Improved estimates of population extinction risk (Harding and McNamara)

An important application of theoretical ecology is in estimation of species extinction risk. Extinction models guide the selection of management regimes for endangered species. Two vital parameters in these models are the mean population growth rate and its variance. However, empirical data on population growth are rarely perfect, but are influenced by random sampling error (induced by e.g. weather conditions). It has been unclear how sampling error influences extinction estimates. In the latest issue of Ecology Letters, McNamara (U. Bristol) and Harding (U. Göteborg) show that sampling error has two opposite effects on estimates of population extinction risk. The errors lead to an exaggerated overall variance, but also introduce negative autocorrelations in the time series of population growth data. Interestingly, these two effects exactly counterbalance. Thus, a practical and efficient way to resolve the problem of dealing with sampling error is to routinely incorporate a measure of between year correlations in estimating population extinction risk.

---

---