By Joseph E. Hightower and Julianne E. Harris.
We simulated and evaluated multistate capture–recapture models to estimate mortality rates using telemetry data. Four field designs were considered: (A) fixed receivers to estimate total instantaneous mortality (Z), (B) manual searches to estimate instantaneous fishing (F) and natural (M) mortality, (C) fixed receivers combined with external high-reward tags to estimate F and M, and (D) manual searches combined with external high-reward tags to estimate M and fishing mortality rates associated with harvest (Fh) and catch-and-release death (Fcr) as well as the probability of death due to catch and release (α). Estimates generally appeared to be unbiased for a simulated study with five periods and releases of telemetered fish at the start of periods 1–4. Compared to estimating Z, larger sample sizes are needed to achieve reliable estimates of component rates (F and M).
Information on seasonality of natural mortality can suggest key mortality sources such as spawning (Waters et al. 2005) or extreme temperature (Ellis 2014), which help us better understand biology, elucidate annual variability in population size, and inform the timing of harvest regulations. Sources of mortality can be evaluated at a very fine temporal scale using telemetry.
Estimates of component rates were more precise when that source of mortality was directly observed (M in design B, F in design C). The field design us- ing fixed receivers and high-reward tags should be especially useful in practice, because manual searches are not required to estimate F and M. Multistate models are useful for clarifying the connection between field observations and ecological processes. Reliable estimates of mortality rates, coupled with information on behavior, habitat use, and movement, make telemetry a highly valuable tool for improving fisheries management and stock assessment.
Members click below for the April 2017 Fisheries magazine’s complete issue. Non-members, join here.