Stiff-grass hedges can resist retard, and disperse concentrated flows
of runoff; trap suspended sediment; and reduce ephemeral gully develop
ment. Flume experiments were conducted at a 5% grade using several com
binations of four grass species, four types of sediment, and eight flo
w rates ranging from 0.33 to 2.66 m(3)/min-m. Sediment trapping result
ed primarily from deposition in the backwater upstream of the grass, r
ather than by filtration in the grass. Sediment was initially deposite
d 1 to 2 m downslope from a hydraulic jump transition that formed at t
he upper edge of the backwater. From this position, a delta of sedimen
t mostly coarser than 125 mu m grew back toward the hydraulic jump unt
il flow depth became shallow enough that bedload transport was initiat
ed and the delta began to advance toward the grass. In the flow zone a
cross the delta, slope steepness approached 1 to 2% and the flow was n
ear critical depth. A steady-state model was developed that describes
sediment trapping in the backwater area as a settling process controll
ed by sediment characteristics, flow rate, and backwater depth at the
grass. This model underpredicted trapping of fine sediment and overpre
dicted trapping of coarse sediment in situations where the delta neare
d the grass.