Abstract
The experiment was conducted during six consecutive years of 1998−2004 at six locations of Bangladesh to assess simple functional yield model for the cropping pattern of Mustard-Boro rice-T. aman rice under nutrient management. Models of N, P, K and S for mustard, boro rice and T. aman rice in the cropping pattern showed quadratic polynomial functions with different coefficients of determination. Response of seed yield of mustard to N (R2=0.68 at p=0.05) was found more favourable as compared to P, K and S (R2=0.52-0.56 at p=0.05). Functional relationship of grain yield of boro rice with N, P, K and S was found significant (R2=0.54-0.79 at p=0.05-0.01). Further, functional relationship of grain yield of T.aman with N, P, K and S was observed significant (R2=0.52-0.80 at p=0.05-0.01). Optimum nutrient levels for mustard (123-22-49-26 kg ha-1 of N-P-K-S), boro rice (145-26-91-21 kg ha-1 of N-P-K-S) and T.aman rice (136-19-72-15 kg ha-1 of N-P-K-S) were determined by using the developed models. Again, economic nutrient levels for mustard (118-16-46-25 kg ha-1 of N-P-K-S), boro rice (140-25-84-21 kg ha-1 of N-P-K-S) and T.aman rice (131-18-65-15 kg ha-1 of N-P-K-S) were also predicted by using the developed models. Yield of crops in Mustard-boro rice-T.aman rice cropping pattern can also be predicted against assumed nutrient levels by using the functional models. Overall, the model predictions of mustard, boro rice and T.aman rice yields were reasonable and showed consistency between observed and predicted yield.
