Late Pleistocene landscape changes and habitat specialization as promoters of population genomic divergence in Amazonian floodplain birds

Abstract

Although vicariant processes are expected to leave similar genomic signatures among codistributed taxa, ecological traits such as habitat and stratum can influence genetic divergence within species. Here, we combined landscape history and habitat specialization to understand the historical and ecological factors responsible for current levels of genetic divergence in three species of birds specialized in seasonally flooded habitats in muddy rivers and which are widespread in the Amazon basin but have isolated populations in the Rio Branco. Populations of the white‐bellied spinetail (Mazaria propinqua), lesser wagtail‐tyrant (Stigmatura napensis) and bicolored conebill (Conirostrum bicolor) are currently isolated in the Rio Branco by the black‐waters of the lower Rio Negro, offering a unique opportunity to test the effect of river colour as a barrier to gene flow. We used ultraconserved elements (UCEs) to test alternative hypotheses of population history in a comparative phylogeographical approach by modelling genetic structure, demographic history and testing for shared divergence time among codistributed taxa. Our analyses revealed that (i) all three populations from the Rio Branco floodplains are genetically distinct from other populations along the Amazon River floodplains; (ii) these divergences are the result of at least two distinct events, consistent with species habitat specialization; and (iii) the most likely model of population evolution includes lower population connectivity during the Late Pleistocene transition (~250,000 years ago), with gene flow being completely disrupted after the Last Glacial Maximum (~21,000 years ago). Our findings highlight how landscape evolution modulates population connectivity in habitat specialist species and how organisms can have different responses to the same historical processes of environmental change, depending on their habitat affinity.

Amazonian birds in more dynamic habitats have less population genetic structure and higher gene flow

“…These prior studies included a mix of highly specialized riverine island species like M. propinqua (Barbosa et al, 2022) and more generalist species like Hypocnemoides melanopogon (Choueri et al, 2017) and Chrysomus icterocephalus (Cadena et al, 2011) that also occur in other riverine habitats. A few of our study species, as well as some species examined in prior studies, did have higher levels of population genetic structure (Luna et al, 2022; Thom et al, 2020). We explore this population differentiation and its causes below.…”

“…A recent study from the Rio Branco in northern Amazonia provides an interesting perspective on this issue. Luna et al (2022) showed that three passerine bird species specialized on white‐water floodplains hold populations on the Rio Branco, and are isolated from populations on other white‐water rivers by the black waters of the Rio Negro. Despite different dispersal abilities, all three species were able to colonize the Rio Branco during the late Pleistocene, but the two species that live on allegedly more dynamic habitats such as sand‐bar scrub ( Mazaria propinqua and Stigmatura napensis ) showed lower population structure than the species that occurs in relatively less dynamic Cecropia ‐dominated flooded forests ( Conirostrum bicolor ).…”

“…We sampled those same populations in this study and found they clustered separately in DAPC analyses (see Appendix S1), but the differentiation was less pronounced than in many floodplain or upland forest species. Prior authors (Luna et al, 2022; Thom et al, 2020) have hypothesized that population differentiation in riverine island birds can be driven by water type, with distinct populations in sediment‐rich “white‐water” versus sediment‐poor, tannin‐heavy “black‐water” systems. Our results are consistent with this idea, with breaks evident at transitions in water type between the Rio Negro and Rio Amazonas and between the Rio Branco and Rio Negro (e.g., Figures S24, S30 and S35).…”

“…There are stretches of Amazonian rivers that lack islands, such as at the mouth of the Rio Tapajós (Irion et al, 2009), and these regions have been hypothesized to cause genetic breaks that are observed in some riverine species (Thom et al, 2020). Similarly, Luna et al (2022) demonstrated that river colour can also represent a meaningful barrier for white-or black-water river specialists, as suggested by Laranjeiras et al (2019).…”

“…Still, it is restricted to the river floodplain and is subdivided by the river itself. Multiple studies found that some Amazonian floodplain forest species had relatively high levels of genetic structure across their distribution but not between opposite banks of the same river (Barbosa et al, 2022; Luna et al, 2022; Silva et al, 2021). Corroborating the influence of habitat association of population history, Sawakuchi et al (2022) found distinct dates of demographic expansion for birds associated with riverine islands and floodplain forests in river margins, and that dates of demographic expansion for river island specialists agree with dates of sediment accumulation on riverine islands throughout Amazonia.…”