A large historical refugium explains spatial patterns of genetic diversity in a Neotropical savanna tree species

Abstract

Background and aims: The relative role of Pleistocene climate changes in driving the geographic distribution and genetic diversity of South American species is not well known, especially from open biomes such as the Cerrado, the most diverse tropical savanna, encompassing high levels of endemism. Here the effects of Quaternary climatic changes on demographic history, distribution dynamics and genetic diversity of Dimorphandra mollis, an endemic tree species widely distributed in the Cerrado, were investigated.

Methods: A total of 38 populations covering most of the distribution of D. mollis were analysed using internal transcribed spacer (ITS) sequences and nuclear microsatellite variation [simple sequence repeats (SSRs)]. The framework incorporated statistical phylogeography, coalescent analyses and ecological niche modelling (ENM).

Key results: Different signatures of Quaternary climatic changes were found for ITS sequences and SSRs corresponding to different time slices. Coalescent analyses revealed large and constant effective population sizes, with high historical connectivity among the populations for ITS sequences and low effective population sizes and gene flow with recent population retraction for SSRs. ENMs indicated a slight geographical range retraction during the Last Glacial Maximum. A large historical refugium across central Brazil was predicted. Spatially explicit analyses showed a spatial cline pattern in genetic diversity related to the paleodistribution of D. mollis and to the centre of its historical refugium.

Conclusions: The complex genetic patterns found in D. mollis are the result of a slight geographical range retraction during the Last Glacial Maximum followed by population expansion to the east and south from a large refugium in the central part of the Cerrado. This historical refugium is coincident with an area predicted to be climatically stable under future climate scenarios. The identified refugium should be given high priority in conservation polices to safeguard the evolutionary potential of the species under predicted future climatic changes.

Keywords: Cerrado; Dimorphandra mollis; Neotropical savanna; climatic changes; coalescent analysis; demographic history; ecological niche modelling; genetic diversity; phylogeography; refugium.

Fig. 1.

Haplotype phylogenetic relationships and geographical distribution for ITS nrDNA based on the sequencing of 155 individuals of Dimorphandra mollis from 32 populations. (A) Geographical distribution of haplotypes. Different colours were assigned for each haplotype according to the key. The size of the circle represents the sample size in each population and the circle sections represent the haplotype frequency in each sampled population. (B) Phylogenetic relationships among haplotypes using a median-joining network. The size of the circumference is proportional to the haplotype frequency. Mutation sites are shown along lines in the network; small black circles are the median vectors. Different colours were assigned for each population according to the key. For details on population codes and localities Table S1.

Fig. 2.

Bayesian cluster of individuals of Dimorphandra mollis for ITS nrDNA (A) and nuclear microsatellites (B). Different colours were assigned to different clusters following the key.

Fig. 3.

Maps of consensus expressing the ensemble climatic suitability for Dimorphandra mollis, and hence its predicted potential distribution across the Neotropics during the (A) LGM (21 ka), (B) mid-Holocene (6 ka) and (C) the present.

Fig. 4.

Spatial distribution of genetic diversity of Dimorphandra mollis in relation to the historical refugium, i.e. areas climatically suitable throughout time (in red). (A) Allelic richness (Ar). (B) Expected heterozygosity (H_E). (C) Haplotype diversity (h). (D) Nucleotide diversity (π). Circumference sizes are proportional to the value of the genetic parameter, following the key.