Abstract: There is a common phenomenon that a discrepancy in female productive success exists in the inflorescence of many plants, especially alpine plants. The non-uniform pollination hypothesis, the architecture effect hypothesis, and the resource competition hypothesis are three hypotheses to explain such patterns of within-inflorescence variation in female reproductive success. However, experiments to evaluate the influence of outside environmental factors on female reproductive success are rare. In this experiment, the effects of population density and defoliation on the female productive success of P. semitorta were investigated. The parameters tested include seed number, seed mass, ovule number per fruit, and individual seed weight. We also choose twenty natural groups of P. semitorta to find whether this plant’s reproduction allocation is restricted by pollen resources. There was no significant discrepancy in the rate of fruit set and the rate of seed set between natural pollinated and artificial supplement pollinated plants. However, population density had a significant effect on seed number per fruit and individual seed weight, while defoliation affected seed numbers, seed mass, and ovule number per fruit. In density and defoliation treatments, fruit position had a marked relationship to all the above parameters, indicating that architecture effects play a role in this plant’s female reproductive success. All the parameters of defoliated individual plants were less than those of the non-defoliated ones. The interaction effect of defoliation and fruit position had a significant influence on seed numbers and ovules per fruit, indicating that defoliation had a negative effect on female reproductive success. We suggest that the architecture effect and resource competition both affect female reproductive success of P. semitorta and that they are the results of plant plasticity.