Urban forest restoration enhances soil microbial functional potential and functional insurance via shifts in β-diversity

BACKGROUND: Forest restoration has primarily been evaluated through changes in aboveground communities, while belowground microbial communities-critical drivers of ecosystem functions-remain less understood. Moreover, studies of soil microbes have focused largely on community structure, which does not necessarily reflect the recovery of functional capacity and stability. METHODS: To determine how forest restoration affects microbial community structure and function and how microbial diversity relates to ecosystem multifunctional potential and stability, we analysed soil microbial communities from 79 urban forest restoration sites across New Zealand, spanning 0-63 years since initial plantings. Shotgun metagenomic sequencing was used to characterize taxonomic composition and functional potential, with diversity quantified using alpha and beta metrics. To evaluate links between diversity and ecosystem function, we assessed ecosystem multifunctional potential (EMF) which describes the ecosystem's capacity to simultaneously provide multiple functions, and we developed a novel functional insurance (FI) index grounded in ecological theory as an indicator of functional stability and resilience. To calculate FI in microbial systems from sequencing data, we quantified functional overlap by estimating over 250 million species-function correlations per sample. RESULTS: Contrary to our expectations, only beta diversity, not alpha diversity, was positively associated with EMF and FI, indicating that community composition and dissimilarity rather than species richness underpins microbial functional capacity and stability. EMF and FI were positively correlated, showing that high functional diversity and functional overlap can co-occur in microbial systems. In addition, archaeal turnover increased with closing forest canopies, contributing to higher EMF and FI, while bacterial turnover was only weakly associated with restoration parameters. Notably, restoration time did not play a role in shaping microbial diversity, EMF and FI. CONCLUSIONS: Our findings demonstrate that microbial compositional turnover, rather than increases in species richness, are critical for restoring soil ecosystem functions. Incorporating microbial functional metrics like the FI index into restoration frameworks that recognise both above and belowground dynamics could promote resilient and multifunctional urban forests.