Herpetofauna under threat: A study of Kogyae Strict Nature Reserve, Ghana

The herpetofaunal species in Ghana are under serious threat from habitat loss and degradation, global climate change, disease and parasitism, introduced invasive species and unsustainable use. The herpetofaunal species of the Kogyae Strict Nature Reserve of Ghana were surveyed using two methods: refuge examination via direct searches, visual/audio surveys and interviews of local residents. Data were obtained during two visits to two study sites in 2012: savanna woodland and riparian forest mosaic. The first visit was in the dry season in January, 2012 for five days and the second visit in the rainy season in June, 2012 for six days. Thirty-six herpetofaunal species were recorded, comprising of 14 amphibians and 22 reptiles. Ten of the reptiles were recorded only from interviews. The savanna woodland recorded a higher number of species (33) than the riparian forest mosaic (22), and there was low similarity in species between the two sites (Sorenson’s similarity index, C S = 0.42). Four species were frequently encountered at both sites in high numbers: Phrynobatrachus natalensis , Phrynobatrachus latifrons , Arthroleptis spp. and Trachylepis affinis . Two lizard ( Varanus niloticus and Varanus exanthematicus) and two snake ( Python sebae and Python regius) species are of both local and global conservation significance. The lizards are categorized under Schedule I (complete protection) and the pythons under Schedule II (partial protection) of the Ghana Wildlife Conservation Regulations. The main threats to the herpetofaunal species of the study area include habitat destruction due to annual bushfires and killing of large reptiles like the monitors as well as snakes. It is recommended that riparian vegetation bordering streams should be particularly targeted to conserve herpetofauna of the study area.


INTRODUCTION
Herpetofauna (amphibians and reptiles) are a diverse but cryptic component of an ecosystem, and can thus serve as excellent bio-indicators of stressed ecosystems (Leduc, 2012). Amphibians are especially sensitive to ecosystem changes because of their biphasic lifestyle which brings them in direct and constant contact with *Corresponding author. E-mail: yahayarousm@gmail.com, Tel: +223-244949109.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License their environment and exchange of gases through their moist, semi-permeable skin, which plays a role in chemical uptake. While reptiles are better protected against the environment with their integument covered with scales, their eggs are still susceptible to metal contaminants (Leduc, 2012) like those of amphibians. Both amphibians and reptiles are ectothermic and rely on environmental conditions to maintain metabolism and other life processes, and also are susceptible to acidification and metal contaminants (Leduc, 2012).
Conservation strategies are often targeted at glamorous taxa such as birds and mammals, neglecting smaller and less conspicuous vertebrates like herpetofauna, which are threatened and are declining more rapidly than birds and mammals (Ramesh, 2013). The main causes of herpetofaunal population declines include habitat loss and degradation, global climate change, disease and parasitism, introduced invasive species and unsustainable use (Gibbons et al., 2000;Stuart et al., 2004).
Wildlife protected areas (PAs) are of economic importance as they contribute to improvement of surrounding communities" living standards. Livelihood support programmes and community-based tourism programmes exist in some communities surrounding some protected areas (IUCN/PACO, 2010), which are of cultural, aesthetic, and spiritual significance, in addition to being sources of wood and medicinal products. In Ghana, PAs are under threat from poaching, bushfires and land conversion due to farming and grazing (IUCN, 2010). There are 23 wildlife PAs in Ghana with a total area of 1,347,600 ha or 5.6% of the country"s total land area (IUCN, 2010). These include seven national parks, six resource reserves, two wildlife sanctuaries, one strict nature reserve, one biosphere reserve and six ramsar sites.
The Kogyae Strict Nature Reserve is the only Strict Nature Reserve in Ghana. Strict Nature Reserve is defined by the International Union for Conservation of Nature (IUCN) as Category Ia: A protected area managed mainly for scientific research and monitoring; an area of land and /or sea possessing some outstanding or representative ecosystems, geological or physiological features and/or species (IUCN, 1994).
Information on herpetofaunal species diversity in Ghana"s protected areas is scanty and not regularly updated. Inventories of species in reserves are important for conservation, monitoring and management as well as the acquisition of baseline data on the distribution and status of even common species (Trakimas, 1999). The forests around Ejura in the Ashanti region have been totally destroyed by human settlements. Located in the area, Kogyae, the only Strict Nature Reserve in Ghana protects an important fragment of riparian forest left on the Afram Plains. This survey was therefore conducted to update existing knowledge regarding herpetofaunal abundance, distribution and diversity in the Kogyae Strict Nature Reserve; provide a preliminary herpetofaunal species list for the reserve; determine existing threats to herpetofauna in the reserve.

Study area
The Kogyae Strict Nature Reserve (KSNR) (07°12"N 01°11"W) (Figure 1), with an area of 388 km², is located in the forest-savanna transition zone. It is bordered by the Afram river and riparian forest along its south-western boundary. There is also transition woodland, a small pocket of dry forest and small rocky hills in the west (Dowsett-Lemaire and Dowsett, 2005). Much of the reserve has lost its status of a "strict nature reserve", due to logging and hunting activities as well as an increasing number of farms encroaching from the south and east (Kyerematen et al., 2014). There are forest remnants, part of the original Kujani forest (Sam and Wilson, 1994) belonging to a dry type of semi-evergreen or deciduous forest with small pockets of Anogeissus leiocarpus, Ceiba pentandra, Cola gigantea, Khaya senegalensis, Milicia excels and Triplochiton scleroxylon.

Methodology
Visual encounter surveys (Rödel and Ernst, 2004) and refuge examination were undertaken either opportunistically or with acoustic searching, by turning over rocks and fallen logs, peeling tree barks, digging through leaf litter, and searching through trees and buildings, rotten tree stumps, tree buttresses, termite mounds and burrows. Care was taken to ensure minimal disturbance of habitats during refuge examination by returning objects moved to their original positions after searching them. A three-man search team positioned themselves five metres apart from each other to search for herpetofaunal species during the day. On sighting a lizard or a snake, a member signaled to the others to converge at a point to surround and capture the specimen. A running animal was chased and pinned down gently with a stick or a snake hook, then hand-picked. Amphibians were surveyed in and around ponds and puddles at the study sites (Heyer et al., 1994).
Interviews were conducted with a cross-section of inhabitants to supplement information obtained from the other methods. The interviews focused on the different types of species commonly found in the study area and some indication of their abundance (commonness and rarity). A manual was shown to the respondents to help them identify the various animals known to occur in the area and to answer questions about those animals. Questions asked included whether the respondent knew a particular animal, and whether he/she had seen that animal before, how often and when the animal was seen. Respondents were not considered "experts" in herpetology, but since most of the inhabitants of the study area were farmers, they were expected to have encountered some herpetofaunal species during their farming activities. The few hunters among them seemed to have more information about wild animals in general.
Animals were recorded based on sightings, captures, calls/sounds and interviews. Such records were considered to be "encounters". There was the possibility of multiple counts since animals were not marked, but to minimize multiple counts, no one particular section of a study site was repeatedly surveyed within each survey period. Amphibian calls were recorded at various water bodies (ponds, pools, puddles, and streams) in the study sites. At each water body, the recorder listened carefully to a chorus and was able to distinguish calls of individual species that were clearly heard. Calls were also recorded and played back later for confirmation. In a chorus, only species presence can be recorded as it is difficult to attach numbers. However, at each water body, individuals begin calling and are later joined by others to form a chorus. Hence, the position of those that begin the calls can be pinpointed, usually along the edges of the water body. Those calling whose positions were clearly identified were recorded. Voucher specimens were euthanized with chloroform, fixed in 10% formalin and preserved in 70% alcohol. General herpetofaunal identification followed Hughes (1988) and Leache et al. (2006), while amphibian species were identified using Rodel (2000), Rodel and Agyei (2003), Rodel et al. (2005) and Onadeka and Rodel (2009). Skink identification was based on Hoogmoed (1974), while Chippaux (1999) and Trape and Mane (2006) were used for snake identification.
The survey sites were visited twice, once in the dry season (5 to 9th January, 2012) and once in the rainy season (16 to 21st June, 2012). There were 11 survey days in total, five days in the dry season and six days in the rainy season.
Species encounters were plotted on an accumulation curve. In Ghana, the dry season falls between October and March, while the rainy season occurs between April and September. The possibility of multiple counts and the fact that the survey was mainly semiquantitative and qualitative, did not allow for rigorous statistical analysis thus only involved a species accumulation curve and Sorenson"s qualitative index. Sorenson"s similarity index (CS) was used to determine the extent of similarity between the sites (Magurran, 2004) as follows: Where, a = number of species at first site, b = number of species at second site c = number of species common to both two sites.

Herpetofaunal abundance and diversity
There were direct encounters of 26 herpetofaunal species comprising of 14 amphibian (two toads and 14 frog) species and 12 reptile (six lizard and six snake) species belonging to 12 families. Seven amphibian families were recorded, notably the Bufonidae, Ranidae, Dicroglossidae, Ptychadenidae, Hyperoliidae, Arthroleptidae and Petropedetidae, There were also three lizard (Agamidae, Scincidae and Gekkonidae) and two snake families (Colubridae and Elapidae) recorded (Table 1). Ten more reptile species were recorded only through interviews of local residents, bringing the total recorded herpetofaunal species at KSNR to 36. This was made up of 22 reptile (eight lizards and 14 snakes) and 14 amphibian species belonging to 15 families ( Table 2). Members of the family Varanidae as well as two snake families, Pythonidae and Viperidae were not directly encountered.
The interviews indicated that monitor lizards; Varanus exanthematicus and Varanus niloticus were regularly encountered, mostly on weekly basis. Bitis gabonica, the largest viper in Ghana, was rarely sighted. The last time one encountered was about five years prior to this survey when it bit a woman who died the same day. The viper was promptly killed by the people and eaten. About two or three individuals of Python sebae were sighted yearly, and these were killed and eaten. Python regius were rarely sighted, with no respondent able to recollect accurately the last time they sighted a Python regius.
There were higher numbers of amphibians encountered than reptiles, with 60 individual encounters each of Phrynobatrachus natalensis, Phrynobatrachus latifrons (both through calls) and Arthroleptis sp. This was followed by Hoplobatrachus oocipitalis (22 encounters), Phrynobatracus francisci (20 encounters), and Afrixalus dorsalis (11 encounters). Five amphibian species were encountered only once: Amnirana galamensis, Ptychadena bibroni, P. longirostris, Leptopelis viridis and Hyperolius concolor. Among the reptiles, Trachylepis affinis and Agama agama were the most encountered, with 58 and 20 individuals respectively. Five reptile species were encountered once, notably the lizards Trachylepis maculilabris and Panaspis togoensis and the snakes Thelothornis kirtlandii, Lamprophis lineatus and Naja nigricollis. Seven more species, all amphibians, were encountered in the savanna woodland (22 species) than in the riparian forest mosaic (15 species) with 11 species common to both sites (Table 1). Eleven species were thus exclusively encountered in savanna woodland, while four species were exclusively encountered in the RFM (Table 1).

Species accumulation curve
The species accumulation curve rose sharply for the first two days, and continued to rise gently to the end of the dry season survey on the fifth day. The curve continued to rise in the rainy season survey and then flattened completely at the end of the rainy season survey ( Figure  2).

Similarity
Overall, Sorenson"s similarity index was 0.42, indicating low similarity between the two sites. There were 21 species common to both sites, with 13 species exclusive to the SW and two species exclusive to RFM (Table 2). In the dry season, 12 herpetofaunal species were recorded in SW (seven amphibian and five reptile species) while 10 were recorded in RFM (three amphibian and seven reptile species) with six species common to both sites. In the rainy season, 13 species were recorded in SW (seven amphibian and six reptile species) while nine species (three amphibian and six reptile species) were recorded in RFM with six species common to both sites (Table 1).
In the SW, the number of amphibian species (seven) was the same for the two seasons but one more reptile species (six) was recorded in the rainy than the dry season (five). Only three species, one frog (Hoplobatrachus occipitalis) and two lizards (A. agama and T. affinis) were common to the two seasons. In the RFM, the number of amphibians was the same for both seasons, but there was one less reptile in the dry season than in the rainy season. Four species (two lizards -Trachylepis quinquetaniata and T. affinis and two snakes -Psammophis sibilans and Paranerita irregularis irregularis) were common to the seasons (Table 1). T. affinis was the only species recorded in all the surveys.

DISCUSSION
Hughes (1988) (Hughes, 1988). The 14 amphibian species recorded in this survey falls within the range normally recorded for amphibian surveys in Ghana. The flattening of the species accumulation curve in this Kogyae survey suggested that the commonest species in the area had been recorded.
According to Omogbai et al. (2002) a population explosion of amphibians occurs during the rainy season, and this consequently influences the populations of snakes. This was not the case in this survey, as there was no observable difference in the number of amphibians recorded in both dry and the rainy seasons. Also, the annual bushfires in the dry season did not appear to have influenced the species numbers, with the number of species remaining similar for both dry and rainy seasons.
Six species, comprising of three skinks (Hemidactylus muriceus, T. maculilabris and P. togoensis) and three snakes (B. gabonica, Desmarestia viridis and Thelotornis kirtlandii), are known forest inhabitants (Hughes and Barry 1969;Hughes 1988). The remnant of the dry semideciduous forest in the Kogyae Strict Nature Reserve could therefore support sustainable populations of existing forest herpetofauna if well-protected.

Conservation issues
Most of the herpetofauna recorded in this survey are listed as "Least Concern" in the IUCN Red List of Threatened Species. However, in the Ghana Wildlife Conservation Regulations (1971), Varanus niloticus and V. exanthematicus are listed in the first schedule (complete protection), while both Python sebae and P. regius in the Second Schedule (partial protection) and all venomous snakes are listed in the fifth schedule (can be killed when their population expands to make them dangerous to humans and/or their livestock). The dry season survey (January 2012) recorded large-scale bushfires in both SW and RFM, but this did not appear to affect the species composition as shown in Table 1.
The main threats to the herpetofaunal populations in the study area are largely anthropogenic, notably the annual bushfires and indiscriminate killing of species like monitors and pythons. The bushfires not only kill the herpetofaunal species, but also destroys their habitats. The monitors and snakes are killed largely for food, while snakes in general are killed because of irrational fears arising out of human superstition. This situation has negative consequences for ecosystem balance (Attuquayefio, 2004). Poverty and lack of formal education are widespread in most parts of rural Ghana. Initiation of education and poverty alleviation programmes will thus go a long way to improve interaction between humans and wildlife in such areas. This study incorporated an outreach component, where the researchers organized a durbar to meet the communities within the study area and sensitize them on the nature and importance of such researches. At such durbars, the communities were educated on the importance of wildlife and the need to protect them.
The estimated 2,000 to 3,000 sacred grooves in Ghana serve important ecological and socio-cultural functions by preserving virgin forests as well as being important refuges for rare and important local biodiversity and a source of herbs for medicinal, social and religious purposes.
Sacred groves are defined as "small patches or islands of remaining original habitat" or "traditionally-p rotected tracts of land of varying sizes that may be as old as mankind" (Attuquayefio and Fobil, 2005). They range in size from hundreds of hectares of forest to single trees or a few stones as ancestral groves, shrines, ancestral forests and burial grounds of different ethnic groups in Ghana (Ntiamoa-Badu, 1995). Sacred groves are considered to be one form of traditional conservation practice in Ghana. Traditional conservation is also practiced in Ghana in the form of taboos (traditional laws) and myths. In some urban and most rural areas in Ghana, taboo days exist for farming, fishing and hunting. Spiritual reasons are given as to why people could not farm, fish or hunt on certain days. Such traditional practices enabled the protection of biological resources from human disturbance and over-exploitation (Attuquayefio and Fobil, 2005).
Different animals are under varying forms of protection based on sacred groves and taboos Snakes, however, except Python regius and to a lesser extent, P. sebae are killed on sight, mainly due to fear and the fact that most people in Ghana erroneously perceive all snakes as venomous and dangerous (Attuquayefio, 2004). In some communities in Northern Ghana, pythons, monitor lizards and crocodiles are tabooed animals which are not killed by the people. For instance, the Paga Crocodile Pond in the Upper East region of Ghana, is a popular tourist attraction. Some rituals are performed by the keepers of the pond, after which the crocodiles are called out of the water for people to play with them, sit on them and take photographs, etc. People in the twin towns of Boabeng and Fiema in the Brong-Ahafo region of Ghana live and interact with monkeys on a daily basis. The monkeys are fed and cared for by the town folk when they visit homes, and they are buried in a cemetery after customary rites are performed when they die (Attuquayefio and Gyampoh, 2010).
Disregard of taboos attracts severe punitive sanctions to culprits and high prices of atonement including making sacrifices and performing certain rites to avert any future mishap (Attuquayefio and Fobil, 2005). Unfortunately, this has not been effective deterrent to the destruction of biological diversity mainly due to rapid population growth, influence of foreign religions and beliefs and increased dependence on western technology.

Conclusion
While many herpetofaunal species thrive in anthropogenically disturbed areas, others do not survive intense habitat destruction because of their restriction to specific microhabitats (Leach et al., 2006). Preserving riparian vegetation bordering streams and rivers will be one of the most effective methods of conserving the herpetofaunal community in the study area. As important inhabitants of tropical ecosystems, amphibians are extremely sensitive to habitat alteration. The composition of herpetofauna assemblages is known to reflect the degree of habitat degradation and destruction. The presence or absence of particular herpetofaunal species could thus form the basis of conservation and management recommendations (Leache et al., 2013). The flattened species accumulation curve indicates that the most common herpetofaunal species had been encountered, and future long-term monitoring programmes could focus on such species. The results of this survey could be useful in baseline monitoring however, there is the need for a more exhaustive study to build upon this study.