Reconciling Carbon Flux Discrepancies in a Desert Environment: Characterizing Influences of Soil Processes Using Automated CO2 Flux Chambers

Dryland ecosystems play a fundamental role in controlling the balance of carbon on a global scale. These ecosystems dominate the inter-annual variability of carbon uptake by terrestrial ecosystems, and this is driven by temperature and precipitation patterns. Despite this global importance, there are some important discrepancies that have been reported between CO2 fluxes measured with differing techniques, particularly between eddy covariance tower measurements of net ecosystem exchange (NEE) of carbon (C) versus on-the-ground methods such as chambers and biomass surveys. The discrepancy is a consequence of instrumentation and measurement technique differences; yet, its biological origins remain unknown. The goal of this thesis is to investigate some of these differences, particularly with respect to the CO2 fluxes coming from soils. The first objective of this thesis was to design and build a set of CO2 automated flux chambers. Using these automated flux chambers, I then investigated (1) the magnitude of soil CO2 fluxes (RS) over time and in comparison to environmental variables such as temperature and moisture; (2) comparisons of RS with estimates of ecosystem respiration (ER) from a co-located eddy covariance tower; and (3) the influence of belowground autotrophic respiration (RA) via a shrub removal experiment. The main findings include: (1) When temperature increases, CO2 flux also increases over the course of spring and responds to rainfall. (2) The magnitude of CO2 efflux from the soil to the atmosphere was similar between the eddy covariance tower and the automated flux chamber measurements. At midnight, a time when photosynthesis does not contribute to NEE measurements, the trends from both measurements matched closely, including a peak in CO2 flux in the middle of March caused by a rain event in both measurements. This result suggests that both types of measurements can be accurately compared to each other. (3) In the shrub removal experiment, soil CO2 efflux did not change as expected; instead it remained similar. Therefore, plants and especially roots did not appear to contribute as much to soil respiration as expected. I believe environmental factors such as temperature, humidity, and pedogenic carbonates may contribute to this result, given the low amount of rainfall during this period. Overall, my results indicate that the discrepancies between tower and ground measurements are less likely to arise from the eddy covariance tower, since they matched the automated flux chamber values relatively well. Instead, the differences are more likely to originate from difficulties in measuring biomass-related fluxes such as belowground production