, depth, and salinity range of the estuary. Although we do not believe that it can reproduce in fresh water, it can osmoregulate as a larva

296 (10-30 psu at 2 hours after fertilization; Nicolini and Penry 2000) and juveniles and adults are commonly found at the freshwater boundary where the animals are routinely exposed to fresh water during some tidal conditions. Although P. amurensis was first found in NB, it quickly spread to the SB in 1988 (Thompson unpublished data), where it became a dominant bivalve by 1990. However, it has not been as interannually persistent in SB as in NB.

Data Sources and Methods Water column chlorophyll a values used in this paper are courtesy of the US Geological Survey and the California Department of water resources. The data are available on the following databases: http://sfbay.wr.usgs.gov/access/wqdata/ http://sarabande.water.ca.gov:8000/~bdtdb/ Biomass estimates of P. amurensis in NB that were used for estimated grazing rate were based on benthic samples loaned to the author by the California Department of Water Resources from the longest sampled benthic station in the estuary. Although only one station has been sampled consistently over the 30 year time period, other spatially intensive surveys of the benthos have shown P. amurensis biomass at this station to be reasonably representative of the biomass at other shallow water locations in Grizzly Bay (Hymanson 1991). Biomass and grazing rate estimates for SB suspensionfeeders represent an average of 41 to 76 stations and were reported in Thompson (1999). Grazing rates are based on measured filtration rates for P. amurensis (Cole et al. 1992) and Venerupis japonica (O’Riordan et al. 1995) and literature values for Mya arenaria and Musculista senhousia and all have been seasonally corrected for temperature, maximum concentration boundary layer development as suggested by O’Riordan et al. (1995), and a feeding frequency of 67%; this percentage represents the intermittent feeding behavior that was observed by the author in the field and in a laboratory flume for these species.

SAN FRANCISCO BAY – A TWO ESTUARY LABORATORY FOR STUDIES OF SUSPENSION-FEEDER PROCESSES P. amurensis has been blamed for the demise of the phytoplankton bloom (Figure 2) in the northern-most bays and has been shown to be capable of consuming the phytoplankton at a sufficient rate to be effectively controlling phytoplankton biomass in these systems (Alpine and Cloern 1992,

297 Thompson et al. in review). Although P. amurensis has also successfully invaded SPB and SB, we have not seen similar, significant declines in phytoplankton biomass (Figure 3) in these systems. The basic mechanisms for phytoplankton growth are similar for all embayments in this nutrient replete but light-limited estuary. The shallow portions throughout the bay are the most productive with chlorophyll a concentrations and primary production

80

Chlorophyll a (Pg/L) g

70

Grizzly Bay

60 50 40 30 20 10 0 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00

20000 16000

3

Freshwater Flow (m m/s)

18000 14000 12000 10000 8000 6000 4000 2000 0 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00

Figure 2. a) Phytoplankton biomass in Grizzly Bay prior to the invasion of P. amurensis in late 1986 and after, and b) hydrograph of same period. Data from state of California database at http://sarabande.water.ca.gov:8000/~bdtdb/

298 being two to three times higher in the shallow water than in the channels (Cloern et al. 1985). Field and modeling studies have shown that turbidity limits phytoplankton growth in the channels, and system-wide blooms cannot be maintained by phytoplankton production in the channel (Cloern 1996, Lucas et al. 2001a, in review). In addition, top down control is less likely to occur in the channel due to the reduced vertical mixing rates. In spite of these similarities in the embayments, the responses of these systems to P. amurensis grazing have been quite different. Mechanisms of Phytoplankton Control in North Bay The three freshest bays (HB, SUB, and GB) have historically had a small magnitude phytoplankton bloom that lasted through summer and fall of most years. The primary production for this system was controlled by a combination of residence time and turbidity. Cloern (1996) suggests that during periods of high freshwater flow, the residence time declines sufficiently in NB to eliminate the annual bloom (see Figure 2, years 1983 and 1986, Figure 4a). With normal freshwater flow, residence time increases

Figure 3. Integrated annual chlorophyll a concentrations in each embayment prior to (first box plot), and after (second box plot) the invasion of P. amurensis into the estuary. *** Integrated chlorophyll a values are significantly different (p”0.001) for the two time periods in an embayment.

299

Chlorophyll a ( Pg/L/bloom) g

a.

7000 6000 5000 4000

p pre-1987

3000 2000 post-1987

1000 0 0

2E+11

4E+11

6E+11

8E+11

Integrated Freshwater Inflow- m3/bloom

Chlorophyll a ( Pg/L/yr)

b.

4000 3500 3000 2500 2000 1500 1000 500 0 0

5E+11

1E+12

1.5E+12

2E+12

2.5E+12

3E+12

3

Integrated Freshwater Inflow- m /yr Figure 4. Relationship between integrated freshwater inflow and phytoplankton biomass in (a) NB (integrated over the bloom period) and (b) SB (integrated annually).

and turbidity declines in late spring, thereby allowing a slow growing bloom to develop. Cloern et al. (1983) also hypothesized that there was likely to be an optimum freshwater flow that would allow phytoplankton growing in the shallows to be tidally exchanged with the deeper water and still be retained in the embayment, thereby allowing the phytoplankton biomass to accumulate. It can be seen that there is now no significant relationship between freshwater flow and productivity since P. amurensis invaded the system (Figure 4a). P. amurensis grazing rates (Figure 5), peak in fall and are at a minimum in the

300 winter and early spring in a pattern that is opposite to that of chlorophyll a concentration. The minimum bivalve biomass in winter and early spring throughout the shallow reaches of NB, SPB, and SB is believed to be due to predation by birds that immigrate into SFB each fall, as part of the Pacific Flyway migration (Poulton et al. 2002). Although benthic grazing appears to be controlling phytoplankton biomass in NB, it is unlikely that the other mechanisms of residence time and light availability are irrelevant processes.

Mechanisms of Phytoplankton Control in South Bay SB phytoplankton blooms are short (usually 2-4 weeks) and have the highest peak chlorophyll a concentrations in the bay (60-80 µg/L). The mechanisms for bloom development in SB have been well studied (Cloern 1996, Thompson et al. in review) and modeled (Cloern 1982, Lucas et al. 1999a, 1999b, in review, May et al. 2003). As in other portions of the bay, light is limiting and blooms are mostly generated in the shallow water during periods of decreased tidal and wind mixing. Persistent stratification, which occurs only during periods with high freshwater flow, tends to increase the magnitude of the bloom (Figure 4b). Similar to what we see in NB, phytoplankton biomass increases during periods when bivalve grazing rates are at their lowest levels, each winter/spring (Figure 5b). Bivalve recruitment occurs in spring in SB and bivalves may grow sufficiently during the bloom, to limit the magnitude and duration of the bloom (see 1995, Figure 5).

A Summary of Differences – Why Do the Phytoplankton Bloom in South Bay? The factors that control turbidity in NB are, in order of importance, freshwater inflow (Schoellhamer 2002), tidal resuspension at fortnightly time scales, and tidal and wind wave resuspension at the diurnal time scale (Ruhl et al. 2001). The high turbidity during the high river flow period (January through April) limits the period for phytoplankton growth to summer and fall, which is the period when P. amurensis populations begin building and peaking in biomass. The SB system has more tidal energy (tidal range of 2m), and as defined by Monbet (1995) is fairly typical of tidally energetic estuaries in that tidal resuspension (at fortnightly and semi-diurnal scales) is the main contributor to turbidity. However, early season freshwater inflow can also bring highly turbid water into SB as can diurnal wind resuspension in summer periods (Schoellhamer 1996). This leaves a short period when the

301 phytoplankton are not light limited in SB and that, by chance, corresponds to the period when shallow-water bivalve suspension-feeders are at their annual minimum biomass.

12 2 9 6 1 3 0

0 1991

ƇChlorophyll a (P g/L)

b.

1992

1993

1994

1995

5

70 60

4

50 40

3

30

2

20 1

10 0 1991

Ƒ Benthic Grazing (m/d)

3

15

Ƒ Benthic Grazing (m/d)

ƇChlorophyll a (P g/L)

a.

0 1992

1993

1994

1995

Figure 5. Phytoplankton biomass (chlorophyll a concentration) and grazing rates of benthic bivalves in (a) NB and (b) SB.

CAN WE PREDICT WHEN A NEW SUSPENSION-FEEDER WILL EXERT TOP-DOWN CONTROL ON THE PHYTOPLANKTON IN A TURBID SYSTEM?

302 The large trophic changes that have resulted from Dreissena sp. and Corbicula sp. invasions in North American and European ecosystems, in addition to the invasion described here, highlight the necessity of understanding which systems are most at risk of large trophic changes from invasive suspension-feeders. My goal is to use what we have learned in these two systems to (1) determine if the CT/RT model developed by Smaal and Prins (1993) is sufficiently sensitive to answer this question and if not, (2) to develop a method to predict the response of phytoplankton in turbid, nutrient rich estuaries to an increase in suspension-feeding. The best method to answer this question is, of course, to develop a multi-dimensional hydrodynamic model that is connected to a biological process model. Given that few systems have such models or the data needed to validate the models, there is a need to develop simple tools that address some of our basic questions.

Figure 6. The clearance time (CT days), residence time (RT days), primary production turnover time (PPT days) relationship developed by Dame and Prins (1998) with new San Francisco Bay data added to the plot.

The ideal requirements for such a tool include that it be simple and intuitive, that it use clear, easily derived parameters that accurately portray the important ecological processes, and that the findings be consistent with our observations. The Smaal and Prins (1993) CT/RT model was described earlier (hereafter referred to as the CR model) and was developed to address questions about when commercially important bivalves had reached the

303 carrying capacity of a system, ie when bivalve consumption was in balance with a combination of system produced and imported primary producers. The major conclusion from that study was that in shallow, closed systems in which bivalve clearance time (CT, the time for suspension-feeders to filter the volume of water in the system) is less than the residence time for the volume of water in the system (RT), the suspension-feeders had the potential of regulating the biomass of primary producers. Dame and Prins (1998) added a measure of the turnover of primary producers (ie. primary production time, PPT, an ecosystem scale measure of the annual rate at which phytoplankton biomass is turned over which they found to be similar to cell doubling time) to the CR model, thereby extending the model to incorporate the rate of growth of the primary producers (hereafter referred to as the CRP model). The 2-D space on this plot is the CR plot and as shown on Figure 6, the drop lines on the parameter points can be used to evaluate these systems using their earlier model. By adding PPT, the relationship now has the advantage of incorporating factors that control phytoplankton growth rate such as nutrient and light limitation and the effect of “new” or imported phytoplankton. One difficulty with a 3-D plot is to know which relationship takes precedence. For example, although the Western Wadden Sea is in the CT/RT space for bivalves exceeding the carrying capacity of a system, the PPT is less than CT and RT and thus may represent a phytoplankton system capable of compensating for some of the consumption losses. To help our interpretations, we can look at this plot as a series of 2-D plots. For example if transport and mixing are not important then the bivalves should be at their maximum biomass if CT=PPT and below their system biomass if CT>PPT. If CT3100

0.8 16.9 10 1.7 .6 20

0.9 29 .3 14.9 19.7 155

12

2.1

11

20

0.6

14

2.4

1.6

2.5

0.6

12

1.2

1800

10

180

North Inlet (NI) Carlingford Lough (CF) Marennes-Oléron (M) Narragansett Bay (NB) Ria de Arosa (RA) North SFB Pre-P. amurensis (NFP) North SFB Post-P. amurensis (NF) South SFB (SF)

annual annual annual annual annual annual

North SFB Pre-P. amurensis (SNFP) North SFB Post-P. amurensis (SNF) South SFB (SSF) Sylt (S) North SFB Pre-P. amurensis (NFPG) North SFB Post-P. amurensis (NFG) South SFB (SFG)

annual

spatial average spatial average spatial average shallow

annual

shallow

12

1.2

3.8

10

0.4

annual growth bloom

shallow

14 0.5 16

1.3 0.7 2.1

1.7 4 >3100

1.7 1.0 20

1 4 155

16

2.1

1.1

20

0.1

7

2.4

2.2

5.0

0.4

2.6 1.2 6.7 10 1.2

3.7 5.8 1278 325 1800

3.1 1.8 7 2.5 10

1.2 3.2 183 130 180

Oosterschelde (O) Western Wadden Sea (W) Delaware Bay (DB) Chesapeake Bay (CB) North SFB Pre-P. amurensis (SNFPG) North SFB Post-P. amurensis (SNFG) ( South SFB (SSFG)

annual annual

bloom bloom

spatial average spatial average spatial average

growth growth growth growth bloom

shallow

40 10 97 22 16

bloom

shallow

16

1.2

7.8

10

0.8

bloom

shallow

7

1.3

22

2.5

8.8

307

Figure 7. Clearance time (CT days), primary production turnover time (PPT days), and tidal mixing ratio (depth/tidal range), for all systems shown in Dame and Prins (1998) and SFB. NB points are shown as solid circles

A plot of a subset of this same data (Figure 8) allows us to bisect the 2-D CT-PPT space to show the relative values of bivalve suspension-feeding rate and primary production rate for a variety of systems. The differences in this plot compared to the previous CRP plot reflect the different uses of the plots. Most of the points in this plot are near the line, where we expect grazing rate to balance primary production rate, or to the left of the line, where the bivalve clearance rate is predicted to be insufficient to control a phytoplankton bloom. Similar to the CRP plot, points to the right of the line represent systems where the bivalves are controlling phytoplankton accumulation (blooms) in the system. The third parameter Z/TR, a mixing parameter, is indicative of large vertical and horizontal mixing when the ratio is §1. Thus any point to the right of the line with a small mixing parameter is likely to be accurately portrayed as a system where benthic consumers control phytoplankton biomass. These types of systems are represented here by the NB system (SNFG and SNF) which is presently controlled by P. amurensis. The other system represented in this space, the Marennes-Oléron system is unique within those systems represented here as the benthic bivalves control the phytoplankton through grazing but are dependent on the resuspended microphytobenthos (Dame and Prins 1998) to maintain their biomass. Thus the mixing parameter in this system (Z/TR=1.7) represents mixing at a rate sufficient to deliver pelagic food to the bivalves and to resuspend the microalgae. Phytoplankton in systems within this CT/PPT sector, but with a mixing parameter near 2 or greater appear to be sufficiently separated from the

308 bottom that the phytoplankton can still accumulate; this is true of the SB system as shown in Figure 5b and represented in Figure 8 as SFG and SF.

Figure 8. Clearance time (CT days), primary production turnover time (PPT days), and tidal mixing ratio (depth/tidal range), for a subset of systems (CT