A preliminary laboratory testing program was conducted to investigate
the potential effect of vegetation on the hydraulic conductivity of so
ils used to construct levee structures. The testing program was conduc
ted using a silty sand soil with a simulated root system. Testing appa
ratus included rigid-wall-double-ring permeameters and flexible-wall p
ermeameters. The simulated roots were made using Balsa wood having a s
quare cross section of 1.65 by 1.65 mm and approximately 12 to 25 mm i
n length. The measured hydraulic conductivity for specimens without si
mulated roots decreased as a function of the moisture content from app
roximately 1 X 10(-3) cm/s to 5 X 10(-4) cm/s. Values obtained using t
he flexible-wall device were less than those measured using the rigid-
wall device by a factor of approximately 2. The addition of 1% simulat
ed roots decreased the measured hydraulic conductivity, k. The reducti
on in k increased from less than 10% at a molding water content of 10%
to approximately 50% at a water content of 25%. Similar behavior was
observed in the case of specimens with 2% simulated roots. For gamma(d
ry) of 10 kN/m(3), the estimated k values were 1 X 10 5 cm/s for the c
ase of 2% roots versus 3 x 10(-4) cm/s for the case of 1% roots. The k
values decreased as the unit weight of the specimens was increased fo
r both cases. In addition, the difference between k values obtained fo
r specimens containing 1% and specimens containing 2% simulated roots
increased as the dry unit weight of the specimens was increased.