Authors

Source

Type

Peer Review

Audience

Pages

Journal Number

Notes

Abstract. This paper discusses the relationship of vegetation phenology and three climate variables (rainfall, minimum and maximum temperature) at 17 savanna sites in Botswana and Zambia. Interactions among climate variables were responsible for the largest variation in vegetation canopy phenology measured by the Normalized Difference Vegetation Index (NDVI) and tree shoot extension from ground observations. The most important determinant of savanna phenology in southern Africa was the interaction betweenminimum and maximum temperature. This observation indicates the need to incorporate minimum and Maximum temperature in modelling the impact of climate change on savanna vegetation. Cluster analysis, using five NDVI metrics, separated the study sitesinto four savanna types with distinct tree:grass ratios. The semi-arid sites were characterized by a co-dominance of the tree and grass components while at more mesic sites, the significance of the grass component was lowest on sites with kalahari sand and mopane woodland. The use of NDVI metrics applied in the study provides an additional technique. that explicitly takes into account the tree:grass ratio during the growing season, for the classification of savanna vegetation types at local scale.