The statistical analysis of animal populations

This thesis is concerned with assessing the spatial variability in animal populations. Chapter 1 is a short general introduction. Chapter 2 contains a discussion of the design of a large field experiment on the effect on the local ecology of the proximity of arable crops (winter barley in this case), semi-natural habitat and mature hedgerow. The counts of the large polyphagous beetle Pterostichus Melanarius recovered from an array of wet pitfall traps from the above experiment is then discussed. New tests for overdispersion and spatial aggregation are given in chapter 3. This extends the work of (Perry and Hewitt, 1991) based on the amount of effort it takes to smooth out counts data so that they are acceptably variable (variance=0). In chapter 4 we proceed to devise indices from Chapter 3's measures so the amount of pattern in different data sets can be compared. In chapter 5 these indices are applied to the counts of P. melanarius from the experiment at Long Ashton that was discussed in chapter 2, with each month analysed separately. A generalised linear model is also fitted to the data. In chapter 6 we look at use of a general host-parasite model by Pacala et al. (1991) to fit a large group of data sets. Some problems related to parameter estimations are considered. A slightly simpler model is then described and shown to be equivalent to a generalised linear model. Finally the distribution of movement distance and survival rates from ring-recovery data on blackbirds is discussed in chapter 7.