How is biodiversity in savanna landscapes structured and effectively conserved?

Globally, we are experiencing a crisis of biodiversity loss. Increasing human dominance of the Earth has led to widespread land-use change, overexploitation of species and natural resources, and climatic change all of which have caused precipitous declines in the abundance and richness of species. Understanding how biodiversity is structured across regions is vital to maintain ecosystem resilience and to identify effective conservation action. Savannas are widespread covering around 20% of the Earth’s surface, playing a vital role in global carbon cycles, and harbouring some of the most endangered and charismatic megafauna. Despite their global importance, patterns of biodiversity and effective conservation actions in savanna landscapes remain understudied in comparison to analyses in moist tropical forests. This thesis investigates broad biodiversity patterns and effective conservation actions in savanna landscapes of Africa. In Chapter two, I use plot-level data from across the savannas of eastern and southern Africa to assess how the alpha, beta, and phylogenetic diversity of savanna tree communities respond to environmental and disturbance gradients. I find that wetter savannas contain more species rich and phylogenetically diverse communities than drier savannas, whilst high taxonomic turnover across the region is largely driven by precipitation regimes. Chapter three built on this by assessing how these same gradients affect the functional composition of tree communities in the region. I collated a functional trait matrix for tree species from a range of sources and used clustering analyses to group species into functional groups. Comparing mean trait values between clusters, I identified three distinct functional syndromes adopted by savanna species to cope with water-limitation and disturbance by fire and herbivory. The relative abundance of these clusters changes in response to environmental and disturbance gradients suggesting that turnover in functional composition of communities is what drives previously identified taxonomic turnover. In combination, chapters two and three demonstrate the importance of climate and disturbance regimes in dictating tree community composition in savanna systems, highlighting the importance of regional studies and plot-level data to identify these patterns. In Chapter four, I assessed the potential for carbon-based payments for ecosystem services to protect habitats on low-intensity farmland in Ghana. I used farm economic data in combination with tree inventory data to show that protection of farmland trees can be achieved at very low carbon breakeven prices (BEP), ranging from US$2.49 - US$6.45 t-1 CO2. Reintroduction and extension of fallowing periods can also be achieved at competitive BEP, US$4.67—US$15.45 t-1 CO2, when combined with tree protection. This chapter proposes a novel approach to conservation within savanna landscapes that moves beyond traditional conservation attempts that aim to conserve pristine habitats or return farmland to woodland.