1. Introduction
Forests play an important role in the mitigation of the effects of climate change. Therefore, climate smart forestry (CSF) may significantly invigorate the positive effects of forests and provide further benefits for sustainable ecosystems (
Nabuurs et al. 2018). Regional environmental conditions as well as traditional forest management may strongly differ; however, the overall integrated survey and development of indicators creates good possibilities for revealing the synergies, evaluating local aspects, and finally building up local strategies for sustaining resilient forests. Providing a guiding framework for future research, practice, and policy, a comprehensive and shared definition of CSF and of the indicators of the “climate-smartness” have been recently suggested by the EU COST CLIMO (
CLImate Smart Forestry in
MOuntain Regions) action (
http://climo.unimol.it/). A total of 29 indicators for assessing CSF were selected; they refer to adaptation and mitigation strategies, considering also the benefits that forests provide to the society. According to
Bowditch et al. (2020), CSF “…is a sustainable adaptive forest management and governance to protect and enhance the potential of forests to adapt to and mitigate climate change. The aim is to sustain ecosystem integrity and functions and to ensure the continuous delivery of ecosystem goods and services while minimising the impact of climate-induced changes on well-being and nature’s contribution to people”. Genetic resources are listed as part of the core group of CSF indicators for “forest biological diversity” (
Bowditch et al. 2020); these indicators contribute in the evaluation of forest health and vitality, which may strengthen adaptation and mitigation measures and are crucial for protecting and maintaining other forest functions and services. Regarding the genetic constitution of species and their populations, studies have revealed that a high level of “standing genetic diversity” preserved in species and their populations is a principal prerequisite to face environmental changes and provide the ability for the populations to survive heterogeneous climatic and spatial conditions (
Vornam et al. 2004;
De Carvalho et al. 2010). While adaptation induced by selection and fixation of new mutations takes comparatively long in species having a long life span, like forest trees, standing variation most probably has already passed through “selective filters” and might have been formerly tested during selection in past environments (
Barrett and Schluter 2008). Moreover, the precondition of fast and effective genetic adaptation in all species is a sufficiently large genetic diversity (
Mátyás and Kramer 2016). Genetic characteristics of living tree populations depend on the historical demography of populations, ancestral dynamics of the effective population size, expansions, declines or divergence at geological time scale (
Hewitt 2004).
Evolution and historical demography of European forest tree species have already been deeply studied by use of molecular markers, revealing different patterns of variation. Range wide studies were performed on European beech (
Fagus sylvatica L.), a keystone tree species of the mountain forest communities, using isozymes (
Belletti and Lanteri 1996;
Comps et al. 2001;
Gömöry et al. 2003) and chloroplast and nuclear DNA markers (
Demesure et al. 1996;
Vornam et al. 2004;
Buiteveld et al. 2007). A combination of palaeobotanical and genetic data have revealed that beech sustained in multiple refugia during the last glacial period, among which the Central European refugium was separated from that of the Mediterranean. Along the mountain chains, the expansion of some of the beech populations was facilitated but not all refugial zones contributed in the colonization of Europe (
Magri et al. 2006;
Magri 2008). Modern populations with their diversity strongly shaped by the multiple glacial and interglacial cycles are likely descendants of the once existing extra-Mediterranean refugial populations, originating from Central Europe and extending toward the North (
Magri et al. 2006).
Several studies on regional levels have evaluated the local patterns of genetic diversity (
Ballian et al. 2012;
Leonardi and Menozzi 1996;
Gömöry et al. 1999;
Cvrčková et al. 2017). Analysing fine-scale spatial distribution of different genotypes in an isolated beech population in Saxony,
Vornam et al. (2004) found a clumped spatial pattern of the genotypes within the stand, up to 30 m distance. This family structure of closely related nearby individuals may be the result of limited or oriented gene-flow (dominant wind direction) or may be due to preferential mating among nearby individuals, if selection by inbreeding depression has not occurred. Hence, local aspects are able to shape the demographic history of populations. More than that, it has been presumed that human activity linked to forest management has an impact on the distribution of the genetic diversity in populations. However, with pairwise comparisons among stands differing in management,
Buiteveld et al. (2007) could not detect clear effects of management (shelterwood system) on the genetic constitution (e.g., allelic richness, rare alleles) of beech populations.
Developing CSF indicators, especially at the local level, requires collection of data from long-term plots as well as from newly established plots, to evaluate the genetic attributes at both the stand and landscape level (
Bowditch et al. 2020). Among other indicators, the genetic characteristics of populations will enable the analyses of trends in CSF and allow the identification of priority areas for adaptation and mitigation. In a similar way, the overall evaluation of the standing genetic diversity provides a tool to monitor forest environmental services.
This study is a report on the molecular genetic variation of beech stands from the established study plots within the COST CLIMO project (
Pretzsch et al. 2021) with the aim to characterize the standing genetic diversity within the plots and provide empirical genetic data for assessing CSF indicators in mountain forest ecosystems. Nuclear microsatellites used in previous studies were screened to characterize the genetic diversity and aspects of the regional distribution of the genetic variation in relation to local environmental conditions.
4. Discussion
High genome-level variability represents the key fitness component that helps populations to survive in different environments (
Scott et al. 2020). Under climate change, when rapid shifts in environmental conditions take place, genetically more variable individuals are presumed to have more chances of survival and successful reproduction. The heterozygosity measure therefore is an important but easy to obtain indicator, predicting the survival potential of individuals and their populations specifically under severe conditions (
Scott et al. 2020).
Continuously changing and newly developing markers and methods are applied for detecting population and individual genetic variation, among which highly variable microsatellite markers are still frequently used. Due to their high polymorphism, even a relatively low number of markers can screen the genome and characterize the genetic potential of populations (
Allendorf et al. 2013). In this study, we analysed the standing genetic diversity of 12 European beech stands distributed across the species’ distribution range. All beech stands of the established CLIMO study plots exhibited high levels of genetic variation based on the six nuclear DNA microsatellite markers. This is in line with former studies performed in natural beech populations, and it is attributed mostly to the high outcrossing rate characteristic for most wind pollinated tree species (
Petit and Hampe 2006). However, observed heterozygosity values detected in our study (
HO: 0.69–0.93) were higher compared with Czech populations (
HO: 0.664–0.754;
Cvrčková et al. 2017) and with other European populations (
HO: 0.560–0.721;
Buiteveld et al. 2007).
The highest diversity values were found at sites from the Balkan Peninsula, in Bosnia and the Southern Carpathians in Romania, where genetic hotspots and several possible glacial refugia were formerly reported (
Gömöry et al. 2010). Moreover, the complex history of the Balkan Peninsula that remained moderately favourable for the survival of temperate forest trees in the past glacial period is expected to have promoted formation and long-term persistence of multiple genetic lineages that are reflected in high levels of allelic and haplotypic richness (
Gömöry et al. 2020).
Despite the fact that in most studies, heterozygote deficiency was reported in beech forest populations (
Vornam 2004;
Buiteveld et al. 2007;
Cvrčková et al. 2017), in our study, most of the stands showed an excess of heterozygotes with the highest
HO value at the Hungarian site (
Fig. 1;
Table 2). Former studies have revealed that for mixed-mating forest tree species, seeds and younger stands often show an excess of homozygotes (
Muona 1989), but with increasing age, this excess is selectively removed (
Buiteveld et al. 2007). As in this study, only mature, old trees were sampled within the study plots, heterozygotes could be more frequent than expected because of this age effect. The highest observed heterozygosity was detected at the Hungarian site, where beech sustains at a low elevation (640 m a.s.l.) and the forest stand is close to the eastern xeric limit of the species. A fine-scale analysis showed that late spring and summer drought strongly affects the limits of beech distribution in the area (
Czúcz et al. 2013). One can presume that abiotic stress acting as a selective force on the genome (massively) favours trees with overall high heterozygosity. Interestingly, a recent study in Swiss stone pine (
Pinus cembra L.) found similar effects when comparing central and marginal stands (
Dauphin et al. 2020). Even though the analysed genomic regions in our study are possibly selectively neutral, they may be associated with more adaptive sites, or adaptation may affect so many sites in the genome that many of those are linked to microsatellites. Given that similar findings hold also for isozymes, it may be assumed that trees with their long lifespan and sedentary lifestyle generally benefit from diversity (i.e., heterozygosity) throughout their genomes. This could simply be due to alleles (especially their promoters) acting differently to environmental cues and due to advantages of the combinations of expressed alleles in certain situations (
Rodríguez-Quilón et al. 2015;
Lindtke et al. 2012). On the other hand, it cannot be excluded that increased observed heterozygosity may result from forest management, as the highest and largest individuals were maintained in these plots. This would point into the same direction that more heterozygous trees are at an advantage regarding growth. However, in our study, tree height and trunk diameter (of the remaining and selected trees) were not significantly correlated with the observed heterozygosity of the individuals (data not shown), similar to previous investigations (
Buiteveld et al. 2007).
Differentiation among study regions was generally weak; however, there was considerable within-region differentiation especially within the Balkan region. A Mantel test (
r = 0.81,
p < 0.001) showed that genetic differentiation increased significantly with the geographical distance among sites, pronouncedly for sites at a distance higher than 750 km. Bayesian clustering using STRUCTURE and TESS identified the most probable number (
K) of groups of populations to be
K = 5, showing a well-separated group south of the Pyrenees in Spain. This is in line with earlier findings based on cpDNA, reflecting a different gene pool for beech in Western Europe (
Magri et al. 2006). Another well-separated group comprised all samples from the Bulgarian site, where a higher number of specific alleles was also detected. The distinct genetic pattern may indicate that the Bulgarian samples originate from a local relict population, or alternatively, they may have a special allele composition because of the stand characteristics and as a consequence of past historical events. Sampled trees are growing just below the tree line at high elevation (1365 m a.s.l.) and are very old (350 years of age). Forest cover in the past was low in this area due to widespread pastures. This type of land use has only declined during the last century, and since then forest cover regenerated naturally (T. Zlatanov, Bulgarian Academy of Sciences, personal communication). Moreover, the location is close to the distribution of the sister species,
Fagus orientalis Lipsky, but no visible signs of hybridization or introgression were formerly reported from this area (
Comps et al. 2001). However, gene flow has been previously detected between the two species in Bulgaria, using genetic markers (
Paule 1995).
Further groups in the Bayesian clustering comprised the rest of the study sites slightly separating samples originating from Central Europe (Germany, northern Italy, and southwestern Poland) from those of the Southern Carpathians and the Balkans, supporting the theory of past multiple refugia and colonization routes reported in earlier studies (
Magri 2008;
Gömöry et al. 2020).
However, many recent studies have shifted their attention from describing genetic parameters towards genetic processes. Habitat quality variables, affected by long-term climatic trends, are expected to explain some genetic variation and may be the drivers of local population demography (
Nicolé et al. 2011). In our study, regression analysis on climate variables showed significant correlation with the number of alleles and Shannon’s information index. Higher average temperature and fewer frost days were associated with higher number of alleles and higher Shannon’s index. Whether this correlation simply reflects the closeness to southern ice age refugia or has some other reason, would require further study.
In conclusion, our study reflects local aspects of the standing genetic diversity in 12 pure CLIMO beech stands along the distribution range of the species. High genetic diversity revealed at these empirical study sites is in accordance with former studies performed on beech populations and reflects the natural character of the studied beech forests where the plots have been established. Besides the overall high level of genetic variation within stands, we also found genetically diverging sites reflecting local stand growth characteristics. The genetic parameters of each stand could be assessed as a resource for CSF indicators, interpreted especially at the local level. We stress the important role of high heterozygosity in response to varying or fluctuating climate conditions; highly heterozygous trees may be more “climate smart” (i.e., in a better position in a changing climate) than others. A high level of standing genetic diversity preserved in populations should be considered indeed a precondition for good performance. Genetic parameters should therefore be taken into account when comparing growth and yield patterns for forest tree populations.