The effect of daily temperature variability on microbial and plant processes in a Chihuahuan Desert ecosystem
van Gestel, Natasja
MetadataShow full item record
Worldwide, the daily temperature range of air (DTRair = Tmax – Tmin) has decreased by 0.07 °C per decade, with a 43% stronger decline for arid and semiarid regions. Although the daily temperature range of soil (DTRsoil) has not been widely measured, it is reasonable to assume that it has decreased at a similar rate and magnitude as DTRair. The role of temperature on plant and soil processes has been extensively studied, but the role of temperature variability on these processes has been largely ignored. In arid systems where DTRsoil is characteristically high, projected additional reductions in DTRsoil may have significant impacts on ecosystem functioning. My dissertation focuses on elucidating the role of high temperature variability on microbial and plant processes in an arid ecosystem, both intra- (i.e. seasonal) and inter-annually. Using a passive temperature manipulation experiment that successfully reduced DTRsoil in a Chihuahuan Desert soil at Big Bend National Park, my field study evaluated the biomass and activity responses of microorganisms in response to year-round reductions in DTRsoil, and subsequent changes to soil nutrient levels. In addition, changes to leaf-level physiology, leaf N content and leaf xylem water status of the dominant and representative plant species of arid landscapes, Larrea tridentata (creosotebush), were measured in response to reduced DTRsoil. To better link below-ground processes to plant responses, I conducted all measurements on the same day. High temperature variability was an important stressor to microbial growth as soil microbial biomass C and N increased in response to reductions in DTRsoil. Reduced DTRsoil benefited both dormant and active microbial populations through increased biomass C and N relative to control plots in both dry (spring) and wet (summer) seasons. In contrast, microbial activity, measured as CO2 evolution from soil in inter-shrub spaces, was more sensitive to soil water content and less sensitive to temperature variability than microbial biomass. Therefore, reductions in DTRsoil generated the largest effects on CO2 evolution in summer, which is the wettest season in Big Bend National Park. Increased microbial biomass reduced soil exchangeable N, most likely because extra N was required for biomass construction. However, soil exchangeable N levels did not always decrease in response to increased microbial biomass, suggesting that mineralization of N from a more stable pool of soil organic matter functioned to replenish depleting levels of soil exchangeable N. Although, I observed changes to belowground dynamics, including soil nutrient status and soil CO2 efflux rates, reductions in leaf [N] in Larrea tridentata did not alter photosynthetic rates in response to reductions in DTRsoil. Lastly, I compared different multiple regression models that utilized daily insolation and air temperature data to predict daily maximum and minimum soil temperatures at two soil depths (0 and 15 cm). Using a weighted average of current and past insolation (to incorporate a “heat” storage effect) in combination with air temperature provided the best fit for observed daily maximum and minimum soil temperature near the soil surface. An analytical solution can then be applied to use the predicted soil surface temperature data to estimate daily maximum and minimum soil temperatures deeper into the soil profile. In summary, my research generated three major findings. First, deserts dominated by Larrea may function temporarily as a source of C, resulting in a positive feedback to rising global temperatures. This imbalance will be sustained as long as the C and energy source (i.e. soil organic matter) continue to fuel higher levels of microbial activity. Second, if additional N incorporated into microbial biomass (labile pool) was derived from a more stable pool, this could increase volatile losses of N and further limit N in this N-limited system, and in turn, affect future primary productivity. Third, my dissertation produced promising results for predicting the soil thermal environment from above-surface conditions in a desert system. Using the same variables, this approach could be used in other arid systems with limited soil temperature data. Predicting future soil thermal regime is necessary to anticipate impacts on ecosystem function.