Direct Effects of Daphnia-Grazing, Not Infochemicals, Mediate a Shift

ARTICLE IN PRESS

Limnologica 37 (2007) 137–145 www.elsevier.de/limno

Direct effects of Daphnia-grazing, not infochemicals, mediate a shift towards large inedible colonies of the gelatinous green alga Sphaerocystis schroeteri Heike Kampea,Ã, Marie Ko¨nig-Rinkeb, Thomas Petzoldtb,Ju¨rgen Benndorfb aInstitute for Biochemistry and Biology, University of Potsdam, Maulbeerallee 2, D-14469 Potsdam, Germany bInstitute of Hydrobiology, Zellescher Weg 40, 01217 Dresden, Germany

Received 30 August 2006; received in revised form 19 December 2006; accepted 5 January 2007

Abstract

The influence of Daphnia galeata hyalina grazing and of infochemicals released by the daphnids on the colony size and growth rate of the colonial gelatinous green alga Sphaerocystis schroeteri (Chlorococcales) was investigated in laboratory batch experiments run for 96 h. High zooplankton grazing pressure was exerted by a final concentration of 100 daphnids L1 in the Daphnia treatments. Infochemicals were obtained by filtration (0.2 mm) of water from D. galeata hyalina cultures (200 ind. L1 exposed for 24 h). This filtrate was added to the S. schroeteri cultures in two concentrations corresponding to 7 and 50 daphnids L1, respectively. The growth rate of S. schroeteri was neither affected significantly by direct Daphnia grazing nor by the presence of Daphnia infochemicals, in comparison to the control. However, the portion of inedible S. schroeteri colonies (diameter450 mm) increased under direct grazing pressure, whereas the Daphnia infochemicals did not influence the colony size significantly. We conclude that the shift in colony size by direct zooplankton grazing denotes an effective defence mechanism against size selective feeding for colonial gelatinous green algae. This effective defence in combination with unchanged growth rates of the larger colonies (under non-limiting nutrient and light conditions) falsifies the assumption of a trade-off between minimising grazing losses and maximising growth by optimising the colony size. r 2007 Elsevier GmbH. All rights reserved.

Keywords: Colonial gelatinous green algae; Zooplankton grazing; Infochemicals; Algal defence mechanism

Introduction 1981; Nizan, Dimentman, & Shilo, 1986), an increase in size by the formation of colonies is a common and Zooplankton grazing is an important loss factor for efficient defence mechanism. The latter mechanism is such types of phytoplankton which do not develop particularly efficient against size selective filter feeders effective defence mechanisms against grazing. In addi- which feed mainly on particles between 2 and 30 mm tion to the bizarre or filamentous shape of cells or (Lehman, 1988; Sterner, 1989). Higher sinking rates, the colonies (Gliwicz & Siedlar, 1980; Hawkins & Lampert, decreasing efficiency of nutrient uptake and utilisation of 1989) and the formation of chemical repellents (Lampert, light cause a decline of net growth rates of algal populations with increasing size (Reynolds, 1984a). ÃCorresponding author. Tel.: +49 331 977 1955. Thus, a trade-off for algal size exists between efficient E-mail address: [email protected] (H. Kampe). grazing resistance and high growth rates. Gelatinous

0075-9511/$ - see front matter r 2007 Elsevier GmbH. All rights reserved. doi:10.1016/j.limno.2007.01.001 ARTICLE IN PRESS 138 H. Kampe et al. / Limnologica 37 (2007) 137–145 green algae are a common example of the colonial Chlorella vulgaris (Boraas, Seale, & Boxhorn, 1998). As organisation of phytoplankton. Reynolds (1984a) a result of this phenotypic plasticity, these algae have a divided these algae into two functional groups: The first fundamental advantage in the phytoplankton commu- comprises the immotile genera Sphaerocystis (synonym nity due to their high competitive ability in the presence Pseudosphaerocystis), Coenochloris and Oocystis, which as well as in the absence of grazers. occur mostly during the early summer period in In the present work we investigate the influence of a mesotrophic lakes. The second group contains the genera filter feeder (Daphnia galeata hyalina) on the growth Eudorina, Pandorina and Volvox, which are motile rate and colony size of an immotile gelatinous green alga species and also occur during the summer period but in (Sphaerocystis schroeteri). The following three hypoth- eutrophic lakes. The functional difference between these eses derived from the literature cited above were tested: two groups results from the motility which allows species (i) the presence of Daphnia increases the colony size of from the second group to regulate their position in S. schroeteri, (ii) the shift in colony size is mediated by relation to the light field (Reynolds, Huszar, Kruk, an infochemical released by Daphnia, and (iii) the Naselli-Flores, & Melo, 2002). Species of the first group growth rate of S. schroeteri is not significantly reduced react sensitively to phosphorus and nitrogen enrichment in Daphnia-induced large colonies when light and because of their lower competitive ability for light and nutrients are available at saturation levels. carbon (Reynolds et al., 2002). Gelatinous green algae have been reported to benefit from high zooplankton biomass overall (Porter, 1975; Sommer, Sommer, Santer, Material and methods &Zo¨llner, 2003; Vanni, 1987) and, because of zooplankton induced nutrient recycling, at unsaturated nutrient levels in particular (Kagami, Yoshida, Gurung, & Starter cultures Urabe, 2002). The functional role of the gelatinous AstrainofSphaerocystis schroeteri (217-1b) from the sheath has been discussed and is controversial. The culture collection of algae (SAG) of the University of secretion of a gelatinous sheath has long been postulated Gottingen was used for the experiments. Light for the as a mechanism for reducing sinking rates (Reynolds, ¨ starter cultures was provided with a 16:8 h light–dark 1984b). Furthermore, the increase in size using gelatine cycle and at 125 (75) mEm2 s1. Cultures were kept in can prevent ingestion by filter-feeding zooplankton (Burns, 1968) with concomitant improved intercellular membrane-filtered (0.2 mm, Schleicher & Schuell) lake water from the mesotrophic Saidenbach Reservoir diffusion in comparison to non-gelatinous colonial forms (Saxony, Germany) enriched with phosphorus at non- (Reynolds, 1984b). Otherwise, the production of gelatine limiting concentrations. Before use for cultivation, the is energy-consuming and also leads to an increase of water was stored for 2 weeks to ensure the decline of sinking velocity with increasing size (Stoke’s law). Porter naturally occurring infochemicals. The experimental (1973, 1975, 1976) reported the viable gut passage of animals were taken from a laboratory culture of Daphnia gelatinous green algae in Daphnia with an additional galeata hyalina reared in the same lake water as was nutrient uptake and enhanced algal growth after the passage. She concluded that the gelatinous sheath used for the experiments. The inoculum of this culture consisted of animals caught in April 2003 in the eutrophic constitutes an effective protection against digestion and Bautzen Reservoir (Saxony, Germany). The culture zooplankton grazing enhances the growth of gelatinous animals were fed thrice weekly with Scenedesmus sp. and green algae. In contrast, Stutzman (1995) found neither additionally once weekly with Sphaerocystis schroeteri. significant increase in mortality nor decrease in the number of moults and clutches of Daphnia feeding on monocultures of Sphaerocystis or Pandorina. However, Experimental conditions the number of offspring per clutch was significantly reduced. In comparison to Diaptomus, Daphnia utilised Tests were run in batch cultures in 300 mL Erlen- the gelatinous green algae more effectively in Stutzman’s meyer flasks with 150 mL culture volume and non- (1995) study. limiting nutrient concentrations (4100 mgPO4- A grazer-mediated unicell-to-colony transformation PL1,44.2 mg inorganic N L1). The initial algal cell and formation of spines, connected with an improved concentration was 23007250 cells mL1. Flasks were protection against grazing, has been reported in some incubated on a rotating table at 18 1C and a light strains of the non-gelatinous green alga Scenedesmus intensity of 125 (75) mEm2 s1, the light–dark cycle (Hessen & Van Donk, 1993; Lampert, Rothaupt, & von was 16:8 h for experiment 1 and 12:12 h for experiment Elert, 1994; Lu¨rling & Van Donk, 1996). After exposure 2. Flasks were shaken every 10 min for 1 min. The to an infochemical released by Daphnia the formerly daphnids used for the grazing experiments were replaced unicellular algae formed colonies. Grazer-induced col- by new animals every 2 days. We used similar sized adult ony formation has also been reported in the green alga animals for all experiments (1.870.2 mm). ARTICLE IN PRESS H. Kampe et al. / Limnologica 37 (2007) 137–145 139

Experiment 1 (low infochemical concentration) solution. The growth rates (d1) were calculated during the exponential period of growth according to A first experiment was intended to differentiate the m ¼ðln N2 ln N1Þ=ðt2 t1Þ, (1) effects of direct Daphnia grazing and of Daphnia infochemicals on the size distribution of Sphaerocystis where N1, N2 are the fluorescence intensities at time t1, schroeteri. The experimental design included the control, t2, respectively. The maximum size of colonies that Daphnia treatment and infochemical treatment, each could be ingested by the daphnids was determined comprising 3 replicates. Control flasks contained neither according to Burns (1968) by daphnids nor infochemicals. The Daphnia treatments y ¼ 22x þ 4:87, (2) contained 15 animals (100 ind. L1). To produce the where y is the diameter (mm) of the largest colony that infochemical water, daphnids (200 ind. L1)wereincu- can be ingested and x is the carapace length (mm) of the bated in membrane-filtered lake water for 24 h. During daphnids. To investigate the morphology of the algae, this time they were fed with Scenedesmus sp. (80%) and additional samples were taken at the end of the Sphaerocystis schroeteri (20%). The solution was filtered experiment and immediately preserved with Lugol’s through 0.2 mm membrane filter (Schleicher & Schuell) solution. and immediately used for the experiment. As a consequence of dilution, the concentration of the infochemical water in the treatment flasks was equivalent to Experiment 3 (Daphnia-grazing) 7 individuals L1 (i.e. low zooplankton abundance). After 48 and 96 h, respectively, the algal size distribution was Subsidiary to experiment 1, a third experiment was determined by measuring the diameter of at least 80 conducted to confirm the effects of direct Daphnia individual algae (colonies as well as single cells) using the grazing on the colony size distribution of Sphaerocystis inverted microscope technique (Utermo¨hl, 1958). schroeteri and to additionally investigate algal growth rates. Compared to experiment 1 the number of replicates was increased from 3 to 5. Experimental Experiment 2 (high infochemical concentration) conditions and the determination of growth rates and algal colony size were equivalent to experiment 2. The In the first experiment the Daphnia abundance direct grazing treatment was run with 15 daphnids per 1 simulating a possible infochemical effect was much treatment (100 ind. L ) and a control without animals. lower (7 ind. L1) than the abundance in the direct grazing treatment (100 ind. L1). Therefore, as a next Statistical analysis step for investigating the indirect infochemical effects, we conducted a second experiment (duration 96 h, Growth rates were compared by means of t-tests. treatments with 5 replicates) simulating higher zoo- Graphical analysis revealed no deviance from normal plankton abundance (50 ind. L1) in the infochemical distribution. However, the Daphnia treatments of treatment. The infochemical water was produced as experiment 3 showed an increased variance compared described for the first experiment, but to avoid a to the control (F-test, po0.001). To account for this, the possible absorption of infochemicals during the filtra- t-test for unequal variances (Welch test) was applied in tion process we used glass fibre filters (Sartorius) instead this case. To test whether Daphnia or infochemical of membrane filters. Infochemicals from fish or Chao- addition would influence the scale, shape or type of the borus are known to be destroyed by microbial degrada- colony size distribution a general comparison of tion after a few days (Loose, von Elert, & Dawidowicz, distributions was employed for each of the three 1993; Tollrian & Dodson, 1999). To avoid such a experiments separately. In a first step (distribution decrease of the infochemical concentration, freshly fitting), a gamma distribution was fitted successfully to produced infochemical water was added after 48 h to the pooled control data using the maximum likelihood the infochemical treatments. Control flasks were in- method for grouped data. In the second step (compar- cubated without infochemicals, but received an equiva- ison step), the empirical size distributions of the lent volume of medium after 48 h. Light was provided treatments were compared against the fitted distribution with a 12:12 h light–dark cycle. of the control using the Chi2-test (p-value in Table 1). To Every 24 h a volume of 4 mL was sampled for the account for possible random errors introduced by the determination of algal growth rates. After a treatment experimental design (i.e. that the observed deviation with ultrasonic waves (GM 70 Sonoplus Bandulin may be a result from the flasks and not the treatment), Electronic), for disintegrating colonies into single cells, an additional nonparametric bootstrap was applied fluorescence intensity was measured at a wavelength of according to the general methodology of Efron and 685 nm (Luminescence Spectrometer, Perkin Elmer, LS Tibshirani (1993). The bootstrap was repeated 1000 50 B). Then samples were preserved with Lugol’s times with random assignment of the control and the ARTICLE IN PRESS 140 H. Kampe et al. / Limnologica 37 (2007) 137–145

Table 1. Chi2 goodness of ﬁt-test of the empirical distribution of colony sizes in the treatments against the ﬁtted gamma distribution of the respective controls

Experiment Chi2 d.f. p-value Bootstrap p-value

Experiment 1 2.238 11 0.998 0.941 Infochemical Daphnia grazing 287.82 11 o0.001 0.023 Experiment 2 5.62 9 0.78 0.365 Infochemical Experiment 3 126.59 11 o0.001 0.017 Grazing

The degrees of freedom (d.f.) and the corresponding p-value account for the number of size classes and not the number of treatments. Bootstrap p-values were estimated to account for a possible random effect of the replicates. treatment replicates to the distribution fitting and the between control+infochemical treatments and Daphnia comparison step. The bootstrap estimate of the p-value treatment (Mann–Whitney U test, p ¼ 0.02). is derived from the empirical distribution function of the bootstrap Chi2 sample as probability of the observed Experiment 2 (high infochemical concentration) Chi2-value. The described procedure should be sensitive against The maximum size of particles that could be ingested any change of the distributions (omnibus test), and to by the daphnids was determined by Eq. (2) to be 45 mm ease interpretation an additional Mann–Whitney U test and rounded up to 50 mm. The distribution of the colony was performed to account for the relative fractions of size was not significantly different for the control and edible ( 50 mm) and inedible (450 mm) colonies. The o the infochemical treatment (Fig. 2A and Table 1). The Mann–Whitney test was applied because of non-normal- largest proportion of algae was found between 8 and ity but equal distribution type of the fractions, and with 20 mm in diameter and consisted mainly of single cells. exact p-values to account for ties in the data. All Between 3% and 21% of the colonies in the control and statistical estimations and tests were performed using between 10% and 29% in the infochemical treatment the open source data analysis system R (R Development were larger than 50 mm in diameter and consequently Core Team, 2006). inedible for daphnids. There was no significant influence of the treatment on the size distribution of the colonies (Table 1). Results The mean of growth rates calculated during the exponential growth phase was 0.52 d1 for the controls and 0.58 d1 for the infochemical treatment (Fig. 3A). Experiment 1 (low infochemical concentration) There were no significant differences in growth rates (t-test, t ¼ 2.3, d.f. ¼ 8, p ¼ 0.0502). After 48 h the algal size distribution was not significantly different between the control and treatments (Fig. 1A and Table 1). The cultures were Experiment 3 (Daphnia grazing) dominated by small colonies, mainly of the size class 20–30 mm. A considerable portion belonged to the size Unlike the findings from experiment 2 (high infochem- classes 30–40 and 8–20 mm (mainly single cells). The ical concentration) the distribution of the colony size, lowest portion of single cells was found in the Daphnia- determined after 96 h, differed significantly (p ¼ 0.017, treatment (3–13%). The size distribution changed see Table 1) between control and direct Daphnia grazing dramatically with time in all samples. Furthermore, (Fig. 2B). The range of the colony sizes increased and in after 96 h (Fig. 1B) there were distinct differences contrast to the gamma distributed control, the treatments between the control and infochemical treatment on the showed a bimodal distribution. In the control flasks one hand and Daphnia treatment on the other. Control 80–99% of the algal particles were in the size fraction and infochemical replicates were dominated by single 8–20 mm, which consists almost exclusively of single cells cells (8–20 mm). All other size classes were negligible. In and only a few very small colonies. The observed colonies contrast, the Daphnia treatment contained distinctly were mainly between 30 and 40 mm in diameter. The fewer single cells and elevated portions of colonies number of large colonies with a diameter X40 mmwas 450 mm (28–57%). The differences in portions of small negligible in the control set. In the Daphnia treatment the edible and large inedible colonies were significant distribution of colony size was very diverse between the ARTICLE IN PRESS H. Kampe et al. / Limnologica 37 (2007) 137–145 141

100 A control 80 infochemical after 48 h 60 Daphnia grazing

rel. frequency (%) 20

0 100 B 80 after 96 h

rel. frequency (%) 20

8-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-10 100-150 150-200 size class (µm)

Fig. 1. Experiment 1: Size class distribution of Sphaerocystis schroeteri after 48 h (A) and 96 h (B) of exposure to low Daphnia galeata hyalina infochemical concentration (equivalent to 7 ind. L1) and direct grazing by Daphnia galeata hyalina (100 ind. L1), control and treatments comprise 3 replicates.

100 control A 80 infochemical

rel. frequency (%) frequency rel. 20

0 100 control B 80 Daphnia grazing

rel. frequency (%) frequency rel. 20

8-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-150 150-200 size class (µm)

Fig. 2. Experiment 2 and 3: Size class distribution of Sphaerocystis schroeteri after 96 h of infochemical treatment (equivalent to 50 ind. L1) (A) and direct grazing of Daphnia galeata hyalina (100 ind. L1) (B), each treatment was run with a separate control; control and treatments comprise 5 replicates. individual replicates. Most of the algae were found in the X50 mm. The largest colonies with a diameter of size class 8–20 mm, but in contrast to the control there 150–200 mm occurred in only one replicate. The portion was a distinct portion of colonies with a diameter of inedible colonies was between 0% and 5% in the ARTICLE IN PRESS 142 H. Kampe et al. / Limnologica 37 (2007) 137–145

1.2 ABAB 1.0 ) 1 − 0.8 (d

0.6

0.4 growth rate 0.2

0.0 control infochem. control grazing

Fig. 3. Experiment 2 and 3: Mean growth rates (with 95% confidence intervals) of Sphaerocystis schroeteri during the exponential growth phase in the infochemical treatment (equivalent to 50 ind. L1) (A) and the treatment with direct grazing by Daphnia galeata hyalina (100 ind. L1) (B), control and treatments with 5 parallels. control and between 3% and 74% in the direct Daphnia increased from a relatively low concentration corre- grazing treatments. Under direct grazing pressure the sponding to 7 individuals L1 (experiment 1) to a high portion of large, inedible algal colonies increased concentration corresponding to 50 individuals L1 (ex- significantly (Mann–Whitney U test, p ¼ 0.016). periment 2). Lu¨rling (2003) investigated the influence of The mean of growth rates calculated during the increasing infochemical concentrations of herbivorous exponential growth phase was 0.65 d1 for the controls zooplankton on the colony formation of Scenedesmus and 0.74 d1 for the Daphnia treatment (Fig. 3B). obliquus. He already observed infochemical-induced Growth rates did not differ significantly between control colony formation at abundances exceeding 5 in- and treatment (F-test: F ¼ 213.3, po0.001, Welch test, dividuals L1 of Daphnia galeata hyalina. The highest t ¼ 1.44, d.f. ¼ 4, p ¼ 0.22). infochemical concentration tested in his study corre- sponded to 40 individuals L1. Therefore, and as the duration of our experiments was long enough to get an algal response (Lampert et al., 1994; Van Donk, Discussion Lu¨rling, Hessen, & Lokhorst, 1997), we expected an algal reaction in our infochemical experiments. Never- The life cycle of Sphaerocystis schroeteri is very theless, even in experiment 2 no effects of Daphnia complex and its morphological shape shows high infochemical on the morphology of Sphaerocystis were diversity. As described by Komarek and Fott (1983), detected. Thus, our hypothesis (ii) assuming a shift in the alga forms colonies of 2–64 cells which are colony size in response to an infochemical released by surrounded by a gelatinous sheath (Ettl and Ga¨rtner Daphnia must be rejected. (1988): 2–32 cells/colony). The species reproduces by Testing hypothesis (ii) might be biased by three division of autospores within the mother colony and problems related to our experimental design. A first releases the daughter colonies by dissolving the mother problem consists in the possibility of different produc- cell wall. Occasionally zoospores with flagella occur tion of infochemicals by different Daphnia clones. To (asexual reproduction). Our laboratory cultures (with- our knowledge, little is known about a clone-specific out zooplankton) were often dominated by single cells production of Daphnia infochemicals. The induction of or small colonies, depending on the age of the cultures algal defence mechanisms seems to depend on the (unpubl. results, H. Kampe, Institute of Hydrobiology), zooplankton species rather than on clones (Lu¨rling whereas colonies consisting of much more than 64 cells 2003). However, even if differences might exist in the were observed in the grazing experiments. production of infochemicals between different Daphnia Our investigations revealed a shift towards large clones, this should not have any consequences for our inedible colonies of Sphaerocystis by direct Daphnia experiments. As all animals used for the experiments galeata hyalina grazing, thus supporting hypothesis (i) were taken from the same culture, clonal differences which postulates that the presence of Daphnia increases between daphnids used for the direct grazing treatments the colony size of S. schroeteri. This effect could only be and daphnids used for the production of the infochem- seen under direct grazing pressure and was not inducible icals were largely excluded. by Daphnia infochemicals. To ensure that the algae A second potential bias might consist in the unknown really were exposed to a substantial amount of Daphnia bacterial by-flora in our experiments. There are several infochemicals, the infochemical concentration was studies on the essential influence of bacteria or intestinal ARTICLE IN PRESS H. Kampe et al. / Limnologica 37 (2007) 137–145 143 enzymes on the production of ‘fish kairomones’ (Lass & These findings support hypothesis (iii) which postulates Spaak, 2003; Ringelberg & Van Gool, 1998). We do not that the growth rate of S. schroeteri is not significantly know whether the bacteria present in our cultures are reduced in Daphnia-induced large colonies when light involved, in analogy to the kairomone production, also and nutrients are available at saturation levels. Thus, in the formation of Daphnia infochemicals. In fact, there under light and nutrient saturation the Daphnia-induced are indications that gut bacteria of Daphnia are involved enlargement of the colonies does not imply a trade-off in the production of infochemicals because antibiotics between reduced growth rate, due to lowered light and inhibited the induction of algal colonies (Van Donk, nutrient availability, and protection against grazing. 1997; Van Holthoon, Van Beek, Lu¨rling, Van Donk, & Under these conditions and when sedimentation can be De Groot, 2003). Furthermore, the feeding activity of neglected as a loss factor, there is a clear ecological Daphnia is decisive for the biological activity of the advantage of forming larger colonies under strong infochemical substance (Lampert et al., 1994; Lu¨rling, grazing pressure. However, under distinct nutrient 2003; Van Donk, 1997). Homogenised algae or daph- and/or light limitation and in habitats where sedimenta- nids did not induce colony formation in Scenedesmus tion is an important loss factor, an ecological disadvan- (Lampert et al., 1994). Since we used the same tage must be expected to explain why S. schroeteri does laboratory cultures of algae and daphnids as inocula not form large colonies permanently. for our experiments major differences in the bacterial We conclude that this colonial green alga is able to assemblages are unlikely. If bacteria were involved in the reduce grazing losses even at high filter-feeding zoo- production of infochemicals in our experiments, they plankton densities by forming large colonies. Our should have exerted marginal effects on the differences laboratory results corroborate analogous findings from in the results only because all animals in the different field studies (Kagami et al., 2002; Porter, 1975; Sommer treatments were of the same origin and, therefore, had et al., 2003). However, we could not define the trigger of the same bacterial by-flora. this reaction. In contrast to experiments with Scenedes- A third confounding effect might exist in the variation mus (Hessen & Van Donk, 1993; Lampert et al., 1994; of the size distribution of S. schroeteri at the start of the Lu¨rling, 2003) our results suggest that Sphaerocystis different experiments, which was unavoidable due to the does not respond to Daphnia infochemicals. Potentially, high morphological diversity in the inoculum cultures. mechanical impulses, direct contact with the filter The portion of single cells and small colonies depended apparatus of Daphnia, or other signals are possible on the age of the culture and probably other physiolo- elicitors of the shift in colony size. Different mechanisms gical conditions. We intended to have similar starting as to how these elicitors can induce colony enlargement conditions by using algal inocula that were 6 days old, are conceivable. In the desmid, Staurastrum, a clumping after culture dilution and identical culture conditions. of cells could be observed when daphnids were present, However, moderate discrepancies in colony sizes could whereas no formation of these mucus globules appeared not be averted. As a consequence of these differences the in Daphnia water alone (Wiltshire, Boersma, & Meyer, start size distributions of S. schroeteri were not 2003). Reynolds and Rodgers (1983) demonstrated the comparable in all experiments. On the other hand, the release of fast growing daughter colonies of Eudorina, different initial conditions regarding the algal inocula which, after a very short time, reached a size that was led to much stronger test criteria, as higher amounts of not ingestible. We assume for S. schroeteri that the single cells in the algal inocula enhance the portion of presence of daphnids inhibits an early break up of the potentially ingestible algae. Despite a portion of mother colonies by any of the elicitors mentioned above, 80–100% single cells in the inocula, the formation and thus leading to larger colonies. In contrast, the colonies growth of large colonies in the grazing treatments was of the control and the infochemical treatment break observed. Thus, considering all the three problems down into single cells when they have reached a certain discussed above we conclude that differences in the size. This morphological adaptation to different envir- algal colony sizes must be related to the different onmental conditions leads to a wider ecological niche Daphnia effects in the treatments rather than on compared to algal species not capable of such an confounding effects. adaptation. The growth rate of Sphaerocystis was not reduced by direct Daphnia grazing in spite of the formation of large colonies. This is in accordance with several other studies Acknowledgements that did not find any differences in algal growth rates between treatments (with large colonies) and controls We wish to thank Susanne Rolinski for critical (single cells) in experiments with Scenedesmus sp. and comments on the manuscript and Mark Clegg for Daphnia infochemicals (Hessen & Van Donk, 1993; linguistic improvements. Two anonymous referees made Lampert et al., 1994; Lu¨rling, 1998, 2003; Lu¨rling, De constructive comments on an earlier version of the Lange, & Van Donk, 1997; Lu¨rling & Van Donk, 1996). manuscript. Financial support was provided by the ARTICLE IN PRESS 144 H. Kampe et al. / Limnologica 37 (2007) 137–145

German Bundesministerium fu¨r Bildung und Forschung Lu¨rling, M. (2003). The effect of substances from different under Grant 02 WT0233. zooplankton species and fish on the induction of defensive morphology in the green alga Scenedesmus obliquus. Journal of Plankton Research, 25, 979–989. Lu¨rling, M., De Lange, H. J., & Van Donk, E. (1997). Changes References in food qualitiy of the green alga Scenedesmus induced by Daphnia infochemical: Biochemical composition and mor- Boraas, M. E., Seale, D. B., & Boxhorn, J. E. (1998). phology. Freshwater Biology, 38, 619–628. Phagotrophy by a flagellate selects for colonial prey: A Lu¨rling, M., & Van Donk, E. (1996). Zooplankton-induced possible origin of multicellularity. Evolutionary Ecology, 12, unicell–colony transformation in Scenedesmus acutus and its 153–164. effect on growth of herbivore Daphnia. Oecologia, 108, 432–437. Burns, C. W. (1968). The relationship between body size of Nizan, S., Dimentman, C., & Shilo, M. (1986). Acute toxic filter-feeding cladocera and the maximum size of particles effects of the cyanobacterium Microcystis aeruginosa on ingested. Limnology and Oceanography, 13, 675–678. Daphnia magna. Limnology and Oceanography, 31, 497–502. Efron, B., & Tibshirani, R. J. (1993). An introduction to the Porter, K. G. (1973). Selective grazing and differential Bootstrap. Boca Raton, FL: Chapman & Hall/CRC. digestion of algae by zooplankton. Nature, 244, 179–180. Ettl, H., & Ga¨rtner, G. (1988). Su¨ßwasserflora von Mitteleu- Porter, K. G. (1975). Viable gut passage of gelatinous green ropa; Chlorophyta II: Tetrasporales, Chlorococcales, Gloeo- algae ingested by Daphnia. Verhandlungen der Internatio- dendrales. Jena: Gustav Fischer Verlag. nalen Vereinigung fu¨r Limnologie, 19, 2840–2850. Gliwicz, M., & Siedlar, E. (1980). Food size limitation and Porter, K. G. (1976). Enhancement of algal growth and algal interfering with food collection in Daphnia. Archiv fu¨r productivity by grazing zooplankton. Science, 192, 1332–1334. Hydrobiologie, 88, 155–177. R Development Core Team. (2006). R: A language and Hawkins, P., & Lampert, W. (1989). The effect of Daphnia environment for statistical computing. Vienna, Austria: R body size on filtering rate inhibition in the presence of a Foundation for Statistical Computing. ISBN:3-900051-07-0, Limnology and Oceanography filamentous cyanobacterium. , URL /http://www.R-project.orgS. 34, 1084–1088. Reynolds, C. S. (1984a). Phytoplankton periodicity: The Hessen, D. O., & Van Donk, E. (1993). Morphological interactions of form, function and environmental varia- changes in Scenedesmus induced by substances released bility. Freshwater Biology, 14, 111–142. from Daphnia. Archiv fu¨r Hydrobiologie, 127, 129–140. Reynolds, C. S. (1984b). The ecology of freshwater phyto- Kagami, M., Yoshida, T., Gurung, T. B., & Urabe, J. (2002). plankton. Cambridge: Cambridge University Press. Direct and indirect effects of zooplankton on algal Reynolds, C. S., Huszar, V., Kruk, C., Naselli-Flores, L., & composition in in situ grazing experiments. Oecologia, Melo, S. (2002). Towards a functional classification of the 133, 356–363. freshwater phytoplankton. Journal of Plankton Research, Komarek, J., & Fott, B. (1983). Das Phytoplankton des 24, 417–428. Su¨ßwassers, 7. Teil, 1. Ha¨lfte, Chlorophyceae, Ordnung Reynolds, C. S., & Rodgers, M. W. (1983). Cell- and colony- Chlorococcales. Stuttgart: E. Schweizerbart0sche Verlags- Eudorina buchhandlung. division in (Chlorophyta: Volvocales) and some Lampert, W. (1981). Toxicity of blue-green Microcystis ecological implications. British Phycological Journal, 18, aeruginosa: Effective defense mechanism against grazing 111–119. pressure by Daphnia. Verhandlungen der Internationalen Ringelberg, J., & Van Gool, E. (1998). Do bacteria, not fish, Vereinigung fu¨r Limnologie, 21, 1436–1440. produce ‘fish kairomone’? Journal of Plankton Research, 20, Lampert, W., Rothaupt, K. O., & von Elert, E. (1994). 1847–1852. Chemical induction of colony formation in a green alga Sommer, U., Sommer, F., Santer, B., & Zo¨llner, E. (2003). (Scenedesmus acutus) by grazers (Daphnia). Limnology and Daphnia versus copepod impact on summer phytoplankton: Oceanography, 39, 1543–1550. Functional compensation at both trophic levels. Oecologia, Lass, S., & Spaak, P. (2003). Chemically induced anti-predator 135, 639–647. defences in plankton: A review. Hydrobiologia, 491, Sterner, R. W. (1989). Resource competition during seasonal 221–239. succession toward dominance by cyanobacteria. Ecology, Lehman, J. T. (1988). Selective herbivory and its role in the 70, 229–245. evolution of phytoplankton growth strategies. In C. D. Stutzman, P. (1995). Food quality of gelatinous colonial Sandgren (Ed.), Growth and reproductive strategies of chlorophytes to the freshwater zooplankters Daphnia freshwater phytoplankton. Cambridge: Cambridge Univer- pulicaria and Diaptomus oregonensis. Freshwater Biology, sity Press. 34, 149–153. Loose, C. J., von Elert, E., & Dawidowicz, P. (1993). Tollrian, R., & Dodson, S. I. (1999). Inducible defenses in Chemically-induced diel vertical migration in Daphnia:A Cladocera: Constraints, costs and multipredator environ- new bioassay for kairomones exuded by fish. Archiv fu¨r ments. In R. Tollrian, & C. D. Harvell (Eds.), The ecology Hydrobiologie, 126, 329–337. and evolution of inducible defenses (pp. 177–202). Princeton, Lu¨rling, M. (1998). Effect of grazing-associated infochemicals NJ: Princeton University Press. on growth and morphological development in Scenedesmus Utermo¨hl, H. (1958). Zur Vervollkommnung der quantitativen acutus Meyen (Chlorophyceae). Journal of Phycology, 34, Phytoplankton-Methodik. Verhandlungen der Internationa- 578–586. len Vereinigung fu¨r Limnologie, 9, 1–39. ARTICLE IN PRESS H. Kampe et al. / Limnologica 37 (2007) 137–145 145

Van Donk, E. (1997). Defenses in phytoplankton against Scenedesmus: A literature overview and further steps grazing induced by nutrient limitation, UV-B stress and towards the chemical characterisation of the Daphnia infochemicals. Aquatic Ecology, 31, 53–58. kairomone. Hydrobiologia, 491, 241–254. Van Donk, E., Lu¨rling, M., Hessen, D. O., & Lokhorst, G. M. Vanni, M. J. (1987). Effects of nutrients and zooplankton size (1997). Altered cell wall morphology in nutrient-deﬁcient on the structure of a phytoplankton community. Ecology, phytoplankton and its impact on grazers. Limnology and 68, 624–635. Oceanography, 42, 357–364. Wiltshire, K., Boersma, M., & Meyer, B. (2003). Grazer- Van Holthoon, F. L., Van Beek, T. A., Lu¨rling, M., Van induced changes in the desmid Staurastrum. Hydrobiologia, Donk, E., & De Groot, A. (2003). Colony formation in 491, 255–260.