| Summary: | Population modelling is recognised as a potentially useful tool for pesticide risk assessment for vertebrate species. The European Food Safety Authority, which provids guidance on how to conduct such risk assessments in the EU, recommends using the woodpigeon (Columba palumbus) as a representative non-target species for a variety of crops, including oilseed rape and cereals. The woodpigeon was extensively studied in the UK for the second half of the twentieth century. The aim of this PhD was to produce a population model of wood pigeons for use in higher-tier pesticide risk assessment with a two-fold purpose: prediction of exposure and population-level effects of this exposure. In the thesis I first discussed how computer simulations are used for these purposes. Then I described the design of the woodpigeon model together with extensive reference to the scientific literature that backed the model and the tests conducted to assure the model credibility. The pattern-oriented evaluation was conducted to choose which foraging strategy allows for the best reproduction of data describing spatial and temporal characteristics of woodpigeon foraging. Memory-based foraging with flocking behaviour outperformed both optimal and random strategies. Next, I evaluated the model with respect to the responses of woodpigeon population to three stressors: nest destruction, shooting and pesticide application. Following the evaluation studies, the model was used to derive and analyse three measures of population-level exposure to pesticides: maximum dose, cumulative amount and duration of exposure. Life stage and pesticide application time proved important determinants of the exposure level. Finally the model was applied to check how the way of modelling exposure influenced the population size and breeding success. The realistically simulated exposure caused long-term damage to the population size and affected the breeding success more than exposure causing all the lethal effects within one day. I discussed the major findings of the thesis in the light of advances in applying population modelling for pesticide risk assessments and possible future application of the woodpigeon model.
|
|---|
