Interface 2003

WiSP: An R library for simulating wildlife surveys
Walter Zucchini, (University of Goettingen),,
David Borchers, (University of St. Andrews),,
Stefan Kirchfeld, (University of Goettingen),, and
Martin Erdelmeier, (University of Goettingen),


WiSP, Wildlife Simulation Package, is a new library of functions written in R (Ihaka and Gentleman, 1996) designed as a tool for introducing students and researchers to methods used in wildlife abundance estimation, and for experimenting with the methods using simulation. The methododology and underlying theory are described in Borchers, Buckland and Zucchini (2002).

WiSP is an object-oriented package that enables users
- to generate and view animal populations with realistically complex spatial and individual characteristics,
- to generate a variety of survey designs (such as quadrat, removal, mark-recapture and line or point transect sampling),
- to survey populations under different assumptions regarding the visibility/catchability of the individual animals,
- to compute point and interval estimates of abundance for each model.

We outline the structure of WiSP and the contents and properties of its main objects emphasising the modularity of the architecture. We demonstrate the capabilities of WiSP and illustrate how it can be used to assess the performance of the estimators, including their sensitivity to specific violations in model assumptions (such as heterogeneity), by means of simulation and graphical displays.

Take me back to the main conference page.