Effects Of Habitat Fragmentation On Diet Composition,Genetic Structure And Sex-biased Dispersal In Piebald Odorous Frog

Habitat fragmentation is a process of ever continuous and big natural habitat changed into isolated and small patches. Habitat fragmentation was considered as one of the most important factors which threated biodiversity severely. As a type of fragmentation, land-bridge islands were treated as natural eco-lab in fragmentation researches. In this study, we analysed two factors which affected the life history traits (diet composition and sex-biased dispersal) of piebald odorous frog (Odorrana schmackeri) and genetic diversity, population structure in Thousand Island Lake, and hope understand the effects of habitat fragmented by water matrix on O.schmackeri. The results of this paper were below:1. We analyzed the diet composition of the piebald odorous frog in Thousand Island Lake, and the influences of landscape-level features on prey categories. Results demonstrate that this species has a wide prey categories (16 categories) and dietary niche (6.92). The diet composition highly overlapped between sexes and among islands, however consumption of caterpillar was differed significantly between sexes as females much larger than males and could eat larger preys. Further, the consumption of Coleoptera, Arachnida and plant seed differed significantly among islands. Different prey has own fragmented effects as different sensitivety, such as patch area affected Hymenoptera and isolation influenced Arachnida.2. Twelve highly polymorphic microsatellite loci were isolated and developed. And ten loci of them were used in 451 frog species from 17 island populations at Thousand Island Lake to analyze the Chinese piebald odorous frog's genetic diversity and genetic structure. The results showed that the genetic diversity of O.schmackeri still high as the mean allele richness and expected heterozygosity were high. Second, the populations experienced minor or no differentiation genetic differentiation as the FST value in most paired population less than 0.05. And the AMOVA illustrated the main variation was occurred in intra-population. Historically, the population size severe declined as beginning save water, but recently there was no bottleneck detected. Population genetic clusters were calculated by two methods and displayed no significant genetic clusters as there were much more gene flows among populations. Last but not the least, the genetic diversity was not influenced by patch characters (island area and island isolation) and there were no significant relationships between island distance and genetic distance.3. Ten microsatellite loci and 382 samples from 14 island populations were used to analyze the pattern of sex-biased dispersal. The results showed there was non-significant male-biased dispersal in this species. But significant female-biased dispersal, non-significant female-biased or male-biased dispersal were found in some populations. Using partial correlation test to detect the relationships between patch characters (island area and island isolation) and the pattern of sex-biased dispersal, the results found the distance which was near in iso-main or iso-big was negative correlated with the pattern after controlling the area effects. That was means that the island which far from the mainland or big island would occur male-biased dispersal, however the island which near from the mainland or big island would female-biased dispersal. The three parameters (sex ratio, breeding sites and inbreeding coefficient) were used to evaluate the three alternative hypotheses and the results illustrated breeding sites that was means avoid resource competition was the best candidate for explaining sex-biased dispersal of O.schmackeri in fragmented Thousand Island Lake.