Skip to main content
Dryad logo

Fire decreases soil enzyme activities and reorganizes microbially-mediated nutrient cycles: a meta-analysis


Zhou, Yong et al. (2022), Fire decreases soil enzyme activities and reorganizes microbially-mediated nutrient cycles: a meta-analysis, Dryad, Dataset,


The biogeochemical signature of fire shapes the functioning of many ecosystems. Fire changes nutrient cycles not only by volatilizing plant material, but also by altering organic matter decomposition—a process regulated by soil extracellular enzyme activities (EEAs). However, our understanding of fire effects on EEAs and their feedbacks to nutrient cycles is incomplete. We conducted a meta-analysis with 301 field studies and found that fire significantly decreased EEAs by ~20-40%. Fire decreased EEAs by decreasing soil microbial biomass and organic matter substrates. Soil nitrogen-acquiring EEA decreased alongside decreasing available nitrogen, likely from fire-driven volatilization of nitrogen and decreased microbial activity. Fire decreased soil phosphorus-acquiring EEA but increased available phosphorus, likely from pyro-mineralization of organic phosphorus. These findings suggest that fire suppresses soil microbes and consumes their substrates, thereby slowing microbially-mediated nutrient cycles (especially phosphorus) via decreased EEAs. These changes can become increasingly important as fire regimes in many ecosystems continue to shift in response to global change.


Study selection

We searched Web of Science database using terms fire* OR burn* AND enzyme AND soil to obtain potential publications for this meta-analysis on December 2, 2020. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses protocol to screen and identify publications to be included into this analysis . Briefly, title, abstract, and methods section of each publication were initially screened for eligibility based on whether the publication included fire impacts on soil EEAs. Eligible publications were further filtered based on whether the publication included at least one field study that examined soil EEAs between burned and unburned plots or treatments . Overall, a total of 301 field studies from 85 study sites were extracted from 71 publications.


Data collection

For each field study, we recorded site locations (latitude and longitude), climatic variables (mean annual precipitation [MAP] and temperature [MAT]), vegetation types, soil clay content, soil types, fire types (wildfire or prescribed fire), fire severity (low, moderate, and high), time since the last fire (years), and the number of fire events. If these data were not reported, we contacted the corresponding authors for additional information. Otherwise, site latitude and longitude were extracted from Google Maps based on the approximate location reported in the publication; MAT and MAP were extracted from the WorldClim database; and soil clay content was extracted from SoilGrids database. We classified vegetation into coniferous forests, deciduous forests, mixed forests, woodlands/shrublands, and savannas/grasslands. Though criteria for estimating fire severity may vary among studies, we used the reported fire severity from studies. If a study reported fire severity as “low to moderate” or “moderate to high”, we made a conservative decision to use the lower estimate. In some cases, if descriptions of post-fire changes in aboveground vegetation and organic matter in soils were reported, we assigned fire severity to these studies based on criteria from Keeley (2009). We classified soil types to orders based on the United States Department of Agriculture soil taxonomy.   


For quantifying fire effects on EEAs, we recorded means, standard deviations (or errors), the number of replicates for EEAs measured in unburned and burned treatments from the top-most soil layers. Soil extracellular enzymes were categorized into C-, N-, and P-acquiring enzymes. Soil C-acquiring enzymes included α-1,4-glucosidase, β-1,4-glucosidase, β-1,4-xylosidase, β -D-cellobiohydrolase, cellulase, invertase, and xylanase; N-acquiring enzymes included β-1,4-N-acetyl-glucosaminidase, asparaginase, leucine aminopeptidase, protease, and urease; and P-acquiring enzymes included both acid and alkaline phosphatase.  We also provided information on soil oxidative and S-acquiring enzymes, though these two categories are not discussed in-depth here. Where available, we tabulated soil pH, water content, organic C (or matter), total N (TN), total P (TP), available N (or total inorganic N), NH4+, NO3-, available P, microbial biomass carbon (SMBC) and nitrogen (SMBN). Following other meta-analyses of fire impacts on soil C and N cycles, soil P cycle, and soil microbial biomass, we treated different sampling dates after fire within the same publication as separate and independent field studies. When results were presented graphically, we used WebPlotDigitizer 4.4 to digitize the data.

Usage Notes

Please refer to the Metadata for more details and cite the above paper when using this data.