On the origin of southern African subtropical thicket vegetation

The origin and affinities of southern African subtropical thicket have been misunderstood and neglected. This formation was only recognised as a biome distinct from savanna and karoo in the mid 1990s. One hypothesis states that it is a young vegetation type, assembled from forest, savanna and karoo elements after Holocene climatic amelioration. Others have suggested an ancient history for thicket. Here we review fossil and phylogenetic data in order to provide a better assessment of the origins of thicket. Albeit patchy, the fossil data are suggestive of a Palaeogene origin for this formation. A review of molecular phylogenetic data of extant thicket lineages indicated three major patterns: (i) ancient Cretaceous elements, including Encephalartos and the Strelitziaceae, (ii) basally branching lineages - many of which dominate contemporary thicket - that evolved in the Eocene (e.g. in the Celastraceae, Sapindaceae, Didiereaceae, Crassulaceae: Cotyledonoideae), and (iii) lineages derived from adjacent biomes that diversified in thicket in association with Neogene climatic deterioration (e.g. Aizoaceae, Asteraceae). We provide a narrative account of the evolution of thicket, which concludes that it is an ancient formation, extending back at least to the Eocene and derived initially from elements in the forest formations that prevailed in Upper Cretaceous and early Palaeogene times. As a biome, thicket is not uniquely southern African, being part of a formation that was globally widespread in the Eocene and which is extant in many parts of the world. Future research on the origins of thicket should focus on providing dates for major dichotomies as a complement to the rapid emergence of molecular phylogenies, as well as data on the genetic variation in populations of taxa categorised as ancient or young, and widespread or range-restricted.