Performance of Trans-Generational Genetic Mark-Recapture (tGMR) Models in Estimating Fish Abundance

The Lincoln-Petersen abundance estimator has recently been adapted using genetic marks to estimate abundance via so-called trans-Generational Mark Recapture (tGMR).  A two-sample Lincoln-Petersen abundance estimate is calculated by: 1) genotyping individuals sampled from the parental generation (marks), 2) genotyping individuals sampled from following generation and using genetic parentage analysis to assign offspring to unique parents (captures) which either matched genotyped parents (recaptures ) or had unidentified parents. Washington Department of Fish and Wildlife has used tGMR to estimate abundance of spawning salmon, but the accuracy and precision of tGMR estimators, which involve different inputs and assumptions than traditional models, have not been fully evaluated.  Biological factors unique to tGMR, such as variance in reproductive success and adult sex ratio, may affect the accuracy and precision of estimates.  We used simulated population data to evaluate the accuracy and precision of tGMR methods.  Simulated populations varied in abundance, sex ratio, and distribution of reproductive success.  Available models accurately estimated abundance with increasing precision with increasing sample sizes.  Models varied in their performance in estimating sampling variance.  Sex ratio and reproductive success affected precision and accuracy of abundance and variance estimates, but did not bias abundance estimates if assumptions of models were met.