Vegetation phenology patterns in semi-arid savannah woodlands of Gonarezhou National Park, Southeastern Zimbabwe

Abstract

Vegetation phenology (VP) patterns of semi-arid savannah woodlands ecosystems are essential for sustainable management and conservation since they are indicators of the health status of protected national parks. However, monitoring VP changes are intermittently carried out in semi-arid savannah woodlands ecosystems despite the links to ecosystem function, status, health and service. The paper analyses VP metric patterns, during the start of the season (SOS), maximum vegetation (PEAK), end of the season (EOS) and length of the growing season (LGS) trends from 1981 to 2015. This study was carried out in Gonarezhou National Park (GNP) in Southeastern Zimbabwe. Shrublands, broadleaved deciduous forested woodlands and mixed cover samples were used to depict existing VP patterns. Derivation of VP using remotely sensed Normalised Difference Vegetation Index (NDVI) data was done using the TIMESAT 3.3 programme. Analysis of variance (ANOVA) was used to test the land cover differences while Mann-Kendall (M-K) and Sen's slope tests were used to analyse time-series trends. Due to differences among the different land cover types in GNP, there are spatial variations in phenology, with the SOS, PEAK and EOS indicating later dates significantly. VP spatial and temporal patterns vary markedly as a result of differences in land cover. Further research in patterns of VP using fine and coarse satellite imagery in GNP is required. VP studies need to be linked to distribution and abundance of large mammalian herbivore species.