A field assessment of the Spalinger–Hobbs mechanistic foraging model: free-ranging moose in winter

Abstract

The mechanistic foraging models introduced by Spalinger and Hobbs in 1992 have been very influential in studies of herbivory at a variety of scales. However, almost no field study has evaluated whether the assumption regarding invariability of parameters with time holds for large herbivores with long foraging bouts, and most studies have obtained the model parameters from very short trials. We used free-ranging moose, Alces alces (L., 1758), to test this assumption of invariability and to compare intake calculated by the Spalinger–Hobbs model using parameters obtained from 10-min trials with intake calculated using data obtained from entire bouts. Our results revealed that the invariance assumption was not fully met: moose increased bite and chew rates and took smaller bites the longer a bite or chew sequence lasted, which resulted in declining intake rates. As a result, the original model misestimated intake by more than double for mountain birch (Betula pubescens ssp. czerepanovii (Orlova) Hämet-Ahti) and by up to 23% for willow (Salix spp.). Compared with data from entire foraging bouts, parameters derived from only the first 10 min of a bout overestimated intake of mountain birch by 31% and underestimated intake of willow by up to 24%. Our results suggest that for herbivores with long foraging bouts, one could modify the model to allow some parameters to vary with time but, more simply, one should parameterize the model using data from entire foraging bouts.