Feeding Ecology of European & North American Perch

HUMBOLDT-UNIVERSITÄT ZU BERLIN Faculty of Agriculture and Horticulture

“Comparative feeding ecology of Eurasian perch (Perca fluviatilis) and North American yellow perch (Perca flavescens)”

Master thesis in the study program: M.Sc. Fishery Science and Aquaculture

submitted by: Linzmaier, Stefan Markus

Supervisors:

PD Dr., Mehner, Thomas Department IV - Biology and Ecology of Fishes Leibniz-Institute of Freshwater Ecology and Inland Fisheries

Prof. Dr., Arlinghaus, Robert Department IV - Biology and Ecology of Fishes Leibniz-Institute of Freshwater Ecology and Inland Fisheries & Department of Animal Science - Integrative Fisheries Management Humboldt-Universität zu Berlin - Faculty of Agriculture and Horticulture

Berlin, 8.11.2013

Contents

Contents ...... i List of Abbreviations...... ii List of Figures ...... iii List of Tables...... v List of Appendices ...... vi List of Equations ...... vii 1 Introduction ...... 1 2 Eurasian Perch and Yellow Perch ...... 9 2.1.1 Systematics & Zoology ...... 9 2.1.2 Distribution ...... 10 3 Material & Methods ...... 11 3.1 Study Area ...... 11 3.1.1 American Lakes ...... 11 3.1.2 German Lakes ...... 12 3.2 Sample Collection ...... 15 3.2.1 Fish ...... 15 3.2.2 Benthic Invertebrates & Zooplankton ...... 15 3.3 Stable Isotopes: Sample Preparation and Analysis ...... 16 3.4 Statistical Analysis ...... 19 3.4.1 Breakpoint Analysis ...... 20 3.4.2 Trophic Position ...... 20 3.4.3 Trophic Niche Metrics ...... 24 3.4.4 Estimated Diet Composition ...... 25 4 Results ...... 28 4.1 Trophic Position & Breakpoint Analysis ...... 28 4.2 Trophic Niche Metrics ...... 33 4.3 SIAR - Mixing Model ...... 34 5 Discussion ...... 41 6 Conclusions ...... 50 7 Summary ...... 52 8 Zusammenfassung ...... 53 9 References ...... 54 10 Annex ...... 72 11 Acknowledgements ...... 80 12 Erklärung ...... 81

List of Abbreviations

YP yellow perch ha hectare (Perca flavescens) µm micrometer EP Eurasian perch mm millimeter (Perca fluviatilis) cm centimeter SI(s) stable isotope(s) m meter TP(s) trophic position(s) mg milligram TL total length g gram e.g. exempli gratia h hour etc. et cetera mL milliliter et al. et alii/aliae °C degree Celcius i.e. id est V-PDB Vienna PeeDee ZP zooplankton Belemnite MZB macrozoobenthos n sample size xC carbon isotope(s) SD standard deviation of mass x GLM generalized linear xN nitrogen isotope(s) model of mass x CD the mean distance Fig. figure to centroid Tab. table MNND the mean nearest Equ. equation neighbour distance YOY young of the year SDNND standard deviation of the nearest neighbour age-x+ in the year x distance after hatching DFO Department of Fisheries USA United States and Oceans (Canada) of America LELF Landesamt für GER Germany Ländliche Entwicklung, d measurement error Landwirtschaft und Flurneuordnung PA Lake Padden p probability value WA Walsh Lake F F-value WI Lake Wilderness df degrees of freedom DO Lake Dölln R2 coefficient VA Lake Väter of determination WU Lake Wucker max. maximum

iii

List of Figures

Figure 1: Schematic figure of the habitat shifts between the littoral and the pelagic habitat during the early ontogeny of P. fluviatilis & P. flavescens (modified from Whiteside et al. (1985))...... 4

Figure 2: Food web scheme for P. fluviatilis & P. flavescens with respect to ontogeny and energy pathways. Trophic links to aquatic consumers are illustrated with solid arrows. Diet shifts between size classes of perch are shown as dashed arrows. The scheme depicts (from top to bottom) the co-ocurring top predators (Esox lucius (upper picture) & Micropterus salmoides (lower picture), the different size classes of P. fluviatilis (upper pictures) & P. flavescens (lower pictures), their most important types of prey (zooplankton, benthic/epiphytic invertebrates and fish) and the bases of the littoral and pelagic food webs (phytoplankton, benthic & epiphytic algae) along their assumed trophic positions (modified from Rasmussen et al. 2008) – pictures: Wisconsin Department of Natural Resources, Bangladesh Fisheries Information, Fish Management in New York State & Fischereiverband für das Land Vorarlberg...... 5

Figure 3: Photographs of adult P. flavescens (a) & P. fluviatilis (b) – pictures: wikimedia.org...... 9

Figure 4: Aerial photographs and position of the studied lake ecosystems in North America and Europe (position and scale of the continents are not drawn to scale): Lake Padden (a), Walsh Lake (b), Lake Wilderness (c), Lake Dölln (d), Lake Väter (e) & Lake Wucker (f) - pictures: Google Maps ...... 14

Figure 5: Preparation of fish tissue samples: Weighing and measuring (a); tissue removal of the dorsal muscle (b); drying of samples in a drying oven at 60°C for 24 h (c); grinding the samples with mortar and pestle (d)...... 17

Figure 6: An example of lake corrected baselines as they are used in this study to depict the δ15N-δ13C relationship in the context of habitats and the consequences for consumer trophic position (TP - i.e. the average position relative to primary producers at which an organism feeds). Primary consumers represent a TP of 2. In spite of similar δ15N-values, sculpins occupy lower TP due to their diet of 13C-depleted profundal prey (modified from Vander Zanden & Rasmussen) - picture of YP: Maryland Dept. of Natural Resources. 22

Figure 7: Trophic positions (TP - i.e. the average position relative to primary producers at which an organism feeds) of P. fluviatilis (green triangles) & P. flavescens (yellow triangles) at increasing total length (TL) and regression of TP and TL with breakpoints (* = significant) in TPs for P. fluviatilis (dotted, green line) & P. flavescens (dotted, yellow line) for all lakes combined...... 28

Figure 8: Box-and-whisker plots of observed mean trophic positions (TP - i.e. the average position relative to primary producers at which an organism feeds) of small (a) and large (b) P. fluviatilis (green) & P. flavescens (yellow). Outliers are plotted as dots with given total length (TL)...... 29 iv

Figure 9: Linear relationships between isotopic signatures (δ15N & δ13C) of muscle tissue from P. flavescens (YP) and total length (TL) in Lake Padden (a & b), Walsh Lake (c & d) and Lake Wilderness (e & f). Yellow triangles represent samples from YP, caught with gill nets, yellow-green triangles are YP caught with hoop nets and brown triangles are YP from the 2009 dataset collected by Eric Larson (University of Washington, unpublished data)...... 32

Figure 10: Linear Relationships between isotopic signatures (δ15N & δ13C) of muscle tissue from P. fluviatilis (EP) and total length (TL) in Lake Dölln (a & b), Lake Väter (c & d) and Lake Wucker (e & f). Dark green triangles represent EP caught with gill nets, light green triangles are EP caught with traps or electrofishing...... 33

Figure 11: Bi-plots of isotopic signatures (δ15N & δ13C) of muscle tissue from P. flavescens (YP) and potential prey items in Lake Padden (a), Walsh Lake (b) and Lake Wilderness (c). YP are grouped by body length (small & large) and by value clusters (1/2), representing groups of similar isotopic composition. The hull area of each size class is outlined by yellow (small) and brown (large) dashed lines. The black dashed line indicates the baseline of primary consumers, which was used for calculating trophic positions of YP. Triangles represent fish, diamonds zooplankton and circles benthic invertebrates...... 36

Figure 12: Bi-plots of isotopic signatures (δ15N & δ13C) of muscle tissue from P. fluviatilis (EP) and potential prey items in Lake Dölln (a), Lake Väter (b) and Lake Wucker (c). EP are grouped by body length (small & large) and by value clusters (1/2), representing groups of similar isotopic composition. The hull area of each size class is outlined by light green (small) and dark green (large) dashed lines. The black dashed line indicates the baseline of primary consumers, which was used for calculating trophic positions of EP. Triangles represent fish, diamonds zooplankton and circles benthic invertebrates...... 38

Figure 13: Results of the SIAR (stable isotope analysis in R) mixing model as mean percentage of zooplankton (yellow), zoobenthos (brown) and fish (green) from the resource use of small and large P. flavescens & P. fluviatilis for all lakes, inferred from stable isotopes. Sample size (n) is given in brackets...... 39

Figure 14: Percentage of zooplankton (yellow), zoobenthos (brown) and fish (green) in the diet of small (a) and large (b) P. flavescens & P. fluviatilis for each lake (PA = Lake Padden, WA = Walsh Lake, WI = Lake Wilderness, DO = Lake Dölln, VA = Lake Väter & WU = Lake Wucker). The diet within each size class is given for each group (1/2) with similar isotopic composition, as defined in preliminary cluster analysis. Sample size (n) is given in brackets...... 40

List of Tables

Table 1: Limnological characteristics and water quality parameters of the studied lakes, measured in June/July (USA) and October/November (Germany). Chemical parameters are the average of one epi- and one hypolimnic water sample per lake. The fish community comprises all fish species caught during sampling in the USA and Germany...... 13

Table 2: Point estimates of the intercepts and point estimates, standard errors (st.err), t- values and confidence intervals of the slopes of each segmented relationship in the fitted model of trophic position and total length for P. flavescens & P. fluviatilis before (reg. 1) and after the breakpoint (reg. 2)...... 28

Table 3: Estimated marginal means, standard error (SE) and 95%-confidence interval (CI) of species trophic position (TP - i.e., the average position relative to primary producers at which an organism feeds) of P. flavescens & P. fluviatilis for each lake and for all sampled small and large perch (bold)...... 30

Table 4: Means ± standard deviation (SD) of total length (TL), trophic position (TP), δ13C- and δ15N-isotopic signatures of muscle tissue from P. flavescens & P. fluviatilis in Lakes of Washington (USA) and Brandenburg (Germany) with the corresponding number of samples (n). TP is given with sample variance (s2)...... 31

Table 5: Trophic niche metrics of P. flavescens & P. fluviatilis: δ15N-range, δ13C-range, hull area, mean distance to centroid (CD), mean nearest neighbour distance (MNND) and standard deviation of the nearest neighbour distance (SDNND)) for each species and size class for each lake...... 34

List of Appendices

Tables

Table A 2: List of samples used for the SIAR mixing model for each lake. Composite samples are marked grey. Every species or taxon is given as identified with the number of analyzed samples (n) and the diet group (type: ZP = zooplankton; MZB = zoobenthos; F = fish). Notes are in parentheses...... 73

Figures

Figure A 1: Results of the SIAR mixing model with percentage mean, median and 95%- confidence intervals for potential contributions of different food source to the isotopic composition of muscle tissue from small and large P. flavescens within Lake Padden (PA - a, b, e & f), Walsh Lake (WA - g & h) and Lake Wilderness (WI – c & d). Contributions are calculated for groups of similar isotopic composition (1/2)...... 76

Figure A 2: Results of the SIAR mixing model with mean percentage, median and 95%- confidence intervals for potential contributions of different food sources to the isotopic composition of muscle tissue from small and large P. fluviatilis within Lake Dölln (DO – a, b & e), Lake Väter (VA – c, d, f & g) and Lake Wucker (WU - h & i). Contributions are calculated for groups of similar isotopic composition (1/2)...... 79

vii

List of Equations

Equation 1: Formula for calculating delta (δ) values from mass ratios of 12C & 13C (R). .. 18

( 푅sample-푅standard) 훿13퐶 = × 1000 푅standard

Equation 2: Formula for calculating delta (δ) values from mass ratios of 14N & 15N (R). .. 18

( 푅sample-푅standard) 훿15푁 = × 1000 푅standard

Equation 3: General formula for calculating the trophic position (TP) of a consumers. .... 21

(훿15푁 - 훿15푁 ) 푡푟표푝ℎ𝑖푐 푝표푠𝑖푡𝑖표푛 = [ consumer baseline ] + λ consumer Δ15N

Equation 4: General formula for calculating the baseline curve derived from primary 15 13 consumer stable isotope (SI) composition (δ Nprim.con. & δ Cprim.con.)...... 22

15 6.43 훿 푁prim.con. = 13 1 + exp[9.67 + (0.356*훿 퐶prim.con.)]

Equation 5: Formula for calculating the residuals of primary consumer stable isotope (SI) 15 13 composition (δ Nprim.con. & δ Cprim.con.)...... 23

15 6.43 푟푒푠𝑖푑푢푎푙 = 훿 푁prim.con.- 13 1 + exp[9.67 + (0.356*훿 퐶prim.con.)]

15 15 Equation 6: Formula for calculating a lake corrected baseline value of δ N (δ Ncorrected). 23

15 6.43 훿 푁corrected. = 13 + 푈resid 1 + exp[9.67 + (0.356*훿 퐶consumer)]

Equation 7: Formula for calculating the trophic positions (TPs) of consumer δ15N composition with respect to lake-corrected primary consumer baselines...... 23

(훿15푁 − 훿15푁 ) 푡푟표푝ℎ𝑖푐 푝표푠𝑖푡𝑖표푛 = [ consumer correct ] + λ consumer Δ15N

1 Introduction

Fishes regulate aquatic food webs by feeding on various organisms, providing profound ecosystem services (Holmlund and Hammer 1999). Investigating food webs calls for detailed information on fish feeding ecology or trophic ecology, which helps to explain structural and functional aspects of ecosystems (Vander Zanden and Vadeboncoeur 2002, Pasquaud et al. 2008). It is crucial to understand community dynamics and species assemblages, especially for economically and ecologically important species (Link and Garrison 2002, Vander Zanden and Vadeboncoeur 2002, Heath 2005). For example, fishes can influence the activity, distribution and abundance of their prey (Carpenter et al. 1987, Wellborn et al. 1996). Moreover, fishes are major members of complex trophic cascades within aquatic ecosystems that influence food webs through top-down and bottom-up controls (Carpenter et al. 1985, Pace et al. 1999, Frank et al. 2005). For instance, piscivory or planktivory can cause cascading trophic effects in lake ecosystems (e.g. Simon and Townsend 2003, Reissig et al. 2006, Bystrom et al. 2007).

Several concepts can help to describe the ecological role of a species. The functional niche and the trophic level or linkages (interactions with other organisms) are key concepts in ecology as they depict resource use and the functional roles of fishes as consumers in aquatic food webs (Paine 1988, Vander Zanden and Vadeboncoeur 2002, Chase and Leibold 2003). The Eltonian or functional niche concept is used to delineate the functional role of a species in the community and its potential resource use (Colwell and Rangel 2009, Devictor et al. 2010). The trophic level concept represents an energy- weighted mean path length leading to a consumer (sensu Vander Zanden and Rasmussen 1999, p. 406). Furthermore, the trophic position (TP), precisely ranks a species in the trophic order (e.g. predators have a trophic level of four, but predator X has a TP of 4.3) (Benke 2011). Such trophic positions are useful to elucidate food web interactions and functionally characterize the resource use (i.e. realized niche) of a species (Post 2002). Moreover, this concept allows for describing cascading trophic interactions, energy flows, the accumulation of pollutants and the horizontal biodiversity of communities (Power 1992, Hairston 1993, Rasmussen et al. 2008, Reiss et al. 2009).

Fishes occupy a range of different functional niches and TPs in aquatic ecosystems, which differ among species as well as among individuals (Helfman et al. 2009, 424 pp). The 2 resource use of fish can generally be described as its dietary guild or feeding strategy (Garrison and Link 2000, Williams and Martinez 2000). Many different feeding strategies apply to fish: phytophagous, detritivorous, planktivorous, benthivorous, piscivorous and omnivorous fish feed at different parts of the food web (i.e. trophic levels). However, changes in food types during ontogeny affect and complicate the feeding ecology of a single species (Werner and Gilliam 1984). Such changes are usually related to increasing gape size, allowing fish to exploit progressively larger prey (Karpouzi and Stergiou 2003). Almost all fish start to feed on zooplankton (ZP) and then undergo ontogenetic diet or niche shifts (Werner and Gilliam 1984). For example, predatory fish shift from zooplanctivory to benthivory to piscivory (Olson 1996). Consequently, individuals in size structured populations can only be assigned to a specific feeding strategy at a certain size (de Roos and Persson 2001, de Roos et al. 2002). Size classes within fish species often form ecologically distinct groups, which differently affect population dynamics and community structure (Werner and Gilliam 1984, Polis et al. 1989, Polis and Strong 1996). The shifts in feeding strategy, whether ontogenetically or environmentally determined, are tantamount to shifts in trophic levels or functional niches (Osenberg et al. 1992, Wainright et al. 1993, de la Moriniere et al. 2003). The growth-related increase in the TP of many fish can even turn interaction with other species from competition to predation (Olson et al. 1995). In addition, the TP and niche space can be altered by several factors, such as interspecific or intraspecific competition (Vander Zanden et al. 1999a, Svanback and Bolnick 2007).

The Eurasian perch (EP - Perca flavescens) and the American yellow perch (YP - Perca fluviatilis) represent good model species for feeding ecology, because many of their traits and characteristics have been studied: diet (e.g. Keast 1977, Persson 1983b, Horppila et al. 2000), ontogenetic diet shifts (e.g. Wahl et al. 1993, Graeb et al. 2006, Bystrom et al. 2012), trophic positioning (e.g. Vander Zanden et al. 1997, Quevedo and Olsson 2006), individual resource use (e.g. Ansari and Qadri 1989, Svanback and Persson 2004), competition (e.g. Fraser 1978, Persson 1983a, Paszkowski 1985), etc. Generally, EP and YP feed on a variety of food sources, using their generalized terminal mouth (Craig 2008). Both are considered omnivorous predatory fish or bentho-pelagic generalists of a medium trophic level, but variability in diet can be high (Webb 1984, Craig 1987, Cabana and Rasmussen 1996, Vander Zanden et al. 1997). 3

Like most fish, perch hatch as very small larvae and then rapidly increase in size and weight many times over (Persson 1983b, Werner and Gilliam 1984, Persson 1987b). Accordingly, their energy requirements change and new resources become available, mainly because of increasing gape size (Mehner and Thiel 1999). For example, size of consumed ZP was smaller in stomachs of YP larvae than in the ZP samples from their habitat, but with increasing gape size, bigger Daphnia species were exploited (Treasurer 1990). Both perch species consume increasingly larger prey as they grow (Keast 1977, 1985, Mittelbach and Persson 1998). More specifically, their diet and TP changes with size and age, while they start to exploit new prey types twice during ontogeny, which are described as ontogenetic diet shifts (Hjelm et al. 2000, Graeb et al. 2006, Quevedo and Olsson 2006). This alternating feeding ecology has been described as life history omnivory (Polis and Strong 1996, Persson and de Roos 2006). In addition, many perch populations exhibit a resource dimorphism between littoral and pelagic morphs with subpopulations that utilize different basal resources (Svanback and Eklov 2002, Parker et al. 2009a, Luek et al. 2010). Hence, different functional groups exist within EP and YP that specialize on particular food items (Werner and Gilliam 1984, Persson and Greenberg 1990a, Claessen and Dieckmann 2002).

Perch ontogeny starts in spring when spawning takes place (Collette et al. 1977). Larvae hatch at about 6 to 7 mm total length (TL) and stay confined to the littoral until their yolk sac is absorbed (Fig. 1). After yolk sac depletion, larvae undergo a pelagic phase accompanied by foraging on ZP (Whiteside et al. 1985, Post and McQueen 1988, Wang and Eckmann 1994). The free swimming larvae start to feed on rotifers (e.g. Keratella sp.) and then include stage one copepodites in their diet. They feed on increasingly larger copepods and finally on cladocerans (Whiteside et al. 1985, Treasurer 1990, Fulford et al. 2006). In general, perch prefer large sized ZP species like daphnids, as soon as they overcome their mouth gape limitation (Furnass 1979, Romare 2000, Shrader 2000, quoted in Brown et al. 2009, MacDougall et al. 2001, quoted in Brown et al. 2009). For instance, at about TL = 9 mm, YP larvae have been shown to start feeding on large Daphnia pulex (Wahl et al. 1993). 4

Figure 1: Schematic figure of the habitat shifts between the littoral and the pelagic habitat during the early ontogeny of P. fluviatilis & P. flavescens (modified from Whiteside et al. (1985)).

Following the pelagic phase, juvenile perch undergo a diet shift towards zoobenthivory (Fig. 2). At about 25 to 30 mm TL, young perch migrate to the littoral and start to feed on various benthic invertebrates (Whiteside et al. 1985, Post and McQueen 1988, Berezina and Strel'nikova 2001). In the littoral environment, perch feed mostly on chironomids, isopods and amphipods (Persson 1983b, Fullhart et al. 2002). The diet shift to zoobenthivory can immediately affect the zoobenthos community, reducing its biomass, abundance and diversity, depending on habitat complexity (Diehl 1992). For example, biomass and abundance of oligochaets and chironomids in a controlled environment was drastically reduced when young-of-the-year (YOY) perch started to feed on benthic prey 30 to 60 days after hatching (Berezina and Strel'nikova 2001).

Most perch perform a second ontogenetic diet shift and become facultative piscivores (Fig. 2). Fish is included in the diet to various degrees at about age-1+, which is why adult perch are considered to be secondary piscivorous (Mittelbach and Persson 1998, Fullhart et al. 2002, Amundsen et al. 2003). Sometimes, very early piscivory can occur in YOY perch at TL < 30 mm, which feed on fish larvae. In EP, this usually results in size bimodality between mainly zooplanctivorous and piscivorous perch (Berezina and Strel'nikova 2001, Beeck et al. 2002, Urbatzka et al. 2008). Generally, perch hunt other fish by chasing them in small troops or packs (Deelder 1951, Nursall 1973). Common prey fish for P. flavescens in their native range include sunfishes (e.g. Lepomis gibbosus), minnows (Cyprinidae), sticklebacks (e.g. Culaea inconsta) and other YP (Knight et al. 5

1984, Fullhart et al. 2002, Parker et al. 2009b). In Lake Washington (Washington, USA), YP primarily feed on benthic-living prickly sculpin (Cottus asper) (McIntyre et al. 2006). Also, EP feed on a variety of fish: cyprinids like roach (Rutilus rutilus) and bream (Abramis brama), but also ruffe (Gymnocephalus cernua), smelt (Osmerus eperlanus), and smaller conspecifics (Persson and Eklov 1995, Beeck et al. 2002).

Figure 2: Food web scheme for P. fluviatilis & P. flavescens with respect to ontogeny and energy pathways. Trophic links to aquatic consumers are illustrated with solid arrows ( ). Diet shifts between size classes of perch are shown as dashed arrows ( ). The scheme depicts (from top to bottom) the co-ocurring top predators (Esox lucius (upper picture) & Micropterus salmoides (lower picture), the different size classes of P. fluviatilis (upper pictures) & P. flavescens (lower pictures), their most important types of prey (zooplankton, benthic/epiphytic invertebrates and fish) and the bases of the littoral and pelagic food webs (phytoplankton, benthic & epiphytic algae) along their assumed trophic positions (modified from Rasmussen et al. 2008) – pictures: Wisconsin Department of Natural Resources, Bangladesh Fisheries Information, Fish Management in New York State & Fischereiverband für das Land Vorarlberg. 6

Especially among EP, cannibalism has been shown to play an important role for population structure and dynamics (Persson et al. 2000, Persson et al. 2004). Exceptionally large individuals are often highly cannibalistic and exhibit increased growth (Lecren 1992, Claessen et al. 2000, Bystrom et al. 2012). The resulting individual variation in morphology allows for larger-sized EP to perform earlier diet shifts or intracohort cannibalism, respectively. Larger-bodied individuals have a larger mouth gape and can therefore switch their diet to invertebrates and fish more easily (Mittelbach and Persson 1998, Bystrom et al. 2012). The cannibalistic perch represent a top-down control for interspecific competitors as well as for smaller perch (Persson et al. 2004). Removal or death of these individuals results in recruitment overcompensation (Ohlberger et al. 2011). For example, more recruits of YOY EP followed the die-off of large age-2+ perch, that inhabited the near shore areas of a lake (Persson et al. 2000).

Diet shifts and the functional niche can be strongly affected by interspecific competition: Inferior foraging success of perch leads to foraging on less profitable prey and has been shown to result in reduced growth or even stunted populations (e.g. Persson 1983a, Persson 1983b, Persson 1988, Persson and Hansson 1999, Schoenebeck and Brown 2010). For example, EP and roach compete for a variety of food sources, whereas roach (R. rutilus) prey more efficiently on ZP than perch (Persson 1987a). Thus, high densities of roach are thought to accelerate the perch ontogenetic diet shift from ZP to macrozoobenthos (MZB) (Persson 1986, Bergman 1990). However, habitat-specific foraging abilities mediate competitor success, and competition might be rather limited (Radke and Eckmann 2001, Haertel et al. 2002, Persson and de Roos 2012). YP, for instance, shifted from ZP to MZB after the removal of zoobenthivorous white suckers (Catostomus commersonii) (Hayes et al. 1992).

Additionally, ontogenetic shifts in perch are highly variable between individuals and localities and can be skewed by individual specialization on certain food items, accompanied by changes in growth, size and morphology (Quevedo and Olsson 2006, Urbatzka et al. 2008). EP have been shown to avoid increasing intraspecific competition by specialization, evidenced by higher niche breadths at high densities of adult perches (Svanback and Persson 2004, Bolnick et al. 2007). For example, the foraging efficiency of pelagic and littoral EP differs: Streamlined, pelagic perch are better adapted to catch ZP while deep bodied, littoral perch perform better with benthic food items (Svanback and Eklov 2003, 2004). Also, small YP individually differed in the number of feeding 7 attempts, number of unsuccessful feeding attempts and timing between feeding attempts when they were fed different prey items (Ansari and Qadri 1989).

The aforementioned factors show that the conditions of diet shifts in perch are not yet fully understood or whether these differ between EP and YP. The functional role of both species in their respective ecosystems has never before been compared across continents. Intercontinental comparisons in fish ecology are rare and in most cases focus on communities (e.g. Mahon 1984, Tonn et al. 1990, Lamouroux et al. 2002). Comparisons at the species level can reveal great differences among species or populations in their native and introduced ranges (Brandner et al. 2013, Budy et al. 2013). This study meant to analyze and describe the functional feeding role or niche of EP and YP from North America and Europe by means of stable isotope (SI) analysis. YP and EP are closely related, but they or their ontogenetic feeding stages, might not play similar roles in the ecosystems they inhabit. Carbon (δ13C) and nitrogen (δ15N) isotope-based calculations of TP, mixing models of resource use and isotopic niche characteristics provide a powerful currency to compare the ecological role and describe the ontogenetic changes of YP and EP in food webs (Vander Zanden and Rasmussen 1996, Semmens et al. 2009).

In a seminal paper on perch biology, Thorpe (1977) concluded that YP and EP fulfill the same ecological role, due to their high degree of similarity in morphological, physiological and behavioral characteristics. However, such commonalities do not necessary mean ecological equality as other factors like behavior, resource abundance and predator distribution influence the diet (Motta et al. 1995). So far, only minor differences could be demonstrated in YP and EP ecology, like a greater physical tolerance to eutrophication and high water temperatures in EP (Hokanson 1977, Leach et al. 1977, Morgan et al. 2003). In general, TP estimates derived from stomach- and SI analyses of YP and EP are characteristic for omnivorous or secondary piscivorous fishes. They range from TP 3.0 to 4.0, which comply with secondary to tertiary consumer levels (Cabana and Rasmussen 1996, Vander Zanden and Vadeboncoeur 2002). However, TP estimates for perch should only be given for a defined size class to account for distinguishable ontogenetic feeding guilds (McIntyre et al. 2006, Quevedo and Olsson 2006). Similar patterns in ontogenetic diet shifts in both perch lead to the first hypothesis:

Hypothesis 1: Similar breakpoints exist within the size related increase of trophic position in YP and EP.

Carbon sources and nitrogen content of perch prey changes, as they feed on different trophic levels and utilize distinct trophic pathways during ontogeny (e.g. McIntyre et al. 2006, Quevedo and Olsson 2006). Generally, TP-estimates (and δ15N-values) are expected to increase with size (Jennings et al. 2002). C13-values should also increase when C13-enriched benthic prey or littoral fish are included in the diet and isotope turnover rates decrease at larger sizes. The size at the breakpoints should be similar in YP and EP and mark the endpoints of a transitional ontogenetic diet shift, representing the limits of the respective feeding guilds.

The resulting subgroups, as defined by the breakpoint, should be used to model the diet of those size classes. SI analysis can be used to investigate their middle-to-long-term feeding patterns via mixing models (Phillips and Gregg 2003). Also, the niche space of populations and the degree of individual specialization can be inferred from SIs (Bearhop et al. 2004, Layman et al. 2007). Studies on stomach contents and mixing models of diet in both perch species showed that they undergo the aforementioned niche shifts from ZP to MZB to fish and consequently occupy similar foraging niches within their respective size classes (Hjelm et al. 2000, Amundsen et al. 2003, McIntyre et al. 2006).

Hypothesis 2: In accordance with Thorpe (1977), YP and EP occupy similar trophic positions and show similar resource utilization during ontogeny.

The resource use in the smaller and larger group therefore should be analogous. Also, convex hull area and other metrics introduced by Layman et al. (2007), which have already been successfully used to study the feeding ecology of perch, should be similar if the two species occupy equivalent ecological niches (Quevedo et al. 2009).

To test these hypotheses, the trophic positions and diets of YP from three North-American lakes in Washington, USA and EP from three lakes in Brandenburg, Germany are estimated from SI samples. 9

2 Eurasian Perch and Yellow Perch

2.1.1 Systematics & Zoology

The genus Perca comprises of three species that resemble each other in many characteristics: The Eurasian perch, P. fluviatilis (Linnaeus 1758 – Fig. 3a), the yellow perch, P. flavescens (Mitchill 1814 – Fig. 3b) and the Balkhash perch, P. schrenkii (Kessler 1874) (Kottelat and Freyhof 2007, Craig 2008). The latter is not dealt with in this study. It has been proposed that the geographically separated P. fluviatilis and P. flavescens represent a single species with three subspecies (Svetovidov and Dorofeeva 1963). Nowadays, both are considered to be distinct species because of the different positioning of their predorsal bone (Collette and Banarescu 1977).

Figure 3: Photographs of adult P. flavescens (a) & P. fluviatilis (b) – pictures: wikimedia.org

YP and EP exhibit minor morphological differences: YP (World Record: 1.91 kg) stay notably smaller than EP (World Record: 4.8 kg). Fishbase (www.fishbase.org) states a maximum size of 500 mm as total length (TL). For EP, the given maximum size is 600 mm as standard length (SL) (Froese and Pauly 2012). In addition, several phenological differences exist: YP are brass-colored and can exhibit five to nine vertical bars. EP are greenish in color, have five to eight vertical bars and exhibit more pronounced red pelvic, anal and caudal fins as well as a more pronounced spot at the second dorsal fin. (Craig 1987).

Genetic analysis of mitochondrial cytochrome-b DNA revealed the phylogenetic relationships in perch. It could be shown that both species are closely related and share a common ancestor (monophyletic), but they differ at 130 sites in their cytochrome-b sequences (Song et al. 1998, Sloss et al. 2004). 10

2.1.2 Distribution

YP and EP exhibit similar tolerance zones in their environment (Thorpe 1977). All species of the genus Perca are distributed in the holoarctic ecozone (Craig 1987). They mainly inhabit mesotrophic to slightly eutrophic water bodies, where especially P. fluviatilis often dominates the fish communities (Mehner et al. 2005, Brown et al. 2009). Both species inhabit and spawn in fresh or sometimes brackish waters of small ponds to large lakes, rivers and estuaries (Scott and Crossman 1973, Kottelat and Freyhof 2007).

The North American P. flavescens is native to the north-eastern part of the American continent (USA, Canada): The Great Lakes and Mississippi River basins from Nova Scotia and Quebec west to Great Slave Lake in Northwest Territories in Canada, and south to Ohio, Illinois and Nebraska in the USA; south in Atlantic drainages to Santee River in South Carolina, USA (Froese and Pauly 2012). Presently, P. flavescens can be found beyond their original distribution west of the continental divide in the states of Washington (introduced in Washington in 1890 as a sport and food fish), Oregon, California and British Columbia, as well as south in New Mexico and Texas (Scott and Crossman 1973).

Native throughout Europe, P. fluviatilis does not occur naturally on the Iberian Peninsula, central Italy and the Adriatic basin. P. fluviatilis is widespread all over Siberia eastwards up to Kolyma and in drainages into the Arctic and the Aralsee. Moreover, this species was introduced to the several southern European countries, Australia and South Africa (Kottelat and Freyhof 2007). The latitude affects certain aspects of their life history traits (Berven and Gill 1983). For example, growth rates, mortality rates, and reproductive investment decrease with latitude (Heibo et al. 2005).

3 Material & Methods

3.1 Study Area

The TPs of fishes in lake ecosystems are dependent on lake size (Vander Zanden et al. 1999b, Post et al. 2000). Thus, six small, mesotrophic, summer-stratified lakes between 12 and 63 ha (also Fig. 4) were chosen for the species comparison: three lakes in North America (USA) and three lakes in Germany (GER). A summary of important limnological lake characteristics is given in table 1. Every lake was considered to be a single sample. One epilimnic (1 m depth) and one hypolimnic water sample (~1 m above the deepest point of each lake) were taken from each lake with a standard water sampler according to Ruttner for total phosphorous and total nitrogen analysis. Also, conductivity, pH, and temperature were measured during sampling events in the epilimnium and hypolimnium by YSI multi-probes (YSI Model 85 / YSI 600)* (Tab. 1).

* Due to different equipment in Washington and Germany, the type and/or brand of an item are given for each country as: (product USA / product GER).

3.1.1 American Lakes

Between June and August 2012, three North American lakes were sampled in western Washington State, USA: Lake Padden (Fig. 4a), Whatcom County (48.7027°, - 122.4529°); Walsh Lake (Fig. 4b), King County (47.4085°, -121.9285°) and Lake Wilderness (Fig. 4c), King County (47.3726°, -122.0344°). The lakes are kettle lakes with glacial origin located in the Pudget Sound basin. The near shore areas are forested and covered by douglas fir (Pseudotsuga menziesii), alder (Alnus spec.), vine maple (Acer circinatum) and black cottonwood (Populus trichocarpa). Mesotrophic Lake Padden and Lake Wilderness exhibit residential clearings at the shoreline, while Walsh Lake lies within the Seattle water protection area with no development in the upstream watershed. YP do not natively inhabit these lakes and have been introduced to this area since the late nineteen hundreds (DFO 2011). Although the records are incomplete, the earliest known stocking of YP into Walsh Lake occurred in 1915-17 (2,000 perch were stocked) (Julian Olden pers. comm. 2012).

3.1.2 German Lakes

Three European lowland lakes in Brandenburg, Germany were sampled in October and November 2012: Lake Dölln (Fig. 4d), Uckermark (52.9946°, 13.5818°); Lake Väter (Fig. 4e), Uckermark (53.0051°, 13.5530°) and Lake Wucker (Fig. 4f), Uckermark (53.0071°, 13.6424°). All lakes are situated in the Baltic lake region and the lake district of Templin-Werbellin. The lakes have a glacial origin and originated in the late Pleistocene. Their near shore areas are forested with alder (Alnus spec.), beech (Fagus sylvatica), birch (Betula spec.) and pine (Pinus sylvestris). The mesotrophic to slightly eutrophic Lake Väter and Lake Dölln have been well studied with respect to hydrogeography, trophic characteristics, submerged macrophytes and fish communites (Radke 1998, Kasprzak et al. 2000). The fish communities are known to be dominated by roach and perch (Radke 1998, Haertel 2001). Generally, there is a high degree of similarity between all of the tested German water bodies with respect to fish community and morphometric characters, like mean and maximum depth (Eckmann 1995, Mehner et al. 2005).

latitude longitude secchi altitude area Zmax Zmean U σ TP TN lake pH fish community (dec. °) (dec. °) (m) (m) (ha) (m) (m) (km) (µS/cm) (µg/l) (µg/l)

USA Padden 48.7027 -122.4529 4.6 137 63.5 18.0 8.2 3.64 100 7.8 23 483 LB, RT, YP, PS, PSc

Walsh 47.4085 -121.9285 4.2 221 24 10.7 N/A 2.05 54 7.0 11 343 LB, YP, PSc, NP

Wilderness 47.3726 -122.0344 5.6 143 27.9 11.6 6.4 2.86 115 8.0 16 366 LB, RT, Ko, YP, PS, GS, PSc

GER Dölln 52.9946 13.5818 1.8 54 25 7.8 4.1 2.19 463 8.1 33 950 Pi, EP, Ro, Ru, Te, Br, WB,

Väter 53.0051 13.5530 2.2 60 12.07 11.5 5.2 1.39 389 6.9 22 780 Pi, EP, Ro, Ru, Te, Br, WB,

Wucker 53.0071 13.6424 3.8 64 22 16.1 8.0 3.15 365 7.0 14 580 Pi, EP, Ro, Ru, Te, Br, WB,

U = shoreline length; Zmax = maximum depth; Zmean = mean depth; σ = conductivity; TP = total phosphorus; and TN = total nitrogen. Abbreviations for fish species – USA: LB = largemouth bass (Micropterus salmoides); RT = rainbow trout (Oncorhynchus mykiss); Ko = kokanee (Oncorhynchus nerka); YP = yellow perch (Perca flavescens); PS = pumpkinseed sunfish (Lepomis gibbosus); green sunfish (Lepomis cyanellus); prickly sculpin (Cottus asper). Abbreviations for fish species – Germany: Pi = pike (Esox lucius); EP = Eurasian perch (Perca fluviatilis); Ro = roach (Rutilus, Rutilus); Ru = rudd (Scardinius erythrophthalmus); Te = tench (Tinca tinca); WB = white bream (Abramis bjoerkna); Br = bream (Abramis brama).

3.2 Sample Collection 3.2.1 Fish

To account for littoral and pelagic differences in feeding ecology and to catch a broad range of size classes, different sampling gears were used. Fish were caught with minnow traps, fyke nets, hook and line, electrofishing and multi-mesh, pelagic gillnets between June and August 2012 (Washington) and between October and November 2012 (Brandenburg). Benthic multi-mesh gill nets were set perpendicular to the shore to catch pelagic fishes; two longer nets in the USA and four shorter nets in Germany (58.52 m x 1.83 m; six panels each being 9.75 m long with mesh-sizes 25.4, 31.75, 38.1, 50.8, 63.5 76.2 mm (stretched mesh) / NORDIC; lxh: 30 m x 1.5 m; twelve panels each being 2.5 m long with mesh-sizes 5, 6.25, 8, 10, 12.5, 16, 19.5, 24, 29, 35, 43 and 55 mm; Lundgrens Fiskredskapsfabrik AB, Stockholm, Sweden). In Washington, one fyke net (with a diameter of 0.76 m, 7.92 m wing length and five hoops with 24.52 cm² openings) and four galvanized, steel minnow traps (41 cm, two 2.54 cm openings and 6.4 mm mesh size) were set in each of four sectors to catch littoral fishes. In Germany, electrofishing was conducted in the littoral zone. Fish were identified to species. Total length (TL in mm) and weight (g) were recorded for each fish (Tanita 1144; d = 0.01 g / Satorius AW- 4202; d = 0.01 g). Later on, fish were grouped into size classes according to the results of the breakpoint models (3.4.1). The fish were bagged, labeled and put on ice for transport. Perch and potential prey fish were selected for analysis. All samples (ZP, benthic invertebrates and fish) were frozen in the laboratory at -20°C prior to preparation for SI analysis, because preservation in formalin or ethanol has been shown to alter SI ratios (Kaehler and Pakhomov 2001, Sarakinos et al. 2002).

3.2.2 Benthic Invertebrates & Zooplankton

To address spatial variability of signatures of rather immobile invertebrates within lakes, lakes were divided into four similarly sized sectors (Syvaranta et al. 2006). Each sector was sampled at two points of the littoral, targeting different microhabitats (wood, macrophytes & stones). Benthic organisms were collected with a D-frame aquatic net (mesh size 500 µm) and hand-picked from stones and wood of the littoral at different depths less than 1 m. Habitat and depth of each netting event were recorded. Larger stones 16 and pieces of wood were examined for benthos and removed from the net sweeps. An Ekman dredge (WILDCO Ponar grab) was used to sample profundal benthos. These samples were not included in the analysis, because they contained very few or no invertebrates at all. ZP were caught by multiple trawls with plankton nets of 100 µm mesh size. The net sweeps and composite ZP samples were bagged, labeled and put on ice for transport. All samples were frozen within 8 h after sampling at -20°C to kill the invertebrates and to store them for analysis.

3.3 Stable Isotopes: Sample Preparation and Analysis

ZP were thawed and filtered through a series of mesh sieves (200 µm, 500 µm (471 µm – USA) and 1000 µm) to remove algae and gain different size fractions. The samples were washed with deionized water and transferred into petri dishes. Each fraction was then examined under a dissecting microscope and sorted with a pair of tweezers. The ZP were divided into groups of cladocerans, calanoids, cyclopoids and Chaoborus spec. due to their different diets (Matthews and Mazumder 2003). All other predatory cyclopoids like Epischura or Mesocyclops and Leptidora spec. were sorted out. Other materials like parts of plant or detrital particles were removed from the samples. Cladocerans were grouped into small and large individuals where possible. A great share of the German ZP samples consisted of very small cladocerans and calanoids (< 500 µm). Hence, they were pooled together in bulk samples.

Benthos samples were thawed and examined for invertebrates under a dissecting microscope. Invertebrates were sorted out and grouped by taxonomic groups as accurately as possible. A minimum of five individuals were pooled for one sample and at least three replicates of each group were created. In case of very large individuals, e.g. large predatory invertebrates, less than five individuals were used for a sample. Only one crayfish was caught in Walsh Lake and thus excluded from the samples.

White muscle tissue samples from 68 YP (20 PA, 19 WA, 15 WI) and 82 EP (32 DO, 30 VA, 20 WU) were used for the analysis. Muscle tissue was chosen because it offers the most robust estimates of long term diet predictability by SIs due to its relatively low metabolic activity (low turnover rate of SI) (Pinnegar and Polunin 1999). 17

The fish were thawed, weighed, measured and descaled at the dorso-ventral area (Fig. 5a). The skin was removed and a cube of about 1 cm edge length was cut out of the dorsal muscle with a scalpel (Fig. 5b). All samples were transferred into aluminum trays and dried for 24 h in a drying oven at 60°C (Fig. 5c). After drying, samples were ground as to a fine powder with mortar and pestle to increase isotopic homogeneity (Fig. 5d). Finally, samples were transferred into Eppendorfer tubes or small glass vials with snap- caps and stored in a desiccator prior to weighing.

To calculate trophic positions and diet composition from SIs, the samples had to be analyzed for the ratios of carbon isotopes (12C, 13C) and nitrogen isotopes (14N, 15N), with 12C and 14N being the more common isotopes (Fry 2006). A micro analytical balance (Sartorius micro / Scientech SA80) was used to weigh about 1.000 mg (± 0.200 mg) from each sample (ZP, MZB and fish). Samples were then weighed into 5*9 mm / 8*5 mm tin capsules (COSTECH Analytical Inc. / Elemental Micoanalysis). After compressing the tin capsule, each sample was re-weighed. Finally the samples were placed in a 96-well 18 polystyrene cell plates (Rotilabo®-microtest plates), which were covered and labeled. Samples were sent to UC Davis SI Facility in California, USA, for 13C and 15N isotope analyses. Altogether, 255 samples (125 invertebrate & 130 fish samples) from Washington and 321 samples (184 invertebrate & 137 fish samples) from Germany were analyzed. In the laboratory, the capsuled samples were broken down into elemental 15 13 components N2 (for N analysis) and CO2 (for C analysis) at 1000°C in the presence of catalyzing chromium oxide and silvered copper oxide in a PDZ Europa ANCA-GSL elemental analyzer. Then oxides were removed and the N2 and CO2 gases were separated in a Carbosieve GC column (65°C, 65 mL/min) with helium carrier gas. Finally, 13C and 15N isotope ratios were analyzed in the connected PDZ Europa 20-20 isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK). The laboratory ran a number of replicates from different laboratory standards with the samples. Those standards resembled the composition of the sample and have been calibrated against the American National Institute of Standards and Technology (NIST) Standard Reference Materials (IAEA-N1, IAEA-N2, IAEA-N3, US.GS-40, and USGS-41). All samples isotope ratios were corrected based on those standards. The long term standard deviation of UC Davis SI laboratory is 0.2‰ for 13C and 0.3‰ for 15N.

Results are expressed relative to international standards (delta (δ) notation), as the heavy to light isotope ratio of the sample (R) to the according ratio of a given standard (δ = 0‰). The standard for carbon is carbon dioxide derived from the calcium carbonate of Vienna PeeDee Belemnite (V-PDB) limestone and atmospheric nitrogen (air) for nitrogen. Results are given as ratios of 13C to 12C (Equ. 1) and 15N to 14N (Equ. 2) relative to international standards:

Equation 1: Formula for calculating delta (δ) values from mass ratios of 12C & 13C (R).

( 푅sample − 푅standard) 훿13퐶 = × 1000 푅standard

Equation 2: Formula for calculating delta (δ) values from mass ratios of 14N & 15N (R).

( 푅sample − 푅standard) 훿15푁 = × 1000 푅standard

Rsample = ratio of the heavy to light isotope in the sample and Rstandard ratio of the heavy to light isotope in V-PDB/air (Sharp 2007). Units are expressed as per mill (‰). 19

Many studies propose a correction of δ13C-values for 13C-depleted lipids, either by chemical extraction or by mathematical models for normalization (Sweeting et al. 2006, Post et al. 2007, Logan et al. 2008). This is done because lipids in the tissue are depleted in 13C, which leads to lower δ13C-values in fatty tissue (Post et al. 2007). Although, many of these models which estimate lipid contents from C/N mass ratios can perform well with fish muscle tissue (e.g. Sweeting et al. 2006, Logan et al. 2008), most of these models work rather poorly on invertebrates (e.g. Kiljunen et al. 2006, Syvaranta and Rautio 2010). Also, chemical lipid extraction may alter δ15N-values and consequently TP estimates (Sweeting et al. 2006, Post et al. 2007). Furthermore, data on lipid content of the samples were not recorded, which would have been necessary to get reasonable estimates of lipid normalized values from mathematical models (Sweeting et al. 2006). Finally, perch dorsal muscle exhibit low lipid contents (2.4 ± 0.4% for P. flavescens and 1.43 ± 0.03% for P. fluviatilis) and low C/N ratios (< 5) (Pinnegar and Polunin 1999, Mairesse et al. 2006, Mackintosh et al. 2012). In this case, lipid normalization does not significantly remove biases in δ13C-values (Post et al. 2007). For example, muscle tissue from P. fluviatilis showed only marginal differences (0.06 ± 0.06‰) between lipid extracted samples and untreated samples (Kiljunen et al. 2006). Hence, the δ13C-values were not corrected for lipids in this study.

3.4 Statistical Analysis

Exploratory analyses were carried out with SPSS version 21 (IBM): Residual plots of TP- data of each species were examined for normal distribution. Homogeny of variances was determined for TP with a non-parametric Levene's test prior to further analysis (Nordstokke and Zumbo 2010). Descriptive statistics were calculated for δ13C-/δ15N- values, TL and TP. Regression analyses were calculated for δ13C- and δ15N-values and TL of both perch species in every lake. Isotope values of different organisms were plotted in bi-plots for each lake together with its specific baseline.

3.4.1 Breakpoint Analysis

The similarity of ontogenetic diet or niche shifts between species was examined by breakpoint analysis of TPs related to length. The three different ontogenetic feeding groups in perch have to be considered when analyzing the prey of perch. Regressions between TL and TP were analyzed for breakpoints, to assess significant threshold values in TP and to determine size classes for subsequent analysis by a mixing model. The procedure was conducted for all YP and EP. First, linear models for the TL against TP were calculated with the statistic program R (RCoreTeam 2012). Those models were then tested for breakpoints with the Davies’ Test implemented in the R-package segmented (Muggeo 2008). The Davies’ Test tests for a non-zero difference-in-slope parameter of a segmented relationship within a generalized linear model (Davies 1987). Such a change in slope is thought of as an indicator of a (ontogenetic) diet change. The perch samples were then grouped into size classes according to the results of the Davies test (small & large). The smallest EP (TL = 62 mm, TP = 3.60) had to be excluded from the dataset because the Davies’ test falsely identified a nonsensical breakpoint before the next larger individual (breakpoint at 62 mm).

In addition, the collected dataset of Walsh Lake and Lake Wilderness – was supplemented by SI data from the year 2009 for this analysis (Eric Larson, University of Washington, unpublished data). The δ13C/δ15N-values of the same taxonomic groups or diet groups from the year 2009 and the present dataset (2012) were tested against each other via Student’s t-test. When primary consumers that were present in both datasets did not differ significantly, perch trophic level was calculated and implemented into the breakpoint analysis.

3.4.2 Trophic Position

The TPs were calculated for each perch in relation to a lake specific baseline. The change in carbon and nitrogen ratios (sometimes complemented by sulfur or hydrogen isotope ratios) can be analyzed for differences between consumers and their diet to calculate the TP and resource use of an individual (Peterson and Fry 1987, Cabana and Rasmussen 1996, Caut et al. 2009). The process of isotopic fractionation makes it possible to infer the TP of an organism in the food web from its δ15N-value and to point towards the dietary 21 sources according to its δ13C-value (France 1995, Post 2002). The 15N isotope accumulates along the food chain within the tissue of consumers (Deniro and Epstein 1981, Checkley and Miller 1989). Therefore, the 15N-content of a consumer represents a temporally integrated measure of its diet, accounting for the energy or mass flow to the consumer (Fry 1988, Hobson and Welch 1992). The TP of any consumer in its respective food web should be calculated by relating its δ15N-value to a properly defined baseline in order to enable cross system comparisons (Cabana and Rasmussen 1996, Vander Zanden and Rasmussen 1999, Post 2002). Primary consumers represent good baseline organisms, because they are more stable and less variable in their isotopic signature than primary producers (Anderson and Cabana 2007).

A general formula (Equ. 3) which is frequently used for calculating TP from δ15N-values is given below (Cabana and Rasmussen 1996, Vander Zanden et al. 1997):

Equation 3: General formula for calculating the trophic position (TP) of a consumers. (훿15푁 − 훿15푁 ) 푡푟표푝ℎ𝑖푐 푝표푠𝑖푡𝑖표푛 = [ consumer baseline ] + λ consumer Δ15N

15 15 The δ Nconsumer-value is the measured δ N-value of the consumer, for which the TP 15 15 should be estimated. The δ Nbaseline is the measured δ N-value of the baseline organism(s) chosen. The λ represents the estimated TP of the baseline consumer (e.g. λ = 2 for a primary consumer). The ratio of the different reaction rates for the light and heavy isotope is expressed as the fractionation factor (Δ) which is given in positive per mill values (Farquhar et al. 1989). The fractionation factor Δ15N resembles the medium- enrichment in 15N from one trophic level to the next higher trophic level or between 15 consumers and their prey, respectively. Typically, Δ N or the degree to which a consumer is enriched in 15N relative to its diet, varies between 2 and 5‰ (for a meta-analysis see: Vanderklift and Ponsard 2003). Many studies and reviews established Δ15N = 3.4‰ as a standard value for trophic interactions of aquatic consumers including P. flavescens (Minagawa and Wada 1984, Vander Zanden et al. 1997, Post 2002).

However, in aquatic environments, the different but interwoven littoral, profundal and pelagic pathways have to be taken into account when TPs are calculated (Vander Zanden and Rasmussen 1999, Vadeboncoeur et al. 2002). In lakes for example, TP can be weighted by importance of benthic and pelagic carbon pathways (Quevedo and Olsson 2006). Vander Zanden and Rasmussen (1999) developed a robust model that accounts for 22 lake specific effects and spatial variability of primary consumers or the baseline, respectively. The model includes the habitat specific 15N-13C relationships within lentic systems to calculate a lake-corrected baseline curve, from which a higher-order consumer’s TP can be derived (Fig. 6). The general baseline curve for aquatic primary consumers is given in equation 4:

Equation 4: General formula for calculating the baseline curve derived from primary 15 13 consumer stable isotope (SI) composition (δ Nprim.con. & δ Cprim.con.).

15 6.43 훿 푁prim.con. = 13 1 + exp[9.67 + (0.356 ∗ 훿 퐶prim.con.)]

Figure 6: An example of lake corrected baselines as they are used in this study to depict the δ15N- δ13C relationship in the context of habitats and the consequences for consumer trophic position (TP - i.e. the average position relative to primary producers at which an organism feeds). Primary consumers represent a TP of 2. In spite of similar δ15N-values, sculpins occupy lower TP due to their diet of 13C-depleted profundal prey (modified from Vander Zanden & Rasmussen) - picture of YP: Maryland Dept. of Natural Resources.

In order to achieve an intercontinentally comparable measure for TP, the model of Vander Zanden & Rasmussen (1999) was used to correct δ15N-values for the aforementioned effects. Firstly, all primary consumers samples were selected from each lake: ZP (bulk), Calanoida, Cladocera, Chironominae, Trichoptera, Gammaridae and Asellidae. Based on their δ15N- and δ13C-values, the residuals for the individual primary consumers were calculated by equation 5:

Equation 5: Formula for calculating the residuals of primary consumer stable isotope 15 13 (SI) composition (δ Nprim.con. & δ Cprim.con.). 15 6.43 푟푒푠𝑖푑푢푎푙 = 훿 푁prim.con. − 13 1 + exp[9.67 + (0.356 ∗ 훿 퐶prim.con.)]

Secondly, the lake-specific mean residual value (Uresid) was calculated as the mean of all primary consumer’s residuals of a given lake. Thirdly, the lake corrected baseline value 15 (δ Ncorrected) for consumers was calculated (Equ. 6):

15 15 Equation 6: Formula for calculating a lake corrected baseline value of δ N (δ Ncorrected).

15 6.43 훿 푁corrected. = 13 + 푈resid 1 + exp[9.67 + (0.356 ∗ 훿 퐶consumer)]

Finally, TP of each perch was calculated by a using the corrected δ15N-baseline value in equation 7. The fractionation factor (Δ) used in the equation 7 was 3.4 for nitrogen (from Vander Zanden and Rasmussen 2001, Post 2002).

Equation 7: Formula for calculating the trophic positions (TPs) of consumer δ15N composition with respect to lake-corrected primary consumer baselines.

(훿15푁 − 훿15푁 ) 푡푟표푝ℎ𝑖푐 푝표푠𝑖푡𝑖표푛 = [ consumer correct ] + λ consumer Δ15N

The TPs of each size class were tested for differences between species and lakes as a dependent variable within a generalized linear model (GLM) in SPSS (ver. 21). A generalized model was chosen because of its robustness, since TPs were not normally distributed in all lakes. The fixed factors in the model were species and lakes, whereas species have been nested within lakes as only one species exists in a given lake. The TL was included in the model as a covariate. Finally, main effects (species, lakes & size) were tested for differences, because full factorial analysis revealed no significant interactions of e.g. size and lake. 24

3.4.3 Trophic Niche Metrics

Trophic niche metrics were calculated for each lake and size class in R with the R-package SIAR - SI-analysis in R (Parnell et al. 2010). N-dimensional SI plots (e.g. two dimensional δ13C-δ15N-bi-plots) can be used to define the trophic niche of consumers. The idea of hypervolumes as descriptors of ecological niches have been implemented in new quantitative metrics and definitions of niche space (Layman et al. 2007, Newsome et al. 2007, Jackson et al. 2011). Studying isotopic niche spaces can give insights into feeding ecology, especially when combined with conventional stomach analysis (Bearhop et al. 2004). The metrics introduced by Layman et al. (2007) can be used to describe trophic niche characteristics in consumers and even ontogenetic niches (e.g. Darimont et al. 2009, del Rio et al. 2009, Olsson et al. 2009, Hammerschlag-Peyer et al. 2011).

For this study trophic niche metrics following Layman et al. (2007) were calculated to describe the niche of both size classes in each perch species within the “δ-space” (δ13C/δ15N-bi-plots). To that end, SIBER (SI Bayesian Ellipses in R) was used, which is part of the SIAR-package for R (Parnell et al. 2010, Jackson et al. 2011). Niche parameters were calculated for each species as mean values for each lake. The metrics comprise the δ13C- and δ15N-range, the smallest convex set of data points (convex hull), the mean distance to centroid (CD), the mean nearest neighbour distance (MNND) and the standard deviation of the nearest neighbour distance (SDNND). The δ15N-range gives a measure of trophic diversity within each perch species. For example, high ranges indicate that many TPs are occupied. High δ13C-ranges can indicate multiple basal resources (i.e. phytoplankton in the pelagic food web - periphyton in the littoral food web) that differ in δ13C. The area covered by the convex hull describes the occupied niche space of a consumer between the most extreme individuals in the bi-plot standard ellipse area (SEA). It has been shown that SEA is a better descriptor for niche space than convex hull area (TA) because it is less influenced by sample size (Syvaranta et al. 2013). The CD is another measure for trophic diversity, which is less influenced by outliers. MNND characterizes the density of individual packing, showing whether individuals exhibit more or less similar trophic ecologies. The SDNND is a degree of evenness in the distribution of trophic niches among individuals. Trophic niche metrics have not been calculated for small YP in Walsh Lake (n = 0), large YP in Lake Wilderness (n = 2) and small EP in Lake Wucker (n = 1) because of limited sample size. Species means of the metrics were compared between the two species with a One-Way ANOVA in SPSS (ver. 21). 25

3.4.4 Estimated Diet Composition

Mixing models of resource use were calculated to substantiate the differences in feeding ecology of the determined size classes. Since dietary niche breadth determined by stomach analysis can be misleading, SIs are ideal to capture long-term dietary differences (Copp 2008). The SI signature of consumer tissues is represents the mixture of SI signatures in all food items consumed in a certain period after being corrected for fractionation (Hobson and Clark 1992, Kelly 2000). The differences in δ15N- and δ13C- signals of prey organisms are integrated in higher-order consumers and can thus be used to calculate relative contributions of distinct food items in isotope mixing models (Fry and Sherr 1984, Vander Zanden and Rasmussen 2001). An estimate of a consumer’s diet can be obtained by comparing distributions of isotope ratios in multidimensional plots (most often two dimensional δ13C/δ15N-bi-plots). Further, diet and shifts in diet can be elucidated by looking at distinct signals of prey items (Phillips and Eldridge 2006).

Many different models exist to calculate potential diets, including different source concentrations, errors from variation in sources and sub-sampled tissues etc. (Hopkins and Ferguson 2012). The isotope mixing models allow transformation of isotopic data into resource contribution values. They are based on linear mixing models with n isotopes and n+1 sources contributing to the sample isotopic composition. To account for greater number of sources (> n+1), Bayesian mixing models have been developed (Moore and Semmens 2008, Semmens et al. 2009).

The predetermined size classes were analyzed with a Bayesian mixing model implemented in the R package SIAR (Parnell et al. 2010). The package construes linear mixing models to infer the diet composition from SI-analysis in consumers and selected food sources. Based on a Bayesian analytical framework, SIAR allows for the incorporation of a number of sources greater than n+1 isotopes analyzed. The mixing model accounts for unequal elemental concentrations in the sources, isotopic variation in the sampled sources (source process error) and the sub-sampled tissue or sample mixture (Hopkins and Ferguson 2012).

Prior to application of the mixing model, perch with marked differences in their δ13C- or δ15N-signals were grouped by cluster analyses, to account for intraspecific niche partitioning within the size classes. The cluster analyses of perch value pairs (δ13C/δ15N) were performed to increase precision in the mixing models. K-means Cluster Analyses 26

(number of clusters = 2) were carried out in SPSS (ver. 21). Resulting groups were modeled separately if they were significantly different. Only groups consisting of more than three perch were considered as distinct group along the δ13C- or δ15N-gradient. The groups are given as the cluster center in δ13C/δ15N-values (Tab. A 1). The resulting F- and p-values are used only for descriptive purposes because the clusters have been chosen to maximize the differences among cases in different clusters and significance levels are not corrected for this (sensu IBM SPSS ver. 21). Some perch could roughly be designated to littoral or pelagic habitat by the fact that they were caught with either electro fishing or hoop nets near the shoreline (littoral) or with gillnets, which were set near shore (littoral) or farther from shore (pelagic) to compare the δ13C-signatures of the respective sites.

The selection of potential food sources for the mixing model was made from the available samples with respect to published dietary data (e.g. Keast 1977, Persson 1983b and others). All sources for which more than two samples were available and that showed great overlap in the δ13C/δ15N-bi-plots were tested for distinct isotopic signals. The δ13C- and δ15N-values of the potential diets were tested separately for differences via post hoc tests (Tukey) in a generalized linear model (fixed factor species, dependent variable: δ13C- or δ15N-values) in SPSS (ver. 21). The samples were pooled if there was no significant difference in both isotopes. Most importantly, Anisoptera in Lake Padden did not differ from the Chinese mystery snail (Bellamya chinensis), which was very abundant in the littoral, and these were pooled together. Chinese mystery snails also inhabited Lake Wilderness, but they were not included in the mixing model due to extreme standard deviations in δ13C (SD=3.54). Additionally, Leptoceridae (Trichoptera) and Libelluidae (Odonata) in Walsh Lake were pooled together due to indiscriminate signals. The bivariate data of prospective food sources then defined the convex hull for which perch a diet could be modeled. Perch outside the convex hull were excluded from mixing models. The four largest fish (310 - 370 mm TL) from Lake Dölln and one large EP (379 mm) from Lake Väter were outside the convex hull defined by the sources and had to be excluded from the mixing model.

Mixing models for the American lakes were calculated for the 2012 dataset. No models were calculated for small YP (TL < 158 mm) in Walsh Lake (n = 0), large YP (TL > 158 mm) in Lake Wilderness (n = 2) or small EP (TL < 169 mm) in Lake Wucker (n = 1) because of limited sample size. The particular food sources used for mixing models in each species and lake are given in table A 2. Fractionation factors were chosen carefully 27 to account for consumer and diet specific variation (Caut et al. 2009). Many SI studies use the fractionation factor Δ13C = 0.4 (1 SD =1.3‰) given in the meta-analysis by Post et al. (2002) (e.g. Pinnegar and Polunin 1999, McCutchan et al. 2003). It has, however, been shown that it is higher for carnivores than for herbivores (Vander Zanden and Rasmussen 2001). Several studies indicate that 13C-enrichment can be much higher than 1‰. For example, in the case of EP, dietary studies produced Δ13C-values > 2.64 ‰ (Vollaire et al. 2007, Banas et al. 2009). The fractionation factors (Δ) used in the mixing model was 3.4 ± 1 for nitrogen (from Vander Zanden and Rasmussen 2001, Post 2002) and a low to medium fractionation of 0.8 ± 1.13 for carbon (Vander Zanden and Rasmussen 2001).

Each computation of SIAR for a given lake species and size consisted of 106 iterations of the Markov Chain Monte Carlo (MCMC). The outputs for each lake are given as mean proportional dietary source estimates. The percent values of single food sources were then attributed to one of the following main diet groups: ZP, MZB and fish. Members of the same diet group were then added up for each lake and size class. In addition, a mean diet contribution for the two size classes of each species was calculated.

4 Results

4.1 Trophic Position & Breakpoint Analysis

The Davies’ test or breakpoint analysis for YP (n = 68) resulted in a non-significant breakpoint estimate at TL = 157.8 mm (p = 0.228) (Fig. 7). There was a significant breakpoint for EP (n = 81) at TL = 169.3 mm (p = 0.021). Subsequently ‘small’ and ‘large’ perch are defined as fish smaller or larger than 158 or 169 mm TL, respectively. The increase in TP up to the breakpoint tended to be higher in YP, but after the breakpoint, the slope was higher for EP (Tab. 2).

4.5

3.5

2.5 P. fluviatilis 2 trophic positiontrophic P. flavescens 1.5 Linear (P. fluviatilis*) 1 Linear (P. flavescens) 0.5

0 0 50 100 150 200 250 300 350 400 total length (mm) Figure 7: Trophic positions (TP - i.e. the average position relative to primary producers at which an organism feeds) of P. fluviatilis (green triangles) & P. flavescens (yellow triangles) at increasing total length (TL) and regression of TP and TL with breakpoints (* = significant) in TPs for P. fluviatilis (dotted, green line) & P. flavescens (dotted, yellow line) for all lakes combined.

Table 2: Point estimates of the intercepts and point estimates, standard errors (st.err), t-values and confidence intervals of the slopes of each segmented relationship in the fitted model of trophic position and total length for P. flavescens & P. fluviatilis before (reg. 1) and after the breakpoint (reg. 2).

species reg. intercept slope st.err. t-value CI (95%) CI (95%) 1 1.940 0.011 0.005 2.41 0.002 0.020 P. flavescens 2 3.547 0.001 0.002 0.23 -0.004 0.006 1 3.036 0.006 0.001 4.86 0.003 0.008 P. fluviatilis 2 3.613 0.002 0.001 4.70 0.001 0.003 29

Small YP and EP had comparable trophic positions (Fig. 8a). Estimated marginal means of both species were about 3.5 (Tab. 3). Small size classes of YP did not differ significantly in their TP from small EP (p = 0.444, Wald Chi-Square = 0.57, df = 1). The TL (p = 0.001, Wald Chi-Square = 10.20, df = 1) and the lakes (p < 0.0001, Wald Chi- Square = 90.75, df = 4) had a significant influence on TP of small YP and EP. Lake Wilderness significantly influenced the course of size related TP of small YP (p = 0.039, Wald Chi-Square = 4.26, df = 1).

TL = 370 mm TL = 320 mm

YP larger than 158 mm and EP larger than 169 mm differed in TP (Fig. 8b). The mean TP of large EP was significantly higher than the position occupied by large YP (p < 0.0001, Wald Chi-Square = 102.96, df = 1 – Tab. 3). Also, TL (p < 0.0001, Wald Chi- Square = 20.64, df = 1) and lakes (p < 0.0001, Wald Chi-Square = 69.75, df = 4) had a highly significant impact on the absolute TP of large YP and EP. Also, two of the largest perch (TL = 320 mm & TL = 370 mm) exhibited very high TPs.

size group species lake n mean TP SE CI (95%) CI (95%) PA 10 3.72 0.08 3.56 3.87 WA 2 N/A N/A N/A N/A P. flavescens WI 15 2.94 0.06 2.82 3.05 3.45 0.07 3.32 3.58 small DO 19 3.74 0.05 3.64 3.85 VA 15 3.42 0.06 3.30 3.54 P. fluviatilis WU 1 N/A N/A N/A N/A 3.53 0.08 3.37 3.70

PA 10 3.56 0.08 3.41 3.71 WA 28 3.92 0.05 3.82 4.02 P. flavescens WI 3 2.90 0.14 2.63 3.18 3.46 0.06 3.35 3.57 large DO 12 4.40 0.07 4.27 4.53 VA 15 4.16 0.07 4.03 4.29 P. fluviatilis WU 19 4.09 0.06 3.98 4.21 4.22 0.04 4.14 4.29

The size range of YP (n = 68) included fish from 63 to 267 mm TL and the EP (n = 81) ranged from 62 to 402 mm TL (Tab. 4). The variation in TP was higher in small (s2 = 0.286) and large YP (s2 = 0.185) compared to small (s2 = 0.070) and large EP (s2 = 0.045). Individual TPs in YP ranged from 2.49 (TL = 127 mm, Lake Wilderness) to 4.34 (TL = 178 mm, Walsh Lake). The lowest observed TP in EP was 3.07 (TL = 67 mm, Lake Väter) and the highest was 4.96 (TL = 370 mm, Lake Dölln).

species lake size group TL δ13C δ15N TP s2 n small 135 ± 11 -25.74 ± 0.43 11.13 ±0.15 3.79 ± 0.07 0.01 10 PA large 224 ± 29 -26.67 ± 0.85 10.58 ± 0.44 3.48 ± 0.21 0.05 10 P. flavescens WA large 215 ± 23 -25.78 ± 2.40 7.37 ± 0.38 3.82 ± 0.42 0.18 19 small 127 ± 11 -29.51 ± 2.80 12.52 ± 0.32 2.99 ± 0.42 0.17 13 WI large 189 -32.54 13.27 2.82 N/A 2 small 95 ± 31 -31.22 ± 0.37 9.17 ± 0.77 3.72 ± 0.23 0.05 19 DO large 247 ± 70 -29.77 ± 1.02 9.51 ± 0.19 4.43 ± 0.29 0.08 12 small 96 ± 26 -30.00 ± 0.61 7.18 ± 0.55 3.37 ± 0.19 0.04 15 P. fluviatilis VA large 299 ± 59 -28.42 ± 0.93 9.05 ± 0.37 4.29 ± 0.16 0.03 15 small 80 -27.99 6.57 3.32 N/A 1 WU large 286 ± 59 -27.77 ± 1.07 10.96 ± 0.61 4.19 ± 0.15 0.02 19 total P. flavescens 130 ± 12 -27.88 ± 2.83 11.92 ± 0.76 3.35 ± 0.52 0.29 23 small total P. fluviatilis 95 ± 28 -30.06 ± 0.89 8.24 ± 1.22 3.56 ± 0.27 0.07 35 total P. flavescens 216 ± 25 -26.50 ± 2.54 8.78 ± 1.96 3.67 ± 0.40 0.19 31 large total P. fluviatilis 280 ± 64 -28.24 ± 1.46 9.70 ± 0.88 4.28 ± 0.22 0.05 46

No general correlations for YP total length and isotopic signatures were found in YP (Fig. 9). YP and all other samples (fish and invertebrates) from Lake Wilderness and Lake Padden had very high δ15N-levels (Tab. 4). A weak positive relationship between δ15N- values and TL was found in Lake Wilderness (R2 = 0.408, p = 0.010, F = 8.97, df = 13 - Fig. 9e). 13C-values negatively correlated with TL in Lake Padden (R2 = 0.297, p = 0.013, F = 7.59, df = 18 - Fig. 9b).

Lake Padden 12.0 δ15N (‰) -24 δ13C (‰) 11.5 -25

11.0 -26

10.5 -27

10.0 -28 a) b) 9.5 -29 100 150 200 250 100 150 200 250

15 8.5 δ N (‰) Walsh Lake -15 δ13C (‰) 8.0 -20 7.5

7.0 -25 6.5 -30 6.0 c) d) 5.5 -35 150 200 250 150 200 250

14.5 δ15N (‰) Lake Wilderness -23 δ13C (‰) YP (gill net) -25 YP (fyke net) 13.5 YP (2009) -27 12.5 -29 -31 11.5 -33 e) f) 10.5 -35 50 100 150 200 250 50 100 150 200 250 total length (mm) total length (mm)

Figure 9: Linear relationships between isotopic signatures (δ15N & δ13C) of muscle tissue from P. flavescens (YP) and total length (TL) in Lake Padden (a & b), Walsh Lake (c & d) and Lake Wilderness (e & f). Yellow triangles represent samples from YP, caught with gill nets, yellow- green triangles are YP caught with hoop nets and brown triangles are YP from the 2009 dataset collected by Eric Larson (University of Washington, unpublished data).

Strong positive relationships were found between total length and δ15N or δ13C-values for EP (Fig. 10). In Lake Dölln (R2 = 0.702, p < 0.0001, F = 68.17, df = 29 - Fig. 10a) and Lake Väter (R2 = 0.776, p < 0.0001, F = 97.07, df = 28 - Fig. 10c). δ15N strongly correlated with TL, but not in Lake Wucker (R2 = 0.166, p = 0.074, F = 3.59, df = 18- Fig. 10e). A positive correlation between δ13C and TL was found for all EP in all lakes (Lake Dölln: R2 = 0.879, p < 0.0001, F = 209.69, df = 29; Lake Väter: R2 = 0.757, p < 0.0001, F = 87.38, df = 28; Lake Wucker: R2 = 0.635, p < 0.0001, F = 34.12, df = 18 - Figs. 10b,d,f). 33

Lake Dölln δ15N (‰) -27 δ13C (‰) 12.0 -28 11.0 -29 10.0 -30 9.0 -31 8.0 -32 a) b) 7.0 -33 50 100 150 200 250 300 350 400 50 100 150 200 250 300 350 400

Lake Väter 10.5 δ15N (‰) -27 δ13C (‰) -28 9.5 -29 8.5 -30 7.5 -31 EP (gill net) EP (other) 6.5 -32 c) d) 5.5 -33 50 100 150 200 250 300 350 400 50 100 150 200 250 300 350 400

10.0 δ15N (‰) Lake Wucker -24 δ13C (‰) -25 9.0 -26 8.0 -27 -28 7.0 -29 e) f) 6.0 -30 50 100 150 200 250 300 350 400 50 100 150 200 250 300 350 400 total length (mm) total length (mm)

4.2 Trophic Niche Metrics

There were no significant differences in trophic niche metrics between the two species of one size class (Tab. A 3). Some trends could be observed: Although, highly variable, the δ15N-range, and therefore the range of occupied trophic niches tended to be lower in YP within lakes than in EP (Tab. 5). By contrast, the δ15N-range showed a trend to be larger for small and large EP than for YP. YP were characterized by a great range of δ13C- 34 signals, especially in Walsh Lake and Lake Wilderness. The δ13C-range was wider for small and large YP than for EP. The occupied hull area was larger in YP. MNND was higher in small and large YP with relatively uneven distribution of individual niche space (SDNND).

species lake size group δ15N-range δ13C-range hull area CD MNND SDNND small 0.46 1.30 0.36 0.38 0.20 0.11 PA large 1.39 2.50 2.06 0.82 0.38 0.38 P. flavescens WA large 1.33 6.96 5.20 2.18 0.32 0.26 WI small 1.13 7.43 5.25 2.60 0.51 0.37 small 2.61 1.12 1.69 0.65 0.19 0.11 DO large 2.00 3.01 2.57 1.04 0.32 0.18 P. fluviatilis small 2.20 1.95 2.56 0.70 0.35 0.22 VA large 0.67 3.25 1.38 0.78 0.27 0.25 WU large 1.26 3.62 2.81 0.95 0.30 0.21

4.3 SIAR - Mixing Model

In most of the lakes, marked differences in δ13C- and δ15N-signals were found. In all lakes, 13C-enriched (littoral) and 13C-depleted (pelagic) groups could be discriminated by K- means Cluster Analyses and mixing models were calculated separately for these groups. Table A 1 gives an overview of the different clusters and resulting groups for each lake. The groups of perch found by K-means Cluster Analyses are marked with different symbols in figures 11 and 12.

The position of all sampled perch in the bi-plot was between two to three trophic levels (~Δ15N = 3.4) above the invertebrates. Figures 11a-c display the trophic postioning of the perch and their prey organisms with the calculated baseline for each lake. 35

a) δ15N (%) Lake Padden small YP 1 12 small YP 2 large YP 1 10 large YP 2 C. asper 8 small C. asper Calanoida Cladocera 6 P. leniusculus Zygoptera 4 Trichoptera Chironominae 2 Gammaridae hull area - small YP Bellamya chinensis hull area - large YP baseline 0 -31 -29 -27 -25 δ13C (%)

Figure 11a

Walsh Lake δ15N (%) b) 9 large YP 1 large YP 2 C. asper 7 small M. salmoides Calanoida 5 Cladocera Anisoptera 3 Zygoptera Leptoceridae

1 Chironominae hull area - large YP Gammaridae baseline -1 -33 -31 -29 -27 -25 -23 δ13C (%)

Figure 11b 36

Lake Wilderness c) δ15N (%) small YP 1 small YP 2 14 large YP C. asper small C. asper 12 Calanoida Cladocera 10 P. leniusculus Anisoptera

8 Zygoptera Leptoceridae Chironominae 6 Gammaridae Bellamya chinensis hull area - small YP 4 baseline -39 -37 -35 -33 -31 -29 -27 -25 -23 -21 δ13C (%)

In general, YP were rather scattered along the horizontal δ13C-axis. In Lake Padden large YP were more depleted in δ15N and δ13C than small conspecifics (Fig. 11a). In Walsh Lake, YP split into δ13C-depleted and δ13C-enriched individuals (Fig. 11b). YP in Lake Wilderness also exhibited a split in δ13C-values, with a 13C-depleted group in contrast to 13C-enriched individuals (Fig. 11c).

EP tended to be ordered along the vertical δ15N-axis. The largest individuals were positioned in the upper right corner of the bi-plots. Small EP and roach had the same position in the bi-plots of Lake Dölln (Fig. 12a) and Lake Väter (Fig. 12b). Small EP exhibited a group of larger 15N enriched individuals in Lake Dölln (Fig. 12a) and Lake Väter (Fig. 12b). The large individuals were scattered along the horizontal δ13C-axis, which was especially distinctive in in Lake Wucker (Fig. 12c). The largest individuals were depleted in 15N and enriched in 13C (e.g. large EP 2 - Fig. 12a). 37

δ15N (%) Lake Dölln a) small EP 1 12 hull area - small EP hull area - large EP small EP 2 10 large EP 1 large EP 2 8 R. rutilus Chaoborus spec. 6 Zooplankton O. limosus (repl.) 4 Zygoptera Trichoptera 2 Ephemeroptera Chironominae 0 Asellidae baseline -2 -37 -35 -33 -31 -29 -27 -25 δ13C (%)

Figure 12a

δ15N (%) Lake Väter b) 11 hull area - small EP small EP 1 hull area - large EP small EP 2 9 large EP 1 large EP 2 7 R. rutilus Chaoborus spec. Zooplankton 5 O. limosus (repl.) Zygoptera 3 Trichoptera Ephemeroptera 1 Chironominae Asellidae baseline -1 -36 -34 -32 -30 -28 -26 -24 13 δ C (%) Figure 12b 38

δ15N (%) Lake Wucker c) 11 small EP hull area - large EP large EP 1 9 large EP 2 R. rutilus 7 Chaoborus spec. Zooplankton 5 O. limosus Zygoptera 3 Trichoptera Ephemeroptera Chironominae 1 Amphipoda baseline -1 -35 -33 -31 -29 -27 -25 δ13C (%)

Similar shares of ZP, MZB and fish resources were found in small YP, EP and large YP, while large EP exhibited lower shares of MZB and higher shares of fish (Fig. 13). The contributions of all sources used in the mixing models explained between 94% (group one Lake Wucker) to 100% of the perch isotopic signals. Detailed results of single source contributions of all groups are given in the annex, figures A 1a-h and figures A 2a-i. 39

100%

80%

60% Fish 40% Zoobenthos Zooplankton 20%

0% P. flavescens P. fluviatilis P. flavescens P. fluviatilis (n=23) (n=35) (n=29) (n=41)

small perch large perch

Some notable differences existed between different clusters within lakes, especially in the ones divided along the 13C-gradient (Fig. 14a&b). Small YP (TL < 158 mm – n = 23) in Lake Padden (22 & 24%) and group one of Lake Wilderness (21%) had similar values for ZP in the while the second group in Lake Wilderness showed more than 50% ZP. Small EP (TL < 169 mm) from the second group in Lake Dölln had a very high contribution of fish resources (35%) in contrast to group one (19%). In Large YP (TL > 158 mm) group one of Walsh Lake (9%) exhibited much lower shares of ZP in the diet than group two (38%), but they had higher shares of MZB. In group one of large EP (TL > 169 mm) in Lake Wucker, fish accounted for 29% of the diet, while group two exhibited 49%.

small perch 100%

80%

60%

40%

20%

a) P. flavescens P. fluviatilis

large perch 100%

80%

60%

40% fish

20% zoobenthos zooplankton 0%

b) P. flavescens P. fluviatilis Figure 14: Percentage of zooplankton (yellow), zoobenthos (brown) and fish (green) in the diet of small (a) and large (b) P. flavescens & P. fluviatilis for each lake (PA = Lake Padden, WA = Walsh Lake, WI = Lake Wilderness, DO = Lake Dölln, VA = Lake Väter & WU = Lake Wucker). The diet within each size class is given for each group (1/2) with similar isotopic composition, as defined in preliminary cluster analysis. Sample size (n) is given in brackets.

5 Discussion

In line with the first hypothesis, the breakpoint analysis revealed a size-related increase of TP in both species and similar breakpoints in TP. In contrast to the second hypothesis, YP and EP markedly differed in the course of size-related trophic positioning and resource use. While YP remained at medium TPs in large size classes using similar resources, the mean TP of EP steadily increased and they became progressively piscivorous. On average small and large size classes of YP relied on the same resources at similar proportions. Further, YP exhibited great individual- and lake-specific variability in TPs and the range of individual δ13C-signals. EP showed more stable patterns and tended to have narrower realized niches.

YP and EP undergo similar ontogenetic niche shifts from ZP to MZB to fish, linked to their respective TPs (Hjelm et al. 2000, Amundsen et al. 2003, McIntyre et al. 2006, Kraemer et al. 2012). Persson et al (2000) coined the term ‘ontogenetic omnivore’ to describe this feeding ecology, in EP. In this study, breakpoints in TP within the range of 160 to 170 mm TL were identified, marking points where TP starts to level off. Up to these points growing perch increasingly consume higher-order organisms (Craig 2008). In the size range of measured perch (62 to 402 mm TL), the breakpoints probably indicate the second trophic niche shift to piscivory (Mittelbach and Persson 1998). The first shift from ZP to MZB generally occurs at smaller sizes (Wu and Culver 1992, Wahlström et al. 2000, Berezina and Strel'nikova 2001, Graeb et al. 2006). At these breakpoints the increment in TP (i.e. slope) is attenuated because the trophic level of the consumed prey, large, predatory invertebrates and fish, does not increase further. The TPs of perch, however, still increases with size, since more higher-order consumers are eaten, but at a lower slope. Simultaneously, δ13C and δ15N increased with TL of EP. This enrichment also reflects the ontogenetic dietary shifts i.e. increasing reliance on littoral resources and fish (Quevedo and Olsson 2006, Quevedo et al. 2009).

These results, nevertheless, question the ecological equivalence of YP and EP as proposed by Thorpe (1977). Both species markedly differed in the course of size related trophic positioning. Up to the breakpoint, they occupied similar TPs, but above the breakpoint, EP further increased in TP, while YP occupied the same TPs as their smaller conspecifics. In Lake Padden, YP even slightly declined in TP. The average TPs of about 3.5 calculated for small perch of either species matches results of other studies (Quevedo and Olsson 42

2006, Browne and Rasmussen 2009). Thus, it can be concluded that small size classes of both perch species take up similar foraging niches or at least utilize analogous resources. Statistical power to detect changes between size classes of YP at the breakpoint was low, due to the high degree of variability in TP. Especially small YP from Lake Wilderness strongly increased the slope and lowered the intercept of the first regression. However, YP remained at mean TPs of about 3.5, even at larger sizes. Other studies on YP also suggest such marginal shifts of less than 0.5 in TP (McIntyre et al. 2006, Bertrand et al. 2011). The TPs and trophic niche characteristics point towards a flexible, generalist feeding strategy or a high degree of omnivory in YP.

The trophic niche metrics (Layman et al. 2007, Jackson et al. 2011) i.e. occupied hull area, δ15N- and δ13C-range support this assertion. All metrics tended to be larger for YP, besides the nitrogen range in small YP. This was also true for MNND in small and large YP, which showed relatively uneven distribution of individual niche space (SDNND). This means that, trophic diversity and realized niche might be wider in YP than in EP. High values in those characteristics are typical for omnivorous fish (e.g. Oreochromis niloticus) (Zambrano et al. 2010). The broad hull area of YP is mostly stretched out by the wide range of δ13C and large intrapopulation variation in δ15N, which is again reflected in differences of more than 1.5 in TP among YP. TPs and the heterogenous resource use of YP could be explained by individual specialization in different niches. Populations of generalist consumers are thought of as assemblages of individual specialists (Bolnick et al. 2007). YP have been shown to differ in individual resource use (Parker et al. 2009a, Roswell et al. 2013). Individual diet strategies are known in EP, promoting different fitness but relieving from intraspecific competition (Beaudoin et al. 1999, Bolnick et al. 2003, Svanback and Bolnick 2007).

The δ15N-levels of YP were expected to increase during their ontogeny (e.g. Kraemer et al. 2012). However, no supporting evidence was found. Individual body condition might have affected δ15N-values (Fuller et al. 2005, Kempster et al. 2007). For example, pumkinseed sunfish (Lepomis gibbosus) exhibited increased δ15N-values with declining condition factor (Colborne and Robinson 2013). In Lake Padden, YP showed an unsuspected pattern and became depleted in 15N. Only the largest fish showed a rise in δ15N-content. Such observations have also been made in Dore Lake, Canada, where littoral YP showed comparable patterns in δ15N-length regressions (Barks et al. 2010). The high δ15N-levels observed in small YP in contrast to large perch could be attributed 43 to either naturally 15N-enriched pelagic food sources (i.e. Chaoborus spec.) or a higher trophic level of these perch (Vander Zanden and Rasmussen 1999, Quevedo et al. 2009). Subsequently, δ15N-values would decrease when MZB are consumed and increase when crayfish and/or fish are eaten. The high variability in δ15N-levels in small size classes, however, could result from early cohort splitting into piscivores and planktivores (Post et al. 1997, Urbatzka et al. 2008).

The range of calculated TPs for EP are within limits determined for other populations (e.g. Quevedo and Olsson 2006, Quevedo et al. 2009, Akin et al. 2011). EP had a significant breakpoint at about a TP of 3.6 and a length of TL = 169 mm. Also, they exhibited a more stable pattern than YP. This ontogenetic diet shift in EP seems to be rather gradual, probably by increasing shares of fish prey. Due to the increase and high TPs of large and very large individuals (max. TP = 4.96), it can be concluded that EP take up the position of a primary piscivorous fish with size and age when prey fish are abundant (Dörner et al. 2003). The strongly piscivorous and cannibalistic perch, so called ‘giants’, have been widely discussed due to their impact on population dynamics and community structure (Claessen et al. 2000, Persson et al. 2003, Persson et al. 2004). Very large YP were not present in the samples and also have not been reported to have similar impacts.

Their very large, piscivorous size class, could be termed ‘ontogenetic specialists’. Modeled dynamics of size structured populations (i.e. perch) assume that fish would gradually shift diets, resulting in either a generalist strategy or two specialist strategies (Claessen and Dieckmann 2002). This complies very well to the observations made for EP, which exhibited size specific diets and TPs. However, adult YP showed a different pattern and behaved more like a generalist. A generalist feeding modality is to be expected if the possible intake rate increases more slowly as body size increases in the first niche than in the second niche (sensu Claessen and Dieckmann 2002). The differences in TPs of large perch becomes clear with regard to the importance of piscivory in both species.

The mixing model outputs corroborate the patterns found within TPs. For YP, such mixing models have been proven to reflect the diet very well, when compared with stomach analysis (McIntyre et al. 2006). Small and large YP exploited the same resources and exhibited a similarly broad diet breadth, which might indicate opportunistic feeding behavior. Piscivory, however, was not prevalent in YP. In YP, the share of fish prey, and 44 correspondingly TP, did not increase with size as it was expected. Fish comprised only a minor share of the resources used by small and large YP. The very abundant sculpins could have been an important prey for YP, as other studies have found extensive feeding on sculpins or gobies (Truemper and Lauer 2005, Truemper et al. 2006, Weber et al. 2011). For example, stomach analyses and isotope mixing models of YP in Lake Washington, USA demonstrate that large perch increasingly consume prickly sculpins (C. asper) (Costa 1979, McIntyre et al. 2006). Other studies have also reported a minor importance of fish prey in the diets of YP (e.g. Jansen and Mackay 1992, Lott et al. 1996, Legler 2008, quoted in Guzzo et al. 2013). YP from acidified lakes, which were restricted in their ability to utilize benthic prey, did not exploit fish to a great extent (Luek et al. 2010). One possible explanation for the low levels of piscivory by comparison to EP might be bulky prey fish. YP prefer smaller prey even if they are not gape limited (Truemper and Lauer 2005). The most abundant prey fish in the studied lakes were centrachids, which is true for many fish communities in North America (Werner et al. 1977, Keast 1978). These rather high bodied fish are often shunned by piscivorous fish (Bolding et al. 1998). The presence of suitable, more slender prey fish might shift the preference in adult YP towards fish prey. Studies on the piscivory of YP emphasize the importance of slender prey fish like brook sticklebacks (Culae inconstans), johnny darters (Etheostoma nigrum) or clupeids (Fish and Savitz 1983, Knight et al. 1984, Fullhart et al. 2002). For example, in Dore Lake, Canada, nine-spined sticklebacks (Pungitius pungitius) seem to represent an ideal diet for age-4+ YP, resulting in over 80% fish in their diet (Barks et al. 2010). In their native range, YP can exploit a wide range of fish prey. In Lake Michigan, USA large YP consume alewife, johnny darter, mottled sculpin, rainbow smelt, round goby, spottail shiner and several other species, most of which are not present in the studied lakes (Truemper et al. 2006).

The observed lake effects seem to be particularly evident for the mean TP of YP. Abiotic factors such as habitat complexity (e.g. abundance of macrophytes) or basin morphometry can influence perch feeding ecology (Diehl 1992, Vander Zanden and Vadeboncoeur 2002). In Lake Wilderness for instance, the lowest degree of piscivory was found, which exhibited very steep slopes and mostly narrow littoral areas. More than 250 perch with a TL < 150 mm and only two bigger individuals were caught. Large pelagic and small littoral zones can create stunted populations under high intraspecific competition (trophic bottleneck) (Persson 1987b, Heath and Roff 1996). Stunted EP feed mainly on MZB and 45 do not perform ontogenetic shifts to piscivory (Ylikarjula et al. 1999). Lake Wilderness’ YP might thus be stunted, but this remains to be confirmed. Further, the importance of zoobenthic or zooplanktonic prey has been shown to be related to morphometric lake characteristics. For example, smaller lakes generally exhibit higher degrees of zoobenthivory (Vander Zanden and Vadeboncoeur 2002). The morphometry of Lake Wilderness, which is characterized by steep slopes and a relatively small littoral zones, might promote zoobenthic and zooplanktonic specialization.

By contrast, large EP had more stable feeding patterns and consumed greater amounts of fish prey. Mixing models predicted high shares of fish prey (> 40%) in large EP. With respect to small EP, the determined breakpoint value could thus be seen as a good estimator for the second ontogenetic shift to piscivory. It has been reported that fish start to be the main food item of either littoral or pelagic EP exceeding 170 to 180 mm TL (Horppila et al. 2000, Heibo et al. 2005). For example, large, subarctic EP fed on more than 75% fish prey (Amundsen et al. 2003). Above the breakpoint, EP seem to have a large enough gape width to cope with profitable shares of the available prey fish species (Mittelbach and Persson 1998). Their gape width exceeds gape height at about TL = 120 mm length and is linearly related to prey length (Dörner and Wagner 2003). More slender prey fish like roach might be available for EP than comparable prey for YP in the sampled lakes. Also, some EP skip the second diet shift and prey preferentially on fish (Dörner et al. 2001). This might explain why two out of the four groups of small perch in Lake Väter and Lake Dölln had higher δ15N-values and larger shares of fish prey than their conspecifics. These groups included mainly larger individuals near the onset of piscivory.

Large EP had distinct positions in the bi-plots. However, some individuals were positioned outside the convex hull, defined by the sources. Therefore, no mixing models could be calculated, for those fish, which were among the largest fish of the samples (310 - 370 mm TL) and particularly high δ13C-values. One possible reason is the existence of a rather 13C-enriched food source that was not part of the sampled prey items. Also, exceptionally large perch are thought to be true piscivores that often have faster growth than their conspecifics. This again can lead to increased incorporation of 13C (Lecren 1992, Zuanon et al. 2006). Additionally, old perch grow slowly and thus have low tissue turnover rates (Tieszen et al. 1983). The diet reflected in the measured signal is probably skewed by accumulated 13C-isotopes. Finally, the carbon fractionation factors selected for the model might be too low, because carbon enrichment could be higher in perch than 46 given standards. Determination of fractionation factors in feeding trails found very high values for 13C in perch (Vollaire et al. 2007, Banas et al. 2009).

Generally, the often recited, intermediate TP of perch is reflected in the high shares of secondary consumers (i.e. invertebrate prey) in the diet (Cabana and Rasmussen 1996, Vander Zanden et al. 1997, Quevedo and Olsson 2006). Interpreting the results of the mixing model, MZB were the most important prey for small size classes of both species and the large YP. Crayfish and odonats made up the highest shares of benthic resources in YP. Smaller invertebrates like chironomids, isopods and amphipods were found to comprise most of the prey in intermediate size classes of YP (Fullhart et al. 2002). Large bodied invertebrates like odonats resemble another important prey and mainly determine trophic level in YP (Keast 1985). For example, odonats alone can account for more than 60% of the diet of age-1+ YP in a comparable lake system (Weidel et al. 2008). Also in EP, stomach analysis confirm that ephemeropterans, chironomids and odonats are the most important invertebrate prey in Lake Dölln and Väter (Radke 1998). The contribution of other invertebrates in small and large YP was similar to EP. Chironomids, however, seem to be underrepresented in SI mixing models of both species, because many studies have emphasized the importance of this prey, especially in smaller perch (Keast 1977, Persson 1983b). For example, in a smaller eutrophic lake up to 80% or 70% amphipods and chironomids (larvae & pupae) were found in the diet of YP at different times of the day (Jansen and Mackay 1992). Chironomid pupae were not found in the MZB samples and thus were not part of the mixing model. Also, chironomids might be more important to size classes < 60 mm.

Some of the patterns found in the importance of MZB for YP might be explained by interspecific competition. Different competition within the fish communities could influence niche segregation (Mahon 1984). Fishes from the family centrarchidae are generalists and omnivorous predators like perch, which have been suspected to limit the trophic impact of YP in North America (Kitchell et al. 1977, Cooke and Philipp 2009). YP and sympatric centrachid species react to the dietary overlap by diet segregation to a certain degree (Keast 1978). The different fish communities of the American lakes probably caused lake effects on TP. Sunfish species populated the littoral areas of Lake Wilderness and Lake Padden. Sympatric pumpkinseed sunfish (L. gibbosus) and bluegill sunfish (Lepomis macrochirus) have been shown to outcompete YP in foraging on MZB in littoral habitats and to induce diet shifts to a ZP dominated diet (Hanson and Leggett 47

1985, 1986, Kaemingk and Willis 2012). Invasive sunfish can even be superior competitors to EP where they occur together (Fobert et al. 2011). In Walsh Lake, however, the only centrachid species is largemouth bass (M. salmoides) (Julian D. Olden pers. comm.). In the absence of sunfish, YP preferentially forage on MZB in the littoral areas (Wall and Blanchfield 2012). This could explain why YP in Walsh Lake utilize littoral, benthic prey to such a great extent.

Although large EP showed a clear tendency towards fish prey, MZB probably makes up for more than half of the diet in small to intermediate EP and about a third in large EP. EP especially consumed high proportions of the crayfish Orconectes limosus. Crayfish are an important trophic-link to higher-order consumers like perch, since they are an important food item to many predatory fish (Momot 1995, Nystrom et al. 1996). For example, large size classes of EP in Australia almost exclusively fed on crayfish (Morgan et al. 2003). The importance of the crayfish (i.e. O. limosus), especially for large EP within the studied lake systems has been emphasized by several stomach analyses (e.g. Radke 1998, Radke and Eckmann 2001, Schulze et al. 2012). Bioenergetic models and stomach analysis suggest that crayfish account for almost half of the annual consumption in EP of Lake Väter, where they comprise half of its MZB biomass (Haertel-Borer et al. 2005). The consumption of crayfish varies over the seasons and is highest in summer, which should be reflected in the tissues sampled in this study (Haertel et al. 2002). However, due to a lack of samples, calculated values for Lake Väter and Lake Dölln were based upon crayfish replicates of Lake Wucker. These lakes do not strongly differ in the isotopic content of samples like the American lakes, but the model remains skewed due to lake-specific isotopic signals. Nevertheless, O. limosus is likely to be an important food items for all sampled populations of EP. The patterns found in Lake Wucker and the American lakes show that crayfish are the only prey organisms which are 15N-enriched and highly 13C-enriched, and can thus explain a great share of EP isotopic composition.

ZP was exploited differently by the classified groups in YP. The importance of calanoids and cladocerans has been shown in Canadian Shield lakes, where they have been the most important prey item throughout the seasons (Johnson 2001). Water transparency potentially could determine zooplanktonic proportions of the diet (Bartels et al. 2012). Transparency was the highest in Lake Wilderness, which might support the high reliance on copepods present in small perch. ZP are often important prey for small YP, but they can be essential for larger perch, too. For example, large ZP like mysid shrimps 48

(Acanthomysis awatchensis) or Leptodora kindtii are intensively fed upon by large YP (Serns and Hoff 1984, McIntyre et al. 2006). Chaoborus spec. have been present in many American zooplankton samples, but did not provide enough material for SI samples.

The degree of zooplanktivory was generally constant in small and large EP. Radke (1998), however, found great differences in the use of ZP among small and large individuals of Lake Väter. The mixing model suggests that Chaoborus spec. is particularly important for large and small EP. Also, L. kindtii is preyed upon in Lake Väter and they probably have a similar isotopic composition as Chaoborus spec.(Haertel 2001, Haertel et al. 2002). Large, predatory, ZP seem to resemble important prey for perch of all size classes, although consumption is strongly determined by availability of ZP (Dörner et al. 2001). In addition, the fish community of Lake Wucker includes vendace (Coregonus albula), a pelagic planktivorous fish that has not been caught during sampling (Kottelat and Freyhof 2007, LELF 2012). Vendace are prey fish for EP where they occur (Beier 2001, Haakana et al. 2007). These prey fish might explain the disproportionately high shares for ZP, which were computed for the large size class of EP in Lake Wucker.

In general, perch are highly mobile and have been shown to shift between pelagic and littoral habitats (Fish and Savitz 1983, Eklöv 1997, Zamora and Moreno-Amich 2002). In YP, stomach- and SI-analyses revealed a high degree of variability up to 95% in resource use of different size classes. YP inconsistently utilized littoral (MZB) and pelagic prey items (ZP) (Browne and Rasmussen 2009). The range of δ13C-signals within a population in comparison with the range of δ13C-signals in their diet can inform about differences in individual variation, like the resource polymorphism in perch (Matthews and Mazumder 2003, Bertrand et al. 2011). Some of the studied perch populations showed potential pelagic and littoral subpopulations, demonstrated by the clustering of value pairs along the 13C-gradient. The wide δ13C-ranges of small and large YP indicate multiple basal resources, i.e. pelagic and littoral resources (Quevedo et al. 2009). Lake Wilderness’s small perch and Walsh Lake’s large perch exhibited two very distinct groups, with high shares of either 13C-enriched, littoral prey or of 13C-depleted ZP and profundal prey (i.e. chironomids) in the diet. The existence of potential subpopulations is corroborated by three perch that were caught with a hoop net in a shallow, littoral area of Lake Wilderness. They had the lowest δ13C-signals of all sampled YP in Lake Wilderness. 49

Also, EP show signs of resource bimodality. Notably the larger individuals, differed in the range of δ13C-signals and share of ZP in the diet, which suggests specialization on littoral or pelagic resources (Svanback and Bolnick 2007, Quevedo et al. 2009). In EP, intraspecific competition is thought to be the driving force of this process (Svanback and Bolnick 2007). However, roach can also influence habitat and diet segregation in EP, since they are superior competitors for ZP (Persson and Greenberg 1990a, Persson and Greenberg 1990b). The marked splits in δ13C-signals and patterns of δ15N among the defined groups could be linked to differentiation into littoral and pelagic or profundal subpopulations (Quevedo et al. 2009, Barks et al. 2010).

6 Conclusions

The new information on perch feeding ecology contributed by this study has several implications. It can help to explain population dynamics of the studied lakes (Persson and Greenberg 1990a, Jansen 1996, Olson et al. 2001, Schoenebeck and Brown 2010). For example, piscivory by YP can promote growth and thus the occurrence of larger size classes, desired by fishermen (Barks et al. 2010). Due to the lack of piscivorous individuals, a poor to moderate growth might thus be predicted for the examined YP populations. Additionally, knowledge on feeding ecology is needed to assess ecosystem impacts since both species have become an issue for ecosystem management (Leunda et al. 2008, Olden et al. 2008, Layman and Allgeier 2012). YP and EP are invasive species and successfully invaded new ecosystems in North America, Southern Europe, South Africa and Australia (Elvira and Almodóvar 2001, Champion et al. 2002, Lintermans 2004, Schade and Bonar 2005, Orrù et al. 2010). They altered local community structures and caused regime shifts (Post and Cucin 1984, Arthington 1991). It was shown that EP become top predators in large size classes. This is why EP can severely affect native fish communities in New Zealand and Australia (Closs et al. 2001, Wilson 2005). Risk assessments state that YP have a potential for “very high impacts”, especially in small lake ecosystems in the Pacific North West of North America (Johnson 2009, DFO 2011). For example, YP have been shown to outcompete the salmonid communities that are native in the water bodies of the region (Fraser 1978, Browne and Rasmussen 2009). Highly variable feeding ecology among populations of different water bodies is thought to promote invasions (Ribeiro et al. 2007). The high variability and the trophic adaptability found in YP might be one reasons for their competitive superiority. Effects of management measures, for instance removal of YP to improve salmonid stocks, are not fully understood. The outcomes differ and might be attributed to this high degree of trophic flexibility of YP (Rumsey et al. 2007, Browne and Rasmussen 2012).

Although, YP and EP live in comparable habitats, continent-specific, physical habitat characteristics or differences in the fish community can have an influence on fish traits (Mahon 1984, Lamouroux et al. 2002). Linking the results, TP estimates and mixing models of diet in adult YP suggest that in the lakes of Western Washington, YP are not as piscivorous as their European relatives. Among large perch, none of the examined YP populations reflected the patterns of EP. They seem to exhibit a broader resource utilization and more omnivorous feeding ecology than EP. The reviewed literature from 51 dietary and SI analysis of both species corroborate these findings. However, the conditions under which large YP exhibit piscivory (e.g. Fullhart et al. 2002, Truemper et al. 2006) or where they do not (e.g. this study, Jansen and Mackay 1992, Lott et al. 1996), remain to be clarified. Also, potential bimodality in resource use of YP has rarely been studied and calls for more investigations (i.e. Parker et al. 2009a). By contrast, adult EP are suggestive of increasing specialization on piscivory. These findings imply different top-down impacts of introduction or removal of either species to the affected ecosystem via trophic cascades (Huryn 1998, Potthoff et al. 2008). Piscivorous fish like EP substantially affect the water quality of lakes (Mehner et al. 2002). The studied EP populations thus probably have greater top-down impacts on the ecosystems than YP. Additionally, trophic cascades in perch could positively affect primary producer levels by increasing nutrient availability (Attayde and Hansson 2001, Leroux and Loreau 2010). Finally, the ecological roles ascribed to YP and EP should be seen with respect to local conditions. In the course of invasions, for instance, resource shifts are possible and feeding ecology of populations might be hard to predict (e.g. Brandner et al. 2013). The functional niche of YP and EP might show great overlap, but the realized niches of different populations probably differ strongly due to competition, predation, the size spectrum or different prey items. The results refute the assumed ecological equivalence of YP and EP and suggest an independent view on the feeding ecology of either species. Top-down interactions and trophic links within food webs of EP and YP should be further examined to understand their complex feeding ecology.

7 Summary

The closely related North American yellow perch (Perca ﬂavescens) and the Eurasian perch (Perca fluviatilis) share many similarities in morphology, life history traits, environmental tolerance and trophic requirements, including ontogenetic diet shifts. Consequently, the functional foraging niches of these species are hypothesized to be similar. However, their feeding ecology has never been compared directly. Knowledge of the ecological and trophic niches of Eurasian perch and yellow perch is needed to better understand their role in ecosystems and their invasive potential. This study aimed at comparing the feeding ecology and characterizing the functional role of yellow perch and the Eurasian perch in North America and Europe by means of stable isotope analysis. The stable isotope ratios of nitrogen and carbon (15N and 13C) represent powerful ecological tools to trace consumer foraging history and determine trophic position. Similarities in the transition of diets and magnitude of trophic position were expected. In contrast to expectations, marked breakpoints in trophic position during growth and an ontogenetic diet shift occurred in Eurasian perch, but not in yellow perch. Also, large Eurasian perch occupied higher trophic levels than adult yellow perch. Eurasian perch became increasingly piscivorous after they reach about 160 mm total length. The species comparison across a similar size range revealed high variability in trophic position among individual yellow perch and a more size related trophic position in Eurasian perch. Large yellow perch seem to be more opportunistic and omnivorous predators than their European counterpart, which exhibited rather stable patterns of resource use. The elevated trophic positioning of Eurasian perch suggests that they may exert a more pronounced top-down trophic control of populated ecosystems, which would influence water quality. High flexibility in diet of yellow perch and strong piscivory in large Eurasian perch might favor their ability to invade ecosystems. The functional niche of YP and EP might show great overlap, but the realized niches of different populations probably differ strongly due to competition, predation, the size spectrum or different prey items, which remain to be clarified.

8 Zusammenfassung

Der Amerikanische Flussbarsch (Perca ﬂavescens) und der Eurasische Flussbarsch (Perca fluviatilis) weisen viele Gemeinsamkeiten auf. Sie ähneln sich in ihren Lebenszyklen, Ansprüchen an ihre abiotische Umgebung und in Bezug auf ontogenetische Nischenwechsel in ihrer Ernährung. Aus diesem Grund wird beiden Arten dieselbe Rolle im Ökosystem zugeschrieben. Funktionelle Eigenschaften beider Barscharten oder deren Rolle im Ökosystem wären somit vergleichbar. Die Nahrungsökologie beider Arten wurde jedoch noch nie interkontinental verglichen. Um ihre Bedeutung für das Ökosystem zu verstehen und ihr Invasionspotenzial besser abschätzen zu können, bedarf es Wissen über die trophische Nische beider Arten. Diese Studie sollte mithilfe stabiler Isotopenanalyse die Ernährungsökologie beider Arten beziehungsweise die sich daraus erschließende Funktion der Fische in den entsprechenden Ökosystemen im interkontinentalen Vergleich betrachten. Mithilfe der stabilen Isotope von Stickstoff und Kohlenstoff (15N und 13C) kann die Nahrungsaufnahme rekonstruiert und die trophische Stellung bestimmt werden. Es wurden ähnliche Wechsel der Ernährung in Abhängigkeit von der Größe der Barsche sowie eine vergleichbare Einnischung beider Arten erwartet. Die Studie konnten dabei deutliche Veränderungen in der trophischen Position während des Wachstums sowie ontogenetische Nischenwechsel bei Eurasischen-, nicht jedoch bei Amerikanischen Flussbarschen feststellen. Adulte, Eurasische Flussbarsche nahmen höhere trophische Positionen als Amerikanische Flussbarsche vergleichbarer Größe ein. Der Vergleich eines ähnlichen Größenspektrums zeigte eine weitaus höhere Variabilität in der trophischen Position der Amerikanischen Flussbarsche und eine deutlich größenabhängige trophische Position der Eurasischen Flussbarsche. Große Amerikanische Flussbarsche scheinen opportunistisch und eher omnivor in ihrer Ernährung zu sein verglichen mit Eurasischen Flussbarschen, welche ab etwa 160 mm Gesamtlänge zunehmend piscivor wurden. Die höhere Position Eurasischer Flussbarsche im Nahrungsnetz impliziert eine stärkere Abwärtssteuerung von besiedelten Ökosystemen, was beispielsweise die Wasserqualität beeinflusst. Wobei die flexible Ernährungsweise von Amerikanischen Flussbarschen und die piscivore Ernährung großer Eurasischer Flussbarsche auch ihre Konkurrenzfähigkeit als Neozooen befördern könnten. Die fundamentalen Nahrungsnischen beider Arten mögen zum Großteil überlappen, jedoch scheint die realisierte Nische der untersuchten Amerikanischen Flussbarsche an unbekannten Standortfaktoren wie Konkurrenz, Größenspektrum und Art der Beuteorganismen angepasst zu sein. 54

9 References

Akin, Ş., C. Şahin, B. Verep, D. Turan, A. M. Gözler, A. Bozkurt, K. Çelik, E. Çetin, A. Aracı, and Đ. Sargın. 2011. Feeding habits of introduced European perch (Perca fluviatilis) in an impounded large river system in Turkey. African Journal of Agricultural Research 6:4293-4307.

Amundsen, P. A., T. Bohn, O. A. Popova, F. J. Staldvik, Y. S. Reshetnikov, N. A. Kashulin, and A. A. Lukin. 2003. Ontogenetic niche shifts and resource partitioning in a subarctic piscivore fish guild. Hydrobiologia 497:109-119.

Anderson, C. and G. Cabana. 2007. Estimating the trophic position of aquatic consumers in river food webs using stable nitrogen isotopes. Journal of the North American Benthological Society 26:273-285.

Ansari, R. and S. Qadri. 1989. Individual variation in the foraging strategies of young yellow perch (Perca flavescens) from the Ottawa River. Hydrobiologia 174:207-212.

Arthington, A. H. 1991. Ecological and genetic impacts of introduced and translocated freshwater fishes in Australia. Canadian Journal of Fisheries and Aquatic Sciences 48:33-43.

Attayde, J. L. and L. A. Hansson. 2001. Fish-mediated nutrient recycling and the trophic cascade in lakes. Canadian Journal of Fisheries and Aquatic Sciences 58:1924-1931.

Banas, D., Y. Vollaire, M. Danger, M. Thomas, C. A. Oliveira-Ribeiro, H. Roche, and Y. Ledore. 2009. Can we use stable isotopes for ecotoxicological studies? Effect of DDT on isotopic fractionation in Perca fluviatilis. Chemosphere 76:734-739.

Barks, P. M., J. L. Doucette, and C. M. Somers. 2010. Lack of angling-sized yellow perch in a Canadian boreal lake: potential influences of growth rate, diet, and predation by double- crested cormorants. Transactions of the American Fisheries Society 139:1029-1040.

Bartels, P., P. E. Hirsch, R. Svanback, and P. Eklov. 2012. Water transparency drives intrapopulation divergence in Eurasian perch (Perca fluviatilis). Plos One 7:e43641.

Bearhop, S., C. E. Adams, S. Waldron, R. A. Fuller, and H. MacLeod. 2004. Determining trophic niche width: a novel approach using stable isotope analysis. Journal of Animal Ecology 73:1007-1012.

Beaudoin, C. P., W. M. Tonn, E. E. Prepas, and L. I. Wassenaar. 1999. Individual specialization and trophic adaptability of northern pike (Esox lucius): an isotope and dietary analysis. Oecologia 120:386-396.

Beeck, P., S. Tauber, S. Kiel, and J. Borcherding. 2002. 0+ perch predation on 0+ bream: a case study in a eutrophic gravel pit lake. Freshwater Biology 47:2359-2369.

Beier, U. 2001. Habitat distribution and size structure in freshwater fish communities: effects of vendace on interactions between perch and roach. Journal of Fish Biology 59:1437-1454.

Benke, A. 2011. Secondary production, quantitative food webs, and trophic position. Nature Education Knowledge 3:26. 55

Berezina, N. A. and A. P. Strel'nikova. 2001. Relationships between the food spectrum of perch fry (Perca fluviatilis L.) and the structure of zoobenthos in experimental mesocosms. Biology Bulletin 28:311-318.

Bergman, E. 1990. Effects of roach Rutilus rutilus on two percids, Perca fluviatilis and Gymnocephalus cernua: importance of species interactions for diet shifts. Oikos 57:241- 249.

Bertrand, M., G. Cabana, D. J. Marcogliese, and P. Magnan. 2011. Estimating the feeding range of a mobile consumer in a river-flood plain system using δ13C-gradients and parasites. Journal of Animal Ecology 80:1313-1323.

Berven, K. A. and D. E. Gill. 1983. Interpreting geographic variation in life-history traits. American Zoologist 23:85-97.

Bolding, B., S. A. Bonar, and M. Divens. 1998. Walleye diet in a shallow impoundment: relative importance of pumpkinseed sunfish and yellow perch. Journal of Freshwater Ecology 13:9-14.

Bolnick, D. I., R. Svanback, M. S. Araujo, and L. Persson. 2007. Comparative support for the niche variation hypothesis that more generalized populations also are more heterogeneous. Proceedings of the National Academy of Sciences of the United States of America 104:10075-10079.

Bolnick, D. I., R. Svanbäck, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L. Forister. 2003. The ecology of individuals: incidence and implications of individual specialization. The American Naturalist 161:1-28.

Brandner, J., K. Auerswald, A. F. Cerwenka, U. K. Schliewen, and J. Geist. 2013. Comparative feeding ecology of invasive Ponto-Caspian gobies. Hydrobiologia 703:113-131.

Brown, T. G., B. Runciman, M. J. Bradford, and S. Pollard. 2009. A biological synopsis of yellow perch (Perca flavescens). Canadian Manuscript Reports of Fisheries and Aquatic Sciences 2883.

Browne, D. R. and J. B. Rasmussen. 2009. Shifts in the trophic ecology of brook trout resulting from interactions with yellow perch: an intraguild predator-prey interaction. Transactions of the American Fisheries Society 138:1109-1122.

Browne, D. R. and J. B. Rasmussen. 2012. Rapid response of brook trout to removal of its intraguild prey, yellow perch. Environmental Biology of Fishes 96:1-12.

Budy, P., G. P. Thiede, J. Lobón-Cerviá, G. G. Fernandez, P. McHugh, A. McIntosh, L. A. Vøllestad, E. Becares, and P. Jellyman. 2013. Limitation and facilitation of one of the world's most invasive fish: an intercontinental comparison. Ecology 94:356-367.

Bystrom, P., M. Huss, and L. Persson. 2012. Ontogenetic constraints and diet shifts in Perch (Perca fluviatilis): mechanisms and consequences for intra-cohort cannibalism. Freshwater Biology 57:847-857.

Bystrom, P., J. Karlsson, P. Nilsson, T. Van Kooten, J. Ask, and F. Olofsson. 2007. Substitution of top predators: effects of pike invasion in a subarctic lake. Freshwater Biology 52:1271- 1280. 56

Cabana, G. and J. B. Rasmussen. 1996. Comparison of aquatic food chains using nitrogen isotopes. Proceedings of the National Academy of Sciences of the United States of America 93:10844-10847.

Carpenter, S., J. Kitchell, J. Hodgson, P. Cochran, J. Elser, M. Elser, D. Lodge, D. Kretchmer, X. He, and C. Von Ende. 1987. Regulation of lake primary productivity by food web structure. Ecology 68:1863-1876.

Carpenter, S. R., J. F. Kitchell, and J. R. Hodgson. 1985. Cascading trophic interactions and lake productivity. Bioscience 35:634-639.

Caut, S., E. Angulo, and F. Courchamp. 2009. Variation in discrimination factors (δ15N and δ13C): the effect of diet isotopic values and applications for diet reconstruction. Journal of Applied Ecology 46:443-453.

Champion, P., J. Clayton, and D. Rowe. 2002. Lake manager’s handbook: alien invaders. New Zealand Ministry for the Environment, Wellington New Zealand.

Chase, J. M. and M. A. Leibold. 2003. Ecological niches: linking classical and contemporary approaches. University of Chicago Press, Chicago USA.

Checkley, D. M. and C. A. Miller. 1989. Nitrogen isotope fractionation by oceanic zooplankton. Deep-Sea Research Part A - Oceanographic Research Papers 36:1449-1456.

Claessen, D., A. M. de Roos, and L. Persson. 2000. Dwarfs and giants: cannibalism and competition in size-structured populations. American Naturalist 155:219-237.

Claessen, D. and U. Dieckmann. 2002. Ontogenetic niche shifts and evolutionary branching in size-structured populations. Evolutionary Ecology Research 4:189-217.

Closs, G. P., B. Ludgate, and R. J. Goldsmith. 2001. Controlling European perch (Perca fluviatilis): lessons from an experimental removal. Pages 37-48 in: G. P. Closs, editor. Managing invasive freshwater fish in New Zealand: proceedings of a workshop hosted by Department of Conservation.

Colborne, S. F. and B. W. Robinson. 2013. Effect of nutritional condition on variation in δ13C and δ15N stable isotope values in Pumpkinseed sunfish (Lepomis gibbosus) fed different diets. Environmental Biology of Fishes 96:543-554.

Collette, B. B., M. A. Ali, K. E. F. Hokanson, M. Nagięć, S. A. Smirnov, J. E. Thorpe, A. H. Weatherley, and J. Willemsen. 1977. Biology of the Percids. Journal of the Fisheries Research Board of Canada 34:1878-1889.

Collette, B. B. and P. Banarescu. 1977. Systematics and zoogeography of fishes of family Percidae. Journal of the Fisheries Research Board of Canada 34:1450-1463.

Colwell, R. K. and T. F. Rangel. 2009. Hutchinson's duality: the once and future niche. Proceedings of the National Academy of Sciences 106:19651-19658.

Cooke, S. and D. P. Philipp. 2009. Centrarchid fishes: diversity, biology and conservation. Wiley- Blackwell, Chichester, UK.

Copp, G. H. 2008. Putting multi-dimensionality back into niche: diel vs. day-only niche breadth separation in stream fishes. Fundamental and Applied Limnology 170:273-280. 57

Costa, H. 1979. The food and feeding chronology of yellow perch (Perca flavescens) in Lake Washington. Internationale Revue der gesamten Hydrobiologie und Hydrographie 64:783-793.

Craig, J. F. 1987. The biology of perch and related fish. Croom Helm, London.

Craig, J. F. 2008. Percid fishes: systematics, ecology and exploitation. Wiley-Blackwell, Chichester.

Darimont, C. T., P. C. Paquet, and T. E. Reimchen. 2009. Landscape heterogeneity and marine subsidy generate extensive intrapopulation niche diversity in a large terrestrial vertebrate. Journal of Animal Ecology 78:126-133.

Davies, R. B. 1987. Hypothesis testing when a nuisance parameter is present only under the alternative. Biometrika 74:33-43. de la Moriniere, E. C., B. J. A. Pollux, I. Nagelkerken, M. A. Hemminga, A. H. L. Huiskes, and G. van der Velde. 2003. Ontogenetic dietary changes of coral reef fishes in the mangrove- seagrass-reef continuum: stable isotopes and gut-content analysis. Marine Ecology Progress Series 246:279-289. de Roos, A. M., K. Leonardsson, L. Persson, and G. G. Mittelbach. 2002. Ontogenetic niche shifts and flexible behavior in size-structured populations. Ecological Monographs 72:271-292. de Roos, A. M. and L. Persson. 2001. Physiologically structured models–from versatile technique to ecological theory. Oikos 94:51-71.

Deelder, C. 1951. A contribution to the knowledge of the stunted growth of perch (Perca fluviatilis L.) in Holland. Hydrobiologia 3:357-378. del Rio, C. M., P. Sabat, R. Anderson-Sprecher, and S. P. Gonzalez. 2009. Dietary and isotopic specialization: the isotopic niche of three Cinclodes ovenbirds. Oecologia 161:149-159.

Deniro, M. J. and S. Epstein. 1981. Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica Et Cosmochimica Acta 45:341-351.

Devictor, V., J. Clavel, R. Julliard, S. Lavergne, D. Mouillot, W. Thuiller, P. Venail, S. Villeger, and N. Mouquet. 2010. Defining and measuring ecological specialization. Journal of Applied Ecology 47:15-25.

DFO. 2011. Science advice from a risk assessment of yellow perch (Perca flavescens) in British Columbia. Canadian Science Advisory Secretariat, Science Advisory Report 2010/081.

Diehl, S. 1992. Fish predation and benthic community structure: the role of omnivory and habitat complexity. Ecology 73:1646-1661.

Dörner, H., S. Berg, L. Jacobsen, S. Hülsmann, M. Brojerg, and A. Wagner. 2003. The feeding behaviour of large perch Perca fluviatilis (L.) in relation to food availability: a comparative study. Hydrobiologia 506:427-434.

Dörner, H., H. Schultz, T. Mehner, and J. Benndorf. 2001. Interaction between prey availability and feeding behaviour of age-1 and age-2 perch (Perca fluviatilis L.) in a biomanipulated lake (Bautzen Reservoir, Germany). Limnologica-Ecology and Management of Inland Waters 31:11-16. 58

Dörner, H. and A. Wagner. 2003. Size-dependent predator–prey relationships between perch and their fish prey. Journal of Fish Biology 62:1021-1032.

Eckmann, R. 1995. Fish species richness in lakes of the northeastern lowlands in Germany. Ecology of Freshwater Fish 4:62-69.

Eklöv, P. 1997. Effects of habitat complexity and prey abundance on the spatial and temporal distributions of perch (Perca fluviatilis) and pike (Esox lucius). Canadian Journal of Fisheries and Aquatic Sciences 54:1520-1531.

Elvira, B. and A. Almodóvar. 2001. Freshwater fish introductions in Spain: facts and figures at the beginning of the 21st century. Journal of Fish Biology 59:323-331.

Farquhar, G. D., J. R. Ehleringer, and K. T. Hubick. 1989. Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40:503-537.

Fish, P. A. and J. Savitz. 1983. Variations in home ranges of largemouth bass, yellow perch, bluegills, and pumpkinseeds in an Illinois lake. Transactions of the American Fisheries Society 112:147-153.

Fobert, E., M. Fox, M. Ridgway, and G. Copp. 2011. Heated competition: how climate change will affect non-native pumpkinseed Lepomis gibbosus and native perch Perca fluviatilis interactions in the UK. Journal of Fish Biology 79:1592-1607.

France, R. L. 1995. Differentiation between littoral and pelagic food webs in lakes using stable carbon isotopes. Limnology and Oceanography 40:1310-1313.

Frank, K. T., B. Petrie, J. S. Choi, and W. C. Leggett. 2005. Trophic cascades in a formerly cod- dominated ecosystem. Science 308:1621-1623.

Fraser, J. 1978. The effect of competition with yellow perch on the survival and growth of planted brook trout, splake, and rainbow trout in a small Ontario lake. Transactions of the American Fisheries Society 107:505-517.

Froese, R. and D. Pauly. 2012. Fishbase. URL: http://www.fishbase.de/Summary/ FamilySummary.php?ID=306. World Wide Web electronical publication.

Fry, B. 1988. Food web structure on Georges Bank from stable C, N, and S isotopic compositions. Limnology and Oceanography 33:1182-1190.

Fry, B. 2006. Stable isotope ecology. 3rd edition. Springer Science + Business Media, LLC, New York, NY, USA.

Fry, B. and E. B. Sherr. 1984. δ13C measurements as indicators of carbon flow in marine and freshwater ecosystems. Contributions in Marine Science 27:13-47.

Fulford, R. S., J. A. Rice, T. J. Miller, F. P. Binkowski, J. M. Dettmers, and B. Belonger. 2006. Foraging selectivity by larval yellow perch (Perca flavescens): implications for understanding recruitment in small and large lakes. Canadian Journal of Fisheries and Aquatic Sciences 63:28-42.

Fuller, B. T., J. L. Fuller, N. E. Sage, D. A. Harris, T. C. O'Connell, and R. E. M. Hedges. 2005. Nitrogen balance and δ15N: why you're not what you eat during nutritional stress. Rapid Communications in Mass Spectrometry 19:2497-2506. 59

Fullhart, H. G., B. G. Parsons, D. W. Willis, and J. R. Reed. 2002. Yellow perch piscivory and its possible role in structuring littoral zone fish communities in small Minnesota lakes. Journal of Freshwater Ecology 17:37-43.

Furnass, T. I. 1979. Laboratory experiments on prey selection by perch fry (Perca fluviatilis). Freshwater Biology 9:33-43.

Garrison, L. P. and J. S. Link. 2000. Dietary guild structure of the fish community in the Northeast United States continental shelf ecosystem. Marine Ecology Progress Series 202:231-240.

Graeb, B. D. S., M. T. Mangan, J. C. Jolley, D. H. Wahl, and J. M. Dettmers. 2006. Ontogenetic changes in prey preference and foraging ability of yellow perch: insights based on relative energetic return of prey. Transactions of the American Fisheries Society 135:1493-1498.

Guzzo, M. M., G. D. Haffner, N. D. Legler, S. A. Rush and A. T. Fisk. 2013. Fifty years later: trophic ecology and niche overlap of a native and non-indigenous fish species in the western basin of Lake Erie. Biological Invasions 15:1695-1711.

Haakana, H., H. Huuskonen, and J. Karjalainen. 2007. Predation of perch on vendace larvae: diet composition in an oligotrophic lake and digestion time of the larvae. Journal of Fish Biology 70:1171-1184.

Haertel-Borer, S. S., D. Zak, R. Eckmann, U. Baade, and F. Hölker. 2005. Population density of the crayfish, Orconectes limosus, in relation to fish and macroinvertebrate densities in a small mesotrophic lake–implications for the lake's food web. International Review of Hydrobiology 90:523-533.

Haertel, S. S. 2001. Quantification of fish predation on zooplankton in a small mesotrophic lake (Großer Vätersee). Dissertation, University of Constance, Germany.

Haertel, S. S., U. Baade, and R. Eckmann. 2002. No general percid dominance at mesotrophic lake conditions: insights from the quantification of predator-prey interactions. Limnologica-Ecology and Management of Inland Waters 32:1-13.

Hairston, N. G. 1993. Cause-effect relationships in energy flow, trophic structure, and interspecific interactions. American Naturalist 142:379-411.

Hammerschlag-Peyer, C. M., L. A. Yeager, M. S. Araújo, and C. A. Layman. 2011. A hypothesis- testing framework for studies investigating ontogenetic niche shifts using stable isotope ratios. Plos One 6:e27104.

Hanson, J. M. and W. C. Leggett. 1985. Experimental and field evidence for inter-and intraspecific competition in two freshwater fishes. Canadian Journal of Fisheries and Aquatic Sciences 42:280-286.

Hanson, J. M. and W. C. Leggett. 1986. Effect of competition between two freshwater fishes on prey consumption and abundance. Canadian Journal of Fisheries and Aquatic Sciences 43:1363-1372.

Hayes, D. B., W. W. Taylor, and J. C. Schneider. 1992. Response of yellow perch and the benthic invertebrate community to a reduction in the abundance of white suckers. Transactions of the American Fisheries Society 121:36-53.

Heath, D. D. and D. A. Roff. 1996. The role of trophic bottlenecks in stunting: a field test of an allocation model of growth and reproduction in yellow perch, Perca flavescens. Environmental Biology of Fishes 45:53-63. 60

Heath, M. R. 2005. Changes in the structure and function of the North Sea fish foodweb, 1973- 2000, and the impacts of fishing and climate. ICES Journal of Marine Science 62:847- 868.

Heibo, E., C. Magnhagen, and L. A. Vøllestad. 2005. Latitudinal variation in life-history traits in Eurasian perch. Ecology 86:3377-3386.

Helfman, G., B. B. Collette, D. E. Facey, and B. W. Bowen. 2009. The diversity of fishes: biology, evolution, and ecology. Wiley-Blackwell, Chichester, UK.

Hjelm, J., L. Persson, and B. Christensen. 2000. Growth, morphological variation and ontogenetic niche shifts in perch (Perca fluviatilis) in relation to resource availability. Oecologia 122:190-199.

Hobson, K. A. and R. G. Clark. 1992. Assessing avian diets using stable isotopes II: factors influencing diet-tissue fractionation. Condor 94:189-197.

Hobson, K. A. and H. E. Welch. 1992. Determination of trophic relationships within a high arctic marine food web using δ13C and δ15N analysis. Marine Ecology Progress Series 84:9-18.

Hokanson, K. E. F. 1977. Temperature requirements of some percids and adaptations to the seasonal temperature cycle. Journal of the Fisheries Research Board of Canada 34:1524- 1550.

Holmlund, C. M. and M. Hammer. 1999. Ecosystem services generated by fish populations. Ecological Economics 29:253-268.

Hopkins, J. B. and J. M. Ferguson. 2012. Estimating the diets of animals using stable isotopes and a comprehensive Bayesian mixing model. Plos One 7:e28478.

Horppila, J., J. Ruuhijärvi, M. Rask, C. Karppinen, K. Nyberg, and M. Olin. 2000. Seasonal changes in the diets and relative abundances of perch and roach in the littoral and pelagic zones of a large lake. Journal of Fish Biology 56:51-72.

Huryn, A. D. 1998. Ecosystem-level evidence for top-down and bottom-up control of production in a grassland stream system. Oecologia 115:173-183.

Jackson, A. L., R. Inger, A. C. Parnell, and S. Bearhop. 2011. Comparing isotopic niche widths among and within communities: SIBER–Stable Isotope Bayesian Ellipses in R. Journal of Animal Ecology 80:595-602.

Jansen, W. A. 1996. Plasticity in maturity and fecundity of yellow perch, Perca flavescens (Mitchill): comparisons of stunted and normal-growing populations. Annales Zoologici Fennici. 33:403-415.

Jansen, W. A. and W. C. Mackay. 1992. Foraging in yellow perch, Perca flavescens: biological and physical factors affecting diel periodicity in feeding, consumption, and movement. Environmental Biology of Fishes 34:287-303.

Jennings, S., J. K. Pinnegar, N. V. Polunin, and K. J. Warr. 2002. Linking size-based and trophic analyses of benthic community structure. Marine Ecology Progress Series 226:77-85.

Johnson, E. E. 2009. A quantitative risk assessment model for the management of invasive yellow perch in Shuswap Lake, British Columbia. Thesis (M. Sc.), School of Resource and Environmental Management - Simon Fraser University, Canada. 61

Johnson, M. W. 2001. Indicators (parasites and stable isotopes) of trophic status of yellow perch (Perca flavescens Mitchell) in nutrient poor Canadian Shield lakes. Thesis (M. Sc.), University of Manitoba, Canada.

Kaehler, S. and E. Pakhomov. 2001. Effects of storage and preservation on the 13C and 15N signatures of selected marine organisms. Marine Ecology Progress Series 219:299-304.

Kaemingk, M. A. and D. W. Willis. 2012. Mensurative approach to examine potential interactions between age-0 yellow perch (Perca flavescens) and bluegill (Lepomis macrochirus). Aquatic Ecology 46:353-362.

Karpouzi, V. S. and K. I. Stergiou. 2003. The relationships between mouth size and shape and body length for 18 species of marine fishes and their trophic implications. Journal of Fish Biology 62:1353-1365.

Kasprzak, P., F. Gervais, R. Adrian, W. Weiler, R. Radke, I. Jäger, S. Riest, U. Siedel, B. Schneider, M. Böhme, R. Eckmann, and N. Walz. 2000. Trophic characterization, pelagic food web structure and comparison of two mesotrophic lakes in Brandenburg (Germany). International Review of Hydrobiology 85:167-189.

Keast, A. 1977. Diet overlaps and feeding relationships between the year classes in the yellow perch (Perca flavescens). Environmental Biology of Fishes 2:53-70.

Keast, A. 1978. Trophic and spatial interrelationships in the fish species of an Ontario temperate lake. Environmental Biology of Fishes 3:7-31.

Keast, A. 1985. The piscivore feeding guild of fishes in small freshwater ecosystems. Environmental Biology of Fishes 12:119-129.

Kelly, J. F. 2000. Stable isotopes of carbon and nitrogen in the study of avian and mammalian trophic ecology. Canadian Journal of Zoology 78:1-27.

Kempster, B., L. Zanette, F. J. Longstaffe, S. A. MacDougall-Shackleton, J. C. Wingfield, and M. Clinchy. 2007. Do stable isotopes reflect nutritional stress? Results from a laboratory experiment on song sparrows. Oecologia 151:365-371.

Kiljunen, M., J. Grey, T. Sinisalo, C. Harrod, H. Immonen, and R. I. Jones. 2006. A revised model for lipid-normalizing δ13C values from aquatic organisms, with implications for isotope mixing models. Journal of Applied Ecology 43:1213-1222.

Kitchell, J. F., M. G. Johnson, C. K. Minns, K. H. Loftus, L. Greig, and C. H. Olver. 1977. Percid habitat: the river analogy. Journal of the Fisheries Board of Canada 34:1922-1935.

Knight, R. L., F. J. Margraf, and R. F. Carline. 1984. Piscivory by Walleyes and Yellow Perch in Western Lake Erie. Transactions of the American Fisheries Society 113:677-693.

Kottelat, M. and J. Freyhof. 2007. Handbook of European freshwater fishes. Vol. 13. Cornol: Publications Kottelat. Berlin, Germany.

Kraemer, L. D., D. Evans, and P. J. Dillon. 2012. The impacts of ontogenetic dietary shifts in yellow perch (Perca flavescens) on Zn and Hg accumulation. Ecotoxicology and environmental safety 78:246-252.

Lamouroux, N., N. L. Poff, and P. L. Angermeier. 2002. Intercontinental convergence of stream fish community traits along geomorphic and hydraulic gradients. Ecology 83:1792-1807. 62

Layman, C. A. and J. E. Allgeier. 2012. Characterizing trophic ecology of generalist consumers: a case study of the invasive lionfish in The Bahamas. Marine Ecology Progress Series 448:131-141.

Layman, C. A., D. A. Arrington, C. G. Montaña, and D. M. Post. 2007. Can stable isotope ratios provide for community-wide measures of trophic structure? Ecology 88:42-48.

Leach, J., M. Johnson, J. Kelso, J. Hartmann, W. Nümann, and B. Entz. 1977. Responses of percid fishes and their habitats to eutrophication. Journal of the Fisheries Board of Canada 34:1959-1963.

Lecren, E. D. 1992. Exceptionally big individual perch (Perca fluviatilis L.) and their growth. Journal of Fish Biology 40:599-625.

Legler, N. D. 2008. Effects of Lake Erie habitat variation and digestion rate on feeding in freshwater fishes. Thesis (M. Sc.), University of Windsor, Canada.

LELF. 2012. Verzeichnis der Gewässer mit einem nachgewiesenen Maränenbestand im Land Brandenburg. Landesamt für Ländliche Entwicklung, Landwirtschaft und Flur- neuordnung. Frankfurt (Oder), Germany.

Leroux, S. J. and M. Loreau. 2010. Consumer-mediated recycling and cascading trophic interactions. Ecology 91:2162-2171.

Leunda, P. M., J. Oscoz, B. Elvira, A. Agorreta, S. Perea, and R. Miranda. 2008. Feeding habits of the exotic black bullhead Ameiurus melas (Rafinesque) in the Iberian Peninsula: first evidence of direct predation on native fish species. Journal of Fish Biology 73:96-114.

Link, J. S. and L. P. Garrison. 2002. Trophic ecology of Atlantic cod Gadus morhua on the northeast US continental shelf. Marine Ecology Progress Series 227:109-123.

Lintermans, M. 2004. Human-assisted dispersal of alien freshwater fish in Australia. New Zealand Journal of Marine and Freshwater Research 38:481-501.

Logan, J. M., T. D. Jardine, T. J. Miller, S. E. Bunn, R. A. Cunjak, and M. E. Lutcavage. 2008. Lipid corrections in carbon and nitrogen stable isotope analyses: comparison of chemical extraction and modelling methods. Journal of Animal Ecology 77:838-846.

Lott, J. P., D. W. Willis, and D. O. Lucchesi. 1996. Relationship of food habits to yellow perch growth and population structure in South Dakota lakes. Journal of Freshwater Ecology 11:27-37.

Luek, A., G. E. Morgan, B. Wissel, J. M. Gunn, and C. W. Ramcharan. 2010. Rapid and unexpected effects of piscivore introduction on trophic position and diet of perch (Perca flavescens) in lakes recovering from acidification and metal contamination. Freshwater Biology 55:1616-1627.

MacDougall, T. M., M. H. Benoit, R. Dermott, O. E. Johannsson, T. B. Johnson, E. S. Millard and M. M. Munawar. 2001. Lake Erie 1998: assessment of abundance, biomass and production of the lower trophic levels, diets of juvenile yellow perch and trends in the fishery. Canadian Technical Report of Fisheries and Aquatic Sciences 2376.

Mackintosh, S. A., A. Perez-Fuentetaja, L. R. Zimmerman, G. Pacepavicius, M. Clapsadl, M. Alaee, and D. S. Aga. 2012. Analytical performance of a triple quadrupole mass spectrometer compared to a high resolution mass spectrometer for the analysis of polybrominated diphenyl ethers in fish. Analytica Chimica Acta 747:67-75. 63

Mahon, R. 1984. Divergent structure in fish taxocenes of north temperate streams. Canadian Journal of Fisheries and Aquatic Sciences 41:330-350.

Mairesse, G., M. Thomas, J. N. Gardeur, and J. Brun-Bellut. 2006. Effects of geographic source, rearing system, and season on the nutritional quality of wild and farmed Perca fluviatilis. Lipids 41:221-229.

Matthews, B. and A. Mazumder. 2003. Compositional and interlake variability of zooplankton affect baseline stable isotope signatures. Limnology and Oceanography 48:1977-1987.

McCutchan, J. H., W. M. Lewis, C. Kendall, and C. C. McGrath. 2003. Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102:378-390.

McIntyre, J. K., D. A. Beauchamp, M. M. Mazur, and N. C. Overman. 2006. Ontogenetic trophic interactions and benthopelagic coupling in Lake Washington: evidence from stable isotopes and diet analysis. Transactions of the American Fisheries Society 135:1312- 1328.

Mehner, T., J. Benndorf, P. Kasprzak, and R. Koschel. 2002. Biomanipulation of lake ecosystems: successful applications and expanding complexity in the underlying science. Freshwater Biology 47:2453-2465.

Mehner, T., M. Diekmann, U. Brämick, and R. Lemcke. 2005. Composition of fish communities in German lakes as related to lake morphology, trophic state, shore structure and human- use intensity. Freshwater Biology 50:70-85.

Mehner, T. and R. Thiel. 1999. A review of predation impact by 0+ fish on zooplankton in fresh and brackish waters of the temperate northern hemisphere. Environmental Biology of Fishes 56:169-181.

Minagawa, M. and E. Wada. 1984. Stepwise enrichment of δ15N along food chains: further evidence and the relation between δ15N and animal age. Geochimica Et Cosmochimica Acta 48:1135-1140.

Mittelbach, G. G. and L. Persson. 1998. The ontogeny of piscivory and its ecological consequences. Canadian Journal of Fisheries and Aquatic Sciences 55:1454-1465.

Momot, W. T. 1995. Redefining the role of crayfish in aquatic ecosystems. Reviews in Fisheries Science 3:33-63.

Moore, J. W. and B. X. Semmens. 2008. Incorporating uncertainty and prior information into stable isotope mixing models. Ecology Letters 11:470-480.

Morgan, D. L., S. J. Hambleton, H. S. Gill, and S. J. Beatty. 2003. Distribution, biology and likely impacts of the introduced redfin perch (Perca fluviatilis)(Percidae) in Western Australia. Marine and Freshwater Research 53:1211-1221.

Motta, P. J., S. F. Norton, and J. J. Luczkovich. 1995. Perspectives on the ecomorphology of bony fishes. Environmental Biology of Fishes 44:11-20.

Muggeo, V. M. R. 2008. segmented: an R Package to Fit Regression Models with Broken-Line Relationships. R News 8/1, 20-25, URL: http://cran.r-project.org/doc/Rnews/.

Newsome, S. D., C. M. del Rio, S. Bearhop, and D. L. Phillips. 2007. A niche for isotopic ecology. Frontiers in Ecology and the Environment 5:429-436. 64

Nordstokke, D. W. and B. D. Zumbo. 2010. A new nonparametric Levene test for equal variances. Psicológica 31:401-430.

Nursall, J. 1973. Some behavioral interactions of spottail shiners (Notropis hudsonius), yellow perch (Perca flavescens), and northern pike (Esox lucius). Journal of the Fisheries Board of Canada 30:1161-1178.

Nystrom, P., C. Bronmark, and W. Graneli. 1996. Patterns in benthic food webs: A role for omnivorous crayfish? Freshwater Biology 36:631-646.

Ohlberger, J., Ø. Langangen, E. Edeline, D. Claessen, I. J. Winfield, N. C. Stenseth, and L. A. Vøllestad. 2011. Stage-specific biomass overcompensation by juveniles in response to increased adult mortality in a wild fish population. Ecology 92:2175-2182.

Olden, J. D., N. L. Poff, and K. R. Bestgen. 2008. Trait synergisms and the rarity, extirpation, and extinction risk of desert fishes. Ecology 89:847-856.

Olson, M. H. 1996. Ontogenetic niche shifts in largemouth bass: variability and consequences for first-year growth. Ecology 77:179-190.

Olson, M. H., D. M. Green, and L. G. Rudstam. 2001. Changes in yellow perch (Perca flavescens) growth associated with the establishment of a walleye (Stizostedion vitreum) population in Canadarago Lake, New York (USA). Ecology of Freshwater Fish 10:11-20.

Olson, M. H., G. G. Mittelbach, and C. W. Osenberg. 1995. Competition between predator and prey: resource-based mechanisms and implications for stage-structured dynamics. Ecology 76:1758-1771.

Olsson, K., P. Stenroth, P. Nystrom, and W. Graneli. 2009. Invasions and niche width: does niche width of an introduced crayfish differ from a native crayfish? Freshwater Biology 54:1731-1740.

Orrù, F., A. Deiana, and A. Cau. 2010. Introduction and distribution of alien freshwater fishes on the island of Sardinia (Italy): an assessment on the basis of existing data sources. Journal of Applied Ichthyology 26:46-52.

Osenberg, C. W., G. G. Mittelbach, and P. C. Wainwright. 1992. Two-stage life histories in fish: the interaction between juvenile competition and adult performance. Ecology 73:255- 267.

Pace, M. L., J. J. Cole, S. R. Carpenter, and J. F. Kitchell. 1999. Trophic cascades revealed in diverse ecosystems. Trends in Ecology & Evolution 14:483-488.

Paine, R. T. 1988. Road maps of interactions or grist for theoretical development? Ecology 69:1648-1654.

Parker, A. D., C. A. Stepien, O. J. Sepulveda-Villet, C. B. Ruehl, and D. G. Uzarski. 2009a. The interplay of morphology, habitat, resource use, and genetic relationships in young yellow perch. Transactions of the American Fisheries Society 138:899-914.

Parker, A. D., D. G. Uzarski, C. R. Ruetz, and T. M. Burton. 2009b. Diets of yellow perch (Perca flavescens) in wetland habitats of Saginaw Bay, Lake Huron. Journal of Freshwater Ecology 24:347-355.

Parnell, A. C., R. Inger, S. Bearhop, and A. L. Jackson. 2010. Source partitioning using stable isotopes: coping with too much variation. Plos One 5:e9672. 65

Pasquaud, S., P. Elie, C. Jeantet, I. Billy, P. Martinez, and M. Girardin. 2008. A preliminary investigation of the fish food web in the Gironde estuary, France, using dietary and stable isotope analyses. Estuarine Coastal and Shelf Science 78:267-279.

Paszkowski, C. A. 1985. The foraging behavior of the central mudminnow and yellow perch: the influence of foraging site, intraspecific and interspecific competition. Oecologia 66:271- 279.

Persson, A. and L.-A. Hansson. 1999. Diet shift in fish following competitive release. Canadian Journal of Fisheries and Aquatic Sciences 56:70-78.

Persson, L. 1983a. Effects of intra-and interspecific competition on dynamics and size structure of a perch Perca fluviatilis and a roach Rutilus rutilus population. Oikos 41:126-132.

Persson, L. 1983b. Food consumption and competition between age classes in a perch Perca fluviatilis population in a shallow eutrophic lake. Oikos 40:197-207.

Persson, L. 1986. Effects of reduced interspecific competition on resource utilization in perch (Perca fluviatilis). Ecology 67:355-364.

Persson, L. 1987a. Effects of habitat and season on competitive interactions between roach (Rutilus rutilus) and perch (Perca fluviatilis). Oecologia 73:170-177.

Persson, L. 1987b. The effects of resource availability and distribution on size class interactions in perch, Perca fluviatilis. Oikos 48:148-160.

Persson, L. 1988. Asymmetries in competitive and predatory interactions in fish populations. Pages 203-218 in: B. Ebenman, editor. Size-structured populations. Springer Berlin - Heidelberg.

Persson, L., P. Bystrom, and E. Wahlstrom. 2000. Cannibalism and competition in Eurasian perch: population dynamics of an ontogenetic omnivore. Ecology 81:1058-1071.

Persson, L., D. Claessen, A. M. de Roos, P. Byström, S. Sjögren, R. Svanbäck, E. Wahlström, and E. Westman. 2004. Cannibalism in a size-structured population: energy extraction and control. Ecological Monographs 74:135-157.

Persson, L. and A. de Roos. 2006. Food-dependent individual growth and population dynamics in fishes. Journal of Fish Biology 69:1-20.

Persson, L. and A. M. de Roos. 2012. Mixed competition–predation: potential vs. realized interactions. Journal of Animal Ecology 81:483-493.

Persson, L., A. M. de Roos, D. Claessen, P. Bystrom, J. Lovgren, S. Sjogren, R. Svanback, E. Wahlstrom, and E. Westman. 2003. Gigantic cannibals driving a whole-lake trophic cascade. Proceedings of the National Academy of Sciences of the United States of America 100:4035-4039.

Persson, L. and P. Eklov. 1995. Prey refuges affecting interactions between piscivorous perch and juvenile perch and roach. Ecology 76:70-81.

Persson, L. and L. A. Greenberg. 1990a. Juvenile competitive bottlenecks: the perch (Perca fluviatilis) - roach (Rutilus rutilus) interaction. Ecology:44-56.

Persson, L. and L. A. Greenberg. 1990b. Optimal foraging and habitat shift in perch (Perca Fluviatilis) in a resource gradient. Ecology 71:1699-1713. 66

Peterson, B. J. and B. Fry. 1987. Stable isotopes in ecosystem studies. Annual Review of Ecology and Systematics 18:293-320.

Phillips, D. L. and P. M. Eldridge. 2006. Estimating the timing of diet shifts using stable isotopes. Oecologia 147:195-203.

Phillips, D. L. and J. W. Gregg. 2003. Source partitioning using stable isotopes: coping with too many sources. Oecologia 136:261-269.

Pinnegar, J. K. and N. V. C. Polunin. 1999. Differential fractionation of δ13C and δ15N among fish tissues: implications for the study of trophic interactions. Functional Ecology 13:225- 231.

Polis, G. A., C. A. Myers, and R. D. Holt. 1989. The ecology and evolution of intraguild predation: potential competitors that eat each other. Annual Review of Ecology and Systematics 20:297-330.

Polis, G. A. and D. R. Strong. 1996. Food web complexity and community dynamics. American Naturalist 147:813-846.

Post, D. M. 2002. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83:703-718.

Post, D. M., C. A. Layman, D. A. Arrington, G. Takimoto, J. Quattrochi, and C. G. Montana. 2007. Getting to the fat of the matter: models, methods and assumptions for dealing with lipids in stable isotope analyses. Oecologia 152:179-189.

Post, D. M., M. L. Pace, and N. G. Hairston. 2000. Ecosystem size determines food-chain length in lakes. Nature 405:1047-1049.

Post, J. R. and D. Cucin. 1984. Changes in the benthic community of a small precambrian lake following the introduction of yellow perch, Perca flavescens. Canadian Journal of Fisheries and Aquatic Sciences 41:1496-1501.

Post, J. R., M. R. S. Johannes, and D. J. McQueen. 1997. Evidence of density-dependent cohort splitting in age-0 yellow perch (Perca flavescens): potential behavioural mechanisms and population-level consequences. Canadian Journal of Fisheries and Aquatic Sciences 54:867-875.

Post, J. R. and D. J. McQueen. 1988. Ontogenetic changes in the distribution of larval and juvenile yellow perch (Perca flavescens): a response to prey or predators? Canadian Journal of Fisheries and Aquatic Sciences 45:1820-1826.

Potthoff, A. J., B. R. Herwig, M. A. Hanson, K. D. Zimmer, M. G. Butler, J. R. Reed, B. G. Parsons, and M. C. Ward. 2008. Cascading food-web effects of piscivore introductions in shallow lakes. Journal of Applied Ecology 45:1170-1179.

Power, M. E. 1992. Habitat heterogeneity and The functional signifiance of fish in river food webs. Ecology 73:1675-1688.

Quevedo, M. and J. Olsson. 2006. The effect of small-scale resource origin on trophic position estimates in Perca fluviatilis. Journal of Fish Biology 69:141-150.

Quevedo, M., R. Svanback, and P. Eklov. 2009. Intrapopulation niche partitioning in a generalist predator limits food web connectivity. Ecology 90:2263-2274. 67

Radke, R. J. 1998. Strukturbildende Prozesse in Fischartengemeinschaften mesotropher Seen des nordostdeutschen Tieflandes. UFO Dissertation 343, Allensbach, Germany.

Radke, R. J. and R. Eckmann. 2001. No general percid dominance under mesotrophic lake conditions: a test of several hypotheses. Limnologica-Ecology and Management of Inland Waters 31:37-44.

Rasmussen, J., J. Gunn, G. Sherwood, A. Iles, A. Gagnon, P. Campbell, and A. Hontela. 2008. Direct and indirect (foodweb mediated) effects of metal exposure on the growth of yellow perch (Perca flavescens): implications for ecological risk assessment. Human and Ecological Risk Assessment 14:317-350.

RCoreTeam. 2012. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL: http://www.R-project.org/.

Reiss, J., J. R. Bridle, J. M. Montoya, and G. Woodward. 2009. Emerging horizons in biodiversity and ecosystem functioning research. Trends in Ecology & Evolution 24:505-514.

Reissig, M., C. Trochine, C. Queimalinos, E. Balseiro, and B. Modenutti. 2006. Impact of fish introduction on planktonic food webs in lakes of the Patagonian Plateau. Biological Conservation 132:437-447.

Ribeiro, F., R. L. Orjuela, M. F. Magalhaes, and M. J. Collares-Pereira. 2007. Variability in feeding ecology of a South American cichlid: a reason for successful invasion in mediterranean-type rivers? Ecology of Freshwater Fish 16:559-569.

Romare, P. 2000. Growth of larval and juvenile perch: the importance of diet and fish density. Journal of Fish Biology 56:876-889.

Roswell, C. R., S. A. Pothoven, and T. O. Hook. 2013. Spatio-temporal, ontogenetic and interindividual variation of age-0 diets in a population of yellow perch. Ecology of Freshwater Fish 22:479-493.

Rumsey, C. R., N. E. Jones, and H. H. Banks. 2007. An attempt to rehabilitate a collapsed brook trout population by introducing F-1 splake to control yellow perch. North American Journal of Fisheries Management 27:1139-1147.

Sarakinos, H. C., M. L. Johnson, and M. J. V. Zanden. 2002. A synthesis of tissue-preservation effects on carbon and nitrogen stable isotope signatures. Canadian Journal of Zoology 80:381-387.

Schade, C. B. and S. A. Bonar. 2005. Distribution and abundance of nonnative fishes in streams of the Western United States. North American Journal of Fisheries Management 25:1386- 1394.

Schoenebeck, C. W. and M. L. Brown. 2010. Potential importance of competition, predation, and prey on yellow perch growth from two dissimilar population types. Prairie Naturalist 42:32-37.

Schulze, T., H. Dörner, U. Baade, and F. Hölker. 2012. Dietary niche partitioning in a piscivorous fish guild in response to stocking of an additional competitor - The role of diet specialisation. Limnologica-Ecology and Management of Inland Waters 42:56-64.

Scott, W. B. and E. J. Crossman. 1973. Freshwater Fishes of Canada. Fisheries Research Board of Canada, Ottawa, Canada. 68

Semmens, B. X., E. J. Ward, J. W. Moore, and C. T. Darimont. 2009. Quantifying inter- and intrapopulation niche variability using hierarchical Bayesian stable isotope mixing models. Plos One 4:e6187.

Serns, S. L. and M. H. Hoff. 1984. Food habits of adult yellow perch - Perca flavescens and smallmouth bass - Micropterus dolomieui in Nebish Lake, Wisconsin, USA with special reference to zooplankton density and composition. Wisconsin Department of Natural Resources Technical Bulletin:1-24.

Sharp, Z. 2007. Principles of stable isotope geochemistry. Pearson Prentice Hall, Upper Saddle River, USA.

Shrader, T. 2000. Effects of invasive yellow perch on gamefish and zooplankton populations of Phillips Reservoir. Oregon Fish Division Information Report 2000/03.

Simon, K. S. and C. R. Townsend. 2003. Impacts of freshwater invaders at different levels of ecological organisation, with emphasis on salmonids and ecosystem consequences. Freshwater Biology 48:982-994.

Sloss, B. L., N. Billington, and B. M. Burr. 2004. A molecular phylogeny of the Percidae (Teleostei, Perciformes) based on mitochondrial DNA sequence. Molecular Phylogenetics and Evolution 32:545-562.

Song, C. B., T. J. Near, and L. M. Page. 1998. Phylogenetic relations among percid fishes as inferred from mitochondrial cytochrome b DNA sequence data. Molecular Phylogenetics and Evolution 10:343-353.

Svanback, R. and D. I. Bolnick. 2007. Intraspecific competition drives increased resource use diversity within a natural population. Proceedings of the Royal Society B-Biological Sciences 274:839-844.

Svanback, R. and P. Eklov. 2002. Effects of habitat and food resources on morphology and ontogenetic growth trajectories in perch. Oecologia 131:61-70.

Svanback, R. and P. Eklov. 2003. Morphology dependent foraging efficiency in perch: a trade- off for ecological specialization? Oikos 102:273-284.

Svanback, R. and P. Eklov. 2004. Morphology in perch affects habitat specific feeding efficiency. Functional Ecology 18:503-510.

Svanback, R. and L. Persson. 2004. Individual diet specialization, niche width and population dynamics: implications for trophic polymorphisms. Journal of Animal Ecology 73:973- 982.

Svetovidov, A. N. and E. A. Dorofeeva. 1963. Sistematicheskie otnosheniya proiskhozhdenie u istoriya rasseleniya evopeisko-aziatskikh i severo-amerikanskikh okunei i sudakov (rody Perca Lucioperca i Stizostedion). Voprosy Ikhtiologii 3:625-651.

Sweeting, C. J., N. V. C. Polunin, and S. Jennings. 2006. Effects of chemical lipid extraction and arithmetic lipid correction on stable isotope ratios of fish tissues. Rapid Communications in Mass Spectrometry 20:595-601.

Syvaranta, J., H. Hamalainen, and R. I. Jones. 2006. Within-lake variability in carbon and nitrogen stable isotope signatures. Freshwater Biology 51:1090-1102. 69

Syvaranta, J., A. Lensu, T. J. Marjomaki, S. Oksanen, and R. I. Jones. 2013. An empirical evaluation of the utility of convex hull and standard ellipse areas for assessing population niche widths from stable isotope data. Plos One 8:e56094.

Syvaranta, J. and M. Rautio. 2010. Zooplankton, lipids and stable isotopes: importance of seasonal, latitudinal, and taxonomic differences. Canadian Journal of Fisheries and Aquatic Sciences 67:1721-1729.

Thorpe, J. E. 1977. Morphology, Physiology, behavior and ecology of Perca fluviatilis L. and Perca flavescens Mitchell. Journal of the Fisheries Research Board of Canada 34:1504- 1514.

Tieszen, L. L., T. W. Boutton, K. G. Tesdahl, and N. A. Slade. 1983. Fractionation and turnover of stable carbon isotopes in animal tissues: implications for δ13C analysis of diet. Oecologia 57:32-37.

Tonn, W. M., J. J. Magnuson, M. Rask, and J. Toivonen. 1990. Intercontinental comparison of small-lake fish assemblages: the balance between local and regional processes. American Naturalist 136:345-375.

Treasurer, J. W. 1990. The food and daily food-consumption of lacustrine 0+ perch, Perca fluviatilis L. Freshwater Biology 24:361-374.

Truemper, H. A. and T. E. Lauer. 2005. Gape limitation and piscine prey size-selection by yellow perch in the extreme southern area of Lake Michigan, with emphasis on two exotic prey items. Journal of Fish Biology 66:135-149.

Truemper, H. A., T. E. Lauer, T. S. McComish, and R. A. Edgell. 2006. Response of yellow perch diet to a changing forage base in southern Lake Michigan, 1984-2002. Journal of Great Lakes Research 32:806-816.

Urbatzka, R., P. Beeck, G. van der Velde, and J. Borcherding. 2008. Alternative use of food resources causes intra-cohort variation in the size distribution of young-of-the-year perch (Perca fluviatilis). Ecology of Freshwater Fish 17:475-480.

Vadeboncoeur, Y., M. J. Vander Zanden, and D. M. Lodge. 2002. Putting the lake back together: reintegrating benthic pathways into lake food web models. Bioscience 52:44-54.

Vander Zanden, M. J., G. Cabana, and J. B. Rasmussen. 1997. Comparing trophic position of freshwater fish calculated using stable nitrogen isotope ratios (δ15N) and literature dietary data. Canadian Journal of Fisheries and Aquatic Sciences 54:1142-1158.

Vander Zanden, M. J., J. M. Casselman, and J. B. Rasmussen. 1999a. Stable isotope evidence for the food web consequences of species invasions in lakes. Nature 401:464-467.

Vander Zanden, M. J. and J. B. Rasmussen. 1996. A trophic position model of pelagic food webs: impact on contaminant bioaccumulation in lake trout. Ecological Monographs 66:451- 477.

Vander Zanden, M. J. and J. B. Rasmussen. 1999. Primary consumer δ13C and δ15N and the trophic position of aquatic consumers. Ecology 80:1395-1404.

Vander Zanden, M. J. and J. B. Rasmussen. 2001. Variation in δ13C and δ15N trophic fractionation: implications for aquatic food web studies. Limnology and Oceanography 46:2061-2066. 70

Vander Zanden, M. J., B. J. Shuter, N. Lester, and J. B. Rasmussen. 1999b. Patterns of food chain length in lakes: a stable isotope study. The American Naturalist 154:406-416.

Vander Zanden, M. J. and Y. Vadeboncoeur. 2002. Fishes as integrators of benthic and pelagic food webs in lakes. Ecology 83:2152-2161.

Vanderklift, M. A. and S. Ponsard. 2003. Sources of variation in consumer-diet δ15N enrichment: a meta-analysis. Oecologia 136:169-182.

Vollaire, Y., D. Banas, T. Marielle, and H. Roche. 2007. Stable isotope variability in tissues of the Eurasian perch, Perca fluviatilis. Comparative Biochemistry and Physiology Part A - Molecular & Integrative Physiology 148:504-509.

Wahl, C. M., E. L. Mills, W. N. McFarland, and J. S. DeGisi. 1993. Ontogenetic changes in prey selection and visual acuity of the yellow perch, Perca flavescens. Canadian Journal of Fisheries and Aquatic Sciences 50:743-749.

Wahlström, E., L. Persson, S. Diehl, and P. Byström. 2000. Size-dependent foraging efficiency, cannibalism and zooplankton community structure. Oecologia 123:138-148.

Wainright, S. C., M. J. Fogarty, R. C. Greenfield, and B. Fry. 1993. Long-term changes in the Georges Bank food web: trends in stable isotopic compositions of fish scales. Marine Biology 115:481-493.

Wall, A. J. and P. J. Blanchfield. 2012. Habitat use of lake trout (Salvelinus namaycush) following species introduction. Ecology of Freshwater Fish 21:300-308.

Wang, N. and R. Eckmann. 1994. Distribution of perch (Perca fluviatilis L.) during their first year of life in Lake Constance. Hydrobiologia 277:135-143.

Webb, P. 1984. Body form, locomotion and foraging in aquatic vertebrates. American Zoologist 24:107-120.

Weber, M. J., J. M. Dettmers, D. H. Wahl, and S. J. Czesny. 2011. Size preferences and behaviors of native yellow perch foraging on invasive round gobies. Journal of Great Lakes Research 37:584-587.

Weidel, B., S. Carpenter, J. Cole, J. Hodgson, J. Kitchell, M. Pace, and C. Solomon. 2008. Carbon sources supporting fish growth in a north temperate lake. Aquatic Sciences 70:446-458.

Wellborn, G. A., D. K. Skelly, and E. E. Werner. 1996. Mechanisms creating community structure across a freshwater habitat gradient. Annual Review of Ecology and Systematics 27:337- 363.

Werner, E. E. and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in size- structured populations. Annual Review of Ecology and Systematics 15:393-425.

Werner, E. E., D. J. Hall, D. R. Laughlin, D. J. Wagner, L. A. Wilsmann, and F. C. Funk. 1977. Habitat partitioning in a freshwater fish community. Journal of the Fisheries Research Board of Canada 34:360-370.

Whiteside, M. C., C. M. Swindoll, and W. L. Doolittle. 1985. Factors affecting the early life history of yellow perch, Perca flavescens. Environmental Biology of Fishes 12:47-56.

Williams, R. J. and N. D. Martinez. 2000. Simple rules yield complex food webs. Nature 404:180- 183. 71

Wilson, G. G. 2005. Impact of invasive exotic fishes on wetland ecosystems in the Murray- Darling Basin. Pages 45-60 in: B. Phillips, editor. Native fish and wetlands of the Murray- Darling Basin: action plan, knowledge gaps and supporting papers Proceedings of a workshop held in Canberra. Murray-Darling Basin Commission, Canberra, Australia.

Wu, L. and D. A. Culver. 1992. Ontogenetic diet shift in Lake Erie age-0 yellow perch (Perca flavescens): a size-related response to zooplankton density. Canadian Journal of Fisheries and Aquatic Sciences 49:1932-1937.

Ylikarjula, J., M. Heino, and U. Dieckmann. 1999. Ecology and adaptation of stunted growth in fish. Evolutionary Ecology 13:433-453.

Zambrano, L., E. Valiente, and M. J. Vander Zanden. 2010. Food web overlap among native axolotl (Ambystoma mexicanum) and two exotic fishes: carp (Cyprinus carpio) and tilapia (Oreochromis niloticus) in Xochimilco, Mexico City. Biological Invasions 12:3061- 3069.

Zamora, L. and R. Moreno-Amich. 2002. Quantifying the activity and movement of perch in a temperate lake by integrating acoustic telemetry and a geographic information system. Hydrobiologia 483:209-218.

Zuanon, J. A. S., A. C. Pezzato, L. E. Pezzato, J. R. S. Passos, M. M. Barros, and C. Ducatti. 2006. Muscle δ13C change in Nile tilapia (Oreochromis niloticus): effects of growth and carbon turnover. Comparative Biochemistry and Physiology Part B - Biochemistry & Molecular Biology 145:101-107.

10 Annex

Table A 1: Clusters of values from P. fluviatilis & P. flavescens with distinct isotopic signals in the δ13C/δ15N-bi-plots identified by K-Cluster analysis. Significant differences in δ13C or δ15N are highlighted (bold; < 0.05 =*; < 0.01=**; < 0.0001=***). center lake size group cluster δ13C δ15N isotope df F sig. n distance (‰) (‰) 1 -25.31 11.14 δ13C 8 25.249 0.001** 4 small 0.72 2 -26.03 11.12 δ15N 8 0.012 0.915 6 PA 1 -25.67 10.74 δ13C 8 14.994 0.005* 3 large 1.44 2 -27.09 10.51 δ15N 8 0.522 0.491 7 1 -23.95 7.53 δ13C 17 97.909 0.000*** 11 WA large 4.38 2 -28.31 7.14 δ15N 17 6.679 0.019* 8 1 -32.31 12.40 δ13C 11 133.163 0.000*** 6 WI small 5.19 2 -27.13 12.65 δ15N 11 2.223 0.164 7 1 -31.28 9.76 δ13C 18 0.321 0.578 10 small 1.26 2 -31.19 8.56 δ15N 18 38.548 0.000*** 10 DO 1 -30.40 10.65 δ13C 10 55.193 0.000*** 8 large 2.12 2 -28.50 11.58 δ15N 10 12.713 0.005** 4 1 -30.21 6.95 δ13C 13 7.351 0.018* 11 small 1.17 2 -29.41 7.80 δ15N 13 13.034 0.003** 4 VA 1 -27.87 9.56 δ13C 13 39.504 0.000*** 10 large 1.67 2 -29.53 9.42 δ15N 13 1.798 0.203 5 1 -25.60 8.70 δ13C 17 64.484 0.000*** 5 WU large 2.15 2 -27.69 9.17 δ15N 17 8.264 0.011* 14

Table A 2: List of samples used for the SIAR mixing model for each lake. Pooled sources for the mixing model are marked grey. Species or taxa are given as identified with the number of analyzed samples (n) and diet group (ZP = zooplankton; MZB = zoobenthos; F = fish). Notes are in parentheses. lake species/taxon type n lake species/taxon type n Cladocera ZP 8 Zooplankton (bulk) ZP 8 Calanoida ZP 1 Chaoborus spec. ZP 3 Gammaridae MZB 4 Chironomidae MZB 4 Chironomidae MZB 4 Asellidae MZB 3 Lepidostomatidae MZB 3 P. fluviatilits (< 120 mm) F 16 PA Cotus asper F 5 Rutilus rutilus F 10 Pacifastacus leniusculus MZB 8 DO Orconectes limosus (replica) MZB 3 Zygoptera MZB 4 Aeshnidae MZB 2 Libellulidae MZB 2 Zygoptera MZB 4 Bellamya chinensis MZB 7 Leptoceridae MZB 2 Cladocera ZP 8 Lepidostomatidae MZB 4 Calanoida ZP 1 Tanypodinae MZB 1 Gammaridae MZB 5 Nepa cinerea MZB 1 Libellulidae MZB 4 Zooplankton (bulk) ZP 8 WA Leptoceridae MZB 2 Chaoborus spec. ZP 4 Chironomidae MZB 5 Chironomidae MZB 4 Aeshnidae MZB 7 Asellidae MZB 3 Zygoptera MZB 4 P.fluviatilits(< 120 mm) F 12 M. salmoides (<35 mm) F 5 Rutilus rutilus F 10 Cotus asper F 3 Orconectes limosus (replica) MZB 3 VA Cladocera ZP 4 Aeshnidae MZB 2 Calanoida ZP 1 Zygoptera MZB 4 Chironomidae (prof.) MZB 2 Leptoceridae MZB 4 Chironomidae (litt.) MZB 2 Lepidostomatidae MZB 3 Pacifastacus leniusculus MZB 13 Haliplidae MZB 2 Zygoptera MZB 4 Corixidae MZB 1 Bellamya chinensis MZB 10 Gyrinidae MZB 1 Libellulidae MZB 4 Zooplankton (bulk) ZP 8 Aeshnidae MZB 6 Chaoborus spec. ZP 3 Cotus asper F 10 Chironomidae MZB 4

WI Gammaridae MZB 4 Asellidae MZB 3 Leptoceridae MZB 6 P. fluviatilits(< 120 mm) F 1 Lepidostomatidae MZB 4 Rutilus rutilus F 5 Orconectes limosus MZB 3 WU Aeshnidae MZB 1 Zygoptera MZB 4 Leptoceridae MZB 3 Lepidostomatidae MZB 3 Tanypodinae MZB 2 Nepa cinerea MZB 2 Corixidae MZB 1 Gyrinidae MZB 1 74

Table A 3: Results of the one-way ANOVA in P. fluviatilis & P. flavescens for the trophic niche metrics (δ15N-range, δ13C-range, hull area, mean distance to centroid (CD), mean nearest neighbour distance, (MNND) & standard deviation of the nearest neighbour distance (SDNND). sum of mean squares df square F sig. δ15N-range 1.008 1 1.008 2.393 0.166 δ13C-range 8.515 1 8.515 1.802 0.221 hull area 2.292 1 2.292 0.838 0.390 CD 1.010 1 1.010 2.025 0.198 MNND 0.011 1 0.011 1.130 0.323 SDNND 0.017 1 0.017 1.962 0.204

P. flavescens TL < 158 mm: a) PA1 small YP 40% 35% 30% 25% 20% 15% 10% 5% 0%

b) PA2 small YP 30% 25% 20% 15% 10% 5% 0%

c) WI1 small YP 30% 25% 20% 15% 10% 5% 0%

d) WI2 small YP 70% 60% 50% 40% 30% 20% 10% 0%

P. flavescens TL > 158 mm: e) PA1 large YP 35% 30% 25% 20% 15% 10% 5% 0%

f) PA2 large YP 35% 30% 25% 20% 15% 10% 5% 0%

g) WA1 large YP 80% 70% 60% 50% 40% 30% 20% 10% 0%

h) WA2 large YP 40% 35% 30% 25% 20% 15% 10% 5% 0%

P. fluviatilis < 169 mm: a) DO1 small EP 35% 30% 25% 20% 15% 10% 5% 0%

b) DO2 small EP 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

c) VA1 small EP 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

d) VA2 small EP 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

P. fluviatilis > 169 mm: e) DO large EP 80% 70% 60% 50% 40% 30% 20% 10% 0%

f) VA1 large EP 70% 60% 50% 40% 30% 20% 10% 0%

g) VA2 large EP 70% 60% 50% 40% 30% 20% 10% 0%

h) WU1 large EP 60% 50% 40% 30% 20% 10% 0%

i) WU2 large EP 70% 60% 50% 40% 30% 20% 10% 0%

11 Acknowledgements

I want to express my thanks to Thomas Mehner who offered a lot of advice and guided me through the process, to Robert Arlinghaus who made this thesis possible and encouraged me to do it, to Julian Olden who accommodated me during my time abroad and supported me from then till now, to Laura Twardochleb who “shared my faith” and provided many useful pointers, to Andy Davis who always lent a hand, to Kristin Scharnweber who spared much time and effort on my questions. Also, I want to thank Alexander Türck, Daniel Hühn, Asia Vogt and Matthias Emmrich from the IGB, the staff of University of Washington’s School of Aquatic & Fishery Sciences, Brigitte Keitz and Brandon Goeller for helping hands and helping thoughts.

12 Erklärung

Hiermit erkläre ich, die vorliegende Masterarbeit selbstständig verfasst und keine anderen als die angegebenen Quellen und Hilfsmittel benutzt zu haben.

Datum: 8.11.2013 Unterschrift: