Occurrence Patterns Of Bird Community And Breeding Ecology Of White-eared Night-heron (Gorsachius Magnificus) In Thousand Island Lake, China

Between 2003 and 2007, we surveyed birds on islands and mainland in Thousand Island Lake region of Chun'an County, Zhejiang Province China, to determine the effects of the habitat loss and fragmentation, the human's disturbance and the size and isolation of islands on the bird communities and species diversity, to understand the relationship between population densities of birds on mainland and their occurrences on islands, to get the occurrence patterns of birds on islands in Thousand Island Lake and to have implications of bird conservation in fragmented habitats. Meantime, we also had field studies on the breeding ecology and nest site selection of White-eared Night-heron (Gorsachius magnificus), one endangered species of the world to have more understandings of its field ecology and implications on its habitat conservation. The main results of our studies both on bird community and on White-eared Night-heron are as follows:In our study, we surveyed the birds on 20 transect lines and 4 samples on mainland and on 38 islands of three type islands (Type I: small islands without disturbance; Type II: small islands with disturbance; Type III: big islands) using transect lines method. Totally 43166 individuals and 137 species (82 residents, 28 summer visitors and 27 winter visitors), including 96 species of forest birds and 41 species of waterbirds, were recorded.The obvious differences on bird species composition and diversity between islands and mainland had been found, but their different degree varied seasonally. Species occurred on islands were mostly common species and their occurrence frequency on islands was partly related with their population size on mainland. For some species, their occurrences were also influenced by their species traits and the landscape characters of islands. The total number of bird species on several small islands was more than on one big island with the similar total area. It confirmed that several small islands had more benefits for bird survive on islands.Based on the data from surveys between 2006 and 2007, we used power function model (S=CA ) to fit the species-area relationship (SAR) birds on mainland system, consisted of 4 samples, and on 4 island systems which are Island system I with 20 islands of Type I, Island system II with 15 islands of Type II, Island system III with all 38 islands and Island system IV with 10 selected randomly islands based on area grads. In breeding season and winter, SAR could be fitted significantly for mainland system, Island system I,II and III except Island system II in winter, and couldn't get the significant fitness models of SAR for Island system IV both in breeding season and in winter. The analyses of variation pattern of Z values of SAR suggested that the models of the island systems in Thousand Island Lake supported the habitat diversity hypothesis of SAR. In Thousand Island Lake, we couldn't find Small Island Effect (SIE) of SAR models in a same spatial and temporal scale. But the species richness of birds in small island systems was more sensitive to the changes of vegetation structure caused by disturbances.Based on the data from surveys between 2003 and 2005, the three modeling approaches, random placement model, power model and stepwise Poisson nonlinear regression, were used for us to find the occurrence patterns of passerines on Type I islands, to determine the patterns of SAR, the effects of area, isolation and shape index of islands on species richness and abundance, and then to find the important factors affecting bird occurrence on small islands. The results documented that the species richness was largely determined by area and SAR could be well fitted by power models even for an instable avian community. The fitted equations for breeding season, winter and all seasons were S = 15.7714A^0.3303, S = 19.1611A^0.5830 and S = 35.7901A^0.3541 respectively. Our results suggested that the SIE was not an important feature affecting the species-area relationship of passerines on the small islands, but the feature of species combination could be an important factor.In Type I islands, we found that the species richness of permanent resident birds and understory birds (passerines) were significantly related with area and isolation, richness of small and medium size birds were significantly related with island area only, while large birds were significantly related with both island area and its isolation. Almost all species abundances were significantly and positively correlated with area, most of species abundances were significantly and negatively correlated with isolation and some of species abundances were significantly correlated with shape index, either positively or negatively. Though the abundances appeared to be more affected by area than isolation, the species abundances in breeding season appeared to be more adversely affected by isolation than in winter. Since some common species occurred on small islands in both breeding season and winter, the same species had different seasonal patterns on the relationship between abundance and landscape variables. However, all species in breeding and non-breeding season were all sensitive to area. The passerine birds in breeding season were more isolation sensitive than in non-breeding season on the small islands. It concluded that the species sensitivity to the landscape characters was related with species traits and had seasonal variations.In breeding seasons between 2004 and 2006, we totally found twenty-nine individuals (sixteen adults and thirteen juveniles) of White-eared Night-heron (Gorsachius magnificus), a highly endangered and an endemic bird species, in Thousand Island Lake, China. It was the biggest breeding population reported in China. White-eared Night-heron had a clutch size ranged from 3 to 5 eggs (n=4), egg size: 6.5±0.2 cm long and 3.5±0.1 cm wide (n=16) and its incubation period was 25.33±1.53 days(n=3).White-eared Night-heron preferred to select its nest sites with high coverage of pine trees (CPT), average height of shrubs (AHS) and short distance to nearest ford (DNF). It suggested that the shelter and food are most important factors affecting its nest site selection and should be paid more attentions to making action plan on habitat management and conservation. In order to improve the breeding success of this species in Thousand Island Lake, logging must be prohibited on islands, especially on small islands to keep higher CPT and human disturbance to the fords as the main food source should be conserved with little during breeding seasons.