I studied how geology, land use and rainfall, correlated with peak flow responses in 110 headwater stream sites during a drought year. Highest discharges were observed in the most developed catchments and in the most poorly drained soils, but specific responses were variable depending on both geology and land disturbance. Redundancy analysis indicated that both surficial geology and land disturbance were important predictors of discharge and that rainfall was in general a poor predictor of discharge. I conclude that responses of headwater streams to individual storms are unpredictable from data generated using GIS, but increased peak flows occur associated with human development, mitigated by surficial geology. The headwater streams that are most vulnerable to flow alterations occur on poorly drained soils, and where urbanization tends to concentrate. Much greater attention to managing water is required if further degradation of stream ecosystems is to be prevented from our future land use.

Over the past century of settlement, the landscapes of the Midwestern United States have experienced extensive anthropogenic modifications in order to convert prior wetlands-lowlands to subsequent fruitful croplands. The hydrologic responses of these landscapes have been significantly altered by the installation of artificial drainage (surface ditches and subsurface tile drains) and the change in natural preferential flow paths (increased cracks or root holes due to land use practices). Changes to peak stream flow behaviors is a result of many different inter-related variables; however, intensified agricultural drainage remains one of the largest suspects. Though the effects of subsurface drainage (primarily in the form of tile drains) on landscape, hydrology, ecology, and economy have been questioned, theories of hydrologic controls continue to be vague at best. Soil-Water-Atmosphere-Plant, known as SWAP, was developed to simulate the interaction of vegetation development with the transport of water, solutes, and heat in the unsaturated zone. It is a one-dimensional, vertically directed model with a domain reaching from a plane just above the canopy to a plane in the shallow saturated zone. In the horizontal direction, the model's main focus is the field scale since most transport processes can be described in a deterministic way. The SWAP model was calibrated and validated for simulating flow regimes of drained and undrained landscapes in Iowa. A new term `flashiness' is used to characterize flow data. The Richards-Baker Flashiness Index quantifies the frequency and intensity of short term changes in streamflow. From the simulated results, the effect of anthropomorphic modifications to a landscape is determined to be strongly influenced by soil structural properties and hydraulic properties, along with rainfall regimes. Adding subsurface drains to soils with lower hydraulic conductivities, such as clay, tends to reduce peak flows during precipitation events. Conversely, adding drainage to soils with higher hydraulic conductivities, such as sand, increases peak flows. During years with heavy precipitation, soils with lower permeability show a `saddle shape' relationship between the flashiness index and the distance between tile drains produces. The lowest point of the `saddle' determines the ideal drain spacing for mitigating flashiness. When the shrinking and cracking of clay soils is considered, macropores dominate water flow pathways into the soil matrix and tile drains have a minimal effect on the flow regime. The volume of macropores at the surface of the soil profile is indirectly proportional to flashiness index. Independent of rainfall regimes, cropping season, and soil type, subsurface flows of drained landscapes always exceed that of undrained landscapes. Continuance of comprehensive studies of artificial subsurface drainage can produce positive impacts on engineering, economic, and ecological environments.

I studied how geology, land use and rainfall, correlated with peak flow responses in 110 headwater stream sites during a drought year. Highest discharges were observed in the most developed catchments and in the most poorly drained soils, but specific responses were variable depending on both geology and land disturbance. Redundancy analysis indicated that both surficial geology and land disturbance were important predictors of discharge and that rainfall was in general a poor predictor of discharge. I conclude that responses of headwater streams to individual storms are unpredictable from data generated using GIS, but increased peak flows occur associated with human development, mitigated by surficial geology. The headwater streams that are most vulnerable to flow alterations occur on poorly drained soils, and where urbanization tends to concentrate. Much greater attention to managing water is required if further degradation of stream ecosystems is to be prevented from our future land use.