Global meta-analysis on the responses of soil extracellular enzyme activities to warming.

Soil enzymes play critical roles in the decomposition of organic matter and determine the availability of soil nutrients, however, there are significant uncertainties in regard to how enzymatic responses to global warming. To reveal the general response patterns and controlling factors of various extracellular enzyme activities (EEA), we collected data from 78 peer-reviewed papers to investigate the responses of extracellular enzyme activities (EEA), including β-1,4-glucosidase (BG), β-d-cellobiosidase (CBH), β-1,4-xylosidase (XYL), leucine amino peptidase (LAP), N-acetyl-glucosaminidase (NAG), urease (URE), phosphatase (PHO), peroxidase (PER), phenol oxidase (POX), and polyphenol oxidase (PPO), to experimental warming. Our results showed that warming treatments increased soil temperature by 1.9 °C on average. The oxidative EEA, calculated as the sum of PER, POX and PPO, was on average stimulated by 9.4% under warming. However, the responses of C acquisition EEA (the sum of BG, CBH and XYL), N acquisition EEA (the sum of LAP, NAG and URE), and P acquisition EEA to warming had large variations across studies. The warming effects on C, N, P acquisition EEA and oxidative EEA tended to increase with soil warming magnitude and duration as well as the mean annual temperature. The response of C acquisition EEA to warming was positively correlated with fungal biomass, while that of P acquisition EEA had positive relationships with fungi: bacteria ratios. The response of oxidative EEA was negatively correlated with the abundance of gram-positive bacterial biomass. Our results suggested that warming consistently stimulated oxidative EEA, but had diverse effects on hydrolytic EEA, which were dependent on the warming magnitude or duration, or environmental factors. The observed relationships between changes in microbial traits and extracellular enzymes suggested that microbial compositions drive changes in enzyme decomposition under warming. Thus, incorporation of microbial modification in biogeochemistry models is essential to better predict ecosystem carbon and nutrient dynamics.