TROPHIC INTERACTIONS IN TEMPERATE LAKE ECOSYSTEMS - A TEST OF FOOD-CHAIN THEORY

In a comparative study covering 11 temperate lake ecosystems of low to intermediate productivity, we tested the predictions of the food chain model developed by L. Oksanen et al. The pelagic habitat of the most unproductive lakes lacked secondary carnivores (piscivores), whereas secondary carnivores inhabited the more productive lakes in agreement with the assumption that food chain length is limited by the productivity of the system. The observed relationships between planktivores (primary carnivores), zooplankton (herbivores), and phytoplankton (primary producers) were in agreement with the predictions of the Oksanen et al. model for three- and four-trophic-link systems. Piscivore relationships were significant only when the biomass of piscivorous perch only was used as the independent variable (piscivorous perch amounted to greater-than-or-equal-to 97% of total piscivore biomass in five out of six four-link systems). The results suggest that gross patterns of food web dynamics in temperate lakes may be collapsed into food chain dynamics. However, as the higher trophic levels of most lakes were dominated by the same two species, an alternative hypothesis is that the simplicity of the food webs studied allowed food chain dynamics to approximate food web dynamics. The food chain model tested gave accurate predictions of trophic-level relationships within the rather narrow productivity range studied. However, previous data suggest that, due to size-structured interactions, secondary carnivores (piscivores) play a minor role in highly productive systems, a circumstance that may explain the postulated increase in mortality rate of zooplankton with increasing productivity. The observed increase in number of trophic levels with productivity does not necessarily imply a causal relationship between productivity and food chain length, as lake productivity is also correlated with structural complexity (in the form of submerged vegetation and an oxygenated hypolimnion). Habitat heterogeneity may thus influence food chain length, and in general it is likely that food chain length will depend on more than one environmental variable.