Grain protein content (GPC) is an important determinant of rice nutritional and eating quality. In this study, 624 cold-region japonica rice accessions were evaluated for GPC and taste value at Harbin Minzhu, Ulanhot, and a saline–alkali site in the Songnen Plain, and 495 accessions with complete phenotypic and genotypic data were used for genome-wide association analysis. GPC showed significant positive correlations across sites and an entry-mean broad-sense heritability of 0.663. Using 1,527,603 high-quality SNPs, GLM-Q identified 56 significant GPC-associated SNPs, mainly on chromosomes 11 and 7, with site-specific patterns. Annotation of GLM-Q significant SNPs yielded 28 non-redundant positional candidate genes. In GEMMA, chromosome 11 signals did not remain genome-wide significant, whereas chromosome 7 associations remained significant. Haplotypes of Os11g0239500, Os11g0239900, Os07g0264100, and Os03g0244200 differed in GPC and taste value. Joint haplotype analysis of Os11g0239500 and Os07g0264100 classified 319 accessions into four types. Type IV had the lowest mean GPC (7.84%) and the highest mean taste value (81.14), whereas Type I had a mean GPC of 8.78% and a mean taste value of 77.45. The identified candidate genes and haplotypes may be useful for screening cold-region japonica rice with low GPC and high eating quality.

Direct seeding of rice reduces labor intensity and cost, helping alleviate labor shortages in cold-region rice production. To investigate the effects of mechanical precision hill-direct seeding versus mechanical transplanting on yield and nutrient accumulation in cold regions, a set of field split-plot experiments were conducted with cultivation method as the main plot and rice variety as the sub-plot. Our comprehensive measurement results indicate that transplanting significantly increased yield by enhancing tiller number, filled grains per panicle, and grain weight per hill, with significant varietal differences observed. No significant difference in 1000-grain weight was found between the two cultivation methods. Except for Zn content, different cultivation methods have no significant effect on other measured nutrients such as N, P, K, Fe, starch, and fat. Transplanting significantly increased effective tiller number (an increase of 2.6 tillers per hill) and filled grains per panicle (an increase of 12.4 grains), with a significant variety–cultivation method interaction. Qijing 2 (QJ2) and Tiandao 261 (TD261) were more suitable for transplanting to achieve high yield potential, whereas Longgeng 3038 (LG3038) and Tianxiangdao 9 (TXD9) obtained relatively high yields under direct seeding. Therefore, appropriate cultivation methods should be selected based on varietal characteristics: transplanting is recommended for high-yield-potential varieties, while simplified direct seeding is advised for varieties tolerant to direct seeding. Overall, this is a comprehensive consideration and rational strategy based on balancing rice yield, revenue, and benefit, as well as ensuring both food security and farmer income of the entire country and society.