(Background and objective) Urbanization drives widespread land use change and poses significant challenges to urban biodiversity conservation. As crucial environmental indicative species, butterflies are widely used to monitor urban habitat quality. Understanding how butterfly taxa respond to environmental changes at the landscape scale is essential for assessing urban ecological health and guiding biodiversity management. Despite this importance, the specific mechanisms governing taxon-specific responses in rapidly urbanizing areas remain poorly understood. (Method) This study employed the Pollard walk method to systematically survey butterfly communities across 8 urban green spaces in Fuzhou, China. Environmental variables were quantified through maximum likelihood classification (MLC) of high-resolution (1?m) remote sensing imagery from the Gaofen-2 (GF-2) satellite. Key environmental drivers were identified using Spearman’s rank correlation analysis and multiple linear regression (MLR), with model selection based on the corrected Akaike Information Criterion (AICc). Differences in butterfly community composition among habitat types were analyzed using principal coordinate analysis (PCoA) and permutational multivariate analysis of variance (PERMANOVA). (Results) The results indicated that: 1) Butterfly diversity was positively correlated with surrounding woodland coverage and the presence of large adjacent mountain patches (P<0.05, R2:0.766), with mountain patches exerting a stronger effect (Coefficient: 0.123>0.023); 2) Landscape heterogeneity significantly increased both species richness and abundance (Richness: P<0.05, R2:0.857; Abundance: P<0.05, R2:0.923), though abundance responses were more pronounced; 3) Grass coverage significantly affected Pieridae richness (P<0.05, R2:0.934) and abundance (P<0.05, R2:0.918), while Nymphalidae and Lycaenidae responded more strongly to landscape heterogeneity; the proportion of grassland and water significantly enhanced Papilionidae diversity (P<0.05, R2:0.544), richness (P<0.05, R2:0.703), and abundance (P<0.05, R2:0.710); 4) Community structure differed markedly between natural habitats (habitat dominated by wood land/ habitat dominated by grass land/ habitat dominated by water body) and artificial surfaces habitats (habitat dominated by sealed surfaces/ habitat dominated by bare land); 5) Nymphalidae and Papilionidae exhibited distinct but overlapping habitat preferences, whereas Pieridae displayed broader niche adaptability. (Conclusion) This study highlighted that butterfly responses to environmental factors varied significantly across taxa, reflecting species-specific habitat preferences, behavioral adaptations, and dispersal capacities. These results highlight the necessity of incorporating taxon-specific ecological knowledge into urban biodiversity conservation and planning. Strategies focusing on enriching the vegetation structure surrounding urban green spaces and maintaining ecological connectivity with adjacent mountain or high-quality habitat patches, as well as conserving natural landscape elements such as wood lands, grass lands, and water bodies within the green spaces, are critical for sustaining butterfly diversity and improving urban ecological quality.