Giant Tree Frog diversification in West and Central Africa: Isolation by physical barriers, climate, and reproductive traits

Abstract

Secondary sympatry amongst sister lineages is strongly associated with genetic and ecological divergence. This pattern suggests that for closely related species to coexist in secondary sympatry, they must accumulate differences in traits that mediate ecological and/or reproductive isolation. Here, we characterized inter‐ and intraspecific divergence in three giant tree frog species whose distributions stretch across West and Central Africa. Using genome‐wide single‐nucleotide polymorphism data, we demonstrated that species‐level divergence coincides temporally and geographically with a period of large‐scale forest fragmentation during the late Pliocene. Our environmental niche models further supported a dynamic history of climatic suitability and stability, and indicated that all three species occupy distinct environmental niches. We found modest morphological differentiation amongst the species with significant divergence in tympanum diameter and male advertisement call. In addition, we confirmed that two species occur in secondary sympatry in Central Africa but found no evidence of hybridization. These patterns support the hypothesis that cycles of genetic exchange and isolation across West and Central Africa have contributed to globally significant biodiversity. Furthermore, divergence in both ecology and reproductive traits appear to have played important roles in maintaining distinct lineages. At the intraspecific level, we found that climatic refugia, precipitation gradients, marine incursions, and potentially riverine barriers generated phylogeographic structure throughout the Pleistocene and into the Holocene. Further studies examining phenotypic divergence and secondary contact amongst these geographically structured populations may demonstrate how smaller scale and more recent biogeographic barriers contribute to regional diversification.

Giant tree frogs exemplify the promise of integrating multiple types of data in phylogeographic investigations

“…At first glance, the investigation into the diversification of Giant Tree Frogs conducted by Jaynes et al (2022) and an all‐star cast of coauthors could be mistaken for just another phylogeography study. The authors are interested in investigating speciation and diversification, and like Hugall et al (2002) and countless papers since they expect that the climatic factors which influence habitat availability over time exert some influence on species formation in this species complex.…”

“…The authors are interested in investigating speciation and diversification, and like Hugall et al (2002) and countless papers since they expect that the climatic factors which influence habitat availability over time exert some influence on species formation in this species complex. Upon closer reading, it is clear that Jaynes et al (2022) design their study to incorporate types of data ranging from morphological, behavioural, genetic, and environmental in a manner that overcomes two of the most substantial sources of uncertainty in phylogeographic research, the many axes of environmental variation that could potentially influence species diversification and the Linnean uncertainty that results from incomplete taxonomy.…”

“…Environmental variation, both in the present across the landscape and historically as the landscape changes through time, is probably an important influence on the evolutionary history of any species complex. Jaynes et al (2022) conduct their investigation with an understanding of the environmental gradients that exist in the present‐day climatic conditions and collect a large number of samples ( n = 230) in part because this allows them to collect frogs from a representative sample of these environmental conditions. While sampling of this magnitude was relatively rare even in the era of single‐locus studies, it is far less common in studies that collect SNP data because sampling a smaller numbers of individuals is an effective way to reduce sequencing costs.…”

“…However, limiting the number of sampling localities makes it more difficult to identify correlations that may be present between patterns of genetic variation and environmental features. In addition to their thorough sampling of the current environment, Jaynes et al (2022) are cognizant of existing biogeographic hypotheses in the region and design their sampling to test the role of hypothesized barriers such as the Dahomey gap and marine incursions as well as suspected locations of refugial forest habitat during the last glacial maximum. While such features would have been of unknown importance to the Giant Tree Frog system prior to the analysis of other data, their sampling allow Jaynes et al (2022) avoid the confirmation bias that can accompany poor sampling schemes (Carstens et al, 2009).…”

“…As with the uncertainty in which aspects of the environmental variation matter to this system, effective research design allows Jaynes et al (2022) to overcome the Linnean uncertainty that exists in this (and nearly all) phylogeographic investigations. Jaynes et al (2022) designed their investigation into three closely related species that reportedly occur in a complex mosaic of allopatric, parapatric, and sympatric populations.…”