There are 7 extant species of sea turtle worldwide, of which 5—the green (Chelonia mydas), hawksbill (Eretmochelys imbricata), leatherback (Dermochelys coriacea), loggerhead (Caretta caretta), and olive ridley (Lepidochelys olivacea)—inhabit the coastal waters of India and its territories (see Shanker and Choudhury 2006). All except the loggerhead turtle are also found as nesting populations, with 1 or more species nesting in all Indian coastal states and island territories.

Including its islands, India has a coastline of ∼ 8000 km, and monitoring of such an extensive area for nesting, courting/mating, and foraging sea turtles is challenging. Nesting turtle distribution and abundance in all coastal states and territories was systematically surveyed in 2000–2002 (see Shanker and Choudhury 2006), 2003–2005 (see Shanker and Andrews 2006), and again, albeit in a more limited process, in 2010–2012 (see Kurian 2013). However, regular and rigorous estimates of nesting population abundance and trends have been limited to relatively large nesting populations, such as the olive ridley turtle mass-nesting events (arribadas) in Odisha (Shanker et al. 2003; Pandav and Choudhury 2006; Kumar et al. 2013) and the Andaman Islands (Namboothri et al. 2015) and leatherback turtles in the Andaman Islands (Swaminathan et al. 2011). Information about the number of nesting turtles and population trends can also be inferred from the number of clutches of eggs that are relocated to hatcheries to mitigate threats to in situ nests (for examples of recent reports, see Laladhas et al. 2017; Phillott and Kale 2018; Arun 2019; Sundaram et al. 2019).

In comparison, studies on in-water populations of turtles in India have been limited to breeding olive ridley turtles in Odisha (Shanker et al. 2002; Behera et al. 2010, 2018a, 2018b; Sivakumar et al. 2010; Tripathy 2013), postnesting leatherback turtles in the Andamans (Namboothri et al. 2012; Swaminathan et al. 2019), and foraging green turtles in the Lakshadweeps (Tripathy et al. 2006a; N. Kale, pers. comm., May 2020). Offshore observations (Andrews et al. 2006) and isolated survey results (Tripathy et al. 2006b) have also been documented, but the overall body of research, especially for populations on the west coast of India, has been restricted by difficulties of capturing turtles in aquatic habitats and the expense of such studies.

Knowledge of the in-water distribution, abundance, and population trends of sea turtles could be enhanced by compiling the ecological knowledge possessed by local fishers. Fishers' ecological knowledge (FEK) about sea turtles encompasses both traditional ecological knowledge, shared across generations, and local experiential or ecological knowledge, based on observations made during the course of an individual's lived experiences in a particular place (see Johannes et al. 2000; Sagarin and Pauchard 2012; Turvey et al. 2014). In India, turtles and tortoises (the names are used interchangeably in the country's mythology and colloquial language) hold great sociocultural importance. Among Hindus, the turtle is commonly associated with the second incarnation of Lord Vishnu, known as the Kurma avatar, and the vehicle or vahana of the Goddess Yamuna. Another Hindu belief is that the world rests on the back of a turtle or tortoise (Krishna 2010). Therefore, observations of sea turtles by coastal fishers in India, of whom ∼ 75% are Hindu (CMFRI 2012), are likely to be remembered.

Surveys of FEK have been successfully used to study the population biology of and/or threats to sea turtles in Bangladesh (Phillott et al. 2015); Brazil (Braga and Schiavetti 2013); Cameroon, Comoros, Jamaica, Malaysia, Nigeria, Sierra Leone, and Tanzania (Moore et al. 2010); El Salvador and Nicaragua (Wedemeyer-Strombel et al. 2019); and the Mediterranean Sea (Maynou et al. 2011; Panagopoulou et al. 2017), among other places. In India, fishers' knowledge about sea turtle nesting species, season, and population trends has been collated (for examples, see Kakodkar 2006; Shanker and Andrews 2006; Shanker and Choudhury 2006; Kurian 2013), but FEK has rarely been compiled to inform understanding of in-water populations.

FEK can contribute to effective research design, management, and conservation plans and may also be a pathway to community engagement with research, management, and conservation initiatives (Drew 2005). For these reasons, we assessed the potential for local FEK of sea turtle distribution, abundance, and population trends to inform and/or complement purposive sea turtle research and monitoring.

METHODS

Study Site. — Sagareshwar Beach (lat 15.85°N, long 73.63°E) is adjacent to the village of Vengurla in the Sindhudurg District of Maharashtra, India. We selected this location to investigate FEK based on recent observations of local fishers with turtle bycatch (N. Kale, pers. comm., September 2018) and previously described fisher–turtle interactions in the District (Giri and Chaturvedi 2006; Kakodkar 2006; Sanaye and Pawar 2009; Andhare and Hatkar 2015).

Interviews. — We conducted a cross-sectional study from 1 to 5 October 2018 using structured, face-to-face interviews. The 14 interviewers (including one of the authors, P.C.) were multilingual, speaking 2–5 languages (including English, Marathi, Hindi, Konkani, and Kannada), so communication with the potential participants could occur in their preferred language. We followed the International Sociological Association Code of Ethics for Research Procedures (https://www.isa-sociology.org/en/about-isa/code-of-ethics) and conducted interviews only after receiving prior, informed consent about the interview purpose, content, and planned dissemination of findings. Verbal consent was obtained, as written consent by research participants is not commonly practiced in India. Potential participants were advised they could decline to answer any questions and were free to end the interview at any time; interviewers were also observant of body language and offered to cease asking questions if the respondent appeared uninterested or distracted. No compensation was offered in return for responding to the survey questions. We deidentified responses in electronic records and ensured individual anonymity when reporting the study findings. A summary of responses, in the form of an infographic written in the state language of Marathi, was shared with the fishing community after completion of the study.

We adapted an existing marine wildlife structured questionnaire for fishers (Pilcher et al. 2017) comprising open- and closed-ended questions, the latter category including dichotomous, multiple-choice, and scaled (Likert-type and semantic differential) questions. Questions collected information about respondent demographics and fishing activities, activities when turtles were seen, observed species and turtle habitats, and perceived abundance and trends in population size. We offered color images of the 5 turtle species found in Indian waters to facilitate species identification. Participants were also provided with maps of the local coastline and asked to indicate locations of passive sightings and/or active interactions (together referred to as “observations”) with sea turtles. After translating questions from English into the regional languages named above, we ensured that the structure and phrasing of the questions minimized the potential for interviewer bias by translator consensus and back-translation for confirmation by one of the authors (A.D.P.).

We used convenience sampling, a nonprobability method that involved approaching possible participants with close proximity and easy accessibility to Sagareshwar Beach, to recruit respondents (Newing 2010). Fishers potentially worked on more than 1 vessel, so we could not restrict interviews to 1 per vessel. Most interviews occurred on the beach in the early morning and late afternoon as fishers sorted their catch and usually took 20–30 min to complete. As the population size of fishers at Sagareshwar Beach was unknown and could not be used to estimate a target sample size, we conducted surveys according to the principle of saturation (Newing 2010).

To minimize nonresponse bias to potentially sensitive questions (Newing 2010), responses were recorded on hard copies of the survey tool instead of using digital recording methods. For example, nonresponse bias would occur if fishers who had observed sea turtles entangled in their fishing gear declined to be recorded answering questions about when and how they observed turtles and could result in an underestimation of observation frequency. We were also aware of the potential for social desirability bias (Newing 2010), in which fishers might provide responses that they thought were favorable to us. These could include claims of having observed many sea turtles in local waters, thereby resulting in an overestimation of abundance. We assessed the potential for both forms of bias by having the interviewers complete a postinterview survey, which included Likert-type scaled questions about perceived respondent interest in the survey questions, comfort in distinguishing among turtle species, and honesty of responses at the end of each individual interview. We also minimized interviewer bias (Newing 2010), in which the tenor of questions may also influence the respondent's answers through awareness and voice control.

Data Analysis. — Responses to closed-ended questions were summarized using descriptive statistics, and responses to open-ended questions were categorized using emergent coding followed by descriptive statistics if appropriate. Relationships among variables such as fishing activities (e.g., number of days fished in high and low seasons, number of months fished per year) and frequency of turtle observations or perceived trend in turtle numbers could not be examined using inferential statistics due to low sample size and resulting counts of < 5 in contingency tables. Instead, data are presented in figures for visual assessment. SPSS version 25 was used to compare means of continuous numerical data (e.g., distance traveled offshore to fish) among categorical variables (e.g., frequency of turtle observations in the past year) using a Kruskal-Wallis test (McKillup 2012).

RESULTS

In total, 93 fishers gave informed consent and contributed responses during the study. An informal, conservative estimate by the authors suggested that 1 or more fishers from > 80% of the vessels on Sagareshwar Beach had been interviewed during the study period. Some fishers declined to participate, and some fishers did not answer all questions; reasons for such included concerns about responses containing potentially sensitive information, unexpectedly having to commence or recommence work, or lack of interest in the study. Hence, the findings presented provide the total number of respondents (n = x) to individual questions and the proportion (%) of responses as indicated or the mean, standard deviation, and range.

Respondent Demographics. — All respondents were male, and fishing was the only occupation of 82.8% (n = 92). Almost equal proportions of fishers were aged between 26 and 50 yrs and 51 and 75 yrs. The largest group of respondents (32.9%, n = 92) had fished in local waters for 11–20 yrs, and all (n = 93) were second- or third-generation fishers in their families (Table 1).

Table 1. Demographic characteristics of fishers who contributed their ecological knowledge about sea turtles at Sagareshwar Beach, Vengurla.

View Fullscreen

Fishing Activities. — Individual fishers at Sagareshwar Beach often used multiple types of gear, the most predominant of which was gill nets (72.5%, n = 91). Other gear types (i.e., purse seines, beach seines, trawl nets, hook and line, bottom longline, and cast nets) were each employed by < 30.0% of respondents (Fig. 1). The waters fished ranged from the shoreline to 100 km offshore, and fishers traveled up to 4 hrs to their fishing areas (Table 2). Fishing activities by the majority of respondents occurred throughout the year except during the months of June and July. The number of days fished each week decreased during the low season compared with the high season (Table 2).

View Figure

• Download as Powerpoint
• Download Figure

Table 2. Fishing activity characteristics of fishers who contributed their ecological knowledge about sea turtles at Sagareshwar Beach, Vengurla.

View Fullscreen

Fishers' Observations and Perceptions of Sea Turtles. — Observations of marine wildlife such as sea turtles, dolphins, whales, whale sharks, other shark species, and sea snakes while fishing were common (97.6%, n = 84). Sea turtles had been seen by the majority of respondents (96.5%, n = 88) during their fishing career and were most commonly observed while fishing, when turtles were accidentally caught in their nets, or coming ashore to lay their eggs. Turtles were also encountered while laying eggs or stranded on beaches (Fig. 2). Respondents (n = 88) most frequently observed olive ridley turtles (70.5%), followed by hawksbill (61.4%), green (38.6%), loggerhead (26.1%), and leatherback (18.2%) turtles in decreasing order of frequency. When respondents were asked how they distinguished between the different sea turtle species, one respondent replied that turtles may be “green and then yellow but striped like a tiger. Some of them are plain like a stone. Some have pointed shells like armour. There are some that are triangular and lopsided. But for us, turtles are turtles” (Anonymous, pers. comm., October 2018, translated from Marathi). Fishers who provided local names for turtles (n = 24) used Kaasav (87.5%), Kasho (8.4%), Kodam (4.2%), and Kachua (4.2%) interchangeably for all species; only 1 respondent provided more than 1 vernacular name (Kaasav and Kasho) (Table 3).

View Figure

• Download as Powerpoint
• Download Figure

Table 3. Vernacular names given to different sea turtles by fishers (n = 24) in the Sindhudurg District of Maharashtra.

View Fullscreen

Respondents were reluctant to indicate locations at which they had observed or interacted with sea turtles while fishing on a map. The respondents who did identify areas where turtles were found (51.8%, n = 85) named the adjacent turtle nesting beach of Wayangani (lat 15.87°N, long 73.61°E). Some respondents (40.0%, n = 55) thought areas utilized by turtles changed over time. Very few (3.6%, n = 83) had observed courting/mating sea turtles.

Less than a third of respondents identified the months when sea turtles were observed, with most sightings reported to occur during August and September and the least from February to May (Fig. 3). When asked about their cumulative observations, the largest cohorts of respondents indicated they had seen sea turtles several times in the last year (57.5%, n = 80) and frequently in their lifetime (47.1%, n = 87). Respondents who observed turtles more frequently in their lifetime had seen more turtles in the last year than fishers who had observed turtles less frequently in their life (Fig. 4). However, the frequency of observing turtles in the last year did not vary with the number of months fished during the year (Fig. 5), number of days fished in the low and high seasons (Fig. 6), or distance fished offshore (Fig. 7; H = 0.909; n = 5, 10, 27, 5; p > 0.05).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

When asked to estimate the number of sea turtles in local waters, equal proportions of fishers (n = 50) indicated a population size of 10–100 or > 100 (21.7% each), but many respondents (39.8%) could not provide an estimate. The majority of respondents (55.8%; n = 77) perceived sea turtle numbers in local waters to have declined; perceptions of population change varied with the combined frequency of turtle sightings during their lifetime and in the last year (Fig. 8). Their reasons for perceived population change are outlined in Table 4.

View Figure

• Download as Powerpoint
• Download Figure

Table 4. Respondents' (n = 37) perceptions of reasons for changes in number of sea turtles in local waters of Sagareshwar Beach, Vengurla.

View Fullscreen

Interviewer's Perceptions of Potential Bias in Survey Responses. — Respondents appeared moderately (42.7%) to very (49.4%) interested when responding to survey questions. Interviewers perceived respondents to be reasonably (44.8%) to very (21.8%) comfortable when distinguishing between sea turtle species and somewhat (33.7%) to very (64.0%) honest when answering questions about observations of sea turtles. Hence, no responses were filtered or discarded.

DISCUSSION

Respondents to our survey were part of a small-scale fishery, operating mainly gill nets in coastal waters. FEK would be informed by multigenerational experiences and the time (number of years and days per week in the high and low seasons) spent fishing in local waters as their main occupation. The decrease in fishing activities during June–July can be attributed to a uniform ban on all fishing activities and fishing vessels in Maharashtra's waters of the Indian Exclusive Economic Zone from 15 June to 31 July (dates subject to annual change) for management and conservation of fisheries resources and sea safety during the southwest monsoon (Rajesh 2013). Based on fisher experience and observed levels of interest during the survey, we believe the findings discussed below to be a reliable measure of FEK about local in-water sea turtle populations.

More than 95% of respondents had observed sea turtles, identifying green, hawksbill, leatherback, loggerhead, and olive ridley turtles from the images provided. Maharashtrian fishers' use of carapace shape and coloration to distinguish among different sea turtle species has been previously documented (Giri and Chaturvedi 2006), and these characteristics, among others, can be reliable indicators of species (Shanker et al. 2003). The presence of all 5 species in local waters was confirmed by local expert Suhas Toraskar (pers. comm. to A.D.P., October 2018, translated from Marathi). The results presented in this study represent the first published record of loggerhead turtles in the district and only the second for the state (see Hatkar et al. 2019); this turtle was distinguished from the similar olive ridley turtle as the “longer olive ridley” (S. Toraskar, pers. comm. to A.D.P., October 2018, translated from Marathi). The other species are more extensively reported from the state. Giri and Chaturvedi (2006) and Andhare and Hatkar (2015) reported that leatherback turtles have not been seen in the area for 10–15 yrs and 20 yrs, respectively, but Kakodkar (2006) and Karve et al. (2020) described more recent encounters. The remaining 3 species have been more recently described and in greater abundance (Giri and Chaturvedi 2006; Kakodkar 2006; Sanaye and Pawar 2009; Andhare and Hatkar 2015). The olive ridley turtle is the most common species of sea turtle along the entire Maharashtra coast (Giri and Chaturvedi 2006). Previous reports suggested that green turtles were more commonly observed in district waters than hawksbill turtles (Venkataraman and Milton 2003; Giri and Chaturvedi 2006; Kakodkar 2006; Andhare and Hatkar 2015). It is possible 1) that respondents in our study or those cited confused hawksbill and green turtles, 2) that the distribution and/or numbers of hawksbill turtles had been previously underestimated, or 3) that the numbers of green turtles have decreased. Species and possibly size class and sex of local sea turtles could be confirmed by providing fishers with waterproof cameras and flexible tape measures and requesting that fishers photograph bycatch or stranded turtles.

Fishers interviewed at Sagareshwar Beach used different local names for sea turtles than those interviewed in a previous study in the Sindhudurg District (Table 3). Kurma/Kurmah is a Sanskrit name and Kachua a Hindi name given to turtles and tortoises (Hinduism does not distinguish between the two taxa) (Krishna 2010). Variation in local names for different species over time may 1) signify an evolution in language over time, 2) signify a change in the prevalence of less common species over time (i.e., leatherback to loggerhead), 3) signify a shift in the cultural significance of sea turtle species among the local fishing community, or 4) be influenced by the language in which the interview was conducted. This topic could be of interest to linguistic anthropologists.

FEK about sea turtle distribution, abundance, and populations trends did not distinguish between species. Sea turtles were most likely to be observed between the months of August and December, when fishing effort is also highest. This corroborates Kakodkar's (2006) findings in which fishers from Sindhudurg District observed sea turtles in the postmonsoon season (after September). As the local nesting season commences in December (Katdare and Mone 2003; Giri and Chaturvedi 2006), turtle sightings at this time could also be attributed to the arrival of breeding adults for courtship and mating. We heard very few observations of mounted pairs of sea turtles despite respondents fishing in waters proximate to known nesting beaches for olive ridley turtles. Little is known about the breeding habitat of olive ridley turtle populations on the west coast on India, all of which are solitary nesters, and most studies on the east coast have focused on arribada populations. Aggregations (“reproductive patches”) of breeding males and females usually form in the 1–2 mo prior to the nesting season (Rajagopalan et al. 1996; Behera et al. 2010; Tripathy 2013). The aggregations associated with arribada populations can be observed in waters adjacent to their nesting grounds, from the nearshore wave zone to < 5 km offshore of Gahirmatha (Rajagopalan et al. 1996; Pandav et al. 2000; Behera et al. 2010) and 2–5 km off Rushikulya (although surveys stopped at 5 km; Tripathy 2013), but olive ridley turtles may also mate far offshore (see Plotkin 2003). The relatively low number of nesting sea turtles in the Sindhudurg District (< 30 nests annually; Giri and Chaturvedi 2006; Sanaye and Pawar 2009; Katdare 2013) could have limited the opportunities for observations of mounted pairs by respondents in our study; alternatively, waters through which respondents traveled or fished may not be used by breeding turtles. Future studies wishing to identify marine habitat important for mating turtles in this region may need to rely on in-water line surveys (Tripathy 2013) or the use of “drones” (unmanned aerial vehicles; Rees et al. 2018) rather than FEK.

We received no information about specific areas for turtle feeding. Respondents may not have observed feeding turtles due to small population sizes or because foraging habitats and diet items suitable for the different sea turtle species may not have been prevalent within the respondent's fishing area (see Bolten 2003; Jones and Seminoff 2013); hence, the 5 species were observed during ranging and foraging movements but not while feeding in specific areas or on specific items.

Sea turtle abundance also could not be assessed from FEK due to the wide disparity in estimates provided and because the majority of respondents indicated they were uncertain. However, the largest cohort of fishers reported reductions in local sea turtle population size and attributed it to fishing activities. This perception is likely based on respondents' observations of turtles while fishing and entangled in nets. Bycatch rates for this area have not been quantified but are under current investigation by the authors. Most respondents who perceived an increase in the local population thought it was a response to reduction in the take of turtles and their eggs.

In another study in this area, Kakodkar (2006) interviewed 8–12 people (fishers, teachers, students, businesspersons, and laypersons) at each of 13 sites in the Sindhudurg District, compared with our 93 fishers at 1 location; we suggest that the greater sample size of fishers with lived ecological knowledge at an individual location in our study revealed more nuanced information about local turtle populations. A broader study encompassing multiple fisheries and locations could provide greater context for the status of turtle populations in south Maharashtra waters. Sea turtles in India are protected by the Wildlife (Protection) Act of 1972, which bans consumption of turtle meat and eggs, and by state fishery regulations that restrict fishing in coastal waters during the turtle breeding season. However, similar to many populations worldwide, sea turtles and their eggs in India are threatened by fisheries-related mortalities and exploitation, respectively (reviewed by Shanker and Choudhury 2006; CMS/IOSEA Secretariat 2014; Phillott and Kale 2018; Manoharakrishnan and Swaminathan 2020). Factors contributing to potential declines in sea turtle populations at Sagareshwar Beach can be inferred from fishers' observations of turtles either while fishing, entangled in nets, laying eggs, or stranded onshore, as reported in this study. These observations suggest that bycatch in fishing gear, potentially leading to stranding, may be a threat to local turtles. As many respondents had also interacted with turtles at adjacent nesting beaches and illegal take of sea turtle eggs has been a previous problem for turtle conservation in the district (Giri 2001; Giri and Chaturvedi 2006; Giri et al. 2006; Kakodkar 2006; Katdare 2013), sale and/or consumption of sea turtle eggs should also be quantified. No fishers reported seeing hunting of turtles, but a turtle has been observed being carried ashore and into a village at Sagareshwar Beach (N. Kale, pers. comm., September 2018), and so the potential for targeted hunting and consumption of bycatch turtles should also be investigated.

We suggest that the FEK shared by fishers during our study indicates the value of such a survey in contributing to knowledge about local sea turtle species and population trends but potentially not distribution and abundance if only small and/or low-density populations are present in local waters. The individual mapping activity did not collect information about turtle distribution in our study, potentially because the fishers interviewed do not use maps or map apps during their fishing activities, or they were reluctant to identify their fishing areas. These reasons could be overcome if small groups of fishers (rather than individuals) were asked to name or indicate locations of turtle observations.

Even if few in number, turtles observed by fishers on the west coast of India contribute to regional management units (RMUs; Wallace et al. 2010) or distinct population segments (DPSs; Seminoff et al. 2015), which are regarded as threatened by the International Union for Conservation of Nature (IUCN; Table 5). Four of the 6 high-risk/high-threat RMUs (Table 5) are also included in the world's 11 most endangered RMUs (Wallace et al. 2011).

Table 5. IUCN status and risk-threat assessment of regional management unit (RMU) or distinct population segment (DPS) to which sea turtles observed by respondents at Sagareshwar Beach, Vengurla, potentially contribute.

View Fullscreen

The IUCN status of these RMUs and DPSs and/or risk and threats scores (Wallace et al. 2011) and the paucity of information about in-water populations of sea turtles on the west coast of India should be considered in association with FEK when determining if more focused investigations should follow. We recommend that systematic in-water population studies and quantification of threats should involve local fishers and their gear in a manner conducive to active engagement of the fishing community with sea turtle research and conservation.