Size-dependent Sex Allocation In A Hermaphrodite Herb With 1-day Flowers, Hosta Ventricosa (Liliaceae)

Sex allocation is one of the most important and success theories in the area of evolution-ecology, it has been used to explain the evolution of breeding systems, sex ratio in population and the formation of selfing. Presently, most studies have focused on the four levels of species, population, individual, inflorescences. The degree of selfing can affect sex allocation of species and populations. Within population, if fitness gain through male and female functions is differently influenced by size or status, it is expected that organisms will adjust sex allocation according to size; Furthermore, sex allocation often varies temporally and spatially among flowers within inflorescences. Ovule or pistil production per flower generally declines from basal(or early-opening)to distal(or late-opening) flowers, whereas pollen or stamen production per flower often remains constant or declines at a lesser degree.Sex allocation theory predicts that there is trade-off between resource allocation to male vs female function or to size and number at multiple levels. For example, ovule production vs pollen production, seed size vs number. However, the size is different among seeds.We examined the sex allocation among plants and flowers within inflorescences in a perennial hermaphrodite species with 1-day flowers, hosta ventricosa (Liliaceae). Flowering progresses sequentially from basal to distal positions within an inflorescence with little flowering overlap among flowers. The results are as follows:(1) There is not a threshold of reproductive in hosta ventricosa.(2) At plant level, the sex allocation is not correlated with plant size significantly, maybe due to the absence of geitonogamy.(3) Within inflorescences ovules per flower remained constant, regardless of floral position. Pollen production per flower, the pollen to ovule ratio of individual flowers and seed to flower ratio declined from basal to distal position.(4) There is no trade-off between ovule production and pollen production per flower at the individual level.(5) No relation between seed size and number at individual level was detected, both negative and positive or neutral relations between seed size and number per fruit within individual plant was detected, there is the weak significant negative relation between seed size and number at the individual fruit level.(6) Variable coefficient of the seed size is 19.1%; seed-size variation is mainly based onintra-individual, inter-individual followed.