Objective In view of the water-limited habitat characteristics in the dry-warm valley of the Lancang River, this study quantitatively analyzed the photosynthetic physiological responses of Sophora davidii across soil moisture gradients and determined its limitation thresholds. Based on these findings, a soil moisture classification system based on photosynthetic productivity was established to provide a theoretical basis for precise water management in regional vegetation restoration.

Method Using a natural drying protocol, five relative soil water content (RSWC) levels (95%, 80%, 65%, 50%, and 35%) were established. Physiological homeostasis was induced via gravimetric moisture control. Leaf gas exchange and light-response curves were quantified using a Li-6400XT system. Nonlinear models were employed to characterize the response of photosynthetic parameters to soil moisture. A soil moisture classification system based on photosynthetic productivity was constructed using key physiological inflection points.

Result (1) With the intensification of soil water deficit (RSWC ≤ 80%), the maximum net photosynthetic rate (P_nmax) and apparent quantum yield (AQY) of Sophora davidii generally declined, with pronounced reductions under lower RSWC conditions. The light compensation point (LCP) shifted from 27.95 to 38.88 µmol/(m²·s), while the light saturation point (LSP) decreased from 1 423.72 to 1 159.56 µmol/(m²·s), substantially narrowing the effective light-utilization range. At 50% and 35% RSWC, pronounced photoinhibition was observed when photosynthetically active radiation (PAR) exceeded 1 200 µmol/(m²·s). (2) Both net photosynthetic rate (P_n) and stomatal conductance (G_s) exhibited a unimodal response to declining soil moisture, peaking at 80% RSWC. The transition from stomatal to non-stomatal limitation in Sophora davidii may occur within the RSWC range of 35%–50%. As RSWC decreased from 80% to 50%, the decline in P_n was primarily driven by stomatal factors. As drought intensified and RSWC fell toward 35%, the intercellular CO₂ concentration (C_i) shifted from decreasing to increasing, while the stomatal limitation value (L_s) shifted from increasing to decreasing. At 35% RSWC, C_i showed an upward trend, whereas L_s declined significantly, suggesting an increasing contribution of non-stomatal limitation with increasing drought severity. (3) Theoretical optima for P_n and water use efficiency (WUE) were identified at 75.6% and 61.8% RSWC, respectively. When RSWC decreased from 80% to 65%, the transpiration rate (T_r) decreased by 26.1%, whereas P_n decreased by only 1.1%, indicating different response magnitudes of P_n and T_r to decreasing soil moisture. The model-extrapolated water compensation point for net photosynthesis was 26.3% RSWC. Based on key indicators such as theoretical optima, the water compensation point for net photosynthesis, and the transition range of photosynthetic limitation, soil moisture availability and photosynthetic productivity were categorized into five hierarchical grades: non-productive and ineffective water, low-productivity and low-efficiency water, moderate-productivity and high-efficiency water, high-productivity and stable-efficiency water, and surplus and low-efficiency water.

Conclusion S. davidii maintained relatively high photosynthetic capacity and water-use efficiency under moderate soil water deficit, indicating its physiological adaptability to the water conditions of dry-warm valley habitats. We propose 50% RSWC as a reference early-warning value for supplemental irrigation, whereas the model-extrapolated value of 26.3% RSWC may serve as a reference physiological threshold under severe soil water deficit, and its ecological applicability requires further experimental validation. This quantitative framework provides a robust physiological metric for assessing the stability of restored vegetation and directly serves the overarching objective of implementing precision irrigation and water management in these fragile ecosystems.