Population estimation using partial double sampling

Abstract

Detectability issues create uncertainty in field surveys of populations. Methods to overcome this problem include mark-recapture methods and double sampling. Partial double sampling involves estimating detectability, p̂, and using this to correct the estimate in areas sampled only once. Results of this study indicate that, if population density is not uniform across sample units, there is a tradeoff for a given sampling effort, with lower population estimation error for double sampling only a portion of the area when spatial variance and/or detectability are lower but lower error for partial double sampling when spatial variance and/or detectability are higher. Thus, whereas standard double sampling increases precision, it does not give adequate accuracy in the face of spatial variation unless the entire area can be surveyed twice, which is more expensive. © 2010 by the Society of American Foresters.