Catch efficiency of trawl nets used in surveys of the yellow squat lobster (Cervimunida johni) estimated by underwater filming records

Abstract

Catch efficiency is an important index to relate observed density to the size of a population in the context of bottom trawl surveys. The estimation of catch efficiency is challenging because it involves independent measures of animals in the path of the bottom trawl. We estimated catch efficiency using an underwater camera system on three trawl fishing vessels used for estimating the density of the yellow squat lobster (Cervimunida johni) off central Chile. During 2015 and 2018, 54 hauls were analyzed, and a total of 20 h of filming were recorded. A total of 4,155 yellow squat lobster individuals were analyzed approaching the net in the path of the trawl, of which 2,330 (56%) were captured and 1,825 escaped underneath the groundrope. The median estimated ranges of catch efficiency varied between 0.81–0.90 for vessel 1, 0.15–0.72 for vessel 2 and 0–0.58 for vessel 3. According to a hierarchical generalized linear model (HGML), fishing vessels and mean depth of hauls showed significant differences in catch efficiency (p<0.05). The results were comparable with estimates available for other crustacean species, and differences among vessels can be associated with differences in rigging configurations. Discussion was focused on the processes affecting catch efficiency and how differential catchability among sampling vessels may bias the construction of spatially explicit density maps and further abundance estimates of yellow squat lobsters in Chile.

Introduction

The yellow squat lobster (Cervimunida johni Porter 1903) is a benthic crustacean belonging to the Munididae family that is distributed along the Chilean continental shelf and upper continental slope between 110 and 350 m, mainly occupying gravel and muddy substrates between Taltal (25$°$19’S) and Mocha Island (38$°$20’S) (Quiroz et al., 2005, Ahumada et al., 2013). It supports an important fishery based on bottom trawlers vessels which managed on a total allowed catch (TAC) basis and also considering an individual transferable quota (ITQ) share base scheme. The TAC is estimated using integrated stock assessment models calibrated with commercial fisheries information as well as fisheries-independent abundance indices from trawl surveys, based on the swept area method and conducted since 1995. These trawl surveys use commercial vessels and their fishing nets to conduct the assessment without prior standardization among vessels due the lack of a sufficient number of scientific vessels being available in the country (Queirolo et al., 2018) and standardized hauls in terms of trawling time and other operational characteristics. The main assumption underlying the swept area method relies on a constant or stationary catchability (q) in time and space (Kimura and Somerton, 2006), and is established despite this parameter being affected by factors other than population density, such as the performance of the sampling gear (Kotwicki et al., 2014, Somerton et al., 2007). This means that any vessel conducting the survey removes the same proportion of the population density with one unit of effort. This assumption can be difficult to meet when using commercial vessels because they can experience differential catchabilities even when using similar standardized nets as a result of differences in rigging configurations. Catchability can be decomposed into two components (Godø, 1994): the availability of individuals in the sampling area (qa), and the catch efficiency, or the proportion of the animals sampled within the swept area (qe) (Dickson, 1993, Kimura and Somerton, 2006). Therefore, population density can be expressed in terms of observed density (CPUA) as D=CPUA/(qa*qe) (Walsh, 1996). In this context, the estimation of catchability can be particularly challenging because it requires a certain knowledge about population density, which is often not available.

However, in relation to yellow squat lobsters, previous experiments revealed that filming cameras attached to the net allow the identification and quantification of individuals approaching the trawling path (Ahumada et al., 2013). Therefore, the proportion of animal sampled within the swept area qe for bottom trawl surveys can be estimated by assessing the animals that approach the trawl net in the path of the trawl and estimating the relative proportion of those animals that enter the net. For this reason, this study was undertaken with the aim of estimating the catch efficiency of three different fishing vessels using underwater cameras with information collected in surveys of yellow squat lobsters conducted in central Chile in 2015 and 2018, allowing to check feasibility of this approach and to verify assumption of stationary catchability among vessels.