Savannas at the forest boundary: an understudied biome at risk

Abstract

Savannas receive a fraction of the policy and research attention of forests and are therefore some of the least known biomes in the world. Fundamental to the protection of these unique systems, and the billion people who rely on them globally, is to gain a new understanding of both their evolutionary history, which has structured diversity within them, and to categorise their present-day diversity. This thesis addresses the historical assembly of savannas globally using the widespread and ecologically variable genus Ziziphus as an exemplar, and provides new data on the flora of poorly studied savannas in Asia.

Chapter two investigates how functional traits enable Ziziphus to occupy several distinct biomes across the tropics. To do this I collated and examined trait and distributional data. I found that the unusually wide distribution of Ziziphus can be explained by its diversity of habits which enable occupation of different biomes spanning continents. Results showed that liana species are restricted to closed forests and the geoxylic habit is found only in open savannas. For Ziziphus species, habit facilitates range size, with trees having larger ranges than shrubs and lianas. Surprisingly, biome is not correlated with range size. Cultivated species have ranges ~10 times that of non-cultivated tree species and with significantly different and broader environmental niches. A diversity of habits, along with the usage of fruits of many Ziziphus species by people, expands the range and environmental occupation of the genus.

Chapter three explores whether geography or ecology structure Ziziphus’ phylogeny to gain a new understanding of the history of the assembly of the savanna biome globally. To do this I created a phylogenetic tree based on chloroplast and nuclear ribosomal internal transcribed spacer (ITS) sequences. I found that at a global scale geography (i.e., the barrier of dispersing amongst continents) structures Ziziphus’ diversification more than ecology (i.e., switches amongst major biomes). Based upon dating the ITS phylogeny, Ziziphus originated in closed biomes in Asia during the Oligocene and subsequently long distance dispersal from Asia to Africa resulted in the novel occupation of two biomes, the tropical savanna biome and the desertic biome. In keeping with the younger age of the savanna biome and the Pliocene radiation of geoxylic species, the youngest species in Ziziphus are found in open biomes, and the most recently evolved habit is geoxylic. Leaf hair is likely to be a key innovation enabling occupancy of drier environments, while occupation of closed biomes is associated with the development of the liana habit.

Chapter four investigates the floristic composition and resilience of ground layer vegetation in Southeast Asian savannas via new floristic surveys in 50 sites across four savanna types. I found that the heterogeneity of Southeast Asian savannas means that systematic botanical survey work that incorporates the diversity of savannas is necessary and urgent. Total plant species richness decreased under increasing tree cover, although patterns differed between functional groups. Crucially, research and management should not conflate monocot and dicot herbs due to their divergent functional responses to shade, as non-graminoid monocots are the most affected by increasing tree cover. Further, Southeast Asian savannas support distinct grass assemblages, with implications for assessments of species diversity and for future management of these ecosystems.

This thesis expands our understanding of the processes that drive diversification and trait evolution in savannas, and characterises the understory floristic diversity of little-known savanna systems in Southeast Asia. Further work is necessary to assess whether trait and evolutionary patterns seen in Ziziphus across the major tropical and subtropical biomes are replicated in other, unrelated and more species-rich genera, and to characterise the non-graminoid communities of Southeast Asian savannas to determine whether they represent distinct associations compared to the dominant woody flora.