Identifying the habitat use in relation with time between 2 bird species in a wetland can help our understanding of how wintering bird species interact and coexist in the same system. The detection probabilities of Bean Goose (Anser fabalis) and Common Crane (Grus grus) were investigated in Yeyahu Wetland Nature Reserve in Beijing over a whole wintering season. Moreover, how use proportion of each type of habitats varied with time for 2 bird species was explored. In the detection probability study, the method of fixed-radius point was used to observe and infer the presence of species. An occupancy model was applied to deal with the occupancy of a sampling unit, accounting for imperfect detection. In the survey for habitat use, the types of habitats that 2 bird species were occupying were investigated by a line- transect method. Then, the method of multinomial logistic regression was applied to compare the proportions that various types of habitats were used during the days after 2 bird species arrived. The parameters of the detection probability model and multinomial logistic regression equation in a Bayesian framework using Markov Chain Monte Carlo (MCMC) simulations were estimated. The detection probability model was used to predict that Bean Goose arrived in Yeya Lake Wetlands in the middle of October and left at the end of March next year, whereas Common Crane arrived at the end of October and left in the middle of April next year, with almost 5 months coexisting in Yeya Lake Wetlands. The result of the multinomial logistic regression showed Bean Goose gradually decreased use proportion of the water but increased use proportion of grassland over time, and use proportion of farmland remained constant at the beginning but declined later. Common
Crane decreased use proportions of the water and grassland but increased that of farmland over time. Common Cranes were more competitive for farmland habitat. It was implied that Bean Goose and Common Crane adjusted and segregated the use proportion of grassland and farmland during their wintering period, which might mitigate interspecific competition and allow for coexistence. It was suggested that adequate habitats of grassland and farmland were probably important requirement enabling habitat segregation and coexistence of Bean Goose and Common Crane and that the overall habitat quality of a wintering site was likely to incorporate the availability of these resources, which provided multiple fitness advantages through secured food supplies and reduced predation risk. Thus, habitats providing adequate food supplies and protection might be unsuitable wintering habitats given an improper agricultural landscape management. Finally, it was proposed that maintenance of habitat diversity and improved farming methods, such as stoping crop residue burning and delaying the time of plowing, could be incorporated in agricultural landscape management plans on habitats of Bean Goose and Common Crane.
