Crop raiding pattern of the savanna elephant (Loxodonta africana) and its association with some key habitat variables in the Red Volta Valley of North-Eastern Ghana

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
This study investigated the degree to which crop-raiding by elephants in localities of the Red Volta Valley is a function of the density/diversity of their natural browse, extent of degradation of their habitat, their local abundance, size of crop field, and the proximity of crop enclaves to the forest reserve. Farm enumerators estimated the size of crop fields and the area damaged by elephants, and graded the growth and quality of affected crops. Incidents were plotted along a 12 month time series, and the location of fields was recorded with a GPS. Quadrats (50mx5m) were assessed for the density, diversity, and distribution of elephant browse. Quadrat area clear-felled, damaged by fire, and mined was estimated to measure habitat degradation. Elephant dung-piles within 1,000m x 10m transects were assessed for the abundance of elephants. Interviews and periodic monitoring of transects along the Ghana-Togo and the Ghana-Burkina Faso frontiers were used to determine trans-frontier movement of elephants. The relationship between crop raiding, and the density and diversity of browse, extent of habitat degradation, size of crop fields, local abundance of elephants, and distance of affected crop enclaves from the forest reserve for each locality was determined using correlation and regression analyses. A functional relationship between crop raiding and habitat variables was determined using stepwise regression. Majority (56%) of raided fields suffered damage to < 21% of the field, and 2-3% of farmers were affected each year. There was no marked variation in the rate and extent of crop damage among locales of the Red Volta Valley (Friedman test: d.f=6, P=0.128). Incidents peaked in October, with about 72% occurring in the harvest season. About 16% of woody species in the Red Volta Valley were categorized as potential elephant browse. Second order jackknife and Michaelis-Menten asymptotic estimators showed that the density and diversity of woody stems was near optimum. About 99% of the vegetation area sampled was burned, and 0.35ha iv of vegetation was clear-felled. The area holds a small population of elephants that seasonally migrate between Ghana and Burkina Faso. The number of elephants per locale ranged between 0 and 3. A significant inverse association between crop raiding and distance of affected crop enclaves from the forest reserve was observed. But the association between crop raiding and density and diversity of browse, and the extent of degradation was not significant. A Stepwise regression model defined the relationship between rate of incident and proximity of fields to the forest as: Y = 25.105 + 3.2 – 9.73X (Y= rate of incident and X= distance of field to the forest boundary). Contrarily to speculations, crop-raiding in the study area is not influenced by the status of browse, extent of degradation, or by the number of farmers (or size of land area cultivated). Thus mitigation measures should include relocating farms away from the forest reserves, while those aimed at reducing the density of farms, or replanting of degraded patches of the forest should be de-emphasized as they will not necessarily reduce crop raiding.
A Thesis submitted to the Department of Wildlife and Range Management of the Kwame Nkrumah University of Science and Technology in partial fulfilment of the requirement for the degree of Master of Philosophy