Genetic Diversity And Mating System Of The Fragmented Populations Of The Endangered Sinocalycanthus Chinensis

Sinocalycanthus chinensis, a national secondary protection endangered species, is used as the study object here. The morphological variation of fruit and seeds and genetic variation of S. chinensis in different habitats, and their relationship with the environmental factors were analyzed. The mating system of differentc S. chinensis was analyzed using ISSR molecular markers. The fitness of S. chinensis progenies in differnet treatments supplyed different pollens were analyzed and the pollen limiation was compared. We aimed to find out the degree and mechanims of inbreeding depression of S. chinensis by combinating the morphological variation, genetic variability and pollen limitation of S. chinensis populations. The results could provide a theoretical basis for the protection and reproduction of S. chinensis. The results are shown in the followings.(1) The variation of the fruits and seeds of Sinocalycanthus chinensis in five different community types show that the phenotypic variation of the fruit and seed exist within and among populations, but the degree of variation in different populations and individuals are quite different. The phenotypic differentiation is significantly among populations. The fruits length, fruits width, fruits weigh, seeds length, seeds width, seeds weight, seeds volume and seed surface area of S. chinensis in DLS Ⅰand DLS Ⅱpopulations were significantly higher than those in the other populations. The fruit width, fruit stalk length, fruit weigh and fruit shell weight in DMS Ⅱ population were the least, while seeds weight, seeds length, seeds volume and seeds surface area were significantly lower than those in the other populations. The variation of deviation of the fruits morphological traits in DLSⅢpopulation were the largest. So did the seeds weight, seeds length, seeds volume and seeds surface area. The variation of deviation of the fruit width, fruit weight, fruit shell weight, the number of full seeds per fruit was the least. The variation of deviation of the morphological traits of seeds in DLS land DLS Ⅱ populations were the least. In total, the fruits in DLS Ⅰ and DLS Ⅱpopulations were larger and heavier, while the fruits in DMS II population was relatively lower. The phenotypic trasits of S. chinensis fruits and seeds in DLSⅢ population were unstable. The phenotypic differentiation coefficient based on the phenotypic trais of the fruits and seeds of S. chinensis were all less than0.5, indicating that most of the variation of fruits and seeds existed whitin poulations. The ANOVA analysis could get the similar results based on the morphological traits of fruits and seeds of S. chinensis populations. The genetic diversity analysis based on ISSR data showed that genetic diversity at population level was relative lower while that at species level was relative higher. The genetic diversity of DMS Ⅱ population was the highest, while that of DLSⅢwas the lowest. AMOVA anlayis and the gene differentiation coffeicient (GSI) indicated aht genetic variation existed among the populations.(2) The pollen limitation of Sinocalycanthus chinensis poulatoins were positive, indicating that pollen limitation existed but varied in different habitats. There is different pollen limitation in different populations and treatments. The pollen limitations in DLSⅣpoulationm was the highest (0.4715), while that in the DMS Ⅰ population was the smallest (0,3359). The different pollen limitation in different treatments in the same population might be due to the inbreeding depression, while the different pollen limitation in different populations might be due tho the heterogeneous of the the environment and the different size of the communities.Although the artificial pollination varied in different populations, the fitness of S. chinensis progenies with artificial pollination (except germination rate) were higher than those pollinated naturally. The fitness of cross-pollinated progenies were greater than the fitness of the self-pollinated progenies in DLS population, while there was no significant difference between treatments in DMS Ⅰ poplation. The germination rate of progenies with naturally pollination were the highest, while those of progenies with cross-pollination was the least. The variation of deviation of the seeds of progenies with naturally pollination were lower than those of progenies with artifical pollination. This indicated that the trade off between the fitness and reproductive success might benefit the reproductive success but decrease the number of seeds.The inbreeding depression index(δ) in different treatments in different populations were different, indicating that the effect of inbreeding depression on the fitess of progenies might varied at different stages and different populations.(3)The multilocus outcrossing rate (tm) of DLS Ⅱ populatoin was the highest, while that of DLS Ⅲ poulation was the lowest. The single locus outcrossing rate estimates (ts) showed the similar results. In addition, the tm and ts at poulation level were0.940±0.002and0.925±0.006, respectively. Parental inbreeding coefficient (f) of DLS Ⅲ population was the largest, while those of DLS Ⅱ and DLS Ⅰ populaiton were realtive lower. The multilocus ovule and pollen frequency (rpm) and single estimate (rps) were also varied in the five populations:rps>rpm in DMS Ⅰ population, and the pattern in the other four populations was exactly on the opposite.