Authors: Matt C. Roby, University of Arizona, Tucson, AZ; and R. L. Scott and D. J. Moore

Title: Biophysical and vegetative controls on soil carbon dioxide efflux in the semiarid southwest U.S.

Authors: Matt C. Roby, University of Arizona, Tucson, AZ; and R. L. Scott and D. J. Moore

Abstract:

Soil efflux of carbon dioxide (CO2) is a major component of total ecosystem respiration, which strongly influences the carbon dynamics of ecosystems. Increased understanding of the controls on soil efflux in water-limited regions could lead to improved predictions of carbon emissions from semiarid ecosystems, which have been shown to influence the interannual variability of the terrestrial carbon sink. When not limited by other factors, soil efflux is often assumed to increase exponentially with temperature. However, in water-limited ecosystems, dynamics in water availability and vegetation activity associated with pulsed precipitation events may complicate the temperature response of soil efflux. Here we combine automated soil chamber and flux tower data in the southwest U.S. to investigate the role of soil temperature, soil moisture, and ecosystem photosynthesis on soil efflux from multiple semiarid sites with contrasting vegetation type and carbon inputs. Data were collected in 2012 from Lucky Hills (Ameriflux site US-Whs; Chihuahuan Desert shrubland), where comparisons were made between soil chambers adjacent to shrubs and those in bare, inter-canopy space. During 2016-2017, soil chambers were deployed next to grass patches at Kendall Grassland (Ameriflux site US-Wkg; warm season grassland). For the 2017 growing season at Kendall Grassland, we added a root-exclusion experiment by placing additional soil chambers in bare, weeded plots with trenched perimeters.

Preliminary results indicate that soil moisture strongly regulated the temperature sensitivity and magnitude of soil efflux. We also found that during the growing season, total soil efflux was greater for sites with higher cumulative ecosystem photosynthesis. When moisture was ample, baseline soil efflux (Rbase) at the grassland was greater for grass plots (Rbase = 0.38 µmol CO2 m-2 s-1; 95% CI: 0.32-0.44), relative to bare soil (Rbase = 0.18 µmol CO2 m-2 s-1; 95% CI: 0.15-0.20), while for the shrubland, Rbase was greater for shrub plots (Rbase = 0.85 µmol CO2 m-2 s-1; 95% CI: 0.64-1.06) than bare soil (Rbase = 0.22 µmol CO2 m-2 s-1; 95% CI: 0.16-0.28). Total soil efflux was 59% and 39% greater for vegetated plots than bare plots at the grassland and shrubland sites, respectively, highlighting the effect of vegetation activity on soil efflux. Accounting for the effects of soil moisture and vegetation improved model performance and the prediction of temporal dynamics in soil efflux (adj.R2 = 0.86), relative to a model based solely on soil temperature (adj.R2 = 0.20). Together, these findings highlight the importance of moisture and vegetative controls on soil efflux in semiarid ecosystems. These results suggest potential changes in the carbon dynamics of southwest US ecosystems in response to changes in soil moisture and vegetation composition associated with changes in climate and/or land use.