Quynh Quach, a master’s student from the Revell Lab at U. Mass. Boston, presented her thesis research on “Phylogeography and Population Structure of Anolis cristatellus on the island of Vieques.” Before Quynh joined the Revell lab, former post-doc Graham Reynolds and former Losos lab undergraduate Tanner Strickland looked at the phylogeography of Anolis cristatellus across Puerto Rico and the Virgin Islands using mitochondrial DNA (in review). Tanner’s work revealed that there was a mitochondrial break on the island of Vieques, just off the coast of Puerto Rico. The mitochondrial data suggested that there were two genetically different groups of A. cristatellus, one on the East and one on the West of Vieques. The only problem was, as we know, mtDNA patterns are not always supported by nuclear whole-genome DNA patterns. In addition, Tanner’s dataset only consisted of 9 samples from Vieques.
When Quynh joined the lab, she wanted to know more about this pattern. Would this division be supported by nuclear genome analyses? Were these lineages anthropogenically introduced? If not, what was the origin of these groups – historical allopatry followed by secondary contact or isolation by distance? So she set out to answer these questions by collecting 300 tail tips from across the island of Vieques, extracting and sequencing both mtDNA and nuclear DNA.
Quynh first constructed a mitochondrial phylogeny to verify the pattern observed by Tanner and Graham. The mtDNA analysis confirmed that there are 2 mtDNA clades on Vieques with strong geographic patterns. The island-wide pattern of mtDNA variation was not what we would expect if anthropogenic introduction were the cause since this would be unlikely to show such a clear East-West pattern with the small contact zone in the middle. So then how did this pattern arise?
To answer that question, Quynh next looked at nuclear DNA using RADseq. She sequenced 48 individuals: 5 from Virgin Islands, 6 from Puerto Rico, and 37 from Vieques, then de novo assembled the genome and called 16,808 SNP’s. She ran STRUCTURE and DAPC analyses on this data and found that the Virgin Island samples form 1 cluster and Puerto Rico and Vieques form a second cluster with 4.1% divergence between the groups. But she wondered, what if we look at just Vieques and specify K=2? When she did this with DAPC and saw a clear geographic pattern similar to what she found with the mtDNA. Finally, she tested whether this represented isolation by distance. She found that there was significantly reduced gene flow between geographically distant individuals, supporting this hypothesis as the most likely cause of the variation.
Lastly, Quynh emphasized that it is important to consider multiple genetic markers and not just rely on mtDNA results. Had the group stopped at their original mitochondrial analysis, they would have reached a very different conclusion.