Abstract
Species delimitation in tree species is notoriously challenging due to shared polymorphisms among species. An integrative survey that considers multiple operational criteria is a possible solution, and we aimed to test it in a species complex of aspens in China. Genetic [four chloroplast DNA (cpDNA) fragments and 14 nuclear microsatellite loci (nSSR)] and morphological variations were collected for 76 populations and 53 populations, respectively, covering the major geographic distribution of the Populus davidiana-rotundifolia complex. Bayesian clustering, analysis of molecular variance (AMOVA), Principle Coordinate Analysis (PCoA), ecological niche modeling (ENM), and gene flow (migrants per generation), were employed to detect and test genetic clustering, morphological and habitat differentiation, and gene flow between/among putative species. The nSSR data and ENM suggested that there are two separately evolving meta-population lineages that correspond to P. davidiana (pd) and P. rotundifolia (pr). Furthermore, several lines of evidence supported a subdivision of P. davidiana into Northeastern (NEC) and Central-North (CNC) groups, yet they are still functioning as one species. CpDNA data revealed that five haplotype clades formed a pattern of [pdNEC, ((pdCNC, pr), (pdCNC, pr))], but most haplotypes are species-specific. Meanwhile, PCA based on morphology suggested a closer relationship between the CNC group (P. davidiana) and P. rontundifolia. Discrepancy of nSSR and ENM vs. cpDNA and morphology could have reflected a complex lineage divergence and convergence history. P. davidiana and P. rotundifolia can be regarded as a recently diverged species pair that experienced parapatric speciation due to ecological differentiation in the face of gene flow. Our findings highlight the importance of integrative surveys at population level, as we have undertaken, is an important approach to detect the boundary of a group of species that have experienced complex evolutionary history.
Highlights
Species is a fundamental unit of biology, but there has been much debate about how to define species (e.g., Sites and Marshall, 2003; De Queiroz, 2007)
When K = 2, the southwestern populations clustered into one group and the northeastern and central populations clustered into the other group, lineage admixture was observed in a few populations where the distribution of the two lineages overlapped (Figure 3)
Under STRUCTURE analysis with K = 3, southwestern populations remained as one group, whereas the northeastern populations formed a separate group from the central populations there was considerable admixture between them (Figure S3)
Summary
Species is a fundamental unit of biology, but there has been much debate about how to define species (e.g., Sites and Marshall, 2003; De Queiroz, 2007). Speciation is a temporally extended process, typically requiring millions of years before total reproductive isolation is achieved (Coyne and Orr, 2004; Seehausen et al, 2014). Of particular concern to efforts to delimit plant species based on DNA markers is lineage sorting (Naciri and Linder, 2015). Lineage sorting renders diverging species reciprocally monophyletic for genetic markers, but until this process is completed, one or both species may appear non-monophyletic for some DNA markers, even if they have achieved complete reproductive isolation (Shaffer and Thomson, 2007; Freeland et al, 2011)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
