Behavioural correlations may cause partial support for the risk allocation hypothesis in damselfly larvae

Prey animals are often confronted with situations that differ in predation risk. According to the risk allocation hypothesis, prey animals should adaptively allocate antipredator behaviour in accordance with the magnitude and frequency of those risk situations. According to the first prediction prey animals should increase foraging in the safe situations and decrease foraging in the dangerous situations as these situations become relatively more dangerous. The second prediction is that with increased time spent in the dangerous situations, progressively more foraging effort is shown in both the dangerous and safe situations, especially in the safer ones. Prey animals may, however, show maladaptive behaviour due to behavioural correlations across risk situations. Here we test for the first time both predictions generated by the risk allocation hypothesis while considering behavioural correlations. We reared larvae of the damselfly Ischnura elegans, from the egg stage, under five rearing risk conditions: (i) in isolation, (ii) in the presence of conspecific larvae, (iii) in the presence of one fish, (iv) in the presence of two fish, and (v) in the presence of two fish for 50% of the time. For each rearing risk condition, we scored their behaviour in the absence and in the presence of fish. In accordance with the first prediction, in the absence of a predator, larvae reared under increasing risk conditions increased their level of foraging. In accordance with the second prediction, in the absence of a predator, larvae that were more frequently exposed to fish during rearing, increased foraging. However, opposite to the predictions from the risk allocation hypothesis, foraging increased both with increasing rearing risk, and with increased predator exposure frequency. The observed positive behavioural correlation of foraging activity across test situations with and without fish, may generate the combination of adaptive patterns in the absence of fish and the maladaptive patterns in the presence of fish. Former studies of the risk allocation hypothesis also found, at best, mixed support, and we hypothesize that behavioural correlations across risk situations, if present, will likely cause partial deviations from model predictions.