Simulating leaf appearance in rice

Most rice (Oryza sativa L.) simulation models assume that only temperature affects leaf appearance rate (LAR). This assumption ignores results from controlled environment studies that show that LAR in rice is not constant with time (calendar days) under constant temperature. The Streck model, which takes into account age effects on LAR, improved the prediction of leaf appearance in winter wheat (Triticum aestivum L.) cultivars compared with the Wang and Engel (WE) model and the phyllochron model but has not been evaluated in rice. The objective of this study was to adapt and evaluate the Streck model to simulate main stem LAR and leaf number in rice. A 4-yr experiment with several sowing dates from 2003-2004 to 2006-2007 was performed at Santa Maria, RS, Brazil. Seven rice cultivars; were used: IRGA 421, IRGA 420, IRGA 417, IRGA 416, BRS 7 (TAIM), BR-IRGA 409, and EPAGRI 109. Plants were grown in 12-L pots during the 4 yr, and in a paddy rice field during the 2006-2007 growing season. Coefficients necessary to run the Streck model, the WE model, and the phyllochron model were estimated with data from five sowing dates of the 2003-2004 growing season and the models were evaluated with independent data from the other three growing seasons. Predictions of the main stem leaf number, represented by the Haun Stage (HS), were better with the Streck model. The RMSE was 0.7, 1.0, and 1.8 leaves, for the Streck model, the WE model, and the phyllochron model, respectively.