Uncertainty of Exploitation Estimates Made from Tag Returns

Abstract

Over 6,000 crappies Pomoxis spp. were tagged in five water bodies to estimate exploitation rates by anglers. Exploitation rates were computed as the percentage of tags returned after adjustment for three sources of uncertainty: postrelease mortality due to the tagging process, tag loss, and the reporting rate of tagged fish. Confidence intervals around exploitation rates were estimated by resampling from the probability distributions of tagging mortality, tag loss, and reporting rate. Estimates of exploitation rates ranged from 17% to 54% among the five study systems. Uncertainty around estimates of tagging mortality, tag loss, and reporting resulted in 90% confidence intervals around the median exploitation rate as narrow as 15 percentage points and as broad as 46 percentage points. The greatest source of estimation error was uncertainty about tag reporting. Because the large investments required by tagging and reward operations produce imprecise estimates of the exploitation rate, it may be worth considering other approaches to estimating it or simply circumventing the exploitation question altogether.

Sensitivity of tag-recovery mortality estimates to inaccuracies in tag shedding, handling mortality, and tag reporting

“…With this function, chronic shedding was near zero for the first 5 months after tagging, and then progressively increased during the next several months before stabilizing at a shedding rate of 31% at 12 months post-tagging. This pattern in chronic shedding, as well as the rates associated with handling mortality and immediate shedding, corresponded to our observations in the lake whitefish study (see Ebener et al, 2010a), and are similar to rates that have been reported in the literature (Ebener and Copes, 1982;Muoneke, 1992;Pierce and Tomcko, 1993;Fabrizio et al, 1996;Hearn et al, 1999;Buzby and Deegan, 1999;Pollock et al, 2001;Miranda et al, 2002;Polacheck et al, 2006;and Taylor et al, 2006). We considered two spatial patterns of reporting rates in our datagenerating model.…”

“…As a result, accurate mortality estimation using a tag-recovery approach at least partly depends on the collection of auxiliary data pertaining to tag shedding (hereafter referred to as shedding), handling mortality, and tag reporting (hereafter referred to as reporting). Each of these factors can be measured in a variety of ways: shedding can be estimated by double tagging or supplemental marking of fish (Pierce and Tomcko, 1993;Fabrizio et al, 1999;Latour et al, 2001;Miranda et al, 2002, Livings et al, 2007; handling mortality can be estimated by withholding samples of tagged fish in tanks, pens, or cages (Pierce and Tomcko, 1993;Latour et al, 2001;Miranda et al, 2002;Taylor et al, 2006); reporting rates can be estimated through the use of highreward tags Pollock et al, 2002;Taylor et al, 2006), planted tags (Hearn et al, 2003), or creel or port surveys (Hearn et al, 1999;Pollock et al, 2002).…”

“…Even when data concerning shedding, handling-mortality, and reporting rates are collected as part of a tagging study, biased mortality estimates may still result if measurements of these rates are not accurate (Miranda et al, 2002). For example, handling mortality may be overestimated if fish held in nets or pens become stressed as a result of biofouling (Ahlgren, 1998;Udomkusonsri and Noga, 2005;Isermann and Carlson, 2008).…”

First evidence of potadromy and partial migration in black basses: shoal bass Micropterus cataractae (Actinopterygii, Centrarchidae) in the Upper Flint River, USA

“…Reporting rates can be problematic when relaying on anglers for recapture information (Miranda et al, 2002). To maximize angler reporting of tags, most rewards given out during the first year were either $20 or $50, to spur interest and participation by anglers (Pollock et al, 2001).…”

Accuracy and precision of fishery and demographic estimates from a spatial tag-recovery model when inter-regional movements are treated as fixed

“…Thus, biologists and managers can prevent a large waste of resources by conducting preliminary evaluations on possible tagging study designs. Recovery of tags is affected by a variety of factors, including tagging levels, issues with the tagging process (e.g., tag shedding, handling mortality), fishery characteristics (e.g., tag non-reporting, level of harvest), and fish population dynamics (e.g., movement patterns, natural mortality) (Fabrizio et al, 1996;Hoeing et al, 1998;Latour et al, 2001;Miranda et al, 2002). Because of this, it is important to examine accuracy and precision of parameter estimates across a range of design features, and stochastic simulations, such as those employed herein, have been identified as useful in this regard .…”