Field assessment of a major QTL associated with tolerance to cold-induced seed coat discoloration in soybean

Abstract

In Hokkaido, the northernmost region of Japan, soybean [Glycine max (L.) Merr.] crops are damaged by cold weather. Chilling temperatures negatively affect seed appearance by causing seed coat discoloration around the hilum region, which is called cold-induced discoloration (CD). An assay for CD tolerance using a phytotron was developed, and two quantitative trait loci (QTLs) associated with CD tolerance were identified. The major QTL was located in the proximal region of the I locus, and structural variation of this locus can serve as a useful DNA marker, called the Ic marker. To use this marker in breeding programs, the effects need to be assessed under field conditions because the Ic marker has been developed solely under phytotron conditions. The aim of this study was thus to assess the effect of the Ic marker under a cool field environment. We confirmed that the Ic allele was highly effective using 27 cultivars and breeding lines including a near-isogenic line grown in the field where severe cold-weather damage occurred. This allele had no negative influence on the agronomic traits in the near-isogenic line. Our results suggest that marker-assisted selection for the Ic allele is effective for improving CD tolerance in breeding programs.

Introduction

In Hokkaido, the northernmost region of Japan, soybean [Glycine max (L.) Merr.] crops are damaged by cold weather (Funatsuki and Ohnishi 2009). Chilling temperatures negatively affect seed appearance, for example, by causing seed coat discoloration around the hilum region, which is called cold-induced discoloration (CD) (Funatsuki and Ohnishi 2009, Morrison et al. 1998, Srinivasan and Arihara 1994, Takahashi and Abe 1994). In Japan, CD damage has mainly been reported in Hokkaido (Srinivasan and Arihara 1994, Takahashi and Abe 1999). Assays for CD tolerance using a phytotron were developed for use in soybean breeding programs (Oka et al. 1989, Srinivasan and Arihara 1994, Takahashi and Abe 1994, Yumoto and Sasaki 1990). In the breeding program in Hokkaido, the plants were transferred into a phytotron and exposed to 14-day chilling-temperature conditions from 7 days after flowering, in accordance with a previous report (Yumoto and Sasaki 1990). There were differences in CD tolerance among the varieties (Yumoto and Sasaki 1990), and a gray-pubescent cultivar, Toyoharuka, was the most tolerant (Tanaka et al. 2015).

Only 20 to 30 breeding lines per year can be tested for CD tolerance in the phytotron assay in Hokkaido because of high running costs and limited space. Thus, soybean breeders have selected chilling-tolerant lines in the field where severe cold-weather damage frequently occurs (Sanbuichi 1979). In the breeding program in Hokkaido, yield trials had been performed in Kamishihoro (43°23′ N, 143°38′ E), a cool site located in the Tokachi area situated in the eastern part of Hokkaido, until 2011 (Yamaguchi et al. 2015a). Some chilling-tolerant cultivars (e.g., Toyohomare, Hayahikari, Toyoharuka, and Toyomizuki) were developed through the field selections there (Tanaka et al. 2015, Yamaguchi et al. 2018, Yumoto et al. 1995, 2000). To evaluate the chilling tolerance under more severe conditions, yield trials have also been conducted in Abashiri (43°90′ N, 144°30′ E), a cool site located in the Okhotsk area situated on the northeastern coast of Hokkaido, since 2010. The frequency of cold-weather damage in Okhotsk is higher than that in Tokachi (Tanaka et al. 2003). The accumulated average temperatures over a recent 10-year period in July, the period of sensitivity to chilling temperature, in Abashiri and Kamishihoro were 560 and 579°C, respectively. In the years of cold weather, cold-induced cracking seeds also appear in addition to CD in the Okhotsk area (Senda et al. 2018, Yamaguchi et al. 2014b, 2015a).

The mechanism underlying the induction of CD has been clarified. Three genetic loci (I, R, and T) are known to control seed coat pigmentation in soybeans (Bernard and Weiss 1973, Palmer and Kilen 1987). In yellow soybean, seed coat pigmentation is inhibited by post-transcriptional gene silencing (PTGS) of chalcone synthase (CHS) genes (Kasai et al. 2009, Senda et al. 2004). CD is caused by the suppression of CHS PTGS by chilling temperature (Kasai et al. 2009). In yellow-hilum cultivars of yellow soybean, an inverted repeat of a CHS truncated sequence was suggested to be the I locus, and its double-stranded CHS RNA transcript is thought to induce CHS PTGS (Kasai et al. 2007, Kurauchi et al. 2011). Toyoharuka has a different allele at the I locus, which was later designated Ic (inhibitor of CD) (Kasai et al. 2009, Ohnishi et al. 2011, Senda et al. 2012, Yamaguchi et al. 2015b). In a previous study, Ohnishi et al. (2011) identified two quantitative trait loci (QTLs) associated with CD tolerance in Toyoharuka using the phytotron assay. The major QTL was located in the proximal region of the I locus on chromosome 8, and structural variation of the I locus can serve as a useful DNA marker, called the Ic marker (Ohnishi et al. 2011). The effect of the Ic marker was confirmed under phytotron conditions using breeding lines with various genetic backgrounds, but some breeding lines with the Ic allele showed lower CD tolerance than Toyoharuka (Ohnishi et al. 2011). It was unclear whether the Ic marker was sufficient to improve CD tolerance in field conditions. Tawny pubescence also darkens the entire seed coat, giving a dirty appearance to yellow-hilum seeds (Cober et al. 1998, Morrison et al. 1998). Previous studies reported that the Ic allele may control the intensity of seed coat discoloration in tawny-pubescent cultivars (Oyoo et al. 2011, Rodriguez et al. 2013).

To use the Ic marker for CD tolerance in breeding programs, the effects need to be assessed under field conditions where severe cold-weather damage frequently occurs because this marker has been developed solely under phytotron conditions. It is also necessary to confirm whether the Ic marker is sufficient to improve CD tolerance in field conditions. The aim of this study was thus to assess the effect of the Ic marker under a cool field environment. Severe cold-weather damage occurred in Abashiri in 2017 in particular. Thus, we assessed the effect of the Ic marker using 27 cultivars and breeding lines grown in Abashiri in 2017. Moreover, we confirmed the effect of the Ic marker using the near-isogenic line (NIL) and evaluated the agronomic traits of this line.

Materials and Methods

Plant materials

All 31 cultivars and breeding lines were developed at the Tokachi Agricultural Experiment Station (TAES), Memuro, Hokkaido, Japan (Tables 1, 2). These are determinate and have a gray pubescence with a yellow hilum. Tokei 1284, an NIL derived from the cultivar Toyomizuki (recurrent parent, I allele) × the breeding line Tokei 1091 (donor parent, Ic allele), was bred by the backcrossing method using the Ic marker (Table 1). The Ic allele of Tokei 1091 originated from Toyoharuka. Tokei 1153 and Tokei 1179, the NILs derived from the cultivar Toyomusume (recurrent parent, I allele) × Toyoharuka (donor parent, Ic allele), were bred by the backcrossing method using the Ic marker (Table 2).

Table 1 Average pigment index (API) of 27 soybean cultivars and breeding lines in the field test and phytotron assay in 2017

Cultivar or line	Generation	Cross combinationa		Ic/I allele	API (0–4)
		P1	P2		Field test	Phytotron assay
Trial A (2nd preliminary yield trial)
Yukihomare	–	Tokei 783	Tokei 780	I	1.04 deb	1.77 b
Toyomizuki	–	Yukihomare	Tokei 930	I	1.05 de	2.88 a
Tokei 1266	F7	Tokei 842	Toyomizuki	I	1.44 cd	NDc
Tokei 1283	F6	Toiku 253	Tokei 1125	I	1.55 c	1.97 bc
Tokei 1287	F7	Chukei 508	Toiku 253	I	2.27 b	2.94 a
Tokei 1290	F6	Tokei 1117	Tokei 1128	I	1.01 e	1.25 c
Tokei 1291	F6	Tokei 1117	Tokei 1128	I	1.00 e	1.49 c
Tokei 1292	F6	Tokei 1120	Chukei 566	I	3.50 a	2.85 a
Tokei 1295	F6	Chukei 552	Toyomizuki	I	0.08 f	1.45 c
Tokei 1296	F7	Chuiku 66	Toiku 254	I	3.16 a	3.50 a
Toiku 266	F7	Toiku 248	Toiku 253	Ic	0.01 f	1.05 cd
Tokei 1284	BC3F6	Toyomizuki(4)	Tokei 1091	Ic	0.00 f	0.10 de
Tokei 1286	F7	Toiku 248	Chukei 552	Ic	0.00 f	0.05 e
Tokei 1289	F6	Toyomizuki	Tokei 1125	Ic	0.04 f	0.01 e
Tokei 1298	F8	Mazowia	Toyoharuka	Ic	0.00 f	0.00 e
				F-value	260.87	34.82
				Significance	***	***
Trial B (1st preliminary yield trial)
2507-13	F5	Toyomizuki	Tokei 1147	I	1.32 b	ND
2404W-38	F6	Toiku 254	Tokei 1120	I	2.31 a	ND
2507-37	F5	Toyomizuki	Tokei 1147	Ic	0.02 c	ND
2507-49	F5	Toyomizuki	Tokei 1147	Ic	0.04 c	ND
2507-81	F5	Toyomizuki	Tokei 1147	Ic	0.00 c	ND
2507-85	F5	Toyomizuki	Tokei 1147	Ic	0.03 c	ND
2501-3	F5	Toiku 255	Tokei 1130	Ic	0.00 c	ND
2501-4	F5	Toiku 255	Tokei 1130	Ic	0.03 c	ND
2509-3	F5	Tokei 1139	Toiku 254	Ic	0.08 c	ND
2509-6	F5	Tokei 1139	Toiku 254	Ic	0.05 c	ND
2509-25	F5	Tokei 1139	Toiku 254	Ic	0.00 c	ND
2514-1	F5	Tokei 1115	Toiku 256	Ic	0.02 c	ND
				F-value	1085.49
				Significance	***

***  indicates significant at the 0.1% level.

^a  Underlined cultivars or lines have the Ic allele. Tokei 1284 was derived from Toyomizuki (recurrent parent) × Tokei 1091 (donor parent) by the backcrossing method using the Ic marker. All Chukei and Chuiku lines were bred at Central Agricultural Experiment Station in Naganuma, Hokkaido, Japan. The cultivar Mazowia was bred in Poland. The other cultivars and lines were bred at Tokachi Agricultural Experiment Station in Memuro, Hokkaido, Japan.

^b  Means followed by a common letter in each trial are not significantly different according to Tukey’s HSD test (P ≥ 0.05).

^c  ND, no data.

Table 2 Average pigment index (API) of the near isogenic lines in the Toyomusume background in the phytotron assay in 2014

Cultivar or line	Generation	Cross combinationa		Ic/I allele	API (0–4) in the phytotron assay
		P1	P2
Toyoharuka	–	Tokei 793	Tokou 6225	Ic	0.00 cb
Toyomusume	–	Tokei 463	Toyosuzu	I	2.75 a
Tokei 1153	BC4F7	Toyoharuka	Toyomusume(5)	Ic	0.95 b
Tokei 1179	BC2F6	Toyomusume(3)	Toyoharuka	Ic	1.37 b
				F-value	36.08
				Significance	***

***  indicates significant at the 0.1% level.

^a  Underlined cultivars or lines have the Ic allele. Tokei 1153 and Tokei 1179 were derived from Toyomusume (recurrent parent) × Toyoharuka (donor parent) by the backcrossing method using the Ic marker.

^b  Means followed by a common letter are not significantly different according to Tukey’s HSD test (P ≥ 0.05).

Molecular marker analysis

Total genomic DNA was extracted from young leaves of three plants by a modified version of the CTAB method (Suzuki et al. 2012). The Ic/I alleles were determined by PCR for Ic markers as described previously (Ohnishi et al. 2011). The Ic marker sequences were 5′-GAG TTT GAA AAA TGT ATT CTT TCT CTT CC-3′ and 5′-GTA TCG CAG ATT CCT CCT GC-3′ for the Ic-specific amplicon and 5′-GCA AAC CAA ATC AAG TAA GAG CG-3′ and 5′-CCC ATT CCT TGA TTG CCT TA-3′ for the I-specific amplicon (Ohnishi et al. 2011).

Evaluation of CD tolerance in the cool field environment

Two yield trials were performed in the field in Abashiri, Hokkaido, Japan (43°90′ N, 144°30′ E). Fertilizer was applied in accordance with Hokkaido fertilization standards (6 N–48 P₂O₅–6 K₂O–8 MgO kg ha⁻¹). In total, 15 and 12 soybeans were evaluated for trials A and B, respectively (Table 1). Seeds were sown on 22nd May, 2017. Each plot consisted of two (trial B) and four (trial A) rows with lengths of 2.5 m; these were spaced 66 cm apart with an 18-cm inter-hill distance with two plants per hill. This gave a plant population density of 16.6 plants m⁻². A randomized complete block design with two replicates was used for each test. Fifty seeds were prepared from the bulk seeds of each plot, and the degree of pigmentation of each seed was visually classified using a pigment index as follows: 0: no pigmentation; 1: pigmented only in the hilum; 2: pigmented in the hilum and slightly around it (pigment extending <1 mm outside of the hilum); 3: pigmented in the hilum and around it on both sides (pigment extending ≥1 mm outside of the hilum); and 4: pigmented in the hilum and around it severely on both sides (pigment extending ≥3 mm outside of the hilum) (Fig. 1). This is a modified version of the index reported in previous studies (Ohnishi et al. 2011, Takahashi and Abe 1994). Pigment indices of 50 seeds from each plot were averaged (average pigment index, API). The JMP 10 statistical package (SAS, Cary, NC, USA) was used for statistical analysis. Two-way analysis of variance (ANOVA) was used to test differences of API among cultivars in each trial. Tukey’s HSD test with a threshold for significance of P < 0.05 was used to confirm the differences among cultivars. “Cultivar” and “replication” were used as the two factors in this case. Three-way ANOVA was also used to test differences of API among Ic/I alleles. Here, “allele”, “trial”, and “replication within trial” were used as the three factors.

Fig. 1

Examples of the pigmentation index. The degree of pigmentation was visually classified in proportion to the pigmented area: 0: no pigmentation; 1: pigmented only in the hilum; 2: pigmented in the hilum and slightly around it (pigment extending <1 mm outside of the hilum); 3: pigmented in the hilum and around it on both sides (pigment extending ≥1 mm outside of the hilum); and 4: pigmented in the hilum and around it severely on both sides (pigment extending ≥3 mm outside of the hilum). White arrows indicate pigmented areas outside the hilum.

Evaluation of CD tolerance by the phytotron assay

The CD tolerances of 14 cultivars and breeding lines in 2017 (Table 1) and four cultivars and NILs in 2014 (Table 2) were evaluated by the phytotron assay, as described previously (Ohnishi et al. 2011). Seeds were sown on 22nd May, 2014, and 17th May, 2017 in Wagner pots (30-cm diameter; Fujiwara Scientific) filled with low-humic andosols. Two pots were prepared for each breeding line. Three plants per pot were grown in an experimental facility under a plastic roof without walls in Memuro, Hokkaido, Japan (42°89′ N, 140°07′ E). Seven days after flowering, plants were transferred into the phytotron and exposed to the following chilling-temperature conditions for 14 days: 18°C day (08:00–18:00) and 13°C night (18:00–08:00), with 55% shading to avoid excessive heating from sunlight. After this 14-day treatment, the remaining flower buds were removed and plants were grown for a further 7 days in the phytotron at higher temperatures (25°C day/20°C night). After this treatment, the pots were returned to the experimental facility and the plants were grown to maturity. The pigment indices of all harvested seeds from each plant were evaluated, and APIs were calculated according to the above (Fig. 1). One-way ANOVA was used to test differences of API among cultivars. Tukey’s HSD test with a threshold for significance of P < 0.05 was used to confirm the differences.

Evaluation of agronomic traits of near-isogenic line

The yield trials of Toyomizuki and Tokei 1284 were performed in the field in Abashiri in 2015 to 2017. The generations of Tokei 1284 were BC₃F₄, BC₃F₅, and BC₃F₆ in 2015, 2016, and 2017, respectively. Seeds were sown on 25th May, 2015, 24th May, 2016, and 22nd May, 2017. Each plot consisted of two (in 2015 and 2016) and four (in 2017) rows with lengths of 2.5 m, which were spaced 66 cm apart with an 18-cm inter-hill distance with two plants per hill. This gave a plant population density of 16.6 plants m⁻². A randomized complete block design with two replicates was used for the experiments. Maturity was defined as the time when >80% of the plants were defoliated and had turned yellow, with pods that rattled when shaken. Before harvesting, main stem length (distance from cotyledonary node to terminal node) was recorded from ten central consecutive plants from each plot. Seed yield was assessed by hand harvest per plot and adjusted to 15% moisture. Combined ANOVA was carried out using the mixed model procedure. “Years” and “replications within year” were considered as random effects, while “cultivars” was considered as a fixed effect.

Results

Occurrence of CD in the field test

The average temperatures in Abashiri in 2017 are shown in Fig. 2. The standard cultivar Yukihomare started flowering on 23rd July. The average temperatures from 2nd to 20th Aug were much lower than those of the 30-year average: the plants were exposed to an 18-day chilling temperature from 10 days after flowering (Fig. 2). This period would correspond approximately to chilling temperature conditions in the phytotron assay, in which the plants were exposed to a 14-day chilling temperature from 7 days after flowering (Yumoto and Sasaki 1990). The seeds of Tokei 1296 (Table 1), a CD-sensitive breeding line grown in the field test, were pigmented in the hilum and also severely around it (Fig. 3).

Fig. 2

Average temperatures after flowering in Abashiri. Flowering refers to the date of 23rd July, 2017, when Yukihomare started flowering (shown as a gray inverted triangle). The 30-year average refers to the average temperatures in 1981 to 2010. The average temperatures from 2nd to 20th Aug, 2017, were much lower than those of the 30-year average (shown as a double-headed bow).

Fig. 3

The seeds of Tokei 1296, a breeding line sensitive to cold-induced discoloration, in the field test in Abashiri in 2017. The average pigment index was 3.16.

Evaluation of CD tolerance in the breeding lines

In the field test of trial A (n = 15), the APIs of the five lines with the Ic allele were significantly lower than those of the nine cultivars and lines with the I allele, with the exception of breeding line Tokei 1295 (Table 1). Only the API of Tokei 1295 was similar to those of the five lines with the Ic allele (API = 0.08). In the field test of trial B (n = 12), the APIs of the ten lines with the Ic allele were significantly lower than those of the two lines with the I allele (Table 1). ANOVA revealed that there was a significant difference between the Ic/I alleles in the field test through the two trials (Table 3). The API of the Ic allele was significantly lower than that of the I allele (Fig. 4).

Table 3 Analysis of variance for the effects of the Ic allele on average pigment index in the field test in 2017

Source	Sum of squares	Degrees of freedom	Mean squares	F-value	P-value	Significance
Model	35.145	4	8.786	19.983	<0.0001	***
Ic/I allele	27.606	1	27.606	62.786	<0.0001	***
Trial	0.062	1	0.062	0.142	0.708	ns
Replication within trial	0.004	2	0.002	0.043	0.996	ns
Error	21.545	49	0.440
Lack of fit	0.082	3	0.027	0.059	0.981
Pure error	21.463	46	0.467
Total	56.689	53

***  and ns indicate significant at the 0.1% level and not significant, respectively.

Fig. 4

Comparison of average pigment index in the field test in 2017. Gray boxes represent least square means with standard errors. *** indicates significant at the 0.1% level.

In the phytotron assay (n = 14), the APIs of the four lines with the Ic allele, with the exception of the breeding line Toiku 266, were significantly lower than those of the nine cultivars and lines with the I allele (Table 1). Only the API of Toiku 266 was slightly higher those of the other four lines with the Ic allele (API = 1.05). The APIs in the phytotron assay were positively correlated with the APIs in the field test (Fig. 5).

Fig. 5

Correlation between average pigment index (API) in the field test and API in the phytotron assay (n = 14). *** indicates significant at the 0.1% level.

Evaluation of CD tolerance and agronomic traits in the near-isogenic line

The APIs of Tokei 1153 and Tokei 1179, the NILs of Toyomusume for the Ic allele, were significantly lower than that of Toyomusume and higher than that of Toyoharuka in the phytotron assay (Table 2). The API of Tokei 1284, the NIL of Toyomizuki for the Ic allele, was significantly lower than that of Toyomizuki in both the field test and the phytotron assay (Table 1). The seed appearance in the field test is shown in Fig. 6. It is clear that the number of seeds pigmented in the hilum in Tokei 1284 was lower than that in Toyomizuki (Fig. 6). These results indicate that the Ic marker would be effective in both the Toyomusume and the Toyomizuki backgrounds.

Fig. 6

The seeds of Toyomizuki and Tokei 1284, the near isogenic line derived from Toyomizuki (recurrent parent, I allele) × Tokei 1091 (donor parent, Ic allele), in the field test in Abashiri in 2017. The average pigment indices of Toyomizuki and Tokei 1284 were 1.05 and 0.00, respectively.

The agronomic traits of Tokei 1284 are shown in Table 4. There were no significant differences in maturity, main stem length, seed yield, and 100-seed weight among cultivars (Table 4). The 100-seed weight of Tokei 1284 was slightly greater than that of Toyomizuki (P = 0.091, Table 4).

Table 4 Comparison of the agronomic traits of Toyomizuki and Tokei 1284 (average in 2015 to 2017)

Cultivar or line	Ic/I allele	Maturity (days)	Main stem length (cm)	Seed yield (t ha−1)	100-seed weight (g)
Toyomizuki	I	133.3	62.3	2.82	32.7
Tokei 1284a	Ic	132.5	62.0	2.88	33.4
	F-value	3.049	0.861	0.179	4.349
	P-value	0.141	0.902	0.690	0.091

^a  The near-isogenic line derived from Toyomizuki (recurrent parent, I allele) × Tokei 1091 (donor parent, Ic allele) by the backcrossing method using the Ic marker.

Discussion

In a previous study, the effect of the Ic marker was confirmed under phytotron conditions using breeding lines with various genetic backgrounds, but not NIL (Ohnishi et al. 2011). In this study, we assessed the effect of the Ic marker under a cool field environment using 27 cultivars and breeding lines with various genetic backgrounds, including the NIL of Toyomizuki for the Ic allele (Table 1), and confirmed that the Ic allele was highly effective in the field (Table 3, Fig. 4). We also assessed the effect of the Ic marker using the NILs in both the Toyomusume and the Toyomizuki backgrounds in the phytotron assay (Tables 1, 2). The chilling tolerance at the flowering stage of Toyomizuki was greater than that of Toyomusume by a seed yield- and maturity-based method (Yamaguchi et al. 2018), suggesting that we can select CD-tolerant lines by the Ic marker without considering chilling tolerance.

The API of Tokei 1295 was significantly lower than those of the other cultivars and lines with the I allele in the field test (Table 1). Some Swiss and Canadian cultivars have a different flowering habit to avoid cold-weather damage (Gass et al. 1996, Kurosaki et al. 2018, Schori et al. 1993). Schori et al. (1993) reported that such cultivars had profuse lateral flowering, and that this was related to these cultivars’ capacity to compensate for a mutilation of the central racemes under cold-weather conditions. Oka et al. (1989) also reported that chilling temperatures immediately after flowering rarely caused CD. Exposure of young pods (5–10 days after flowering) to chilling temperatures results in CD (Oka et al. 1989, Takahashi and Abe 1994, Yumoto and Sasaki 1990). The parental line of Tokei 1295 is Chukei 552 (Table 1), and Chukei 552 has Tohoku 74, an elite line developed in Tohoku situated in the northeast of Japan, in its pedigree. It is thus possible that the flowering habit of Tohoku 74 differs from that of Hokkaido cultivars. In trial A, Tokei 1295 matured 9 days later than the standard cultivar Yukihomare, and the latest among the 15 cultivars and breeding lines (data not shown). It is possible that Tokei 1295 flowered later than the other breeding lines. In either case, the API of Tokei 1295 was 1.45 in the phytotron assay, which was similar to those of the other cultivars and lines with the I allele (Table 1). Thus, we speculate that the flowering habit and/or the flowering stage of Tokei 1295 might differ from those of the other breeding lines, which allowed it to avoid the cold-weather damage in the field.

In the phytotron assay, the APIs of the NILs in the Toyomusume background were significantly higher than that of Toyoharuka (Table 2). The second QTL associated with CD tolerance in Toyoharuka was identified on chromosome 14 (Ohnishi et al. 2011), and the NILs had the Toyomusume alleles at this locus (data not shown). Thus, we suspect that the lower CD tolerance in these NILs may be attributable to the lack of this QTL.

Among the breeding lines with the Ic allele, the API of Toiku 266 was slightly high in the phytotron assay (Table 1). Previous studies also reported that some maturity-related genes or QTLs were associated with CD tolerance (Benitez et al. 2004, Githiri et al. 2007, Takahashi and Abe 1994, 1999). We consider that Toiku 266 may have the susceptible alleles of these minor QTLs including the second QTL reported by Ohnishi et al. (2011). In the field test, the APIs of all 15 lines including Toiku 266 with the Ic allele were less than 0.10 (Table 1). The APIs in the phytotron assay were positively correlated with the APIs in the field test, but the APIs in the phytotron assay in each line tended to be higher than those in the field test (Fig. 5, Table 1). These results suggest that the effects of the minor QTLs may be low in the field conditions. Thus, we believe that we can select the CD-tolerant lines at the field level only by marker-assisted selection of the Ic allele. The second QTL on chromosome 14 was detected only in the Ic background in the phytotron assay (Ohnishi et al. 2011). Further studies of the second QTL are needed for pyramiding of the Ic allele and this QTL in case more severe cold-weather damage occurs.

The 100-seed weight of Tokei 1284 was slightly greater than that of Toyomizuki (Table 4). In a previous study, Ikeda et al. (2009) reported a QTL controlling seed development at low temperature on chromosome 8, located in the proximal region of the I locus. This QTL might affect the 100-seed weight of Tokei 1284 because the average temperature after flowering was low in 2017 (Fig. 2). Heavier seeds are preferred for boiled-bean processing in Japan (Kato et al. 2014, Yamaguchi et al. 2014a, 2015b). Therefore, we consider that the effect of the Ic allele on the 100-seed weight is not disadvantageous for breeding programs in Japan.

In summary, we confirmed the effect of a major QTL for CD tolerance in field conditions. The Ic allele had no negative influence on the agronomic traits in the NIL. Our results suggest that marker-assisted selection for the Ic allele is effective for improving CD tolerance in breeding programs.

Author Contribution Statement

NY designed the research, conducted the phytotron experiment, analyzed the data, and wrote the manuscript. SH conducted the field experiment. DH conducted the DNA marker experiment. All authors read and approved the manuscript.

Acknowledgments

We thank Edanz (www.edanzediting.co.jp) for editing the English text of a draft of this manuscript. This work was supported by grants from the Ministry of Agriculture, Forestry, and Fisheries of Japan (Genomics-based Technology for Agricultural Improvement, SFC1006, and Research Program on Development of Innovative Technology, 26098C).

Literature Cited

•  Benitez,  E.R.,  H.  Funatsuki,  Y.  Kaneko,  Y.  Matsuzawa,  S.W.  Bang and  R.  Takahashi (2004) Soybean maturity gene effects on seed coat pigmentation and cracking in response to low temperatures. Crop Sci. 44: 2038–2042.
•  Bernard,  R.L. and  M.G.  Weiss (1973) Qualitative genetics. In:  Caldwell,  B.E. (ed.) Soybeans: Improvement, Production, and Uses, 1st edn. American Society of Agronomy, Madison, WI, pp. 117–149.
•  Cober,  E.R.,  G.R.  Ablett,  R.I.  Buzzell,  B.M.  Luzzi,  V.  Poysa,  A.S.  Sahota and  H.D.  Voldeng (1998) Imperfect yellow hilum color in soybean is conditioned by II rr TT. Crop Sci. 38: 940–941.
•  Funatsuki,  H. and  S.  Ohnishi (2009) Recent advances in physiological and genetic studies on chilling tolerance in soybean. Jpn. Agric. Res. Q. 43: 95–101.
•  Gass,  T.,  A.  Schori,  A.  Fossati,  A.  Soldati and  P.  Stamp (1996) Cold tolerance of soybean (Glycine max (L.) Merr.) during the reproductive phase. Eur. J. Agron. 5: 71–88.
•  Githiri,  S.M.,  D.  Yang,  N.A.  Khan,  D.  Xu,  T.  Komatsuda and  R.  Takahashi (2007) QTL analysis of low temperature induced browning in soybean seed coats. J. Hered. 98: 360–366.
•  Ikeda,  T.,  S.  Ohnishi,  M.  Senda,  T.  Miyoshi,  M.  Ishimoto,  K.  Kitamura and  H.  Funatsuki (2009) A novel major quantitative trait locus controlling seed development at low temperature in soybean (Glycine max). Theor. Appl. Genet. 118: 1477–1488.
•  Kasai,  A.,  K.  Kasai,  S.  Yumoto and  M.  Senda (2007) Structural features of GmIRCHS, candidate of the I gene inhibiting seed coat pigmentation in soybean: implications for inducing endogenous RNA silencing of chalcone synthase genes. Plant Mol. Biol. 64: 467–479.
•  Kasai,  A.,  S.  Ohnishi,  H.  Yamazaki,  H.  Funatsuki,  T.  Kurauchi,  T.  Matsumoto,  S.  Yumoto and  M.  Senda (2009) Molecular mechanism of seed coat discoloration induced by low temperature in yellow soybean. Plant Cell Physiol. 50: 1090–1098.
•  Kato,  S.,  T.  Sayama,  K.  Fujii,  S.  Yumoto,  Y.  Kono,  T.Y.  Hwang,  A.  Kikuchi,  Y.  Takada,  Y.  Tanaka,  T.  Shiraiwa et al. (2014) A major and stable QTL associated with seed weight in soybean across multiple environments and genetic backgrounds. Theor. Appl. Genet. 127: 1365–1374.
•  Kurauchi,  T.,  A.  Kasai,  M.  Tougou and  M.  Senda (2011) Endogenous RNA interference of chalcone synthase genes in soybean: formation of double-stranded RNA of GmIRCHS transcripts and structure of the 5′ and 3′ ends of short interfering RNAs. J. Plant Physiol. 168: 1264–1270.
•  Kurosaki,  H.,  S.  Ohnishi,  S.  Yumoto,  S.  Shirai and  I.  Matsukawa (2018) Improvement of the cold weather tolerance to low temperature during the flowering stage by introducing the lateral blooming trait in soybeans (Glycine max (L.) Merr) in Hokkaido. Jpn. J. Crop Sci. 87: 304–311.
•  Morrison,  M.J.,  L.N.  Pietrzak and  H.D.  Voldeng (1998) Soybean seed coat discoloration in cool-season climates. Agron. J. 90: 471–474.
•  Ohnishi,  S.,  H.  Funatsuki,  A.  Kasai,  T.  Kurauchi,  N.  Yamaguchi,  T.  Takeuchi,  H.  Yamazaki,  H.  Kurosaki,  S.  Shirai,  T.  Miyoshi et al. (2011) Variation of GmIRCHS (Glycine max inverted-repeat CHS pseudogene) is related to tolerance of low temperature-induced seed coat discoloration in yellow soybean. Theor. Appl. Genet. 122: 633–642.
•  Oka,  H.,  M.  Takahashi and  L.M.  Wang (1989) Influence of low temperature on browning of yellow hilum soybeans. Rep. Hokkaido Branch, Crop Sci. Soc. Jpn. and Hokkaido Branch, Jpn. Soc. Breed. 29: 22.
•  Oyoo,  M.E.,  E.R.  Benitez,  H.  Kurosaki,  S.  Ohnishi,  T.  Miyoshi,  C.  Kiribuchi-Otobe,  A.  Horigane and  R.  Takahashi (2011) QTL analysis of soybean seed coat discoloration associated with II TT genotype. Crop Sci. 51: 464–469.
•  Palmer,  R.G. and  T.C.  Kilen (1987) Qualitative genetics. In:  Wilcox,  J.R. (ed.) Soybeans: Improvement, Production, and Uses, 2nd edn. American Society of Agronomy, Madison, WI, pp. 135–209.
•  Rodriguez,  T.O.,  F.R.  Rodas,  M.E.  Oyoo,  M.  Senda and  R.  Takahashi (2013) Inverted repeat of chalcone synthase 3 pseudogene is associated with seed coat discoloration in soybean. Crop Sci. 53: 518–523.
•  Sanbuichi,  T. (1979) Studies on cool weather tolerance in soybean breeding. Rep. Hokkaido Pref. Agric. Exp. Stn. 28: 1–57.
•  Schori,  A.,  A.  Fossati,  A.  Soldati and  P.  Stamp (1993) Cold tolerance in soybean (Glycine max L. Merr.) in relation to flowering habit, pod set and compensation for lost reproductive organs. Eur. J. Agron. 2: 173–178.
•  Senda,  M.,  C.  Masuta,  S.  Ohnishi,  K.  Goto,  A.  Kasai,  T.  Sano,  J.S.  Hong and  S.  MacFarlane (2004) Patterning of virus-infected Glycine max seed coat is associated with suppression of endogenous silencing of chalcone synthase genes. Plant Cell 16: 807–818.
•  Senda,  M.,  T.  Kurauchi,  A.  Kasai and  S.  Ohnishi (2012) Suppressive mechanism of seed coat pigmentation in yellow soybean. Breed. Sci. 61: 523–530.
•  Senda,  M.,  M.  Kawasaki,  M.  Hiraoka,  K.  Yamashita,  H.  Maeda and  N.  Yamaguchi (2018) Occurrence and tolerance mechanisms of seed cracking under low temperatures in soybean (Glycine max). Planta 248: 369–379.
•  Srinivasan,  A. and  J.  Arihara (1994) Soybean seed discoloration and cracking in response to low temperatures during early reproductive growth. Crop Sci. 34: 1611–1617.
•  Suzuki,  T.,  M.  Sato and  T.  Takeuchi (2012) Evaluation of the effects of five QTL regions on Fusarium head blight resistance and agronomic traits in spring wheat (Triticum aestivum L.). Breed. Sci. 62: 11–17.
•  Takahashi,  R. and  J.  Abe (1994) Genetic and linkage analysis of low temperature-induced browning in soybean seed coats. J. Hered. 85: 447–450.
•  Takahashi,  R. and  J.  Abe (1999) Soybean maturity genes associated with seed coat pigmentation and cracking in response to low temperatures. Crop Sci. 39: 1657–1662.
•  Tanaka,  Y.,  K.  Tomita,  S.  Yumoto,  H.  Kurosaki,  H.  Yamazaki,  C.  Suzuki,  I.  Matsukawa,  T.  Tsuchiya,  K.  Shirai and  M.  Tsunoda (2003) A new soybean variety ‘Yukihomare’. Bull. Hokkaido Pref. Agric. Exp. Stn. 84: 13–24.
•  Tanaka,  Y.,  S.  Shirai,  S.  Yumoto,  I.  Matsukawa,  S.  Hagihara,  H.  Kurosaki,  H.  Yamazaki,  C.  Suzuki,  S.  Ohnishi and  M.  Tsunoda (2015) A new soybean variety ‘Toyoharuka’. Bull. Hokkaido Pref. Agric. Exp. Stn. 99: 47–60.
•  Yamaguchi,  N.,  T.  Sayama,  H.  Yamazaki,  T.  Miyoshi,  M.  Ishimoto and  H.  Funatsuki (2014a) Quantitative trait loci associated with lodging tolerance in soybean cultivar ‘Toyoharuka’. Breed. Sci. 64: 300–308.
•  Yamaguchi,  N.,  H.  Yamazaki,  S.  Ohnishi,  C.  Suzuki,  S.  Hagihara,  T.  Miyoshi and  M.  Senda (2014b) Method for selection of soybeans tolerant to seed cracking under chilling temperatures. Breed. Sci. 64: 103–108.
•  Yamaguchi,  N.,  H.  Kurosaki,  M.  Ishimoto,  M.  Kawasaki,  M.  Senda and  T.  Miyoshi (2015a) Early-maturing and chilling-tolerant soybean lines derived from crosses between Japanese and Polish cultivars. Plant Prod. Sci. 18: 234–239.
•  Yamaguchi,  N.,  F.  Taguchi-Shiobara,  T.  Sayama,  T.  Miyoshi,  M.  Kawasaki,  M.  Ishimoto and  M.  Senda (2015b) Quantitative trait loci associated with tolerance to seed cracking under chilling temperatures in soybean. Crop Sci. 55: 2100–2107.
•  Yamaguchi,  N.,  S.  Ohnishi and  T.  Miyoshi (2018) Screening for chilling-tolerant soybeans at the flowering stage using a seed yield- and maturity-based evaluation method. Crop Sci. 58: 312–320.
•  Yumoto,  S. and  K.  Sasaki (1990) Seed coat staining caused by treatment of cool temperature in soybeans with yellow hilum. Rep. Hokkaido Branch, Crop Sci. Soc. Jpn. and Hokkaido Branch, Jpn. Soc. Breed. 30: 39.
•  Yumoto,  S.,  I.  Matsukawa,  Y.  Tanaka,  H.  Kurosaki,  M.  Tsunoda,  T.  Tsuchiya,  K.  Shirai,  K.  Tomita,  K.  Sasaki,  M.  Kamiya et al. (1995) A new soybean variety ‘Toyohomare’. Bull. Hokkaido Pref. Agric. Exp. Stn. 68: 33–49.
•  Yumoto,  S.,  Y.  Tanaka,  H.  Kurosaki,  H.  Yamazaki,  C.  Suzuki,  I.  Matsukawa,  T.  Tsuchiya,  K.  Shirai,  K.  Tomita,  K.  Sasaki et al. (2000) A new soybean variety ‘Hayahikari’. Bull. Hokkaido Pref. Agric. Exp. Stn. 78: 19–37.