Embedding Population Dynamics Models in Inference.pdfVIP

Statistical Science 2007, Vol. 22, No. 1, 44–58 DOI: 10.1214/088342306000000673 c Institute of Mathematical Statistics, 2007 Embedding Population Dynamics Models in Inference Stephen T. Buckland, Ken B. Newman, Carmen Fern´andez, Len Thomas and John Harwood 7 0 0 2 Abstract. Increasing pressures on the environment are generating an g ever-increasing need to manage animal and plant populations sustain- u ably, and to protect and rebuild endangered populations. Effective man- A agement requires reliable mathematical models, so that the effects of 8 management action can be predicted, and the uncertainty in these 2 predictions quantified. These models must be able to predict the re- sponse of populations to anthropogenic change, while handling the ] E major sources of uncertainty. We describe a simple “building block” M approach to formulating discrete-time models. We show how to esti- . mate the parameters of such models from time series of data, and how t to quantify uncertainty in those estimates and in numbers of individu- a t als of different types in populations, using computer-intensive Bayesian s [ methods. We also discuss advantages and pitfalls of the approach, and 1 give an example using the British grey seal population. v Key words and phrases: Hidden process models, filtering, Kalman fil- 6 9 ter, matrix population models,