A Global Meta-analysis on the Impact of Human Activities on Soil Microbial Diversity and Carbon Cycling

Changes to the structure or functioning of the soil microbial community could alter the way it metabolises aboveground organic inputs, with significant potential implications for plant nutrient availability, the carbon cycle, and other aspects of soil health. Human activities have been shown to alter microbial diversity and activity in various study sites. However, different ecosystems respond differently to the same disturbance, so we need to identify the globally common patterns.

We perform a meta-analysis of the effects of five human activities – land use change, ecosystem restoration, pollution, pesticide use and fertiliser addition – on microbial diversity (measured as Shannon index of the catabolic diversity) and activity (measured as soil basal respiration). From an initial 693 records from Web of Science, we short-list 177 studies covering 924 datapoints across all six inhabited continents. For each of the five human activities, we identify treatment-control pairs from this dataset, and calculate their log response ratios (‘lRR’, the logarithm of the ratio of the treatment diversity or activity to the control value). From these lRRs, we calculate an overall effect size and confidence interval under a robust variance estimation meta-regression model. We also check for publication bias and any changes in reported effect size over time.

Our dataset did not significantly differ from a random sampling of land points on the earth along various climatic and edaphic axes. Median catabolic diversity in our dataset was 2.57 (with 95% of readings in the range 0.90 - 4.43) and median respiration activity was 1.63 μg CO2 g−1 h−1 (with 95% of readings between 0.12 and 150). Among human activities, fertiliser addition and ecosystem restoration increased diversity (by +12.9% and +8.4% respectively) and activity (+38.9% and +73.5%), while land use change reduced diversity (by 1.5%) and activity (by 21.0%). The effects of pollution and pesticide use were not statistically significant. We found no significant effect of publication bias, and no consistent trends in reported effect size over time.

Greater diversity generally improves ecosystem efficiency, so we expected an increase in diversity to lead to greater carbon assimilation by microbes and a decrease in respiration activity. However, we found human activities to cause changes in the same direction for both diversity and activity. Also, the increase in respiration activity in response to ecosystem restoration is almost three times the reduction in activity due to land use change, even after accounting for the different baselines. This suggests that restored ecosystems might use carbon less efficiently compared to intact ones.

Our results show that land use intensity has a negative impact on soil microbial diversity and activity, whereas nutrient addition has a positive effect. Soil microbes mediate how much carbon and other nutrients remain in soil and how much is lost to the atmosphere or other pools. Therefore, learning how humans alter their community structure and functioning will help in better understanding current global problems like soil nutrient deficiencies and climate change.