Aim: Off-reserve conservation is a major contributor to China biodiversity conservation efforts, biodiversity conservation being achieved within afforestation and low-intensity agriculture in fragmented landscapes. Functional trait is more strongly related to ecological processes than taxonomic diversity and reflects ecosystem functioning and species responses to environmental changes. In this study, we selected five habitats that differ in degree of disturbance to explore the effects of land use on ant community compositions, traits distributions and functional diversity change. We assessed how habitat disturbance affects the ant community compositions and traits distributions and asked if ant functional diversity respond to disturbance at the local scale? Location: Lüchun County, Yunnan Province, southwest China. Methods: Pitfall traps were used to survey ant communities. Additionally, we measured four ant morphological traits (eyes diameter, distance between eyes, femur length of the hind-leg and Weber’s length) to assess the functional traits distributions and functional diversity. Shade plot of ant relative abundance was used to explore species distribution amongst different habitats. Kernel density plot was used to explore ant traits distribution patterns amongst different habitats. Non-metric multi-dimensional scaling ordination, based on ant Weber's length, was used to explore the ant traits compositions amongst different habitats. The fourth corner model was used to evaluate the association between ant traits and environmental variables. The FRic, RaoQ and FEve indices were selected as three complementary measures of the multivariate functional traits space and functional redundancy of different habitats. Results: We collected 14258 ants, representing 89 species, 40 genera and seven subfamilies. Aphaenogaster schurri and Tetramorium ciliatum were the common species of secondary forest; P. sagei, P. pieli, Cardiocondyla wroughtonii, Recurvidris nuwa, Tapinnoma melanocephalum, Monomorium pharaonis and M. orientale were the common species in plantations; and Iridomyrmex anceps and Cardiocondyla nuda were the common species in managed farms. Ants had medium eye diameters, narrow distances between eyes, medium leg lengths and smaller body sizes in greatly-disturbed habitats; and ants had an increasing eye diameter and narrowing of the space between eyes, while the leg length and Weber’s length became shorter in moderately-disturbed habitats. Ant trait composition, based on Weber’s length, showed significantly differences amongst five habitats. The fourth corner analysis indicated that ant species traits were significantly correlated with environmental variables. The functional diversity of secondary forest, lac plantation and lac plantation-corn agroforest were higher than those in dryland farm and rice paddy. Functional diversities were significantly negatively correlated with bare ground cover and significantly positively correlated with leaf-litter cover, leaf-litter thickness and plant cover. Main conclusion: Our results indicated that ant traits distribution patterns were affected by land-use changes, followed by anthropogenic disturbance pressures at the local scale. Ant traits compositions in greatly-disturbed habitats also differed from the habitats with less disturbance. It is unfavourable for the survival of the large body-size ants in more open habitats with more anthropogenic disturbance. Compared with secondary forest, dryland farm and rice paddies were less resistant and more vulnerable and lac plantations had approximately functional diversity of ant communities, suggesting that lac plantations might be resistant as secondary forest to species loss.