Socio-ecological dynamics of woodlands, forests and fire in southeast Angola

Angola, a country that endured a 13-year war for independence followed by a 27-year civil war, remains one of the most understudied countries in Africa. This study develops in southeast Angola, in the watershed of the Okavango Delta and the Zambezi River, an area that has been deemed one of the largest areas of undisturbed miombo in Africa. The highlands of southeast Angola, one of the country’s most isolated and still heavily land-mined regions, contains well-preserved vegetation with scattered human populations that depend on natural resources. Because of its ecological richness and low population density, the region has been targeted for conservation interventions but the potential impact on livelihoods has not been investigated. Using a socio-ecological systems framework, this research investigates how human practices, knowledge systems, and historical events interact with ecological patterns and processes such as vegetation types, structure, floristics, and fire regimes, in a setting where human-environment relationships are primarily governed by customary practices rather than state governance. To achieve this, I begin this thesis by presenting an overview of the wooded ecosystems-livelihoods context, including savanna and miombo ecology, conservation approaches and ethical considerations during this study (Chapter 1). I then present three chapters with the main questions and results from my research. For these, I executed three main tasks, each linked to one chapter. First, I mapped vegetation types according to local classifications (Chapter 2), then I typified the vegetation bridging local classifications and Western scientific floristic approaches (Chapter 3) and lastly, I studied the effects of the war on fire dynamics in the area (Chapter 4). I end this thesis with the contributions that emerge overall across the chapters. In Chapter 2, I address the challenges of solely using remote sensing for classifying vegetation, highlighting the potential oversimplification of complex landscapes when following this approach. This chapter incorporates participatory mapping in five rural villages to recognise local knowledge to classify woody ecosystems. My findings suggest that incorporating place-based knowledge can offer a deeper understanding of the landscape. The results reveal differences between local classifications and remote sensing ones, emphasizing the need for inclusive conservation approaches that align with local ecological knowledge. I argue that landscapes are complex socio-ecological systems and that their understanding can benefit from ground-based approaches that consider not only the physical features but also the social and cultural values of the land. Chapter 3 focuses on woodland ecology, and challenges existing classifications of southeast Angola as a homogeneous miombo woodland. Through quantitative ecological surveys (establishment of 26 plots) and learnings from local knowledge, the study identifies five structurally and floristically distinct woodland formations, including a unique dry forest dominated by Cryptosepalum exfoliatum pseudotaxus. The latter differs greatly from miombo woodlands, as circumscribed in the literature. These findings call for a re-evaluation of current maps of Angolan vegetation, prompting the need for updated maps of African evergreen dry forests that are currently identified only in western Zambia. I argue that an accurate identification of the diverse vegetation formations in Angola is fundamental for their proper management, given that fire is widely used in miombo ecosystems but is likely not appropriate in evergreen forests. Chapter 4 explores the impact of the Angolan civil war on fire and forest dynamics in southeast Angola, since local peoples indicate a decrease in fire events due to war, evident in satellite-derived time series. This chapter employs remote sensing, historical data, and interviews to reconstruct war-time (1985-2002) and post-war (2003-2018) fire dynamics, revealing the role of social dynamics in shaping fire regimes. The study challenges the idea that ecological and climatic factors are the main determinants of fire regimes, and highlights the intricate relationship between fire, war, and socio-ecological systems, showing that war can restrict the use of fire and change the fire dynamics of an entire ecosystem. It also advocates for interdisciplinary approaches to understand the complex socio-ecological dynamics of war. Overall, this thesis highlights the importance of recognising local knowledge, challenging existing landscape classifications, and considering historical and socio-cultural contexts in understanding and managing vegetation and associated ecosystems in the highlands of southeast Angola. The findings emphasize the need for inclusive and multi-disciplinary approaches to vegetation classification, conservation and landscape management. Lastly, this study calls for ethical and reflective research practices that centre local voices and challenge dominant conservation narratives that often overlook lived experiences and place-based knowledge.