Testing methods to estimate abundance in a Magellanic penguin colony

VILLANUEVA, M. C.; BERTELLOTTI, M.; DELL’ARCIPRETE, P.; PARMA, A.

Tasmania, Australia

Conferencia; 6th International Penguin Conference; 2007

stimating population sizes is essential for many ecological analyses. However, in many cases it is difficult to establish accurately the limits of the distribution of a population or the number of individuals. Although several sampling methods have been used to establish the size of Magellanic penguin colonies, their relative performance has never been assessed. In this study we tested the efficiency of three sampling methods frequently applied in Patagonia using geostatistical tools and simulation-estimation trials. We generated a virtual colony with spatial parameters estimated for a real colony of Magellanic penguins located at Peninsula Valdes, applying kriging interpolation techniques. On this virtual colony three sampling methods were applied using a range of sampling fractions from 1% to 25% of the area covered: random sampling with quadrats of 100 m2, systematic sampling with quadrats of 100 m2, and systematic sampling with fixed-width transects. Simple design-based estimators were used assuming that the area occupied by the colony was known. Systematic sampling with quadrats gave the most accurate estimates for any given sampling fraction.