Methods

Two independent methods were used to assess marine turtle deaths in Bahrain waters: direct monitoring of dead marine turtles (from 18 September 2007 to the 15 January 2009) and survey of fishermen (conducted at the end of each quarter in 2008).

Monitoring Dead Turtles

Public participation to report carcasses was solicited via posters and pamphlets. Prizes were provided for reporting dead turtles and for retrieving turtle carcasses from the sea. The value of the prizes was relatively low to encourage the participation but not to encourage turtle mortalities. People were reminded that marine turtles are protected in Bahrain waters by legislation (Decree No. 3 for 2003). In addition, the study team also monitored dead turtles by random visits to coastal and sea areas. To avoid duplication, carcasses were removed from the inspection sites immediately after data collection. Data collected included the following: postmortem decomposition stage (Limpus 2007; Flint et al. 2009: I—comatose, II—fresh, III—medium, IV—advanced, V—high-advanced, and VI—bones); photographs; species (field identification and confirmed from photograph); sex (mature males only); curved carapace length (CCL) in centimeters; and GPS coordinates of the inspection or collection site. Coordinates of offshore carcasses were determined from navigation charts based on fishermen's description of natural and human-introduced land marks.

Fishermen Questionnaire Survey

The sea area was divided into five subareas, whereas the coastal line was divided into four zones (Fig. 1). At-sea Areas 4 and 5 were separated by the main island of Bahrain; Areas 2 and 4 were separated by Fasht Al Adhom (Fig. 1). Area 3 was defined as the shallow Fasht Al Jarim which was separated from Area 2 by deeper water. Area 1 was the deepest offshore area.

View Figure

• Download as Powerpoint
• Download Figure

Fishermen from the eastern and western coasts spend most of their time and effort in Areas 4 and 5, whereas fishermen from the northern coast spend more time and effort in Areas 1, 2, and 3 (Abdulqader 2006). Similar numbers (mean  =  302, SD  =  5.4) of fishermen were randomly interviewed at the end of each quarter of 2008, which provided similar representation among coastal areas (mean  =  75.5, SD  =  15.1). This represents between 3.6% and 3.8% of total number of Bahrain's fishermen (Al-Radhi et al. 2004).

Interviewees were requested to provide the total number of dead turtles during the previous three months by month and area. Because data included a considerable number of zero observations, the nonparametric Kurskal-Wallis Test was used to determine sighting differences in areas and time periods (Zar 1974). Data were analyzed by SPSS (Version 17) statistical program.

Results

A total of 57 dead marine turtles were inspected and collected during the last quarter of 2007, including 52 (91%) green turtles (Chelonia mydas), 3 (5%) hawksbill turtles (Eretmochelys imbricata), 1 (2%) loggerhead turtle (Caretta caretta), and 1 (2%) olive ridley turtle (Lepidochelys olivacea). A total of 88 carcasses were inspected and collected in 2008, including 63 (72%) green turtles, 22 (25%) hawksbills, and 3 (3%) loggerhead turtles. In addition, the study team was not able to inspect some carcasses reported by the public (16 in 2007 and 34 in 2008).

Numbers of inspected dead turtles were not similar throughout the study period (Table 1). Number of green turtle carcasses inspected in the fourth quarter of year 2007 almost equals the total of carcasses inspected in 2008. From the 52 inspected, 58% were found in last decomposition stage. In the first two quarters of 2008, low numbers of green turtle carcasses were found. The number increased by an order of magnitude in the third quarter, wherein about 63% of these cases were found in decomposition Stages II to IV. The number of green turtle carcasses reduced during the fourth quarter of 2008, and the proportion in the earlier stages of decomposition reduced to 50%.

Table 1. Numbers of marine turtle carcasses inspected in each quarter of the study and the numbers in decomposition Stages II, III, and IV.

View Fullscreen

Low numbers of green dead turtles were inspected in February (n  =  3), March (n  =  2), April (n  =  1), May (n  =  2), June (n  =  1), September (n  =  4), and November (n  =  2) of 2008. No cases were found in December 2007 and 2008 or in January 2008. The numbers of inspected dead green turtles increased in July (n  =  17), August (n  =  19), and October (n  =  12). In July and August 2008, more than 60% of the green turtles (22 of 36) inspected were found in early decomposition stages (Stages II to IV).

Hawksbill turtle carcass numbers indicate a different pattern; highest number was found in fourth quarter of 2008 (Table 1). About 77% of these cases were in decomposition Stages II to IV, and all (100%) remaining cases inspected in different time periods were also found in these stages. Highest number of dead hawksbill turtles was found in October 2008 (n  =  7), 6 of these turtles were in decomposition Stages II and III. Lower numbers were found for December 2007 (n  =  3), February (n  =  2), March (n  =  3), April (n  =  1), July (n  =  1), August (n  =  2), November (n  =  1), and December (n  =  2) 2008 and for January 2009 (n  =  3). No cases were found in September, October, and November 2007 and in January, May, June, and September 2008.

Four dead loggerhead turtles were inspected during the study, 2 in the fourth quarter of 2008 and 1 each in the first quarter of 2007 and 2008 (Table 1). Monthly, 1 was inspected in December 2007, 1 in March 2008, and 2 in November 2008. Loggerhead turtles inspected in December 2007 and November 2008 were in decomposition Stages III and IV, respectively, whereas the carcass inspected in March 2008 was in the final decomposition stage. One dead olive ridley turtle was collected in December 2007; this turtle was in decomposition Stage IV (Table 1).

The locations of inspected marine turtle carcasses in 2007 and 2008 (Figs. 2, 3, respectively) showed higher numbers of dead green turtles in the areas north of Fasht Al Adhom than in southern areas where the main shrimp trawling grounds are found, also identified here as Areas 3 and 4 (Fig. 1). A few green turtle carcasses were found in the western area and north of Fasht Al Adhom (Areas 2 and 5), whereas no dead marine turtles were inspected from deeper offshore area (Area 1, Fig. 1).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

All inspected hawksbill carcasses were from the northern areas 2 and 3 (Figs. 1–3), with higher numbers in Area 3. Six and one inspected hawksbill turtles were found dead in big wire traps in 2008 and 2007, respectively, as reported by the fishermen who collected these turtles. These turtles were caught in Area 2.

All four inspected dead loggerhead turtles (CCL  =  46, 87, 92, 93 cm) were collected from the northern area (Area 3). The single olive ridley turtle (CCL  =  63 cm) was collected dead from Area 2.

CCL frequency distribution for carapace-intact dead green and hawksbill marine turtles indicates that turtles of nonreproductive and reproductive size inhabit the area around Bahrain (Fig. 4). Based on the minimum recorded CCL class (81–85 cm) for reproduction (Miller 1989; Al-Merghani et al. 2000), the size class frequency for green turtle carcasses indicates that 59 (69.4%) were smaller than the minimum recorded reproductive size class and 26 (30.6%) were in the reproductive size range. At least 18 (72%) of the recovered hawksbill carcasses were below minimum recorded reproductive size (based on 56–60 cm; Miller 1989; Al-Merghani et al. 2000). One of the four loggerhead carcasses (25%) was below minimum reproductive size (based on 81–85 cm; Limpus and Limpus, 2003); the other three specimens were of mature size. The CCL of the single olive ridley carcass was greater than the minimum breeding size (based on 56–60 cm; Qureshi 2006).

View Figure

• Download as Powerpoint
• Download Figure

The Questionnaire

For months and sea areas, 94% to 100% of the interviewees did not report seeing dead marine turtles. Most of remaining interviewees reported seeing 1 to 3 dead turtles during the last three months, and a few observed 4, 6, and 7 dead turtles. Higher percentages of interviewees reported sighting 1 or 2 dead turtles in August and September (5.6% and 5.4%, respectively). Area 3 had the highest percentages (5.3% and 4.3%, respectively) for the third and fourth quarter surveys. All of the interviewees who fished in Area 1 reported zero sightings of dead turtles.

The nonparametric Kruskal-Wallis Test indicated significant statistical differences in the number of dead marine turtle sightings in the different sea areas (X²  =  65.219, df  =  4, Asymp. Sig. < 0.001, n  =  6040), months (X²  =  48.884, df  =  11, Asymp. Sig. < 0.002, n  =  1208), and quarters (X²  =  14.674, df  =  3, Asymp. Sig. < 0.001, n  =  3624). The Kruskal-Wallis Test ranked Area 3 as the most important area, whereas Areas 1 and 5 were ranked as the least important areas (Area Ranked Order 3, 4, 2, 5, 1). The Kurskal-Wallis Test ranked year-quarters in the following order of importance: third (July through September), fourth (October through December), first (January through March), and second (April through June) quarter. When the numbers of dead turtles inspected by this study were used as the second criterion to rank the order of importance of the year-quarters the result was the same order of importance.

Discussion

Although no significant marine turtle nesting occurs on Bahrain's islands, the large seagrass areas and reefs provide important foraging habitat (Miller and Abdulqader 2009). The weather is characterized by cool winters (mean 18°C, January) and hot summers (mean 37°C, July and August). The wind in the vicinity of Bahrain is generally from the NNW; however during April, July, August, and September, an important component comes from the NNE and E (Windfinder.com: wind statistics of Bahrain from June 2005 to May 2011). Fishing is one of the major traditional industries in the country, and 29 direct fisheries have been identified (Abdulqader 2006). Most do not have seasonal closures; trawling and some gillnetting are exceptions. Trawling is banned for four months each year, usually from mid-March to mid-July. The primary trawling grounds are located to the north and east of the main island Areas 3, 4 (Figs. 1–3).

Abdulqader (2010) reported that an average of over 300 turtles was caught each season in shrimp trawls, and it is reasonable to assume that a proportion of these turtles die accordingly. Elsewhere in the world, shrimp trawl nets have been identified as important source of mortality to marine turtles (e.g., Poiner and Harris 1996; Robins 1995; Guinea and Whiting 1997; Garcia-Gaudillo et al. 2000; Epperly 2003; Casale et al. 2004; Chan 2006; Sasso and Epperly 2006; Cox et al. 2007).

The deaths of 57 and 88 marine turtles in Bahrain's water during 2007 and 2008, respectively, likely represent minimum mortalities because additional mortalities probably occurred that were not inspected or sank before being washed ashore (e.g., 16 and 34 carcasses reported but not inspected in 2007 and 2008, respectively).

Between 72 and 91% of turtle carcasses inspected in 2008 and 2007 were green turtles. In 2008, the highest numbers of carcasses were inspected during July, August, and October. Most of these were found in early decomposition Stages II–IV, indicating that most of these mortalities had occurred within the respective months. These months coincide with the start of the trawling season (in 2008 from mid-July).

Further, the results of questionnaire identified the third quarter (July through September 2008) as the most important time period in respect to dead turtle sightings. The lowest incidence of observed dead turtles reported occurred in the second quarter, before the opening of the trawling season. These data suggest that shrimp trawl fishery is responsible for most green turtle mortalities in Bahrain waters.

Further evidence that the shrimp trawl fishery is responsible for green turtle mortalities in Bahrain comes from the distribution of the locations of these dead turtles. Almost all green turtle carcasses were collected from locations down wind and close to shrimp fishing grounds (Figs. 2, 3). In contrast, no dead turtles were reported in Area 1, whereas one and two dead turtles were reported in the Area 5 (Figs. 1–3). Similarly, the questionnaire indicated that dead turtles were not sighted in Area 1 throughout 2008. Areas 1 and 5 were ranked as being the least important. In addition, Areas 1 and 5 do not contain important shrimp fishing grounds. This implicates the shrimp trawl fishery as an important source of mortality of the green turtle population.

In December 2007 and 2008 and January 2008 and 2009, no dead green turtles were inspected in Bahrain waters. This might be attributed to environmental factors (e.g., temperature) or to seasonal changes in the source of threat (i.e., fishing). Abdulqader (2010) found that the mean monthly catch rates in Bahrain's shrimp trawl fishery were not significantly different. In Bahrain waters, temperature continues to decline through December into January (Abdulqader 1999). The decline in the water temperature may have forced green turtles to move to other areas away from the source of the threat or the reduced temperature may have reduced the activity of the turtles (Higgins 2003). Temperature is known to have significant impact on marine turtles (Mrosovsky 1980; Spotila et al. 1997; Higgins 2003).

All dead hawksbill turtles were collected from the northern areas (Areas 2 and 3). The highest number of dead hawksbill turtles was found in October 2008 where 77% of these were found in decomposition Stages II to IV. This indicates that this species has different mortality pattern than the green turtles. Further, seven of the inspected dead hawksbill turtles (28%) were found dead in wire traps. This provides strong evidence that fisheries using wire traps are responsible for many of the hawksbill turtle mortalities in Bahrain waters. Similarly the wire trap fisheries was responsible for the death of at least one olive ridley turtles in Bahrain.

All loggerhead turtle carcasses were collected from Area 3 which includes part of Bahrain shrimp fishing area. This may indicate that shrimp trawls contribute to the mortality of loggerhead turtles. Elsewhere in the world, loggerhead turtles were captured in shrimp trawls (Casale et al. 2007; Cox et al. 2007; Wallace et al. 2008). Other fisheries (e.g., gillnets and longline fisheries) may be contributing in turtle mortalities in Bahrain waters; however, data are lacking.

Summary

It is confirmed that four marine turtle species are present in the territorial waters of Bahrain. Relatively large populations of green and hawksbill turtles are known to forage in Bahrain waters (Miller and Abdulqader 2009). The four loggerhead carcasses confirm that loggerheads are present in Bahrain waters, at least in low numbers. The mortality of one olive ridley in a wire trap represents the first record for the species in Bahraini waters.

Timing of marine turtle mortalities and the distribution of carcasses suggest that the shrimp trawl fishery was responsible for most of green and loggerhead turtle mortalities in Bahrain waters. Dead hawksbill turtles were found in lower numbers and in different time periods. Seven hawksbill turtles and one olive ridley turtle carcasses were recovered from wire traps used by fin-fish fisheries. It appears that fin-fish wire traps were responsible for hawksbill and olive ridley turtle mortalities in Bahrain waters.

Recommendations

Recognizing that Bahrain territorial waters provide foraging habitat for four species of marine turtles and to reduce the impact of fisheries on these turtle populations the following recommendations are made: (1) introduce legislations regulating the fisheries and requiring the use of technology that minimizes the impact of fishing gear on marine turtles, with an emphasis on shrimp trawl nets and wire fish traps; (2) provide support for the use and training of fishermen in the use of modified fishing gears; (3) conduct an on-board observer program on shrimp trawl vessels; (4) conduct a questionnaire-based survey to assess the capture of marine turtles in the wire traps, gillnets and longlines fisheries; and (5) resume the monitoring of marine turtle mortalities on a continual basis.