INTERACTIONS BETWEEN CHEMICAL, PHYSICAL, AND BIOLOGICAL PROCESSES DURING DESERTIFICATION OF GROUNDWATER-DEPENDENT SEMI-ARID GRASSLANDS

Desertification is estimated to cost $26 billion per year through loss of agricultural production, water reserves, and air quality. Semi-arid grasslands are prone to desertification through many factors including groundwater pumping. Groundwater pumping below the root-zone of groundwater-dependent vegetation leads to a decrease in vegetation cover exposing bare soil to wind erosion. Desertification of semi-arid grasslands can lead to a permanent change in vegetation state. Identifying when and where ecological changes are irreversible is problematic, requiring observations of a new ecological state that favors the continued process of wind erosion and depletion of soil resources. To determine biological, physical, and chemical processes affecting desertification in semi-arid groundwater-dependent grasslands, I examined hydrological and ecological factors across groundwater-dependent meadows in Owens Valley, California. I developed and compared empirical, process-based, and mechanistic models that predict mass transport. I found that scaled gap size explains 56% of the variation in total horizontal flux (Q), and the process-based model predicts Q better than the mechanistic model indicating the importance of scaled gap size in wind erosion modeling of heterogeneous vegetation. I explored the role of landscape connectivity of bare soil in enhancing Q and quantifying the magnitude of desertification across the landscape using circuit theory and Qrule. I found that landscapes that were more connected than neutral landscapes with the same bare soil cover were associated with groundwater decline during the drought and greater Q. This is consistent with the idea that the enhanced formation of connected pathways is evident at plots that arrived at a particular bare-soil cover via groundwater decline and wind erosion, rather than another process. I analyzed vegetation structure and monitored Q, air quality data, and PM10 emissions to evaluate the relationship between meadow degradation and air quality. I found that management practices have generated a new mid-valley meadow source of PM10 pollution. These results provide information and tools for resource managers of groundwater-dependent semi-arid grasslands to identify areas degraded by wind erosion, producing Q and PM10, and prone to desertification. Managers can use the information and tools to better gauge a well-field's health and adjust the amount pumped from wells