Are survival rates for northern spotted owls biased?

The northern spotted owl (Strix occidentalis caurina (Merriam, 1898)) is listed as threatened in both Canada and the United States. As part of a 1998–2004 study of habitat usage, we attached radio tags to 197 northern spotted owls. Owls that died or emigrated from the study areas could be identified with high certainty. The long-term data we obtained enabled us to estimate survivorship using multiple statistical methods. Using a pooled data set, we estimated annual survivorship at 0.927. Using a year-by-year analysis, we obtained some variation in survival by year, but the same overall mean. Using a staggered-entry cohort approach, we obtained an estimate of 0.934. Mean annual survival estimated by program MARK was 0.927. These estimates are outside the confidence intervals of prior studies that used capture–recapture methods. Capture–recapture methods are based on the assumption that birds remain within a demographic study area, but our data suggest that owls may disperse or remain undetected within a study area often enough that capture–recapture methods may overestimate mortality. Our results imply that the true finite population growth rate, λ, may be higher than estimated in prior studies that used capture–recapture methods.