The Structure and Dynamics of an Insularized Large Herbivore Community in Lake Nakuru National Park, Kenya
Lake Nakuru National Park was fenced all round in 1987, making it an ecological island. Its isolation started earlier this century when man began utilizing the catchment for urban development, commercial farming and livestock rearing. The isolation has resulted in loss of migration opportunities, low predation pressure and increased herbivore density. Having been so modified by man, the park is now in dire need for active management which requires that accurate information be readily available on all aspects of its ecology. This study aimed at understanding the structure of large herbivore populations and their interactions with grazing resources. Field work began in March 1991 involving a series of 12 monthly and six bimonthly surveys along a predetermined road transect network. The method produces reliable density estimates, when strip width is reasonably large. During the surveys, data were gathered on numbers, age structure, sex composition and the spatial distribution of large herbivore species. The quality and quantity of herbaceous plants were measured on eight study plots after every two months throughout the study. The effects of grazing on species composition and standing crop accumulation were documented.on five vegetation exclosures. ; Seventeen large herbivore species were found along tfltntansects during the study. Overall herbivore density differed significantly among species and.vegetation types, but not among ~ seasons. Of nine species encountered in all surveys, no seasonal differences were found in the density of buffalo iSvncetus caffer), Grant's gazelle (Gazella grant!), impala (Aepyceros melampus), Thomson's gazelle (Gazella thomsoni!), Defassa waterbuck (Kobus el/ipsiprymnus defassa), warthog r(Phacochoerus aethiopicus) and zebra (Equus burchelii. Only the eland (Taurotragus oryx) had density estimates differing significantly among seasons. Differences were significant among habitats for all species, except the buffalo and eland. Differences in age and sex composition were not significant among seasons for the giraffe, Grant's gazelle and zebra. They were, however, significant for all other species. Detrended Correspondence Analysis (DCA) showed clear seasonal patterns of association between the species, whereby big bodied, bulk and mixed feeders, namely the buffalo, eland, Grant's gazelle, rhinoceros and zebra were grouped into one large and distinct cluster. Two predominantly grazer species with a preference for short grass, the Thomson's gazelle and warthog, fell into a smaller but nevertheless distinct cluster. This clustering was more diffuse for dry than wet season data, suggesting greater habitat selectivity and increased ecological separation of the herbivore species during dry periods, when availability of vital resources is more localized. This finding was further supported by DCA based on vegetation types, which placed open habitats together in the dry seasons, as was the case for wooded or bushed habitats. Data for waterbuck fell in neither of the clusters, possibly because, being the most numerous large herbivore species in the park, it competitively displaced other species. Herbaceous plant standing crop differed significantly between wet and dry seasons, and among vegetation types. The differences were more pronounced on grazed than ungrazed plots. Standing crop was correlated with rainfall accumulated over the eight days preceding sampling, indicating the need for proper timing of field measurements. It was highest after the long rains in 1991, thereafter decreasing to a minimum in the middle of a dry season in February 1992. The analysis of nutritive value of 23 most common grass and sedge species indicated clear seasonal fluctuations in percent protein content, with some species falling below the theoretical minimum required for maintenance byanimals during the dry season. The ;,;- temporal fluctuations in availability and nutritive value of food are profoundly manifested in 'periodic losses of body condition among the park's large~erbivores. Herbaceous standing crop inside vegetation exclosures remained consistently higher than outside after the first few months of sampling, fluctuating with two annual cycles. Relative abundances of green and dry material differed over time inside all the exclosures, these being the same throughout the study outside three of the five exclosures. Ratios of different plant parts differed with time both inside and outside all the exclosures. Species composition snowed only rnarqinal changes, with the most common grass species, among them Themeda triandra and Hyparrhenia lintoni, slightly declining in importance over the two years of field work. This study discusses the implications of ecological isolation to the interactions between plant and animal species and puts forward some recommendations for the management of wildlife resources in the park. This study recommends a prescribed burn management programme, emphasizing the removal of each season's production to stimulate fresh growth. It suggests that some woody habitats be opened up to enable the park continue accommodating its remarkably large grazer community, alongside an expanding black rhinoceros population. Depending on habitat attributes and management priorities, it is recommended that the populations of constituent herbivore species be manipulated to fit within inherent environmental constraints.