Abstract: Despite Ethiopia is among the few African countries with high mammal species diversity, its mammalian resource is highly deteriorating and there is little information about the mammalian resource. As a result, a study was carried out to determine the species composition, distribution and relative abundance of large wild mammals during dry and wet seasons from July 2010 to February, 2011 in and around Wondo Genet Forest. The study area was stratified into four habitat types (natural forest, plantation, wooded grassland and agro forestry) based on land-use and vegetation cover of the study area. Stratified systematic sampling method was used to lay line transects in each habitat type. Species identification and total counting of individuals were made along each transect. A total of 19 species of large wild mammals were recorded in the area. The commonest mammals were: Papio anubis. Chlorocebus aethiops, Colobus guereza and Tragelaphus scriptus meneliki. In terms of abundance, Papio anubis (48%) and Colobus guereza (32%) were the most abundant species, while Felis serval and Felis caracal were the least abundant that each contributed only less than 1% of the total observations. The highest Simpson’s Index of diversity (1-D) was obtained in the agro forestry habitat (0.769), while wooded grassland had the lowest (0.383). The seasonal variation in abundance (number) of individuals of large wild mammals was significantly different (χ2 = 153.309, df =1, p<0.05). However, the seasonal variation in species composition of large wild mammals was not significantly different (χ2 = 0.125, df = 1, p>0.05). Despite the study area is home for different species of mammals, livestock grazing, agriculture and human encroachments are very evident in the areas, putting an enormous strain on the flora and fauna. Hence, there is a need for urgent conservation measures to save the area.
INTRODUCTION
Ethiopia is among the few African countries with high mammal species diversity (Yalden et al., 1996). The country currently possesses 284 species of mammals (Bekele et al., 1993; Cole et al., 1993; Bekele and Corti, 1997; Abune, 2000), of which 31 are endemic. The high level of endemicity of mammals in Ethiopia is attributed to the large extent of highlands, isolated from the rest part of Africa and the variations in temperature and rainfall among different habitats (Yalden, 1983; Yalden and Largen, 1992).
However, due to the expansion of human settlement and agriculture that lead to habitat destruction, the natural habitats range of many wildlife species have increasingly become smaller. As a result, the wildlife populations are forced to occupy isolated habitat areas that are often found in National Parks (Tedla, 1995; Hundessa, 1997; Stephens et al., 2001; Mamo and Bekele, 2011; Girma et al., 2012).
The Ethiopian highlands are among the most densely populated agricultural areas in Africa (Evangelista et al., 2007). This in turn has led to formation of many forest fragments in most part of Ethiopia (Hundessa, 1997). However, the significance of small fragments of wildlife habitats that exist outside protected areas in maintaining diverse groups of wildlife species in Ethiopia is poorly understood (Hundessa, 1997; Girma et al., 2012).
The present study area consists of fragments of forest in which part of it is protected by Wondo Genet College of Forestry and Natural Resources. The study area was thought to be home to different wildlife including large mammals, like Tragelaphus buxtoni, Tragelaphus scriptus meneliki, Phacochoerus africanus, Sylvicapra grimmia and Papio anubis (Erikssion and Stern, 1987). Nevertheless, no published information exists about mammalian species type, their abundance and distribution of the study area. The area was known to support 118 species of birds, of which 7 were endemics, 3 inter-African migrants and 6 inter-continental migrants (Sim, 1979). Although the area is thought to be a home for diverse wildlife species including reptiles, amphibians and insects; there are barely few scientific investigations carried on the multifaceted aspects wildlife species in the study area.
Most part of the forest was relatively protected after the establishment of the college in the late 1970’s, although different human activities and anthropogenic pressures appear to be unabated. As far as the existing threats to wildlife of the study area continue to prevail, the survival of these mammals would be negatively affected. The impact of human pressure on wildlife of the area was noted by Dessie (2007) and Mamo (2007), they have reported among the wild mammals, Tragelaphus buxtoni was locally extinct. Therefore, the aim of the present study is to describe the species composition, distribution and relative abundance of large wild mammals along systematically laid transects in the study area.
MATERIALS AND METHODS
Study area: The study area is situated in the southern part of Ethiopia at about 250 km south of the capital city and located between 38°36’30” and 38°39’ 0” E longitude and 7°5’30” and 7°7’30” N latitude (Fig. 1). The total area coverage of the study area is 958 ha. The lower elevation areas are mostly covered by agriculture, the college infrastructure and plantation forests, while the uplands are covered by the natural forests and wooded grassland. For this study, the area was stratified into four habitat types based on altitudinal ranges and vegetation types. Accordingly, natural forest, plantation, wooded grassland and agro forestry habitat types were identified (Fig. 1). Natural forest represents valley areas and middle altitude areas (2000-2200 m.a.s.l.) with dominant tree species of: Celtis africana, Albizia gummifera, Croton macrostahyus and Millettia ferruginea. The plantation zone occupies relatively lower altitudinal elevations areas (1778-2000 m.a.s.l.) acting as a buffer zone between the College infrastructure areas and natural forest. The dominant planted tress species are Gravillea robusta, Cupressus lusitanica and Eucalyptus spp. The wooded grassland habitat covers the upland areas of the study area with rugged topography (2200-2670 m.a.s.l.) and dominated by grassland intermixed with scattered stands of tress predominantly Syzygium spp. and Dodonaea viscosa and scattered Erica scrub at the summit of mount Abaro. The agro forestry land consists of coffee plantation, undifferentiated grassland and cultivated land that occupies the lower altitudes areas (1747-1775 m.a.s.l.) of the study area. The topography of the study area is characterized by high and rugged mountainous ridges, slightly raised slopes, gently rolling steep hills and frequently intersected gorges created by streams and rivers.
The study area falls under humid montane climate having a bimodal rainfall during the long rainy season (June to October) and the second shorter rainy season (March to April). The total amount of annual rainfall varies between 700-1400 mm with an average rainfall of 1,200 mm. The mean monthly temperature is 19.5°C.
METHODS
Sampling design: Line transects on each habitat type (natural forest, plantation, wooded grassland and agro forestry land) were established clustered into three blocks. These were: Block-I (wooded grassland); Block-II (natural forest and plantation) and Block-III (agro forestry) (Fig. 2). A total of 30 transect lines were systematically established representing each habitat and used to estimate the abundance and distribution of large wild mammals. The length of each transect was 0.55 km and the width varied based on the visibility of each habitat type. As a result, in the natural forest (200 m), wooded grassland (400 m), plantation forest (400 m) and agro forestry land (400 m) transects width was used. The total length and the width of transects in each of the four habitat types were varied based on the habitat types (open or thick) and the size (area) of the of each four habitat types. Hence, records of the number of individuals of mammals of transects laid in similar habitat types were pulled together for analysis.
Data collection: Data collection was carried out in July and August, 2010 for the wet season and January-March, 2011 for the dry season. Large wild mammals identifications and recordings were carried out by direct observations aided with binoculars. Data collection was carried out walking along each transects and directly counting all the individuals of every species sighted using unaided eyes and 7x400 binoculars.
| Fig. 1: | Location map of the study area (NASA Landsat Program, 2003; Landsat ETM+ scene L71168055_05520051203.ETMGLS, 2005, SLC-Off, USGS, Sioux Falls, 03/12/2005) |
| Fig. 2: | Zonation of the study area and transects layout (10 transects in each of the three blocks) (NASA Landsat Program, 2003; Landsat ETM+scene L71168055_05520051203. ETM-GLS, 2005, SLC-Off, USGS, Sioux Falls, 03/12/2005) PF: Plantation forest, NF: Natural forest, WG: Wooded grassland, AL: Agricultural land |
Field guide of Kingdon (Kingdon, 2003) was used for identification of the mammals. While walking quietly and gently along each transects, the following data on any large mammalian individual observed was recorded; species, number, sex, distance of observations and altitude using GPS were recorded. Survey of the large wild mammals was carried out in the morning (07:00-11:00 a.m.) and late in the afternoon (03:00-05:00 p.m.). Ten days and nights were spent per session in each habitat to census the mammals. They were two sessions during per both dry and wet seasons. Hence, each transect was visited 20 days per season to estimate the abundance and distribution of large mammals. As much as possible, the observer traveled against the direction of wind to minimize disturbances (Yimer, 2008).
Indirect evidences such as fecal droppings, feeding marks, digging or territorial markings, animal parts and other tangible evidences as well as occasional direct observations were used to confirm the presence of nocturnal mammals. However, the numerical abundances of such nocturnal wild mammals were not estimated.
Data analysis: SPSS software version 15.0 (Chi-square test), Simpson’s diversity index (D) and Shannon-Weaver diversity Index (H’) were used to analyze the data. Non-parametric Chi-square test was used to compare the seasonal variations in species composition and abundance of individuals among habitats at 0.05 levels of significance. Furthermore, the age distribution (Adult, Sub-adult and Young) variations were also compared seasonally among different habitats using Chi-square test at 0.05 levels of significance. Records of number of individuals of mammals in the line transects that fall in the same habitats were also summed together and treated as a record in one habitat. The species richness and diversity of mammals in each habitat types were also computed using Simpson’s diversity index (D) and Shannon-Weaver diversity index (H).
RESULTS
A total of 19 species of large wild mammals were recorded and identified in the study area (Table 1). The most frequently sighted species were: Tragelaphus scriptus meneliki, Papio anubis, Chlorocebus aethiops and Colobus guereza (Table 1, Fig. 3-4).
Seasonal variations were observed in species composition and abundance of large wild mammals among different habitats. The highest number of species was recorded in the natural forest during the dry season (Table 1). Wooded grassland and plantation habitats had also considerably high number of species during both dry and wet seasons. Wooded grassland and plantation had similar number of species during the dry season. Agro forestry contains the least number of species (Table 1, Fig. 3-4). The highest number of individuals of large wild mammals was recorded in the natural forest habitat during the dry season and the lowest was recorded in agro forestry habitat during the same season (Table 1, Fig. 4).
| Table 1: | Seasonal abundance and distribution of large wild mammals among different habitats in the study area |
| Agf: Agro forestry; Nf: Natural forest, Pf: Plantation forest, Wg: Wooded grassland | |
| Fig. 3: | Wet season distribution of large wild mammals among different habitats in the study area (NASA Landsat Program, 2003, Landsat ETM+scene L71168055_05520051203.ETMGLS2005, SLC-Off, USGS, Sioux Falls, 03/12/2005), 1: O. afer, 2: P. anubis, 3: T. scriptus meneliki, 4: C. guereza, 5: S. grimmia, 6: P. larvatus, 7: C. crocuta, 8: C. civeta, 9: C. aureus, 10: P. pardus, 11: I. albicauda, 12: H. cristata, 13: P. capensis, 14: F. serval, 15: C. aethiopis, 16: F. silvestris, 17: P. africanus |
Seasonal variations in abundance (number) of individuals between seasons in agro forestry habitat (Wet = 81 and Dry = 40) (χ2 = 13.893, df = 1, p<0.05), natural forest (wet = 1454 and dry = 2275) (χ2 = 180.757, df = 1, p<0.05) and plantation forest (wet = 1006 and dry = 1233) (χ2 = 23.014, df = 1, p<0.05) were significantly different, but not in the wooded grassland habitat (Wet = 517 and dry = 558) (χ2 = 1.564, df =1, p>0.05) (Fig. 3, 4).
From a total of 7164 individuals of large wild mammals recorded during the study period, 4106 (57%) were observed during the dry season and 3058 (43%) during the wet season. The seasonal variation in abundance (number) of individuals of large wild mammals was significantly different (χ2 = 153.309, df = 1, p<0.05). Out of a total of 19 species recorded in the study area, 17 occurred during the dry season and 15 during the wet season (Table 1, Fig. 3-4). However, the seasonal variation in species composition of large wild mammals was not significantly different (χ2 = 0.125, df=1, p>0.05). Among all the families of species of large mammals recorded in the study area, the family Felidae outnumbered all other by comprising four species while most other families were represented by single species each (Table 1).
The highest large mammal Simpson’s Index (1-D) of diversity was obtained in the agro forestry habitat (0.769), while the wooded grassland had the lowest diversity (0.383) (Table 2). Likewise, the highest species evenness was obtained in the agro forestry habitat (0.758), but the lowest was in natural forest habitat (0.381).
| Table 2: | Diversity indices of large wild mammals in different habitats of the study area |
| Fig. 4: | Dry season distribution of large wild mammals among different habitats in the study area (NASA Landsat Program, 2003, Landsat ETM+scene L71168055_05520051203. ETMGLS2005, SLC-Off, USGS, Sioux Falls, 03/12/2005), 1: O. afer, 2: P. anubis, 3: T. scriptus meneliki, 4: C. guereza, 5: S. grimmia, 6: P. larvatus, 7: C. crocuta, 8: C. civeta, 9: C. aureus, 10: P. pardus, 11: I. albicauda, 12: H. cristata, 13: P. capensis, 14: F. serval, 15: C. aethiopis, 16: F. silvestris, 17: P. africanus |
Majority of the observed mammals species had large proportions of adult age group except the most abundant primate species of P. anubis, C. aethiopis and C. guereza that have relatively higher proportion of sub-adults and young than other species (Table 3). The number and proportion of adult, sub adult and young in the observed species were 5635 (78.6%), 383 (5.4%) and 1146 (16%), respectively (Table 3). The variation in age distribution was significantly different (χ2 = 6744.388, df = 2, p<0.05). The number and proportion of age group varied between seasons. The seasonal variation in the number of adult (Wet = 2491 and Dry = 3144) (χ2 = 75.672, df = 1, p<0.05) and sub adult (Wet = 511 and Dry = 635) (χ2 = 239.710, df = 1, p<0.05) were significantly different, but not the number of young (Wet = 40 and Dry = 3144) (χ2 = 0.551, df = 1, p>0.05).
| Table 3: | Seasonal age distribution of large wild mammals of the study area |
DISCUSSION
During the study period, a total of 19 species of large wild mammals were identified. This result can be compared with similar studies in different parts of Ethiopia that have used similar line transect techniques, if the number of species recorded is considered. For example, in the Simien Mountains National Park (SMNP), Dunbar (1978) recorded 12 species of large mammals, Bekele (1988) recorded 25 species in the Harena forest, Kasso et al. (2010) identified 21 species in Chilalo-Glama Forest Priority Area and Girma et al. (2012) recorded 18 species in Kaka and Hunkolo Fragments. Outside Ethiopia, Mian (2003), has recorded 21 mammalian species in wintering habitat of Houbara in Pakistan.
Two out of seven large endemic mammals of Ethiopia were also recorded namely Tragelaphu Tragelaphus scriptus meneliki and Tragelaphus buxtoni in the present study area. This shows how the area is important in terms of keeping unique world biota. Although, Dessie (2007) and Mamo (2007) had revealed that T. buxtoni was locally extinct, this study sighted the animal indirectly by its droppings. The illusive and shy behaviors of the animal, the small number of individuals (maximum 10) and human disturbances in the area might have contributed for not direct sighting of the animal. Moreover, the species is thought to migrate to the nearby forested area of Munesa Shashamane forest, thought to have better number of mountain nyala population, through a State Forest Plantation using as wildlife corridor, which in turn could also made sighting infrequent. So this study will act as baseline information to study activity pattern, reasons of migration and habitat preferences of T. buxtoni and some other species to the nearby forested area of Munesa Shashamane forest. The relatively higher number of T. scriptus meneliki might be attributed to the relatively thick underground cover of the forest and surrounding grazing land that is assumed to be ideal for the species to secure cover and food. Similar studies have noted that the number of this subspecies was considerably high in most highland Afro-montane remnant forests of Ethiopia with thick underground cover (Bekele, 1988; Nievergert et al., 1998; Williams, 2003; Girma et al., 2012).
Anubis baboon (P. anubis) was the most abundant species in the study area. The species is known to be widely distributed in Africa in wide variety of habitats from savanna grass land to up land Afro-montane forest. This might be attributed to the feeding behavior that it is adapted to feed on variety of food items (Johnson et al., 2012). P. anubis is also known to prefer, like many primates, habitats in an altitudinal ranges between 1500 and 2500 m.a.s.l., which is the attitude range of the study area. Similar observation was made by Dunbar and Dunbar (1974) that P. anubis distribution, was ranging from the valley bottom, found at an altitude of 1700 m.a.s.l. to the plateau which rises over the valley floor to an altitude of 2300 m.a.s.l. in Bole Valley of Ethiopia. UNESCO (2008) had also noted similar observation about the abundance of P. anubis between the altitude of 1800 and 2600 m.a.s.l. in the upland humid forest vegetation of Bonga Forest of Ethiopia. Generally, it is known that primates (particularly families of Colobidae and Cercopithecidae ) need forested areas with tall trees (Kingdon, 2003). In this regard, Wondo Genet forest is among the few, perhaps the only remnant forest in the central rift valley with tall giant tress such as Aningeria spp., Ficus spp., Millettia ferruginea (Dessie, 2007), which are inhabited by arboreal mammalian species such as C. guereza, C. aethiopis and P. anubis.
The number of individuals of large wild mammals recorded during the dry season surpassed the number recorded during the wet season. The possible explanations for this could be: high number of livestock and people were encroaching more during the wet season than dry season, particularly in the wooded grassland habitat, where the area is thought to be ideal habitat for the T. s. meneliki and P. africanus. Literatures show that, livestock and pack animals and human encroachments often reduce the foraging opportunities of the wild mammals, which in turn reduce the mammals opportunities of being sighted (Dinakaran and Anbalagan, 2007; Gundogdu, 2011; Hassani et al., 2008). In addition, human activity such as collecting of firewood and harvesting of grass were often high during the wet season, thus reducing the sighting of mammals. Furthermore, the excessive rain during wet season, which obviously boost up regeneration and growth of herbaceous and ground vegetation might have provided thick cover for the animals, which makes sighting of them difficult. Similarly, Ajibade et al. (2011), noted that excessive rain during wet season can be attributed to poor viewing of grivet monkey due to vegetation outgrowth at Kainji Lake National Park in Nigeria. The overall effect of these factors had presumably contributed to the low number of individuals of the mammals during the wet season than the dry season.
Several studies in different localities have revealed the adverse effect of livestock and pack animals encroachments and human settlement on the abundance and distribution of wild mammals, which in turn would reduced the sighting opportunities of the mammals (Mathooko and Kariuki, 2000; Maan and Chaudhry, 2001; Stephens et al., 2001; Busby, 2005; Bonnington et al., 2007; Al-Wadie, 2002; Subramanian et al., 2005; Dinakaran and Anbalagan, 2007; Yimer, 2008; Gundogdu, 2011; Hassani et al., 2008; Girma et al., 2012).
The highest species number recorded in the natural forest habitat during the dry season was presumably due to the movement of large wild mammals from the peripheral part of the study area (Wooded grassland and agro forestry land) towards the inner and relatively protected part the (natural forest) in search of food and cover. During the dry season, peripheral or marginal habitats like wooded grassland and agro forestry land were relatively dry that tends to reduce the amount of water and palatable grasses and vegetation. Hence, mammals could be forced to move towards a relatively wet vegetation (natural forest) in the center of the study area. Earlier studies in different parts of Ethiopia have revealed that species diversity is often high in areas where there are sufficient food and water sources (Ayele, 2008; UNESCO, 2008; Yimer, 2008). Similarly, studies carried out in different parts of the world have also noted that mammalian distribution and their habitat associations are often correlated mainly with the better availability of water, foraging opportunities and protection (Cloudsely-Thompson, 1969; Afolayan, 1976; Bailey, 1984; Reed, 1998; Mekonnen et al., 2011; Yaba et al., 2011). A study on the influence of seasonality on distribution and movements of large mammals at Yankari National Park has found out that food, water, cover, minerals and burning practices were the main factors affecting the movements and distribution of animals especially during the dry season (Afolayan, 1976).
The insignificant seasonal variation in abundance of individuals of mammals observed in the wooded grassland habitat might be attributed to the behaviors of the animals like feeding and movement that are displayed year round despite the season. In addition, this area is the most inaccessible area for human activities and live stock encroachments that were found to vary seasonally in other habitats of the study area. However, because of its inaccessibility these factors were found to be relatively constant in the wooded grassland habitat, it could possible contributed to the insignificant variation. Furthermore, some nocturnal and territorial mammals like leopard were noticed to stay in the same habitat despite the food fluctuations. As a result, seasonal changes in abundance for such mammalian species appeared to be minimal (Stephens et al., 2001).
Particularly, top canopy tree layer dwelling animal like Black and White Colobus Monkey (C. guereza) was mainly found in the natural forest dominated by tree species like Aningeria spp., Ficus spp. and Millettia ferruginea. This could be probably due to relatively surplus fruits and leaves and better protection provide by those tall tree species. Since the forest habitat is relatively less disturbed, it might have also contributed as a refuge for the shy behavior of the Colobus monkey. Fetene et al. (2011) has also revealed that Black and White Colobus Monkey selected riverine and large trees as their best habitat. On the contrary, Vervet monkey (C. aethiopis) was only observed in the plantation forest while it was totally absent in the natural forest of the study area. This could be mainly as a result of social behavior of the animal. C. aethiopis is generally known as highly social animals that often interact with human being (Homo sapiens). This species was observed in the human residential areas of the study area often collecting fruits from fruit trees like Persea americana in home garden areas and scavenging on discarded food leftover. Hence, because of thus food sources, in the study area the species almost abandoned their original natural forest habitat and dwell in plantation forests established surrounding the built-up areas. Similar observations were made in different parts of Africa that Vervet monkey selected secondary growth areas (Hartshorn, 1987; Chapman, 1985).
The lowest number of individuals of species like F. caracal and C. civeta could be due to their nocturnal behavior and their need for densely forested habitat, cover and caves for breeding that all could made the sighting of the animals difficult.
Even though, most of the individuals of the mammals observed were adults and sub adults, the number of young or juveniles in primates and herbivores were relatively high especially in the wet season. This indicates populations of such species promising to increase in the future. The relatively higher number of young individuals of primates was probably due to relatively abundant food supply (in the natural forest, fruit plantation and human food leftover around the human residential area), better protection and year round reproductive behavior of the species as far as there is abundant food in the study area. This is in line with different studies in different localities (Brown, 1969; Kingdon, 2003; UNESCO, 2008). The significant statistically variation in age distribution was the result of a combination of factors. Even though, they are considerably large young individual representatives in primates and other fewer species, the number of young was significantly reduced in other species. On one hand, this might have resulted from degradation of the habitat (deforestation, fire, livestock grazing and perhaps illegal hunting) that have resulted in poor cover and foraging opportunities, hence limiting their breeding success (Haule, 1997; Stephens et al., 2001; Busby, 2005; Bonnington et al., 2007). Thirdly, it might be as a result of their natural behavior, most species prefer to hide their juveniles at early age in the deep forest making the sighting of the individuals difficult, especially during the wet season where vegetation regeneration is booming (Ajibade et al., 2011).
CONCLUSION
Despite being small fragment the study area harbors diverse mammalian species including the endemics like Mountain nyala and Menelik bushbuck. The forest might also serve as main migration route/corridor for large mammals to Arsi Mountains like Munesa Forest. Different habitat types in the study area are important as preferred habitats at least to one of the species recorded in the study area. Hence, the populations of different species of mammals are distributed among different habitats of the study area. Despite the importance of the localities as home for the endangered and endemic mammals, it is not yet legalized as wildlife refuge area. At the same time livestock grazing, agriculture and human encroachments are very evident in the areas, putting an enormous strain on the flora and fauna. There is a need for urgent conservation measures to save the area.
ACKNOWLEDGMENTS
Our special gratitude goes to the Department of Wildlife and Eco-tourism, Hawassa University for providing fund and facilities. We also duly acknowledge our students of 2003 graduates for their input in the data collection process. We are also very grateful to Ato Tigneh Eshete for his help in mapping. We also appreciate the inputs and valuable comments made by reviewers.