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

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