INDEPENDENT VARIABLES FOR PREDICTING INTAKE RATE OF MAMMALIAN HERBIVORES - BIOMASS DENSITY, PLANT-DENSITY, OR BITE SIZE

Contemporary models of herbivore functional response are typically dri ven by one of three potential measures of forage availability: biomass density (g/m2), prey (plant, tiller, ramet) density (no./m2), or bite size (g/bite). The choice between these measures is not always obviou s, but can have a large influence on the shape of the functional respo nse. Furthermore, under most conditions two or more of these variables are confounded, hampering statistical tests to separate their effects . We designed a set of two, parallel, factorial experiments to separat e the effects of plant density, biomass density, and bite size on inta ke rates of lemmings. Plant density was varied by changing the distanc e between plants, and biomass density was varied by changing both plan t density and plant size. Changes in bite size were imposed on lemming s by offering them plants that varied in height and mass. In our first experiment, we crossed bite size with biomass density, and in the sec ond, we crossed bite size with plant density. Biomass and plant densit y had no effect on intake rates (p > 0.2), but we observed 2-3 fold in creases in intake rate in response to increased bite size (p < 0.0001) . Intake rates we observed corresponded closely with those predicted b y a model of functional response driven by bite size alone, indicating that the time needed to process a bite in the mouth exceeded that nec essary to travel between plants, even though biomass densities in our feeding trials were much lower than those typical of natural habitats. Our results emphasize that the choice of an independent variable can have a profound influence on the interpretation of studies of herbivor e functional response.