TY - JOUR
AU - Cavender-Bares,, Jeannine
AB - Abstract Heritable variation in polygenic (quantitative) traits is critical for adaptive evolution and is especially important in this era of rapid climate change. In this study, we examined the levels of quantitative genetic variation of populations of the tropical tree Quercus oleoides Cham. and Schlect. for a suite of traits related to resource use and drought resistance. We tested whether quantitative genetic variation differed across traits, populations and watering treatments. We also tested potential evolutionary factors that might have shaped such a pattern: selection by climate and genetic drift. We measured 15 functional traits on 1322 1-year-old seedlings of 84 maternal half-sib families originating from five populations growing under two watering treatments in a greenhouse. We estimated the additive genetic variance, coefficient of additive genetic variation and narrow-sense heritability for each combination of traits, populations and treatments. In addition, we genotyped a total of 119 individuals (with at least 20 individuals per population) using nuclear microsatellites to estimate genetic diversity and population genetic structure. Our results showed that gas exchange traits and growth exhibited strikingly high quantitative genetic variation compared with traits related to leaf morphology, anatomy and photochemistry. Quantitative genetic variation differed between populations even at geographical scales as small as a few kilometers. Climate was associated with quantitative genetic variation, but only weakly. Genetic structure and diversity in neutral markers did not relate to coefficient of additive genetic variation. Our study demonstrates that quantitative genetic variation is not homogeneous across traits and populations of Q. oleoides. More importantly, our findings suggest that predictions about potential responses of species to climate change need to consider population-specific evolutionary characteristics. Introduction Current climate change is predicted to cause many species to suffer demographic declines or become locally extinct unless they are able to migrate or to adapt in situ to new conditions (Davis and Shaw 2001, Chen et al. 2011). In order for populations to track climate change and heterogeneous environmental conditions via evolution by natural selection, heritable variation in polygenic (quantitative) traits must be present (Lynch and Hill 1986, Pease et al. 1989, Falconer et al. 1996, Lynch and Walsh 1998, Etterson 2004). We use the term ‘quantitative genetic variation’ to refer to genetic variation in quantitative traits that is quantified based on the phenotypic resemblance among relatives. Populations and species might differ in the degree of quantitative genetic variation and such differences are usually attributed to evolutionary forces such as genetic drift and natural selection (Conner et al. 2003, Sherrard et al. 2009, Brousseau et al. 2013, Miranda-Jácome et al. 2014, Stock et al. 2014). Genetic drift and other stochastic processes are hypothesized to erode quantitative genetic variation in small populations with restricted gene flow (Billington 1991, Ellstrand and Elam 1993, White et al. 2007, Shaw and Etterson 2012). The magnitude of these effects is observable using neutral molecular markers (Eckert et al. 2008, van Heerwaarden et al. 2009, Hoffmann and Sgrò 2011, Berger et al. 2013). In some instances, genetic diversity is curtailed in a way that is detectable by both neutral markers and quantitative genetic analyses (Hewitt 1996, Svenning and Skov 2005, Hatziskakis et al. 2011, Linares 2011). For example, in a study on Quercus suber, Ramírez-Valiente et al. (2014a) showed that traits such as branchiness and growth architecture were associated with the cpDNA lineage of the population. Limited genetic variation within species has been argued to be a widespread phenomenon (Stone et al. 2011). However, in general, patterns identified in neutral or weakly selected loci usually differ from those observed from quantitative traits (Pfrender et al. 2000, Fraser and Bernatchez 2001, Reed and Frankham 2001, Sagnard et al. 2002, Ouborg et al. 2006, Frankham 2010). Natural selection is considered to be the primary factor that shapes the levels of quantitative genetic variation in natural populations. It is generally accepted that directional and stabilizing selection erodes quantitative genetic variation in traits that are closely associated with fitness (Fisher 1930) and examples of this process have been found in many organisms (Mousseau and Roff 1987, Stearns 1992, Futuyma 2009, Flatt and Heyland 2011, Sniegula et al. 2016). Strong selection exerted by climatic change is expected to decrease quantitative genetic variation in traits important for climate change adaptation (i.e., functional traits), and as a result, limit the evolutionary potential of species (Armbruster et al. 1998, Hewitt 2000, Davis and Shaw 2001, Etterson and Shaw 2001, Etterson 2004). Predictions about levels of quantitative genetic variation in a given population are complex since the phenotypic expression of genetic variation is also environment- and trait-dependent. As an example of this complexity, previous studies have shown that novel or stressful environments can trigger the expression of cryptic genetic variation caused by deleterious alleles or stress-sensitive mutations that have not been removed by natural selection (Kondrashov and Houle 1994, Rutherford and Lindquist 1998, Collins and Bell 2004, Le Rouzic and Carlborg 2008). Conversely, other studies have observed that different genotypes respond similarly under stressful conditions whereas they differ when exposed to more favorable conditions (Bubliy et al. 2001, Wilson et al. 2006, Kristensen et al. 2015). Moreover, populations that experience unpredictable environmental heterogeneity may also evolve towards an inflexibility of phenotypic expression in response to environmental variation (i.e., canalization) (Valladares et al. 2002). In this study, we tested for differences in quantitative genetic variation among functional traits, populations and watering treatments in seedlings of the neotropical oak, Quercus oleoides Cham. and Schlect. This species inhabits seasonally dry tropical forests from Costa Rica to Mexico. Briefly, a greenhouse experiment was conducted in which seedlings of five populations that vary in rainfall patterns were grown under two contrasting watering treatments, as reported in Ramírez-Valiente and Cavender-Bares (2017). We used data on growth rates, specific leaf area, leaf lamina area, leaf thickness and gas exchange from the prior study, and measured a suite of new morphological, physiological and anatomical traits related to resource use and drought resistance for the current study. These traits include vein length per unit leaf area (VLA), stomatal density, stomatal length, stomatal pore index (SPI), non-photochemical quenching (NPQ) and electron transport rate (ETR). Leaf thickness and SLA, which is the projected area per unit dry weight, are related to photosynthetic capacity, growth rates and resistance to desiccation (Reich et al. 1997, Wright et al. 2004). The VLA is related to leaf hydraulic conductance and rates of gas exchange per unit leaf area (Sack and Scoffoni 2013) and is also associated with stomatal traits such as density of stomata, stomatal length and stomatal pore index, which collectively influence leaf hydraulic conductance (Sack et al. 2003). Non-photochemical quenching is a good indicator of heat dissipation of excitation energy in the antenna system in photosystem II (PSII), a mechanism that is important under unfavorable conditions such as drought (Cavender-Bares and Bazzaz 2004). Electron transport rate is related to CO2 fixation rates and growth rates in oaks (Cavender-Bares and Bazzaz 2004). Previous studies have demonstrated the adaptive value of these traits for several plant species including oaks when growing under different water availability (Dudley 1996, Donovan et al. 2007, Ramírez-Valiente et al. 2014b, 2015a). A strong pattern of population divergence has been previously reported in Q. oleoides for growth and leaf morphology in seedlings growing under contrasting watering regimes (Ramírez-Valiente et al. 2015b, 2017, Ramírez-Valiente and Cavender-Bares 2017). Climate and particularly the index of moisture of the populations were suggested as the main selective factors driving population differentiation (Ramírez-Valiente et al. 2017, 2018). Here, we extend the analyses to examine differences among populations in levels of quantitative genetic variation, which we have not done previously. We hypothesized that traits demonstrated to be under long-term selection (growth, SLA, leaf thickness, leaf area) would exhibit lower quantitative genetic variation as a result of strong selective pressures (Ramírez-Valiente et al. 2018). We further expected that traits that show greater plasticity in response to environmental variation, such as gas exchange traits, would maintain high levels of quantitative genetic variation within populations compared with other traits (Thuiller et al. 2008, Lande 2009, Nicotra et al. 2010). Our rationale is that plastic physiological adjustments that confer adaptive advantages under environmental stress would buffer individuals from selection and, consequently, reduce the amount of genetic variation that is culled from the population (Brock and Weinig 2007, Gage et al. 2017). This expectation would only hold if all genotypes have the same plastic response (i.e., no genotype by environment interaction, see also comments below). We also expected that populations of Q. oleoides would exhibit different levels of quantitative genetic variation (Donohue et al. 2000). Specifically, we expected that harsh environmental conditions such as low precipitation and high temperatures during the dry season would produce strong selection that would reduce quantitative genetic variation within populations (Sgrò and Hoffmann 1998, Husby et al. 2011). In addition, we expected that temporally fluctuating selection, as occurs in regions with high climate seasonality, would favor the maintenance of quantitative genetic (Barton and Keightley 2002). This hypothesis was based on previous findings of contrasting trait–fitness relationships in dry and wet seasons in Q. oleoides seedlings growing in field conditions (Ramírez-Valiente et al. 2017). In addition, we tested whether genetic drift could have played a role in determining the levels of quantitative genetic variation within populations. Quercus oleoides exhibits a complex evolutionary history with a clear genetic disjunction between southern and central populations based on neutral genetic markers (Cavender‐Bares et al. 2011, 2015). Although Q. oleoides harbors high levels of genetic diversity throughout its whole distribution range, populations isolated and restricted to a small area in Costa Rica exhibit lower molecular genetic diversity than populations from the rest of the distribution range (Cavender‐Bares et al. 2011, 2015). If Costa Rica populations exhibit reduced levels of neutral genetic and quantitative genetic variation, this would indicate an important role of genetic drift and demographic processes on shaping quantitative genetic variation within populations. Finally, we examined whether the amount of quantitative genetic variation that was expressed differed when plants were exposed to contrasting water availability. On one hand, natural populations of Q. oleoides are subjected to both dry and rainy seasons suggesting that watering treatments would not strongly impact expression of quantitative genetic variation. On the other hand, our previous study revealed that some traits, particularly growth, exhibited genotype by environment interactions in response to the watering treatments (Ramírez-Valiente et al. 2018). Thus, an impact of the watering treatments on the expressed variation could be also expected. Elucidating how quantitative genetic variation is structured across populations and traits and how environment affects its expression is essential to understanding the evolutionary potential of species in response to ongoing climate change (Etterson 2000, 2004, Davis et al. 2005, Johnson and Barton 2005). Materials and methods Seed collection and experimental design Acorns were collected in early 2013 from 101 open-pollinated mother trees randomly selected within five populations of Q. oleoides (Table 1). Mother trees were separated by more than 150 m, which is sufficient distance to avoid family structure (Deacon and Cavender-Bares 2015). The focal populations represent a wide range of the precipitation and soil moisture availabilities found within the species distribution (Table 1, Figure 1). Acorns were stored at 4 °C until synchronous sowing in greenhouse facilities at the University of Minnesota in June 2013. A total of 12,160 acorns (11–96 per maternal family) were randomly sown in deep pots with mix of 60% LC8 growing mix (70–80% Canadian Sphagnum peat moss, 20–25% perlite, 5–10% vermiculite, Sushine®, Sun Gro Horticulture, Bellevue, WA), 20% perlite and 20% vermiculite. Greenhouse temperatures were set at tropical conditions (28 °C-daytime average, 18 °C-nighttime average). In October 2013, 84 families (10–25 per maternal family depending on availability) were randomly selected for transplantation into 6.2 dm3 pots with a 50% mix of sand and LC8 growing mix soil (Table 1). Table 1. Latitude (Northern hemisphere), longitude (Western hemisphere), altitude and climatic characterization of the five focal populations of this study. n is the number of open-pollinated families sampled, P is precipitation and T is temperature. Code Population n Latitude Longitude Altitude (m) P (mm) T (°C) TA Las Tablas 18 14° 00′ 25′ 87° 02′ 04′ 933 1014 22.2 MZ Macuelizo 10 13° 59′ 51′ 87° 02′ 41′ 1030 1017 21.7 SG Sabana Grande 12 13° 48′ 54′ 87° 14′ 55′ 1104 1185 21.2 SE Santa Elena 26 10° 53′ 30′ 85° 34′ 54′ 278 1776 24.9 RI Rincón 19 10° 46′ 42′ 85° 21′ 36′ 779 2683 22.0 Code Population n Latitude Longitude Altitude (m) P (mm) T (°C) TA Las Tablas 18 14° 00′ 25′ 87° 02′ 04′ 933 1014 22.2 MZ Macuelizo 10 13° 59′ 51′ 87° 02′ 41′ 1030 1017 21.7 SG Sabana Grande 12 13° 48′ 54′ 87° 14′ 55′ 1104 1185 21.2 SE Santa Elena 26 10° 53′ 30′ 85° 34′ 54′ 278 1776 24.9 RI Rincón 19 10° 46′ 42′ 85° 21′ 36′ 779 2683 22.0 Table 1. Latitude (Northern hemisphere), longitude (Western hemisphere), altitude and climatic characterization of the five focal populations of this study. n is the number of open-pollinated families sampled, P is precipitation and T is temperature. Code Population n Latitude Longitude Altitude (m) P (mm) T (°C) TA Las Tablas 18 14° 00′ 25′ 87° 02′ 04′ 933 1014 22.2 MZ Macuelizo 10 13° 59′ 51′ 87° 02′ 41′ 1030 1017 21.7 SG Sabana Grande 12 13° 48′ 54′ 87° 14′ 55′ 1104 1185 21.2 SE Santa Elena 26 10° 53′ 30′ 85° 34′ 54′ 278 1776 24.9 RI Rincón 19 10° 46′ 42′ 85° 21′ 36′ 779 2683 22.0 Code Population n Latitude Longitude Altitude (m) P (mm) T (°C) TA Las Tablas 18 14° 00′ 25′ 87° 02′ 04′ 933 1014 22.2 MZ Macuelizo 10 13° 59′ 51′ 87° 02′ 41′ 1030 1017 21.7 SG Sabana Grande 12 13° 48′ 54′ 87° 14′ 55′ 1104 1185 21.2 SE Santa Elena 26 10° 53′ 30′ 85° 34′ 54′ 278 1776 24.9 RI Rincón 19 10° 46′ 42′ 85° 21′ 36′ 779 2683 22.0 Figure 1. View largeDownload slide (a) Distribution of Quercus oleoides (shaded green area) based on our collection herbarium records (green triangles). (b) Location of the sampled populations established in the greenhouse experiment at the University of Minnesota. (c) Monthly index of moisture (Im) averaged across maternal families for the five studied populations (data from WorldClim database; Hijmans et al. 2005). Index of moisture was calculated as Im = P − PET, where P is annual precipitation and PET potential evapotranspiration (see text for details). Values below zero indicate water deficit. TA: Las Tablas (Honduras), MZ: Macuelizo (Honduras), SG: Sabana Grande (Honduras), SE: Santa Elena (Costa Rica), RI: Rincón de la Vieja (Costa Rica). Figure 1. View largeDownload slide (a) Distribution of Quercus oleoides (shaded green area) based on our collection herbarium records (green triangles). (b) Location of the sampled populations established in the greenhouse experiment at the University of Minnesota. (c) Monthly index of moisture (Im) averaged across maternal families for the five studied populations (data from WorldClim database; Hijmans et al. 2005). Index of moisture was calculated as Im = P − PET, where P is annual precipitation and PET potential evapotranspiration (see text for details). Values below zero indicate water deficit. TA: Las Tablas (Honduras), MZ: Macuelizo (Honduras), SG: Sabana Grande (Honduras), SE: Santa Elena (Costa Rica), RI: Rincón de la Vieja (Costa Rica). A greenhouse experiment was set up following a randomized complete block design with a total of seven blocks. Seedlings were grown for 6 months (October 2013–March 2014) under well-watered conditions and maintained at tropical temperatures and 12 h photoperiod. At this point, two watering treatments were implemented: well-watered and dry. Half of the plants per family were subjected to 22–30% volumetric soil moisture (well-watered treatment) whereas volumetric soil moisture was reduced in the other half until reaching 9–13% for the duration of the experiment (dry treatment). Plants were maintained under these conditions for 100 days. On average, predawn leaf water potential (ψPD) of seedlings was −0.42 ± 0.04 MPa in the well-watered treatment and −1.18 ± 0.04 MPa in the dry treatment, indicating that watering treatments were effective in controlling the water status of plants (see Ramírez-Valiente and Cavender-Bares 2017 for more experimental details). Quantitative traits Growth We measured stem height, basal diameter, and number of leaves in all plants of the experiment at the beginning (t = 0 days) and end (t = 100 days) of the watering treatments. We used these measurements to estimate initial (t = 0) and final total biomass (t = 100) using allometric equations constructed by harvesting a subsample of plants throughout the experiment (Ramírez-Valiente and Cavender-Bares 2017). Biomass growth (g) was calculated as (Mfinal − Minitial) where Mfinal is the estimated biomass at the final of the experiment and Minitial is the estimated biomass at the beginning of the experiment. Chlorophyll fluorescence and gas exchange Mid-experiment, 45–55 days after initiating the watering treatments, dark- and light-acclimated chlorophyll fluorescence measurements were performed on sun leaves of three to eight seedlings per maternal family per treatment (n = 1195 plants). Measurements were taken in the youngest fully expended leaves at solar noon with a photosynthetically active radiation (PAR) of 1500 μmol m−2 s−1. A PAR value of 1500 μmol m−2 s−1 was used to match maximum ambient light levels in the greenhouse. Dark-acclimated measurements were taken at predawn (4:00–5:30 a.m.). Saturating light for both dark- and light-acclimated measurements was set at 2500 μmol m−2 s−1 although saturating level was lower according to previous tests. For further analysis, we calculated: Stern-Volmer non-photochemical quenching, NPQ [(Fm – Fm′)/Fm′] and electron transport rate, ETR (assuming a leaf absorbance of 0.8 and equal photon excitation of PSII and PSI, Cavender-Bares and Bazzaz 2004). During the same period, we measured gas exchange (photosynthesis and stomatal conductance) in three to five seedlings per maternal family within each treatment (n = 695 plants). Measurements were taken from 10:30 a.m.–2:00 p.m. on seven sunny days, using a Licor 6400 (LI-COR, Lincoln, NE, USA). Measurements were logged only when stomatal steady state was achieved (i.e., when photosynthesis and stomatal conductance remained stable; Coefficient of Variation <1). Leaves were exposed to an atmospheric CO2 concentration of 400 p.p.m., relative humidity of 45–65% and saturating light of 1500 μmol m–2 s–1. Photosynthetic rate and stomatal conductance were determined on both an area and mass basis using SLA data for the same leaves (see details below). Intrinsic water-use efficiency was also calculated by dividing the photosynthetic rate by stomatal conductance (A/gs). Leaf morphology The youngest fully expanded leaf that developed after the watering treatments were imposed was collected between Days 45 and 55 of the experiment (n = 176 plants). Thickness was measured using a micrometer (accuracy ±0.004 mm). Fresh leaves were scanned to obtain leaf size and main vein length and then dried at 60 °C for 7 days to estimate dry leaf mass. For further analyses, we calculated two important functional morphological traits for plants: specific leaf area (SLA) and vein length per unit leaf area (VLA) as leaf area/leaf mass and main vein length/leaf area, respectively (Reich et al. 1997, Sack and Scoffoni 2013). Stomatal analyses were performed in dried leaves on impressions from abaxial nail varnish peels because Q. oleoides is hypostomatic. Both 20× and 40× microscope images were taken to estimate density and aperture length of the stomata, respectively. Initial pre-tests showed that averaged values across three and five stomata for aperture length were highly correlated (R2 = 0.99, P < 0.001, n = 50). Averaged stomatal density was also highly correlated when one or four lamina impressions were used (R2 = 0.99, P < 0.001, n = 40), indicating that stomatal density was homogeneous across the lamina. Thus, aperture length was measured in three stomata and density was calculated using one image for the remaining seedlings. Stomatal pore index (SPI), which is related to leaf hydraulic conductance, was calculated as the stomatal density × pore length2 (Sack et al. 2003). Quantitative genetic variation within populations Quantitative genetic variation within populations was estimated by three quantitative genetic parameters for each combination of trait, population and treatment: additive genetic variance (VA), coefficient of additive genetic variation (CVA) and narrow-sense heritability (h2). Variance parameters were obtained by performing Bayesian mixed models applying inverse Wishart priors. Markov Chain Monte Carlo (MCMC) chains were run a minimum of 5.5 million times, sampling every 5000 and with a burn in period of 500,000 iterations (see Rodríguez‐Quilón et al. 2016 for a similar approach). The Bayesian mixed model implemented was: Yijn=μ+Mn+Bi+Fj+Εijn (1) where Yijn is the observed value for the trait considered in the nth seedling of the jth open-pollinated family in the ith block, μ the global mean, Mn the fixed effect of initial biomass in the nth individual, Bi the fixed effect of the ith block, Fj the random effect of the jth family and Eijn is the residual error term. Date of measurement was included as a fixed factor and time of day as a covariate in the model for chlorophyll fluorescence and gas exchange traits. Initial biomass, which was highly dependent on seed size, emergence time and early growth (data not shown), was included in the model as a covariate to account for maternal effects (Falconer et al. 1996). We calculated the additive genetic variance (VA) as: VA=12θ×VF where VF denotes the inter-family variance within each population and θ the coancestry coefficient. Previous studies on oaks growing in similar systems have shown that the percentage of full-sibs is usually low in progenies originating from open-pollinated trees in natural populations (Ramírez-Valiente et al. 2014c). For this reason, we used θ = 0.125 to estimate VA. CVA was calculated as: CVA(%)=100×VAμ where VA and μ represent, respectively, the additive genetic variance and the mean of the target trait for a population under a given treatment. Standard error of CVA was calculated following García‐González et al. (2012): σ[CVA]≈σ[VA]2μVA where σ[CVA] is the standard error of CVA, σ[VA] is the standard error of VA, μ is the mean and VA is the square root of the additive genetic variance. The narrow-sense heritability on an individual basis for each population was estimated as follows: h2=VAVP=4VFVF+VRVA represents the additive genetic variance, VP is the total phenotypic variance, VF is the inter-family variance within each population and VR is the residual variance (Falconer et al. 1996). Standard errors for VA and h2 were calculated using the total of 1000 estimates resulting from running Bayesian mixed models with MCMC. Since maternal environment might potentially influence the expression of traits, maternal half-sib designs might overestimate values of the additive genetic variance. Consequently, our estimates of coefficient of additive genetic variation (CVA) and heritability (h2) represent the upper bound measures for the additive coefficient of additive additive variation (CVA) and narrow-sense heritability (h2) (Falconer et al. 1996). Bayesian mixed models were implemented with MCMCglmm package in R 3.2.1 (Hadfield 2010, R Development Core Team 2015). For further analyses, we particularly focused on the coefficient of additive genetic variation (CVA) since mean-standardized metrics such as CVA are considered the best estimates of the quantitative genetic variation and better reflect the evolvability of the population (Houle 1992, Hansen et al. 2011, García‐González et al. 2012). To test for differences in CVA among populations, traits and treatments, we performed a three-way ANOVA (see Stirling et al. 2002 for a similar procedure). We also tested for differences in quantitative genetic variation among populations and treatments per each trait separately by comparing additive genetic variance in mixed-models using likelihood-ratio tests (see Appendix S1 available as Supplementary Data at Tree Physiology Online for details) (Etterson 2004). These analyses were carried out using STATISTICA 10.0 (Statsoft, Tulsa, OK, USA) and SAS University Edition (SAS/STAT® Software, SAS Institute, Inc., Cary, NC, USA), respectively. Molecular data and genetic diversity We genotyped a total of 119 individuals (with at least 20 individuals per population, Table 2) using 11 previously published microsatellite loci: QpZAG 1/2, QpZAG 1/5, QpZAG 9, QpZA15, QpZAG 16, QpZAG 36, QpZAG 46, QpZAG 102, QpZAG 110 (Steinkellner et al. 1997), QrZAG 11 and QrZAG 30 (Kampfer et al. 1998). See Ramírez-Valiente et al. (2018) for more details. We estimated the number of alleles, allelic richness and expected heterozygosity for each population using GenAlex 6.5 (Peakall and Smouse 2012). To make estimates of allelic richness comparable across populations with different sample sizes, we standardized allelic richness of each locality to our smallest sample size. We used the rarefaction procedure implemented in the program HP-RARE for this purpose (Kalinowski 2005). Table 2. Results of the analysis of variance for the coefficient of additive genetic variation (CVA). Degrees of freedom (df), F-values and tests and P-values are shown. df F P Trait 14 23.03 <0.001 Population 4 4.06 0.006 Treatment 1 8.75 0.005 Trait × Population 56 1.00 0.503 Trait × Treatment 14 1.60 0.107 Population × Treatment 4 2.66 0.042 df F P Trait 14 23.03 <0.001 Population 4 4.06 0.006 Treatment 1 8.75 0.005 Trait × Population 56 1.00 0.503 Trait × Treatment 14 1.60 0.107 Population × Treatment 4 2.66 0.042 Note: Significant values are in bold face (P < 0.05). Table 2. Results of the analysis of variance for the coefficient of additive genetic variation (CVA). Degrees of freedom (df), F-values and tests and P-values are shown. df F P Trait 14 23.03 <0.001 Population 4 4.06 0.006 Treatment 1 8.75 0.005 Trait × Population 56 1.00 0.503 Trait × Treatment 14 1.60 0.107 Population × Treatment 4 2.66 0.042 df F P Trait 14 23.03 <0.001 Population 4 4.06 0.006 Treatment 1 8.75 0.005 Trait × Population 56 1.00 0.503 Trait × Treatment 14 1.60 0.107 Population × Treatment 4 2.66 0.042 Note: Significant values are in bold face (P < 0.05). Relationships between quantitative genetic variation, neutral genetic diversity and climate In order to test for associations between CVA and climate, we first performed a Principal Component Analysis (PCA) to summarize multidimensional climate variation across populations. Nineteen bioclimatic variables were included in the PCA from Worldclim database (Hijmans et al. 2005). Then, the PCA factor scores were extracted for each population and linear regressions were performed between PCA factors and CVA. To determine if neutral and quantitative genetic variation covaried, we tested for an association between CVA and standardized allelic richness (AR) as a metric of genetic diversity of neutral markers within populations. To account for potential confounding effects of covariation in climate and population structure on quantitative genetic variation within populations, we conducted generalized linear mixed models where CVA was the dependent variable, the matrix of pairwise neutral genetic differentiation for 11 nuclear microsatellites was used as a random effect and the PCA factors were included as fixed factors in the models (see Stone et al. 2011, Sinclair et al. 2015, Sato and Kudoh 2017 for similar approaches). The pairwise neutral genetic differentiation in molecular markers was estimated using an AMOVA implemented in Arlequin 3.5 (Excoffier and Lischer 2010). We used the ‘MCMCglmm’ package in R 3.2.1 (Hadfield 2010, R Development Core Team 2015). Mean and 95% confidence intervals were estimated for Generalized Linear Mixed Model (GLMM) parameters of each effect factor. Linear regressions and GLMM were performed per each treatment separately because of the population-by-treatment interaction observed for CVA (see Results section). Finally, to analyze whether CVA in one environment was predicted by CVA in the other environment, we performed linear regressions between CVA trait values in one treatment against the CVA trait values in the other treatment. Similar regressions were conducted for h2. These analyses were performed using STATISTICA 10.0 (Statsoft). Results Quantitative genetic variation The coefficient of additive genetic variation (CVA) was highly variable across traits, populations and watering treatments (Table 2, Table S1 and Figure S1 available as Supplementary Data at Tree Physiology Online). Differences among traits were particularly noteworthy. Area- and mass-based stomatal conductance (gs,area, gs,mass), growth and area-based photosynthetic rate (Aarea) had the highest CVA, whereas SLA, leaf thickness or stomatal length had the lowest CVA (Figure 2). Figure 2. View largeDownload slide Coefficient of additive genetic variantion (CVA) and standard errors for fifteen traits measured in 1-year-old seedlings of Quercus oleoides growing in a greenhouse. Same letters indicate homogenous group of traits. SLA: specific leaf area, VLA: vein length per area SPI: stomatal pore index, ETR: electron transport rate (ETR), NPQ: non-photochemical quenching, Aarea: area-based photosynthetic rate, gs,area: area-based stomatal conductance, Amass: mass-based photosynthetic rate, gs,mass: mass-based stomatal conductance, WUE: intrinsic water-use efficiency (A/gs). Figure 2. View largeDownload slide Coefficient of additive genetic variantion (CVA) and standard errors for fifteen traits measured in 1-year-old seedlings of Quercus oleoides growing in a greenhouse. Same letters indicate homogenous group of traits. SLA: specific leaf area, VLA: vein length per area SPI: stomatal pore index, ETR: electron transport rate (ETR), NPQ: non-photochemical quenching, Aarea: area-based photosynthetic rate, gs,area: area-based stomatal conductance, Amass: mass-based photosynthetic rate, gs,mass: mass-based stomatal conductance, WUE: intrinsic water-use efficiency (A/gs). There were also differences in CVA among populations, but these differences were environment-dependent, as revealed by a significant population-by-environment interaction (Table 2, Table S2, available as Supplementary Data at Tree Physiology Online, Figure 3). In the well-watered treatment, Macuelizo had higher CVA than Sabana Grande and Santa Elena (Figure 3). In the dry treatment, Rincón had higher CVA than Sabana Grande and Santa Elena (Figure 3). Only Rincón showed differences in CVA between treatments, with the dry treatment exhibiting the highest CVA (Figure 3). In fact, if Rincón was excluded from the analyses, the treatment effect was no longer significant (P = 0.083). Figure 3. View largeDownload slide Coefficient of additive genetic variation (CVA) and standard errors for five Quercus oleoides populations in dry (red) and well-watered (blue) treatments. Same letters indicate homogenous populations within treatments. The asterisk (*) indicates differences between treatments. TA: Las Tablas, MZ: Macuelizo, SG: Sabana Grande, SE: Santa Elena and RI: Rincón. Figure 3. View largeDownload slide Coefficient of additive genetic variation (CVA) and standard errors for five Quercus oleoides populations in dry (red) and well-watered (blue) treatments. Same letters indicate homogenous populations within treatments. The asterisk (*) indicates differences between treatments. TA: Las Tablas, MZ: Macuelizo, SG: Sabana Grande, SE: Santa Elena and RI: Rincón. Although the trait-by-treatment interaction was not significant in the ANOVA, likelihood-ratio tests showed differences between treatments in traits related to stomatal morphology (density, length and pore index), stomatal conductance (gs,area, gs,mass) and photosynthesis (ETR, NPQ and Aarea) for some populations (Table S3 and Figure S1 available as Supplementary Data at Tree Physiology Online). All of these traits, except NPQ, had higher CVA in the dry treatment compared with the well-watered treatment. Trait expression in the well-watered and dry treatments was correlated for all populations except for Rincón (Figure S2 available as Supplementary Data at Tree Physiology Online). These treatment differences were attributable primarily to higher CVA for area- and mass-based stomatal conductance (gs,area, gs,mass) in the dry treatment compared with the well-watered treatment (Figure S1 available as Supplementary Data at Tree Physiology Online). For any given population, the regression slopes across environments were not significantly different from one and the intercept was not significantly different from zero. Estimates of heritability (h2) were also highly variable across populations, traits and watering treatments (Figure S3 available as Supplementary Data at Tree Physiology Online). Correlations between h2 measured in the well-watered and dry treatments were low and non-significant for all populations (Figure S4 available as Supplementary Data at Tree Physiology Online). Genetic diversity in neutral markers All microsatellite markers were polymorphic and the number of alleles ranged from 9 (QpZAG46 and QpZA15) to 26 (QpZAG102). On average, populations from Costa Rica (Rincón and Santa Elena) harbored considerably lower levels of expected heterogygosity (HE), allelic richness (A) and standardized allelic richness (AR) than populations from Honduras (Las Tablas, Macuelizo and Sabana Grande, Table 3). Table 3. Number of sampled individuals (N) and genetic diversity parameters using 11 nuclear microsatellite loci for five Quercus oleoides populations. He, expected heterozygosity, A, allelic richness and AR, standardized allelic richness. Standard errors are included in parentheses Population N He A AR Las Tablas 31 0.752 (0.051) 8.73 (0.87) 7.71 Macuelizo 26 0.697 (0.057) 8.91 (1.00) 7.91 Sabana Grande 22 0.740 (0.054) 9.27 (1.15) 8.69 Santa Elena 20 0.544 (0.076) 5.64 (0.70) 5.42 Rincon 20 0.642 (0.047) 5.45 (0.43) 5.27 Population N He A AR Las Tablas 31 0.752 (0.051) 8.73 (0.87) 7.71 Macuelizo 26 0.697 (0.057) 8.91 (1.00) 7.91 Sabana Grande 22 0.740 (0.054) 9.27 (1.15) 8.69 Santa Elena 20 0.544 (0.076) 5.64 (0.70) 5.42 Rincon 20 0.642 (0.047) 5.45 (0.43) 5.27 Table 3. Number of sampled individuals (N) and genetic diversity parameters using 11 nuclear microsatellite loci for five Quercus oleoides populations. He, expected heterozygosity, A, allelic richness and AR, standardized allelic richness. Standard errors are included in parentheses Population N He A AR Las Tablas 31 0.752 (0.051) 8.73 (0.87) 7.71 Macuelizo 26 0.697 (0.057) 8.91 (1.00) 7.91 Sabana Grande 22 0.740 (0.054) 9.27 (1.15) 8.69 Santa Elena 20 0.544 (0.076) 5.64 (0.70) 5.42 Rincon 20 0.642 (0.047) 5.45 (0.43) 5.27 Population N He A AR Las Tablas 31 0.752 (0.051) 8.73 (0.87) 7.71 Macuelizo 26 0.697 (0.057) 8.91 (1.00) 7.91 Sabana Grande 22 0.740 (0.054) 9.27 (1.15) 8.69 Santa Elena 20 0.544 (0.076) 5.64 (0.70) 5.42 Rincon 20 0.642 (0.047) 5.45 (0.43) 5.27 Relationships between quantitative genetic variation, neutral genetic diversity and climate Principal component analysis (PCA) grouped bioclimatic variables (sensu Hijmans et al. 2005) in two PCA factors. ‘PCA factor 1’ explained 55.4% of the total climatic variation in the studied populations and ‘PCA factor 2’ explained 38.4% (Figure 4). The PCA also showed that Honduran populations (Las Tablas, Macuelizo and Sabana Grande) have similar climate and very different from both Costa Rican populations (Rincón and Santa Elena) which also differ among each other (Figure 4). Figure 4. View largeDownload slide Scores on factor 1 and 2 from the Principal Component Analysis (PCA) of five populations of Quercus oleoides: Las Tablas, Macuelizo, Sabana Grande, Santa Elena and Rincón. The x- and y-axes are normalized with zero corresponding to the mean PCA score. Small white dots represent the values of the bioclimatic variables for the factor coordinates. Lower-case letters indicate the abbreviations of the bioclimatic variables obtained from Worldclim database (Hijmans et al. 2005). ap: annual precipitation (12), at: annual mean temperature (1), iso: isothermality (3), mdr: mean diurnal range (2), pcq: precipitation of the coldest quarter (19), pdm: precipitation of the driest month (14), pdq: precipitation of the driest quarter (17), ps: precipitation seasonality (15), pwaq: precipitation of the warmest quarter (18), pweq: precipitation of the wettest quarter (16), pwm: precipitation of the wettest month (13), tar: temperature annual range (7), tcm: minimum temperature of the coldest month (6), tcq: mean temperature of the coldest quarter (11), tdq: mean temperature of the driest quarter (9), ts: temperature seasonality (4), twaq: mean temperature of the warmest quarter (10), tweq: mean temperature of the wettest quarter (8), twm: maximum temperature of the warmest month (5). In parenthesis, the number of the bioclimatic variable from Worldclim database. Figure 4. View largeDownload slide Scores on factor 1 and 2 from the Principal Component Analysis (PCA) of five populations of Quercus oleoides: Las Tablas, Macuelizo, Sabana Grande, Santa Elena and Rincón. The x- and y-axes are normalized with zero corresponding to the mean PCA score. Small white dots represent the values of the bioclimatic variables for the factor coordinates. Lower-case letters indicate the abbreviations of the bioclimatic variables obtained from Worldclim database (Hijmans et al. 2005). ap: annual precipitation (12), at: annual mean temperature (1), iso: isothermality (3), mdr: mean diurnal range (2), pcq: precipitation of the coldest quarter (19), pdm: precipitation of the driest month (14), pdq: precipitation of the driest quarter (17), ps: precipitation seasonality (15), pwaq: precipitation of the warmest quarter (18), pweq: precipitation of the wettest quarter (16), pwm: precipitation of the wettest month (13), tar: temperature annual range (7), tcm: minimum temperature of the coldest month (6), tcq: mean temperature of the coldest quarter (11), tdq: mean temperature of the driest quarter (9), ts: temperature seasonality (4), twaq: mean temperature of the warmest quarter (10), tweq: mean temperature of the wettest quarter (8), twm: maximum temperature of the warmest month (5). In parenthesis, the number of the bioclimatic variable from Worldclim database. Linear regressions showed weak associations between local climate at the source of the populations and the coefficient of additive genetic variation (CVA). Only ‘PCA factor 2’ was positively associated with CVA in the dry treatment. This was mainly due to the fact that Costa Rican populations significantly differed in CVA in the dry treatment, with the mild and mesic population, Rincón, showing the highest CVA and the warm and more xeric population, Santa Elena, having the lowest CVA (Figure 5). There was no association between standardized allelic richness (AR) and CVA (Figure 6). Figure 5. View largeDownload slide Relationships between the coefficient of additive genetic variation (CVA) of populations estimated in the well-watered treatment (blue) and dry treatment (red) with factors 1 and 2 from the principal component analysis. TA: Las Tablas, MZ: Macuelizo, SG: Sabana Grande, SE: Santa Elena and RI: Rincón. Figure 5. View largeDownload slide Relationships between the coefficient of additive genetic variation (CVA) of populations estimated in the well-watered treatment (blue) and dry treatment (red) with factors 1 and 2 from the principal component analysis. TA: Las Tablas, MZ: Macuelizo, SG: Sabana Grande, SE: Santa Elena and RI: Rincón. Figure 6. View largeDownload slide Relationships between the coefficient of additive genetic variation (CVA) of populations estimated in the well-watered treatment (blue) and dry treatment (red) with standardized allelic richness (AR) estimated using eleven nuclear microsatellites. TA: Las Tablas, MZ: Macuelizo, SG: Sabana Grande, SE: Santa Elena and RI: Rincón. Figure 6. View largeDownload slide Relationships between the coefficient of additive genetic variation (CVA) of populations estimated in the well-watered treatment (blue) and dry treatment (red) with standardized allelic richness (AR) estimated using eleven nuclear microsatellites. TA: Las Tablas, MZ: Macuelizo, SG: Sabana Grande, SE: Santa Elena and RI: Rincón. The GLMM analyses showed that population genetic structure did not affect the associations between CVA and PCA factors (Table 4). Table 4. Results of the Generalized Linear Mixed Model (GLMM) testing the effects of the neutral genetic structure and PCA factors for climate variation on the coefficient of additive genetic variation (CVA) in the well-watered and dry treatments. CVA well-watered CVA dry Mean 95% LCI 95% UCI P Mean 95% LCI 95% UCI P Neutral structure 6.881 0.000 22.569 0.234 3.472 0.000 11.184 0.226 PCA factor 1 −0.044 −0.843 0.713 0.918 0.079 −0.515 0.624 0.799 PCA factor 2 −0.576 −1.627 0.620 0.320 1.204 0.362 1.939 0.003 CVA well-watered CVA dry Mean 95% LCI 95% UCI P Mean 95% LCI 95% UCI P Neutral structure 6.881 0.000 22.569 0.234 3.472 0.000 11.184 0.226 PCA factor 1 −0.044 −0.843 0.713 0.918 0.079 −0.515 0.624 0.799 PCA factor 2 −0.576 −1.627 0.620 0.320 1.204 0.362 1.939 0.003 Means with 95% lower and upper confidence intervals (LCI and UCI) are shown for the posterior distribution of estimated coefficients for variance components for random factors (genetic distance) and fixed factors (PCA factor 1, PCA factor 2) and obtained from generalized linear mixed models. Bold values highlight a significant deviation from coefficients of zero at P < 0.05. Table 4. Results of the Generalized Linear Mixed Model (GLMM) testing the effects of the neutral genetic structure and PCA factors for climate variation on the coefficient of additive genetic variation (CVA) in the well-watered and dry treatments. CVA well-watered CVA dry Mean 95% LCI 95% UCI P Mean 95% LCI 95% UCI P Neutral structure 6.881 0.000 22.569 0.234 3.472 0.000 11.184 0.226 PCA factor 1 −0.044 −0.843 0.713 0.918 0.079 −0.515 0.624 0.799 PCA factor 2 −0.576 −1.627 0.620 0.320 1.204 0.362 1.939 0.003 CVA well-watered CVA dry Mean 95% LCI 95% UCI P Mean 95% LCI 95% UCI P Neutral structure 6.881 0.000 22.569 0.234 3.472 0.000 11.184 0.226 PCA factor 1 −0.044 −0.843 0.713 0.918 0.079 −0.515 0.624 0.799 PCA factor 2 −0.576 −1.627 0.620 0.320 1.204 0.362 1.939 0.003 Means with 95% lower and upper confidence intervals (LCI and UCI) are shown for the posterior distribution of estimated coefficients for variance components for random factors (genetic distance) and fixed factors (PCA factor 1, PCA factor 2) and obtained from generalized linear mixed models. Bold values highlight a significant deviation from coefficients of zero at P < 0.05. Discussion The key finding of our study was that quantitative genetic variation was not uniformly distributed across functional traits and populations of Q. oleoides. Gas exchange traits and growth exhibited strikingly higher levels of quantitative genetic variation compared with traits related with leaf morphology, anatomy and photochemistry. Although populations differed in quantitative genetic variation, their rank was dependent on the watering treatment. Interestingly, populations with a shared evolutionary history, similar neutral diversity and geographically closed exhibited differences in quantitative genetic variation. This result reinforces previous findings that neutral molecular variation is not an adequate predictor of quantitative genetic variation, the substrate of evolutionary change. Overall, our results point to the need for more studies on quantitative genetic variation in understudied tropical tree species to help us understand the potential for evolution of populations under the ongoing climate change in these regions. Trait differences in quantitative genetic variation The coefficient of additive genetic variation was highly variable across traits. Based on the theoretical prediction that selective pressures erode quantitative genetic variation, we hypothesized that traits under selection would exhibit low quantitative genetic variation (Mousseau and Roff 1987, Stearns 1992, Futuyma 2009, Flatt and Heyland 2011, Sniegula et al. 2016). Analogously, we expected that traits that are largely plastic would maintain high levels of quantitative genetic variation within populations since their environmental responsiveness would buffer them from natural selection (Thuiller et al. 2008, Lande 2009, Nicotra et al. 2010). Consistent with our hypothesis, we observed that most gas exchange traits, which are characterized by high plasticity (see previous studies on Q. oleoides, Ramírez-Valiente and Cavender-Bares 2017, Ramírez-Valiente et al. 2017) exhibited high CVA (Figure 2). Consistently, SLA and leaf thickness, two traits that were previously reported to be under long-term directional selection (Ramírez-Valiente et al. 2018), had very low CVA (Figure 2). However, in contrast to our hypothesis, leaf lamina area and growth, which were also observed to be under natural selection in this species (but see Ramírez-Valiente et al. 2018), exhibited moderately high CVA (Figure 2). Recent studies question the expectation that selection always causes an erosion of genetic variation. For example, evidence from artificial selection experiments on protein and oil content in maize have documented that quantitative genetic variation was maintained in a breeding population over >100 generations (Moose et al. 2004). No empirical evidence on the effects of long-term selection on quantitative genetic variation within populations exists for forest tree species. Nevertheless, it is generally argued that, some forms of balancing selection—including overdominance and temporally or spatially divergent selection—can contribute to the rise or maintenance of quantitative genetic variation of traits in natural populations of long-lived species (Goldstein and Holsinger 1992, Waxman and Peck 1999, Barton and Keightley 2002, Bürguer and Gimelfarb 2002, Turelli and Barton 2004, Yeaman and Jarvis 2006, Siepielski et al. 2009, Brousseau et al. 2013). Population differences in quantitative genetic variation Our study revealed that populations of Q. oleoides differed in the coefficient of additive genetic variation (Figure 3). Previous studies have shown intraspecific differences in quantitative genetic variation among populations distributed along the species range (Rehfeldt et al. 1999, Etterson and Shaw 2001, Etterson 2004, Caruso et al. 2005, Ramírez-Valiente et al. 2011, 2014c, Kristensen et al. 2015, Sniegula et al. 2016), between cohorts from the same population (Hoffmann and Schiffer 1998) and between native and exotic conspecific populations in close vicinity (Ramírez-Valiente and Robledo-Arnuncio 2015). The amount of quantitative genetic variation within populations usually depends upon the interaction between evolutionary processes, including natural selection, gene flow and genetic drift (Bürger and Lande 1994, Falconer et al. 1996, Shaw and Etterson 2012). Our results revealed that CVA was positively associated with ‘factor 2’ from the PCA, indicating that populations from environments with greater annual and summer precipitation and lower seasonality in precipitation (sensu Hijmans et al. 2005) had higher CVA in the dry treatment (Figures 4 and 5). The meaning of this result is difficult to ascertain. On one hand, these results are consistent with our hypothesis that harsh environments reduce quantitative genetic variation due to strong selection pressures. On the other hand, the results contradict expectations that temporal variation in environmental conditions promotes the maintenance of quantitative genetic variation (Barton et al. 2002). Overall, the relationship between CVA and ‘PCA factor 2’ was relatively weak (R2 = 0.80, P = 0.041). This finding can be largely attributed to the fact that the two Costa Rican populations, Rincón and Santa Elena, which are geographically proximal but characterized by contrasting elevations and climates, exhibited different levels of quantitative genetic variation. Genetic drift and stochastic processes are also hypothesized to modulate quantitative genetic variation, particularly in small populations with restricted gene flow (Billington 1991, Ellstrand and Elam 1993, White et al. 2007, Shaw and Etterson 2012). Costa Rican populations of Q. oleoides occur at the periphery of the species’ geographic range and are isolated from the central populations (Cavender‐Bares et al. 2011, 2015). These southernmost populations harbor lower levels of genetic diversity (HE) and standardized allelic richness (AR) than Honduran populations in neutral molecular markers in agreement with the hypothesis of lower genetic diversity in peripheral populations (Table 3). Previous studies also revealed that Costa Rican populations are genetically distinct from other populations of the species as a result of their isolation and posterior genetic drift after the formation of the Nicaraguan depression (Cavender‐Bares et al. 2011, 2015). Although these findings point to an important role of genetic drift in shaping genetic structure and diversity for neutral markers in Q. oleoides, we did not detect any relationship between standardized allelic richness (AR) or genetic structure and coefficient of additive genetic variation (CVA) in any treatment (Table 4, Figure 6). This suggests that genetic drift has had a limited impact on the additive genetic variance of the traits and populations studied here. The absence of an association between neutral genetic diversity and quantitative genetic variation has frequently been explained by differential impacts of natural selection on variation in molecular markers vs quantitative genetic variation, higher sampling error for quantitative traits and environmentally dependent expression of quantitative genetic variation—or a combination of these mechanisms (Pfrender et al. 2000, Reed and Frankham 2001, Leinonen et al. 2008, Mittell et al. 2015). Environmental dependence of quantitative genetic variation Our analyses revealed differences between treatments in the coefficient of additive genetic variation. However, post-hoc tests showed that this was largely due to differences between treatments for the Rincón population (Figure 3). We hypothesized that watering treatments would not strongly affect the expression of quantitative genetic variation because all populations of Q. oleoides experience two watering seasons, dry and wet, over the year. The observations (i) that watering treatment had an impact on the Rincón population, the most mesic locality during the dry season, and (ii) high CVA in the dry treatment are consistent with the idea that stressful environments may trigger the expression of cryptic genetic variation caused by deleterious alleles or stress sensitive mutations that have not been removed by natural selection (Kondrashov and Houle 1994, Rutherford and Lindquist 1998, Collins and Bell 2004, Le Rouzic and Carlborg 2008, Chirgwin et al. 2015). Although our results are consistent with this interpretation, further research should be conducted to ascertain how watering treatments affect the expression of genetic variance in Q. oleoides populations. Likelihood ratio tests also showed significant differences in quantitative genetic variation between treatments for some traits (Figure S1 available as Supplementary Data at Tree Physiology Online). Interestingly, significant differences between treatments were observed for traits relating to stomata (morphology and conductance) and photosynthesis (ETR and area-based photosynthesis) and in all cases quantitative genetic variation was higher in the dry treatment (Figure S1 available as Supplementary Data at Tree Physiology Online). These results raise the question of why high CVA is more frequently observed in dry conditions. The well-watered treatment varied between 22% and 28% volumetric soil moisture content. This range of soil moisture is unlikely to limit physiology of Q. oleoides seedlings. In contrast, the volumetric soil moisture varied between 9% and 13% in the dry treatment. Although we do not know the field capacity of the soil, this might mean a variation of the relative extractable soil water is between 20% and 30% (Edwards and Jarvis 1982). This magnitude of variation in soil moisture is likely to induce more variation in the physiological responses compared with the well-watered treatment. Moreover, sensitivity to this rather strong drought might be governed by threshold responses (Craine et al. 2013), which might be different at intraspecific level. Similar to our results, Sherrard et al. (2009) observed that environmental stress increased CVA mainly in gas exchange traits in Avena barbata. Earlier studies, however, found evidence for the opposite result (e.g., Johnson et al. 1990) or evidence for no variation under different water availabilities (e.g., Donovan and Ehleringer 1994). It is possible that these mixed results across studies are due to differences in the severity of the drought treatment. Overall, severe drought impacts physiology quite differently than intermediate drought in Q. oleoides (Center 2015). The way drought stress is applied might also influence levels of quantitative genetic variation. For example, if the dry treatment is imposed by cycles of drought for a given number of days and then the plants are watered to field capacity, the physiological responses of plants are expected to differ from a dry treatment in which the plants are watered more frequently but only given small quantities of water, such that they never reach the field capacity, as was done in this experiment. Finally, it is important to consider that maternal effects may influence additive genetic variance and therefore CVA in our study. Maternal effects can inflate genetic variance within populations, by making the offspring of a given mother more similar than expected by genetic factors alone (Falconer et al. 1996, Mousseau and Fox 1998). Estimates of actual additive genetic variance in quantitative traits need therefore to avoid maternal effects by performing experiments with paternal half-sib families or by cultivating plants for several generations under common environments. The former is difficult, particularly in remote regions of Central America (and under conditions where seeds transport across political borders can fail), and the latter is prohibitive for long-lived plants. An alternative is to include a proxy for maternal effects, such as seed size or initial biomass, as a covariate in the analysis (Etterson 2004, White et al. 2007). In this study, we included initial biomass as covariate, a variable that is affected by seed mass, germination time and early growth, all considered to be important mechanisms of transmission of maternal effects in plants (Herman and Sultan 2011). This likely limited the bias in estimates of CVA and h2 caused by maternal effects, but we cannot completely rule out their potential influence in the calculated quantitative genetic parameters. In conclusion, our study shows that quantitative genetic variation measured in 1-year-old seedlings of Q. oloeides varied across traits and populations. Gas exchange traits, growth and leaf lamina area exhibited high coefficients of additive genetic variation (CVA). Climate was associated with CVA but only weakly. Genetic structure and diversity in neutral markers did not relate to coefficients of additive genetic variation. Interestingly, geographically proximal populations within Honduras (Macuelizo and Sabana Grande) and within Costa Rica (Rincón and Santa Elena) exhibited contrasting CVA in the well-watered and dry treatments, respectively. These findings suggest that predictions about potential evolutionary responses to climate change will be population-specific and differ across species ranges. Acknowledgments We thank Esaú Zuniga and Marileth de los Angeles Briceño for seed collection in Honduras and Costa Rica and for facilitating the permitting processes to collect, export and import acorns. We would like to thank Rubén Ramírez, Xiaojing Wei, Chris Park, Matthew Kaproth, Sydney Schiffner, Natalie McMann, Tatiana Dimugno, Nolan Radziej and all people who participated and collaborated in seed sowing, transplanting, setting up the irrigation system and helped in taking measurements of gas exchange, chlorophyll fluorescence, growth, leaf morphology and anatomy. We also thank the Plant Growth Facilities staff, particularly Roger Meissner and Pamela Warnke, for their technical support during the development of the greenhouse experiment. Finally, we would like to thank Paco García-González for important suggestions and comments on the previous versions of the manuscript. The project was funded by the National Science Foundation IOS 0843665 to J.C.-B. and J.E. Conflict of interest None declared. References Armbruster P , Bradshaw WE , Holzapfel CM ( 1998 ) Effects of postglacial range expansion on allozyme and quantitative genetic variation of the pitcher-plant mosquito, Wyeomyia smithii . Evolution 52 : 1697 – 1704 . Google Scholar Crossref Search ADS PubMed Barton NH , Keightley PD ( 2002 ) Multifactorial genetics: Understanding quantitative genetic variation . Nat Rev Genet 3 : 11 – 21 . Google Scholar Crossref Search ADS PubMed Berger D , Postma E , Blanckenhorn WU , Walters RJ ( 2013 ) Quantitative genetic divergence and standing genetic (co)variance in thermal reaction norms along latitude . Evolution 67 : 2385 – 2399 . Google Scholar Crossref Search ADS PubMed Billington HL ( 1991 ) Effect of population size on genetic variation in a dioecious conifer . Conserv Biol 5 : 115 – 119 . Google Scholar Crossref Search ADS Brock MT , Weinig C ( 2007 ) Plasticity and envrionment-specific covariances: an investigation of floral-vegetative and within flower correlations . Evolution 61 : 2913 – 2924 . Google Scholar Crossref Search ADS PubMed Brousseau L , Bonal D , Cigna J , Scotti I ( 2013 ) Highly local environmental variability promotes intrapopulation divergence of quantitative traits: an example from tropical rain forest trees . Ann Bot 112 : 1169 – 1179 . Google Scholar Crossref Search ADS PubMed Bubliy OA , Loeschcke V , Imasheva AG ( 2001 ) Genetic variation of morphological traits in Drosophila melanogaster under poor nutrition: isofemale lines and offspring–parent regression . Heredity 86 : 363 – 369 . Google Scholar Crossref Search ADS PubMed Bürger R , Lande R ( 1994 ) On the distribution of the mean and variance of a quantitative trait under mutation-selection-drift balance . Genetics 138 : 901 – 912 . Google Scholar PubMed Bürguer R , Gimelfarb A ( 2002 ) Fluctuating environments and the role of mutation in maintaining quantitative genetic variation . Genet Res 80 : 31 – 46 . Google Scholar Crossref Search ADS PubMed Caruso CM , Maherali H , Mikulyuk A , Carlson K , Jackson RB ( 2005 ) Genetic variance and covariance for physiological traits in Lobelia: are there constraints on adaptive evolution? Evolution 59 : 826 – 837 . Google Scholar PubMed Cavender-Bares J , Bazzaz FA ( 2004 ) From leaves to ecosystems: using chlorophyll fluorescence to assess photosynthesis and plant function in ecological studies. In: Papageorgiou GC, Govindjee (eds) Chlorophyll a fluorescence . Springer , Dordrecht , pp 737 – 755 . Cavender‐Bares J , González‐Rodríguez A , Pahlich A , Koehler K , Deacon N ( 2011 ) Phylogeography and climatic niche evolution in live oaks (Quercus series Virentes) from the tropics to the temperate zone . J Biogeogr 38 : 962 – 981 . Google Scholar Crossref Search ADS Cavender‐Bares J , González‐Rodríguez A , Eaton DA , Hipp AA , Beulke A , Manos PS ( 2015 ) Phylogeny and biogeography of the American live oaks (Quercus subsection Virentes): a genomic and population genetics approach . Mol Ecol 24 : 3668 – 3687 . Google Scholar Crossref Search ADS PubMed Center A ( 2015 ) Physiological and fitness consequences of seasonal rainfall variation in neotropical live oak seedlings (Quercus oleoides): Implications for global change. PhD Thesis, College of Biological Sciences, University of Minnesota Saint Paul, MN, USA . Chen IC , Hill JK , Ohlemüller R , Roy DB , Thomas CD ( 2011 ) Rapid range shifts of species associated with high levels of climate warming . Science 333 : 1024 – 1026 . Google Scholar Crossref Search ADS PubMed Chirgwin E , Monro K , Sgro CM , Marshall DJ ( 2015 ) Revealing hidden evolutionary capacity to cope with global change . Glob Chang Biol 21 : 3356 – 3366 . Google Scholar Crossref Search ADS PubMed Collins S , Bell G ( 2004 ) Phenotypic consequences of 1,000 generations of selection at elevated CO2 in a green alga . Nature 431 : 566 – 569 . Google Scholar Crossref Search ADS PubMed Conner JK , Franks R , Stewart C ( 2003 ) Expression of additive genetic variances and covariances for wild radish floral traits: comparison between field and greenhouse environments . Evolution 57 : 487 – 495 . Google Scholar Crossref Search ADS PubMed Craine JM , Ocheltree TW , Nippert JB , Towne EG , Skibbe AM , Kembel SW , Fargione JE ( 2013 ) Global diversity of drought tolerance and grassland climate-change resilience . Nat Clim Chang 3 : 63 – 67 . Google Scholar Crossref Search ADS Davis MB , Shaw RG ( 2001 ) Range shifts and adaptive responses to Quaternary climate change . Science 292 : 673 – 679 . Google Scholar Crossref Search ADS PubMed Davis MB , Shaw RG , Etterson JR ( 2005 ) Evolutionary responses to changing climate . Ecology 86 : 1704 – 1714 . Google Scholar Crossref Search ADS Deacon NJ , Cavender-Bares J ( 2015 ) Limited pollen dispersal contributes to population genetic structure but not local adaptation in Quercus oleoides forests of Costa Rica . PLoS One 10 : e0138783 . Google Scholar Crossref Search ADS PubMed Donohue K , Hammond Pyle E , Messiqua D , Heschel MS , Schmitt J ( 2000 ) Density dependence and population differentiation of genetic architecture in Impatiens capensis in natural environments . Evolution 54 : 1969 – 1981 . Google Scholar Crossref Search ADS PubMed Donovan LA , Ehleringer JR ( 1994 ) Potential for selection on plants for water-use efficiency as estimated by carbon isotope discrimination . Am J Bot 81 : 927 – 935 . Google Scholar Crossref Search ADS Donovan LA , Dudley SA , Rosenthal DM , Ludwig F ( 2007 ) Phenotypic selection on leaf water use efficiency and related ecophysiological traits for natural populations of desert sunflowers . Oecologia 152 : 13 – 25 . Google Scholar Crossref Search ADS PubMed Dudley SA ( 1996 ) Differing selection on plant physiological traits in response to environmental water availability: a test of adaptive hypotheses . Evolution 50 : 92 – 102 . Google Scholar Crossref Search ADS PubMed Eckert CG , Samis KE , Lougheed SC ( 2008 ) Genetic variation across species’ geographical ranges: the central–marginal hypothesis and beyond . Mol Ecol 17 : 1170 – 1188 . Google Scholar Crossref Search ADS PubMed Edwards WRN , Jarvis PG ( 1982 ) Relations between water content, potential and permeability in stems of conifers . Plant Cell Environ 5 : 271 – 277 . Ellstrand NC , Elam DR ( 1993 ) Population genetic consequences of small population size: implications for plant conservation . Ann Rev Ecol Syst 24 : 217 – 242 . Google Scholar Crossref Search ADS Etterson JR ( 2000 ) Evolutionary potential of the annual legume, Chamaecrista fasciculata, in relation to global warming. PhD thesis, University of Minnesota, Minneapolis, MN. Etterson JR ( 2004 ) Evolutionary potential of Chamaecrista fasciculata in relation to climate change. II. Genetic architecture of three populations reciprocally planted along an environmental gradient in the great plains . Evolution 58 : 1459 – 1471 . Google Scholar Crossref Search ADS PubMed Etterson JR , Shaw RG ( 2001 ) Constraint to adaptive evolution in response to global warming . Science 294 : 151 – 154 . Google Scholar Crossref Search ADS PubMed Excoffier L , Lischer HE ( 2010 ) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows . Mol Ecol Res 10 : 564 – 567 . Google Scholar Crossref Search ADS Falconer DS , Mackay TF , Frankham R ( 1996 ) Introduction to quantitative genetics (4th edn) . Trends Genet 12 : 280 . Google Scholar Crossref Search ADS Fisher RA ( 1930 ) The genetical theory of natural selection . Clarendon Press, Oxford . Flatt T , Heyland A ( 2011 ) Mechanisms of life history evolution: the genetics and physiology of life history traits and trade-offs . Oxford University Press , Oxford . Frankham R ( 2010 ) Challenges and opportunities of genetic approaches to biological conservation . Biol Conserv 143 : 1919 – 1927 . Google Scholar Crossref Search ADS Fraser DJ , Bernatchez L ( 2001 ) Adaptive evolutionary conservation: towards a unified concept for defining conservation units . Mol Ecol 10 : 2741 – 2752 . Google Scholar Crossref Search ADS PubMed Futuyma D ( 2009 ) Evolution . Sinauer Associates Inc. , Sunderland, MA . Gage JL , Jarquin D , Romay C et al. . ( 2017 ) The effect of artificial selection on phenotypic plasticity in maize . Nat Commun 8 : 1348 . Google Scholar Crossref Search ADS PubMed García‐González F , Simmons LW , Tomkins JL , Kotiaho JS , Evans JP ( 2012 ) Comparing evolvabilities: common errors surrounding the calculation and use of coefficients of additive genetic variation . Evolution 66 : 2341 – 2349 . Google Scholar Crossref Search ADS PubMed Goldstein DB , Holsinger KE ( 1992 ) Maintenance of polygenic variation in spatially structured populations: roles for local mating and genetic redundancy . Evolution 46 : 412 – 429 . Google Scholar Crossref Search ADS PubMed Hadfield JD ( 2010 ) MCMC methods for multi-response generalized linear mixed models: the MCMCglmm R package . J Stat Soft 33 : 1 – 22 . Google Scholar Crossref Search ADS Hansen TF , Pélabon C , Houle D ( 2011 ) Heritability is not evolvability . Evol Biol 38 : 258 – 277 . Google Scholar Crossref Search ADS Hatziskakis S , Tsiripidis I , Papageorgiou AC ( 2011 ) Leaf morphological variation in beech (Fagus sylvatica L.) populations in Greece and its relation to their post-glacial origin . Bot J Linn Soc 165 : 422 – 436 . Google Scholar Crossref Search ADS Herman JJ , Sultan SE ( 2011 ) Adaptive transgenerational plasticity in plants: case studies, mechanisms, and implications for natural populations . Front Plant Sci 2 : 102 . Google Scholar Crossref Search ADS PubMed Hewitt GM ( 1996 ) Some genetic consequences of ice ages, and their role in divergence and speciation . Biol J Linn Soc 58 : 247 – 276 . Google Scholar Crossref Search ADS Hewitt GM ( 2000 ) The genetic legacy of the Quaternary ice ages . Nature 405 : 907 – 913 . Google Scholar Crossref Search ADS PubMed Hijmans RJ , Cameron SE , Parra JL , Jones PG , Jarvis A ( 2005 ) Very high resolution interpolated climate surfaces for global land areas . Int J Climatol 25 : 1965 – 1978 . Google Scholar Crossref Search ADS Hoffmann AA , Schiffer M ( 1998 ) Changes in the heritability of five morphological traits under combined environmental stresses in Drosophila melanogaster . Evolution 52 : 1207 – 1212 . Google Scholar Crossref Search ADS PubMed Hoffmann AA , Sgrò CM ( 2011 ) Climate change and evolutionary adaptation . Nature 470 : 479 – 485 . Google Scholar Crossref Search ADS PubMed Houle D ( 1992 ) Comparing evolvability and variability of quantitative traits . Genetics 130 : 195 – 204 . Google Scholar PubMed Husby A , Visser ME , Kruuk LE ( 2011 ) Speeding up microevolution: the effects of increasing temperature on selection and genetic variance in a wild bird population . PLoS Biol 9 : e1000585 . Google Scholar Crossref Search ADS PubMed Johnson DA , Asay KH , Tieszen LL , Ehleringer JR , Jefferson PG ( 1990 ) Carbon isotope discrimination: potential in screening cool-season grasses for water-limited environments . Crop Sci 30 : 338 – 343 . Google Scholar Crossref Search ADS Johnson T , Barton N ( 2005 ) Theoretical models of selection and mutation on quantitative traits . Phil Trans R Soc Lond B Biol Sci 360 : 1411 – 1425 . Google Scholar Crossref Search ADS Kalinowski ST ( 2005 ) Hp‐rare 1.0: a computer program for performing rarefaction on measures of allelic richness . Mol Ecol Res 5 : 187 – 189 . Kampfer S , Lexer C , Glössl J , Steinkellner H ( 1998 ) Characterization of (GA)n microsatellite loci from Quercus robur . Hereditas 129 : 183 – 186 . Google Scholar Crossref Search ADS Kondrashov AS , Houle D ( 1994 ) Genotype-environment interactions and the estimation of the genomic mutation rate in Drosophila melanogaster . Proc R Soc Lond B Biol Sci 258 : 221 – 227 . Google Scholar Crossref Search ADS Kristensen TN , Overgaard J , Lassen J , Hoffmann AA , Sgrò C ( 2015 ) Low evolutionary potential for egg‐to‐adult viability in Drosophila melanogaster at high temperatures . Evolution 69 : 803 – 814 . Google Scholar Crossref Search ADS PubMed Lande R ( 2009 ) Adaptation to an extraordinary environment by evolution of phenotypic plasticity and genetic assimilation . J Evol Biol 22 : 1435 – 1446 . Google Scholar Crossref Search ADS PubMed Le Rouzic A , Carlborg Ö ( 2008 ) Evolutionary potential of hidden genetic variation . Trends Ecol Evol 23 : 33 – 37 . Google Scholar Crossref Search ADS PubMed Leinonen T , O’Hara RB , Cano JM , Merilä J ( 2008 ) Comparative studies of quantitative trait and neutral marker divergence: a meta‐analysis . J Evol Biol 21 : 1 – 17 . Google Scholar Crossref Search ADS PubMed Linares JC ( 2011 ) Biogeography and evolution of Abies (Pinaceae) in the Mediterranean basin: the roles of long‐term climatic change and glacial refugia . J Biogeogr 38 : 619 – 630 . Google Scholar Crossref Search ADS Lynch M , Hill WG ( 1986 ) Phenotypic evolution by neutral mutation . Evolution 40 : 915 – 935 . Google Scholar Crossref Search ADS PubMed Lynch M , Walsh B ( 1998 ) Genetics and analysis of quantitative traits , Vol. 1 . Sinauer , Sunderland, MA . Miranda-Jácome A , Montaña C , Fornoni J ( 2014 ) Environmentally dependent expression of heritable variation on early recruitment traits induced by light conditions and provenance in the columnar cactus Pilosocereus leucocephalus . Pers Plant Ecol Evol Syst 16 : 322 – 330 . Google Scholar Crossref Search ADS Mittell EA , Nakagawa S , Hadfield JD ( 2015 ) Are molecular markers useful predictors of adaptive potential? Ecol Lett 18 : 772 – 778 . Google Scholar Crossref Search ADS PubMed Moose SP , Dudley JW , Rocheford TR ( 2004 ) Maize selection passes the century mark: a unique resource for 21st century genomics . Trends Plant Sci 9 : 358 – 364 . Google Scholar Crossref Search ADS PubMed Mousseau TA , Roff DA ( 1987 ) Natural selection and the heritability of fitness components . Heredity 59 : 181 – 197 . Google Scholar Crossref Search ADS PubMed Mousseau TA , Fox CW ( 1998 ) The adaptive significance of maternal effects . Trends Ecol Evol 13 : 403 – 407 . Google Scholar Crossref Search ADS PubMed Nicotra AB , Atkin OK , Bonser SP et al. . ( 2010 ) Plant phenotypic plasticity in a changing climate . Trends Plant Sci 15 : 684 – 692 . Google Scholar Crossref Search ADS PubMed Ouborg NJ , Vergeer P , Mix C ( 2006 ) The rough edges of the conservation genetics paradigm for plants . J Ecol 94 : 1233 – 1248 . Google Scholar Crossref Search ADS Peakall ROD , Smouse PE ( 2012 ) GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and researchdan update . Bioinformatics 28 : 2537e2539 . Google Scholar Crossref Search ADS Pease CM , Lande R , Bull J ( 1989 ) A model of population growth, dispersal and evolution in a changing environment . Ecology 70 : 1657 – 1664 . Google Scholar Crossref Search ADS Pfrender ME , Spitze K , Hicks J , Morgan K , Latta L , Lynch M ( 2000 ) Lack of concordance between genetic diversity estimates at the molecular and quantitative-trait levels . Conserv Genet 1 : 263 – 269 . Google Scholar Crossref Search ADS R Development Core Team ( 2015 ) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.r-project.org. Ramírez-Valiente JA , Robledo-Arnuncio JJ ( 2015 ) Exotic gene flow affects fitness trait values but not levels of heritable trait variation in the southernmost population of Scots pine (Pinus sylvestris L. var. nevadensis) . Biol Conserv 192 : 331 – 342 . Google Scholar Crossref Search ADS Ramírez-Valiente JA , Cavender-Bares J ( 2017 ) Evolutionary trade-offs between drought resistance mechanisms across a precipitation gradient in a seasonally dry tropical oak (Quercus oleoides) . Tree Physiol 37 : 889 – 901 . Google Scholar Crossref Search ADS PubMed Ramírez-Valiente JA , Valladares F , Huertas AD , Granados S , Aranda I ( 2011 ) Factors affecting cork oak growth under dry conditions: local adaptation and contrasting additive genetic variance within populations . Tree Genet Genom 7 : 285 – 295 . Google Scholar Crossref Search ADS Ramírez-Valiente JA , Alia R , Aranda I ( 2014a) Geographical variation in growth form traits in Quercus suber and its relation to population evolutionary history . Evol Ecol 28 : 55 – 68 . Google Scholar Crossref Search ADS Ramírez-Valiente JA , Valladares F , Sánchez-Gómez D , Delgado A , Aranda I ( 2014b) Population variation and natural selection on leaf traits in cork oak throughout its distribution range . Acta Oecol 58 : 49 – 56 . Google Scholar Crossref Search ADS Ramírez-Valiente JA , Valladares F , Aranda I ( 2014c) Exploring the impact of neutral evolution on intrapopulation genetic differentiation in functional traits in a long-lived plant . Tree Genet Genom 10 : 1181 – 1190 . Google Scholar Crossref Search ADS Ramírez-Valiente JA , Valladares F , Delgado A , Nicotra AB , Aranda I ( 2015a) Understanding the importance of intrapopulation functional variability and phenotypic plasticity in Quercus suber . Tree Genet Genom 11 : 35 . Google Scholar Crossref Search ADS Ramírez-Valiente JA , Koehler K , Cavender-Bares J ( 2015b) Climatic origins predict variation in photoprotective leaf pigments in response to drought and low temperatures in live oaks (Quercus series Virentes) . Tree Physiol 35 : 521 – 534 . Google Scholar Crossref Search ADS PubMed Ramírez-Valiente JA , Center A , Sparks JP , Sparks KL , Etterson JR , Longwell T , Pilz G , Cavender-Bares J ( 2017 ) Population-level differentiation in growth rates and leaf traits in seedlings of the neotropical live oak Quercus oleoides grown under natural and manipulated precipitation regimes . Front Plant Sci 8 : 1 – 14 . Google Scholar Crossref Search ADS PubMed Ramírez-Valiente JA , Deacon N , Center A , Sparks JP , Sparks KL , Etterson JR , Longwell T , Pilz G , Cavender-Bares J ( 2018 ) Natural selection and neutral evolutionary processes contribute to genetic divergence in leaf traits across a precipitation gradient in the tropical oak Quercus oleoides . Mol Ecol 27 : 2176 – 2192 . Google Scholar Crossref Search ADS PubMed Reed DH , Frankham R ( 2001 ) How closely correlated are molecular and quantitative measures of genetic variation? A meta-analysis . Evolution 55 : 1095 – 1103 . Google Scholar Crossref Search ADS PubMed Rehfeldt GE , Ying CC , Spittlehouse DL , Hamilton DA ( 1999 ) Genetic responses to climate in Pinus contorta: niche breadth, climate change, and reforestation . Ecol Monogr 69 : 375 – 407 . Google Scholar Crossref Search ADS Reich PB , Walters MB , Ellsworth DS ( 1997 ) From tropics to tundra: global convergence in plant functioning . Proc Nat Acad Sci USA 94 : 13730 – 13734 . Google Scholar Crossref Search ADS PubMed Rodríguez‐Quilón I , Santos‐del‐Blanco L , Serra‐Varela MJ , Koskela J , González‐Martínez SC , Alía R ( 2016 ) Capturing neutral and adaptive genetic diversity for conservation in a highly structured tree species . Ecol Appl 26 : 2254 – 2266 . Google Scholar Crossref Search ADS PubMed Rutherford SL , Lindquist S ( 1998 ) Hsp90 as a capacitor for morphological evolution . Nature 396 : 336 – 342 . Google Scholar Crossref Search ADS PubMed Sack L , Scoffoni C ( 2013 ) Leaf venation: structure, function, development, evolution, ecology and applications in the past, present and future . New Phytol 198 : 983 – 1000 . Google Scholar Crossref Search ADS PubMed Sack L , Cowan PD , Jaikumar N , Holbrook NM ( 2003 ) The ‘hydrology’of leaves: co‐ordination of structure and function in temperate woody species . Plant Cell Environ 26 : 1343 – 1356 . Google Scholar Crossref Search ADS Sagnard F , Barberot C , Fady B ( 2002 ) Structure of genetic diversity in Abies alba Mill. from southwestern Alps: multivariate analysis of adaptive and non-adaptive traits for conservation in France . For Ecol Manage 157 : 175 – 189 . Google Scholar Crossref Search ADS Sato Y , Kudoh H ( 2017 ) Fine-scale frequency differentiation along a herbivory gradient in the trichome dimorphism of a wild Arabidopsis . Ecol Evol 7 : 2133 – 2141 . Google Scholar Crossref Search ADS PubMed Sgrò CM , Hoffmann AA ( 1998 ) Effects of temperature extremes on genetic variances for life history traits in Drosophila melanogaster as determined from parent‐offspring comparisons . J Evol Biol 11 : 1 – 20 . Google Scholar Crossref Search ADS Shaw RG , Etterson JR ( 2012 ) Rapid climate change and the rate of adaptation: insight from experimental quantitative genetics . New Phytol 195 : 752 – 765 . Google Scholar Crossref Search ADS PubMed Sherrard ME , Maherali H , Latta RG ( 2009 ) Water stress alters the genetic architecture of functional traits associated with drought adaptation in Avena barbata . Evolution 63 : 702 – 715 . Google Scholar Crossref Search ADS PubMed Siepielski AM , DiBattista JD , Carlson SM ( 2009 ) It’s about time: the temporal dynamics of phenotypic selection in the wild . Ecol Lett 12 : 1261 – 1276 . Google Scholar Crossref Search ADS PubMed Sinclair FH , Stone GN , Nicholls JA et al. . ( 2015 ) Impacts of local adaptation of forest trees on associations with herbivorous insects: implications for adaptive forest management . Evol App 8 : 972 – 987 . Google Scholar Crossref Search ADS Sniegula S , Golab MJ , Drobniak SM , Johansson F ( 2016 ) Seasonal time constraints reduce genetic variation in life‐history traits along a latitudinal gradient . J Anim Ecol 85 : 187 – 198 . Google Scholar Crossref Search ADS PubMed Stearns SC ( 1992 ) The evolution of life histories , Vol. 249 . Oxford University Press , Oxford . Steinkellner H , Lexer C , Turetschek E , Glössl J ( 1997 ) Conservation of (GA)n microsatellite loci between Quercus species . Mol Ecol 6 : 1189 – 1194 . Google Scholar Crossref Search ADS Stirling DG , Réale D , Roff DA ( 2002 ) Selection, structure and the heritability of behaviour . J Evol Biol 15 : 277 – 289 . Google Scholar Crossref Search ADS Stock AJ , Campitelli BE , Stinchcombe JR ( 2014 ) Quantitative genetic variance and multivariate clines in the Ivyleaf morning glory, Ipomoea hederacea . Phil Trans R Soc Lond B Biol Sci 369 : 20130259 . Google Scholar Crossref Search ADS Stone GN , Nee S , Felsenstein J ( 2011 ) Controlling for non-independence in comparative analysis of patterns across populations within species . Phil Trans R Soc Lond B Biol Sci 366 : 1410 – 1424 . Google Scholar Crossref Search ADS Svenning JC , Skov F ( 2005 ) The relative roles of environment and history as controls of tree species composition and richness in Europe . J Biogeogr 32 : 1019 – 1033 . Google Scholar Crossref Search ADS Thuiller W , Albert C , Araújo MB et al. . ( 2008 ) Predicting global change impacts on plant species’ distributions: future challenges . Persp Plant Ecol Evol Syst 9 : 137 – 152 . Google Scholar Crossref Search ADS Turelli M , Barton NH ( 2004 ) Polygenic variation maintained by balancing selection: pleiotropy, sex-dependent allelic effects and G × E interactions . Genetics 166 : 1053 – 1079 . Google Scholar Crossref Search ADS PubMed Valladares F , Balaguer L , Martínez‐Ferri E , Pérez‐Corona E , Manrique E ( 2002 ) Plasticity, instability and canalization: is the phenotypic variation in seedlings of sclerophyll oaks consistent with the environmental unpredictability of Mediterranean ecosystems? New Phytol 156 : 457 – 467 . Google Scholar Crossref Search ADS van Heerwaarden J , Hellin J , Visser RF , Van Eeuwijk FA ( 2009 ) Estimating maize genetic erosion in modernized smallholder agriculture . Theor Appl Genet 119 : 875 – 888 . Google Scholar Crossref Search ADS PubMed Waxman D , Peck JR ( 1999 ) Sex and adaptation in a changing environment . Genetics 153 : 1041 – 1053 . Google Scholar PubMed White TL , Adams WT , Neale DB ( 2007 ) Forest genetics . Cabi , Cambridge, MA. Wilson AJ , Pemberton JM , Pilkington JG , Coltman DW , Mifsud DV , Clutton-Brock TH , Kruuk LB ( 2006 ) Environmental coupling of selection and heritability limits evolution . PLoS Biol 4 : e216 . Google Scholar Crossref Search ADS PubMed Wright IJ , Reich PB , Westoby M et al. . ( 2004 ) The worldwide leaf economics spectrum . Nature 428 : 821 – 827 . Google Scholar Crossref Search ADS PubMed Yeaman S , Jarvis A ( 2006 ) Regional heterogeneity and gene flow maintain variance in a quantitative trait within populations of lodgepole pine . Proc Biol Sci 273 : 1587 – 1593 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Evolutionary potential varies across populations and traits in the neotropical oak Quercus oleoides
JO - Tree Physiology
DO - 10.1093/treephys/tpy108
DA - 2019-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/evolutionary-potential-varies-across-populations-and-traits-in-the-vdhpUdrGY0
SP - 427
VL - 39
IS - 3
DP - DeepDyve
ER -