This article was originally published in a journal published by Elsevier, and the attached copy is provided by Elsevier for the author’s benefit and for the benefit of the author’s institution, for non-commercial research and educational use including without limitation use in instruction at your institution, sending it to specific colleagues that you know, and providing a copy to your institution’s administrator. All other uses, reproduction and distribution, including without limitation commercial reprints, selling or licensing copies or access, or posting on open internet sites, your personal or institution’s website or repository, are prohibited. For exceptions, permission may be sought for such use through Elsevier’s permissions site at: http://www.elsevier.com/locate/permissionusematerial Fisheries Research 85 (2007) 68–73 Demersal fish density in the upwelling ecosystem off Colombia, Caribbean Sea: Historic outlook Camilo B. Garc´ıa a,∗, Luis O. Duarte b,c, Jairo Altamar b, Luis M. Manjarres´ b a Departamento de Biolog´ıa, CECIMAR/INVEMAR, Universidad Nacional de Colombia, A.A. 1016, Santa Marta, Colombia b Laboratorio de Investigaciones Pesqueras Tropicales, Universidad del Magdalena, Cra. 32 #22-08, Santa Marta, Colombia c Programa de Postgrado en Oceanograf´ıa, Departamento de Oceanograf´ıa, Universidad de Concepci´on, Casilla 160-C, Concepci´on, Chile Received 14 July 2006; received in revised form 11 December 2006; accepted 21 December 2006 Abstract We conducted a review of historic demersal fish biomass estimates (1970–2001) in the upwelling area off northern Colombia. A clear pattern of decline in demersal fish biomass has been identified. Total demersals and most selected species groups showed a negative trend in mean biomass. Partial recovery in biomass by year 2001 as compared to 1970 was detected, concomitantly with reductions in fishing effort by the shrimp trawling fleet. Triggerfishes and puffers showed an increasing trend in biomass probably as a result of good survivorship as discards from the shrimp fishery. These findings are discussed in the context of current general concerns on the health of world marine ecosystems. © 2007 Elsevier B.V. All rights reserved. Keywords: Biomass trends; Tropical fish; Upwelling; Colombia; Caribbean Sea 1. Introduction 1980–1984 period, declining afterwards to the end of their study period (1990–1994). In Colombian Caribbean waters the only Deterioration of ecosystems should ultimately be reflected published register of a decline in historic landings is that of in steady reductions of biomass (Pitcher, 1998). Recent global Garc´ıa and Solano (1995) with regard to Tarpon atlanticus. assessments of marine fish biomass conclude exactly that. For The present work compiles and presents historic changes instance, Myers and Worm (2003) estimate that large preda- (1970–2001) in demersal fish biomass in the upwelling area off tory fish biomass today is only about 10% of pre-industrial Colombia, Caribbean Sea (Fig. 1). Information from scientific levels. Watson and Pauly (2001) showed that global landings cruises which had the general purpose of estimating biomass lev- tended to decline by 0.36 million tonnes per year since 1988 els of demersal fishes, was compiled and analysed for biomass after correcting for misreporting to FAO. trends. Although there is a general concern about the conserva- tion status of Western Atlantic marine ecosystems (Jackson, 2. Methods 2001), regional or local assessments of fish biomass trends in the Caribbean Sea have been scarce. The only regional study A total of 13 cruises have been conducted in the last 30 years known to us that quantitatively explores catch trends is that of in the study area, several of them in the same year. Temporal Baum et al. (2003) that reviewed the status of sharks in the coverage of the cruises is uneven. Cruises were conducted in north Atlantic including the Caribbean Sea. They show that 1970 (3 cruises), 1988 (4 cruises), 1992 (2 cruises), 1995 (2 coastal and oceanic shark populations have rapidly declined. cruises), 1996 (1 cruise) and 2001 (1 cruise) yielding a total Caddy et al. (1998) found for FAO statistical area 31 including of 237 tows. Tow speed and duration, and depth, that may the Caribbean that landings of shelf dependent species (dem- influence composition of capture (Garc´ıa et al., 1998; Winger ersal fish plus commercialAuthor's benthic invertebrates) peaked in thepersonalet al., 1999; Manjarres´ et al.,copy 2001), varied short as well as mesh size (Table 1). In order to assure comparability among cruises, data are expressed as biomass density, i.e., kg per 2 ∗ Corresponding author. Tel.: +57 13165237; fax: +57 13165310. km , estimated by the swept area method. Because all research E-mail address: [email protected] (C.B. Garc´ıa). vessels were otter trawlers using similar scientific fishing gears 0165-7836/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.fishres.2006.12.003 C.B. Garc´ıa et al. / Fisheries Research 85 (2007) 68–73 69 Fig. 1. The upwelling ecosystem off Colombia showing tows carried out during the sampling dates. Data recorded in the 1990s (1992–1996) were grouped for the analysis. and simple random sampling was applied in all cruises with gies (e.g., Calamus penna), spadefishes (only Chaetodipeturs the only constraint posed by bottoms not amenable to trawling, faber), boxfishes (e.g., Acanthostracion quadricornis), grunts we assume a catchability of 1 in this calculation as we do not (e.g., Haemulon aurolineatum) and other demersals (remain- expect it to change over time in the context of this study. ing species, e.g., Polydactilus virginicus, Priacanthus arenatus, In the analysis species groups were preferred over individ- etc.). Species that were caught in the demersal nets but are ual species in order to reduce stochastic fluctuations. These considered of pelagic habit were excluded from analysis. groups constitute the bulk of demersal fish biomass, are of par- In order to picture time trends arithmetic mean biomass per ticular interest due to their alleged sensitivity to overfishing, year for each group were plotted against year. Data recorded like rays and sharks (e.g., Casey and Myers, 1998; Stobutzki et in the 1990s were grouped for this plot. Nonparametric 95% al., 2002; Baum et al., 2003), or support a long standing fish- confidence intervals of the biomass were derived by the bias- ing pressure, like snappers (Manjarres´ et al., 2004a,b). Species corrected and accelerated bootstrap method (Efron, 1987). groups included are: total demersals, snooks (e.g., Centropomus Lineal trends with time were sought by correlating (Pear- undecimalis), mojarras (e.g., Eucinostomus spp.), croakers (e.g., son’s coefficient) year and yearly mean biomass (no grouping Micropogonias furnieri), goatfishes (e.g., Upeneus parvus), of data recorded in the 1990s). Correlation between historic sharks (e.g., Rhizoprionodon porosus), triggerfishes (e.g., Bal- effort (days at sea) exerted by the shrimp fleet and yearly mean istes capriscus), puffers (e.g., Diodon holacanthus), snappers biomass was also investigated. Significance of correlation was (e.g., Lutjanus analis), rays (e.g., Dasyatis americana), por- tested via a bootstrap procedure that tests the null hypothesis Table 1 Historical scientific cruises carried out in the upwelling ecosystem off Colombia, Caribbean Sea Year Vessel Headrope length (m) Codend mesh (mm) Towing velocity (knots) Towing duration (h) Depth (m) Number of tows Source 1970 Choco´ 43.0Author's and 54.0 36 personal 2.5 0.1–2.0 copy 13–134 32 1 1988 Fridtjof Nansen 31.0 20 1.0–4.5 0.2–0.6 11–151 96 2 1992 Malpelo/Ancon´ 33.3/26.6 50/54 3.4–3.9 0.5–3.0 25–89 21 3 and 4 1995 Ancon´ 26.6 45 1.7–4.4 0.5–0.6 13–90 33 5 1996 Malpelo 33.3 50 2.9–4.5 0.5–0.6 14–146 16 5 2001 Ancon´ 26.6 45 2.8–3.8 0.5 10–88 39 6 Source:1,Testaverde and R´ıos (1972);2,Strømme and Saetersdal (1989);3,Quintero (1992);4,Zuniga˜ and Escobar (1993);5,Manjarres´ et al. (2005a,b,c);6, Manjarres´ (2002). 70 C.B. Garc´ıa et al. / Fisheries Research 85 (2007) 68–73 that the observed correlation value is to be expected by chance. Snooks were scarce, probably due to their rather near coast- Mean biomass of year 1970, the 1990s (data grouped for this estuarine habit (Orrell, 2002), and hence were not amenable to analysis) and 2001 was contrasted to each other via a bootstrap statistical analysis. Nevertheless, it is noticeable that in practical procedure that tests the null hypothesis that the observed differ- terms they disappear in the 1990s (Fig. 2). ence between means is to be expected by chance. Routines for Triggerfishes and puffers share the feature that year 2001 these tests were written in Resampling Stats®. biomass is significantly higher than 1970 biomass level (p < 0.05 in both cases; for puffers the contrast is against 1988 as in 3. Results 1970 puffers were registered in only one tow). For triggerfishes biomass level in 2001 is also significantly higher than in the Temporal change in mean annual biomass and their boot- 1990s (p < 0.05), whereas for puffers it is not (p > 0.05). This strap confidence intervals is shown in Fig. 2. The biomass of suggests that the increase in biomass for triggerfishes found in total demersals is high in 1970, gets somewhat reduced in 1988, 2001 has been abrupt, whereas in the case of puffers the increase declines severely in the 1990s and shows some recovery in 2001. has been more uniform. This pattern is followed by a number of the groups but not all For other demersals mean biomass in 1970 is significantly of them, suggesting that not all demersal fish respond the same higher than in the 1990s and 2001 (p < 0.05), whereas mean way to the vagaries of time.