Three decades of woodland cover change in Hwedza, Zimbabwe
View/ Open
Kerry Smith_Dissertation.docx (20.87Mb)
Date
31/08/2021Author
Smith, Kerry
Metadata
Abstract
Land cover is critical to people’s livelihoods and sustainable development strategies. In Zimbabwe, loss of natural vegetation threatens sustainable management, but the cause of this change is disputed. I used remote sensing and interview data to investigate the patterns, causes and implications of land cover change in Hwedza, Zimbabwe between 1990 and 2020. I focused on changes in woodland cover, due to the importance and limited availability of the provisioning services they provide. I used Landsat data to generate a timeseries of land cover and applied a Hidden Markov Chain Model (HMM) to reduce erroneous change. Key informant interviews were used to gather information on the processes and implications of change. I found that in the last three decades, land cover transitioned from a trajectory of net woodland loss to net woodland gain. Exogeneous socio-economic factors, such as increase in off-farm employment, were more important than endogenous socio-ecological forces in explaining this change, so sustainable management strategies which include socioeconomic objectives are likely to be most successful in tackling woodland cover loss. This agrees with growing global evidence that woodland regrowth often occurs due to innovation and opportunity rather than resource scarcity, highlighting the importance of demand-side measures in natural resource management. Integration of remote sensing and interview data allows the complementary nature of these data streams to be utilised and evaluated. In doing so, it is possible to challenge the narrative that farm resettlement has led to deforestation, and provide a wholistic understanding of land cover change in Hwedza.