Assessing Ecosystem Service Benefits of Improved Soil Management Practices at the Field and Watershed Scales
Abstract
Poor soil management limits the soil functions of water infiltration and storage and enhances water and soil losses. Amidst recurring droughts and declining water availability, improved soil management practices such as no-tillage (NT) and cover crops (CC) could play a critical role in improving soil ecosystem services while sustaining agriculture and mitigating negative impacts of climate change. The goal of this Dissertation research was to quantify the potential soil and water conservation benefits of adopting improved soil management practices at the field and watershed scales using hydrologic models and Unmanned Aerial System (UAS)- based measurements in the Texas Blackland Prairies and Rolling Plains regions. This was achieved by using the Precision Agricultural Landscape Modeling System (PALMS) and Agricultural Policy / Environmental eXtender (APEX) models and developing regression models based on UAS data.
PALMS field-scale simulations indicated a substantial increase in plant available water in the soil under NT over conventional tillage (CT), especially during dry years. At the watershed scale, APEX was trained to follow the Curve Number-Precipitation relationship, which was developed based on rainfall-runoff measurements from CT and NT fields at Riesel, TX. Changing management on all croplands of the Bushy Creek Watershed from CT to NT resulted in a 25% and 57% reduction in average annual runoff and sediment loss, respectively at the watershed outlet, most of which occurred during the months with high intensity rainfall. Simulated average reduction in annual runoff and sediment loss with NT management under projected future climate ranged between 27–33% and 55–59%, respectively, compared to CT management under historic climate.
The effects of CCs and tillage management on crop growth were assessed using the UAS imageries collected during cotton growing season from a long-term CC/NT experimental field at Chillicothe, TX. Vegetation indices from plant pixels were found to be highly effective in distinguishing between the effects of different soil management practices. The UAS imageries were also found to be useful for predicting soil water content by developing regression models. Overall, this research revealed several benefits of improved soil management practices at field and watershed scales under current and projected future climatic conditions.
Description
Keywords
Texas Blackland Prairies, macropores, surface roughness, surface
disturbance, no-tillage, APEX, PALMS, CMIP6, climate change, machine learning, soil moisture, UAS, Riesel, Brushy Creek, Micasense, multispectral, extreme rainfall, curve number-precipitation relatioship, cover crop