Size Bias in Line Transect Sampling: A Field Test

Abstract

An important problem in line transect sampling is that objects or point clusters of objects of different sizes have different sighting probabilities. In a recent paper Drummer and McDonald (1987, Biometrics 43, 13-21) develop a bivariate sighting function. Their function is dependent on perpendicular distance and object size. One important special case is an extension of the exponential power series sighting function first proposed by Pollock (1978, Biometrics 34, 475-478). In this note empirical evidence is given for this model based on a field test of line transect sampling theory. Beer cans were used to simulate point clusters of objects with cluster sizes 1, 2, 4, and 8. To achieve approximately equal precision of parameter estimates, equal numbers of each cluster size were taken.