School Seminars and Colloquia

Estimating Species Abundance from Presence-Absence Maps:Leveraging Spatial Dependence Information

Statistics Seminar

by Associate Professor Wen-Han Hwang


Institution: Department of Applied Mathematics, National Chung Hsing University, Taichung, Taiwan
Date: Tue 26th August 2008
Time: 1:15 PM
Location: Rm 213, Richard Berry Bldg, The University of Melbourne

Abstract: A standard method for estimating species abundance from a presence-absence
map (PAM) or occurrence data in field grid cells is based on the negative
binomial distribution (NBD). However, estimating the NBD clumping
parameter using a PAM is not possible without external information. This
paper develops an alternative, the mixed Gamma-Poisson model, to estimate
species
abundance from a PAM. The proposed model is a simple generalisation of the
NBD model that characterises spatial dependence in the abundance
distribution across cells. Under this framework, the clumping parameter and,
consequently, species abundance can be estimated using a three-map
technique. The new approach sheds light on an outstanding controversy
about
the standard methodology in the literature. A 50-hectare stem-mapped tree
dataset from Panama is analysed to demonstrate the value of the proposed
method.

For More Information: Contact: Guoqi Qian g.qian@ms.unimelb.edu.au