The objective of this research was to test selection methods in order to increase female flowers percentage of castor bean plants (Ricinus communis L.) from population FCA-UNESP-PB. Selection of plants was done in Botucatu, State of São Paulo, during the out-of-season crop of 2007. Plants with primary raceme strictly female were selected by the mass selection and plants that presented sex reversion were selfed. Evaluations were taken in out-of-season crop of 2008 in Botucatu and São Manuel, State of São Paulo, when the following treatments were compared: mass selection method; mass selection method with selfing; and control (racemes of plants collected at random, without selection). The traits evaluated were female flowers percentage on primary raceme (%), grain yield (kg ha-1) and seeds oil content (%). The experimental design was in randomized block design with 30 replications. The results were analyzed using analysis of variance of each individual environment and both environments together, by the F test at 1% probability level. From the results obtained, it is concluded that mass selection method with selfing presented the highest female flowers percentage on primary raceme, with 18% of phenotypic gain in Botucatu and 29% in São Manuel, State of São Paulo. The selection methods presented distinguished behavior comparing the two environments for grain yield, with mass selection with selfing presenting the lowest yield. The oil content was not affected by selection methods and environments.
Ricinus communis L.; mass selection; selfing; sexual expression
