VARIATION IN SEED CHARACTERS IN NEMOPHILA-MENZIESII - EVIDENCE OF A GENETIC-BASIS FOR MATERNAL EFFECT

A growing body of evidence indicates that phenotypic selection on juve nile traits of both plants and animals may be considerable. Because ju venile traits are typically subject to maternal effects and often have low heritabilities, adaptive responses to natural selection on these traits may seem unlikely. To determine the potential for evolutionary response to selection on juvenile traits of Nemophila menziesii (Hydro phyllaceae), we conducted two quantitative genetic studies. A reciproc al factorial cross, involving 16 parents and 1960 progeny, demonstrate d a significant maternal component of variance in seed mass and additi ve genetic component of variance in germination time. This experiment also suggested that interaction between parents, though small, provide s highly significant contributions to the variance of both traits. Suc h a parental interaction could arise by diverse mechanisms, including dependence of nuclear gene expression on cytoplasmic genotype, but the design of this experiment could not distinguish this from other possi ble causes, such as effects on progeny phenotype of interaction betwee n the environmental conditions of both parents. The second experiment, spanning three generations with over 11,000 observations, was designe d for investigation of the additive genetic variance in maternal effec t, assessment of paternal effects, as well as further partitioning of the parental interaction identified in the reciprocal factorial experi ment. It yielded no consistent evidence of paternal effects on seed ma ss, nor of parental interactions. Our inference of such interaction ef fects from the first experiment was evidently an artifact of failing t o account for the substantial variance among fruits within crosses. Th e maternal effect was found to have a large additive genetic component , accounting for at least 20% of the variation in individual seed mass . This result suggests that there is appreciable potential for respons e to selection on seed mass through evolution of the maternal effect. We discuss aspects that may nevertheless limit response to individual selection on seed mass, including trade-offs between the size of indiv idual seeds and germination time and between the number of seeds a mat ernal plant can mature and their mean size.