Adaptive potential and genomic vulnerability of keystone forest tree species to climate change: a case study in Scots pine

doi:10.21203/rs.3.rs-4376686/v1

Download PDF

Article

Adaptive potential and genomic vulnerability of keystone forest tree species to climate change: a case study in Scots pine

https://doi.org/10.21203/rs.3.rs-4376686/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

A better understanding of the possible adaptive response and genomic vulnerability of forest trees is needed to properly assist future forest management and develop adequate resilience strategies to changing environments. Scots pine (Pinus sylvestris L.), a keystone species with extensive distribution and a broad ecological niche, is expected to be directly impacted by climate change due to fitness loss and genetic maladaptation on a large spatial scale. Despite extensive studies that have clarified the broad-scale history and genetic structure of the species, understanding the genetic basis for the local adaptation and genomic vulnerability of Scots pine remains incomplete. Here, we used thousands of genotyped SNP markers in 39 natural populations (440 trees) along a broad latitudinal gradient of species distribution to examine molecular signatures of local adaptation. Specifically, this landscape genomics approach aimed to assess fine-scale patterns of SNPs associated with environmental gradients, predict vulnerability to climate change using genomic offset, and evaluate the adaptive response of populations to projected climate shifts. The variation of outlier SNPs, which exhibits selection signatures between genetically very similar populations in the distribution range, was highly correlated with mean temperature, a key limiting factor for the growth and survival of tree species. Furthermore, our simulation results indicated a high genomic vulnerability on a large spatial scale in P. sylvestris, with the time frame required to close the offset gap by natural selection estimated to be in the range of hundreds of years. The results improve our understanding of Scots pine's adaptive capacity and provide insights for management approaches to mitigate the impacts of climate change on temperate forest ecosystems. By evaluating adaptive responses, the study adds to the discussion on the long-term sustainability of forest ecosystems in the face of ongoing environmental change.

Earth and environmental sciences/Environmental sciences/Environmental impact

Biological sciences/Evolution/Population genetics

Biological sciences/Plant sciences/Plant genetics

genomic offset

resilience strategies

local adaptation

forest management

assisted migration

Many keystone temperate forest tree species have a large distribution range and occupy habitats across steep climatic gradients. In response to varying selection pressures across time and heterogeneous environments, they not only maintain a high level of background genetic variation (orders of magnitude higher than in perennial plants), show considerable levels of phenotypic plasticity but also signatures of local adaptation due to natural selection, despite extensive gene flow ^1–5. These mechanisms likely evolved to ensure persistence in the face of environmental upheavals and are of crucial importance in long-lived organisms with limited mobility and a lack of behavioral responses to changing environments ^6,7. Consequently, from a short-term perspective, the adaptive potential of trees is almost entirely reliant on standing genetic variation. Studies have shown that in rapidly changing environments, populations are more likely to adapt when potentially beneficial alleles are already present within them ^8,9. However, the exact relationship between the adaptive potential and the magnitude of predicted environmental change is often unknown.

Increasing pressure on forest ecosystems due to environmental change raises questions about tree conditions and performance considering their specific life history traits (longevity, long generation time), complex demographic histories, and population structures. Some levels of resilience to large-scale environmental perturbations are already known for those species, as glacial cycles during the Pleistocene shaped their genetic structure, especially in the Northern Hemisphere ^2,6,10. However, the magnitude and tempo of current climatic shifts are unprecedented in the historical record ¹¹, and many studies predict that substantial shifts in distribution ranges of species might be required to follow the optimal climatic envelopes ^12–14. Another possible way to cope with increasing selective pressure due to a rapidly changing environment is the adaptive response of populations, but still less is known about underlying genetic architecture and potential, confounding factors (i.e. plastic response) and limits of such adaptive response ^5,15. Therefore, a better understanding of the link between the geographical distribution of adaptive genetic variation and its molecular signatures, along with the associations between genomes and the environment, is essential. These insights could inform decisions regarding the management of both economically important commercial forests and natural forests of great ecological value, that will face increased pressure from environmental changes.

Scots pine (Pinus sylvestris L.) is an outcrossing and wind-pollinated species that naturally forms random mating populations across extensive areas of Europe and Asia. Given its wide distribution range and significant commercial importance, Scots pine stands are one of the most ecologically and socially valuable forest trees globally. The genetic variation observed in P. sylvestris arises from several interrelated factors, including changes in population ranges and sizes during glacial periods, gene flow between populations, and natural selection responding to adaptation to local environmental conditions ¹⁶. Numerous molecular genetic studies have provided insights into the broad-scale history of Scots pine populations in Europe and Asia, contributing to their current geographic range and the observed phenotypic variability established during and after the species' postglacial recolonization ^17–19. These studies have revealed the spatial structure of several genetic groups within the species ^20–25 and the persistence of a distinctive genetic lineage, despite putative gene flow across large geographical areas ²⁶. This research provided a comprehensive understanding of standing natural variation within Scots pine. However, much less is known about the adaptive potential of this species across its range and consequently, the relative climate vulnerability of its populations ²⁷.

In the face of ongoing environmental changes it is particularly important to consider the possibility of adaptive responses of populations affected by shifting climatic conditions as shown across various study systems ^28–33. Species distribution models under various climate change scenarios indicate that many forest tree species will alter their geographical distribution to align with their ecological niche ^7,12,34–36. The massive decline of coniferous forests in the temperate zone of Europe is predicted by distribution models ^12,37,38. Scots pine, in particular, might suffer not only from the direct effects of climate change on its distribution but also from increased competition with deciduous species such as oaks and beeches. The predicted mismatch between suitable climatic niches and the genetic composition of species will likely impact silviculture and forestry on a large scale, requiring the appropriate management plans that should accommodate changing climates ³⁹. Assisted migration might be needed to counter these challenges ⁴⁰. However, such approaches would require revision of existing rules and guidelines on the origin of reproductive materials at national scales.

Estimating the adaptive potential of Scots pine across its range is crucial to evaluate the climate vulnerability of wild populations and identify those that might exhibit greater resilience to changing environments. This information can guide strategies for enhancing forest management and conservation, ensuring the continued health of socially and economically significant forests. Here, we utilize data from thousands of polymorphic loci located in the coding regions of P. sylvestris sampled across an environmental gradient in Europe, combined with environmental data to 1) identify genomic regions correlated with the environmental factors driving the local adaptation; 2) calculate the genomic offset for each population to predict climate change vulnerability; and 3) assess the extent and timeframe of the adaptive response to predicted climate, considering the species' biology. Finally, based on our results, we discuss the impact of genomic offset on managing genetic resources for forestry in the face of ongoing environmental change.

Plant material and sampling

We sampled 39 populations of Scots pine, covering a large part of the species distribution (latitudes 49.1°N-69.8°N) and well reflecting the steep environmental gradient found within the species range (Fig S1., Table S1). Those include populations from northern Finland (FIN1-21; FIN) to southern Poland (POL1-15; POL) and three populations from the Baltic Region (LVA, LTU, EST). For each population needles were collected from 5 to 23 (median = 10) randomly selected matured trees that were at least 50 m apart. In total, we analyzed 440 individuals.

SNP genotyping and data processing

Genomic DNA was extracted from needles using a Genomic Mini AX Plant kit following the manufacturer's instructions (A&A Biotechnology, Poland). The quantity of DNA was measured by Qubit 4 fluorometer, using the Broad Range (BR) Assay Kit and DNA was diluted to the working concentration of 40 ng/µl. We acquired the single nucleotide polymorphism (SNP) genotypes of all individuals using a custom Axiom PineGAP SNP array (Affymetrix, Thermo Fisher Scientific, Santa Clara, CA, USA). The processing was carried out at the Bristol Genomics Facility (Bristol, UK). The array consisted of 49,829 SNP markers, and a detailed description of its composition and design is described in Perry et al. (Perry et al., 2020). Genotypes were called using Axiom Analysis Suite software (Applied Biosystems, Waltham, MA, USA), following the manufacturer's recommendations. We changed the default quality filters to more stringent, adjusting the sample threshold configuration parameters as follows: QC call_rate was set to 95%, Average cut-rate for passing samples to 95%, and Cr-cutoff to 95%. Initially, our set of SNPs was composed of genotypes assigned to BestAndRecomended (52.79%) category by Axiom Analysis Suite software and we exported those SNPs to ped file format. Then, several filtering steps were performed to extract high-confidence, informative and polymorphic SNPs in PLINK v.1.07 (Purcell et al., 2007). First, we removed 93 SNPs with BLAST hits to Pinus taeda mitochondrial genome (NC_039746.1) and Pinus sylvestris chloroplast genome (NC_035069.1). Next, SNPs with allele frequency (MAF) < 0.01, and loci/individuals with more than 10% missing data were discarded. Finally, we conducted LD pruning to remove SNPs that were highly linked (LD r² > 0.7). This final set of SNPs (6995) was used for all analyses in this study.

Population structure and genetic diversity

We calculated observed (Ho) and expected heterozygosity (He) and fixation index F_IS in each population and overall using hierfstat package in R ^41,42 R Core Team 2023). Genetic differentiation among populations was determined using the pairwise F_ST calculated according to Weir and Cockerham ⁴³ in adegenet ^44,45. To inspect the spatial pattern of genetic diversity we examined population structure using LEA R package ⁴⁶ and function snmf() that computes least-squares estimates of ancestry proportions and ancestral allelic frequencies and evaluates the quality of fit of the statistical model to the data by using a cross-validation technique ⁴⁷. We tested for ancestral clusters ranging from K = 1 to K = 40, using 10 replications for each K, to determine cross-entropy. The results were visualized using POPHELPER Structure Web App v1.0.10 ⁴⁸. Population-level ancestry coefficients were calculated by custom R script and visualized on a map using ggplot2 ^49,50. The distribution of genetic variance was also assessed by principal component analysis (PCA) in adegenet R package. We first performed PCA utilizing the complete dataset and all of the sampled populations to reveal potential population subdivisions, elucidating high-level patterns of genetic variation. Next, we looked at PCAs within each region (excluding populations from Latvia, Lithuania and Estonia). This localized approach was aimed at unraveling the fine-scale genetic structure patterns and potentially uncovering genetic signatures unique to each region's ecological context, that might have remained concealed within the broader PCA. Finally, to assess the impact of outlier SNPs on the identified genetic structure, we performed supplementary PCA analyses encompassing two scenarios: one including the full dataset of SNPs and another excluding the SNPs found to be outliers (see section 2.4. Outlier detection for details).

To understand the interplay of genetic differentiation, geographical distances, and environmental factors, we employed a Mantel test approach. First, we tested the correlation between populations genetic distances (as linearized F_ST values) and geodistances to account for patterns of Isolation by Distance (IBD). Geographic coordinates were transformed into cartesian coordinates using geoXY() function from SoDA and Euclidean distances were calculated using dist() function before performing the Mantel test in vegan with 999 permutations to assess statistical significance ⁵¹. Similarly, we performed the Mantel test on environmental distances to determine the degree of isolation by environment (IBE). Environmental distances were calculated as Euclidean distances based on the first two PCs performed on a set of 8 selected variables (Fig. S9A, see Outliers detection section for details on variable selection).

Outlier detection

We conducted outlier detection analyses using three different methods: F_ST outlier scan, detection based on population structure, and genotype-environment association (GEA) method. First, we used OutFLANK v0.2 ^52,53 to compute a trimmed distribution of F_ST values, which approximate the F_ST for neutral markers (not subjected to selection), and we compared this to the F_ST distribution for all loci in the dataset. Then, using qvalue R package ⁵⁴ we performed the q-value procedure at 0.05 false discovery rate (FDR) to identify loci that show significantly high/low F_ST values and thus are likely subjected to selection.

Next, we used pcadapt v4.3.3 R package ⁵⁵ to detect outliers that are related to population structure using robust Mahalanobis distance. In this method, the population structure is first ascertained with PCA, and then Cattell’s graphical rule is used to choose the number of principal components (K) to identify SNPs involved in local adaptation. Compared to other outlier detection methods, pcadapt produces fewer false positive results and is less computationally demanding (Luu, Bazin et al. 2016). When using methods that utilize only genomics data for outliers detection (e.g. OutFLANK and pcadapt; hereafter we refer to those as genomic outliers scans) we followed the same procedure using all populations, representing a full environmental gradient and separately for populations from Poland (POL) and Finland (FIN) to track both large and fine-scale patterns of local adaption in Scots pine.

Finally, to detect linear relationships between genetic variations and multivariate environmental variables, allowing us to link outlier loci with environmental conditions we used redundancy analyses (RDAs) in vegan. We performed RDA using the population-level allele frequencies as a dependent matrix and two independent matrices, an environmental matrix, and a geographic distance matrix based on distance-based Moran's eigenvector maps (dbMEMs) which represent a spatial decomposition of distances between populations. To properly represent environmental variation at species’ distribution we initially included 25 environmental variables (e.g. temperature, precipitation, soil PH and phenological indices; see Supplementary Table S3 for a list of all variables). Because highly correlated predictors and multicollinearity could lead to biased estimates of model coefficient ⁵⁷, we retained only 8 variables with correlation coefficient |r| < 0.7 and Variance Inflation Factors (VIF) < 10 ⁵⁸ using pairs.panels() and vif.cca() functions in vegan (Table S3, Fig. S9A). The second independent matrix was created from 10 dbMEMs retained after the forward selection step in adespatial R package, from the initial set of 38 variables ⁵⁹. We followed our RDA analysis with partial redundancy analysis (pRDA) to discriminate between the pure effects of environment and geography and their combined effects. We assessed all the models using anova.cca() function in vegan and 999 permutations. To determine which SNPs could be assigned as putative adaptive variants we used a cutoff point of ± 2.5 standard deviation from the distribution of individual SNP loadings in the ordination space for the constrained axes explaining most of the variance. We performed RDA using all SNPs, as well as two sets including outlier SNPs detected by genomic outlier scans, and all three methods (including GEA). When following all the outlier detection methods to minimize the false positive results and retain only a robust set of loci that are putatively under selection, we only considered SNPs detected by all three outlier methods as potential adaptive variants (PAV).

Genomic vulnerability assessment

To evaluate the risk of adaptive pressure for Scots pine we assessed the current adaptive potential of its populations as well as the projected environmental changes that those populations will likely face in the future following the genomic vulnerability approach, by calculating RONA ^60,61. First, for each sampling location, we downloaded future bioclimatic variables (2050; average for years 2041–2060) for the climatic model BCC-CSM2-MR projected under three representative concentration pathways (RCPs): RCP2.6, RCP4.5 and RCP8.5 climate changes, from WorldClim Version2 ^62,63) at a spatial resolution of 30 seconds (~ 1 km²). These pathways model the possible future climatic conditions differing in the level of control over global carbon emissions and range from relatively mild to severe projected changes (RCP2.6 < RCP4.5 < RCP8.5, respectively).

We calculated a linear relationship between PAV and environmental variables with the strongest association to those outlier loci, and we also cheeked the correlation between genotype and said bioclimatic variables using a generalized linear model (GLM) with a binomial error, to model the probability of having one of two alleles in a given loci. Next, we used the combination of current allele frequencies per population for PAV and the projected future environmental conditions to predict the shift in alleles over time under changing climatic conditions, using AlleleShift R package ⁶⁴, for all PAVs and each RCP scenario. The genomic vulnerability score was calculated for each population as the mean difference between current and future expected allele frequencies, and extrapolated values of genomic vulnerability were visualized for the studied transect, using Inverse Distance Weighting (IDW) implemented in gstat R package ⁶⁵. To illustrate the timeframe for populations to track the predicted changes in allele frequencies in future environmental conditions, we calculated the number of generations required for a genetic shift of such magnitude with natural selection being sole driver of this change. We simulated scenarios with different strengths of selection coefficients, we set all SNPs to be dominant (h = 0), and set the change of alleles frequency (∆p) to be equal maximum and minimum value of genomic vulnerability score, using custom-made R script. We transformed the time from a number of generations into calendar years by assuming an average generation time of 20–25 years in P. sylvestris ^66,67.

Genetic diversity and patterns of population structure

The overall level of observed heterozygosity Ho was 0.3149, while expected heterozygosity He was 0.3115. The diversity levels recorded for individual populations were found to be similar across the studied transect and within the regions (Fig. 1, Fig S1., Table S1, Table S2). Following the high outbreeding rate of this species, no indication of inbreeding was found in our dataset, as the values of the F index were close to 0 (Fig. 1, Table S2). Pairwise F_ST across all populations was low with a global estimate of 0.014. The divergence was higher between Poland and Finland (F_ST = 0.019) and within regions, the F_ST was twofold higher in Poland when compared to Finland (0.017 and 0.008, respectively). This result was mainly driven by populations POL2, POL6, POL14 and POL15 which showed the highest values across all pairwise comparisons, leading to significantly elevated levels of F_ST within the population in Poland, where it was comparable to F_ST recorded between geographical regions (Fig. S2). The Mantel test results showed that despite the low level of divergence between populations, an increase in F_ST was statistically significantly correlated with both genetic and environmental distances between populations (Fig. S3). However, the strength of those correlations differed and indicated that isolation by distance pattern (IBD) is stronger than isolation by environment (IBE) (R² = 0.27 and R² = 0.17, respectively, both with p < 0.001, Fig. S3 A, B).

The genetic structure analysis in LEA showed a cline-like pattern of ancestry strongly associated with latitude but discriminating Polish from Finnish lineages (Fig. 2A, B). The cross-entropy criterion showed two significant drops at K = 4 and K = 10, however increasing K values beyond 4 did not indicate any additional structure, but rather reflected the pattern already present at K = 4, so we refer here to this value (Fig. S4, S5). The ancestry more prevalent within Finnish populations (blue color on the LEA plot) is decreasing substantially from north to south, and in Poland is lower than 20% (per population). In contrast, the Polish populations are characterized by a higher ancestry proportion of two clusters (yellow and orange on the LEA plot), the latter being very rare in the north. Interestingly, the population from Stołowe Mts. (POL2, POL13-15) shows not only the highest proportion of this ancestry but some trees from this region are so diverged that they form their own cluster (marked red in the LEA plot). Principal component analysis performed at all SNPs (6995) showed a consistent pattern with the LEA results of population structure driven by geography (Fig. 2C). We also looked at the fine-scale patterns of population structure within regions but did not find any such pattern in Finland and in Poland. The most distinctive were populations from Stołowe Mts., which is consistent with earlier analyses (Figure S6 A, B). To assess how the outlier loci (possibly under selection) impact the population structure we excluded SNPs detected by genomic outliers scans (16 SNPs) and separately PAV (6 SNP; see Outlier detection section) and in both cases we retrieved the same grouping pattern as with all SNPs (Figure S7). Interestingly, when we used only outlier loci in PCA (16 and 6) we recovered a similar pattern but with lower discriminating power (Fig S7C,D), and no such pattern was present, when we randomly selected the same number of SNPs as outlier loci and run 10 000 permutations (see Supplementary Materials Fig. S7, for details).

Genetic-environmental associations

Using 6995 SNPs as input when looking for loci that show departure from neutrality three different methods including OutFLANK, pcadapt, and RDA detected 55, 39 and 30 SNPs, respectively. To provide robust inference and avoid reporting false positives we consider only 6 SNPs (PAV) detected by all three methods as systematically linked with local adaptation in P. sylvestris (Fig. 3A). The frequency of those SNPs showed a clinal pattern (SNP 2236 and 5721 are the exception – as they are almost fixed in one region and segregating in the other, Fig. 3B). Moreover, all of them were found in the RDA analysis to be strongly correlated with annual mean temperature, and allele frequencies at those SNPs showed a linear relationship with this environmental variable (Fig. 3C, Fig S9). Results of GLM analysis with a binomial error provided additional support for this relationship and additionally highlighted the association of individual genotypes with temperature for those SNPs (Figure S8). It is worth noting that among SNPs selected as outliers by RDA (39), all were strongly correlated with two variables linked to temperature: annual temperature (34 SNPs) and temperature seasonality (5 SNPs) (Fig. S9C).

We also wanted to contrast weak interregional population structure that likely reflects the background neutral variation with SNPs showing departure from neutrality, to find genetic variants, that may be uniquely favored in those two pine groups. We found 18 such outlier SNPs in Poland and 8 in Finland (Fig. S10), their spatial distribution showed a less clear pattern and interestingly those SNPs were not included in the outlier set for the whole transect. Although, we found several SNPs with excess F_ST within regions, not all of those SNPs were formally selected as outliers by our criteria (selected by all outlier detection methods). Nevertheless, this pattern was especially pronounced within Polish populations (Stołowe Mts. vs rest of populations), where the mean F_ST for all SNPs was 0.013, and 31 SNPs had F_ST higher than 0.1 with maximum F_ST = 0.46. In comparison between Finnish populations (North vs South), 13 SNPs had F_ST > 0.05 with mean F_ST = 0.005).

The RDA analysis was used also to explain the partitioning of the population genetic variation between impacts of environment and geography. After removing the highly correlated variables only 8 environmental (mean_temp, mean_dr, temp_s, perc_dry_m, perc_wet_q, ph, wet and carbon) and 10 geographic dbMEM were included in the model (Figure S9A, Table S3). Together, environment and geography explained 17.8% of the total genetic variance (R² = 0.038 and 0.113, p < 0.001; “combined fractions” in Table S4). Partial redundancy analysis showed, that, 11.4% of variance was confounded between the two matrices, and that when constraining one matrix with the other, environment explained much less observed genomic variation than geography (0.1% vs 6.3%, respectively, “individual fractions” in Table S4). We then performed RDA using outlier SNPs (16 SNPs detected by genomic outliers scans and 6 PAV) and found that the contribution of environment and geography jointly explained the majority of variation, especially in the case of PAV leaving only 15.8% of variation unexplained. Surprisingly, in all of those analyses, the explanatory power of environment alone was smaller, than the exclusive impact of geography (0.1% vs 6.3% for all SNPs; 1.1% vs 17% for genomic scans; 5% vs 10.6% for PAV, Table S4).

Predicting genomics offset for future climate change

Based on the contemporary genotype-environment association found for SNPs denoted as potentially adaptive, we could predict the amount of genetic change in a population required to keep up with future climate change. The most important factor driving the allele frequency change in our study was mean annual temperature, which is predicted to increase substantially for all studied populations, regardless of the RCP scenario (with the highest increase under the RCP 8.5 scenario) and the temperature in the north is predicted to increase relatively higher than in the middle latitudes – which is also reflected in our data, and the spatial distribution of genomic vulnerability (GV) values (Fig. 4A, B, C). The predicted allele shifts were found in all populations for all PAV loci, and were consistent across three RCPs, with only a small increase between scenarios, which resulted in a similar trend for GV (RCP2.6 < RCP4.5 < RCP8.5; Fig. S11). The highest GV values were reported for FIN4 (0.21–0.30) and the lowest for POL9 (0.06–0.09). Generally, the GV increase mirrored the temperature trend (with the Northernmost population having the highest values), with some exceptions in Finland and Poland (Fig. 4C). The mean GV values for Finland and Poland were as follows: 0.114 and 0.106 for RCP2.6, 0.145 and 0.111 for RCP4.5, and 0.176 and 0.120 for RCP8.5.

Because GV is expressed as an averaged change of frequency of adaptive alleles we calculated the time required for a shift in allele frequencies when only natural selection is driving such change. We obtained estimates for the time required for such change under selective pressure ranging from extremely high (s = 0.9) to moderately low (s = 0.1), and assuming moderate starting allele frequencies (p = 0.4). For maximum GV under RCP 8.5 (worst case scenario of warming) our time was bound between 3–29 generations (75–725 years) and for minimum GV value it was 1–7 generations (25–225 years), with slightly lower estimates for milder climatic change scenario modeled in RCP2.6 (Fig. 4D, E). Although we assumed for simplicity the same starting allele frequencies for populations in Poland and Finland, this was not often the case – for some PAV, many populations from Poland had frequencies ~ 0.7–0.8. To account for these differences, we additionally show the relationship between starting allele frequencies, the strength of selection and given (∆p). Regardless of the strength of selection, the required time could be substantially longer when starting from low or approaching high allele frequencies (Fig. S12).

Our work contributes to a better understanding of the mechanisms and processes driving the distribution of genetic variation in Scots pine, a keystone tree in Eurasian forests and economically important species widely used in sylviculture. Our findings reveal that temperature is the primary factor influencing levels of genetic variation and driving the species' local adaptation, offering insights for predicting genomic offset in this species. We discuss the patterns of geographic distribution of alleles under selection along environmental gradients and provide data to advance active management and practice to safeguard the resilience of Scots pine populations in the face of ongoing climate change.

The analyzed populations were previously included in a study addressing neutral genetic variation and demographic history of Scots pine ²⁶. These populations span a vast environmental gradient, representing one of the best-known examples of adaptive variation in frost tolerance and phenology within this species ¹⁶. Despite differences in environmental conditions, population histories, and some distinct genetic characteristics, our analysis revealed that the populations share much of their neutral genetic variation, with very similar allelic frequency spectra observed at most loci. The background neutral variation is also quite uniformly distributed across the studied range (mean F_ST of 0.014), indicating low between-population genetic differentiation. Our results corroborate other reports of “missing variation” within the species with such a broad distribution range, which is likely attributed to the combined effects of multiple range shifts experienced in the past, large effective population sizes and high level of gene flow (Pyhäjärvi et al., 2020; Tyrmi et al., 2020; Wachowiak et al., 2022a; Milesi et al., 2023). Despite an overall similar genetic background in the analyzed species distribution, some signatures of divergence were observed in our data in mountain regions of southern Poland. Recent results of neutral genetic structure based on mtDNA variation suggest that Stołowe Mts in Central Sudetes could harbor a unique genetic lineage of Scots pine, probably predating the last glacial event ⁶⁸. Patterns of genetic variation in other forest tree species in this area, for instance, Pinus mugo ⁶⁹ and Rhododendron ferrugineum ⁷⁰, support the presence of such microrefugia. Postglacial recolonization also played a significant role in shaping the observed allele frequency gradients in Scots pine along the studied transect, as indicated by the cline-like pattern in ancestry found by LEA and PCA analysis. The among-population variation could, in general, be partitioned between the effects of isolation by distance (IBD) and isolation by environment (IBE). These spatial processes are not mutually exclusive and can interact to reduce gene flow among ecologically divergent populations ^71,72. In P. sylvestris, IBD was found to be a stronger driver of genetic divergence for neutral loci, but the genomic signature of IBE was especially strong for outlier loci, where only 15% of the variation was unexplained. This could be attributed to the unequal strength of gene flow affecting different parts of the genome. Certain regions such as those containing loci involved in adaptation or closely linked outliers, may be selectively constrained and less influenced by gene flow, while the remainder of the genome is more susceptible ⁷³. A similar pattern was also described in other pine species, with comparable ranges and demographic histories shaped by glaciation ^74–76.

Scots pine, similar to other keystone forest tree species was the subject of numerous quantitative and population genetic studies that aimed to identify the genetic basis of adaptive trait variation including phenology, cold tolerance, responses to drought or waterlogging ^6,16,77–82. Variations in these traits exhibit clinal patterns across the species range, and common garden experiments have further confirmed their high heritability and signatures of local adaptation ^83,84. Some quantitative genetic studies suggested the presence of genes with a strong phenotypic effect related among others to cold tolerance and phenology (reviewed in e.g. ⁸⁵. However, association genetic studies have been less conclusive, indicating the presence of adaptive genes of relatively low individual phenotypic effect ^{27,77,86–88}.

The polygenic character of many traits related to environmental gradients usually results in weak allelic frequency spectra shifts that can be detected by outlier detection approaches alone ⁵². To overcome this issue, we combined complementary methods, examining different aspects of the association between genotype and environment. Specifically, we analyzed patterns of differentiation based on F_ST, SNPs underlying population structure, and allele frequencies associated with the multivariate environmental predictors. By selecting outliers that were concordant among all three methods, we aimed to ensure both high detection power and a low false discovery rate. These identified loci contribute to overall patterns of genetic variation across the species range, as indicated in our multivariate analysis. Notably, randomly selected and putatively neutral SNPs failed to explain the observed structure in these pines. The results of our population genomic approach provide additional support to local adaptation in Scots pine being primarily influenced by selection acting on numerous loci of small effect.

Increasing evidence supporting polygenic architecture of local adaptation was recently provided for non-model organisms, including forest tree species like poplars, oaks and beeches ^32,89,90. Climate-associated traits with a complex genetic architecture, shaped in response to environmental gradients through multilocus selection were described in other species including among others humans ^91,92, fruit flies ⁹³ and Arabidopsis ⁹⁴. In the case of species with a high level of gene flow selection must be high enough to overcome its homogenizing effect to maintain local adaptation (Tigano and Friesen 2016; e.g., Chavarria-Pizarro et al. 2019). In our study, the variation in all SNPs under selection was highly correlated with mean temperature and possibly other environmental variables highly correlated with temperature, but not with precipitation. Temperature serves as a key limiting factor for tree species, especially young individuals, influencing growth patterns and phenology, and ultimately impacting tree performance and survival ^95–97. It seems that significant selective pressure exerted by temperature may be strong enough to maintain local adaptation despite gene flow in Scots pine.

Unfortunately, due to the genome complexity of conifers in general, and the lack of a good-quality, annotated reference genome for Scots pine, the molecular function of identified climate-associated SNPs remains unknown. However, given the rapid decay of linkage disequilibrium within the conifers genome (LD drops to R2 < 0.2 at ~ 200 bp, ) and the reduced genome representation method for genotyping used here (SNP array), with only a small fraction of the whole genome likely represented, we can expect a limited number of SNP outliers detected by all three genome-environment association methods ^87,98. Although scarce, the variants detected in our study must be either causative polymorphisms or nearby loci ⁸⁵, and further empirical validations would be needed to confirm their role and fitness effect. In addition to putative structural variation in genes underlying local adaptation, variants involved in regulatory mechanisms, gene expression patterns, or epigenetic changes might be equally important ^99,100. Further studies, that will take advantage of the availability of the complete genome sequence of Scots pine and methods in functional genomics and plant molecular biology are needed to fully explore those possibilities.

In a rapidly changing environment, forest tree populations face limited options for survival. These options include tracking the shifting climatic niche through migration, relying on pre-existing genetic variation and selection to adapt to novel conditions in their current location ⁷. Life history traits, especially pollen and seed dispersal rates, serve as limiting factors for effective migration, and the actual rates are quite low for many tree species ^101,102. A significant discrepancy between actual and predicted rates of migration to track future conditions has been reported, raising concerns about the ability of tree species to respond solely through natural migration ^12,34,103. Furthermore, our predicted GV values, which could be interpreted as the allele frequency change required for adaptation to future climate conditions, are relatively high for analyzed groups, with slightly higher mean values in Finland across all RCP scenarios. Considering estimates of possible allele frequency shifts in the range of 0.01 per generation at neutral and adaptive loci in a different tree species, our estimates are an order of magnitude higher. Closing the perceived offset gap by natural selection alone will thus be challenging for those populations ¹⁰⁴.

Our simulation results support these findings by illustrating that the duration required for allele shift in Scots pine populations, encompassing both minimum and maximum GV values in our dataset, ranged from 25 to 725 years. This timeframe was notably influenced by the strength of selection, with a higher selection coefficient resulting in quicker shifts, and the initial allele frequency, wherein shifts occurred more rapidly at intermediate frequencies. While there is a possibility that climate change-induced mismatches could lead to heightened mortality rates (indicative of high selection pressures), a meta-analysis investigating selection strength within natural populations delivered a mean selection coefficient of s = 0.135. The scarcity of studies reporting very high selection coefficients (s > 0.5) suggests that the time required for allele frequency changes may need to be long ⁸. Thus it is evident, that standing genetic variation in many places may become maladapted to rapidly changing environments and alternative approaches as compared to natural regeneration might be needed to ensure the resilience of Scots pine in the face of climate change ^7,40,105.

Such alternative approaches may involve forest-assisted migration (FAM) that in principle involves transferring existing genotypes from one place within the species range to an area of interest, where the projected environmental condition match those already present in the donor place ^106–109. The presented research contributes to the development and implementation of FAM in Scots pine, as it provides an evaluation of current genomic offset and indicates the optimal allelic frequency spectra in populations to ensure their resilience. Overall, our vulnerability assessment predicts, that Finnish populations, especially those located above the Arctic Circle will be more harmed, as a consequence of increased warming of the Arctic region ¹¹⁰. On the other hand, the results of species distribution changes in response to future conditions have generally predicted a substantial northward range shift in the case of Scot pine, and more optimal conditions to grow at higher latitudes ¹². This apparent contradiction may be reconciled by considering the need for turnover: pines currently growing in affected areas will likely be replaced by those migrating from lower latitudes. At smaller spatial scales, within individual regions, there is a slight variation, and the relationship between latitude and genomic vulnerability is weaker, particularly in Poland. However, for all populations studied here, the magnitude of genomic offset to future conditions, even under the mild prognosis of RCP2.5, is significant enough to suggest their insufficient adaptive response without external assistance.

Before any FAM implementation, empirical validation of the robustness of genomic prediction, such as through common garden experiments, seems to be of critical importance ¹¹¹. Subsequently, careful selection of the potential source population for assisted migration suited for each location and adequate policy adjustments are needed ¹⁰⁵. Especially challenging, given the vast range of Scots pine, are the existing regulations regarding the maintenance of seed zones and transfer of reforestation material in different countries. The seed zones were established to maintain unique ecotypes and/or to protect locally adapted populations, thus careful analysis of the potential revision of such regulations is necessary ^112,113. Furthermore, long-term monitoring and adaptive management strategies are crucial to evaluate the efficacy of FAM interventions. Despite the challenges, it seems that assisted climate-adaptive strategies must be considered soon, as lack of action will likely lead to significant losses, both ecologically and economically. Successful implementations of various assisted migration approaches in other forest tree species like black ash, red pine, and eastern white pine (Palik et al. 2022) and their frameworks could serve as valuable guidelines for developing solutions suited for Scots pine and could help mitigate those costs ^114–116.

Author Contribution

B. Ł: Conceptualization, Data curation, Formal analysis, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. W. W.: Conceptualization, Investigation, Funding acquisition, Project administration, Resources; Supervision; Validation, Writing – original draft, Writing – review & editing.

Acknowledgement

The research was funded by the Polish National Science Centre (Grant No. 2020/37/B/NZ9/01496 and 2017/27/B/NZ9/00159)

Data Availability

Data is provided within the manuscript or supplementary information files. The data presented in this study are available on request from the corresponding author.

Hamrick, J. L. & Godt, M. J. W. EffectsN of life history traits on genetic diversity in plant species. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 351, 1291–1298, doi:DOI 10.1098/rstb.1996.0112 (1996).
Hamrick, J. L. Response of forest trees to global environmental changes. Forest Ecology and Management 197, 323–335, doi:10.1016/j.foreco.2004.05.023 (2004).
Nybom, H. Comparison of different nuclear DNA markers for estimating intraspecific genetic diversity in plants. Molecular ecology 13, 1143–1155, doi:10.1111/j.1365-294X.2004.02141.x (2004).
Petit, R. J. & Hampe, A. Some evolutionary consequences of being a tree. Annual Review of Ecology Evolution and Systematics 37, 187–214, doi:10.1146/annurev.ecolsys.37.091305.110215 (2006).
Franks, S. J., Weber, J. J. & Aitken, S. N. Evolutionary and plastic responses to climate change in terrestrial plant populations. Evolutionary Applications 7, 123–139, doi:10.1111/eva.12112 (2014).
Alberto, F. J. et al. Potential for evolutionary responses to climate change - evidence from tree populations. Glob Chang Biol 19, 1645–1661, doi:10.1111/gcb.12181 (2013).
Aitken, S. N., Yeaman, S., Holliday, J. A., Wang, T. & Curtis-McLane, S. Adaptation, migration or extirpation: climate change outcomes for tree populations. Evol Appl 1, 95–111, doi:10.1111/j.1752-4571.2007.00013.x (2008).
Thurman, T. J. & Barrett, R. D. H. The genetic consequences of selection in natural populations. Molecular ecology 25, 1429–1448, doi:10.1111/mec.13559 (2016).
Barrett, R. D. & Schluter, D. Adaptation from standing genetic variation. Trends Ecol Evol 23, 38–44, doi:10.1016/j.tree.2007.09.008 (2008).
Davis, M. B. & Shaw, R. G. Range shifts and adaptive responses to Quaternary climate change. Science 292, 673–679, doi:10.1126/science.292.5517.673 (2001).
Intergovernmental Panel on Climate, C. Climate Change 2022 – Impacts, Adaptation and Vulnerability. (Cambridge University Press, 2023).
Dyderski, M. K., Paz, S., Frelich, L. E. & Jagodzinski, A. M. How much does climate change threaten European forest tree species distributions? Glob Chang Biol 24, 1150–1163, doi:10.1111/gcb.13925 (2018).
Pecchi, M. et al. Species distribution modelling to support forest management. A literature review. Ecol Model 411, doi: 108817 10.1016/j.ecolmodel.2019.108817 (2019).
Thuiller, W. et al. Niche-based modelling as a tool for predicting the risk of alien plant invasions at a global scale. Global Change Biology 11, 2234–2250, doi:10.1111/j.1365-2486.2005.001018.x (2005).
Nogues-Bravo, D. et al. Cracking the Code of Biodiversity Responses to Past Climate Change. Trends Ecol Evol 33, 765–776, doi:10.1016/j.tree.2018.07.005 (2018).
Savolainen, O., Pyhäjärvi, T. & Knürr, T. Gene flow and local adaptation in trees. Annual Review of Ecology Evolution and Systematics 38, 595–619, doi:10.1146/annurev.ecolsys.38.091206.095646 (2007).
Soranzo, N., Alia, R., Provan, J. & Powell, W. Patterns of variation at a mitochondrial sequence-tagged-site locus provides new insights into the postglacial history of European Pinus sylvestris populations. Molecular ecology 9, 1205–1211, doi:10.1046/j.1365-294x.2000.00994.x (2000).
Cheddadi, R. et al. Imprints of glacial refugia in the modern genetic diversity of. Global Ecology and Biogeography 15, 271–282, doi:10.1111/j.1466-822x.2006.00226.x (2006).
Semerikov, V. L. et al. Colonization history of Scots pine in Eastern Europe and North Asia based on mitochondrial DNA variation. Tree Genetics & Genomes 14, 8, doi:10.1007/s11295-017-1222-0 (2018).
Naydenov, K., Senneville, S., Beaulieu, J., Tremblay, F. & Bousquet, J. Glacial vicariance in Eurasia: mitochondrial DNA evidence from Scots pine for a complex heritage involving genetically distinct refugia at mid-northern latitudes and in Asia Minor. BMC Evol Biol 7, 233, doi:10.1186/1471-2148-7-233 (2007).
Naydenov, K. D., Naydenov, M. K., Tremblay, F., Alexandrov, A. & Aubin-Fournier, L. D. Patterns of genetic diversity that result from bottlenecks in Scots Pine and the implications for local genetic conservation and management practices in Bulgaria. New Forest 42, 179–193, doi:10.1007/s11056-010-9245-5 (2011).
Pyhajarvi, T. et al. Demographic history has influenced nucleotide diversity in European Pinus sylvestris populations. Genetics 177, 1713–1724, doi:10.1534/genetics.107.077099 (2007).
Dering, M. et al. Tertiary remnants and Holocene colonizers: Genetic structure and phylogeography of Scots pine reveal higher genetic diversity in young boreal than in relict Mediterranean populations and a dual colonization of Fennoscandia. Diversity and Distributions 23, 540–555, doi:10.1111/ddi.12546 (2017).
Wachowiak, W. et al. Phylogeography of Scots pine in Europe and Asia based on mtDNA polymorphisms. Journal of Systematics and Evolution 61, 315–327, doi:10.1111/jse.12907 (2023).
Bruxaux, J. et al. Scots pine – panmixia and the elusive signal of genetic adaptation. New Phytologist n/a, doi:https://doi.org/10.1111/nph.19563 (2024).
Łabiszak, B. & Wachowiak, W. Mid-Pleistocene events influenced the current spatial structure of genetic diversity in Scots pine (Pinus sylvestris L.). Journal of Systematics and Evolution n/a, doi:https://doi.org/10.1111/jse.13013.
Hall, D. et al. Divergent patterns between phenotypic and genetic variation in Scots pine. Plant Communications 2, 100139, doi:https://doi.org/10.1016/j.xplc.2020.100139 (2021).
Filipe, J. C. et al. Signatures of natural selection in a foundation tree along Mediterranean climatic gradients. Molecular ecology 31, 1735–1752, doi:10.1111/mec.16351 (2022).
Tigano, A. et al. Genomic vulnerability of a freshwater salmonid under climate change. Evolutionary Applications, doi:10.1111/eva.13602 (2023).
Tigano, A., Colella, J. P. & MacManes, M. D. Comparative and population genomics approaches reveal the basis of adaptation to deserts in a small rodent. Molecular ecology 29, 1300–1314, doi:10.1111/mec.15401 (2020).
Marková, S. et al. Local adaptation and future climate vulnerability in a wild rodent. Nature Communications 14, doi: 784010.1038/s41467-023-43383-z (2023).
Sang, Y. P. et al. Genomic insights into local adaptation and future climate-induced vulnerability of a keystone forest tree in East Asia. Nature Communications 13, doi: 6541 10.1038/s41467-022-34206-8 (2022).
Koo, K. A. et al. Potential climate change effects on tree distributions in the Korean Peninsula: Understanding model & climate uncertainties. Ecol Model 353, 17–27, doi:10.1016/j.ecolmodel.2016.10.007 (2017).
Zhu, K., Woodall, C. W. & Clark, J. S. Failure to migrate: lack of tree range expansion in response to climate change. Global Change Biology 18, 1042–1052, doi:https://doi.org/10.1111/j.1365-2486.2011.02571.x (2012).
Sittaro, F., Paquette, A., Messier, C. & Nock, C. A. Tree range expansion in eastern North America fails to keep pace with climate warming at northern range limits. Global Change Biology 23, 3292–3301, doi:https://doi.org/10.1111/gcb.13622 (2017).
Iverson, L. R. & McKenzie, D. Tree-species range shifts in a changing climate: detecting, modeling, assisting. Landscape Ecol 28, 879–889, doi:10.1007/s10980-013-9885-x (2013).
Hanewinkel, M., Cullmann, D. A., Schelhaas, M. J., Nabuurs, G. J. & Zimmermann, N. E. Climate change may cause severe loss in the economic value of European forest land. Nat Clim Change 3, 203–207, doi:10.1038/Nclimate1687 (2013).
Schueler, S. et al. Vulnerability of dynamic genetic conservation units of forest trees in Europe to climate change. Global Change Biology 20, 1498–1511, doi:10.1111/gcb.12476 (2014).
Oliver, T. H., Smithers, R. J., Beale, C. M. & Watts, K. Are existing biodiversity conservation strategies appropriate in a changing climate? Biol Conserv 193, 17–26, doi:https://doi.org/10.1016/j.biocon.2015.10.024 (2016).
Aitken, S. N. & Whitlock, M. C. Assisted Gene Flow to Facilitate Local Adaptation to Climate Change. Annu Rev Ecol Evol S 44, 367-+, doi:10.1146/annurev-ecolsys-110512-135747 (2013).
Goudet, J. HIERFSTAT, a package for R to compute and test hierarchical statistics. Molecular Ecology Notes 5, 184–186, doi:10.1111/j.1471-8286.2004.00828.x (2005).
R: A Language and Environment for Statistical Computing (2020).
Weir, B. S. & Cockerham, C. C. Estimating F-Statistics for the Analysis of Population Structure. Evolution 38, 1358–1370, doi:10.1111/j.1558-5646.1984.tb05657.x (1984).
Jombart, T. & Ahmed, I. adegenet 1.3-1: new tools for the analysis of genome-wide SNP data. Bioinformatics 27, 3070–3071, doi:10.1093/bioinformatics/btr521 (2011).
Jombart, T. adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24, 1403–1405, doi:10.1093/bioinformatics/btn129 (2008).
Frichot, E. & François, O. LEA: An R package for landscape and ecological association studies. Methods in Ecology and Evolution 6, 925–929, doi:10.1111/2041-210x.12382 (2015).
Frichot, E., Mathieu, F., Trouillon, T., Bouchard, G. & François, O. Fast and Efficient Estimation of Individual Ancestry Coefficients. Genetics 196, 973-+, doi:10.1534/genetics.113.160572 (2014).
Francis, R. M. pophelper: an R package and web app to analyse and visualize population structure. Mol Ecol Resour 17, 27–32, doi:10.1111/1755-0998.12509 (2017).
Wickham, H. ggplot2. Wires Comput Stat 3, 180–185, doi:10.1002/wics.147 (2011).
Wickham, H. ggplot2 Elegant Graphics for Data Analysis. (Springer International Publishing, 2016).
Oksanen, J. et al. vegan: Community Ecology Package. (2022).
Whitlock, M. C. & Lotterhos, K. E. Reliable Detection of Loci Responsible for Local Adaptation: Inference of a Null Model through Trimming the Distribution of FST. The American Naturalist 186, S24-S36, doi:10.1086/682949 (2015).
Whitlock, M. C. & Lotterhos, K. OutFLANK: Fst outliers with trimming. (2014).
Storey, J. D., Bass, A. J., Dabney, A. & Robinson, D. qvalue: Q-value estimation for false discovery rate control. (2022).
Prive, F., Luu, K., Vilhjalmsson, B. J. & Blum, M. G. B. Performing Highly Efficient Genome Scans for Local Adaptation with R Package pcadapt Version 4. Mol Biol Evol 37, 2153–2154, doi:10.1093/molbev/msaa053 (2020).
Luu, K., Bazin, E. & Blum, M. G. pcadapt: an R package to perform genome scans for selection based on principal component analysis. Mol Ecol Resour 17, 67–77, doi:10.1111/1755-0998.12592 (2017).
Dormann, C. F. et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27–46, doi:10.1111/j.1600-0587.2012.07348.x (2013).
Zuur, A. F., Ieno, E. N. & Elphick, C. S. A protocol for data exploration to avoid common statistical problems. Methods in Ecology and Evolution 1, 3–14, doi:10.1111/j.2041-210X.2009.00001.x (2010).
Dray, S. et al. adespatial: Multivariate Multiscale Spatial Analysis. (2023).
Rellstab, C. et al. Signatures of local adaptation in candidate genes of oaks (Quercus spp.) with respect to present and future climatic conditions. Molecular ecology 25, 5907–5924, doi:10.1111/mec.13889 (2016).
Bay, R. A. et al. Genomic signals of selection predict climate-driven population declines in a migratory bird. Science 359, 83-+, doi:10.1126/science.aan4380 (2018).
Fick, S. E. & Hijmans, R. J. WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. Int J Climatol 37, 4302–4315, doi:10.1002/joc.5086 (2017).
Wu, T. et al. The Beijing Climate Center Climate System Model (BCC-CSM): the main progress from CMIP5 to CMIP6. Geosci. Model Dev. 12, 1573–1600, doi:10.5194/gmd-12-1573-2019 (2019).
Kindt, R. AlleleShift: An R package to predict and visualize population-level changes in allele frequencies in response to climate change. PeerJ, doi:10.7717/peerj.11534 (2021).
Gräler, B., Pebesma, E. J. & Heuvelink, G. B. M. Spatio-Temporal Interpolation using gstat. R J. 8, 204 (2016).
Boratyński, A. in Scots pine biology (eds S. Białobok, A. Boratyński, & W. Bugała) 33–44 (Instytut Dendrologii Polskiej Akademii Nauk, 1993).
Wachowiak, W., Palme, A. E. & Savolainen, O. Speciation history of three closely related pines Pinus mugo (T.), P. uliginosa (N.) and P. sylvestris (L.). Molecular ecology 20, 1729–1743, doi:10.1111/j.1365-294X.2011.05037.x (2011).
Wachowiak, W. et al. Phylogeography of Scots pine in Europe and Asia based on mtDNA polymorphisms. Journal of Systematics and Evolution 61, 315–327, doi:10.1111/jse.12907 (2022).
Zukowska, W. B. & Wachowiak, W. Nuclear microsatellite markers reveal the low genetic structure of Pinus mugo Turra (dwarf mountain pine) populations in Europe. Plant Syst Evol 303, 641–651, doi:10.1007/s00606-017-1395-x (2017).
Suchan, T., Malicki, M. & Ronikier, M. Relict populations and Central European glacial refugia: The case of Rhododendron ferrugineum (Ericaceae). Journal of Biogeography 46, 392–404, doi:https://doi.org/10.1111/jbi.13512 (2019).
DeWoody, J., Trewin, H. & Taylor, G. Genetic and morphological differentiation in Populus nigra L.: isolation by colonization or isolation by adaptation? Molecular ecology 24, 2641–2655, doi:https://doi.org/10.1111/mec.13192 (2015).
Papadopulos, A. S. et al. Evaluation of genetic isolation within an island flora reveals unusually widespread local adaptation and supports sympatric speciation. Philos Trans R Soc Lond B Biol Sci 369, doi:10.1098/rstb.2013.0342 (2014).
Tigano, A. & Friesen, V. L. Genomics of local adaptation with gene flow. Molecular ecology 25, 2144–2164, doi:10.1111/mec.13606 (2016).
Nadeau, S. et al. Contrasting patterns of genetic diversity across the ranges of Pinus monticola and P. strobus: A comparison between eastern and western North American postglacial colonization histories. American Journal of Botany 102, 1342–1355, doi:10.3732/ajb.1500160 (2015).
Nadeau, S., Meirmans, P. G., Aitken, S. N., Ritland, K. & Isabel, N. The challenge of separating signatures of local adaptation from those of isolation by distance and colonization history: The case of two white pines. Ecology and Evolution 6, 8649–8664, doi:10.1002/ece3.2550 (2016).
Xia, H. H. et al. Combining mitochondrial and nuclear genome analyses to dissect the effects of colonization, environment, and geography on population structure in. Evolutionary Applications 11, 1931–1945, doi:10.1111/eva.12697 (2018).
Kujala, S. T. et al. Genetic heterogeneity underlying variation in a locally adaptive clinal trait in Pinus sylvestris revealed by a Bayesian multipopulation analysis. Heredity 118, 413–423, doi:10.1038/hdy.2016.115 (2017).
Salmela, M. J., Cavers, S., Cottrell, J. E., Iason, G. R. & Ennos, R. A. Spring phenology shows genetic variation among and within populations in seedlings of Scots pine (Pinus sylvestris L.) in the Scottish Highlands. Plant Ecol Divers 6, 523–536, doi:10.1080/17550874.2013.795627 (2013).
Hurme, P., Repo, T., Savolainen, O. & Paakkonen, T. Climatic adaptation of bud set and frost hardiness in Scots pine (Pinus sylvestris). Canadian Journal of Forest Research 27, 716–723, doi:DOI 10.1139/cjfr-27-5-716 (1997).
Donnelly, K., Cavers, S., Cottrell, J. E. & Ennos, R. A. Cryptic genetic variation and adaptation to waterlogging in Caledonian Scots pine, Pinus sylvestris L. Ecology and Evolution 8, 8665–8675, doi:10.1002/ece3.4389 (2018).
Wachowiak, W., Perry, A., Donnelly, K. & Cavers, S. Early phenology and growth trait variation in closely related European pine species. Ecol Evol 8, 655–666, doi:10.1002/ece3.3690 (2018).
Semerci, A. et al. Morphological and physiological responses to drought stress of European provenances of Scots pine. European Journal of Forest Research 136, 91–104, doi:10.1007/s10342-016-1011-6 (2017).
Giertych, M. in Developments in Plant Genetics and Breeding Vol. 3 (ed Maciej Giertych) 87–101 (Elsevier, 1991).
Pyhajarvi, T., Kujala, S. T. & Savolainen, O. 275 years of forestry meets genomics in Pinus sylvestris. Evol Appl 13, 11–30, doi:10.1111/eva.12809 (2020).
Neale, D. B. & Savolainen, O. Association genetics of complex traits in conifers. Trends Plant Sci 9, 325–330, doi:https://doi.org/10.1016/j.tplants.2004.05.006 (2004).
Tyrmi, J. S. et al. Genomics of Clinal Local Adaptation in Pinus sylvestris Under Continuous Environmental and Spatial Genetic Setting. G3 (Bethesda) 10, 2683–2696, doi:10.1534/g3.120.401285 (2020).
Wachowiak, W., Balk, P. A. & Savolainen, O. Search for nucleotide diversity patterns of local adaptation in dehydrins and other cold-related candidate genes in Scots pine (Pinus sylvestris L.). Tree Genetics & Genomes 5, 117–132, doi:10.1007/s11295-008-0188-3 (2009).
Kujala, S. T. & Savolainen, O. Sequence variation patterns along a latitudinal cline in Scots pine (Pinus sylvestris): signs of clinal adaptation? Tree Genetics & Genomes 8, 1451–1467, doi:10.1007/s11295-012-0532-5 (2012).
Yuan, S. et al. Genomic vulnerability to climate change in Quercus acutissima, a dominant tree species in East Asian deciduous forests. Molecular ecology 32, 1639–1655, doi:10.1111/mec.16843 (2023).
Niels, A. M. et al. Genomic variation of a keystone forest tree species reveals patterns of local adaptation and future maladaptation. bioRxiv, 2023.2005.2011.540382, doi:10.1101/2023.05.11.540382 (2023).
Field, Y. et al. Detection of human adaptation during the past 2000 years. Science 354, 760–764, doi:10.1126/science.aag0776 (2016).
Sella, G. & Barton, N. H. Thinking About the Evolution of Complex Traits in the Era of Genome-Wide Association Studies. Annu Rev Genomics Hum Genet 20, 461–493, doi:10.1146/annurev-genom-083115-022316 (2019).
Barghi, N. et al. Genetic redundancy fuels polygenic adaptation in Drosophila. PLOS Biology 17, e3000128, doi:10.1371/journal.pbio.3000128 (2019).
Hancock, A. M. et al. Adaptation to Climate Across the Arabidopsis thaliana Genome. Science 334, 83–86, doi:10.1126/science.1209244 (2011).
Niinemets, Ü. & Valladares, F. Tolerance to shade, drought, and waterlogging of temperate Northern Hemisphere trees and shrubs. Ecol Monogr 76, 521–547, doi:10.1890/0012-9615(2006)076[0521:Ttsdaw]2.0.Co;2 (2006).
Pallardy, S. G. & Kozlowski, T. T. (2008). Physiology of woody plants. (Elsevier, Amsterdam, 2008).
Ryan, M. G. Temperature and tree growth. Tree Physiology 30, 667–668, doi:10.1093/treephys/tpq033 (2010).
Dvornyk, V., Sirviö, A., Mikkonen, M. & Savolainen, O. Low Nucleotide Diversity at the pal1 Locus in the Widely Distributed Pinus sylvestris. Mol Biol Evol 19, 179–188, doi:10.1093/oxfordjournals.molbev.a004070 (2002).
Verhoeven, K. J. F., vonHoldt, B. M. & Sork, V. L. Epigenetics in ecology and evolution: what we know and what we need to know. Molecular ecology 25, 1631–1638, doi:https://doi.org/10.1111/mec.13617 (2016).
Sork, V. L. Genomic Studies of Local Adaptation in Natural Plant Populations. Journal of Heredity 109, 3–15, doi:10.1093/jhered/esx091 (2017).
Meier, E. S., Lischke, H., Schmatz, D. R. & Zimmermann, N. E. Climate, competition and connectivity affect future migration and ranges of European trees. Global Ecology and Biogeography 21, 164–178, doi:10.1111/j.1466-8238.2011.00669.x (2012).
McLachlan, J. S., Hellmann, J. J. & Schwartz, M. W. A framework for debate of assisted migration in an era of climate change. Conservation Biology 21, 297–302, doi:10.1111/j.1523-1739.2007.00676.x (2007).
Malcolm, J. R., Markham, A., Neilson, R. P. & Garaci, M. Estimated migration rates under scenarios of global climate change. Journal of Biogeography 29, 835–849, doi:DOI 10.1046/j.1365-2699.2002.00702.x (2002).
Dauphin, B. et al. Genomic vulnerability to rapid climate warming in a tree species with a long generation time. Global Change Biology 27, 1181–1195, doi:10.1111/gcb.15469 (2021).
Palik, B. J., Clark, P. W., D'Amato, A. W., Swanston, C. & Nagel, L. Operationalizing forest-assisted migration in the context of climate change adaptation: Examples from the eastern USA. Ecosphere 13, doi: e426010.1002/ecs2.4260 (2022).
Nagel, L. M. et al. Adaptive Silviculture for Climate Change: A National Experiment in Manager-Scientist Perships to Apply an Adaptation Framework. J Forest 115, 167–178, doi:10.5849/jof.16-039 (2017).
Park, A. & Talbot, C. Information Underload: Ecological Complexity, Incomplete Knowledge, and Data Deficits Create Challenges for the Assisted Migration of Forest Trees. BioScience 68, 251–263, doi:10.1093/biosci/biy001 (2018).
Williams, M. I. & Dumroese, R. K. Preparing for Climate Change: Forestry and Assisted Migration. J Forest 111, 287–297, doi:10.5849/jof.13-016 (2013).
Pedlar, J. H. et al. Placing Forestry in the Assisted Migration Debate. BioScience 62, 835–842, doi:10.1525/bio.2012.62.9.10 (2012).
Rantanen, M. et al. The Arctic has warmed nearly four times faster than the globe since 1979. Communications Earth & Environment 3, 168, doi:10.1038/s43247-022-00498-3 (2022).
Capblancq, T., Fitzpatrick, M. C., Bay, R. A., Exposito-Alonso, M. & Keller, S. R. Genomic Prediction of (Mal)Adaptation Across Current and Future Climatic Landscapes. Annual Review of Ecology, Evolution and Systematics 51, 245–269, doi:https://doi.org/10.1146/annurev-ecolsys-020720-042553 (2020).
McKenney, D., Pedlar, J. & O’Neill, G. Climate change and forest seed zones: Past trends, future prospects and challenges to ponder. The Forestry Chronicle 85, 258–266, doi:10.5558/tfc85258-2 (2009).
Karasov-Olson, A. et al. Co-development of a risk assessment strategy for managed relocation. Ecological Solutions and Evidence 2, e12092, doi:https://doi.org/10.1002/2688-8319.12092 (2021).
Janowiak, M. K. et al. A Practical Approach for Translating Climate Change Adaptation Principles into Forest Management Actions. J Forest 112, 424–433, doi:10.5849/jof.13-094 (2014).
Brandt, L. et al. A framework for adapting urban forests to climate change. Environmental Science & Policy 66, 393–402, doi:https://doi.org/10.1016/j.envsci.2016.06.005 (2016).
Swanston, C. W. et al. Forest Adaptation Resources: climate change tools and approaches for land managers. 2nd ed. (U.S. Department of Agriculture, Forest Service, Northern Research Station, 2016).

No competing interests reported.

SupplementarymaterialsBWW.docx

Download PDF

Reviewers agreed at journal
27 Oct, 2024
Reviewers agreed at journal
24 Oct, 2024
Reviews received at journal
05 Sep, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers agreed at journal
26 Aug, 2024
Reviewers invited by journal
23 Aug, 2024
Editor assigned by journal
21 Aug, 2024
Editor invited by journal
11 May, 2024
Submission checks completed at journal
08 May, 2024
First submitted to journal
06 May, 2024

You are reading this latest preprint version

Adaptive potential and genomic vulnerability of keystone forest tree species to climate change: a case study in Scots pine

Status:

Version 1

Abstract

Figures

Introduction

Material and Methods

Results

Discussion

Declarations

Author Contribution

Acknowledgement

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1