Poster Presentation: An expert system for bush encroachment in Namibia

Publication Year:

An online expert system for rangeland management was developed for semiarid savannas in central Namibia, with particular emphasis on encroachment by Acacia mellifera subsp. detinens. The Expert System consists of 21 questions and 22 decisions. Decisions are divided into three types: adaptive, reactive and ongoing good management. Adaptive decisions are related to seasons of high rainfall, as this is a critical window both in terms of a hazard (transition to bushy thickened state if no fire is applied) and opportunity (transition back to an open savanna if fire is applied). Reactive decisions (or symptom treatment) refer to ways in which existing bush thicket problems occur. The Expert System should be applied to different camps, and used three times a year (hot wet season, cold dry season and cold wet season). The questions require mostly quualitative "high"/"medium"/"low" and "yes"/"no" answers. This is because absolute thresholds are complex and not yet determined, and farmers often have their own perception of what constitutes "bush encroachment". The Expert System uses wiki technology (as is used for wikipedia, for example), which allows a high level of interaction between user and administrator. It has embedded links to explanations, photographs and additional information. An additional booklet has also been developed, because computer and internet access is still mostly limited to established commercial farmers, and internet bandwidth is limited. Both the online and the booklet versions have an easy-to-follow one page decision tree for easy reference in the field. The decision pathway is represented as in a dichotomous key.

Meeting Name:
45th Annual Congress of the Grassland Society of Southern Africa, Control of Invasive Woody Species. Tuesday, 20 July 2010
Item Type:
vegetation level, plant species, meteorological research, research policy, research centre, nuclear research centre, research project, tree, vegetation type, vegetation cover, vegetation, market research, research of the effects, environmental research, ecosystem