In order to understand the effects of water depth and food resources on the habitat use of waterbirds, we surveyed individual numbers, water depth of foraging and resting sites, food availability (density and biomass of winter buds of submerged macrophytes for bud-feeders) of four waterbird species (Grus leucogeranus, Cygnus columbianus, Ciconia boyciana and Platalea leucorodia) using plot methods in winter 2012 ~ 2013 in Shahu Lake, a sub-lake of Poyang Lake. A total of 152 plots were set in the lake with each plot being 150 m ? 150 m. The buds were collected in October before arrival of waterbirds and waterbirds were counted in January when large amounts of waterbirds concentrated in the lake. Distribution patterns of winter buds in different water depth in October were analyzed using Kolmogorov-Smirnov Z test. We compared water depth between foraging and resting sites of the 4 species using Mann-Whitney U test, respectively. We further compared water depth, density and biomass of winter buds in foraging sites and water depth in resting sites between Grus leucogeranus and Cygnus columbianus (both mainly forage on buds) using Mann-Whitney U test, respectively. We also compared water depth between two piscivorous waterbird species in foraging sites and resting sites, respectively. Regression analysis were used to assess the relationship between foraging bird numbers and water depth for the 4 species, as well as between foraging bird numbers and biomass of winter buds for 2 bud-feeding species. Results indicated that winter buds distributed in an average water depth of (124.2 ± 12.0) cm in October (Fig. 2). Water depth at foraging sites of all the four species were deeper than that at resting sites (Grus leucogeranus: Z = 11.96, Cygnus columbianus: Z = 4.69, Ciconia boyciana: Z = 14.44, Platalea leucorodia: Z = 29.33, P < 0.01 for all) (Table 1). Density of winter buds at foraging sites was not significantly different (Z = 0.6, P = 0.55). For the two bud-feeding waterbird species, Grus leucogeranus roosted in shallower water and foraged in shallower water with lower biomass of winter buds than that of Cygnus columbianus, and the former had shallower foraging depth that the latter (water depth at foraging sites: Z = 8.56; biomass of winter buds: Z = 2.93, foraging depth: Z = 14.69, water depth at resting sites: Z = 4.34, P < 0.05 for all). For the two piscivorous waterbird species, foraging depth, water depth at foraging sites and water depth at resting sites of Ciconia boyciana were deeper than that of Platalea leucorodia (water depth at foraging sites: Z = 10.60, foraging depth: Z = 9.35, water depth at resting sites: Z = 8.47, P < 0.01 for all). Regression analysis indicated significant quadratic relationships between the individual numbers of foraging Grus leucogeranus (R2 = 0.39, P < 0.05), Ciconia boyciana (R2 = 0.31, P < 0.05), Platalea leucorodia (R2 = 0.29, P < 0.05) and water depth at study site (Fig. 3). The water depth with the highest densities of the three species were 23.9, 33.0, and 22.6 cm, respectively. Individual number of foraging Grus leucogeranus (R2 = 0.43, P < 0.01) and Cygnus columbianus (R2 = 0.54, P < 0.05) increased with the increase of biomass of winter buds. Three wader species distributed only within a certain range of water depth and species with same diets distributed at different water depth to reduce overlap of spatial niche.