Invasion Histories Differ Markedly Among the Four Anolis Lizard Species Introduced to Bermuda
Islands are the recipients of numerous invasive amphibians and reptiles, often resulting in negative impacts on native biodiversity. In the Caribbean, Anolis lizards, a well-studied system for their adaptive radiation, are also commonly introduced outside of their native range, which has resulted in reshaping species-area and species-isolation relationships. We used mitochondrial DNA (mtDNA) sequence data in a phylogeographic framework to reconstruct and compare invasion histories of four non-native Anolis species (A. grahami, A. extremus, A. leachii, and A. sagrei) established in Bermuda over the past 120 years. Our findings support different invasion histories for the four species, including the number of native-range sources, genetic diversity, secondary introductions and opportunities for intraspecific hybridization between previously isolated lineages. The extent of population genetic structure in the native range and the mode of introduction (i.e., intentional vs. unintentional) may influence patterns of the invasion history and genetic diversity for introduced Anolis lizards in Bermuda. These results suggest that human-mediated introductions can create substantial variation in invasion dynamics, which in turn may influence fates of species in new environments.Abstract
Sampling locations for introduced Anolis species in (A) Bermuda and their respective native ranges (black dots) of (B) Antigua and Barbuda for A. leachii, (C) Barbados for A. extremus, (D) Jamaica for A. grahami, and (E) Cuba and Florida (non-native) for A. sagrei. Pie charts indicate haplotype frequency in introduced-range populations and are color-matched to putative native-range source localities (i.e., those native-range locations with the least divergent haplotypes compared to haplotypes sampled in Bermuda). Percentages above these native-range sources localities indicate minimum pairwise sequence divergence between haplotypes from introduced populations and that native-range population. Within each panel, the legend shows colors associated with native-range haplotypic variation for each introduced Anolis species, locations in the native range with the most similar haplotypes to those sampled in Bermuda. The inset map shows the location of Bermuda (A) and all possible native and non-native (for A. sagrei) range source areas in the region (B–E) with red arrows showing introduction events.
Median-joining networks for each introduced Anolis species in Bermuda. (A) For A. leachii, the network was derived from 81 ND2 sequences. All individuals in Bermuda share the same haplotype (BM-leaH1). One haplotype from Barbuda, BAR-H1, was identical to a haplotype sampled on Antigua, ANT-H1. (B) For A. grahami, the network was derived from 88 ND2 sequences; introduced haplotypes in Bermuda (BM-gra) cluster within east-central haplogroups (EC1, EC2, EC4) from the native range in Jamaica. (C) For A. sagrei, three independent networks were constructed, each using the subset of ND2 sequences corresponding to the native-range clade with which the three haplotypes on Bermuda clustered. BM-sagH1 from Bermuda was identical to FL-BigPine364 from Florida (originally derived from the Western Cuba lineage). (D) For A. extremus, the network was derived from 37 cytB sequences. Only haplotypes belonging to the Central lineage in Barbados were used along with the two haplotypes found in Bermuda. The size of circles corresponds to the number of sampled individuals sharing that haplotype. All haplotypes from Bermuda are shown in the same olive-green color across networks, and different colors are used for the closely related native-range haplotypes. Lines represent mutation steps.
Contributor Notes
