Field measurements give biased estimates of functional response parameters, but help explain foraging distributions

Abstract

Mechanistic insights and predictive understanding of the spatial distributions of foragers are typically derived by fitting either field measurements on intake rates and food abundance, or observations from controlled experiments, to functional response models. It has remained unclear, however, whether and why one approach should be favoured above the other, as direct comparative studies are rare. The field measurements required to parameterize either single or multi-species functional response models are relatively easy to obtain, except at sites with low food densities and at places with high food densities, as the former will be avoided and the second will be rare. Also, in foragers facing a digestive bottleneck, intake rates (calculated over total time) will be constant over a wide range of food densities. In addition, interference effects may depress intake rates further. All of this hinders the appropriate estimation of parameters such as the 'instantaneous area of discovery' and the handling time, using a type II functional response model also known as 'Holling's disc equation'. Here we compare field- and controlled experimental measurements of intake rate as a function of food abundance in female bar-tailed godwits Limosa lapponica feeding on lugworms Arenicola marina. We show that a fit of the type II functional response model to field measurements predicts lower intake rates (about 2·5 times), longer handling times (about 4 times) and lower 'instantaneous areas of discovery' (about 30-70 times), compared with measurements from controlled experimental conditions. In agreement with the assumptions of Holling's disc equation, under controlled experimental settings both the instantaneous area of discovery and the handling time remained constant with an increase in food density. The field data, however, would lead us to conclude that although handling time remains constant, the instantaneous area of discovery decreased with increasing prey densities. This will result into highly underestimated sensory capacities when using field data. Our results demonstrate that the elucidation of the fundamental mechanisms behind prey detection and prey processing capacities of a species necessitates measurements of functional response functions under the whole range of prey densities on solitary feeding individuals, which is only possible under controlled conditions. Field measurements yield 'consistency tests' of the distributional patterns in a specific ecological context.