Measuring Patch Assessment Abilities of Desert Granivores

We develop two criteria for measuring patch assessment ability. First, we examine the ability of foragers to equalize benefits and costs at manipulated resource patches. Second, we compare patch utilization patterns of four possible foraging strategies (prescient, fixed time, Bayesian, and rate assessor) with actual foraging patterns. Experiments with several desert rodent and avian species suggest that Merriam's kangaroo rat may obtain the best estimate of patch quality, followed by the round—tail ground squirrel and Arizona pocket mouse. Kangaroo rats exhibited both a prescient and Bayesian strategy. Pocket mice and ground squirrels exhibited both a fixed—time and Bayesian strategy. Gambel's Quail appeared to be the least sophisticated forager and exhibited only a fixed—time strategy. The fixed—time strategy was observed most frequently in the low variance environment where patch differences were relatively minor. In general, increased patch variation led to poorer patch estimates but allowed employment of sophisticated foraging strategies. Avian group foragers did not obtain better estimates of patch quality than solitary foragers.