Efficacy of depletion models for estimating abundance of endangered fishes in streams

Capture probability of fishes in Central European (Hungary) wadeable lowland streams

“…Using a hierarchical multinomial N-mixture model on double-pass electrofishing data, the study estimated the magnitude and uncertainty of capture probability and abundance for seven common fish species in wadeable Pannonian lowland streams. Similarly to many other studies (Benejam et al, 2012;Harris et al, 2016;Hedger et al, 2018;Peterson & Cederholm, 1984;Riley et al, 1993;Riley & Fausch, 1992), data showed that efficacy of electrofishing is not perfect, single-pass sampling resulted in a catch less than the true population size, and capture probability varied across the species (Mollenhauer et al, 2018;Price & Peterson, 2010;Reid et al, 2009;Stewart et al, 2019), although its stochastic fluctuations were too wide to identify sharp differences between the species. Stochastic fluctuation of capture probability is likely enhanced when stream reaches with very different environmental characteristics are sampled.…”

“…The model used in this study can be extended for allowing separate estimations of capture probability for the passes of a data set with three or more passes, and then testing of the decline in catchability becomes possible. In that case, even if the catchability proved to be constant, the increased number of passes would likely improve the estimations of the model (Riley & Fausch, 1992;Stewart et al, 2019).…”

“…Because the main goal of this study was to estimate the extent of capture probability and not the investigation of the factors influencing that, random effects were used as surrogates for the environmental covariates. Nonetheless, our model can easily be developed by merging covariate terms into the log-linear and the logit-linear regression equations to assess the influential role of habitat features or sampling circumstances on capture probability and abundance (Joseph et al, 2009;Kéry & Schaub, 2012;Mollenhauer & Brewer, 2017;Smith et al, 2012;Som et al, 2018;Stewart et al, 2019).…”

Estimating the Effects of Environmental Variables and Gear Type on Panfish Catch and Bycatch in Reservoirs

“…We restricted the table of frequency categories for analysis to those categories containing more than five individuals, and we used Cochran-Mantel-Haenszel (CMH) test statistics (χ 2 ) to draw inferences while controlling for lake, with the Bonferroni correction used to reduce the chance of a type I error (Agresti 2002). We used a multispecies Bayesian mixture model to link the individuals collected from the temporally replicated samples to a latent abundance state at a site within each lake to quantify detection and true local abundance (Stewart et al 2017(Stewart et al , 2019. Therefore, we consider the experimental design where the observed catch histories represent a sequence of replicated counts of unmarked fish (y) for panfish species j at each site i = 1, 2,. .…”

“…We used a multispecies Bayesian mixture model to link the individuals collected from the temporally replicated samples to a latent abundance state at a site within each lake to quantify detection and true local abundance (Stewart et al 2017, 2019). Therefore, we consider the experimental design where the observed catch histories represent a sequence of replicated counts of unmarked fish ( y ) for panfish species j at each site i = 1, 2,…, R within k = 1, … , K lakes for a total of t = 1, …, T replicate sampling occasions (Stewart and Long 2016).…”