Phylogeography: the Nanling Range is not a simple barrier to
gene flow
Although a modest mountain chain with respect to elevation (Körner et
al. 2017), the Nanling Range is an effective barrier to dispersal of
relatively sedentary taxa such as amphibians and many plant species
(Chen et al. 2017, Wang et al. 2017). But, for vagile taxa like birds,
the efficacy of the range as a barrier has until now yet been
emphasized. Our phylogeographic examination indicates that the mountains
have little effect on gene flow of populations of the 5 sylvioid bird
species whose phylogeography we compared: Alcippe hueti ,Leiothrix lutea , Pterorhinus pectoralis , Staphida
torqueola , and Ixos mcclellandii . Indeed, the range may act more
as a conduit than a barrier for these birds. This effect appears to have
been the case even during the dramatic climatic (and consequent
environmental) shifts from the LGM forward.
Although the 5 species move readily past the mountains, their population
structures vary, suggesting differences in their dispersal pathways and
histories of demographic changes. For example, only one species,A. hueti , display a significant correlation between genetic and
geographic distances and, hence, IBD. For this species, genetic distance
correlates best with penalized cost-distance rather than spheric or
non-penalized distances (Supporting information). This bird more likely
took routes flying under rather than over preferred elevation.
Elevational niche gradient is known to have driven variable seasonal
migration in birds (Williamson and Witt 2021, Céspedes Arias et al.
2022), and it could also shape birds’ behavior of activity range and
direction of colonization. The closely related species to A.
hueti , A. fratercula (Zou et al. 2007, Song et al. 2009) is
distributed in the mountains in Yunnan, southwesten China. The specific
preference for certain elevation range could be the echo of the its
origination from the higher mountains. Another species, P.
pectoralis showed correlation of genetic distance to geographic
distance in part of the scenarios (Supporting information). Moreover, it
was also the only species with clustered populations in the genetic
clustering analysis (Supporting information), indicating its divergence
was gradual over time as well as distance. Combined to the negative
Tajima’ D (Table 2) which suggests a selective sweep or population
expansion after a bottleneck or founder event, and threshold timing,
this species likely went through population expansion recently, and most
likely after the LGM period.
The 3 other species differ from A. hueti and P. pectoralisand from one another in their genetic patterns. Leiothrix luteadisplays little population clustering and no IBD. This unusual pattern
likely results from a history of anthropogenic transport. At one time,
this species was a popular cage bird in China, and only recently has it
been legally protected. As such, it was transported in unpredictable
ways by pet traders and owners across the country. Staphida
torqueola displays hardly any variation across its range. This species
has a Tajima’s D close to zero, indicating a stable demographic history
and no significant selection (Tajima 1989, Rozas et al. 2017), a pattern
that may have resulted from redistribution of ancestral groups by
population contraction from a continuous genetic pool, masking its
geneflow patterns.
A general pattern in all 5 species is that no obvious shift in genotype
is evident north-to-south across the mountains (Figure 2). This
observation agrees with the BARRIER analysis (Figure 5), which
identified relatively few barriers in the mountains and even fewer
configured west-to-east, thus inhibiting north-south dispersal. Another
general pattern is the apparently small effect imposed by differences in
bioclimatic (or related) variables and elevation on bird divergence. In
the three of the species with little or no IBD (I. mcclellandii ,L. lutea , S. torqueola ), dispersal may have been more
strongly influenced by habitat (IBE) than distance, given that in the
absence of IBD some other factors must be at play to differentiate
populations (Wright 1943, Nosil et al. 2005). However, we found no
correlation between genetic distances and environmental variable
differences in these species (Supporting information). Although no
significant habitat barrier was detected, the rugged terrain of the
mountains (an environmental factor that is difficult to quantify) likely
influenced the route taken by birds to some degree.