Coastal wetland is one of the important avian habitats, and the wetland habitat changes caused by human activities have had a certain influence on bird community and indicator species in different habitats. Reed swamp, farmland, fishpond, and tidal flat were formed as four types of disturbed habitats which resulted from different human activities in Xinyanggang estuary of Yancheng Nature Reserve. According to a thorough field study, we sampled 3 three-kilometer-long lines in each habitat to investigate the species and numbers of wintering birds. Then we utilized Multi-response Permutation Procedures (MRPP)、Indicator Species Analysis (ISA)、Two-way Cluster Analysis and Detrended Correspondence Analysis (DCA) to analyze the indicator species, bird community structure and gradient change of wintering birds in different disturbed habitats. Our results showed that: 1) A total of 67 bird species were recorded in these four types of habitat, and fishpond had the maximum amount among different habitats, which was counted out 37 species. The D' diversity index of tidal flat was the highest in the distinct duration in different habitats respectively, as well as the H' diversity index of reed swamp. Compared with other habitats, the evenness index of reed swamp and tidal flat was also higher. The differences between habitats are related to the composition of bird community. 2) According to MRPP analysis, there was no significant difference of bird community structure between reed swamp and farmland (P > 0.05), while it was the opposite between other two habitats of the rest combinations. Grus grus could serve as the indicator species in farmland; The indicator species in fishpond consist of Larus crassirostris、Larus ridibundus、Larus vegae and Ardea cinerea; Meanwhile, the indicator species in tidal flat were made up of Calidris alpina、Pluvialis squatarola and Numenius arquata; The last but not the least, Gallinula chloropus served as the indicator species in reed swamp. 3) Two-way Cluster Analysis classified 12 observation conditions into 4 groups, and it was highly consistent with different types of habitats. Birds could be clustered into 7 groups on the 40% of the rest information. Further more, birds could be divided into three groups on axis 1 of the DCA analysis using 12 observation conditions as variables: birds adapting the open water, birds depending on the aquatic vegetation and birds depending on the terrestrial vegetation. The results of our study showed that bird diversity could be influenced by habitats heterogeneity, human disturbance and tidal water level; and different habitats had obvious indicator species respectively; in order to adapt to the changes in habitats, bird community structure appeared a certain gradient change. In summary, we should strengthen the protection of reed swamp and other original habitats on which wintering birds depend much, and then take the impact of human activities on the diversity of birds into consideration when we reclaim farmland and fishpond as well as introduce Spartina alterniflora into naked flat.