Evaluating the selective performance of the encircling gillnet used in tropical estuarine fisheries from Colombia

Abstract

Experimental estimates of gear selectivity by age used for estimating stock size and measuring fishing impact have been scarcely documented in small-scale fisheries from Latin-America and the Caribbean. In order to improve stock assessment and fisheries management in such fisheries, an experimental approach was designed to estimate the selective performance of the encircling gillnet “boliche” used in the tropical small-scale fishery of the Ciénaga Grande de Santa Marta, Colombia. Similar to covered cod-end hauls, an experimental net was enclosed by a control small-mesh net in 70 hauls. Both contact and available-selection processes affecting the selectivity of the “boliche” were modelled by fitting two selection curves: the logistic (a retention curve) and the encounter-probability (probability that a fish contacts the net) models. The encounter-probability model gave the most appropriate results in terms of the deviance statistic and the lowest Akaike's Information Criterion. Selectivity parameters were estimated for the main targeted fish species, highlighting differential avoidance reactions to the “boliche” for Eugerres plumieri (0.28), Diapterus rhombeus (0.24), Elops saurus (0.22), Micropogonias furnieri (0.13) and Cathorops spixii (0.16), but not Mugil incilis (all fish retained). To provide an overall (aggregated) effect of the “boliche” on the fish assemblage, the multi-species selection curve showed 61% retention at 15 cm, indicating that 39% of the multi-specific component may avoid the “boliche”.

Introduction

Millar and Fryer (1999) recently clarified the conceptualization of fishing gear selection processes and detailed the statistical framework to assess and obtain unbiased results in size-selectivity experiments. Estimates of gear selectivity by age and length-based population models are well documented in industrial fisheries, and used for estimating stock size and establishing management measures directed to reduce by-catch and discards (ICES, 1996, Caddy, 1999). However, the selective performance of fishing gears used in small-scale fisheries has been scarcely documented (Erzini et al., 1996, Stergiou et al., 1996, Stergiou and Erzini, 2002), particularly, in tropical estuaries. The proper interpretation of survey data is relevant in such fisheries, which have an overwhelming social value as sources of both food and income for local fishers (Blaber, 1997). These multi-species and multi-fleet fisheries are generally open-access with low operating costs, which make fish resources more susceptible to overfishing. In this setting, one critical point to improve management requires considerable technical changes to the gear to increase its selectivity (Caddy and Cochrane, 2001).

Based on the statistical framework developed by Millar and Fryer (1999) and following further developments by Zuur et al. (2001), this paper assesses the selection processes affecting catch efficiency of the encircling gillnet. This enclosure device is used in small-scale tropical and inshore fisheries from Latin America and the Caribbean, Africa and South East Asia (Arias, 1988, Nedelec, 1990, Trinidad et al., 1993), and is called “boliche” in the Ciénaga Grande de Santa Marta (CGSM), Colombia. The CGSM is one of the most important small-scale fisheries in this country due to its fish production, which supplies the northern Caribbean market. Of the fishing gears employed in this fishery, the “boliche” is one of the most intensively used, contributing up to 40% of the commercial landings in 1994 (Santos-Martínez and Viloria, 1998), and has been adapted as sampling gear in comprehensive fishing surveys (Rueda, 2001, Rueda and Defeo, 2001). The primary aim of this paper is to quantify the catch efficiency of the “boliche” in terms of probability of capture of fish species, based on an experimental approach. A secondary objective was to combine species-specific selectivity curves to demonstrate the overall effect of the “boliche” on a multi-specific basis.