Climate sensitivity of Cryptomeria japonica in two contrasting environments: Perspectives from QTL mapping

RESEARCH ARTICLE Climate sensitivity of Cryptomeria japonica in two contrasting environments: Perspectives from QTL mapping Hideki MoriID*, Kana Yamashita, Shin-Taro Saiki, Asako Matsumoto, Tokuko Ujino-Ihara Forestry and Forest Products Research Institute, Tsukuba, Ibaraki, Japan * morih@ffpri.affrc.go.jp Abstract Long-lived forest tree species experience a wide range of environmental conditions through- out their lifespan. Evaluation of the underlying growth and development mechanisms of these species is essential to predict tree growth under climate change. This study investi- gated climate sensitivity to temperature, precipitation, dry periods, and the associated geno- mic regions in Cryptomeria japonica, Japan’s most commercially important tree. We used tree rings and common garden experiments with three clonal replicates planted in two con- trasting environments in Kyushu (Kumamoto site) and Honshu (Chiba site), Japan. Tree growth showed a significant negative correlation with the dry period (>4 days) in March of the year of tree-ring formation at the Chiba site. In contrast, temperature and precipitation had little influence on tree growth. Quantitative trait locus (QTL) analysis was performed to investigate climate sensitivity to dry periods at the Chiba site, revealing 13 significant QTLs. One QTL showed a substantially large contribution to the overall climate sensitivity, accounting for 12.4% of the total phenotypic variation. The phenotypic variance explained (PVE) by other QTLs ranged from 0.9% to 2.9%, and the total PVE by all QTLs was 35.6%.

These findings indicate that the tree population at the Chiba site could be vulnerable to drought in early spring and that the QTL showing the greatest impact on climate sensitivity may be closely related to genes associated with tolerance or adaptation to drought stress. The QTLs identified in this study could be useful for molecular breeding, forest manage- ment, and predicting the growth of C. japonica under a changing climate. 1. Introduction Longed-lived forest tree species, with their limited dispersal ability, experience various envi- ronmental changes throughout their lifespan. As extreme climatic events such as droughts and heatwaves are expected to increase as a result of ongoing climate change [1], evaluating the underlying genetic mechanisms that control growth and development of these species under changing climatic conditions is essential to understand the local adaptation abilities of tree species to predict tree growth, which is necessary for the effective forest management and breeding programs [2,3]. PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 1 / 14 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Mori H, Yamashita K, Saiki S-T, Matsumoto A, Ujino-Ihara T (2020) Climate sensitivity of Cryptomeria japonica in two contrasting environments: Perspectives from QTL mapping. PLoS ONE 15(1): e0228278. https://doi. org/10.1371/journal.pone.0228278 Editor: Filippos A. Aravanopoulos, Aristotle University of Thessaloniki, GREECE Received: October 3, 2019 Accepted: January 12, 2020

Published: January 28, 2020 Copyright: © 2020 Mori et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: This study was supported by the Research Grant #201421 of the Forestry and Forest Products Research Institute to AM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.

The growth responses of tree species to climate changes have been evaluated by multiple studies [4–6]. These studies often targeted tree species adapted to dry or extremely cold cli- mates (e.g., Pinus, Larix), such as temperate and boreal forests, and successfully identified cli- matic variables affecting tree growth. Although the number of studies conducted outside those climatic regions is rapidly increasing [7–10], climate sensitivity of tree species adapted to warm and humid climate remains relatively understudied, probably due to their weak growth response to interannual climatic variation. However, the impacts of climate change (e.g., droughts and heatwaves) are expected to rapidly increase globally [1], and the negative impacts of such events (e.g., seasonal drought; [2]) on tree species in humid climates are also likely to increase. Evaluating the mechanisms underlying these species’ growth response to varying cli- matic conditions changes is necessary to predict the impact of climate change in these forest ecosystems. Previously, empirical research of climatic factors was limited on tree growth and consisted of dendrochronological studies, often using mature or old-growth trees to investigate the impact of climate variation over a large time scale (>100 years; [11–13]). In contrast, short- term climatic variation (e.g., decades) may also have a significant impact on tree growth, par- ticularly during the initial growth phase (from seedlings to young trees; [14–17]). It is during

this initial stage of growth that trees exhibit the largest response to environmental conditions [18]. Additionally, the initial growth of trees has been studied as an important trait in tree breeding [19–21]. Evaluating climatic factors which affect initial tree growth is important not only for the fundamental knowledge on growth response of tree species but also for forest management and molecular breeding under changing climate. Phenotypic traits such as wood growth and properties are influenced by environmental and genetic factors [22]. Therefore, investigating the underlying genetic mechanisms of tree growth response to climate change is essential to gain a comprehensive understanding of cli- mate sensitivity of tree species [23]. In this context, genetic analyses such as quantitative trait locus (QTL) mapping and association mapping have been conducted on many forest species to identify important genome regions (QTLs) and genes that influence complex traits, primar- ily growth (stem growth and bud phenology), wood properties (wood density and chemistry) [24], and climate sensitivity (i.e., growth response to climate change; [17]). Identification of the important genome region or genes in terms of climate sensitivity of tree species would be useful for predicting the performance of genotypes under climate change (genomic selection; [25,26]). Cryptomeria japonica (sugi) is a wind-pollinated, allogamous, evergreen conifer species [27]. Sugi is the most widely planted and commercialized tree in Japan, with its straight trunk

and soft wood, and it has many diverse uses, such as in house construction, ship building, and pulp production [28]. Vegetative propagation of C. japonica is easily conducted by cutting shoots, thereby allowing the replication of clones in common garden experiments across mul- tiple environments. A large number of genetic markers have been developed for C. japonica [29,30], which allows the identification of QTLs strongly associated with genes. These charac- teristics make C. japonica suitable for the evaluation of QTLs associated with climate sensitivity. Several studies have evaluated genomic regions that affect the growth and wood properties of C. japonica [30–32]. Long-term growth patterns of C. japonica have been evaluated using tree rings [33,34] and other methods (e.g., dendrometer; [35,36]). However, few studies have assessed the genetic mechanisms of growth patterns linked with climatic variability in C. japonica. In this study, we aimed to evaluate the growth patterns during temporal variations in temperature, precipitation, and dry period using tree rings (climate sensitivity traits; Housset et al. 2018); we also attempted to identify the genomic regions (QTLs) influencing climate Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 2 / 14

sensitivity in C. japonica using common garden experiments in two contrasting environments in Japan. 2. Materials and methods 2.1. Plant material, common garden experiment, linkage map The plant material, common garden experiment, and linkage map previously described by [32] were used in this study. The C. japonica population used in this study comprised 139 genotypes, which were obtained from crossing between two F1 progenies: YI96 (female parent) and YI38 (male parent). YI96 and YI38 were obtained by crossing local varieties in the Kyushu region of Japan, ‘Yabukuguri’ × ‘Iwao’ and ‘Yabukuguri’ × ‘Kumotooshi’, respectively [37,38]. Common garden experiments were established for the 139 genotypes with three clonal repli- cates in two different environments in Japan: the Chiba site (35˚20052.5500N, 140˚1047.9200E) and the Kumamoto site (32˚41058.3600N, 130˚45017.6400E). Trees at both sites were planted in 2005 and were harvested for phenotypic measurement in 2015 (Kumamoto) and 2016 (Chiba). The Kumamoto site in the Kyushu region represents the original climatic condition of the mapping progeny, which was derived from crossing local varieties of this region. We used two linkage maps based on 858 genetic markers (YI96: heterozygous; YI38: homozygous) to con- struct the YI96 linkage map and 916 genetic markers (YI38: heterozygous; YI96: homozygous) to construct the YI38 linkage map. Genetic markers with large numbers of missing values (>1%) were excluded. The average distance between adjacent markers in the YI96 and YI38 maps was 1.47 and 1.56 cM, respectively.

2.2. Tree-ring data Wooden boards with a width of approximately 4 cm were placed 80 cm above the ground from the tree trunk of all trees in the sampled populations. All wooden boards included both pit and bark and were used to acquire scanned images using a conventional scanner. For each tree, two rectangular sections containing both pit and bark were selected from both sides of the scanned image. The selected sections were then used to measure tree-ring width. We aver- aged the width between two pairs of tree rings for each tree, which were then converted into basal area increment (BAI) chronologies. To control for the effects of age- and size-related growth, BAI chronologies were detrended using a modified Hugershoff model [39]. An auto- regressive model (ar1) was used to reduce the temporal autocorrelation in the BAI chronolo- gies, and the residuals were extracted (hereafter, relative BAI; S1 Fig). For each clone, Tukey’s biweight robust mean was calculated to obtain annual averages of the relative BAI. The above analyses were performed using R version 3.6.0 [40] using R packages “measuRing” [41] and “dplR” [42]. 2.3. Climate data Climate data (i.e., temperature and precipitation) of the two study sites during the study period were sourced from AMeDAS (Automated Meteorological Data Acquisition System of the Japan Meteorological Agency; https://www.jma.go.jp/jma/en/Activities/amedas/amedas.html). At the Chiba and Kumamoto sites, the mean annual temperature is 15.7˚C and 16.4˚C, respec-

tively, and the mean annual precipitation is 1614.6 mm and 2285.5 mm, respectively. Both sites are characterized by contrasting climatic conditions in the growing season, particularly precipitation patterns. For instance, Kumamoto receives on average 2.5-fold more rainfall than Chiba during the growing season (May to August; [32]). We examined the effect of con- tinuous dry (non-raining) days (hereafter, dry period) obtained from the precipitation data for Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 3 / 14

each month (S2 Fig). The magnitude of drought stress on sampled populations can change with the length of the dry period [i.e., short dry periods (<1 day) may have no effect on tree growth]; thus, dry periods were accumulated for each month with varying thresholds, ranging from 1 to 7 days. As setting a higher threshold of dry days produces a larger ratio of zero data (i.e., raining days) per month, the largest threshold was set as 7 days to make the ratio of zero data less than 50% (S3 Fig) to avoid inaccurate estimation of climate sensitivity (described below). Climate data described above were linearly detrended prior to the analysis. 2.4. Climate sensitivity traits Based on the methods of [17], climate sensitivity traits were obtained by calculating correlation coefficients between climate data (i.e., temperature, precipitation, and dry period) and relative BAI scores. Monthly temperature, precipitation, and dry period during the experimental years were used for the climate sensitivity traits. Correlation coefficients were calculated for the months starting from June of the year prior to tree-ring formation until September of the year of tree-ring formation. 2.5. QTL analysis For each study site, QTL analyses of climate sensitivity traits in the sampled population were conducted using the Bayesian Lasso linear regression model: y ¼ μI þ Xβ þ e where y is climate sensitivity traits, μ is the overall mean, I is the identity matrix, X is the geno- type of each marker, β is the effect of the genetic marker on climate sensitivity traits, and e is

the error. Heterozygous and homozygous marker genotypes were converted to 0 and 1, respec- tively. The Bayesian Lasso model computation was performed using the Empirical Bayesian Lasso algorithm (EBlasso; [43]) with three-level hierarchical prior distribution (normal, expo- nential, and gamma prior; [44]): bi � Nð0; siÞ; ði ¼ 1; 2; . . . ; kÞ s2 i � expðlÞ; l � gammaða; bÞ where k is the number of genetic markers and a and b are regularization parameters that control the level of shrinkage of the marker effects (β) toward zero. The regularization parameters were determined using five-fold cross validation. The EBlasso was conducted using R package “EBglmnet” [44]. Genetic markers with non-zero effects and P < 0.01 were considered as significant QTLs. The phenotypic variance explained (PVE) by i-th QTL was estimated as PVEi ¼ ^b2 i sxi =sy, where ^b is the estimated value of parameter β, σx is the vari- ance of the marker genotype, and σy is the variance of climate sensitivity traits. PVE by all QTLs was estimated as PVEall ¼ s^y =sy, where s^y is the variance of the estimated value of the climate sensitivity traits, which is obtained from a multiple linear regression using significant QTLs as explanatory variables [45]. The prediction accuracy of significant QTLs was tested by correlating the observed and predicted values using five-fold cross validation. The results of the QTL analysis were validated with composite interval mapping (CIM) using R package “qtl” [46]. Climate sensitivity of Cryptomeria japonica

PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 4 / 14

3. Results 3.1. Climate sensitivity traits Trees at the Chiba site exhibited a significant negative growth response to dry periods (with threshold of 4 days) in March of the year of tree-ring formation (Figs 1 and 2). This significant response to drought gradually decreased with larger thresholds (>5 days; Fig 1). On the other hand, trees at the Kumamoto site did not exhibit a strong growth response to dry periods (Figs 1 and 2) despite the fact that the number of dry days in both sites was not significantly different (S4 Fig). The maximum number of clones with significant correlations at the Kumamoto site was 23, which was recorded in October of the year of tree-ring formation, whereas the maximum number of clones with significant correlations at the Chiba site was 50, which was recorded in March of the year of tree-ring formation as mentioned above (Fig 1). The growth response to drought for the Chiba population at the site varied with the set threshold, whereas the corre- sponding response at the Kumamoto site appeared to depend on the genotypes (Figs 1 and 2). In contrast, growth response to monthly temperature and precipitation were relatively weak for the mapping populations at both sites (S5 Fig). The maximum number of clones with significant correlations was 23 and 25 for temperature and precipitation, respectively, which were both found in July of the year prior to tree-ring formation at the Chiba site. Similar to the climate sensitivity to drought, the tree population at the Kumamoto site showed varying

responses to temperature and precipitation, whereas the trees at the Chiba site showed positive or negative responses to temperature and precipitation (S5 Fig). Because a significant drought response was found at the Chiba site (Figs 1 and 2), we pro- ceeded with QTL analyses on the climate sensitivity traits (with threshold of 4 days) in March of the year of tree-ring formation at this site. Fig 1. Climate sensitivity of growth to dry period at the two sites. Climate sensitivity in terms of significant correlation between growth and dry period at Kumamoto (upper panels) and Chiba (lower panels) sites are shown with varying thresholds (from 1 to 7 days). Panel columns represent the thresholds. Positive and negative correlations are shown with open and closed bars, respectively. Month with lower case represents the month of the year prior to tree-ring formation (from June to December), whereas month with upper case represents the month of the year of tree-ring formation (from January to September). https://doi.org/10.1371/journal.pone.0228278.g001 Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 5 / 14

3.2. QTL analysis QTL analysis on sensitivity to drought in March of the year of tree-ring formation for the Chiba trees revealed 13 significant QTLs (Fig 3; Table 1). Of these, one QTL located in linkage group 3 of the YI38 map had a significant effect compared with the remaining 12 QTLs (Figs 3 and 4; Table 1). The PVE of the strongest QTL was 12.4%, while that of other QTLs ranged from 0.9% to 2.9%. The PVE by all QTLs combined was 35.9%. Climate sensitivity differed by 0.14 in average between the marker genotype of the strongest QTL (Fig 4). A significant posi- tive correlation (r = 0.57, P < 0.001) was found between the observed and predicted values of significant QTLs (S6 Fig). Identified QTLs were validated with CIM (S7 Fig); most were identi- fied at similar positions in the linkage maps, and the logarithm of odds (LOD) score of the QTL with the largest contribution was 11.9. 4. Discussion 4.1. Climate change sensitivity The growth response at the Chiba site indicates that those trees have a high sensitivity to drought stress during the early spring. This may be a result of the seasonal variation of gas exchange in C. japonica: [47] reported that whole plant evaporation increased significantly from March to April, possibly because this species requires a lot of water for gas exchange (i.e., photosynthesis) during early spring. High water demand during a dry period decreases the plant’s water potential; the water deficit further causes dehydration and cavitation in leaf and stem, resulting in cell turgor [48–50], which inhibits photosynthesis and metabolic activity

[51–53]. The response of C. japonica stem growth to water deficit was investigated previously using irrigation experiments [54,55]. The negative growth response during dry periods in early spring is also consistent with the seasonal growth patterns of C. japonica [35,36,56,57]; studies Fig 2. Heatmaps of growth response to dry period with a threshold of 4 days. The climate sensitivity traits of mapping populations at the Kumamoto (left panel) and Chiba (right panel) sites are shown as correlation coefficients between relative basal area increment and monthly cumulative dry period with a threshold of 4 days. Open circles indicate significant correlations (P < 0.05). Month notation is the same as in Fig 1. https://doi.org/10.1371/journal.pone.0228278.g002 Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 6 / 14

reported that the stem diameter growth of C. japonica occurred mostly between April and June. [36] investigated the radial growth patterns of 16 C. japonica populations across Japan, and reported that trees started to grow from April to May, peaking until around June, and decreased to almost zero after June. This tendency was also reported in other studies [35,56,57], indicating that our observations may reflect the seasonality of both growth patterns and gas exchange traits related to water demand of C. japonica, suggesting that stem diameter growth is probably influenced by water availability during early spring. Fig 3. Effect and location of significant QTLs for drought sensitivity at the Chiba site. Panel columns represent linkage groups. Panel rows represent two types of linkage maps based on the marker types (upper: YI96 map; lower: YI38 map). Short vertical lines at the bottom of each panel represent marker positions. https://doi.org/10.1371/journal.pone.0228278.g003 Table 1. List of significant QTLs identified for drought sensitivity at the Chiba site. Marker name Linkage group Linkage map Position (cM) Beta PVEa (%) m_1124 LG03 YI38 map 140.943 −0.171 12.4 m_1534 LG08 YI38 map 58.521 −0.083 2.9 m_1640 LG10 YI38 map 47.554 −0.080 2.7 m_0325 LG04 YI96 map 83.624 −0.076 2.5 m_1331 LG06 YI38 map 22.273 0.071 2.1 m_0149 LG02 YI96 map 105.401 −0.066 1.8 m_0708 LG10 YI96 map 0.000 −0.065 1.8 m_0788 LG10 YI96 map 104.723 0.061 1.6 m_1034 LG03 YI38 map 23.201 −0.060 1.5 m_1343 LG06 YI38 map 41.999 0.059 1.5

m_0031 LG01 YI96 map 33.557 0.054 1.2 m_0063 LG01 YI96 map 90.525 0.049 1.0 m_0817 LG11 YI96 map 18.205 0.046 0.9 a Phenotypic variance explained https://doi.org/10.1371/journal.pone.0228278.t001 Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 7 / 14

On the other hand, the mapping populations at the Kumamoto site exhibited a significantly lower growth response than populations at the Chiba site although there was no significant dif- ference in dry period between the two sites. Furthermore, the mapping population at the Chiba site tended to show either positive or negative growth in response to climatic factors, whereas the growth response of the mapping population in Kumamoto site was both positive and negative depending on the genotype. The differences observed at the two sites may be due to their transfer from their original habitat to a different environment, which is known to neg- atively impact tree growth (e.g., Abies sachalinensis, [58]; Pinus contorta, [59]). As mentioned earlier, the tree populations in this study were derived from a local variety from the Kyushu region, where the Kumamoto site is located. Indeed, decreased tree growth at the Chiba site compared with that at the Kumamoto site was also previously reported [32]; the study found that the mean stem diameter of trees at the Kumamoto site was larger than at the Chiba site Fig 4. Difference in climate sensitivity between the marker genotype of the strongest QTL identified for drought sensitivity at the Chiba site. https://doi.org/10.1371/journal.pone.0228278.g004 Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 8 / 14

despite the Kumamoto trees being 1 year younger. This indicates that Kumamoto is more suit- able for tree growth and is probably more robust to drought conditions than the Chiba site, which may explain the contrasting growth patterns observed in our study. Another possible explanation might be that there was a significant difference in precipitation between the two sites during the growing season: the amount of rainfall from May to August was 2.5-fold higher in Kumamoto than in Chiba. The Chiba population therefore experienced more drought stress during the growing season than the Kumamoto population. This might have caused a delayed effect of drought stress in early spring at the Chiba site [60,61], which partly explains the lack of a significant growth response to drought at the Kumamoto site. A relatively intense growth response to temperature and precipitation was found in July of the year prior to tree-ring formation at the Chiba site; however, the sensitivity to temperature and precipitation was relatively weak compared to sensitivity to drought conditions, as indi- cated by the number of significant correlations (i.e., temperature: 23; precipitation: 25; and dry period: 50). This indicates that climatic factors influencing the growth of C. japonica may be difficult to identify using monthly temperature and precipitation data alone. The lack of sensi- tivity to temperature can partially be explained by the wide distribution range of C. japonica, which is widely adapted to various environments. Natural forests of this species range from the

Aomori prefecture (40˚420N, 140˚120E) to Yakushima Island (30˚150N, 130˚300E) [27,62]. As the study sites were spread over a wide latitudinal range (within the species’ natural distribu- tion range), the temporal variation in temperature observed may not sufficiently influence tree growth. However, the observed lack of sensitivity to precipitation could be due to more extreme climatic events, such as a large amount of rainfall during a short period (e.g., typhoons). Such climatic events can significantly increase monthly precipitation but have a limited influence on water availability for tree growth. This indicates that water availability for the growth of C. japonica cannot be fully explained by monthly precipitation alone and high- lights the importance of using other indicators more likely to reflect water availability and its influence on tree growth [63]. This may be particularly important for tree species adapted to humid climates because such species’ climate sensitivity is often difficult to assess compared with those adapted to drier habitat (e.g., Pinus, [5]; Larix, [4]). 4.2. Climate sensitivity QTLs The QTL located in linkage group 3 in the YI38 map contributed significantly to the climate sensitivity of C. japonica (PVE = 12.4%). The PVE magnitude of the QTL is considered to be relatively high; in comparison, the generally reported effect of QTLs in other tree species is often less than 5% (reviewed in [24]). The large contribution of this QTL to climate sensitivity

indicates that it plays a significant role in tree growth under drought conditions. Additionally, this QTL could be closely related to genes associated with growth response to drought stress. Key genes in drought tolerance and adaptation of plant model species (e.g., Arabidopsis thali- ana, Populus) and other forest tree species (e.g., Pinus, Picea) have been reported in previous research (reviewed in [64]), and further research with similar approaches on C. japonica is essential to identify the same genes. Furthermore, identifying genes associated with drought stress is of importance for future forest management as climate change is expected to increase the frequency of extreme climatic events, such as drought and heatwaves [1]. Apart from the strongest QTL, there were a number of significant QTLs identified with a relatively small contribution to overall climate sensitivity (<3%). This implies that drought sensitivity in C. japonica is influenced by a number of genes, each with a different effect on the overall climate sensitivity phenotype. This observation was consistent with the findings of pre- vious QTL mapping studies on the growth traits of major tree species (e.g., Pinus taeda, [65]; Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 9 / 14

Eucalyptus globulus, [66]). These studies indicate that, similar to growth, climate sensitivity is also influenced by many QTLs with relatively small effect sizes. It is important to note that the contribution to climate sensitivity by all identified QTLs (35.9%) was almost three-fold larger than the strongest QTL alone (12.4%), and the intermediate prediction accuracy was found using identified QTLs (r = 0.57, P < 0.001). This indicates that the 13 QTLs identified in this study might be useful in predicting the drought sensitivity of C. japonica. Likewise, using geno- mic data to predict climate change sensitivity is important to inform future forest management practices, particularly for long-lived forest tree species. 5. Conclusions This study found that dry periods (longer than 4 days) strongly influenced the growth response of C. japonica mapping population at the Chiba site. Moreover, we found one QTL with a large contribution to overall climate sensitivity, possibly related to a key gene associated with growth response to drought stress. In contrast, we did not find a significant growth response to drought stress at the Kumamoto site, which was the original habitat of the mapping progenies. This indi- cates that climate sensitivity of the study population is probably influenced by surrounding environmental conditions, which highlights the need for further research on climate sensitivity of C. japonica under various conditions. This includes research on life history stages, climate

(i.e., experiments covering wider geographical locations), and other environmental conditions (e.g., soil conditions; [67]). The epigenetic effects on genotypes might also be important for the growth response observed in the present study [68, 69]. Additionally, because our study was conducted using local varieties from the Kyushu region, additional studies capturing wider genetic variation (e.g., populations from natural forests) with different approaches—such as genome-wide association study (GWAS)—are necessary to generalize the findings of this study. It is also important to note that a further study using the mapping progenies derived from the parents in different provenances (e.g. Honshu region), could provide significant insights into the growth response of Cryptomeria japonica found in the present study. Furthermore, it is pos- sible that the climate adaptation QTLs have been fixed in the Kyushu region, and were mono- morphic in the mapping progenies. This could be tested by crossing C. japonica trees in and outside the Kyushu region. Finally, it is important to note that, while heavy rainfall events (100 mm/day) are increasing, the number of rainfall days in Japan has decreased over the past 4 decades [70]. If this tendency continues in the future, the frequency of dry periods > 4 days (which had significant influence on tree growth at the Chiba site) may also increase. Although C. japonica is currently planted or naturally distributed in humid climates, the negative impact of climate change on the growth of C. japonica via short-term droughts may become more

severe, which requires more research on the climate sensitivity of this species in the near future. Supporting information S1 Fig. Clonal averages of raw (A, B) and relative (C, D) basal area increment (BAI) at the Kumamoto site (A, C) and Chiba site (B, D). (PDF) S2 Fig. Continuous dry (non-raining) days at (A) the Chiba site and (B) the Kumamoto site. The horizontal dashed line indicates the threshold of 4 days of dry period. (PDF) S3 Fig. Varying thresholds and the ratio of zero data (no dry period in a given month). The horizontal dashed line represents 50% of the ratio of zero data. (PDF) Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 10 / 14

S4 Fig. Frequency of dry days in March of the year of tree-ring formation at (A) the Chiba site and (B) the Kumamoto site. (PDF) S5 Fig. Growth response to temperature (right panels) and precipitation (left panels) at the Kumamoto site (upper panels) and the Chiba site (lower panels). Positive and negative correlations are shown with open and closed bars, respectively. Month notation is the same as in Fig 1. (PDF) S6 Fig. Prediction accuracy of the 13 identified QTLs. (PDF) S7 Fig. Results of the QTL analysis using composite interval mapping. The horizontal dashed lines represent the logarithm of odds (LOD) scores of two and three. (PDF) S1 Table. Tree ring width of the mapping populations at the two sites. (CSV) S2 Table. Marker genotypes and positions. (CSV) Acknowledgments We thank Drs. H. Iwata and S. Goto for their helpful advice during this study. Author Contributions Conceptualization: Hideki Mori. Data curation: Hideki Mori. Formal analysis: Hideki Mori. Funding acquisition: Asako Matsumoto. Investigation: Hideki Mori. Methodology: Hideki Mori, Kana Yamashita. Project administration: Hideki Mori. Resources: Kana Yamashita, Asako Matsumoto. Supervision: Tokuko Ujino-Ihara. Visualization: Hideki Mori. Writing – original draft: Hideki Mori, Shin-Taro Saiki. Writing – review & editing: Hideki Mori, Kana Yamashita, Shin-Taro Saiki, Asako Matsu- moto, Tokuko Ujino-Ihara. References 1. IPCC. Climate change 2013: The physical science basis. Contribution of working group i to the fifth

assessment report of the intergovernmental panel on climate change. Cambridge, United Kingdom; New York, NY, USA: Cambridge University Press; 2013. Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 11 / 14

2. Allen CD, Macalady AK, Chenchouni H, Bachelet D, McDowell N, Vennetier M, et al. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. For Ecol Manage. 2010; 259: 660–684. 3. Gray LK, Hamann A, John S, Rweyongeza D, Barnhardt L, Thomas BR. Climate change risk management in tree improvement programs: Selection and movement of genotypes. Tree Genet. Genomes. 2016;12. 4. Carrer M, Urbinati C. Long-term change in the sensitivity of tree-ring growth to climate forcing in Larix decidua. New Phytol. 2006; 170: 861–872. https://doi.org/10.1111/j.1469-8137.2006.01703.x PMID: 16684244 5. Leonelli G, Pelfini M, Battipaglia G, Cherubini P. Site-aspect influence on climate sensitivity over time of a high-altitude Pinus cembra tree-ring network. Clim Change. 2009; 96: 185–201. 6. Tei S, Sugimoto A, Yonenobu H, Ohta T, Maximov TC. Growth and physiological responses of larch trees to climate changes deduced from tree-ring widths and δ13C at two forest sites in eastern Siberia. Polar Sci. 2014; 8: 183–195. 7. Brookhouse M. Eucalypt dendrochronology: Past, present and potential. Aust. J. Bot. 2006; 54: 435. 8. Wils THG, Sass-Klaassen UGW, Eshetu Z, Bra¨ uning A, Gebrekirstos A, Couralet C, et al. Dendrochro- nology in the dry tropics: The Ethiopian case. Trees. 2010; 25: 345–354. 9. Pumijumnong N. Dendrochronology in Southeast Asia. Trees. 2012; 27: 343–358. 10. Gebrekirstos A, Bra¨uning A, Sass-Klassen U, Mbow C. Opportunities and applications of dendrochro- nology in Africa. Curr Opin Environ Sustain. 2014; 6: 48–53.

11. Frank D, Esper J. Characterization and climate response patterns of a high-elevation, multi-species tree-ring network in the European Alps. Dendrochronologia. 2005; 22: 107–121. 12. Carrer M. Individualistic and time-varying tree-ring growth to climate sensitivity. PLoS One. 2011; 6: e22813. https://doi.org/10.1371/journal.pone.0022813 PMID: 21829523 13. Rahman M, Islam M, Wernicke J, Bra¨uning A. Changes in sensitivity of tree-ring widths to climate in a tropical moist forest tree in bangladesh. Forests. 2018; 9. https://doi.org/10.3390/f9010026 14. Tange T. Ecophysiological study on the growth of Cryptomeria japonica planted trees (in Japanese with English summary). Bulletin of The University of Tokyo Forests. 1995; 93: 65–145. 15. Batllori E, Camarero JJ, Ninot JM, Gutie´ rrez E. Seedling recruitment, survival and facilitation in alpine Pinus uncinata tree line ecotones. Implications and potential responses to climate warming. Glob Ecol Biogeogr. 2009; 18: 460–472. 16. Zhu K, Woodall CW, Ghosh S, Gelfand AE, Clark JS. Dual impacts of climate change: Forest migration and turnover through life history. Glob Chang Biol. 2013; 20: 251–264. 17. Housset JM, Nadeau S, Isabel N, Depardieu C, Duchesne I, Lenz P, et al. Tree rings provide a new class of phenotypes for genetic associations that foster insights into adaptation of conifers to climate change. New Phytol. 2018; 218: 630–645. https://doi.org/10.1111/nph.14968 PMID: 29314017 18. Nakashizuka T, Iida S, Masaki T, Shibata M, Tanaka H. Evaluating increased fitness through dispersal:

a comparative study on tree populations in a temperate forest, Japan. Ecoscience. 1995; 2: 245–251. 19. Gaspar MJ, Borralho N, Gomes AL. Comparison between field performance of cuttings and seedlings of Eucalyptus globulus. Ann For Sci. 2005; 62: 837–841. 20. Sixto H, Salvia J, Barrio M, Ciria M, Cañellas I. Genetic variation and genotype-environment interactions in short rotation Populus plantations in southern Europe. New For. 2011; 42: 163–177. 21. Berlin S, Hallingba¨ ck HR, Beyer F, Nordh N-E, Weih M, Ro¨ nnberg-Wa¨stljung A-C. Genetics of pheno- typic plasticity and biomass traits in hybrid willows across contrasting environments and years. Ann Bot. 2017; 120: 87–100. https://doi.org/10.1093/aob/mcx029 PMID: 28449073 22. Villemereuil P de, Gaggiotti OE, Mouterde M, Till-Bottraud I. Common garden experiments in the geno- mic era: new perspectives and opportunities. Heredity. 2016; 116: 249–54. https://doi.org/10.1038/hdy. 2015.93 PMID: 26486610 23. Trujillo-Moya C, George J-P, Fluch S, Geburek T, Grabner M, Karanitsch-Ackerl S, et al. Drought sensi- tivity of Norway spruce at the species’ warmest fringe: quantitative and molecular analysis reveals high genetic variation among and within provenances. G3. 2018; 8: 1225. https://doi.org/10.1534/g3.117. 300524 PMID: 29440346 24. Neale DB, Kremer A. Forest tree genomics: Growing resources and applications. Nat Rev Genet. 2011; 12: 111. https://doi.org/10.1038/nrg2931 PMID: 21245829 25. Jannink J-L, Lorenz AJ, Iwata H. Genomic selection in plant breeding: from theory to practice. Brief

Funct Genomics. 2010; 9: 166–177. https://doi.org/10.1093/bfgp/elq001 PMID: 20156985 26. Crossa J, Pe´rez-Rodrı´guez P, Cuevas J, Montesinos-Lo´ pez O, Jarquı´n D, Campos G de los, et al. Genomic selection in plant breeding: methods, models, and perspectives. Trends Plant Sci. 2017; 22: 961–975. https://doi.org/10.1016/j.tplants.2017.08.011 PMID: 28965742 Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 12 / 14

27. Tsumura Y. Cryptomeria. In: Kole C, editor. Wild crop relatives: Genomic and breeding resources: For- est trees. Berlin, Heidelberg: Springer Berlin Heidelberg; 2011. pp. 49–63. 28. Ohba K. Clonal Forestry with Sugi (Cryptomeria japonica). In: Ahuja M-R, Libby WJ, editors. Clonal For- estry II: Conservation and Application. Berlin, Heidelberg: Springer Berlin Heidelberg; 1993. pp. 66– 90. 29. Ueno S, Moriguchi Y, Uchiyama K, Ujino-Ihara T, Futamura N, Sakurai T, et al. A second generation framework for the analysis of microsatellites in expressed sequence tags and the development of EST- SSR markers for a conifer, Cryptomeria japonica. BMC Genomics. 2012; 13: 136. https://doi.org/10. 1186/1471-2164-13-136 PMID: 22507374 30. Hiraoka Y, Fukatsu E, Mishima K, Hirao T, Teshima KM, Tamura M, et al. Potential of genome-wide studies in unrelated plus trees of a coniferous species, Cryptomeria japonica (Japanese cedar). Front Plant Sci. 2018; 9: 1322. https://doi.org/10.3389/fpls.2018.01322 PMID: 30254658 31. Uchiyama K, Iwata H, Moriguchi Y, Ujino-Ihara T, Ueno S, Taguchi Y, et al. Demonstration of genome- wide association studies for identifying markers for wood property and male strobili traits in Cryptomeria japonica. PLoS One. 2013; 8: e79866. https://doi.org/10.1371/journal.pone.0079866 PMID: 24260312 32. Mori H, Ueno S, Ujino-Ihara T, Fujiwara T, Yamashita K, Kanetani S, et al. Mapping quantitative trait loci for growth and wood property traits in Cryptomeria japonica across multiple environments. Tree Genet. Genomes. 2019; 15: 43.

33. Fujisawa Y, Ohta S, Tajima M. Wood characteristics and genetic variations in sugi (Cryptomeria japon- ica), 2: Variation in growth ring components among plus-trees clones and test stands. Mokuzai Gak- kaishi. 1993. 34. Fujisawa Y. Variation in moisture contents of heartwood among clones and test stands in Sugi (Crypto- meria japonica). Mokuzai Gakkaishi. 1995; 41: 249–255. 35. Yamashita K, Okada N, Kamo K. Application of the wire dendrometer for monitoring the radial growth of trees (in Japanese with English summary). Mokuzai Gakkaishi. 2006; 52: 8–18. 36. Nishizono T, Zushi K, Hiroshima T, Toyama K, Kitahara F, Terada F, et al. Latitudinal variation in radial growth phenology of Cryptomeria japonica D. Don trees in Japan. Forestry. 2018; 91: 206–216. 37. Tani N, Takahashi T, Iwata H, Mukai Y, Ujino-Ihara T, Matsumoto A, et al. A consensus linkage map for sugi (Cryptomeria japonica) from two pedigrees, based on microsatellites and expressed sequence tags. Genetics. 2003; 165: 1551–1568. PMID: 14668402 38. Moriguchi Y, Uchiyama K, Ueno S, Ujino-Ihara T, Matsumoto A, Iwai J, et al. A high-density linkage map with 2560 markers and its application for the localization of the male-sterile genes ms3 and ms4 in Cryptomeria japonica D. Don. Tree Genet. Genomes. 2016; 12: 57. 39. Fang K, Gou X, Peters K, Li J, Zhang F. Removing biological trends from tree-ring series: testing modi- fied Hugershoff curves. Tree Ring Res. 2010; 66: 51–59. 40. R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation

for Statistical Computing; 2019. 41. Lara W, Bravo F, Sierra CA. MeasuRing: An R package to measure tree-ring widths from scanned images. Dendrochronologia. 2015; 34: 43–50. 42. Bunn AG. A dendrochronology program library in R (dplR). Dendrochronologia. 2008; 26: 115–124. 43. Cai X, Huang A, Xu S. Fast empirical Bayesian LASSO for multiple quantitative trait locus mapping. BMC Bioinformatics. 2011; 12: 211. https://doi.org/10.1186/1471-2105-12-211 PMID: 21615941 44. Huang A, Liu D. EBglmnet: a comprehensive R package for sparse generalized linear regression mod- els. Bioinformatics. 2016; btw143. 45. Huang A, Xu S, Cai X. Whole-genome quantitative trait locus mapping reveals major role of epistasis on yield of rice. PLoS One. 2014; 9: e87330. https://doi.org/10.1371/journal.pone.0087330 PMID: 24489897 46. Broman KW, Wu H, Sen S, Churchill GA. R/qtl: QTL mapping in experimental crosses. Bioinformatics. 2003; 19: 889–890. https://doi.org/10.1093/bioinformatics/btg112 PMID: 12724300 47. Saito T, Kumagai T, Tateishi M, Kobayashi N, Otsuki K, Giambelluca TW. Differences in seasonality and temperature dependency of stand transpiration and canopy conductance between Japanese cypress (hinoki) and Japanese cedar (sugi) in a plantation. Hydrol Process. 2017; 31: 1952–1965. 48. Tyree MT, Sperry JS. Vulnerability of xylem to cavitation and embolism. Annu Rev Plant Biol. 1989; 40: 19–36. 49. Davis SD, Ewers FW, Sperry JS, Portwood KA, Crocker MC, Adams GC. Shoot dieback during pro-

longed drought in Ceanothus (Rhamnaceae) chaparral of California: a possible case of hydraulic failure. Am J Bot. 2002; 89: 820–828. https://doi.org/10.3732/ajb.89.5.820 PMID: 21665682 Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 13 / 14

50. Nardini A, Salleo S, Raimondo F. Changes in leaf hydraulic conductance correlate with leaf vein embo- lism in Cercis siliquastrum L. Trees. 2003; 17: 529–534. 51. Yoshimura K, Saiki S-T, Yazaki K, Ogasa MY, Shirai M, Nakano T, et al. The dynamics of carbon stored in xylem sapwood to drought-induced hydraulic stress in mature trees. Sci Rep. 2016; 6: 24513. https:// doi.org/10.1038/srep24513 PMID: 27079677 52. Saiki S-T, Ishida A, Yoshimura K, Yazaki K. Physiological mechanisms of drought-induced tree die-off in relation to carbon, hydraulic and respiratory stress in a drought-tolerant woody plant. Sci Rep. 2017; 7: 2997. https://doi.org/10.1038/s41598-017-02959-8 53. Kono Y, Ishida A, Saiki S-T, Yoshimura K, Dannoura M, Yazaki K, et al. Initial hydraulic failure followed by late-stage carbon starvation leads to drought-induced death in the tree Trema orientalis. Commun Biol. 2019; 2: 8. 54. Abe H, Nakai T. Effect of the water status within a tree on tracheid morphogenesis in Cryptomeria japonica D. Don. Trees. 1999; 14: 124–129. 55. Abe H, Nakai T, Utsumi Y, Kagawa A. Temporal water deficit and wood formation in Cryptomeria japon- ica. Tree Physiol. 2003; 23: 859–863. https://doi.org/10.1093/treephys/23.12.859 PMID: 12865252 56. Nanami S, Kawaguchi H, Yamakura T. Seasonal patterns of stem diameter growth in woody species with different life forms and geographical distributions at Mt. Mikasa, Nara, Japan. Applied Forest Sci- ence. 2010; 19: 27–34. 57. Jyske T, Fujiwara T, Kuroda K, Iki T, Zhang C, Jyske TK, et al. Seasonal and clonal variation in cellulose

microfibril orientation during cell wall formation of tracheids in Cryptomeria japonica. Tree Physiol. 2014; 34: 856–68. https://doi.org/10.1093/treephys/tpu008 PMID: 24633653 58. Ishizuka W, Goto S. Modeling intraspecific adaptation of Abies sachalinensis to local altitude and responses to global warming, based on a 36-year reciprocal transplant experiment. Evol Appl. 2011; 5: 229–244. https://doi.org/10.1111/j.1752-4571.2011.00216.x PMID: 25568044 59. Montwe´ D, Isaac-Renton M, Hamann A, Spiecker H. Cold adaptation recorded in tree rings highlights risks associated with climate change and assisted migration. Nat Commun. 2018; 9: 1574. https://doi. org/10.1038/s41467-018-04039-5 PMID: 29686289 60. Huang M, Wang X, Keenan TF, Piao S. Drought timing influences the legacy of tree growth recovery. Glob Chang Biol. 2018; 24: 3546–3559. https://doi.org/10.1111/gcb.14294 PMID: 29729065 61. Jiang P, Liu H, Piao S, Ciais P, Wu X, Yin Y, et al. Enhanced growth after extreme wetness compen- sates for post-drought carbon loss in dry forests. Nat Commun. 2019; 10: 195. https://doi.org/10.1038/ s41467-018-08229-z PMID: 30643121 62. Hayashi Y. Taxonomical and phytogeographical study of Japanese conifers (in Japanese). Tokyo: Norin-Shuppan; 1960. 63. DeSoto L, Varino F, Andrade J, Gouveia CM, Campelo F, Trigo RM, et al. Different growth sensitivity to climate of the conifer Juniperus thurifera on both sides of the Mediterranean Sea. Int J Biometeorol. 2014; 58: 2095–2109. https://doi.org/10.1007/s00484-014-0811-y PMID: 24659114

try of Agriculture, Forestry and Fisheries, Ministry of Land, Infrastructure, Transport and Tourism, Japan Meteorological Agency. Synthesis report on observations, projections and impact assessments of climate change “Climate Change in Japan and its Impacts”. 2018. Climate sensitivity of Cryptomeria japonica PLOS ONE | https://doi.org/10.1371/journal.pone.0228278 January 28, 2020 14 / 14