Gradient Effects of Vegetation Cover and Carbon Sequestration in Highway Corridors: A Case Study of Shandong Province, China

Highway corridors are increasingly being discussed not only as zones of ecological disturbance but also as components of regional green infrastructure with potential carbon sequestration functions, yet their long-term evolutionary characteristics and multi-scale associated factors remain insufficiently understood. Using multi-source time-series data from 2000 to 2023, we developed an analytical framework integrating the CASA model, Random Forest, and geographically weighted regression (GWR). To ensure methodological rigor, we implemented a Spatial K-fold Cross-Validation strategy and incorporated Partial Dependence Analysis (PDA) to identify non-linear thresholds. The results indicate that: (1) Vegetation carbon sequestration within Shandong’s highway corridors increased significantly, with total sequestration rising from 5.54 × 10 6 t in 2000 to 1.55 × 10 7 t in 2023, representing an average annual growth rate of approximately 5.0%. This growth transitioned from a relatively stable phase to a more rapid growth phase. (2) A clear distance-related ecological pattern was observed. Statistical tests (Kruskal–Wallis H test) confirmed that vegetation carbon sequestration exhibited a significant non-monotonic gradient ( p < 0.05 ), with a stable peak zone observed 50–100 m from the roadbed. This peak zone is associated with a spatial “trade-off” pattern between the attenuation of traffic-related stressors and roadside ecological management. (3) The observed spatial pattern was associated with a nonlinear coupling of natural background conditions and human disturbance. Precipitation and temperature were the dominant associated factors, while PDA further identified a critical precipitation threshold (~750 mm) and localized tipping points for human interference, with a distinct road-disturbance-sensitive zone evident within 200–500 m. The results suggest that high-standard ecological design and active restoration measures are associated with lower ecological disturbance and higher vegetation carbon sequestration performance in some highway corridors. However, these relationships should be interpreted cautiously, as they may also be influenced by differences in climate background, topography, land-use context, and road construction history. These findings provide empirical evidence to inform differentiated ecological restoration and low-carbon management of traffic corridors.