MODELING A RYE COVER CROP AND SUBSEQUENT SOYBEAN YIELD

Soybean [Glycine mar (L.) Merr.] can be no-till seeded in the spring f ollowing a killed rye (Secale cereale L.) cover crop. A shortage of ra in can affect soybean yields, depending on the timing of killing and t he amount of mulch produced. Our main objective was to incorporate the effect of a rye mulch on soil water content into a soybean growth mod el (SOYGRO). The use of a wheat (Triticum aestivum L.) growth model (C ERES-Wheat) for rye mulch production calculations and SOYGRO for predi cting yields under no-mulch and mulch conditions were evaluated. The w ater balance subroutine of SOYGRO was modified to account for the effe ct of a rye mulch on soil water content due to mulch albedo, rain inte rception, and transmissivity of solar radiation. Data from previous ex periments at Delhi (Psammentic Hapludalf soil), and Woodstock (Typic H apludalf) in Ontario, Canada, were used to evaluate these modification s and yield predictions. Mulch production was described well by CERES- Wheat (r(2) = 0.96). The modified SOYGRO model improved soil water con tent calculations relative to the original model (r(2) = 0.67 vs. 0.58 , root mean square error = 0.028 vs. 0.064 m(3) m(-3)). Simulated soyb ean yields were within 1 SD of measured values. In agreement with held measurements, the simulated effect of mulch did not increase soybean yields, and modification of SOYGRO to include mulch effects did not ch ange yield predictions (r(2) = 0.83 vs. 0.81) when no prolonged drough t occurred. The simple mulch model incorporated into SOYGRO was adequa te to model the effect of rye mulch on soybean growth. The use of CERE S-Wheat and the modified SOYGRO model in sequence under hypothetical d rier weather conditions was demonstrated.