Data of graded restoration induces asynchronous recovery of soil multifunctionality and biodiversity in degraded alpine meadows

The restoration of degraded alpine meadows on the Qinghai-Tibet Plateau (QTP) is a critical global priority. However, current restoration strategies often prioritize severely degraded areas, potentially neglecting more cost-effective interventions in moderately degraded systems and creating mismatches in the recovery of different ecosystem components. This study tests a graded management framework designed to align intervention intensity with degradation level, assessing its efficacy in restoring both ecosystem structure and function, with a particular focus on soil processes. We established a restoration experiment in alpine meadows across a degradation gradient: light (LD, fencing only), moderate (MD, fencing + reseeding + fertilization), and severe (SD, intensive reseeding + fertilization), compared with non-degraded (ND) controls. Over two years, we quantified vegetation diversity, aboveground biomass, soil microbial communities (bacteria, fungi, nematodes), and 16 soil physicochemical properties to calculate soil multifunctionality. An economic cost-benefit analysis complemented the ecological assessment. We found that aboveground biomass and major soil functions recovered rapidly, even exceeding ND levels in some treatments. In contrast, plant diversity and soil microbial community composition, particularly bacteria and nematodes, remained significantly impaired. This asynchrony was most pronounced in SD grasslands. LD and MD grasslands achieved soil multifunctionality comparable to ND sites, whereas SD sites showed persistent functional deficits. Economically, interventions on LD grasslands yielded a 10.4% return, compared to a -29.5% loss for SD grasslands. Graded management effectively restores soil functions but creates a clear trade-off: productivity responds directly to resource subsidies, while biodiversity recovery lags, constrained by dispersal limitations and soil legacies. We demonstrate that prioritizing preventative interventions in LD/MD grasslands is ecologically and economically optimal. For SD systems, future strategies must address fundamental soil biotic thresholds, potentially through targeted soil amendments, to synchronize recovery trajectories.