In parts of the upper reaches of the Red River Basin in Texas, streamflow is characterized by levels of salinity that limit its usefulness for most purposes. Large dissolved solids and dissolved chloride concentrations are caused primarily by flow from natural salt springs in tributaries to the Red River. To reduce downstream salinity in the Wichita River, a dam in the South Wichita River downstream of an area of salt springs (designated salinity source area VIII) diverts low flows (which are the most saline) to a manmade brine lake for evaporation.

Statistical tests on salinity data for the South Wichita River, Lake Kemp, and the Wichita River for the period October 1982–September 1992 were done to determine the effects on downstream salinity of low-flow diversions from the South Wichita River that began in May 1987.

Salinity in the South Wichita River downstream of the low-flow diversion structure was (statistically) significantly less during the 65-month period of record after diversion than during the 55- month period of record before diversion. Wilcoxon rank-sum tests yielded strong evidence that discharge-weighted dissolved solids and dischargeweighted dissolved chloride concentrations, as well as discharge-weighted specific conductance, were significantly less after diversion.

Whether salinity in Lake Kemp had a significant downward trend during the period of record August 1989–August 1992 could not be determined conclusively from observed salinity data. Mann-Kendall trend tests yielded weak evidence that volume-weighted dissolved solids and dissolved chloride concentrations in Lake Kemp tended to decrease with time. However, serial correlation in the time series of salinity data could have adversely affected the test results.

The significant effects of low-flow diversions on salinity in the South Wichita River are not discernible in the Wichita River downstream from Lake Kemp. Although salinity was significantly less downstream from Lake Kemp after diversion, the decrease probably is mostly a result of dilution of Lake Kemp by large inflows of (assumed) low-salinity water that occurred in the spring of 1989 rather than an effect of diversion.