Hermaphroditic perennial organisms that produce both female and male sexual organs should allocate their reproductive resource into their two sex functions, i.e., their sex allocation, such as to maximize lifetime reproductive success in the face of potential trade-offs among survival, growth, and reproduction. In this thesis, I addressed several key questions related to sex
allocation strategies and their fitness implications in wild populations of the perennial protogynous herbaceous plant, Pulsatilla alpina (Ranunculaceae).
By considering both the life history and sexual selection, I found that the species possesses a size- and time-dependent sex allocation strategy, characterizing it as both gender-diphasic and andromonoecious. Small individuals usually produce only one or few early-mating male flowers, likely maximizing their male reproductive success by capitalizing on the high mate
opportunity in the early season as a result of the species’ protogynous phenology, whereas larger individuals allocate absolutely and proportionally more to their female function.
I explicitly examined the dependency of reproductive success on sex allocation to female and male functions, a key question in the studies of the evolution of sexual systems. The results
indicate that the reproductive success through one sex depends on the sex allocation also to the opposite sex in a way that goes beyond a simple allocation trade-off. This dependency is likely the result of seed discounting or sexual interference, notably because allocation to male function increased self-fertilization and compromised female function due to strong inbreeding
depression. This result thus points to an important link between the mating system and sex allocation strategy. Furthermore, the mapping of reproductive success on a fitness landscape as a function of both female and male allocation illustrates a novel and potentially powerful way to apply key notions derived from sex-allocation theory to real populations by overcoming
difficulties associated with currency conversion between sexual functions.
Lastly, I found that the height of floral stalks in P. alpina, an ancillary trait enhancing seed dispersal by wind, is a costly trait strongly associated with female allocation. Interpretation of
the costs and benefits of allocation to stalk height further help to explain the gender-diphasic and andromonoecious sex allocation strategy in this and other similar species. Moreover, the
height of stalks was under disruptive selection by pollinators and seed predators in terms of components of female reproductive success, pointing to likely context-dependent phenotypic
selection.
Taken together, the results of my research demonstrate an unusual empirical examination of sex allocation theories incorporating life history, sexual selection, and mating systems together in P. alpina. Moreover, they provide not only adaptive explanations to the evolution of sex allocation strategies of gender-diphasy and andromonoecy but also valuable insights into the general understanding of sex allocation in hermaphroditic plants.