How the African house gecko (Hemidactylus mabouia) conquered the world

Abstract

Alien species are among the greatest threats to biodiversity, but the evolutionary origins of invasiveness remain obscure. We conducted the first range-wide sampling of Hemidactylus mabouia from more than 120 localities across Africa, Madagascar and the Neotropics to understand the evolutionary history of one of the most widely distributed, invasive vertebrates in the world. We used a multi-locus phylogeny, species delimitation, fossil-calibrated timetree, ancestral area reconstruction and species distribution models (SDMs) to determine how many putative species-level lineages are contained within H. mabouia , the timing and tempo of diversification, and the origins of commensality—providing insights into the evolutionary origins of invasiveness. Our analyses suggest ‘ H. mabouia ’ originated in the Miocene in the Zambezian biogeographic region and includes as many as 20 putative species-level lineages, of which only Hemidactylus mabouia sensu stricto is invasive and widely distributed, including all Neotropical records. Zambezia is the hotspot for diversity within the group with 14 species in southeastern Zambezia. SDMs suggest that H. mabouia was able to establish in the Neotropics due to habitat suitability, and globalization and the slave trade probably allowed it to cross the Atlantic. Distribution models for the H. mabouia complex overpredict the range of the invasive H. mabouia sensu stricto —highlighting the importance of taxonomy in invasive species management.

DNA barcoding of the National Museum of Natural History reptile tissue holdings raises concerns about the use of natural history collections and the responsibilities of scientists in the molecular age

“…Our results show that both H. mabouia ss and H. mercatorius occur in the Caribbean, where the presence of H. mercatorius (in at least Cuba and both the British and US Virgin Islands) represents a novel finding. We also note that Agarwal et al [34] did not detect H. mercatorius in the Gulf of Guinea, as reported by Rocha et al [36]. Studies of the H. mabouia-mercatorius complex to date have used different mtDNA loci (e.g., 16S [37], ND2 [34], and COI [39] and this study), which prohibits direct comparison of their results.…”

“…The authors suggested that the H. mabouia complex may be composed of more than 20 putative species, ranging throughout sub-Saharan Africa to the Caribbean, with H. mabouia sensu stricto (ss) ranging from Ghana through the Congolian and northern Zambezian regions, the South African and extreme southern Zambezian regions, and the neotropics. Hemidactylus mabouia ss was found to be sister to H. mercatorius, and this H. mabouia ss + H. mercatorius clade was nested within several clades of H. mabouia sensu lato (sl), which included several additional samples from the Zambezian and South African regions [34]. Our dataset includes some of the same H. mabouia specimens sequenced by Agarwal et al [34] from the Republic of Congo, which represent H. mabouia ss, as does our sample from Puerto Rico and one (of two) from the US Virgin Islands.…”

“…Hemidactylus mabouia ss was found to be sister to H. mercatorius, and this H. mabouia ss + H. mercatorius clade was nested within several clades of H. mabouia sensu lato (sl), which included several additional samples from the Zambezian and South African regions [34]. Our dataset includes some of the same H. mabouia specimens sequenced by Agarwal et al [34] from the Republic of Congo, which represent H. mabouia ss, as does our sample from Puerto Rico and one (of two) from the US Virgin Islands. Our results show that both H. mabouia ss and H. mercatorius occur in the Caribbean, where the presence of H. mercatorius (in at least Cuba and both the British and US Virgin Islands) represents a novel finding.…”

“…Hemidactylus. The H. mabouia-mercatorius complex ranges through Madagascar and Sub-Saharan Africa to the neotropics, including northern South America and the Caribbean [34]. We sequenced 15 specimens in this complex: eight from the Republic of Congo (USNM 576055-57, USNM 576116, USNM 576122, USNM 584224-25, and USNM 584229); two from Cuba (USNM 317834-35); two from the US Virgin Islands (USNM 577978-79); one from the British Virgin Islands (USNM 577977); one from Puerto Rico (USNM 577893); and one from Madagascar (USNM-HT 061).…”

“…The other US Virgin Islands sample (USNM 577978), the one from Puerto Rico, and those from the Republic of Congo were placed in a H. mabouia clade. The global distribution of this species complex was assessed recently using the mtDNA locus ND2 and several nuclear loci [34]. Although the taxonomic sampling of that study was dense regarding H. mabouia specimens from sub-Saharan Africa, it contained only a few samples from the Caribbean and only three H. mercatorius specimens from Madagascar.…”