:rHESls Date This is to certify that the thesis entitled Factors Affecting the Local Distribution and Abundance of the Senegal Kob in the Comoe National Park, Ivory Coast presented by Jonathan H. Gilbert has been accepted towards fulfillment of the requirements for Master of Science degree in Fisheries and Wildlife 6:04 m Major professor 11-9-84 0—7 639 MS U is an Affirmative Action/Equal Opportunity Institution IV1ESI_} RETURNING MATERIALS: Place in book drop to LjBRARJES remove this checkout from n your record. FINES will _ be charged if book is returned after the date stamped below. FACTORS AFFECTING THE LOCAL DISTRIBUTION AND ABUNDANCE OF THE SENEGAL KOB IN THE COMOE NATIONAL PARK: IVORY COAST by JONATHAN H. GILBERT A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1984 ABSTRACT FACTORS AFFECTING THE LOCAL DISTRIBUTION AND ABUNDANCE OF THE SENEGAL KOB IN THE COMOE NATIONAL PARK. IVORY COAST BY JONATHAN H. GILBERT Senegal kob (5222§,§gb_522) habitat preferences. home ranges and local distribution patterns were studied from May 1981-June 1982 in the Comoe National Park. Ivory Coast. Seasonal habitat preferences for 5 Rob social catagories revealed that the ecotone between grasslands and wooded savannas was the most preferred habitat type during both seasons. Grasslands were also a preferred habitat type for all social groups. but preference for this habitat type declined during the dry season except for solitary males. Male and female mean home range areas were similar in size but dry season home range areas were significantly larger than rainy season areas. Clusters of male reproductive territories were found on short-grass savannas bordering the Comoe River during the rainy season but were temporarily abandofia for a portion of the dry season. Rainy season kob distributions were clumped around grass savannas containing male territories. Kob were more diapersed throughout the study area during the dry season. Dedicated to Judy ii ACKNOWLEDGMENTS The people who rendered invaluable assistance during this study are too numerous to name. however. I wish to thank a few individuals. Dr. Harold H. Roth provided the Opportunity to study the Senegal kob in the Ivory Coast and made available the necessary equipment and lodging. To him I owe my sincerest gratitude. Dr. George A. Petrides. my major professor. whose patience and creativeness helped guide me through the difficult analysis and manuscript preparation portions of the thesis. The Minestére des Eaux et PorSts and Direction des Parcs Nationaux kindly granted permission to use the Comoe National Park for my study. Special thanks is due to N'Guesson Michele. the Director des Parcs Nationaux for his generous donation of gasoline. Dr. Michael Muhlenberg and Bernd Steinhaur offered valuable field suggestions and provided the results of their aerial hippopotamus survey. N'Dri Koffi. Coulibally Juliette and Harouna Duattera gave both essential assistance and companionship in the field The Payne family provided me with a touch of America during my long stay in Ivory Coast. My family's continued support was much appreciated. iii Many thanks to Judy Weed for assistance in data analysis. figure preparation and typing of the manu- script. Finally a special thanks to the peOple of Bouna who accepted me as one of them. iv TABLE OF CONTENTS Page LIST OF TABLES 00.00.IIIIOOOIIIIIIIIIIIIIOOIO LIST OF FIGURES I o O O I I I O O O I I I I I O I I I I I I I I I I I I I INTRODUCTION 0 O O 0 I I I I I I I I I I O O I I I I I I 0 I I I I I I I I I CHAPTER 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I l 5 AbStraCt IIIIIIIOIIIIIIIIIIIIIIIIIIIIIIII 6 IntrOd UCtion I I I I I I I I I I I I I I I I I I I I I I I I I I I O 7 Study area description .................. 9 MethOds IIIIIIOIIIIIIIIIIIIIIIIIIIIIIIIII 1“ Results IIIIIIOIIIIIIIIIIIIIIIIIIIIIIIIII 20 Rainy Season I I I I I I I I I I I I I I I I I I I I I I I I 25 Dry season I I I I I O I I I I O I I I I I I I I I I I I I I I 27 Seasonal differences ................ 28 Discussion .............................. 30 Rainy season I I I I I I I I I I I I I I I I I I I I I I I I 30 Dry season .......................... 31 CHAPTERzIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 34 Abstract ................................ 35 IntrOdUCtion OIIIIIIIIIIIIIIIIIIIIIIIIIII 36 ”IethOds IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 39 Results IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 1+1 Home range IIIIIIIIIIIIIIIIIIIIIIIIII “1 Male reproductive territories ....... #3 Daily movements ..................... 47 Discussion .............................. 5a Territoriality ...................... 54 Home range and daily movements ...... 56 CHAPTER3IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 60 AbStraCt I I 0 0 I I I I I I I I I I I I I I I I I I I I O I I I I I I I 61 Introduction ............................ 62 raethoas I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 6“ Resu1ts I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 66 Rainy season ........................ 66 Dry season I I I I I I I I I I I I I I I I I I I I I I I I I I 66 Page CHAPTER3 (contI) IIIIIIIIIIIIIIIIIIIIIIIII DiSCUSSion IIIIIIIIIIIIIIIIIIIIIIIIIIIII 75 Rainy season ....................... 75 Dry season .IIIIIIIIIIIIIIIIIIIIIIII 76 Management implications ................ 78 Comoe National Park ................ 78 Game utilization schemes ........... 80 LIrI|ERATY~IRE CITED IIOIIIIIIIIIIIIIIIIIIIIIIII 86 vi Chapter 1. Chapter 1. Chapter LIST OF TABLES 1. Habitat preferences of the Senegal kob. Comoé National Park. Ivory Coast. Seasonal differences in areas sampled during road strip counts in the Comoé National Park. Ivory Coast. 1981-1982. Percentages of grass cover and kob visibility ratings of # habitat types in the Comoé National Park. Ivory Coast. 1981-1982. Percentage composition of herbaceous species in 4 habitat types in the Comoe National Park. Ivory Coast. 1981-1982. Numbers of kob in 5 social groups (and arcsin- transformed prOportions of these numbers) with preference ratios (PR) for # habitat types in the Comoé National Park. Ivory Coast. 1981-1982. Seasonal comparisons of preference ratings for 4 social groups in 4 habitat types. Comoe National Park. Ivory Coast. 1981-1982. 2. Home range and territoriality of the Senegal kob. Mean home range areas (standard errors) for adult males. adult females and subadult male kob. Comoe National Park. Ivory Coast. 1981-1982. 3. Local distribution and density of the Senegal Rob in the Comoe National Park. Ivory Coast. Monthly and mean seasonal density estimates for mornings and evenings in the study area of the Comoe National Park. Ivory Coast. 1981-1982. vii LIST OF FIGURES Chapter 1. Habitat preferences of the Senegal kob. 1. 2. Chapter 1. 2a. 2b. Comoé National Park. Ivory Coast. Study area in the Comoé National Park. Ivory Coast. Habitat type map of the study area showing the mosaic pattern of the different vegetative formations in the Comoe National Park. Ivory Coast. 1981-1982. Monthly average grass heights in four habitat gges i3 the Comoe National Park. Ivory Coast. 1 -19 2. 2. Home range and territoriality of the Senegal kob. Schematic of the territorial ground (TC) on the Plaine de Gansé. Comoe National Park. Ivory Coast. 1981-1982. Rainy season home ranges and daily movements for an adult male (RT 500) and an adult female (Rg 590). Comoé National Park. Ivory Coast. 19 1. Dry season home ranges and daily movements for an adult male (RT 500) and an adult female (RT 590). Comoé National Park. Ivory Coast. 1981-1982. Chapter 2. Local distribution and density of the 1a&b. Senegal kob in the Comoé National Park. Ivory Coast. Observations of kob made during morning (a) and evening (b) rainy season (June. July and August) road strip counts in the Comoé National Park. Ivory Coast. 1981-1982. viii Chapter 3. (cont.) 2a&b. 3a&b. Observations of Rob made during morning (a) and evening (b) dry season (December. January and February) road strip counts in the Comoé National Park. Ivory Coast. 1981-1982. Observations of other large mammals made during dry season (a) and rainy season (b) road strip counts in the Comoe National Park. Ivory Coast. 1981-1982. ix INTRODUCTION The Senegal kob (gppppigpp_§pp Erxleben) is the predominant large ungulate of the Comoé National Park. Ivory Coast. West Africa. They are restricted to a limited area along the major waterways and around permanent waterholes. Roth et a1. (1979) estimated that 60,000 kob inhabit the valley of the Comoe and Iringou rivers. The kob's effects on the range because of the large population size in a small area. must be significant. It has been recently noted that large areas of eroded land are common in occupied kob habitat. Roth et a1. (1979) have suggested that these eroded areas are the result of locally heavy kob grazing pressure and that the park may have reached the kob carrying capacity. The kob are a principal tourist attraction in the park. They are easily observed because of their relative abundance and are rapidly becoming accustomed to the vehicular traffic. Buechner (1971) documented kob behavior which includes ritualized combats between males in defence of their reproductive territories and a complex series of pre and post-coital displays. The kob. therefore offer the tourist an opportunity to view and photograph an animal in the wild and to observe this unique repertoire of behaviors. The government of the Ivory Coast has recognized the importance of the kob to the Comoe National Park both ecologically and as a tourist attraction. The Ministére des Eaux et Forets and the Direction des Parcs Nationaux are interested in developing a management plan for the park which would encompass both the ecology as well as the socio-economic dimensions of management. In order to do so. however. an understanding of the ecology of the major plant and animal species as well as the socio-economic impacts of tourist development is imperative. Laugenie (1975) has offered a detailed analysis of the economics surrounding the exploitation of the Comoe National Park by tourists. As of 1981. however. only preliminary studies have been undertaken on specific aspects of the park's ecology. The kob have been selected as the principal Species in a game utilization scheme proposed by the West German foreign assistance agency. Deutische Gesellschaft fur Technische Zusammerarbeit (GT2). for an area adjacent to the park (Roth and Mfihlenberg 1980). The plan there calls for three levels of game management. The first involves a controlled hunting scheme in which a few select village hunters. chosen by the village. would be licensed by the government to take only those animals which could be consumed or marketed in the immediate area. The second plan proposed is an extensive game ranch where a multiple species wildlife ranch would be maintained. POpulation growth rates would be max- imized by eliminating predators and by providing the necessary habitat requirements. POpulation surpluses would be harvested and the meat and hides marketed throughout the country. The third prOposal involves domestication of the kob on enclosed. intensively farmed ranches. These game utilization projects could help provide necessary protein missing from the diets of many Ivorians. They could do so by using land which could otherwise not be suitable for livestock production because of tse-tse fly and therefore trypanisomiasis infestations. Although some investigations into the kob's ecology have been completed (Geerling and Bokdam 1971; Muhlenberg 1978: Roth et a1. 1979: Prauser 19793 Feiler 1980; Muhlenberg and Roth 1981). many aspects essential to the successful implementation of the game utilization projects remain unknown. u I was asked by the Ivorian government and by Dr. H.H. Roth. director of the German research team to undertake a year-long study on the local distrib- ution of the kob in the Comoé’National Park. They felt that important data were lacking. particularly concerning rainy season habitat preferences and home ranges. These data were needed to supplement existing data so that a complete management plan could be formulated. In addition. habitat requirements of the kob were desireable so that planning could continue on the proposed game utilization schemes. The study objectives were: to investigate seasonal changes in habitat preferences; to calculate the rainy and dry season home range areas: to study local kob distribution patterns: and to relate seasonal variations in all of the above components to environ- mental changes. Chapter 1 HABITAT PREFERENCES OF THE SENEGAL KOB. COMOE NATIONAL PARK. IVORY COAST ABSTRACT Habitat preferences were determined for 5 social categories of Senegal kob in the Comoe National Park. from June 1981 to May 1982. Grass savannas and the grass savanna/wooded savanna ecotone were preferred during the rainy season. Dry-season use patterns indicated a shift from the grass savannas to the wooded savannas. although the ecotone remained a preferred habitat. Seasonal changes in vegetative density and forage and water availability were associ- ated with changes in habitat selection patterns. Habitat preference dffects on reproductive stategies are discussed. INTRODUCTION In West Africa. the range of the Senegal kob (gpppp k. 52p) extends from the forest/savanna tran- sition zone (Bouliére et al. 1975) to the sahel region (Puche 1976: Roster 1981). Although wooded savannas are the dominant vegetation type in much of the region. kob are mostly seen on the more-Open grass savanna types along the major streams (UNDP/FAO 1974; Sodeinde 1983). Geerling and Bokdam (1971) first suggested that these grass savannas are preferred by kob in the Comoe National Park. Ivory Coast. Roth et al. (1979) reported that different social catagories of kob in the Comoe National Park exhibited different habitat preferences. They determined that grass savannas were preferred by territorial males and harems but the use of this habitat diminished during the dry season as forage became dry. Wooded savannas. on the other hand. were selected by the non-territorial bachelor males during the dry season. As part of a further investigation into seasonal variations in the distribution of the kob in the Comoe National Park. habitat selection patterns were studiei for 12 months beginning in June 1981. The study objectives were to determine vegetative charac- teristics of 4 major habitat types occupied by the kob. to calculate habitat preference ratios for each of 5 types of kob social groups and to interpret seasonal and daily changes in these ratios. STUDY AREA DESCRIPTION The Comoe National Park is situated in northeastern Ivory Coast. West Africa. between latitudes 805' and 905' n. and longitudes 301' and 404' w. A 75 km2 study area on the east bank of the Comoé River (Figure 1) was selected for study. It contained a large kob population. stands of the major vegetation types and a road system passable the year around. Within the study area. the Comoe River is boardered by a continuous strip of narrow riverine forest (Figure 2) 5 to several hundres meters wide. Cynometra megaIOphylla. and Cpl; cordifolia dominate the tree canOpy. with the understory composed principally of Qialum guineense and Lecanodiscus cupanoides (Roth et al. 1979). Adjoining this gallery forest is a discontinuous band of grass savanna dominated by Brachiaria jubata. Vetevaria fulvibarbis and Sporobolus pyramidalis grasses (Cézar 1978). There are scattered Ierminalia spp. and Mitrggyna Spp. trees. Further from the river wooded savannas contain tree and grass Species which vary depending on location and soil morphology (Cézar 1978). Interspersed throughout are forest islands ranging in area from a few hectares to several square kilometers. Dominant tress there 10 .325 to». .33; 3.3.32 sets". 2: e. a... as...» air. / 2 0.03 .- \ . e 0054 his.“ E \~A\\.\\ N 33.; O :1 52... $2.39 3% f e A >518...- Jsis .l|.\ .300 to». $3000.— . l x .. . g , /\ / ._\../.L o ?\ loll-0’. o \\ \ ). . _o\-o/ ... a) \ ’o/ / \>. \- 11 Fig. 2. Habitat type map of the study area showing the mosaic pattern of the different vegetative formations in the Comoé National Park. Ivory Coast. 1981-1982. 13 include Celtis zenkeri. Christiana africana and ChlorOphora excelsa (Cézar 1978: Roth et al. 1979). Mean annual rainfall in the park is 1100-1600 mm. The major rainy season extends from June to October. Precipitation at other seasons is slight. METHODS Rainfall patterns during 1981 and 1982 followed the 15 year average. The 1982 rainy season. however. began early. There. May data were included in the 1982 rainy season analysis. Four habitat types frequented by the kob and described by Roth et al. (1979) were distinguished as: dense wooded savanna (30-70% tree canOpy cover). open wooded savanna (5-30% tree canOpy cover). grass savanna ((5% tree canOpy cover) and the ecotone between the grass and both wooded savannas. This edge. approx- imately 20 meters wide. was identified by the sharp increase in tree density from 10-20 trees/ha on the grass plains to 200+ trees/ha on the wooded savannas. Forests were not sampled because previous studies indicated that kob rarely use these habitat types (Geerling and Bokdam 1971: Roth et al. 1979). To quantify vegetative characteristics. seven straight-line transects were established randomly. Three were in the grass savanna. two in the open wooded savanna. and one each in the dense wooded savanna and the ecotone. Each was 500 m long with square-meter sampling stations at 25 meter intervals. Transect starting-points were marked with steel stakes. Compass headings were followed to maintain direct routes. 14 15 Transects were traversed monthly except for September. At each sampling station. average grass height was measured and the percentage of grass cover estimated. A density board (Wright 1938) was used to calculate monthly visibility ratings for kob in each habitat type. This rating was the percentage of times per month that at least one board section below 1.0 meter was visible. This figure was the mean height at the shoul- der given for the kob by Dorst and Dandelot (1980). Grass species composition was assessed by spot- plot census (Osborn 1947). On 4 transects in October 1981 when most of the grasses were in flower. a 25 m tape was strung between each of the sampling stations. The gramineae species touching the tape at each meter- point was identified. If no plant was touched by the tape. the miss was recorded but. to increase sample size. the nearest gramineae species was identified. Herbarium specimens were compared to ensure accurate field identification. Two road-strip counts of kob were established both to estimate population density and to monitor habitat selection patterns. No point on either census route was farther than 2.5 km from the river and all routes passed through habitats believed to be available to kob. Two counts per day were conducted once per month 16 on each strip. The first was between 7:00 and 11:00 am and the second between 2:00 and 6:00 pm. During the census. a Land Rover was driven at an average Speed of 20 km/hr with an observer and a driver. on opposite sides of the car. Spotting animals. As each animal was sighted the habitat type was noted and the perpendicular distance from the census line at which the animal disappeared was measured with a range finder. In cases where the animals did not move sufficiently far to disappear from view. the perpendicular distance at which it would have disappeared was estimated. All such distances were recorded as disappearing distances. The sex and age classes of each kob seen were assessed. Three classes of males (adult. subadult and immature) were distinguished by horn develOpment while the hornless females were placed either in adult or immature class based on body size (Roth et al. 1979). The sexes of calves. kob less than 6 months old. could not be determined in the field. Social catagories observed during the strip census were identified largely by the system of Petrides (unpublished). Male herds involved either single adult males seen alone or with calves or bachelor groups of more than one adult or subadult male. 17 In both cases. no adult females were present. Female herds were those which involved no adult or subadult males. although calves or immature males were sometimes present. Several adult females attended by one adult male were classed as harems. Pairs included one adult male and one adult female. For the purpose of habitat preference calculations. however. pairs were included with harems. Mixed herds were female kob with more than one adult or subadult male present. the proportion of animals in each social group as observed in each habitat type was calculated and the data pooled by season. The area sampled was calculated monthly by multiplying the strip lengths traversed in each habitat type by the mean kob disappearing distance within that habitat. That area was multiplied by the mean monthly kob visibility rating expressed as a decimal to give the actual area in which kob could be seen. This adjustment reduced the area sampled when vegetation was dense but had no effect when visibility remained good. The total area sampled was determined for each season (Table 1) and the proportion of the total calculated for each habitat type. Habitat selection was determined by use-avail- ability analysis (Neu et al. 1974). Chi-square tests 18 Table 1. Seasonal differences in areas sampleda during road strip counts in the Comoé National Park, Ivory Coast, 1981-1982. Arcsin- Area Sampled Proportion of transformed Season Habitat km ) Area Sampled Proportion Crass Savanna 17.48 0.46 0.427 Rainy Ecotone 1.12 0.03 0.100 n = 12 Open Hooded 16.58 0.43 0.410 Dense Hooded 3.04 0.08 0.161 Crass Savanna 16.50 0.39 0.385 Drv Ecotone 1.30 , 0.03 0.100 r : 12 Cpen Wooded 19.84 0.46 0.427 Dense Wooded 5.08 0.12 0.203 a. Area calculation - see text. 19 of independence were used to challenge the hypotheses that kob distribution was neither contingent upon season nor upon morning and evening census hours. Goodness-of-fit analysis determined whether or not habitat utilization occurred merely in proportion to availability. PrOportions of each habitat which were available and preportions of each social group observed in these habitats were modified by arcsin transformations (Steele and Torrie 1980) in order to reduce their variability. Preference ratios were calculated as described by Petrides (1975) though using the transformed data. Standard errors of these ratios were calculated and the Bonferroni-t statistic was used at the 95% confidence level to compare preference values between seasons (Gill 1978). RESULTS Both in the rainy and dry seasons. grass cover was least in the wooded savanna. increasing through the ecotone to the grass savannas (Table 2). The wooded savannas had a clumped distribution of grasses whereas the grasses on the grass savannas were distributed in a more regular fashion. The ecotone had more variability in grass distributions. In some places a carpet of short grass covered the ground and at other times. grasses were taller and more clumped. Although the grasses in the wooded savannas averaged taller. total soil coverage was patchy. Grass savannas had a more even distribution of grasses and maintained a high percentage of grass cover throughout the year. regardless of the mean grass height. Maximum grass growth and flowering was achieved in September/October (Figure 3) when mean heights were 30 cm. 40 cm. 78 cm and 110cm for the grass savanna. ecotone. Open-wooded savanna and dense-wooded savanna. respectively. These height differences were in part a result of grazing but also were affected by the morphology of the different Species. Exclosure results (Gilbert unpublished) indicated that grass Species untouched by the larger herbivores grew to only 75-100 cm on the grass plains. while in the wooded savannas 20 21 Table 2. Percentages of grass cover and kob visibility ratings of 4 habitat types in the Comoé National Park. Ivory Coast. 1981-1982. Rainy Season a Dry Season Habitat Type Cover Visibility Cover Visibilitya (i) (i (i (% Grass savannas 60.0 100.0 59.6 100.0 Ecotone 49.5 90.0 39.7 95.0 Open wooded savanna 34.4 60.0 18.0 79.0 Dense wooded savanna 39.0 58.0 19.0 77.0 a. Visibility of kob. see text. 22 .32183 £30 to». .fua 3:232 on00 2: 5 non: .232. 53 5 3:22. :29 3225 3.2.3: .n .3“. :2 .E< .5! do“. £2. .25 so: .30 .Eom 63¢ 32. 25.. \— eoa...» 322...:- to: u::c>u. $0100! 0231 «.u acau>un voted! some flu Scope». «\— ucsu>un anus. N.— 130004 3 as an ac on E. E5 22»: 23 they often exceeded 200 cm. AS the grass dried their heights decreased until January. when all habitat types had similar minimal mean grass heights (ca. 5 cm) following the bush fires. Kob visibility ratings on the grass savannas remained high at all seasons (Table 2). Slight seasonal reductions in visibility. though. were recorded in the ecotone and severly-reduced visibility occurred in the wooded savanna when grasses grew tall during the rainy season. The grass savanna had a relatively homogeneous grass-species composition. There. Vetivaria fulvibarpig. Braghiarig jubapg. Andropogon africanus and Sporobolus pyramidalis collectively comprised 858 of 1000 samples of the herbaceous community (Table 3). Grass species were most numerous in the wooded savannas. concuring with the results of Roth et al. (1979) who found the same species dominant on the grass savannas. They found. however. that gramineae Species' composition in the wooded savannas was much more variable than was determined in this study. The small number of ecotone (1) and wooded savanna (2) transects censused in this study perhaps were not representative of all wooded savanna cover types. Chi-square tests of independence (Table 4) indicated that habitat utilization for each herd type varied 24 Table 3. Percentage composition of herbaceous species in 4 habitat types in the Comoe National Park, Ivory Coast, 1981-1982. Grass Wooded Species Savanna Ecotone Savanna Andropogon africanus 10.2 0 0 A. ascinoides 0.1 59.7 0 A.cwmfimdfia 0 12 A. gyanus O 9.0 A. schirensis 0 0 5.2 Brachiaria jubata 31.8 4.4 0 Ctenium newtonii 0 0 0.5 Cyperaceae 4.95 2.4 0.5 Ligitaria spp. 4.65 O 0.9 Elymondra androphila 0 O 14.2 Fimbrystylus ovata 0.95 8.1 0.5 Forbaceae 1.4 0.7 5.2 Hypartynia cyanescens 0 0 3.8 h. mutlca 0 0 1.9 H. rufa O 1.7 2.8 H. smithiana O 3.0 O 1!. spp. O O 11.3 Hyperthelia dissoluta 0 0 5.2 Imperatat cylindrica 0 0.7 0.5 Loudentia arcendinacea O 0 0.5 Panicum spp. 0 2.4 23.6 Paspalum orbiculare 0 2.2 0 Schizachyrium plattyphylum 0.3 O O S. sanguineum 0 0 3.3 Sporobolus pyramidalis 8.45 7.1 1.4 Succulents 1.3 0 O Vetivaria fulvibarbis 35.4 7.4 O non-identified 0.5 0.7 8.0 Totals 100.0 100.0 100.2 25 Tabla .. Vulbor! o! kob 1n Soocul group. (and "com-unofor-c oroooruono of thou (Io-on) Huh prctcronu rouos I") (or a molt.“ noo- lo tho Co-oo' '(otlonol nu. lwrv Coat. 1901-1902. Ova—Hoodoo Moo-Hoodoo Gnu Sovom tcotono Savanna Sumo b c Annul- a Ant-o1. . Annal- . Ant-.1. . ‘1’qu ol loom“ [2 too: (or Soclal Crouoo 5“"! ‘l’loo (”Wont”) PI (vfooortlon) n (orowrzloo) Pl (proponlon) n lat-y vo 017 an '- ll Mono-ob“: *- uCHQIOI I‘lny M ll]? lo.’ 7. 0“, ,7 o”, ) .6" 10.4.35“ Hordl (.707) (.161) (-125) (.001) 119.1" PH 270 |.10 76 2.25 160 .000 0 0 200.0009 (.209) .225) (.165) (0.0) 015.12“ On 01 17) .:61 155 2.71 :71. .090 127 .20 919.7" (.291) (.271) (.10l) (.226) 0.65. PM 9‘.‘ .NS 01 2.66 102 .906 69 1.00 “0.02" (.23?) (.260) (.121) (.101) 0. 3171."..- 3.11.1“ ‘ AH 69 t J] 1 3.0.1 ‘1...“ (.556) 1.)0 (.129) 1.29 (.291) .715 (.110) .601 1.90 I... 02.02“ m o? 0 so ‘a 0. 2“ r '24 1.‘ 9 10 (O (”90) 1.27 (.151) 1.51 (.112) .270 (.162) .700 1.51 I... 59.50“ ("1 no 10 1.1 6 _. “Jr-cu ~..1'.nx .3“. 1-11 1.55 1):. 1.15 75 .120 20 .690 1102.02“ 1.00" («‘15) (I83) (00") ‘02.!” 1'! )« 1.00 112 2.51 201 .000 12 .694 ‘N-““ (.~.‘7) .251) (.161) (.001) 10.57“ r m )72 117 2‘5 29 (.105) 1.05 (.220) 2.20 (.101) .090 (.115) .‘9‘ 1.07 o... ’31-)“ "5 ¢ 10’ 2M) )6 -. . 1 \‘1 176 1..") :9 1.29 100 .010 .‘5 .700 30.2)" U11“ (-|-‘°) (.112) (.115) 56.05" PH 2.") .920 I) 1.29 292 1.11 25 .700 00.1700 (.192) (.129) (.656) (.115) 219.”“ Drv Al 79 .650 29 1.51 206 1.2. u .910 (“-07" (.251) (.151) (.517) (.130) 51.00“ m 106 1.26 105 I.” 221 I.” u .900 002.10“ .2”) (.200) (.627) (.196) E. ’(1~.¢r.‘ (Lunvs M 116 25 0 o 211.)" :wrn (.672) (.220) (0.0) (0.0) 1101.10“ H 19 0 26 (c 11.22“ 417.0“ m m m 1.11 32 1.“ m .000 n .310 12 "a “.38“ (.610) (.1“) (.157) (.l“) ' m 269 1.16 12 1.06 107 .690 22 .MO 1.1.05“ (.515) (.10A) (.291) (.129) I. Proxcrsncu ratlo v.1. \Jl('0‘.l°¢ av dlvldlng (ho orctlrtrmfol‘od proponlon ol onto-lo nun-(unsur-ml ornpmum at UM .woluolu Noon.“ (hblo 1). o( non oocul group doorvoo lo «cu kiln: 97” iv tho 5. (Palomar: In: one-(Mn: the anorunun o! "non and norm-4.17 on (not dlurlbuuom ulchlo no 6 h‘luu. c. Owl-sour. toot ualnlno coo-two loo Mouth-luau an canto-1 to (1.12:: ovollablluy. 0. lo (loo-ebony otfocu won «and. Mac. proporuooo no noun-co tattoo won coll-clot“ an. oooloi “to It- auto. .0 wool»; counts. S. ‘0 proloronco ruloo won calculatod duo to tho rolulvo lair-gm ol um Mu doomtlooo. 9 P’ .05 N P( .01 a... .‘fiot 11‘1”“:th 26 by season. Goodness of fit comparisons. furthermore. disclosed that social group distributions differed significantly from the distribution of available habitats in all seasons (Table 4). Rainy Season The ecotone was consistently preferred by all social categories at all times of the day. Preference ratios for the ecotone were highest during the afternoons for both bachelor herds and harems. Female herds and single adult males. on the other hand. showed no daily differences in their ecotone preference ratios (Table 4). Preferences for grass savannas by all social groups were evident during the morning counts. Afternoon shifts away from this habitat type occurred. however. for bachelor herds. female herds and harems. Both wooded savannas were avoided during the mornings of the rainy season. The use of the open wooded savanna increased in the afternoons. but it was not selectively chosen except by female herds. They were the only social category to show a slight preference for either wooded savanna type during the rainy season. All social groups. except single adult males. exhibited significantly different utilization patterns during mornings and evenings of the rainy season (Table 4). Mixed herds comprised less than 5% of all categories 2? seen during the rains and were not observed in 3 habitat types. Their wet-season preference ratios were not calculated. Dry Season All social groups preferred the ecotone over all other habitat types both mornings and evenings during the dry season (Table 4). Bachelor herds avoided the grass savannas while single adult males and mixed herds gave them some preference both mornings and evenings. Grass savannas were avoided by female herds in the morning but preferred by them in the afternoons. Harems used them only in pr0portion to their availability. Both wooded savanna types were used in prOportion to their availability or slightly preferred by bachelor and female herds. They were avoided by harems. single adult males and mixed herds. Neither single adult males nor harems exhibited significantly different habitat selection patterns during the mornings or evenings of the dry season. though bachelor herds. female herds and mixed herds did. 28 Seasonal Differences Significant decreases in grass savanna preference ratios were detected for bachelor herds. harems and female herds from rainy to dry season mornings (Table 5). Bachelor herds and harems were the only groups to increase their preference of the ecotone from rainy to dry season. These two herd types and the female herds showed a significant increase in morning Open- savanna preference ratings from rainy to dry seasons. Bachelor males were the only herd type to show a significant increase in dense wooded savannas from the rainy to dry season. The preference ratings for single adult males did not vary seasonally for any habitat type. Seasonal trends in habitat utilizations were less apparent during the afternoons. 29 Table 5. Seasonal comparisons of preference ratings for 4 social groups in 4 habitat types, Comoe National Park, Ivory Coast, 1981-1982. Rainy Season Dry Season Herd Type Habitat (std. error) (std. error) Bonferroni-t Bachelor Herds grass 1.56 (.303) 0.76 (.044) 16.69* AM ecotone 1.43 (.098) 2.73 (.165) 6.77* open wood 0.31 (.023) 0.89 (.042) 12.22* dense wood 0.49 (.075) 1.20 (.078) 6.52* grass 1.10 (.051) 0.75 (.058) 4.48* PM ecotone 2.25 (;183) 2.66 (.219) 1.44 open wood 0.84 (.051) 0.99 (.057) 1.32 dense wood 1.00 (.098) ---- Single Males grass 1.30 (.094) 1.27 (.080) 0.28 ecotone 1.29 (.203) 1.53 (.222) 0.80 open wood 0.72 (.067) 0.77 (.052) 0.65 dense wood 0.61 (.111) 0.70 (.106) 0.72 Harems grass 1.55 (.029) 1.05 (.032) 11.58* AM ecotone 1.75 (.100) 2.28 (.106) 3.64* open wood 0.32 (.021) 0.89 (.029) 15.92* dense wood 0.49 (.049) 0.56 (.039) 1.19 grass 1.00 (.053) 1.05 (.320) 0.81 PM ecotone 2.51 (.158) 2.28 (.106) 1.21 open wood 0.88 (.043) 0.89 (.029) 0.19 dense wood 0.49 (.061) 0.56 (.039) 0.91 Female Herds grass 1.20 (.047) 0.65 (.054) 7.68* AM ecotone 1.29 (.132) 1.53 (.172) 1.11 open wood 0.81 (.046) 1.26 (.560) 6.21* dense wood 0.70 (.078) 0.91 (.091) 1.75 grass 0.92 (.048) 0.74 (.052) 2.61 PM ecotone 1.29 (.140) 2.80 (.204) 7.60* open wood 1.11 (.051) 1.00 (.053) 1.50 dense wood 0.70 (.081) 0.96 (.089) 2.16 Standard error = JE2'u(1-P'u)/n7P'a * P<.05 DISCUSSION Rainy Season Good forage on Open areas near permanent water were found to be important in regulation Uganda kob (Egbgg k. thomasi) distributions in East Africa (Buechner 1974). Many grass savanna areas provided these conditions during the rainy season in the Comoé National Park and there. large concentrations of kob were observed with many male reproductive territories. Single adult males remained on the grass savannas during rainy-season afternoos but bachelor herds. harems and female herds moved away from these areas then. These movements may have been the result of higher temperatures which forced the animals to seek shade on sunny afternoons. The ecotone also appeared to be an important habitat type for kob during the rains. Although small in area. the animals utilized this habitat heavily. Frequent observations were make of kob resting in the shaded portions of the ecotone and of increased use of this area by bachelor herds and harems during the afternoos. Since the wooded savannas were not sought out. it seemed that kob chose the ecotone because it provided both protection from the hot sun and yet provided good visibility. 30 31 The general avoidance of wooded savannas during the rainy season seemed to be a result of unfavorable conditions created by tall coarse grasses. Nevertheless. certain areas of the wooded savannas which were grazed well into the rainy season remained occupied. In these areas grass heights had been lowered by heavy dry-season grazing pressure. That grazing maintained the grasses at an early phenological stage and evidently provided palatable forage and good visibility. By the end of the rainy season. however. grazing pressure no longer suppressed grass growth and the use of these areas diminished. 21;); Season The use of grass savannas during the dry season decreased for all social categories apparently because of decreased forage quality. Grass became parched and brown as the rains ceased. In contrast to the wooded savannas. vegetative regrowth began there only after new rains. Many observations of kob traversing the grass plains during this season were made as they traveled to and from the river. Dry-season use of the grass savannas during the dry months evidently was largely incidental. The ecotone was heavily selected for resting cover even during the dry season. This habitat component 32 was most important to the kob and yet could not exist without the adjacent grasslands. An increase in the use of open wooded savannas during the dry season was associated with the occurance of fire. It cleared the dry grass. increased visibility and stimulated vegetative regrowth even prior to the rains. The greenest regrowth appeared to be preferred for grazing and these areas were maintained into the rainy season. Mixed herds were frequently observed on recently-burned portions. evidently attracted by favorable grazing conditions. Habitat use has also been reported in the Uganda kob to vary seasonally as grass grows (Leuthold 1966) and as food availability changes (Jarmon 1974). A change in food availability apparently resulted in a reduction in the number of single reproductive territories in the East African subspecies (Leuthold 1966). Variations in habitat selection caused by changing environmental characteristics in the Comoe National Park appeared to influence kob reproductive strategy there. Concentrations of kob and reproductive activities on the grass savannas increased during the rainy season. The number of reproductive territories grew and their size became smaller (chapter 2). 33 With the onset of the dry season. the kob moved from the grass savanna into the wooded savanna. There was an associated decrease both in the number of reproductive territories and in the size of the social groupings seen on these areas. Though single adult males continued to prefer the grass savannas. the number of males defending territories declined following the rainy season. Female herds and harems also decreased their use of grass savannas then. In contrast to conditions in East Africa. where little or no seasonal variations occurred on the territorial grounds (Buechner 1966) there was a virtual cessation of reproductive activity on the open area territorial grounds during the dry season. Single territories (Leuthold 1966) were not seen to be established in the wooded savannas of the Comoe National Park following burning. Neither did harem groups form there. Yet neonates did occur in the Comoe National Park the year around. Matings. therfore. must have been continuous. It seems evident that the reproductive strategy of kob must change seasonally as habitat selection patterns changed. Chapter 2 HOME RANGE AND TERRITORIALITY OF THE SENEGAL KOB. COMOE NATIONAL PARK. IVORY COAST 34 ABSTRACT Home ranges of 14 Senegal kob (Hgbgg k. 52b) were studied from May 1981 to May 1982 in the Comoe National Park. Ivory Coast. Dry season home ranges of both sexes were significantly larger than their rainy season home ranges. There were no differences detected between male and female home range areas. When resources. eSpecially forage and water were abundant home ranges were small and bachelor groups were segregated from female herds. When forage became reduced these social groups intermixed as they moved to grazing areas. Male reproductive territories were defended on slightly elevated grassy-plains near permanent water. Territory size and retention time were inversely related to bachelor herd size and availability of forage resources. 35 INTRODUCTION In West Africa there are no quantifiable data on the home range characteristics of the Senegal kob (ggbgg‘g. £22). Home range. the area traversed in an animal's normal food gathering and reproductive activities (Burt 1943). is affected by both habitat characteristics and by the Species' resource requirements. Knowldge of the individuals' home range and its' activities within that home range can help determine the animal- habitat interactions and assess the habitat quality. The Uganda kob (figbgg g. thomasi) is reported to be philopatric with males and females remaining within 1 km of their home leks more than 90% of the time (Buechner 1971). In Nikolo-koba National Park. Senegal. Riney (1982) found that kob were loyal to their small home range. He observed a herd in poor condition which remained in an area though the grasslands there became overgrown with brush and nearby there was favorable habitat. with kob in good health. The Uganda kob is also known to defend reproductive territories (Buechner 1961: Leuthold 1966) Modha and Eltringham 1976). Buechner (1971). in the Toro Game Reserve. Uganda. observed kob leks and the well-developed repertoire of kob behavioral diaplays. 36 37 Some confusion exists. however. as to whether the Senegal kob defends it‘s reproductive territories or its harem groups for mating. Pouche (1978) and Koster (1981) both mention territory formation in Park W. Niger and Sodiende (1983). in the Kainji Lake National Park. Nigeria. reported on harem groups formed there. In Senegal. Dekeyser (1956) described harem group formation and Riney (Leuthold 1966) and Montfort (1974) reported on territorial formation. While working in the Comoé National Park. Ivory Coast. too. Montfort saw clusters of territories resembling leks. Geerling and Bokdam (1971) observed harem groups throughout their study area in the Comoe National Park but male reproductive territories were seen only on the grass savannas. Roth et al. (1979) also working in the Comoé National Park. described reproductive territories defended from May to November by dominant males. They observed that territories are marked by visual cues and that therefore. territory formation occurred primarily on the more Open grass savannas. As part of an investigation into seasonal variations in the distribution of the Senegal kob in the Comoe National Park. individual home ranges were studied. The objectives were to study the mean seasonal and annual home range areas for adult and subadult males 38 and adult females) to document adult male and female daily movement and activity patterns by season. and relate these movements to environmental characteristics; and to analyze male reproductive territories on two territorial grounds. METHODS In March and April of 1981. 14 adult and 4 subadult males and 9 adult female kob were immobilized using a "capchur" gun (Parker Chemical Co.) and 3 ml projectile syringes. A neurochemical inhibitor. R33799 (Jassen Laboratories. Belgium) was used as the immobilizing agent in conjunction with 50-100 mg of the neuroleptic drug. Azaperone. The dose used was 7.qg of R33799/kg body weight for males and 5-6,qg of R33799/kg bodyweight for females (Kupper et a1. 1982). An antidote. Cyprenorphine. was administered intravenously to revive the animals. All immobilized kob were weighed and body measurements taken (Kfipper et a1. 1982) for later study. One male and one female died during the immobilization of 27 individuals. In order to subsequently identify individual kob. colored collars were attached to 9 adult males. 4 subadult males and 7 adult females. Radio-transmitters (Dav-Tron Inc. LT 605 collars. 35.5-35.7 MHz) were placed on 4 adult males and 1 female. Due to the age of the collars. however. the transmitter batteries rapidly became depleted rendering radio telemetry impossible. The collars remained visible. however. and permitted individual identification. 39 40 Study areas were throughly searched each month for collared individuals and relocations of these marked kob provided the basis for home range determination. Sightings made during other census or observational periods also contributed useful mapped locations. After 1 year. seasonal relocations were mapped. Rainy season. dry season and annual home range limits were measured from the smallest convex polygons formed by joining the outermost seasonal relocation.points (Wittenberger 1981). Home range areas were measured for those individuals seen more than 5 times during the period concerned. Home range areas of the several age-sex groups were compared for the two seasons using the t-statistic at P I .10 probability level. In order to investigate daily movement patterns. marked individuals were followed during several day-long observation.periods. All movements and activities were documented and plotted on home range maps. The relative position of individual territories and physical characteristics of a male territorial ground were mapped. The size and number of territories were compared for two territorial areas and related to the existing environmental variables. The months that the territorial grounds were occupied and the amount of time that individual males retained their territories were also examined. RESULTS Home Range Only 14 of the original 25 marked individuals (7 adult males. 5 adult females and 2 subadult males) yielded data sufficient to determine a seasonal or annual home range. One adult female was known to have died shortly after immobilization. Five others (2 adult females. 1 adult male and 2 subadult males) were never resighted. The other 5 adult males and 1 adult female offered fewer than 5 observation and their home ranges were not plotted. The mean annual home range areas were 168.2 ha for adult males. 134.0 ha for adult females and 330.0 ha for subadult males (Table 1). Male and female annual home ranges were not significantly different from each other. Subadultrnales had significantly larger annual home ranges than adults (P<0.05). Seasonally. males and females had significantly larger dry season home ranges than rainy season home ranges. but no differences were detected between sexes for either season (Table 1). 41 42 Table 1. Mean home range areas (standard errors) for adult males. adult females and subadult male kob. Comoé National Park. Ivory Coast. May 1981 - May 1982. Age/Sex Dry Season Rainy Season Annual Adult Male 80.4 ha.“ 24.5 ha. 168.2 ha. (211.03) (16.23) (92.55) n=5 n=4 n=5 Adult Females 96.5 ha.* 37.3 ha. 134.0 ha. (“5.14) (10.7) (35.0) n=4 n=3 n=3 Subadult Males 88.0 ha. 25.0 ha. 330.0 ha. 3 n=1 n=1 n=2 * Rainy vs. dry season adult female mean home range areas. P< .10. HRainy vs. dry season adule male mean home range areas. P < .05. 8. Adult vs. subadult male mean annual hane range areas. P<. .05. 43 Male Reproductive Territories In the Comoé National Park. dominant males defend reproductive territories from April through November. Groups of these territories. termed territorial grounds (TG) (Leuthold 1966) are situated on slightly elevated. well-drained areas with good visibility near permanent water. All known TG's were located on grass savannas with seasonal streams which bordered the Comoé River. Grasslands on poorly drained moist soils did not support territorial grounds. Most TG's had mineral licks nearby but the necessity of these licks as prerequisites for territory formation could not be ascertained. TG's typically contained 5-15 circular or elliptical territories ranging from 100-200 meters in diameter. Territories on plains which had poorer grass cover and where bachelor herds were less frequently observed were larger. These larger areas were completely defended but each had an activity center approximately 10 meters in diameter on which the grasses remained short and trampled. Territories on the TG's were oriented so as to provide both drinking sites and access to the wooded savanna/plain ecotone for most territorial males (Figure 1). Territorial males which occupied territories containing water resources tolerated the occasional presence of territorial males which did not have direct access to them. 44 Fig. 1. Schematic of the territorial ground (TG) on the Plaine de Gansé. Comoé National Park. Ivory Coast. 1981-1982. (For habitat type classification see Chapter 1. Figure 2.) (+5 . O a t I I i: : O c z I: 3 - I o i. O b = c 9! N \ ' 0 N 1.2 Fig. 1. 46 Bachelor males obtained reproductive territories by diSplacing territorial males through combat. Once aquired. territories were defended mostly by ritualized fighting in the manner described by Buechner (1971). The small number of marked individuals which occupied territories prevented a calculation of rates at which occupancies turned-over. Males on the Plaine de Lola. a TC less frequented by a bachelor herd. however. retained their territories for longer periods than males on the Plaine de Gansé. a TG with a larger bachelor herd. Male G34 defended his territory on the Plaine de Lola steadily for 8 months while male RT 665 occupied 3 territories on the Plaine de Gansé for separate periods of 2 weeks. 10 days and 1 week. Leuthold (1966) described a second form of territory (single territories. ST's) in Uganda. These were larger reproductive territories situated in marginal habitat between TG's. Some evidence exists for the formation of temporary ST's in the Comoé National Park. After the bush fires adult males were observed mating in the more Open areas of the wooded savanna. However. the small number of individually identifiable kob and the short duration of the mating behavior in the wooded savanna prevented the verification of either ST defence or of harem group formation in that vegetative type. 4? Dail Movements Rainy season Before 10 am most bachelor males were concentrated in a relatively small portion of the grass savannas (Figure 1). These males regularly occupied specific areas separate from the male reproductive territories. While bachelors were tolerated for short periods of time in male territories. they were never allowed to remain. eSpecially if females were present. Grazing activity among the bachelor males increased later in the morning and they began to move towards the edge of the plain (FigureZa). The bachelors tended to move together in a long procession. By early afternoon. most of the bachelor herd had moved into the wooded savanna grazing areas a short distance from the plain. As they entered the savanna the large herd broke up into segments which regularly changed in composition. These males returned to the plains by 7-8 pm and there they passed the night. Throughout the rainy season bachelor male movements were only slightly changed. They Spent less time in the grazing areas as the grasses there grew tall and coarse. These males continued to leave the plains during the afternoons but remained in the ecotone or in the more degaged portions of the wooded savannas. When they did leave they returned earlier (4-5 pm) than previously. 48 Fig..2am Rainy season home ranges and daily move- ments for an adult male (RT 500) and an adult female (RT 590). Comoé National Park. Ivory Coast 1981. For habitat type class- ifications see Chapter 1. Figure 2. (+9 «fix fin N6 «é Fig. 2a. 50 The female herds formed less-cohesive and less permanently organized groups. These herds split and rejoined apparently in a random fashion. During the morning hours females were found grazing. mostly in the male territories on the grass savannas (Figure 21a). although the wooded savannas did provide the females with some forage. Female herds moved back and forth from the grass- lands to wooded savannas. As the day grew hot. grazing activity by females would decrease and they would move off the plains to the wooded savannas. Although some overlap of male and female rainy season home ranges was observed. activity centers. as indicated by daily movements were distinct. Dry season Bachelor males did not have easily identified bachelor grounds present during the dry season. They were attracted to the new grass stimulated by the fires and spent most of the day grazing and resting in the wooded savannas (Figure 2b). As progressively more of the wooded savannas were burned. the kob moved to take advantage of the tender grass regrowth. The females also were attracted to the green grass regrowth in the wooded savannas and they passed most of the day in this habitat type. This attraction to burned portions resulted in a greater amount of 51 Fig. 2b. Dry season home ranges and daily movements for an adult male (RT 500) and an adult female (RT 590). Comoé National Park. Ivory Coast 1981-82. For habitat class- ifications see Chapter 1. Figure 2. 52 N.- Fig. 2b. 53 overlap in male and female dry season home range activity centers (Figure 2b). These kob crossed the grass savannas frequently during this season as they visited the Comoé River to drink. These areas. therefore. were included in the dry season daily movements. DISCUSSION Territoriality Territorial grounds Reproductive behavior in the Senegal kob was found to be similar to that of the Uganda kob as described by Mokda and Eltringham (1976) but different from the lek behavior reported in other portions of East Africa by Buechner (1961) and Leuthold (1966). Male Senegal kob in the Comoe National Park held territories which were much larger than those reported in the Toro Game Reserve. Uganda (Buechner 1961) but were similar in size to the territories in Queen Elizabeth National Park. Uganda (Mokda and Eltringham 1976). Size alone cannot be used to distinguish lek territories from other reproductive territorial systems. Perhaps more important is the behavior of the female. Females are attracted to lek territories by elaborate male displays (Wittenberger 1981). In the Toro Game Reserve. female kob in estrus were attracted to male territories Specifically for mating (Buechner 1961). In the Comoé National Park. however. female Senegal kob moved through male territories in herds. apparently attracted to the forage there. The territorial males took advantage of their presence to search for estrus females. 54 55 Modha and Eltringham (1976) and Leuthold (1966) hypothesize that territory size is inversly proportional to kob density. Observations of the Senegal kob indicate that not only the size and number of territories but also the duration of territory occupation is related to the density of the bachelor herd. Where the numbers of males is large (e.g. Plaine do Gansé) there are more. smaller territories per TG and the length of stay on any one territory by an individual adult male is shortened. Besides competition between adult males. the condition of the grass forage on the TG appeared to affect territory size. It seemed that territorial males defend larger territories when percent grass cover is low (Plaine de Lola) because apparently only larger territories provided enough grass to attract females. Single territories Leuthold (1966) stated that the kob social system may be temporarily disrupted by adverse environmental conditions. He indicated that while maintaining territories on the TG. ST's may form during the dry season as the kob disperse to insure a continued high reproductive rate. In contrast. dry season use of TG's in the comoé National Park was greatly reduced and it was during this time that all observations were made 56 of reproductive behavior off the TG. Neonates were observed throughout the year in the Comoe National Park. although some peak in parturition was observed from November to March. The Senegal kob must. therefore. form ST's or other reproductive systems to maintain parturition that would temporarily be eliminated by TC abandonment. Home Range and Daily Movements Rainy season The availability of required resources appears to play a significant role in determining the home range characteristics of the Senegal kob in the Comoe National Park. During the rainy season. when most resources were abundant. kob home ranges were small. As resources. especially forage. became scarce and more dispersed during the dry season. the home ranges increased in size. It has been noted that kob require water daily. Water was scarce during the dry season but permanent water sources were abundant. enough not to limit kob distributions. Both males and females appear to be equally influenced by resource availability as indicated by the fact that their reSpective seasonal home ranges did not differ in size. The avoidance of certain habitat characteristics also was evident. As the rainy season progressed the males. in particular. spent less time in the wooded 57 savannas. The kob's neglect of this habitat type (chapter 1) may have been because the decreased visi- bility reduced their ability to detect predators. Other factors such as the large numbers of tse-tse flies present or the possible decreased palatability of the forage there may also have been influential. The avoidance of the closed forest might be associated with the dearth of grass or reduced visibility. Fear of predators was thought to be the determining factor because the few times kob did pass through the forest to the river they were in large aggregations. Wittenberger (1981) has hypothesised that antelope group together as a anti-predatory mechanism. During the height of the rainy season. social groups became more segregated when kob occupied the open grasslands. Bachelor males were tolerated to a lesser extent in the male territories. Females would infre- quently enter the bachelor grounds. Although mating must occur the year around. it seemed that territorial reproductive activity increased at that time. Whether the high numbers of kob present or social pressures re- sulting from more-concentrated reproductive animals or both caused herd types to segregate was not learned. The result. however. was that bachelor males were restricted to a small portion of the plain. These bachelor grounds then became trampled and some became seriously eroded. Reduced amount of forage apparently forced bachelors 58 to search more widely for food. deepite their avoidance of the wooded savannas at that time. Their preference then for the ecotone (chapter 1) may be a consequence of their avoidance of the wooded savannas and their inability to find sufficient forage on.the grass savannas. Females. as a result of male territoriality. had much freer access to good grazing on the plains. Dominant males defended territories in areas of good grass cover and the females regularly visited those territories. Evidently females foraged on the plains and retired to the wooded savannas to rest. When they went to the wooded savannas. however. it was to a portion separate from bachelor males. Rainy season overlap between bachelor males and female home ranges thus was limited. Dry season Kob of all types increased their use of wooded savannas during the dry season (chapter 1). This happened after the grass had become dry on the plains and fires stimulated fresh grass regrowth in the wooded savannas. Both sexes were attracted to recently burned areas so that an increased overlap of the larger dry season home ranges and a reduced herd-type segregation resulted. Innate diapersal (Riney 1982) of young and subadults are not generally considered in assessing home ranges 59 (Wittenberger 1981). Buechner (1972) and Leuthold (1966) concluded that young kob wander more than adults in order to effect an exchange of genetic material between papulational demes. Marked subadult Senegal kob also were observed to move relatively large distances before settling into a more—sedentary lifestyle in a new location. Whether this dispersal was a mechanism to reduce inbreeding or had another function was not learned. CHAPTER 3 LOCAL DISTRIBUTION AND DENSITY OF THE SENEGAL KOB IN THE COMOE NATIONAL PARK. IVORY COAST 60 ABSTRACT Seasonal density and distribution of the Senegal kob (ggbgs k. 52b) were studied between May 1981 and May 1982 in the Comoe National Park. Ivory Coast. Rainy season results revealed a highly clumped distrib- ution with kob concentrated on grass savannas containing territorial grounds. Kob diapersed into the wooded savannas during the dry season when forage became scarce. Rainy season morning density was signifcantly higher than at any other time (P 0.10). The effects of habitat preferences and home range characteristics on kob distributional patterns are discussed. 61 INTRODUCTION The Senegal kob (£222§.£3 52b) is the predominant large ungulate in the Comoe National Park. Ivory Coast. Roth et a1. (1979) estimated that 60.000 kob occupy the Open grasslands there near the rivers and around permanent water holes. Recently. large eroded portions of these grass savannas have been observed and Roth et al. (1979) suggested that kob overgrazing is the cause. The kob. because of their beauty and large numbers. are an important tourist attraction in the park. The animals are easily approached and offer the tourist an Opportunity to view and photograph. often at close range. a unique wild animal. The Direction des Parcs Nationaux of the Ivory Coast is interested in developing a long-term management plan for the park. The ecological relationships of the kob must be understood in order to formulate such a plan. In addition. kob have been selected as the principal Species for a game utilization plan (Roth et al. 1981) proposed for an area adjacent to the national park. The plan there calls for 3 levels of management: local hunting. extensive game ranching. and an inten- sively-managed domestication program. 62 63 Some studies of kob ecology in the Comoe National Park have been completed (Geerling and Bokdam 1971) Roth et al. 1979: Feiler 1981: Preuser 19793 Muhlenberg and Roth 1980). Many aspects essential to the successful implementation of the game utilization projects. however. remained unknown. At the request of the Ivory Coast government. this study was undertaken to obtain seasonal data in home ranges. habitat preferences and local distrib- utions. This paper reports on seasonal differences in kob densities and distribution from May 1981 to May 1982. and draws on knowledge of kob habitat prefer- ences and home range characteristics to explain these differences. Management implications for the national park and game utilization schemes are discussed. METHODS Two road strip routes were established on which to estimate kob p0pulation densities and assess seasonal distributions. Two counts were conducted one day each month on each strip. The first of these was made between 7:00 and 11:00 am and the second between 3:00 and 6:00 pm. During the counts. a Land Rover was driven at an average Speed of 20 km/hr with an observer and driver seated on opposite sides of the car. As each animal was sighted the perpendicular distance to the animal was measured with a range finder. In cases where more than one animal was present. the perpendicular distance to the middle of the herd was measured. A computer program LINETRAN. using 9 census est- imators as developed by Gates (1969) and modified by Koster (in press). was used to calculate monthly density estimates. The Kelker Index (Kelker 1945). was the LINETRAN estimator chosen because Robinette (1974) found that it provided an unbiased density estimate when animals were visible but not flushed and when perpendicular distances from the census line were measured. Mean seasonal estimates of kob density were obtained by averaging the monthly estimates given by the LINETRAN calculators. Density estimates were 64 65 compared with the t-statistic at the P = 0.10 prob- ability level. To study seasonal changes (in distributional patterns). morning and evening sightings for June. July and August (rainy season) and December. January and February (dry season) counts were mapped. The location of other herbivores seen during the counts were also plotted seasonally. In March. 1981. an - aerial count of the hipp0potamus (HippOpotamus amphibius) p0pu1ation in the Comoe River was undertaken and the position of individuals were mapped (Muhlenberg. personal communication). RESULTS Rainy Season Kob distribuion along the census routes was most clumped during rainy season mornings when the kob occupied grass savannas containing male territorial grounds (TG) (Figure 1a). Afternoon patterns indicated an attachment to the TG on the grass savannas but kob then also were in wooded-savanna areas surronding the TG (Figure 1b). Mean rainy-season density estimates were 78.0 kob/km2 and 52.9 kob/km2 for mornings and evenings. reSpectively (Table 1). 0mm During the dry season. fewer kob were associated with the TGS and were more dispersed throughout the study area (Figure 2a). Whereas the Ganse. Lola and Kongo'TGs were heavily occupied during the rainy season. few sightings were made there during the dry season. Contrasting the situation during the rainy season. morning and evening dry-season distributions were similar (Figures 2a and 2b). Mean dry-season density estimates were 54.55 kob/km2 for the mornings and 45.84 kob/km2 during the evenings (Table 1). Kob densities on the routes traversed were significantly higher (P<10.10) on rainy-season mornings than on rainy-season afternoons and much higher than at any time of day during the dry season. 66 67 Fig. 1a&b. Observations of kob made during morning (a) and evening (b) rainy season (June. July and August) road strip counts in the Comoé National Park. Ivory Coast. 1981-82. (T0 = territorial ground) 1 dot = indiv- idual observation) 68 TMflel. 69 Monthly and mean seasonal density estimates for mornings and evenings in the study area of the Comoe National Park. Ivory Coast. 1981-1982. Mean Season Time Date Densit Std. Error' Seasonal Density (kob/m (Std. Error) 5-81 51.39 0.299 6-81 95.54 20.270 1 M 7-81 95.32 9.530 78.70 ab 8-81 60.23 0.756 (9.48) 5-82 91.00 10.000 Rainy 5-81 29.69 2.320 6-81 61.39 6.100 1 PM 7-81 74.98 16.440 52.92 a 8-81 60.28 10.470 (8.23) 5-82 38.25 0.717 11-81 32.12 1.170 1-82 49.31 0.709 54.55 b AM 2-82 43.47 1.100 (7.73) 3—82 70.25 4.690 4-82 48.38 4.190 Dry 12-81 33.81 2.170 1-82 53.01 0.288 43.81 PM 2—82 51 . 54 9. 190 (4 . 25) 3-82 44.51 12.450 4-82 51.76 0.420 1. Means fgllowed by Similar letters are Significantly different P 1.5 km). Some marked adult males moved from one TG to another at this time. No male. however. was observed to hold territories on more than one TG. Male G36.was observed to hold a territroy on Plaine de la Boucle during the rains and later was with a bachelor herd near the Plaine de Gansé. 3 km away. but his relationships and the reason for his movement was not clear. During the dry season. lone adult males still showed a preference for grass savannas. Many tended to 76 77 remain there. though less territorial defense was observed. All other social groups reduced their util- ization of the grasslands and increased their occupation of the wooded savannas. Grass regrowth in the wooded savannas seemed to have attracted this increased utilization. Kob preference for the ecotone also remained high. There they could find forage and good visibil- ity. During the dry season. there was increased home range size. a reduced concentration on open grass savannas and a significantly lower density over the study area as a whole (Table 1). In this connection. it must be remembered that the study area was selected primarily because of the large kob population occuring there. Density estimates for that area did not represent that of the park as a whole. though they remain useful in making future comparisons. 78 Management Implications Comoé National Park National parks are established to preserve natural conditions for both flora and fauna. Management. therefore. should be undertaken only to restore natural conditions when those conditions are threatened. Territorial male kob in the Como; National Park require Open grasslands with grasses of short stature for reproductive territory formation during the rainy season. These grass savannas are also preferred then by all other herd types as well. At this time kob apparently avoid the dense vegetation of wooded savannas. It is not known whether it is fire..soil conditions or some other factor which is responsible for the absence of tree Species on the plains. If soil morphology is the controlling element. grassland vegetation should persist regardless of occasional fires. If it is fire that suppresses tree regeneration. however. then grasslands there may not be maintained in the absence of occasional burnings. “If it is decided that burning is necessary to preserve natural conditions. then fires should not be so frequent that they deplete the grass cover and expose the landscape to erosion. Fire probably is an integral component of all African grassland ecosystems. Most. if not all. savannas there are thought to be a fire-maintained climax. 79 Fire. therefore. could be used as a method of management which does not change the natural conditions and facilitates game-viewing. Alternate areas of wooded savannas could be burned in order to maintain a fresh supply of new growth throughout the dry season. Fire control procedures must be established and local research on the needs for and the effects of different fire policies would be desirable. Overgrazing on some grass savannas was observed and possibly could become serious. Evidences were seen. however. that kob have evolved a mechanism to mitigate this potential problem. Social groups segregate on the plains. Territorial males and females. the reproductive components of the p0pulation. occupy the areas of highest percentage grass cover. Bachelor males frequent the poorest grasslands on the plains and supplement their forage by maintaining grazing areas in the wooded savannas. This apparently assures that the reproducing animals have the best resources available to them. Buechner (1971) hypothesized that territorial grounds were a spacing mechanism which prevented overutilization of the grass resources. This mechanism apparently distributed population domes throughout the occupied range and reduced the liklihood of locally- heavy overgrazing. In the Comog National Park. Ivory Coast. however. grasslands are distributed in a mosaic 80 pattern throughout the kob range and the animals are associated with these grasslands. The findings of this study indicates that social group segregation within the demes tends to prevent over-exploitation of the forage resources on these plains. Though it is reported that population demes in Uganda distribute themselves to reduce overgrazing. kob in Ivory Coast seem to segregate by social group within the demo to achieve the same results. _§gmg_Utilization Schemes Each of the three game utilization schemes planned for the Comoe region. local hunting. game ranch and intensive domestication. requires a different form of management. Local hunting. the least intensive plan. does not require habitat management. The pr0posed management will concentrate on hunter regulation and harveSt rates. The habitat requirements of kob. as determined during this study are not applicable to hunter/harvest management. This form of management. therefore. will not be discussed. The most intensive game utilization pr0posal. domestication. may not be feasible with our present understanding of kob ecology. Kob evidently have a complex reproductive system which is sensitive to environmental conditions. Any interference with this system through forced enclosures or over-crowding 81 could disrupt matings. A more detailed investigation into the reproductive ecology of the kob plus test confinements of mature animals are needed before intensive planning for domestication are warranted. The results of this study could be useful. however. for the establishment of the third pr0posed scheme. an extensive game ranch. The goal of any ranch is to maximize reproduction and to minimize losses through disease. predation or otherwise. A principal reason for ranching wild ungulates in Africa is that they are naturally immune. or at leaSt.reSistant.‘to many of the diseases prevalent throughout the savanna regions there. As indicated. kob have a complex mating system. It is essential to provide for this system. if a successful game ranch is to be established. The central feature of the kob reproductive strategy is the territorial grounds. These are located on slightly- elevated. well-drained short-grass savannas. near permanent water. Providing this habitat type. however. does not necessarily insure the formation of a territorial ground. These grounds are traditional in nature. Members of p0pulation subunits return to their home T0 for mating. Once a TG is established. and perhaps eSpecially if the game ranch has several TG's on it initially. then kob should remain in the immediate area. 82 Kob reproduction. however. is adaptable. It was found by Leuthold (1966) and again in this study that if environmental conditions temporarily disrupt activity on the TG kob nevertheless will adjust the reproductive strategy to maintain matings. Yet it must be emphasized that the TC is the central component of the kob mating System. Although they are the site of territory formation and a preferred habitat type. grass savannas are not the only habitat required by the Senegal kob. Grass savannas in the region are limited in size. possibly because of edaphic factors. and apparently do not provide resources in adequate amounts to sustain large kob populations. Forage often is limited on these plains and most segments of the kob population there rely on wooded savannas. at least to some extent. for grazing. The ecotone between the grass and wooded savannas is a highly-preferred habitat type. This edge cannot exist without the proper juxtapositioning of grass and wooded savannas and. therefore. any ranching scheme must include areas where this edge is abundant. While many small grassy Openings increase the amount of edge. they might reduce the chances of territorial ground formation. Ecotone must be provided while maintaining grass savannas large enough to support 83 one or more territorial grounds. Kob cannot go long periods without drinking. It is necessary. therefore. to establish or maintain permanent drinking sites. Since most grass savannas occur near permanent streams. water management may be unnecessary. In the Comoe National Park. isolated pockets of kob have been observed far from the river but around permanent water holes (i.e. Mare de Kakpin. Mare de Gadipiére). It may be possible. therefore. to increase potential kob habitats by creating watering sites. Additional resource and habitat requirements of these p0pulations especially with regard to mating systems must be assessed. Buechner (1971) has provided a detailed analysis of harvesting strategies for the Uganda kob in relation to their ecology. Some modification of these strategies may be possible for the Senegal kob because of differ- ences in their ecology. The Uganda kob maintains an extremely high reproductive rate of 1.25 calves/adult female/year. Buechner suggests that this is a result of continuous reproductive activity on the TG and of a relatively short post-partum anestrus period. The Senegal kob of the Comoé National Park apparently do not maintain this continuous high birth rate. Roth et a1. (1979) documented a peak in parturition from November through March which resulted from a 84 peak in TG reproductive activity between March and September. Since the number of calves/adult female/year is less in Ivory Coast than in Uganda. the percentage of animals harvested perhaps also must be less. Buechner (1971) also found that a large portion of the bachelor males belonged to the reproductive portion of the p0pu1ation. He observed a rapid turn- over rate of dominant males on the TG. The territorial male kob of the Comoé National Park. however. seems to retain their territories for longer periods than in Uganda and fewer males may contribute to matings. Bachelor male dry-season reproductive roles in Ivory Coast wooded savannas is unknown. If it is assumed that dominant males are mating in the wooded savannas as well as on the TG. however. then perhaps the re- productive role of the bachelors is minimal. In that case. more bachelor males could be harvested than Buechner felt desirable in Uganda. Buechner believed that few females should be har- vested and only 101 of the bachelors could be culled without adversly impacting the p0pu1ation. It is recommended in Ivory Coast. however. that while few females should be taken. nevertheless. a larger percentage of bachelors (ca. 15%) probably safely could be har- vested. If a density of kob could be maintained which 85 approaches the density found in this study area of the Comog National Park. then it is believed that 3-5 kob/ka/year could be harvested annually on a sustained basis. LITERATURE C ITED 86 LITERATURE CITED Bouliere. F.. E. Minner. and R. Vautlaux. 1974. Les grandes mammiferes de la region de Lamto. Cote d' Ivoire. Mammalia 38(3): 433-437. Buechner. H.K. 1961. Territorial behavior in Uganda kob. Science 133(3454): 698-699. -------- 1964. Behavioral adaptions of mammals to environments. Territoriality as a behavioral adaption to environments in Uganda kob. Proc. 16th Inter. Cong. 2001. 3: 59-64. -------- 1971. Implications of social behavior in the management of Uganda kob. Presented Inter. Symp. on the Behav. of Ungulates and its relation to management. Univer. of Calgary. Alberta. Canada. Buechner. H.K.. R. Schloeth. 1965. Ceremonial mating behavior in Uganda kob (Adenota kob thomasi Neumann). Zeit Tierpsychol. 22: 209-225. Burt. W.H. 1943. Territoriality and home range concepts as applied to mammals. J. Mammology. 34: 346-352. Cezar. J. 1978. Vegetation. flore et valeur pastorale des savannes du Parc National de la Comoé. Publ. Centre de Reserches Zootechniques de Minankro. Bouake. Ministere de la Reserche Scientifique. Cote d' Ivoire. Dekeyser. P.L. 1956. Le Parc National du Nikolo-Koba. III. Mammifere Memoires IFAN. No. 48. Dakar. Senegal. Dorst. J.. and P. Dandelot. 1980. A field guide to the larger mammals of Africa. Houghton-Mifflin Co. Boston. 28? ppl. Feiler. W. 1981. Study bf the seasonal modifications of habitat preferences and social structures of the kob antilope in the Comoe National Park. Ivory Coast. Diploma. University of Wurzburg. Gates. C.E.. 1967. Simulation study of estimators for the transect method. Biometrics 25: 317-328. 87 Geerling. C. and J. Bokdam. 1971. The Senegal kob. Adenota kob kob (Erxleban) in the Comoe National Park. Ivory Coast. Mammalia (Paris) 35(1): 17-24. Gill. J.L. 1978. Design and analysis of experiments in the animal and medical sciences. Iowa Univer. Press. Ames. 409 pp. Jarman. P.T. 1974. The social organization of antilope in rgéation to their ecology. Behavior 48(3-4): 215- 7o Kelker. G.H. 1945. Measurements and interpretation of forces that determine populations of managed deer. PhD. thesis. Univer. of Mich. Ann Arbor. 422 pp. Koster. S. 1981. Survey of the vegetative and ungulate p0pulations in Park W. Niger. M.S. thesis. Mich. State Univer. East Lansing. Kupper. Von W.. M. Wolters and J. Gilbert. 1982. Immobilisierung von freilebenden Schwartzfu moor- antilOpen. Praktische Tierarzt. 11: 994-996. Leuthold. W. 1966. Behavior of Uganda kob. Behavior 28: 214-257. Monfort. N. 1974. Quelque examples de structure social chez les Ongules Africains. Zoologie et Assistance Technique. Ed. J.C. Ruwet. FULREAC. Univer. of Liege. France. Muhlenberg. M. and H. Roth. 1981. Récapitulation des resultats des Travaux réalizés ce jour sur les anti10pes Cobes de Buffon dans le Parc National de la Comoé. FGU Kronberg GMBH. Kronberg. West Germany. Neu. G.W.. R. Beyers and J.M. Peek. 1974. A technique for analysis of utilization-availability data. J. Wildl. Manage. 38: 541-545. Osborn. B. 194?. Determining range utilization by frequency tallies. J. Soil Water Conserv.. 2(1): 51-55. Petrides. G.Au-and Wendell G. Swank. 1965. POpulation densities and the Range-carrying capacity for large mammals in Queen Elizabeth National Park. Uganda. Zoologica Africans 1(1): 209-225. 88 Petrides. G.A. 1975. Principal foods versus preferred foods and their relation to stocking rate and range condition. Biol. Conserv. 7: 161-169. Preuser. N. 1979. Study of food and habitat preference of the Senegal kob in the Comoe National Park. Ivory Coast. Diploma University of Heidelberg. West Germany. Puche. R.M. 1976. Seasonal distribution and reproduction in Artiodactyla from Southwestern Niger. Nigerian Riney. T. 1982. Study and movements of large mammals. John Wiley and Sons. NY. RObinet‘te. WoCo' CeCo LOVEIGSS. Dvo Jones. 1974. Field Techniques of strip census methods. J. Wildl. Manage. 38 (4). 81-96. Roth. H.H.. M. Mfihlenberg. P. Roben and W. Bartlott. 1979. Etat actuel des Parcs Nationaux de la Comoé et de Tai ainsi que de la Reserve d'Azagny et propositions visant a leur conservation et leur developpement aux fins de promotion du tourisme. FGU-KRONBERG GMBH. Vol. IV. Kronberg. West Germany. Roth. H.H.. M. Muhlenberg and P. Goedicke. 1981. Rapport Special sur les possibilities d'utilization rationelle des populations d'animaux sauvages dans les zone de savanne de Cote d'Ivoire en vue de la production controlée de viande degiber. FGU Kronberg GMBH. Kronberg. West Germany. Soleinde. F. 1983. Habitat preferences and aSpects of general ecology of the Senegal kob in the Kainji Lake National Park. Nigeria. M.S. thesis. Mich. State Univers. East Lansing. MI 74 pp. Steel. R.G. and J.H. Torrie. 1980. Principles and procedures of statistics: A biometric approach. McGraw-Hill Book Co. New York. 633 pp. UNDP/FAO. 1974. Kainji Lake Research Project. An ecological survey of the Borgu Game Reserve. FI: SF/NIR 24 Tech. Report 24. Wight. R.M. 1938. Field and laboratory techniques in wildlife management. Univer. Mich. Press. Ann Arbor. MI. 107 pp. Wittenberger. J.F. 1981. Animal social behavior. Duxbury Press. Boston. 722 pp. ll11|WI)111(1))!(11111111111111(1)1111