Ecological hierarchy of foraging in a large herbivore: the plains bison perspective in tallgrass prairie

Abstract

Foraging decisions by native grazers in fire-dependent landscapes reflect fire-grazing interactions. I assessed behavioral responses associated with the attraction of grazers to recently burned areas at multiple spatial scales. (a) I focused on feeding in the area between steps in a foraging bout – the feeding station – where forage quality and vegetation architecture underlie these fine-scale decisions. The ‘forage maturation hypothesis’ (FMH) predicts the distribution of large herbivores based on the temporal dynamics of forage quality and quantity, but does not address herbivore responses to inter-patch variation caused by fire-induced increases of forage quality. The ‘transient maxima hypothesis’ (TMH) also predicts variable forage quality and quantity, but in response to intermittent disturbance from fire. I described the effects of variable spring burn history to bison foraging and their spatio-temporal distribution at Konza Prairie. Forage attributes met predictions of the TMH to explain how forage maturation affects foraging behavior across watersheds with varying burn frequency. At sites burned in the spring after several years without burning, intake rate increased with increasing vegetation biomass at a greater rate during the early growing season than during the transitional mid-summer period. This foraging behavior occurred in response to a non-equilibrial pulse of high quality resource that set the stage in the burned area, and was then retained by repeated grazing over the growing season. Thus, bison responded increased forage resource availability resulting from transient maxima in infrequently-burned watersheds burned that spring and they intensely used these areas until forage availability and forage regrowth was not possible. (b) At the patch scale, bison selected areas of low-to-moderate grass cover in which to feed and avoided areas of high forb cover in the growing season. During the dormant season, however, bison selected feeding-sites with uniformly high canopy cover in watersheds that were not burned. (c) At the landscape-scale, infrequently burned watersheds (compared to watersheds that were not burned) provided the strongest significant predictor of bison space use in all early growing- and transitional-season months. (d) The probability of habitat selection was driven by availability of high foliar, protein and low-to-intermediate herbaceous biomass throughout the growing season. These results explain the hierarchy of foraging by a dominant consumer in an experimental landscape by linking two prominent hypotheses, TMH-FMH, proposed to explain spatial variation in forage quality and quantity at local and landscape scales.