Geoscience Reference
In-Depth Information
14
Mapping Pesticide
Partition Coefficients By
Electromagnetic Induction
Dan B. Jaynes
ContentS
14.1 Introduction ......................................................................................................................... 233
14.2 Materials and Methods........................................................................................................ 234
14.3 Results ................................................................................................................................. 236
14.4 Author's Note ...................................................................................................................... 239
References ...................................................................................................................................... 239
14.1 IntRodUCtIon
One of the difficulties in predicting the fate and leaching risk of field-applied pesticides is that their
transport properties are affected by soil properties that can vary greatly across the landscape. For
example, the affinity of a chemical to sorb to the soil matrix has been shown to vary spatially within
a field (Cambardella et al., 1994; Rao et al., 1986; Wood et al., 1987) and even within the same soil
map unit (Novak et al., 1997). Yet Loague et al. (1990) found that sorption can be the predominant
process contributing to the variability of pesticide mobility across the landscape.
The partitioning of an herbicide between the solution and soil phases is typically represented by
Freeze and Cherry (1979):
s = K
d
c
(14.1)
where
s
is the sorbed concentration on the soil (mg kg
−1
),
c
is the solution concentration (mg L
−1
), and
K
d
is the partition coefficient (L kg
−1
). Standard methods for measuring
K
d
(Novak et al., 1994) are
time consuming and require costly specialized equipment. Due to the expense and time required to
measure partition coefficients for combinations of specific soils and chemicals, the coefficients are
often estimated from related soil properties. In particular, organic carbon content is commonly used
because this soil fraction strongly interacts with nonionic herbicides (Bailey and White, 1970). The
relationship between soil organic carbon mass fraction, SOC (kg kg
−1
) and
K
d
can be expressed as
K
d
=
K
oc
SOC
(14.2)
where
K
oc
(L kg
−1
) is the coefficient of proportionality. Although K
d
is not perfectly correlated with
SOC because of additional sorption to the clay fraction (Laird et al., 1992), the K
oc
approach has
proven to be satisfactory for many purposes (Rao and Davidson, 1980). However, if we wish to char-
acterize the spatial variability of K
d
over a field or larger unit, even this approach is cumbersome
233
