The Effect of Observer Variability on Bird Census Results Obtained by a Territory Mapping Technique

Abstract

This study was carried out to determine whether experienced ornitohologists could replace each other in a census programme using the territory mapping technique. We studied the variation between different observers carrying out a 10-visit mapping census in the same plot and in the same period of time. The test was repeated in two consecutive seasons in a plot with about 50 territorial males. The coefficient of variation for the total counts of territories was 14 and 10% respectively in the two seasons. The variation between the observers decreased with increasing sample size in good agreement with the theoretical expectation. Consequently, it could be concluded that census data obtained from different observers were comparable, at least so long as the process of species map evaluation did not introduce any additional variation.