Ancient DNA Suggests Single Colonization and Within-Archipelago Diversification of Caribbean Caviomorph Rodents

Abstract

Reconstructing the evolutionary history of island biotas is complicated by unusual morphological evolution in insular environments. However, past human-caused extinctions limit the use of molecular analyses to determine origins and affinities of enigmatic island taxa. The Caribbean formerly contained a morphologically diverse assemblage of caviomorph rodents (33 species in 19 genera), ranging from ∼0.1-200 kg and traditionally classified into three higher-order taxa (Capromyidae/Capromyinae, Heteropsomyinae, Heptaxodontidae). Few species survive today, and the evolutionary affinities of living and extinct Caribbean caviomorphs to each other and to mainland taxa are unclear: are they monophyletic, polyphyletic, or paraphyletic? We use ancient DNA techniques to present the first genetic data for extinct heteropsomyines and heptaxodontids, as well as for several extinct capromyids, and demonstrate through analysis of mitogenomic and nuclear datasets that all sampled Caribbean caviomorphs represent a well-supported monophyletic group. The remarkable morphological and ecological variation observed across living and extinct caviomorphs from Cuba, Hispaniola, Jamaica, Puerto Rico, and other islands was generated through within-archipelago evolutionary radiation following a single Early Miocene overwater colonisation. This evolutionary pattern contrasts with the origination of diversity in many other Caribbean groups. All living and extinct Caribbean caviomorphs comprise a single biologically remarkable subfamily (Capromyinae) within the morphologically conservative living Neotropical family Echimyidae. Caribbean caviomorphs represent an important new example of insular mammalian adaptive radiation, where taxa retaining “ancestral-type” characteristics coexisted alongside taxa occupying novel island niches. Diversification was associated with the greatest insular body mass increase recorded in rodents, and possibly the greatest for any mammal lineage.

Total evidence tip-dating phylogeny of platyrrhine primates and 27 well-justified fossil calibrations for primate divergences

“…This clade was also found by Marivaux et al (2016), but not by Kay (2015) or , even though we used the latter's morphological matrix here; however, all of these previous studies were based on parsimony analysis of a morphological matrix with a molecular scaffold enforced, which precludes synergistic interactions between morphological and molecular data and which does not take into account temporal information, unlike the approach used here. The Caribbean clade receives weak support in our analyses (BPP of 0.50 in the mandibular analysis, and 0.58 in the maxillary analysis), but it is of interest because it implies that the presence of these genera in the Caribbean may be the result of a single dispersal event, as also recently suggested for Caribbean caviomorph rodents based on DNA evidence (Woods et al, 2021; although the presence of oryzomyin muroids and an apparent geomorph still indicates multiple dispersals by rodents from mainland South America to Caribbean landmasses; Marivaux et al, 2021). Of the three Caribbean taxa, DNA sequence data is currently only available for Xenothrix (Woods et al, 2018); obtaining molecular data (e.g., ancient DNA or protein sequences) from Antillothrix, Paralouatta, and also Insulacebus (Cooke et al, 2011), will allow rigorous testing of this hypothesis.…”

“…As noted above, monophyly of the Caribbean genera Xenothrix, Antillothrix and Paralouatta raises the possibility that they may be the result of a single dispersal event from the South American mainland; if so, our divergence estimates permit a wide range of ages for this event, spanning from the earliest Miocene to the late Pliocene (22.9-3.5 Ma). Although poorly constrained, this nevertheless overlaps with the inferred timing of dispersal of caviomorph rodents to Caribbean landmasses based on the molecular clock (node-dating) analysis of Woods et al (2021), which is 21.7-7.1 Ma. These estimates are therefore permissive of synchronous dispersals by platyrrhines and caviomorphs from mainland South America to Caribbean landmasses.…”

“…Node-dated molecular clock estimates for the age of crown Caviomorpha are markedly older, typically mid-to-late Eocene or earliest Oligocene (Sallam et al, 2009;Rowe et al, 2010: table 1; Meredith et al, 2011;Patterson, 2012, 2015;Voloch et al, 2013;Álvarez et al, 2017;Woods et al, 2021). Several of these studies (Voloch et al, 2013;Upham and Patterson, 2015;Álvarez et al, 2017) have calibrated the deepest divergences within crown Caviomorpha based on fossil taxa from Contamana in Peru, which have been dated to ~41 Ma and which have been referred to crown clades by some authors (Antoine et al, 2012;Boivin et al, 2017).…”

Comprehensive total evidence phylogeny of chinchillids (Rodentia, Caviomorpha): Cheek teeth anatomy and evolution

“…Elasmodontomys and Amblyrhiza) were not included in the analysis because a much wider taxon sampling would be necessary to test their inclusion in Chinchilloidea (as was made by MacPhee, 2011; Marivaux et al, 2020) and also because one of the goals of our analysis was to test the phylogenetic position of late Eocene-early Miocene basal chinchilloids, not Quaternary highly derived caviomorphs like Elasmodontomys and Amblyrhiza. Moreover, Elasmodontomys is recovered in the octodontoid family Capromyidae in recent works (see Woods et al, 2021).…”

“…Busker et al, 2019; but see Busker et al, 2020), and even several authors recognize the giant hutias (Heptaxodontidae) as closely related to chinchilloid families (e.g. MacPhee, 2011;Marivaux et al, 2020; but see Woods et al, 2021).…”