Procedures for Estimating Tag Loss Rate and Spawning Escapement in a Mark-recapture Study of Metolius River Kokanee Oncorhynchus nerka
Abstract
Mark-Recapture/Resight procedures were used to estimate the 2007 spawning escapement of kokanee Oncorhynchus nerka migrating from Lake Billy Chinook into the Metolius River, Jefferson County, Oregon. An initial analysis of data, which included an adjustment for a presumed 25% tag loss rate (without confidence limits), provided an escapement estimate of 101,854 +/- 11,151 (+/-10.9%; 95% confidence interval). Here we present an expanded analysis of the data to look for effects of the survey protocols, and we have re-estimated abundance, incorporating the additional uncertainty associated with estimation of tag loss rate based on recapture (resight) counts of double-tagged fish. The resight procedures in project involved subdivision of the spawning area in the Metolius into five sections, each of which was surveyed once in each of three different weeks. A two-way analysis of variance of these data indicated no significant effect on the proportion of marked fish among observed spawners (overall average = 2.42%) for either survey section or survey week. Among the 3,298 fish tagged in 2007, 491 received a second tag. A total of 59 of these double-tagged fish were resighted – 35 with both tags and 24 with a single tag. A binomial likelihood analysis of these data indicated a tag loss rate of 6%+/-9.9%. Incorporation of the uncertainty associated with tag loss estimation into a hypergeometric likelihood procedure to calculate escapement yielded an estimate of 106,630+/-16,393 (+/- 15.4%). Several scenarios based on the 200 ta were run using these likelihood procedures, to illustrate the effects on precision of estimate from an increase or decrease in the total number of tagged fish, in the number of double-tagged fish, and in the total number of fish observed during the surveys, as well from an increase or decrease in observed tag loss rate. The most dramatic improvement in precision of the abundance estimate per unit change in 2007 tagging/survey effort could have been obtained from increasing the proportion of double-tagged fish – to increase certainty of the tag-loss estimate. We propose adoption of the bi al-hypergeometric likelihood estimator procedures to incorporate tag loss rate and its uncertainty into a Mark-Recapture/Resight analysis for estimation of abundance, and provide suggestions for modifications of the Metolius kokanee tagging and survey procedures for spawning escapement estimation in future years.
Authors
Citation
Galbreath, P. F. and S-Y. Hyun. 2008. Procedures for estimating tag loss rate and spawning escapement in a mark-recapture study of Metolius River Kokanee Oncorhynchus nerka. Columbia River Inter-Tribal Fish Commission Technical Report 08-7. Portland, OR. 32p.