Exploring the variation and coordination of functional leaf traits across small-scale natural environmental gradients contributes to our understanding of plant adaptation strategies in response to localized desert environmental changes. This study uses Nitraria sphaerocarpa as case, a species widespread in the desert-oasis transition zone of the Hexi Corridor, to analyze the characteristics of leaf functional traits at different stages of sand dune stabilization and their responses to soil factors. The results indicate: (1) There are significant differences in specific leaf area (SLA) and leaf nitrogen content (LNC) of Nitraria sphaerocarpa across stabilized, semi-stabilized, and mobile sand dunes, in the order of stabilized>semi-stabilized>mobile. The tissue density (TD) and leaf dry matter content (LDMC) of Nitraria sphaerocarpa on stabilized sand dunes are significantly lower than those on mobile sand dunes (P < 0.05). (2) The degree of fleshiness (DOF) of Nitraria sphaerocarpa on stabilized, semi-stabilized, and mobile sand dunes is negatively correlated with LDMC, and SLA is negatively correlated with TD. Additionally, on stabilized and mobile sand dunes, DOF is negatively correlated with TD and SLA is negatively correlated with LDMC, while DOF is positively correlated with SLA and LDMC is positively correlated with TD (P < 0.05). (3) The Nitraria sphaerocarpa on stabilized sand dunes has leaves with higher SLA and LNC, while those on mobile sand dunes have higher TD and LDMC. Leaves on semi-stabilized sand dunes exhibit trait variations in between. (4) The soil carbon content, soil nitrogen content, soil phosphorus content, soil compaction, and soil moisture collectively explain 80.76% of the variation in leaf functional traits. This study demonstrates that environmental variations determine the allocation strategies of Nitraria sphaerocarpa, with those on stabilized sand dunes adopting a "fast investment-return" strategy, and those on mobile sand dunes a "slow investment-return" conservative strategy. The results provide insights into assessing the resource allocation strategies of desert plants at the intraspecific trait level.