• Energy Research

The fate of peripheral forest tree populations is of particular interest in the context of climate change. These populations may concurrently be those where the most significant evolutionary changes will occur ; those most facing increasing extinction risk ; the source of migrants for the colonization of new areas at leading edges ; or the source of genetic novelty for reinforcing standing genetic variation in various parts of the range. Deciding which strategy to implement for conserving and sustainably using the genetic resources of peripheral forest tree populations is a challenge. Here, we review the genetic and ecological processes acting on different types of peripheral populations and indicate why these processes may be of general interest for adapting forests and forest management to climate change. We particularly focus on peripheral populations at the rear edge of species distributions where environmental challenges are or will become most acute. We argue that peripheral forest tree populations are ‘‘natural laboratories” for resolving priority research questions such as how the complex interaction between demographic processes and natural selection shape local adaptation ; and whether genetic adaptation will be sufficient to allow the long-term persistence of species within their current distribution. Peripheral populations are key assets for adaptive forestry which need specific measures for their preservation. The traditionally opposing views which may exist between conservation planning and sustainable forestry need to be reconciled and harmonized for managing peripheral populations. Based on existing knowledge, we suggest approaches and principles which may be used for the management and conservation of these distinctive and valuable populations, to maintain active genetic and ecological processes that have sustained them over time.

The fate of peripheral forest tree populations is of particular interest in the context of climate change. These populations may concurrently be those where the most significant evolutionary changes will occur ; those most facing increasing extinction risk ; the source of migrants for the colonization of new areas at leading edges ; or the source of genetic novelty for reinforcing standing genetic variation in various parts of the range. Deciding which strategy to implement for conserving and sustainably using the genetic resources of peripheral forest tree populations is a challenge. Here, we review the genetic and ecological processes acting on different types of peripheral populations and indicate why these processes may be of general interest for adapting forests and forest management to climate change. We particularly focus on peripheral populations at the rear edge of species distributions where environmental challenges are or will become most acute. We argue that peripheral forest tree populations are ‘‘natural laboratories” for resolving priority research questions such as how the complex interaction between demographic processes and natural selection shape local adaptation ; and whether genetic adaptation will be sufficient to allow the long-term persistence of species within their current distribution. Peripheral populations are key assets for adaptive forestry which need specific measures for their preservation. The traditionally opposing views which may exist between conservation planning and sustainable forestry need to be reconciled and harmonized for managing peripheral populations. Based on existing knowledge, we suggest approaches and principles which may be used for the management and conservation of these distinctive and valuable populations, to maintain active genetic and ecological processes that have sustained them over time.

# Low but significant evolutionary potential for growth, phenology and reproduction traits in European beech [https://doi.org/10.5061/dryad.x0k6djhrn](https://doi.org/10.5061/dryad.x0k6djhrn) Phenotypic and genotypic data used in the paper Low but significant evolutionary potential for growth, phenology and reproduction traits in European beech. ## Description of the data and file structure Data is provided in two separate files, phenotypic and genotypic data. Phenotypic data consists of 16 traits recorded for 251 adult Fagus sylvatica trees growing in a single location. The data file contains 17 columns: 1. ID: sample identification number (same as sample in the genotype data file for adult trees, can be connected to a single tree coordinate through the genotype data file) 2. DBH-2018: Diameter at 1.30 m (cm) measured in 2018 3. GL-2018: Length of the growing season (days) in 2018 4. GL-2019: Length of the growing season (days) in 2019 5. LBB-2018: Length of budburst (GDD5, i.e. temperature sum of Growing Degree Days with base temperature of 5ºC) between budburst stage 2 and stage 5 in 2018 6. LBB-2019: Length of budburst (GDD5, i.e. temperature sum of Growing Degree Days with base temperature of 5ºC) between budburst stage 2 and stage 5 in 2019 7. BB-JD100-2018: Bud phenological stage at Julian day 100 (1-5 scale) in 2018 8. BB-JD100-2019: Bud phenological stage at Julian day 100 (1-5 scale) in 2019 9. SPSS-2018: Spring phenology score sum, i.e. sum of daily phenology scores during the observation period in 2018 10. SPSS-2018: Spring phenology score sum, i.e. sum of daily phenology scores during the observation period in 2019 11. FLW-2018: Male flowering stage at Julian day 114 (1-3 scale) in 2018 12. FT-JD108-2018: Fruit abundance at Julian day 108 (1-4 scale) in 2018 13. FT-JD217-2018: Fruit abundance at Julian day 217 (1-4 scale) in 2018 14. LLS-2018: Length of leaf senescence (days) in 2018 15. LLS-2019: Length of leaf senescence (days) in 2019 16. LS-JD287-2018: Leaf senescence stage at Julian day 287 (1-4 scale) in 2018 17. LS-JD267-2019: Leaf senescence stage at Julian day 267 (1-4 scale) in 2019 For LLS, NA means that the individual did not reach the final stage of senescence at the time of the last observation. Genotypic data (16 nSSR loci) is provided for 251 adult Fagus sylvatica trees, 436 seeds collected from 22 adult trees and 400 saplings from two different regeneration events, all growing on the same plot. For each individual, coordinates (carthesian coordinates) and elevation are provided. Saplings/seedlings were collected at 4 locations in two different years, therefore, 50 saplings/seedlings share the same geographic coordinate. Genotypic data file contains 35 columns: 1. Sample: sample identification number (same as ID in the genotype data file for adult trees) 2. Phase: ontogenetic stage, either tree (sampled in 2016), sapling (sampled in 2016, multiyear cohort aged five to nine years), seedling (sampled in 2019, one-year-old seedlings) or seed (sampled in 2016) 3.-32. Allele lengths for 16 SSR loci named as in Lefevre et al. 2012, 2 columns per locus (bp) 3. X: latitude in carthesian coordinates (m) 4. Y: longitude in carthesian coordinates (m) 5. Base elevation: Elevation of a tree above the sea level (m) Missing SSR data in the genotype file is given as zero. NA means not applicable, i.e. not used in the study. References: Lefevre, S., Wagner, S., Petit, R. J. & De Lafontaine, G. (2012). Multiplexed microsatellite markers for genetic studies of beech. Mol. Ecol. Resour., 12, 484-491, 10.1111/j.1755-0998.2011.03094.x Local survival of forest tree populations under climate change depends on existing genetic variation and their adaptability to changing environments. Responses to selection were studied in European beech (Fagus sylvatica) under field conditions. A total of 1,087 adult trees, seeds, one-year-old seedlings, and established multiyear saplings were genotyped with 16 nuSSRs. Adult trees were assessed for phenotypic traits related to growth, phenology and reproduction. Parentage and paternity analyses were used to estimate effective female and male fecundity as a proxy of fitness and showed that few parents contributed to successful regeneration. Selection gradients were estimated from the relationship between traits and fecundity, while heritability and evolvability were estimated using mixed models and the breeder’s equation. Larger trees bearing more fruit and early male flowering had higher total fecundity, while trees with longer growth season had lower total fecundity (directional selection). Stabilising selection on spring phenology was found for female fecundity, highlighting the role of late frosts as a selection driver. Selection gradients for other traits varied between measurement years and the offspring cohort used to estimate parental fecundity. Compared to other studies in natural populations, we found low to moderate heritability and evolvability for most traits. Response to selection was higher for growth than for budburst, leaf senescence or reproduction traits, reflecting more consistent selection gradients across years and sex functions, and higher phenotypic variability in the population. Our study provides empirical evidence suggesting that populations of long-lived organisms such as forest trees can adapt locally, even at short-time scales.

# Low but significant evolutionary potential for growth, phenology and reproduction traits in European beech [https://doi.org/10.5061/dryad.x0k6djhrn](https://doi.org/10.5061/dryad.x0k6djhrn) Phenotypic and genotypic data used in the paper Low but significant evolutionary potential for growth, phenology and reproduction traits in European beech. ## Description of the data and file structure Data is provided in two separate files, phenotypic and genotypic data. Phenotypic data consists of 16 traits recorded for 251 adult Fagus sylvatica trees growing in a single location. The data file contains 17 columns: 1. ID: sample identification number (same as sample in the genotype data file for adult trees, can be connected to a single tree coordinate through the genotype data file) 2. DBH-2018: Diameter at 1.30 m (cm) measured in 2018 3. GL-2018: Length of the growing season (days) in 2018 4. GL-2019: Length of the growing season (days) in 2019 5. LBB-2018: Length of budburst (GDD5, i.e. temperature sum of Growing Degree Days with base temperature of 5ºC) between budburst stage 2 and stage 5 in 2018 6. LBB-2019: Length of budburst (GDD5, i.e. temperature sum of Growing Degree Days with base temperature of 5ºC) between budburst stage 2 and stage 5 in 2019 7. BB-JD100-2018: Bud phenological stage at Julian day 100 (1-5 scale) in 2018 8. BB-JD100-2019: Bud phenological stage at Julian day 100 (1-5 scale) in 2019 9. SPSS-2018: Spring phenology score sum, i.e. sum of daily phenology scores during the observation period in 2018 10. SPSS-2018: Spring phenology score sum, i.e. sum of daily phenology scores during the observation period in 2019 11. FLW-2018: Male flowering stage at Julian day 114 (1-3 scale) in 2018 12. FT-JD108-2018: Fruit abundance at Julian day 108 (1-4 scale) in 2018 13. FT-JD217-2018: Fruit abundance at Julian day 217 (1-4 scale) in 2018 14. LLS-2018: Length of leaf senescence (days) in 2018 15. LLS-2019: Length of leaf senescence (days) in 2019 16. LS-JD287-2018: Leaf senescence stage at Julian day 287 (1-4 scale) in 2018 17. LS-JD267-2019: Leaf senescence stage at Julian day 267 (1-4 scale) in 2019 For LLS, NA means that the individual did not reach the final stage of senescence at the time of the last observation. Genotypic data (16 nSSR loci) is provided for 251 adult Fagus sylvatica trees, 436 seeds collected from 22 adult trees and 400 saplings from two different regeneration events, all growing on the same plot. For each individual, coordinates (carthesian coordinates) and elevation are provided. Saplings/seedlings were collected at 4 locations in two different years, therefore, 50 saplings/seedlings share the same geographic coordinate. Genotypic data file contains 35 columns: 1. Sample: sample identification number (same as ID in the genotype data file for adult trees) 2. Phase: ontogenetic stage, either tree (sampled in 2016), sapling (sampled in 2016, multiyear cohort aged five to nine years), seedling (sampled in 2019, one-year-old seedlings) or seed (sampled in 2016) 3.-32. Allele lengths for 16 SSR loci named as in Lefevre et al. 2012, 2 columns per locus (bp) 3. X: latitude in carthesian coordinates (m) 4. Y: longitude in carthesian coordinates (m) 5. Base elevation: Elevation of a tree above the sea level (m) Missing SSR data in the genotype file is given as zero. NA means not applicable, i.e. not used in the study. References: Lefevre, S., Wagner, S., Petit, R. J. & De Lafontaine, G. (2012). Multiplexed microsatellite markers for genetic studies of beech. Mol. Ecol. Resour., 12, 484-491, 10.1111/j.1755-0998.2011.03094.x Local survival of forest tree populations under climate change depends on existing genetic variation and their adaptability to changing environments. Responses to selection were studied in European beech (Fagus sylvatica) under field conditions. A total of 1,087 adult trees, seeds, one-year-old seedlings, and established multiyear saplings were genotyped with 16 nuSSRs. Adult trees were assessed for phenotypic traits related to growth, phenology and reproduction. Parentage and paternity analyses were used to estimate effective female and male fecundity as a proxy of fitness and showed that few parents contributed to successful regeneration. Selection gradients were estimated from the relationship between traits and fecundity, while heritability and evolvability were estimated using mixed models and the breeder’s equation. Larger trees bearing more fruit and early male flowering had higher total fecundity, while trees with longer growth season had lower total fecundity (directional selection). Stabilising selection on spring phenology was found for female fecundity, highlighting the role of late frosts as a selection driver. Selection gradients for other traits varied between measurement years and the offspring cohort used to estimate parental fecundity. Compared to other studies in natural populations, we found low to moderate heritability and evolvability for most traits. Response to selection was higher for growth than for budburst, leaf senescence or reproduction traits, reflecting more consistent selection gradients across years and sex functions, and higher phenotypic variability in the population. Our study provides empirical evidence suggesting that populations of long-lived organisms such as forest trees can adapt locally, even at short-time scales.

AbstractLocal survival of forest tree populations under climate change depends on existing genetic variation and their adaptability to changing environments. Responses to selection were studied in European beech (Fagus sylvatica) under field conditions. A total of 1087 adult trees, seeds, 1‐year‐old seedlings and established multiyear saplings were genotyped with 16 nuSSRs. Adult trees were assessed for phenotypic traits related to growth, phenology and reproduction. Parentage and paternity analyses were used to estimate effective female and male fecundity as a proxy of fitness and showed that few parents contributed to successful regeneration. Selection gradients were estimated from the relationship between traits and fecundity, while heritability and evolvability were estimated using mixed models and the breeder's equation. Larger trees bearing more fruit and early male flowering had higher total fecundity, while trees with longer growth season had lower total fecundity (directional selection). Stabilizing selection on spring phenology was found for female fecundity, highlighting the role of late frosts as a selection driver. Selection gradients for other traits varied between measurement years and the offspring cohort used to estimate parental fecundity. Compared to other studies in natural populations, we found low to moderate heritability and evolvability for most traits. Response to selection was higher for growth than for budburst, leaf senescence or reproduction traits, reflecting more consistent selection gradients across years and sex functions, and higher phenotypic variability in the population. Our study provides empirical evidence suggesting that populations of long‐lived organisms such as forest trees can adapt locally, even at short‐time scales.

AbstractLocal survival of forest tree populations under climate change depends on existing genetic variation and their adaptability to changing environments. Responses to selection were studied in European beech (Fagus sylvatica) under field conditions. A total of 1087 adult trees, seeds, 1‐year‐old seedlings and established multiyear saplings were genotyped with 16 nuSSRs. Adult trees were assessed for phenotypic traits related to growth, phenology and reproduction. Parentage and paternity analyses were used to estimate effective female and male fecundity as a proxy of fitness and showed that few parents contributed to successful regeneration. Selection gradients were estimated from the relationship between traits and fecundity, while heritability and evolvability were estimated using mixed models and the breeder's equation. Larger trees bearing more fruit and early male flowering had higher total fecundity, while trees with longer growth season had lower total fecundity (directional selection). Stabilizing selection on spring phenology was found for female fecundity, highlighting the role of late frosts as a selection driver. Selection gradients for other traits varied between measurement years and the offspring cohort used to estimate parental fecundity. Compared to other studies in natural populations, we found low to moderate heritability and evolvability for most traits. Response to selection was higher for growth than for budburst, leaf senescence or reproduction traits, reflecting more consistent selection gradients across years and sex functions, and higher phenotypic variability in the population. Our study provides empirical evidence suggesting that populations of long‐lived organisms such as forest trees can adapt locally, even at short‐time scales.