An assessment of the interactions between climatic conditions and genetic characteristic on the agricultural performance of soybeans grown in Northeast Asia

Glycine max, commonly known as soybean or soya bean, is a species of legume native to East Asia. The interactions between climatic conditions and genetic characteristic affect the agricultural performance of soybean. Therefore, an investigation to identify the main elements affecting the agricultural performances of 11 soybeans was conducted in Northeast Asia, China [Harbin (45°12′N) Yanji (42°53′N) Dalian (39°30′N) Qingdao (36°26′N)] Republic of Korea [Suwon (37°16′N) and Jeonju (35°49′N)]. The days to flowering (DTF) of soybeans with the e1-nf and e1-as alleles and the E1e2e3e4 genotype, except Keumgangkong, Tawonkong, and Duyoukong, was relatively short compared to soybeans with other alleles. Although DTF of the soybeans was highly correlated to all climatic conditions, days to maturity (DTM) and 100-seed weight (HSW) of the soybeans showed no significant correlation with any climatic conditions. The soybeans with a dominant Dt1 allele, except Tawonkong, had the longest stem length (STL). Moreover, the STL of the soybeans grown at the test fields showed a positive correlation with only day length (DL) although the results of our chamber test showed that STL of soybean was positively affected by average temperature (AVT) and DL. Soybean yield (YLD) showed positive correlations with latitude and DL (except L62-667, OT89-5, and OT89-6) although the response of YLD to the climatic conditions was cultivar-specific. Our results show that DTF and STL of soybeans grown in Northeast Asia are highly affected by DL although AVT and genetic characteristic also affect DTF and STL. Along with these results, we confirmed that the DTM, HSW, and YLD of the soybeans vary in relation to their genetic characteristic.


Introduction
Variability in climate conditions is a major fluctuating factor for estimating crop 46 development. Although the influence of climate on crops depends on geographic location and production conditions [1], interactions between genetic characteristic and climatic conditions,  115 To complement the results of the field tests, the DTF and stem length (STL) of 116 soybeans grown in walk-in chambers (Koito manufacturing Co., Ltd., Tokyo, Japan) were 117 investigated. We selected four soybeans based on their DTF tested in the field: Hannamkong, 118 Tawonkong, OT93-28, and OT89-5. The chambers were controlled with six combinations of 119 temperatures (18, 23, and 28 o C) and DLs (10 and 15 h) under a fixed relative humidity of 65% 120 from sowing to harvesting of the soybeans. as the weight of 100 seeds measured under 13% moisture content; 4). stem length (STL) 129 defined as the length from the cotyledonary node to the top of the main stem; and 5). yield 130 (YLD, kg/10 a) calculated as the total weight (kg) of seeds harvested in 1 m 2 × 1000. 131 Regional climatic conditions at the test fields during the overall growth periods of the 132 soybeans, including DL, average temperature (AVT), accumulated temperature (ACCT), and 133 precipitation (PRC), were investigated using weather data obtained from the Chinese 134 Meteorological Administration and the Korean Meteorological Administration. We examined 135 the data for each climatic element by month then classified the data largely into two soybean 136 growth periods: vegetative (early) and reproductive (late) periods. This classification is because the soybean requires different cultural environmental conditions for these developmental stages 138 [35]. The early period from May to July is commonly the vegetative stage of the soybeans, 139 while the late period from August to October is, on average, the reproductive stage of the 140 soybeans grown in the six test fields. 141 Analysis of genetic relationships 142 We conducted genotyping of the soybeans to estimate the correlation between genetic 143 characteristic and agricultural performances. Genotyping of the soybeans occurred using the 144 allele-specific DNA markers ( between agricultural performance of the soybean and/or climatic conditions were analyzed with 159 SAS 9.2 (SAS Institute Inc., Cary, NC, USA) and used to assess the degree of correlation.

160
Multiple comparisons of the soybeans within the experimental test fields were performed using 161 the least significant difference (LSD; p>0.05; total degrees of freedom (DF)=30 or 24).

164
Correlation between climatic conditions 165 DL of the test fields showed a highly positive correlation with latitudes of the fields; a 166 higher latitude results in a longer DL (Fig 2, and S1 Table). Differences in late DL between the 167 test fields were very small and differences in ACCT between the fields were also low ( Table   168 3). AVT showed a negative correlation with the latitudes. PRC and AVT also were negatively 169 correlated with DL (Fig 2, and S1 Table). In Particular, latitude and DL showed highly positive  Table). PRC in Suwon (37°16′N) was the highest among the regions and Harbin showed 172 the lowest PRC. AVT was the highest in Qingdao (36°26′N) and Dalian (39°30′N) had the 173 highest ACCT (Fig 2).    (Table 4).  Genetic relationships and genotypes 201 The genetic trees of the soybeans were based on SSR markers that are significantly 202 associated with the agronomic traits. These trees were differentially constructed by traits (Fig   203   4). The tree based on SSR markers related to DTF had 5 subclades at a genetic distance of 0.5 204 ( Fig 4A). The subclade I, the subclade that Hannamkong is a part of, was very different 205 genetically from the other subclades ( Fig 4A). Soybeans in subclades II and III showed the 206 shortest DTF (Fig 4A and Table 4). Genotypes performed using four major E genes (E1 to E4) 207 and stem growth habit locus (Dt1) were significantly different between each soybean (Table   208 5). The DTF of soybeans that had predominantly E3 and E4 alleles was more than 55 days;

209
Tawonkong only had the E4 allele and had a DTF of 55.3 days (Tables 4 and 5). The tree 210 associated with DTM also formed five subclades at a genetic distance of 0.5 ( Fig 4B).

211
Hannamkong, which had the longest DTF and the shortest DTM, was categorized in the 212 subclade III in the tree related to DTM (Table 4 and Fig 4B). The tree associated with STL 213 showed six subclades at a genetic distance of 0.5 (Fig 4C). In the tree associated with STL, the 214 subclade VI (including L62-667, OT89-5, and OT93-28) had the longest STL among the 215 subclades on average; however, the subclade V (including OT94-37 and OT89-6) showed the 216 shortest STL (Fig 4C and  Tawonkong, and Duyoukong, was located furthest, genetically, from the other subclades ( Fig  4D). The tree associated with YLD also showed three subclades and subclade I, including    Table   233 6). STL showed positive correlations with latitude and DL; however, a negative correlation 234 with late PRC was observed (Fig 6, and Table 6). DTM and HSW showed no significant correlation with any of the climatic conditions (Fig 7 and 8, and Table 6). YLD also showed 236 positive correlations with latitude and DL, except in L62-667, OT89-5, and OT89-6 (Fig 9 and 237 Table 6). The soybeans were largely separated into two groups by yield without reference to 39, and OT94-51 (Fig 9).    Fig 10B). Nonetheless, each soybean grown at 23 and 28°C under 10 h DL had a 278 similar STL (Fig 10).  an effect on DTF, the effect requires further elucidation. As shown in Table 6, although DTF 315 of all the soybeans were highly correlated with all climatic conditions, we could not assume 316 that PRC and ACCT were correlated with DTF as the data for late PRC and late ACCT were 317 measured after R1. The water needs of soybean gradually increase during flowering and 318 through the grain filling stages [8,11]. For the chamber tests, we confirmed that DTF was 319 negatively and positively correlated with AVT and DL, respectively; the effect of temperature 320 on DTF was not significant when it was over 23°C (Fig 10). However, the average of the early 321 AVT at the six fields was 22.6°C (Table 3). Based on these results, we concluded that although 322 the DTF of soybean grown in Northeast Asia is affected by its genetic characteristic, AVT, and 323 DL, the effect of DL is more significant than the others.

324
Between the subclades of the tree structure shown in Fig 4B, and the data for the DTM 325 (R1 to R8) shown in Table 4, no correlation was found. In addition, as can be seen in Fig 7 and 326 Table 6, the DTM was not significantly correlated with any climatic conditions. As shown in 327 Table 4, Hannamkong showed the longest DTF and the shortest DTM. Moreover, the average 328 of the DTM of soybeans including OT94-37, OT89-6, and OT94-39 was the longest, as is seen 329 in subclade I. In the tree associated with DTF, these soybeans were classified into subclades III and IV, which had the shortest average of DTF (Table 4, and Fig 4A and 4B). These results

331
showed that the internal and external influences on DTM of soybean require further elucidation.  Table). Therefore, these results showed that STL of soybean is more affected by its 346 genetic characteristic and DL than AVT, as with DTF.

Effect of climatic conditions on yield components 348
In the genetic tree associated with HSW, 3 subclades with a similar HSW were formed, 349 as also seen in the tree associated with YLD (Fig 4D and 4E). Many researchers have reported conditions. This indicates that YLD of soybean is mostly affected by its genetic characteristic.

356
The soybeans that had a relatively longer DTF (over 55 days) showed higher correlations with 357 DTF than others which had relatively short DTFs (under 47 days). In addition, our study 358 showed that YLD has a positive correlation with DTF (Fig 3). These results indicate that DTF