Does prey preference change as a result of prey species being presented together? Analysis of prey selection by the predatory mite Typhlodromus pyri (Acarina: Phytoseiidae).

Prey-selection behaviour of the phytoseiid mite Typhlodromus pyri Scheuten was analysed with a Markovtype model of feeding-state dynamics and feeding-state dependent searching behaviour (Sabelis 1981, 1986, 1989; Metz and Van Batenburg 1985a, b). All behavioural characteristics of the predator which are independent of the feeding state were represented by one parameter. The remaining feeding-state dependent characteristics were represented by a function of the feeding state, with one parameter. The best parameter values to describe a predator-prey interaction were determined by fitting the model to the predation rates in monocultures. Under the assumption that the parameter values are not dependent on the composition of prey species supply, the diet of the predators in mixed cultures was predicted from parameters estimated in monoculture experiments.Two prey types, apple rust mite (Aculus schlechtendali (Nalepa)) adults and European red spider mite (Panonychus ulmi (Koch)) larvae were studied. A large discrepancy was observed between calculated and experimentally determined predation rates of T. pyri in mixed cultures: the predators actually killed 3-7 times more P. ulmi larvae than was predicted by the model.The large difference between observed and predicted predation rates in mixed cultures cannot be explained by changes in the behaviour of the prey species as a result of being together. Therefore, it seems likely that the prey selection behaviour of the predator was different when prey species were presented together than when presented singly. Apparently the predatory mite T. pyri prefers P. ulmi to S. schlechtendali.