Adapting the CSM-CROPGRO model for pigeonpea using sequential parameter estimation

Abstract

Pigeonpea (Cajanus cajan (L.) Millsp.) is an important crop in Asia, Africa, Latin America and Caribbean. Despite pigeonpea's global importance there is a dearth of crop simulation models available for studying pigeonpea growth. The objectives of this study were to adapt the CSM-CROPGRO model for simulating pigeonpea growth and development through parameter modification and to illustrate the use of a sequential parameter estimation technique using a dataset from Gainesville, FL in 1984 and a dataset from India in 2003. The sequential approach to parameter estimation using a hybrid Metropolis–Hastings–Gibbs algorithm worked well at estimating physiologically plausible values for parameters with good correspondence to measured data. Reasonable results were obtained despite the use of approximations for measurement errors for the Gainesville dataset, which contained only treatment means. This study demonstrated that CROPGRO can be used to simulate the growth and development of pigeonpea. However, further testing of CROPGRO with more extensive pigeonpea datasets should be undertaken to confirm the accuracy of the parameter estimates developed from this study. Further theoretical and practical research into the parameter estimation approach is needed.