Soil heterotrophic respiration is insensitive to changes in soil water content but related to microbial access to organic matter

• We partitioned soil respiration into its autotrophic and heterotrophic components.
• Autotrophic and total soil respiration increased with irrigation.
• Autotrophic respiration and total soil respiration were correlated with soil water content.
• Heterotrophic respiration was insensitive to irrigation and soil water content.
• Heterotrophic respiration was correlated with soil organic matter protection.

While most models describing soil organic matter dynamics relate the turnover of soil organic matter to its chemical structure, there is increasing evidence that microbial accessibility to the organic matter is predominant in controlling its stability and its decay. This finding has major implications for forecasting the impacts of land management and climate change on soil organic carbon cycling and there is a need for further research to understand the role of microbial accessibility. Further progress is problematic because of difficulties in partitioning total soil respiration (RS) into its components soil heterotrophic respiration (RH) and soil autotrophic respiration (RA) in undisturbed ecosystems. To overcome this problem, we used an improved natural abundance of 13C technique to partition RS into RH and RA at one and six months after irrigation and addition of nitrogen fertiliser treatments was imposed in a grassland ecosystem.Addition of nitrogen had no significant effect on the components of soil respiration compared with the values for the control. Irrigation resulted in differential effects on RH and RA, leading to strong differences in the proportion of RH contributing to RS (fRH) with (mean ± standard error) 0.65 ± 0.09 and 0.30 ± 0.06 for the non-irrigated and irrigated plots, respectively, six months after the treatments began. RS and RA were strongly correlated and both showed strong linear relationships with gravimetric soil water content (WS) resulting in strong differences between irrigation treatments. Soil temperature (TS) and the amount of dissolved organic carbon were also correlated with both RS and RA. In contrast, RH was not related to variation in WS and TS but was correlated with the amounts of particulate organic matter carbon, a labile fraction of carbon defined as physically unprotected, and specific surface area, an indicator of soil protection capacity. There were, however, no differences in the labile carbon fraction and specific surface area between the irrigation treatments. We showed that RH was insensitive to irrigation and remained constant throughout the entire experimental period with a mean value of 1.7 ± 0.2 μmol m2 s− 1.Our findings support for the growing evidence that microbial accessibility to substrate, rather than its chemical structure, is predominant in regulating soil organic matter decomposition.