University of Ghana http://ugspace.ug.edu.gh DISTRIBUTION AND ECOLOGY OF NESTING SEA TURTLES IN GHANA. BY BENDORF TETTEY AMITE YE Thesis p resen ted to the D epartm en t of O ceanography and F isheries, U niversity of G hana, in p a rtia l fulfillm ent for the aw ard o f a M as te r o f Philosophy D egree in F isheries Science October, 2002 University of Ghana http://ugspace.ug.edu.gh DEDICATION This work is affectionately dedicated to the entire AMITE YE FAMILY That they may live to appreciate the power of knowledge and pursue it to its highest pinnacle. University of Ghana http://ugspace.ug.edu.gh DECLARATIO N I hereby declare tha t this w ork is a product of my own research w ork and th a t no p a r t has ever been presented as thesis elsewhere. BENDORF TETTEY AMITE YE (Student) D ate:................................................ / < 1 / A. K. ARMAH G. A. DARPAAH (SUPERVISOR) (MEMBER OF SUPERVISORY COMMITTEE) Date:.. I t f .L . Date: . . / / . !!’. . ........... .................. f m i ■■ — m i .................................................rrt— - Dr. P. KJ OFORI-DANSON ( h e a d d f d e p a r t m e n t ) Date:.........•2'o c 2- 11 University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENTS My sincere gratitude goes to my supervisor and lecturer Mr. A.K. Armah o f the Department o f Oceanography and Fisheries, University o f Ghana, whose supervision and useful suggestions led to the successful completion of this work. My thanks go to the NGO, Resource and Environment Development Organisation (REDO) and the Global Environment Facility - Small Grants Programme (GEF-SGP) for providing me with financial support to undertake this research work. I am also indebted to Mr. George Darpaah, Mr. George Wiafe and Mr. F.K.E Nunoo (members o f the supervisory committee) o f the Department o f Oceanography and Fisheries for their invaluable suggestions, assistance and reading through my work. I wish to express my thanks to the Caribbean Conservation Corporation, Gainesville, Florida, for selecting me to participate in the 1999 Green Turtle Monitoring Program at Tortuguero, Costa Rica, where I was introduced to modem sea turtle research techniques. I would also like to register my special thanks to Dr. Isaac K. Quaye o f the Noguchi Memorial Medical Research Institute, University o f Ghana, for his encouragement. I am \ery grateful to REDO members and others at Old Ningo, New Ningo and Prampram for assisting me with my fieldwork. Finally, to Mr. and Mrs. E. K. Amiteye, my parents, 1 say thanks so much, for your encouragement and financial support throughout the master’s program. - B.T. AMITEYE University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DEDICATION ............................................................................... 1 DECLARATION............................................................................... « ACKNOWLEDGEMENTS................................................................ ™ TABLE OF CONTENTS....................................................................... iv LIST OF FIGURES........................................................................... vii LIST OF TABLES............................................................................. viii LIST OF PLATES............................................................................. ix ABSTRACT....................................................................................... x CHAPTER ONE: GENERAL INTRODUCTION AND OBJECTIVE OF STUDY 1. INTRODUCTION 1 1.1. LEGAL STATUS OF SEA TURTLES 3 1.2. ECONOMIC IMPORTANCE OF SEA TURTLES 3 1.3. THREATS TO SEA TURTLES 6 1.3.1. NATURAL ENEMIES 6 1.3.2. HUMAN DIMENSIONS 7 1.4 OBJECTIVE OF STUDY 11 CHAPTER TWO: LITERATURE REVIEW 2.0 OCCURRENCE OF SEA TURTLES 12 2.1 GLOBAL DISTRIBUTION 12 2.2. OCCURRENCE IN AFRICA 17 iv University of Ghana http://ugspace.ug.edu.gh 1 o 2.3. OCCURRENCE IN GHANA lO 2.4 REPRODUCTION IN SEA TURTLES 19 2.4.1 AGE AT FIRST MATURITY 20 2.4.2 COURTSHIP BEHAVIOUR 21 2.4.3 NESTING PERIODICITY 21 2.4.4 NEST SITE FIDELITY 22 CHAPTER THREE: MATERIALS AND METHODS 3.0 STUDY AREA 24 3.1 DATA COLLECTION 25 3.1.1 USE OF QUESTIONNAIRES 25 3.1.2 SEDIMENT GRANULOMETRY 25 3.1.3 FIELD OBSERVATIONS AT INTENSIVE STUDY AREA 27 CHAPTER FOUR: 4.0 DISTRIBUTION OF SEA TURTLES IN GHANA 30 4.1 INTERVIEWS BY QUESTIONNAIRE METHOD 30 4.1.1 INTRODUCTION 30 4.1.2 MATERIALS AND METHOD 30 4.1.3 RESULTS 31 4.1.4 DISCUSSION 34 4.2. BEACH SEDIMENT CHARACTERISTICS AND DISTRIBUTION OF SEA TURTLES IN GHANA 38 4.2.1 INTRODUCTION 38 4.2.2 MATERIALS AND METHOD 40 V University of Ghana http://ugspace.ug.edu.gh 4.2.3 RESULTS 4.2.4 DISCUSSION 49 CHAPTER FIVE: 5.0 REPRODUCTIVE ACTIVITY OF SEA TURTLES IN THE INTENSIVE STUDY AREA 52 5.1 INTRODUCTION 52 5.2 MATERIALS AND METHOD 54 5.2.1 COLLECTION OF BIOMETRIC DATA 56 5.3 RESULTS 59 5.3.1 NESTING SEASONALITY AND RELATIVE OCCURRENCE 59 5.3.2 EMERGENCE PERIOD, CLUTCH SIZE AND PERCENTAGE HATCHING SUCCESS 59 5.3.3 BIOMETRIC MEASUREMENTS 74 5.4 DISCUSSION 83 CHAPTER SIX GENERAL DISCUSSION, CONCLUSION AND RECOMMENDATIONS 88 LITERATURE CITED 103 vi University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure Page 1 Map of Ghana showing various sampled stations 26 2 Map o f intensive study area, showing the various zones 28 3 Map showing sea turtle nesting activity along the coast o f Ghana 35 4 Cumulative plot for sediment from Old Ningo beach 45 5 Grain size versus sorting coefficient 46 6 Dendrogram showing classification analysis o f sand samples from various beaches 48 7 Method of measuring CCLmin and CCWmax 58 8 Relative occurrence of each nesting species in the intensive study area 60 9 Relative percentage of each nesting species in each zone o f the intensive study area 61 10 Total number o f nests recorded in each zone during the study period 62 11 Percentage false crawls and nests destroyed by dogs in each zone 63 12 Sea turtle nesting activity at intensive study area 64 vii University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table Page 1 Sampled beaches, showing sediment type and nesting activity 44 2a Data on nesting activities in zone A 66 2b Data on nesting activities in zone B 67 2c Data on nesting activities in zone C 68 2d Data on nesting activities in zone D 69 2e Data on nesting activities in zone E 70 2f Data on nesting activities in zone F 71 3 Emergence period and hatching success o f Olive ridley 72 4 Emergence period and hatching success o f Green turtle 73 5 Emergence period and hatching success for Leatherback 73 6 Biometric measurements for Olive ridley and Leatherback turtles 81 7 T-test on novel method used in estimating clutch size 82 viii University of Ghana http://ugspace.ug.edu.gh LIST OF PLATES 1 Interviewing fishermen at New Town on the border with Cote D ’Ivoire 32 2 Interviewing fishermen mending their nets at Senya Bereku 32 3 Collecting sea turtle eggshells at zone F (Old Ningo) 57 4 Excavating Leatherback nest at zone D (New Ningo) 57 5 Leatherback turtle returning to sea after nesting on the beach at Ada 75 6 Olive ridley turtle nesting at zone F (Old Ningo) 76 7 Measuring curved carapace length (CCLmin) o f an Olive ridley turtle at zone F (Old Ningo) 77 8 Eggs of Leatherback turtle during nesting at zone F (Old Ningo) 77 9 Olive ridley laying eggs at zone F (Old Ningo) 78 10 Leatherback turtle ‘camouflaging’ after nesting at Ada 79 11 One of several carapace of Leatherback turtles found on the beach at Anloga 80 12 Effect o f sea erosion at zone E (New Ningo) 97 13 Sand winning activity at zone F (Old Ningo) 97 14 Debris washed ashore at Prampram beach 98 15 Recreational activity at Prampram beach 98 ix University of Ghana http://ugspace.ug.edu.gh ABSTRACT Information on the occurrence and reproductive biology of sea turtles in Ghana is very scanty, therefore there is a need for detailed scientific studies on which their conservation and protection could be based. The present study has shown that the Olive ridley, Leatherback and Green turtles are the three main species that continue to nest on the beaches along the coast o f Ghana. The Olive ridley having the highest relative abundance o f 91% at the intensive study area followed by the Leatherback turtle with 6% and the Green turtle 3 %. The nesting season begins in August and extends to March, sometimes early April. It peaks in October for the Olive ridley and between December and January for the Green and Leatherback turtles. The Olive ridley turtles arrive first, followed by the green and Leatherback turtles. The nesting period o f the Green and Leatherback turtles overlaps that o f the Olive ridley. Sea turtles nest in a wide range o f beach sediment type along the coast o f Ghana. Beach sand with median grain size between 0.2-0.4 mm (medium to coarse sand) and sorting coefficient between 0.4-0.7 (well-sorted to moderately well sorted) are those preferred by nesting turtles along the coast o f Ghana. Based on the above grain characteristics, the potential sea turtle nesting sites in Ghana may include beaches along Princess Town to Busua, Senya Bereku to Accra, Prampram to Old Ningo and Anloga to Denu. Estimated average clutch size for Olive ridley, Green and Leatherback turtles were 86, 83 and 81 respectively. However, those counted for the Olive ridley and Leatherback were 100 (SD=6.64) and 85 (SD=10) respectively. Emergence period for the Olive ridley was 52 days, University of Ghana http://ugspace.ug.edu.gh Green turtle 56 days and Leatherback 63 days. Percentage hatching success was very high among all the three nesting species with Olive ridley recording the highest with 92.40%.The percentage hatching success for the other two, Green turtle and leatherback were respectively 89.38 and 88.32. Destruction o f sea turtle eggs by dogs at the intensive study area was very high, but this problem was not widespread along the entire coast o f Ghana. Percentage false crawls where found to be higher within zones closer to human habitation. xi University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE GENERAL INTRODUCTION AND OBJECTIVE OF STUDY 1.0 INTRODUCTION The loss of the world’s biodiversity is of a major concern to the world’s nations. This concern is embodied in the Convention on Biological diversity negotiated by the United Nations at Rio de Janeiro in 1992. The Convention aims to protect the world’s biological resources from further extinction, or at least slow down the rate of this decline. Nonetheless, the rate of biodiversity loss is increasing (Pearce and Moran, 1994) as a result of natural causes, climatic changes, over-exploitation and habitat destruction by man. It is for this reason that some organisms including sea turtles have been listed as either endangered or threatened species by the World Conservation Union (IUCN) (Groombridge, 1982). According to Marquez (1990), sea turtles were common and lived with the dinosaurs in the Cretaceous era (130 million years ago) and their fossil record even extends back to about 200 million years. They form an important component of marine ecosystems (Pritchard, 1997), and reach an average age o f 150 years; some reaching monstrous sizes and weigh about a ton (907 kg) (Groves and Hunt, 1980). Taxonomically, sea turtles belong to two families which are sub-divided into six genera. Presently only eight species of marine turtles exist world wide. These are: Green turtle Chelonia mydas (Linnaeus, 1758) Black turtle Chelonia agassizi (Bocourt, 1868) Flatback Natator depressus (Garman, 1880) ] University of Ghana http://ugspace.ug.edu.gh Loggerhead Caretta caretta (Linnaeus, 1758) Hawksbill Eretmochelys imbricata (Linnaeus, 1766) Olive ridley Lepidochelys olivacea (Eschscholtz, 1829) Kemp’s ridley Lepidochelys kempii (Garman, 1880) Leatherback Dermochelys coriacea (Vandelli, 1761) However, some authorities classify the Black and Green turtles as the same species. This is because the systematic status and nomenclature of the Black turtle remains uncertain. Recent genetic evidence is said to support an Atlantic-Mediterranean versus Indian-Pacific grouping, while morphological and behavioural data suggest an east Pacific species or subspecies (Eckert et al., 1999). Various arguments have been presented in favour and against the designation of the Black turtle as a full species within the genus Chelonia (Bowen and Karl, 1996; Pritchard and Mortimer, 1999). In this study, the full species concept is based on the traditional criteria o f the degree of morphological divergence and probable existence of reproductive isolation mechanisms (Pritchard and Mortimer, 1999). Sea turtles belong to the Kingdom Animalia, Subphylum Vertebrata and Class Reptilia. They are separated into two families namely: a) Dermochelyidae, that contains the leatherback Dermochelys coriacea as the only living species. b) Cheloniidae, with two subfamilies: i) Chelonini, which includes the green (Chelonia mydas), the flatback (Natator depressus) and the hawksbill {Eretmochelys imbricata) turtles, ii) Carettini, which includes the loggerhead (Caretta caretta), the Olive or Pacific ridley {Lepidochelys olivacea) and Kemp’s or Atlantic ridley (L. kempi) 2 University of Ghana http://ugspace.ug.edu.gh (Butterworth, 1989). A review of Ghanaian species and their status is given in Chapter 2 of this thesis. 1.1 LEGAL STATUS OF SEA TURTLES All sea turtles have been listed by IUCN as endangered or threatened except Natator depressus. However, the Convention on International Trade in Endangered Species of Flora and Fauna (CITES) lists all marine turtles on its Appendix I. That means, international sale of sea turtles and their products is illegal among the 142 signatory nations as of September, 1997 (Anon, 1997, Lagueux, 1998). CITES was adopted in Washington, on March 3, 1973 and signed by Ghana on February 12, 1976. Other treaties and conventions such as the Convention on Biological Diversity; Convention on the Conservation of Migratory species o f Wild Animals (Bonn Convention or CMS, adopted in Bonn on June 23, 1979); African Convention on the Conservation of Nature and Natural Resources (adopted at Algeria on Sept. 15, 1968 and signed by Ghana on October 9, 1969) also protect sea turtles. In Ghana, sea turtles are protected by the Wildlife Conservation Regulations, 1971, L. I 685 (Armah and Amlalo, 1998) which has its IUCN number as 841340 (H-971043000) (Navid, 1982). 1.2 ECONOM IC IM PORTANCE OF SEA TURTLES Sea turtles as a natural resource are used by local communities in various ways just as other natural resources are harvested to meet dietary, medical, cultural, economical and religious human needs and wants (Lagueux, 1998; Robison and Redford, 1991; Robison, 1994; Jorgenson, 1993; Freese, 1997). Sea turtles are exploited for their eggs, meat, shell, skin and other products. All species of sea turtles suffer from the collection of eggs and hunting for meat, which is a significant source of protein in 3 University of Ghana http://ugspace.ug.edu.gh many protein- deficient coastal regions o f the tropics. To many communities the meat is a delicacy. The eggs are eaten not only for its nutritional value, but also for its purported aphrodisiac qualities. Green turtles are liked for their meat and cartilage as well as their shells and bodies, which are stuffed and sold as curios for the tourist trade. The bones are used to make house ware items, decorative artwork, ornaments, jewellery, and fertiliser (Mark et a l, 1982 and Canin, 1989). They also have aesthetic value. For example, Pacific ridley skin is used to fashion boots, shoes, handbags and other accessories, and oil rendered from the carcass is used for everything from skin lotion additives to boat caulking agents. Hawksbills shells are highly desired as a decorative ornament, or the scutes from their shells used to craft jewellery, combs and eyeglass frames. Oil from sea turtle is believed to cure several ailments. Sea turtle oil is used as medicine or tonic by the Aboriginal and Torres Strait Islanders in coastal areas of northern Australia. The oil is used for protection against evil spirits in Kenya (Wamukoya et al., 1996) and as medicine for sore muscles and hernia in Tanzania (Howell and Mbindo, 1996). Among the Mauritian Creole community, blood from turtle is a traditional cure for asthma in children (Mangar and Chapman, 1996). The bone of "turtle tail" is grated into tea and believed to be a powerful aphrodisiac and the meat is said to make one strong so it is eaten by members o f the local football teams in Mauritius (Mangar and Chapman, 1996). However, in Ghana, when mixed with honey, the oil from sea turtle is used to treat convulsion in children among some local communities in the Volta region. In some coastal communities, sea turtles have been object o f ancient ritual practices, whiles in some areas they are venerated as 4 University of Ghana http://ugspace.ug.edu.gh sacred animals (Marquez, 1990). In Ghana, the people of Old Ningo regard sea turtles as a totem and therefore revere them (Armah et al., 1997). Islamic precept considers sea turtle meat unclean (Fretey and Fourmy, 1996) hence not eaten by some Moslems. In Zanzibar, some fishermen believe that Islam prohibits the consumption of turtle meat since animals that live on both land and water are considered impure (Khatib et al., 1996). Muslims in Indonesia and Thailand also avoid turtle meat, but eat turtle eggs in large quantities (Hill, 1991). Sea turtle meat is used as bait for lagoon fishing. In Nicaragua, Loggerhead turtle meat is not consumed but harvested and used as shark and lobster trap bait, due to its strong flavour (Lagueux, 1998). Shells of the Hawksbill turtles are used by artisans in the Seychelles for handicrafts (combs, jewellery cases, trinkets etc.) which are sold to tourists. In some communities in the Seychelles, turtle shells are used as water containers, and when polished, the shells are used as trophies. Exploitation of sea turtles for international trade had existed for centuries. Marine turtle fishery industry existed in Mexico until it was bought from private owners in June 1980 (Romero, 1980). The leather industries in Italy, France, Germany and the United Kingdom all formerly produced a variety of leather products made from sea turtles. Hawksbill shell was widely used in Europe for the production of expensive eyeglass frames and other items. In Nicaragua, Hawksbills are still harvested by some communities for their scutes (Lagueux, 1998). In the 1980s, Japan was ranked as the world's foremost importer of sea turtle products, with 86% emanating from Latin America / Caribbean and Asia /Pacific regions. Only 12% of the overall total was imported from the Africa / India ocean nations. 5 University of Ghana http://ugspace.ug.edu.gh Sea turtle shells also serve as substrate for some sessile barnacles. Specimens o f barnacle Chelonibia testudinaria were collected from the back o f a green turtle (Mustaquim and Javed, 1993). Sea anemones Adamsia, Metridium and Balanus barnacles have been found on olive ridleys. Hawksbill turtle is the only known spongivore marine reptile (Meylan, 1988). Even, strictly spongivore vertebrates include only a small number of teleostean fishes (Marquez, 1990) hence hawksbill turtles might therefore play an important role in checking the populations of sponges. 1.3 THREATS TO SEA TURTLES 1.3.1 NATURAL ENEMIES: Sea turtles face a lot of threats within the environment in which they find themselves. For example, turtle eggs are susceptible to terrestrial predation by birds, ghost crabs (Ocypode sp.), racoons, armadillos and ants. Black vultures (Coragyps atratus) and turkey vultures (Cathartes aura) prey on both eggs and hatchlings, whiles the Yellow crowned night herons (Nyctanassa violacea) eat sea turtle hatchlings (Troeng, 1997). Coatis (Nasua narica) have been observed feeding on green turtle eggs. Wild pigs and dogs also destroy the eggs. This is so high in some areas such as Gahirmatha beach (India) where only four or five hatchlings out of some 1000 eggs survive. Hatchlings also suffer from predation by bony fish and sharks (George, 1997), dolphins and octopus. Parasitic infections have been reported in sea turtles. Two species of leeches are known to parasitize sea turtles. These are Ozobranchus branchiatus found on only green turtles that inhabit tropical waters, and O. margoi found on most species of the 6 University of Ghana http://ugspace.ug.edu.gh sea turtles (George, 1997). Some barnacles are said to damage the underlying shells and skin of turtles allowing bacterial or fungal pathogens to enter. 1.3.2 HUMAN DIMENSIONS Human activities affecting sea turtles range from destruction of their nesting beaches through their foraging habitats to intensive harvesting at high seas, basically, touching every stage of their life cycle. Beach armouring, beach nourishment, beach cleaning and beach driving as well as sand mining greatly cause sea turtle habitat alteration and loss. Beach armouring is the construction of hardened structures meant to protect dune property from erosion. It involves erecting vertical or inclined concrete sea defence walls, wooden walls, rock revetments and sandbag/sand tube structures (Lutcavage et al., 1997). Groynes and jetties meant to control long shore sand movement also present barriers to nesting turtles. Beach armouring is greatest where coastal development is most widespread such as the United States of America and the Mediterranean. Beach nourishment which involves mechanical dumping or pumping of sand onto eroded beaches, is very expensive, however, it is undertaken in developed countries, where the economic value of the beach makes it worthwhile. This is found especially in some parts of Europe and the United States. Some nourished beaches have high clay, silt and shell content making them too compact for nest excavation by turtles. Wave action usually causes the formation of steep escarpments that prevent sea turtles from reaching the upper beach. Properties such as sorting, moisture content, reflection and conduction of 7 University of Ghana http://ugspace.ug.edu.gh artificially nourished substrate differ from that o f the natural beach. These properties go a long way to affect the architecture of the egg chamber, incubation temperature, gas exchange and water uptake for development of clutch thereby resulting in diminished egg and hatching survivorship. The process of nourishment which involves lighting, sand spreading, and pumping activities could disturb nesting females and produce disorientation and mortality in hatchlings. Sand mining operations also have detrimental effect on sea turtle nesting. This involves the removal of large quantities of sand from beaches for production of concrete and other construction activities. This activity destroys the beaches and eliminates nesting. According to Lutcavage et a l, (1997), this has been recognised as a threat to sea turtle populations, primarily in the Mediterranean. A similar situation is evident at Old Ningo, in Ghana. Beach cleaning and beach driving also pose threat to sea turtles. Heavy vehicles can crush developing eggs and pre-emergent hatchlings. The tyre ruts can trap hatchlings, allowing them to become exhausted or taken by predators. Vehicles operated at night can also disturb nesting females and crush emerging hatchlings crawling toward the sea. Beach cleaning may uncover and destroy nests. Raking may leave ruts or ridges that may disrupt hatchlings’ sea finding behaviour. Artificial lighting on nesting beaches is detrimental to sea turtles. It disrupts critical behaviours such as nest-site choice and the nocturnal behaviour of both hatchling and nesting females. Both associative and direct experimental evidence showed that artificial lighting on the beaches deters sea turtles from nesting. Hatchlings move 8 University of Ghana http://ugspace.ug.edu.gh toward artificial light source rather than the sea, and succumb to exhaustion, dehydration, and predation. Boat collisions have also been identified as causes o f death o f sea turtles. Sea turtle stranding data from the U.S.A, Gulf of Mexico and Atlantic coasts, Puerto Rico and U.S Virgin Islands have shown that between 1986 and 1993, about 9% of living and dead stranded sea turtles had propeller or other boat strike injuries (Lutcavage et al 1997). Also, between 1991 and 1993, in Florida (USA) where coastal boating is popular, frequency of boat injuries was 18% of 2,156 strandings. Sea turtles are known to associate with offshore oil and gas platforms and can be harmed by underwater explosives detonated to remove platforms no longer in use. These explosions can cause capillary damage, disorientation, and loss o f motor control in sea turtles. Those near detonation sites may sustain fatal injuries. About 50 to 500 sea turtles are estimated to be killed each year from explosive platform removal in the Gulf of Mexico (Lutcavage et a l, 1997). Although, specific information is not available, it is believed that petroleum seismographic canons and military manoeuvres involving explosives also have the potential to harm sea turtles. Direct evidence of sea turtles being seriously harmed by oil spills has been showed by instances such as the oiled loggerhead nests and jeopardised hatchlings and adult female in Tampa Bay, Florida. The Mexican Ixtoc blow-out oil reached Rancho Nuevo which is the only known major nesting beach o f the Kemp's ridley, and oiled sea turtles were subsequently seen in Texas coastal waters (Fritts and McGehee, 9 University of Ghana http://ugspace.ug.edu.gh 1981). In 1983, 180 hawksbill turtles were known to have been killed off the islands of Jana and Karan, near Saudi Arabia (Lutcavage et al., 1997). The list of materials found in turtles’ digestive tracts is extraordinary. Plastic bags, beads, pellets, line, rope, strapping, pieces from bottles and hard pieces o f unknown origin are common items ingested by sea turtles. They also ingest latex balloons, aluminium, paper, cardboard, styrofoam, rubber, string, cigarette filters, wax, cellophane, fishhooks, charcoal and glass (Plotkin and Amos, 1990). Plastics are also the most common type of debris sea turtles become entangled in and ingest. Fishing gear, particularly, monofilament line may account for about 68% of all entanglement cases (National Research Council, 1990; O'Hara and Iudicello, 1987). Large numbers of turtles are being lost to shrimp fishing operations (Nunoo and Evans, 1997). Incidental capture of sea turtles in shrimps trawls is reported to account for more deaths than all other sources of human activities combined (National Research Council, 1990; Lutcavage et al., 1997). Sea turtles are also killed when incidentally captured in purse seine, gill nets, and various types of untended fishing gear. Lobster and crab pots as well as hook and line fishing also cause entanglement, mutilation and debilitation. Sea turtle forcibly submerged in any type o f restrictive fishing gear would eventually suffer fatal consequences from prolonged anoxia or seawater infiltration of the lung. Sea turtles are also vulnerable to capture in pelagic long line, paired trawl, and gill net fisheries. 10 University of Ghana http://ugspace.ug.edu.gh 1.4 OBJECTIVE OF STUDY This has been a source of increasing concern for both fisheries managers and the industry. Although, in Ghana, the law makes it an offence punishable by fine, imprisonment or both, to capture or kill sea turtles, several cases of poaching have been reported among many coastal fishing communities. As Ghana is currently trying to boost its tourism potential, there has been an increase in coastal development such as beach resorts for recreational activities in some cases at the expense o f sea turtle nesting habitats. Information on turtle occurrence and especially reproductive biology are very scanty. There is therefore no opportune time than now to conduct detailed studies on the occurrence and reproductive biology of sea turtles in Ghana to provide scientific data on which their conservation and protection could be based. The primary objective of this study is to: i) determine the present distribution of sea turtles on the Ghanaian coastline, ii) identify suitable nesting beaches along the entire coast, iii) determine the emergence period, hatching success and duration o f the nesting season for each species, iv) estimate mean clutch size of each nesting species, v) collect biometric data for each species, and vi) estimate the extent of destruction of nests by predators (dogs). 11 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO LITERATURE REVIEW 2.0 OCCURRENCE OF SEA TURTLES 2.1 GLOBAL DISTRIBUTION Most of the eight species of sea turtles surviving presently are unevenly distributed in all the three tropical oceans, except for three species. These three species have relatively restricted distributions (Pritchard, 1997). They are the flatback (Natator depressus) in Northern Australia, Kemp's ridley (Lepidochelys kempii) in the Gulf of Mexico and North Atlantic, and black turtle (Chelonia agassizii) in the Eastern Pacific. Loggerhead (Caretta carreta) is found in temperate and subtropical waters world­ wide (Me Diarmid, 1978). The loggerhead has been discribed by Pritchard (1997) as having an "antitropical" distribution; and this has fragmented its overall range into well-separated enclaves in the northern and southwestern Indian Ocean, eastern Australia, Japan, south eastern United States, the Mediterranean, and Southern Brazil. This fragmentation is said to bring the loggerhead into contact with industrial and development stresses ranging from massive incidental capture by shrimpers to recreational development of nesting beaches. In the United States nesting of loggerheads occurs on suitable beaches from North Carolina to southern Florida (McDiarmid, 1978). This species also nests in the western Caribbean, Cuba, Jamaica, the Dominican Republic and Puerto Rico. Nesting 12 University of Ghana http://ugspace.ug.edu.gh sites have been reported in Mexico (Hendrickson, 1982; Sternberg, 1981). Scattered nesting occur along the eastern coasts and offshore cays o f Belize, Guatemala, Honduras, Nicaragua and Colombia, and the western coast and Islands o f Venezuela (Groombridge, 1982). According to Gudynas (1980) loggerheads are common coastal species in the Atlantic areas o f Rocha in Uruguay, but records o f its nesting is not known. Loggerheads occur in the waters of India and Sri Lanka (Frazier, 1982), along the Chinese coast, where they appear to nest on the Xisha Islands in the South China seas (Chu-chien, 1982). Nesting has also been reported on all the southern Islands o f Japan (Limpus, 1982) as well as the Great Barrier Reef area. C. carretta is known to be fairly common around the Mediterranean (Di Palma, 1978) where large numbers nest along the coast of Turkey to Israel. Smaller numbers also nest on the Northwest coast of Cyprus, and Zakynthos in Greece (Groombridge, 1982). The largest known nesting population of C. caretta is said to occur on the Masirah Island in Oman, where 30,000 females nest annually. Green turtle (Chelonia mydas) is also a circumglobal species, with most o f its nesting and feeding grounds lying within the tropics (Pritchard, 1997). It occurs along the West Coast of North America (Anon, 1980; Groombridge, 1982), Mexico (Sternberg, 1981), El Salvador, Honduras, Nicaragua (Cornelius, 1982), and Costa Rica (Sternberg, 1981) where the nesting colony is large and the most important in the Caribbean (Pritchard, 1997). Other major nesting colonies occur on the Galapagos, the Barrier Reef Islands (Australia), d'Entrecasteaux Reef (New Caledonia) and Oceanic Islands such as the Ascension Island (Pritchard, 1997). They also occur in the 13 University of Ghana http://ugspace.ug.edu.gh Uruguayan waters (Gudynas, 1980) and Aves Island, Venezuela (Groombridge, 1982). In Australia C. mydas rookies occur around the area of the Great Barrier Reef o f which the coral sea, southern Papua New Guinea, New Caledonia, the Torres strait and the Northern Territory serve as feeding grounds. The French Frigate shoals in the Hawaiian archipelago of the central Pacific are important breeding sites for this species. Nesting green turtles have been observed on some beaches in Pakistan, India, Thailand and Malaysia. They occur along the coasts of Taiwan, China and Japan. C. mydas is common in Yemen, Oman (Ross and Barwani, 1982), Bahrain, G ulf coast of Saudi Arabia and Iran. They have also been reported from the Black, Marmara and Aegean seas (Geldiay et al., 1982). Hawksbill turtle (Eretmochelys imbricata) is circumtropical and nests on tropical or sub-tropical beaches in the Atlantic, Indian and Pacific oceans. This species is said to be uncommon in the eastern pacific region and nest sporadically along the Central American mainland from central Mexico to Ecuador. It is known to be widespread through the Indian Ocean, with most nesting on islands such as the Suakin archipelago in the Red sea, islands off the Arabian peninsula and the Gulf, and oceanic island including the Seychelles, Lakshadweeps, Maldives right to Southern India. Occasional nesting has been reported from Thailand, eastern part o f Malaysia, Indonesia, northern coast of Australia and also sparse breeding in Hawaii (Pritchard, 14 University of Ghana http://ugspace.ug.edu.gh 1979, 1982; Sternberg, 1981) and Costa Rica. E. imbricata occurs in the Mediterranean regions but no nesting record is available. Kemp’s ridley or Atlantic ridley (Lepidochelys kempii) has been classified by IUCN as critically endangered. According to Pritchard, (1997), L. Kempii remains the rarest sea turtle in the world and apart from few isolated cases, Racho Nuevo, in southern part of Tamaulipas, Mexico is the only known nesting site for this species. Major known feeding ground for the adult kemp's ridley include the crab-rich shallow waters off Louisiana and the Tabasco-Campeche area of Mexico (Groombridge, 1982). Olive ridley (Lepidochelys olivacea) is a circumglobal species, present in tropical regions of the Atlantic, Indian and Pacific Oceans. Record of nesting of this species is known along the Pacific coast of Central America, particularly, from North Mexico, through Guatemala, El Salvador, Honduras and Nicaragua to Panama. Synchronized aggregations ("arribadas") are said to o c c u t in Oaxaca, Mexico and Nancite site, Costa Rica with over 200,000 nests per year. Shanker and Mohanty (1999) reports that between March 24-April 1,1999, 200,000 turtles nested on two islands which are the current nesting sites at Gahirmatha (India). L. olivacea does not nest in the Caribbean, the only known nesting in the West Atlantic occurs in Surinam, with small numbers in Guyana. Minor to moderate nesting is known to occur on Masirah Island (Oman), Lakshadweep Islands, the Andaman Islands, Sri Lanka, Malaysia, Burma, Papua New Guinea and northern i f University of Ghana http://ugspace.ug.edu.gh Australia (Sternberg, 1981). Olive ridley is probably the most abundant sea turtle around the world. Though it is known principally to inhabit the northern hemisphere, with 20 °C isotherms as its distributional boundaries, non-breeding Olive ridleys outside the 20 °C isotherms have been observed as far as the G ulf of Alaska during warm weather such as the El Nifio phenomena in the eastern Pacific (Marquez, 1990). Leatherback (Dermochelys coriacea) is also a circumglobal species and nests on beaches of tropical seas in the Atlantic, Indian and Pacific oceans and occasionally in subtropical regions and the Mediterranean. It forages widely and regularly in temperate waters. Dense Leatherback nesting occurred along the Pacific Coast o f Mexico, specifically in Michoacan, Guerrero and Oaxaca. Also Playa Naranjo in Costa Rica is said to be an important nesting beach. However, according to Pritchard (1997), the Terengganu colony (Malaysia) has collapsed and a serious decline has been documented in both Pacific Mexico and Costa Rica; resulting from beach slaughter, egg collection and serious incidental captures by fishing gear in the open sea. The Atlantic colonies include those found in Trinidad, Surinam and French Guiana (Pritchard, 1997). Nesting on the Atlantic coast of the Americas has been recorded from the east coast of Florida in the north, through the Caribbean region to Espirito Santo Brazil in the south (Sternberg, 1981). This species is occasionally found in the Mediterranean. Small- scale nesting occur on the island o f Tobago, but much larger aggregations are known to nest on north and east coasts of Trinidad (Carr et al., 1982; Sternberg, 1981). 16 University of Ghana http://ugspace.ug.edu.gh Large aggregations of leatherbacks exist in the Union Territories of the Andaman and Nicobar Islands whereas small-scale nesting occur in Lakshadweep Islands, parts of India and in Sri Lanka (Bhaskar, 1979, Frazier, 1982). Nesting is also reported in many parts of Southeast Asia, Australia and Papua New Guinea. These include Burma, Thailand, Malaysia, Indonesia, China seas and as far north as the Yellow Sea (Chu-chien, 1982). 2.2 OCCURRENCE IN AFRICA The History of turtles in Africa dates back to the early Jurassic period (60 million years). Along the African continent, large scale nesting o f loggerhead turtle have been reported in Southern Africa, especially on Paradise Island, Mozambique and on the Tongaland coast of South Africa (Frazier, 1982; Hughes, 1982). The southern part o f Madagascar around Fort Dauphin is said to be an important nesting site for the Loggerheads (Pritchard, 1979; Sternberg, 1981; Frazier, 1982; Hughes, 1982). Nesting of C. caretta occurs also in the Eastern Atlantic. These include Namibia (Hughes, 1982) and Morocco (Brongersma, 1982; Sternberg, 1981). C. mydas feeds and nests throughout the Seychelles; nests on the Comoros, Tanzania and is common on the eastern coast of Somalia, Red Sea coast o f Egypt (Frazier, 1982) and the Dhalak Archipelago around Ethiopia. Reports o f occurrence in Namibia and South Africa but not nesting (Hughes, 1982) is available. E. imbricata is known to occur south to Angola, Mozambique, Tanzania and Kenya. L. olivacea nests in good numbers in Northern Mozambique, smaller numbers in 17 University of Ghana http://ugspace.ug.edu.gh Tanzania and iarely in South Africa. Minor nesting occur in Angola and northern part of Namibia. Leatherback occurs around the Republic of Congo (Zaire) and Angola (Hughes, 1982, Brongersma, 1982). In the Indian Ocean nesting occurs in southern Africa, mainly along the Tongaland coast of northern Natal, South Africa (Hughes, 1982) and Mozambique (Frazier, 1982). In West Africa loggerheads have been reported in Senegal, Guinea Bissau, Sierra Leone, Ghana and the Congo. Nesting records of C. mydas is available from Mauritania, Senegal, Sierra Leone (Turtle Island); Fernando Po, and Angola. However, historical records of its occurrence in Liberia, Principe and Sao Thome, the Congo and Democratic Republic o f Congo (Zaire) is available (Brongersma, 1982). E. imbricata is believed to nest on Cape Verde Island, Senegal, Mauritania and probably other areas. L. olivacea nests on beaches from Senegal to Angola. MinoT and solitary nesting of leatherback occurs in Senegal, Liberia, Ivory Coast Togo and Ghana (Marquez, 1990). 2.3 OCCURRENCE IN GHANA Documentation o f marine turtles in Ghana dates back to Irvine (1947). Five species of marine turtles are known to occur in Ghana's territorial waters. These include the Loggerhead, Green, Hawksbill, Olive ridley and the Leatherback turtles (Irvine, 1947; Toth and Toth, 1974; Brongersma, 1982; Armah et a l, 1997). The Olive ridley turtle 18 University of Ghana http://ugspace.ug.edu.gh is known to show the highest relative abundance in Ghana (Carr and Campbell, 1995). Minor nesting of the Leatherback turtle in Ghana is recorded by Marquez (1990). Seventy percent (70%) of Ghana's 535 km coastline presents suitable sites for turtle nesting. The shoreline from Prampram to the Volta estuary, a distance o f about 65 km, however serves as the main turtle sites (Carr and Campbell, 1995). The local Ghanaian names are given as follows: Leatherback (Ga: Gbosange; Fante : Puhuru; Ewe: Agbosege ), Hawksbill ( Ga : Hala akoo ; Fante : Puhuru ; Ewe : K]o ), The same names are used for Loggerheads, Olive ridley and Green turtles in the various dialects ( Ga : Hala ; Fante : Puhuru ; Ewe : Klo ) ( Armah et al., 1997). 2.4 REPRODUCTION IN SEA TURTLE According to Miller (1997), reproduction in marine turtle occurs within three general constraints. These are: 1) Nesting must occur during conditions that are "conducive to adult activity". 2) Nesting must occur during conditions that facilitate embryonic development and survival. 3) Hatchlings must emerge into conditions that "are conducive to their survival". Within these constraints, sea turtles share a number o f general reproductive characteristics. For instance, all species exhibit: a) Iteroparous reproduction (Hirfh, 1980; NRC, 1990) with possible exception of the Kemp's ridley (NRC, 1990), b) Stereotyped nesting behaviour (Hendrickson, 1982), c) Laying of relatively large numbers of eggs several times during the reproductive period (Hirth, 1980, Van Buskirk and Crowder, 1994; Miller, 1997). 19 University of Ghana http://ugspace.ug.edu.gh d) Relatively strong attachment to a particular location for nesting (Bjomdal et al., 1985; Limpus et a l, 1992) but inter- and intra- specific variations exists. 2.4.1 AGE AT FIRST MATURITY There have been lots of speculations about the age at first maturity. Whilst some authors estimate values as low as 6 years, others state between 8 and 13. According to Marquez (1990), studies using average sizes instead of the smallest sizes of turtles estimated age at first maturity for Green turtles between 25 and 30 or more years. Limpus (1990), using a laparoscope (a fiberoptic instrument used to examine visually the interior of the peritoneal cavity) followed the maturation (changes in oviduct and ovaries, testis and epididymis) of C. caretta in foraging areas o f eastern Australia. His results showed that Hawksbill, Green and Loggerhead turtles typically begin breeding at just less than a few centimetres longer than the minimum breeding size and still immature or just starting into puberty. Presumably, other species show similar patterns of maturation. This shows that size is not a reliable indicator of maturity or breeding. Sea turtles must have access to beaches with deep, loose sand that are above high tide for nesting. They nest predominantly during the warmer months. Some beaches are known to host year-round nesting. The incubation o f sea turtle eggs is dependent on the temperature of the sand, hence, the duration at which the embryos develop and the hatching period show wide variations. Cooler sand has been found to produce more males, with warmer sand producing a higher ratio of females (Caribbean Conservation Corporation, 1996). 20 University of Ghana http://ugspace.ug.edu.gh 2.4.2 COURTSHIP BEHAVIOUR Male and female exhibit courtship behaviours prior to mating. In the male, precopulatory behaviours include head bobbing, position in the water column, head to head bumps, nuzzling, biting, movement of flippers. During courtship, a female receives bites (nips) to her flippers, neck and head that leave open sores for several weeks to heal. The male mounts the female usually at the surface with a lot o f splashing and hooks onto her carapace using the enlarged claws on his front flippers and the large claws on his hind flippers to hold himself in place. The male curls his long tail to bring their cloaca into contact. His penis is erected into her cloaca. The shape of the penis with a bifurcation of the sperm duct at the tip allows the transfer of semen into each oviduct without passing through the environment of the female's cloaca. According to Wood and Wood (1980) an increase in fertility correlated with the duration of the coupling period. Genetic studies have shown that males mate with several females and females mate with several males (Miller, 1997). 2.4.3 NESTING PERIODICITY Generally, female sea turtles do not reproduce every year, with the exception o f L. kempi. Males of some species (e.g. C. caretta) however, may mate every year or every two years. While in foraging areas, sea turtles accumulate the energy reserves required to support vitellogenesis over a variable period of several years depending, in part, on the quality and quantity of food available. 21 University of Ghana http://ugspace.ug.edu.gh Several clutches of eggs are laid at approximately two weeks intervals. This greatly reduces the likelihood o f all eggs being lost during the breeding period. The ecological consequences of this behaviour require elucidation, but it is clear that the turtles are spreading their reproductive effort through time which reduces impact o f an unpredictable environment on hatchling production (Eckert, 1987). When a set of exogenous (e.g. photoperiod) and endogenous (e.g. hormone levels and / or fat levels) factors interact, the consequence is reproduction. The timing of reproduction in marine turtles follows behind period o f ample food availability during which the turtles accumulate fat reserves (one to several years), complete vitellogenesis (1 0 -1 2 months) and migrate to the breeding and nesting area. 2.4.4 NEST SITE FIDELITY All eight species nest on beaches around the world. The species migrate from foraging grounds to mating grounds a short distance away. After mating, the females then migrate to nesting areas while males return to foraging grounds. After a reproductive period, the females return to the foraging grounds and prepare for the next reproductive period, a few to several years in the future. Tag-retum data have established that some turtles migrate over distances greater than 2600 km, however, most travel less than 1000 km. All species migrate to varying degrees except the N. depressus that does not migrate beyond the continental shelf of Australia. Carr (1975) made the distinction between regional discrimination (philopatry) and fine-scale homing to beach (site fidelity) within the region for sea turtles. It is now well established that marine turtles migrate between their foraging 22 University of Ghana http://ugspace.ug.edu.gh areas and their nesting areas with a high degree of accuracy (Limpus et al., 1992; Limpus and Miller, 1993), Although they may not necessary return to the beach o f their birth, genetic studies have demonstrated that breeding sea turtles return to the region o f their birth (Miller, 1997; Bowen et a l, 1992). Once a turtle returns to the region of its birth and selects a nesting beach, it will tend to renest in relatively close proximity (0 to 5 km) during subsequent nesting attempts within that nesting season. Limpus et al. (1984) reported that N. depressus returned to the nesting beach with a high degree o f accuracy regardless of whether they were returning after an "unsuccessful" nesting (or) in later breeding season. Green turtles are known to show a high degree o f nest site fidelity. Bose and Le Gall (1986) reported that most renesting attempts were within 200 m of the previous attempt, with a range o f about 600 m over a three-month nesting season. 23 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE MATERIALS AND METHODS 3.0 STUDY AREA Ghana is located in West Africa on the Gulf of Guinea. The coastline is generally low lying; not more than 200m above sea level (Armah and Amlalo, 1998) and is characterised by sandy beaches, rocky beaches and rocky headlands. The continental shelf is narrow. It extends between 20 km and 35 km except off Takoradi where it reaches 90 km. The coastal waters are characterised by two seasonal upwellings with varying annual intensities. The bottom is varied consisting of sandy, muddy and rocky portions. The entire 550 km coast of Ghana was investigated, visiting a total of 23 beaches with average interval of about 21 km. The beaches along Prampram through New Ningo to Old Ningo in the Greater Accra region, east o f Accra were selected as an intensive study area. This was based on preliminary investigations involving visits to beaches and interviews. This area has been described as the prime nesting site o f marine turtles. This confirmed earlier assertions by Toth and Toth (1974) and Carr and Campbell (1995). The shoreline o f the intensive study area is characterised by gently sloping fine to coarse sandy beaches. Partially submerged rocks are observed intermittently in some parts of the sublittoral zone and the intertidal. The vegetation along the beach is of the strand type and includes Sporobolus sp., Ipomea pes-caprae, Sesuvium portulacastrum, Cyperus maritimus, Canavalia rosea 24 University of Ghana http://ugspace.ug.edu.gh among many others. Cocos nucifera, however, dominates the entire coastline in size and height. 3.1 DATA COLLECTION Three major methods were employed in data collection viz questionnaires, sediment granulometry and field observations of nesting turtles. 3.1.1 USE OF QUESTIONNAIRES Questionnaires were administered along the entire coast of Ghana; from New Town on the extreme west near Half Assini to Denu on the east (Figure 1). This was to find out the perceptions of the fishers of sea turtles especially their knowledge on nesting locations. Fishermen were interviewed in groups between 5-15 men at a time on the canoe landing beaches. Within the intensive study area comprising Prampram, New Ningo and Old Ningo 50 questionnaires were administered. The questionnaires were analysed using the computer program 'Statistical Package for Social Sciences' (SPSS). 3.1.2 SEDIMENT GRANULOMETRY In order to establish whether sediment type could influence choice of nesting locations, sand samples were collected at about 40 cm deep from all the beaches visited and analysed. Sub-samples were weighed from each sand sample, washed in distilled water for 10 minutes using a 63 pm mesh sieve to remove the silt-clay fraction as outlined by Gray (1981). The washed sub-samples were dried at 100°C overnight (about 16 hours). 25 University of Ghana http://ugspace.ug.edu.gh Figure 1 Map of Ghana showing various sampled stations University of Ghana http://ugspace.ug.edu.gh 100g of each sub-sample was weighed on an electronic balance and shaken by a mechanical shaker with series of screens. The mesh sizes o f the screens were 1mm, 710jtm, 500nm, 355nm, 250nm, 125jim, and 63jim arranged in a decreasing geometric scale. The amount of sand retained on each sieve was weighed and the median grain size of each sample determined from a cumulative plot. 3.1.3 FIELD OBSERVATIONS AT INTENSIVE STUDY AREA The approximately 15 km beach from Prampram to Old Ningo was divided into six zones (Zone A- F, Figure 2). These beaches were patrolled daily from August 1998 to April 2000. Patrols were conducted on foot, just after sunrise for best viewing o f crawls. This is because track signs begin to deteriorate as the sun dries out the sand, and the crisp shadows that facilitate track identification are lost. Complete counts o f nesting and non-nesting emergences (false crawls) were done. Track of each individual turtle was identified by its shape (design), measured and nest marked with pegs as well as taking co-ordinates using a Global Position System (GPS). Dates of nesting and hatching were recorded to estimate the emergence period. All nests were left to develop in situ, and eggshells from hatched eggs were gathered by excavating each nest four days from the hatched date. The number o f eggs that were not hatched in each nest were counted and recorded. To estimate the clutch size, eggshells were gathered from each nest, sun-dried and weighed on an electronic balance to estimate the clutch size o f each species. The total weight of the eggshells from each nest was divided by the average weight of a small sub-sample of hatched eggs. This enabled estimation of hatching success by 27 University of Ghana http://ugspace.ug.edu.gh 8s Nc M3O Ml S J1MZ ce 1es £ 2 V 0r40 c-- © D«. fS CuJ O3fc University of Ghana http://ugspace.ug.edu.gh calculating the percentage ratio of hatched eggs to the clutch size. The number of freshly laid clutches destroyed by dogs in each zone were recorded to estimate the total percentage of sea turtle eggs destroyed by dogs during the study period. During the 1999/2000 nesting season, carapace lengths for nesting females were measured from the precentral scute in the carapace midline to the posterior margin of the postcentrals. Carapace widths were also measured over the curve across the widest part of the carapace perpendicular to the longitudinal body axis. Counting of eggs as they were laid was also done and filmed with a night vision video camera on two occasions. 29 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR 4.0 DISTRIBUTION OF SEA TURTLES IN GHANA 4.1 INTERVIEW BY QUESTIONNAIRE 4.1.1 INTRODUCTION In areas where sea turtle conservation programmes are desirable but lacking, conservationists and sea turtle researchers are faced with identifying potential study sites in the absence o f field data. Some of the techniques for sea turtle habitat surveys outlined by Diez and Ottenwalder (1999) include interviewing potentially knowledgeable residents and preliminary surveys to confirm the presence o f crawls, nesting pits, or egg shells on the beaches. After Irvine (1947), other authors have also documented the occurrence of sea turtles in Ghana’s coastal waters (Toth and Toth, 1974; Brongersma 1982, Groombridge, 1982; Carr and Campbell, 1995 and Armah et al., 1997). Five species are known to exist in Ghana’s coastal waters with the Olive ridley being the most common. However, no detailed studies on the distribution of sea turtles had been conducted in Ghana in the recent past. 4.1.2 MATERIALS AND METHOD Questionnaires were administered to fishermen at various fishing communities along the entire coast of Ghana, about 21 km interval. A total of 97 fishermen in groups made up between 5 and 15 people were interviewed at a time. The fishermen were interviewed in local dialects: Fante, Ga, Dangme and Ewe. Series of questions 30 University of Ghana http://ugspace.ug.edu.gh addressing observations on sea turtles, such as the occurrence and seasonality of nesting were posed to obtain basic information. Care was taken not to bias the responses of interviewees. At the intensive study area, a total of 50 questionnaires were administered to the fishermen at the canoe landing beaches. This was to compare their observations with the data collected from the field. 4.1.3 RESULTS 4.1.3.1 Entire coast Interview conducted among local fishing communities along the entire coast of Ghana showed that Olive ridley is still the most common sea turtle seen (Plates 1 and 2). Respectively, 31.8%, 30.3%, and 26.4% of the respondents are very familiar with the Olive ridley, Leatherback and Green turtle respectively. About 10.1% of fishermen interviewed said they have ever seen the Hawksbill at sea while 1.4% believe they had seen the Loggerhead. The interview revealed that the nesting season starts in late August, peaks in December and ends in March. However, visits to some of the nesting beaches showed that the turtles stop nesting before mid-January. Of the fishermen interviewed, 54.6% were of the view that the number o f turtles they see these days is decreasing. However, 45.4% maintained that the numbers are almost the same in that they see turtles regularly as in the recent past during the nesting season. 31 University of Ghana http://ugspace.ug.edu.gh Plate 1 Interviewing fishermen at New Town on the boarder with Cote D ’ivore. Plate 2 Interviewing fishermen mending their nets at Senya Bereku 32 University of Ghana http://ugspace.ug.edu.gh Along the coast, 59.9% of fishermen interviewed have tasted and still eat sea turtle meat, of which 36.2% claimed they enjoy the meat very much, 3.4% do not enjoy it but eat when given, while 13.1% were indifferent. The turtle meat is rarely sold on the local markets. Fishmongers who buy the turtles, smoke the meat for sale at markets hinterland. The percentage of respondents who have tasted and still eat sea turtle eggs was 69.5. As high as 79.6% of the fishermen interviewed confirmed that sea turtles are still landed by some fishermen along the coast. Only 1% of those interviewed said they were aware that sea turtles play important role in the marine ecosystem. Out of the interviewees, 88.9% suggested that dogs must be killed at the beaches in order to protect sea turtle eggs. Destruction o f sea turtle eggs by dogs and pigs is not a major problem along most of the nesting beaches, only 15.2 % o f the fishermen interviewed along the entire coast said they are aware that dogs and pigs destroy sea turtle eggs. Concerning prohibition by law, 31.4% of those interviewed said they were aware that it was illegal to capture sea turtles or collect their eggs. A higher proportion o f 45.7% of the total respondents expressed willingness to join any sea turtle management committee, 53.3% were indifferent but would participate if given incentives, while only 1.0% said they would not join. The questionnaire evaluation revealed that other potential nesting sites in Ghana are PrincessTown, Elmina, Gomoa-Fete, besides Prampram, Old Ningo, Akplabanya, 33 University of Ghana http://ugspace.ug.edu.gh Ada, Anloga and Keta. Also, sporadic nesting occurs on most of the sandy beaches along the Ghana shoreline (Figure 3). 4.1.3.2 Intensive Study Area At the intensive study area, 41.2% of those interviewed were very familiar with the Olive ridley, 35.3% with Leatherback and 17.6% with the Green turtle. From the questionnaire evaluation, the nesting season was placed between September and February. At Old Ningo, New Ningo and Prampram, 80% the fishermen believe the number of nesting turtles have increased in recent times. At Old Ningo and New Ningo, none of the fishermen admitted ever tasting sea turtle meat. However, at Prampram 50 % of those interviewed had tasted sea turtle meat. According to the fishermen, no turtle have ever been landed by any of them on the beach in recent times (the past two years). On the threat posed by dogs to turtle nests, 40.6% o f the interviewees at Prampram, 87.5% at New Ningo and 100% at Old Ningo agree that destruction o f sea turtle eggs by dogs is a major problem in their local communities. 4.1.4 DISCUSSION The aim of the questionnaire, inter alia, was to ascertain: i) nesting sites ii) familiarity of coastal fishers with species of sea turtle in Ghana iii) extent to which they are used as food by coastal dwellers and its legality iv) potential and actual non-human predators o f turtle eggs v) willingness of coastal communities to conserve sea turtles. 34 University of Ghana http://ugspace.ug.edu.gh OcC M V W) ^ VK .5 rCa ^■*' «Sn ccj t *0 ~ h to ^ ^ / « j : U nou CJ 2f o [> u w wo c vn cj OB E O* ccc. jy 5OX) E University of Ghana http://ugspace.ug.edu.gh Interviews conducted among the various fishing communities have shown that the Olive ridley, Leatherback and the Green turtle all nest along the entire coast o f Ghana. The Olive ridley being the most common, followed by the Leatherback, then the Green turtle. Although earlier records show that the Hawksbill turtle is very rare along the coast of Ghana (Irvine, 1947; Toth and Toth, 1974), 10.1 % of the total respondents claim they have seen it at sea in less than five years. The presence o f very few individual Hawksbills in the Ghanaian waters cannot be ruled out. The Ghanaian waters could have served as migratory route for some individual hawksbill turtles in the past. This, however, needs confirmation. Hawksbills are not entirely absent in West Africa as they are reported to nest in relatively appreciable numbers on the volcanic island of Bioko (Tomas et al., 1999), and probably on Cape Verde Island, Senegal and Mauritania (Marquez, 1990). Of the fishermen interviewed, 59.9 % have tasted and most o f them still eat sea turtle meat, while 15.2 % o f those interviewed were aware that it was illegal to capture sea turtles or collect their eggs. Also, 79.6% agree that sea turtles are still landed on their beaches. This implies that poaching activities will continue if education on the need to conserve this resource as well as enforcement of the law that protects sea turtles in Ghana is not intensified in the coastal fishing communities. Destruction of sea turtle eggs by dogs and pigs is not a major problem along the sea turtle nesting beaches with the exception of the stretch from Prampram to Old Ningo where the problem is with the dogs and Lekpoguno where pigs destroy the eggs. 36 University of Ghana http://ugspace.ug.edu.gh It is however, encouraging that as much as 45.7% of the interviewees expressed willingness to join any sea turtle management committee that will be established in their local communities. On the other hand, 53.3% of the respondents said they would join if given some incentives. This is also a good sign, in that with some form of incentives, an overwhelming majority will be attracted to help protect sea turtles in their local communities. The fact that most o f the fishermen in the intensive study area are o f the opinion that the numbers of nesting sea turtles on their beaches have increased in the past few years is good news for both conservationists and fisheries managers. It is not really clear as to whether this assumption is true or not. A complete estimation and monitoring of nesting sea turtle populations along the entire coast of Ghana is therefore necessary in the near future. If it is true that no fisherman had landed any sea turtle on the beach at Prampram in the past two years as the fishermen claimed then it means that earlier education efforts by the NGO, Resource and Environmental Organization (REDO) which began in 1996 at Prampram, has been successful (Armah et al., 1997). It also has promoted the compliance of the Wildlife Conservation Regulations, 1971; L.I 685 which protects sea turtles in Ghana. Most of the fishermen suggested that dogs found destroying sea turtle eggs at the beach should be shot dead. Others are o f the view that the dogs should be poisoned. The shooting of dogs at the beach will be very dangerous in that one might mistake a human poacher for a dog. 37 University of Ghana http://ugspace.ug.edu.gh The poisoning of dogs at the beach will not be easy to implement. A better option would be to protect nests with wire mesh enclosures ensuring that the mesh sizes are big enough to allow the passage of hatchlings. In such cases the beaches should be patrolled every morning to free any stranded hatchling. 4.2 BEACH SEDIMENT CHARACTERISTICS AND DISTRIBUTION OF SEA TURTLES IN GHANA. 4.2.1 INTRODUCTION Laboratory studies have shown that temperature, gas diffusion and available moisture are the three main factors that influence the survival o f reptilian eggs (Packard and Packard, 1988). However, these factors are complicated by the physical characteristics of the substrate. On a typical sandy beach the coarsest particles are found at the top of the beach and graded down to the finest sediments at the water line. This is because the upper shore o f the beach is dry where there is much windblown sand, because coarse sands drain rapidly, whereas at the lower shore the sediments are wet. According to Gray (1981), coarse sediments occur at the upper shore because as the waves break on the beach the heaviest particles sediment out first. The finer particles remain in suspension longer and are carried seaward on the wave backwash. Several factors, including current speed, the roughness of the sediment and the length of time when conditions are still for the particles to settle down account for the type of deposit. Sediments are not made up o f uniform particle size. They contain mixtures of grain sizes. The degree of mixing of the different types can be represented by a sorting coefficient, which is a measure o f the uniformity of the particle sizes. Well-sorted sediments tending towards homogeneity are typical of 38 University of Ghana http://ugspace.ug.edu.gh high wave and current activity (high-energy areas), whereas poorly sorted sediments are heterogeneous and are typical of low wave and current activity (low-energy areas). In other words, well-sorted substrates are composed of uniform particle sizes, whereas poorly sorted substrates are composed o f particles with a wide range o f sizes. Wave action and current velocity are therefore the two most important factors determining the grain-size distribution and sorting coefficients o f nearshore sediments. The geographic location, rates of precipitation and characteristics of the water table may have tremendous effect on both water availability and gas diffusion in the beach sand which may influence hatching success. According to Buchanan and Kain (1971), two sediments with disparate grain size characteristics will, in most cases also show demonstrable and related differences in many other physical and biological properties, such as bulk density, capillarity, thixotrophy, permeability, oxygenation, plasticity, content and nature of organic matter and bacterial count. In a study on nesting beaches at Ascension Island, Mortimer (1982) showed that sea turtles do not use sand texture as criteria in their choice o f beach. This was because there was no correlation between percent hatching success and nesting density. However, sand texture and hatching success were related. Furthermore, some turtles even laid in sand that produce 0 % hatching success. Also when a turtle fails in her effort to construct a nest, she usually tries again only a few meters away from her aborted nest hole instead of returning to sea to look for a new beach. Some turtles were said to spend the night on the beach digging as many as 12 trail nest holes (Mortimer, 1981; Mortimer and Carr, 1987). After failing to nest, turtles normally re- 39 University of Ghana http://ugspace.ug.edu.gh emerge onto the same beach the following night (Mortimer and Portier, 1989). A study of beaches around the world had provided further evidence that green turtles do not use sand texture as criteria in their choice of nesting beach. Mortimer (1982) therefore defines "Good" sand as one that provides proper temperature, allows adequate gas exchange and provides the eggs with sufficient moisture. Two parameters of particle size distribution are mean particle diameter and sorting coefficient. Particle size distribution and particle shape interacts to determine the pore spaces between the sand grains. The amount of pore space and the shape of the pores in the sand affect both gas diffusion and water conductivity. That is, very small pores can have a capillary effect and retain water. The amount of water in the sand affects gas diffusion, while the rate o f gas diffusion in turn affects water availability, since water can be removed in its gaseous phase. Too much gas can cause desiccation o f the sea turtle eggs. Chemically, some minerals o f sand are hydrophilic and water adheres to their surface making water less available to the eggs. On the other hand, hydrophobic sand minerals repel water making it available to the eggs (Mortimer, 1982). Several sea turtle studies have characterised nesting beaches based on features such as grain size, beach profile and vegetation (Diez and Ottenwalder, 1999). 4.2.2 MATERIALS AND METHOD Sand samples were collected at about 40cm deep from all the beaches visited. Sub­ samples were weighed ftom each sand sample, washed in distilled water for 10 minutes using a 63 [im mesh sieve to remove the silt-clay fraction as outlined by Gray (1981). The washed sub-samples were dried at 100°C overnight (about 14 hours). 40 University of Ghana http://ugspace.ug.edu.gh 100g o f each sub-sample was weighed on an electronic balance and fractionated in a mechanical shaker with series of screens. The mesh sizes o f the screens were 1mm, 710nm, 500nm, 355]um, 250^im, 125nm, and 63]um arranged in a decreasing geometric scale. The amount of sand retained in each sieve was weighed and the median grain size of each sample determined from a cumulative plot. The cumulative frequency was plotted showing the percentage of sediment (from coarsest to finest) on the ordinate and the sieve size on the abscissa. The intercepts o f the 5, 16, 50, 84, and 95 percentiles with the cumulative curve were used to calculate the sorting coefficient using the Inclusive Graphic Standard Deviation (Oi). Where Qi= (0g4 . 0 16)/ 4 + (e 95 - a 5)/ 6.6 The sorting classes produced by this index is given as: Under 0.35 very well sorted 0.35-0.50 well sorted 0.50-0.71 moderately well sorted 0.7I-I.00 moderately sorted 1.00-2.00 poorly sorted 2.00-4.00 very poorly sorted Over 4.00 extremely poorly sorted The Inclusive Graphic standard Deviation (Oi) was used in the calculation o f the sorting coefficient because according to Gray (1981), this formula covers over 90% of the distribution. 41 University of Ghana http://ugspace.ug.edu.gh Sediment type was characterised using the phi ( 0) scale or grain size as below: Grain size phi (0) Type of sediment (A) Type o f Sediment(B) (mm) scale (Classification by student) 256 -8 Cobble 64 -6 Cobble 16 -4 Pebble 4 -2 Pebble 2 -1 Granule 1 0 Very coarse sand Coarse (C) 0.5 1.0 Coarse sand 0.25 2.0 Medium sand Medium (M) 0.125 3.0 Fine sand Fine (F) 0.0625 4.0 Very fine sand 0.031 5.0 Coarse silt 0.0039 8.0 Silt 0.002 9.0 Silt 0.00006 14 Clay (Buchanan and Kain, 1971; Gray, 1981). Finally, the Biay-Curtis similarity index was used to discriminate between stations with similar sediment characteristics. The software, PRIMER, was used to generate a dendrogram. 42 University of Ghana http://ugspace.ug.edu.gh 4.2.3 RESULTS The median grain size of the various sand samples collected from the beaches along the entire coast o f Ghana is shown in Table 1. Sediment obtained from Essiama was very fine with median grain size of 0.06mm, and recorded 3.99 on the phi-scale (Table 1). Fifteen (15) stations (65.22 %) out o f the 23 stations sampled had sediments ranging between medium sand and coarse sand. Sediment from Keta and Denu were coarse and that from Bortianor was rated as very coarse sand. Stations with fine or veiy fine sand constituted 34.78 of the shoreline. Results obtained from the particle size analysis of the sediment from the various sampled stations showed that about 86.96% of the total have sorting coefficient between 0.4 and 0.9 i.e moderately to well-sorted sediment (Figure 4). Three stations have very well sorted (< 0.35) sediments. These were Essiama, PrincessTown and Akentechi. It must be emphasized that in the case of Essiama, the value was zero. Salt Pond, Apam and Denu have sorting coefficient between 0.35 and 0.5 and classified as well sorted. Ten stations (43.48%) have sorting coefficient between 0.5 and 0.71, These were moderately well sorted. Seven stations (30.44%) have moderately sorted sediments (Figure 5). 43 University of Ghana http://ugspace.ug.edu.gh Table 1 Sampled beaches, showing sediment type and nesting activity STATION Ssial No. Sorting coeff Medan gain size PH scale Sand type Nesting activity NEWTOMM 1 0.63 0.13 3.00 F 0 HALF ASSIN1 2 0.76 0.25 200 M 1 EKWE 3 0.8 0.25 200 M 0 ESSIAMA 4 0 0.06 3.99 F 0 AXIM 5 0.88 0.13 3.00 F 0 PRINCES TO/VN 6 0.3 0.13 300 F 2 AKBSfTECH 7 0.3 0.13 3.00 F 2 BUSUA BEACH 8 0.55 0.13 3.00 F 2 TAKCRAQ 9 0.88 0.13 3.00 F 0 KOIVEmA 10 0.88 0.25 200 M 1 CAPE COAST" 11 0.68 0.36 1.49 MC 0 SALT POND 12 0.48 0.25 200 M 1 APAM 13 0.48 0.13 3.00 F 1 WllsNEBA 14 0.68 0.36 1.49 MC 1 SENVABEREKU 15 0.68 0.25 200 M 1 BORTIANCR 16 0.88 0.71 0.49 C 2 PR/WPRAM 17 0.63 0.25 200 M 4 OLD NINGO 18 0.63 0.36 1.49 MC 4 AKPLABANYA 19 0.63 0.36 1.49 MC 3 ADA 20 0.55 0.36 1.49 MC 3 AMjOGA 21 0.88 0.5 1.00 C 3 KETA 22 0.55 0.25 200 M 3 DBJU 23 0.48 0.5 1.00 C 3 No nesting activity 0 Sandtvoe Sporadic nesting activity 1 C -Coarse Moderate nesti ng activity 2 MC =MedumOoeree Hgh nesting activity 3 M = IVtedum Very high nesting activity 4 F = Fine 44 University of Ghana http://ugspace.ug.edu.gh SIEVE SIZE (mm) Figure 4 Cumulative plot for sediment from Old Ningo beach 45 % SEDIMENT University of Ghana http://ugspace.ug.edu.gh 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Median grain size(mm) Figure 5. Grain size versus sorting coefficient 46 Sorting coefficient University of Ghana http://ugspace.ug.edu.gh The dendrogram generated using the Bray-Curtis similarity index showed clear groupings and overlaps o f sampled sediments (Figure 6). For example, Denu(23), Anloga(21) and Bortianor (16) clustered to form one group. Keta (22), Salt Pond (12), Prampram (17), Senya Bereku(15), Takoradi (9) and Axim (5) formed a second group. Three other groupings with distinctive sediment properties are shown on the dendrogram. In general, the results suggest a west to east trend o f beach sediments with increasing mean grain size. No clear trend for sorting was evident. 47 University of Ghana http://ugspace.ug.edu.gh MOTimis sunno-ma CLASSIFICATION ANALYSIS Figure 6 Dendrogram showing classification analysis of sand samples fiom vaiious bcaches University of Ghana http://ugspace.ug.edu.gh 4.2.4 DISCUSSION Long stretches of sandy beaches occur from the extreme west bordering Cote d ’ivore to Axim, and from Prampram to Aflao on the east (Armah and Amlalo, 1998). However, the occurrence o f several kilometers o f sandy beaches does not necessarily guarantee the existence of suitable nesting sites. Sea turtles nest on beaches with a range of physical characteristics of beach sediment. These include mean particle size and sorting coefficient. Mortimer (1982) found that green turtles nest in sands having a wide variety o f textures. At Ascension Island, sands that produced the highest rates of hatching success and which were the easiest for nesting females to construct nests had mean particle diameters ranging from 0.4 to 1.0 mm and sorting coefficient between 0.5 and 1.1. Other factors such as patterns o f precipitation, characteristic of water table, mineral composition and the shape of the particles may introduce additional variables, which could affect hatching success. As shown in figure 5, most (82.61%) of the beaches sampled have mean grain sizes between 0.1 and 0.4mm. Most of these beaches therefore might not be expected to support moderate to high nesting activities. For example, at Old Ningo, a town with high nesting activity, the mean grain size was 0.36 mm, and sorting coefficient was 0.63 which is within the range obtained at Ascension Island by Mortimer (1982). From interviews conducted, sporadic sea turtle nesting activities occur along some of the beaches between New Town and PrincessTown. At Essiama the sediment was very fine and wet making it very compact. This is certainly not good for nest construction. One fisherman said he had once witnessed a turtle trying to dig a nest but abandoned it half way and returned to sea. Moderate nesting of Olive ridley and 49 University of Ghana http://ugspace.ug.edu.gh Leatherbacks occur on beaches from PrincessTown to Dixcove, as well as sporadic nesting by the Green turtle (Figure 3). No nesting is known around Takoradi, but sporadic to moderate nesting occurs between Takoradi and Winneba. Moderate nesting occurs beyond Winneba through Bortianor to some beaches around Accra such as Gbegbeseh and Sakumono. Some hatchlings were picked from Sakumono beach in January, 2000 by national service personnel from the Department of Oceanography and Fisheries (University of Ghana) during a field trip. Most of the nesting activities occur from Prampram to Ada, and after the Volta estuary to Denu. Green turtle nesting along the coast o f Ghana was found to be low, however, constant monitoring of this species would provide a clearer picture. It was surprising that most o f the beaches (medium to coarse) provide nesting habitat for the Leatherback. The nesting population of this species is certainly higher than previously anticipated. Sea turtles are said to have difficulty constructing suitable nests in beaches composed of coarse, dry sand. The tendency to dig multiple nest holes and to re-emerge on successive nights also has been reported at the coarse-grained beaches o f French Frigate shoals, on the Great Barrier Reef of Australia (Bustard and Tognetti, 1969). On the other hand, sea turtles do not normally construct more than one egg chamber on the fine-grained nesting beach at Tortuguero, Costa Rica before laying their eggs. At Bortianor, Anloga and Denu where the sand textures were coarse, these areas turned out to be good sea turtle nesting sites. 50 University of Ghana http://ugspace.ug.edu.gh According to Packard and Packard (1988), sea turtles do not actually use sand texture for choice o f nesting beach but may rely on temperature, moisture and respiratory gases that are considered the most important physical variables that affect the survival of reptilian embryos. This assertion is not supported by the result from this study because no nests were seen in areas composed of very fine (Essiama) and fine (Axim) in the course of this study. Temperature was not considered in this study due to lack of appropriate soil thermometer. One soil thermometer was used during the 1997/98 nesting but was discontinued because it was too short and could not measure temperatures deeper than 30 cm. However, the average temperature measured was 29°C. 51 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE 5.0 REPRODUCTIVE ACTIVITY OF SEA TURTLES IN THE INTENSIVE STUDY AREA 5.1 INTRODUCTION The nesting seasons for sea turtles appear to vary with species and geographic location. With the Olive ridley, this occurs mainly in summer and autumn, however, variations exist from place to place. In Mexico and Central America it is known to extend from June or July to November or December. In Surinam, from April to September, in Cape Verde and Senegal from May to August, and in Northeastern India it occurs from February to June (Marquez, 1990). Nesting season of Green turtle varies in time among distant and near localities due to their wide distribution range. For instance, In the Southeastern Atlantic Ocean, it occurs from November to February in the Gulf of Guinea and in the Ascension Island it occurs between February and April. Nesting by the Green turtles in the Western Indian Ocean occurs throughout the year, and peaks from February to April in Aldabra Island, but around Seychelles and the Comoro Archipelago, it occurs from May to August (Marquez, 1990). In the Eastern Atlantic the Leatherback turtle nest from June to August in Senegal and from October to February in South Africa. In the Western Atlantic Ocean nesting is 52 University of Ghana http://ugspace.ug.edu.gh from March to July. In Colombia, French Guiana and Surinam nesting occurs from March to July, but in Guyana it is between January and March (Marquez, 1990). Within-season, re-nesting periodicity of sea turtles occurs, as they often lay more than one complement o f eggs during each nesting season. According to Marquez (1990), successive nesting within the same season occurs at two weeks intervals, most green turtles lay between 2 and 5 clutches. Some may lay only once or more than five times. For Green turtles, the consecutive nesting emergences are given as follows: Tortuguero (Costa Rica) 10.5 days, French Frigate Shoals (Hawaii) 12.1 days, Surinam 13.4 and Heron Island (Australia) 13.4 days (Mortimer and Carr, 1987). In Tortuguero, the average number of clutches laid by Green turtles is estimated as 2.8 (Mortimer and Carr, 1987). Kemp’s ridley nest only 1.5 times per season, however they nest nearly every year. Clutch sizes of sea turtles vary from species to species and locality to locality. The number of eggs laid by each Olive ridley ranges from a couple o f dozens to more than 155. The mean clutch size for Olive ridley is around 109 eggs with variations among localities. For example, in Mexico the mean size is 105.3, Honduras 108.3 eggs, Costa Rica (Nancite) 105 eggs, and India 113 eggs. Small clutches may also occur as a result of interrupted nesting or the last clutch deposited by an individual in the season. For Green turtle, the mean clutch size ranges from 84.6 eggs in the Solomon Islands to 144.4 eggs in Southeast Africa. Can and Hirth (1962) reported that successive clutch sizes laid by nesting green turtles decreased during the season. Four or five clutches are laid by the Leatherback each season with clutch sizes ranging from 61 to 126 eggs. The Hawksbill is known to have the highest mean individual fecundity 53 University of Ghana http://ugspace.ug.edu.gh among sea turtles. Figures ranging from 93 to 223 eggs for Hawksbill have been reported from Yucatan in Rio Lagartos, and from 51 to 211 in Virgin Islands (Marquez, 1990). Age and size of sea turtle, time of the season as well as distance of migration also have influence on the clutch size. Eggs of Olive ridley take between 45 to 65 days to hatch. The incubation period which correlates strongly with temperature and humidity varies from one nesting beach to another. In Escobilla, Mexico it is known to change along with the season. For example eggs laid between August and September hatch between 47 to 58 days. However in northern beaches such as Sinaloa, it takes between 49 and 62 days. Incubation period for Hawksbill eggs is between 47 to 75 days and that for Leatherback eggs varies between 50 to 78 days depending on temperature and humidity (Marquez, 1990). 5.2 MATERIALS AND METHOD The prime nesting beaches along Prampram through New Ningo to Old Ningo in the Greater Accra region, east of Accra were selected for intensive study of reproductive behaviour. This area stretches about 14 km and lies within the core nesting site o f sea turtles in Ghana (Toth and Toth, 1974; Carr and Campbell, 1995; Armah et al., 1997). The intensive study area was divided into six zones using physical characteristics such as the fort, lagoons and outfall along the beaches. Each zone stretches about 2 km except foT zone F that is approximately 3.5 km (Figure 2). Daily patrols were conducted from August 1998 to January 2000 in the morning. All turtle tracks in each zone were identified according to species using the track width 54 University of Ghana http://ugspace.ug.edu.gh and designs. The number o f each species was then recorded. The numbers o f false crawls as well as freshly laid nests destroyed by dogs were also recorded. Important diagnostic features used to differentiate turtle tracks by species include track width, body pit depth, and whether the diagonal marks made by the front flippers are symmetrical or asymmetrical (Pritchard and Mortimer, 1999). A symmetrical track is formed when the front flippers move together synchronously to pull the turtle over the surface of the sand, resulting in a track in which the right and left halves are almost mirror images. An asymmetrical track is formed when the front flippers move alternately (right, left, right, left, etc.) to carry the turtle forward. Species were identified based on species identification protocols outiined by Pritchard et al. (1983). Lepidochelys olivacea has tracks, which are very lightly cut, with alternating (asymmetrical) oblique marks by the forelimbs. Tail drag mark is lacking or inconspicuous. Track width is between 70-80 cm. Tracks o f Chelonia mydas are deeply cut, with symmetrical diagonal markings made by the forelimbs. It has straight, central tail drag marks present, either as solid or broken line. Track width is typically between 100-130 cm. Dermochelys coriacea tracks were identified from their very deep and broad, with symmetrical diagonal marks made by the forelimbs, and usually with a deep incised median groove formed by dragging the tail. The width is between 150-230 cm. Tracks of Caretta caretta are moderately deeply cut, with alternating (asymmetrical) diagonal marks made by the forelimbs. Typically no tail drag mark. Track width is between 70-90cm. 55 University of Ghana http://ugspace.ug.edu.gh False crawls (non-nesting emergences) were determined by careful examination of the entire turtle crawl. Examples of false crawls made by sea turtles include extensive wandering with no body pitting or digging; U-shaped crawl to the high tide line; considerable sand disturbance with evidence of body pitting and digging with a smooth-walled chamber and no evidence of covering. Also crawls where the relative lengths of the emerging and returning tracks of the turtle are the same. A thorough understanding of the nesting behaviour of each species is, however, critical to accurate interpretation of false crawl signs. To estimate the emergence period and clutch size, nests were marked with wooden pegs, and dates on which eggs were laid and hatched were recorded. A method for estimating clutch size after emergence from nests was devised. With this method, marked nests were excavated a few days (3 to 5 days) after the eggs were hatched and all eggshells from each nest collected (Plates 3 and 4), sun dried, and weighed. The total weight of all eggshells from each nest was divided by the average weight o f each eggshell to estimate the clutch size. Eggs that did not hatch were also counted to estimate the percentage hatching success. 5.2.1 COLLECTION OF BIOMETRIC DATA Periodic visits to the beach at night were undertaken and sea turtles were measured to estimate the average size of each species. Both minimum curved carapace length (CCLmin) and maximum curved carapace width (CCWmax) o f each turtle encountered were measured (Figure 7) and recorded. The measurement was done when the turtle had just finished laying its eggs. Nest depths and track widths were measured for both the Olive ridley and Leatherback. A simple but effective method 56 University of Ghana http://ugspace.ug.edu.gh Plate 3 Collecting sea turtle eggshells at zone F (Old Ningo) Plate 4 Excavating Leatherback nest at zone D (New Ningo) 57 University of Ghana http://ugspace.ug.edu.gh was used to measure the nest depth. This was done by placing a 50 cm ruler horizontally across the edge of the nest and dipping a graduated thin rod into the turtle nest and reading the value where the rod touches the ruler. This was done just before the first egg was laid, or after the first egg had dropped. The measurements were taken quickly and carefully to avoid any disturbance to the nesting turtle. Eggs were also counted as they dropped by the aid of a touch light and also a night vision camera. Data recorded from the clutch size estimates and that counted were subjected to t-test. This was to determine the reliability of estimates made from total eggshell weights. Figure 7 M ethod of measuring CCLmin and CCWmax. 58 University of Ghana http://ugspace.ug.edu.gh 5.3 RESULTS 5.3.1 NESTING SEASONALITY AND RELATIVE OCCURRENCE. Figures 8 and 9 show the relative occurrence and percentage respectively of nesting species in the entire intensive study area and within the identified sub-zones/ segments. Other aspects of nesting activities such as total number of nests, percent false crawls and seasonality are presented graphically in figures 10, 11, and 12 respectively. Results obtained from analysis of data collected during the study period, spanning the period o f August 1998 to April, 2000 showed that Olive ridley (L. olivecea) accounted for about 91% nests. Green turtles (3%) and Leatherback turtles (6%) nest mainly from November to February. However, very few tracks (about seven per month) were recorded for Green turtles in August, September and October 1998 along the whole stretch of the beach from zones A-F. Also, two tracks o f loggerhead turtle were recorded in December, 1998 at zone C. Average percentage false crawls by the nesting turtles for zones ‘A ’ to ‘F ’ were respectively 12.65, 28.24, 0, 7.23, 13.38 and 9.29 (Tables 2a-f). 5.3.2 EMERGENCE PERIOD, CLUTCH SIZE AND PERCENTAGE HATCHING SUCCESS The mean emergence period for Olive ridley was found to be 52 days (SD=2.12; Table 3), Green turtle 56 days (SD=2.57; Table 4) and Leatherback turtle 62 days (SD=3.5; Table 5). Average clutch size estimates for Olive ridley, Green and Leatherback turtles were 86 (SD=12), 83 (SD=11) and 81 (SD=18) respectively 59 University of Ghana http://ugspace.ug.edu.gh Leatherback 6% Green turtle 3% ■ Olive ridley ■ Green turtle □ Leatherback Olive ridley 91% Figure 8. Relative occurrence of each nesting species in the intensive study area 60 University of Ghana http://ugspace.ug.edu.gh loo%- 90% 80%- 70% 60% 50% ■ Hawksbill 40% □ Loggerhead 30% ■ Green turtle □ Leatherback 20% ■ Olive ridley 10% Figure 9. Relative percentage of each nesting species in each zone of the intensive study area 61 % OF SEA TURTLE TRACKS University of Ghana http://ugspace.ug.edu.gh 2500 2000 1500 1000 500 0 Figure 10. Total number of nests recorded in each zone during the study period 62 NO. OF NESTS University of Ghana http://ugspace.ug.edu.gh 100% 80% 60% 40% □ No dest by dogs ■ fa lse crawls I Total tracks 20% “T 1 ' i ' i I i A B C D E F ZONES Figure 11. Percentage false crawls and nests destroyed by dogs in each zone 63 NO. OF SEA TURTLE TRACKS University of Ghana http://ugspace.ug.edu.gh 600 500 400 300 200 100 III MONTHS □ O live ridley ■ G reen turtle □ Leatherback ■ Loggerhead B H aw ksb ill Figure 12 Sea turtle nesting activity at intensive study area 64 NO. OF SEA TURTLE TRACKS University of Ghana http://ugspace.ug.edu.gh (Tables 3, 4 and 5). Also, percentage-hatching success was higher in Olive ridley 92.40 (SD=4.56) than Green turtle 89.38 (SD=6.64) and Leatherback 88.32 SD=2.70) as shown in tables 3, 4 and 5. 65 University of Ghana http://ugspace.ug.edu.gh CO S S Ogl 0s0 gO) 5 8* 8 —8: —S. —8. •8 .8 •S | S ^ 8 8 8 00 . a a . — i n • _* _* . a v — , T i • fv. o o o oi S ^ j o o r v : c o 2 J Q o o o o o o r " ^ ^ O l f l l f l f M C O ^ O O O O O O W t D W i n f S I O O O roCO s O M(1 inIf) o s OOITJCMOOOOOO cd co in o o o co o o o o o o a j cm c o d o o £CM cm c o (Bi nf f i J lTtnrOOOOQOt t l l l i nr - rOOQ CMCO r5t ^0i D( p(TDfiT o f -l Ti nf T( O- wOJ nD na no N( M! (M0 0n 'n- ^nC[ MO fiO- mn an a> o o o o a o o o a a o o o o o o o o o o o COCO o o o o o o o o o o o o o o o o o o o o o o o o o o - < t ^ ( D O O O o o o o o o ^ ! eCMo r « - o Q co cM®j, Srr\® 5*-r-p. ) p p S p 0) o n h >~O O£ z mz ^mK ££> q;l Z> j G E L I - > o - , f f i K r K « - _ LU < U J < o a < w“■0O. OzZO oOl U3> i n N o 10 o o t o o o o o o o C—D 1■ ^- i—> ’O—> o otv. co oo o < ̂ ' t—> f—1 ® ® ® ^ N oOd Od Od Qd d ®0 *S: '°2. 0ON « ®<° 0° 0° <\l OO dO tcos n i f l O O ^ f f l O O O O O O I / l N O N P I ' - r O O £ o o o o o o o o o o o o o o o o o o o o o CO o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o ▼-t- O C N I O O O O O O O O O O O O O O O O O I f i CO CO 05 CM •f- T - CNJ CO M o a a o a 3 oo oooo oqp oa. . ) f f i ® o j r o 2 ® 0 ! 2 a i a ) f f l ® 5CM P S A Ql a O L L l < l i J < q - < 3 3 Z ) L U O O U j 5 cc in 3 LD o z C/3 0 Z Q - 5 U - S < S - ) - > < W 0 2 : q ^ < CL 2 < Olive ridley Green turtle Leatherback Loggerhead Hawksbill Total False crawl % False crawl Dest. By dogs % Dest.by Dogs University of Ghana http://ugspace.ug.edu.gh O O O O O O O O O O O O O O O O O O O O O o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o C\J ( M ^ o o o o o o c N c o T r e M f - o o o o o>CM o o o o o o o o o o o o o o o o o o o o o 03CO O O O O C M O O O O O O O O O O O O O O O O CM o o o o o 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 T- O O O O O l T i O O O O O O O O O O O O O O O t - ’̂ - C M ’ - O ^ J - O O O O O O C M C O ' J ’ C M ' - O O O O CM co O ra CD =) 5 i y O o< ( 0 O Z mQ <- >i iUj <. Sa<. S< -D>D- 5D> o o o o o o o o o o o o o o o o o o o o o o 0C3O o o o o o o o o o o o o o o o o o o o o o o <3500000000 k ! c \ C N i < N < r 3 i n i n ^ o o o o o o T - c N j o o o o o o COCM (' JO(nCi0f 3l fSfNl (.Ml DNC( DM W! O' -O a o n 'OJ ^WQi n( !E DO C( N0 iOn )' tTN- n_ 001 ■o ( ooJ Peoooooo) 0? 2J 2? ®S? o>2? o) 0>a>a> o O OaM a> p p0 o raH o LI -ra LU < CL “ < = I O3 PUhwJ oO >o Um <5< Q ~> < z CSL­ o u. 2 < »- Olive ridley Green turtle Leatherback Loggerhead Hawksbill io ta I False crawl % False crawl Dest By dogs % Dest.by Dogs University of Ghana http://ugspace.ug.edu.gh O Nm oNi '- tCnMNc Ns i Oi nOo oOoOo Oo al iDO nC cMqt ioOo oBcBo c^ oOoOo (C0O cvi csi cd f\i d o ’ d o' o ’ t>-‘ oj cm" iri in o o p«i 0>-(M(M10NrOOOOO(M^(MtK)»-rOO £J 03 CM 0m5 c0 5 o 0 0c oO o/ fgK g g g gs g s i Xs c■HJ qO - cCO qCci q* g<«•i* . ca CO n cn• iO O O O O O T - f l O r * r t O K’ CO n cn O O O O T- CO 00 WO) Ntrj r- co ID o CM G 00 m 05 Sto co o a o o CO fO T - CM CD O ) o CM r r s N CD CO N CO o C0M0 o o o o o o o o o o o o o o o o o o o o o ho- o o o o o o o o o o o o o o o o o o o o o o o o a o un a o o o o a o o a ^ J E o i o o toCO C O lO oO ^ C D O O O O O O C M C M ^ - O O t- O O ccno 00 CD 00 00 CO CO CD CD T - a) T- IX. CM LO 00 h - N - CD T t co o a o o CM ^ (O N to O sT—- a) CM 51)^’ p) 0) SJ ) 5) p) ?>p) P>o) p) 0) p) P) P) ® p po p n 0 >o o cLUo at 0a.: > ■Z. UCLJ P s re Z3 D => CJ O> o CD O' ♦o-*LU Q < uj < 111 <£L L “ 3 “ 5 c/d o z a k Olive ridley Green turtle Leatherback Loggerhead Hawksbill Total False crawl % False crawl Dest. By dogs % Dest.by Dogs University of Ghana http://ugspace.ug.edu.gh CO o00o o o > N - m r - kCD o 1 ." lo CO O)CO 00 COGO CM aO>l 00 o CD CO oCM CM CO m CO O r - t- a o o o cm o CD o G ) oo co Clco i n co cm CM CM CM CN t - -r- O O CM N « ^ N O O CM ID Lfi t- Q t- CO ° B M P cnfvJ“ NCV, 0 ^ Q 0 rCM! C:M ?CM® CO^ CM( D tD COCM O O O Q Q O O Q Q O a O O Q Q Q O a O O O O O O O O O O O O O O O O O O O O O O O O O o o o o CD CM CO O O O O O O T j - O O Tj- O ▼- CO T- CN CO CM o o CM C n C M O O O O C M C M C M O C S l T - O O O CO t - CM CM CM CM CM CM CO CM CM N 00 $ 03 P 5 o « p n - - z a , ? ¥ r f Z j Wf li ,- > U 7 ( D a : Q: 2 n 3< 0U)J0« < i