Stoichiometry of fishes and their prey: implications for nutrient recycling

Ecology, Sept, 1997 by Daniel E. Schindler, Lisa A. Eby

INTRODUCTION

Integrating the approaches of population/community ecology and ecosystem ecology remains an important but difficult goal of current ecological research (Jones and Lawton 1995). Studies of ecosystems have traditionally evaluated material and energy flows, while population and community ecology have evaluated ecological systems largely within the context of natural selection and phylogeny (Wedin 1995). In many ecosystems the biota represent major storage pools of the nutrients that limit system productivity, and consumerfood interactions often regulate the cycling and distribution of essential nutrients in various ecosystem components (DeAngelis 1992). Consumer-food interactions can therefore be evaluated in terms of the flows of essential nutrients between ecosystem components that must conform to the constraints of mass balance and stoichiometry (Sterner 1995). Evaluating material flows (e.g., of limiting nutrients) between ecosystem components that result from food web interactions represents a first step towards melding the approaches of population/community and ecosystem ecology (DeAngelis 1992, Jones and Lawton 1995).

Phytoplankton productivity in lakes is generally limited by the availability of phosphorus (P; Schindler 1977). At seasonal scales, availability of nitrogen (N) can also limit phytoplankton growth, either alone or as a colimiting nutrient with P (e.g., Goldman 1972, Vincent et al. 1984, Elser et al. 1988). Thus, understanding the biogeochemical cycles of both N and P is important to understanding lake ecosystem dynamics.

In lakes, the biota are both major storage pools of N and P, and critical sources of these nutrients needed to support algal production (Barlow and Bishop 1965, Kitchell et al. 1979, Lehman 1980, Carpenter et al. 1992a). Changes in food web structure can alter patterns of nutrient storage and recycling in lake ecosystems (Boers et al. 1991, Carpenter et al. 1992b, Schindler et al. 1993). Because the relative supply rates of N and P have important implications for resource-based phytoplankton competition (Tilman et al. 1982), alterations of lake food webs that lead to changes in the stoichiometry of nutrient recycling have direct implications for phytoplankton community composition (Elser et al. 1988, Sterner 1989) and water quality (Smith 1983).

Fishes have important effects on lentic nutrient cycles because they alter zooplankton communities through size-selective predation (Brooks and Dodson 1965), which, in turn, affects the rates of N and P recycling by this important component of pelagic food webs (Peters and Rigler 1973, Elser et al. 1988). Fish also excrete nutrients that can be important to phytoplankton, especially in lakes where fishes feed on benthos, which results in translocation of nutrients from benthic to pelagic habitats (Lamarra 1975, Brabrand et al. 1990, Reinertsen et al. 1990, Boers et al. 1991, Carpenter et al. 1992a, b, Schindler et al. 1993, 1995, Leavitt et al. 1994, Vanni 1995). Although most studies have concentrated on P fluxes from fish excretion, N excretion is also a potentially important nutrient source from fish communities (Kraft 1992, Vanni 1995). However, our understanding of the effects of fishes on nutrient cycling in lakes remains incomplete.

Because of the implications for phytoplankton community interactions and allocation patterns of N and P among ecosystem components, a new "stoichiometric approach" (Sterner 1990, Sterner et al. 1992) has been introduced that stresses the implications of elemental ratios of nutrients recycled by consumers. This approach evaluates a consumer-food interface in terms of the relative proportions of carbon (C), N, and P in the consumer and its food. Each of these elements is provided in the food, ingested by the consumer, and accumulated at rates that maintain a homeostatic elemental balance in consumer tissue. Mass balance can be applied to estimate the ratios of excreted nutrients. This theory predicts that the stoichiometry of recycled nutrients is largely dependent on the imbalance between the homeostatic elemental composition of the consumer and its food.

A second important aspect of the "stoichiometric approach" accounts for the fact that the growth rate of a consumer can be limited by its accumulation rate of a mineral rather than of energy. If a consumer has a higher growth requirement for a mineral (e.g., P) than is provided in its food, the consumer's growth will become limited by the availability of that element. Because zooplankton are generally richer in N and P than phytoplankton (i.e., higher N:C and P:C ratios), zooplankton growth can be retarded in the presence of poor quality algae (i.e., with low N:C or P:C; Sterner and Hessen 1994). Although the generality of mineral growth limitation in zooplankton is currently debated (Brett 1993, Hessen 1993, Urabe and Watanabe 1993) it has been demonstrated in laboratory experiments (Urabe and Watanabe 1992, Sterner 1993, Sterner and Robinson 1994, Rothhaupt 1995). Mineral limitation of growth is important to both the consumer and to the environment because the N:P ratio of excreted nutrients changes dramatically when one of these two minerals becomes limiting to growth. For example, if P limits growth of a consumer, then excretion of P is low relative to N and the resulting N:P supply ratio (i.e., the ratio of excreted nutrients) is high. Alternatively, if N limits consumer growth, N excretion. Therefore, the N: P supply ratio will both be low (Sterner 1990).