Modelling the habitat requirements and demography of a population of roan antelope Hippotragus equinus

Abstract

The aim of this study was to model the roan-habitat relationships and the demography of the roan population in the Nylsvlei Nature Reserve. Habitat variables were measured at three sites in each of eight plant communities over 24 months. These were species composition, dry weight contribution per species, grass density, above-ground standing crop, total grass height, grass tuft height, grass canopy cover, woody plant density, woody plant canopy cover,% moisture,% N, % NDF, % ADP, % IVOMD, % OM, % Ca, % P, % Kand% Na. Species density, species diversity, veld condition and biomass concentration were calculated from the above variables. Differences in habitat variables between seasons and between plant communities were tested with general linear modelling. They were further analyzed with correspondence analysis and logistic · regression analysis to identify those variables that are most important in a plant community. Seventeen habitat parameters were isolated with logistic regression analysis of which only 11 were used for constructing seasonal Habitat Suitability Index models. These models predict the suitability of habitats and the roan densities in each plant community. Regression analysis between HSI output and breeding herd densities indicated high correlations, suggesting some validity of these models. These models need further testing with independent field data. The roan population and its distribution was monitored monthly for 24 months. A spatial segregation was found between the roan breeding herd and batchelor males. The floodplain with its high forage quality (mainly Na) and availability (high biomass concentration) is an important summer feeding area utilized by both social groups. The drying of the floodplain in autumn with a corresponding decrease in forage quality and availability forces roan out of this plant community and into more suitable areas. During winter the breeding herd selected the Aristida bipartita - Setaria sphacelata Grassland Variation probably because of the high Ca content and the large amounts of forage. The adjacent Tristachya rehmannii - Digitaria monodactyla and the Rhus leptodictya - Combretum apiculatum Variation were selected to a lesser extent. The Eragrostis pa/lens - Burkea africana Savanna was also utilized to a large extent and may serve as a calving area in winter because of the cover provided. The high density and negative rate of increase of this population indicates that they have reached carrying capacity. The demography of the roan population was analyzed with a Population Viability Analysis (PV A) to predict its persistence and to identify those population parameters that negatively influence population growth. The PV A revealed that female calf mortality is an important parameter influencing population viability. This aspect needs to be monitored and controlled to ensure population persistence. Other important parameters identified were sub-adult female mortality and calf sex ratio. The harvesting and supplementation of animals will increase the genetic variation through the exchange of genetic stock. A management plan was compiled with recommendations for improving the habitat for roan and to apply population management to ensure maximum population persistence.