Large underestimations of warming-induced soil carbon emissions from oversimplistic Q10 indicator

Abstract

The sensitivity of soil microbial respiration to climate warming is a major source of uncertainty in predicting soil carbon (C) emissions to the atmosphere and their feedback to climate change. One key issue is the persistent misuse of Q₁₀, the factor by which respiration rate is multiplied for a 10^oC increase in temperature, as an indicator of the temperature sensitivity. Despite ample empirical and theoretical evidence that Q₁₀ is temperature-dependent, most publications on the topic continue to measure and conceptualise Q₁₀ as being independent of temperature. Here, we analyse a published dataset of temperature incubations of soil microbial respiration across a global latitudinal gradient and project the resulting sensitivities onto global maps under four climate change scenarios. We reveal that omitting to account for the temperature dependence of Q₁₀ leads to an underestimation of global soil C emissions from 2015 to 2100 ranging from 5.5 ± 2.4 PgC to 10.4 ± 6.9 PgC across different climate change scenarios. Moreover, beyond uncertainties in predictions of global soil C emissions, modelling inaccuracies are geographically skewed, with large underestimations at high latitudes (of up to 34 MgC ha^-1) and overestimations in warmer regions (of up to 12 MgC ha^-1). The disparate regional patterns have large implications for land stewardship, as management efforts could overlook soil C losses in vulnerable septentrional areas while unnecessary interventions could be recommended in tropical regions where soil C sequestration may not be as pressing.