Accounting for Tag Loss and Its Uncertainty in a Mark–recapture Study with a Mixture of Single and Double Tags
Unless accounted for in the estimation, tag loss will cause mark–recapture methods to overestimate the true abundance of a closed population and to underestimate the associated uncertainty. Current methods of accounting for tag loss require all marked individuals to be double-tagged. We present a new model that fully accounts for tag loss and allows for the use of a mixture of single- and double-tagged individuals, thus simplifying implementation in the field. Treating abundance, tag loss rate, and capture probabilities as free parameters, we estimated those parameters and their uncertainty by using maximum likelihood. Whereas existing methods assume that a double-tagged animal does not lose both tags, the new model allows the animal to lose both tags. The new model’s performance was assessed and compared with that of other estimators (modified Petersen and Seber–Felton) via simulation. As expected, estimates from the new model were less biased and more precise than estimates from the other models. The model was used to estimate the abundance of the kokanee Oncorhynchus nerka population in the Metolius River, Oregon, during 2007. Abundance was estimated at 102,970 fish (SE = 8,930), tag loss rate was estimated at 0.27 (SE = 0.05), the capture probability for the first sample (tagging) was 0.03 (SE = 0.00), and the capture probability for the second sample (recovery) was 0.11 (SE = 0.01). The new model uses all of the information from single- and double-tag data, provides unbiased abundance estimates in the presence of tag loss for a closed population, and has less-stringent field requirements that make it easier to employ than other methods.
Hyun, S-Y., J.H. Reynolds, and P.F. Galbreath. 2012. Accounting for tag loss and its uncertainty in a mark–recapture study with a mixture of single and double tags. Transactions of the American Fisheries Society 141(1):11-25. Online at https://doi.org/10.1080/00028487.2011.639263.