Filter strips are widely prescribed to reduce contaminants in surface runof
f from agricultural fields. This study compared performance of different fi
lter strip designs on several contaminants and evaluated the contributing p
rocesses. Different vegetation types and widths were investigated using sim
ulated runoff event on large plots (3 m x 7.5 or 15 m) having fine-textured
soil and a 6 to 7% slope. Filter strips 7.5 and 15 m,vide downslope greatl
y reduced concentrations of sediment in runoff (76-93%) and contaminants st
rongly associated,with sediment (total P, 55-79%; permethrin, 27-83% [(3-ph
enoxyphenyl) methyl (+/-)-cis, trans-3-(2,2-dichloroethenyl)-2,2-dimethylcy
clopropanecarboxylate]). They had less effect on concentrations of primaril
y dissolved contaminants [atrazine, -5-43% (2-chloro-4-ethylamino-6-isoprop
ylamino-s-triazine); alachlor, 10-61% [2-chloro-2'6'-diethyl-(methoxymethyl
) acetanilide]; nitrate, 24-48%; dissolved P, 19-43%; bromide, 13-31%], Dil
ution of runoff by rainfall accounted for most of the reduction of concentr
ation of dissolved contaminants, Infiltration (36-82% of runoff volume) sub
stantially reduced the mass of contaminants exiting the filter strips. Doub
ling filter strip width from 7.5 to 15 m doubled infiltration and dilution,
but did not improve sediment settling, Young trees and shrubs planted in t
he lower one-half of otherwise grass strips had no impact on filter perform
ance. Compared with cultivated sorghum [Sorghum bicolor (L.) Moench] grass
clearly reduced concentrations of sediment and associated contaminants in r
unoff, but not volume of runoff and concentration of dissolved contaminants
. Settling, infiltration, and dilution processes can explain performance di
fferences among pollutant types and biter strip designs.