Modelling Seed Yield Estimation in Jatropha Curcas Plantations in Gujarat

Jatropha curcas Linn. is considered as a promising source of bio-diesels since the plant grows well in semi-arid marginal soils and is fairly resistant to pests and diseases. The seed parameters are highly variable and the growth and yield are inter-related to each other. The yield potential of seed is affected by many factors. The parameters like height, diameter, crown, age and density are good tools to predict the seed yield in J. curcas. Attempts have been made to develop models to predict the seed yield in Jatropha curcas. Both, the linear and non-linear equations were used to develop the model. However, linear models suffered with the problems of abnormal distribution and residual autocorrelation and hence non-linear model was found suitable to estimate the seed yield based on the age and growth parameters. The model developed is based on the dataset collected from two plantations at a particular location having density of 2500 plants/ha. For wider applicability of the model, larger dataset is required which should have data from different locations and from different planting densities and age groups.