Biogeographic Variation in Temperature Sensitivity of Decomposition in Forest Soils

Jinquan Li,Ming Nie,Elise Pendall,Peter B. Reich,Junmin Pei,Nam Jin Noh,Ting Zhu,Bo Li,Changming Fang
DOI: https://doi.org/10.1111/gcb.14838
IF: 11.6
2019-01-01
Global Change Biology
Abstract:Determining soil carbon (C) responses to rising temperature is critical for projections of the feedbacks between terrestrial ecosystems, C cycle, and climate change. However, the direction and magnitude of this feedback remain highly uncertain due largely to our limited understanding of the spatial heterogeneity of soil C decomposition and its temperature sensitivity. Here we quantified C decomposition and its response to temperature change with an incubation study of soils from 203 sites across tropical to boreal forests in China spanning a wide range of latitudes (18 degrees 16 ' to 51 degrees 37 ' N) and longitudes (81 degrees 01 ' to 129 degrees 28 ' E). Mean annual temperature (MAT) and mean annual precipitation primarily explained the biogeographic variation in the decomposition rate and temperature sensitivity of soils: soil C decomposition rate decreased from warm and wet forests to cold and dry forests, while Q(10-MAT) (standardized to the MAT of each site) values displayed the opposite pattern. In contrast, biological factors (i.e. plant productivity and soil bacterial diversity) and soil factors (e.g. clay, pH, and C availability of microbial biomass C and dissolved organic C) played relatively small roles in the biogeographic patterns. Moreover, no significant relationship was found between Q(10-MAT) and soil C quality, challenging the current C quality-temperature hypothesis. Using a single, fixed Q(10-MAT) value (the mean across all forests), as is usually done in model predictions, would bias the estimated soil CO2 emissions at a temperature increase of 3.0 degrees C. This would lead to overestimation of emissions in warm biomes, underestimation in cold biomes, and likely significant overestimation of overall C release from soil to the atmosphere. Our results highlight that climate-related biogeographic variation in soil C responses to temperature needs to be included in next-generation C cycle models to improve predictions of C-climate feedbacks.
What problem does this paper attempt to address?