Land-use and climate effects on soil respiration quantified with a landscape sensor network

Abstract:

Land-use change alters the magnitudes and variability of soil respiration (R_s). However, the importance of ecosystem and landscape drivers of R_s remain poorly understood from an empirical and mechanistic standpoint and likely vary across climate gradients. To address this knowledge gap we asked, what regulates the spatial and temporal variation of R_s across a human dominated region? From a landscape perspective, climate and land-use are hypothesized to be key drivers of R_s. From a soil physiological perspective, variability in temperature, moisture, and substrate availability regulate R_s. According to an inverse metabolic activity hypothesis, systems with higher metabolic activity will have less temporal variability in R_s than those with lower rates. Alternatively, soil substrate availability may drive sensitivity of R_s to water inputs. To quantify variability in R_s and test hypotheses of its regulation we deployed an R_s sensor network beginning in November 2013 with nodes distributed across three land-use types ─ lawn, agriculture, and wildland ─ at three sub-regions spanning a coastal to inland to desert climate gradient (total 9 sites, 3 land-uses x 3 sub-regions). At each node we are measuring soil R_s using the flux gradient approach, which includes soil state CO₂ sensors, temperature, and moisture measured at three depths and at five minute intervals.

We analyzed the data for the winter sampling period, which is southern California's rainy season. The mean of R_s was lowest at the coastal and highest in the desert sub-regions for both lawn (3.99 and 4.7 μmol CO₂ m^-2 s^-1) and wildland (0.23 and 0.49 μmol CO₂ m^-2 s^-1)  land-uses. R_s was the highest in the inland sub-region for agricultural land-uses (4.1 μmol CO₂ m^-2 s^-1). Lawn and wildland land-uses had increasing coefficient of variation in R_s across the coastal to desert climate gradient, while agriculture had decreasing variation. Sites that had higher mean fluxes and soil organic matter, a proxy for substrate availability, also had higher variability, supporting the hypothesis on substrate regulation of R_s sensitivity to water inputs. As urban centers spread in the arid and semi-arid Southwest, the likely consequences of land-use change will be increased magnitude and variance in R_s as the result of increased soil nutrient and water inputs.