Intercropping is one of the most vital practice to improve land utilization rate in China that has limited arable land resource. However, the traditional intercropping systems have many disadvantages including illogical field lay-out of crops, low economic value, and labor deficiency, which cannot balance the crop production and agricultural sustainability. In view of this, we developed a novel soybean strip intercropping model using maize as the partner, the regular maize-soybean strip intercropping mainly popularized in northern China and maize-soybean relay-strip intercropping principally extended in southwestern China. Compared to the traditional maize-soybean intercropping systems, the main innovation of field lay-out style in our present intercropping systems is that the distance of two adjacent maize rows are shrunk as a narrow strip, and a strip called wide strip between two adjacent narrow strips is expanded reserving for the growth of two or three rows of soybean plants.  The distance between outer rows of maize and soybean strips are expanded enough for light use efficiency improvement and tractors working in the soybean strips.  Importantly, optimal cultivar screening and increase of plant density achieved a high yield of both the two crops in the intercropping systems and increased land equivalent ratio as high as 2.2.  Annually alternative rotation of the adjacent maize- and soybean-strips increased the grain yield of next seasonal maize, improved the absorption of nitrogen, phosphorus, and potasium of maize, while prevented the continuous cropping obstacles.  Extra soybean production was obtained without affecting maize yield in our strip intercropping systems, which balanced the high crop production and agricultural sustainability.

Grain number per-panicle is one of the most important components for rice yield. Spikelets on the primary and secondary branches determine the grain number per-panicle in rice. In this study, we identified a natural mutant, gnp4, lack of lateral spikelet on the secondary branches in the field condition. In addition, the Gnp4 and Lax1-1 double mutant showed dramatically reduced secondary branches and spikelets in panicle at reproductive stage, and tillers at vegetative stage. By map-based cloning approach, and using four F2 segregating populations, the Gnp4 gene was finally mapped to a 10.7-kb region on the long arm of chromosome 4 in rice. In this region, only one gene was predicted, and genomic DNA sequencing of the 10.7-kb region showed no nucleotide differences between the mutant and wild type. Interestingly, we found that the methylation level of several cytosines in the promoter CpG islands region of the predicted gene in gnp4 were different from the wild type. Thus, we propose that the DNA methylation changes at these sites may induce to decrease expression level of Gnp4, consequently, resulting in phenotypic variation.