Inflexible wasps : the aphid parasitoid Lysiphlebus testaceipes does not track multiple changes in habitat profitability

Bayesian updating is a form of learning by which individuals can adapt their behaviour to uncertain environments. In the particular situation of optimal foraging in a patchy habitat, animals may estimate the profitability of the patches they encounter to update their estimate of the profitability of the habitat as a whole, and adjust their foraging decisions accordingly. Recent studies have suggested that parasitoid insects may have such abilities, but were based on too few patches for clear conclusions on Bayesian updating. We studied how females of an aphid parasitoid update their estimate of habitat profitability, by recording their behaviour on a series of seven consecutive host patches, each serving both as a potential cue for habitat profitability and as a means to measure the behavioural response to previously collected information. The experiment was designed to test specific predictions based on Bayesian updating with different memory dynamics. The parasitoids clearly perceived the profitability of the patches, but the effect of previous experience on current patch use was different from our predictions. The reinterpretation of previous studies in the light of our results suggests that foraging parasitoids give more weight to early experience.