Historical data for nine selected streamgages in southwest and south-central Kansas were used in an assessment of streamflow characteristics and trends. This information is required by the U.S. Fish and Wildlife Service and the Kansas Department of Wildlife, Parks and Tourism to assist with the effective management of Etheostoma cragini (Arkansas darter) habitats and populations in the State. Changing streamflow conditions, such as a reduction or elimination of streamflow, may adversely affect the Arkansas darter. Priority basins for the Arkansas darter represented by the selected streamgages include the Cimarron River, Rattlesnake Creek, the North Fork Ninnescah River, the South Fork Ninnescah River, the Medicine Lodge River, and the Chikaskia River.

Streamflow conditions were assessed using annual streamflow characteristics computed for the period of record for each of the selected streamgages. Specific streamflow characteristics computed were mean discharge, mean base flow, 90th-percentile flow, 10th-percentile flow, minimum 7-day mean flow, minimum 28-day mean flow, number of days of flow less than 1 cubic foot per second, and number of zero-flow days.

Two of the priority basins had statistically significant decreases in annual mean discharge during the period of record. In the Cimarron River Basin, there was a pronounced multidecadal decrease in the magnitude and variability of annual mean discharge. Concurrently, the percentage of the annual mean discharge that was contributed by base flow increased. In the Rattlesnake Creek Basin, there was a pre-1985 decrease in annual mean discharge. Typically, in these two basins, significant decreases were indicated for mean base flow, 90th-percentile flow, 10th-percentile flow, minimum 7-day mean flow, and minimum 28-day mean flow. No significant trend in annual mean discharge was indicated for the North Fork Ninnescah, South Fork Ninnescah, Medicine Lodge, and Chikaskia River Basins. For the Medicine Lodge and Chikaskia River Basins as well as the downstream part of the South Fork Ninnescah River Basin, a significant increase in mean base flow and 10th-percentile flow was indicated. Also, for the latter two basins, a significant increase was indicated for minimum 7-day mean flow.

Factors investigated to explain long-term trends in annual mean discharge, or lack thereof, included precipitation and groundwater withdrawals. Annual precipitation in the study area varied substantially from 1951 to 2013 with no pronounced long-term trend. Thus, a precipitation-related explanation for the significant decrease in annual mean discharge in the Cimarron River and Rattlesnake Creek Basins was not supported. Because the most pronounced decreases in annual mean discharge were in the basin with the largest groundwater-level declines (that is, the Cimarron River Basin), both in terms of magnitude and areal extent, it is likely that groundwater withdrawals were a primary, if not dominant, causative factor.

The occurrence of extremely low-flow (less than 1 cubic foot per second) and zero-flow days varied by basin and year. Typically, such days occurred in the summer and autumn for all basins.