seaturtle.org : MTN : ARCHIVES : Sign In

We present data collected from dead sea turtles discovered on the beaches of southern Gabon and northern Congo during daylight foot and ATV beach patrols in 2005 and 2006. Dead turtles were recorded in both years, particularly olive ridley turtle carcasses. While the cause of death could not be determined with certainty in most cases due to decomposition, of those turtles examined by a veterinarian (n=17), 18% appeared to have sustained trauma of a potentially anthropogenic origin. We suggest that the most likely source of such trauma is interactions with fishing vessels and/or gears.

Table 1. Numbers of dead turtles encountered in September and October 2005 and number per kilometer of beach.LNP = Loango National Park; NP = National Park

2005. During September and October 2005, 98 dead turtles were encountered in 145 km of beach from Loango National Park (LNP), the beach between Sette Cama and the town of Gamba, Mayumba NP in Gabon, and Conkouati NP in northern Congo (Table 1 and Figure 1). No dead turtles were reported from Pongara NP in the north of Gabon where an allied research team was operating, using similar methods.

The highest densities of dead turtles were found in Conkouati NP, St. Catherine, and Mayumba NP. Olive ridley turtles were by far the most frequently encountered species, and in Mayumba NP, 55 individuals were discovered during a single patrol in mid-October. No trained veterinary staff was available at this time, so a detailed evaluation of the cause of death could not be undertaken. However, of the two turtles discovered in the Loango South patrol, one had evidence of a puncture wound through the plastron. The patrol team in Mayumba noted that 24 of the 55 olive ridley carcasses showed wounds to the limbs, carapace, or head that could have been sustained pre-mortem. It should be noted, however, that the advanced state of decomposition in many of these animals, and the action of scavengers (e.g. sharks and/or crabs) may have caused extensive post-mortem damage. One fresh leatherback carcass showed clear signs of extensive damage from shark bites.

During a subsequent patrol of the same area of Mayumba NP in early November 2005, only 4 carcasses were encountered, suggesting that the higher numbers of dead turtles during October was related to a wave of turtles arriving in the area and encountering the same or similar cause of death.

Figure 1. Map of the study areas

2006. During the 2006-07 nesting season, a count of dead turtles was made in Mayumba NP between the 5 and 8 October 2006. A patrol was undertaken over 83.5 km of beach from Mayumba town southwards to the Congo border. The distance covered included the same 50 km section patrolled during the 2005-06 season.

In total, 10 dead leatherbacks and 32 dead olive ridleys were encountered. By comparison, over the same 50 km stretch patrolled in the previous season, 9 dead leatherbacks and 15 dead olive ridleys were observed. Interestingly, 15 of the 32 olive ridley carcasses in 2006 were discovered in the first 5 kilometers north of the Congo border, whereas in 2005, carcasses were more evenly distributed.

Following the discovery of this wave of dead turtles, a veterinarian was asked to examine the stranded carcasses for possible causes of death. Between the 16 and 24 October 2006, patrols were undertaken in the 40 kilometers immediately south of Mayumba town, stretching into the Mayumba National Park. In total, 17 adult turtle carcasses were encountered during these patrols (Table 2), all but one of which was already desiccated or decomposing.

Table 2. Dead turtles encountered during veterinarian patrols, October 2006 (sex determined from tail lengths).

Of the 17 carcasses, one olive ridley and two leatherbacks showed wounds consistent with fisheries or boat related interactions. The two leatherbacks showed pre-mortem amputations of a fore-flipper, while a third leatherback had multiple fractures of the cranium consistent with human induced trauma (e.g., during removal from net). It is possible then that 50% (3) of the leatherbacks evaluated may have died due to fisheries/boat related activity.

Two turtles found in a less decomposed state were given either full and partial necropies. No amputations, wounds, or other significant gross findings were found in either case. All other carcasses were too decomposed for more than visual, external observations (the internal organs having largely liquified).

These findings demonstrate a level of mortality among sea turtles on the coast of central and southern Gabon, and northern Congo. Given the conservation status of sea turtles worldwide, this is of concern, and discovering the cause of death is of considerable importance. Several hypotheses can be proposed to explain some or all of the deaths noted, including, a) poisoning from chemical/ hydrocarbon pollutants in water or on land; b) lethal damage caused by offshore oil exploration (seismic surveys); c) natural mortality (viruses, bacteria, other pathogens etc.); d) shark or other predator attack; e) drowning in industrial or artesanal fishing nets (turtles may or may not be subsequently mutilated by fishermen, facilitating their removal from the net).

The advanced state of decomposition of most of the turtles encountered hindered a full study of the cause of death. Thus it was impossible to examine hypotheses ‘a’, ‘b’, and ‘c’. It should be noted, however, that on 8 October 2005, a moderate slick of crude oil washed up in the Mayumba NP. The slick was swiftly dissipated, and was judged not to be severe in the patrol area. Of course, the oil and the turtles themselves may have collided under more damaging circumstances some distance from the beach where they were later discovered. Indeed, the south to north current may have brought many of the turtles a long way from their point of death to their beaching site. However no significant oiling was noted on any of the carcasses.

During the study period, a seismic exploration vessel was operating in the coastal waters opposite Mayumba NP. Little is known of the effects of seismic exploration on sea turtles (McCauley et al. 2000), hindering an examination of this as a potential cause of turtle mortality in West Africa.

Several carcasses showed signs of limb amputation that could have been caused by shark attack. Shark bites may have been sustained by living turtles, leading to death from infection or bloodloss, or sustained post-mortem. However, our examinations suggest that most turtles did not encounter sharks in either state.

The advanced state of decomposition of many of the carcasses also hindered investigation of hypotheses ‘d’ and ‘e’. However, the large number of dead turtles encountered in the south of Gabon in comparison to the lack of carcasses in the north hints at a possible anthropogenic cause of death. The combination of uninjured and mutilated carcasses suggests that in some cases turtles may have died from drowning and/or pre- or post-mortem injury from a sharp object. Fisheries related mortality could explain drownings, as turtles caught in static or trawled nets may be held beneath the surface too long. The observed injuries could have been caused by interactions with fisheries, either through propeller strikes (although we observed none of the characteristic wounds seen in turtles that have been struck by propellers), or through mutilation by fishermen extricating dead or live turtles from their nets. The injuries observed in both seasons were more consistent with the precise use of a knife than with a random mechanical collision. During the 2005-06 season, 5 - 7 industrial trawlers were noted fishing in and around Mayumba NP. These vessels are also commonplace in the coastal waters around Gamba, Loango south, and St Catherine. Furthermore, in November 2005, park staff reported the illegal presence of 15 small fishing vessels close to the coast within Mayumba NP. These boats deployed nets parallel to the coast from vessel to vessel, less than a kilometer from the shore, thus creating an unbroken net barrier to any turtle approaching the coast to nest. While these vessels were subsequently seized, they are known to have been fishing illegally within the Conkouati-Douli NP on the other side of the Congo/Gabon border during October.

We suggest that the most likely hypothesis to explain the majority of the turtle deaths is incidental capture and subsequent mutilation by fishermen present in the area at exactly the moment of greatest concentration of olive ridley turtles along the coast. The distribution of carcasses northwards from this suggested ‘epicenter’ is then explained by the carcasses drifting with the prevailing coastal current, with the possibility of further turtles being killed in fisheries interactions closer to their point of discovery.

Whether or not the rate of mortality can be considered as high for this population cannot be judged with confidence. Certainly it was high relative to a patrolled area in the north of the country, but the lack of accurate data on the number of nesting olive ridley turtles in the region is regrettable. An effort to have teams on the beach earlier in the season will be made in the future. In addition, the number of stranded carcasses observed may be a poor indicator of actual mortality levels, due to the local action of sea currents (Epperly et al. 1996). It is also possible that sea currents themselves are responsible for creating the impression of a ‘strandings hotspot’ by concentrating carcasses in time and space that may have died across a much wider area (Hart et al. 2006).

In future years, it will be beneficial to have veterinary staff present on the main beaches from the start of October in order to perform full necropsies on fresh carcasses, using strict criteria to distinguish anthropogenic from natural causes of death. Indeed the creation of a ‘stranding network’ throughout the area, with personnel trained in necropsy and sampling techniques, is recommended. It is hoped that this can be accomplished in part with the assistance of Gabon’s new turtle conservation partnership, a loose grouping of NGOs and government agencies, conceived with the aim of standardizing methods, producing a national dataset, and producing a united voice for sea turtle conservation.

Both Conkouati-Douli NP and Mayumba NP now have patrol vessels to guard against illegal fishing, but a strict insistence on sizeable fines for vessels apprehended is of critical importance if enforcement is to be effective in the long-term. Some turtle mortality may also result from interactions with deep-sea fisheries such as the long-line fishery for pelagic predatory fish. On-board observers are the best means to evaluate bycatch impacts and the measures designed to mitigate them. However Gabon is still lacking a full observer program to accurately determine the numbers of turtles captured in either the pelagic or the coastal trawler fishery. As a precautionary measure, Gabon is currently considering the deployment of turtle excluder devices (TEDS) for its shrimp industry, and it is hoped that reductions in bycatch in this sector will lead to the adoption of TEDS throughout the trawler fleet.

Acknowledgements: We wish to acknowledge assistance and support from the Mayumba National Park Conservator and National Focal Point for Marine Turtles Madame Solange Ngouessono, the National Council for National Parks, the field technicians of Ibonga, WWF, Aventures Sans Frontières and the Wildlife Conservation Society, Angela Formia of the Gabon Sea Turtle Partnership, and two anonymous reviewers of an earlier draft of this manuscript.

EPPERLY, S.P., J. BRAUN, A.J. CHESTER, F.A. CROSS, J.V. MERRINER, P.A. TESTER & J.H. CHURCHILL. 1996. Beach strandings as an indicator of at-sea mortality of sea turtles. Bulletin of Marine Science 59: 289-297.

FRETEY, J. 2001. Biogeography and Conservation of Marine Turtles of the Atlantic Coast of Africa/Biogeographie et conservation des tortues marines de la cote Atlantique de l’Afrique. CMS Technical Series Publication No. 6, UNEP/CMS Secretariat, Bonn, Germany.

FRETEY, J., & N. GIRADIN. 1988. La nidification de la tortue luth, Dermochelys coriacea (Vandelli, 1761) (Chelonii, Dermochelyidae) sur les côtes du Gabon. Journal of African Zoology 102: 125-132.

HART, K.M., P. MOORESIDE & L.B. CROWDER. 2006. Interpreting the spatio-temporal patterns of sea turtle strandings: Going with the flow. Biological Conservation 129:283-290.

SOUNGUET, G. P., A. FORMIA, & R. PARNELL. (In press). Assessment of leatherback nesting in Gabon by aerial survey. Proceedings of the 25th Annual Symposium on Sea Turtle Biology and Conservation.

MCCAULEY, R.D, J. FEWTRELL, A.J. DUNCAN, C. JENER, M.N. JENNER, J.D. PENROSE, R.I.T. PRINCE, A. ADHITYA, J. MURDOCH, & K. MCCABE. 2000. Marine seismic surveys – a study of environmental implications. APPEA Journal 2000:692-708.