Estimating Abundance from Counts in Large Data Sets of Irregularly-Spaced Plots using Spatial Basis Functions

摘要

Monitoring plant and animal populations is an important goal for bothacademic research and management of natural resources. Successful management ofpopulations often depends on obtaining estimates of their mean or total over aregion. The basic problem considered in this paper is the estimation of a totalfrom a sample of plots containing count data, but the plot placements arespatially irregular and non randomized. Our application had counts fromthousands of irregularly-spaced aerial photo images. We used change-of-supportmethods to model counts in images as a realization of an inhomogeneous Poissonprocess that used spatial basis functions to model the spatial intensitysurface. The method was very fast and took only a few seconds for thousands ofimages. The fitted intensity surface was integrated to provide an estimate fromall unsampled areas, which is added to the observed counts. The proposed methodalso provides a finite area correction factor to variance estimation. Theintensity surface from an inhomogeneous Poisson process tends to be too smoothfor locally clustered points, typical of animal distributions, so we introduceseveral new overdispersion estimators due to poor performance of the classicone. We used simulated data to examine estimation bias and to investigateseveral variance estimators with overdispersion. A real example is given ofharbor seal counts from aerial surveys in an Alaskan glacial fjord.