Identification Of Factors Influencing Habitat Selection Of Red-crowned Crane In Yancheng Nature Reserve

Knowledge of key habitat factors affecting the habitat secleciton of endangered species is the first step in habitat conservation and restoration. Howerver, scale dependence, spatial autocorrelation, and multicollinearity of habitat factors hinder the correctness of identification of these key factors. In order to solve these three problems, this paper considers two scales(microhabitat and landscape), adds spatial variables from cubic trend-surface analysis, and uses two statistical techniques(variation partitioning & hierarchical partitioning) to identify the key factors affecting the occurrence of red-crowned crane in Yancheng Nature Reserve. The main results are as follows:(1) Landscape is the key scale limiting the distribution of red-crowned crane. Landscape variables explains 52.9% variation in bird presence/absence data. Its joint explained variation with microhabitat and spatial variables is much larger than its pure explained variation.(2) Vegetation factor in landscape scale is the most significant habitat factor influencing red-crowned crane occurrence. Its explained variation occupies 39.9% of total variation in bird data. Disturbance factor and shelter factor explain 17.4% and 8.6% variation, respectively.(3) Among all the explanatory variables, Seepweed area percentage within 500 hm2 sampled circle, water area percentage within 50 hm2 sampled circle, and shelter distance from sampled points are most crucial environmental variables cotrolling the spatial pattern of red-crowed crane. These three variables explain 4.12%, 3.34%, and 3.04% bird distribution variation, respectively.(4) Spatial structure of red-crowned crane explains 33.7% bird presence/absence variation, of which 25.0% is the joint explained variation of spatial variables and landscape variables.Several conclusions and conservation measures are put forward based on the results. Firstly, spatial structue and resource configuration at landscape scale, especially vegetation type significantly affects the presence/absence of red-crowned crane. Among these vegetation type, Seepweed is the favorate habitat type of red-crowned crane due to its abundant food source and suitable rest sites, whereas high coverage and height decrease cordgrass’ s suitability for red-crowned crane. Therfore, restriction of seepweed patch reclamation and control of cordgrass invation are recommended. Secondly, the influence of human disturbance on the habitat seclection of red-crowned crane is relatively small, but it has positive effect on bird occurrence. This indicates red-crowned crane tend to select the area with human activity. The reason for this may be the deterioration of natural habitat quality, especially the lack of food source. Thirdly, shelter is also an non-ignorable factor influencing the distribution of red-crowned crane and thus marginal reed should be retained during the havest time. Forthly, water area within bird’s habitat was seriously insufficient and therefore became one significant variable limiting the spatial distribution of red-crowned crane. Eco-hydraulic engineering is in urgent need to increase the accessibility of water for the birds.In addition to these consevational recommendations, our results also show the spatial structure of bird and landscape variables are highly correlated. Hence the spatial pattern of landscape factors can reflect the habitats birds like or dislike. Finally, the results of variation partitioning and hierarchical partitioning are well consistent, and hierarchical partitioning can also examine whether the variables removed in stepwise selection affect the occurrence of red-crowned crane.