Avian points counts are commonly used to inventory bird species composition and abundance. When designing a study to inventory birds using point counts, point plot size and count duration are 2 variables that must be considered based on project goals. We conducted double-observer point counts on the Camp Dawson Collective Training Area, Preston County, West Virginia, during 2000 and 2001 to ascertain how detection probabilities and abundance estimates are affected by different plot sizes and durations. We conducted point counts from dawn to 1000 hours, and sub-tallied birds into different distance (0-<50 m, 50-100 m, and >100 m) and time (0-<3 minute, 3-5 minute, and >5-10 minute) intervals. We used program DOBSERV to generate species-specific detection rates and abundances for the different distance and time intervals. Detection probabilities were greatest for 50-m radius point counts and least for unlimited-radius counts (P < 0.05). Three minute counts produced the greatest detection rates whereas probabilities were least for 10-minute counts (P < 0.05). Abundance estimates were least for 50-m radius point counts and greatest for unlimited-radius counts (P < 0.05). Three-minute counts produced the lowest abundances and 10-minute counts produced the greatest abundance estimates (P < 0.05). Lower detection probabilities for 10-minute counts likely occurred because a majority of birds had already been recorded at the beginning of the count. Low detection probabilities with unlimited-radius point counts can be attributed to the increasing difficulty of detecting birds at longer distances. Abundance estimates were greater for 10-minute and unlimited-radius point counts because more time and area is available in which to detect birds. Project objectives will ultimately dictate the count duration and plot size to use when conducting avian point counts.
