Field performance on grain yield and quality and genetic diversity of overwintering cultivated rice (Oryza sativa L.) in southwest China

Understanding the field performance on grain yield and quality and the genetic diversity of overwintering (OW) cultivated rice (Oryza sativa L.) across main crop (MC) and ratooning crop (RC) is the premise to make strategies for the future OW rice variety improvement in rice production. The present field experiments were conducted in RC of 2016, in MC of both 2017 and 2018, and RC in 2019 to identify genotypes OW rice that perform stable in terms of grain yield and quality across different climate conditions. The grain yield plant-1 (GYP) and its components in six genotypes of OW rice exhibited significant difference across the 4 years (P ≤ 0.05), the maximum GYP in OW6 rice was harvested (60.28 g) in MC of 2017, but the minimum GYP in OW1 rice was harvested (33.01 g) in MC of 2018. Within six genotypes of OW rice, four grain shape traits displayed a relative small significant difference, four grain quality traits exhibited a relative small significant difference except for chalkiness rate (CR), there 226 pairs of significant PCC values between GYP and its components were calculated in all tested rice and varied from six in OW6 to eleven in OW1, there 130 pairs of significant PCC values among the four grain shape traits were calculated and ranged from twenty-one in OW1, 3, 5 to twenty-three in OW2, there 118 pairs of significant PCC values among the four grain quality traits were calculated and ranged from seventeen in OW2 to twenty-three in OW1. The numbers, directions, and size of PCC values for the grain yield and quality characters in all tested rice displayed a series of irregular variations. Six genotypes of OW rice were apparently distinguished by employing 196 pairs of simple-sequence repeats (SSRs) markers and exhibited abundant genetic diversity at the DNA level. Data from this study provide an extensive archive for the future exploration and innovation of overwintering cultivated rice variety.

perennial Dongxiang wild rice. Hu et al 13 and Zhang et al 14 have succeeded in developing perennial rice through selecting O. longistaminate as gene donors. The field performance on both GYP and its components in perennial rice genotypes were examined and easily influenced by both genotype and external environment factors 15,16 . Up to now, the genes or QTLs underlying perennial trait from O. longistaminata have been successfully transferred into common cultivated rice and even perennial rice variety would be commercially released to farmers as result of reduced labor costs 17 . Meanwhile, the genes (QTLs) underlying overwintering traits was also hidden in the existing cultivated rice variety 18 . However, the field performance on both grain yield and quality and even genetic diversity of overwintering cultivated rice has been largely unknown. Currently, it is necessary to perform the study on field performance variations in grain yield and quality of overwintering cultivated rice across different environmental conditions before we integrate multi-genes (QTLs) underlying overwintering traits into a novel overwintering rice variety adapted to Chongqing in southwest China.
To this end, six genotypes of overwintering (OW) rice identified during the natural snowy winter season in Chongqing, southwest China, were grown to evaluate the field performance variations in grain yield and quality under different climate condition. The present objectives were (i) to evaluate field performance variations in grain yield and quality of OW rice across main crop (MC) and ratooning crop (RC), (ii) to identify genotypes OW rice that perform stable in terms of grain yield and quality across different climate conditions, and examine the genetic diversity of all tested rice using SSR marker, and (iii) to identify excellent OW cultivated rice resources involved in elucidating the molecular mechanisms about OW characteristics and the development of novel OW cultivated rice variety.
In RC of 2016, rice stub of all tested rice were retained with 30 cm, survived through natural cold-winter season, germinated and flowering and being harvested for the grain yield and quality evaluated. In MC of 2017, seeds collected from six OW cultivated rice were sown on 15 March 2016 and 35-day-old seedlings of all tested rice were transplanted into four-row plots with six plants per row, 20 cm between plants within each row, and 23 cm between rows. Similarly, a parallel test was performed in both MC of 2018 and RC of 2019, on 10 March 2018, seeds of all tested rice were sown and 35-day-old seedlings of all OW cultivated rice were also transplanted. In RC of 2019, all OW cultivated rice could germinate after being harvested for grain yield and quality evaluated (Supplementary Fig. 2-4).
A special rice compound fertilizer (450 kg urea ha -1 ) occupied more than 45% of total nutrients was made up of 12% N, 18% P 2 O 5 , and 15% K 2 O and applied at the basal stage. Nitrogen (180 kg urea ha -1 , 46% N) was applied 2 weeks after transplanting seedlings. The water management strategy adopted was shallow water at the transplanting stage and flooding midseason with drainage-reflooding-moist intermittent irrigation. Weed control, pest management, and disease treatment were carried out to avoid grain yield loss. DNA extraction and PCR amplification. Six genotypes of OW cultivated rice and two sequenced rice varieties of Nipponbare and 93-11 were sampled at stage of rice tillers. The rice genomic DNA was extracted referred to the method cetyltrimethylammonium bromide (CTAB) described by Nei 19 . A total of 196 pairs of simple-sequence repeats (SSR) markers were employed to evaluate the genetic diversity of six genotypes of OW cultivated rice and two sequenced rice (data not shown). PCR amplifications were performed referred to the protocol described by Greer et al. 20 . DNA products were separated by 8% polyacrylamide gel electrophoresis (PAGE).

Measurements. Grain yield.
A random sample of five plants per plot within a single genotype of OW rice for each replication across RC of 2016, MR of both 2017 and 2018, and RC of 2019 was collected to measure the following phenotypic values according to the method described by Shen 21 : HD, days to heading; PH, plant height (cm); PP, panicles plant -1 ; PL, panicle length (cm); FGP, fulling grains panicle -1 ; EGP, empty grains panicle -1 ; SP, spikelets panicle -1 ; GW, 1000-grain weight (g), GYMP, grain yield of major panicle (g); and GYP, grain yield plant -1 (g) were measured. Grain shape traits including GL, grain length (mm); GW, grain width (mm); and GT grain thickness (mm) were measured with a Mitutoyo absolute digimatic caliper (Model 500-173). Three derived traits were calculated including GSR, grain setting rate (%); GSD, grain-setting density; and LWR, length to width ratio. The phenotypic data for each trait in five plants within individual genotype with three replicates were calculated for statistical analysis.
Grain quality. Four grain quality traits including CR, chalkiness rate (%); AC, amylose content (%); GC, gel consistency (mm); and ADV, alkali digestion value (class) were measured at the Rice Product Quality Inspection & Supervision Testing Center, Ministry of Agriculture. China National Rice Research Institute (CNRRI), Hangzhou, China. Chalkiness version 2.0 software was used to measure CR according to the method described in NY14788 and GB/T17891-1999 22 ; the spectrophotometric method according to the Chinese Ministry of Agriculture's NY/T 2639-2014 was used to measure AC. Two methods described in GB/T22294-2008 and NY/T83-2017 were used to measure GC and ADV, respectively 23 .   T   7.15  30  22  33  25  37  27  33  25   7.16  35  25  36  27  38  29  27  24   7.17  34  25  37  27  38  29  30  25   7.18  34  25  37  27  39  28  30  25   7.19  30  25  35  28  40  29  31  26   7.20  36  25  36  28  39  30  32  27   7.21  38  29  37  27  40  30  34  27   7.22  38  29  39  29  40  31  36  24   7.23  37  28  38  29  39  28  30  25   7.24  39  28  39  29  40  32  33 (Table 2). Within individual genotype of OW rice, the GYP almost displayed significant difference across the 4 years (Fig. 2). The maximum GYP in genotype of OW6 rice was harvested (60.28 g) in MC of 2017. However, the minimum GYP in genotype of OW1 rice was investigated (33.   www.nature.com/scientificreports/ and displayed a relatively regular variation. However, the GSR in two genotypes of OW2 and OW3 were more than 80% across the 4 years. However, the GSR in the remaining four genotypes of OW rice partially displayed more than 80% across the 4 years. The remaining seven grain yield characters of PL, FGP, EGP, SP, GSD, GW and GYMP in all tested rice displayed a series of irregular significant difference across the 4 years. In summary, the field performances on GYP and its component in all tested rice expressed rather unstable under different climate conditions.  (Table 4). Within individual genotype of OW rice, the four grain shape traits also exhibited small differences across the 4 years (Fig. 3). The GL in four genotypes of OW1, OW2, OW3, and OW6 displayed nonsignificant difference across the 4 years, but exhibited small significant difference in two genotypes of OW4 and OW5. The GW in genotypes of OW3 and OW6 exhibited small differences across the 4 years, but exhibited relatively stable in genotypes of OW1, OW2, OW4, and OW5. The GT in OW1, OW4, and OW6 exhibited nonsignificant differences across the 4 years. The LWR in all tested rice except for OW5 exhibited relative stable across the 4 years.

PCCs between GYP and its components in OW rice.
Overall, the field performances on four grain shape traits in partial tested OW rice displayed significant difference and expressed relatively stable across the 4 years.  (Table 6). In particularly, a series of significant difference on CR was simultaneously observed in all tested rice across the 4 years ( Fig. 4; Supplementary Fig. 5). Within individual genotype of OW1 rice, The AC exhibited nonsignificant difference and expressed rather stable across the 4 years, but displayed significant difference on the remaining five genotypes of OW rice. The ADV in genotypes of both OW5 and OW6 displayed nonsignificant difference, but exhibited significant difference in four genotypes of OW1, OW2, OW3, and OW4. The GC displayed nonsignificant difference in three genotypes of OW1, OW4, and OW5, but displayed a relatively small significant difference in three genotypes of OW2, OW3, and OW6. Overall, a series of significant differences on the four grain quality traits were partially observed in all tested rice.
PCCs for grain quality traits in OW rice. Altogether 118 pairs of significant positive PCCs for the four grain quality traits exhibiting theirs various numbers, directions, and size in the six genotypes of OW rice across the 4 years (   Genetic diversity of OW rice. Six genotypes of OW rice, Nipponbare, and 93-11 were roughly split into four groups at the genetic distance of 0.38 by employing a total of 196 pairs of SSR markers. However, only nine pairs of SSR markers displayed abundant polymorphism, which apparently revealed the molecular differences between six genotypes of OW rice and two sequencing rice varieties ( Fig. 5 and 6, Table 8). Genotype of OW1 was separately split into the group I at the genetic distance of 0.384. Genotypes of OW2, OW3, OW5, and Nipponbare were divided into the group II at the genetic distance of 0.40, the group II was split into two subgroups, Genotypes of OW2, OW3, and Nipponbare were split into the 1st subgroup II at the genetic distance of 0.42, OW5 was separately divided into the 2nd subgroup II, Genotypes of OW2, OW3, and Nipponbare could be apparently distinguished at the genetic distance of 0.43. Genotype of OW6 was separately split into the group III at the genetic distance of 0.40. Genotypes of OW4 and 93-11 were divided into the group IV; it was clear to distinguish between 93-11 and OW4 at the genetic distance 0.15. Overall, six genotypes of OW rice, Nipponbare and 93-11 could be apparently distinguished by nine pairs of SSR markers and exhibited abundant genetic diversity at DNA level.

Discussion
Overwintering cultivated rice resource screened from the existing cultivated rice variety. Developing new genotype of OW rice that combined high yield and good quality across the four seasons will help increase farms' income by through labor costs saved. However, to screen OW cultivated rice resource was the premise to perform the OW cultivated rice variety breeding project. The present six genotypes of OW cultivated rice resource could survive through natural snowy winter seasons and germinated in the following spring and be harvested after paddy cultivation in the following autumn, which might contain a series of genes or QTLs underlying the OW characteristics, and was prior to be considered as excellent OW rice resources involved into understanding the molecular mechanisms about the OW characteristic 12 . In particularly, the present six genotypes of OW cultivated rice displayed abundant genetic diversity at the DNA level and were apparently dis- www.nature.com/scientificreports/ tinguished by employing a total of 196 pairs of SSR markers. However, only nine pairs of SSR markers displayed abundant polymorphism, which apparently revealed the molecular differences between six genotypes of OW rice. More interesting, it may be feasible to integrate OW genes into the current backbone parent of super hybrid rice to develop a novel genotype of OW cultivated rice variety for the future agricultural production and ecological restoration 24,25 . Therefore, special attention should be given to identification of OW cultivated rice resource in the future rice genetics and breeding project.

Susceptible GYP and its components in OW rice. For a long time, rice breeders and cultivators have
devoted themselves to developing new genotypes of rice with relatively stable field performance on grain yield and quality under different rice cropping systems 26,27 . However, in different genotypes of rice exhibit significant differences in their field performance regarding grain yield and quality traits under different cultivation ecosystem 28,29 . Understanding the field performance variation of GYP and its components in OW rice across the four seasons was the foundation of performing the future OW cultivated rice breeding projects and even commercially released to the farmers 15,16 . Therefore, the present study evaluated the field performance variations on GYP and its components in different genotype of OW rice for the precise identification of stable agronomical traits about OW rice across the 4 years. We observed that the major determinants of GYP, including HD, PP, SP, and GW contributed unequally to GYP in all tested rice across the 4 years. PP in MC of 2018 exhibited a significantly positive correlation with GYP in OW6 but two significantly negative correlations with GYP in OW6 of MC 2017 and RC 2019, respectively. The positive main determinants of GYP in the same OW rice may become negatively contributing components of GYP across the 4 years. These findings were not in agreement with the       31 . This phenomenon resulted from a series of significant field performance variations occurred in GYP and its components in six genotypes of OW rice. The field performance on both GYP and its components in all tested rice expressed rather unstable and displayed significant difference across the 4 years and were easily affected by both external environmental factors. Three important reasons could be applied to explain the reason for that. Firstly, a complicated regulatory network might exist in the grain yield and its components of OW cultivated rice and was strongly influenced by both genotype and environmental factors. Secondly, all genotypes of OW rice identified through natural snowy winter conditions exhibited significant difference on GYP and its components and might possess a set of unique patterns of grain yield and quality related to traits. Thirdly, the grain yield and quality characters in six genotypes of OW rice were controlled by both major gene and polygene and easily affected by multi-external environmental factors of weather conditions, transplanting time, planting density, and fertilization. More interestingly, genotypes OW2 and OW3 exhibited nonsignificant variations in GSR as result of being insensitive to daily Max T (°C) during each ripening period even if the significant difference on daily Max T (°C) occurred at the same period across the 4 years. This finding is in agreement with the previous report by Ishimaru et al 32  Relatively stable grain shape and grain quality traits in OW rice. The GL, GW, GT, and LWR are important indicators for grain weight and final yield judgment and influence the yield and quality 33 . Grain shape traits have been widely accepted as complex traits controlled by multiple genes with small genetic effects 34,35 . In all tested rice across the 4 years, the field performance on GL in OW4 and OW5 and GW in OW2 and OW6 only displayed a relative small significantly different, the GT in OW4 and OW5 and the LWR in OW5 exhibited a small significantly different. The four grain shape traits expressed relative stable in partial OW cultivated rice across the 4 years. This result also suggests that seasonal variation had small impact on grain shape variations. Similar results were reported by Wan et al. 36 and Fan et al. 37 . The grain shape traits of OW rice exhibited a relatively stable field performance through four different seasons and should not be given more attention in the future development of the OW cultivated rice variety.
Four grain quality traits including CR, AC, ADV, and GC have been widely considered as important indicators of grain quality for the new rice variety breeding 38 . Among them, the CR was the most important determinant of grain appearance quality in rice production 39,40 . These previous studies have reported that the CR was a complex quantitative trait that was controlled by polygenes and easily influenced by environmental factors 41,42 . In the present study, the CR in all tested OW rice exhibited significant difference and expressed rather unstable across the 4 years, the CR in all tested rice might be affected by daily Max T, °C during each ripening period for the rice grain filling 43,44 . The other three quality traits in all tested rice exhibited a relative small significant difference across the 4 years and not were easily affected by both genotype and genotype × environment, which might be regulated by the Waxy gene 45,46 . The numbers, directions, and size of PCC values for the four grain quality traits displayed a series of irregular variations and were strongly affected by both genotype and genotype × environment. Similar results have been reported in previous studies 47,48 . Consequently, special attention should be given to the chalkiness rate in the future OW cultivated rice improvement program. Meanwhile, the remaining three grain quality traits displayed nonsignificant difference on partial tested rice across the 4 years, this was suggested that some gain quality traits in some OW cultivated rice variety could expressed rather stable and were insensitivity www.nature.com/scientificreports/ to external environmental factors across the 4 years. Consequently, we should be given more attention to the stable grain quality traits in the future development of the OW cultivated rice variety.