A predictive model of chickpea (Cicer arietinum L.) yield

Chickpea evapotranspiration (ET) varied considerably among different irrigation regimes and yield was linearly related to cumulative water use. The results of irrigation experiments carried out at Lincoln University, Canterbury during 1998-99 and 1999-2000 were re-assessed in order to develop a simple empirical model for predicting seed yield. The general form of this model is based on functions of water use and green area index (GAI), which were used to predict total intercepted radiation. Two years of field data for chickpea, cultivar Sanford, were used to develop empirical relationships among water use, total intercepted radiation, GAI, total dry matter, pods/m(2) and seed yield. The model derived was successful in accurately predicting total intercepted PAR. A comparison of measured and modelled estimates of total intercepted PAR for three cultivars (Sanford, Dwelley and B-90) was significant (r(2) = 0.79 - 0.97). Seed yield of the three cultivars, both rainfed and irrigated, was predicted reasonably well. The root mean square deviations (RMSD) were 10.9 g/m(2) and 61.4 g/m(2) for rainfed and irrigated chickpeas, about 4% and 14% of the mean seed yields, respectively. Future work requires the inclusion of climate parameters, subroutines for crop specific processes and testing over a wider range of environments and cultivars.