Research indicates that, owing to the enhanced grain-filling rate of tetraploid rice, its yield has notably improved compared to previous levels. Studies conducted on diploid rice have revealed that optimal planting density and fertilization rates play crucial roles in regulating rice yield. In this study, we investigated the effects of different nitrogen application and planting density treatments on the growth, development, yield, and nitrogen utilization in tetraploid (represented by T7, an indica–japonica conventional allotetraploid rice) and diploid rice (Fengliangyou-4, represented by FLY4, a two-line super hybrid rice used as a reference variety for the approval of super rice with a good grain yield performance). The results indicated that the highest grain-filling rate of T7 could reach 77.8% under field experimental conditions due to advancements in tetraploid rice breeding. This is a significant improvement compared with the rate seen in previous research. Under the same conditions, T7 exhibited a significantly lower grain yield than FLY4, which could be attributed to its lower grain-filling rate, spikelets per panicle, panicle number m⁻², and harvest index score. Nitrogen application and planting density displayed little effect on the grain yield of both genotypes. A higher planting density significantly enhanced the leaf area index and biomass accumulation, but decreased the harvest index score. Compared with T7, FLY4 exhibited a significantly higher nitrogen use efficiency (NUE_g), which was mainly due to the higher nitrogen content in the straw. Increasing nitrogen application significantly decreased NUE_g due to its minimal effect on grain yield combined with its significant enhancement of nitrogen uptake. Our results suggest that the yield and grain-filling rate of T7 have been improved compared with those of previously tested polyploid rice, but are still lower than those of FLY4, and the yield of tetraploid rice can be further improved by enhancing the grain-filling rate, panicle number m⁻², and spikelets per panicle via genotype improvement.

Soil salinization is a major environmental constraint that poses a significant threat to global agricultural productivity and food security. Coix lacryma-jobi L., a minor cereal crop that is valued for its nutritional and medicinal properties, displays moderate susceptibility to salinity stress. Although exogenous treatments have been demonstrated to enhance plant resilience against various biotic and abiotic stresses, the potential of nano-silicon (NaSi), melatonin (MT), and their combined application in mitigating salinity-induced damage, particularly in relation to the medicinal properties of this medicinal and edible crop, remains poorly understood. This study investigated the effects of exogenous NaSi and MT application on Coix under salinity stress using two varieties with contrasting salinity tolerances. The plants were subjected to salinity stress and treated with NaSi, MT, or a combination of both. The results revealed that salinity stress significantly impaired the agronomic traits, physiological performance, and accumulation of medicinal compounds of Coix. Exogenous MT application effectively alleviated salinity-induced damage to agronomic and physiological parameters, exhibiting superior protective effects compared to NaSi treatment. Strikingly, the combined application of MT and NaSi demonstrated synergistic effects, leading to substantial improvements in growth and physiological indices. However, the medicinal components were only marginally affected by exogenous treatments under both control and salinity-stressed conditions. Further clarification of the molecular mechanisms underlying salinity stress responses and exogenous substance-induced effects is critical to achieving a comprehensive understanding of these protective mechanisms.

Panicle architecture is a key agronomic trait determining rice yield. Identifying and characterizing the regulatory genes controlling panicle morphogenesis is essential for breeding improved rice varieties. To clarify the regulatory pathways underlying panicle architecture and identify novel candidate genes, we performed RNA sequencing (RNA-seq) of 1 cm and 15 cm young panicles from Zhonghua 11 (ZH11). Numerous differentially expressed genes were identified, primarily enriched in metabolic and cellular processes. Among the differentially expressed transcription factors, we identified a previously uncharacterized Homeodomain-Leucine Zipper (HD-Zip) family gene, PaniCle Development 8 (PCD8; LOC_Os08g37580), which was highly expressed in young panicles. CRISPR/Cas9-mediated knockout of PCD8 (PCD8-KO) significantly reduced grain number per panicle, seed setting rate, and grain width while increasing grain length, grain length-to-width ratio, and chalkiness, indicating that PCD8 regulates panicle morphogenesis. Analysis of the 3,000 Rice Genomes (3 K-RG) population identified four major PCD8 haplotypes. Haplotype4 (Hap4), a temperate japonica-specific haplotype associated with long grains, carries a G/A substitution at - 2464 bp in the promoter. Haplotype3 (Hap3) and Hap4 likely evolved from Haplotype1 (Hap1), and all four haplotypes are closely related. PCD8 showed lower nucleotide diversity in temperate japonica and indica than in wild rice. Neutrality tests indicated balancing selection in wild rice and positive selection in indica, whereas reduced diversity in temperate japonica mainly reflects domestication bottlenecks rather than strong locus-specific selection. DNA affinity purification sequencing (DAP‑seq) identified candidate downstream targets of PCD8, and electrophoretic mobility shift assays (EMSA) confirmed direct binding to selected targets. Multi-omics analyses demonstrate that PCD8 is a novel transcription factor regulating panicle morphogenesis through downstream gene networks, providing a promising candidate for high-yield rice breeding.

A high-quality rice population structure serves as the foundation for achieving high yields, with plant spacing and the number of seedlings per hill playing key roles in regulating this structural development. However, the combined effects of seedling number per hill and plant spacing on yield formation and resource allocation in rice ( Oryza sativa L.) are not yet fully understood. This study aimed to assess the effects of coordinated changes in seedling number per hill and plant spacing on source-sink characteristics and yield in rice, while maintaining a constant seedling density. The study evaluated the response of four rice varieties to combined variations in seedling number per hill and plant spacing, all under a baseline seedling density of 60 × 10 ⁴ ha ^- ¹, over a two-year field experiment. Linear models, mixed-effects models, ridge regression, and structural equation modeling were used to examine the effects of these factors on rice growth, biomass allocation, nitrogen and carbon accumulation, and yield. Results showed that increasing seedlings per hill and plant spacing reduced yield and above-ground biomass (AGB) but increased the harvest index (HI). Single seedlings with narrow spacing produced the highest yield (7.27 t ha ^- ¹), AGB (16.4 t ha ^- ¹), and the lowest HI (0.46). The combined effects of seedling number per hill and plant spacing negatively affected tiller number, number of effective panicles, leaf area index, and nutrient accumulation, while positively influencing specific leaf area and panicle biomass allocation. Organic carbon accumulation (43.52%) and biomass accumulation (31.32%) were major contributors to yield variation. Path analysis indicated that 96% of yield variance was accounted for, highlighting the importance of panicle biomass, organic carbon accumulation, and the source-sink balance between stems and panicles. In conclusion, single seedlings with narrow spacing (12.93 cm spacing with 1 seedling per hill) optimize the source-sink balance in rice, compensating for a lower harvest index by enhancing biomass accumulation, thereby increasing yield.

Rice stem internal structure is a critical micro-phenotype influencing lodging resistance and yield; however, its analysis remains constrained by labor-intensive manual methods. Here, we present a publicly available dataset of 686 rice stem cross-sections, with 21,027 large vascular bundles (LVBs) and 19,342 small vascular bundles (SVBs) manually annotated. Five deep learning architectures were systematically evaluated, among which UNet-VGG16 achieved the best performance with a mean intersection over union (mIoU) of 87.4% (82.49% for LVBs and 74.41% for SVBs). An improved model, UNet-ECA-Bio, further raised mIoU to 89.32% and SVB IoU to 78.97% by integrating Efficient Channel Attention (ECA) and biologically informed class weighting using an image-level dataset. Leveraging these high-accuracy phenotypic predictions, genome-wide association studies (GWAS) indicated concordance between annotated and predicted traits, with SNP overlap rates of 96% (LVB count: 1,217/1,262), 43% (SVB count: 6/14), 98% (stem area: 122/124), and 100% (cavity area: 3/3) at −log10(p) ≥ 6. Meanwhile, compared with manual annotation (estimated 10–30 minutes per image), the proposed approach processed all 686 images within 10 minutes, representing a >600-fold increase in throughput. We further developed a user-friendly software tool, “Rice_Stem_Pre_V1.1.exe,” for automated phenotyping of 14 stem traits, providing a cost-effective platform for genetic studies of lodging resistance and yield improvement.