Classification and mapping of the woody vegetation of Gonarezhou National Park, Zimbabwe

Abstract

Within the framework of the Great Limpopo Transfrontier Conservation Area (GLTFCA), the purpose of this study was to produce a classification of the woody vegetation of the Gonarezhou National Park, Zimbabwe, and a map of its potential distribution. Cover-abundance data of woody species were collected in 330 georeferenced relevés across the Park. These data were used to produce two matrices: the first one using the cover-abundance values as collected in five height layers and the second one based on merging the layers into a single cover value for each species. Automatic classifications were produced for both matrices to determine the optimal number of vegetation types. The two classification approaches both produced 14 types belonging to three macro-groups: mopane, miombo and alluvial woodlands. The results of the two classifications were compared looking at the constant, dominant and diagnostic species of each type. The classification based on separate layers was considered more effective and retained. A high-resolution map of the potential distribution of vegetation types for the whole study area was produced using Random Forest. In the model, the relationship between bioclimatic and topographic variables, known to be correlated to vegetation types, and the classified relevés was used. Identified vegetation types were compared with those of other national parks within the GLTFCA, and an evaluation of the main threats and pressures was conducted.

Conservation implications: Vegetation classification and mapping are useful tools for multiple purposes including: surveying and monitoring plant and animal populations, communities and their habitats, and development of management and conservation strategies. Filling the knowledge gap for the Gonarezhou National Park provides a basis for standardised and homogeneous vegetation classification and mapping for the entire Great Limpopo Transfrontier Conservation Area.