Habitat loss and shift of the population centroid will make migration process more difficult. The habitat loss also deteriorates the connectivity between adaptation areas (Finch et al., 2017). The migratory birds have less food on the flyways and have to fly longer distance to achieve the migration. These conditions do not only make the migration process harder, but also retains many birds from completing the process. Migratory species require appropriate survival conditions throughout their whole migration process, including the breeding process, the wintering process and the migration between the 2 process (Rushing et al., 2016). Climate change may affect the 3 process all and lead species to spend more energy and time to migrate. With the increasing of distance and the unstable of food, the difficulty of survival for migratory species will increase (Finch et al., 2017; Rushing et al., 2016; Saino et al., 2011). Especially for the long-distance migration, the risk of death will increase a lot (Bauer & Hoye, 2014; Kirby et al., 2008; Yong et al., 2018).

The importance of environment variables

The percentage contribution of different variables can be seen by the jackknife test in the MaxEnt model. The determination of the leading factor for the migratory birds is derived from the comparison of the contribution rate of the environmental variables involved in the model.
The result of estimating the importance of variables showed that\(t_{\min}\) was the most crucial environmental variable for distribution of migratory birds, accounting for 23.02% importance. The LUCC, alt, bio1 and NDVI also made massive contributions to the model, accounting for 15.2%,18.64%,12.97% and 12.22% importance respectively. The sum of the 5 essential variables accounted for 82.05% of the cumulative contribution rate. The individual contribution of these variables can be seen in Fig 3a.