We performed field experiments on the forage rice cultivation system to examine suitable operating conditions and conducted comparative studies with the customary transplanting system, by interseeding rice in wheat stand using no-till direct seeder with V-shaped furrow for the double cropping operation system.

In the wheat cropping system the treading of wheat plants was carried out once, and interseeding of rice was carried out from mid-February to the initial growth of the wheat flag leaf. These improves the establishment of seedling of rice and minimized the decrease in growth of wheat, with consequent increased in dry matter yields of forage rice and wheat. Moreover, when the customary transplanting system and no-till direct interseeding of rice in wheat stand system were compared in the field experiments, the operating hours were found to be 5.9 hours for the interseeding system compared to 28.7 hours for the custom system during May and June, when there are serious working competition. This resulted in a significant reduction in operating hours. The mowed dry matter yield of rice and wheat was 24.7t/ha for the interseeding system and was equivalent to the yield obtained with the custom system.

The study aimed to show the grass species suitable for grazing during the late autumn period in the Kanto-Koshinetsu region. For this, the change in productivity of several types of forage crops was investigated in the two regions. One site was located at an altitude of 1000m, and the other at 320m. Two varieties of oat (Avena sativa L.), one of barley (Hordeum vulgare L.), one of rye (Secale cereale L.), and two of Italian ryegrass (Lolium multiflorum Lam.) were sown from August to September at both sites. Crop production was measured during October to December. In both sites, the production of rye (Rai Taro) was the highest throughout the study period. The mid-October when to begin late autumn grazing, production of all the grass varieties except the Italian ryegrass at the 1000m site was more than 100g/m² ; all seeds were sown in late August. At 320m altitude, the production of all crops in mid-October was less than 35g/m², regardless of the time of seeding. Thus, because the production was too low to begin grazing at the 320m site from mid-October, it is necessary to delay the start of grazing.

In order to extend the stocking period in fall and winter, the best species and seeding time should be selected, and grazing management should be planned, to enable more animals to be stocked for a longer period. Therefore, we de-veloped a predictive model for herbage production of oat, rye, and Italian ryegrass sown in late summer or early fall. To construct a growth curve, data collected from both a field experiment, which measured the early growth of the species, and a database, which contains information on recommended varieties of the species, were used. Temperature at each seed-ing day was included as a covariate, and it was derived by multiple regression of the mean yearly temperature at the sowing place and the seeding date. Although the constructed model showed low accuracy, it could predict a trend in ob-served mass for days after seeding and temperatures on the seeding date. The model could also yield a stable estimate for cross-validation of independent unknown data. Thus, it can be concluded that the model aids in approximate estimation, which could contribute to the development of a grazing man-agement plan for extending the stocking period in fall and winter using winter annuals.