LAUSR

Abstract Castor bean ( Ricinus communis L.) seed oil has unique characteristics and can be used as feedstock for many biobased products. Under greenhouse conditions, a previous recurrent selection for high-oil castor seeds from a base population resulted in a new population with an increased mean oil content from 50.33 to 54.47% dry weight, average single seed weight from 0.44 to 0.54 g, and correlation between oil content and weight from moderate ( r  = 0.43, p   0.0001) to strong ( r  = 0.77, p   0.0001). Here, we describe field trials to evaluate the oil content and seed weight of the new (Test), and base (Control) populations. Duplicated Tests and Controls were planted and the trials were repeated for two consecutive years. From planting to seed harvesting, no difference in growth pattern was observed in either of the populations. Test seeds had a mean oil contents of 54.01% dry weight, an average single seed weight of 0.54 g, and strong correlations between oil content and weight ( r  = 0.71 to 0.72, p   0.0001), similar to the new population obtained from the greenhouse study. Correspondingly, Control seeds had a mean oil content of 50.5% dry weight, an average single seed weight of 0.44 g, and moderate correlation between oil content and weight ( r  = 0.43 to 0.44, p   0.0001), comparable to previous data on the base population. The frequency distribution of oil content and seed weight in field- and greenhouse-grown castor seeds were similar. Our results confirm that recurrent selection for high-oil content seeds is an effective approach to increase the mean oil content of a castor population.