Introduction

Large marine vertebrates such as marine mammals, sea turtles and chondrichthyans are highly vulnerable to non-natural mortalities resulting from anthropogenic activities, especially fisheries (Lewison et al. 2004; Read et al. 2006; Žydelis et al. 2009). This is a result of the mainly k-selected life history displayed by these species groups: comparatively long-life, high natural survivorship, slow growth, late maturity and low fecundity. Chondrichthyans, comprising the chimeras and elasmobranchs (sharks and batoids), marine mammals (specifically cetaceans and sirenians) and sea turtles represent some of the most threatened animal groups (Online Resource 1). Using IUCN Red List criteria, both marine mammals and sea turtles represent a relatively small number of species (92 and 7, respectively) with high levels of vulnerability. In contrast, chondrichthyans combine large numbers of species (546 batoids, 475 sharks and 46 chimeras) and high levels of vulnerability with the highest proportions of Data Deficient and lowest of Least Concern status of any vertebrate class (Dulvy et al. 2014; Hoffmann et al. 2010; IUCN 2016). These species provide vital marine ecosystem services at various levels. As apex and meso-predators across a number of food webs they affect community structure and dynamics (Heithaus et al. 2008; Kiszka et al. 2015), and as grazers impact seagrass systems and nutrient cycling (Aragones et al. 2006; Burkholder et al. 2013; Preen 1995). Therefore, the loss of vulnerable marine megafauna has potential consequences for ecosystem structure and function, with implications and impacts across multiple spatiotemporal scales.

The complex interrelationships between marine megafauna and human impacts on the marine ecosystem make simultaneously managing the use of marine resources and protection of these species especially challenging. Fisheries are widely considered the greatest threat to vulnerable marine megafauna (Dulvy et al. 2014; Lewison et al. 2004; Read et al. 2006; Wallace et al. 2010). Many are non-target species, widely perceived to be of low value and are often viewed as a nuisance by fishermen, especially in industrial fisheries. In contract, others, mainly elasmobranchs, are targeted in a range of coastal and oceanic fisheries, particularly for their fins and other products including meat and gill plates (Couturier et al. 2012; Musick 2005). For many fisheries, particularly small-scale fisheries in developing nations, vulnerable marine megafauna species may constitute both target and non-target catch. Indeed, their categorisation as target or by-catch species may vary on a fisher-by-fisher and trip-by-trip basis. As such we herein refer to their presence in the fisheries simply as ‘catch’.

The multi-gear nature of many fisheries, the perceptions of many vulnerable marine megafauna as either a nuisance or of low value, together with the illegality of catching certain species and the sometimes-secretive nature of fishermen mean that catch is largely under-reported and data are sparse in many regions, making accurate estimation of global catch exceedingly difficult. However, available estimates indicate that catches are likely unsustainable, with an estimated 0.53–0.82 million marine mammals, 0.85–8.5 million sea turtles and 63–273 million sharks caught worldwide annually (Read et al. 2006; Wallace et al. 2010; Worm et al. 2013). Gillnet and line fisheries account for the majority of marine mammal, elasmobranch and sea turtle catch (Lewison et al. 2004; Read et al. 2006). These fishing methods are relatively inexpensive, simple and effective with widespread usability.

While vulnerable marine megafauna interactions with industrial and commercial fisheries have received some attention, less is known of the magnitude and mechanisms of interaction with small-scale fisheries, herein defined as those fisheries operating either for subsistence or for income generation (artisanal) but not as part of a commercial company, particularly in the developing regions. Globally, small-scale fisheries include some 50 million fishers (FAO 2016c), more than 95% of fishers worldwide (Pauly 2006). They are especially prevalent in the developing regions of South and Central America, Africa and the Indo-Pacific. Given their prevalence and widespread occurrence the environmental impacts of small-scale fisheries are likely significant, though they are often overlooked (e.g. Hawkins and Roberts 2004; Moore et al. 2010; Salas et al. 2007). With continued unregulated exploitation, small-scale fisheries can negatively impact the abundance, distribution and species composition of vulnerable taxa (Pinnegar and Engelhard 2008), including vulnerable marine megafauna. Thus, small-scale fisheries may lead to declines of these key species with consequences for the broader food web and ecosystem, including other species that are critical to local livelihoods.

In developing regions small-scale fisheries are of considerable socio-economic importance, particularly in rural areas where they are important contributors to the local economy (Béné 2006; Pauly 2006) and to food security. In these regions elasmobranchs, sea turtles and marine mammals were historically important sources of human sustenance and remain so in many areas (Robards and Reeves 2011; Vannuccini 1999). Elasmobranchs are most important in this respect, but hunts still exist for both sea turtles and marine mammals, often in spite of national or international laws and regulations banning these practices (Cerchio et al. 2009; Hart et al. 2013; Humber et al. 2014; Kasuya 2007; Riedmiller 2013). Vulnerable marine megafauna are an important source of income, both in fisheries and increasingly from ecotourism activities (Cisneros-Montemayor et al. 2013; O’Connor et al. 2009). However, there is a lack of information regarding the non-monetary, including cultural, value of vulnerable marine megafauna to fishers, which has implications both for decision-making regarding catch and conservation of these species and for the full understanding of their societal value.

Our aim in this paper is to review existing knowledge and status regarding vulnerable marine megafauna interactions with marine small-scale fisheries in the southwestern Indian Ocean (SWIO) region. Data were gathered from a range of sources including, but not limited to, information requests from relevant government departments in SWIO nations, scientific and non-governmental organisations (NGO) publications and reports, international and national annual reports and databases. We discuss the likely implications of the current situation, highlighting vital knowledge gaps that need to be addressed and challenges for future research and management across the region.

Southwestern Indian Ocean profile

The SWIO, as considered in this review, consists of 8 countries (and their Economic Exclusive Zones) with broadly comparable fisheries: Comoros, Kenya, Madagascar, Mauritius, Mayotte and La Réunion (France), Mozambique, the Seychelles and Tanzania (including Zanzibar) (Fig. 1, Online Resource 2). The region’s human population is projected to more than double from 155 million to 357.3 million by 2050 (WB 2016b), with coastal cities and island nations experiencing particularly high rates of population increase (Online Resource 2). Food security and income generation are therefore major policy drivers requiring sustainable solutions built on sound management practices. Coastal communities have traditionally relied on marine fishes (including elasmobranchs) as their main sources of protein, with some also making use of marine mammals and sea turtles for sustenance or as bait (Church and Palin 2003; Humber et al. 2011; Razafindrakoto et al. 2008). Marine fisheries (including mariculture) account for 0.5–30% of Gross Domestic Product (GDP), with island nations particularly reliant (Online Resource 2). However, other marine income-generating activities also contribute to local economies, particularly marine-tourism activities, worth around $3.95billion, including recreational fishing, whale and dolphin-watching and dive tourism (Amir and Jiddawi 2001; Divetime 2016; Gallagher and Hammerschlag 2011; Obura 2017; O’Connor et al. 2009; O’Malley et al. 2013; Pérez-Jorge et al. 2016). Some countries in the SWIO, like the Seychelles, are focusing on expanding their other food security sectors (e.g. mariculture) to reduce reliance on vulnerable fisheries especially in the face of climate change (Stead et al. 2015).

Fig. 1
figure 1

Sources: ESRI (2014), VLIZ (2014)

The southwestern Indian Ocean: Comoros, Kenya, Madagascar, Mauritius, Mayotte and La Réunion (France), Mozambique, Seychelles and Tanzania (mainland and Zanzibar).

Full size image

The SWIO (part of FAO Fishing Area 51) has among the highest marine species richness worldwide (Tittensor et al. 2010; Worm and Branch 2012). This diversity is threatened by increasing anthropogenic pressure, especially from fisheries, and limited management effectiveness where it exists (Mora et al. 2009; Worm et al. 2013). The area contains a high diversity of vulnerable marine megafauna species but there is large uncertainty regarding status, catch and trends of many of these at both a global (Online Resource 1) and regional scale (Kiszka 2015; Kiszka and van der Elst 2015). These uncertainties, coupled with the high proportion of vulnerability and mostly decreasing trends (Online Resource 1) in assessed species is a major concern.

The risk to vulnerable marine megafauna species in the SWIO is further exacerbated by the continued expansion of the dominant yet largely poorly documented and unregulated small-scale fisheries. Small-scale fisheries officially account for 75–85% of marine landings across SWIO nations (Pauly and Zeller 2015), with annual landings reportedly 345,000–390,000 mt as of 2014 (FAO 2016b). However, independent estimates suggest gross under-reporting of landings and effort, with total SWIO landings estimated to average 165% greater than reported figures for 1950–2010 (Fig. 2; Pauly and Zeller 2015). Although these are retrospective estimates, they provide an improved assessment of the landings magnitude and serve as a useful reference point.

Fig. 2
figure 2

Data sources: FAO (2016b), Pauly and Zeller (2015)

Official and estimated landings data for SWIO nation’s fisheries between 1950 and 2013.

Full size image

Small-scale fisheries of the southwestern Indian Ocean: main features and data quality

Currently the SWIO small-scale fisheries employ more than 495,000 fishers operating 150,000 assorted vessels across the SWIO (Table 1), with the largest fleets in Madagascar and Mozambique. However, this does not account for many of the unlicensed small-scale fisheries fishers, which are of substantial number (Teh and Sumaila 2013). Unlicensed and open-access fishing is a major issue for SWIO small-scale fisheries, with direct implications for the assessment of catch and socio-economic value of these fisheries and so inhibiting effective stock management. Additionally, small-scale fisheries support various other livelihoods, including: auctioneers, fish mongers, middlemen, gear repairers and fish fryers among others.

Table 1 Marine small-scale fishery vessel, fisher and gear data for South Western Indian Ocean nations.
Full size table

In terms of fisher participation the regional fisheries are dominated by handlines (Table 1), with simplicity, ease of use and affordability as likely drivers. However, more advanced gears are increasingly used. Specifically, fisheries in Kenya, Madagascar, Mozambique, the Seychelles and Tanzania (including Zanzibar) are using longline (demersal and pelagic) and gillnet (drift and bottom set) gears (Table 1), mostly targeting sharks and pelagic fishes.

Assessing general landings trends in the SWIO small-scale fisheries is challenging. Whilst long-term data sets are available through the FAO (Fig. 2) and national reports, the validity and quality of these are questionable given the lack of standardised and systematically collected historical data, particularly regarding effort, and unlicensed fishing. However, if we consider only data from recent years (Fig. 3; KMALF 2015; MFR 2012; SFA 2015; L’Institut Français de Recherche pour l’Exploitation de la Mer unpublished data; Tanzania Ministry of Livestock and Fisheries Development unpublished data; Seychelles Fishing Authority unpublished data) landings appear relatively stable in most nations, with a marked overall regional increase (Fig. 2). Decreasing trends are seen in the official data from Mauritius and La Réunion, reflecting the declining effort (vessel numbers and fisher days respectively) in these fisheries (MFR 2010; L’Institut Français de Recherche pour l’Exploitation de la Mer unpublished data). Conversely a rapid increase in official landings has been observed for Mozambique, likely driven by improvements in both monitoring programmes and proper extrapolation of data to the national level (Doherty et al. 2015).

Fig. 3
figure 3

Data sources Pauly and Zeller (2015)

SWIO small-scale fisheries landings 2000–2010. Note different scales on the y-axes.

Full size image

Compounding the issues regarding monitoring efficacy and accuracy is the widespread commonality of national and international migrant fisheries in the region (WIOMSA 2011). These catches may be taken in one nation and declared in another, declared in both or in neither. Undeclared transhipment of catches to neighbouring markets is also common. For example catches in Zanzibar are often landed, compiled and shipped directly to markets in mainland Tanzania or southern Kenya, often following a seasonal pattern (Fowler et al. 2005; Wanyonyi et al. 2016; A. Temple personal observation). These may have significant impact on landings data and could have consequences for stock management.

The biggest stumbling block in the monitoring and management of SWIO small-scale fisheries is the lack of standardised data and the relatively poor resolution of landings and effort data available. The variability in basic recording metrics (Online Resource 3) hinders the comparability and summation of data at the regional scale, with only simple measures feasible for use i.e. effort can only be regionally derived through vessel count data. Measures and definitions of gear type, effort metrics and vessel types differ between countries and even within countries between years, whilst the breakdown of data by geographic region, gear and vessel types is often inconsistent over time. Data reports have variable formats and contents, and are often unclear as to whether data presented are that observed or whether they have been extrapolated to country level. Most notably data are inconsistent among reports. For example, there are a number of years (1990, 1993, 1994, 2000 and 2007) where artisanal handline and troll line numbers reported by the Seychelles to the Indian Ocean Tuna Commission (IOTC) are greater than the numbers reported by the Seychelles Fishing Authority for its entire artisanal fleet (IOTC 2016a; SFA 2001–2013). Similarly Kenya reports 165 mt of all elasmobranchs landed from artisanal fisheries in its 2007 statistical bulletin, yet reports 174 mt of sharks alone from the same fishery and year to the IOTC (IOTC 2016a; KMALF 2008). Clearly, these issues must be addressed, both at national and the regional levels, if small-scale fisheries are to be sustainably managed and vital livelihoods protected across coastal areas of the SWIO region.

Marine mammal interaction with small-scale fisheries

There is limited information available on marine mammal populations and their interaction with the small-scale fisheries of the SWIO (Kiszka 2015; Kiszka et al. 2009), but where data exists there is evidence of both targeted and incidental catch. Catches, mostly incidental, have been documented in the Comoros and Mayotte (Kiszka et al. 2007, 2010; Poonian et al. 2008; Pusineri et al. 2013; Pusineri and Quillard 2008), Zanzibar and Tanzania (Amir 2010; Amir et al. 2002; Muir and Kiszka 2012), Kenya (Kiszka 2012), Madagascar (Cerchio et al. 2009; Razafindrakoto et al. 2004, 2008), Mozambique (Guissamulo and Cockcroft 1997; Kiszka 2012) and La Réunion (Kiszka et al. 2009). To date no marine mammal catch has been reported in Mauritius or the Seychelles (Kiszka et al. 2009). Published studies identify coastal gillnet fisheries (both drift and set nets) as the main threat to marine mammals across the region, although interactions have also been documented in longline fisheries (Kiszka et al. 2009, 2010).

Understanding the true impacts of these catches requires data on capture rates and abundance estimates for any given population investigated. To date population abundances have only been estimated for two cetacean species in restricted areas of the SWIO: Indian Ocean humpback (Sousa plumbea) and Indo-Pacific bottlenose dolphins (Tursiops aduncus) in the Menai Bay Conservation Area off the south coast south of Unguja Island, Zanzibar and in the Kisite-Mpunguti Marine Protected Area (MPA), Kenya (Meyler et al. 2012; Pérez-Jorge et al. 2015, 2016; Stensland et al. 2006); and T. aduncus off the south-west of Mauritius (Webster et al. 2014), around Mayotte (Pusineri et al. 2014) and La Réunion (Dulau et al. 2017). Capture rate estimates are only available for Zanzibar, with these showing unsustainable levels of fisheries mortality for both species (Amir 2010; Amir et al. 2002). Numbers of the dugong (Dugon dugon) have significantly reduced across the SWIO region with only relict populations remaining in the region, the largest of which exists in Mozambique (WWF-EAME 2004). Evidence of ongoing catches has led to serious concern for the future viability of this species (Kiszka 2015; Kiszka et al. 2007). Of further concern is the on-going illegal hunt for marine mammals in Madagascar (Cerchio et al. 2009; Razafindrakoto et al. 2008) and Tanzania (Riedmiller 2013), possibly in Mayotte (Kiszka et al. 2009) and likely other parts of the region.

Currently there are no annual statistics relating to the catch or landings of marine mammals in the SWIO region. Minimal attention is given to these species as a component of the fisheries at a national level and there is likely an inherent reluctance to report any such catch given its illegality. Yet, this is hardly a problem restricted to this region, rather it is one at the global level.

Sea turtles interaction with small-scale fisheries

Five species of sea turtles are known to occur in the SWIO, but green (Chelonia mydas), loggerhead (Caretta caretta) and hawksbill turtles (Eretmochelys imbricata) are the most common and widely distributed in the region (Bourjea 2015). Sea turtles have attracted both long-term and intensive studies in the SWIO relative to other vulnerable marine megafauna species. Nevertheless, there are still major data gaps e.g. unreliable nesting data and a lack of species abundance estimates, partially as a result of their highly mobile and complex life history preventing comprehensive population level assessment for most species in the SWIO region (Bourjea 2015). However, qualitative global assessments rank loggerhead, leatherback (Dermochelys coriacea) and olive ridley (Lepidochelys olivacea) as high risk species in the western Indian Ocean, with olive ridley and green considered to face the greatest levels of threat to survival (Wallace et al. 2011, 2013).

Three gear-types have been identified as catching substantial numbers of sea turtle, namely gillnets, prawn/shrimp trawls and longlines (Bourjea et al. 2008; FAO 2010; Wallace et al. 2013). Yet, in most countries of the region, the extent and impact of fisheries on sea turtles is poorly known, except for open ocean fisheries (Bourjea et al. 2014). Both incidental and targeted catch of sea turtles appear widespread in small-scale fisheries (for review see Bourjea (2015) and Bourjea et al. (2008)), with the threat posed by gillnet and line gears across the region to sea turtles well established (Bourjea et al. 2008; Kiszka et al. 2010; Poisson and Taquet 2001; Poonian et al. 2008). However, there are no annual statistics of note for sea turtle capture in the SWIO. Kenya is the only reporting nation, for which it has reported 0 mt/year since 1964 (FAO 2016b). Whilst other sources of quantitative data are sparse, and for areas where data exist they are rarely comprehensive, it appears annual regional small-scale fisheries catch is in the order of tens or even hundreds of thousands (Table 2), representing a serious threat to the survival of sea turtles in the SWIO. This is compounded further by alterations and destruction of nesting beaches in some areas, sizeable egg poaching activities and hunting of nesting females, which are common across the region (Bourjea 2015).

Table 2 Existing numerical data for sea turtle catch in the small-scale fisheries of the South Western Indian Ocean.
Full size table

Elasmobranch interaction with small-scale fisheries

In 2014, 34 countries across all fisheries scales reported 105,969 mt of elasmobranch landings originating from the western Indian Ocean region, of which only 17,663 mt were landed by SWIO nations (Fig. 4; FAO 2016b). Of this a disproportionate amount (89.6%) was accounted for by Tanzania (including Zanzibar) and Madagascar, which together account for 62% of known SWIO small-scale vessels (Table 1). It is therefore unlikely that the reporting reflects the true proportional contribution of SWIO nations. The regional estimate is likely an underestimate, resulting from under-reporting of landings, illegal fishing and discards, and is consistent with the under-reporting of other landings in the SWIO region (Pauly and Zeller 2015). Despite the high level and year-on-year increase in landings across the SWIO (Fig. 4) little independent research has been undertaken on these fisheries.

Fig. 4
figure 4

Data Source: FAO (2016b)

Total landings data for elasmobranchs caught by South Western Indian Ocean nations 1970–2013. Zanzibar and Tanzania are reported separately after 2000.

Full size image

There is a notable imbalance (>80% of peer-reviewed papers) between studies focussing on elasmobranchs in the industrial and semi-industrial fisheries e.g. (Fennessy 1994; Huang and Liu 2010; Kiszka and van der Elst 2015; Romanov 2002) and the small-scale fisheries (Molina and Cooke 2012). Published information for small-scale fisheries is generally sparsely quantified and limited to target species (e.g. Marshall 1997b; McVean et al. 2006; Schaeffer 2004). Grey literature, in the form of unpublished theses, governmental and consultancy reports, data collected by NGO’s and other such works exist. However, much of this information is not easily accessible and in some cases remains confidential. This highlights a need to better the flow of information to responsible organisations and into the public domain.

A variety of species are regularly taken in the region’s fisheries, with the most commonly reported being blue (Prionace glauca) and silky (Carcharhinus falciformis) sharks from the industrial longline and purse seine fisheries (Smale 2008). Concurrently, a number of species are known to appear in the region’s small-scale fisheries (Table 3). The species listed are influenced by ease of identification and observation bias. As such they are unlikely to accurately reflect fisheries composition but do provide evidence for a level of regional species homogeneity. Most elasmobranch landings in the SWIO region are not identified beyond basic taxonomic level and are simply grouped as “sharks and/or rays” (FAO 2016b). Thus, the data cannot support effective stock management, at either local or regional levels. This is a major constraint to decision-making on fisheries management measures as there is no reliable data for population dynamics, given that that small-scale fishers are reporting significant declines in elasmobranch abundance and catch this demonstrates a clear information gap (Kiszka and van der Elst 2015). Specific data is required to properly document both which species, and in what volume, elasmobranchs are interacting with SWIO small-scale fisheries. Below, current understanding of elasmobranch catch and landings in SWIO small-scale fisheries is summarised by country:

Table 3 Chondrichthyan species common amongst the reported small-scale fisheries landings in the SWIO, La Réunion has been omitted as no data is available.
Full size table

Comoros

Since first being reported in 1994, elasmobranch landings in the Comoros have dwindled from 230 to 19 mt by 2013 (FAO 2016b), despite the apparent increases in fishing effort (IOTC 2016a; Soilihi 2014). Import of dried shark meat from Madagascar (Cooke 1997) is one possible driver, with Comorian fisheries known to be unable to meet the domestic demand for dried fish products (WB 2016a). Alternatively, these declines may reflect stock collapses. Blue shark is a major constituent, accounting for 26% of the 19.97 mt landings in 2012 (Soilihi 2014). Other commonly reported catch includes a variety of oceanic and coastal species; primarily oceanic whitetip (Carcharhinus longimanus), silky, grey reef (C. amblyrhynchos) and hammerhead (Sphyrna spp.) sharks (Maoulida et al. 2009).

Based on reported landings and gear composition from IOTC National Reports (Soilihi 2014), effective catch per unit effort (CPUE) for sharks is much higher in both static (775 kg/vessel/year) and trolled (830 kg/vessel/year) handlines than in net fisheries (2.20 kg/vessel/year). Indeed, handlines account for nearly 96% of the reported blue shark landings. Given the seemingly greater CPUE for sharks in the handline fisheries specific scrutiny should be placed on these in any future assessments. Conversely, the lack of reporting of batoid catches and/or landings combined with their known susceptibility to net gears means the potential contribution of these gears to elasmobranch catches should not be overlooked.

Kenya

Elasmobranch landings have fluctuated in recent years, dipping as low as 165 mt in 2007, peaking at 373 mt in 2012 and reported at 293 mt in 2014 (Ndegwa 2015). Curiously, between 2011 and 2013 Tana River province had the highest contribution to the total elasmobranch landings, despite it having the lowest overall reported fisheries landings (KMALF 2015). This is possibly the result of the much greater longline prevalence in this area, although use of this gear type has subsequently dramatically decreased (KMALF 2014b).

Landings are of elasmobranchs are not reported to species level. However, recent studies suggest that hammerheads, tiger (Galeocerdo cuvier), blacktip reef (Carcharhinus melanopterus), whitetip reef (Triaenodon obesus) and grey reef sharks feature prominently, together with a number of batoids including Mobulid rays (Mobula spp.), spotted eagle (Aetobatus cf. ocellatus) and bluespotted fantail rays (Taeniura lymma) as well as large wedgefish (Rhynchobatus spp.) (Ndegwa 2015; J. Kiszka unpublished data). Other rays belonging to Dasyatidae and Myliobatidae families also appear common, having been recorded in catch assessment surveys (State Department of Fisheries and the Blue Economy unpublished data).

Madagascar

Sharks are exploited throughout Madagascan waters. However, catch and landing data are mostly available for the west coast, particularly in the southwestern (Toliara) and northwestern (Mahajunga) regions, where conditions are more favourable for fishing (Cooke 1997; Cripps et al. 2015; McVean et al. 2006; Robinson and Sauer 2013). Data relating to the volume of elasmobranch catch are scarce. Traditional fishing (i.e. from sail-powered dug-out canoes) is estimated to produce ≥85% of the total shark catch (Le Manach et al. 2011; Randriamiarisoa and Rafomanana 2005) and likely for the vast majority of batoid catch. In some areas elasmobranchs may account for as high as 50–60% of the overall catch (Andriantsoa and Randriamiarisoa 2013; de Feu 1998). Elasmobranch landings across all Madagascan fisheries was reported as 5650 mt in 2014 (FAO 2016b), yet estimates suggest small-scale fisheries landings alone are in the region of 7500 mt/year (Le Manach et al. 2011). Shark catches are reportedly decreasing in Madagascar (Cooke 1997; McVean et al. 2006) possibly a response to declining shark fin demand (Whitcraft et al. 2014) and/or intensive overfishing.

A breakdown of species composition is not available for the small-scale fisheries, however a number of case studies have been undertaken. These studies list hammerheads, silky, tiger, spottail (Carcharhinus sorrah), sliteye (Loxodon macrorhinus), whitetip reef, blacktip reef and grey reef sharks as common species in a number of areas (Andriantsoa and Randriamiarisoa 2013; Cooke 1997; Robinson and Sauer 2013; Short 2011; Smale 1998). Information regarding batoid catch is more limited, but spotted eagle, thornback (Raja clavata) and bluespotted fantail rays, various guitarfish (Rhinobatidae spp.) and large wedgefish (Rhynchobatus spp.) are also captured (Cooke 1997), with Mobula spp. reported anecdotally (Heinrichs et al. 2011).

Mauritius

Mauritius reported artisanal landings of 0.456 mt of elasmobranchs in 2013 (Albion Fisheries Research Centre unpublished data), representing around 0.8% of the total fisheries landings by weight. In contrast, interview surveys suggest around 6000 elasmobranchs are caught annually (Poonian 2015). Whilst interview surveys are likely an unreliable way to estimate total catch effectively, the magnitude of difference suggests official reports are substantial underestimates.

The reported landings in 2013 is a 97% decrease from the 16.725 mt in 2000 and follows general declines in landings across the Mauritian fisheries since the mid-2000’s (FAO 2016b). Over this period around 85% of elasmobranch landings originated from line gears (representing 0.95% of the total line fisheries landings), whereas landings from net gears was disproportionally lower at around 7% of total elasmobranch landings (representing 0.1% of the total net fisheries landings). Basket traps account for the remaining portion, probably impacting smaller, benthic species, and may represent a threat to small batoids in particular. Hammerhead and tiger sharks are by far the most common species reported (Poonian 2015). Further, Mobula spp. were the only batoid reported in the catch. Ease of identification for these species likely heavily biases interview responses and so inevitably overestimates their importance in the fisheries.

Mayotte

Traditionally fishers have exploited the species-rich lagoon surrounding the island. However, decreasing catches in reef habitats (Guézel et al. 2009) and modernisation of fishing gears have resulted in a shift towards offshore pelagic resources, evidenced by the proliferation of pelagic teleost and elasmobranch species in their landings (FAO 2016b). It seems likely that significant loss of coastal/inshore elasmobranch species may have already taken place and remaining lagoon-based small-scale fisheries may be continuing to impact on or at least hindering the recovery of these stocks.

The shift towards a pelagic fishery means both competition and shared stock resources with the regional industrial fisheries. An investigation of the expanding longline small-scale fisheries of Mayotte revealed high abundance of silky, blue, scalloped hammerhead (Sphyrna lewini) and oceanic whitetip sharks in the catch, with pelagic stingrays (Pteroplatytrygon violacea) being the only batoid representative (Kiszka et al. 2010). The study showed that elasmobranchs comprised 24.6% of the catch. Whilst the fisheries themselves are dominated by handlines (both static and trolled) (Table 1) the composition is potentially similar to that of the longlines and so these may reflect the fisheries as a whole. Most, if not all, elasmobranch catch is discarded, of which 16.1% were dead (Kiszka et al. 2010). If this is representative of the fisheries as a whole, elasmobranchs discarded as dead would represent 5% of total catch of pelagic species (880 mt landed in 2014). This suggests that the reported landings of elasmobranchs (11 mt in 2014), are a significant underestimate of the true catch, perhaps by 75% or more.

Mozambique

Significant improvement in the monitoring and estimation of small-scale fisheries landings has been made in Mozambique (Dias and Afonso 2011; Doherty et al. 2015). Partial elasmobranch disaggregation in overall fisheries data is available through the FAO, with shortfin mako (Isurus oxyrinchus) and blue shark landings available separately in 2014 accounting for 26.3% of elasmobranch landings (FAO 2016b). Previously, separate landings data for copper (Carcharhinus brachyurus), silky and oceanic whitetip sharks have also been provided. However, at least for the small-scale fisheries in some provinces, data are available for a further 15 shark species and for spotted eagle ray (Instituto Nacional de Investigação Pesqueira unpublished data). This represents a significant step forward in the fisheries landings data resolution compared to other SWIO nations, though further disaggregation of species, and batoids in particular, is desirable. This is especially true for Mobula spp., for which Mozambique has a targeted sustenance fishery (Marshall et al. 2011) and possible gill plate trade (Heinrichs et al. 2011).

Over 98.8% of the 854 mt of elasmobranch landings reported in 2014 originates from small-scale fisheries (Instituto Nacional de Investigação Pesqueira unpublished data). The majority comes from the dominant beach seine fishery, of which 92% of landings were reportedly spotted eagle ray, clearly highlighting this fishery as a specific threat to this species. Whilst the beach seine fishery is the most important component by virtue of its size (accounting for 46.4% of small-scale fisheries landings in 2014), much higher CPUE rates for elasmobranchs are seen in the bottom set gillnet sector (Instituto Nacional de Investigação Pesqueira unpublished data). This further emphasizes the apparent threat this gear poses to elasmobranchs at a regional level.

The validity and accuracy of both the FAO and the official Mozambique small-scale fisheries datasets, is difficult to assess. Aside from the probable significant under-reporting, there are serious discrepancies both between and within data sets. The small-scale fisheries elasmobranch landings data are often much larger than that of the total reported amounts through the FAO, in the case of 2011 by over 200 mt (FAO 2016b; Instituto Nacional de Investigação Pesqueira unpublished data). There are also significant variations between years e.g. in 2013 bottom set gillnet landings were reported as 534.1 mt, yet in 2012 and 2014 only 142.6 and 141.1 mt were reported, respectively. In addition, the only known estimate of shark catch in the small-scale fisheries was 2186 mt from 1993, far in excess of both current and historical landings figures for the whole fisheries sector (Sousa et al. 1997). It is clear that these inconsistencies and the identification of their underlying drivers must be resolved as an urgent priority.

La Réunion

La Réunion does not currently record the landings of elasmobranchs in its small-scale fisheries; therefore, no official estimates exist (L’Institut Français de Recherche pour l’Exploitation de la Mer personal communication). The fisheries are dominated by longline and trolled handline (~77.5% of catch) and so it is possible that elasmobranchs, particularly sharks, are an important constituent of the catch. The restricted shelf system around La Réunion is thought overfished and has been exploited historically (Le Manach et al. 2015). Since 2011, tiger and bull sharks (Carcharhinus leucas) have been the cause of increasing numbers of reported attacks on bathers and surfers, resulting in increasing efforts to reduce numbers of these species (Lemahieu et al. 2017; L’Institut de Recherche pour le Développement personal communication). However, no statistics are currently available and the magnitude of culling is unknown.

Seychelles

Fishing for elasmobranchs, primarily for finning, has been ongoing since the 1920s. Since declines in the 1950s, elasmobranchs have shifted towards incidental catch (Fowler et al. 2005), though targeted fisheries still exist. In 2014 elasmobranchs accounted for approximately 1% of total small-scale fisheries landings (SFA 2015). Landings appear to be seasonal, peaking during the months of July and August (Online Resource 6). This pattern is likely driven by the increased catches of hammerhead sharks during this time (particularly S. lewini and S. mokarran), a fishery that is believed to be sustainable (Nevill et al. 2007). Breakdown of effort and landings by gear in the official reports are insufficient to allow for analyses of historical CPUE across the fisheries. A lack of species level identification in landings data has been identified as impeding effective management in the Seychelles (Nevill et al. 2007). It is suggested that the most commonly caught are spottail and grey reef sharks in inshore waters (Fowler et al. 2005), though interviewed fishers reported tiger and sandbar sharks (Carcharhinus plumbeus) as most common (Nevill et al. 2007).

Currently there is an encouraged expansion of small-scale fisheries towards targeting pelagic finfish. This includes a development fund providing access to loans for purchasing/upgrading to longlines (Seychelles Fishing Authority personal communication). This could increase pressure on oceanic elasmobranchs, with catch composition mirroring that of the semi-industrial pelagic longliners, which report high landings of silvertip (Carcharhinus albimarginatus) and oceanic whitetip sharks (Nevill et al. 2007). Given the current lack of detailed catch statistics these developments need to be monitored closely.

Tanzania (including Zanzibar)

The catch of elasmobranchs in small-scale fisheries is significant and, at least in Zanzibar, shows signs of overexploitation and partial collapse (Jiddawi and Shehe 1999). Information regarding the catch composition of these fisheries is limited. In mainland small-scale fisheries, 11 species are commonly caught (Barnett 1997). These are predominantly requiem sharks, both oceanic and coastal (including coral reef associated), alongside hammerhead, milk (Rhizoprionodon acutus) and whitetip reef sharks. Large wedgefish (Rhynchobatus spp.) have also been reported. In Zanzibar, at least 21 elasmobranch species are caught (Barrowclift et al. 2017). A market sampling survey in 2004 identified milk, grey reef and black tip reef sharks as the most common species, with various wedgefish including bottlenose and/or whitespotted and bowmouth (Rhina ancylostoma) and Zanzibar guitarfish (Acroteriobatus zanzibarensis) also present in relatively high numbers (Schaeffer 2004). Other batoids were not recorded, however it is known that various Dasyatidae species dominate the batoid catch (Barrowclift et al. 2017). Mobula spp. are caught in both mainland and Zanzibari small-scale fisheries (Heinrichs et al. 2011; A. Temple unpublished data).

In mainland small-scale fisheries, shark catches are common through most of the year, reduced only when the weather restricts fishing activity (Barnett 1997). In Zanzibar catches of sharks appear seasonal, being highest during the north-east monsoon, particularly between January and May (Barnett 1997; Schaeffer 2004). However, there is insufficient information available to suggest that elasmobranch abundance is related to season. More likely is that seasonal weather precludes the use of certain gears and/or fishing locations, so impacting elasmobranch captures.

Use and value

Elasmobranchs are generally considered a target species throughout SWIO small-scale fisheries (Cooke 1997; Jiddawi and Shehe 1999; Maoulida et al. 2009; Poonian 2015; Wekesa 2013). Often they are taken as a desired constituent of a multi-species fishery also targeting moderate-to-large pelagic or reef fish species, rather than in a dedicated elasmobranch fishery. Finning is relatively rare in the region’s small-scale fisheries, with the majority of catch landed whole and fully utilised (Wekesa 2013; A. Temple personal observation). However, finning does occur in parts of Mozambique and Madagascar where dedicated fisheries with formal processing and export markets exist (Cripps et al. 2015; Pierce et al. 2008). These practices present further difficulties in documenting and assessing catch. Currently, there is little evidence of demand for Mobula spp. gill plates emanating from the SWIO region (Cisneros-Montemayor et al. 2013), with the exception of small exports from Mozambique (Dent and Clarke 2015). However, should this change it will likely increase fisheries pressure exerted on these species. Where export markets for elasmobranch products do exist, shark and wedgefish fins hold substantially higher value relative to the rest of the body (Online Resource 4). However, readily available data on landings volume and value are limited e.g. the only reports to the FAO were from Madagascar and Seychelles in 2013 (FAO 2016b). Further, there is a lack of information regarding supply chains and the contribution from small-scale fisheries to these.

As a source of protein elasmobranch meat is relatively cheap (Online Resource 4) in comparison with teleosts (e.g. MFR 2004, 2012), and may form an important nutritional component in the diets of those supplied by and dependant on small-scale fisheries. Lack of access to cold storage facilities requires that meat is often either sold fresh locally or air/salt dried in preparation for sale at distant markets or export (Cripps et al. 2015). Value of elasmobranch products depends on a number of factors including perceived quality (species and preparation related), route of sale and cultural and religious influences, such as Ramadan, which affect both supply and demand (Barrowclift et al. 2017; Cripps et al. 2015). It is therefore important to further our understanding of elasmobranch value in SWIO small-scale fisheries, and their markets and drivers, if we are to assess the socio-economic importance of these fisheries and their component species. Ultimately, a proper understanding of the socio-economic value of the fisheries is vital to the design and implementation of any successful management strategy.

Fisheries policy and management in the southwestern Indian Ocean: implications for vulnerable marine megafauna in small-scale fisheries

Formal governance arrangements in the fisheries sector are a fundamental component in the sustainable use of fisheries stocks. Good governance, policies and resultant effective management of fisheries stand to create a strong platform from which sustainable species harvest can be achieved and controlled. It is therefore vital that we understand both the principles upon which policies are built and what management is in place to achieve these goals with regard to vulnerable marine megafauna in the SWIO.

International

Many SWIO nations are party to international fisheries-specific agreements that have implications for both general fisheries policy and management and vulnerable marine megafauna specifically (Online Resource 5). Primary amongst these is the United Nations Fish Stocks Agreement (UNFSA) 1995, which obligates that parties undertake ecosystem-based and precautionary approaches to migratory fish stock management (UN 2016). This agreement increases the responsibility of nations over their fisheries and their enforcement of laws within them, and strengthens the roles of regional fisheries bodies. Given the overlap in vulnerable marine megafauna species between SWIO nations (Table 3; Bourjea 2015; Kiszka 2015; Kiszka and van der Elst 2015), and absence of stock delineations for the majority of species, there is a clear need to consider these as shared resources until clarification can be achieved. Concurrently, the signatory status, ratification and accession to the Port State Measures Agreement by many SWIO nations (FAO 2016a), signifies increasing regional efforts to tackle illegal, unlicensed and unreported (IUU) fisheries. These fisheries are rife in the SWIO region (Agnew et al. 2009) and so likely have substantial impacts on vulnerable marine megafauna species and marine ecosystems in general. Better control and documentation of these IUU fisheries will therefore be vital in managing both vulnerable marine megafauna fisheries and the fisheries as a whole in the SIWO region.

At a regional level, SWIO marine environmental policies are largely outlined by a number of conventions and agreements. The most ubiquitous of these is the Nairobi Convention 2010 (formally, the Amended Convention of the Protection, Management and Development of the Marine and Coastal Environment of the Western Indian Ocean) to which all SWIO nations are party. Broadly, the convention places onus on the party states to work, both individually and co-operatively, in an effort to sustainably maintain, manage and develop their marine and coastal ecosystems (UNEP 2010). It highlights a recognition and willingness of SWIO nations to view the marine environment as an inter-linked and shared resource. This outlook is pivotal to any meaningful management of the region’s fisheries, including those which catch vulnerable marine megafauna.

There are also specific international agreements dealing with vulnerable marine megafauna to which SWIO nations are contracted. The Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) (CITES 2016) make international trade of any marine mammal or sea turtle illegal. Further, trade of 17 species of elasmobranch (12 of which are known to occur in the SWIO) are controlled to varying extents. In this regard these species receive some level of protection through control over commercial exploitation for international trade. The Indian Ocean Cetacean Sanctuary, designated and established in 1979 by the International Whaling Commission (IWC), covers the entirety of the Indian Ocean south to 55°S and prohibits commercial whaling. However, it does not provide protection for smaller cetaceans nor does it identify critical habitats for cetaceans (IWC 1980), a significant roadblock to achieve effective implementation. Currently Kenya and Tanzania are the only SWIO nation members of the IWC. The Conservation of Migratory Species Memoranda of Understandings (CMS MoUs) seeking to encourage protection of and promote stock recovery for sea turtles, dugongs and sharks, have been effective since 2001, 2007 and 2010 respectively. Both the sea turtle and dugong agreements boast wide coverage, with only La Réunion absent as a signatory on the dugong MoU, as the species is absent from its waters (CMS 2016a, b). Conversely, few SWIO nations have signed the CMS MoU for sharks (Online Resource 5), raising some concern over the political willingness of SWIO nations to sustainably manage these species, though it must be considered that this MoU has only been created very recently (CMS 2016c). Those SWIO countries yet to sign the CMS MoUs should be encouraged too do so, as the commission and these agreements represent a pathways for facilitating the conservation (in its widest sense) of vulnerable marine megafauna in the region.

National directed management

At a national level there are varying degrees of directed management relating to vulnerable marine megafauna species. Dugongs and cetaceans are protected by law throughout the SWIO and for sea turtles, intentional catch, egg poaching and sale is widely prohibited (Table 4). Further protection is offered through anti-disturbance regulations in some areas, both site-specific (in various MPAs) and nationally in the case of Mozambique (Marine and Coastal Environment Regulation, Decree 45/2006). Only the Seychelles do not offer complete protection, with green and hawksbill protected but not loggerhead, leatherback or olive ridley, despite their presence (although not nesting) in the SWIO region (Frazier 1984; Remie and Mortimer 2007). A National Plan of Action (NPOA) for sea turtles has been implemented by France, covering Mayotte, the French dispersed islands (Tromelin, Glorieuses, Juan de Nova, Bassas da India and Europa) and La Réunion, and a NPOA is under discussion for Tanzania (Igulu and El Kharousy 2015).

Table 4 Legal status and related punishments regarding vulnerable marine megafauna in the small-scale fisheries of the South Western Indian Ocean
Full size table

Conversely, current regulations regarding elasmobranchs are very limited for the small-scale fisheries of the region (Table 4) and where they do exist their effectiveness is often questioned. For example, as of 2016, licences are no longer being distributed for shark fin export from Zanzibar (Ministry of Agriculture, Livestock and Fisheries, personal communication), effectively making the trade illegal. However, post-harvest removal of fins from landed elasmobranchs to supply international trade (including from CITES listed species) still occurs in spite of these restrictions (A. Temple personal observation). Unlicensed shipping of shark fins to Kenya is thought common (Fowler et al. 2005) and may be continuing to provide the export route out of the country. In spite of the general lack of regulations for the elasmobranch small-scale fisheries, NPOAs have been or are currently in development for most SWIO nations (Table 4), with Kenya and Mozambique projecting completion by the end of 2017 (Chacate and Mutombene 2015, Wekesa 2014), suggesting widespread recognition of the threats faced by these species and a movement towards addressing the issues. As with other regulations pertaining to many small-scale fisheries in the region, implementation is challenged by management authorities’ lack of infrastructure and resources to ensure compliance, compounded by the generally open-access nature of these fisheries.

National indirect management

There are also some regulations and initiatives which likely indirectly impact interactions between small-scale fisheries and vulnerable marine megafauna in the SWIO (see http://www.wiofish.org for comprehensive information on prohibited fisheries and gear restrictions in the SWIO). For example, in Mauritius, the Fisheries and Marine Resources Act 2007 prohibits the use of certain fishing gears. Most relevant to vulnerable marine megafauna is the ban on the use of driftnets (defined there as a net exceeding 250 m in length, fitted with floats or weights to make it hang vertically in the water column) with a fine of up to 20,000 USD if found in breach of the Act. Further it prohibits the use of shorter net gears (“large nets” and “gillnets” less than 250 m in length) for 5 months of the year (October–February, 500,000 MUR/~14,000 USD fine) and licences may only be issued for use of 10 “large nets” and 5 “gillnets” at any one time in the lagoon waters of Mauritius. Given these nets form the principal gear threats to vulnerable marine megafauna at the global scale, this likely has a major impact on their interactions with the fisheries in Mauritius. In Kenya (Fisheries Act Cap 378, Kenya Gazette Notice No. 7565) and Tanzania (Fisheries Act Regulations, 2003) the use of mono-filament nets is prohibited, though compliance is poor (KMALF 2014b; A. Temple personal observation), resultant changes in gear use could impact the catches of vulnerable marine megafauna species susceptible to alternate gears whilst protecting those species susceptible to the mono-filament nets. Beach seine nets are prohibited in Kenya, Tanzania and Comoros, which reduces fishing pressure on some coastal elasmobranchs, but again, ensuring compliance is problematic.

Discussion

This review highlights the severely limited understanding of vulnerable marine megafauna and their interactions with the small-scale fisheries of the SWIO resulting from a lack of robust data. Yet, where evidence exists, there are indications of population declines due to fisheries interactions (Amir et al. 2002; Cooke 1997; FAO 2016b; Jiddawi and Shehe 1999; McVean et al. 2006; Muir and Kiszka 2012). Furthermore, it is clear that at both national and regional levels, current small-scale fisheries monitoring practices and management are insufficient to assess and ensure the long-term sustainability of small-scale fisheries and the vulnerable marine megafauna species impacted by them. Therefore, there is a clear regional priority to collect much more information on small-scale fisheries characteristics, species catch, landings and composition, as well as data regarding vulnerable marine megafauna gear-interactions. These data are required to undertake a robust and detailed analysis at both regional and area-specific spatial scales, without which informed, evidence-based management and facilitating policies cannot be achieved effectively. As such, and in the absence of truly precautionary management, vulnerable marine megafauna must be considered at high risk of on-going overexploitation at both the national and regional scale in the SWIO.

Improved monitoring and assessment of small-scale fisheries both at national and regional levels is critical in achieving sustainable harvest of fisheries resources, including vulnerable marine megafauna. It is probable that the majority of these fisheries stocks are shared given: the highly mobile, transboundary nature of many species (e.g. Table 3); few geographical barriers in the SWIO; migrant fishing; and the notion of Economic Exclusive Zones being at best flexible when regarding small-scale fisheries. In light of this there is a clear need for SWIO nations to begin identification and delineation of stocks and to devise a joint strategy defining protocols for collection and reporting of small-scale fisheries data. At the heart of any such strategy must be a consensus on minimum data requirements (landings, effort, gear composition and their breakdowns), standardised metrics and methodologies used for collecting the data. Standardised reporting procedures and formats would also be of great benefit, allowing data to be compiled and synthesised with greater ease. Beyond any agreed minimum, nations should be encouraged to collect further data as far as is feasible, particularly where these data address specific issues or interests of each party or the region. Where such issues may have wider applicability an open discussion regarding metrics and methods would benefit all parties. Ultimately such changes would aid the understanding and management of small-scale fisheries and assist the decision-making processes, with implications for the long-term regional sustainability of the sector. However, currently there is no regional body with the ability to make binding decisions on such issues, the Indian Ocean Tuna Commission’s (IOTC) mandate is too restricted and the Southern Indian Ocean Fisheries Agreement’s (SIOFA) membership only has partial SWIO coverage. Perhaps best placed to facilitate such changes is the South West Indian Ocean Fisheries Commission (SWIOFC), through its role as an advisory body, in conjunction with expanding the mandate of the IOTC and through the ability of IOTC and SIOFA to make binding decisions (van der Geest 2017).

CPUE is a fundamental measure used to monitor stock health and fisheries sustainability (Maunder and Punt 2004; Sparre and Venema 1998), yet for most SWIO small-scale fisheries such data cannot be generated. To create an accurate and usable CPUE time series, effort data must accommodate changes in fisheries dynamics, behaviour and power. To this end measurement of fishing effort requires four main data types: gear type including specificity (e.g. mesh size, hook size, mono/multifilament etc.); fishing mode (e.g. active or passive) to allow proper categorisation of the fisheries; gear characteristics (e.g. net dimensions, number of hooks etc.); and active fishing effort (e.g. soak times, trawl speeds/distance etc.). Currently data collected for small-scale fisheries in SWIO nations consistently lacks one or more of these aspects (Table 1, Online Resource 3). For example, in five nations (Comoros, Kenya, Madagascar, Tanzania mainland and Zanzibar) effort is either recorded and not compiled or not recorded at all. Whereas, in others fishing effort is given in trips, sets, active vessels and fisher-days (Online Resource 3), none of which are suited for accurate estimation of fishing effort due to large potential variation within these measures e.g. trips may last for hours or days. Thus the utilisation of fisheries data for informing management objectives, targets and strategies is restricted.

Some of these effort monitoring weaknesses could be addressed through relatively minor changes to current monitoring protocols. Detailed gear specifications and fishing modes could be incorporated into national census/frame surveys in which data on gear type are already routinely collected (Table 1). Active fishing effort data are however more difficult to obtain, especially given the informal nature of many small-scale fisheries. It is perhaps inevitable that in the short term active fishing effort data may need to be generated through declarations by fishers along with evidence-based assumptions until a more formalised system is possible. Alternatively, CPUE estimates could be generated through fisheries-independent data (Sparre and Venema 1998), however this can be costly and so is likely unfeasible for most SWIO nations.

Regarding vulnerable marine megafauna in SWIO small-scale fisheries, generally catch and landings data are relatively poor and where available are often lacking in both species composition and catch-by-gear data. This lack of information severely limits the ability to identify and manage at-risk species and stocks, including assessment of gear and area-specific threats. Despite the general paucity of data, it is clear that in a number of areas catches are in decline and some populations are known to be overexploited (Amir et al. 2002; Cooke 1997; FAO 2016b; Jiddawi and Shehe 1999; McVean et al. 2006; Muir and Kiszka 2012; Nevill et al. 2007). Thus, there is an urgent need for proper assessment of vulnerable marine megafauna in SWIO small-scale fisheries. Catch and landings data which do exist show substantial numbers of large oceanic shark species in the small-scale fisheries (Andriantsoa and Randriamiarisoa 2013; Barnett 1997; Cooke 1997; FAO 2016b; Kiszka et al. 2010; Maoulida et al. 2009; Ndegwa 2015; Nevill et al. 2007; Poonian 2015; Robinson and Sauer 2013; Schaeffer 2004; Smale 1998; Soilihi 2014), indicating increasing competition for resources with the industrial fisheries. Increasing competition for these stocks is concerning for two main reasons: increased pressure on stocks that are already thought to be overharvested and at high extinction risk (García et al. 2008); and it indicates a shift towards fishing further offshore by the small-scale fisheries sector, a phenomenon seen elsewhere in instances where inshore stocks may have become depleted.

Comparatively little information on the marine mammal, sea turtle and batoid and chimera catch components of the fisheries are available, though given the general confinement of chimera to deeper waters (Kyne and Simpfendorfer 2007) they are unlikely affected by small-scale fisheries in the SWIO region. There is therefore a clear need to address this data vacuum. With regards to marine mammals and sea turtles specifically, understanding of catch is further limited as a result of their legal status, creating a reluctance to declare catches. Governmental departments are therefore poorly placed to generate these data and so they should represent a priority focus for collaborative work with independent researchers and NGOs.

Currently the data regarding SWIO small-scale fisheries interactions with vulnerable marine megafauna are unable to support management aimed at safe sustainable exploitation of these resources, yet there is clear evidence that vulnerable marine megafauna species are at risk from these fisheries. As such, and in accordance with the UN Fisheries Stocks Agreement 1995 and the Nairobi Convention, a precautionary conservation-minded approach is mandated to safeguard vulnerable marine megafauna until such time as robust evidence-based management strategies for sustainable exploitation can be achieved. If such measures are not taken, SWIO nations will be failing in their duty of care to both to the fishers and communities that rely on these resources and to the vulnerable marine megafauna themselves by failing to protect the biodiversity of their marine environment.

At the national scale, effective management strategies require a proper understanding of the human elements of the fisheries (Gray 2005; Kooiman et al. 2005). Fishers face an increasing variety of changing socio-economic conditions related to overexploitation, climate change, globalization, and conservation of marine biodiversity. Understanding the socio-economic importance of vulnerable marine megafauna species across stakeholder groups, including perceptions and attitudes (both cultural and individual) towards these species, as well as how fishers will respond to potential ecosystem and institutional changes is critical to better managing these fisheries, achieving fishers acceptance of and compliance with management strategies and improving the livelihoods of those dependent on fisheries supply chains (Daw et al. 2012). Further, effective and appropriate enforcement is vital for sustainable management practices to be implemented, whether this be establishment or community driven, and presents a significant challenge for SWIO governments. Without this, dissent and non-compliance (Peterson and Stead 2011) can become wide-spread issues (Hauck 2008; Keane et al. 2008; Nielsen and Mathiesen 2003) and management strategies can be rendered ineffective. This review highlights that a significant gap in information exists in describing existing and shifting dependency of fishers in SWIO small-scale fisheries which, if collected alongside the much needed ecological data on fisheries, could provide better context for introducing effective management measures for vulnerable marine megafauna.

In the face of the numerous data gaps and taking into context infrastructure and the resource constraints present in SWIO nations, there is a clear need to identify appropriate low-cost methods to assess the magnitude of vulnerable marine megafauna catch in small-scale fisheries and to mitigate these when they are unsustainable. Critical assessment of various data collection methods (e.g. vessel-based observer programs, interview surveys, landing site data collection), incorporating time and cost factors, is vital in facilitating informed decision making, through which challenges can addressed in both the short and long term using the appropriate methodological tools. Further, for species at highest risk, precautionary mitigation strategies need to be considered. Sea turtles, marine mammals and several species of elasmobranchs are of primary concern in this regard and catch mitigation methods (e.g. turtle excluding devices, weak links for nets and acoustic alarms) are already available for some of them (e.g. Barlow and Cameron 2003; Gilman et al. 2006; Ward et al. 2008). However, many of these methods are costly, and so there is a clear need and an opportunity to develop minimum-cost methods that are feasible for implementation in small-scale fisheries in the SWIO and globally. As a Regional Fisheries Authority, the SWIOFC and its working groups are well placed to facilitate the promotion and co-ordination of these initiatives, and to undertake and/or guide regular assessments of small-scale fisheries in the SWIO, in order to address questions of sustainability of vulnerable megafauna, and of the fisheries themselves.

Conclusions

This review highlights information needed to reconcile vulnerable marine megafauna conservation with small-scale fisheries demands and where it is lacking in the SWIO. Here the SWIO acts as both subject and case study in the broader issue of marine species conservation goals and the needs of those communities that rely on them, especially in data poor and developing regions. In addressing these issues, it is essential that solutions be built properly upon principles that balance environmental and human (economic and social) needs and are grounded in realism. Both funding and the time-scales in which to find effective solutions are limited, particularly in developing regions, and so research must be strictly prioritised towards practical and goal-oriented outputs that properly account for and engage stakeholders. Given the potentially dire situation for several vulnerable marine megafauna species it is critical to address priority baseline data gaps and their associated challenges. Governments, NGOs, independent researchers and research institutions and other stakeholders must act collaboratively to achieve common goals and ensure implementation of findings into effective evidence-based management, thus facilitating for a sustainable future for vulnerable marine megafauna species, the marine ecosystem and the associated livelihoods in coastal communities.