Enhancing the ability of CERES-Maize to compute light capture

Recently it has been proposed to use the relationship between average intercepted photo-synthetically active radiation (IPAR) around silking and total number of seeds per plant as the basis to improve kernel number prediction in CERES-Maize. However, there has been no previous evaluation of the accuracy of IPAR predictions in the model. The objectives of this work were to evaluate CERES-Maize predictions of IPAR around silking by testing their components incident PAR, light extinction coefficient (k), and leaf area index (LAI), and to develop alternative methods to simulate PAR and k. Measured IPAR was averaged over a thermal time window of 250 before to 100 growing degree-days after silking and compared with model predicted IPAR averaged over the same thermal time window. Independent data sets were used to develop and test new relationships to predict fraction of PAR to solar radiation as a function of incident total solar radiation, and canopy extinction coefficient as a function of the crop phenological age. The new relationships were incorporated into CERES-Maize and the new IPAR predictions were compared with measured values. We found that the common assumption of predicting PAR as 50% of the solar radiation overestimated PAR for our conditions in Iowa, where a value of 43% worked better. The extinction coefficient changed with crop development, reaching a peak of 0.66 at silking, and being lower early and late in the season. The best IPAR predictions were obtained when the new procedures to convert solar radiation into PAR and estimate k were coupled with the original leaf area model of CERES-Maize. (C) 2003 Elsevier Science Ltd. All rights reserved.