submerged macrophytes modify food web interactions and
TRANSCRIPT
Submerged macrophytes modify food web interactionsand stability of lake littoral ecosystems
Kirsi Vakkilainen
Department of Ecological and Environmental SciencesFaculty of BiosciencesUniversity of Helsinki
Finland
Academic dissertation in environmental ecology
To be presented, with the permission of the Faculty of Biosciences of the University ofHelsinki, for public criticism in the Auditorium of Lahti Science and Business Park,
Niemenkatu 73, Lahti, on December 9th, 2005, at 12 a.m.
Lahti 2005
Supervised by
Prof. Timo KairesaloDepartment of Ecological and Environmental SciencesUniversity of HelsinkiLahti, Finland
Reviewed by
Prof. Jouko SarvalaDepartment of BiologyUniversity of TurkuTurku, Finland
Prof. Mari WallsDepartment of BiologyUniversity of TurkuTurku, Finland
Opponent
Research professor Erik JeppesenDepartment of Freshwater EcologyNational Environmental Reseach InstituteSilkeborg, Denmark
ISBN 9521027509 (nid.)ISBN 9521027517 (PDF)ISSN 12391271
YliopistopainoHelsinki 2005
CONTENTSABSTRACT 1LIST OF ORIGINAL PAPERS 2THE AUTHOR’S CONTRIBUTION 2ABBREVIATIONS AND CONCEPTS USED IN THE THESIS AND THEIRDEFINITIONS 31. INTRODUCTION 4
1.1 Aquatic ecosystem functioning 4Role of trophic structure 4Aquatic ecosystem responses to nutrient enrichment 5
1.2 Littoral ecosystems 6Community structure in littoral habitats 6Relative roles of macrophytes and zooplankton in controlling phytoplankton
biomass 71.3 Ecosystem regime shifts 91.4 Changes in the functioning of shallow aquatic ecosystems with climate 91.5 Fish stock management, an efficient restoration measure in LakeVesijärvi, southern Finland – implications for fishmediated habitat coupling 10
2. OBJECTIVES OF THE PRESENT STUDY 133. MATERIAL AND METHODS 14
3.1 Study sites 143.2 Design of the experiments 163.3 Sampling procedure, sample and data analyses 17
4. RESULTS AND DISCUSSION 184.1 Fishzooplankton interactions within macrophyte beds 18
Alternative food sources for fish in the littoral habitat and consequences forfishzooplankton interaction 20
Diel vertical migration of littoral zooplankton in response to fish predation 21Major effects of fish relative to nutrients in structuring the zooplankton
community 224.2 Stability of littoral ecosystems and the control of phytoplankton biomasswithin macrophyte beds 22
Confounding effects of macrophytes on phytoplankton 22Littoral ecosystem resistance and resilience to phosphorus enrichment 24Littoral ecosystem resistance to phosphorus and nitrogen enrichment 25Role of plantassociated animals in controlling phytoplankton and periphyton 26Control of phytoplankton biomass by zooplankton grazers: geographical
differences 27Role of zooplankton grazing in explaining the variation between chlorophyll
a and total phosphorus 274.3 Field mesocosm experiments and their applicability for studying aquaticecosystem functioning 30
5. CONCLUSIONS AND FUTURE RESEARCH NEEDS 316. ACKNOWLEDGEMENTS 327. REFERENCES 33
ABSTRACT
Community structure and food web interactions determine ecosystem functioningand stability such as resistance to nutrient loading. Littoral habitats with submergedvegetation serve as major regulators of nutrient dynamics in lake ecosystems throughhabitat coupling. Deeper understanding of how trophic structure and biotic interactionswithin littoral communities are connected with water quality facilitates the designing ofrestoration measures for lakes suffering from eutrophy. The present study focuses on thestructure and functioning of lake littoral ecosystems, especially plankton communities.Special interest is focused on the effects of submerged macrophytes in modifying foodweb interactions with respect to changes in planktivorous fish stocks and productivity.A series of enclosure and mesocosm experiments of factorial design were run in thelittoral zone of Lake Vesijärvi, southern Finland, and also in five other lake littoral areasalong a climatic gradient across Europe.
The results emphasised the general importance of consumer control over resourcecontrol in determining zooplankton community structures in shallow lake systems.Roach (Rutilus rutilus (L.)) preferred feeding on cladocerans even as they declined inabundance, while also using lesspreferred, mainly nonanimal food sources asalternative food. Thus, littoral subsidies stabilise planktivorous and omnivorous fishpopulations, which may intensify rather than dampen the strength of the interactionsbetween fish and zooplankton. Plantassociated and small euplanktonic cladocerans, onthe other hand, benefited from macrophytic refuge and could control phytoplanktonwhen fish were not abundant (< 4 g fresh weight m2 or < 2.54 ind. m2).
Community composition, primarily the abundance of submerged macrophytes andcladoceran grazers, played the decisive role in determining the stability of littoralecosystems and resistance to nutrient enrichment, recorded as lower biomass ofphytoplankton and periphyton. A macrophyte, Elodea canadensis Michx, hampered thegrowth and toxin production of cyanobacteria and regulated phytoplanktonzooplanktoninteractions. The mechanisms behind the control of phytoplankton growth bymacrophytes were obviously complex and related to interactions and nutrient recyclingwithin the food web, including heterotrophic processes. Wide yeartoyear variation inlittoral community composition, determined largely by prevailing weather conditions,resulted in variable outcome of perturbations imposed by nutrient loading. Suchvariable conditions may cause oscillations in the habitat coupling between littoral andpelagial zones through, for instance, variable recruitment and foraging behaviour offish.
The phytoplankton biomass was positively related to the total phosphorusconcentration even in the presence of efficient grazers, reflecting the positive responseof adjacent trophic levels and the ultimate, quantitatively important role of productivitycompared with consumer regulation in determining the phytoplankton biomass.However, the abundance of large crustacean grazers explained reasonably well thevariance between productivity and algal biomass, even better than did the mere numberof trophic levels. This result may be explained by factors such as heterogeneity withintrophic levels and within the littoral habitat. The only consistent geographical patternalong the climatic gradient of Europe was the reduced role played by large crustaceangrazers (> 0.5 mm) in controlling phytoplankton biomass in the southernmost locationin Valencia, Spain, compared with the other sites. Although food web managementappears to be a useful measure in northern temperate locations, nutrient control may bemore important in southern lakes.
1
LIST OF ORIGINAL PAPERS
This thesis is based on the following papers, which in the text are referred to by theirRoman numerals:
I. Kornijów, R., Vakkilainen, K., Horppila, J., Luokkanen, E. & Kairesalo, T.2005: Impacts of a submerged plant (Elodea canadensis) on interactionsbetween roach (Rutilus rutilus) and its invertebrate prey communities in alake littoral zone. Freshwater Biology 50: 262276.
II. Vakkilainen, K., Horppila, J., Väisänen, A., Sivonen, K. & Kairesalo, T.Submerged macrophytes increase the resistance of lake littoral tocyanobacterial blooming and toxin production after phosphorus enrichment.Manuscript.
III. Hietala, J., Vakkilainen, K. & Kairesalo, T. 2004: Community resistance andchange to nutrient enrichment and fish manipulation in a vegetated lakelittoral. Freshwater Biology 49: 15251537.
IV. Vakkilainen, K., Kairesalo, T., Hietala, J., Balayla, D., Bécares, E., Van deBund, W., Van Donk, E., FernándezAláez, M., Gyllström, M., Hansson, L.A., Miracle, M.R., Moss, B., Romo, S., Rueda, J. & Stephen, D. 2004:Response of zooplankton to nutrient enrichment and fish in shallow lakes: apanEuropean mesocosm experiment. Freshwater Biology 49: 16191632.
Papers I, III and IV are reproduced by the kind permission of Blackwell Publishing Ltd.
THE AUTHOR’S CONTRIBUTION
I. KV, RK and EL set up the experiment and carried out field measurementsand sampling. EL and KV planned the statistical analyses. KV analysed thezooplankton and phytoplankton samples, performed the data analyses,interpreted the results and drew the figures. KV wrote the paper incooperation with TK and RK. RK and TK planned the experiment and TKsupervised the work.
II. TK, JHo and KV planned the experiment. KV wrote the paper, conducteddata analyses and interpreted the results. TK supervised the work.
III. JHi and KV took charge of setting up the experiments, field sampling andmeasurements. KV contributed to the laboratory measurements (excludingthe nutrient analyses), analysed zooplankton samples, wrote and interpretedthe zooplankton results, drew all the figures and contributed to the generaldiscussion. TK supervised the work.
IV. KV and JHi collected the data from the different laboratories and planned thedata analyses. KV analysed the data, interpreted the results and wrote thepaper. TK supervised the work.
In addition to the results of the original papers, the thesis also includes unpublishedadditional material analysed by the author.
2
ABBREVIATIONS AND CONCEPTS USED IN THE THESIS AND THEIRDEFINITIONS:
FW = fresh weight, synonym fresh mass used in papers III and IV
DW = dry weight, synonym dry mass used in papers III and IV
C = carbon, P = phosphorus, TP = total phosphorus, N = nitrogen
TSS = total suspended solids
chl a = chlorophyll a, algal photosynthetic pigment (Its concentration has been used as asurrogate of phytoplankton and periphyton biomass.)
euplanktonic = permanently planktonic species as opposed to plantassociated speciesliving primarily on plant surfaces (cf. Hutchinson 1967)
edible and inedible algae for zooplankton = algal forms having the greatest axial lengthdimensions (GALD) of < 50 and > 50 µm, respectively
DHM = diel horizontal migration, DVM = diel vertical migration
ANOVA = analysis of variance
3
1. INTRODUCTION
1.1 Aquatic ecosystem functioning
Role of trophic structure
Predicting how ecosystemfunctioning is determined by the trophicstructure is one of the major challengesmet in ecology (Jones & Lawton 1995,Polis & Winemiller 1996). In aquaticecosystems, the discovery of the roleplayed by predation in shapingzooplankton community structure withconsequences on the efficiency ofherbivory (Hrbácek et al. 1961, Brooks &Dodson 1965) inspired a number ofstudies on how the effects of highertrophic levels are reflected in the structureand biomass of lower trophic levels.Comparisons between lake ecosystemswith differing food web length ormanipulations of the top trophic levelsshowed evidence of the generally strongvertical structuring of aquatic food webs(e.g. Carpenter et al. 1985, Carpenter &Kitchell1993, Salonen et al. 1990,Hansson 1992, Sarnelle 1992, Rask et al.1996, Brett & Goldman 1997, Horppila etal. 1998, Jeppesen et al. 1998b, Järvinen& Salonen 1998). In general, the trophicstructure of lake food webs can bedefined as the partitioning of biomass intodistinct trophic levels and is basicallydetermined by available resources but isalso regulated by dynamic feedback fromhigher trophic levels (Carpenter et al.1985, Arditi & Ginzburg 1989, Power1992, Carpenter & Kitchell 1993, Polis1999). It is now well recognised thatdirect and indirect interactions withinlake communities and food chain lengthmay explain much of the variance inphytoplankton biomass that cannot beexplained by variation in ecosystemproductivity (Carpenter et al. 1985,Kerfoot & Sih 1987, Carpenter &Kitchell 1993, Sarvala et al. 1998).
The relative importance of resourceand predator control in shapingcommunity structure is viewed ascomplementary, not contradictory. Forinstance, the zooplankton community isexpected to respond to changes both inavailable food resources and in predationpressure, reflecting the “sandwiched” roleof zooplankton in aquatic food webs(Gliwicz 2002, Jeppesen et al. 2002a).However, the two forces operate ondifferent time scales. The effect ofpredation is immediate, while it requiresmore time to transform enhancedproductivity into new biomass, and thistime lag is dependent on the generationtime of organisms (Reynolds 1994,Gliwicz 2002). Therefore, the responsesof the zooplankton community tend to bemore vigorous to manipulations of fishthan nutrients. Planktivorous fish keep theabundance of (large) herbivorouszooplankton, the intermediate consumer,at low levels and thus benefit prey livingtwo trophic links lower, i.e.phytoplankton via reduced grazing andvia enhanced nutrient regeneration(Carpenter et al. 1985). This trophiccascade from higher trophic levels can betraced back to the traditional food webhypothesis (Hairston et al. 1960, Fretwell1977, Oksanen et al. 1981), whichpredicts that ecosystem productivityultimately determines food chain lengthand that biomasses along the food chainexhibit discontinuous, alternating changeswith productivity.
Even though the food chain modelswith their modifications (e.g. Scheffer1991) have successfully explained muchof the dynamics observed in aquaticecosystems, these patterns may notalways occur under natural conditions.Strong predator effects at the top of thefood web (fishzooplankton) are widelydocumented, while at lower trophic levels(zooplanktonphytoplankton) responses tofish manipulations are more variableacross the productivity gradient (DeMelo
4
et al. 1992, Leibold et al. 1997).Increased productivity, on the other hand,often results in increased abundance ofadjacent trophic levels and responsesunrelated to the number of trophic levelsin the food web (Arditi & Ginzburg 1989,Hansson 1992, Mazumder 1994, Brett &Goldman 1997, Leibold et al. 1997,Hansson et al. 1998b). Such variation wasattributed to factors such as omnivory(Diehl & Feiβel 2000), compensation orsubsidized prey (Leibold et al. 1997, Paceet al. 1998), efficiency of consumers atexploiting their prey (Arditi & Ginzburg1989, Power 1992, Abrams & Walters1996) and habitat heterogeneity includingrefuges (Timms & Moss 1984, Persson etal. 1996). Trophic level heterogeneity hasbeen addressed as one of the mainlimitations of simple food chain models,which are based on abstractions of thetrophic levels represented, in effect, by asingle keystone species (Osenberg &Mittelbach 1996, Leibold et al. 1997,Persson et al. 2001).
In aquatic ecosystems, largebodiedDaphnia, which are highly sensitive tofish predation, have generally been usedas the keystone grazer and theirabundance as a relevant indicator forpredicting the grazing impact ofzooplankton on phytoplankton (Hansson1992, Sarnelle 1992, Cyr & Curtis 1999,Persson et al. 2001, Jeppesen et al. 2003).Zooplankton may also affectphytoplankton biomass via alteredrecycling of nutrients. Nutrientregeneration of planktonic animals isimportant especially in pelagicecosystems (Järvinen & Salonen 1998,Hudson et al. 1999, Tarvainen et al.2002, Vanni 2002). For example,Daphnia have a remarkably constant andlow carbontophosphorus (C:P) ratiocompared with the higher variation in theseston C:P ratio and may therefore act assinks for P by incorporating substantialamounts of P per body mass (Andersen &Hessen 1991, Salonen et al. 1994).However, there is so far little evidence for
the role of zooplanktondrivenstoichiometry in changing phytoplanktonbiomass (Cyr & Curtis 1999). Therelative importance of grazing andnutrient regeneration may often bedifficult to interpret and is dependent onconditions such as time scale, lake depthand trophic state (Benndorf et al. 2002).
Aquatic ecosystem responses tonutrient enrichment
An increase in ecosystemproductivity removes the nutrientconstraints that tend to reduce thevariation between predatorpreypopulation interactions (DeAngelis et al.1989). Nutrient enrichment may also leadto a succession toward invulnerable orpredatorresistant prey such asfilamentous or even toxic cyanobacteria,thus uncoupling the dynamics ofzooplankton from the dynamics ofphytoplankton (McQueen et al. 1986,Larsson & Dodson 1993, Abrams &Walters 1996). In general, however,efficient grazers reduce the sensitivity ofphytoplankton biomass to nutrientenrichment (DeAngelis et al. 1989,Cottingham et al. 2004). Much ecologicalinterest is focused on different stabilityproperties of ecosystems with respect totheir responses to perturbations,especially nutrient enrichment (reviewedin DeAngelis et al. 1989). Two centralfeatures of ecosystem stability includeresistance, the ability of an ecosystem toresist perturbations, and resilience, theability and the rate at which an ecosystemreturns toward a steadystate equilibriumfollowing perturbation (Carpenter et al.1992, Wetzel 2001). Understanding ofecosystem responses to nutrient loadingalso has major implications for ecosystemmanagement and restoration.Eutrophication due to nonpoint sourceloading of nutrients, especially P,continues to be one of the mostwidespread water quality problems and is
5
often difficult to manage and regulate(Wetzel 2001, Räike et al. 2003).
Community structure plays acentral role with respect to ecosystemresponses to nutrient inputs (Carpenter etal. 1992, Jeppesen et al. 1998c). Inaddition, habitat heterogeneity togetherwith flexible, adaptive behaviour oforganisms may result in confoundingeffects that may interact withproductivity, thus affecting communityand ecosystem dynamics (Persson et al.1996). Whole lakescale experimentationhas revealed uncertainties in predictingthe functioning of lake ecosystems, thussuggesting drawbacks related to aquaticfood web research, which has beencarried out mainly in pelagic habitats.Planktivorous and omnivorous fishpopulations are stabilized by alternativefood resources (subsidies) in the littoralecosystem and may become uncoupledfrom the dynamics of zooplankton prey(Schindler & Scheuerell 2002). Switchingof prey preference may reduce thepredation pressure on zooplankton(Perrow et al. 1999) and thereby dampentheir population fluctuations. However,anthropogenic disturbances, includingeutrophication, alter habitat connectionsand thus the flows of nutrients and energyin lake ecosystems (Schindler &Scheuerell 2002).
1.2 Littoral ecosystems
Community structure in littoralhabitats
Littoral habitats with submergedmacrophytes and associated periphyticcommunities serve as important sinks fornutrients that enter the lake and as majorregulators of nutrient dynamics in lakeecosystems through habitat coupling(Sarvala et al. 1982, Wetzel 1990, 2001,Wetzel & Søndergaard 1998, Schindler &Scheuerell 2002). Shallow lake areaslacking thermal stratification comprise
the majority of the earth’s freshwaterareas worldwide and therefore, in general,littoral areas overwhelmingly dominatethe pelagic (Wetzel 1990). They aresensitive to eutrophication and resilient tomanagement due to internal loading viaeffective exchange of nutrients betweenwater and sediment. Resuspension maybe an important process even in relativelydeep lakes (KoskiVähälä et al. 2000) andsometimes more important than nutrientrecycling by fish (Tarvainen et al. 2002).Abundant submerged vegetation reduceswater movement and sedimentresuspension and thus P regeneration(Horppila & Nurminen 2003). Nutrientregulation by macrophytes can be of theutmost importance in suppressingphytoplankton growth and cyanobacterialblooms during the growing season(reviewed in Søndergaard & Moss 1998).The existence of allelopathy also seemspossible but has been a subject ofongoing debate, since its significanceunder field conditions is still too poorlyevidenced (reviewed in Gross 2003).Furthermore, given the presence ofmacrophytes, changes in phytoplanktonbiomass throughout the growing seasoncannot be explained by allelopathy(Balayla & Moss 2004).
In addition to the proximity ofbottom sediment, submerged vegetationcreates a structurally rich habitat. Itsustains a high diversity of different lifeforms from those attached to surfaces tothose swimming or floating freely in thewater column and to those exhibiting bothof these ways of life. Thus, the structureas well as trophic interactions ofmacrophyteassociated littoral food websare inherently complex (Kairesalo 1980,Scheffer et al. 1993, Kornijow &Kairesalo 1994, Jeppesen et al. 1998c,Scheffer 1999) (Fig. 1). High habitatstructural complexity increases the spatialseparation and availability of refuges thuslimiting populations, influencingconsumerresource dynamics andweakening the strength of trophic
6
cascades through foraging efficiency ofpredators (Persson et al. 1996, Polis et al.2000, Shurin et al. 2002). For instance,the foraging ability of roach Rutilusrutilus (L.) decreases with habitatcomplexity and macrophyte coverage(Persson 1987, Diehl 1993, Moss et al.1998). One of the fascinating roles ofsubmerged macrophytes is their potentialfor providing refuge for zooplanktonagainst fish predation, thus hampering thetrophic cascade (Timms & Moss 1984,Carpenter et al. 1985, Lauridsen et al.1996, Jeppesen et al. 1998a, Kairesalo etal. 1998, 2000, Scheffer 1999).
Relative roles of macrophytes andzooplankton in controllingphytoplankton biomass
In plant beds, both euplanktonicand littoral, plantassociated cladoceranspecies may occur in dense swarms andplay an important role in controllingphytoplankton biomass (Kairesalo 1980,Walls et al. 1990, Jeppesen et al. 1998c,1999, Nurminen et al. 2001, Balayla &Moss 2004). Zooplankton grazing,especially that of Daphnia, is of centralimportance in enhancing water clarity andthe availability of light for macrophytes.On the other hand, several studiesreported negative effects of submergedmacrophytes on zooplankton, especially
Figure 1. Schematic overview of interactions between different constituents of a littoralfood web. The arrows show the processes of interest in this thesis. Food web links areillustrated by thin black arrows and spatial refuge by broader grey arrows, while resourcecompetition and allelopathic interactions are illustrated by open arrows. (Modified fromVan Donk & Van de Bund 2002.)
7
Daphnia, due to low food supply, highdaytime pH, lowoxygen conditions,invertebrate predation or plant chemicals(Hasler & Jones 1949, Pennak 1973,Kairesalo 1980, Dorgelo & Heykoop1985, Jeppesen et al. 1998a, Blindow etal. 2000, Burks et al. 2000, 2002).Littoral phytoplankton may be low inbiomass (Schriver et al. 1995) or inquality (Kairesalo 1980, Smiley &Tessier 1998) compared with pelagicphytoplankton.
In contrast, other studies haverevealed that small mobile flagellateswith high growth rates appear successfulin the presence of macrophytes and serveas highquality food for zooplankton(Søndergaard & Moss 1998).Zooplankton, including Daphnia, mayalso use alternative food sources in thelittoral zone by browsing on plant,sediment and other surfaces (Horton et al.1979, Jeppesen et al. 2002b). In addition,macrophyte beds may be rich in bacteriathat can be utilized as food by manycladoceran species (Søndergaard et al.1998). Factors such as species, densityand the growth phase of macrophytes alsoaffect conditions in the vegetation and itsquality as habitats for zooplankton. Bymigrating horizontally or verticallybetween openwater habitats or sedimentsurface and macrophyte beds, cladoceransmay avoid constant susceptibility to thepossible adverse effects of macrophyteswhile still benefiting from the refuge(Timms & Moss 1984, Walls et al. 1990,Lauridsen et al. 1996, Smiley & Tessier1998, Burks et al. 2002, Jeppesen et al.2002b).
Macrophytes may also modify fishzooplankton interactions throughsustaining a habitat with multiplealternative food sources for fish (Persson1987, Diehl 1993, Moss et al. 1998,Perrow et al. 1999, Schindler &Scheuerell 2002). Generalist feedingbehaviour, i.e. switching of preypreference, may reduce the predationpressure on zooplankton (Perrow et al.
1999) and thereby dampen theirpopulation fluctuations. On the otherhand, it was suggested that suchbehavioural flexibility may have only ashortterm stabilizing effect, dependingon the predator’s behaviour and foodpreferences (Murdoch & Bence 1987).For instance, in the littoral zone of LakeVesijärvi, southern Finland, thepercentage share of zooplankton in thediets of roach decreases and theproportion of benthos and macrophytesincreases with the low availability ofplanktonic food (Horppila & Kairesalo1992) and with the increasing size of fish(Horppila 1994). However, theconsequences of such plasticity in foodpreference for zooplanktonphytoplankton interactions and forfunctioning of the littoral ecosystem havebeen less quantified (reviewed inJeppesen et al. 1998c).
The spatial heterogeneity producedby submerged vegetation, coupled withthe movement of organisms betweenopenwater, vegetated and benthicenvironments may provide importantcompensatory mechanisms. They work asa buffering mechanism in shallow lakesand littoral ecosystems, where theoutcome of perturbation, such as nutrientloading, results from variable responsesof organisms from different habitats thatmay partially or completely compensatefor, or strengthen, each other’s effect(Scheffer et al. 1993, Jeppesen et al.1998c, Scheffer 1999). Thus, it may bedifficult to estimate the relativeimportance of macrophytes and fishpredation in structuring the littoralinvertebrate community (Jeppesen et al.1998c). The structuring role ofsubmerged macrophytes in controllingphytoplankton growth is generallyacknowledged, although the mechanismsbehind this phenomenon are still arguableand partly contradictory (cf. Jeppesen etal. 1998c, 1999). The debate stems fromthe ultimate complexity of macrophyterelated mechanisms and differences in
8
their relative role, which may vary withlake morphometry, plant communitycomposition, trophic state of the lake aswell as climate (Jeppesen et al. 1999,Scheffer et al. 2001).
1.3 Ecosystem regime shifts
Natural populations alwaysfluctuate, even when factors such asseasonality or climatic variation aredisregarded. Multiple variables changingat different rates with different directionsof feedbacks were attributed tocatastrophic regime shifts in ecosystems(Scheffer & Carpenter 2003). These shiftsare variable and are dependent on severalfactors such as lake morphometry,temperature and the predominance ofmacrophytes (GenkaiKato & Carpenter2005). Shallow lakes provide among thebestdocumented regime shifts, so calledalternative stable states, with either aturbid, phytoplanktondominated state ora clearwater, plantdominated state,depending on their community structure,through complex feedback mechanisms(Irvine et al. 1989, Scheffer et al. 1993,cf. Chapter 1.2). In deep stratified lakes,alternative stable states were related to Precycling by fish and/or from anoxicsediments (Carpenter et al. 1992,Horppila et al. 1998, GenkaiKato &Carpenter 2005). On the other hand, theexistence of real, sustained alternativestable states in lake ecosystems has beendebated (Sarvala et al. 2000, Van deBund & Van Donk 2002).
Macrophytedominated ecosystemsmay not be affected by ongoingeutrophication until a threshold is reachedat which a large switch, or regime shift,occurs leading to a takeover byphytoplankton (Scheffer et al. 1993,Scheffer & Carpenter 2003, Folke et al.2004). Algal blooms hamper macrophyticgrowth through shading and gainadvantage over nutrient regeneration viasediment resuspension. Trophic cascades
further increase the vulnerability of theecosystem to nutrient input and aretypically manifested by the loss of toppredators, i.e. piscivorous fish, leading toa resilient system with abundantplanktivores, smallbodied zooplanktonand abundant phytoplankton (reviewed inFolke et al. 2004). Mere reduction inexternal nutrient loading may not besufficient to lead to the recovery ofmacrophytes and revert a turbid lake to aclearwater state but major perturbations,such as fish manipulation, are needed todecrease internal nutrient loading(Scheffer et al. 1993, Moss et al. 1996,Horppila et al. 1998, Mehner et al. 2002).
Such ecosystem resilience has beena matter of extensive interest, becausethese alternative stable states can also beused as a conceptual basis forsuccessfully managing shallow lakeecosystems (Moss et al. 1996, Mehner etal. 2002). Indeed, restoration ofeutrophicated, turbid lakes has been mostsuccessful in shallow lakes, wheremacrophytes can colonise large bottomareas, thus increasing the resistance tonutrient loading (Mehner et al. 2002).However, techniques for restoration havebeen developed largely in northerntemperate locations, while it is less clearhow efficiently measures such asmanipulation of the fish community mayoperate at lower latitudes (Stephen et al.2004).
1.4 Changes in the functioning ofshallow aquatic ecosystems withclimate
All biochemical and biologicalprocesses such as population cycles,biomass turnover times and trophicinteractions have rates of operation thatare temperaturedependent (McCauley &Murdoch 1987, Lehman 1988, Petchey etal. 1999). Such factors are likely to affectcommunities, especially in shallowaquatic ecosystems (cf. Hargeby et al.
9
2004). Increasing temperature mayfacilitate the recolonisation ofmacrophytes in shallow lakes, dependingon the efficiency of zooplankton atproducing clearwater phases (Scheffer etal. 2001). It may also induce lakes to bemore susceptible to nutrient loading(GenkaiKato & Carpenter 2005).
Production of zooplanktonincreases with temperature, whilebiomass accumulation is more dependenton resource availability (Shuter & Ing1997). Similarly, enhanced primaryproductivity at higher temperatures (cf.Petchey et al. 1999) becomes morecontrolled by nutrient turnover rate andavailability. Accordingly, with risingtemperature the inherently faster growthrate of algae compared with that ofzooplankton may lead to reduced controlof algal biomass by zooplankton. Inaddition, the threshold food level ofzooplankton increases with watertemperature and, for metabolic reasons,with increasing animal size (Lehman1988). The available evidence suggeststhat the size structure of zooplanktonassemblages shifts from largebodied(efficient) grazers toward smaller (lessefficient) grazers (Moore et al. 1996).Fish predation may further suppress largezooplankton since the activity and capturesuccess of several planktivorous fish,such as roach, increase with temperature(Persson 1986). With increasingtemperature, effects at the resource basemay play a major role in the functioningof food webs, while the cascading effectsof fish via zooplankton grazers could playa minor role in controlling algal biomass.Thus, in southern locations manipulationof the fish community may not be asappropriate a technique for lakerestoration as in northern temperatelocations.
1.5 Fish stock management, anefficient restoration measure inLake Vesijärvi, southern Finland –implications for fishmediatedhabitat coupling
In Lake Vesijärvi, culturaleutrophication due to industrial anddomestic waste water began in the early20th century and was manifested bymassive cyanobacterial blooms thatpersisted even after sewage diversion(Keto & Sammalkorpi 1988). In the early1990s, the cyanobacterial bloomssuddenly disappeared and water claritymarkedly increased following the drasticdecrease of external loading and a fiveyear mass removal of fish, mainly roachand smelt Osmerus eperlanus (L.) from172 kg ha1 to less than 30 kg ha1
(Horppila et al. 1998, Kairesalo et al.1999). A new increase in planktivorousfish in the pelagic zone was prevented bycontinued management fishing andstocking of predatory fish, especiallypikeperch Sander lucioperca (L.),(Kairesalo et al. 1999, Ruuhijärvi et al.2005).
The decline in the ecosystemproductivity, including reduced internalloading and transport of P from thelittoral zone by roach rather than reducedplanktivory and enhanced herbivory wereconsidered the decisive factors in thedecline of cyanobacteria and in shiftingthe lake ecosystem into the clearwaterstate (Horppila et al. 1998, Kairesalo etal. 1999). The appearance ofcyanobacteria in highly productive lakesoften cannot be explained based on asimple herbivorephytoplanktoninteraction (reviewed in Persson et al.1996). Cyanobacteria tend to inhibitzooplankton feeding due to their colonialmorphology through mechanicalinterference and/or direct toxicity, withconsequences for the growth ofzooplankton, especially Daphnia(Lampert 1987, Hietala et al. 1995). Afterthe disappearance of cyanobacteria in
10
Lake Vesijärvi, the phytoplanktonbiomass was dominated by more ediblealgal forms such as cryptophytes anddiatoms (Keto & Tallberg 2000). Lowdeviation in chlorophyll a (chl a), despitea relatively wide variation in Pconcentrations, suggest enhanced controlof phytoplankton biomass by grazingcompared with the situation before thelake shifted to its present state (Fig. 2)(Kairesalo & Vakkilainen 2004). Aninverse relationship also prevailedbetween chl a concentration andcladoceran body length (Fig. 2) but not
biomass (Vakkilainen et al. unpublisheddata). Cladoceran body length is animportant predictor of phytoplanktonbiomass and water clarity (Carpenter etal. 1998, Stemberger & Miller 2003, butsee Sarvala et al. 1998) as well as asurrogate of sizeselective predation andplanktivorous fish abundance (Soranno etal. 1993). In Lake Vesijärvi, the bodylength of the cladoceran communityremained remarkably stable during the1990s, reflecting negligible changes inplanktivorous fish stocks during thatperiod. In the early 21st century, the bodylength increased (Fig. 3) with changes inthe relative abundances of species(Vakkilainen & Kairesalo 2005). Thiscoincided with the intensified fisherymanagement initiated in 2000 and withthe collapse of the smelt stock in 20022003 (Ruuhijärvi et al. 2005). Suchchanges in the food web structure wereattributed to relatively high transparencycoinciding with a deficiency inhypolimnetic oxygen even at 710mdepths, i.e. at euphotic zone, in latesummers 2002 and 2003 (Fig. 4). Thedecreased hypolimnetic water volumeavailable for zooplankton to hide fromplanktivorous fish may partly explain theconcomitant drastic reduction in
0
4
8
12
16
0,4 0,45 0,5 0,55
cladoceran length (mm)
chla
(µg
L1)
19912004
0
10
20
30
0 10 20 30 40 50 60
TP (µg L1)
chla
(µg
L1)
1983199219932004
Figure 2. Relationship between epilimnetic chlorophyll a (chl a) and (a) total phosphorus(TP) concentrations and (b) cladoceran mean body length (MaySeptember means + 1 SE)in the Enonselkä Basin of Lake Vesijärvi. Note the different scales of chl a concentrations.
11
0,4
0,5
0,6
1990
1992
1994
1996
1998
2000
2002
2004
year
leng
th (m
m)
epilimnionhypolimnion
Figure 3. Densityweighed body length(growing season means + 1 SE) ofcladocerans in the epilimnion (010 m) andhypolimnion (1029 m) of the EnonselkäBasin, Lake Vesijärvi in 19992004.
2004
0
5
10
15
20
25
30
1.5.
1.6.
2.7.
2.8.
2.9.
3.10
.
dept
h,m < 1 mg O2 l1
1 % of surfaceirradiance
2003
0
5
10
15
20
25
30
1.5.
1.6.
2.7.
2.8.
2.9.
3.10
.
dept
h,m
2002
0
5
10
15
20
25
30
1.5.
1.6.
2.7.
2.8.
2.9.
3.10
.
dept
h (m
)
Figure 4. Compensation depth (1 % of surface irradiance, measured with a LiCor LI1400 sensor) and the depth of oxygen deficient hypolimnetic water (< 1 mg O2 l1) of the
Enonselkä Basin, lake Vesijärvi during the growing season in 2002, 2003 and 2004.
cladoceran biomass from ca. 190 to 80 µgC l1 within two weeks in both 2002 and2003 (Vakkilainen et al., unpublisheddata). With high increasing transparencythe risk of predation may increase despiteof potentially low share of planktivorousfish, in accordance with conclusions byJeppesen et al. (2003). These phenomenain Lake Vesijärvi were accompanied byincidental late summer blooms ofcyanobacteria in 2002 and 2003. Duringthe rainy and windy summer 2004,oxygen conditions were improved (Fig. 4)and no algal blooms were recorded.
Before the restoration of LakeVesijärvi, the maximum depth ofsubmerged vegetation was 2 m butthereafter macrophytes colonised thebottom down to 4m depth (Venetvaara &Lammi 1995). In the Enonselkä Basin,the potential area for the growth ofsubmerged macrophytes increased fromca. 4.8 km2 to 8.5 km2. Such structuralchanges within the littoral areas alsosuggested altered regulation of nutrientsand food web dynamics. Both roach andperch Perca fluviatilis (L.) effectivelyutilise different habitats and diverse foodresources in Lake Vesijärvi (Horppila etal. 2000) and thus littoral and benthic
resources are an important subsidy forfish populations (Horppila et al. 1998,Kairesalo et al. 1999, cf. Schindler &Scheuerell 2002, Vadeboncoeur et al.2002). Differential use of habitats by fishmay be an important factor affectingwater quality (Holopainen et al. 1992,Horppila et al. 1998). Wide variation inthe recruitment of roach in Lake Vesijärvisuggested the presence of variableconditions in the littoral zone (Horppila etal. 1998). Similar observations were donein a number of other lakes, suggesting themajor importance of littoral and benthicfactors such as submerged macrophytesand benthic feeding fish in affecting theoutcome of food web restoration ofeutrophic lakes (Hansson et al. 1998a).Considerations of lake restoration throughfood web management, which formerlyfocused primarily on pelagic food chains,i.e. phytoplanktonzooplanktonfishinteractions (Carpenter et al. 1985,Carpenter & Kitchell 1993, Reynolds1994, Moss et al. 1996), have now goneforward in attempts to integrate wholelake ecosystem processes (Hansson et al.1998a, Schindler & Scheuerell 2002,Vadeboncoeur et al. 2002).
12
2. OBJECTIVES OF THEPRESENT STUDY
Alterations in planktonic food websfollowing fish manipulation trigger other,mainly benthic and littoral, processes inlakes (Hansson et al. 1998a). However,compared with extensive studies in thepelagic zone of lakes, trophic interactionswithin the littoral zone are less wellexplored (cf. Jeppesen et al. 1998c).From a practical point of view, deeperunderstanding of how the trophicstructure and biotic interactions withinthe littoral community are connected withwater quality facilitates the designing ofproper restoration measures. Moreover,they may play a crucial role in thestability of the improved state afterrestoration. The zooplanktonphytoplankton link is affected by severalconfounding factors such as variation innutrients, climate, algal edibility andmacrophytic vegetation, which have beenthe subjects addressed in controlledexperiments to identify underlyingmechanisms (e.g. DeMelo et al. 1992,Jeppesen et al. 1998c).
The present study focuses on thestructure and functioning of the littoralcommunity. Special interest is focused onthe effects of submerged macrophytes inmodifying food web interactions withrespect to changes in planktivorous fishstocks and productivity (Fig. 1).Generalist feeding behaviour andswitching of prey preference by fish aswell as macrophytic refuges forzooplankton against fish predation areexpected to have major implications forthe dynamics of littoral predatorpreyinteractions. The submerged vegetationwith its associated biota can be assumedto increase the resistance of the littoralecosystem to nutrient loading, whilecontrasting effects are expected with theincreasing abundance of planktivorousfish. Given the major role of nutrientavailability in the functioning of foodwebs at low latitudes, the cascading
effects of fish via zooplankton grazers areexpected to increase with increasinglatitude.
The main questions addressed in thisthesis are:
1. Do submerged macrophytes modifythe structure and functioning of thelittoral community through providingzooplankton with a refuge against fishpredation and a habitat rich inalternative food sources foromnivorous fish, thus changing thecascading effects of fish viazooplankton grazers?
2. Do submerged macrophytes increaselake littoral resistance and resilienceto nutrient enrichment?
3. How much is the resistance of thelittoral ecosystem against nutrientenrichment modified by fish,especially through cascading impactsvia zooplankton, and does the relativeimportance of resource availabilityand predation change along withdifferent climatic conditions?
Mesocosms are laboratory or fieldenclosures, i.e. enclosed subsystems thatcan be controlled and replicated.Mesocosm experiments are a powerfulexperimental tool for studying ecologicalmechanisms such as planktonic food webinteractions and chemical effects up to afew months in duration (Schindler 1990,Carpenter & Kitchell 1993). This was theapproach used in this thesis. Thus, thefunctioning of the littoral ecosystem wasstudied by conducting series of controlledfield mesocosm experiments of factorialdesign to answer the above questions.
13
3. MATERIAL AND METHODS
3.1 Study sites
The field experiments wereperformed at 1.01.7meter water depth inshallow lakes or lake littoral zones. Themain study site (IIV) was situated in theshallow Kilpiäistenpohja Bay (waterdepth < 3 m), which lies adjacent to theEnonselkä Basin of Lake Vesijärvi,southern Finland (Table 1, Fig. 5).
Kilpiäistenpohja Bay is characterised byabundant submerged vegetation, mainlyElodea canadensis Michx., Myriophyllumspp. and Ceratophyllum demersum L.,covering large areas of the bottom of theopen water area together with floatingleaved macrophytes, e.g. Nuphar lutea L.(Fig. 5). Freely floating Lemna trisulca L.is a characteristic species in the bay andforms thick flocks through the entirewater column. The emergent vegetation isdominated by Phragmites australis (Cav.)Trin. ex Steud. Roach and perch are the
Figure 5. The bathymetric map of the Enonselkä Basin of Lake Vesijärvi showing thelocation of Kilpiäistenpohja Bay as well as the sampling point (denoted by an asterisk),where longterm water quality and plankton monitoring have been conducted (Chapter1.5). The striped area denotes the urban, mainly built areas of the City of Lahti. Themagnification of the Kilpiäistenpohja Bay (right) shows the location of the main studyarea. It also illustrates the different vegetation zones identified from an aerial photo(Copyright FMKartta Oy) using optical resolution parameters. The zones are interpretedas follows: black = emergent vegetation, dark grey = floatingleaved vegetation, light grey= submerged vegetation, white = no identified vegetation.
14
most abundant fish species in the littoralarea (Kairesalo et al. 1999, Horppila et al.2000). The biomass of northern pike Esoxlucius (L.) decreased with eutrophication(Keto & Sammalkorpi 1988) but hassince increased due to stocking of pikefingerlings. Since 1997, yearly pikecatches in Lake Vesijärvi have averaged1.5 kg ha1 (Ruuhijärvi et al. 2005).
The other study sites (IV) weresituated across Europe in Sweden,England, The Netherlands, northern Spain(Leon) and southern Spain (Valencia)(Fig. 6) and were chosen to representshallow mesotrophichypertrophic lakesin which submerged vegetation had beenabundant during the years immediatelypreceding the experiments (Table 1).
Table 1. General characteristics of the study lakes.
Lake Location pH TP(µg L1)
Dominant submerged macrophyte taxa
Vesijärvi Finland61o02’ N, 25o39’ E
7.57.9
3060 Lemna trisulca L., Elodea canadensisMichx., Myriophyllum spp.,Ceratophyllum demersum L.
Krankesjön Sweden55o42’ N, 28o19’ E
8.2 25 Myriophyllum spp., Chara spp.
Little Mere England53o20’ N, 2o24’ E
7.3 184 Potamogeton berchtoldii Fieber, Elodeacanadensis, Ceratophyllum demersum
Naardemeer the Netherlands52o30’ N, 5o01’ E
8.59.6
30 Chara spp.
Sentiz Leon, northern Spain42o33’ N, 5o22’ E
8.49.0
100 Myriophyllum alterniflorum L.
Xeresa Valencia, southern Spain39o6’ N, 0o20’ E
8.78.9
1722 Chara spp.
Figure 6. Map showing the locations of the fieldexperiments along the European climatic gradient.
15
3.2 Design of the experiments
All the experiments had fullyfactorial design with two factors andaimed at studying the littoral communityresponses to a submerged macrophyteElodea canadensis (I, II), planktivorousfish (I, III, IV), weekly P and nitrogen(N) enrichment (III, IV) or P enrichmentsconducted either once at the beginning ofthe experiment (pulse enrichment) orweekly (press enrichment) (II) (Table 2).The experiments were conducted in 1994and 1996 in rectangular enclosures of 2 x2.5 m (I, II) or in 1998 and 1999 in
cylindrical mesocosms of 1m diameter(III, IV) (Table 2). In 1998 and 1999, theexperiments were conductedconcomitantly in the different study sitesalong the European climatic gradient(Fig. 6). Thus, twelve parallelexperiments (collectively theInternational Mesocosm Experiment,IME) were carried out in order to studylargescale betweenyear variation in theeffects of planktivorous fish and nutrientenrichment on littoral communities(Stephen et al. 2004). Inevitably, the fishspecies differed between locations but,however, taxon is considered less
Table 2. Summary of the experimental designs. Hereafter, in the text the treatments are referredto as denoted in this table.
Time ofexperiment andsize ofmesocosms
Treatment 1(levels in brackets)
Treatment 2(levels in brackets)
No. ofreplicates
Paper
129 August,1994(1) (4 weeks)area 5 m2
vol. 7.5 m3
45yearold (16 cm) roach:nofish = 0 g FW m2
highfish = 69 g FW m2 (1.9ind. m2)
Elodea canadensis:noElodea = 0 g DW m2
sparseElodea = 17.5 g DW m2,denseElodea = 52.5 g DW m2
3 I
25 June – 31 July,1996 (5 weeks)area 5 m2
vol. 7.5 m3
Phosphorus:control = 0 mg P l1
pulse = one initial Penrichment of 0.04 mg P l1
press = five weekly Penrichments of 0.04 mg P l1
Elodea canadensis:noElodea = 0 g DW m2
Elodea = 105 g DW m2
3 II
8 June – 13 July,1998 (5 weeks;III) or 8 June – 20July, 1998 (6weeks; IV)area 0.8 m2
vol. 0.71.0 m3
Weekly nutrient enrichments:(0 mg P + 0 mg N l1,0.1 mg P + 1 mg N l1,0.5 mg P + 5 mg N l1,1.0 mg P +10 mg N l1)
Locally appropriateplanktivorous fish species of 510cm length(2):fishfree = 0 g FW m2
lowfish = 4 g FW m2 (2.54ind. m2)highfish = 20 g FW m2 (720ind. m2)
3 III, IV
29 June – 10August, 1999(7 weeks; III) or29 June – 4August, 1999 (6weeks; IV)area 0.8 m2
vol. 0.71.0 m3
Weekly nutrient enrichments: (0 mg P + 0 mg N l1,0.03 mg P + 0.3 mg N l1,0.06 mg P + 0.6 mg N l1,0.09 mg P + 0.9 mg N l1,0.15 mg P + 1.5 mg N l1,0.30 mg P + 3.0 mg N l1)
Locally appropriateplanktivorous fish species of 510 cm length(3):fishfree = 0 g FW m2
lowfish = 4 g FW m2 (2.54ind. m2)highfish = 20 g FW m2 (720ind. m2)
2 III, IV
(1) The enclosures were set up two weeks before the fish manipulation, on 14 July, to allow themto recover from disturbance.(2, 3) In the experiment conducted in Lake Vesijärvi, 69cm juvenile roach(2) and 810cm juvenileperch(3) were used.
16
determinative than biomass or size withrespect to zooplanktivory (Williams &Moss 2003). Compromises about themost appropriate nutrient enrichmentlevels across locations were less easy tomake and drawbacks could not beavoided. During the third week of theexperiment in 1998 in Finland, high fishmortality occurred at the two highestnutrient levels, probably due to the highpH (> 9) of the water. Therefore, theresults of these nutrient treatments wereexcluded from the data analyses and thenutrient range was narrowed in 1999(Table 2). This and other accidentalevents, such as the breakup of theSwedish experiment in windy summer1998, resulted in the exclusion of somedata, described in more detail in paper IV.
The mesocosms, made ofwatertight, clear polyethylene (PE)plastic, were open to the atmosphere andthe low ends were sealed into the lakebottom sediment so that the water was incontact with the sediment. Theexperimental area was covered with a netto prevent the entry of birds. The fish andmacrophytes used in the experimentswere collected from a nearby lake area.
3.3 Sampling procedure, sampleand data analyses
Samples for water chemistry, chl a,phytoplankton, zooplankton, microcystins(only in 1996; II) and total suspendedsolids (TSS; only in 1998) were collectedweekly at 911 a.m. with a transparentPerspex core (length 1.5 m, innerdiameter 4.04.5 cm) from the surface tothe bottom of each enclosure. Specialcare was taken not to disturb thesediment. A total of 510 subsampleswere taken from different places of themesocosms, pooled and mixed. In asimilar way, additional samples weretaken weekly also from the lake outsidethe enclosures in 1994, 1998 and 1999.
The first samples were taken justbefore the introduction of fish and/or firstnutrient addition. The shortterm impactof Elodea on roachzooplanktoninteraction was studied in 1994 by takingan additional set of samples three daysafter the introduction of roach (I). On 2829 June, 1998, three weeks after the onsetof the experiment, samples were alsotaken between 11 p.m. and 1 a.m. toexamine the possible daytime aggregationof Daphnia and other crustaceanzooplankton close to the sediment, aswell as their diel vertical migration(DVM). Concomitantly with the samplingprocedure, the temperature, oxygen andsecchi depth were measured in the field.On the last week of the experiment in1999, vertical profile of light penetrationinto water was measured also with aLiCor LI1400 4 sensor (LiCor Ltd.,Lincoln, NE, U.S.A.). The standardmethods used in physical and chemicaldeterminations, microscopic planktonanalyses as well as data analyses aredescribed in detail in each individualpaper.
Samples for TSS (additional resultsin this thesis) were filtered through dried(105ºC), weighed Whatman GF/F filters(Whatman International, Maidstone, UK).After drying (105ºC, 2 hours), the filterswere weighed for the amount of TSS. On45 August, 1999, samples of epiphyticmacroinvertebrates (additional results inthis thesis) were taken from macrophytebeds of each enclosure using a coresampler (length 32 cm, diameter 13 cm;Kornijów 1998). Living animals wereidentified, weighed and their biomass wascalculated per macrophytic biomass.
The additional time series resultspresented in this thesis were analysedusing repeated mixed procedure of theSAS package, version 6.12 (StatisticalSystem Institute Inc., USA). Furtherinformation about the mixed model isgiven in papers I and II. Other statisticaltests were done using univariate analysisof variance (ANOVA, followed by
17
Tukey’s test) or Pearson’s bivariatecorrelation test of SPSS for Windows(version 10.0). When the assumptions forANOVA were not met, logtransformation was used to stabilizeheterogeneous variances or to normalizethe distribution of residuals.
4. RESULTS ANDDISCUSSION
4.1 Fishzooplankton interactionswithin macrophyte beds
In shallow lakes and littoral zones,edge zones between plant beds and openwater as refuges for Daphnia wereconsidered an important bufferingmechanism to changes in fish predationpressure (Schriver et al. 1995, Lauridsenet al. 1996, Jeppesen et al. 1998a, Mosset al. 1998). Consequently, during a 3day period after roach introduction thebiomass of Daphnia longispina O. FMüller was reduced more in the noElodea enclosures than in the Elodeaenclosures (I). However, Elodea also hada strong negative impact on Daphnia,thus complicating the estimation of therefuge effect (I; Fig. 1). In another
experiment without fish treatments in1996, no negative effects of Elodea onDaphnia were evident (II). Burks et al.(2000) demonstrated that the growth ofDaphnia magna Straus was stronglysuppressed when exposed to chemicalsfrom both E. canadensis and roach.Schriver et al. (1995) found positiveeffects of macrophytes but a negativeinteraction of fish and macrophytes onzooplankton biomass. Chemicalcommunication plays a major role inaquatic habitats (Larsson & Dodson 1993,Burks et al. 2002, Gross 2003) and theresponses of plankton to differentcombinations of numerous chemicalsignals may be multifaceted especially inlittoral habitats (Larsson & Dodson1993). Such communication may becomepronounced in enclosed systems that donot allow largescale horizontal migrationof Daphnia (I) (cf. Burks et al. 2002).The repellent properties of macrophyteson Daphnia and other zooplankton(Hasler & Jones 1949, Pennak 1973,Dorgelo & Heykoop 1985, Blindow et al.2000, Burks et al. 2000, 2002) arecertainly dependent on variableconditions within plant beds and one ofthe key factors appears to be the presenceof fish.
0,4
0,6
0,8
1
0 7 14 21 28
time, days
noElodeasparseElodeadenseElodea
0,4
0,6
0,8
1
0 7 14 21 28
time, days
leng
th (m
m)
Elodea NSfish F 1,12 = 11.24, P = 0.006Elodea *fish NStime F 1,83 = 15.38, P < 0.001Elodea *time NSfish*time F 1,83 = 15.38, P < 0.001Elodea *fish*time NS
Figure 7. Treatment means (+ 1 SE) over the course of the experiment for the body lengthof euplanktonic cladocerans in the Elodea enclosures with or without fish (I). The F ratioswith significance levels for the main effects and interactions of Elodea, fish and timeobtained in the mixed model are given for each variable.
18
0,2
0,3
0,4
0,5
0,6
0,7
0 25 50 75 100Macrophyte biomass, g DW m2
Leng
th (m
m)
Fishfree Lowfish HighfishFishfree Lowfish Highfish
Figure 8. Relationships between thedensityweighed mean length ofeuplanktonic filterfeeding cladocerans andthe total biomass of macrophytes at the endof the experiment in 1999 in the differentfish treatments (n = 12, except in highfishtreatments n = 11).
Fish introduction generally led tothe reduction in largebodied, filterfeeding cladoceran biomass, especiallyDaphnia, which became virtually absentwith fish regardless of the abundance ofmacrophytes (I, III, IV). This resultaddresses the generally high predationrisk in shallow lakes (Jeppesen et al.2003) also suggested by the negligiblenumbers of daphnids in the open lakearea outside enclosures (05 ind. l1;unpublished data) and their notableincrease when enclosed and protectedfrom fish predation (I, II). Theineffectiveness of the Elodea refuge wasalso evidenced by the equally decreasingbody size of euplanktonic cladocerans inall Elodea treatments with 16cm roach(Fig. 7) (I). In contrast, small cladoceranssuch as Ceriodaphnia, a genus welladapted to plant beds (Jeppesen et al.1998a), benefited from the refugeprovided by macrophytes, especiallythick floating flocks of Lemna trisulca,against 510 cm perch (III). The refugeeffect was also indicated by the slightlyincreasing body size of euplanktoniccladocerans with macrophyte biomasstowards the end of the experiment in thefish treatments (Fig. 8).
In the larger panEuropean dataset,the variability of zooplankton responsesto fish manipulations was influenced bysubstantial variation in communitystructure and macrophytic refuges amongsites. For instance, the refuge effect forlarge crustacean grazers by the abundantMyriophyllum spp. was evident in Leon(IV, FernándezAláez et al. 2004). Charabeds have also been considered as goodrefuges for zooplankton against fishpredation (Diehl 1988, Jeppesen et al.1998a). However, Chara spp. at the studysites in Valencia and The Netherlands didnot prevent efficient foraging of fish onlarge crustacean grazers (IV) (Van deBund & Van Donk 2004, Romo et al.2004). Even the lowest densities of fish(1.92.5 ind. m2) used in the experimentsof this thesis were near the threshold
density (25 ind. m2), at which the refugeeffect of macrophytes may be partially ortotally lost for cladocerans (Jeppesen etal. 1998a). However, such threshold fishdensity is probably shaped by the fishspecies and size/ageclass, makinggeneralisations difficult.
In contrast, plantassociatedcladocerans benefited of macrophyticrefuge against fish predationmacrophytes. The biomass and body sizeof Sida crystallina O. F. Müller andseveral species of chydorids increased inElodea beds and were even moreabundant and larger in the presence thanin the absence of 16cm roach (I). Lemnatrisulca also appeared to provide refugefor plantassociated cladocerans(chydorids and Simocephalus vetulus O.F.Müller) against 510cm perch predation(III), indicated by the positive correlationbetween the final macrophyte biomassand cladoceran body size (Pearson; r =0.53, P = 0.009; unpublished data). Inaddition to plantassociated cladocerans,the biomass of epiphytic
19
b.
0
20
40
60
80
100
Fishfree
Lowfish
Highfish
Bio
mas
s, (m
g g
1)
***a.
0
20
40
60
80
100
Fishfree
Lowfish
Highfish
Bio
mas
s (m
g g1
)
**
Figure 9. Mean (+ 1 SE) biomass ofepiphytic macroinvertebrates (DW perDW macrophytic biomass), excludinggastropods, in the different fish treatmentson (a.) 7 July, 1999 and (b.) 4 August,1999. Nutrient enrichment did not affectthe macroinvertebrate biomass (ANOVA;P > 0.05). The results from the nutrienttreatments were combined and, thus, eachbar represents the mean of 12 replicates.Significant differences between the fishtreatments (ANOVA; P < 0.05) are shownby the results of the Tukey’s test in thefigure (**P < 0.01, *** P < 0.001).
macroinvertebrates increased in Elodeabeds (I; Fig. 2). Thus, within macrophytebeds, a shift from the dominance ofplanktonic communities towards that ofplantassociated communities can beexpected, with important consequencesfor competitive interactions of differentfish species through food availability(Diehl 1993, Kornijów & Kairesalo1994).
Alternative food sources for fish in thelittoral habitat and consequences for
fishzooplankton interaction
According to the general view,roach are not only selective but alsogeneralists shifting to the prey categorythat is most rewarding as a result of boththe properties of an individual preyspecies and the density of the preypopulation (Townsend et al. 1986,Gliwicz 2002). In the enclosureexperiment in Lake Vesijärvi, however,with the decreasing availability ofcladoceran food roach switched tofeeding on mainly detritus and plantmaterial (I; Fig. 2) instead of usingmacroinvertebrates as an alternative foodsource as has been shown in laboratorystudies (Horppila & Kairesalo 1992).Roach had a high search image forcladocerans and preferred them evenwhen these became smaller and scarcer.In addition, roach had no influence on thebiomasses of even the most eatentrichopteran and ephemeropteran larvae.Thus, the overall foraging of roach formacroinvertebrates was low and theunavailability of their preferredcladoceran food induced roach to increaseconsumption of less nutritiousdetritus/plant food. This result is inagreement with observations by Perssonand Greenberg (1990). The capability forusing alternative, lesspreferred foodsources may serve as lifesupportingmeans for roach populations to manage
over periods of low cladoceranabundance. Thus, they are stabilised bythe alternative food resources in littoralhabitats and can maintain high predationpressure on zooplankton (cf. Schindler &Scheuerell 2002, Vadeboncoeur et al.2002, Jeppesen et al. 2003). Among themany factors affecting predatorpreydynamics, the high preference forcladocerans is not likely to be astabilising effect (Murdoch & Bence1987, cf. Schindler & Scheurell 2002).On the other hand, in perchdominatedsystems, macroinvertebrates may play amore important role as alternative preyknowing that perch are generally superiorto roach in foraging formacroinvertebrates in vegetation (Persson1987, Diehl 1988). For instance, in theexperiment conducted in 1999 (Table 2),epiphytic animals such as chironomid andephemeropteran larvae were relativelyabundant within macrophyte beds (ca. 28g m2; unpublished data). They wereefficiently preyed by the juvenile perch,
20
evidenced by their decreasing biomasseswith fish (Fig. 9; unpublished data),which apparently contributed to therefuge effect observed for cladoceransagainst fish predation (III) (cf. previouschapter). These results emphasise theimportant role of submerged vegetation inmodifying the foraging behaviour ofdifferent fish species and, consequently,affecting the predation pressure onzooplankton.
Diel vertical migration of littoralzooplankton in response to fish
predation
Littoral zooplankton may undergoDVM between the lake bottom and openwater in response to fish predation (Wallset al. 1990, Jeppesen et al. 2002b).Consequently, in mesocosms with 510cm roach, higher biomasses ofeuplanktonic cladocerans (Bosmina,Ceriodaphnia and Diaphanosoma) wereobserved at night than during day (Fig.10). However, the number of individualswas low and the variation betweenreplicates high. The dominating
cladoceran, predaceous Polyphemuspediculus L. (cf. III), exhibited a similarbehavioural pattern in the lowfishtreatments (4 g FW m2), while acontrasting trend was observed in thefishfree treatments (Fig. 10). P.pediculus typically occurs in denseswarms in the littoral zone and isconsidered a dayactive animal migratingto the bottom during night (Butorina1986). Apparently, the risk of fishpredation can modify such behaviour. Inthe highfish treatments (20 g FW m2),the numbers of Polyphemus were reducedto low levels, unlike in the earlierexperiment, where they were evenpositively affected by the presence of 16cm roach (I). The discrepancy obtainedhere may reflect changes in the feedingpreferences of roach with size. Despitetheir large body size, P. pediculus is notconsistently preferred by roach due totheir active escape reaction (Moss et al.1998).
Thus, even in shallow watersplankton abundance may change due toDVM, especially when exposed to fishpredation, and this should be taken intoaccount when planning experiments. In
Polyphemuspediculus
0
20
40
60
80
100
Fishfree
Lowfish
Highfish
num
ber o
f ind
ivid
uals
l1
daynight
Euplanktoniccladocerans
0
5
10
15
20
25
Fishfree
Lowfish
Highfish
Den
sity
(ind
. L1
)
Figure 10. Mean (+ 1 SE) density of euplanktonic cladocerans and Polyphemus pediculusin the different fish treatments during day and night three weeks after the beginning of theexperiment in 1998. Each bar represents the mean of 6 replicates, i.e. the results from bothnutrient treatments were combined.
21
general, however, zooplankton maybenefit less from DVM in response to fishpredation in shallow than in deeper,stratified lakes (Sarvala et al. 1998). Thisis likely due to light penetrating down tothe lake bottom, making verticalmigration less advantageous (Burks et al.2002).
Major effects of fish relative tonutrients in structuring thezooplankton community
The panEuropean mesocosmexperiment revealed remarkable yeartoyear and geographical differences inzooplankton community structure andbiomass (IV). This was partly attributedto differing weather conditions innorthern and central Europe between thestudy years: summer 1998 was cool andwindy, whereas summer 1999 was warm(IV; Table 1). However, some consistentpatterns were evident. Fish had greaterinfluence than nutrients in regulatingzooplankton biomass, especially therelative abundances of differentfunctional groups of zooplankton. Thebiomass of large crustacean grazers (> 0.5mm) declined with the presence of fish(IV; Figs. 1 and 2, Table 2), thusaddressing their sensitivity to predatormanipulations, which is not a novelconclusion (Hrbácek et al. 1961, Brooks& Dodson 1965, Brett & Goldman 1997,Mehner et al. 2002, Shurin et al. 2002).At many sites, the biomass of predatory(raptorial) zooplankton, predominantlycyclopoid copepods, was also reduced byfish (IV; Figs. 2 and 3).
However, the responses of totalzooplankton biomass to nutrientenrichment were mainly positive (IV;Figs. 1 and 3). For instance, in 1998nutrient enrichment increased thebiomasses of small crustaceans (< 0.5mm) largely independently on thepresence or absence of fish (IV; Table 2,Fig. 3). Macrophytes appeared to weaken
the effects of fish in some of theexperiments. Small zooplankton speciesapparently benefited from the mostlynegative responses of their predators andlarger competitors to the presence of fish.Such compensation suggests a ratiodependent functional response and isconsistent with the view that nutrientenrichment leads to proportionalincreases at all trophic levels (Arditi &Ginzburg 1989, Leibold et al. 1997). Italso emphasises the importance ofcompositional changes within trophiclevels and reveals the limitation of simplefood chain and ratiodependent models incapturing such changes, as pointed out byLeibold et al. (1997), Hulot et al. (2000)and Persson et al. (2001).
4.2 Stability of littoral ecosystemsand the control of phytoplanktonbiomass within macrophyte beds
Confounding effects of macrophyteson phytoplankton
Zooplankton grazing in vegetationplays a central role in controllingphytoplankton (Schriver et al. 1995,Jeppesen et al. 1998a, 1999). However,Meijer et al. (1999) and Blindow et al.(2000) suggested that more directinteractions between macrophytes andplanktonic algae, rather than mere grazingof algae by zooplankton, are probably thedecisive factors in controlling algaldensities in lakes with dense submergedvegetation. Submerged macrophytes areregarded as P and N sinks during theiractive growth, thus suppressing thegrowth of phytoplankton, especiallycyanobacteria (reviewed in Søndergaard& Moss 1998). In addition, P retentionincreases within rooted macrophyte bedssuch as Elodea canadensis through rootoxygen release suitable for the formationof complexes of P with iron (Fe) (Hupfer& Dollan 2003). However, several studiesshowed that macrophytes do not
22
0
20
40
60
80
0 20 40 60 80 100 120
Lemna coverage (%)Lemna biomass (g DW m2)
Chl
a (µ
g L1
)
chl a : Lemna coverage chl a : Lemna biomass
chl a : Lemna coverage chl a : Lemna biomass
Figure 11. Relationships between the timeweighted average chl a concentration and thecoverage and biomass of Lemna trisulca(Pearson; P > 0.05) (n = 36).
0
2000
4000
6000
8000
10000
noElodea
sparseElodea
denseElodea
biom
ass
(µg
C L
1)
16400
Figure 12. Mean (+ 1 SE) biomass ofcyanobacteria in the different Elodeaenclosures with or without fish during themiddle of the experiment on 14 August, 1994.
necessarily affect nutrient availabilityfor phytoplankton and that nutrientcompetition between macrophytes andphytoplankton may be relativelyunimportant (Schriver et al. 1995,Beklioglu & Moss 1996, Van Donk &Van de Bund 2002). In agreement withthese observations, the results of thisthesis suggested that macrophytes andtheir associated periphytic communitiesdid not appreciably affect nutrientconcentrations in the water and wereunlikely to have caused nutrientlimitation of phytoplankton (I; Table 1)(II; Fig. 1, Appendix A) (III; Fig. 3).
Macrophytes may have negativeeffects on phytoplankton throughshading, but phytoplankton canovercome the problem of light limitationthrough buoyancy and motility(Søndergaard & Moss 1998). Theabundant floating flocks of Lemnatrisulca reduced photon flux density by9095% immediately below the flocksbut only 040% at the bottom due toscattering of light (III). The surfacecoverage as well as the final biomass ofLemna correlated weakly with chl aconcentration (Fig. 11), suggestingminor effects on phytoplankton throughshading. Moreover, the phytoplanktonwas dominated by mobile, flagellatedalgae such as Chlamydomonas andVolvox (III).
The dense beds of Elodeacanadensis also showed consistently lowchl a concentrations (I; Fig. 1) andcyanobacterial biomasses (Fig. 12), evenwhen the zooplankton biomass wasrelatively low, suggesting thatphytoplankton production was directlyor indirectly constrained by Elodea.Even though fish had a negligible effecton the total biomass and clearance rate ofeuplanktonic grazers, approximately 2fold higher phytoplankton biomasses withslightly more edible algae such ascryptophytes and chlorococcales wereobserved in all Elodea treatments withroach as compared to the corresponding
fishfree enclosures (I). Thus, fish had apositive cascading effect onphytoplankton biomass via zooplanktongrazers that addressed the role of grazingin controlling phytoplankton withinmacrophyte beds. Even though nodifferences in dissolved nutrientconcentrations between the treatments
23
could be measured, this cascade was alsoevident via nutrient regeneration by fish(I). Enhancement of P and/orphytoplankton with fish was attributed toP inputs by fish in several previousexperiments (e.g. Horppila & Kairesalo1992, Järvinen & Salonen 1998, Vanni2002), also emphasising the role ofresource control relative to predatorcontrol.
Littoral ecosystem resistance andresilience to phosphorus enrichment
Littoral planktonic communitiesreadily responded to both pulse and pressP enrichments, although the major sinkfor the added P was well oxygenatedsurface layers of the sediment (II).However, periphytic chl a concentrationand accumulation of P by periphytonwere low even with press P enrichment(II; Table 1, Appendix B), thuscontrasting earlier observations of themajor role of periphytic communities inthe nutrient dynamics of littoral areas (cf.Wetzel and Søndergaard 1998).
Planktonic responses to Penrichment differed between Elodeatreatments and noElodea treatments. Penrichment resulted in higher biomassesof cyanobacteria (Anabaena circinalisRabenhorst and A. spiroides Klebahn),greater concentrations of microcystinsand lower biomass of cladocerans(predominantly Daphnia longispina andSida crystallina) in the absence, asopposed to the presence, of Elodea (II;Figs. 2, 3). Thus, measured ascyanobacterial biomass, noElodeatreatments had a lower resistance to pressand pulse P enrichments than Elodeatreatments (II; Fig. 4). Both Elodeatreatments were also resilient, returning toreference conditions after pulse P
enrichment, except for one noElodeaenclosure, which had a low biomass offilterfeeding cladocerans and in whichhigh cyanobacterial biomasses developedeven after pulse P enrichment. Thecalculated return time was slightly lowerin Elodea treatments (16 + 7 (SE) days, n= 3) than in noElodea treatments (21 + 4days, n = 2).
Grazing by large cladocerans,which increased towards the end of theexperiment, probably contributed to therecovery of the littoral ecosystem.Evidently, however, Elodea was able toregulate the growth of Anabaena directlyor indirectly (I, II), thus strengthening theresistance and resilience of the systemagainst P enrichment (II). Organiccompounds released by macrophytes mayinhibit the growth of cyanobacteria (cf.Gross 2003) and stimulate bacterialproduction, increasing thereby the uptakeof P by bacteria at the expense ofphytoplankton (Søndergaard & Moss1998, Søndergaard et al. 1998, cf.Järvinen & Salonen 1998). This indirect,plant mediated P pathway may also havecontributed to the suppression ofAnabaena through feeding on Prichbacteria by cladocerans, especially by D.longispina (cf. Kankaala 1988, Salonen etal. 1994, Jeppesen et al. 1996, Järvinen &Salonen 1998). Daphnia and several othercladocerans also recycle nutrients withhigh N:P ratios and may thus reduce thecompetitive success of cyanobacteria withlow optimal N:P ratios (MacCay & Elser1998). Moreover, they release nutrients inorganic rather than in inorganic forms,thus supporting heterotrophic rather thanautotrophic processes in ecosystems(Anderson et al. 2005). The results of theenclosure experiment in 1996 showedthat submerged plants regulatephytoplanktonzooplankton interactionsin lake littoral (II).
24
0
0,002
0,004
0,006
0,008
lake fishfree lowfish highfish
TSS
(mg
L1)
*
Figure 13. Mean (+ 1 SE) values of timeweighted averages of total suspended solid(TSS) concentrations in the different fishtreatments (n = 3) as well as in the nearbylake area outside the enclosures. Onlynonenriched treatments are included in thefigure. The asterisk shows the result of theTukey’s test.
Macrophytes lowered the total toxinproduction of cyanobacteria but did notaffect the specific toxicity. Toxinproducing strains of cyanobacteria arefavored by high concentrations of P(Rapala et al. 1997) and by direct orindirect exposure to zooplankton (Jang etal. 2003). In this experiment, however, Penrichment did not evidently increase themassspecific toxicity and no correlationwas found between toxicity ofcyanobacteria and zooplankton biomass.This, in fact, was to be expected sincezooplankton biomasses and speciescomposition did not differ drasticallybetween the treatments. This result alsosuggests tolerance against toxiccyanobacteria in Daphnia and otherzooplankton, in accordance withobservations by Gustafsson and Hansson(2004).
Littoral ecosystem resistance tophosphorus and nitrogen enrichment
The central role of the littoralcommunity structure in determining theresistance to nutrient enrichment withboth N and P was demonstrated also intwo consecutive years (1998 and 1999)with contrasting abundance ofmacrophytes and cladocerans (III). Whenmacrophytic growth was poor, filterfeeding cladocerans were scarce (initialbiomass < 10 µg C l1) and thezooplankton community was dominatedby cyclopoid copepods and rotifers,enrichment provoked a turbid water statewith high planktonic (predominantlychlorophytes and cryptophytes) andperiphytic algal biomasses, masking theeffects of fish on algal biomasses (III;Table 2, Figs. 1, 2). Under the lowtemperature conditions prevailing in earlysummer, the start of macrophyte growthwas delayed, and further suppressed bythe development of planktonic andperiphytic algal biomasses in the enrichedmesocosms. A clearwater state with low
phytoplankton biomass occurred only inunenriched mesocosms without fish orwith low fish biomass (4 g FW m2).Roach further increased turbidity insidethe mesocosms, as evidenced by thehigher amounts of total suspended solids(ANOVA; F2,8 = 5.70, P = 0.046; Fig. 13;unpublished results), while the clearancerate of cladocerans was equally low in allfish treatments (III; Fig. 2). Even thoughweather conditions were cool and windy,turbidity was low in the open lake area
outside the enclosures (Fig. 13). Thus,fishinduced resuspension is an importantfactor causing turbidity in lake littoralzones, in accordance with severalprevious studies (Horppila & Kairesalo1992, Horppila 1994, Horppila &Nurminen 2003).
When macrophytes (mainly Lemnatrisulca) were abundant and thezooplankton community was dominatedby filterfeeding cladocerans(Diaphanosoma and Ceriodaphnia; initialbiomass > 100 µg C l1) (III; Fig. 2), aclearwater state with low phytoplanktonbiomass prevailed even under the highest
25
levels of nutrient enrichment (III; Fig. 1).The cladoceran grazing rate increasedwith nutrients and controlled thephytoplankton biomass, but the cascadingeffect of fish shown by the reducedgrazing rate and increased chl aconcentration became evident betweenthe low and highfish treatments (Tukey;P = 0.001 and P = 0.039, respectively)(III; Fig. 1). Only fish but not nutrientsaffected the phytoplankton biomass (III;Table 5); however, at the two highestnutrient levels the threshold fish biomasswas even above 20 g FW m2 alsosuggesting a high growth rate ofcladocerans with nutrients. Ceriodaphniaare small (< 0.7 mm in this experiment)but they can have high massspecificgrazing rates (Mourelatos & Lacroix1990). They, as well as Diaphanosoma,are efficient microfiltrators ofbacterioplankton (DeMott 1986) and thuswell adapted to macrophyte beds(Jeppesen et al. 1998a). Similar plantmediated nutrient pathways via bacteriaand heterotrophic processes, as suggestedby observations with Elodea (II) in theprevious chapter, probably also operatedin systems with abundant Lemna. Inaddition, the capability of Ceriodaphniafor incorporating high amounts of Prelative to N is comparable to that ofDaphnia (Hessen & Lyche 1991). Thus,nutrient cycling by cladocerans may alsohave controlled phytoplankton growth.
Thus, ecosystem resistance may behigh and the sensitivity of phytoplanktonbiomass to nutrient input low withabundant, relatively small cladocerans inaccordance with the observations by e.g.Helminen and Sarvala (1997) andCottingham et al. (2004). However, morepronounced effects of zooplankton onphytoplankton are often seen with largeDaphnia as the key grazer (cf. Carpenteret al. 1985, Carpenter & Kitchell 1993,Mazumder 1994, Cyr & Curtis 1999).Furthermore, analyses of major species,not only keystone species, with carefulconsiderations about their interactions
within the community may provideinsights into ecosystem functioning (cf.Hulot et al. 2000), an important aspectespecially in littoral ecosystems.
Role of plantassociated animals incontrolling phytoplankton and
periphyton
Plantassociated cladocerans, lessvulnerable to fish predation thaneuplanktonic species (I, III) (Chapter 4.1),can have high grazing impact onphytoplankton (Jeppesen et al. 1998a,Balayla & Moss 2004). Sida is anespecially efficient filter feeder (Balayla& Moss 2004), while chydorids scrape onperiphyton and are considered lesseffective grazers (Lövgren & Persson2002). The biomasses of thesecladocerans increased with macrophytesand were likely even higher thanobserved, since water samplingunderestimated the numbers of plantassociated species (IIII). Elodea alsosustained rich assemblages of herbivorousand detrivorous macroinvertebrates (I).Thus, Elodea harboured herbivorescapable of controlling not onlyphytoplankton but also periphyton, whichare qualitatively different forms ofprimary producers. Grazing wasconsidered as a possible explanation forthe slight accumulation of periphyticbiomass on plastic strips despite Penrichment (II). This addresses theimportance of taking into considerationthe trophic level heterogeneity in littoralecosystems (cf. Lövgren & Persson2002). The complexity of littoral habitatsinvolving different forms of primaryproducers also addresses the limitedcapability of the food chain theory forconsidering such heterogeneity.Separating the two communities is alsodifficult since they are linked viapredators and behavioural patterns. Forinstance, several plantassociated
26
cladocerans such as Sida also swim freelyat night (Walls et al. 1990).
Control of phytoplankton biomass byzooplankton grazers: geographical
differences
Increase in the biomass of rotiferswas generally associated with an increasein chl a, indicating the low ability ofthese specialised suspension feeders tocontrol total phytoplankton biomass (IV;Table 3). On the other hand, throughproviding food for fish larvae, rotifersmay maintain a high recruitment ofzooplanktivorous fish and therefore therole of rotifers is emphasised under turbidconditions. At many sites, the biomass ofsmall crustaceans also had a positive orinsignificant correlation with chl a (IV;Table 3). In contrast, the biomasses oflarge crustacean grazers were inverselyrelated to the chl a concentrations, exceptat the highest temperature (close to 30 oC)in Valencia where chl a was unrelated tothe biomass of large crustacean grazers(IV; Table 3) and the overall biomass oflarge cladocerans was low compared withthat at other study sites (IV; Fig. 2). Thisresult suggests that the role of grazing incontrolling phytoplankton biomass byespecially large grazers was important atall sites except in Valencia. The dataset,however, was insufficient to showwhether there could be a thresholdbetween temperature regimes at whichthe functioning of ecosystems wouldmarkedly change (cf. Scheffer &Carpenter 2003). Increase in the biomassof small grazers and rotifers wasgenerally associated with an increase inchl a (IV; Table 3), indicating the lowability of small zooplankton to controltotal phytoplankton biomass.
At the high temperatures inValencia, nutrient enrichment apparentlyled to inharmoniously more rapid growthrates of algae compared with those of
zooplankton and thus disrupted topdowncontrol of algae (cf. Arditi & Ginzburg1989, Power 1992). In addition, thecrustacean zooplankton biomass wasefficiently reduced by fish, supporting thestatement that with increasingtemperature the herbivore control willfurther weaken with respect to resourcecontrol if the disruption of trophicregulation is interfered with predators(Power 1992). Apart from the reducedrole of large crustacean grazers in thesouthernmost location in Valencia, noconsistent geographical patterns wereobserved in the responses of zooplanktoncommunities to nutrient and fishmanipulation. Other potentialgeographical differences were probablymasked by the high yeartoyear variationin prevailing weather conditions. Suchenvironmental variability may greatlyaffect the structure and functioning oflittoral communities (cf. III).
Role of zooplankton grazing inexplaining the variation between
chlorophyll a and total phosphorus
A steep slope between chl a andtotal phosphorus (TP) concentrationsuggests a cascading effect of fishthrough herbivorous zooplankton onphytoplankton in systems with threetrophic levels (Hansson 1992) and can befound in nonstratified lakes lacking largezooplankton grazers (Mazumder 1994,Sarvala et al. 1998). The results from thehighfish treatments (20 g FW m2) of thepanEuropean experiment agree withthese predictions and were evidenced bythe generally low biomass of largegrazers and consistently high biomass ofphytoplankton with increasing nutrients(IV; Fig. 4e). However, the slope in thelowfish treatments was less steep andsimilar to that in the fishfree treatments.Perhaps small crustacean grazers andplantassociated cladocerans thatbenefited from the macrophytic refuge
27
and were not heavily preyed upon by fishprobably played at least a complementaryrole in controlling phytoplanktonbiomass, as was shown in the Finnishmesocosm experiment in 1999 (III)(Chapter 4.2, pages 2526). Thus,zooplankton controlled phytoplanktonwhen fish were not abundant (< 4 g FWm2 or 2.54 ind. m2). Cladoceran bodylength is an important predictor ofphytoplankton biomass and water clarity(Carpenter et al. 1998, Stemberger &Miller 2003) and was suggested also by
observations in Lake Vesijärvi (Chapter1.5). On the other hand, in some systems,e.g. in shallow, nonstratified LakePyhäjärvi with planktivorous fishbiomass of < 30 kg ha2 (i.e. 3 g m2),total cladoceran biomass consisting ofrelatively small individuals explainedmuch of the variation in chl aconcentration (Sarvala et al. 1998).
Despite the scarcity of the keystonegrazer Daphnia, the total abundance oflarge crustacean grazers (predominantlyDiaphanosoma, Sida, Simocephalus,
1
10
100
1000
10 100 1000 10000TP (µg L1)
Chl
a (µ
g L
1)
<8090 µg>8090 µg<8090 µg>8090 µg
b.a.
1
10
100
1000
10 100 1000 10000
TP (µg L1)
chla
(µg
L1)
<30 %
>30 %
<30 %
>30 %
1
10
100
1000
10 100 1000 10000
TP (µg L1)
chla
(µg
L1)
c.
f ishfree low fish highf ish
f ishfree low fish highf ish
Figure 14. Relationship between concentrations of chlorophyll a (chl a) and total phosphorus(TP) in enclosures and mesocosms with a biomass of large grazers below and above (a.) 30%and (b.) 8090 µg DW l1 of the total zooplankton biomass as well as with (c.) three fishdensities. Data are from experiments carried out in Lake Vesijärvi in 1994, 1996, 1998 and1999 and at five other shallow lake sites across Europe in 1998 and 1999. Note the logarithmicpresentation of both axes. Panel a. also shows the regressions from Mazumder (1994) fornorthern temperate nonstratified lake ecosystems having (lower dashed line) or lacking (upperdashed line) large Daphnia. The circle in panel c. shows the data points from fishfree, highnutrient treatments in Valencia with negligible biomass of large grazers. The arrow shows howthe slope of the fishfree regression line changes when excluding these data points (see text forfurther explanation).
28
Eudiaptomus) explained the relationshipbetween chl a and TP better than did themere number of trophic levels. Whenabundant (> 8090 µg DW l1; IV), largegrazers controlled phytoplankton biomasseven under hypertrophic conditions (up to1600 µg TP L1). Otherwise, the chl aconcentration increased steeply withincreasing TP concentration (IV; Fig. 4d).This threshold biomass is close to that(100 µg DW l1) observed by Hansson(1992). However, even strongerdifferences in the TP:chl a ratio werefound in systems with sparse/abundantDaphnia (Sarnelle 1992). In reality,however, hypertrophic conditions withlow biomass of fish and high biomass oflarge zooplankton are rare except afterfish kills or in effectively biomanipulatedlakes with abundant submergedvegetation (cf. Sarvala et al. 1998, 2000).
Phytoplankton control by grazinghas been shown to be consistently low,when Daphnia constitutes < 20% ofzooplankton biomass, while above thisthreshold grazing potential is high(Brooks & Dodson 1965, Mazumder1994). The empirically observedthreshold proportion of large grazers inthe panEuropean data set was similar inmagnitude (ca. 30%) although notdirectly comparable, due to the lack oflarge Daphnia in most systems. Theratios between the concentrations of TPand chl a may be sensitive to further datapoints. However, when the dataset wassupplemented with the results obtained inthe experiments conducted in 1994 and1996 in Kilpiäistenpohja Bay (I, II)(Table 2), the TP:chl a ratios remainedessentially similar (Fig. 14a, b). Bothgrazer systems had smaller slopes thanthose obtained by Mazumder (1994) fromseveral northern temperate nonstratifiedlake ecosystems and enclosureexperiments under contrasting grazercommunities (Fig. 14 a). This may reflectalso other mechanisms besides grazing inthe control of phytoplankton in shallowlittoral ecosystems such as those related
to macrophytic regulation ofphytoplankton growth. The low resistanceof the systems in the southernmostlocation in Valencia to nutrientenrichment even in the absence of fishcan be discerned as a separate group ofdata points (denoted by a circle in Fig.14c). When excluding these data points,the slope became even less steep (denotedby an arrow in Fig. 14c).
On the other hand, thephytoplankton biomass was positivelyrelated to the TP concentration even inthe presence of efficient grazers, thusreflecting the positive response ofadjacent trophic levels, inconsistentlywith traditional foodchain theory. Thisobservation is in agreement with theresults of several earlier studies frommesocosm to wholelake scale (Hansson1992, Mazumder 1994, Brett & Goldman1997, Leibold et al. 1997, Hansson et al.1998c, Sarvala et al. 1998, Persson et al.2001). Ultimately, ecosystemproductivity is quantitatively moreimportant than consumer regulation indetermining the biomass of primaryproducers. Nevertheless, the results ofthis thesis showed that zooplanktongrazers play an important role incontrolling phytoplankton, in accordancewith several earlier studies (e.g.Carpenter et al. 1985, 1993, Hansson1992, Sarnelle 1992, Sarvala et al. 1998,2000, Cyr & Curtis 1999, Jeppesen et al.1999, 2003, Persson et al. 2001).Community structure, especially theabundance of zooplankton grazerstogether with macrophytes, is of crucialimportance in determining the stability ofthe littoral ecosystem as well as itsresistance to nutrient loading. Thus, foodweb management can be a useful tool foraffecting water quality of northern andtemperate eutrophic lakes.
29
4.3. Field mesocosm experimentsand their applicability for studyingaquatic ecosystem functioning
Experimental mesocosm systemsmay produce results quite comparablewith natural systems (Sarnelle 1992,Mazumder 1994). Nevertheless, theyinvolve the problem of both spatial andtemporal scale that exclude some of theecosystem processes and tend to increasethe strength of food web interactions andthe trophic cascade (Schindler 1990,Carpenter & Kitchell 1993). Thisphenomenon was also revealed by severalresults of the experiments in this thesisand has been discussed in the text. It alsoaddresses the cautiousness called for ininterpreting their results and generalisingthem across larger scales.
Littoral community structures,especially the abundance of macrophytesand efficient zooplankton grazers, appearto be driven largely by prevailing weatherconditions (III, IV, Stephen et al. 2004).Longterm variation in lake food webstructure and water quality is dependenton climatic fluctuations, stronglydetermined by spring and summer windand temperature conditions, that affect thevariation in fish yearclasses as well asthe establishment of submergedvegetation and thus indirectlyzooplankton and phytoplanktondevelopment (cf. Sarvala et al. 1998,Hargeby et al. 2004). Thus, multiplestresses from climatic and biotic variablesmodify patterns within a lake ecosystemin different years (III, IV) (cf. Salonen etal. 1990, Hargeby et al. 2004). Smallscale experiments can be helpful inexplaining the mechanisms behind suchlargerscale patterns as well as ecosystemresponses to perturbations in more detail(Scheffer & Carpenter 2003). Forinstance, complex direct and indirectways, through which submergedmacrophytes affect littoral nutrient andplankton dynamics (II) would probablybe impossible to document without
controlled experiments. In addition,studying the responses of littoralecosystems to perturbation is complicatedby the patchy, heterogeneous nature ofthese habitats. Such practical constraintsdetermine the experimental design.
Differences in the startingconditions, i.e. community structure, arecrucial to the outcome of perturbationssuch as nutrient enrichment (II, III,Stephen et al. 2004). Even in the absenceof environmental heterogeneity, the initialstate and variation in communityassemblages across landscapes ofcommunity patches can result in variableoutcomes (Drake et al. 1996). Suchobservations emphasise the limitationsand reproducibility of fieldexperimentation (Polis and Winemiller1996, Moss et al. 2004). High deviationof several results obtained in theexperiments of this thesis also addressesthe need to have preferably more thantwo or three replicates of each treatment.On the other hand, the basicunderstanding of how ecosystemsrespond to perturbations is increased bythe knowledge about how different initialconditions may lead to different finalstates. Field experiments can be apowerful way to show alternativeattractors of systems (II, III) (Scheffer &Carpenter 2003). This, in turn, haspractical implications for the managementand restoration of lake ecosystems inorder to increase their resistance andresilience to anthropogenic perturbations.
30
5. CONCLUSIONS ANDFUTURE RESEARCH NEEDS
Initial community composition,primarily the abundance of submergedmacrophytes and cladoceran grazers,played the decisive role in determiningthe stability of littoral ecosystems andresistance to nutrient enrichment,recorded as lower biomass ofphytoplankton. When the abundance ofmacrophytes and cladocerans was low, Penrichment favoured cyanobacteria, whileP and N enrichment favouredchlorophytes and cryptophytes as well asperiphytic algae. Wide yeartoyearvariation in littoral communitycomposition was determined largely byprevailing weather conditions. Suchvariable conditions may cause oscillationsin the habitat coupling between littoraland pelagial zones, for instance, throughvariable recruitment and foragingbehaviour of fish.
A submerged macrophyte Elodeacanadensis hampered the growth andtoxin production of cyanobacteria andregulated phytoplanktonzooplanktoninteractions, although littoral ecosystemseven without macrophytes could beresilient and recover from instantaneous Ploading when cladoceran grazers wereabundant. The mechanisms underlyingthe control of cyanobacterial growth bymacrophytes proved to be complex andapparently involved direct and indirecteffects through allelopathy as well asaltered nutrient cycling which was alsoincluenced by heterotrophic processes ofthe microbial compartment. Moredetailed experiments and techniques areneeded to reveal such processes and theirrelative role in controlling phytoplanktongrowth in field conditions.
The results emphasised the generalimportance of consumer control overresource control in determiningzooplankton community structure inshallow lake systems. The results are inagreement with the general view that
(large) herbivores respond strongly topredator manipulations in aquatic foodwebs. However, different growth formsof submerged macrophytes were ofvariable value as refuges for zooplanktonagainst fish predation. Especially plantassociated and small euplanktoniccladocerans benefited from macrophyticrefuge and were able to controlphytoplankton when fish predationpressure was not high (< 2.54 fish m2 or< 4 g FW m2). The results suggested thatanalyses of major cladoceran species, notonly keystone species, with carefulconsiderations about their interactionswithin the community may provideinsights into littoral ecosystemfunctioning. The refuge value of differentmacrophytic species and growth forms ina fluctuating environment both in spaceand time varies, being a subject forfurther studies.
Roach preferred feeding oncladocerans even as they declined inabundance, but also used lesspreferred,mainly nonanimal food sources asalternative food. Switching of preypreference and, consequently, reductionin the predation pressure on cladoceranscould not, however, be evidenced. Thus,littoral subsidies may maintainomnivorous and planktivorous fishpopulations such as roach underconditions with low cladoceranabundance and this support may intensifyrather than dampen the strength of theinteractions between fish andzooplankton. In a long term, on the otherhand, fish assemblages and the fishmediated coupling of pelagic and littoralprocesses change with macrophyticvegetation also affecting planktondynamics. More evidence for such longterm patterns in littoral zones would bevaluable to increase our understanding ofecosystem functioning and couplingbetween littoral and pelagial underfluctuating environmental conditions.
Geographical differences in theresponses of zooplankton communities to
31
nutrient and fish manipulation wereprobably masked by wide yeartoyearvariation in the prevailing weatherconditions. The only consistentgeographical pattern was the reduced roleof large crustacean grazers in thesouthernmost location in Valenciacompared with the other sites. Thus,although food web management may be auseful measure in northern and temperatelocations, nutrient control may be moreimportant in southern lakes. Thisconclusion needs further researchespecially in warm weather conditions inorder to predict how climate changeshould be taken into consideration whenplanning restoration measures foreutrophicated lakes.
The phytoplankton biomass waspositively related to TP concentrationeven in the presence of efficient grazers.This result reflects the positive responseof adjacent trophic levels and theultimate, quantitatively important role ofproductivity compared with consumer
regulation in determining the primaryproducer biomass. However, theabundance of large crustacean grazersexplained reasonably well the variancebetween productivity and algal biomass,even better than did the mere number oftrophic levels. This may be explained byconfounding factors such as heterogeneitywithin the habitat and trophic levels,often difficult to disentangle from eachother.
In summation, the abundance ofzooplankton grazers and macrophyteswas of crucial importance in determiningthe stability of the littoral ecosystemevidenced as high resistance to nutrientloading. In northern temperate eutrophiclakes, this beneficial community structuremay be achieved and maintained bycontrolling the planktivorous andomnivorous fish populations that aresupported by both pelagic and littoralresources and may contribute to thetrophic state throughout the lake.
6. ACKNOWLEDGEMENTS
Analogously to food web theory,belief in the accomplishment of this workhas varied, exhibiting relativelydiscontinuous, alternating (high/low/high)changes with the increase in the amountof samples, Excel sheets, variousstatistical analyses and manuscriptversions. Apart from constructing fieldexperiments and interpreting theircomplicated results, the SIL congress,among other things, showed thatpractically nothing is impossible – not
even stepping across the threshold ofmaking a thesis.
I thank my supervisor TimoKairesalo for his patience and for sharinghis wide knowledge of aquatic, especiallylittoral ecosystems. Jouko Sarvala andMari Walls reviewed the thesis and gavevaluable comments. I appreciate allwonderful colleagues, both present andpast, at the Research and EducationCenter of Lahti and, later, at theDepartment of Ecological andEnvironmental Sciences for theirinvaluable support during the differentstages of this work. I warmly rememberthe cooperation and insightful discussions
32
with Jukka Horppila, Eira Luokkanen,Ryszard Kornijów, Jaana Hietala, HeliKarjalainen, Heikki Setälä, RauniStrömmer, Mirva Nykänen, Anne Ojala,Erika Heikkinen, Tiina Käki, AnuVäisänen, Leena Nurminen and SakuAnttila – just to mention a few. Thelimnological phenomena of LakeVesijärvi as well as diverse matterscovering environmental management andvarious aspects of life have been subjectsof fruitful discussions with Juha Keto, towhom I am most grateful. I warmlyremember the collaboration with theSWALE people, especially the splendidcoordinator prof. Brian Moss.
I wish to acknowledge the technicalstaff of the City of Lahti and of theDepartment of Ecological andEnvironmental Sciences for theirinvaluable contribution to theestablishment of the studies. Easy accessto Kilpiäistenpohja Bay and the wellequipped facilities in Neopoli contributedto many practical matters. This studywould not have been possible without the
financial support from the Academy ofFinland, the European Commission, LakeVesijärvi II project and the Departmentof Ecological and EnvironmentalSciences.
The spirit was born with my earlyhydrobiological experiences at LammiBiological Station, among its warmhearted people and atmosphere.Experimental manipulations andzooplanktonoriented studies became mydestiny. Special thanks to Marko Järvinenand Martti Rask for immemorialcollaboration.
I wish to express my deepestgratitude to my family, Juha, Aino andAapo for your patience andunderstanding. Juha also helped me toedit pictures and taught me how to usePhotoShop. Relatives, especially Riitta,Maare and Antero, as well as neighboursand friends have given their greatsupport. Thank you so much! Finally, this“drawing book” (värityskirja, as 5yearold Aapo calls it) is finished.
7. REFERENCES
Abrams, P. A. & Walters, C. J. 1996:Invulnerable prey and the paradox ofenrichment. –Ecology 77: 11251133.
Andersen, T. & Hessen, D. O. 1991: Carbon,nitrogen and phosphorus content offreshwater zooplankton. – Limnol.Oceanogr. 36: 807814.
Anderson, T. R., D. O. Hessen, J. J. Elser, andJ. Urabe. 2005: Metabolic stoichiometryand the fate of excess carbon and nutrientsin consumers. –Am. Nat. 165: 115.
Arditi, R. & Ginzburg, L.R. 1989: Coupling inpredatorprey dynamics: ratiodependence.–J. Theor. Biol. 139: 6975.
Balayla, D. J. & Moss, B. 2004: Relativeimportance of grazing on algae by plantassociated and openwater microcrustacea(Cladocera). – Arch. Hydrobiol. 161: 199224.
Beklioglu, M. & Moss, B. 1996: Mesocosmexperiments on the interaction of sedimentinfluence, fish predation and aquatic plantswith the structure of phytoplankton andzooplankton communities. –Freshw. Biol.36: 315325.
Benndorf, J., Boing, W., Koop, J. & Neubauer,I. 2002: Topdown control ofphytoplankton: the role of time scale, lakedepth and trophic state. –Freshw. Biol. 47:22822295.
Blindow, I., Hargeby, A., Wagner, B. M. A. &Andersson, G. 2000: How important is the
33
crustacean plankton for the maintenance ofwater clarity in shallow lakes withabundant submerged vegetation? –Freshw.Biol. 44: 185197.
Brett, M. T. & Goldman, C. R. 1997: Consumerversus resource control in freshwaterpelagic food webs. –Science 275: 384386.
Brooks, J. L. & Dodson, S. I. 1965: Predation,body size and composition of plankton. –Science 150: 2835.
Burks, R. L., Jeppesen, E. and Lodge, D. 2000:Macrophyte and fish chemicals suppressDaphnia growth and alter life history traits.–Oikos 88: 139147.
Burks, R. L., Lodge, D.M., Jeppesen, E. &Lauridsen, T. L. 2002: Diel horizontalmigration of zooplankton: costs andbenefits of inhabiting the littoral. –Freshw.Biol. 47: 343365.
Butorina, L. G. 1986: On the problem ofaggregations of planktonic crustaceans[Polyphemus pediculus (L.), Cladocera]. –Arch. Hydrobiol. 105: 355386.
Carpenter, S. R., Cole, J. J., Kitchell, J. F. &Pace, M. L. 1998: Impact of dissolvedorganic carbon, phosphorus and grazing onphytoplankton biomass and production inexperimental lakes. – Limnol. Oceanogr.43: 7380.
Carpenter, S. R. & Kitchell, J. F. (eds.) 1993:The trophic cascade in lakes. – CambridgeUniversity Press. 385 p.
Carpenter, S. R., Kitchell, J. F. & Hodgson, J.R. 1985: Cascading trophic interactionsand lake productivity. – BioScience 35:634639.
Carpenter, S. R., Kraft, C. E., Wright, R., He,X., Soranno, P. A. & Hodgson, J. R. 1992:Resilience and resistance of a lakephosphorus cycle before and after foodweb manipulation. – Am. Nat. 140: 781798.
Cottingham, K. L., S. Glaholt, and A. C.Brown. 2004: Zooplankton communitystructure affects how phytoplanktonrespond to nutrient pulses. – Ecology 85:158171.
Cyr, H. & Curtis, J. M. 1999: Zooplanktoncommunity size structure and taxonomiccomposition affect sizeselective grazing innatural communities. –Oecologia 118: 306315.
DeAngelis, D. L., Mulholland, P. J., Palumbo,A. V., Steinman, A. D., Huston, M. A. &Elwood, J. W. 1989: Nutrient dynamicsand foodweb stability. –Annu. Rev. Ecol.Syst. 20: 7195.
DeMelo, R., France, R. & McQueen, D. J.1992: Biomanipulation: hit or myth? –Limnol. Oceanogr. 37: 192207.
DeMott, W. R. 1986: The role of taste in foodselection by freshwater zooplankton. –Oecologia 69: 334340.
Diehl, S. 1988: Foraging efficiency of threefreshwater fishes: effects of structuralcomplexity and light. –Oikos 53: 207214.
Diehl, S. 1993: Effects of habitat structure onresource availability, diet and growth ofbenthivorous perch, Perca fluviatilis. –Oikos 67: 403414.
Diehl, S. & Feiβel, M. 2000: Effects ofenrichment on threelevel food chains withomnivory. –Am. Nat. 155: 200218.
Dorgelo, J. & Heykoop, M. 1985: Avoidance ofmacrophytes by Daphnia longispina. –Verh. Int. Verein. Limnol. 22: 33693372.
Drake, J.A., Huxel, G.R. & Hewitt, C.L. 1996:Microcosms as models for generating andtesting community theory. – Ecology 77:670677.
FernándezAláez, M., FernándezAláez, C.,Bécares, E., Valentin, M., Goma, J. &Castrillo, P. 2004: A 2year experimentalstudy on nutrient and predator incluenceson food web constituents in a shallow lakeof northwest Spain. – Freshw. Biol. 49:15741592.
Folke, C., Carpenter, S., Walker, B., Scheffer,M., Elmquist, T., Gunderson, L. & Holling,C. S. 2004: Regime shifts, resilience andbiodiversity in ecosystem management. –Annu. Rev. Ecol. Syst. 35: 557581.
Fretwell, S. D. 1977: The regulation of plantcommunities by food chains exploitingthem. – Perspectives in Biology andMedicine 20: 169185.
GenkaiKato, M. & Carpenter, S. R. 2005:Eutrophication due to phosphorus recyclingin relation to lake morphometry,temperature and macrophytes. – Ecology86: 210219.
Gliwicz, Z. M. 2002: On the different nature oftopdown and bottomup effects in pelagicfood webs. –Fresw. Biol. 47: 22962312.
Gross, E. M. 2003: Allelopathy of aquaticautotrophs. –Crit. Rev. Plant Sci. 22: 313339.
Gustafsson, S. & Hansson, L.A. 2004:Development of tolerance against toxiccyanobacteria in Daphnia. – Aquat. Ecol.38: 3744.
Hairston, N. G., Smith, F. E. & Slobodkin, L.B. 1960: Community structure, population
34
control and competition. – Am. Nat. 94:421425.
Hansson, L.A. 1992: The role of food chaincomposition and nutrient availability inshaping algal biomass development. –Ecology 73: 241247.
Hansson, L.A., Annadotter, H., Bergman, E.,Hamrin, S. F., Jeppesen, E., Kairesalo, T.,Luokkanen, E., Nilsson, P.Å., Sndergaard,Ma., & Strand, J. 1998a: Biomanipulationas an application of foodchain theory:constraints, synthesis andrecommendations for temperate lakes. –Ecosystems 1: 558574.
Hansson, L.A., Brönmark, C., Nyström, P.,Greenberg, L., Lundberg, P., Nilsson, P.A., Persson, A., Pettersson, L. B., Romare,P. & Tranvik, L. J. 1998b: Consumptionpatterns, complexity and enrichment inaquatic food chains. –Proc. R. Soc. Lond.265: 901906.
Hargeby, A., Blindow, I. & Hansson, L.A.2004: Shifts between clear and turbid statesin a shallow lake: multicausal stress fromclimate, nutrients and biotic interactions. –Arch. Hydrobiol. 161: 433454.
Hasler, A. D. & Jones, E. 1949: Demonstrationof the antagonistic action of large aquaticplants on algae and rotifers. –Ecology 30:359364.
Helminen, H. & Sarvala, J. 1997: Responses ofLake Pyhäjärvi (southwestern Finland) tovariable recruitment of the majorplanktivorous fish, vendace (Coregonusalbula). –Can J. Fish. Aquat. Sci. 54: 3240.
Hessen, D. O. & Lyche, A. 1992: Inter andintraspecific variations in zooplanktonelement composition. – Arch. Hydrobiol.121: 343353.
Hietala, J., Reinikainen M. & Walls, M. 1995:Variation in life history responses ofDaphnia to toxic Microcystis aeruginosa. –J. Plankton Res. 17: 23072318.
Holopainen, I. J., Tonn, W. M., Paszkowski, C.A. 1992: Effects of fish density onplanktonic communities and water qualityin a manipulated forest pond. –Hydrobiologia 243/244: 311321.
Horppila, J. 1994: The diet and growth of roach(Rutilus rutilus (L.)) in Lake Vesijärvi andpossible changes in the course ofbiomanipulation. –Hydrobiologia 294: 3541.
Horppila, J. & Kairesalo, T. 1992: Impact ofbleak (Alburnus alburnus) and roach(Rutilus rutilus) on the quality,
sedimentation and internal nutrient loading.–Hydrobiologia 243/244: 323331.
Horppila, J. & Nurminen, L. 2003. Effects ofsubmerged macrophytes on sedimentresuspension and internal phosphorusloading in Lake Hiidenvesi (southernFinland). –Water Research 37: 44684474.
Horppila, J., Peltonen, H., Malinen, T.,Luokkanen, E. & Kairesalo, T. 1998: Topdown or bottomup effects of fish: issues ofconcern in biomanipulation of lakes. –Restoration Ecology 6: 2028.
Horppila, J., Ruuhijärvi, J., Rask, M.,Karppinen, C., Nyberg, K. & Olin, M.2000: Seasonal changes in the diets andrelative abundances of perch and roach inthe littoral and pelagic zones of a largelake. –J. Fish Biol. 56: 5172.
Horton, P. A., Rowan, M., Webster, K. E. &Peters, R. H. 1979: Browsing and grazingby cladoceran filter feeders. –Can. J. Zool.57: 206212.
Hrbácek, J., Dvorakova, M., Korinek, V. &Prochazkova, L. 1961: Demonstration ofthe effect of the fish stock on the speciescompostion of zooplankton and theintensity of metabolism of the wholeplankton association. – Verh. Int. Verein.Limnol. 14: 192195.
Hudson, J. J., Taylor, W. & Schindler, D. W.1999: Planktonic nutrient regeneration andcycling efficiency in temperate lakes. –Nature 400: 659661.
Hulot F. D., Lacroix G., LescherMoutoué F. &Loreau M. 2000: Functional diversitygoverns ecosystem response to nutrientenrichment. –Nature 405: 340344.
Hupfer, M. & Dollan, A. 2003: Immobilisation ofphosphorus by ironcoated roots ofsubmerged macrophytes. – Hydrobiologia506509: 635640.
Hutchinson, G. E. 1967: A treatise on limnology.Volume II. Introduction to lake biology andthe limnoplankton. – John Wiley & Sons,Inc, New York, pp. 1115.
Irvine, K., Moss, B. & Balls, H. 1989: The loss ofsubmerged plants with eutrophication II.Relationships between fish and zooplanktonon a set of experimental ponds, andconclusions. –Freshw. Biol. 22: 89107.
Jang, M.H., Ha, K., Joo, G.J. & Takamura, N.2003: Toxin production of cyanobacteria isincreased by exposure to zooplankton. –Freshw. Biol. 48: 15401550.
Jeppesen, E., Jensen, J. P., Jensen, C., Faafeng,B., Hessen, D. O., Søndergaard, Ma.,Lauridsen, T., Brettum, P. &
35
Christoffersen, K. 2003: The impact ofnutrient state and lake depth on topdowncontrol in the pelagic zone of lakes: a studyof 466 lakes from the temperate zone to thearctic. –Ecosystems 6: 313325.
Jeppesen, E., Jensen, J. P. & Søndergaard, Ma.2002a: Response of phytoplankton,zooplakton and fish to reoligotrophication:An 11 year study of 23 Danish lakes. –Aquatic Ecosystem Health & Management5: 3142.
Jeppesen, E., Jensen, J. P., Søndergaard, Ma. &Lauridsen, T. 1999: Trophic dynamics inturbid and clearwater lakes with specialemphasis on the role of zooplankton forwater quality. – Hydrobiologia 408/409:217231.
Jeppesen, E., Lauridsen, T. L., Kairesalo, T. &Perrow, M. 1998a: Impact of submergedmacrophytes on fishzooplanktoninteractions in lakes. – In: Jeppesen, E.,Søndergaard, Ma., Søndergaard, Mo. &Christoffersen, K. (eds.) The structuringrole of submerged macrophytes in lakes:91114. SpringerVerlag, New York.
Jeppesen, E., Søndergaard, M., Jensen, J. P.,Mortensen, E., Hansen, A.M. &Jorgensen, T. 1998b: Cascading trophicinteractions from fish to bacteria andnutrients after reduced sewage loading: An18year study of a shallow hypertrophiclake. –Ecosystems 1: 250267.
Jeppesen, E., Søndergaard, M., Jensen, J.P.,Mortensen, E. & Sortkjaer, O. 1996:Fishinduced changes in zooplanktongrazing on phytoplankton andbacterioplankton: a longterm study inshallow hypertrophic Lake Søbygaard. –J.Plankton Res. 9: 16051625.
Jeppesen, E., Søndergaard, Ma., Søndergaard,Mo. & Christoffersen, K. (eds.) 1998c: Thestructuring role of submerged macrophytesin lakes. – SpringerVerlag, New York.423 p.
Jeppesen, E., Søndergaard, Ma., Søndergaard,Mo., Christoffersen, K., TheilNielsen, J. &Jürgens, K. 2002b: Cascading trophicinteractions in the littoral zone: anenclosure experiment in shallow LakeStigsholm, Denmark. – Arch. Hydrobiol.153: 533555.
Jones, CG & Lawton, JH. (eds.) 1995. Linkingspecies & ecosystems. Chapman & Hall.335 s.
Jones, J. I. & Sayer, C. D. 2003: Does the fishinvertebrateperiphyton cascade precipitate
plant loss in shallow lakes? –Ecology 84:21552167.
Järvinen, M. & Salonen, K. 1998: Influence ofchanging food web structure on nutrientlimitation of phytoplankton in a highlyhumic lake. –Can. J. Fish. Aquat. Sci. 55:25622571.
Kairesalo, T. 1980: Diurnal fluctuations withina littoral plankton community inoligotrophic Lake Pääjärvi, southernFinland. –Fresw. Biol. 10: 533537.
Kairesalo, T., Kornijów, R. & Luokkanen, E.2000: Trophic cascade structuring aplankton community in a stronglyvegetated lake littoral. –Verh. Int. Verein.Limnol. 27: 17631767.
Kairesalo T., Laine, S, Luokkanen, E., Malinen,T. & Keto, J. 1999: Direct and indirectmechanisms behind successfulbiomanipulation. –Hydrobiologia 395/396:99106.
Kairesalo, T., Tátrai, I. & Luokkanen, E. 1998:Impacts of water weed (Elodea canadensisMichx) on planktonfish interactions in thelake littoral. – Verh. Int. Verein. Limnol.26: 18461851.
Kairesalo, T. & Vakkilainen, K. 2004: LakeVesijärvi and the city of Lahti (southernFinland): comprehensive interactionsbetween the lake and the coupled humancommunity. –SILnews 41: 15.
Kankaala, P. 1988: The relative importance ofalgae and bacteria as food for Daphnialongispina (Cladocera) in a polyhumiclake. –Freshw. Biol. 19: 285296.
Kerfoot, W. C. & Sih, A. (eds.) 1987:Predation. Direct and indirect impacts onaquatic communities. – University Press ofNew England. 386 p.
Keto, J. & Sammalkorpi, I. 1988: A fadingrecovery: A conceptual model for LakeVesijärvi management and research. –Aqua Fennica 18: 193204.
Keto, J. & Tallberg, P. 2000: The recovery ofVesijärvi, a lake in southern Finland: waterquality and phytoplankton interpretations.–Bor. Env. Res. 5: 1526.
Kornijów, R. 1998: Quantitative sampler forcollecting invertebrates associated withsubmersed and floatingleaved macrophytes.–Aquat. Ecol. 32: 241244.
Kornijów, R. & Kairesalo, T. 1994: Elodeacanadensis sustains rich environment formacroinvertebrates. – Verh. Int. Verein.Limnol. 25: 22702275.
KoskiVähälä, J., H. Hartikainen, and T.Kairesalo. 2000. Resuspension in regulating
36
sedimentation dynamics in Lake Vesijärvi. –Archiv für Hydrobiologie 148: 357381.
Lampert, W. 1987: Laboratory studies onzooplanktoncyanobacteria interactions. –New Zealand Journal of Marine andFreshwater Research 21: 483490.
Larsson, P, & Dodson, S. 1993: Chemicalcommunication in planktonic animals. –Arch. Hydrobiol. 129: 129155.
Lauridsen, T. L. & Lodge, D. M. 1996:Avoidance by Daphnia magna of fish andmacrophytes: Chemical cues and predatormediated use of macrophyte habitat. –Limnol. Oceanogr. 41: 794798.
Lauridsen. T., Pedersen, L. J., Jeppesen, E. &Søndergaard, Ma. 1996: The importance ofmacrophyte bed size for cladocerancomposition and horizontal migration in ashallow lake. –J. Plankton Res. 18: 22832294.
Lehman, J.T. 1988: Ecological principlesaffecting community structure andsecondary productivity by zooplankton inmarine and freshwater environments. –Limnol. Oceanogr. 33: 931945.
Leibold, M. A., Chase, J. M., Shurin, J. B. &Downing, A. L. 1997: Species turnover andthe regulation of trophic structure. –Annu.Rev. Ecol. Syst. 28: 467494.
Lövgren, J. & Persson, L. 2002: Fishmediatedindirect effects in a littoral food web. –Oikos 96: 150156.
Mazumder, A. 1994: Phosphoruschlorophyllrelationships under contrasting herbivoryand thermal stratification: predictions andpatterns. –Can. J. Aquat. Sci. 51: 390400.
McCauley, E. & Murdoch, W. W. 1987: Cyclicand stable populations: plankton asparadigm. –Am. Nat. 129: 97121.
McCay, N. A. & Elser, J. J. 1998: Nutrientrecycling by Daphnia reduces N2 fixationby cyanobacteria. –Limnol. Oceanogr. 43:347354.
McQueen, D. J., Post, J. R. & Mills, E. L. 1986:Trophic relationships in freshwater pelagicecosystems. –Can. J. Fish. Aquat. Sci. 43:15711581.
Mehner, T., Benndorf, J., Kasprzak, P. &Koshel, R. 2002: Biomanipulation of lakeecosystems: successful applications andexpanding complexity in the underlyingscience. –Freshw. Biol. 47: 24532466.
Meijer, ML., de Boois, I., Scheffer, M.,Portielje, R. & Hosper, H. 1999:Biomanipulation in shallow lakes in TheNetherlands: an evaluation of 18 casestudies. –Hydrobiologia 408/409: 1330.
Moore, M. V., Folt, C. L. & Stemberger, R. S.1996: Consequences of elevatedtemperatures for zooplankton assemblagesin temperate lakes. –Arch. Hydrobiol. 135:289319.
Moss, B., Kornijów, R. & Measey, G. J. 1998:The effect of nymphaeid (Nuphar lutea)density and predation by perch (Percafluviatilis) on the zooplankton communitiesin a shallow lake. –Freshw. Biol. 39: 689697.
Moss, B., Madgwick, J. & Phillips, G. 1996: Aguide to the restoration of nutrientenriched shallow lakes. EnvironmentAgency, Broads Authority & EuropeanUnion Life Programme, Norwich, U.K.
Moss, B., Stephen, D., Balayla, D. M., Bécares,E., Collings, S. E., FernándezAláez, C.,FeernándezAláez, M., Ferriol, C., Garcia,P., Gomá, J., Gyllström, M., Hansson, L.A., Hietala, J., Kairesalo, T., Miracle, M.R., Romo, S., Rueda, J., Russell, V., StåhlDelbanco, A., Svensson, M., Vakkilainen,K., Valentin, M., Van de Bund, W. J., VanDonk, E., Vicente, E. & Villena, M. J.2004: Continentalscale patterns of nutrientand fish effects on shallow lakes: synthesisof a panEuropean mesocosm experiment.–Fresw. Biol. 49: 16331649.
Mourelatos, S. & Lacroix, G. 1990: In situfiltering rates of Cladocera: effect of bodylength, temperature and food concentration.–Limnol. Oceanogr. 35: 1101111.
Murdoch, W. W. & Bence, J. 1987: Generalpredators and unstable prey populations. –In: Kerfoot, W. C. & Sih, A. (eds.)Predation. Direct and indirect impacts onaquatic communities: 1730. UniversityPress of New England.
Nurminen, L., Horppila, J. & Tallberg, P. 2001:Seasonal development of the cladoceranassemblage in a turbid lake: the role ofemergent macrophytes. Arch. Hydrobiol.151: 127140.
Oksanen, L., Fretwell, S. D., Arruda, J. &Niemelä, P. 1981: Exploitation ecosystemsin gradients of primary productivity. –Am.Nat. 118: 240261.
Osenberg C. W. & Mittelbach, G. G. 1996. Therelative importance of resource limitationand predator limitation in food chains – In:Polis, G. A. & Winemiller, K. O. (eds.)Food webs: integration of patterns anddynamics: 134148. Chapman and Hall,New York.
Pace, M. L., Cole, J. J. & Carpenter, S. R. 1998:Trophic cascades and compensation:
37
differential responses of mirozooplanktonin wholelake experiments. – Ecology 79:138152.
Pennak, R. W. 1973: Some evidence for aquaticmacrophytes as repellents for a limneticspecies of Daphnia. – Int. Revue Ges.Hydrobiol. 60: 569576.
Perrow, M. R., Jowitt, A. J. D., Stansfield, J. H.& Phillips, G. L. 1999: The practicalimportance of the interactions betweenfish, zooplakton and macrophytes inshallow lake restoration. – Hydrobiologia395/396: 199210.
Persson, L. 1986: Temperatureinduced shift inforaging ability in two fish species, roach(Rutilus rutilus) and perch (Percafluviatilis): implications for coexistencebetween poikilotherms. – J. Anim. Ecol.55: 829839.
Persson, L. 1987: Effects of habitat and seasonon competitive interactions between roach(Rutilus rutilus) and perch (Percafluviatilis). –Oecologia 73: 170177.
Persson, L., Bengtsson, J., Menge, B. A. &Power, M. E. 1996: Productivity andconsumer regulation – concepts, patternsand mechanisms. In: Polis, G. A. &Winemiller, K. O. (eds.) Food webs:integration of patterns and dynamics: 396434. Chapman and Hall, New York.
Persson, L. & Greenberg, L.A. 1990: Juvenilecompetitive bottlenecks: the perch (Percafluviatilis) – roach (Rutilus rutilus)interaction. –Ecology 71: 4456.
Persson, A., Hansson, L.A., Brönmark, C.,Lundberg, P., Pettersson, L. B., Greenberg,L., Nilsson, P. A., Nyström, P., Romare, P.& Tranvik, L. 2001: Effects of enrichmenton simple aquatic food webs. – Am. Nat.157: 654669.
Petchey, O. L., McPhearson, T., Casey, T. M.& Morin, P. J. 1999: Environmentalwarming alters foodweb structure andecosystem function. –Nature 402: 6972.
Polis, G. A. 1999: Why are parts of the worldgreen? Multiple factors controlproductivity and the distribution ofbiomass. –Oikos 86: 315.
Polis, G. A., Sears, A. L. W., Huxel, G. R.,Strong, D. R., Maron, J. 2000: When is atrophic cascade a trophic cascade? –TREE15: 473475.
Polis, G. A. & Winemiller, K. O. (eds.) 1996:Food webs: integration of patterns anddynamics. – Chapman and Hall, New York.434 p.
Power, M. E. 1992: Topdown and bottomupforces in food webs: do plants haveprimacy? –Ecology 73: 733746.
Rapala, J., Sivonen, K., Lyra, C. & Niemelä, S.I. 1997: Variation of microcystins,cyanobacterial hepatotoxins, in Anabaenaspp. as a function of growth stimuli. –Appl. Env. Microbiol. 63: 22062212.
Rask, M., Järvinen, M., Kuoppamäki, K. &Pöysä, H. 1996: Limnological responses tothe collapse of the perch population in asmall lake. – Ann. Zool. Fennici 33: 517524.
Reynolds, C. S. 1994: The ecological basis forthe successful biomanipulation of aquaticcommunities. – Arch. Hydrobiol. 130: 133.
Romo, S. Miracle, M. R., Villena, M.J., Rueda,J., Ferriol, C. & Vicente, E. 2004:Mesocosm experiments on nutrient andfish effects on shallow lake food webs in aMediterranean climate. –Freshw. Biol. 49:15631607.
Ruuhijärvi, J., Malinen, T., AlaOpas, P. &Tuomaala, A. 2005: Fish stocks of LakeVesijärvi – from nuisance to flourishingfishery in 15 years. – Verh. Int. Verein.Limnol. 29: in print.
Räike, A., Pietiläinen, O.P., Rekolainen, S.,Kauppila, P., Pitkänen, H, Niemi, J.,Raanteland, A. & Vuorenmaa, J. 2003:Trends of phosphorus, nitrogen andchlorophyll a concentrations in Finnishrivers and lakes in 19752000. – TheScience for the Total Environment 310: 4759.
Salonen, K., Jones, R. I., De Haan & James, M.1994: Radiotracer study of phosphorusuptake by plankton and redistribution in thewater column of a small humic lake. –Limnol. Oceanogr. 39: 6983.
Salonen, K., Järvinen, M., Kuoppamäki, K. &Arvola, L. 1990: Effects of liming on thechemistry and biology of a small humiclake. – In: Kauppi, P., Anttila, P. &Kenttämies, K. (eds.) Acidification inFinland: 11471167. SpringerVerlag,Berlin, Heidelberg.
Sarnelle, O. 1992: Nutrient enrichment andgrazer effects on phytoplankton in lakes. –Ecology 73: 551560.
Sarvala, J., Helminen, H., Saarikari, V.,Salonen, S. & Vuorio, K. 1998: Relationsbetween planktivorous fish abundance,zooplankton and phytoplankton in threelakes of differing productivity. Hydrobiologia 363: 8195.
38
Sarvala, J., Kairesalo, T., Koskimies, I.,Lehtovaara, A., Ruuhijärvi, J. & VähäPiikkiö, I. 1982: Carbon, phosphorus andnitrogen budgets of the littoral Equisetumbelt in an oligotrophic lake. –Hydrobiologia 86: 4153.
Sarvala, J., Ventelä, A.M., Helminen, H.,Hirvonen, A., Saarikari, V., Salonen, S.,Sydänoja, A. & Vuorio, K. 2000:Restoration of the eutrophicatedKöyliönjärvi (SW Finland) through fishremoval: wholelake vs. mesocosmexperiences. –Bor. Env. Res. 5: 3952.
Scheffer, M. 1991: Fish and nutrients interplaydetermines algal biomass: a minimalmodel. –Oikos 62: 271282.
Scheffer, M. 1999: The effect of aquaticvegetation on turbidity: how important arethe filter feeders? – Hydrobiologia408/409: 307316.
Scheffer, M. & Carpenter, S.R. 2003:Catastrophic regime shifts in ecosystems:linking theory to observation. – TREE 18,648656.
Scheffer, M., Hosper, S. H., Meijer, ML.,Moss, B. & Jeppesen, E. 1993: Alternativeequilibria in shallow lakes. –TREE 8: 275279.
Scheffer, M., Straile, D., Van Nes, E. H. &Hosper, H. 2001: Climatic warming causesregime shifts in lake food webs. –Limnol.Oceanogr. 46: 17801783.
Schindler, D. E. & Scheuerell, M. D. 2002:Habitat coupling in lake ecosystems. –Oikos 98: 177189.
Schindler, D. W. 1990: Experimentalperturbations of whole lakes as tests ofhypotheses concerning ecosystem structureand function. –Oikos 57: 2541.
Schriver, P., Bøgestrand, J., Jeppesen, E. &Søndergaard, Ma. 1995: Impact ofsubmerged macrophytes on fishzooplanktonphytoplankton interaction:largescale enclosure experiments in ashallow eutrophic lake. –Freshw. Biol. 33:255270.
Shurin, J. B, Borer, E. T., Seabloom, E. W.,Anderson, K., Blanchette, C. A., Broitman,B, Cooper, S. D. & Halpern, B. S. 2002: Acrossecosystem comparison of the strengthof trophic cascades. – Ecology Letters 5:785791.
Shuter, B. J. & Ing, K. K. 1997: Factorsaffecting the production of zooplankton inlakes. –Can. J. Fish. Aquat. Sci. 54: 359377.
Smiley, E. A. & Tessier, A. J. 1998:Environmental gradients and the horizontaldistribution of microcrustaceans in lakes. –Fresw. Biol. 39: 397409.
Soranno, P. A., Carpenter, S. R. & He, X. 1993:Zooplankton biomass and body size. In: S.R. Carpenter & Kitchell, J. F. (Eds.) Thetrophic cascade in lakes: 172188.Cambridge University Press.
Stemberger, R. S. & Miller, E. K. 2003:Cladoceran body length and secchi disktransparency in northeastern U.S. lakes. –Can. J. Fish. Aquat. Sci. 60: 14771486.
Stephen, D., Balayla, D. M., Bécares, E.,Collings, S. E., FernándezAláez, C.,FernándezAláez, M., Ferriol, C., García,P., Gomá, J., Gyllström, M., Hansson, LA., Hietala, J., Kairesalo, T., Miracle, M.R., Romo, S., Rueda, J., StåhlDelbanco,A., Svensson, M., Vakkilainen, K.,Valentín, M., Van de Bund, W., Van Donk,E., Vicente, E., Villena, M. J. & Moss, B.2004: Continentalscale patterns of nutrientand fish effects on shallow lakes:introduction to a panEuropean mesocosmexperiment. – Freshw. Biol. 49: 15171524.
Søndergaard, Mo. & Moss, B. 1998: Impact ofsubmerged macrophytes on phytoplanktonin shallow freshwater lakes. – In: Jeppesen,E., Søndergaard, Ma., Søndergaard, Mo. &Christoffersen, K. (eds.) The structuringrole of submerged macrophytes in lakes:115132. SpringerVerlag, New York.
Søndergaard, Mo., J. TheilNielsen, K.Christoffersen, L. Schlüter, E. Jeppesen,and Ma. Søndergaard. 1998.Bacterioplankton and carbon turnover in adense macrophyte canopy. – In: Jeppesen,E., Søndergaard, Ma., Søndergaard, Mo. &Christoffersen, K. (eds.) The structuringrole of submerged macrophytes in lakes:250261. SpringerVerlag, New York.
Tarvainen, M., Sarvala, J. & Helminen, H.2002: The role of phosphorus release byroach [Rutilus rutilus (L.)] in the waterquality changes of a biomanipulated lake.Freshw. Biol. 47: 23252336.
Timms, R.M. & Moss, B. 1984: Prevention ofgrowth of potentially dense phytoplanktonpopulations by zooplankton grazing, in thepresence of zooplanktivorous fish in ashallow wetland ecosystem. – Limnol.Oceanogr. 29: 472286.
Townsend, C.R., Winfield, I. J., Peirson, G. &Cryer, M. 1986: The response of youngroach Rutilus rutilus to seasonal changes in
39
abundance of microcrustacean prey: a fielddemonstration of switching. – Oikos 46:372378.
Vadeboncoeur, Y., Zanden, M. J. & Lodge, D.M. 2002: Putting the lake back together:reintegrating benthic pathways into lakefood web models. –BioScience 52: 4454.
Vakkilainen, K. & Kairesalo, T. 2005:Zooplankton community responses to therehabilitation of the eutrophic LakeVesijärvi, Finland: changes in thecladoceran body size in 19992003. –Verh.Int. Verein. Limnol. 29: 488490.
Walls, M., Rajasilta, M., Sarvala, J. & Salo, J.1990: Diel changes in horizontalmicrodistribution of littoral Cladocera. –Limnologica 20: 253258.
Van de Bund, W. & Van Donk, E. 2002: Shortterm and longterm effects ofzooplanktivorous fish removal in a shallowlake: a synthesis of 15 years of data fromLake Zwemlust. –Freshw. Biol. 47: 23802387.
Van de Bund, W. & Van Donk, E. 2004:Effects of fish and nutrient additions onfoodweb stability in a charophytedominated lake. –Freshw. Biol. 49: 15651573.
Van Donk, E. & Van de Bund, W. 2002:Impact of submerged macropytes includingcharophytes on phyto and zooplanktoncommunities: allelopathy versus othermechanisms. –Aquat. Bot. 72: 261274.
Vanni, M. J. 2002: Nutrient cycling by animals infreshwater ecosystems. – Annu. Rev. Ecol.Syst. 33: 341370.
Venetvaara, J. & Lammi, E. 1995: Vesijärvenkasvillisuuden nykytila ja viimeaikaisetmuutokset. (In Finnish; “The current stateand recent changes in the macrophyticvegetation of Lake Vesijärvi”) – In:Sammalkorpi, I., Keto, J., Kairesalo, T.,Luokkanen, E., Mäkelä, M., Vääriskoski, J.& Lammi, E. (eds.) Vesijärviprojekti 19871994: 101106. Vesi ja YmpäristöhallinnonJulkaisuja – sarja A, Finnish EnvironmentInstitute.
Wetzel, R. G. 1990: Landwater interfaces:Metabolic and limnological regulators. –Verh. Int. Verein. Limnol. 24: 624.
Wetzel, R. G. 2001: Limnology. Lake and riverecosystems. 3rd edition. Academic Press,NY, USA. 1006 p.
Wetzel, R. G. & Søndergaard, Mo. 1998: Roleof submerged macrophytes for themicrobial community and dynamics ofdissolved organic carbon in aquaticecosystems. – In: Jeppesen, E.,Søndergaard, Ma., Søndergaard, Mo. &Christoffersen, K. (eds.) The structuringrole of submerged macrophytes in lakes:133148. SpringerVerlag, New York.
Williams, A. E. & Moss, B. 2003: Effects ofdifferent fish species and biomass onplankton interactions in a shallow lake. –Hydrobiologia 49: 331346.
40