Evaluation of a Mark–Recapture Method for Estimating Mortality and Migration Rates of Stratified Populations
Authors:
- Robert M. DorazioRDRobert M. Dorazio
- Paul J. RagoPRPaul J. Rago
- 1 February 1991
- journal article
- Published by Canadian Science Publishing in Canadian Journal of Fisheries and Aquatic Sciences
- Vol. 48 (2) , 254-260
Abstract
We simulated mark–recapture experiments to evaluate a method for estimating fishing mortality and migration rates of populations stratified at release and recovery. When fish released in two or more strata were recovered from different recapture strata in nearly the same proportions, conditional recapture probabilities were estimated outside the [0, 1] interval. The maximum likelihood estimates tended to be biased and imprecise when the patterns of recaptures produced extremely "flat" likelihood surfaces. Absence of bias was not guaranteed, however, in experiments where recapture rates could be estimated within the [0, 1] interval. Inadequate numbers of tag releases and recoveries also produced biased estimates, although the bias was easily detected by the high sampling variability of the estimates. A stratified tag–recapture experiment with sockeye salmon (Oncorhynchus nerka) was used to demonstrate procedures for analyzing data that produce biased estimates of recapture probabilities. An estimator was derived to examine the sensitivity of recapture rate estimates to assumed differences in natural and tagging mortality, tag loss, and incomplete reporting of tag recoveries.Keywords
This publication has 3 references indexed in Scilit:
- The Relationship between Harvest and Survival Rates of Mallards: A Straightforward Approach with Partitioned Data SetsThe Journal of Wildlife Management, 1983
- Parameter estimates from mark‐recapture experiments on two populations subject to migration and deathPopulation Ecology, 1972
- The two-sample capture-recapture census when tagging and sampling are stratifiedBiometrika, 1961