Heterotrophic respiration from soil increases with atmospheric carbon dioxide and temperature

Background/Question/Methods

Soils are the largest carbon pools in most ecosystems, but global change models retain substantial uncertainty about how soils will respond to long-term increases in temperature and CO₂. These uncertainties are primarily based on conflicting experimental findings, underscoring the need for multifactorial experiments under natural conditions to build and validate robust process-based models. In this study, we manipulated air temperature and CO₂ levels to characterize changes in soil respiration, which is the most sensitive indicator of short-term changes in soil carbon. We predicted that heat would stimulate soil heterotrophs and thus increase respiration, and that this effect would be stronger under elevated CO₂ because of increased litter inputs.

 We used infrared heaters (ambient + 3.5 °C) and free air carbon dioxide enrichment (FACE) technology (550 ppm CO₂) to simulate climate conditions predicted for 2050 in a central Illinois corn-soybean rotational agro-ecosystem. We monitored heterotrophic (R_het) and total soil respiration, then calculated autotrophic respiration (R_aut) as R_total - R_het. We used these observations to parameterize the DAYCENT dynamic SOM model to predict changes in soil C under future climate scenarios.

Results/Conclusions

Heat significantly increased heterotrophic respiration by 10-30% across two growing seasons and the intervening winter. Heat decreased autotrophic respiration by 5-40% during the growing season. Increasing atmospheric CO₂ did not increase summer heterotrophic respiration, but there was a lag effect of CO₂ that resulted in higher heterotrophic respiration from litter decomposition during winter. There was no significant change in total soil respiration during any season.

 DAYCENT correctly predicted changes in respiration under our experimental treatments, even when run with randomly selected weather data from years prior to the experiment. Modeled R_het accounted for over 60% of observed variation in measured R_het, lending confidence that model predictions of soil C cycling are robust in the near term. Because R_het originates from breakdown of stored carbon while R_aut comes from current photosynthate, increases in R_het may indicate losses of SOM even when matched by R_aut decreases. This underscores the need to partition flux measurements rather than relying on R_total to indicate SOM behavior. If temperature increases continue to stimulate observed R_het, continued climate change is likely to result in loss of soil organic matter, causing reductions in soil fertility and positive feedbacks to warming.