Abstract Elemental composition influences the dynamic balance between plant growth and community succession, thereby influencing the synergistic recovery of vegetation and soil on exposed slopes. However, existing research lacks a systematic understanding of the dynamic patterns in elemental allocation strategies and multi-element coupling mechanisms during artificial vegetation restoration. This theoretical gap constrains the transition from short-term community reconstruction to long-term functional stability. In this study, we comprehensively investigated the coupling relationships of 30 elemental indices in roots and leaves across five restoration stages (1, 3, 7, 12, and 17 years). We analysed single-element, element-element relationships, and multi-element networks to uncover patterns in elemental allocation and interactions. Restoration duration significantly affected the concentration changes and accumulation rates of N, K, and Zn in both roots and leaves, reflecting functional specialization-driven allocation strategies between organs. Notably, plant element networks showed a critical shift in hubs from root to leaf elements at 7 years, primarily driven by soil organic carbon, total nitrogen, and nitrate nitrogen. Furthermore, soil available potassium and ammonium nitrogen exhibited significantly decreased with network edge density, thereby highlighting their regulatory roles in elemental coordination. In short, restoration age shapes root-leaf elemental accumulation and allocation strategies, while the interaction network architecture between root and leaf elements is predominantly governed by soil nutrient availability. Understanding multi-element coupling mechanisms helps identify stage-specific limiting factors, optimize phased vegetation management, and accelerate functional recovery in degraded ecosystems. Integrating plant elemental dynamics into restoration planning can enhance long-term vegetation stability and resilience, supporting sustainable land rehabilitation.
Geng et al. (Wed,) studied this question.