Effects of mesh size and towing speed on the multispecies catch rates of historical swept area surveys
Introduction
In general, research survey cruises for assessing demersal fish stocks follow specific protocols that try to minimize the effect of the variables that are not considered in the calculations inherent to the swept area method. It is, however, relatively common that some technical and operational characteristics vary between historical surveys in a region. One of these technical variables is the mesh size used at the codend of the net. This variable is crucial in determining the escape or loss of fish through the nets (Ragonese et al., 2001, Weinberg et al., 2002, Weinberg and Kotwicki, 2008). A second operational variable to take into account is towing speed (Dahm et al., 2002, Weinberg et al., 2002), which may vary significantly over a survey, due to many uncontrollable factors, generally environmental, that modify the functioning of the net (Weinberg et al., 2002, Herrmann, 2005a, Herrmann, 2005b, Duarte and Cuello, 2006). The impact of this variable, as well as that related to the calculation of the swept area, is essential in three basic processes of bottom trawling: (i) horizontal herding, stimulated by sand clouds and the bridles; (ii) vertical herding, in response to stimuli such as the headrope or boat noise; and (iii) fish loss or escapement under the footrope (Weinberg et al., 2002, Weinberg and Kotwicki, 2008). Generally, the effect of these aspects is analysed by a video camera attached to the trawl (Somerton and Weinberg, 2001), since it is difficult to define a measurable variable that can, by itself, reflect the effect of the horizontal herding or the vertical herding.
Other factors that generally vary between historical surveys are tow duration and net size. Although tow duration is a variable considered in the swept area method (Ye et al., 2005, Catalán et al., 2006), it has been proposed that this variable also has an influence on the size structure of the catch, on the basis that the largest individuals are able to swim ahead of the net for longer periods of time (Sparre and Venema, 1998). However, several studies have failed to find a significant relationship between tow duration and catch size distribution (Godo et al., 1990, Walsh, 1991, Somerton et al., 2002). According to Godo et al. (1990), this is due to the fact that at the beginning of the tow, before a school is established at the mouth of the net inducing an alert reaction in the fishes, the effect of exhaustion is less important than that of surprise.
With respect to net size, the horizontal opening of the trawl net is considered through the calculation of the swept area (Ye et al., 2005, Catalán et al., 2006, Bergstad et al., 2008). Regarding the vertical opening of the trawl net, in controlled experiments with two different heights of vertical opening, Johnson et al. (2008) did not find statistically significant differences in the number of individuals of the main fish taxa that were caught, or the structure and composition of the assemblages. In contrast, the mean catch per unit effort (CPUE) was greater in the net with a larger opening height, though this trend was most clear for certain pelagic fish taxa that form large shoals.
The ability to use fish catch rates obtained from scientific surveys as relative abundance indices depends on the feasibility of removing effects other than abundance, a process that has been described as “standardization of catch/effort” (Maunder and Punt, 2004). Therefore, we should analyze the effect on the catch rates of those factors that influence the trawl performance as a prerequisite for properly calculating estimates of relative abundance that reflect changes in population distribution and density (Weinberg and Kotwicki, 2008).
It is known that behaviour and swimming performance of different species can vary greatly (Winger et al., 2004, Queirolo et al., 2012) and therefore it would be recommendable to analyse the effect of the mesh size and towing speed at the species level. However, it is also known that analysis of the status of tropical demersal fish resources is usually made at multispecies level, supplemented with analysis of species composition, due to the extremely high species diversity that characterizes this type of surveys in tropical areas (Blaber et al., 1994, Sparre and Venema, 1998). In fact, the fish bycatch of the shrimp trawl fishery operating in the Colombian Caribbean Sea (CCS) is composed of 175 taxa distributed in 58 families (Viaña et al., 2004), while the database of the surveys carried out in the CCS between 1988 and 2001 reports a total of 311 taxa of demersal fish (Duarte and Cuello, 2006). Therefore, the objective of this study was to analyze the effect of the mesh size and towing speed on the multispecies catch rates in historical demersal fish assessment surveys performed in the CCS. Both generalized linear models (GLMs) and generalized additive models (GAMs) offer a powerful tool for achieving this objective, since these models allow more flexible linear predictors as well as automatically control the parametric complexity (Venables and Dichmont, 2004). However, we chose GLMs because a significantly better fitting was not obtained using GAM models, in spite that this technique includes local smoothers as terms.
Section snippets
Data sources
We used data collected from eight fishery-independent trawl surveys for the assessment of demersal fish stocks conducted in the CCS during the period 1988–2001, which were all based on the swept area method. Specifically, the study was based on the four surveys carried out in 1988 under the framework of the NORAD-UNDP/FAO programme (Strømme and Sætersdal, 1989), the three in 1995 and 1996 under the INPA-VECEP/UE programme (Manjarrés et al., 2005a, Manjarrés et al., 2005b, Manjarrés et al., 2005c
Effects of the mesh size and towing speed
The variables mesh size and towing speed had a minimal effect on the multispecies catch rates of demersal fish, compared to the spatial and temporal variables considered in the models. The combined relative contribution of these two variables to the overall deviance (49.10% in model 1 and 37.01% in model 2) was no more than 3% in either case. In model 1, the deviance explained by mesh size was even less than that explained by towing speed, the reverse being true in model 2 (Table 4). Both the
Effect of the mesh size on the catch rates
The assessment of the effect of the mesh size using GLMs led to different results to those obtained by an independent assessment of the effect of this variable, confirming the importance of taking into account the combined effect of all factors that may potentially influence catch rates. Using GLMs, we found that the mesh size does not have a very strong effect on the multispecies catch rates from the demersal fish surveys carried out over the last two decades in the CCS. Specifically, the GLMs
Acknowledgments
This study was undertaken with the support of the projects “Dinámica espacio-temporal del ecosistema de afloramiento del área Bocas de Cenizas-Punta Espada y sus implicaciones para un régimen de pesca responsable”, (COLCIENCIAS Grant 3135-09-11245), in addition to the Universidad del Magdalena, INCODER, and the Universidad Nacional de Colombia. This is also the contribution nº75 from the CEIMAR Journal Series. We would also like to express our appreciation to Jairo Altamar, Luis Duarte Casares
References (64)
- et al.
The Guajira upwelling system
Cont. Shelf Res.
(2005) - et al.
Demersal fish on a mid-ocean ridge: distribution patterns and structuring factors
Deep-Sea Res. II: Top. Stud. Oceanogr.
(2008) - et al.
Spatial and temporal changes of coastal demersal assemblages in the Gulf of Cadiz (SW Spain) in relation to environmental conditions
Deep-Sea Res. II: Top. Stud. Oceanogr.
(2006) - et al.
On the influence of towing speed and gear size on the selective properties of bottom trawls
Fish. Res.
(2002) - et al.
Demersal fish density in the upwelling ecosystem off Colombia, Caribbean Sea: historic outlook
Fish. Res.
(2007) - et al.
Effect of tow duration on length composition of trawl catches
Fish. Res.
(1990) - et al.
Application of generalized linear modeling to catch rate analysis of Western Mediterranean fisheries: the Castellón trawl fleet as a case study
Fish. Res.
(1999) - et al.
Generalized linear and generalized additive models in studies of species distributions: setting the scene
Ecol. Model.
(2002) - et al.
A generalized linear mixed model analysis of a multi-vessel fishery resource survey
Fish. Res.
(2004) Effect of catch size and shape on the selectivity of diamond mesh cod-ends. I. Model development
Fish. Res.
(2005)
Effect of catch size and shape on the selectivity of diamond mesh cod-ends. II. Theoretical study of haddock selection
Fish. Res.
Sampling estuarine fish and invertebrates using demersal otter trawls: effects of net height, tow duration and diel period
Fish. Res.
A new improved method to compute swept area estimates of biomass from commercial catch rate data: application to Namibian orange roughy (Hoplostethus atlanticus)
Fish. Res.
Standardizing catch and effort data: a review of recent approaches
Fish. Res.
Analysis of catch per unit effort by multivariate analysis and generalised linear models for deep-water crustacean fisheries off Barcelona (NW Mediterranean)
Fish. Res.
Standardization of catch and effort data in a spatially-structured shark
Fish. Res.
The effect of speed through the water on footrope contact of a survey trawl
Fish. Res.
Can changes in tow duration on bottom trawl surveys lead to changes in CPUE and mean size?
Fish. Res.
MLZs, GAMs and GLMMs: an overview of theory for applications in fisheries research
Fish. Res.
Factors influencing net width and sea floor contact of a survey bottom trawl
Fish. Res.
The effect of trawl speed on the footrope capture efficiency of a survey trawl
Fish. Res.
Use of generalized linear models to analyze catch rates having zero values: the Kuwait driftnet fishery
Fish. Res.
Constructing abundance indices from scientific surveys of different designs for the Torres Strait ornate rock lobster (Panulirus ornatus) fishery, Australia
Fish. Res.
Distribution, biomass and community structure of demersal fishes of the Gulf of Carpentaria, Australia
Aust. J. Mar. Freshwater Res.
An Introduction to Generalized Linear Models
Hacia un esquema integrado para el almacenamiento y análisis de información oceanográfica, ecológica y pesquera: Sistema de Información Evaluación y Ecología Pesquera, SIEEP
Intropica
Sistema de información evaluación y ecología pesquera, SIEEP (versión 1.1). Actualización para estudios de la captura acompañante en pesquerías de camarón
Intropica
Variabilidad circadiana de la tasa de captura y la estructura de tallas en camarones e ictiofauna acompañante en la pesquería de arrastre del Mar Caribe de Colombia
Invest. Mar.
Effects of bridle length on catch rates of hake
Cruise Report No 2/2000
Tasas de captura estandarizadas como índice de abundancia relativa en pesquerías: enfoque por Modelos Lineales Generalizados
Invest. Mar.
Cited by (5)
A review of bycatch reduction in demersal fish trawls
2021, Reviews in Fish Biology and FisheriesOptimization remould and experiment on DX-180 fiber reinforced plastic trawler for scallop
2017, Nongye Gongcheng Xuebao/Transactions of the Chinese Society of Agricultural EngineeringGeneralisation of Mood's Test to Factorial Designs to Detect Heterogeneity of Variances
2015, Australian and New Zealand Journal of Statistics