Integrated Population Modelling, a natural tool for population dynamics
Friday Seminar by Jean-Dominique Lebreton.
Human activities have become a key factor in the functioning of all ecological systems. In this context, population studies often attempt to predict or understand the trajectory of a population based on a variety of information, ranging from indices of population size or censuses to detailed demographic information based on marked individuals. Moreover, the available information is often incomplete. Traditionally, estimates of demographic parameters based on capture-recapture data are fed into a matrix model and the output of this model, in particular the predicted growth rate, is compared with the observed change in numbers. Rather than resorting to this ad hoc approach, integrated modelling combines intimately the two types of information above in a probabilistic model coupling a state equation (A population model iterating over time a population vector ), an observation equation (relating the population size estimates and the population vector) and the probabilistic model for the capture-recapture observation. Technically, in the simplest cases, Kalman filtering is a widely available technique to fit the resulting integrated model.
I present a collaborative work (Gauthier et al. Ecology in press) that uses this combined approach to analyze the population dynamics of a hunted species, the Greater Snow Goose (Chen caerulescens atlantica), and to examine the extent to which it can improve previous demographic population models. The matrix population model in the state equation included fecundity and regression parameters relating adult survival and harvest rate estimated in a previous capture-recapture study. The observation equation combined the output from this model with estimates from an annual spring photographic survey of the population. Integrated modelling appears as a promising approach to forecast population change because it incorporates survey information in a formal way compared with ad hoc approaches that either neglect this information or require some parameter or model tuning.
The CEES seminar room has a coffee-machine – it is therefore recommended that you come a bit earlier and get yourself a good cup of coffee (for the price of 3 NOK).