Field-based autochamber measurements of soil net CO2 exchange in experimentally warmed biological soil crusts

Background/Question/Methods

Dryland ecosystems cover >40% of Earth’s terrestrial surface and, due to this large spatial extent, arid and semiarid soils maintain large stocks of carbon (C). Many dryland ecosystems have robust biological soil crust communities made up of mosses, lichens, cyanobacteria, and heterotrophs living at the soil surface. These communities are a fundamental component of dryland ecosystems, and are critical to dryland C cycling because they fix and respire significant amounts of CO₂. In dryland soils dominated by biocrusts, C is concentrated close to the surface, potentially increasing its vulnerability to climate change. Due to the prevalence of biocrusts and to the large stocks of C, this vulnerability is relevant to C balance at the global scale. Furthermore, loss of organic matter in dryland soils can result in loss of soil fertility and soil structure in these ecosystems. To examine the potential effects of warming temperatures on C balance in biocrusts, we warmed biocrusts on the Colorado Plateau, USA by 2°C above ambient using infrared heaters and monitored net ecosystem exchange (NEE) of CO₂ once per hour for 1 year and 9 months using 10 automated flux chambers (5 control and 5 warmed plots).

Results/Conclusions

Soils in control and ambient+2°C plots were net CO₂ sources to the atmosphere, with control plots losing 109±14 g C m^-2 (mean±SE) over the measurement period and the warmed plots losing 130±15 g C m^-2. However, there was substantial variability among the five replicate blocks, with paired plots having NEE differences (warmed-control) of -19, -3, 14, 39, and 72 g C m^-2 over the measurement period, for an overall difference of 20±44 g C m^-2 (mean±95%CI). The differences between warmed and control plots were more pronounced when soils were not excessively hot, cold, or dry. While not statistically significant, this difference indicates a 73% chance that warmed plots lost more C than control plots. Based on these data, we cannot rule out a substantial effect of warming on C balance in these soils, and variability in warming effects and/or ecosystem heterogeneity are likely important. Greater statistical power would permit more accurate quantification of effect size. The consistently negative C balance even in control plots suggests that (1) the ecosystem is losing C from soil organic matter or carbonate sources; and/or (2) losses are balanced by plant inputs from adjacent areas via processes such as root turnover or rhizodeposition.