Journal of Biogeography Insight into an island radiation: the Tarentola geckos of the Cape Verde archipelago ForJournal: Journal Peer of Biogeography Review Manuscript ID: draft Manuscript Type: Special Issue Date Submitted by the Author: Complete List of Authors: Vasconcelos, Raquel; CIBIO-UP Centro de Investigação em Biodiversidade e Recursos Genéticos, ICETA Carranza, Salvador; Institue of Evolutionary Biology (CSIC-UPF), Animal Phylogeny and Systematics Harris, D.J.; CIBIO-UP Centro de Investigação em Biodiversidade e Recursos Genéticos, ICETA Tarentola, Cape Verde Islands, cytochrome b, 12S rRNA, Key Words: Phylogeny, Biogeography, Island radiation Page 1 of 30 Journal of Biogeography 1 1 2 3 4 Original article 5 6 7 8 Insight into an island radiation: the Tarentola geckos of the 9 10 Cape Verde archipelago. 11 12 13 R. Vasconcelos, CIBIO-UP, Centro de Investigação em Biodiversidade e Recursos Genéticos, Campus Agrário de Vairão, R. 14 Padre Armando Quintas, 4485-661 Vairão, Portugal; Faculdade de Ciências da Universidade do Porto Pr. Gomes Teixeira, 15 4099-002 Porto, Portugal. E-mail: [email protected] 16 17 S. Carranza, Institute of Evolutionary Biology (CSIC-UPF) CMIMA - Passeig Marítim de la Barceloneta, 37-49, E-08003 18 Barcelona, España. E-mail: [email protected] 19 D.J. Harris, CIBIO-UP, Centro de Investigação em Biodiversidade e Recursos Genéticos, Campus Agrário de Vairão, R. Padre 20 Armando Quintas, 4485-661For Vairão, Portugal. Peer Faculdade de CiênciasReview da Universidade do Porto Pr. Gomes Teixeira, 4099- 21 002 Porto, Portugal. E-mail: [email protected] 22 23 24 Corresponding author: 25 Salvador Carranza, Institute of Evolutionary Biology (CSIC-UPF) CMIMA - Passeig Marítim de la Barceloneta, 37-49, E-08003 26 Barcelona, España. E-mail: [email protected] 27 28 29 Running head: Cape Verdian Tarentola island radiation 30 31 ABSTRACT 32 Aim To reassess relationships of Tarentola from the Cape Verde Islands By including specimens from all 33 34 islands, including distinct suBspecies never previously analysed genetically. To adequately determine 35 for the first time variation within forms By sequencing over 400 new specimens, decreasing the 36 37 possiBility of any cryptic form Being overlooked and resolving some of the issues raised in earlier works. 38 Extensive within-island sampling was also used to shed light on distriButions and to make comparisons 39 with age and geological features of the islands including size, altitude and haBitat availaBility to explain 40 41 genetic diversity. 42 Location An oceanic archipelago situated in the Atlantic Ocean Belonging to the Macaronesian 43 44 Biogeographic region and located around 500 km off the coast of Senegal: the Cape Verde Islands 45 (latitudes 14° – 17° N and longitudes 22° – 25° W). 46 47 Methods A total of 405 specimens of Tarentola were collected in 276 localities from nine islands and 48 three islets within three groups with very different geological histories, topography, climate and 49 haBitats. Mitochondrial cytochrome b gene and 12S rRNA partial sequences were oBtained and analysed 50 51 using Bayesian and Maximum Likelihood methods and networks to infer molecular diversity, 52 demographic features and phylogeographic patterns. Correlations with geological and ecological 53 54 features of the islands were performed to relate these variaBles with genetic variaBility. 55 Results The total picture of the phylogenetic relationships of the Cape Verdian Tarentola was presented 56 57 for the first time, the positions of new forms revealed and some rearrangements of previously 58 hypothesised relationships. Contrary to what was expected and despite the large sample size, low 59 60 intraspecific diversity was found using the cyt b gene fragment. Star-like haplotype networks and statistical tests suggested the past occurrence of a rapid demographic and geographical expansion over Journal of Biogeography Page 2 of 30 2 1 2 3 most of the islands examined. Size, altitude and haBitat availaBility of the islands are positively 4 5 correlated with genetic variaBility . 6 Main Conclusions Biogeographic patterns have in general a high concordance with phylogenetic Breaks 7 and with the three eco-geographical island groups. Volcanism, haBitat availaBility, intimately linked with 8 9 orography and present and historical size of the islands appear to Be the main factors explaining the 10 genetic diversity of Tarentola from the Cape Verdes. 11 12 Keywords 13 Tarentola , Cape Verde Islands, cyt b, 12S, phylogeny, biogeography, island radiation 14 15 16 17 18 19 20 For Peer Review 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 3 of 30 Journal of Biogeography 3 1 2 3 INTRODUCTION 4 Islands are ideal model systems to study evolution and phylogeography, and through time have Become 5 6 known as “natural laBoratories” for this purpose (Williamson, 1981). Oceanic archipelagos, such as the 7 Galapagos, Hawaii, Madeira and Canary islands, permit testing of various evolutionary hypotheses (Losos & 8 9 Ricklefs, 2009). Gene flow Between the island and the mainland or among islands may Be practically non- 10 existent, allowing fixation of genetic variation and so differentiation of populations can occur By geographical 11 isolation. The sea depth isolates each island into a discrete evolutionary unit (Emerson, 2002). Once the ages 12 13 of the islands are known, phylogeography of taxa in archipelagos can Be analysed within a known timeframe. 14 Additionally, island populations have higher risk of extinction than mainland populations (Frankham, 1997); 15 even though they generally present a depauperate numBer of species relative to similar mainland areas, the 16 17 numBer of endemic species is usually very high (Whittaker, 1998), as is the risk of introduction of exotic 18 species (World Conservation Monitoring Centre, 1992). As a result, investigation and protection of endemic 19 island species is particularly important. 20 For Peer Review 21 The Cape Verde Islands are an oceanic archipelago located approximately 500-800 km from the West 22 African coast. The archipelago is volcanic (the last eruption occurred on Fogo in 1995), and has never Been 23 connected to the mainland (Mitchell-Thomé, 1976). It has ten main islands, plus several islets, that are 24 25 politically separated in Windward and Leeward Islands But eco- and topologically divided in north-western, 26 eastern and southern islands (Fig. 1). They are arranged in a horse-shoe shape with its concavity facing 27 28 westwards. The islands are Between 6 to 26 million years old (My) and the youngest ones are on the tips of 29 the arc (see Griffiths et al. , 1975; Grunau et al. , 1975; Stillman et al. , 1982; Mitchell et al. , 1983; Carracedo, 30 1999; Torres et al. , 2002; Plesner et al. , 2002). During sea level fluctuations in the Pleistocene (1.6 My) some 31 32 of the islands of the north-western group (S. Vicente, Sta. Luzia, Branco and Raso islets) were most likely 33 linked, and possiBly also Boavista with Maio. Elsewhere the water channels are very deep so it is highly 34 improBaBle that the other islands were ever connected By land (Morris, 1998). Their sizes and topographies 35 36 vary dramatically; Santiago is the largest (around 1,000 km 2) and Raso islet (<6 km 2) among the smallest, 37 Fogo is the highest (near 2,800 m) and Santa Maria islet the flattest. 38 The fossil record on volcanic islands is generally scarce, thus in most cases no direct evidence of past 39 40 evolutionary events is availaBle. However, organisms may show evidence of past population decline and 41 growth in their genetic variation (Beheregaray et al. , 2003). In this way reptiles can Be considered model 42 organisms as they are non-volant, decreasing migration importance in the evolutionary process, are easy to 43 44 handle in the field and are present in all the Cape Verde islands. 45 Although Darwin, during his voyage on the Beagle, considered the Cape Verde islands to Be “utterly 46 47 sterile” (Darwin, 1845), he noticed the lizards in the more humid valleys. There are approximately 27 48 currently recognized native lizard taxa, all endemic to the Cape Verdes (100% endemicity), which are divided 49 into three genera: the skinks Mabuya and the geckos Tarentola and Hemidactylus . The genus Tarentola is 50 51 Biogeographically interesting Because it arrived to the Cape Verde islands approximately seven million years 52 ago, from a propagule that dispersed from the western Canary Islands situated 1500 Km to the north 53 (Carranza et al. , 2000). The endemic Tarentola have Been previously studied phylogenetically (Carranza et al. , 54 55 2000, 2002; Jesus et al. , 2002) and estimates of relationships indicated possiBle cryptic species and paraphyly 56 of some species. However, in all these studies not all the islands of the archipelago where included and 57 therefore not all taxa. Moreover, those studies were Based on a very small numBer of samples per taxa and 58 59 island, studying only the deep divergent lineages and not assessing intraspecific variation. Such information is 60 valuaBle, as Tarentola on the Canary Islands show consideraBle intraspecific variation, possiBly associated with island sizes, volcanic activity, ecological niche availaBility, or a comBination of these factors (González et Journal of Biogeography Page 4 of 30 4 1 2 3 al. , 1996; Juan et al ., 2000; GüBitz et al. , 2005). Assessing intraspecific variation may uncover additional 4 cryptic lineages, and is highly relevant for any conservation assessment (Schwartz et al. , 2006; Kahindo et al. , 5 6 2007). 7 8 9 The aim of this study was to reassess relationships of Tarentola from the Cape Verde islands By including 10 specimens from all islands, including distinct suBspecies never previously analysed genetically.