Is gene flow promoting the reversal of pleistocene divergence in the Mountain Chickadee (Poecile gambeli)?

Abstract

The Pleistocene glacial cycles left a genetic legacy on taxa throughout the world; however, the persistence of genetic lineages that diverged during these cycles is dependent upon levels of gene flow and introgression. The consequences of secondary contact among taxa may reveal new insights into the history of the Pleistocene's genetic legacy. Here, we use phylogeographic methods, using 20 nuclear loci from regional populations, to infer the consequences of secondary contact following divergence in the Mountain Chickadee (Poecile gambeli). Analysis of nuclear data identified two geographically-structured genetic groups, largely concordant with results from a previous mitochondrial DNA (mtDNA) study. Additionally, the estimated multilocus divergence times indicate a Pleistocene divergence, and are highly concordant with mtDNA. The previous mtDNA study showed a paucity of sympatry between clades, while nuclear patterns of gene flow show highly varied patterns between populations. The observed pattern of gene flow, from coalescent-based analyses, indicates southern populations in both clades exhibit little gene flow within or between clades, while northern populations are experiencing higher gene flow within and between clades. If this pattern were to persist, it is possible the historical legacy of Pleistocene divergence may be preserved in the southern populations only, and the northern populations would become a genetically diverse hybrid species.

Figure 1. Distribution of Poecile gambeli.

Breeding distribution of the Mountain Chickadee throughout its range. Different colors represent the three major, widespread morphological groups. The star in eastern California represents the location of a population from that showed both eastern and western mtDNA haplotypes. Dots indicate sampling localities used in this study.

Figure 2. Sampling points and genetic structure in Poecile gambeli.

Sampling localities, mtDNA assignment of all populations in circles [east or west (red or blue, respectively) from [14]] and STRUCTURE results for all individuals. Each bar represents the assignment probability of an individual to western (blue) or eastern (red) groups. Small white or black circles represent localities (in addition to genetically-sampled populations) used for ecological niche modeling of mitochondrial lineages. Population localities: (1) Siskiyou Co. California, USA (N = 8), (2) Ventura Co. California, USA (N = 8), (3) San Diego Co. California, USA (N = 8), (4) British Columbia, Canada (N = 5), (5) Kootenai Co. Idaho, USA (N = 7), (6) Nye Co. Nevada, USA (N = 8), (7) Coconino Co. Arizona, USA (N = 8), and (8) Routt, Pueblo and Fremont Cos. Colorado, USA (N = 8). Throughout the discussion, populations referred to as northern (or southern) indicate populations 1, 4, 5 (or 2, 3, 6, 7, 8).

Figure 3. Summary of genetic diversity, mismatch distributions, and neutrality tests.

Genetic diversity by locus (A) and by population (B). For both (A) and (B), bars indicate haplotype diversity and dots represent nucleotide diversity. In (B), nucleotide diversity is x100 to match the scale of haplotype diversity. In (C), % of loci per population that are significantly different than the null hypothesis: Bars and dots indicate deviations from mismatch distributions and Tajima’s D neutrality index, respectively.

Figure 4. Haplotype networks.

Median-joining haplotype networks for all 20 loci used in this study. Haplotypes are sized proportional to frequency (within each marker). Hash marks indicate haplotypes separated by two or more mutations. Blue and red indicate samples with western (pops. 1–3 in Fig. 2) or eastern (pops. 4–8) mtDNA haplotypes, respectively. Gray indicates outgroup samples. Points with an “X” through them (e.g. in Pg12) are inferred haplotypes that are absent in the sampling.

Figure 5. Patterns of migration and divergence between populations in Poecile gambeli.

Results from coalescent-based isolation with migration (IMa) analyses. Population pie charts shown as a proportion of clade assignment (from Fig. 2). Effective migration (2 Nm) represented by arrows between populations (see migration key). Inset shows divergence time between clades (pairwise comparisons for inset marked with an asterisk on map, asterisks and posterior distributions of inset are colored to indicate population comparisons). The single comparison that reached a plateau in divergence time estimates is marked with a double asterisk. Throughout the discussion, populations referred to as northern (or southern) indicate populations 1, 4, 5 (or 2, 3, 6, 7, 8).

Figure 6. Distributional modeling of Poecile gambeli mitochondrial lineages for contemporary and Last Glacial Maximum conditions.

Potential niche space models for contemporary (A) and LGM (B) periods. Blue indicates space of western clade, red of eastern clade, and black of areas of overlap between clades. In (B), white space outlined with thick black is indicative of estimated glaciated areas.