Were sea level changes during the Pleistocene in the South Atlantic Coastal Plain a driver of speciation in Petunia (Solanaceae)?

Abstract

Background: Quaternary climatic changes led to variations in sea level and these variations played a significant role in the generation of marine terrace deposits in the South Atlantic Coastal Plain. The main consequence of the increase in sea level was local extinction or population displacement, such that coastal species would be found around the new coastline. Our main goal was to investigate the effects of sea level changes on the geographical structure and variability of genetic lineages from a Petunia species endemic to the South Atlantic Coastal Plain. We employed a phylogeographic approach based on plastid sequences obtained from individuals collected from the complete geographic distribution of Petunia integrifolia ssp. depauperata and its sister group. We used population genetics tests to evaluate the degree of genetic variation and structure among and within populations, and we used haplotype network analysis and Bayesian phylogenetic methods to estimate divergence times and population growth.

Results: We observed three major genetic lineages whose geographical distribution may be related to different transgression/regression events that occurred in this region during the Pleistocene. The divergence time between the monophyletic group P. integrifolia ssp. depauperata and its sister group (P. integrifolia ssp. integrifolia) was compatible with geological estimates of the availability of the coastal plain. Similarly, the origin of each genetic lineage is congruent with geological estimates of habitat availability.

Conclusions: Diversification of P. integrifolia ssp. depauperata possibly occurred as a consequence of the marine transgression/regression cycles during the Pleistocene. In periods of high sea level, plants were most likely restricted to a refuge area corresponding to fossil dunes and granitic hills, from which they colonized the coast once the sea level came down. The modern pattern of lineage geographical distribution and population variation was established by a range expansion with serial founder effects conditioned on soil availability.

Figure 1

Plant material. Map of the sampling sites for the two Petunia integrifolia subspecies, where black circles (numbers 1 to 30) represent Petunia integrifolia ssp. depauperata samples and black squares correspond to Petunia integrifolia ssp. integrifolia. More information on collection sites is available in Additional file 2. Right side: representatives of two Petunia integrifolia subspecies, general view of individual and flower detail.

Figure 2

Evolutionary relationships among haplotypes of Petunia integrifolia subspecies. (A) Bayesian phylogenetic tree with posterior probabilities (PP > 0.9) shown above the branches and ages indicated below branches for selected nodes in thousands of years (Kya). Confidence intervals are presented in parenthesis; selected nodes marked with “a” indicate the higher probability of Most Recent Common Ancestor origin in Petunia integrifolia ssp. integrifolia group and “b” and “c” indicate Center and Southern Groups of Petunia integrifolia ssp. depauperata, respectively (see Table xx for probabilities); (B) Haplotype median-joining network. Sizes of the circles are proportional to the overall frequency of the haplotypes, and the color within each circle represents different genetic groups, according to the colors on the left and the Bayesian inference. Crossed lines represent inferred differences between haplotypes. The haplotype marked by asterisk (H19) belongs to Petunia integrifolia spp. depauperata population 17 (more details in the main text).

Figure 3

Geographical groups identified by SAMOVA analyses. Each group is indicated by a border style. The colors displayed in the populations correspond to genetic groups found in the phylogeny and the median-joining haplotype network. The gray dots indicate Petunia integrifolia ssp. integrifolia populations. The black dots highlighted with larger fonts are the populations with haplotypes from different genetic groups. Lines represent SAMOVA groups according the legend.

Figure 4

Historical changes in the effective size. Bayesian skyline plot showing the effective population size fluctuation throughout time for Petunia integrifolia ssp. depauperata (solid line, median estimations; grey area, confidence interval).