Abstract

A study was conducted by the U.S. Geological Survey in cooperation with the Missouri Department of Conservation to estimate the effects of existing and proposed impoundments, land-cover changes, and reported water uses on streamflows in the 5,410-square mile upper Osage River Basin. The hydrologic model Hydrologic Simulation Program-FORTRAN (HSPF) was calibrated and validated to current (1995–2004 water years) regulation and water-use conditions, and scenarios were developed to evaluate differences for the same 10-years of record under pre-settlement, and proposed impoundment conditions. Analyses included quantification of changes in the magnitude, frequency, timing, and duration of streamflows under each simulation scenario. Streamflows from the simulations were used in conjunction with known streamflow-fish habitat relations to quantify effects of altered flows on fish-habitat area at selected Marais des Cygnes and Marmaton River locations.

The cumulative effects of impoundments and land-cover changes were determined to substantially alter streamflows in the upper Osage River Basin model simulations spanning pre-settlement to proposed future conditions. The degree of streamflow alteration varied between major subbasins. Streamflows in the Marais des Cygnes River Basin were altered between pre-settlement and current conditions, primarily by major impoundments, with smaller changes expected with proposed regulation. Streamflows in the Little Osage River Basin were relatively unchanged between pre-settlement and current conditions with land-cover changes (primarily the conversion of native prairies to cultivated land) affecting flows more than the few current impoundments in this basin. The current peak flows in the Marmaton River Basin generally were higher than pre-settlement or proposed scenario peak flows. Of the three major subbasins, the Marmaton River Basin is likely to be the most affected by proposed impoundments.

Declines in monthly minimum streamflows under a proposed impoundment scenario at the Marais des Cygnes River near the Kansas-Missouri state line, Kansas, were greatest for the lowest 10 percent of corresponding observed flows and during the driest years (2000, 2001 water years); that is, the greatest percent declines in flows under proposed conditions generally occurred during the lowest current/observed flow periods. In a small headwater basin in the Marmaton River Basin, simulated declines in minimum flows were small (generally less than 6 cubic feet per second and less than 1 cubic foot per second for 1- and 3-day scenarios), but resulted in 10 to 18 additional zero flow days for the 10-year simulation for the proposed scenarios relative to current simulated conditions. Reductions in minimum monthly flows as a result of additional impoundments generally were less than 5 cubic feet per second at the Marmaton River near Marmaton, Kansas, and resulted in 6 additional zero flow days. The greatest declines between proposed and current flows at the Marmaton River near the Kansas-Missouri state line, Missouri, generally occurred in the lower 50 percentile of the distribution of current simulated flows and during the drier simulation years (2001–2003). Proposed conditions resulted in declines in the 0-10 percentile flow values for the 1-, 3-, and 7-day durations. July, August, and October had the largest declines in proposed low flows relative to current simulated low flows for the 10-year simulation at this site.

The flood frequency for the Marais des Cygnes River near the Kansas-Missouri state line was unchanged between observed and proposed conditions for the 10-year simulation, but was 450 percent greater under the pre-settlement scenarios compared to observed conditions. Flood frequency generally was greatest for the current condition scenarios in the Marmaton River Basin and least for the proposed conditions, although the effects of regulation on flood frequency decreased downstream from the Kansas-Missouri state line with substantial streamflow contributions from less regulated tributaries. The flood frequencies of the proposed scenarios were 54 to 60 percent less than current conditions at the Marmaton River near the Kansas-Missouri state line, whereas at a downstream location the flood frequencies under proposed conditions were 39 to 45 percent less than current conditions.

Minimum annual 1-, 7-, or 14-day selected fish habitat availability under a proposed scenario declined at three Marais des Cygnes River reporting sites by more than 10 percent compared with observed conditions, for one or more years, for each of nine seasonal fish habitat categories. Declines in minimum habitat availability under proposed conditions were at or near 100 percent for one or more years for summer flathead catfish, fall flathead catfish, fall channel catfish, and fall stonecat habitat categories at one Marais des Cygnes River location, and for summer flathead catfish, summer channel catfish, and fall flathead catfish at another Marais des Cygnes River location. Declines in annual 1-, 7-, or 14-day minimum habitat also were greater than 10 percent for one or more years for all categories at both Marmaton River reporting locations except for spring paddlefish habitat, which generally remained unchanged between current and proposed scenarios at one location. Declines in 1-, 7-, or 14-day proposed minimum habitat availability were at or near 100 percent for one or more years for slenderhead darter, summer flathead catfish, fall channel catfish, and fall stonecat habitat categories at one Marmaton River location and for spring suckermouth minnow, spring slenderhead darter, summer channel catfish, summer stonecat, and fall flathead catfish habitat at another.

Simulations of the Marmaton River Basin (1995 through 2004) indicated that the effects of a conversion of cultivated row crops back to pre-settlement native prairie soils, simulated using an increase in the infiltration model parameter, accounted for a greater difference in total runoff between pre-settlement and current/proposed scenarios than other changes in land cover or from impoundments. The simulated increase in soil infiltration capacity under native prairie conditions also resulted in lower peak flows for the pre-settlement model scenario compared with the current or proposed scenarios. Evaporative water losses increased with the addition of impoundments, and while these increases did not have a substantial effect on the total runoff from the basin they could account for simulated declines in low flows. The greater detention associated with proposed impoundments resulted in longer hydrograph recessions and lowered peak flows; this varied with modifications in simulated impoundment design.