Evaluating observer effects in point count studies: Using stereo-microphones to investigate the role of species identification errors on Breeding Bird Survey (BBS) point counts and microphone arrays to estimate changes in bird behaviour in response to an observer's presence.

摘要

I evaluated the role of observer effects in point count survey and analysis methods used in North American avian monitoring programs. Skilled observers listening to recordings of Breeding Bird Survey (BBS) routes differed significantly in the number of species detected. On average 11% of the species reported were errors and 23% of the species were missed on the recordings. Field observers reported similar number of species as listeners but fewer than were on the recordings. Site occupancy models failed to correctly account for identification errors and tended to overestimate species richness by 1.0--8.7 species and positively bias between 69%--100% of species specific occupancy estimates. Tests using a microphone array and acoustic localization system with mean localization error of 4.19 +/- 0.52m showed no difference in the mean number of individuals, singing position or onset of singing behaviour of birds in response to an observer conducting a point count.