Testing models of bee foraging behavior through the analysis of pollen loads and floral density data

Authored by Philippe Marchand, Alexandra N Harmon-Threatt, Ignacio Chapela

Date Published: 2015

DOI: 10.1016/j.ecolmodel.2015.06.019

Sponsors: National Science and Engineering Research Council of Canada (NSERC)

Platforms: Fortran

Model Documentation: ODD Mathematical description

Model Code URLs: http://www.sciencedirect.com.ezproxy1.lib.asu.edu/sdfe/arp/media/1-s2.0-S0304380015002689-mmc2.pdf

Abstract

The composition of social bees' corbicular pollen loads contains information about both the bees' foraging behavior and the surrounding floral landscape. There have been, however, few attempts to integrate pollen composition and floral landscape to test hypotheses about foraging behavior. Here, we present an individual-based model that generates the species composition of pollen loads given a foraging model and a spatial distribution of floral resources. We apply this model to an existing dataset of inflorescence counts and bumble bee pollen loads sampled at different field sites in California. For two out of three sites, a foraging model consisting in correlated random walks with constant preferences for each plant species provides a plausible fit for the observed distribution of pollen load content. Pollen load compositions at the third site could be explained by an extension of the model, where different preferences apply to the choice of an initial foraging patch and subsequent foraging steps. Since this model describes the expected level of pollen load differentiation due solely to the spatial clustering of conspecific plants, it provides a null hypothesis against which more complex descriptions of behavior (e.g. flower constancy) can be tested. (C) 2015 Elsevier B.V. All rights reserved.

Tags

Protocol Rules Bumblebees Approximate bayesian computation Constancy