Assessing BMP effectiveness: multiprocedure analysis of observed water quality data

Observed water quality data obtained from eight stream monitoring locations within Richland-Chambers Watershed in north central Texas were analyzed for trends using box-and-whisker plots, exceedance probability plots, and linear and Mann-Kendall statistical methods. Total suspended solids decreased at seven out of eight stations, and at two of these stations, the decrease was significant. Mixed results were obtained for nitrogen across the stations. A nonsignificant and significant increase in nitrite plus nitrate nitrogen (nitrite+nitrate N) was noticed in two stations each, whereas at the other four stations showed nonsignificant decrease. The results of organic nitrogen (Org N) was similar to nitrite+nitrate N except that the two stations that showed significant increase in nitrite+nitrate N showed nonsignificant decrease in Org N. Mixed results were also noticed for orthophosphorus (Ortho P) including nonsignificant decrease at two stations, significant decrease and increase at one station each, and nonsignificant increase in four stations. In general, total phosphorus (TP) decreased at all stations, significantly at some, except one station where it increased significantly. Decreasing trends in sediment, Org N, Ortho P, and TP were likely related to implementation of best management practices (BMPs). Increasing trends in dissolved constituents including Ortho P and nitrite+nitrate N were likely due to increased surface residue as a result of some BMPs such as conservation tillage.