PSMC analysis of effective population sizes in molecular ecology and its application to black‐and‐white Ficedula flycatchers

Climatic fluctuations during the Quaternary period governed the demography of species and contributed to population differentiation and ultimately speciation. Studies of these past processes have previously been hindered by a lack of means and genetic data to model changes in effective population size (N_e) through time. However, based on diploid genome sequences of high quality, the recently developed pairwise sequentially Markovian coalescent (PSMC) can estimate trajectories of changes in N_e over considerable time periods. We applied this approach to resequencing data from nearly 200 genomes of four species and several populations of the Ficedula species complex of black‐and‐white flycatchers. N_e curves of Atlas, collared, pied and semicollared flycatcher converged 1–2 million years ago (Ma) at an N_e of ≈ 200 000, likely reflecting the time when all four species last shared a common ancestor. Subsequent separate N_e trajectories are consistent with lineage splitting and speciation. All species showed evidence of population growth up until 100–200 thousand years ago (kya), followed by decline and then start of a new phase of population expansion. However, timing and amplitude of changes in N_e differed among species, and for pied flycatcher, the temporal dynamics of N_e differed between Spanish birds and central/northern European populations. This cautions against extrapolation of demographic inference between lineages and calls for adequate sampling to provide representative pictures of the coalescence process in different species or populations. We also empirically evaluate criteria for proper inference of demographic histories using PSMC and arrive at recommendations of using sequencing data with a mean genome coverage of ≥18X, a per‐site filter of ≥10 reads and no more than 25% of missing data.     

Low‐coverage genomic data resolve the population divergence and gene flow history of an Australian rain forest fig wasp

Population divergence and gene flow are key processes in evolution and ecology. Model‐based analysis of genome‐wide data sets allows discrimination between alternative scenarios for these processes even in nonmodel taxa. We used two complementary approaches (one based on the blockwise site frequency spectrum [bSFS], the second on the pairwise sequentially Markovian coalescent [PSMC]) to infer the divergence history of a fig wasp, Pleistodontes nigriventris. Pleistodontes nigriventris and its fig tree mutualist Ficus watkinsiana are restricted to rain forest patches along the eastern coast of Australia and are separated into The Northern population is to the north of the Southern populations by two dry forest corridors (the Burdekin and St. Lawrence Gaps). We generated whole genome sequence data for two haploid males per population and used the bSFS approach to infer the timing of divergence between northern and southern populations of P. nigriventris, and to discriminate between alternative isolation with migration (IM) and instantaneous admixture (ADM) models of postdivergence gene flow. Pleistodontes nigriventris has low genetic diversity (π = 0.0008), to our knowledge one of the lowest estimates reported for a sexually reproducing arthropod. We find strongest support for an ADM model in which the two populations diverged ca. 196 kya in the late Pleistocene, with almost 25% of northern lineages introduced from the south during an admixture event ca. 57 kya. This divergence history is highly concordant with individual population demographies inferred from each pair of haploid males using PSMC. Our analysis illustrates the inferences possible with genome‐level data for small population samples of tiny, nonmodel organisms and adds to a growing body of knowledge on the population structure of Australian rain forest taxa.     

Genetic variation in nuclear and mitochondrial markers supports a large sex difference in lifetime reproductive skew in a lekking species

Sex differences in skews of vertebrate lifetime reproductive success are difficult to measure directly. Evolutionary histories of differential skew should be detectable in the genome. For example, male‐biased skew should reduce variation in the biparentally inherited genome relative to the maternally inherited genome. We tested this approach in lek‐breeding ruff (Class Aves, Philomachus pugnax) by comparing genetic variation of nuclear microsatellites (θ_n; biparental) versus mitochondrial D‐loop sequences (θ_m; maternal), and conversion to comparable nuclear (N_e) and female (N_ef) effective population size using published ranges of mutation rates for each marker (μ). We provide a Bayesian method to calculate N_e (θ_n = 4N_eμ_n) and N_ef (θ_m = 2N_efμ_m) using 95% credible intervals (CI) of θ_n and θ_m as informative priors, and accounting for uncertainty in μ. In 96 male ruffs from one population, N_e was 97% (79–100%) lower than expected under random mating in an ideal population, where N_e:N_ef = 2. This substantially lower autosomal variation represents the first genomic support of strong male reproductive skew in a lekking species.     

Late Pleistocene climate change promoted divergence between Picea asperata and P. crassifolia on the Qinghai–Tibet Plateau through recent bottlenecks

Divergence during the early stage of speciation can be driven by a population bottleneck via reduced gene flow and enhanced lineage sorting. In this study, we aimed to examine whether such bottlenecks occurred during the initial speciation of two closely related spruce species Picea asperata and P. crassifolia occurring on the Qinghai–Tibet Plateau (QTP). We analyzed sequences of three chloroplast, two mitochondrial DNA fragments and a further 13 nuclear loci from 216 individuals of the two species. Both species showed a low level of genetic diversity in contrast to other congeners occurring in the QTP and adjacent regions. The estimated population sizes of P. asperata and P. crassifolia are less than the ancestral population size before splitting. These results together with multiple statistical tests (Tajima's D, Fu and Li's D* and F*) suggest that these two species underwent recent bottlenecks. Based on approximate Bayesian computation (ABC), we also determined that the period of the population shrinkage was consistent with the interspecific divergence during the late Pleistocene. The reduced population sizes and the divergent selection may together have triggered the initial divergence under high gene flow between these two species. Our results therefore highlight the importance of climatic oscillations during the late Pleistocene in promoting speciation through changing demographic sizes of the ancestral species on the QTP and in adjacent regions.     

Effective population size as a driver for divergence of an antimicrobial peptide (Hymenoptaecin) in two common European bumblebee species

Social insects are the target of numerous pathogens. This is because the high density of closely‐related individuals frequently interacting with each other enhances the transmission and establishment of pathogens. This high selective pressure results in the rapid evolution of immune genes, which might be counteracted by a reduced effective population size (N_e) lowering the effectiveness of selection. We tested the effect of N_e on the evolutionary rate of an important immune gene for the antimicrobial peptide Hymenoptaecin in two common central European bumblebee species: Bombus terrestris and Bombus lapidarius. Both species are similar in their biology and are expected to be under similar selective pressures because pathogen prevalence does not differ between species. However, previous studies indicated a higher N_e in B. terrestris compared to B. lapidarius. We found high intraspecific variability in the coding sequence but low variability for silent polymorphisms in B. lapidarius. Estimates of long‐ and short‐term N_e were three‐ to four‐fold higher N_e in B. terrestris, although the species did not differ in census population sizes. The difference in N_e might result in less efficient selection and suboptimal adaptation of immune genes (e.g. hymenoptaecin) in B. lapidarius, and thus this species might become less resistant and more tolerant, turning into a superspreader of diseases.     

Phylogeography of two closely related species of Allium endemic to East Asia: Population evolution in response to climate oscillations

This study investigated the effects of climate oscillations on the evolution of two closely related Allium species, A. neriniflorum and A. tubiflorum. We sequenced three cp DNA (cpDNA) fragments (rps16, rpl32‐trnL, and trnD‐trnT, together approximately 2,500 bp in length) of two closely related Allium species, with samples from 367 individuals in 47 populations distributed across the total range of these species. The interspecific and intraspecific divergence times of the two species were in the Quaternary glaciation. The population divergence was high for the cpDNA variation, suggesting a significant phylogeographic structure (N_ST = 0.844, G_ST = 0.798, p < 0.05). Remarkable ecological differentiation was also revealed by Niche models and statistical analyses. Our results suggest the speciation event of the two species was triggered by violent climatic changes during the Quaternary glaciation.     

Population genetic structure,genetic diversity,and natural history of the South American species of Nothofagus subgenus Lophozonia (Nothofagaceae) inferred from nuclear microsatellite data

The effect of glaciation on the levels and patterns of genetic variation has been well studied in the Northern Hemisphere. However, although glaciation has undoubtedly shaped the genetic structure of plants in the Southern Hemisphere, fewer studies have characterized the effect, and almost none of them using microsatellites. Particularly, complex patterns of genetic structure might be expected in areas such as the Andes, where both latitudinal and altitudinal glacial advance and retreat have molded modern plant communities. We therefore studied the population genetics of three closely related, hybridizing species of Nothofagus (N. obliqua, N. alpina, and N. glauca, all of subgenus Lophozonia; Nothofagaceae) from Chile. To estimate population genetic parameters and infer the influence of the last ice age on the spatial and genetic distribution of these species, we examined and analyzed genetic variability at seven polymorphic microsatellite DNA loci in 640 individuals from 40 populations covering most of the ranges of these species in Chile. Populations showed no significant inbreeding and exhibited relatively high levels of genetic diversity (H_E = 0.502–0.662) and slight, but significant, genetic structure (R_ST = 8.7–16.0%). However, in N. obliqua, the small amount of genetic structure was spatially organized into three well‐defined latitudinal groups. Our data may also suggest some introgression of N. alpina genes into N. obliqua in the northern populations. These results allowed us to reconstruct the influence of the last ice age on the genetic structure of these species, suggesting several centers of genetic diversity for N. obliqua and N. alpina, in agreement with the multiple refugia hypothesis.     

Contrasting genetic patterns between two coexisting Eleutherococcus species in northern China

Climate oscillations are the key factors to understand the patterns in modern biodiversity. East Asia harbors the most diverse temperate flora, largely because an extensive terrestrial ice cap was absent during repeated Pleistocene glaciation–interglacial cycles. Comparing the demographic histories of species that are codistributed and are close relatives may provide insight into how the process of climate change influences species ranges. In this study, we compared the spatial genetic structure and demographic histories of two coexisting Eleutherococcus species, Eleutherococcus senticosus and E. sessiliflorus. Both species are distributed in northern China, regions that are generally considered to be sensitive to climatic fluctuations. These regions once hosted temperate forest, but this temperate forest was replaced by tundra and taiga forest during the Last Glacial Maximum (LGM), according to pollen records. Using three chloroplast DNA fragments, we assessed the genetic structure of 20 and 9 natural populations of E. senticosus and E. sessiliflorus, respectively. Extremely contrasting genetic patterns were found between the two species; E. sessiliflorus had little genetic variation, whereas E. senticosus had considerably higher levels of genetic variation (15 haplotypes). We speculated that a recent severe bottleneck may have resulted in the extremely low genetic diversity in E. sessiliflorus. In E. senticosus, populations in Northeast China (NEC) harbored all of the haplotypes found in this species and included private haplotypes. The populations in NEC had higher levels of genetic diversity than did those from North China (NC). Therefore, we suggest that both the NC and NEC regions can sustain LGM refugia and that lineage admixture from multiple refugia took place after the LGM elevated the local genetic diversity in NEC. In NEC, multiple genetic hot spots were found in the Changbai Mountains and the Xiaoxing'an Range, which implied that multiple locations in NEC may sustain LGM refugia, even in the Xiaoxing'an Range.     

Population genomics reveals a fine‐scale recombination landscape for genetic improvement of cotton

Recombination breaks up ancestral linkage disequilibrium, creates combinations of alleles, affects the efficiency of natural selection, and plays a major role in crop domestication and improvement. However, there is little knowledge regarding the variation in the population‐scaled recombination rate in cotton. We constructed recombination maps and characterized the difference in the genomic landscape of the population‐scaled recombination rate between Gossypium hirsutum and G. arboreum and sub‐genomes based on the 381 sequenced G. hirsutum and 215 G. arboreum accessions. Comparative genomics identified large structural variations and syntenic genes in the recombination regions, suggesting that recombination was related to structural variation and occurred preferentially in the distal chromosomal regions. Correlation analysis indicated that recombination was only slightly affected by geographical distribution and breeding period. A genome‐wide association study (GWAS) was performed with 15 agronomic traits using 267 cotton accessions and identified 163 quantitative trait loci (QTL) and an important candidate gene (Ghir_COL2) for early maturity traits. Comparative analysis of recombination and a GWAS revealed that the QTL of fibre quality traits tended to be more common in high‐recombination regions than were those of yield and early maturity traits. These results provide insights into the population‐scaled recombination landscape, suggesting that recombination contributed to the domestication and improvement of cotton, which provides a useful reference for studying recombination in other species.     

High‐resolution insight into recombination events at the SD1 locus in rice

Recombination during meiosis plays an important role in genome evolution by reshuffling existing genetic variations into fresh combinations with the possibility of recovery of lost ancestral genotypes. While crossover (CO) events have been well studied, gene conversion events (GCs), which represent non‐reciprocal information transfer between chromosomes, are poorly documented and difficult to detect due to their relatively small converted tract size. Here, we document these GC events and their phenotypic effects at an important locus in rice containing the SD1 gene, where multiple defective alleles contributed to the semi‐dwarf phenotype of rice in the ‘Green Revolution’ of the 1960s. Here, physical separation of two defects allows recombination to generate the wild‐type SD1 gene, for which plant height can then be used as a reporter. By screening 18 000 F₂ progeny from a cross between two semi‐dwarf cultivars that carry these different defective alleles, we detected 24 GC events, indicating a conversion rate of ~3.3 × 10^?4 per marker per generation in a single meiotic cycle in rice. Furthermore, our data show that indels and single‐nucleotide polymorphisms (SNPs) do not differ significantly in GC rates, at least at the SD1 locus. Our results provide strong evidence that GC by itself can regain an ancestral phenotype that was lost through mutation. This GC detection approach is likely to be broadly applicable to natural or artificial alleles of other phenotype‐related functional genes, which are abundant in other plant genomes.     

Genetic connectivity among swarming sites in the wide ranging and recently declining little brown bat (Myotis lucifugus)

Characterizing movement dynamics and spatial aspects of gene flow within a species permits inference on population structuring. As patterns of structuring are products of historical and current demographics and gene flow, assessment of structure through time can yield an understanding of evolutionary dynamics acting on populations that are necessary to inform management. Recent dramatic population declines in hibernating bats in eastern North America from white‐nose syndrome have prompted the need for information on movement dynamics for multiple bat species. We characterized population genetic structure of the little brown bat, Myotis lucifugus, at swarming sites in southeastern Canada using 9 nuclear microsatellites and a 292‐bp region of the mitochondrial genome. Analyses of F_ST, Φ_ST, and Bayesian clustering (STRUCTURE) found weak levels of genetic structure among swarming sites for the nuclear and mitochondrial genome (Global F_ST = 0.001, P < 0.05, Global Φ_ST = 0.045, P < 0.01, STRUCTURE K = 1) suggesting high contemporary gene flow. Hierarchical AMOVA also suggests little structuring at a regional (provincial) level. Metrics of nuclear genetic structure were not found to differ between males and females suggesting weak asymmetries in gene flow between the sexes. However, a greater degree of mitochondrial structuring does support male‐biased dispersal long term. Demographic analyses were consistent with past population growth and suggest a population expansion occurred from approximately 1250 to 12,500 BP, following Pleistocene deglaciation in the region. Our study suggests high gene flow and thus a high degree of connectivity among bats that visit swarming sites whereby mainland areas of the region may be best considered as one large gene pool for management and conservation.     

Contrasting responses to Pleistocene climate changes: a case study of two sister species Allium cyathophorum and A. spicata (Amaryllidaceae) distributed in the eastern and western Qinghai–Tibet Plateau

It has been hypothesized that species occurring in the eastern and the western Qinghai–Tibet Plateau (QTP) responded differently to climate changes during the Pleistocene. Here, we test this hypothesis by phylogeographic analysis of two sister species, Allium cyathophorum and A. spicata. We sequenced two chloroplast DNA (cpDNA) fragments (accD‐psaI and the rpl16 intron) of 150 individuals, and the nuclear (ITS) region of 114 individuals, from 19 populations throughout the distributional ranges of these species. The divergence between the two species was dated at 779 ‐ 714 thousand years before the present and was likely initiated by the most major glaciation in the QTP. Analysis of chlorotype diversity showed that A. spicata, the species occurring in the western QTP, contains much lower genetic diversity (0.25) than A. cyathophorum (0.93), which is distributed in the eastern QTP. Moreover, multiple independent tests suggested that the A. spicata population had expanded recently, while no such expansion was detected in A. cyathophorum, indicating a contrasting pattern of responses to Pleistocene climate changes. These findings highlight the importance of geographical topography in determining how species responded to the climate changes that took place in the QTP during the Pleistocene.     

Exome resequencing reveals signatures of demographic and adaptive processes across the genome and range of black cottonwood (Populus trichocarpa)

Extant variation in temperate and boreal plant species has been influenced by both demographic histories associated with Pleistocene glacial cycles and adaptation to local climate. We used sequence capture to investigate the role of these neutral and adaptive processes in shaping diversity in black cottonwood (Populus trichocarpa). Nucleotide diversity and Tajima's D were lowest at replacement sites and highest at intergenic sites, while LD showed the opposite pattern. With samples grouped into three populations arrayed latitudinally, effective population size was highest in the north, followed by south and centre, and LD was highest in the south followed by the north and centre, suggesting a possible northern glacial refuge. F_ST outlier analysis revealed that promoter, 5′‐UTR and intronic sites were enriched for outliers compared with coding regions, while no outliers were found among intergenic sites. Codon usage bias was evident, and genes with synonymous outliers had 30% higher average expression compared with genes containing replacement outliers. These results suggest divergent selection related to regulation of gene expression is important to local adaptation in P. trichocarpa. Finally, within‐population selective sweeps were much more pronounced in the central population than in putative northern and southern refugia, which may reflect the different demographic histories of the populations and concomitant effects on signatures of genetic hitchhiking from standing variation.     

Isolation by instability: Historical climate change shapes population structure and genomic divergence of treefrogs in the Neotropical Cerrado savanna

Although the impact of Pleistocene glacial cycles on the diversification of the tropical biota was once dismissed, increasing evidence suggests that Pleistocene climatic fluctuations greatly affected the distribution and population divergence of tropical organisms. Landscape genomic analyses coupled with paleoclimatic distribution models provide a powerful way to understand the consequences of past climate changes on the present‐day tropical biota. Using genome‐wide SNP data and mitochondrial DNA, combined with projections of the species distribution across the late Quaternary until the present, we evaluate the effect of paleoclimatic shifts on the genetic structure and population differentiation of Hypsiboas lundii, a treefrog endemic to the South American Cerrado savanna. Our results show a recent and strong genetic divergence in H. lundii across the Cerrado landscape, yielding four genetic clusters that do not seem congruent with any current physical barrier to gene flow. Isolation by distance (IBD) explains some of the population differentiation, but we also find strong support for past climate changes promoting range shifts and structuring populations even in the presence of IBD. Post‐Pleistocene population persistence in four main areas of historical stable climate in the Cerrado seems to have played a major role establishing the present genetic structure of this treefrog. This pattern is consistent with a model of reduced gene flow in areas with high climatic instability promoting isolation of populations, defined here as “isolation by instability,” highlighting the effects of Pleistocene climatic fluctuations structuring populations in tropical savannas.     

Phylogeography of northern Dolly Varden Salvelinus malma malma based on analysis of mitochondrial DNA

The northern Dolly Varden, Salvelinus malma malma, is a typical representative of arctic fauna distributed in northeastern Asia and northwestern North America. Because its spawning habitats were affected by Pleistocene glacial advances over most of its natural range, S. m. malma is among the most interesting objects of phylogeographic and microevolutionary studies. We reconstructed the genealogy of mtDNA haplotypes from 27 Alaskan and Asian populations to study the influence of glacial and geological vicariance events on the contemporary population genetic structure, phylogeographic subdivision and distribution of the northern Dolly Varden. Analysis of restriction site states in three PCR‐amplified mtDNA regions (ND1/ND2, ND5/ND6, Cytb/D‐loop; 47% of the mitochondrial genome) resolved 75 haplotypes in 436 fish. Similar patterns of subspecific variation apparent from hierarchical diversity and nested clade analyses of mtDNA haplotypes identify weak spatial differentiation and low levels of divergence. Our results suggest that (1) demographic history has been influenced by historical range expansions and recent isolation by distance, (2) present populations from Asia and North America were colonized from one main Beringian Refugium, and (3) that this taxon's ancestral population probably experienced a bottleneck in the Beringian Refugium during the late Pleistocene (Wisconsin) glacial period.     

Molecular data and ecological niche modelling reveal the Pleistocene history of a semi‐aquatic bug (Microvelia douglasi douglasi) in East Asia

This study investigated the Pleistocene history of a semi‐aquatic bug, Microvelia douglasi douglasi Scott, 1874 (Hemiptera: Veliidae) in East Asia. We used M. douglasi douglasi as a model species to explore the effects of historical climatic fluctuations on montane semi‐aquatic invertebrate species. Two hypotheses were developed using ecological niche models (ENMs). First, we hypothesized that M. douglasi douglasi persisted in suitable habitats in southern Guizhou, southern Yunnan, Hainan, Taiwan and southeast China during the LIG. After that, the populations expanded (Hypothesis 1). As the spatial prediction in the LGM was significantly larger than in the LIG, we then hypothesized that the population expanded during the LIG to LGM transition (Hypothesis 2). We tested these hypotheses using mitochondrial data (COI+COII) and nuclear data (ITS1 + 5.8S+ITS2). Young lineages, relatively deep splits, lineage differentiation among mountain ranges in central, south and southwest China and high genetic diversities were observed in these suitable habitats. Evidence of mismatch distributions and neutrality tests indicate that a population expansion occurred in the late Pleistocene. The Bayesian skyline plot (BSP) revealed an unusual population expansion that likely happened during the cooling transition between LIG and LGM. The results of genetic data were mostly consistent with the spatial predictions from ENM, a finding that can profoundly improve phylogeographic research. The ecological requirements of M. douglasi douglasi, together with the geographical heterogeneity and climatic fluctuations of Pleistocene in East Asia, could have shaped this unusual demographic history. Our study contributes to our knowledge of semi‐aquatic bug/invertebrate responses to Pleistocene climatic fluctuations in East Asia.     

Genome‐wide association study identifies variants in the CAPN9 gene associated with umbilical hernia in pigs

Pig umbilical hernia (UH) affects pig welfare and brings considerable economic loss to the pig industry. To date, the molecular mechanisms underlying pig UH are still poorly understood. To identify potential loci for susceptibility to this disease, we performed a genome‐wide association study in an Erhualian × Shaziling F₂ intercross population. A total of 45 animals were genotyped using Illumina Porcine SNP60 BeadChips. We observed a SNP (rs80993347) located in the calpain‐9 (CAPN9) gene on Sus scrofa chromosome 14 that was significantly associated with UH (P = 1.97 × 10^?10). Then, we identified a synonymous mutation rs321865883 (g.20164T>C) in exon 10 of the CAPN9 gene that distinguished two affected individuals (CC) from their normal full‐sibs (TC). Finally, quantitative polymerase chain reaction was explored to investigate the mRNA expression profile of the CAPN9 gene in 12 tissues in Yorkshire pigs at different developmental stages (3, 90 and 180 days). CAPN9 showed high expression levels in the gastrointestinal tract at these three growth stages. The results of this study indicate that the CAPN9 gene might be implicated in UH. Further studies are required to establish a role of CAPN9 in pig UH.     

Climate,physiological tolerance and sex‐biased dispersal shape genetic structure of Neotropical orchid bees

Understanding the impact of past climatic events on the demographic history of extant species is critical for predicting species' responses to future climate change. Palaeoclimatic instability is a major mechanism of lineage diversification in taxa with low dispersal and small geographical ranges in tropical ecosystems. However, the impact of these climatic events remains questionable for the diversification of species with high levels of gene flow and large geographical distributions. In this study, we investigate the impact of Pleistocene climate change on three Neotropical orchid bee species (Eulaema bombiformis, E. meriana and E. cingulata) with transcontinental distributions and different physiological tolerances. We first generated ecological niche models to identify species‐specific climatically stable areas during Pleistocene climatic oscillations. Using a combination of mitochondrial and nuclear markers, we inferred calibrated phylogenies and estimated historical demographic parameters to reconstruct the phylogeographical history of each species. Our results indicate species with narrower physiological tolerance experienced less suitable habitat during glaciations and currently exhibit strong population structure in the mitochondrial genome. However, nuclear markers with low and high mutation rates show lack of association with geography. These results combined with lower migration rate estimates from the mitochondrial than the nuclear genome suggest male‐biased dispersal. We conclude that despite large effective population sizes and capacity for long‐distance dispersal, climatic instability is an important mechanism of maternal lineage diversification in orchid bees. Thus, these Neotropical pollinators are susceptible to disruption of genetic connectivity in the event of large‐scale climatic changes.     

Interglacial microrefugia and diversification of a cactus species complex: phylogeography and palaeodistributional reconstructions for Pilosocereus aurisetus and allies

The role of Pleistocene climate changes in promoting evolutionary diversification in global biota is well documented, but the great majority of data regarding this subject come from North America and Europe, which were greatly affected by glaciation. The effects of Pleistocene changes on cold‐ and/or dry‐adapted species in tropical areas where glaciers were not present remain sparsely investigated. Many such species are restricted to small areas surrounded by unfavourable habitats, which may represent potential interglacial microrefugia. Here, we analysed the phylogeographic structure and diversification history of seven cactus species in the Pilosocereus aurisetus complex that are restricted to rocky areas with high diversity and endemism within the Neotropical savannas of eastern South America. We combined palaeodistributional estimates with standard phylogeographic approaches based on two chloroplast DNA regions (trnT‐trnL and trnS‐trnG), exon 1 of the nuclear gene PhyC and 10 nuclear microsatellite loci. Our analyses revealed a phylogeographic history marked by multiple levels of distributional fragmentation, isolation leading to allopatric differentiation and secondary contact among divergent lineages within the complex. Diversification and demographic events appear to have been affected by the Quaternary climatic cycles as a result of isolation in multiple patches of xerophytic vegetation. These small patches presently harbouring P. aurisetus populations seem to operate as microrefugia, both at present and during Pleistocene interglacial periods; the role of such microrefugia should be explored and analysed in greater detail.