Drivers of genomic landscapes of differentiation acrossPopulusdivergence gradient

Abstract

Investigating genome-wide variation patterns along a speciation continuum is of central importance to understand the evolutionary processes contributing to lineage diversification. To identify which forces have shaped the genomic landscapes in Populus, we resequenced 201 whole-genomes from eight closely related species, with pairs of species at different stages along the speciation continuum. Using population structure and identity by descent analyses, we show extensive introgression between some species pairs, especially those with parapatric distributions. Inference of historical changes in effective population sizes support species-specific demographic trajectories, including recent population expansions in species characterized by broad present-day distributions. We observe highly conserved genomic landscapes, either focusing on within-species (genetic diversity: π and recombination rate: ρ) or among-species variation (relative divergence: FST and absolute divergence: DXY ). Independent of the stage across the divergence continuum, we recovered positive correlations between the pair π and ρ and the pair DXY and ρ across all species pairs, which is consistent with a substantial contribution of linked selection in shaping these genomic landscapes. However, the positive correlations between π and DXY became weaker as the overall divergence level (da) increased, suggesting that background selection is not the only factor at play. Positive correlations between FST and DXY in all species pairs, regardless of the rate of gene flow, also indicates the high FST could be due to divergent sorting of ancient polymorphism before speciation. Our study showcases the importance of investigating genomic patterns on multiple species across the speciation continuum to better understand the genomic landscapes of diversity and differentiation.

Speciation without gene-flow in hybridising deer

“…We inferred ρ estimates that are consistent with those in other mammals (Jensen-Seaman et al, 2004), their correlations with divergence would suggest Odocoileus are well defined species and reflects a shift in recombination landscape between WTD & MD (Fig. 5; (Burri, 2017; Shang et al, 2021).…”

“…We suggest this absence could be a result of species being far along on the speciation continuum such that patterns of AS (specifically high F ST ) are no longer detectable or meaningful, which is consistent with late stages of ecological speciation (Feder 2012). Recombination rates, demography and selection can influence the patterns investigated here (Shang et al 2021), and thus future studies could integrate those factors to confirm the position of WTD & MD on the speciation continuum.…”

“…Scanning the genome for just F ST peaks when speciation with gene flow is ongoing, however, is problematic as other mechanisms can create similar F ST profiles such as genetic drift, global adaptation, or simply reduced genetic diversity due to background selection (Cruickshank and Hahn 2014; Burri 2017; Ravinet et al 2017; Booker et al 2021). Joint genome scans including F ST , d xy and π facilitate a better depiction of the processes of selection and ecological speciation (Cruickshank and Hahn 2014; Ravinet et al 2017; Campbell et al 2018; Shang et al 2021). Of note, Shang et al (2021) recently conceptualized the expected patterns F ST , d xy and π under the main modes of selection; this approach allows for neutral patterns of evolution to be disentangled from selection based on contrasting patterns of the aforementioned summary statistics.…”

“…Joint genome scans including F ST , d xy and π facilitate a better depiction of the processes of selection and ecological speciation (Cruickshank and Hahn 2014; Ravinet et al 2017; Campbell et al 2018; Shang et al 2021). Of note, Shang et al (2021) recently conceptualized the expected patterns F ST , d xy and π under the main modes of selection; this approach allows for neutral patterns of evolution to be disentangled from selection based on contrasting patterns of the aforementioned summary statistics.…”

“…Shang et al (2021) recently conceptualised the expected patterns FST, dxy and p under four main modes of selection: divergence with gene flow, allopatric selection (i.e., positive selection post species split), balancing selection and background selection. Using pairs of Populus species along the speciation continuum, they investigated the genomic landscape to test clear predictions on the behaviour of FST, dxy and p. Including correlations of p and FST to recombination rate (ρ) can help place a species on the continuum, with for example the correlation between FST and ρ is expected to become stronger with increasing divergence (Burri, 2017;Shang et al, 2021). This approach ultimately allows for disentangling selection patterns from a neutral background based on those statistics, and helps decipher the evolutionary history of species.…”