Information regarding the population status of sea turtle nesting aggregations along the Atlantic coast of Africa is relatively scarce in the literature. Long-term data sets are essential to understanding population trends (Franklin 1989; Magurran and Queiroz 2010), but most of the conservation and research projects in the region are young and/or hindered by financial, political, and logistical problems that make such data sets difficult to obtain (Formia et al. 2003). Five of the 7 sea turtle species have been reported in the Gulf of Guinea waters and are known to nest on the Atlantic coast of Africa (Fretey 2001; Formia et al. 2003). These species have been classified by the International Union for Conservation of Nature (IUCN) Red List as follows: hawksbill (Eretmochelys imbricata) turtle as critically endangered; green (Chelonia mydas) turtles as endangered; and loggerhead (Caretta caretta), olive ridley (Lepidochelys olivacea), and leatherback (Dermochelys coriacea) turtles as vulnerable (IUCN 2015). Until recently, the global status of leatherback turtles in the IUCN Red List was listed as critically endangered. This global status has now been changed to vulnerable and the 7 subpopulations in the world have been assessed separately (IUCN 2015). The Southeast Atlantic population is currently listed as Data Deficient, highlighting the lack of information available on population trends of sea turtles in this region.

Nesting leatherback turtles occur from Mauritania to Angola (Márquez 1990; Butynski 1996; Fretey et al. 2007; Weir et al. 2007; Fitzgerald et al. 2011), with one of the world's largest leatherback aggregations in Gabon (36,185–126,480 clutches annually; Witt et al. 2009). It has been reported that the leatherback populations in this region are either stable or increasing (Girondot et al. 2002; Dutton et al. 2005; Godgenger et al. 2009). Green turtles have been reported to nest in all Gulf of Guinea countries (from Mauritania southward) and Corisco Bay has been shown to be an important foraging ground for this species (Formia 2002; Formia et al. 2003), with another foraging population in Banc D'Arguin, Mauritania (Godley et al. 2010). The largest known green turtle nesting colony occurs in Guinea Bissau (7000–29,000 clutches annually; Catry et al. 2009). Nesting olive ridley turtles occur from Senegal to Angola, including the islands of Bioko, São Tomé, Corisco, Mbanye, and Hoco, with one of the most important olive ridley nesting populations in Gabon (2370–9814 clutches annually; Fretey et al. 2004, 2012; Bernardo and Plotkin 2007; Weir et al. 2007; Godgenger et al. 2009; Metcalfe et al. 2015). Hawksbills on the Atlantic coast of Africa are present in the waters of all countries from Mauritania to Angola including the islands of Bioko, São Tomé, and Principe (Fretey et al. 2002; Rader et al. 2006; Weir et al. 2007; Fitzgerald et al. 2011). Loggerheads have been known to nest in Mauritania, Senegal, and Sierra Leone, and the largest aggregation occurs in Cape Verde (Brongersma 1995; Ross 1995; López-Jurado et al. 2000; Monzón-Argüello et al. 2010; Marco et al. 2011).

Bioko Island's southern coast has been classified as one of the most important nesting areas for 4 of the 5 species of sea turtles (leatherback, green, olive ridley, and hawksbill turtles) within the Gulf of Guinea (Castroviejo et al. 1994; Fretey et al. 2007; Tomás et al. 2010). First published proof of sea turtles nesting on Bioko Island dates back to Eisentraut's visit in 1963 (Eisentraut 1964). Based on interviews with the local people from the village of Ureca, it was recorded that possibly 4 species of turtles nest on the 19-km stretch of Bioko Island beaches. During his visit, Eisentraut observed green turtles nesting and a hawksbill nest hatching on the nesting beach near Ureca. Dr. H. Wermuth suggested that the other 2 species might be olive ridley and the loggerhead turtle (Eisentraut 1964). However, based on descriptions given to Butynski and Koster in 1989 and the fact that the local people seem unfamiliar with a loggerhead, it is more likely that the leatherback was the fourth species (Butynski and Koster 1989).

Systematic sea turtle surveys on Bioko Island started in the 1996–1997 nesting season, when 18 men and 16 women from the nearby village of Ureca were trained by Asociación Amigos de Doñana (AAD) to survey the southern beaches (Tomás et al. 2010). Data collection continued for 2 nesting seasons (1996–1997 and 1997–1998) followed by a 2-yr hiatus. Since 2000, the same Urecan community members continue to survey the nesting beaches under the direction of Bioko Biodiversity Protection Program (BBPP) in partnership with the Universidad Nacionál de Guinea Ecuatorial (UNGE) during the dry season, from September to April (Rader et al. 2006; Tomás et al. 2010; Fitzgerald et al. 2011). Based on the data collected in 1996–1998 and 2000–2004, it has been suggested that Bioko Island has the second highest number of nesting leatherback and green turtles in West Africa after Gabon and Guinea Bissau, respectively (Rader et al. 2006; Tomás et al. 2010). The number of nesting olive ridleys seems to be increasing and hawksbill nesting females are critically low (Rader et al. 2006; Tomás et al. 2010). It is clear that Bioko Island represents an important nesting ground for sea turtles in the Gulf of Guinea; however, the short-term nature of previous studies limits our understanding of population trends on Bioko.

Current threats affecting sea turtles on Bioko Island are similar to those facing sea turtles in other parts of the world. Sea turtle eggs and meat have been used for human consumption, with nightly adult turtle take estimated as high as 200–300 green turtles in the 1940s (Castroviejo et al. 1994). In the 1980s, most common turtles in the market were still green turtles (100 turtles/night) for their meat and hawksbill turtles for their shells (Castroviejo et al. 1994). As a result of efforts of AAD (1996–1998) and more recently BBPP (2000–2014), and the fact that there are smaller numbers of nesting green turtles on the beach, the take of adult green turtles has decreased to 50 turtles/yr and there has been virtually no egg take during this period (Fitzgerald et al. 2011). Incidental captures in commercial fisheries operating in the Gulf of Guinea, specifically bycatch in long lines and gillnets, is a serious threat for resident turtles and nesting adults (Witt et al. 2011; Fossette et al. 2014). The Gulf of Guinea is also the focus of extensive and rapidly increasing oil exploration, which poses the potential risk of pollution and habitat destruction (Witherington et al. 2009). Although loss of suitable nesting beaches due to coastal developments, pollution, and climate change are increasingly serious threats for turtles throughout the world (Formia et al. 2003; Hamann et al. 2010), the nesting beaches on Bioko Island's southern coast have remained undeveloped because of the surrounding rugged topography and lack of safe harbors. The completion of a paved road in November 2014 has created direct overland access between the nesting beaches and the Island's second-largest city for the first time (Fig. 1). The number of turtles and eggs taken is now expected to increase as a result of the easy access, making the establishment of long-term population trends prior to beach development critically important.

View Figure

• Download as Powerpoint
• Download Figure

Long-term data sets are essential to understanding the population dynamics and ecology of an organism (Franklin 1989; Magurran and Queiroz 2010). In order to estimate population size in marine turtles, monitoring efforts have often been confined to a small portion of their demography (i.e., nesting females) because of logistic and funding constraints. These assessments usually rely on the number of nests and/or the number of tracks left by the nesting female turtles (Whiting et al. 2013). Although not ideal, these estimates can provide a basic assessment of the population abundance and detect trends in the population over time (Whiting et al. 2013).

This study further confirms the importance of Bioko Island's nesting beaches for the Southeast Atlantic by analyzing trends in sea turtle encounters over 14 yrs, and presenting results of the first extensive tagging program on the island. Specifically, the objective of this study is to evaluate temporal trends in the annual number of sea turtle encounters (leatherback, green, olive ridley, and hawksbill turtles) from 2000 to 2014, and to estimate the number of females (leatherback and green turtles) nesting seasonally on Bioko Island's southern beaches. These long-term data fill a critical need for sea turtle conservation in a data-deficient, yet globally significant, area.

METHODS

RESULTS

Regression of total annual encounters suggested that leatherback encounter rates decreased (leatherback, R² = 0.59; F_1,66 = 5.22; p = 0.025) and the total number of olive ridley encounter rates increased (olive ridley, R² = 0.26; F_1,66 = 24.56; p < 0.001) annually on Bioko Island from the 2000–2001 to 2013–2014 seasons. There was no statistically significant linear trend in the number of green (R² = 0.001; F_1,66 = 0.072; p = 0.79) and hawksbill (R² = 0.009; F_1,66 = 0.58; p = 0.45) turtle encounter rates from the 2000–2001 to 2013–2014 seasons.

A total of 43,860 leatherback, 16,778 green, 1731 olive ridley, and 85 hawksbill turtle encounters were recorded on the southern beaches of Bioko Island from the 2000–2001 to 2013–2014 seasons. The annual number of encounters ranged from 350 to 6353 for leatherback, 248 to 2571 for green, 66 to 191 for olive ridley (Table 1 and 2), and 2 to 17 for hawksbill turtles (Fig. 2). Because only the total number of turtle encounters were available during the 2000–2001 to 2007–2008 nesting seasons, and not the number of nests, these data have been presented in a separate table (Table 2) from the data collected during the 2008–2009 to 2013–2014 nesting seasons (Table 1). During the 2008–2009 to 2013–2014 nesting seasons, 25.8% of all leatherback, 28.5% of all green, and 18.4% of all olive ridley encounters were false crawls (Table 1). The annual proportion of false crawls ranged from 18.4% to 34.3% for leatherbacks, from 10.5% to 42.8% for greens, and from 10.9% to 26.1% for olive ridley turtles (Table 1).

Table 1. Number of leatherback, green, and olive ridley encounters, nests, percentage of false crawl (FC) and estimated number of nesting females on Bioko Island's southern beaches from the 2008–2009 to 2013–2014 seasons.

View Fullscreen

Table 2. Number of leatherback, green, and olive ridley encounters and estimated number of nesting females on Bioko Island's southern beaches from the 2000–2001 to 2007–2008 seasons.

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

All 5 nesting beaches were surveyed during the dry season, with the annual number of days patrolled ranging from 83 to 210, with a mean of 144.61 ± 12.82 d/beach. Each ‘survey' includes all of the patrols required to estimate the number of turtles encountered on a single day on all beaches along the southern coast of Bioko Island. The estimated number of encounters per survey for leatherback turtles was 20.52 ± 2.71 ranging from 4.00 to 38.67 annually, green turtles 8.27 ± 1.28 ranging from 2.83 to 21.46, and olive ridley 1.29 ± 0.16 ranging from 0.61 to 2.73 (Fig. 3).

View Figure

• Download as Powerpoint
• Download Figure

Leatherback encounter rates were higher on Beach E followed by Beaches C and D, whereas green turtles were encountered mainly on Beaches A and B followed by Beach C (Fig. 3). Olive ridley and hawksbill turtles were encountered mainly on Beach C (Fig. 3). There was a significant annual difference (p < 0.001) in the seasonal pattern of encounters of leatherback, green, and olive ridley turtles per month (Fig. 4). Peak nesting time for leatherback, green, and olive ridley turtles occurred mainly in the months of December and January (Fig. 4).

View Figure

• Download as Powerpoint
• Download Figure

A total of 755 leatherback turtles have been PIT-tagged in 2008 (January–February) and from the 2008–2009 to 2013–2014 seasons on Beach E. In addition, 17 leatherbacks were PIT-tagged on Beach A and 18 on Beach B during the 2013–2014 season. Two leatherback turtles that were tagged on Beach E in 2013–2014 season were also observed to nest on Beach B. From the 790 tagged leatherbacks, only 48 turtles were recorded to have returned from previous years with an average remigration interval of 3.27 ± 0.20 yrs. To calculate ECF and CI, 209 PIT-tagged turtles were considered that fit our criteria (successful nest, > 3 nests, interval < 18 d, see Methods). Mean ECF was 5.5 ± 0.11 nests/yr and CI was 10.6 ± 0.07 d. Based on calculated ECF in this study, the estimated number of leatherback females (No. of nests/ECF) ranged from 42 to 444 annually. However, the authors recognize that the extremely low leatherback tag return (6.1%) and proximate nesting grounds in the region may add an error to the approach.

A total of 279 green turtles have been flipper-tagged on Beach A (156 turtles) and Beach B (123 turtles) in the 2013–2014 nesting season. From 279 flipper-tagged turtles, 29 were observed to nest on both Beach A and B, and 117 turtles were observed to nest twice or more on the beach. The OCF was 2.63 ± 0.08 and the estimated number of green turtles ranged from 63 to 649 during the 2008–2009 to 2013–2014 seasons (Table 1). The estimated number of olive ridley turtles based on ECF of 3 ranged from 22 to 53 during the 2008–2009 to 2013–2014 seasons (Table 1). The estimated number of nesting turtles during the 2000–2008 seasons ranged from 214–266 (2007–2008) to 759–942 (2001–2002) nesting leatherback turtles, 207–325 (2004–2005) to 523–819 (2001–2002) nesting green turtles, and 16–20 (2001–2002) to 47–57 (2006–2007) nesting olive ridley turtles (Table 2).

During this 14-yr study 2 leatherback sea turtles were recorded with tags from other study sites. On 20 November 2008 a leatherback with the PIT-tag number (296 807 769) was recorded on Beach E. The origin of this tag has not been identified. On 20 February 2014 a leatherback with flipper-tags (KUD6933, KUD6932) was observed to nest on Beach E. The inconel, KUD series tags were used by PROTOMAC to tag turtles on Praia Grande on Principe Islands, the Democratic Republic of São Tomé and Príncipe (Loureiro et al. 2011).

DISCUSSION

Bioko Island's southern beaches are an important nesting site for both leatherback and green turtles. The annual number of leatherback encounters (350–6353) are second highest in the Gulf of Guinea region after Gabon and green turtle encounters (248–2571) are the second highest after Guinea–Bissau (Catry et al. 2002, 2009; Witt et al. 2009; Tables 1 and 2). Historically, Bioko Island's nesting beaches have been mainly visited by green and hawksbill turtles (Eisentraut 1964; Butynski and Koster 1989). In the past, there has been debate and confusion about whether leatherback turtles even existed on Bioko Island (Eisentraut 1964; Butynski and Koster 1989). Regardless, there are more leatherback encounters (43,860) than green (16,778), olive ridley (1731), and hawksbill turtles (85) combined during this 14-yr study period. In addition, the ECF of 5.5 ± 0.11 clutches per female leatherback per season with a clutch interval of 10.6 ± 0.07 d is similar to other published estimates (Reina et al. 2002; Hitipeuw et al. 2007; Sarti Martínez et al. 2007; Tapilatu et al. 2013). Based on clutch frequencies calculated in this study, the estimated number of leatherback females ranged from 42 to 444, green turtles from 63 to 649, and olive ridley turtles from 22 to 53 annually (Table 1).

Although annual fluctuations are evident, linear trends suggest an overall decline in the number of leatherback encounters from the 2000–2001 to 2013–2014 seasons on Bioko Island. This decline is in contrast to nearby leatherback nesting beaches of Gabon where they are thought to be either stable or increasing (Girondot et al. 2002; Dutton et al. 2005; Godgenger et al. 2009). There was no significant trend in the number of green turtle encounter rates from the 2000–2001 to 2013–2014 seasons. The trends for other green turtle populations in this region are still largely unknown. Olive ridley and hawksbill turtle encounters are low on Bioko Island. Only 1731 olive ridley and 85 hawksbills were recorded on the southern beaches of Bioko Island during the study period. Previous data suggested that olive ridley populations are increasing (Rader et al. 2006; Tomás et al. 2010), which coincides with the long-term data presented here.

Nesting on Bioko Island occurs annually from September through April. Nesting peaks during December and January each year for leatherback, green, and olive ridley turtles (Fig. 4), following the same patterns as turtles nesting in Gabon and Congo (Godgenger et al. 2009; Witt et al. 2009). When comparing annual fluctuations in the number of leatherback encounters on Bioko Island with the number of leatherback clutches calculated from an aerial survey in Gabon (Witt et al. 2009) and beach surveys in Congo (A. Girard, pers. comm., April 2011), similar patterns emerge. The number of clutches for all 3 areas decreased after 2002–2003 and began to increase again after the 2005–2006 season, followed by a drastic decline during the 2008–2009 season (Witt et al. 2009; A. Girard and A. Formia, pers. comm., April 2011). One possible explanation for these fluctuations may be related to the productivity of the foraging grounds. Leatherbacks usually migrate to pelagic foraging areas where their prey is more abundant than on tropical coasts (Davenport 1998; Hays et al. 2004; Caut et al. 2008). The similar patterns between Bioko, Gabon, and Congo suggest that these nesting populations may be sharing foraging areas and changes in the productivity of the foraging grounds may be affecting recruitment (Wallace et al. 2006; Saba et al. 2008). Another possible explanation is that these fluctuations are due to annual fluctuations of offshore fishing efforts in the region. For instance, the declines observed in all nesting populations during the 2008–2009 season was correlated with substantially higher levels of offshore fishing in the region in 2008 (A. Formia and A. Girard, pers. comm., April 2011).

Spatial distribution of nesting sites on Bioko's southern beaches differ among the 4 species, with leatherbacks tending to nest on the eastern beaches and green turtles nesting predominantly on the westernmost beaches (Rader et al. 2006; Tomás et al. 2010). Small numbers of both leatherbacks (n = 2) and green turtles (n = 29) nested on more than one nesting beach during the 2013–2014 nesting season. Beach selection in sea turtles is thought to depend mainly on offshore cues and beach characteristics (Mortimer 1982; Provancha and Ehrhart 1987; Horrocks and Scott 1991; Wood and Bjorndal 2000). Beaches A and B, where mainly green turtles are encountered, are high wave energy, steep beaches, with a short (20–30-m) distance from the high tide line to the vegetation line and a steep rocky cliff within 5 m behind the vegetation, and contain large boulders and rocks throughout the beach and offshore. Beaches C, D, and E on the other hand, where leatherback turtles are encountered, are low wave energy, open, and wide (30–60 m), with more gentle slopes (30°) and little to no offshore obstructions closer to Beach D and E. To reduce the probability of injury and energy expenditure, leatherbacks usually prefer to emerge near areas with deep water and limited shallow-water obstructions such as rocks, sand bars, and coral reefs (Pritchard 1971; Eckert 1987). Physical and chemical characteristics of the sand may also play a significant role in nest site selection and warrant further investigation; however, initial investigations showed no difference in sand grain size on Beach B compared with Beach E, where the greatest number of leatherback encounters occur (Suss 2013).

This study represents the first extensive tagging program to take place on Bioko Island. A total of 790 leatherbacks were tagged on Beach E during this study. Only 48 of these individuals (6.1%) have returned to nest, with an average remigration interval of 3.3 yrs. The remigration interval is more similar to Atlantic leatherbacks that return to the nesting grounds every 2–3 yrs, as compared with the Pacific leatherback, where remigration interval is usually 4 yrs (Santidrián Tomillo et al. 2007; Caut et al. 2008). The proportion of remigrants seen on Bioko Island is drastically lower than the proportion observed elsewhere, which is usually between 14% and 52.5% (Spotila et al. 2000; Reina et al. 2002; Dutton et al. 2005; Hilterman and Goverse 2007; Santidrián Tomillo et al. 2007). We are confident that the observed low proportion of remigrants on Bioko Island is not due to tag loss or faulty PIT-tag readers. AVID tags and the universal PIT-tag readers are used worldwide and are highly reliable (Dutton and McDonald 1994; Reina et al. 2002). This low proportion of remigrants on Bioko Island could be explained by the fact that we were mainly tagging leatherbacks on one nesting beach (Beach E), when in fact a significant number of leatherbacks are also nesting on Beaches D and C. Another reason for this could be high mortality of adults due to incidental capture by fisheries (Fossette et al. 2014). Recent studies show that leatherbacks nesting on the West–Central African coast have a high susceptibility to potential bycatch in longline fishing in the southern–equatorial Atlantic (Witt et al. 2011; Fossette et al. 2014). Variation in number of leatherback turtles seen on Bioko may be also attributed to turtle take in local artisanal fisheries (Tomás et al. 2001; Formia et al. 2008). Satellite telemetry studies would provide a better overview of patterns of movement of Bioko Island leatherbacks (Witt et al. 2011).

Another possible explanation for the low proportion of remigrants is that Bioko Island leatherbacks may be part of the same populations that nest elsewhere, mainly in Gabon and Congo. Fretey and Lescure (1998) suggested that the leatherbacks could maintain fidelity to a nesting region, but they could also potentially nest in areas quite distant from year to year. This would be consistent with the similar annual fluctuations in encounters mentioned above. Over the 7 yrs of our tagging efforts, we did not encounter any leatherbacks with either PIT-tags or flipper-tags from Gabon. To our knowledge, PIT-tag readers used in Gabon are not universal and would not have been able to detect the tags used on Bioko leatherbacks (A. Formia, pers. comm., April 2011). We did however detect a leatherback turtle with tags from Praia Grande on Principe Islands, the Democratic Republic of São Tomé and Príncipe. Genetic analyses of the population structure of sea turtles in the Gulf of Guinea region will be necessary to further investigate this possibility.

Until recently, the global status of leatherback turtles in the IUCN Red List has been critically endangered. This global status was recently changed to Vulnerable and the 7 leatherback subpopulations in the world have been assessed separately (IUCN 2015). The Southeast Atlantic population, where Bioko Island's turtles are found, has been assessed as Data Deficient. Hence, the data presented here are critical for appropriate assessment of this subpopulation, and will have significant conservation implications given the importance of this rookery in the Gulf of Guinea region.

Turtle and egg take on Bioko Island has been generally limited because of the protection provided by geographic isolation. However, the construction of a paved road from the second largest city to the nesting beaches now threatens the existence of these populations. For instance, during the 2007–2014 nesting seasons there were reports of 3 poached leatherback turtles on the southern beaches of Bioko Island, whereas ≥ 12 leatherbacks were poached on the nesting beaches during the 2014–2015 nesting season alone. There are daily reports of people using the road to access the beaches and quickly transport both meat and eggs to markets in larger cities. In addition, small hunting camps have already been built on Beach D and are occupied on a daily basis. The ease of access and cost-effectiveness of transporting turtles to markets via the new road will undoubtedly elevate pressures on these nesting beaches beyond previous levels. In the past, agreements with members of the local village of Ureca and creation of job opportunities associated with conservation nongovernmental organizations were sufficient to decrease take on the nesting beaches. The newly gained access to the nesting beaches creates the opportunity for individuals outside of the local village to become involved in the turtle trade, making conservation efforts with the Urecanos less effective.

In Equatorial Guinea, the National Institute of Forestry Development and Protected Area Management and Ministry of Fisheries and the Environment manage the protected area where these 4 species of turtle nest on a 19-km stretch of beach. The government of Equatorial Guinea has issued law 8/1988 (regulating hunting of wildlife) and presidential decree 183/87 (regulating fishing), which offer protection for sea turtles. The enforcement of these current legislations by the government of Equatorial Guinea is urgently needed if the threats to this nesting population are to be reduced. Short-term but effective solutions, such as implementation of checkpoints on the newly constructed road, monetary fines, property seizure (hunting equipment, cars, fire arms, etc.), and beach guards may safeguard Bioko Island's nesting turtles in the wake of recent and ongoing nesting beach development.