Elephant utilization of the acacia tortilis woodlands in Lake Manyara National Park Tanzania

A study of elephant habitat utilization was carried out in the A. tortilis woodland of lake Manyara National Park during the dry seasons of August to December 1981 and July 1982. The study involved the survey of the vegetation, studies of habitat selection within different density stands of mature A. tortilis and food species selectivity. The A. tortilis woodland was subdivided into four strata based on the density of mature A. tortilis trees. Quantitative and qualitative characteristics of the vegetation were determined by use of the Point Centre Quarter method and several plot sampling techniques. Of the different strata the high density stratum had 21.6 A. tortilis trees/ha, low density 7.8 trees/ha, very low density 2.3 trees/ha. The mixed woodland stratum around Bagayo river had a mixture of tall trees and shrubs. It had a density of 8.2 A. tortilis trees/ha. Tree density in the different strata varies inversely with density of A. tortilis trees. The density of mature A. tortilis tree is still declining. However, the number of damaged and killed trees is very small compared to observations of the past 16 years. The mortality rate of the A. tortilis trees has been maintained at a constant of around 6 per cent annum. The very low density stratum has the lowest cover or shade value of all the and river banks have spreading mats of low grasses dominated by cynodon plectostachyus. A model showing how to predict the total population size of A. tortilis as well as the density of each size class on several consecutive years has been developed. The inputs in the model are the previous years population size, the size specific annual mortality rate, the annual growth rate per size class and the proportion of the total population that is the regenerating class for the next year. Models of the natural phenomena are difficult to perfect because of the many unpredictable factors that influence the systems. Elephant selection of habitat in the A. tortilis woodland was investigated by use of point sample technique. Selection is determined by time of the day, areas with abundant shade are favoured during hot hours of the day otherwise areas with riverine vegetation as well as lake shore grasses are the ones selected for feeding purposes. When data are averaged over all time categories the mixed woodland stratum in significantly selected for compared to the A. tortilis strata. Assessment of day time activity pattern used the point sample technique as well as the continuous observation method and it was observed that morning hours are used mainly for travelling from the escarpment through different habitats of the woodland, with elephant feeding as they walk. The major activity at noon is resting. In the evening elephants feed heavily, mainly on the lake shore grasses before they start moving up the escarpment in the late hours of the day. Elephant feeding selectivity was also investigated by both point sample technique (first feed method) and the continuous observation method. The main diet constituent is grass even in the dry season. Elephants show a big preference for A. tortilis leaves branches and fruits in all strata. Among the shrubs Maeruatriphylla, Acalyphafruticosa, Cordiasinensis and C. oyvalis are eaten more often than others. Osmium suae though found in plenty is not fed on as much. Certain species including Barleriaernthemoides, Calotropisprocera, Justiciacordata and Veprisuguenensis are avoided. Typical of bulk feeders, elephants feed on a wide variety of plant species but this does not mean that they are indiscriminate feeders. This study has concluded that at present there is enough food resources to support the elephant population. The reduction of A. tortilis tree mortality and the prolific regeneration suggests that the future of the A. tortilis woodland is not as bleak as had been indicated by earlier studies. However, expansion of the Park boundaries is desirable and continued monitoring programmes are required as indispensable information to success of future research and Park management.