Yvonne Allen

Interannual patterns of planktivory 1987- 1989: an analysis ofvertebrate and invertebrate planktivores

After a period of relative stasis, the crustacean zooplankton of Lake Mendota underwent a pronounced change in species composition between 1987 and 1988 (Lathrop and Carpenter, Ch. 9). This change was concordant with a dramatic die-off of cisco (Vanni et al. 1990; Rudstam et al., Ch. 12), indicating that changes in predation rates on different zooplankton groups may have been responsible for the observed shifts in zooplankton species composition. To determine the degree to which varying predation pressure resulted in changes in zooplankton abundance patterns, we examined the feeding relationships of the dominant zooplanktivores in the pelagic region of the lake and estimated the ability of these predators to regulate the dominant herbivorous zooplankton species. We were interested in determining the relative importance of various predators and in quantifying both interannual and seasonal dynamics of zooplanktivores and their prey.

Piscivores and Their Prey

Not suprisingly, Lake Mendota’s fishes have been the subject of a multitude of studies since the days of Birge and Juday. However, these studies have focused almost entirely on yellow perch (Perca flavescens), white bass (Morone chrysops), and cisco (Coregonus artedii), probably because of their abundance or importance to the fishery. When this study began, relatively little was known about Lake Mendota’s piscivore community, including abundance, reproductive success, growth rates, diet, and distribution. Hence, we had some fundamental questions that needed to be addressed before we could predict the course of the biomanipulation experiment.

An in situ test of the effects of food quality on Daphnia population growth

In years of low planktivory by fish, midsummer declines and low population abundances of Daphnia pulicaria in Lake Mendota, WI, USA have been attributed to poor food quality and low food abundance. This hypothesis has been proposed because of midsummer blooms of blue-green algae in this eutrophic urban lake. We tested the hypothesis by performing in situ food manipulations during the midsummer decline of the D. pulicaria population, and during the early autumn low population period. In July, animals held in clear plastic vials containing GF/F filtered lake water exhibited poor survivorship, ceased producing offspring after three days, and had low lipid reserves. Daphnids fed whole-lake water or filtered water enriched with Chlamydomonas survived at a significantly higher rate and produced abundant offspring. In September, there were no significant differences in survivorship among treatments, but animals fed water enriched with Chlamydomonas produced significantly more offspring than animals in the other treatments. The hypothesis of poor food quality causing the midsummer decline of the Daphnia population is not supported by our experiments. Because predation by juvenile fish does not appear to be important in regulating the midsummer Daphnia population during these years, alternate hypotheses for the midsummer decline are suggested. These include increased predation by invertebrate predators such as Leptodora kindtii, a life history shift by daphnids during midsummer to production of resting eggs, and a combination of low levels of planktivory and deteriorating feeding conditions acting together to cause the population decline.

Herbivory, Nutrients, and Phytoplankton Dynamics in Lake Mendota, 1987–89

It is becoming increasingly clear that lake plankton communities are regulated by both predation and resources. Top predators, through effects on herbivores, can regulate phytoplankton community structure, biomass, and primary productivity (Carpenter et al. 1985; Carpenter and Kitchell 1988; Vanni and Findlay 1990). Increase in potential limiting nutrients, such as nitrogen and phosphorus, can stimulate phytoplankton production and biomass, and the ratio of limiting nutrients can influence community structure (Schindler 1977; Smith 1983). Recently, lake ecologists have realized that when attempting to explain phytoplankton dynamics, both nutrients and herbivory must be considered. Often the two interactions simultaneously influence phytoplankton, although their relative strengths may vary seasonally (e.g. Sommer et al. 1986; Vanni and Temte 1990). Furthermore, resource effects and herbivore effects may interact in complex ways to influence phytoplankton (Leibold 1980; Sterner 1990; Carpenter et al. 1991)—a case of the interactions themselves interacting (Roughgarden and Diamond 1986).

Food Web Structure of Lake Mendota

Although some classic studies (Hrbacek et al. 1961; Brooks and Dodson 1965) have shown strong effects of planktivorous fish on lower trophic levels, it is only in the last decade that effects of interactions among trophic levels on lake ecosystems have come to the forefront of limnological research (Andersson 1984; Carpenter et al. 1985; McQueen et al. 1986; Northcote 1988; Gulati et al. 1990) A present challenge is to understand the interplay bewtween food web effects and nutrient loading (Benndorf 1988; Persson et al. 1988; Vadas 1989), a challenge that will require comparing experimental manipulations in lakes with varying nutrient loadings (Carpenter and Kitchell 1988; Carpenter et al. 1991). The objective of the food web manipulation in eutrophic Lake Mendota is to establish a large population of piscivorous fish (walleye and northern pike) and, through a cascade of trophic interactions, reduce the planktivorous fish, increase the herbivorous zooplankton, decrease algae, and increase water transparency (Kitchell, Ch. 1). In this chapter we describe the structure of the open-water food web in Lake Mendota and the dramatic changes in the planktivorous fish community that occured during the summer of 1987. The following chapters in this section describe and interpret the behaviour of different trophic levels during the first 3 years of the food web manipulation (1987–89).

Modeling the Lake Mendota ecosystem: Synthesisand evaluation of progress

The models presented in the three preceding chapters were planned as elements of an integrated ecosystem approach from phosphorus to fishes. The modeling problem was broken into three parts in order to maximize our rate of progress and make best use of the people involved. The modules—piscivory, planktivory, and herbivory-algae-nutrients—have fundamentally different time scales yet strong vertical interactions (Figure 22.1). Within a given nutrient and weather regime, differences in return time cause the upper modules to act as constraints on lower ones (O’Neill et al. 1986). Piscivore dynamics have return times of years (Post and Rudstam, Ch. 19). Stock and harvest policies as well as resource levels must be considered in modeling piscivory. Planktivory by fishes has return times of years, while that by the zooplankter Leptodora has return times of weeks (Luecke el at., Ch. 20). Herbivory, algal growth, and nutrient fluxes have rapid dynamics and short return times of a few days (Vanni et al., Ch. 21).