Small-scale fisheries make significant contributions to nutrition, food security, and coastal economies, particularly in low-income countries (FAO, 2019). Yet there is often a lack of data documenting their contribution, and importantly their composition and change over time (Jackson et al., 2001; Andrew et al., 2007; Jacquet and Pauly, 2008). Fishing activities are increasing worldwide both as a result of a growing population and increased per capita consumption (Cai and Leung, 2017). This growing demand places additional pressure on the health of fish stocks. Although there is evidence of improving trends in the stock abundance of well managed fisheries, this is not the case for fisheries in much of the global south where effective governance, monitoring and management are often lacking (Costello et al., 2016; Hilborn and Hilborn, 2019).

In addition to the threats of overfishing, climate change is expected to impact fish and fisheries through a complex suite of linked processes (Brander, 2010). Larger declines are predicted particularly in tropical regions and therefore likely to disproportionately impact the communities that rely on the ocean for their livelihoods (Lam et al., 2016; Ding et al., 2017).

Madagascar is listed as the eleventh poorest country in the world (International Monetary Fund, 2019) and more than 77% of its people live on less than USD $1.90 per day (Alkire et al., 2019). Similar to many countries in the global south, the country lacks the institutional and governance capacity to monitor and manage the socially dynamic, multi-species and often geographically remote fisheries effectively. Therefore, policy and management often remain largely uninformed, ineffective and unable to identify - let alone respond to - changes in fisheries resources and sustainability (Le Manach et al., 2011; Le Manach et al., 2012).

Madagascar’s marine capture fisheries production was estimated to be around 125,000 tonnes in 2017 (FAO, 2019). A study published in 2011 estimated that Madagascar’s small-scale marine fisheries contributed as much as 65% of total fish production (Le Manach et al., 2011), and this same study highlighted that Madagascar’s national fisheries statistics were under-reported by up to 200% between the 1950s and 2000’s (Le Manach et al., 2011). In a country where coastal regions have historically had a higher incidence of poverty than the rest of the country as a whole (Horning, 2008; Moran et al., 2008), the contribution of small-scale fisheries to the national economy and food security is likely to be greatly undervalued.

In much of Madagascar, and particularly in the south-western regions of Atsimo Andrefana and Menabe, fishing and associated activities, including aquaculture and fish processing, are the principal, and often only, source of income for many people (Laroche et al., 1997; Lilette, 2006). This stretch of coastline is home to the Vezo, a semi-nomadic ethnic group of traditional fishers (Astuti, 1995; Sanders, 2005; Lilette, 2006; Cripps and Gardner, 2016). With up to 95% of households in these coastal communities being reliant on fishing as their primary source of income (Barnes-Mauthe et al., 2013), effective fisheries management is vital to ensure the livelihoods and food security of people most at risk from overfishing as well as other human-induced and external threats e.g. climate change.

In 1991, Laroche et al. (1997) surveyed the small-scale fisheries in coastal cities of Toliara, Morombe and Morondava. They found diverse catches, comprising of pelagic, reef and coastal species, with differing species composition across the three main coastal towns. Fisheries in Toliara and Morombe focused on reefs, while in Morondava it primarily focused on sandy coastal areas, with some hand line fishers targeting scattered coral banks. The town of Morondava, also the focus of the current study, had the highest diversity of catches, with a greater composition of high-value and higher trophic level species than the other towns. Despite lower catch rates, particularly noted for handlines, fishers were still able to generate a good income from targeting high-value species in the region in 1991.

Laroche et al. (1997) also highlighted that fishers consistently reported that they were already adapting their activities to mitigate impacts of declining catches on their lives. These adaptations included increasing effective fishing effort (e.g. more hooks, longer nets, and smaller mesh sizes), developing technical solutions (e.g. using longer lasting nylon nets instead of cotton), and seeking supplementary income sources (e.g. farming or dockwork (boutry) in the nearby port). The authors reflected on the reasons for fisheries declines, including changes in the market forces across all of the towns, while specifically focusing on the competition with the commercial shrimp-trawl fishery in Morondava. They concluded that the fisheries were overexploited and in need of improved management and called for the integration of traditional management with government policy, and attribution of exclusive access rights to traditional fisher groups. They also noted that the use of marine sanctuaries could be particularly useful for maintaining species diversity and abundance and suggested that management would be most effective if associated with strategies for supporting livelihood diversification and stabilising population growth.

Since Laroche et al. (1997), a handful of isolated studies have been conducted on Madagascar’s small-scale fisheries, notably those involving finfish (Davies et al., 2009; Doukakis et al., 2009; Brenier et al., 2011; Gough et al., 2020), marine invertebrates (McVean et al., 2005; Barnes and Rawlinson, 2009; Bemiasa, 2009), sharks (McVean et al., 2006; Cripps et al., 2015; Humber et al., 2017a) and marine turtles (Walker and Roberts, 2005; Humber et al., 2011; Humber et al., 2017b), as well as monitoring methods (Behivoke et al, 2021). However, the majority of these studies provide only temporal ‘snapshots’ of fishery status; and none have assessed medium term trends in small-scale finfish fisheries. The current deficiency of accurate fisheries information underlines the need to improve understanding of temporal trends in small-scale fisheries in order to develop recommendations for adaptive management.

Building upon the analysis presented in Gough et al. (2020) that provided evidence of overfishing in the region in 2011, the current study compares landings from 2011 with those from the same fishing communities in 1991 (Laroche et al., 1997). We look at what had changed over the two decades between the two studies and interpret findings with a view to guiding future research and management that will safeguard fisheries livelihoods and human wellbeing.

In 2011, surveys were conducted 2-4 days per week (overall 34% of total fishing days in Avaradrova and 40% of fishing days in Betania). Data collectors observed a total of 18,304 boats going fishing, with an average of 68 and 77 boats per day in Avaradrova and Betania respectively, which was estimated by the data collectors to represent 60% of the total number of boats out fishing on any given day. They collected samples from a total of 1,555 fishing trips (570 in Avaradrova and 985 in Betania) which indicates that samples were taken for between 6 and 8% of the active boats. The number of fishers on a single fishing trip ranged from 1 to 11 and catch surveys recorded 72,234 kg of fish landed, comprising a total of 447,585 individual fish from 32 families. A further 2,158 kg of crustaceans and 41 kg of holothurians was also landed but excluded from this analysis as the study by Laroche et al. focused solely on finfish. In the 1991 study, monitoring was carried out on 22 days per month, with 10 pirogues sampled for catch each day (Laroche et al., 1997). Although Laroche et al. reports recording the number of pirogues, these data are not published and so we do not know what proportion of the fleet, or consequently, the catch this sample represented in the paper.

This paper provides a useful snapshot of the temporal change in the small-scale fishery of Morondava focusing on the changes that occurred between 1991 and 2011. The main results of the comparison point towards overexploitation of the resources revealing that; 1) the catch composition had changed between the two studies, with fewer high-trophic level species in catches in 2011 than in 1991 ( Figures 4A, B and Table 1 ) the estimated total monthly landings had dramatically increased between the two years. The seven-fold increase in catches appear to be explained by the substantial increase in the population of fishers between the two study periods ( Figure 1C ) and diversification of fishing gears, signified by the introduction of gears such as beach seines and long lines that were not reported in the 1991 study. We observed that gillnet and handline fishers were fishing further afield, spending more time fishing ( Figure 5B ) and returning with a slightly higher catch weight per fisher for a single fishing trip (CPUE) ( Figure 5D ). We examine each of these findings below and discuss what may be contributing to these results and what they may mean for the fishery and fishers of Morondava.

In 2011, Morondava fishers continued to catch a large diversity of demersal and pelagic species from offshore reef and sandy lagoon habitats, similar to that observed in 1991 by Laroche et al. (1997). While the number of fish families remained similar, some fish families that were present in 1991 became reduced or absent in the 2011 study. The reduction in the proportional contribution from families such as Istiophoridae (sailfish), and Serranidae (grouper) as well as sharks and rays was notable, as well as that from Lethrinidae (emperor) and Nemipteridae (sea bream), again larger and slow maturing species that show high vulnerability to fishing pressure (Cheung et al., 2005). Since many of these species are from high trophic levels, it explains the observed decline in median trophic level (MTL) of the catches. The 0.35 decline in trophic level (TL) observed in the Morondava fishery between 1991 and 2011 is higher than the reported global average of 0.1 TL per decade (Pauly et al., 1998). This provides evidence that the fishery may have been experiencing the phenomenon commonly referred to as ‘fishing down marine food webs’ and is considered indicative of a fishery under unsustainable exploitation (Pauly et al., 1998). A similar trend has been documented in the small-scale fishery of Toliara, southwest Madagascar (Brenier et al., 2011) as well as in a number of other small-scale tropical coastal fisheries (McClanahan and Mangi, 2004; Mangi and Roberts, 2006). Although change in trophic level could be explained by other factors besides resource overexploitation (Branch et al., 2010; Sethi et al., 2010) in this case the change is most likely associated with the change in catch composition (see below).

Essington et al. (2006) additionally defined ‘fishing through’ marine food webs describing how the MTL may decline due to the sequential addition of low TL catches rather than simply the decline of high TL species. The increased contribution of some low TL families, such as Clupeidae (sardine) in 2011 may reflect changes in markets and targeted fish species but is also likely a reflection of the introduction of new fishing gears, in particular the increased use of beach seines in the region over the 20 year period. The increased abundance of these species in catches has also been noted in a similar comparison conducted in the small-scale fishery of Toliara a region ca. 400km from the current site (Brenier et al., 2011).

Conversely, there may be some ‘fishing up’ the marine food web (Branch et al., 2010; Stergiou and Tsikliras, 2011) that is not captured by the overall change in TL as a number of higher TL families that were absent from the 1991 study were present in our 2011 surveys (although they may have been included in Laroche’s ‘other’ family group). Again, these are also likely to have appeared as a result of the introduction of new fishing gears and sites ( Figure 3 ), for example Lobotidae (tripletails), being caught primarily by longline fishers at offshore sites, and Trichuridae (cutlassfish), and Chirocentridae (wolf herring) being caught by small mesh gillnets in inshore waters.

Fishers in Avaradrova, particularly those using handline, were fishing for longer periods ( Figure 5B ) and more frequently at offshore sites in 2011 than in 1991 ( Figure 2 ). This may reflect the targeting of some of these larger bodied high trophic level fish families found predominantly in offshore areas, such as Scombridae (tuna and mackerel), or reef fishes such as those of the Serranidae (grouper), Lutjanidae (snapper) and Lethrinidae (emperor) families found on and around offshore reefs. Since these were some of the main families targeted in 1991 and their contribution to catches has declined ( Figure 3 ), it is likely that this shift may be due to reduced catches of these species on nearshore reefs, driving fishers further afield. In addition, the diversification of fishing gears, such as the introduction of shark nets and longlines to target sharks and scombrids, as well as fishing further afield in deeper or previously less exploited waters may also be partially maintaining catches and masking the decline in higher trophic level value (Morato et al., 2006). The fish families caught by small-scale fishers are targeted for both local consumption and export markets. Increases in fishing effort across these families is therefore likely linked to both the requirements of feeding a growing local population and meeting the growing demands for export. A socio-economic survey carried out in the region in 2011 reported that 63% of fishers identified a large decrease in the abundance of fish in catches, and that they subsequently fished further away (26%) or more frequently (7%) to compensate for the changes (Dewar, 2012). These behavioural responses have been observed in numerous other small-scale fisheries around the tropics (Padilla et al., 2003; Eder, 2005; Sadovy, 2005; Fabinyi, 2010) and are often driven by declining catches and can sometimes mask trophic declines (Stergiou and Tsikliras, 2011).

Beyond the absence of some fish families in catches, intensive fishing has been demonstrated to lead to imbalances in ecosystem function, which has ramifications for community structure overall (Jennings and Kaiser, 1998; Jennings et al., 1999). Fishery removals may induce indirect trophic (food web) interactions, and thus fishing can have many effects on community structure (Botsford et al., 1997). These interactions become increasingly more complex in multispecies fisheries where multiple gear types are deployed (Jennings et al., 1999; Jennings, 2005) as observed in this study.

In addition to the changes in trophic composition, the overall average monthly landings for the Morondava landing sites in 2011 were estimated to be 700% higher than estimated by Laroche in 1991 ( Table 1 ). It is likely that the increase in landings is due in part to increases in total fishing effort, particularly the number of fishers operating within the fishery. The population of Morondava has more than doubled from Laroche’s estimate of 40,000 to an estimated 120,000 in 2011. However, the population of fishers has increased more than 6-fold in this same period. This high growth is likely due to a combination of high fertility rates as well as the migration of fishers from communities to the south, who travel north as fisheries resources decline further south, attracted by the growth of export markets for shark fin and sea cucumbers and more plentiful resources further north (Cripps and Gardner, 2016). There is also a strong flow of agro-pastoralists and others from inland communities towards the coast as crop yields decline and communities face rising food and physical insecurity (Marikandia, 2001; Chaboud, 2006; Bruggemann et al., 2012). Migrants therefore have variable experience and expertise in fishing livelihoods depending on where they originated. Small-scale fisheries continue to make an important contribution to the livelihoods, income and food generation of the Vezo people living in the coastal town of Morondava and the wider south-west region of Madagascar. Yet with increasing numbers of people operating in the fishery, fishers often cite the number of fishers as being an important contributor to the changes they have seen in their catches over time, alongside the presence of an industrial shrimp-trawl fishery (Jones, 2011; Dewar, 2012).

Higher total monthly landings are, in part, explained by the diversification of fishing gears. Our results show the presence of shark nets, longlines, beach seines and mosquito nets in the Morondava fishery in 2011 ( Figure 2 and Table 2 ), yet these gears were absent in the records from 1991. The absence of these gears in the 1991 study however, may not indicate absence in the fishery, but may be a result of the limited geographic and temporal scope of the original study. While these gears may not have been routinely employed in the fishing communities of Betania and Avaradrova, it is likely that these gears were being employed in the wider Menabe region in 1991 as they were reported to be common at the time in other Vezo fisheries, e.g. Toliara and Morombe (Laroche and Ramananarivo, 1995; Laroche et al., 1997). Other anthropological studies in the Menabe region also note the use of mosquito nets (Astuti, 1995) and shark nets (Sanders, 2005). However, it is likely that these gear types were not used as extensively in the study villages at the time of the original study or because they were used seasonally and outside of the Laroche et al. (1997) study period (Apr-Oct). Beach seines and Mosquito nets for example, were used between October and December, and mosquito nets were previously reported to be used only seasonally to target a small shrimp known locally as patsa (Acetes erythraeus) (Le Reste, 1970) rather than finfish. Shark nets, which were used in the present study between August and December, and Longlines (used all year round), may have previously only been more commonly used by migrant fishers operating on the nearby islands of Nosy Andriangory and Nosy Andramitaroky (Sanders, 2005).

Additionally, the slightly higher catches observed for gillnets and handlines may also suggest that they have improved in efficiency between the two studies. Observations were already reported in 1991 that fishers were using nets with smaller mesh sizes and increased numbers of hooks to compensate for declining catches (Laroche et al., 1997). However, since the 1997 paper does not report on net lengths or mesh sizes, it is not possible to quantify the change in more precise measures of fishing effort. Fisher interviews conducted in 2011 showed that they had been increasingly deploying nets with mesh sizes as small as 10 mm, and with numerous lengths of nets spliced together, forming barriers of up to 1 km in length in attempts to maintain catches (Dewar, 2012). This phenomenon of “technology creep” may also have allowed fishers to maintain or increase catch rates despite a declining ecological condition, through the use of increasingly effective and often destructive fishing gears (Rijnsdorp et al., 2006; Eigaard et al., 2014).

The authors of the 1991 study already noted that the fisheries of Morondava were showing signs of overexploitation and were in need of management intervention (Laroche et al., 1997). Yet it is notable that with reduced abundance of high trophic levels and a much larger monthly catch, overfishing has continued in these fisheries since the original study, with further evidence of overfishing highlighted by a wider analysis of fisheries catches in the region (Gough et al., 2020). The 1997 paper specifically noted the contribution that the commercial shrimp-trawl fishery made to declining catches through large bycatch rates and high discards, and this fishery continues to operate in the same area as the small-scale fishery and continues to be cited by local fishers to be contributing to fisheries declines in the region (Jones, 2011; Dewar, 2012). However, it is not possible to determine the impact of the trawl fishery on the small-scale fisher catches without investigating the trawl fishery itself.

We were unable to perform statistical comparisons due to the absence of raw data from 1991, but wherever possible we have noted where variation in 2011 overlaps with 1991. This highlights an issue around access to data needed in order to inform fisheries management. Data deficiency is often cited in respect to small scale fisheries, and this is often due to data not being collected. However, this study highlights that even where data exists it may not be accessible to fisheries managers in a form that enables further analysis. This is changing however with increasing innovation in fisheries data collection and management (Bradley et al., 2019) including new tools and methods that could be incorporated to allow more precision in understanding changes to effective fishing effort (Behivoke et al, 2021) and a growing trend in sharing of data for transparency, reproducibility and verification purposes (Pendleton et al., 2019).

Due to both studies providing a snapshot of the fishery across a single year we are unable to assess the contribution that inter-annual variation makes to the differences observed. While additional data have not been collected from the villages of Betania and Avaradrova since 2011, socio-economic assessments conducted in 2018 in 10 villages to the south of Morondava reported that 71% of fishers reported declining catches. Of those reporting declines, 31% attributed this to increasing numbers of fishers, and a further 21% to growing use of destructive fishing gears and methods (Blue Ventures, 2018, unpublished data). The villages in the 2018 study were located not more than 60 km south of Morondava and so it is likely that these results reflect the ongoing situation of overfishing and fisheries declines across the region, particularly given the management situation in the study villages has not changed dramatically since the 2011 study.

The original Laroche et al. (1997) paper highlighted the need for integrating traditional management with government policy and attributing exclusive access rights to traditional fisher groups and noted that the use of marine sanctuaries could be particularly useful for recovery of overexploited populations and for maintaining species diversity and abundance. While there isn’t yet any formal management in place in the Morondava fishery where this study was conducted, since 2003 a growing network of locally managed marine areas, which also include a number of permanent no-take-zones, have been established across the country. There is also growing evidence from the Toliara region that these no-take-zones are helping the recovery of fish biomass (Gilchrist et al., 2020).

A number of fishing villages to the south of the study location have been trialling periodic closures of their mangrove mud crab fisheries with strong social support from fishers (Rocliffe et al, in review), similar management could be supported in the communities engaged in this study. Additionally, the MIHARI (MItantana Harena Ranomasina avy eny Ifotony – translated as marine resource management at the local level) network, a civil society network established in 2012 by locally managed marine area leaders and supporting NGOs, put forward three position statements on three key issues, documented as motions during the 2017 national LMMA forum. These motions advocate for the formalisation of local management (MIHARI network, 2017b), the regulation of gears used in small-scale fisheries (MIHARI network, 2017c) and finally, to address the issue of spatial overlap with commercial shrimp-trawl operations and, the instalment of an exclusive fishing zone for small-scale traditional fishers (MIHARI network, 2017a). These motions are a significant first step towards LMMA leaders having a greater voice in management of their fisheries. These motions are under consideration by the Government, particularly as it aims to triple the area in marine protected areas (World Parks Congress, 2014), and would contribute to the management of the fisheries included in the study.

In conclusion this comparison between the fisheries data collected in 1991 and 2011 highlights a concerning trend in the small-scale fisheries of Morondava and given the lack of management in the region since the study it is likely that this trend has continued. A substantial increase in estimated total monthly landings for the fishery, as well as changes in family and trophic composition of catches supports other evidence of overfishing within the region’s fisheries (Brenier et al., 2011; Gough et al., 2020). It is likely that technological advancement, increases in effective fishing effort, and employment of less selective methods such as beach seining is contributing to larger catches in addition to masking some of the effects of overfishing by enabling fishers to continue to catch target species despite their reduced abundance. Failure to respond to the challenge of managing this and other comparable small-scale fisheries in western Madagascar may have enduring consequences for local food security in one of the poorest countries in the world, particularly as climate change is expected to cause further fisheries declines in the tropics. Regular monitoring, in addition to improvements in the precision of monitoring including the addition of length based measures (Gough et al., 2020) and improvements in monitoring effective fishing effort (Behivoke et al., 2021) would all support improvements in understanding the trends in the fishery and address some of the limitations of the current study.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

Data used in this study were collected by Blue Ventures which has an accord de siege with the Madagascar Government. Data were collected as part of an ongoing project to assess small-scale fisheries in the region of Menabe with a view to improving their management. All fish were sampled as part of the legal fishery with permission granted from each fisher. No animals were harmed for the purposes of this study (normal daily catches of a working fishery were observed), and participant’s names were not recorded. We obtained verbal consent from participants before conducting landings surveys due to low literacy levels in the region. During verbal consent, participants were informed about the survey, its purpose, and how the data would be used. Ethical approval for this research has been given by the Ethics committee at the University of Exeter.

CG, EZ, MA and FM contributed to the conception and design of the study, as well as data collection. CG and FM conducted analyses and CG, AH, AB and BG developed the first drafts of the manuscript and all authors contributed to further manuscript revisions and read and approved the final version.

Funding for this study was providedby Louvain Development and the European Union’s Regional Coastal Management Programme of the Indian Ocean Countries (ReCoMaP/127769/M/ACT/RSA/005).

Authors CG, AH, and MA are or have been employed by the NGO Blue Ventures, and EZ was employed by Louvain Coorporation.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.