The Variation Pattern Of Floral Sexual Durations And Its Adaptive Value In Protandrous Aconitum Gymnandrum

The sexual phases determines the durations of flowering times and it plays an important role in the process and success of dichogamous plant reproduction. The female and male durations appear sequentially, the environmental factors(such as temperature and precipitation) and plant characteristics(eg. flower attraction features and flowering time, etc.) will affect the durations of male and female phases in dichogamous plants. The flowering time plays an important role in the mating opportunities of male and female functions and final reproductive success in plants. The mating environment hypothesis predicts that the differences of flowering time in protandrous species can change individual’s phenotypic gender and the mating environment within population, finally affect the optimal allocation of resources to sexual functions. To study the function of the male and female phases and the variation patterns of phenotypic gender with flowering time in dichogamous plants, and the impact of this change on the success of female reproduction, we observed the natural and artificial protandrous Aconitum gymnandrum populations(ranunculaceae), analyzed the variation pattern of male and female phases of flowers within populations and inter-individual phenotypic gender, and the correlation of the variation and flower features and female reproductive success, and discuss the fitness significance of the variation of sexual phases through phenotypic selection analysis.The results show that the environmental temperature is significantly related to the male and female durations, but the correlation is not consistent even opposite among populations. It indicated that there are other environmental factors also influence male and female durations. The single floral lifespan is longer with flowering time delayed in the flower. At the individual level, the positive correlation between first flowering time and male duration, the negative correlation between first flowering time and flower length in two populations of the same year. The first flowering time is significantly negative correlation to female phase, individuals of flowering earlier have a longer female duration, but this variation is different in three populations. And the negative correlation of male durations and flower length is significant, the larger flowers have a shorter male duration.It showed a shift from male-biased to female-biased gender during flowering season in A. gymnandrum population, because most of individuals had only male-phase flowers at the beginning of flowering time and only female-phase flowers at the end. Therefore, mean phenotypic gender of individuals shifted from femaleness to maleness with flowering time. Our results support the mating environment hypothesis, i.e. male-biased floral sexual ratio(mating environment) early in protandrous A. gymnandrum population leads to female-biased phenotypic gender of individuals flowered early and thus female-biased temporal sex allocation in early-flowering individuals and early flowers within inflorescences in comparison with the late-flowering individuals and late flowers.The significant linear selection differential and gradients on floral traits were estimated respectively through female fitness which was represented by pollen number and seed number respectively were different in the different populations and estimated under the different female fitness components. Flowering earlier is always subjected to selection, The individuals with earlier flowering time product more seeds. The individuals of earlier flowering and shorter male duration have higher female fitness, that is early-flowering individuals can shorten the male duration. The selection differential of male duration is negative and the selection differential of female duration showed opposite trends in different populations. Generally, more flowers and larger flower(petal length is higher) are advantageous to improve the individual female fitness, these selections were consistent in different populations. But nonlinear selection analysis showed that there were significant correlation between sexual durations and floral traits. The results also showed that individuals with the earlier flowering time and the higher petal length have a bigger female fitness.