Vegetation phenology dynamics as an indicator of energy and productivity functions in semi-arid savannah protected areas: a case study of Gonarezhou National Park in south-eastern Zimbabwe

Abstract

Spatial and temporal patterns of vegetation productivity in semi-arid savanna national parks are influenced by differences in land cover and changes in time series trends. The main purpose of this paper is to analyse patterns of vegetation productivity metrics of base value, peak value, amplitude, and small and large integrals in Gonarezhou National Park (GNP) in south-eastern Zimbabwe from 1981 to 2015. Three sample sites comprising shrublands, deciduous broadleaved forested woodlands and mixed cover (shrublands, broadleaved deciduous forested woodlands and grasslands) were selected to show existing patterns of vegetation productivity for GNP. We used remotely sensed Normalised Difference Vegetation Index (NDVI) data which was further processed in the TIMESAT 3.3 program to derive productivity metrics. We then tested differences in land cover using analysis of variance and changes in time-series trends using Mann–Kendall and Theil–Sen’s tests. We note significant differences in land cover (P < 0.01) in selected samples. There are significant downward trends in the base value in shrublands (P < 0.01) and broadleaved deciduous forested woodlands (P = 0.04). Significant upward trends in the amplitude in the shrublands (P < 0.01) and mixed cover areas (P = 0.01) were noted. However, there are no changes in vegetation productivity, as indicated by the peak value and large and small integral indices. Shrublands are becoming vulnerable in terms of energy and vegetation productivity and need constant monitoring. Long-span coarse-resolution images are important stepping stones in providing a baseline for further studies from moderate and fine-resolution imagery. Research on vegetation productivity using fine-resolution imagery is more suitable for GNP.

Keywords:

• shrublands
• south-eastern Zimbabwe
• Gonarezhou National Park
• vegetation productivity metrics
• land cover

ACKNOWLEDGEMENTS

We extend our gratitude to Great Zimbabwe University for allowing the use of their facilities and equipment. We thank the Zimbabwe Parks and Wildlife Authority and Frankfurt Zoological Society for providing access to the study area. Mr G. Mutowo and Mr T. Mawere provided technical advice and expertise on processing the images. Mr J. Chirima and Mr T. Makoni advised us on statistical analysis.

DATA AVAILABILITY STATEMENT

http://ecocast.arc.nasa.gov

DATA DEPOSITION

Website: http://www.uz.ac.zw/science/geography/

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors .

Additional information

Notes on contributors

Talent Murwendo

Talent Murwendo, Great Zimbabwe University, Department of Physics, Geography and Environmental Science. Wildlife management, global environmental change, geoinformation applications. Email: [email protected] or [email protected]

Amon Murwira

Amon Murwira, University of Zimbabwe, Department of Geography and Environmental Science. His main works are in earth observation, geographic information systems (GIS), remote sensing (RS) and global positioning systems (GPS) in resource detection, environmental monitoring and management. Geo-database development and applications in land, agriculture and forestry, spatial information acquisition and management, environmental systems analysis and monitoring, and vulnerability analysis. E-mail: [email protected] or [email protected]

Mhosisi Masocha

Mhosisi Masocha, University of Zimbabwe, Department of Geography and Environmental Science. His main interests are in applied geoinformation science, environmental remote sensing applications, and mathematical modelling. He has refereed for a number of journals. [email protected] or [email protected]