Implementing thinning practices can enhance the growth of plantation forests and improve soil health. Nevertheless, the impacts of thinning applications on soil quality, large-diameter timber production of Castanopsis hystrix, and ecosystem multifunctionality are poorly understood. Therefore, we chose two sample plots, unthinned (control) and thinned, to investigate productivity and ecosystem multifunctionality after thinning for six years. Results revealed that thinning significantly reduced the soil’s bulk density, enhanced large-diameter timber growth, and undermined ecosystem multifunctionality in the short term compared to control (unthinning) treatment. Compared to the control, the thinning treatment considerably enhanced the soil organic carbon (0–30 cm soil layer) and tree diameter at breast height (20–30 cm), and enhanced shrub leaf nitrogen (N), shrub root N, herb aboveground N, Gram-positive bacteria (0–10 cm soil layer), and Gram-positive bacteria (20–30 cm soil layer) contents by 29.61%, 65.29%, 44.61%, 274.35%, and 323.44%, respectively. Furthermore, the thinning application could improve the N and P resorption efficiency more than control. Furthermore, compared with control, thinning treatment maximized decomposition and nutrient cycling function by 11.81% and 143.40%, respectively. Moreover, total PLFA content significantly impacts carbon stocks, wood production, and water regulation functions. In conclusion, this study underscores the considerable potential of thinning in augmenting large-diameter timber production by stimulating the positive effects of forest stands. These findings provide valuable insights for ecosystem multifunctionality elevation and the judicious application of thinning to improve forestry productivity, facilitating sustainable development in the forestry sector.

Monoculture plantation systems face increasing challenges in sustaining ecosystem multifunctionality (EMF) under intensive management and climate change, with long-term functional trajectories remaining poorly understood. Although biodiversity–EMF relationships are well-documented in natural forests, the drivers of multifunctionality in managed plantations, particularly age-dependent dynamics, require further investigation. This study examines how stand development influences EMF in Castanopsis hystrix L. plantations, a dominant subtropical timber species in China, by assessing six ecosystem functions (carbon stocks, wood production, nutrient cycling, decomposition, symbiosis, and water regulation) of six forest ages (6, 10, 15, 25, 30, and 34 years). The results demonstrate substantial age-dependent functional enhancement, with carbon stocks and wood production increasing by 467% and 2016% in mature stand (34 year) relative to younger stand (6 year). Nutrient cycling and water regulation showed intermediate gains (6% and 23%). Structural equation modeling identified plant diversity and microbial community composition as direct primary drivers. Tree biomass profiles emerged as the strongest biological predictors of EMF (p < 0.01), exceeding abiotic factors. These findings highlight that C. hystrix plantations can achieve high multifunctionality through stand maturation facilitated by synergistic interactions between plants and microbes. Conservation of understory vegetation and soil biodiversity represents a critical strategy for sustaining EMF, providing a science-based framework for climate-resilient plantation management in subtropical regions.

In the face of increasingly complex urban challenges, a critical question arises: can urban ecosystems maintain resilience, vitality, and sustainability when confronted with external threats and pressures? Taking Kunming—a plateau-mountainous city in China—as a case study, this research constructs an urban ecosystem resilience (UER) assessment model based on the DPSIR (Driving forces, Pressures, States, Impacts, and Responses) framework. A total of 25 indicators were selected via questionnaire surveys, covering five dimensions: driving forces such as natural population growth, annual GDP growth, urbanization level, urban population density, and resident consumption price growth; pressures including per capita farmland, per capita urban construction land, land reclamation and cultivation rate, proportion of natural disaster-stricken areas, and unit GDP energy consumption; states measured by Evenness Index (EI), Shannon Diversity Index (SHDI), Aggregation Index (AI), Interspersion and Juxtaposition Index (IJI), Landscape Shape Index (LSI), and Normalized Vegetation Index (NDVI); impacts involving per capita GDP, economic density, per capita disposable income growth, per capita green space area, and per capita water resources; and responses including proportion of natural reserve areas, proportion of environmental protection investment to GDP, overall utilization of industrial solid waste, and afforestation area. Based on remote sensing and other data, indicator values were calculated for 2006, 2011, and 2016. The entire-array polygon indicator method was used to visualize indicator interactions and derive composite resilience index values, all of which remained below 0.25—indicating a persistent low-resilience state, marked by sustained economic growth, frequent natural disasters, and declining ecological self-recovery capacity. Forecasting results suggest that, under current development trajectories, Kunming’s UER will remain low over the next decade. This study is the first to integrate the DPSIR framework, entire-array polygon indicator method, and Grey System Forecasting Model into the evaluation and prediction of urban ecosystem resilience in plateau-mountainous cities. The findings highlight the ecosystem’s inherent capacities for self-organization, adaptation, learning, and innovation and reveal its nested, multi-scalar resilience structure. The DPSIR-based framework not only reflects the complex human–nature interactions in urban systems but also identifies key drivers and enables the prediction of future resilience patterns—providing valuable insights for sustainable urban development.