An 18-month field experiment was conducted to evaluate the effectiveness of
grass filter strips in removing sediment and various nitrogen species from
runoff. Runoff was collected from six 3.7 m wide experimental plots with 2
4.7 m long runoff source areas. Two plots had 8.5 m filters, two plots had
4.3 m filters, and two plots had no filters. Runoff was analyzed for total
suspended solids (TSS), total Kjeldahl nitrogen (TKN), filtered TKN (FTKN),
NH4+-N, and NO3--N. The Mann-Kendall nonparametric test for trend (changes
in filter effectiveness over time) indicated that there were no trends in
the yields and concentrations of TSS, NO3--N, NH4+-N, TKN, and FTKN for the
8.5 m filter over time. For the shorter 4.3 m filters, there were signific
ant upward trends in TKN yield and downward trends in TSS, NH4+-N, and FTKN
concentrations, indicating that trapping efficiency may have started chang
ing with time. The Kruskal-Wallis test indicated that the 8.5 m filters red
uced median yields and concentrations of TSS and all N species, but the 4.3
m filters only reduced the median yields and concentrations of TSS, NH4+-N
, TKN, and the median concentration of FTKN. The 8.5 and 4.3 m filters redu
ced contaminate yields and concentrations from 42 to 90 percent and from 20
to 83 percent, respectively.