Clustering and abundance estimation for Neyman-Scott models and line transect surveys

This paper considers the estimation of clustering parameters and mean species intensity based on likelihood theory for the simplified Neyman-Scott Poisson model, with observations taken from line transect surveys with a Gaussian detection function. The estimators and accompanying standard error expressions are tractable and easy to calculate, and, coming from likelihood methods, often will have high efficiency. Such properties compare favourably with those of existing K-function methods which however are semiparametric in nature and less reliant on specific parametric assumptions. The likelihood analysis reveals auxiliary information which could be used to check the form of the detection function. Clustering; Cramer-Rao bound; Detection function; K-function; Line transect survey; Maximum likelihood; Neyman-Scott line transect distribution; Neyman-Scott Poisson process.