PARENT SESSION
Oral Session 16: Statistics and Biometrics.
Presiding: E Garton and T Simons
Monday, August 2, 8:00 AM to 11:30 AM, Meeting Room D 139.
Experimental analysis of detection probabilities on avian point count censuses.
Simons, Theodore1, 2, Pollock, Ken2, Alldredge, Mat 2, 1 USGS Cooperative Fish and Wildlife Research Unit, Raleigh, NC2 North Carolina State University, Raleigh, NC
ABSTRACT- We have developed a system for simulating avian census conditions when most birds are identified by sound, for example most Breeding Bird Surveys. The system uses a laptop computer to control a set of MP3 players placed at known locations around a census point. The system can realistically simulate a known population of songbirds under a range of factors that affect detection probabilities, so that the performance of various field and analytical methods can be evaluated. We anticipate that our analyses will provide field biologists and natural resource managers with a clearer understanding of the biases and precision of current sampling methods, and the improvements expected if additional information needed to estimate detection probabilities is collected. Validation experiments are incorporating traditional methods for estimating detection probabilities such as distance sampling, and new approaches that incorporate information from multiple observers, the time sequence of observations, and combined methods. Preliminary data from over 600 experimental counts conducted in the winter 2003 and spring 2004 indicate that estimates from unlimited radius counts underestimated true populations, while counts restricted to 50 m fixed-radius plots significantly overestimated populations. Our goal is to find new applications of theory and sampling methodologies that will result in practical improvements in the quality of bird census data.
Key words: point-count, accuracy, avian, validation