Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Nov 7;112(11):e70122. doi: 10.1002/ajb2.70122

Dissimilar climatic niche is predictive of contrasting historical demographic changes and altitudinal shifts in related oak species (Quercus)

Ricardo Gaytan‐Legaria 1,2, Ken Oyama 3, Octavio Rojas‐Soto 4, Antonio González‐Rodríguez 1,✉
PMCID: PMC12640476  PMID: 41199650

Abstract

Premise

Comparative surveys allow us to characterize the influence of specific factors on population genetic diversity and structure. We conducted a comparative phylogeographic study for three Mexican oak species to identify how their climatic niche preferences and breadth may have influenced historical demography and range shifts during Pleistocene climatic oscillations.

Methods

We estimated genetic diversity and structure for Quercus deserticola, Q. glaucoides, and Q. peduncularis. We inferred historical demographic changes using approximate Bayesian computation and used ecological niche models to determine present potential distribution of the species and used past climatic scenarios to estimate range and altitudinal shifts. We also measured the niche breadth of each species and evaluated niche similarity among species.

Results

We identified differences in population history, which we related to the climatic niche of individual species. For Q. deserticola, we inferred a historical bottleneck consistent with the interglacial refugia hypothesis. Quercus glaucoides, which is characterized by a narrow niche breadth, had high levels of genetic structure based on plastid DNA. Quercus peduncularis had high genetic diversity and low structure. We found correlations between niche breadth and values of genetic structure and diversity. Interglacial contraction and glacial expansion in the three species differed in magnitude, with Q. deserticola exhibiting the most drastic contraction during the interglacial.

Conclusions

Mexican oak species responded differently to historical climatic changes since they have distinct distributions in geographic and climatic space. Levels and patterns of genetic variation agreed with the population history of each species inferred using niche modeling.

Keywords: comparative phylogeography, ecological niche, Fagaceae, niche breadth, niche divergence, range shifts

The effects of Pleistocene glacial–interglacial climatic oscillations on the distribution of population genetic diversity have been widely documented across distinct taxa and regions (Hewitt, 2000, 2004). The geographical contraction of species that inhabited Holarctic regions into southern refugia during glacial periods is an example of how Pleistocene climatic changes had an influence on historical demography and shaped the genetic structure and diversity of populations (Schönswetter et al., 2005; Nieto Feliner, 2011). At lower latitudes, such as the Mexican highlands, climatic conditions during glacial periods were not as severe as in the temperate zone (Sandel et al., 2011), allowing populations of many species to persist in situ with moderate altitudinal migration and relatively constant effective population size (Ramírez‐Barahona and Eguiarte, 2014; Mastretta‐Yanes et al., 2015; Peñaloza‐Ramírez et al., 2020).

However, phylogeographical patterns of species from the Mexican transition zone (MTZ) and Neotropical regions indicate that species had differential responses to Pleistocene fluctuations according to their climatic affinities or to the geographical configuration of the areas where they were distributed (Cabanne et al., 2016; Ornelas et al., 2019; Van Els et al., 2021). The MTZ is the area where the Neotropical and Nearctic regions overlap, corresponding to the moderate to high‐elevation highlands of Mexico, Guatemala, Honduras, El Salvador, and Nicaragua (Morrone, 2020). It is one of the most biologically diverse regions of the world (Barthlott et al., 1996; Marshall and Liebherr, 2000), due to a complex geological history, the interaction between Neotropical and Nearctic biotas and heterogeneous climatic conditions. Within the MTZ, some species distributed in xerophytic areas show signals of demographic expansions and contractions during glacial cycles (Cornejo‐Romero et al., 2017; Aguirre‐Planter et al., 2020) but at different time, because some species show patterns in agreement with the glacial refugia hypothesis (Sosa et al., 2009; Ornelas et al., 2018), but others are consistent with the interglacial refugia hypothesis (Aguirre‐Planter et al., 2020; Contreras‐Negrete et al., 2021). However, under both hypotheses, refugia are crucial in maintaining genetic diversity in those species (Ruiz‐Sanchez et al., 2012; Loera et al., 2017; Contreras‐Negrete et al., 2021). In the case of species associated with the tropical dry forest in the MTZ, high levels of genetic structure have been found, resulting from substantial shifts in potential distribution through climatic changes (Pennington et al., 2004; Castillo‐Chora et al., 2021; Rivera‐Ortíz et al., 2023). By contrast, in tropical cloud forests and temperate forests, species probably maintained fairly constant effective population sizes and distribution ranges because they may instead have experienced altitudinal migrations that promoted alternate periods of relative population isolation followed by secondary contact, allowing connectivity and retention of high genetic diversity (Ramírez‐Barahona and Eguiarte, 2014; Ramos‐Ortíz et al., 2016; Sosa et al., 2016; Ornelas et al., 2019). Thus, the responses to climatic changes in the Quaternary for species distributed in the MTZ likely depended partly on their climatic affinities and distribution patterns, with species within the same ecosystem types showing similar, common patterns.

Oaks (Quercus L., Fagaceae) are widely distributed in temperate and tropical regions of the northern hemisphere and are particularly dominant in several biomes such as temperate deciduous forests, subtropical evergreen forests, and oak–pine forests (Nixon, 2006). The MTZ has a high richness of Quercus species, likely related to the highly diverse environments to which they are adapted (Hipp et al., 2018). Specifically, species in the Mexican white and red oak clades (Erythromexicana and Leucomexicana, respectively) are distributed in cloud, temperate, and tropical forests and in arid zones (Valencia‐Ávalos, 2004; Rodríguez‐Correa et al., 2015; Hipp et al., 2018). The diversification of oak clades offers an opportunity to infer the effects of Pleistocene climatic changes on species that are closely related, but that occur in distinct environments (Hipp et al., 2019; Cavender‐Bares et al., 2018).

Previous phylogeographic studies of oak species from the Holarctic region have found that the geographic distribution of their genetic diversity reflects drastic contractions of populations into southern refugia during glaciations and postglacial range expansions (Petit et al., 2005; Bagnoli et al., 2016), resulting in reduced genetic diversity and high genetic structure in plastid DNA markers. In contrast, most (but not all) oaks in the Mexican highlands are characterized by higher genetic diversity and moderate population differentiation, indicating less drastic effects of climatic change, with populations experiencing altitudinal rather than latitudinal displacements (González‐Rodríguez et al., 2004; Peñaloza‐Ramírez et al., 2020). However, patterns across species may vary depending on ecological requirements and niche breadth (Gaytán‐Legaria et al., 2023).

Comparative phylogeography studies explore geographical patterns of genetic diversity across multiple co‐distributed taxa (Arbogast and Kenagy, 2001), and we can evaluate whether genetic structure is associated with recently derived differences among closely related species (Gutiérrez‐Garíca and Vázquez‐Domínguez, 2011). Here in a comparative phylogeographic study of three oak species in the Leucomexicana clade (Q. deserticola Trel., Q. glaucoides M. Martens & Galeotti, and Q. peduncularis Née; Hipp et al., 2019) that differ in niche breadth (Gaytán‐Legaria et al., 2023) and in specific climatic niches, we evaluated whether these differences can be related to contrasting population histories and genetic diversity patterns that may be responses to recent geological climate change episodes. These species are endemic to the MTZ but have dissimilar distributions and environmental requirements (Figure 1). Quercus deserticola is distributed mainly in the Trans‐Mexican Volcanic Belt and in part of the Oaxacan province of the Sierra Madre del Sur at elevations above 2000 m. Quercus glaucoides is found in the transition zones between temperate and seasonally dry tropical forests in the Trans‐Mexican Volcanic Belt, the Sierra Madre del Sur, and the margins of the Balsas Basin. Finally, Q. peduncularis occurs in the transition zones between temperate and cloud forests from central Mexico to Honduras.

Figure 1.

Figure 1

Open in a new tab

Geographic and climatic distribution of (A) Quercus deserticola, (B) Q. glaucoides, and (C) Q. peduncularis. Potential distribution was generated from an ecological niche model per species; density plots indicate the scores per species in the first two principal components (D, E) in a PCA of the uncorrelated bioclimatic variables.

The main aims of our study were to (1) compare the effects of Pleistocene climate changes on the genetic diversity and structure of these three closely related oak species distributed in distinct climatic conditions, (2) estimate the geographic ranges and altitudinal shifts through Pleistocene climatic changes of the three oak species and the consistence with historical demographic dynamics, and (3) associate niche breadth of oak species with their patterns of genetic diversity and structure.

MATERIALS AND METHODS

Study system

According to the phylogeny inferred by Hipp et al. (2019), Q. glaucoides and Q. deserticola are sister species, while Q. peduncularis is part of a sister clade to those two species. Quercus deserticola is distributed in sites with a mean annual temperature (range) of 16.4°C (12–23°C), mean annual precipitation of 786 mm (500–1100 mm), and an elevation range of 2000 to 2800 m above sea level (m a.s.l.), mainly in the Sierra Madre del Sur and the Trans‐Mexican Volcanic Belt. In comparison, Q. glaucoides occurs at sites with a mean annual temperature of 21.4°C (17°C–26°C), mean annual precipitation of 944 mm (500–1200 mm) and a prolonged and marked dry season, at elevations of 600 to 2200 m a.s.l. of the Trans‐Mexican Volcanic Belt, the Balsas Basin and the Sierra Madre del Sur. Finally, Q. peduncularis, is distributed from central Mexico to Honduras, in areas with mean annual temperature of 20.5°C (10–24°C) and mean annual precipitation of 1284 mm (500–2800 mm) at elevations between 340 and 1940 m a.s.l. in the Chiapas Highlands, the Veracruzan province, the Balsas Basin, the Sierra Madre del Sur and the Trans‐Mexican Volcanic Belt.

Samples and DNA sequencing

Leaves were collected from 202 individuals from 12 populations of Q. deserticola and 11 of Q. peduncularis throughout their distribution (Appendix S1: Table S1), except for inaccessible areas (for security reasons) from the Guerreran district of the Sierra Madre del Sur. In addition, we incorporated samples from 21 populations of Q. glaucoides analyzed previously (Gaytán‐Legaria et al., 2023) into this study. We sampled 3–12 oaks per population, trying to keep a minimum of 20 m between sampled trees to avoid including highly related individuals. DNA was extracted as described by Doyle and Doyle (1987) and diluted to a final concentration of 20 ng/µL.

The plastid trnC‐trnD intergenic region was amplified by PCR using three pairs of primers (trnC‐ycf6R, ycf6‐psbMR, psbMF‐trnD; Shaw et al., 2005) in a final volume of 20 µL using Taq PCR Master Mix (Qiagen, Hilden, Germany) and the manufacturer's instructions. Thermal cycling programs were described by Gaytán‐Legaria et al., (2023). PCR products were purified and then sequenced by Psomagen (Rockville, MD, USA). We then assembled, base‐called, and aligned sequences with the software PhyDE‐1 V0.9971 (Müller et al., 2010).

We also amplified nine nuclear simple sequence repeats (nSSRs) loci (quru‐GA‐0C11, quru‐GA‐0M05, quru‐GA‐0C19, quru‐GA‐2F05, quru‐GA‐0M07, quru‐GA‐1C08, QpZAG36, QrZAG39, and QpZAG110) that were developed for Q. rubra, Q. petraea, and Q. robur (Steinkellner et al., 1997; Kampfer et al., 1998; Aldrich et al., 2003), in samples of Q. deserticola and Q. peduncularis. For Q. glaucoides, we used previous data from Gaytán‐Legaria et al., (2023) and new amplifications of the quru‐GA‐OM07 and quru‐GA‐1C08 loci. We used fluorescently labelled (VIC, FAM, PET) forward primers. The Platinum Multiplex PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) was used for the 6‐µL reaction mixture containing 3 µL of Platinum 2X, 1 µL of template DNA (10 ng/µL), 0.3 µL of each primer (10 µM) and 1.4 µL of deionized water. Four groups of primers were used for multiplex PCR reactions according to the color of fluorescent tags and fragment sizes (Appendix 1: Table S2). The temperatures of reactions were as in Gaytán‐Legaria et al., (2023). We analyzed PCR products in an ABI‐PRISM 3300 Avant sequencer (Thermo Fisher) and determined fragment sizes using the program GeneMarker 2.6.4 (SoftGenetics, State College, PA, USA) using Gene‐scan‐600 Liz as a size standard.

Plastid DNA data analysis

We estimated the number of haplotypes (h) and the haplotype diversity (Hd) per population and species using Arlequin 3.5 (Excoffier and Lischer, 2010). To evaluate phylogeographic structure, we estimated the genetic differentiation indexes G ST and N ST (for unordered and ordered alleles, respectively) and corresponding haplotypic diversity at the population (h S and v S) and global (h T, v T) levels, using 10,000 permutations to test for differences between the G ST and N ST values in the software Permut 1.2.1 (Pons and Petit, 1996). To characterize the genetic structure in the three species, we performed analyses of molecular variance (AMOVA) considering (1) three groups according to species designation and (2) three separate analyses to infer intraspecific structure. In the first case, overall genetic differentiation among populations (F ST) was partitioned into the among species component (F CT) and the among‐population within‐species component (F SC). We conducted the AMOVAs in Arlequin 3.5 (Excoffier and Lischer, 2010) using 10,000 permutations to test for the significance of the genetic differentiation statistics.

We reconstructed relationships among haplotypes with a statistical parsimony network approach implemented in TCS 1.2.1 (Clement et al., 2000) using a connection limit of 95% probability. Gaps in the alignment were treated as a fifth evolutionary state, considering contiguous gaps as a single insertion or deletion event (Simmons and Ochoterena, 2000).

Genetic diversity and structure in nSSRs

We tested the presence of null alleles or genotyping errors due to stuttering or large allele dropout using the software micro‐checker (Van Oosterhout et al., 2004) with a confidence interval of 95%. We used the software FreeNA (Chapuis and Estoup, 2007), which implements the expectation‐maximization algorithm (Dempster et al., 1977) to estimate the frequency of null alleles per locus and population. We also estimated the total and pairwise genetic differentiation (F ST) using Weir's method (Weir, 1996) with and without the ENA correction (excluding null alleles; Chapuis and Estoup, 2007). The analysis was run with a bootstrap over loci with 10,000 replicates to obtain F ST values with a 95% confidence interval.

To estimate inbreeding levels in the three oak species while considering null alleles, we used the software INest 2.0 (Chybicki and Burczyk, 2009) based on the Bayesian individual inbreeding model (IIM). The analysis considers the effect of null alleles (n), inbreeding (f), and genotyping errors (b) on the homozygosity values, by implementing the full model (nfb), and its comparison with the alternative models (nf, nb, fb) using the deviance information criterion (DIC).

We used the software GenAlEx 6.5 (Peakall and Smouse, 2012) to obtain the values of mean number of alleles per locus (N A), mean effective number of alleles per locus (N AE), mean observed heterozygosity (H O), mean expected heterozygosity (H E), mean unbiased expected heterozygosity (uHE) and the fixation index (F) per population. To determine possible associations of the genetic diversity values with geographic variables such as latitude, longitude, and elevation, we carried out Pearson correlation analyses in R version 3.0.3 (R Core Team, 2014) for each species. To evaluate the association of genetic differentiation with geographic distances for each species, we ran a Mantel test to evaluate isolation by distance using the R package adegenet (Jombart, 2008) using 1000 permutations to test the significance of the relationship between genetics and geography.

To evaluate genetic differentiation within the three species, we performed AMOVAs considering the same groups described for the plastid DNA analysis above. We also used a Bayesian clustering algorithm implemented in structure 2.3.1 software (Pritchard et al., 2000) to infer genetic structure and admixture among populations. First, we ran an analysis using all populations of the three species to test for the correspondence of genetic structure with species designation. We then analyzed each species separately to test for intraspecific structure. We carried out each analysis using the admixture model with a burn‐in period of 100,000 and 1,000,000 MCMC replicates. We determined the number of potential K values according to the maximum number of populations for each analysis (K = 5 for the analysis with the three species; K = 12 for Q. deserticola, K = 20 for Q. glaucoides, and K = 11 for Q. peduncularis). For each K value, we performed 10 iterations. We used Evanno's method (Evanno et al., 2005) to determine the most likely K value for each analysis. We used CLUMPP (Jakobsson and Rosenberg, 2007) to align replicates for the most likely K values and then illustrated the results as a bar plot and a map. We also performed a discriminant analysis of principal components (DAPC) implemented in the R package adegenet (Jombart, 2008) to test for the correspondence of genetic clusters with species designation and for each species separately to infer intraspecific structure. We used the Bayesian information criterion (BIC) to estimate the most likely K value for the comparison among all populations and per species. The cross‐validation method was used to obtain the optimal number of principal components to retain in the DAPC. The membership for each individual from DAPC was plotted to compare with the structure results.

To identify discontinuities in gene flow over the distribution of each oak species, we used Monmonier's algorithm as implemented in BARRIER 2.2 (Manni et al., 2004). We based the analysis on an F ST pairwise matrix per species, estimated using Arlequin with 1000 bootstraps, to localize on a map the locations of areas with maximum among‐population differentiation.

Historical demography

We used two methods to test for historical demographic changes in the three oak species. The first uses Fu's F S (Fu, 1997) and Tajima's D (Tajima, 1989) neutrality tests in the program Arlequin 3.5 (Excoffier and Lischer, 2010). The second evaluates alternative historical demography scenarios using an approximate Bayesian computation (ABC) analysis, as implemented in DIYABC 2.1.0 (Cornuet et al., 2014). For the ABC analysis, five competing scenarios were hypothesized: (1) constant population size; (2) a recent bottleneck; (3) a recent expansion; (4) historical expansion followed by a recent bottleneck; and (5) historical bottleneck followed by a recent expansion. Both data sets (plastid DNA and nSSRs) were combined, simulating two million data sets for each of the five scenarios. We used 1% of the simulated data closest to the observed data to estimate the posterior distribution of historical parameters and the relative posterior probabilities of each scenario through logistic regression. The timing of demographic events was estimated by assuming an average generation time of 100 years (Cavender‐Bares et al., 2011; Li et al., 2019). The prior distributions of historical parameters are presented in the supplementary material (Appendix S1: Table S3).

Ecological niche model

We ran ecological niche models (ENMs) to estimate the current potential distribution for each of the three oak species and compare their suitable area across past climatic scenarios, that is, the last interglacial (LIG ≈ 120 k years ago [kya]), the last glacial maximum (LGM ≈ 21 kya), and the mid‐Holocene (MH ≈ 6 kya). We obtained occurrence records for the three oak species (114,168, and 159, respectively, for Q. deserticola, Q. glaucoides, and Q. peduncularis) from the Herbario Nacional of the Universidad Nacional Autónoma de México (MEXU) and the Herbarium of the Centro Regional del Bajío of the Instituto de Ecología, A. C. (IEB). We eliminated duplicate occurrences and thinned remaining occurrences to be spaced at least 4 km from each other using the R package spThin (Aiello‐Lammens et al., 2015) to avoid undue aggregation of the records (Hijmans, 2012). To reduce the possibility of including erroneous records, we used the distribution and altitudinal ranges of the oak species proposed by Valencia‐Ávalos (2004) as filtering criteria. After this filtering, we retained a total of 31, 60, and 77 occurrences for Q. deserticola, Q. glaucoides, and Q. peduncularis, respectively.

We constructed ENMs using 19 bioclimatic variables derived from monthly temperature and precipitation from 1960 to 1990 (Hijmans, 2012) from the WorldClim global climate data (http://www.worldclim.org/data/v1.4) with a resolution of 30 arcsec (≈1 km2). We excluded four of these 19 variables (Bio8, Bio9, Bio18, and Bio19) because they presented spatial anomalies in the form of odd discontinuities between neighboring pixels (Escobar et al., 2014; Booth, 2022). We reduced redundancy among remaining variables using a jackknife test implemented in Maxent 3.4.1 (Phillips and Dudík, 2008). We used Spearman's rank correlation values below 0.85 as a criterion to select the final set of variables to construct the ENMs (van Steenderen and Sutton, 2024). For Q. deserticola, the final set of variables included Bio1, Bio5, Bio6, Bio11, Bio13, Bio15, and Bio17. For Q. glaucoides, the final variables were Bio2, Bio3, Bio4, Bio6, Bio7, Bio10, Bio16, and Bio17. For Q peduncularis, the variables were Bio2, Bio3, Bio4, Bio5, Bio6, Bio7, and Bio16.

Using a spatial restriction criterion, we created a minimum convex polygon of the occurrence points for each species; we then applied a buffer of 200 km to delimit the calibration area for each species as done by Rojas‐Soto et al., (2024). The calibration areas obtained were used as masks to reduce overprediction in the suitability areas and to perform better model validations. To build the ENMs based on the maximum entropy algorithm, we used Maxent 3.4.1 (Phillips and Dudík, 2008). With R package kuenm (Cobos et al., 2019), we explored 90 possible combinations of three feature class combinations (linear = l, quadratic = q, and product = p) and 15 regularization multipliers (from 0.1 to 6). We partitioned occurrence records per species, using 80% for model calibration and 20% for model validation. We ran 100 replicates of the model with a convergence threshold of 10–5 and 500 iterations, with the extrapolation and clamping options disabled to avoid artificial overprediction on non‐analogous past climatic conditions. We obtained 90 models from the combinations of the parameters evaluated based on three criteria: (1) the partial receiver operating characteristic (ROC) curve (a modification of the traditional ROC curve), (2) omission rate and (3) the Akaike information criterion (AIC). Once we selected the optimal parametrization, we then re‐ran Maxent to generate the final models. We evaluated model performance using the AUC and partial ROC values. We used the minimum training presence value as a threshold to obtain final binary maps per species. We transferred these ENMs into three past climatic scenarios (i.e., MH, LGM, and LIG) under the Community Climate System Model ver. 4 (CCSM4; Gent et al., 2011) global circulation model. To infer elevation range shifts for each species between periods, we generated 5000 random points within the suitable area predicted for each climatic scenario from the LIG to the present. We then extracted the elevation value for each point and generated a box plot and a density plot using the R package ggplot2 (Wickham, 2016) to compare elevation shifts.

Niche conservatism and niche breadth

We measured niche overlap among the three oak species using a PCA‐environment approach (Broennimann et al., 2012), constructing the environmental space using bioclimatic variables from Worldclim 2.1 (Fick and Hijmans, 2017) with a variance inflation factor below 10 from the sum of the calibration area of the three species (Bio2, Bio3, Bio8, Bio9, Bio13, Bio14, Bio15, Bio18, and Bio19). We used two principal components to define the environmental space to measure the overlap among species using Schoener's D (Schoener, 1968) and Hellinger's I (Van der Vaart, 1998) ranging from 0 (no overlap) to 1 (complete overlap). We performed an identity (equivalency) test to compare the resulting values with random values using the R package ecospat (Di Cola et al., 2017), with two hypotheses (conservatism and divergence) for each pair of species using 100 replicates. We conducted ANOVAs followed by a Tukey HSD test in R to identify which bioclimatic variables present significant differences among Quercus species.

We estimated the niche breadth of the three species by calculating Levins' B2 (Levins, 1968) values in environmental space, using the R package ENMTools (Warren et al., 2018, 2021). We used the same set of uncorrelated bioclimatic variables for this procedure as in the niche comparison analysis. We estimated an ecological niche model for each species using general linear models (GLM) implemented in ENMTools and measured the niche breadth in terms of the geographic range and the environmental range as proportions of the sum of the accessibility areas of the three species. We also included data on niche breadth for other oak species (González‐Rodríguez et al., 2004; Tovar‐Sánchez et al., 2008; Ramos‐Ortíz et al., 2016; Peñaloza‐Ramírez et al., 2020) within the MTZ, and performed a linear regression analysis to evaluate the relationship of genetic diversity (H E, H O, H S, and H T) and structure (F ST, G ST, and N ST) indices to the geographic and environmental niche breadth.

RESULTS

Plastid DNA diversity, distribution, and relationship of haplotypes

We analyzed 362 sequences of the plastid region trnC‐trnD with an alignment length of 2109 bp from 44 populations. Of these 362 sequences, 103 were obtained for Q. deserticola, 175 for Q. glaucoides, and 84 for Q. peduncularis. We found a total of 26 haplotypes across the three oak species (Figure 2): seven in Q. deserticola, 18 in Q. glaucoides, and eight in Q. peduncularis. Quercus peduncularis had the lowest haplotype diversity (Hd = 0.54), followed by Q. deserticola (Hd = 0.615) and then Q. glaucoides (Hd = 0.86) (Table 1). Haplotypes H1, H2, and H5 were shared among the three species, with H2 being the most frequent. In Q. deserticola, this haplotype (H2) was found in 55.3% of the individuals and 75% of the populations. In Q. glaucoides, it was present in 30.28% of the individuals and 38.1% of the populations, mainly in the Balsas Basin. In Q. peduncularis, this haplotype was present in 70.23% of individuals and 91% of populations, predominantly in the Chiapas Highlands and Veracruz province. The second most frequent haplotype in Q. deserticola and Q. glaucoides was H1, present in 25.24% of the individuals of Q. deserticola and 11.42% of Q. glaucoides. In contrast, this haplotype only occurred in 3.57% of Q. peduncularis individuals. Thirteen of the haplotypes present in Q. glaucoides were not shared with the other two species, while in Q. peduncularis, five haplotypes were not shared, and in Q. deserticola, three haplotypes were not shared.

Figure 2.

Figure 2

Open in a new tab

Geographic distribution of haplotypes from the plastid region trnC‐trnD in (A) Quercus deserticola, (B) Q. glaucoides, and (C) Q. peduncularis and (D) a haplotype network of all haplotypes (haplotype networks for each species are also presented in individual maps).

Table 1.

Genetic diversity statistics for populations of Quercus deserticola, Q. glaucoides, and Q. peduncularis at plastid DNA and nuclear microsatellites (nSSRs). Number of individuals analyzed for SSRs (N SSR), mean number of alleles (N A), mean number of effective alleles (N AE), mean observed heterozygosity (H O), mean expected heterozygosity (H E), rarefied allelic richness (AR), fixation index (F), number of individuals sequenced for the trnC‐trnD region (N cp), number of haplotypes (h), haplotype diversity (Hd), Tajima's D, Fu's F S, for populations of Q. deserticola, Q. glaucoides, and Q. peduncularis. *P < 0.05. Please note that for population 23 nuclear genetic diversity statistics were not calculated given the very small sample size.

Pop Species N SSR N A N AE H O H E uHE F AR N CP h Hd D F S
1 Q. deserticola 12 5.778 3.987 0.579 0.691 0.725 0.156 2.74 12 1 0 0 0
2 12 5.111 3.586 0.591 0.682 0.721 0.136 2.7 11 5 0.818 2.2* 3.832
3 9 5.444 3.662 0.648 0.669 0.71 –0.009 2.72 9 1 0 0 0
4 5 4.222 3.143 0.533 0.644 0.739 0.185 2.79 4 1 0 0 0
5 13 6.333 3.827 0.466 0.66 0.693 0.231 2.7 10 2 0.533 1.95 12.14*
6 5 4 2.972 0.639 0.585 0.653 –0.12 2.55 5 1 0 0 0
7 9 5.222 3.645 0.428 0.69 0.735 0.412 2.75 9 1 0 0 0
8 10 5.222 3.634 0.558 0.699 0.741 0.175 2.77 10 1 0 0 0
9 9 5 3.36 0.467 0.582 0.624 0.186 2.52 8 1 0 0 0
10 8 5.111 3.687 0.498 0.674 0.725 0.245 2.75 8 2 0.571 2.1* 9.42*
11 11 4.889 3.389 0.619 0.624 0.66 0.016 2.57 11 1 0 0 0
12 6 4.444 3.411 0.526 0.633 0.699 0.155 2.68 6 1 0 0 0
Total 109 5.065 3.525 0.546 0.653 0.702 0.15 2.687 103 7 0.615 0.889 7.996
13 Q. glaucoides 9 5.889 3.956 0.432 0.712 0.756 0.384 4.78 6 1 0 0 0
14 10 6.556 4.312 0.574 0.735 0.774 0.213 4.93 10 2 0.2 –1.839 6.119*
15 10 5.556 4.087 0.465 0.67 0.709 0.273 4.37 10 1 0 0 0
16 9 5.556 3.606 0.44 0.663 0.702 0.334 4.35 9 1 0 0 0
17 10 5.333 3.725 0.55 0.655 0.694 0.139 4.31 8 4 0.642 0 −0.073
18 12 5.889 3.978 0.492 0.707 0.739 0.309 4.39 12 1 0 0 0
19 10 5.667 3.798 0.556 0.702 0.739 0.199 4.41 10 1 0 0 0
20 9 6.111 3.972 0.481 0.702 0.745 0.326 4.76 9 1 0 0 0
21 12 6.556 4.172 0.587 0.745 0.779 0.219 4.88 9 2 0.222 0 −0.263
22 10 5.444 3.738 0.51 0.672 0.719 0.265 4.52 10 1 0 0 0
23 — — — — — — — — 3 2 0 0 0
24 9 5.111 3.475 0.489 0.653 0.694 0.271 4.22 9 1 0 0 0
25 10 6 3.876 0.5 0.719 0.757 0.311 4.63 8 1 0 0 0
26 9 5.444 3.877 0.519 0.65 0.688 0.244 4.36 9 2 0.222 –1.088 −0.26
27 9 5.444 3.757 0.638 0.66 0.703 0.032 4.43 9 1 0 0 0
28 11 5.889 3.425 0.519 0.654 0.687 0.189 4.3 9 3 0.555 –0.382 0.909
29 10 4.889 3.399 0.325 0.634 0.668 0.445 3.91 4 1 0 0 0
30 10 5.333 3.291 0.444 0.633 0.667 0.269 4.04 4 1 0 0 0
31 10 5.667 3.739 0.578 0.683 0.72 0.122 4.33 10 2 0.466 1.033 2.052
32 10 4.889 3.435 0.5 0.635 0.672 0.186 4.02 7 1 0 0 0
33 8 3.889 2.696 0.477 0.518 0.56 0.058 3.47 10 1 0 0 0
Total 197 5.556 3.716 0.504 0.67 0.709 0.24 4.37 175 18 0.86 –0.836 −1.56
34 Q. peduncularis 7 4.222 3.365 0.532 0.657 0.712 0.157 2.65 7 1 0
35 12 6.222 4.14 0.602 0.698 0.731 0.128 2.78 8 1 0
36 4 4 3.374 0.704 0.662 0.792 –0.102 3.02 4 1 0
37 6 4.222 3.076 0.524 0.584 0.649 0.113 2.57 6 2 0.533 0.85 0.625
38 11 6.444 4.533 0.602 0.732 0.768 0.174 2.9 11 2 0.181 –1.128 −0.41
39 9 5.444 4.313 0.626 0.729 0.781 0.12 2.91 9 2 0.222 –1.362 0.671
40 8 5.444 3.435 0.46 0.687 0.734 0.319 2.74 8 1 0
41 10 5.556 3.53 0.66 0.658 0.697 –0.021 2.65 9 2 0.222 –1.088 −0.263
42 5 3.444 2.706 0.511 0.562 0.637 0.099 2.5 4 2 0.5 –0.71 1.098
43 12 6.667 4.48 0.56 0.684 0.716 0.191 2.76 10 2 0.355 0.019 1.52
44 12 5.889 4.097 0.497 0.705 0.736 0.293 2.79 12 3 0.44 −0.85 −0.724
Total 96 5.232 3.732 0.571 0.669 0.723 0.134 2.752 87 9 0.54 –0.741 –4.22
Open in a new tab

Genetic structure at plastid DNA

There was no significant differentiation among the three species for plastid DNA (F CT = 0.012, P = 0.24) (Appendix 1: Table S4). The genetic structure inferred within Q. deserticola (F ST = 0.49, P < 0.0001) and Q. peduncularis (F ST = 0.62, P < 0.0001) was relatively high, and for Q. glaucoides it was very high (F ST = 0.882, P < 0.0001). There was no evidence of phylogeographic structure for any of the species, according to the Permut results. By contrast, in Q. deserticola and Q. peduncularis, genetically more distant haplotypes are found in proximity, according to the lower values of N ST (0.572 and 0.464, respectively, for the two species) than G ST (0.726 and 0.572, respectively). Analyzing all sequences of the three species together also did not reveal phylogeographic structure (N ST = 0.723 and G ST = 0.79).

Genetic diversity and structure in the nSSR data

The micro‐checker results did not provide any evidence of genotyping errors or large allele dropout, but did suggest the presence of null alleles at several loci in the three species. The null allele frequency, estimated using FreeNA, ranged from 0 to 0.38 in Q. deserticola, from 0 to 0.35 in Q. glaucoides, and from 0 to 0.36 in Q. peduncularis. Considering the effects of the null alleles, F ST values (considering the ENA correction) for Q. deserticola, Q. glaucoides, and Q. peduncularis were 0.08, 0.033, and 0.081, respectively, while the uncorrected values were 0.074, 0.027, and 0.077. The three species showed relatively high genetic diversity values (Table 1). For Quercus deserticola, the overall value of the mean number of effective alleles (N A) was 3.53, the overall mean corrected expected heterozygosity (uHE) was 0.702, and the mean allele richness (AR) was 2.69. In Q. glaucoides, the corresponding values were N AE = 3.72, uHE = 0.709, and AR = 4.37, and in Q. peduncularis N AE = 3.73, uHE = 0.723, and AR = 2.75. The analysis with INest indicated an effect of null alleles on the estimation of inbreeding levels, since for the three species the nfb model had a lower DIC value than the fb model (6276.2 vs. 6340.5; 10,891.35 vs. 11,038.54; and 5850.107 vs. 5870.492, for Q. deserticola, Q. glaucoides, and Q. peduncularis, respectively). The revised F IS values were much lower than the estimates without correcting for null alleles and close to zero (0.28 vs. 0.03; 0.32 vs. 0.05; 0.28 vs. 0.04).

We found significant negative correlations of latitude with uHE (r = –0.62, P = 0.003) and N AE (r = –0.7, P = 0.0004) in Q. glaucoides. In Q. peduncularis, we found negative and significant correlations of uHE (r = –0.61, P = 0.05) and N AE (r = –0.75, P = 0.011) with elevation (Appendix S1: Figure S1). Nonsignificant trends of genetic diversity and geographic variables were observed in Q. deserticola. The Mantel tests indicated significant isolation by distance only for Q. glaucoides (r = 0.168, P = 0.04, Appendix S1: Figure S2).

Within the three species, population differentiation values were low but significant, with F ST = 0.069 (P < 0.0001), 0.035 (P < 0.0001), and 0.1 (P < 0.0001) in Q. deserticola, Q. glaucoides, and Q. peduncularis (Appendix S1: Table S4), respectively. Genetic differentiation among the three species was low but significant (F CT = 0.018, P < 0.0001), and the average genetic differentiation among populations within species was slightly higher (F ST = 0.032, P < 0.0001). The structure analyses indicated a most likely value of K = 2 using Evanno's method, when the three species were included together, with Q. deserticola and Q. peduncularis individuals mainly assigned to one cluster and Q. glaucoides individuals mainly assigned to the second one. However, an erroneous value of K = 2 is a well‐known phenomenon (Janes et al., 2017). Therefore, we assumed a value of K = 3, which follows the a priori species designation (Figure 3). At the intraspecific level, in Q. deserticola, the most likely K value was 3 (Figure 4). Individuals from populations 1 and 2, situated in the Oaxacan district of the Sierra Madre del Sur, had a higher proportion of assignment to the first cluster. In contrast, populations 5 and 7 in the Trans‐Mexican Volcanic Belt had the highest proportion of assignment to cluster 2, and populations of the western part of the Trans‐Mexican Volcanic Belt had a high proportion of assignment to the third cluster. Populations of the central part of the Trans‐Mexican Volcanic Belt (4, 6, and 8) imply admixture between the three clusters (Figure 4). The most likely K value for Q. glaucoides was 2, and without a clear geographic distribution pattern of the genetic groups. In Q. peduncularis, the most probable K was 2 (Figure 4), with populations distributed in the Chiapas Highlands harbouring a higher proportion of the first cluster. In contrast, the second cluster was represented in populations of the Veracruz province (41 and 43) and population 44, located in the western part of the Trans‐Mexican Volcanic Belt in the northern part of the distribution of this species.

Figure 3.

Figure 3

Open in a new tab

Proportion of genetic groups in each population and their geographic distribution for K = 3 in three Quercus species (Q. deserticola, Q. glaucoides, and Q. peduncularis) based on nine nuclear microsatellites.

Figure 4.

Figure 4

Open in a new tab

Genetic substructure in the three Quercus species. Dashed lines in the maps indicate the location of the first three main genetic breaks.

Similar genetic structure results were found in the DAPC for the three species (Appendix S1: Figure S3). At the intraspecific level, the most likely K value was 3 according to the BIC analysis for Q. deserticola and 2 for Q. peduncularis. For Q. glaucoides, the most likely K value was 10 with no clear population segregation. Those results showed patterns of genetic structure comparable to the structure analysis (Appendix S1: Figure S3).

The geographic locations of the three most important genetic discontinuities for each species, based on the Barrier results, are shown in Figure 4. For Q. deserticola, the first barrier was in the center of the Trans‐Mexican Volcanic Belt, the second one separated the Oaxacan district from the Balsas Basin, and the third discontinuity was in the western part of the Trans‐Mexican Volcanic Belt. For Q. glaucoides, the first barrier separated the San Bartolo Coro population (population 30) in the center of the Trans‐Mexican Volcanic Belt from the rest, the second barrier separated the Chichihualco population (population 17) in the Guerreran district of the Sierra Madre del Sur, and the third barrier was located in the central valleys of Oaxaca. In the case of Q. peduncularis, the first barrier was in the Isthmus of Tehuantepec, and the other two were along the Chiapas Highlands.

Demographic history

Tajima's D had a negative overall value for Q. glaucoides (D = –0.836, P = 0.22) and for Q. peduncularis (D = –0.741, P = 0.2) and a positive value for Q. deserticola (D = 0.889, P = 0.847), but none of those values were significant. At the population level, only populations 2 and 10 of Q. deserticola showed positive and significant values (Table 1). Likewise, Fu's F S was negative (F S = –1.56, P = 0.354) and nonsignificant for Q. glaucoides and negative and significant for Q. peduncularis (F S = –4.226, P = 0.028), suggesting a recent population expansion. For Q. deserticola, a positive and significant value was found (F S = 7.996, P = 0.024), suggesting a recent population bottleneck. At the population level, populations 5 and 10 of Q. deserticola and population 14 of Q. glaucoides showed positive and significant values (Table 1). The ABC results for Q. deserticola indicated a higher probability for scenario 5, a historical bottleneck at approximately 187 kya, followed by a recent expansion at 17 kya. For Q. glaucoides, a higher probability was also identified for scenario 5, with a historical bottleneck at approximately 180 kya followed by a recent expansion at 15,300 years ago. Although historical demographic changes in both species indicate the same scenario, they differed greatly in demographic values (Appendix S1: Table S5), and Q. deserticola had lower population size values (Appendix S1: Table S5). In the case of Q. peduncularis, the most probable scenario was a recent population expansion (scenario 3), with the time of the expansion at 18.7 kya, at the end of the LGM.

Changes in historical potential distribution

The ENMs for the three species showed good performance, according to values of AUC (>0.8) and AUC ratio (>1.5), meaning that the models are better than random and are statistically descriptive of the climatic niche of the three species (Appendix S1: Figure S4). The prediction of the current distribution of the three species corresponds with the areas where the species are known to occur (Valencia‐Ávalos, 2004; Valencia et al., 2017).

The current potential distribution of Q. deserticola (Figure 5) includes the Sierra Madre Occidental, the Trans‐Mexican Volcanic Belt, and the Oaxacan district of the Sierra Madre del Sur. During the MH, a larger distribution area was predicted in the same areas, while in the LGM the potential distribution in the Trans‐Mexican Volcanic Belt contracted mainly in the western part, but areas in the Oaxacan district of the Sierra Madre del Sur are predicted to have expanded. During the LIG, the distribution of the species is inferred to have been drastically reduced to areas of the Oaxacan district and the southern Sierra Madre Occidental. The climatically stable areas of this species from the present to the LIG (Figure 5) are distributed in the western part of the Trans‐Mexican Volcanic Belt and the Sierra Madre del Sur. In contrast, the elevational shifts (Figure 6) are not as drastic, although the altitudinal range during the LIG was above 1200 m a.s.l. compared with the LGM, when the species was distributed below 1000 m a.s.l.

Figure 5.

Figure 5

Open in a new tab

Potential geographic distribution of Quercus deserticola, Q. glaucoides, and Q. peduncularis during present, middle Holocene (MH), Last Glacial Maximum (LGM) and Last Interglacial (LIG), and stable areas through HM to present (pink areas), LGM to present (red areas), and LIG to present (black areas).

Figure 6.

Figure 6

Open in a new tab

Potential altitudinal distribution of Quercus deserticola, Q. glaucoides, and Q. peduncularis during the present (current), the middle Holocene (MH), Last Glacial Maximum (LGM), and Last Interglacial (LIG).

The current potential distribution of Q. glaucoides (Figure 5) comprises the Trans‐Mexican Volcanic Belt, the Sierra Madre del Sur, and the Balsas Basin. During the MH, the potential distribution area in general is predicted to have contracted, while during the LGM, suitable areas are predicted to have been located in the Balsas Basin, the western part of the Trans‐Mexican Volcanic Belt, and a part of the Sierra Madre del Sur. Distribution areas in the Oaxaca province and the eastern part of the Balsas Basin predicted in the current potential distribution are not inferred in the LGM. The potential distribution of Q. glaucoides during the LIG is inferred to have been reduced to small patches in the Sierra Madre del Sur of the Oaxacan and Guerreroan provinces, some parts of the Balsas Basin, and the western part of the Trans‐Mexican Volcanic Belt. The climatically stable areas in Q. glaucoides from the present to the LIG are inferred to have been restricted to small patches in the northwestern part of the Sierra Madre del Sur and the central part of the Balsas Basin (Figure 5). Quercus glaucoides likely had a pattern of multiple areas with stable climatic conditions. Its altitudinal range shifts through time are more accentuated than in Q. deserticola due to a stronger displacement toward lowlands in the LGM and a stronger shift to middle elevations during the LIG (Figure 6).

The current potential distribution of Q. peduncularis (Figure 5) includes some parts of the Trans‐Mexican Volcanic Belt, the Sierra Madre del Sur, the southern part of the Sierra Madre Oriental, the Chiapas Highlands, and areas in Guatemala, El Salvador, and Honduras. During the MH, an expansion of those areas is inferred to have occurred. For the LGM, the potential distribution of the species contracted in Mexico, with suitable areas predicted only in the Guerreroan and Jaliscan districts of the Sierra Madre del Sur and in the Chiapas Highlands, while the potential distribution in Central America expanded. During the LIG, the potential distribution is inferred to have been drastically reduced to small areas in the Sierra Madre del Sur, the Chiapas Highlands, and Central America. From the present to the LIG (Figure 5), the climatically stable areas occur mainly in northwestern Sierra Madre Del Sur and in the Chiapas Highlands. The altitudinal range shifted downward from the HM to the LGM; during the LIG, the altitudinal range was predicted to be at higher elevations than at present (Figure 6).

Niche conservatism and niche breadth

Niche comparison tests revealed significant niche divergence in all comparisons (Appendix S1: Table S6). The species pair with the least overlap is Q. deserticola–Q. peduncularis, while the pair with the highest overlap is Q. glaucoides–Q. peduncularis. Significant differences in all nine bioclimatic variables among oak species were found in the ANOVA (Appendix S1: Figure S5). Bio3 (isothermality), Bio8 (mean temperature of wettest quarter), and Bio9 (mean temperature of the driest quarter) differed in Q. deserticola, indicating that this species occurs at sites with lower temperatures compared to the other two. For Q. peduncularis, differences were found in Bio2 (mean diurnal range), Bio13 (precipitation of wettest month), Bio14 (precipitation of driest month), Bio18 (precipitation of warmest quarter), and Bio19 (precipitation of coldest quarter), indicating that it occurs at sites with higher precipitation. Finally, for Q. glaucoides, the Bio15 (precipitation seasonality) parameter indicates that this species is found at sites with a more marked dry season (Appendix S1: Figure S5).

The highest value of environmental niche breadth was for Q. peduncularis (0.435), followed by Q. deserticola (0.382) and Q. glaucoides (0.184). The highest geographic range breadth was for Q. glaucoides (0.385), followed by Q. peduncularis (0.345) and Q. deserticola (0.259). The correlation between niche breadth with observed heterozygosity (H O ) across 11 oak species was positive and significant (r = 0.665, P = 0.025, Appendix S1: Figure S6), and niche breadth with N ST was negative and significant (r = –0.772, P = 0.041, Appendix S1: Figure S5). We did not observe a correlation between climatic niche and geographic breadth (r = 0.231, P = 0.46) or other genetic parameters (P > 0.05).

DISCUSSION

In this study, we compared phylogeographical patterns among three closely related oaks from the Leucomexicana clade to infer the influence of climatic niche affinity and breadth, and their interactions with the effect of climatic changes during the last 130 kya, on the levels and distribution of genetic diversity.

Nuclear and plastid genetic diversity and structure

Discrepancies in the patterns of genetic diversity and structure between nuclear and cytoplasmic genomes are common not only in oak species but also more broadly in Fagaceae, as a result of incomplete lineage sorting, hybridization, and differential dispersal capacities (Zhou et al., 2022). In general, nuclear diversity is consistent with species boundaries in oaks (Cavender‐Bares et al., 2015; Kremer and Hipp, 2020), but commonly with low values of intraspecific and interspecific genetic differentiation due to extensive gene flow (Peñaloza‐Ramírez et al., 2010). Because it is generally maternally inherited, plastid DNA reflects signals of demographic changes and historical events of colonization and dispersal through seeds (Petit et al., 2005; Petit and Vendramin, 2007), and variation in this genome is not necessarily reflective of species assignment (e.g., Fazekas et al., 2009), but rather of geographic regions or major clades (Simeone et al., 2016; Pham et al., 2017; Kremer and Hipp, 2020). The nuclear microsatellite data here indicated low but significant differentiation among the three species, and genetic clusters inferred using structure and DAPC were congruent with species assignment, even though admixture was also predicted in populations from areas where species come into proximity. Conversely, shared plastid haplotypes were found among the three species, and the plastid DNA genetic structure was not congruent with species assignment. The plastid data showed high population differentiation, but without clear geographic distribution patterns. This complex distribution of plastid DNA variation could be the result of ancient hybridization, incomplete lineage sorting or both (Cavender‐Bares et al., 2015; Zhou et al., 2022), although here the retention of ancestral polymorphisms or ancient introgression seem more plausible explanations than contemporary hybridization because nuclear microsatellites did not show signals of extensive gene flow across species boundaries (Gugger and Cavender‐Bares, 2013).

Despite the sharing of common haplotypes among species, plastid DNA variation showed differing levels of structuring within species, with Q. glaucoides having the highest structure and Q. deserticola the least. These contrasts may be related to the physiographic peculiarities of the area that each species inhabits, combined with their demographic responses to historical climate changes (Petit et al., 2002). In Q. deserticola, a severe bottleneck followed by a recent demographic expansion could explain the lower genetic differentiation as found in a previous study (Rodríguez‐Gómez et al., 2018) and for other plant species (Contreras‐Negrete et al., 2021; Zorrilla‐Azcué et al., 2021). In Q. glaucoides, we found elevated plastid haplotype diversity, richness, and number of private haplotypes, and the highest genetic structure, suggesting low historical connectivity among populations through seed exchange. Thus, apparently, this species did not experience a bottleneck as severe as Q. deserticola. Quercus glaucoides probably maintained populations in multiple refugia that later served as sources of local expansions, which, coupled with a narrow climatic niche, resulted in very limited gene flow among localities (Gaytán‐Legaria et al., 2023). In Q. peduncularis, the species with the broader distribution, the levels of genetic structure at plastid DNA are moderate, which is common in temperate oak species located in the Mexican transition zone (MTZ; Ramos‐Ortíz et al., 2016; Peñaloza‐Ramírez et al., 2020).

Effects of late Pleistocene climatic changes on the potential distribution and altitudinal shifts

The climatic instability during the late Pleistocene appears to have had different effects on the biota within the MTZ, changing the abundance of some groups and producing altitudinal movements of others (Caballero et al., 2022). In the three species that we considered, we found differences in potential distribution areas and climatic niche divergence, and—as we expected—we also observed differences in potential distribution changes from the LIG to the present.

During the LIG (ca 130 kya), the distribution range of Q. deserticola may have been markedly reduced compared to the present range because the species is inferred to have occupied only 3% of the current potential distribution area. Such a geographic restriction is reflected in the decline of the effective population size inferred in the ABC analysis. The suitable areas for this species during the LIG were restricted to the Oaxacan district of the Sierra Madre del Sur and the western part of the Trans‐Mexican Volcanic Belt. Notably, the two populations located near the areas that maintained suitability are the ones showing the highest genetic diversity for the nuclear and plastid DNA data sets, suggesting that those areas acted as a climatic refugium through time. Climatic refugia for other arid‐adapted species have been found in the Tehuacán Valley in the Sierra Madre del Sur (Aguirre‐Planter et al., 2020), reinforcing the importance of climatically stable areas in maintaining biodiversity (Keppel et al., 2012). An expansion of the potential distribution area toward the Trans‐Mexican Volcanic Belt occurred with colder temperatures during the LGM. At the transition to the Holocene, this species is inferred to have reached its maximum potential distribution area and then suffered a slight reduction under the current climate. A marked range contraction of Q. deserticola inferred for the LIG is congruent with the interglacial refugia pattern noted for other plant species with xerophytic affinities that are distributed in tropical latitudes (Bonatelli et al., 2014; Perez et al., 2016; Contreras‐Negrete et al., 2021).

A contraction of the potential distribution area is also inferred in Q. glaucoides, but not as severe as in Q. deserticola. The climatically suitable areas for this species in the LIG appear to have been more widespread in the Sierra Madre del Sur and surrounding areas of the Balsas Basin. From the LGM to the present, the potential distribution area of Q. glaucoides was displaced from the Sierra Madre del Sur and the Balsas Basin to northern and eastern areas. This probable migration process might explain the decreased genetic diversity observed with latitude and the significant isolation by distance pattern.

Finally, a contraction of the potential distribution was also inferred in Q. peduncularis during the LIG, although with a less severe effect than in the other two oak species. After the LIG, it was followed by an expansion of the distribution area until the present day, with marked altitudinal displacement, which is supported by the ABC results and the association of genetic diversity indices with elevation. Altitudinal migration during Pleistocene climatic oscillations has also been noted in temperate and tropical cloud forest species (Ramírez‐Barahona and Eguiarte, 2014; Ornelas et al., 2016, 2019).

Notably, the three species were inferred to have had low suitable areas during the LIG when dry and warm conditions prevailed (Railsback et al., 2015). Congruent with these observations, the pollen records in Lake Chalco, located in central Mexico, indicate a considerable decline of Quercus pollen and a dominance of Pinus L. in the transition from glacial (140–130 kya) to interglacial conditions (~126 kya) (Lozano‐García et al., 2022).

The role of niche breadth in closely related species

We infer that Q. peduncularis is the oak species with the broadest climatic niche and larger range size in our study. In contrast, Q. deserticola and Q. glaucoides, are inferred to have narrower niches but not comparatively reduced range sizes. When several Mexican oak species were included in a regression, niche breadth (but not range size) is significantly and negatively associated with population differentiation in plastid DNA, suggesting that populations with narrower niche breadth tend to have lower connectivity among populations. We also observed a positive correlation of niche breadth with H O, confirming previous results (Gaytán‐Legaria et al., 2023) indicating that the degree of ecological specialization influences the geographic distribution, and population size, which consequently affects the patterns of genetic diversity and gene flow.

AUTHOR CONTRIBUTIONS

R.G.‐L. and A.G.‐R. conceived the ideas; R.G.‐L. and A.G.‐R. collected the data; R.G.‐L. analyzed the data; and R.G.‐L. and A.G.‐R. wrote the manuscript with the assistance of K.O. and O.R.‐S.

Supporting information

Appendix S1. Supplemental figures and tables.

Figure S1. Linear correlation of latitude and altitude per species with values of population genetic diversity.

Figure S2. Correlations between geographic and genetic distances (Nei's D) for (a) Q. deserticola, (b) Q. glaucoides and (c) Q. peduncularis.

Figure S3. Discriminant analysis of principal components for (a) all individuals, (b) Q. deserticola, (c) Q. glaucoides and (d) Q. peduncularis.

Figure S4. Receiver operating characteristic (ROC) curves for the ecological niche models of Q. deserticola, Q. glaucoides, and Q. peduncularis.

Figure S5. Boxplots of bioclimatic variables and their distribution per oak species.

Figure S6. Relationship of niche breadth with (a) coefficient of differentiation for ordered alleles (N ST /R ST) and (b) observed heterozygosity (H O).

Table S1. Geographic information of populations collected of the three Quercus species.

Table S2. Characteristics of the nine microsatellite loci used to evaluate genetic diversity in Q. deserticola, Q. glaucoides and Q. peduncularis.

Table S3. Prior distributions of 11 parameters used in the DIYABC historical demography analysis for Q. deserticola, Q. glaucoides, and Q. peduncularis.

Table S4. Analysis of molecular variance (AMOVA) based on plastid DNA and nuclear microsatellites (nSSR) among the three oak species and within species.

Table S5. Posterior distribution of historical demography parameters for selected scenarios in ABC models for each species.

Table S6. Niche comparison tests among the three oak species studies.

AJB2-112-e70122-s001.docx (1,013.5KB, docx)

ACKNOWLEDGMENTS

We thank J. Llanderal‐Mendoza for technical support in microsatellite analysis and Goretty Mendoza and Oscar De Luna for field support. We also thank Hernando Rodríguez for the samples provided for this study. Comments by two anonymous reviewers greatly contributed to improve previous versions of the manuscript. This paper constitutes a partial fulfillment of the Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de Mexico (UNAM) of R.G.‐L., who was supported by a scholarship from the Consejo Nacional de Humanidades, Ciencia y Tecnología (CONHACyT; CVU 743141). This work was funded by Consejo Nacional de Humanidades, Ciencia y Tecnología (CONHACyT 240136), Dirección General de Asuntos del Personal Académico (DGAPA)‐Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica (PAPIIT) IV201015, IN207417, IN210020, and l'Arboretum des Pouyouleix.

Gaytan‐Legaria, R. , Oyama K., Rojas‐Soto O., and González‐Rodríguez A.. 2025. Dissimilar climatic niche is predictive of contrasting historical demographic changes and altitudinal shifts in related oak species (Quercus). American Journal of Botany 112(11): e70122. 10.1002/ajb2.70122

DATA AVAILABILITY STATEMENT

The microsatellite database used in this work is available in the Dryad repository (DOI: 10.5061/dryad.37pvmcvxp). Chloroplast DNA sequences are available in GenBank (accessions PQ846109–PQ846474).

REFERENCES

  1. Aguirre‐Planter, E. , Parra‐Leyva J. G., Ramírez‐Barahona S., Scheinvar E., Lira‐Saade R., and Eguiarte L. E.. 2020. Phylogeography and genetic diversity in a southern North American Desert: Agave kerchovei from the Tehuacán‐Cuicatlán Valley, Mexico. Frontiers in Plant Science 11: 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aiello‐Lammens, M. E. , Boria R. A., Radosavljevic A., Vilela B., and Anderson R. P.. 2015. spThin: an R package for spatial thinning of species occurrence records for use in ecological niche models. Ecography 38: 541–545. [Google Scholar]
  3. Aldrich, P. R. , Jagtap M., Michler C. H., and Romero‐Severson J.. 2003. Amplification of North American red oak microsatellite markers in European white oaks and chinese chestnut. Silvae Genetica 52: 176–179. [Google Scholar]
  4. Arbogast, B. S. , and Kenagy G. J.. 2001. Comparative phylogeography as an integrative approach to historical biogeography. Journal of Biogeography 28: 819–825. [Google Scholar]
  5. Bagnoli, F. , Tsuda Y., Fineschi S., Bruschi P., Magri D., Zhelev P., Paule L., et al. 2016. Combining molecular and fossil data to infer demographic history of Quercus cerris: insights on European eastern glacial refugia. Journal of Biogeography 43: 679–690. [Google Scholar]
  6. Barthlott, W. , Lauerm W., and Placke A.. 1996. Global distribution of species diversity in vascular plants: Towards a world map of phytodiversity (globale verteilung der artenvielfalt höherer pflanzen: Vorarbeiten zu einer weltkarte der phytodiversität). Erdkunde 50: 317‐327. [Google Scholar]
  7. Bonatelli, I. A. S. , Perez M. F., Peterson A. T., Taylor N. P., Zappi D. C., Machado M. C., Koch I., et al. 2014. Interglacial microrefugia and diversification of a cactus species complex: phylogeography and palaeodistributional reconstructions for Pilosocereus aurisetus and allies. Molecular Ecology 23: 3044–3063. [DOI] [PubMed] [Google Scholar]
  8. Booth, T. H. 2022. Checking bioclimatic variables that combine temperature and precipitation data before their use in species distribution models. Austral Ecology 47: 1506–1514. [Google Scholar]
  9. Broennimann, O. , Fitzpatrick M. C., Pearman P. B., Petitpierre B., Pellissier L., Yoccoz N. G., Thuiller W., et al. 2012. Measuring ecological niche overlap from occurrence and spatial environmental data. Global Ecology and Biogeography 21: 481–497. [Google Scholar]
  10. Caballero, M. , Lozano García S., and Ortega‐Guerrero B.. 2022. Paleoenvironmental change in Central Mexico during the last 20,000 years. In Alcocer J. [ed.], Lake Alchichica limnology, 33–50. Springer Nature, Cham, Switzerland. [Google Scholar]
  11. Cabanne, G. S. , Calderón L., Trujillo Arias N., Flores P., Pessoa R., d'Horta F. M., and Miyaki C. Y.. 2016. Effects of Pleistocene climate changes on species ranges and evolutionary processes in the Neotropical Atlantic Forest. Biological Journal of the Linnean Society 119: 856–872. [Google Scholar]
  12. Castillo‐Chora, V. D. J. , Sánchez‐González L. A., Mastretta‐Yanes A., Prieto‐Torres D. A., and Navarro‐Siguënza A. G.. 2021. Insights into the importance of areas of climatic stability in the evolution and maintenance of avian diversity in the Mesoamerican dry forests. Biological Journal of the Linnean Society 132: 741–758. [Google Scholar]
  13. Cavender‐Bares, J. , Gonzalez‐Rodriguez A., Pahlich A., Koehler K., and Deacon N.. 2011. Phylogeography and climatic niche evolution in live oaks (Quercus series Virentes) from the tropics to the temperate zone. Journal of Biogeography 38: 962–981. [Google Scholar]
  14. Cavender‐Bares, J. , González‐Rodríguez A., Eaton D. A. R., Hipp A. A. L., Beulke A., and Manos P. S.. 2015. Phylogeny and biogeography of the American live oaks (Quercus subsection Virentes): a genomic and population genetics approach. Molecular Ecology 24: 3668–3687. [DOI] [PubMed] [Google Scholar]
  15. Cavender‐Bares, J. , Kothari S., Meireles J. E., Kaproth M. A., Manos P. S., and Hipp A. L.. 2018. The role of diversification in community assembly of the oaks (Quercus L.) across the continental US. American Journal of Botany 105: 565‐586. [DOI] [PubMed] [Google Scholar]
  16. Chapuis, M. P. , and Estoup A.. 2007. Microsatellite null alleles and estimation of population differentiation. Molecular Biology and Evolution 24: 621–631. [DOI] [PubMed] [Google Scholar]
  17. Chybicki, I. J. , and Burczyk J.. 2009. Simultaneous estimation of null alleles and inbreeding coefficients. Journal of Heredity 100: 106‐113. [DOI] [PubMed] [Google Scholar]
  18. Clement, M. , Posada D., and Crandall K. A.. 2000. TCS: a computer program to estimate gene genealogies. Molecular Ecology 9: 1657–1659. [DOI] [PubMed] [Google Scholar]
  19. Cobos, M. E. , Peterson A. Townsend, Barve N., and Osorio‐Olvera L.. 2019. Kuenm: an R package for detailed development of ecological niche models using Maxent. PeerJ 7: e6281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Contreras‐Negrete, G. , Letelier L., Piña‐Torres J., and González‐Rodríguez A.. 2021. Genetic structure, phylogeography and potential distribution modeling suggest a population expansion in the mesquite Prosopis laevigata since the last interglacial. Plant Systematics and Evolution 307: 22. [Google Scholar]
  21. Cornejo‐Romero, A. , Vargas‐Mendoza C. F., Aguilar‐Martínez G. F., Medina‐Sánchez J., Rendón‐Aguilar B., Valverde P. L., Zavala‐Hurtado J. A., et al. 2017. Alternative glacial–interglacial refugia demographic hypotheses tested on Cephalocereus columna‐trajani (Cactaceae) in the intertropical Mexican drylands. PLoS One 12: 1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Cornuet, J. M. , Pudlo P., Veyssier J., Dehne‐Garcia A., Gautier M., Leblois R., Marin J. M., and Estoup A.. 2014. DIYABC v2.0: A software to make approximate Bayesian computation inferences about population history using single nucleotide polymorphism, DNA sequence and microsatellite data. Bioinformatics 30: 1187–1189. [DOI] [PubMed] [Google Scholar]
  23. Dempster, A.P. , Laird M., and B Rubin, D. 1977. Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, B, Methodological 39: 138. [Google Scholar]
  24. Di Cola, V. , Broennimann O., Petitpierre B., Breiner F. T., D'Amen M., Randin C., Engler R., et al. 2017. ecospat: an R package to support spatial analyses and modeling of species niches and distributions. Ecography 40: 774–787. [Google Scholar]
  25. Doyle, J. J. , and Doyle J. L.. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15. [Google Scholar]
  26. Escobar, L. E. , Lira‐Noriega A., Medina‐Vogel G., and Townsend Peterson A.. 2014. Potential for spread of the white‐nose fungus (Pseudogymnoascus destructans) in the Americas: use of Maxent and NicheA to assure strict model transference. Geospatial Health 9: 221–229. [DOI] [PubMed] [Google Scholar]
  27. Evanno, G. , Regnaut S., and Goudet J.. 2005. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14: 2611‐2620. [DOI] [PubMed] [Google Scholar]
  28. Excoffier, L. , and Lischer H. E. L.. 2010. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10: 564–567. [DOI] [PubMed] [Google Scholar]
  29. Fazekas, A. J. , Kesanakurti P. R., Burgess K. S., Percy D. M., Graham S. W., Barrett S. C., Newmaster S. G., et al. 2009. Are plant species inherently harder to discriminate than animal species using DNA barcoding markers? Molecular Ecology Resources 9: 130‐139. [DOI] [PubMed] [Google Scholar]
  30. Feliner, G. N. 2011. Southern European glacial refugia: A tale of tales. Taxon 60: 365–372. [Google Scholar]
  31. Fick, S. E. , and Hijmans R. J.. 2017. WorldClim 2: new 1‐km spatial resolution climate surfaces for global land areas. International Journal of Climatology 37: 4302–4315. [Google Scholar]
  32. Fu, Y. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147: 915–925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Gaytán‐Legaria, R. , Oyama K., Ruiz‐Sánchez E., and González‐Rodríguez A.. 2023. The role of niche breadth in oak phylogeography: Quercus glaucoides as a study case. Journal of Biogeography 50: 1653–1667. [Google Scholar]
  34. Gent, P. R. , Danabasoglu G., Donner L. J., Holland M. M., Hunke E. C., Jayne S. R., Lawrence D. M., et al. 2011. The community climate system model version 4. Journal of Climate 24: 4973–4991. [Google Scholar]
  35. González‐Rodríguez, A. , Bain J. F., Golden J. L., and Oyama K.. 2004. Chloroplast DNA variation in the Quercus affinis–Q. laurina complex in Mexico: geographical structure and associations with nuclear and morphological variation. Molecular Ecology 13: 3467–3476. [DOI] [PubMed] [Google Scholar]
  36. Gugger, P. F. , and Cavender‐Bares J.. 2013. Molecular and morphological support for a Florida origin of the Cuban oak. Journal of Biogeography 40: 632–645. [Google Scholar]
  37. Gutiérrez‐Garíca, T. A. , and Vázquez‐Domínguez E.. 2011. Comparative phylogeography: designing studies while surviving the process. BioScience 61: 857–868. [Google Scholar]
  38. Hewitt, G. 2000. The genetic legacy of the Quaternary ice ages. Nature 405: 907–913. [DOI] [PubMed] [Google Scholar]
  39. Hewitt, G. M. 2004. Genetic consequences of climatic oscillations in the Quaternary. Philosophical Transactions of the Royal Society, B: Biological Sciences 359: 183–195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Hijmans, R. J. 2012. Cross‐validation of species distribution models: removing spatial sorting bias and calibration with a null model. Ecology 93: 679‐688. [DOI] [PubMed] [Google Scholar]
  41. Hipp, A. L. , Manos P. S., González‐Rodríguez A., Hahn M., Kaproth M., McVay J. D., Avalos S. V., and Cavender‐Bares J.. 2018. Sympatric parallel diversification of major oak clades in the Americas and the origins of Mexican species diversity. New Phytologist 217: 439–452. [DOI] [PubMed] [Google Scholar]
  42. Hipp, A. L. , Manos P. S., Hahn M., Avishai M., Bodénès C., Cavender‐Bares J., Crowl A. A., et al. 2019. Genomic landscape of the global oak phylogeny. New Phytologist 226: 1198‐1212. [DOI] [PubMed] [Google Scholar]
  43. Jakobsson, M. , and Rosenberg N. A.. 2007. CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23: 1801–1806. [DOI] [PubMed] [Google Scholar]
  44. Janes, J. K. , Miller J. M., Dupuis J. R., Malenfant R. M., Gorrell J. C., Cullingham C. I., and Andrew R. L.. 2017. The K= 2 conundrum. Molecular Ecology 26: 3594‐3602. [DOI] [PubMed] [Google Scholar]
  45. Jombart, T. 2008. adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24: 1403–1405. [DOI] [PubMed] [Google Scholar]
  46. Kampfer, S. , Lexer C., Glossl J., and Steinkellner H.. 1998. Characterization of (GA)n microsatellite loci from Quercus robur . Hereditas 129: 183–186. [Google Scholar]
  47. Keppel, G. , Van Niel K. P., Wardell‐Johnson G. W., Yates C. J., Byrne M., Mucina L., Schut A. G. T., et al. 2012. Refugia: identifying and understanding safe havens for biodiversity under climate change. Global Ecology and Biogeography 21: 393–404. [Google Scholar]
  48. Kremer, A. , and Hipp A. L.. 2020. Oaks: an evolutionary success story. New Phytologist 226: 987–1011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Levins, R. 1968. Evolution in changing environments. Monographs in population biology, 2. Princeton University Press, Princeton, NJ, USA. [Google Scholar]
  50. Li, Y. , Zhang X., and Fang Y.. 2019. Landscape features and climatic forces shape the genetic structure and evolutionary history of an Oak species (Quercus chenii) in East China. Frontiers in Plant Science 10: 1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Loera, I. , Ickert‐Bond S. M., and Sosa V.. 2017. Pleistocene refugia in the Chihuahuan Desert: the phylogeographic and demographic history of the gymnosperm Ephedra compacta . Journal of Biogeography 44: 2706–2716. [Google Scholar]
  52. Lozano‐García, S. , Torres‐Rodríguez E., Figueroa‐Rangel B., Caballero M., Sosa‐Nájera S., Ortega‐Guerrero B., and Acosta‐Noriega C.. 2022. Vegetation history of a Mexican Neotropical basin from the late MIS 6 to early MIS 3: the pollen record of Lake Chalco. Quaternary Science Reviews 297: 107830. [Google Scholar]
  53. Manni, F. , Guérard E., and Heyer E.. 2004. Geographic patterns of (genetic, morphologic, linguistic) variation: how barriers can be detected by using Monmonier's algorithm. Human Biology 76: 173–190. [DOI] [PubMed] [Google Scholar]
  54. Marshall, C. J. , and Liebherr J. K.. 2000. Cladistic biogeography of the Mexican transition zone. Journal of Biogeography 27: 203‐216. [Google Scholar]
  55. Mastretta‐Yanes, A. , Moreno‐Letelier A., Piñero D., Jorgensen T. H., and Emerson B. C.. 2015. Biodiversity in the Mexican highlands and the interaction of geology, geography and climate within the Trans‐Mexican Volcanic Belt. Journal of Biogeography 42: 1586–1600. [Google Scholar]
  56. Morrone, J. J. 2020. The Mexican Transition Zone: a natural biogeographic laboratory to study biotic assembly. Springer Nature. Switzerland AG. [Google Scholar]
  57. Müller, J. , Müller K., Neinhuis C., and Quandt D.. 2010. PhyDE. Website: www.Phyde.de/
  58. Nixon, K. C. 2006. Global and neotropical distribution and diversity of oak (genus Quercus) and oak forests. In Kappelle M. [ed.], Ecology and conservation of neotropical montane oak forests, 3–13. Springer‐Verlag, Berlin, Germany. [Google Scholar]
  59. Ornelas, J. F. , Ortiz‐Rodriguez A. E., Ruiz‐Sanchez E., Sosa V., and Pérez‐Farrera M. Á.. 2019. Ups and downs: genetic differentiation among populations of the Podocarpus (Podocarpaceae) species in Mesoamerica. Molecular Phylogenetics and Evolution 138: 17–30. [DOI] [PubMed] [Google Scholar]
  60. Ornelas, J. F. , Gándara E., Vásquez‐Aguilar A. A., Ramírez‐Barahona S., Ortiz‐Rodriguez A. E., González C., Saules M. T. M., and Ruiz‐Sanchez E.. 2016. A mistletoe tale: postglacial invasion of Psittacanthus schiedeanus (Loranthaceae) to Mesoamerican cloud forests revealed by molecular data and species distribution modeling. BMC Evolutionary Biology 16: 1–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Ornelas, J. F. , Licona‐Vera Y., and Vásquez‐Aguilar A. A.. 2018. Genetic differentiation and fragmentation in response to climate change of the narrow endemic Psittacanthus auriculatus . Tropical Conservation Science 11: 1–15. [Google Scholar]
  62. Peakall, R. , and Smouse P. E.. 2012. GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research‐an update. Bioinformatics 28: 2537–2539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Peñaloza‐Ramírez, J. M. , González‐Rodríguez A., Mendoza‐Cuenca L., Caron H., Kremer A., and Oyama K.. 2010. Interspecific gene flow in a multispecies oak hybrid zone in the Sierra Tarahumara of Mexico. Annals of Botany 105: 389–399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Peñaloza‐Ramírez, J. M. , Rodríguez‐Correa H., González‐Rodríguez A., Rocha‐Ramírez V., and Oyama K.. 2020. High genetic diversity and stable Pleistocene distributional ranges in the widespread Mexican red oak Quercus castanea Née (1801) (Fagaceae). Ecology and Evolution 10: 4204–4219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Pennington, R. T. , Lavin M., Prado D. E., Pendry C. A., Pell S. K., and Butterworth C. A.. 2004. Historical climate change and speciation: neotropical seasonally dry forest plants show patterns of both Tertiary and Quaternary diversification. Philosophical Transactions of the Royal Society, B: Biological Sciences 359: 515–538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Perez, M. F. , Bonatelli I. A. S., Moraes E. M., and Carstens B. C.. 2016. Model‐based analysis supports interglacial refugia over long‐dispersal events in the diversification of two South American cactus species. Heredity 116: 550–557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Petit, R. J. , and Vendramin G. G.. 2007. Plant phylogeography based on organelle genes: an introduction. In Weiss S. and Ferrand N. [eds.], Phylogeography of southern European refugia, 23–97. Springer, Dordrecht, Netherlands. [Google Scholar]
  68. Petit, R. J. , Duminil J., Fineschi S., Hampe A., Salvini D., and Vendramin G. G.. 2005. Comparative organization of chloroplast, mitochondrial and nuclear diversity in plant populations. Molecular Ecology 14: 689–701. [DOI] [PubMed] [Google Scholar]
  69. Petit, R. J. , Csaikl U. M., Bordács S., Burg K., Coart E., Cottrell J., Van Dam B., et al. 2002. Chloroplast DNA variation in European white oaks. Phylogeography and patterns of diversity based on data from over 2600 populations. Forest Ecology and Management 156: 5–26. [Google Scholar]
  70. Pham, K. K. , Hipp A. L., Manos P. S., and Cronn R. C.. 2017. A time and a place for everything: phylogenetic history and geography as joint predictors of oak plastome phylogeny. Genome 60: 720–732. [DOI] [PubMed] [Google Scholar]
  71. Phillips, S. J. , and Dudík M.. 2008. Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography 31: 161–175. [Google Scholar]
  72. Pons, O. , and Petit R. J.. 1996. Measuring and testing genetic differentiation with ordered versus unordered alleles. Genetics 144: 1237–1245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Pritchard, J. K. , Stephens M., and Donnelly P.. 2000. Inference of population structure using multilocus genotype data. Genetics 155: 945–959. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. R Core Team . 2014. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Website: https://www.r-project.org [Google Scholar]
  75. Railsback, L. B. , Brook G. A., Ellwood B. B., Liang F., Cheng H., and Edwards R. L.. 2015. A record of wet glacial stages and dry interglacial stages over the last 560 kyr from a standing massive stalagmite in Carlsbad Cavern, New Mexico, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 438: 256–266. [Google Scholar]
  76. Ramírez‐Barahona, S. , and Eguiarte L. E.. 2014. Changes in the distribution of cloud forests during the last glacial predict the patterns of genetic diversity and demographic history of the tree fern Alsophila firma (Cyatheaceae). Journal of Biogeography 41: 2396–2407. [Google Scholar]
  77. Ramos‐Ortíz, S. , Oyama K., Rodríguez‐Correa H., and González‐Rodríguez A.. 2016. Geographic structure of genetic and phenotypic variation in the hybrid zone between Quercus affinis and Q. laurina in Mexico. Plant Species Biology 31: 219–232. [Google Scholar]
  78. Rivera‐Ortíz, F. A. , Sanabria‐Urbán S., Prieto‐Torres D. A., Navarro‐Sigüenza A. G., Arizmendi M. del C., and Oyama K.. 2023. Phylogeography of Ara militaris (Military Macaw): implications for conservation. Diversity 15: 1035. [Google Scholar]
  79. Rodríguez‐Gómez, F. , Oyama K., Ochoa‐Orozco M., Mendoza‐Cuenca L., Gaytán‐Legaria R., and González‐Rodríguez A.. 2018. Phylogeography and climate‐associated morphological variation in the endemic white oak Quercus deserticola (Fagaceae) along the Trans‐Mexican Volcanic Belt. Botany 96: 121–133. [Google Scholar]
  80. Rodríguez‐Correa, H. , Oyama K., MacGregor‐Fors I., and González‐Rodríguez A.. 2015. How are oaks distributed in the Neotropics? A perspective from species turnover, areas of endemism, and climatic niches. International Journal of Plant Sciences 176: 222–231. [Google Scholar]
  81. Rojas‐Soto, O. , Forero‐Rodríguez J. S., Galindo‐Cruz A., Mota‐Vargas C., Parra‐Henao K. D., Peña‐Peniche A., Piña‐Torres J., et al. 2024. Calibration areas in ecological niche and species distribution modelling: unravelling approaches and concepts. Journal of Biogeography 51: 1416‐1428. [Google Scholar]
  82. Ruiz‐Sanchez, E. , Rodriguez‐Gomez F., and Sosa V.. 2012. Refugia and geographic barriers of populations of the desert poppy, Hunnemannia fumariifolia (Papaveraceae). Organisms Diversity and Evolution 12: 133–143. [Google Scholar]
  83. Sandel, B. , Dalsgaard B., Davies R. G., Gaston K. J., Sutherland W. J., and Svenning J.‐C.. 2011. The influence of late Quaternary climate‐change velocity on species endemism. Science 334: 660–664. [DOI] [PubMed] [Google Scholar]
  84. Schoener, T. W. 1968. Anolis lizards of Bimini: resource partitioning in a complex fauna. Ecology 49: 704–726. [Google Scholar]
  85. Schönswetter, P. , Stehlik I., Holderegger R., and Tribsch A.. 2005. Molecular evidence for glacial refugia of mountain plants in the European Alps. Molecular Ecology 14: 3547–3555. [DOI] [PubMed] [Google Scholar]
  86. Shaw, J. , Lickey E. B., Beck J. T., Farmer S. B., Liu W., Miller J., Siripun K. C., et al. 2005. The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany 92: 142–166. [DOI] [PubMed] [Google Scholar]
  87. Simeone, M. C. , Grimm G. W., Papini A., Vessella F., Cardoni S., Tordoni E., Piredda R., et al. 2016. Plastome data reveal multiple geographic origins of Quercus group Ilex . PeerJ 4: e1897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  88. Simmons, M. P. , and Ochoterena H.. 2000. Gaps as characters in sequence‐based phylogenetic analyses. Systematic Biology 49: 369‐381. [PubMed] [Google Scholar]
  89. Sosa, V. , Ruiz‐Sanchez E., and Rodriguez‐Gomez F. C.. 2009. Hidden phylogeographic complexity in the Sierra Madre Oriental: the case of the Mexican tulip poppy Hunnemannia fumariifolia (Papaveraceae). Journal of Biogeography 36: 18–27. [Google Scholar]
  90. Sosa, V. , Ornelas J. F., Ramírez‐Barahona S., and Gándara E.. 2016. Historical reconstruction of climatic and elevation preferences and the evolution of cloud forest‐adapted tree ferns in Mesoamerica. PeerJ 4: e2696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  91. van Steenderen, C. J. , and Sutton G. F.. 2024. Climate covariate selection influences MaxEnt model predictions and predictive accuracy under current and future climates. Ecological Modelling 498: 110872. [Google Scholar]
  92. Steinkellner, H. , Fluch S., Turetschek E., Lexer C., Streiff R., Kremer A., Burg K., and Glössl J.. 1997. Identification and characterization of (GA/CT)n‐microsatellite loci from Quercus petraea . Plant Molecular Biology 33: 1093–1096. [DOI] [PubMed] [Google Scholar]
  93. Tajima, F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585–595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  94. Tovar‐Sánchez, E. , Mussali‐Galante P., Esteban‐Jiménez R., Piñero D., Arias D. M., Dorado O., and Oyama K.. 2008. Chloroplast DNA polymorphism reveals geographic structure and introgression in the Quercus crassifolia × Quercus crassipes hybrid complex in Mexico. Botany 86: 228–239. [Google Scholar]
  95. Valencia‐Ávalos, S. 2004. Diversidad del género Quercus (Fagaceae) en México. Botanical Sciences 75: 33‐53. [Google Scholar]
  96. Valencia, A. S. , Flores‐Franco G., Jiménez‐Ramírez J., and Mora‐Jarvio M.. 2017. Distribution and diversity of Fagaceae in Hidalgo, Mexico. Botanical Sciences 95: 660–721. [Google Scholar]
  97. Van der Vaart, A. W. 1998. Asymptotic statistics. Cambridge University Press, Cambridge, UK. [Google Scholar]
  98. Van Els, P. , Zarza E., Rocha Moreira L., Gómez‐Bahamón V., Santana A., Aleixo A., Ribas C. C., et al. 2021. Recent divergence and lack of shared phylogeographic history characterize the diversification of neotropical savanna birds. Journal of Biogeography 48: 1124–1137. [Google Scholar]
  99. Van Oosterhout, C. , Hutchinson W. F., Wills D. P. M., and Shipley P.. 2004. micro‐checker: Software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes 4: 535–538. [Google Scholar]
  100. Warren, D. L. , Beaumont L. J., Dinnage R., and Baumgartner J. B.. 2018. New methods for measuring ENM breadth and overlap in environmental space. Ecography 42: 444–446. [Google Scholar]
  101. Warren, D. L. , Matzke N. J., Cardillo M., Baumgartner J. B., Beaumont L. J., Turelli M., Glor R. E., et al. 2021. ENMTools 1.0: an R package for comparative ecological biogeography. Ecography 44: 504–511. [Google Scholar]
  102. Weir, B. S. 1996. Genetic data analysis II. Sinauer, Sunderland, MA, USA. [Google Scholar]
  103. Wickham, H. 2016. ggplot2: Elegant graphics for data analysis. Springer‐Verlag, NY, NY, USA. Website: https://ggplot2.tidyverse.org [Google Scholar]
  104. Zhou, B. F. , Yuan S., Crowl A. A., Liang Y. Y., Shi Y., Chen X. Y., An Q. Q., et al. 2022. Phylogenomic analyses highlight innovation and introgression in the continental radiations of Fagaceae across the Northern Hemisphere. Nature Communications 13: 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  105. Zorrilla‐Azcué, S. , González‐Rodríguez A., Oyama K., González M. A., and Rodríguez‐Correa H.. 2021. The DNA history of a lonely oak: Quercus humboldtii phylogeography in the Colombian Andes. Ecology and Evolution 11: 6814–6828. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix S1. Supplemental figures and tables.

Figure S1. Linear correlation of latitude and altitude per species with values of population genetic diversity.

Figure S2. Correlations between geographic and genetic distances (Nei's D) for (a) Q. deserticola, (b) Q. glaucoides and (c) Q. peduncularis.

Figure S3. Discriminant analysis of principal components for (a) all individuals, (b) Q. deserticola, (c) Q. glaucoides and (d) Q. peduncularis.

Figure S4. Receiver operating characteristic (ROC) curves for the ecological niche models of Q. deserticola, Q. glaucoides, and Q. peduncularis.

Figure S5. Boxplots of bioclimatic variables and their distribution per oak species.

Figure S6. Relationship of niche breadth with (a) coefficient of differentiation for ordered alleles (N ST /R ST) and (b) observed heterozygosity (H O).

Table S1. Geographic information of populations collected of the three Quercus species.

Table S2. Characteristics of the nine microsatellite loci used to evaluate genetic diversity in Q. deserticola, Q. glaucoides and Q. peduncularis.

Table S3. Prior distributions of 11 parameters used in the DIYABC historical demography analysis for Q. deserticola, Q. glaucoides, and Q. peduncularis.

Table S4. Analysis of molecular variance (AMOVA) based on plastid DNA and nuclear microsatellites (nSSR) among the three oak species and within species.

Table S5. Posterior distribution of historical demography parameters for selected scenarios in ABC models for each species.

Table S6. Niche comparison tests among the three oak species studies.

AJB2-112-e70122-s001.docx (1,013.5KB, docx)

Data Availability Statement

The microsatellite database used in this work is available in the Dryad repository (DOI: 10.5061/dryad.37pvmcvxp). Chloroplast DNA sequences are available in GenBank (accessions PQ846109–PQ846474).

Articles from American Journal of Botany are provided here courtesy of Wiley

RESOURCES