Contrasting mechanisms underlie short- and longer-term soil respiration 1 responses to experimental warming in a dryland ecosystem

Soil carbon losses to the atmosphere through soil respiration are expected to rise with ongoing temperature increases, but available evidence from mesic biomes suggests that such response disappears after a few years of experimental warming. However, there is lack of empirical basis for these temporal dynamics in soil respiration responses, and for the mechanisms underlying them, in drylands, which collectively form the largest biome on Earth and store 32% of the global soil organic carbon pool. We coupled data from a 10 year warming experiment in a biocrust‐dominated dryland ecosystem with laboratory incubations to confront 0–2 years (short‐term hereafter) versus 8–10 years (longer‐term hereafter) soil respiration responses to warming. Our results showed that increased soil respiration rates with short‐term warming observed in areas with high biocrust cover returned to control levels in the longer‐term. Warming‐induced increases in soil temperature were the main drivers of the short‐term soil respiration responses, whereas longer‐term soil respiration responses to warming were primarily driven by thermal acclimation and warming‐induced reductions in biocrust cover. Our results highlight the importance of evaluating short‐ and longer‐term soil respiration responses to warming as a mean to reduce the uncertainty in predicting the soil carbon–climate feedback in drylands.
This research was funded by the European Research Council (ERC Grant agreements 242658 [BIOCOM] and 647038 [BIODESERT]). M.D. is supported by a FPU 592 fellowship from the Spanish Ministry of Education, Culture and Sports (FPU-593 15/00392). PGP is supported by a Ramón y Cajal grant from the Spanish Ministry of 594 Science and Innovation (RYC2018-024766-I). S.A. acknowledge the Spanish MINECO 595 for financial support via the DIGGING_DEEPER project through the 2015-2016 596 BiodivERsA3/FACCE‐JPI joint call for research proposals. F.T.M. and S.A. 597 acknowledge support from the Generalitat Valenciana (CIDEGENT/2018/041). C.C-D. 598 acknowledge the supported from the European Research Council (ERC Grant 647038 599 [BIODESERT]). 600 Competing interests 601 The authors declare no competing financial interests. 602 603 Author 604 F.T.M. designed the field study and wrote the grant that funded the work. F.T.M, P.G.P 605 and M.D. developed the original idea of the analyses presented in the manuscript. M.D. 606 performed the statistical analyses, with inputs from F.T.M and P.G.P. M. D., S.A., C.C.-607 D., B.G. and V. O. conducted the field and laboratory work. All authors contributed to 608 data interpretation. M.D. wrote the first version of the manuscript, which was revised by 609 all co-authors. 610 611 Data Sharing and Data Accessibility 612 The data that support the findings of this study and the R code are openly available in 613 Figshare at