Recent assembly of the Cerrado, a neotropical plantdiversity hotspot, by in situ evolution of adaptationsto fireMarcelo F. Simona,b,1, Rosaura Gretherc, Luciano P. de Queirozd, Cynthia Skemae,2, R. Toby Penningtone,and Colin E. HughesaaDepartment of Plant Sciences, University of Oxford, South Parks Rd, Oxford, OX1 3RB, United Kingdom;bEmbrapa Recursos Gene´ ticos e Biotecnologia,PqEB, Caixa Postal 02372, 70770–917, Brasília-DF, BrazilcDepartamento de Biología, Universidad Auto´ noma Metropolitana-Iztapalapa, Apdo. Postal 55–535,09340 Me´ xico, D.F. Me´ xico;dDepartamento de Cieˆ ncias Biolo´ gicas, Universidade Estadual de Feira de Santana, Km 03, BR 116, Campus, 44031–460, Feira deSantana-BA, Brazil; andeRoyal Botanic Garden Edinburgh, 20a Inverleith Row, Edinburgh, EH3 5LR, United KingdomEdited by Mary Arroyo, University of Chile, Santiago, Chile, and approved October 12, 2009 (received for review March 27, 2009)The relative importance of local ecological and larger-scale histor-ical processes in causing differences in species richness across theglobe remains keenly debated. To gain insight into these ques-tions, we investigated the assembly of plant diversity in theCerrado in South America, the world’s most species-rich tropicalsavanna. Time-calibrated phylogenies suggest that Cerrado lin-eages started to diversify less than 10 Mya, with most lineagesdiversifying at 4 Mya or less, coinciding with the rise to dominanceof flammable C4 grasses and expansion of the savanna biomeworldwide. These plant phylogenies show that Cerrado lineagesare strongly associated with adaptations to fire and have sistergroups in largely fire-free nearby wet forest, seasonally dry forest,subtropical grassland, or wetland vegetation. These findings implythat the Cerrado formed in situ via recent and frequent adaptiveshifts to resist fire, rather than via dispersal of lineages alreadyadapted to fire. The location of the Cerrado surrounded by adiverse array of species-rich biomes, and the apparently modestadaptive barrier posed by fire, are likely to have contributed to itsstriking species richness. These findings add to growing evidencethat the origins and historical assembly of species-rich biomes havebeen idiosyncratic, driven in large part by unique features ofregional- and continental-scale geohistory and that different his-torical processes can lead to similar levels of modern speciesrichness.The uneven distribution of species diversity across the globeand the occurrence of biodiversity hotspots with high con-centrations of species are well established (1–3). However, theunderlying causes of differences in species diversity and c om-position among biomes and the processes that have promptedac cumulation of high species diversity in some areas remainpoorly understood (4–6). Correlations bet ween species richnessand annual energy input, water supply, and physiog raphic com-plexity suggest that climatic and other environment al factors aremajor determinants of species diversity (7). Indeed, it has beenshown that these factors can accurately predict the locations ofmost global diversit y hotspots and acc ount for the latitudinalgradient of species richness (7). However, such insights c ontrib-ute little to our understanding of the historical assembly ofspecies-rich biomes and the larger-scale evolutionary processesthat have generated global patterns of diversity (4, 8, 9). Littleis known about the historical and geographical assembly ofspecies-rich biomes (5), in terms of when, how quickly, and fromwhere the species and lineages that make up different biomeshave been recruited and how they subsequently evolved in situ.This lack of data on geographical and temporal patterns ofspecies diversification, especially in the tropics where mostdiversit y resides, makes it difficult to assess why there are somany species in these areas, to what extent variation in diversitycan be attributed to regional and continental-scale historicalc ontingencies (4, 10, 11), and to compare patterns amongdif ferent species-rich biomes (9, 12).Recent discussion has highlighted the potential role of phy-logenetic niche conservatism in shaping regional species pools(8, 13) and explaining diversit y gradients (11). Prominent ex-amples of large-scale n iche or biome conservatism have beendocumented for the tropical–temperate biotas (11), mangroves(14), southern hemisphere extratropical biomes (13), A ndeangrasslands (8, 15), and seasonally dry tropical forests (16–18).However, the extent to which this tendency to retain ancestralec ology across lineages has influenced species composition inthe most species-rich tropical biomes is unknown.Recent insights into the historical assembly of species diversityand biomes have c ome f rom time-calibrated phylogenies forbiome-specific lineages (5, 9, 19–23). Phylogenies have potentialfor reconstructing transitions from precursor to modern biotasand identifying the underlying factors that drive these processes(4, 11, 24). However, the sparse sampling of lineages and speciesin such studies (5, 24) has limited these insights to a fewwell-studied areas such as the Cape Floristic Region (12, 20).The Cerrado. To address these questions, we have investigatedplant diversification in the Cerrado of South America by usinga comparative phylogenetic approach. The Cerrado is a floris-tically diverse savanna that covers more than two million km2ofCentral Brazil and parts of Boliv ia and Paraguay (Fig. 1). Thisregion has been recognized as a global biodiversit y hotspot withmore than 10,000 plant species, of which 44% are endemic (1, 25,26). Despite the floristic and global c onservation import ance ofthe Cerrado, little is known about the origins and diversificationof its flora. Indeed, hypotheses about Cerrado origins range fromthe early Cretaceous (27), wherein Cerrado lineages were sug-gested to be possible precursors of the adjacent Amazonian andAtlantic rain forests, to the Holocene (28). Here we evaluatethese alternatives, but we particularly focus on the hypothesisthat the origin of the biome coincided w ith the rise to dominanceof flammable C4 grasses within the past ten million years (6, 29).In common with other tropical savannas, the Cerrado is domi-Author contributions: M.F.S. and C.E.H. designed research; M.F.S., R.G., L.P.d.Q., C.S., R.T.P.,and C.E.H. performed research; M.F.S., C.S., and C.E.H. analyzed data; and M.F.S., R.G.,L.P.d.Q., C.S., R.T.P., and C.E.H.