Modelling foraging behaviour in the insect predator Notonecta maculata using the individuals approach

Predation is considered to significantly alter dynamics and diversity of aquatic communities. The current thesis was therefore aimed at assessing mechanisms behind the predation by the backswimmer Notonecta maculata in the context of ecology, ethology and physiology. Moreover, a foraging model was developed, parameterized and evaluated on the basis of individual foraging behaviour, for application in ecology and ecotoxicology. An iterative process of model development thereby provided insights into several research questions. In a first step the ecological and ecotoxicological relevance of backswimmer predation was examined. Therefore, population dynamics of Daphnia magna was observed under predation pressure, xenobiotic stress and a combination of both predation and chemical disturbance. The experiment revealed that size selective predation of the backswimmers did alter population demography, which in turn led to higher vulnerability of D. magna population to toxic stress. The results mentioned above have raised the question to what extend the functional response and size selectivity of juvenile backswimmers changes with predator instar. Therefore, ecological parameters that were considered influential in prey selection were quantified experimentally and subsequently led to the parameterization of an individuals model to be used for the prediction of predation rates. Testing the model against independent data revealed that prey selection in backswimmer might not be determined by active preference, but by passive selectivity mediated by variations in the components of predation. Application of the model revealed that predation rates in juvenile backswimmers are reduced under dark conditions compared to light. These differences are accounted for by the presence/absence of light and are not regulated by any endogenous circadian rhythm. Behavioural observations led to the conclusion that in the dark backswimmers probably detect prey items at a shorter distance as compared to light and that the resulting reduction in the encounter rate is compensated by gaining a higher amount of food per item during a lengthier, more thorough handling of the prey. Comparing model output to independent functional response data did support this hypothesis. In order to account for variations in food availability, and thus allow for long term prediction of predation rates, the model was extended by a physiological component. Therefore, behavioural observations on the components of predations were correlated with gut state of the predator. Moreover, the empirical results showed that food-deprived backswimmers attack their prey more frequently and that they extract more biomass from a single prey item than their well-fed conspecifics do. Model application provided further evidence for compensatory feeding under variable environmental conditions. In the current study, proximate factors in the foraging behaviour were identified and implemented in a simulation model. The model developed here is able to adequately predict size selectivity and long term predation rates of juvenile backswimmers under laboratory conditions. A major future task will be the generalization of mechanistic rules that will allow predictions for a broad range of species