Abstract
The influence of pelagic herbivores on the competitive arena of their prey is examined in three models of herbivore physiology. One hallmark of pelagic herbivory is the rapid and tight cycling of nutrients among algae, herbivores, and the dissolved phase. An implication of this cycling is that the ratio of elements released by the herbivore is a "supply ratio" as defined in resource-competition theory. The nitrogen (N)-to-phosphorus (P) supply ratio depends on the assumptions made regarding grazer physiology. Three alternative models are considered: (1) a constant, mass-specific basal matabolic release of N and P; (2) a constant efficiency of accumulating N and P in grazer biomass; and (3) an anjusting efficiency of N and P accumulation that maintains a homeostatic ratio of N to P (N:P) in the bodies of the grazers. All three models predict a positive correlation between the N:P in the algal pool and that released by the animals. Previously published data provided strong supportive evidence for this relationship. Studies that have used the N-to-P concentration ratio as a proxy for the N-to-P supply ratio are therefore reinforced. The model of basal metabolism is rendered unrealistic by a lack of relationship between the N:P released by grazers and that in the zooplankton pool. The other two models differ only in whether grazers maintain a homeostasis of the N:P in their bodies. A review of the literature together with an analysis of independently gathered data indicate that homeostasis of the N:P is a reasonable constraint but that homeostasis is not strict. Constant accumulation predicts a direct proportionality between the N:P released and the N:P in the algal pool. Homeostasis predicts a curvilinearity in this relationship (though it may be only a shallow bend for some realistic parameters). The curvilinearity is such that imbalances of the N:P in the food compared with that in the zooplankton are accentuated. Comparison of the N:P released calculated from alternatives 2 and 3 with measured values showed good agreement with both models, but especially for the homeostatic one. The homeostatic model is thus the most realistic. Homeostatic grazers appear to establish an unstable equilibrium in the N:P in the algal pool and thus, presumably, in the nutrient-limitation patterns of the algae. Whichever element appears in the food in low relative proportion will be recycled by the grazer with relatively low efficiency. The model of strict homeostasis thus predicts a role for grazers not predicted by other theoretical approaches: grazers cause the nutrient-limitation patterns of their algal prey to diverge. Algal communities receiving their nutrients primarily from homeostatic grazers are expected to become limited by either N or P, but not by both. Suggestive evidence from the literature for such dynamics is cited.