Introduction
Archipelagos provide unique opportunities for phylogeographic analyses. They are geographically isolated and dispersal-mediated gene flow among individual islands is limited for some taxa (Emerson & Hewitt, 2005), such as terrestrial squamate reptiles. Located in the Tropical Eastern Pacific 960 km off the coast of Ecuador, the Galápagos archipelago has long been recognized as a model system for evolutionary and biogeographic studies (Darwin, 1859; Emerson, 2002; Grant & Grant, 2014; Losos & Ricklefs, 2009; Parent, Caccone, & Petren, 2008).
At present, the Galápagos archipelago includes 123 islands, only 13 of which are larger than 10 km2 (Snell, Stone, & Snell, 1996). Among the larger islands, Floreana is one of the few inhabited islands, and lies on the southern limit of the archipelago (Fig. 1). Only three years before Darwin’s historical arrival to the Galápagos in 1835, the first human colony of the Galápagos was established in Floreana under the term of Ecuador’s first president, Juan José Flores, after whom the island was named in Spanish. The new inhabitants introduced domestic animals, mice and rats, and the domestic animals established feral populations. Floreana’s terrestrial biodiversity was thereby impacted; several vertebrate species including Floreana’s giant tortoise (Chelonoidis niger Quoy & Gaimard, 1824), racer snake (Pseudalsophis biserialis Günther, 1860) and mockingbird (Mimus trifasciatus Gould, 1837) went extinct in the following 40-50 years (Broom, 1929; Steadman, 1986). Of these, Floreana’s racer snake and mockingbird still survive on adjacent islets (Grant, Curry, & Grant, 2000; Thomas, 1997), such as Champion and Gardner, which lack introduced mammals. In contrast, the Floreana lava lizardsMicrolophus grayii Bell, 1843 and the Floreana leaf-toed geckosPhyllodactylus baurii Garman, 1892 have survived human impacts both on Floreana and adjacent islets, although both species are frequently predated by feral and domestic cats in Floreana (Paula Castaño pers. obs. 2018; Rodriguez & Fessl, 2016; Steadman 1986). A population viability analysis of Floreana lava lizards showed that lack of recruitment was the primary factor restricting population size on Floreana and suggested that eradication of feral cats and invasive rodents might have a positive effect (Rodriguez & Fessl, 2016).
Floreana has an area of 173 km2 and is surrounded by several satellite islets, including Caldwell, Enderby, Champion, Gardner and Watson. With an estimated age of 1.5–2.3 million years (Geist, Snell, Snell, Goddard, & Kurz, 2014; White, McBirney, & Duncan, 1993), Floreana is of intermediate age of islands within the archipelago. Noting that some Galápagos islands are broadly connected by shallow seabed, some authors have proposed that Pleistocene changes in sea level had a strong influence on the evolution of land-locked vertebrates by allowing contact among populations from different islands during glaciation periods (Ali & Aitchison, 2014; Geist et al., 2014; Poulakakis, Russello, Geist, & Caccone, 2012). For example, Fernandina, Isabela, Pinzón, Rábida, Santiago, Baltra and Santa Fe are connected by shallow water (<200 m); therefore, these islands were possibly connected by land bridges during Pleistocene glaciations, when sea level fell to between -90 and -130 m below the present level, for 5–10 kyr intervals (Ali & Aitchison, 2014). In contrast, other major islands are isolated from one another by a deeper (>500 m) ocean floor and were thus not connected to other large islands in the recent past. Nonetheless, these islands were presumably connected to satellite islets separated by depths <200 m. According to this paleogeographical model, Floreana was connected to the satellite islets of Caldwell, Champion and Enderby, whereas Gardner and Watson were separated and formed a single islet during Pleistocene glaciations (Fig. 1; Geist et al., 2014). Consequently, present land organisms with limited dispersal abilities from Floreana should be more similar genetically to those from Caldwell, Champion and Enderby than they are to organisms from Gardner and Watson due to gene flow during Pleistocene low sea-level periods.
If the paleogeographic model holds true, it has great implications for the conservation of Floreana’s terrestrial vertebrates. Plans currently exist to implement an invasive rodent and feral cat eradication program using a rodenticide bait containing the second-generation anticoagulant brodifacoum (Island Conservation, 2013). Brodifacoum is highly toxic to mammals and birds; therefore, it could represent a serious threat for non-target wildlife species through primary or secondary exposure pathways (Broome, et al., 2014; Rueda et al., 2016). Fisher, Campbell, and Castaño (2017) identified that Floreana lava lizards and Floreana leaf-toed geckos were at low risk (island population may be reduced by up to 10%) if exposed to brodifacoum, in agreement with a previous study showing low toxicity of this rodenticide in the fence lizardSceloporus occidentalis (Weir et al., 2015), as well as findings from previous successful invasive rodent eradications implemented in the Galápagos Islands (Rueda et al., 2019). However,
However, as both species are endemic to Floreana and are considered evolutionary significant units (Young, 2016), mitigation actions (e.g. captive holding in situ or translocation from satellite islets) should be considered if the invasive rodent and feral cat eradication program is to be implemented. Thus, knowing the genetic affinities among lizard populations of Floreana and surrounding islets is of great importance for their conservation. Moreover, if a future goal is to recover Floreana’s extirpated endemics using the closest genetic counterparts, then reintroduction strategies should consider translocating individuals from islets containing populations recently (Pleistocene) connected with Floreana.
The aim of this study was to test the hypothesis that haplotypes of Floreana lava lizards and Floreana leaf-toed geckos from Floreana are more similar to those from Champion than they are to those from Gardner, thereby supporting the Pleistocene paleogeographic model. If the invasive rodent and feral cat eradication program on Floreana Island is implemented, our results will help identify the most appropriate risk mitigation action for both lizard species.