A closer look at the bycatch of medium-sized and large sharks in the northern Catalan coast (north-western Mediterranean Sea): Evidence of an ongoing decline?

1. Historically, the Mediterranean Sea supported a rich shark fauna. Presently, however, populations of most shark species have largely due to intense fishing pressure. of trawlers, drifting longliners and longliners bycatch selected shark species — Alopias vulpinus , Cetorhinus maximus , Galeorhinus galeus , Hexanchus griseus , Isurus oxyrinchus , Mustelus Prionace glauca Squalus acanthias (89.7%) 2017, 89 of griseus confidence interval (CI) = 145, 34), of G. galeus (95% CI = 30, A. vulpinus

general decline of pelagic sharks in the Mediterranean Sea (Megalofonou, 2005;Ferretti et al., 2008;Fortibuoni et al., 2016;Colloca et al., 2020) and medium-sized and large shark species are seldom reported in the catch list of bottom trawlers operating in the western Mediterranean Sea (Massutí & Moranta, 2003;Massutí & Reñones, 2005), which suggests a severe reduction of the rich shark fauna reported by the 19 th century authors (Colloca et al., 2017).
The Mediterranean Sea hosts the highest proportion of threatened species due to unregulated fishing (Colloca et al., 2017). In a recent report published by the IUCN (Dulvy et al., 2016), 39 of the 73 assessed species of Chondrichthyans were found to be regionally threatened in the Mediterranean Sea, with overfishing identified as the main driver of decline and local extinction. Twenty-three of these 39 species were sharks; 12 fell under the category of 'Critically Endangered', six were considered 'Endangered', and five were 'Vulnerable.' Furthermore, 10 species were still listed as 'Data Deficient'. In the Spanish Mediterranean Sea, there remains a paucity of information regarding shark populations (García- Cortés & de la Serna, 2002;de la Serna et al., 2002;Massutí & Moranta, 2003;Massutí & Reñones, 2005), with recent research mostly focusing on the trophic ecology of a few species (Navarro et al., 2014;Albo-Puigserver et al., 2015;. In general, estimating parameters such as abundance or distribution of rare species is time-consuming and requires extensive funding (Anad on et al., 2009). In this context, the local ecological knowledge (LEK) held by fishing communities, taken as "a cumulative body of knowledge, practice and belief evolving by adaptive processes and handed down through generations by cultural transmission, about the relationship of living beings (including humans) with one another and with their environment" (Berkes, Colding & Folke, 2000: p.1252 has been used to obtain information about abundance and population trends of a wide range of species (Ferguson, Williamson & Messier, 1998;S aenz-Arroyo et al., 2005;Bender et al., 2014;Damalas et al., 2015;Colloca et al., 2020), as well as other biological and ecological information (Johannes, Freeman & Hamilton, 2000;Silvano et al., 2006;Stacey et al., 2012;Giovos et al., 2019). While some scientists and managers are rather sceptical about its reliability (Shackeroff & Campbell, 2007;Brook & McLachlan, 2008), LEK has proved to be a source of high-quality and low-cost information (Anad on et al., 2009) and numerous studies working on fisheries have shown that combining scientific assessments with LEK can improve management decisions (Garcia-Quijano, 2007;Silvano & Valbo-Jørgensen, 2008;Bender et al., 2014;Carothers et al., 2014).
The purpose of this study was to update the information about the presence of some medium-sized and large shark species inhabiting the Costa Brava, in north-eastern Spain (Figure 1) Figure 1). Its maritime relief is distinctive, with a narrow continental shelf and the shelf break found just a few kilometres offshore. The continental shelf and slope are dissected by three submarine canyons (Cap de Creus, Palam os, and Blanes) from north to south (Figure 1). They represent hotspots for marine life and are used as fishing grounds, mostly by the red shrimp (Aristeus antennatus) and Norway lobster (Nephrops norvegicus) fisheries.

| Questionnaire-based survey
Interviews were conducted from October 2016 to July 2017 at six of the 11 fishing ports in the area. These ports were selected because F I G U R E 1 Location of the Costa Brava (top, right; source: Google), surveyed ports and topographic representation of the marine relieve of the Costa Brava, along with its three main canyons, Cap de Creus (1), Palam os (2), and Blanes (3). Source: Institut Cartogràfic de Catalunya y GRC en Geociències Marines de la U.B. www.icc.es they represented the only ones that were used as a base by either longliners or bottom trawlers. Information was compiled by interviewing fishermen with the aim of covering at least 30% of the fleet of each fishing gear.
The interviews were carried out using a modified version of the questionnaires previously used to assess the bycatch of other species of megafauna in the Spanish Mediterranean Sea (Carreras, Cardona & Aguilar, 2004;Alvarez de Quevedo et al., 2010), which were adapted for sharks. At the start of each interview, the participants were given some identification forms and pictures of each shark species and then responded verbally to a set of questions concerning biological aspects of the species, catch frequency, factors influencing their bycatch, and the population trend of the different species.
Most specifically, fishermen were asked to identify the shark species they had seen and captured and to indicate the month of the For those species under any form of legal protection and whose population numbers were labelled as 'decreasing' in the interviews, potential factors responsible for the decline of these species were divided into three main categories: biological, fishing-induced, and other anthropogenic impacts. The combination of factors was also considered in the analysis and those that did not conform to any of these categories were classified as 'others'.

| Fishing effort and shark bycatch estimations
The fishing fleet includes three different types of vessels using three different fishing gears (bottom trawlers, bottom longliners, and drifting longliners), so a stratified approach was used to estimate total shark bycatch (Greenwood, 1996). Fishing effort was defined as the number of months in which each vessel operated on an annual basis, as in other studies using the same approach (Carreras, Cardona & Aguilar, 2004;Alvarez de Quevedo et al., 2010). According to Spanish fishing regulations, all bottom trawlers in the region are obligated to cease fishing for 2 months at different times of the year in each port, while longline vessels are not subject to this temporal closure. During the season, every bottom trawler operates from 7 am to 6 pm, from Monday to Friday, with 2-3 tows per day of work.
Total fishing effort with gear z (E z ) was calculated as: where E zi is the effort from vessels from the ith port operating with gear z.
In its turn, E zi was calculated as follows: where E ozi is the effort reported by the interviewed fishermen from port i operating with gear z, n zi is the number of registered vessels based in port i that used gear z, and n ozi the number of vessels from port i that used gear z and whose crew was interviewed. Total shark catch with gear z (C z ) was calculated as: where C oz is the number of sharks caught, as reported by fishermen with gear z; E z is the total fishing effort with gear z; and E oz is the effort reported by fishermen with gear z. Total shark catch by the whole fleet was the sum of the C z values of the three gears. The 95% confidence intervals of estimated catch were calculated with the procedure detailed by Greenwood (1996) for stratified sampling.

| Fishermen's perception on the evolution of shark populations
The questionnaires asked fishermen about their perception on the evolution of the population size of each species, with only three possible answers: increasing, stable and decreasing. A chi-squared test was carried out using IBM SPSS Statistics (Version 24) to assess whether the frequencies of the three possible answers (increase, stable, and decline) differed from those expected by chance (all three answers had the same probability by chance).

| Fleet coverage and occurrence of shark captures
A total of 42 full interviews were carried out and a minimum coverage of 30% was accomplished for each fishing gear-bottom trawlers (31 interviews; 40.8% of the fleet), bottom longliners (9 interviews; 45% of the fleet), and drifting longliners (2 interviews; 33.3% of the fleet; Table 1). The number of interviews for each fishing gear and fishing port are summarized in Table 1

| Fishermen's perception on the evolution of shark populations
All gears combined, results show that respondents mainly considered sharks to be suffering a decrease in abundance over time or, to a lesser extent, to remain at a stable number of individuals ( Figure 3).
The distribution of answers about the population trends of A. vulpinus, G. galeus, I. oxyrinchus, and P. glauca did not depart from that expected by chance (Table 3) and hence fishermen's answers were uninformative. The opposite was true for C. maximus and S. acanthias, both considered by respondents to be declining more often than expected by chance (Table 3). Likewise, the frequency to which respondents considered that the bycatch of H. griseus and Mustelus spp. is stable was also higher than that expected by chance (Table 3).

| Factors influencing shark captures
Respondents considered overlap between fishing grounds and shark habitat as the main reason for bycatch, followed by the attraction of sharks drawn by the fishing activity ( Figure 4).
Together, they accounted for the 82.3% of all the answers given by respondents.
At the time of the study, four species were legally protected by Spanish legislation-A. vulpinus, I. oxyrinchus, C. maximus, and G.
galeus-and each of them were differently affected by each factor ( Figure 5). According to respondents, fishing-induced mortality was regarded as the main factor causing the decline in I. oxyrinchus questionnaire-based surveys are one of the most practical methods to assess the bycatch of rare species (Godley et al., 1998). The compilation of data using such a practical low-cost methodology provided the present study with some insight that may pave the way for future species-specific studies and encourage local and regional authorities to develop new conservation strategies in order to tackle the issues facing shark populations off the Costa Brava.
Sharks are highly vulnerable to overfishing and bycatch because of their very low rates of population increase (Camhi et al., 2009 (Casey & Kohler, 1992;Peristeraki et al., 2008;Cortés et al., 2010;Mejuto et al., 2013). The fact that only a few specimens were captured off the Costa Brava might be further evidence of the severe decline of these large predatory species in the Mediterranean Sea, already reported by Ferretti et al. (2008). Bottom trawlers were involved in all the captures of C. maximus reported in this study.
C. maximus is also a pelagic species found at the top of the water column, but captures of this species can sometimes occur as some individuals get entangled in a trawl net when it is being hauled.
G. galeus was mainly captured by bottom trawlers but also by bottom longliners. Again, this is as expected for this species, taking into account that this shark usually inhabits waters deeper than those of the three species before-mentioned but shallower than those of bottom-dwelling sharks, such as H. griseus, making it susceptible to being caught by both trawl nets and bottom longlines.
Not a single capture of Mustelus spp. was recorded by the interviewed fishermen from 2016 to 2017. Mustelus spp. have suffered a drastic decline in the Mediterranean Sea not only in abundance but also in terms of spatial distribution, as its distribution range has diminished in recent decades due to heavy fishing according to (Colloca et al., 2017). The absence of captures of Mustelus spp. concurs with Colloca's findings and suggests the possibility that populations of these medium-sized sharks might no longer exist in the Costa Brava.
In the case of P. glauca and S. acanthias, the extremely high variability in all reported captures from 2016 to 2017 meant it was not possible to give a reliable estimation of their total captures. This high variability stems from the fact that only a few vessels reported captures: two for P. glauca, which reported having captured 50 and 200 individuals respectively, and one for S. acanthias, which reported having captured 90 sharks. Nonetheless, these results reveal that both species still occur in the region, although probably restricted to a few areas.
Hexanchus griseus was the most consistently bycaught shark by the fleet, yet almost exclusively by bottom trawlers. This species is most usually found on the continental slopes in the Mediterranean Sea (Stefanescu, Lloris & Rucabado, 1993;d'Onghia et al., 2004), Overlap between fishing grounds and species distribution has been identified in some studies as a major cause of shark bycatch (Perez & Wahrlich, 2005;Queiroz et al., 2016). Similarly, respondents in this study also acknowledged overlap as the main reason why shark bycatch occurs along with the belief that sharks get attracted to fishing activity.
Even though respondents considered there was no protected species the decline of which could be ascribed to a single factor, some factors did seem to be more relevant in explaining the dwindling numbers of some of these shark species. For example, fishermen perceived fishing-induced factors are greatly responsible for the decline of I. oxyrinchus. This coincides with other studies highlighting the dramatic decline of the species in the Mediterranean Sea (Ferretti et al., 2008;Colloca et al., 2017;Colloca et al., 2020), and moderate decline in the Atlantic Ocean (Baum et al., 2003) due to intense fishing. Other anthropogenic impacts, such as changes in water temperature due to global warming, pollution of the marine environment or water impoverishment, represented the main reasons perceived by respondents as to why C. maximus and A. vulpinus are disappearing. The results obtained for C. maximus agree with those of Sims & Reid (2002) and Cotton et al. (2005), who linked the decrease in shark catches with a decrease in the abundance of copepods and sea surface temperature respectively rather than fishing practices and thus differs from other studies that suggest C. maximus is mostly affected by overfishing (Stevens et al., 2000). Fishermen also attributed the depletion of A. vulpinus to other anthropogenic impacts, thus differing from quite a good number of studies that have considered A. vulpinus to suffer from intense fishing pressure (Baum et al., 2003;Ferretti et al., 2008;Goldman et al., 2009). Fishermen did not identify fishing as the main driver of decline for G. galeus either, as opposed to findings from other studies that attributed the depletion of this species to a long history of exploitation in target fisheries (Walker et al., 2006;Pondella & Allen, 2008). According to respondents, G. galeus seems to be declining for multiple reasons, with biological factors being much more relevant in this than in the other species. Some respondents stated that populations of G. galeus were less abundant in the Costa Brava because they had dispersed to other areas, yet no study has ever attributed the disappearance of G. galeus in a given area to migration to the best of our knowledge. In The designation of MPAs is a powerful approach to tackle the problem of biodiversity loss (Davidson & Dulvy, 2017) given the consequent range of benefits for marine wildlife (García-Charton et al., 2008). Amongst other top consumers, sharks have been used as focal species to designate MPAs given their population traits and distribution (Hooker & Gerber, 2004). Unfortunately, the limitation on data regarding spatial distribution of many shark species might lead to certain MPAs not being as effective in protecting sharks as they should be. Therefore, it is highly important to have reliable information for a given species when identifying areas of ecological importance and again fishermen could contribute to this by broadening the data baseline for many sharks. In a recent study, Giménez et al. (2020) assessed the usefulness of some current MPAs concerning the protection of certain species of demersal elasmobranchs and concluded that only a small part of them was actually relevant for the demersal shark community, since the majority of them were distributed outside the existing MPAs. This conclusion is also relevant for the species studied here, as current MPAs are too small to protect viable populations.
In any case, keeping a synergistic relationship with the fishing sector can for instance engage fishermen in providing continuous data on any kind of megafauna. This has proved to be fruitful in the past in some fisheries (Ticheler, Kolding & Chanda, 1998;Obura et al., 2002;Yochum, Starr & Wendt, 2011). Data on shark abundance or distribution would be of high value especially in areas like the Costa Brava, where the status of some sharks is still far from being wellknown. Fishermen could not only provide data on protected species when these got bycaught, but also on commercial species such as H. griseus, Mustelus spp., P. glauca, and S. acanthias, which are not regulated under a TAC/quota management scheme in the Mediterranean Sea. This constant flow of information coming from fishermen could help monitor the population status of such species as well, making it easier to detect changes in abundance over time and therefore allowing scientists and managers to develop conservation measures to reverse any detected decline.