Incidental Capture and Mortality of Olive Ridley Turtles (Lepidochelys olivacea) in Commercial Trawl Fisheries in Coastal Waters of Orissa, India
ABSTRACT
A shrimp trawl fishery study was conducted in the coastal waters of Orissa, India, from November 2001 to March 2002; 76 trawls in water of 6–35 fathoms resulted in the capture of 26 olive ridley turtles (Lepidochelys olivacea). Most turtle captures occurred within 5 km of the shoreline and at a depth of less than 10 fathoms, emphasizing the need for greater protection of these nearshore habitats. Data revealed a strong relationship between tow time and mortality of turtles.
Sea turtles face threats at every stage of their life cycles from natural and human-induced factors—both on land where females nest and hatchlings emerge and in offshore waters where they spend their life as hatchlings, juveniles, and adults. Of all known causes, fishery-related mortality is, by far, the most important cause of mortality of adult sea turtles and is considered the single largest threat to sea turtle populations globally. Five species of endangered sea turtles inhabit India's coastal waters, of which four are known to occur in Orissa: olive ridley (Lepidochelys olivacea), green (Chelonia mydas), hawksbill (Eretmochelys imbricatta), and leatherback (Dermochelys coriacea). All species occurring in India are legally protected in Schedule I of the Indian Wildlife (Protection) Act, 1972.
Orissa has three major mass nesting grounds for L. olivacea: Gahirmatha, Devi River mouth, and Rushikulya River mouth. There is global concern regarding the future of olive ridleys in Orissa, particularly due to large-scale mortality of sea turtles from incidental capture-related mortality (Rajagopalan et al. 1996; Pandav et al. 1998; Shanker and Mohanty 1999).
Sea turtle strandings related to shrimp fisheries on beaches of the southeastern United States was first documented in the early 1970s (Ulrich 1978; Murphy and Hopkins 1989). Hillestead et al. (1982) documented that sea turtles are caught in fisheries around the world, especially in trawl nets, and also reported that incidental or accidental capture in fishing nets was emerging as a global problem for sea turtle survival. Magnusson et al. (1990) emphasized that drowning in trawl nets in the Gulf of Mexico and along the US Atlantic coast is the major cause of sea turtle mortality due to human activities; this was also previously noted by Henwood and Stuntz (1987) who demonstrated a strong positive relationship between tow time and incidence of sea turtle mortality in the shrimp fishery.
Gove et al. (2001) highlighted the impact of shallow-water shrimp trawlers operating in Safola Bank, which caught between 1932 and 5436 sea turtles every year; these figures placed the shrimp fishery as one of the main sources of marine turtle mortality in central Mozambique. Guinea and Whiting (1997) inferred that flatback, loggerhead, and olive ridley turtles were being caught in Queensland waters. Dash and Kar (1990) and Pandav et al. (1997) have suggested that uncontrolled, mechanized fishing in areas of high sea turtle concentration has resulted in large-scale mortality of olive ridleys during the last two decades in Orissa. However, no quantified information about sea turtle and shrimp fishery interactions is available from the coastal waters off Orissa, and information on actual capture levels is relatively scarce.
The intent of this study was to estimate the incidental captures of sea turtles in the shrimp fishery off Orissa coastal waters. The rationale of the study was to determine which species were affected, what size classes, where captures occurred, and catch per unit effort (CPUE) of sea turtles in stratified study zones. We also wanted to determine the relationship between tow time and turtle mortality and to assess the relationship between sea turtle strandings and shrimp trawling patterns. The study also recommends conservation measures to safeguard sea turtle populations of Orissa while they are in offshore waters.
Study Area
Orissa, an important maritime state on the east coast of India, has a coastline of 480 km and a continental shelf area of 24,000 km forming 8% and 4.5%, respectively, of the nation's total area. Orissa accounts for an annual production of 6858 tons of shrimp and 125,000 tons of other marine fisheries. A total of 545,109,387 USD are earned per annum by export of marine products from Orissa (Anonymous 1997). High sea turtle concentrations occur in Brahmini-Bhaitarini, Devi, and Rushikulya river mouths. Thus, the coastline between Brahmini-Baitarini and the Devi river mouths was chosen for sampling (Fig. 1). For easy accessibility to these areas, Paradip was selected as the base station to carry out the experimental trawling. The selected coastline was classified into three zones of 35, 55, and 60 km (Fig. 1):(Zone I) the coastal waters of Gahirmatha that extend from Dhamra River to Barunei, the mouth of River Hansua and forms the eastern boundary of the Bhitarkanika Wildlife Sanctuary, (Zone II) the coastal waters of Paradip that stretch from Barunei to the mouth of Jatadhara, near Paradip, the only natural harbor along the Orissa coast where the Mahanadhi enters Bay of Bengal, and (Zone III) the coastal waters of Devi that stretches from the Jatadhra River mouth to the Kadua River mouth. Operation of mechanized and nonmechanized boats in Orissa (Anonymous 1997) is given in Table 1.
Citation: Chelonian Conservation and Biology 5, 2; 10.2744/1071-8443(2006)5[276:ICAMOO]2.0.CO;2
Methods
Data collection was carried out between November 2001 and March 2002, and was conducted both at sea and along the coast. Stranded turtles on shore were counted and marked with white paint every 2 weeks from November 2001 to March 2002, and the size class and sex were recorded. Offshore data collection took place between January 2002 and March 2002. A fishing trawler (M.V. Telaban) was hired from the Orissa Fisheries Department. The authors also accompanied other private vessels to document the incidental capture of sea turtles in the private vessels. For this study, otter board single net trawling was used uniformly, because such trawlers are common along the Orissa coast. Trawling speed was maintained between 2 to 4.5 knots/h. Trawl time was limited to 1–2 hours in the research vessel to avoid sea turtle mortality due to drowning and 3–4 hours in the observer vessels, where the fishing protocols duplicated those of private vessels. The geographic locations of all the sampling areas (start and end position of the trawler) were taken with Garmin 12 Global Positioning System (GPS). Other parameters, such as start and end times of trawls and depth of the trawl, state of the sea, weather condition, and turtle captures and sightings, were recorded. Turtles caught in the net were kept aboard the vessel, and their physical condition was examined (i.e., alive, comatose, or dead). All turtles were identified by species and sex; and their curved carapace length (CCL), curved carapace width (CCW), straight carapace length (SCL), and straight carapace width (SCW) were measured, after which they were released at sea. The activity of comatose turtles was studied until they recovered. CPUE also was calculated according to the methods suggested by Jamir (1999). For each trawl sampled, effort (E) was standardized to reflect hours towed with a 50-m net using the formula:
Where nets = number of nets towed in a trawler, length = head rope length of the net, min = min fished.
Where R = CPUE, T = number of turtles captured, E = Effort (standardized to 50-m net hours).
To determine the relationship between tow time and mortality, turtles captured were classified as alive, comatose, or dead (Magnusson et al. 1990). Because comatose turtles are already in a stressed condition, cannot swim actively, and will eventually drown, comatose turtles were counted as dead turtles. Tow duration of the trawls that resulted in turtle captures was plotted against the physical condition of the turtles to check the relationship between tow time and mortality. To document the size class of the turtles both caught in the net and stranded on shore, morphometric measurements of the turtles were analyzed using statistical software (SPSS-PC). The relationship between sea turtle stranding and shrimp fisheries was determined by correlating the number of stranded turtles with the number of shrimp vessels operating in each zone.
Results and Discussion
Seventy-six trawls were carried out from January to March 2002, of which 52 were conducted in the research vessel and 24 in observer vessels. Of the 26 turtles caught, 21 were caught by the research vessel and 5 by the observer vessels. All turtles caught during the study were olive ridleys, of which 13 were males and 13 were females. Morphometric measurements of the turtles captured in trawl nets by the study vessels are given in Table 2. Trawls per zone, depth of trawls, duration of trawls, trawl distance, and distance of trawl from the nearest shoreline are given in Table 3.
During the study, 20 turtles (71.9%) were captured less than 5 km from the shore, and the remaining 6 sea turtles (28.1%) were recorded at varied distances beyond the 5-km line (Fig. 2). As observed during the study, day trawlers operating from fishing bases in Orissa trawl mostly within a distance of 5 km from the shoreline. Although the Orissa Marine Fishing Regulation Act prohibits mechanized trawling within 5 km from shoreline, the trawl operators in Orissa continue to operate in the prohibited zones. This may result in large-scale captures of turtles in nearshore areas.
Citation: Chelonian Conservation and Biology 5, 2; 10.2744/1071-8443(2006)5[276:ICAMOO]2.0.CO;2
Of the 26 turtles captured, 18 were captured at depths of less than 10 fathoms. All the captures in zone I were between 6 to 8 fathoms; in zone II, captures occurred between 8 to 18 fathoms; and in zone III, most captures were recorded in less than 10 fathom depth (Fig. 3), which was well supported when CPUE was calculated and analyzed. CPUE was calculated for every trawl in each study zone and was comparatively much higher in areas of less than 10 fathom depth and lower in areas of 14 to 35 fathoms depth, where normally turtles do not spend much time.
Citation: Chelonian Conservation and Biology 5, 2; 10.2744/1071-8443(2006)5[276:ICAMOO]2.0.CO;2
Most turtles caught on trawls of 1–2 hours were alive (mortality rate 11.1%). Trawls of more than 2 hours duration had a higher mortality rate of 66.6% (Fig. 4), which supports the study of Henwood and Stunz (1987), and Magnusson et al. (1990), who depicted the death rate of sea turtles in trawl nets at near zero until the tow time exceeds 60 minutes; mortality rose with increasing tow time to around 50% for tow times in excess of 200 minutes. Most trawls for the research vessel were restricted to 60-minute duration, and none exceeded 200 minutes. However, in the observer vessel, the minimum trawl duration was 180 minutes and maximum was 6 hours 45 minutes.
Citation: Chelonian Conservation and Biology 5, 2; 10.2744/1071-8443(2006)5[276:ICAMOO]2.0.CO;2
A total number of 11,593 (2714 male, 7147 female, 1732 unknown) stranded olive ridley turtles were counted between Paradip to Sonapur (Fig. 5). Mortality rate was higher during December and January, which may be because of the peak fishing season along the Orissa coast. Most dead females encountered during February and March were egg-bearing. The Paradip and Konark coasts accounted for the highest mortality, which can be attributed to high abundance of turtles and high intensity of trawling operations. Although sea turtles also are abundant on the Rushikulya coast due to the presence of the mass nesting site, strandings were much lower compared to other coastal areas due to less intensity of fishing in that area. Sea turtle strandings are mostly the result of a complex interaction between sea turtles and trawl or other fishing gear mortalities. Stranding occurs when a carcass starts to decompose, becomes inflated, and floats. Wind and currents bring a few of them to shore, but most dead turtles never reach the shore because they are either eaten by scavengers, break apart, or burst and sink. Stranding has been used as an index of sea turtle mortality from shrimping activities (Caillouet et al. 1991) but is subject to error due to many factors including wind, currents, etc. Murphy and Hopkins-Murphy (1989) reported only 6 of 22 tagged loggerhead carcasses at sea turned up as strandings. Thus, sea turtle strandings represent only a small fraction of actual mortality.
Citation: Chelonian Conservation and Biology 5, 2; 10.2744/1071-8443(2006)5[276:ICAMOO]2.0.CO;2
Conclusions
This study concludes that more than 70% of turtle captures in shrimp trawl fisheries in Orissa occurred within 5 km from shore. These were in generally shallow waters, less than 10 fathoms. A steep increase occurred in the mortality rate when the duration of the trawl was extended beyond 2 hours. Strandings were higher in zones where there is heavy marine fishing and high sea turtle abundance. These findings will help in the management of the sea turtle crisis that exists today. Because the peak turtle mating and breeding period coincides with the main fishing season (October–May), the turtles are most vulnerable at this time. Strict enforcement of the Orissa Marine Fishing Regulation Act in nearshore areas and in areas of high sea turtle concentration, and mandatory use of Turtle Excluder Devices (TEDs) in all fishing zones would reduce turtle mortality.
Study area on the coast of Orissa, India.
Distance from shore at which turtle captures occurred.
Depth at which captures occurred.
Relationship between tow time and mortality.
Sea turtle strandings and spatiotemporal pattern of the shrimp fishery.
