Next Article in Journal
Aboveground Biomass and Endogenous Hormones in Sub-Tropical Forest Fragments
Previous Article in Journal
The Potential of Green Schoolyards for Healthy Child Development: A Conceptual Framework
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 14
Issue 4
10.3390/f14040659
forests-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Genetic Differentiation of Budburst Timing in Fagus crenata Populations along a Spatial Gradient in Late Frost Timing in the Hakkoda Mountains, Northern Japan

by
Saki Sugimoto
1,2,* and
Kiyoshi Ishida
3
1
The United Graduate School of Agricultural Science, Iwate University, Morioka 020-0066, Japan
2
Tohoku Regional Forest Office, Forestry Agency, Akita 010-8550, Japan
3
Faculty of Agriculture and Life Science, Hirosaki University, Hirosaki 036-8561, Japan
*
Author to whom correspondence should be addressed.
Forests 2023, 14(4), 659; https://doi.org/10.3390/f14040659
Submission received: 14 January 2023 / Revised: 14 March 2023 / Accepted: 15 March 2023 / Published: 23 March 2023
(This article belongs to the Section Forest Ecology and Management)
Browse Figures
Review Reports Versions Notes

Abstract

:
We studied the genetic differentiation in budburst timing among Fagus crenata populations along spatial gradients in late frost timing in the Hakkoda Mountains, northern Japan, by focusing on last fatal frost day and topography. For budburst timing, we analyzed interpopulation variations in habitats, genetic variations in a nursery, and the relationships between these variations and environmental conditions in the habitats. Analyses of interpopulation variation showed that the day and the temperature sum of budburst positively correlated with the last fatal frost day in the habitats. Analyses of genetic variation showed significant genetic variations among provenances and families for both traits. For all provenances, the heritability for these traits were 0.7–0.8. The genetic variations were significantly associated with variations in the last fatal frost day among the provenances, suggesting that natural selection due to late fatal frost causes genetic differentiation in the traits along the spatial gradient in late frost timing. These results demonstrate that late frost is a key factor driving genetic differentiation of leaf-out phenology within a regional tree population.

1. Introduction

The timing at which leaf out occurs is becoming earlier in many temperate tree species because of climate warming [1,2]. Changes in spring phenology are expected to degrade forest ecosystems by decreasing productivity and phenological mismatches in biological interactions [3,4]. Understanding the mechanisms underlying the adaptation of tree species to environmental seasonality is essential for predicting the effects of ongoing climate change on forest ecosystems. Previous studies on the adaptive significance of leaf-out phenology for temperate tree species showed that natural selection imposed by late frost and competition influence the optimal day of budburst for each population, that is, trees that budburst before the optimal day suffer from late frost damage, whereas trees that budburst after the optimal day have disadvantages in competing for light under the same condition [5,6]. For tree species that budburst in early spring, regional variation in late frost timing (the final day of late frost in spring) or late frost frequency often causes genetic differentiation along altitudinal or latitudinal gradient for budburst timing (day and temperature sum of budburst; we defined “temperature sum of budburst” as accumulated temperature required for budburst): such species exhibit delayed budburst in regions where late frost occurs later compared to regions where late frost occurs early [7,8]. Similarly, if late frost regime varies locally according to landscape components, similar genetic differentiation of leaf-out phenology would be observed on a local scale within a metapopulation [9]. For tree populations with extensive gene flow, such local genetic differentiation can be regarded as possessing high adaptability to environmental variation in which focal traits are involved. This is because gene flow between differentiated populations sustains high genetic diversity within populations, which facilitates adaptation to environmental changes [10,11]. Hence, studies on genetic differentiations involved in local adaptation help in the evaluation of the adaptability of tree populations to environmental change. However, few studies have focused on local genetic differentiation in leaf-out phenology within a regional population in temperate trees [9,12].
Late frost tends to occur in specific topographies such as basins, valleys, and plateaus [13,14]. For instance, late frost occurs more frequently in basins than in hillside slopes, because temperature inversion is likely to occur in basins due to radiation cooling [15]. Therefore, it is predicted that tree species with early budburst will exhibit genetic differentiation in budburst timing along spatial gradient with topographic variation. In addition, high genetic diversity within a metapopulation may increase heritability (a parent–offspring regression value that represent the similarity between parents and offspring) in budburst timing, which determines the rate of evolutionary response to directional selection for traits.
The deciduous tree species Siebold’s beech (Fagus crenata Blume), predominantly found in temperate forests in northern Japan, is susceptible to late frost damage due to its tendency to leaf out early in the spring [16]. Previous studies have reported that this species exhibits genetic differentiation along latitudinal gradients for the temperature sum of leaf out [17,18]. However, the relationship between local genetic variation and late frost timing has not been studied in this species. Our recent study on the leaf-out phenology of Fagus crenata populations in the Hakkoda Mountains, northern Japan, showed that populations that inhabit basins where late frost occurred in late spring exhibited a larger value and lower phenotypic plasticity for temperature sum of budburst compared to other populations on surrounding hillside slopes [19]. This study also revealed that the safety margin of budburst timing (the time lag between the last fatal frost and the day of budburst) is long enough to prevent late frost damage for any population, even though there is a large interpopulation variation in last frost timing. These results suggest that there is a local adaptation in the leaf-out phenology to late frost timing at each habitat. However, the relationship between the late frost regime and genetic variation in leaf-out phenology has not yet been studied. Information on this relationship in species will contribute to the elucidation of the mechanisms underlying local adaptation to late frost regime in temperate tree species. Furthermore, measuring the genetic differentiation among populations for the timing of leaf out would aid in assessing the adaptability of the species in ongoing climate change because the risk of late frost damage is increasing owing to climate warming in Asia and Europe [20,21].
This study aimed to investigate genetic differentiation in budburst timing of Fagus crenata populations along spatial gradient in late frost regime in the Hakkoda Mountains, northern Japan, by focusing on late frost timing. We posed the following four questions on budburst timing: (1) Is interpopulation variation in habitats related to spatial variation in last frost timing? (2) To what extent does genetic variation exist between populations and families? (3) To what extent is this heritable? (4) Are the patterns of genetic differentiation among populations explained by spatial variation in late frost timing? To address these questions, we analyzed interpopulation variation in budburst timing in habitats on the basis of data from observations [19] for leaf-out phenology at six sites with different altitudes and topography (four hillside slopes; two basins). Furthermore, we observed the budburst of saplings originating from five of the six study sites for three years in a nursery in order to estimate genetic variation and heritability of the traits. We then analyzed the relationship between genetic variation in the traits with provenance and between spatial variations in late frost timing and altitude in the habitats.

2. Methods

2.1. Study Species

Siebold’s beech (Fagus crenata) is a deciduous tree of old growth forests in the cool temperate region and is distributed from Kyushu to Hokkaido, Japan [22]. The species is monoecious, wind-pollinated, and budbursts from April to May in northern Japan [23,24].

2.2. Study Area

We investigated six study sites at different topographies and altitudes in natural forests in the Hakkoda Mountains, northern Japan (Figure 1, Table 1). The species was observed between altitudes of 200 and 1200 m. Of the six study sites, four (S1–S4) were located on hillside slopes, while the other two (B1 and B2) were located at a basin named “Tashirotai”. This basin covers 16 km2 of flat land, with an altitude ranging from 540 to 600 m. The study sites were located at least 1.1 km apart from each other. According to AMeDAS data of the weather station (Sukayu Station; altitude 890 m; 40°38.9′ N, 140°50.9′ E; Japan Meteorological Agency, https://www.data.jma.go.jp/obd/stats/etrn/ accessed on 30 May 2022; Figure 1), the mean annual temperature for 2011–2019 was 5.3 °C.

2.3. Field Survey

2.3.1. Air Temperature

We set a data logger (HOBO Pro v2, Onset Ltd., Burlington, VT, USA) on the trunk of a canopy tree for each study site 5 m above the ground. The air temperature was measured at 10 min intervals for 2010–2019. The logger was set in a sunshade box (Solar Radiation Shield, Onset, Ltd., Burlington, VT, USA). Further description of the measurement method is available in the work of Sugimoto and Ishida [19].

2.3.2. The Timing of Budburst in the Habitat

We observed budbursts of 30–59 canopy trees per population from 2015 to 2019 for the six populations (hillside slopes: S1, S6, S7, S9; basin: B1, B2). We arbitrarily selected study trees from canopy trees within 1 ha of each population. In 2015, we observed 30 trees in each population and we added 29 study trees to the sample in 2016 for populations that exhibited large variations in leaf-out phenology.
We observed budbursts twice a week from March to June. The entire canopy of the study trees was observed by binoculars. We evaluated the percentage of budburst for each study tree in reference to 11 stages: 0% (no budburst), 10%–90% (evaluation in 10% increments), and 100% (all buds exhibited budburst). We could not determine the percentage of budburst for the S1 population in 2015 and S7 population in 2018 because budburst had progressed when we started the observation.
For the analyses of interpopulation variation in the timing of budburst, we defined “the day of budburst” as the day when the percentage of budburst reached 10%. Thus, the temperature sum of budburst denotes the temperature sum of 10% budburst. We calculated the temperature sum of budburst by assuming that the threshold temperature was 5 °C and that the starting date was January 1, in accordance with previous studies [5].

2.4. Common Garden Experiment

In 2015, when the beech trees exhibited masting in most areas of the Hakkoda Mountains, we selected 10 masting trees from the canopy trees studied for budburst timing at each of five (hillside slopes: S1, S6, S7, and S9; basin: B1) of the six sites and used the trees as mother trees for the experiment. Thus, the provenances of the seeds corresponded to the five sites. Fifty seeds were collected under the canopy of each mother tree from October to November. The seeds were sown in nursery boxes and overwintered in a nursery in Hirosaki City, Aomori Prefecture (40°35.9′ N, 140°28.26′ E; altitude 52 m). From April to May of the following year, we transplanted the germinated seeds into the nursery after leaf expansion and cultivated them for 3 years to grow the seedlings until they were large enough that the spring phenology was less susceptible to snowpack. Approximately 98% of the seedlings died due to insect damage, disease, and summer drought over the three years. Afterward, the 44 saplings that survived were dug up in November 2018 and transplanted into the Chitose nursery of Hirosaki University in the same city (40°34.8′ N, 140°28.9′ E; altitude 83 m) in April 2019. The saplings were planted in two rows (1 m between rows with 40 cm spacing between saplings in the row) in the Chitose nursery. We did not set blocks within the planting plot because the size of each family (offspring, i.e., open-pollinated seeds from the mother tree) was small. Budbursts of saplings were observed for three years (2019–2021). The numbers of saplings and mother trees (i.e., families) per provenance were 3–13 and 1–8, respectively (Table 2). The height of the saplings ranged from 40 to 230 cm in 2020. Budbursts of saplings were observed twice per week from early April to early June. The day and temperature sum of budburst of each sapling were calculated in the same manner as was performed for the canopy trees. The temperature sum of budburst was calculated using the AMeDAS data (Hirosaki Station; altitude 30 m; 40°36.7′ N, 140°27.3′ E; Japan Meteorological Agency, https://www.data.jma.go.jp/obd/stats/etrn/ accessed on 30 May 2022). According to the AMeDAS data, the mean values of the annual temperature, annual precipitation, and the maximum snow depth for 2019–2021 were 11.4 °C, 1235.3 mm, and 75 cm, respectively.

2.5. Data Analysis

2.5.1. Relationship between the Timing of Budburst of Canopy Trees and Environmental Conditions in Habitat

We performed the model selection based on AIC using linear mixed models (LMMs) for the day and temperature sum of budburst to analyze the effects of late frost timing, topography, and altitude on the budburst timing for canopy trees at the six sites. For late frost timing, we used “the last fatal frost day”, which we defined as the final day when the minimum temperature fell below −3 °C in spring, because a temperature of −3 °C or less is fatal for leaves of F. crenata in spring [19]. For the model selection, we used full models, in which explanatory variables were “altitude”, “topography”, and “the last fatal frost day (mean of nine seasons measurements)”, and the random effect was “year of the observation”. We also performed the same analyses for four (S1, S6, S7, and S9 on the hillside slopes; hereafter, referred to as “the hillside slope populations”) of the six populations, using LMMs with the same variables. We used R statistical software ver.4.0.1 [26] for all the analyses. The packages lme4 [27] and lmerTest [28] were used for LMM, and the package MuMIn [29] was used for the model selection.

2.5.2. Genetic Variation in the Timing of Budburst of Saplings in the Nursery

We performed nested ANOVAs for the day and temperature sum of budburst of saplings in the nursery to evaluate genetic variation among and within populations for the five provenances (S1, S6, S7, S9, and B1). The following mixed model was used for each of the nested ANOVAs: Yijk = μ + yk + Pi + Fj(i) + εijk, where μ is the mean trait value, yk and Pi are fixed effects for the observation year and provenance, respectively; Fj(i) is the random effect for family nested within provenance; and εijk is the residual effect. For the mixed model, variance components were calculated for each explanatory variable using the restricted maximum likelihood method with the statistical package VCA [30]. Furthermore, we analyzed genetic variation among and within the hillside slope populations by the same method using four (S1, S6, S7, and S9 on the hillside slopes; hereafter, referred to as “the hillside slope provenances”) of the five provenances. To show the effects of explanatory variables in which the confounding effect between provenance and family (i.e., effect of S6 provenance with a single family) was removed for the nested ANOVAs, the genetic variation was analyzed with the same method used for provenances with multiple families (four provenances other than S6).
Furthermore, we calculated the expected values for the day and temperature sum of budburst of each of the five provenances and calculated the p value for the difference between B1 provenance (hereafter referred to as “the basin provenance”) and the other hillside slope provenance using the LMMs: the explanatory variable was “provenance”, and the random effects were “year of the observation” and “family”.

2.5.3. Heritability of the Timing of Budburst

We estimated the heritability (the narrow sense heritability), h2, of the beech populations for the day and temperature sum of budburst from the regression coefficients of mother tree–family regressions for these trait values. According to quantitative genetics, h2 is equal to twice the regression coefficient of the parent–offspring regression for single parents when the offspring is half-sibs [31]. Therefore, assuming that the family was half-sibs, we calculated the h2 values based on the regressions between measurements of the mother trees in habitats and measurements of the families (mean value per family) in the nursery for all populations, as well as for the four hillside slope populations. To estimate the h2 values, we estimated the regression coefficients using the LMMs for the day and temperature sum of budburst, in which the response variable was “family value (mean value per family) of each observation year”, the explanatory variable was “mother tree value (mean value for five years measurements)”, and the random effect was “year of the observation”. For these regressions, the trait values for each of the mother trees and the families were standardized using the mean values and standard deviations, as performed in the study of leaf phenology of Quercus petraea [9], because leaf-out phenology generally differs between mature trees and saplings owing to environmental differences surrounding winter buds and due to ontogenetic change [32,33].

2.5.4. Relationship between the Timing of Budburst of Saplings in the Nursery and Environmental Conditions in the Provenances

We analyzed the relationship between the budburst timing of saplings in the nursery and environmental conditions in the provenances for the five provenances, using LMMs for the day and temperature sum of budburst: the explanatory variables were “altitude” and “the last fatal frost day”, and the random effect was “year of the observation”. In these analyses, “topography” (basin or hillside slopes) was not included as an explanatory variable because we could use only one provenance for the basin. We performed the same analyses for the four hillside slope provenances using LMMs with the same variables.

3. Results

3.1. Relationships between Timing of Budburst of Canopy Trees and Environmental Conditions in the Habitat

Analysis of the day of budburst of canopy trees showed that the effects of last fatal frost day, topography, and altitude were all significant (Table S1). According to the selected LMM for all populations, trees that inhabited sites with later last frost exhibited later budburst (Figure 2a). In addition, trees that inhabited higher sites and basins exhibited later budburst compared to those that inhabited lower sites and hillside slopes, respectively. Similarly, for the temperature sum of budburst, effects of last frost day, topography, and altitude were all significant (Table S1). The selected LMM showed that trees that inhabited sites with a later last frost exhibited larger values, that is, a high temperature requirement for budburst (Figure 2b). Regarding the effect of topography on the temperature sum of budburst, the selected LMM showed that trees in the basin exhibited larger values than those in the hillside slopes. However, there was a negative correlation between the temperature sum of budburst and altitude, meaning that the value tended to decrease as the altitude increased. In addition, the LMMs selected for the four hillside slope populations showed similar results for both traits (Table S1).

3.2. Genetic Variations in Nursery for the Timing of Budburst

The nested ANOVAs for the day and temperature sum of budburst of saplings showed that the value of both traits differed significantly among provenances and among families nested within provenances (Table S2). In addition, the nested ANOVAs for the four hillside slope provenances and those for the four provenances with multiple families showed similar results (Tables S2 and S3).
Although the LMM for the day and temperature sum of budburst showed that the basin provenance (B1 saplings) exhibited the largest values for both traits among provenances, differences between the basin provenance and the two hillside slope provenances (S1 and S7) were small and not significant (Table 3). According to the expected values of the LMMs, the largest difference among provenances for the day and that for the temperature sum were 9 days and 66 degree days, respectively.

3.3. Heritability for the Timing of Budburst

The LMMs that used the standardized trait values for all populations showed that the day and temperature sum of budburst of the families in the nursery were significantly positively correlated with those of mother trees in the habitat (Figure 3, Table S4). The heritability estimates (twice the slope of parent–offspring regression) for the day and temperature sum of budburst were similar, 0.72 and 0.75, respectively. In addition, the LMMs that used the standardized trait values for the four hillside slope populations showed similar results: the estimates for the day and temperature sum of budburst were 0.75 and 0.68, respectively (Table S4).

3.4. Relationships between the Timing of Budburst in the Nursery and Environmental Conditions in the Provenances

Analyses of the day and temperature sum of budburst of saplings showed that the effect of the last fatal frost day measured at the provenance was significant, whereas that of altitude was not significant for either trait value (Table S5). According to the selected LMMs, saplings from provenance, where the last fatal frost day occurred later, exhibited later budburst and a larger temperature sum of budburst compared to others (Figure 4a,b). In addition, the analysis of the two trait values of saplings from the four hillside slope provenances showed similar results for the temperature sum of budburst but not for the day of budburst; the effect of the last fatal frost day was significant for the temperature sum of budburst, but not for the day of budburst (Figure 4a,b, Table S5).

4. Discussion

4.1. Interpopulation Variation in Habitats

Analysis of budburst timing of Fagus crenata canopy trees observed within the habitat of the Hakkoda Mountains showed that the day and temperature sum of budburst were not only associated with topography (hillside slope and basin) but were also correlated with the last fatal frost day. These trait values were positively correlated with the last fatal frost day, indicating that the species exhibited later budburst in habitats where the last fatal frost day was later. These results are consistent with those of our recent study [19], suggesting that late frost is an abiotic factor that imparts natural selection in the budburst timing. In addition, the analysis showed a positive correlation between trait values and the last fatal frost day for the four hillside slope sites, even at the same altitude. This suggests that the spatial gradient in the late frost regime is caused not only by topographic variation between hillside slopes and the basin but also by topographic variations such as valleys, depressions, and small flatlands within hillside slopes. In general, temperature inversion that causes late frosts tends to occur not only in basins but also in valleys and flatlands [13,15]. Furthermore, the day and temperature sum of budburst correlated with altitude. However, these trait values differed in the variety of the correlation, that is, the correlation for the day was positive, while that for the temperature was negative. The positive correlation between the day of budburst and altitude is consistent with the nature of temperate tree species, which require high temperatures in spring for budburst [9,12].

4.2. Genetic Variation and the Heritability

Nested ANOVAs, based on the common garden experiment, showed significant genetic variations among provenances and among families within provenances in the day and temperature sum of budburst of saplings. The provenance with the largest value (i.e., the latest budburst timing) among the five provenances was the basin provenance (B1) for both traits. These results indicate that budburst timing is genetically differentiated among populations and that there is substantial genetic diversity for traits within populations as well. In addition, nested ANOVAs showed significant variation among the four hillside slope provenances for both traits. This suggests that the genetic differentiation of these traits among populations is caused not only by environmental variations due to topographic variation between hillside slopes and the basin [19], but also by other environmental variations independent of the basin. Hence, these results are consistent with our findings on the interpopulation variation of canopy trees.
Analyses of the mother tree–family regression, based on the standardized values of the day and temperature sum of budburst, showed that the values of families in the nursery were positively correlated with the values of the mother trees in the habitats for both traits. The heritability was estimated to be 0.7–0.8 for these traits for all populations. Similar estimates were obtained for the hillside slope populations. According to previous studies, the heritability for the day of budburst is generally high; the estimates for broadleaved tree species and coniferous species were 0.25–1.07 and 0.42–0.87, respectively [9,34], whereas the heritability for other traits such as size and shape of trees was estimated to be 0.19–0.26 for several species (estimates from 67 studies; [35]). Given these previous studies, the heritability estimated in this study can be regarded as high, suggesting that the timing of budburst is highly heritable for the species. However, as we calculated the heritability from the mother tree–family regression based on the assumption that the family was half-sibs, the heritability may have been overestimated due to the violation of the assumption. In addition, it is likely that the heritability of the traits is smaller than our indicated by our estimates in studies with larger spatial gradient in environmental conditions because the leaf-out phenology of saplings is generally affected by microclimates, such as snowpack and shading, resulting in an increase in the environmental variances [32,36]. As heritability is proportional to the rate of evolutionary response to natural selection per generation of the focal population, the heritability estimated in this study implies that the populations have adaptability to changing environmental seasonality with natural selection, at least in terms of the timing of budburst.

4.3. Local Adaptation to Spatial Variation in Late Frost Timing

The LMMs for saplings from all provenances showed that the day and temperature sum of budburst in the nursery positively correlated with the last fatal frost day in the provenances. However, the correlations between the trait values and the altitudes of the provenances were not significant. These results suggest that genetic differentiation among provenances in budburst timing is caused by natural selection due to late frost damage as the last fatal frost day implies the final day, when the minimum temperature falls below the lethal temperature of the species in spring. Hence, it is likely that natural selection due to late frost causes genetic differentiation of the traits, as mentioned above, resulting in local adaptation to the spatial variation in the timing of the last fatal frost day for the species. This is consistent with the results for canopy trees in which the day and temperature sum of budburst are positively correlated with the last fatal frost day, corroborating the idea of Sugimoto and Ishida [19] that trees which budburst earlier than the last fatal frost day may have been excluded by the last fatal frost from the populations, resulting in adaptation to climate in basins where late fatal frost is delayed in spring; however, the topography on hillside slopes and local altitude may also influence spatial variation in the timing of the last fatal frost day. Similarly, the positive correlation between the day and temperature sum of the budburst of canopy trees and the last fatal frost day found in this study can be explained by natural selection due to late fatal frost. Accordingly, our findings on genetic differentiation are consistent with the conventional view that natural selection due to late frost and competition with other trees drives the evolution of the timing of leaf out [5,6]. To date, very few studies have reported local genetic variation of leaf-out phenology which is caused by natural selection due to late frost for tree species. For Quercus petraea populations, which inhabit the Pyrenees, successive genetic variation along the altitudinal gradient of the day of budburst has been observed, and the genetic variation may have been caused by altitudinal variation in late frost timing [9]. However, other environmental factors, such as shortening of the growing season with increasing altitude, may have been involved in the genetic differentiation of Q. petraea populations.
Genetic variation in the phenotypic plasticity of budburst timing in response to the chilling duration might be involved in local adaptation to spatial variation in late frost timing, as suggested by Sugimoto and Ishida [19]. According to a common garden experiment that used F. crenata populations, the populations exhibited genetic variation along latitude in the phenotypic plasticity of the temperature sum of budburst in response to the chilling duration [18]. However, no study has been conducted on this species to elucidate whether genetic variation in phenotypic plasticity of budburst timing is involved in the adaptation to spatial variations in environmental conditions such as climate. Planting experiments using Populus fremontii showed that phenotypic plasticity in the timing of bud formation and budburst was genetically differentiated, and that the genetic differentiation reflects the adaptation of the species to regional variations in climate [37]. Further studies on the adaptive significance of genetic differentiation in the phenotypic plasticity of phenological traits are necessary to improve our understanding of the mechanisms driving local adaptation in budburst timing.

5. Conclusions

Our analyses revealed genetic differentiation in budburst timing of Fagus crenata populations along spatial gradient in late frost timing in the Hakkoda Mountains. This suggests that natural selection due to late fatal frost causes genetic differentiation in traits, resulting in local adaptation to late frost timing at each habitat, the result being tree budburst at a later period in habitats where late fatal frost is delayed in spring. These findings demonstrate that late frost is a key factor driving the genetic differentiation of leaf-out phenology within a regional tree population. However, we could not identify whether genetic variation in the phenotypic plasticity of budburst timing is involved in local adaptation to late frost timing. Furthermore, the heritability of traits was estimated to be high, as shown in studies on other tree species. If the leaf-out phenology of this species is highly adaptable to changing environmental seasonality, as suggested by the estimated heritability, its adaptation to the ongoing climate change might be possible, at least in terms of leaf-out phenology. Further studies on natural selection due to late frost are necessary in order to assess the adaptability of the leaf-out phenology of temperate tree species in the ongoing process of climate change.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/f14040659/s1, Table S1: Results of selected LMMs for the timing of budburst of canopy trees in the habitat; Table S2: Nested ANOVAs for the timing of sapling budburst in the nursery; Table S3: Nested ANOVAs for the timing of sapling budburst using the four provenances with multiple families; Table S4: Results of the LMMs of the relationships between families and mother trees for standardized values of budburst timing; Table S5: Results of the selected LMMs for the timing of sapling budburst in the nursery.

Author Contributions

Conceptualization and methodology, S.S. and K.I.; data collection and curation, S.S. and K.I.; statistical analysis and testing of the model, S.S. and K.I.; writing—original draft preparation, S.S and K.I.; review and editing, S.S. and K.I.; project administration, K.I.; funding acquisition, K.I. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by JSPS KAKENHI (grant numbers: 18K05737 and 25450199).

Data Availability Statement

Not applicable.

Acknowledgments

We thank the members of the Forest Ecology Laboratory, faculty of agriculture and life science, Hirosaki University, for their help in fieldwork. We sincerely thank Shinzi Akada, Shigeta Mori, Akira Yamao, Kazuhiko Masaka, Kenichi Yoshimura, and Sachinobu Ishida for advice on this study. We conducted the field survey with permission from Ministry of Environment and the Forestry Agency.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Menzel, A.; Sparks, T.H.; Estrella, N.; Koch, E.; Aasa, A.; Ahas, R.; Alm-kübler, K.; Bissolli, P.; Braslavská, O.; Briede, A.; et al. European phenological response to climate change matches the warming pattern. Glob. Chang. Biol. 2006, 12, 1969–1976. [Google Scholar] [CrossRef]
  2. Peñuelas, J.; Filella, L. Responses to a Warming World. Science 2001, 294, 793–795. [Google Scholar] [CrossRef] [PubMed]
  3. Memmott, J.; Craze, P.G.; Waser, N.M.; Price, M.V. Global warming and the disruption of plant-pollinator interactions. Ecol. Lett. 2007, 10, 710–717. [Google Scholar] [CrossRef] [PubMed]
  4. Parmesan, C.; Hanley, M.E. Plants and climate change. Complexities and surprises. Ann. Bot. 2015, 116, 849–864. [Google Scholar] [CrossRef]
  5. Hänninen, H. Boreal and Temperature Trees in A Changing Climate; Springer Business Media: Dordrecht, The Netherlands, 2015. [Google Scholar]
  6. Fu, Y.H.; Zhang, X.; Piao, S.; Hao, F.; Geng, X.; Vitasse, Y.; Zohner, C.; Peñuelas, J.; Janssens, I.A. Daylength helps temperate deciduous trees to leaf-out at the optimal time. Glob. Chang. Biol. 2019, 25, 2410–2418. [Google Scholar] [CrossRef]
  7. Kreyling, J.; Thile, D.; Nagy, L.; Jentsch, A.; Huber, G.; Konnert, M.; Beierkuhnlein, C. Late frost sensitivity of juvenile Fagus sylvatica L. differs between southern Germany and Bulgaria and depends on preceding air temperature. Eur. J. For. Res. 2011, 131, 717–725. [Google Scholar] [CrossRef]
  8. Eysteinsson, T.; Karlman, L.; Fries, A.; Martinsson, O.; Skúlason, B. Variation in spring and autumn frost tolerance among provenances of Russian larches (Larix Mill.). Scand. J. For. Res. 2009, 24, 100–110. [Google Scholar] [CrossRef]
  9. Alberto, F.; Bouffier, L.; Louvet, J.M.; Lamy, J.B.; Delzon, S.; Kremer, A. Adaptive responses for seed and leaf phenology in natural populations of sessile oak along an altitudinal gradient. J. Evol. Biol. 2011, 24, 1442–1454. [Google Scholar] [CrossRef]
  10. Petit, R.J.; Hampe, A. Some evolutionary consequences of being a tree. Annu. Rev. Ecol. Evol. Syst. 2006, 37, 187–214. [Google Scholar] [CrossRef] [Green Version]
  11. Savolainen, O.; Pyhäjärvi, T.; Knürr, T. Gene flow and local adaptation in trees. Annu. Rev. Ecol. Evol. Syst. 2007, 38, 595–619. [Google Scholar] [CrossRef]
  12. Vitasse, Y.; Delzon, S.; Bresson, C.C.; Michalet, R.; Kremer, A. Altitudinal differentiation in growth and phenology among populations of temperate-zone tree species growing in a common garden. Can. J. For. Res. 2009, 39, 1259–1269. [Google Scholar] [CrossRef] [Green Version]
  13. Sakai, A.; Larcher, W. Frost Survival of Plants; Springer: Berlin/Heidelberg, Germany, 1987. [Google Scholar]
  14. Vitasse, Y.; Klein, G.; Kirchner, J.W.; Rebetez, M. Intensity, frequency and spatial configuration of winter temperature inversions in the closed La Brevine valley, Switzerland. Theor. Appl. Climatol. 2017, 130, 1073–1083. [Google Scholar] [CrossRef]
  15. Geiger, R. The Climate Near the Ground; Harvard University Press: Cambridge, MA, USA, 1950. [Google Scholar]
  16. Kojima, H.; Mariko, S.; Nakamura, T.; Hayashi, I. Bud burst process and late-frost experimens on Fagus crenata and Quercus mongolica ssp. crispula. Veg. Sci. 2003, 20, 55–64. [Google Scholar] [CrossRef]
  17. Hashizume, H.; Lee, J.H.; Yamamoto, F. Variation in the flushing time of Fagus crenata Blume among provenances and families. J. Jpn. For. Soc. 1996, 78, 363–368, (In Japanese with English summary). [Google Scholar] [CrossRef]
  18. Osada, N.; Murase, K.; Tsuji, K.; Sawada, H.; Nunokawa, K.; Tsukahara, M.; Hiura, T. Genetic differentiation in the timing of budburst in Fagus crenata in relation to temperature and photoperiod. Int. J. Biometeorol. 2018, 62, 1763–1776. [Google Scholar] [CrossRef]
  19. Sugimoto, S.; Ishida, K. Interpopulation variation in leaf out phenology of Fagus crenata along topographic variation associated with the late frost regime in the Hakkoda Mountains, northern Japan. Ecol. Res. 2022, 2022, 1–14. [Google Scholar] [CrossRef]
  20. Vitasse, Y.; Schneider, L.; Rixen, C.; Christen, D.; Rebetez, M. Increase in the risk of exposure of frost and fruit trees to spring frosts at higher elevations in Switzerland over the last four decades. Agric. For. Meteorol. 2018, 248, 60–69. [Google Scholar] [CrossRef]
  21. Zohner, C.M.; Mo, L.; Renner, S.S.; Crowther, T.W. Late-spring frost risk between 1959 and 2017 decreased in North America but increased in Europe and Asia. Proc. Natl. Acad. Sci. USA 2020, 117, 12192–12200. [Google Scholar] [CrossRef]
  22. Nakashizuka, T. Fagus crenata. In The Trees of Japan Ι; Editorial committee of the Trees of Japan, Ed.; J-FIC: Tokyo, Japan, 2009; pp. 577–590, (In Japanese. The English title is the author’s tentative translation). [Google Scholar]
  23. Maruyama, K. Comparative studies on the phenological sequences among different tree species and layer communities—Ecological studies on natural beech forest (33). Bull. Niigata Univ. For. 1979, 12, 19–41, (In Japanese with English summary). [Google Scholar]
  24. Tomita, M.; Seiwa, K. Influence of canopy tree phenology on understorey populations of Fagus crenata. J. Veg. Sci. 2004, 15, 379–388. [Google Scholar] [CrossRef]
  25. Takahashi, K. On the snow scale for measuring maximum snow depth. Seppyou 1968, 30, 11–14, (In Japanese with English summary). [Google Scholar] [CrossRef] [Green Version]
  26. R Development Core Team. R Version 4.0.1; A language and environment for statical computing; R Foundation for Statistical Computing: Vienna, Austria, 2021. [Google Scholar]
  27. Bates, D.; Mächler, M.; Bolker, B.; Walker, S. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 2015, 67, 1–48. [Google Scholar] [CrossRef]
  28. Kuznetsova, A.; Brockhoff, P.B.; Christensen, R.H.B. Lmer test package: Tests in linear mixed effects models. J. Stat. Softw. 2017, 82, 1–26. [Google Scholar] [CrossRef] [Green Version]
  29. Barton, K. MuMIn: Multi-Model Inference. R Package Version 1.43.17. 2020. Available online: https://CRAN.R-project.org/package=MuMIn (accessed on 30 May 2022).
  30. Schuetzenmeister, A. VCA: Variance Component Analysis. R Package Version 1.4.5. 2022. Available online: https://cran.r-project.org/web/packages/VCA/ (accessed on 30 May 2022).
  31. Falconer, D.S.; Mackay, T.F.C. Introduction to Quantitative Genetics, 4th ed.; LongmanGroup Ltd.: Essex, UK, 1996. [Google Scholar]
  32. Osada, N.; Hiura, T. Intraspecific differences in spring leaf phenology in relation to tree size in temperature deciduous trees. Tree Physiol. 2019, 39, 782–791. [Google Scholar] [CrossRef]
  33. Vitasse, Y. Ontogenic changes rather than difference in temperature cause understory trees to leaf out earlier. New Phytol. 2013, 198, 149–155. [Google Scholar] [CrossRef]
  34. Howe, G.T.; Aitken, S.N.; Neale, D.B.; Jermstad, K.D.; Wheeler, N.C.; Chen, T.H. From genotype to phenotype: Unraveling the complexities of cold adaptation in forest trees. Can. J. Bot. 2003, 81, 1247–1266. [Google Scholar] [CrossRef]
  35. Cornelius, J. Heritabilities and additive genetic coefficients of variation in forest trees. Can. J. For. Res. 1994, 24, 2. [Google Scholar] [CrossRef]
  36. Vitasse, Y.; Baumgarten, F.; Zohner, C.M.; Kaewthongrach, R.; Fu, Y.H.; Walde, M.G.; Moser, B. Impact of microclimatic conditions and resource availability on spring and autumn phenology of temperate tree seedlings. New Phytol. 2021, 232, 537–550. [Google Scholar] [CrossRef]
  37. Cooper, H.F.; Grady, K.C.; Cowan, J.A.; Best, R.J.; Allan, G.J.; Whitham, T.G. Genotypic variation in phenological plasticity: Reciprocal common gardens reveal adaptive responses to warmer springs but not fall frost. Glob. Chang. Biol. 2018, 25, 187–200. [Google Scholar] [CrossRef] [Green Version]
Forests 14 00659 g001 550
Figure 1. Location of study sites in the Hakkoda Mountains, northern Japan. Black and red abbreviations denote the study sites on the hillside slopes (S1, S6, S7, and S9) and in the basin (B1, B2), respectively. “Sukayu” indicates the location of Sukayu Station of AMeDas. The red dashed line shows the outline of flat land at the bottom of the basin.
Figure 1. Location of study sites in the Hakkoda Mountains, northern Japan. Black and red abbreviations denote the study sites on the hillside slopes (S1, S6, S7, and S9) and in the basin (B1, B2), respectively. “Sukayu” indicates the location of Sukayu Station of AMeDas. The red dashed line shows the outline of flat land at the bottom of the basin.
Forests 14 00659 g001
Forests 14 00659 g002 550
Figure 2. Relationships between last fatal frost day and timing of budburst of canopy trees: day of budburst (a) and temperature sum of budburst (b). Points for the trait values and those for the last fatal frost days indicate the mean values per tree for the 5 observation years (2015–2019) and the mean values per site for the 9 observation years (2011–2019). The black lines and black dashed lines denote the expected values of the LMMs for altitudes of 500 m and 900 m on the hillside slopes, respectively. The red lines denote the expected LMMs values of the LMMs for an altitude of 500 m in the basin.
Figure 2. Relationships between last fatal frost day and timing of budburst of canopy trees: day of budburst (a) and temperature sum of budburst (b). Points for the trait values and those for the last fatal frost days indicate the mean values per tree for the 5 observation years (2015–2019) and the mean values per site for the 9 observation years (2011–2019). The black lines and black dashed lines denote the expected values of the LMMs for altitudes of 500 m and 900 m on the hillside slopes, respectively. The red lines denote the expected LMMs values of the LMMs for an altitude of 500 m in the basin.
Forests 14 00659 g002
Forests 14 00659 g003 550
Figure 3. Relationships between families in the nursery and mother trees in the habitat for standardized values of the day and temperature sum of budburst. Figures show the relationships between families in the nursery and mother trees in the habitat for the day (a) and temperature sum (b) of budburst for all populations. For both traits, points denote the mean value per family for three years (2019–2021) and the mean value per mother tree for five years (2015–2019). The dotted lines denote the expected values of LMMs for the family value. The standardized values of families were calculated using the mean and SD of all families for each observation year. The standardized values for mother trees were calculated using the mean and SD of all mother trees for the mean values per mother tree over the five years. The expected values (standard error) of the mother–offspring regression for the day and temperature sum of budburst for all populations were 0.36 (0.11) and 0.38 (0.11), respectively, and those for the four hillside slope populations were 0.37 (0.13) and 0.34 (0.13), respectively.
Figure 3. Relationships between families in the nursery and mother trees in the habitat for standardized values of the day and temperature sum of budburst. Figures show the relationships between families in the nursery and mother trees in the habitat for the day (a) and temperature sum (b) of budburst for all populations. For both traits, points denote the mean value per family for three years (2019–2021) and the mean value per mother tree for five years (2015–2019). The dotted lines denote the expected values of LMMs for the family value. The standardized values of families were calculated using the mean and SD of all families for each observation year. The standardized values for mother trees were calculated using the mean and SD of all mother trees for the mean values per mother tree over the five years. The expected values (standard error) of the mother–offspring regression for the day and temperature sum of budburst for all populations were 0.36 (0.11) and 0.38 (0.11), respectively, and those for the four hillside slope populations were 0.37 (0.13) and 0.34 (0.13), respectively.
Forests 14 00659 g003
Forests 14 00659 g004 550
Figure 4. Relationships between the last fatal frost day in the provenances and the timing of budburst of saplings: the day (a) and temperature sum (b) of budburst. Points for the trait value and the last fatal frost day denote mean values per sapling for the 3 observation years (2019–2021) and mean values per provenance for the 9 observation years (2011–2019), respectively. The solid and dotted lines denote the expected values of the LMMs for all provenances and for the four hillside slope provenances, respectively. The slopes of the solid lines for the day and temperature sum of budburst were 0.43 and 3.37, respectively, and those of the dotted lines were 0 and 2.75, respectively. Black and red denote hillside slope and basin provenances, respectively.
Figure 4. Relationships between the last fatal frost day in the provenances and the timing of budburst of saplings: the day (a) and temperature sum (b) of budburst. Points for the trait value and the last fatal frost day denote mean values per sapling for the 3 observation years (2019–2021) and mean values per provenance for the 9 observation years (2011–2019), respectively. The solid and dotted lines denote the expected values of the LMMs for all provenances and for the four hillside slope provenances, respectively. The slopes of the solid lines for the day and temperature sum of budburst were 0.43 and 3.37, respectively, and those of the dotted lines were 0 and 2.75, respectively. Black and red denote hillside slope and basin provenances, respectively.
Forests 14 00659 g004
Table
Table 1. Summary of the study sites.
Table 1. Summary of the study sites.
SitesTopographyAltitude (m)Snow Depth a (cm)Dominant Species
S1Hillside slope900379Fagus crenata and Abies mariesii
S6Hillside slope620262Fagus crenata and Magnolia obovate
S7Hillside slope600287Fagus crenata
S9Hillside slope450179Fagus crenata and Quercus crispula
B1Basin570247Fagus crenata and Quercus crispula
B2Basin560248Fagus crenata and Quercus crispula
a Mean value of the maximum snow depth per season from 2011–2019, measured using snow depth gauges, according to Takahashi [25].
Table
Table 2. Summary of the saplings used for the common garden experiment.
Table 2. Summary of the saplings used for the common garden experiment.
ProvenanceNumber of
Saplings		Number of
Mother Trees		Height (cm)
Mean (Range)		Diameter at the Stem Base (mm)
Mean (Range)
Hillside slope
S111676.4 (39.8−116.5)22.3 (18.4−32.2)
S631140.5 (72.3−229.7)24.6 (17.9−30.8)
S7138115.6 (75.8−164.2)29.2 (18.9−41.5)
S963151.7 (92.7−225.0)24.3 (19.6−31.0)
Basin
B1118110.1 (59.3−147.3)24.6 (18.0−31.2)
The height and diameter at the stem base of the saplings were measured in 2020. The number of mother trees is equal to the number of families per provenance.
Table
Table 3. Expected values of the LMMs for the timing of budburst of saplings in the nursery.
Table 3. Expected values of the LMMs for the timing of budburst of saplings in the nursery.
Number of
Mother Trees		Number of
Saplings per Mother Tree		The Day of BudburstTemperature Sum of Budburst
Provenance (DOY)(Degree Days)
(Topography) a EstimateStd. Errorp ValueEstimateStd. Errorp Value
B1 (b)81–2127.62.1–200.317.0–
S1 (s)61–5125.41.80.24183.316.30.31
S6 (s)13118.83.3<0.05134.029.4<0.05
S7 (s)81–4125.21.70.19183.415.10.28
S9 (s)31–3119.42.2<0.01137.419.9<0.01
a “b” and “s” in the parentheses denote the basin and the hillside slopes, respectively. The p value indicates the significance of the difference from B1 provenance.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Sugimoto, S.; Ishida, K. Genetic Differentiation of Budburst Timing in Fagus crenata Populations along a Spatial Gradient in Late Frost Timing in the Hakkoda Mountains, Northern Japan. Forests 2023, 14, 659. https://doi.org/10.3390/f14040659

AMA Style

Sugimoto S, Ishida K. Genetic Differentiation of Budburst Timing in Fagus crenata Populations along a Spatial Gradient in Late Frost Timing in the Hakkoda Mountains, Northern Japan. Forests. 2023; 14(4):659. https://doi.org/10.3390/f14040659

Chicago/Turabian Style

Sugimoto, Saki, and Kiyoshi Ishida. 2023. "Genetic Differentiation of Budburst Timing in Fagus crenata Populations along a Spatial Gradient in Late Frost Timing in the Hakkoda Mountains, Northern Japan" Forests 14, no. 4: 659. https://doi.org/10.3390/f14040659

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop