A temperature-based model of direct-, water-seeded rice (Oryza sativa) stand establishment in California

We used data sets from 23 field and greenhouse experiments to develop, validate, and evaluate a temperature-based model of stand establishment for direct-, water-seeded rice (Oryza sativa) culture. Data were split for model estimation and validation. The best performing stand density model used seed rate and total degree-days (adjusted R-2 = 65%). We detected no problems with this model's assumptions or statistical adequacy, and alternate models performed worse or were invalidated. The cultivars used in validation account for almost 75% of the rice currently grown in California. The model is therefore widely applicable and is useful either alone or as a submodel in rice growth and management models. Using long-term degree-day averages for California rice growing areas, we evaluated the effects of planting dates and rates on achieving optimal stand densities. These results showed why typical grower practices have been successful, but suggested ways that planting date and seeding rates might be adjusted to increase stand densities when weed problems are anticipated. (C) 1998 Elsevier Science B.V.