Research article Unexpected synchronous differentiation in Mediterranean and Canarian Cistus (Cistaceae) Beatriz Guzma ´n à , Pablo Vargas Beatriz Guzma ´n, Real Jardı ´n Bota ´nico, CSIC, Plaza de Murillo 2, 28014 Madrid, Spain article info Article history: Received 2 August 2008 Received in revised form 23 April 2009 Accepted 8 September 2009 Keywords: Diversification rates Hotspots Molecular clock Phylogeography Plastid haplotypes Radiation abstract Diversification rates of insular oceanic lineages have been historically considered to be higher than those of mainland counterparts. In order to test this hypothesis in a group of Canarian endemics (Cistus), we estimated phylogeny and conducted both network analyses of haplotypes and molecular clock estimates using plastid DNA sequences of the trnS trnG and the trnK matK. Net differentiation rates in the Canarian purple-flowered lineage parallel those of the mainland (Mediterranean) purple-flowered sister group, as revealed by a similar number of species (five vs. three), haplotypes (seven vs. eight) and haplotype clades (four vs. four). Splitting dates in the Canarian (0.33 70.14/0.28 70.16 Ma) and Mediterranean (0.66 70.18/0.38 70.19Ma) lineages reveal concordance in tempo of diversification. All these results are interpreted as evidence of two synchronous evolutionary histories given that Cistus ancestors had to cope with factors promoting differentiation in two hotspots of plant diversity: the Macaronesian and Mediterranean floristic regions. Analysis of Canarian populations suggests that a geographical rather than ecological differentiation was primarily responsible for early stages of a non- adaptive radiation of the Canarian lineage. Differentiation patterns of mainland populations appear to be more complex. & 2009 R¨ ubel Foundation, ETH Z ¨ urich. Published by Elsevier GmbH. All rights reserved. Introduction The canonical hypothesis of differentiation in oceanic islands predicts speciation associated with radiation (Carlquist, 1967; Baldwin et al., 1998), which is a rapid proliferation of species from a single ancestor that implies a great morphological and physiological divergence among lineages in response to environ- mental (adaptive radiation) or non-environmental (non-adaptive radiation) causes (Schluter, 2000). Scholars traditionally accepted in the 20th century that island lineages often diversify in relatively short periods of time compared with their continental lineages (Carlquist, 1965, 1974; Sang et al., 1994; Kim et al., 1996; Crawford and Stuessy, 1997; Francisco-Ortega et al., 1997; Baldwin et al., 1998; Grant, 1998). This may be due to a considerable habitat heterogeneity, island age, island size, isolation from source areas, number of empty ecological niches and generally lower competition compared with mainland habitats. These factors have been suggested to have contributed to adaptive radiation of remarkable insular lineages (e.g., Hawaiian silversword alliance, Baldwin and Robichaux, 1995; Echium, Garcı ´a-Maroto et al., 2009; Hawaiian lobelioids, Givnish et al., 1996; Sonchus, Kim et al., 1996; Argyranthemum, Francisco-Ortega et al., 1997). Alternative models of evolution in islands have been postu- lated in more recent years: (1) the niche pre-emption hypothesis argues that endemic groups with many species should be more effective at excluding repeat colonizations by relatives than groups represented by fewer species, because larger groups should pre-empt more of the habitat space available for colonization (Silvertown, 2004); (2) higher species diversity leads to a higher rate of diversification (Emerson and Kolm, 2005); (3) the colonization window hypothesis considers that island estab- lishment rather than continental extinction is the prime determi- nant of spatio–temporal relationships of plant groups and predicts that opportunities for colonization have been temporally con- strained to discrete waves of colonization (Carine, 2005); (4) anagenetic speciation, which implies divergence through time without further specific differentiation, appears to be significant in the origin of endemic species in most archipelagos (Stuessy et al., 2006); (5) early and recurrent colonization of islands are expected to be related to plant groups with diaspora displaying long-distance dispersal mechanisms (Vargas, 2007); and (6) the island immaturity-speciation pulse, in which the opportunities for speciation have a broadly predictable relationship to the life cycle of oceanic islands (Whittaker et al., 2007) (see also Stuessy, 2007; Whittaker et al., 2008). The characteristics of Cistus offer an ideal opportunity to test previous models of evolution of the Canarian flora. The 21 species of Cistus are frutescent and suffrutescent shrubs primarily Contents lists available at ScienceDirect journal homepage: www.elsevier.de/ppees Perspectives in Plant Ecology, Evolution and Systematics 1433-8319/$ - see front matter & 2009 R¨ ubel Foundation, ETH Z ¨ urich. Published by Elsevier GmbH. All rights reserved. doi:10.1016/j.ppees.2009.09.002 à Corresponding author. Tel.: +34 914203017; fax: +34 914200157. E-mail address: bguzmanasenjo@gmail.com (B. Guzma ´ n). Perspectives in Plant Ecology, Evolution and Systematics 12 (2010) 163–174