Comparison of Transects and Circular-Plots for Estimating Bobolink Densities

Abstract

We compared the effectiveness of line transect and variable circular-plot techniques for estimating bobolink (Dolichonux oryzivorus) densities in 2 grasslands in New York [USA]. Population sizes were known precisely because of intensive netting, color marking, and observation. Both techniques usually overestimated density, and mena relative biases ranged from .apprx. 30 > 100%. Line transects consistenly provided density estimates with smaller biases and higher correlations with true densities than did variable circular-plots. Density estimates derived from afternoon counts were usually more accurate than those derived from morning counts. Both counting techniques provide reasonable indices of abundance but overestimate bobolink density because of undetected movement of the birds toward the observer.