Abstract The objective of this field trip is to examine variability in stream systems in west-central Ohio at different time and space scales. Scales of study range from watershed changes over 10 3 –10 4 years as drainage networks are established during glacial retreat, to reach-scale changes in tributary streams in response to human impacts and land use change over 10 1 –10 2 years, to diurnal and event-driven changes in water quality over 10 –2 to 10 –3 years. Drainage network changes in the Little Miami River and Mad River systems occurred through stream capture and were dependent on the location of early meltwater channels relative to ice lobe position and relict bedrock topography. At the reach scale, channel morphology (width, depth, slope, shape, and pattern) is dependent on mean discharge of water and sediment to the reach. Tributary streams to the Little Miami River, Mad River, and Buck Creek illustrate the impact that historic changes in land use, water and sediment discharge, channelization, and straightening of stream reaches have had on channel morphology and vertical stability. At the cross-section scale, flow characteristics, including stream stage and physical water quality parameters (temperature, pH, specific conductivity, oxidation-reduction potential, dissolved oxygen, and turbidity), are being measured on Buck Creek and Beaver Creek. Though some characteristics change diurnally in response to internal stream processes, event-based changes in response to stormflow reflect source area contributions of runoff and sediment.

The central questions of karst hydrology concern (1) recharge, storage, and flow of groundwater in contemporary aquifers, (2) identification of geologic constraints on groundwater storage and flow paths, and (3) understanding of how aquifers evolve through time and the relation of this evolution to the evolution of interconnected surface-water basins. Karst aquifers generally display matrix, fracture, and conduit permeability with contrasts in effective hydraulic conductivity of many orders of magnitude. Dispersed recharge into the matrix and fracture permeability provides most of the storage and a slow-response flow system, while point recharge into the conduit system provides quick flow and little storage. A current question is how to describe these components of permeability and the interchange of groundwater between them. Groundwater basins can be delineated by tracer studies and geologic boundary conditions. Progress is being made on the quantitative, fluid mechanics description of conduit flow. The evolution of karst aquifers is a mainly chemical process, with some transport of insoluble clastics by high-velocity conduit waters. The equilibrium carbonate chemistry has been well established for a long time, including accurate values for equilibrium constants. Conduits, shafts, and solutionally modified fractures are the result of differential dissolution rates. Geochemically satisfactory descriptions of dissolution kinetics have been established, so that the evolution of karst aquifers through time can be accurately described. Overall, a comprehensive model for karst aquifer behavior seems to be within sight.

The Spring Mill Lake watershed is located in the Mitchell Plateau, a karst area that developed on Mississippian carbonates in southern Indiana. Spring Mill Lake is a reservoir built in the late 1930s and is located in Spring Mill State Park. Within the park, groundwater from subsurface conduits issues as natural springs and then flows in surface streams to the lake. From 1998 to 2002, surface and subsurface hydrology and water quality were investigated to determine the types and sources of potential contaminants entering the lake. Water samples collected during base flow and a February 2000 storm event were analyzed for selected cations, anions, trace elements, selected U.S. Environmental Protection Agency (EPA) primary and secondary drinking-water contaminants, nitrogen isotopes, suspended solids, Escherichia coli , and pesticides. All of the water samples met the EPA drinking-water standards for inorganic constituents, except those collected at five sites in August 1999 during a drought. Nitrate nitrogen (NO 3 -N) concentrations were highest during base-flow conditions and displayed a dilutional trend during peak-flow periods. The NO 3 -N concentrations in water samples collected during the 2001 spring fertilizer applications tended to increase from early to late spring. All of the δ 15 N values were low, which is indicative of either an inorganic source or soil organic matter. Storm discharge contained increased concentrations of total suspended solids; thus, storms are responsible for most of the sediment accumulation in the lake. E. coli levels in 24% of the samples analyzed contained a most probable number (MPN) greater than 235/100 mL, which is the maximum acceptable level set for recreational waters in Indiana. E. coli does appear to be a potential health risk, particularly at Rubble spring. The sources of E. coli found at this spring may include barnyard runoff from a horse barn or wastes from a wastewater treatment facility. The pesticides atrazine, metolachlor, acetochlor, and simazine were detected during the spring of 2001. Atrazine, metolachlor, ace-tochlor, and simazine are used to suppress weeds during corn and soybean production. Additional sources of atrazine and simazine may result from application to right-of-ways, orchards, and managed forest areas.

Storms create stresses on karst systems that can alter the pathways and travel-times of water, solutes, and sediment. Flow contribution during storms is not only a matter of activation of new conduits, but is also a complex combination of water from conduits, enlarged fractures, and fractured matrix. In order to obtain evidence of pathway changes, we sampled three karst springs of varying size and maturity using data loggers for conductivity and water level, and storm water samplers for suspended sediment. The largest spring (Arch Spring) had the lowest conductivity of the three springs, indicating mainly conduit pathways at base flow. The high conductivity of base flow at the Nolte and Bushkill Springs pointed to contributions from slower-moving water in the fractured matrix. During storms, Arch Spring showed a consistent pattern of conductivity with a slight increase, then a large decrease, indicating an initial fracture flush of high-conductivity water, then passage of low-conductivity water from the precipitation. During storms, the conductivity of the middle-sized spring (Nolte Spring) either dropped immediately, or increased sharply then declined as storm water reached the spring. The smallest spring (Bushkill Spring) had a predictable conductivity pattern, with a sharp decrease and gradual recovery, suggesting shorter paths during storms than base flow. Sediment concentrations during storms were lowest at Nolte Spring and higher at Bushkill and Arch Springs, indicative of the fast flow through conduits or enlarged fractures suggested by the latter two springs during storms. The storm-water pathways vary from spring to spring and from storm to storm. These data show the importance of continuous monitoring to understand spring behavior.