Fitting ecological dynamic models to data

This set of lectures aims to provide an introduction to fitting and inference with population dynamic models. The basic motivation is that often we know enough about a biological system to formulate models describing how we think it works, but not quite enough to pin down the precise numerical values of all the parameters (or in some cases functions) that it contains. In this circumstance it’s useful to be able to make inferences about the unknown parts of the model by fitting to data. The structure of the lectures is as follows. I’ll start by reviewing the basic theory of linear statistical models, and then give an example of how this theory can be used directly to fit simple structured population models of the type popularised by Caswell. This example highlights some interesting general problems. Having proposed solutions to these I’ll then proceed to develop a slightly more interesting continuous counterpart of the matrix model approach, to further illustrate the applications of strictly linear theory. The final part generalizes further to deal with the non-linear models that usually arise when modelling on the basis of biological mechanisms.