Predation on rotifers by the suspension-feeding Calanoid copepod Diaptomus pallidus

Transcription

1 Limnol. Oceanogr., 31(2), 1986, , by the American Society of Limnology and Oceanography, Inc. Predation on rotifers by the suspension-feeding Calanoid copepod Diaptomus pallidus Craig E. Williamson and Nancy M. Butler Department of Biology, Lehigh IJniversity, Bethlehem, Pennsylvania Abstract Predation on rotifers by the small suspension-feeding Calanoid copepod Diaptomus pallidus was examined in order to quantify the effects of prey density, prey type, and the presence of algal food resources on ingestion rates, and to determine whether ingested rotifer biomass could be utilized to enhance the survival and reproduction of the copepods. Clearance and ingestion rates of D. pallidus on rotifers were times greater than on algae presented at the same concentration and similar to the maximum rates previously reported for more carnivorous cyclopoid copepods preying on rotifers. The survival and reproduction of D. palhdus were substantially enhanced by the addition of rotifers to a threshold algal diet. When presented with a natural assemblage of plankton, D. pallidus preferentially ingested certain rotifer species over others. Predation on rotifers by such diaptomids may form an important trophic link in freshwater planktonic food webs. The importance of rotifers in freshwater planktonic ecosystems has become increasingly apparent over the past several years. Although cladocerans and copepods are the dominant zooplankters in many limnetic environments (Porter 1977), at times rotifers may account for most of the community grazing (Bogdan and Gilbert 1982) and contribute substantially to the biomass and productivity (Makarewicz and Likens 1979) of zooplankton communities. Many species of invertebrate predators prey heavily on rotifers (Williamson 1983) and may subsequently increase the energy available to vertebrate predators as they package these small prey into larger particles (the invertebrate predators themselves) which are more available to the vertebrates (Confer and Blades 1975). The greater size and visibility of the invertebrate predators as food for the vertebrates more than compensates for the energy lost due to the presence of an additional trophic level. The most widespread and abundant invertebrate predators of planktonic rotifers are the copepods (Williamson 1983). Previous research on cyclopoid copepods has demonstrated the intense and selective nature of their predation on natural assemblages of rotifers (Brand1 and Fernando 1979; Williamson 1984). Less information is available on consumption of rotifers by Calanoid copepods, although several in- 1 Supported by NSF grant BSR to C.E.W. vestigators have reported the remains of rotifers in the guts of calanoids (Anderson 1967; Grygierek 1971; Maly and Maly 1974; Hairston 1979). Recent advances in our understanding of marine Calanoid feeding behavior have demonstrated that large particles are captured actively by coordinated movements of the mouthparts in response to chemical or mechanical signals given off by the particles in the current field near the copepod (Alcaraz et al. 1980; Koehl and Strickler 1981; Paffenhijfer et al. 1982). This suggests that even some smaller (N 1 mm) freshwater diaptomids previously thought to be totally herbivorous might prey on small, weakswimming animal prey such as rotifers and thus enhance the transfer of energy between trophic levels. The omnivorous nature of many species of marine calanoids has been well documented and supports this hypoth- esis (Mullin 1966; Robertson and Frost 1977; Landry 198 1; Paffenhijfer and Knowles 1980). We examine here the potential importance of predation by the freshwater suspension-feeding Calanoid Diapto- mus pallidus Herrick on several species of sympatric rotifers. The specific questions which we ask are: Does D. pallidus ingest rotifers, and if so, is this ingestion incidental or substantial? Can D. pallidus utilize ingested rotifer biomass to enhance its own survival and reproduction? And is predation by D. pallidus on different rotifer species uniform or selective? We thank R. Stemberger for advice on 393

2 394 Williamson and Butler the isolation and culture of the cryptomonads and rotifers, L. Forcina for assistance in several stages of the study, and Mr. and Mrs. F. B. Griswold for permitting us access to Whiteacre Pond and letting us use their boat. B. Hargreaves provided suggestions at several stages of this project. Methods The three primary organisms used were the Calanoid copepod D. pallidus, the rotifer Synchaeta oblonga Ehrenberg, and the naked flagellate Cryptomonas rejlexa Skuja. The three species coexist in the primary study site, Whiteacre Pond (40 35 N, W), a shallow (1 m deep) pond constructed by damming of the Black River. The C. reflexa was isolated from Whiteacre Pond and cultured in 0.2-pm filter-sterilized, modified MBL medium at 20 C and a 12 : 12 L/D cycle (73-PEinst cool-white fluorescent light) (Stemberger 1981). Synchaeta oblonga was also isolated from Whiteacre Pond and grown in sterile modified MBL medium under the same light and temperature conditions with C. reflexa as the sole food. Large numbers of organisms were dried to constant weight in an oven at 60 C and 380 mm of mercury vacuum with a slight purge to prevent condensation and weighed on a Cahn electrobalance. Except where otherwise stated, both C. reflexa (32- pm cell length) and S. oblonga ( 155 -pm body length) used in the experiments were derived from cultures in exponential growth phase in which the dry weights per single individual were 6 x 1 Op4 and 3.1 x 1 Oe2 pg. Diaptomus pallidus adults were collected from either Whiteacre Pond or Lake Nockamixon (40 28 N, W) and prefed for at least 2 days on high densities (1,800-6,000 pg dry wt liter-l) of C. reflexa. Except where noted, all experiments were performed in modified MBL medium in the dark at 20 C and started at the same time of day. None of the experimental organisms showed any adverse reactions to the MBL medium. The potential ingestion rates of D. pallidus on plant vs. animal food were examined in a series of functional response experiments with C. reflexa and S. oblonga. Eight bottles ( ml volume) were filled with a given density of food organisms and two adult male and two adult female D. pallidus were added to each of four of the bottles; the remaining four served as controls. All eight bottles were rotated in the dark at 1 rpm for 6 h. Differences in prey density between the experimental and control bottles at the end of the incubation were used to calculate ingestion and clearance rates: WC, - we, I = TN V(ln Wet - In W,,) F= TN where Z is the ingestion rate (pg dry wt copepod-l d-l), F is the clearance rate (ml copepod-l d-l), W,, and W,, are the prey densities (pg dry wt vessel-l) in the control and experimental vessels at the end of the incubation, T is the duration of the experiment (days), and N is the number of copepods per experimental vessel. Rates are expressed as the means and standard errors of the four experimental replicates. The following food concentrations were tested for each food separately: C. reflexa, 28.0, 442, and 5,874 pg dry wt liter-l; S. oblonga, 2.5, 11.5, 19.7, and 43.8 pg dry wt liter-. All food densities for all experiments are expressed as the average densities to which the copepods were exposed during the experiment. They account for prey depletion during the experiment and were computed with the following equation (after Landry 1981): -& wet- we0 Wn WC, - ln Wee) where d is the mean food density (pg dry wt liter-l), W,, and W,, are the final and initial food densities (pg dry wt vessel- ) in the experimental vessels, and V is the volume of the experimental vessel (liters). This equation assumes a constant exponential rate of decrease in food in the experimental vessels during the experiment. To cover a wide range of food densities, we had to use experimental vessels (narrownecked Wheaton bottles) which ranged in volume from 66 to 315 ml. We therefore examined the effect of bottle size on the feeding rate of D. pallidus. The design was

3 Diaptomus predation on rotifers 395 identical to that described above for the functional response experiments except that prey and predator densities were kept constant (1,000 prey ml-l, predators liter-l) and bottles of 66 and 315 ml were used. There were no significant differences (P = 1.00, Kruskal-Wallis nonparametric rank test) between ingestion rates in the two. Once we determined that D. pallidus did feed on rotifers when they were the only food offered, we examined feeding rates on rotifers in the presence of low and moderate densities of algae. The design was the same except that the density of S. oblonga at the start of the experiment was 12.6 pg dry wt liter-l in all three treatments, while the starting densities of C. reflexa were 0, 60, and 600,ug dry wt liter-. Ingestion rates of D. pallidus on rotifers in natural assemblages of plankton were determined by filling twelve 71 O-ml Wheaton bottles with whole water samples collected from Whiteacre Pond on 17 June Adult female copepods were collected from Whiteacre Pond on the same day with a 363-pm-mesh net and 20 were added to each of six bottles. All 12 bottles were rotated at 1 rpm under room light conditions for 48 h at 26 C. The contents of each bottle were then strained through 20-pm mesh and placed in 10% Formalin. Differences in the density of each prey species in the control and experimental bottles at the end of the experiment were used to calculate ingestion and clearance rates and mean prey densities with the equations given above. Both attached and unattached eggs were counted for the two rotifers which carry their eggs externally [KerateZZa cochlearis (Gosse) and Polyarthra remata (Skorikow)]. The eggs of S. oblonga were not counted, due to their low abundance and to the difficulty of identifying the eggs which are carried internally and then released. The selectivity index of Vanderploeg and Scavia (1979a,b) was used to quantify se- lective feeding by D. pallidus on the three abundant species of rotifers. This index (E*) ranges from - 1.O (least preferred prey) to + 1.O(most preferred prey); a value of zero indicates no preference. The null hypothesis that feeding is in proportion to the availability of each prey species in the experi- mental vessels was tested with the x2 statistic (Sokal and Rohlf 1981; Lechowicz 1982; Pearre 1982) with the equation: x2i = (4 ;;4)* I where 1i is the mean ingestion rate (prey predator-l d-l) of D. pallidus on prey species i, 4 is the mean ingestion rate of D. pallidus on all prey species other than i, It is the sum of Ii plus rj, and r is the ratio of the mean proportion of individuals of species i to the mean proportion of individuals of all other species available to D. pallidus. The mean proportion of each species available was derived from values of d The next question to be answered was whether D. pallidus can use ingested rotifer biomass to enhance its own survival and reproduction when algal food densities are low. Williamson et al. (1985) found that the threshold food concentration, above which D. pallidus produced eggs and below which it did not, was around 132 pg dry wt liter- of C. reflexa. We chose a concentration of,120 pg dry wt liter-l of C. reflexa as a low food level at which the reproductive rate of D. pallidus would be limited. Mature D. pallidus adults were fed 120 hg dry wt liter-l of C. reflexa with and without daily additions of rotifers and with rotifers as the sole food source. The survival and reproduction of D. pallidus in all three treatments were measured. Adult copepods were collected from Lake Nockamixon on 20 May 1985 and prefed for 9 days on 6,000 hg dry wt liter-l of C. reflexa at 20 C. This generated a population that produced only resting eggs and had a nutritionally more uniform history than freshly collected animals. Two males and two ovigerous, nongravid females were placed in each of fifteen 50-ml beakers containing 30 ml of MBL medium with the appropriate food density. A flow-through system (Williamson et al. 1985) with a turnover rate of five times the experimental chamber volume per day was used which maintains relatively constant, uniform concentrations of algal food at this flow rate. In the treatment with no algal food the flow served to constantly provide fresh medium. Rotifers (S. oblonga) were added in

4 250, 01 Williamson. DIAPTOMUS ON CRYPTOMONAS. Dn ON SYNCHAETA o MESOCYCLOPS ON BRACHIONUS IO IO2 I03 I04 FOOD CONCENTRATION (vg DRY WT liter-l) Fig. 1. Ingestion (A) and clearance (B) rates for Diaptomus pallidus on algae (Ctyptomonas reflexa) and rotifers (Synchaeta oblonga) and for Mesocyclops edax on rotifers (Brachionus calyciflorus). All values are expressed as means + SE (N = 4 replicates per mean for Diaptomus, 3 for Mesocyclops). Points without error bars indicate SE smaller than the symbol itself..(mesocyclops data from Williamson 1984.) batches of 400 per beaker per day to two of the three treatments. The 48-pm mesh between the experimental beakers and the overflow vessels retained all of the copepods but not all of the rotifers. However, smaller mesh sizes tended to clog. At the end of 24 h, we counted the number of rotifers in the overflow vessels and in the beakers and es- timated the ingestion rate of D. pallidus in the treatment with rotifers alone. If we assume that all attrition was due to ingestion, D. pallidus ingested 50 rotifers copepod- d-l (1.6 pg dry wt copepod-l d-l). Actual ingestion rates were probably somewhat lower than this due to natural mortality of the rotifers. Data from the functional response of D. pallidus feeding on C. reflexa and Butler (Fig. 1 B) were used to estimate the ingestion rate of the copepods on algae alone: this estimate was 2.2 pg dry wt copepod-l d-l. The ingestion rates of the copepods in the rotifers plus algae are not known and could not be estimated because S. oblonga also feeds readily on C. reflexa. The nature of the functional response curves (Fig. 1 A, B) and the relatively low ingestion rates on rotifers only and algae only indicate that ingestion rates of D. pallidus in the combined treatment were substantially higher than in either of the other two. This omnivory experiment was run for 2 weeks at 20 C in the dark. Adult survivorship and the number of eggs released per female were scored in each beaker every other day. Adult copepods were scored as dead and removed if they had stopped actively filtering. The abundance of rotifers in Whiteacre Pond during summer 1983 was estimated so that we could assess the potential importance of rotifers in the diet of D. pallidus in nature. Replicate samples were taken weekly from 16 June to 24 August at a depth of 0.5 m, or both 0.25 and 0.75 m, with a modified Schindler trap equipped with a 48- pm-mesh net, preserved in 10% Formalin, and counted in a Sedgwick-Rafter chamber. Results The ingestion rates of D. pallidus increased with increasing food concentration for both C. reflexa and S. oblonga over the range of food densities tested (Fig. 1A). At similar food concentrations both ingestion and clearance rates were substantially higher on the rotifers than on the algae (Fig. 1 A, B). For instance, at a food concentration of 28.4 pg dry wt liter-, ingestion rates were 6.2 times greater on Synchaeta (3.35 hg dry wt copepod-l d-l) than on Cryptomonas (0.54 pg), and clearance rates were 5.5 times greater on Synchaeta (126 ml copepod- d-l) than on Cryptomonas (23 ml). The ingestion and clearance rates of Diaptomus on Synchaeta decreased with an increase in the concentration of alternative food sources. At a rotifer concentration of pg dry wt liter- of Synchaeta, ingestion rates on the rotifers decreased from 1.61 to 0.37 pg dry wt copepod- d-l as the

5 Diaptomus predation on rotifers Table 1. Diaptomus pallidus feeding rates on the rotifer Synchaeta oblonga at various densities of the cryptomonad Cryptomonas reflexa. Mean ingestion kg dry wt copepod-l d-l) and clearance (ml copepod- d-l) rates and their standard errors are calculated from N = 4 replicates. Cryptomonas Synchaeta (pg dry wt liter- ) Ingestion rate Clearance rate f tO f Cryptomonas concentration increased from 0 to 600 pg dry wt liter-l (Table 1). Additionally, feeding rates of Diaptomus feeding on Synchaeta in natural prey assemblages were lower than those for Diaptomus feeding on Synchaeta alone (Table 2, Fig. 1). If we can assume that the dry weight of S. oblonga is similar in nature to what it is in culture (3.1 x 1O-2 I,cg individual- ), then at a Synchaeta density of 4.8 pg dry wt liter-, ingestion and clearance rates on this rotifer are 1.O hg dry wt copepod- d-l and 153 ml copepod- d-l in the absence of other prey (Fig. l), but 0.25 pg and 58 ml in natural prey assemblages (Table 2). Feeding by Diaptomus on different rotifer species is a selective process (Table 2). In the presence of three sympatric rotifers, Diaptomus showed a distinct positive selection for S. oblonga (E* := +0.40), negative selection for P. remata (E* = -0.21), and a strong negative selection for K. coch- Zearis (E* = - 1.OO)(Table 2). Although Keratella was not eaten in this experiment, its eggs were ingested, as were those of PoZyarthra (Table 2). Diaptomus is apparently able to utilize ingested rotifer biomass to enhance both its survival and its reproduction when algal food densities are low (Fig. 2). At threshold densities of algae, 40% of the adult Diaptomus survived for 2 weeks; 95% survived when rotifers were added to this (Fig. 2A). Similarly, egg production virtually stopped after 4 days in the threshold algal treatment, while sustained egg production continued throughout the experimental period when rotifers were added (Fig. 2B). The survival and reproduction of Diaptomus fed only rotifers were similar to those observed for Diaptomus fed only algae (Fig. 2). The high levels of survival and reproduction over the first few days of the experiment are due to the high preexperimental food densities. The mean rotifer densities calculated from our weekly sampling of Whiteacre Pond (Table 3) suggest that the densities of the small soft-bodied rotifer species vulnerable to ingestion by D. pallidus are high enough to account for substantial enhancement of the diet of this copepod in nature. The extreme fluctuations in the flow-through rate of Whiteacre Pond with rainfall precludes any meaningful assessment of the impact of D. pallidus on rotifer community structure through correlation analyses. However, the mean density of adult D. pallidus in Whiteacre Pond over the 11 -week sampling period was 17 individuals liter-l (max = 68 individuals liter-*). At clearance rates of up to 58 ml copepod-l d-l on S. oblonga (Table Table 2. Diaptomus pallidus feeding rates on natural assemblages of rotifers expressed as ingestion rate (I = prey copepod-l d-l; P(H) = probability that differences observed between experimental and control treatments are due to chance as determined by Kruskal-Wallis nonparametric rank test), clearance rate (F = ml copepod-* d-l), and selectivity [E* of Vanderploeg and Scavia (1979a,b) ranging from (least preferred prey) to (most preferred prey) where 0.0 indicates no preference]. Feeding was significantly nonrandom on all prey species (P -=c 0.001, x2 test); SE calculated from N = 6 replicates. Synchaeta Polyarthra Prey P. remata eggs Keratella oblonga remata cochlearis K. cochlearis eggs Nauplii Other zooplankton No. liter-l 1.54x x lo x x lo3 1.37x x x 102 Ingestion rate I I!I SE P(Ir) F+ SE E* <O.OOl 58.Ok Ok2.30 co * lo.of0.9 <O.OOl 22.Ok * f

6 398 Williamson and Butler o CRYPTOMONAS GYNCHAETA * CRY PTOMONAS AND SY NCHAETA A 100 B 100 i- 90 > A 70 T ot 70 a 2 60 v) 60 > 2 a 50 i? w 5o s 40 g c 30 a 20 g 20 IO 5 IO I IO I2 I4 SACS DAY DAY Fig. 2. Percent survival (A) and cumulative egg production (B) for adult Diaptomus pallidus fed rotifers (Synchaeta oblonga, I = 1.6 pg copepod- d-l), cryptomonads (Cryptomonas reflexa, I = 2.2 pg copepod-l d-l), or a combination of both foods at densities eauivalent to the sum of the individual treatments. Sacs--eggs carried by females at the end of the experiment: 2), D. pallidus at average density could remove all of the S. oblonga from a volume of water equivalent to the total volume of the pond in a single day. Clearance rates on other rotifer species may be much lower (Table 2), with such species being correspondingly less vulnerable to ingestion by D. pallidus. Discussion Diaptomus pallidus is a small (- 1 mmmetasome length) suspension-feeding calanoid capable of ingesting particles as small as 0.5- X 2.0-pm bacteria (Friedman 1980). Predatory behavior has not been examined in this or other small diaptomids as they are commonly believed to be obligate herbivores. The predation of D. pallidus on rotifers reported here thus has important implications for the trophic dynamics of freshwater zooplankton communities. Although D. pallidus is commonly considered an herbivore, its feeding rates (clearance and ingestion) were five to six times greater on rotifers than on algae (Fig. 1 A, B). It is unlikely that this is due to C. reflexa being a suboptimal algal food for D. pallidus because the clearance rates reported here (23 ml copepod-l d-l) are comparable to the maximum rates previously observed for similar sized diaptomids feeding on algae (2 1.6 ml copepod-l d- for Diaptomus oregonensis feeding on natural phytoplankton: Richman et al. 1980; 19 ml copepod-l d-l for Diaptomus minutus feeding on Cryptomonas sp.: Bogdan and Gilbert 1984; 25 ml copepod-l d-l for Diaptomus sicilis feeding on 12-pm-diam Chlamydomonas sp.: Vanderploeg et al. 1984). The higher feeding rates on rotifers are probably the result of the ability of the copepods to detect, attack, and capture the rotifers at a greater distance than the algae. In extensive behavioral observations on the interactions between D. pallidus and several rotifer species (Williamson in prep.), the copepods actively orient and even pounce short distances to capture rotifers which become entrained in their incoming feeding currents. These responses are elicited when a rotifer touches the first antennae of the copepod or even before any physical contact is made. No such active orientation or pouncing responses of the copepods are apparent when

7 Diaptomus predation on rotifers 399 they are feeding on algae. This suggests a much greater reactive volume and consequently an increase in the clearance rate for D. pallidus when feeding on rotifers. These observations are consistent with the widely recognized ability of calanoids to use two different feeding mechanisms for capturing large and small food particles (Conover 1966; Gauld 1966; Landry 1980; Strickler 1982; Price et al. 1983). The predation rates of D. pallidus on rotifers reported here are similar to maximum rates for the more predatory cyclopoid copepods preying on rotifers. Cyclopoids prey on several species of rotifers, with the maximum rates observed for Mesocyclops edax preying on Brachionus cazycijlorus (Williamson 1983, 1984). For comparative purposes the functional response data for M. edax preying on B. CaZyczjZorus (Williamson 1984) were converted to a dry weight basis (B. calyciflorus dry wt = 0.2 pg individual-l: Stemberger and Gilbert in press) and plot- ted with the D. pallidus data (Fig. 1). The result is a remarkably close correspondence between both magnitude and shape of these curves. Ingestion rates for both copepods on rotifers increase in a similar manner with food concentration over the range of food densities tested, while clearance rates tend to decrease as food concentration increases. The apparent reduction in clearance rate for D. pallidus at the lowest food concentration tested (Fig. 1B) suggests a threshold feeding response (sensu Frost 1975) and has implications for optimal foraging theory (Lam and Frost 1976; Lehman 1976). However, the reduction is not significant (P = 0.087, Kruskal-Wallis nonparametric rank test). To further examine whether this reduced feeding rate was real, we added an experiment with D. pallidus at food concentrations of 1.O and 11.3 pg dry wt liter-l of S. oblonga. The design was the same as before except that five rather than four replicates were used, spring water was used as the medium, and the temperature was kept at 25 C. The clearance rates were 180 (SE = 46) and 112 (SE = 22) ml copepod-l d-l for the low and high food concentrations. This indicates the lack of a threshold feeding re- sponse in D. pallidus feeding on S. oblonga over the range of prey densities studied and Table 3. Rotifer densities in Whiteacre Pond over an 11 -week period during summer The high mean density of Filinia was due exclusively to a very dense bloom of this rotifer the week of 20 July. This rotifer was generally present at densities < 100 liter-. Polyarthra Synchaeta Keratella Keratella Filinia major) Brachionus tremula, pectinata) terminalis spp. (remata, spp. (oblonga, stylata, cochlearis americana angularis Mean Maximum (No. liter- ) 2.32 x lo2 1.96x lo x lo x lo x X X x lo x lo3 1.24~ lo2 5.02x lo x lo2 is consistent with the absence of a threshold response in Eudiaptomus gracilis feeding on algae at very low densities (Muck and Lampert 1980). Diaptomus pallidus can utilize rotifer biomass to enhance its own survival and reproduction. Furthermore, the ingestion rates at which survival and reproduction were enhanced in the laboratory (50 rotifers copepod- d-l) are comparable to those for D. pallidus feeding on natural assemblages of rotifers (50 rotifers and 18 rotifer eggs copepod-l d-l: Table 2). The importance of food in limiting the growth and reproduction of herbivorous zooplankton in freshwater ecosystems (Neil1 1978; Kerfoot and DeMott 1980; Lampert and Schober 1980) and the higher assimilation of animal than plant food for some suspension-feeding calanoids (Corner et al. 1976) suggest that ingestion of rotifers by diaptomid copepods may form an important trophic link in aquatic food webs. Diaptomus pallidus preys selectively on different rotifer species. This was shown in our experiments with natural prey assemblages in which S. oblonga was preferred to P. remata, and K. cochlearis was not consumed at all (Table 2). The defenseless S. oblonga is apparently very vulnerable to Calanoid predation, while the hard lorica and spines of K. cochlearis and, to a lesser extent, the jumping escape response of P. remata appear to be effective in reducing predation rates. This pattern is similar to that observed for the cyclopoid M. edax preying on Brachionus, Polyarthra, and Ke-

8 400 Williamson and Butler ratella (Gilbert and Williamson 1978; Williamson and Gilbert 1980). Two previous field studies support our findings that Keratella and other spined, loricate rotifers are not vulnerable to diaptomid predation. Porter et al. (1979) noted no significant reduction in the densities of Keratella or Kellicottia in enclosures with enriched crustacean zooplankton (Daphnia galeata mendotae, D. minutus, Cyclops scutifer, and Epischura lacustris) from those in which the crustaceans were eliminated, although algae and ciliate densities were significantly reduced. Neil1 (1984) noted no significant differences in densities of K. cochlearis or Kellicottia longispina between enclosures from which Diaptomus kenai and Diaptomus leptopus had been removed and those in which these diaptomids were present. Some of the larger, more predatory diaptomids may consume the spiny, loricate rotifers, but the limited information available suggests that they too prefer small, softbodied rotifers. Keratella cochlearis loricas have been observed in the guts of late copepodite and adult Diaptomus shoshone (Anderson 1967; Maly and Maly 1974), while Diaptomus arcticus feeds on K. longispina, but prefers S. oblonga (Anderson 1970). The selective nature of this predation by diaptomids on rotifers suggests that these calanoids may have an impact on rotifer community structure in nature. Some preliminary experiments with other Diaptomus spp. also support the general importance of rotifers in the diet of other suspension-feeding diaptomids. Adult Diaptomus spatulocrenatus (total body length 1.5 mm) females were collected from nature and placed in 5 ml of filtered lake water in 5-cm plastic Petri dishes. Each experimental dish contained two copepods and 20 individuals of a potential prey species. Two control dishes with identical numbers of prey but no copepods were incubated with two experimental dishes under constant light at 13 C for 10 h. Potential prey species included Kellicottia bostoniensis, K. cochlearis, Polyarthra sp., Synchaeta pectinata, and copepod nauplii, all freshly collected from the same lake as the copepods, and Brachionus calyciflorus from a laboratory culture. All 20 of the Brachionus, 6-l 0 of the Poly- arthra, 2-4 of the Kellicottia, l-4 of the Synchaeta, and none of the Keratella or nauplii were consumed by the D. spatulocrenatus. In similar experiments with Diaptomus birgei (metasome length 1.4 mm for females, 1.1 mm for males) for 12 h in the dark at 25 C, adult females consumed all 20 S. oblonga offered while adult males consumed 14 of the 20 S. oblonga. The freshwater calanoids might be compared most appropriately with the marine calanoids, where most species are omnivores with tendencies toward herbivory or carnivory (Anraku and Omori 1963; Paffenhijfer and Knowles 1980). Along this gradient, small suspension feeders like D. pal- Zidus, which exhibit a distinct reduction in their predation rates as algal density increases (Table I), would be classified as omnivores with a tendency toward herbivory, while calanoids like E. Zacustris, whose predation rates do not change in the presence of algal food (Wong 198 l), would be classified as being more carnivorous. The ability of calanoids to consume both plant and animal food would seem to be most adaptive, given the frequent and often intense seasonal fluctuations in phytoplankton and zooplankton population densities characteristic of freshwater environments. One of the more interesting aspects of the trophic interactions between suspensionfeeding diaptomids and rotifers is that they may be competitive interactions as well as predator-prey interactions. We have successfully cultured both D. pallidus and S. oblonga through several generations with C. reflexa as the sole food source. These two species clearly have the potential to exploit the same types of food resources in nature. Thus, the predation of D. pallidus on S. oblonga might be viewed as a mechanism of direct interference competition. However, more data on resource availability as well as the spatial and temporal patterns of utilization of these resources in nature are necessary before competition can be demonstrated in these interactions. References ALCARAZ, M.,G.-A. PAFFENH~FER,AND J.R. STRICK- LER Catching the algae: A first account of visual observations on filter-feeding calanoids. Am.

9 Diaptomus predation on rotifers 401 Sot. Limnol. Oceanogr. Spec. Symp. 3: New England. ANDERSON, R. S Diaptomid copepods from two mountain ponds in Alberta. Can. J. Zool. 45: Predator-prey relationships and predation rates for crustacean zooplankters from some lakes in western Canada. Can. J. Zool. 48: ANRAKU, M., AND M. OMORI Preliminary survey of the relationship between the feeding habit and the structure of the mouth-parts of marine copepods. Limnol. Oceanogr. 8: BOGDAN, K. G., AND J. J. GILBERT Seasonal patterns of feeding by natural populations of Keratella, Polyarthra, and Bosmina: Clearance rates, selectivities, and contributions to community grazing. Limnol. Oceanogr. 27: ,AND Body size and food size in freshwater zooplankton. Proc. Natl. Acad. Sci. 81: BRANDL,Z.,ANDC. H. FERNANDO Theimpact of predation by the copepod Mesocyclops edax (Forbes) on zooplankton in three lakes in Ontario, Canada. Can. J. Zool. 57: CONFER, J. L., AND P. I. BLADES Omnivorous zooplankton and planktivorous fish. Limnol. Oceanogr. 20: 57 l-579. CONOVER, R. J Feeding on large particles by Calanus hyperboreus (Krbyer), p. 187-l 94. In H. Barnes [ed.], Some contemporary studies in marine science. Allen and Unwin. CORNER, E. D.,R.N. HEAD,~. C. KILVINGTON, AND L. PENNYCUICK On the nutrition and metabolism of zooplankton. 10. Quantitative aspects of Calanus helgolandicus feeding as a carnivore. J. Mar. Biol. Assoc. U.K. 56: FRIEDMAN, M. M Comparative morphology and functional significance of copepod receptors and oral structures. Am. Sot. Limnol. Oceanogr. Spec. Symp. 3: 185-l 97. New England. FROST, B. W A threshold feeding behavior in CalanuspaciJicus. Limnol. Oceanogr. 20: GAULD, D. T The swimming and feeding of planktonic copepods, p In H. Barnes ted.], Some contemporary studies in marine science. Allen and Unwin. GILBERT, J. J., AND C. E. WILLIAMSON Predator-prey behavior and its effect on rotifer survival in associations of Mesocyclops edax, Asplanchna girodi, Polyarthra vulgaris, and Keratella cochlearis. Oecologia 37: GRYGIEREK, E Some data on the role of food in the biology of Eudiaptomus zachariasi Poppe. Ekol. Pol. 19: HAIRSTON, N. G., JR The adaptive significance of color polymorphism in two species of Diaptomus (Copepoda). Limnol. Oceanogr. 24: KERFOOT, W.C., AND W.R. DEMOTT Foundations for evaluating community interactions: The use of enclosures to investigate coexistence of Daphnia and Bosmina. Am. Sot. Limnol. Oceanogr. Spec. Symp. 3: New England. KOEHL, M. A., AND J. R. STRICKLER Copepod feeding currents: Food capture at low Reynolds number. Limnol. Oceanogr. 26: 1062-l 073. LAM, R. K., AND B. W. FROST Model of copepod filtering response to changes in size and concentration of food. Limnol. Oceanogr. 21: LAMPERT, W., AND U. SCHOBER The importance of threshold food concentrations. Am. Sot. Limnol. Oceanogr. Spec. Symp. 3: New England. LANDRY, M. R Detection of prey by Calanus paczjicus: Implications of the first antennae. Limnol. Oceanogr. 25: Switching between herbivory and carnivory by the planktonic marine copepod Calanus paczjkus. Mar. Biol. 65: LECHOWICZ, M. J The sampling characteristics of electivity indices. Oecologia 52: LEHMAN, J. J The filter-feeder as an optimal forager, and the predicted shapes of feeding curves. Limnol. Oceanogr. 21: 50 l MAKAREWICZ, J. C., AND G. E. LIKENS Structure and function of the zooplankton community of Mirror Lake, New Hampshire. Ecol. Monogr. 49: MALY, E. J., AND M. P. MALY Dietary differences between two co-occurring Calanoid copepod species. Oecologia 17: MUCK, P., AND W. LAMPERT Feeding of freshwater filter-feeders at very low food concentrations: Poor evidence for threshold feeding and optimal foraging in Daphnia longispina and Eudiaptomus gracilis. J. Plankton Res. 2: MULLIN, M. M Selective feeding by Calanoid copepods from the Indian Ocean, p In H. Barnes [ed.], Some contemporary studies in marine science. Allen and Unwin. NEILL, W. E Experimental studies on factors limiting colonization by Daphnia pulex Leydig of coastal montane lakes in British Columbia. Can. J. Zool. 56: Regulation of rotifer densities by crustacean zooplankton in an oligotrophic montane lake in British Columbia. Oecologia 61: 175-l 81. PAFFENH~FER, G.-A., ANDS.C. KNOWLES Omnivorousness in marine planktonic copepods. J. Plankton Res. 2: , J. R. STRICKLER, AND M. ALCARAZ Suspension-feeding by herbivorous Calanoid copepods: A cinematographic study. Mar. Biol. 67: PEARRE, S., JR Estimating prey preference by predators: Uses of various indices and a proposal of another based on x2. Can. J. Fish. Aquat. Sci. 39: PORTER, K. G The plant-animal interface in freshwater ecosystems. Am. Sci. 65: 159-l 70. -, M. L. PACE, AND J. F. BATTEY Ciliate protozoans as links in freshwater planktonic food chains. Nature 277: PRICE, H.J.,G.-A. PAFFENH~FER,AND J.R. STRICKLER Modes of cell capture in Calanoid copepods. Limnol. Oceanogr. 28: 116-l 23. RICHMAN,S.,S. A. BOHON, ANDS. E. ROBBINS

10 402 Williamson and Butler Grazing interactions among freshwater Calanoid copepods. Am. Sot. Limnol. Oceanogr. Spec. Symp. 3: New England. ROBERTSON, S. B., AND B. W. FROST Feeding by an omnivorous planktonic copepod A&&us divergens Bradford. J. Exp. Mar. Biol. Ecol. 29: 23 l-244. SOKAL, R. R., AND F. J. ROHLF Biometry, 2nd ed. Freeman. STEMBERGER, R. S A general approach to the culture of planktonic rotifers. Can. J. Fish. Aquat. Sci. 38: AND J. J. GILBERT. In press. Body size, food concentration and population growth in planktonic rotifers. Ecology. STRICKLER, J. R Calanoid copepods, feeding currents and the role of gravity. Science 218: VANDERPLOEG, H. A., AND D. SCAVIA. 1979a. Calculation and use of selectivity coefficients of feeding: Zooplankton grazing. Ecol. Modeling 7: ,AND b. Two electivity indices for feeding with special reference to zooplankton grazing. J. Fish. Res. Bd. Can. 36: , AND J. R. LIEBIG Feeding rate of Diaptomus sicilis and its relation to selectivity and effective food concentration in algal mixtures and in Lake Michigan. J. Plankton Res. 6: WILLIAMSON, C. E Invertebrate predation on planktonic rotifers. Hydrobiologia 104: Laboratory and field experiments on the feeding ecology of the cyclopoid copepod, Mesocyclops edax. Freshwater Biol. 14: ,N. M. BUTLER,AND L. FORCINA Food limitation in naupliar and adult Diaptomus pallidus. Limnol. Oceanogr. 30: 1283-l , AND J. J. GILBERT Variation among zooplankton predators: The potential of Asplanchna, Mesocyclops, and CycIops to attack, capture, and eat various rotifer prey. Am. Sot. Limnol. Oceanogr. Spec. Symp. 3: New England. WONG, C. K Predatory feeding behavior of Epischura Iacustris (Copepoda, Calanoida) and prey defense. Can. J. Fish. Aquat. Sci. 38: Submitted: 2 April 1985 Accepted: 3 October 1985