Introducing selfisher: open source software for statistical analyses of fishing gear selectivity

There is a need to improve fishing methods to select for certain sizes and species while excluding others. Experiments are conducted to quantify selectivity of fishing gears and how variables such as gear design (e.g., mesh size, mesh shape), environmental parameters (e.g., light, turbidity, substrate) or biological parameters (e.g., fish condition) alter selectivity; the resulting data need to be analyzed using specialized statistical methods in many cases. Here, we present a new tool for analyzing this type of data: an R package named selfisher. It allows estimating multiple fixed effects (e.g., fish length, total catch weight, environmental variables) and random effects (e.g, haul). A bootstrapping procedure is also provided. We demonstrate its use via four case studies, including (A) covered codend analyses of four gears, (B) a paired gear study with numerous covariates, (C) a catch comparison study of unpaired hauls of gillnets and (D) a catch comparison study of paired hauls using polynomials and splines. This software will make it easier to model selectivity, teach statistical methods, and make analyses more repeatable.

Automated summary

There is a need to improve fishing methods to select for certain sizes and species while excluding others. A new tool called selfisher, which is an R package, has been developed to analyze data related to the selectivity of fishing gears. The tool allows for the estimation of fixed and random effects, and includes a bootstrapping procedure. It has been demonstrated through four case studies, showing its potential to model selectivity, teach statistical methods, and improve the repeatability of analyses.