Relative Energetic Economy of Cleistogamous Selfing in Three Populations of the Perennial Ruellia humilis

Premise of research. The maintenance of mixed selfing and outcrossing is an enduring mystery in evolutionary biology. Cleistogamy, in which individuals produce both potentially outcrossing chasmogamous (CH) and obligately selfing cleistogamous (CL) flowers, provides an ideal framework in which to study the evolutionary forces maintaining mixed mating. Despite their promise for providing insight into mating system evolution, few studies of cleistogamous species have comprehensively investigated the relative costs and benefits of CL selfing.Methodology. We quantified the necessary components to calculate the relative energetic cost of reproduction of each flower type-CH and CL-in a greenhouse study using three populations of the perennial Ruellia humilis Nutt (Acanthaceae). These components included flower dry mass, fertility (i.e., fruit set), seed mass per fruit, and number of seeds per fruit. We also measured pollen-ovule ratios for both flower types as another measure of relative energetic investment. Last, we tracked phenology of the two flower types and used the proportion of CH flowers to estimate maximum potential outcrossing rates.Pivotal results. We found that the energetic cost of reproduction via CL flowers was four to 10 times lower than that of reproduction via CH flowers. This energetic economy derived both from the reduced mass and from the increased fertility of CL flowers. Pollen-ovule ratios in CL flowers were one-third to one-half those of CH flowers, providing additional support for their energetic economy. Maximum potential outcrossing rates in these populations based on CH flower production were between 43% and 61%, but CH flowers can autogamously self at rates between 33% and 75%.Conclusions. These results suggest that selection should favor only CL flowers due to their greater energetic economy. Additional work in this species demonstrates that neither inbreeding depression nor heterosis can provide an advantage strong enough to overcome the energetic benefits of CL reproduction. Further research on the developmental genetics of the two flower types and potential genetic constraints on the loss of CH flowers is needed.