Zoological Studies

Assessing Avian Point-count Duration and Sample Size Using Species Accumulation Functions

Hau-Jie Shiu and Pei-Fen Lee (2003) We used 2 species accumulation functions to fit observed data and determine appropriate point-count durations and sample sizes for bird surveys in 3 vegetation types during both breeding and non-breeding seasons. We conducted fieldwork in a grassland, a coniferous forest, and a broadleaf forest in subtropical Taiwan from October 1999 to January 2000 and from March to June 2000. In almost all cases, both the linear dependence model and the Clench model fit the observed data very well. We calculated the duration of point counts required to reach 80% of the predicted asymptote. The point-count duration required in the breeding season was 1 min for the grassland, 2 min for the coniferous forest, and 3 min for the broadleaf forest. During the non-breeding season, the required duration was 3 min for the grassland, 9 min for the coniferous forest, and 6 min for the broadleaf forest. The required samples depend on the goals of a survey, the season, and vegetation types. Our results demonstrate that bird detections vary between seasons and vegetation types. Therefore, when determining the durations and sample sizes of pointcounts, investigators should consider characteristics of the detection patterns of the bird communities. We show that asymptotic models can be used to examine these patterns in an objective and rigorous way.