Particles and vegetation: Implications for the transfer of particle-bound organic contaminants to vegetation

Abstract

This paper presents a comprehensive review of the mechanisms responsible for the transfer of atmospheric particulate deposition and soil particulate re-suspension onto vegetation. The nature of atmospheric aerosols and dry/wet particulate deposition are reviewed, together with information from the literature on radionuclides as tracers of the air particle/soil particle to vegetation transfer processes. Information from these fields is used to make inferences about the potential significance of these pathways in supplying particle-bound semi-volatile organic chemicals (e.g. polycyclic aromatic hydrocarbons, polychlorinated dibenzo-p-dioxins and dibenzofurans, polychlorinated biphenyls) to vegetation. Retention of compounds on particles brought to the above-ground plant surfaces is discussed. In the absence of definitive field/experimental studies, calculations are made drawing on the literature data to estimate the contributions of atmospheric and soil particle-bound organic contaminants to the plant concentration. These show that depending on the site-specific, species-specific and compound-specific scenarios considered, particulate-bound inputs may be negligible or may dominate the supply of organic contaminants to the above-ground portion of plants. However, field/experimental studies and direct measurements are needed to provide reliable quantitative data on this topic.

Full text

Ž.
The Science of the Total Environment 246 2000 207᎐236
Particles and vegetation: implications for the transfer of
particle-bound organic contaminants to vegetation
Kilian E.C. Smith, Kevin C. JonesU
En¨ironmental Science Department, Institute of En¨ironmental and Natural Sciences, Lancaster Uni¨ersity, Lancaster LA1
4YQ, UK
Received 27 February 1999; accepted 21 October 1999
Abstract
This paper presents a comprehensive review of the mechanisms responsible for the transfer of atmospheric
particulate deposition and soil particulate re-suspension onto vegetation. The nature of atmospheric aerosols and
dryrwet particulate deposition are reviewed, together with information from the literature on radionuclides as
tracers of the air particlersoil particle to vegetation transfer processes. Information from these fields is used to make
inferences about the potential significance of these pathways in supplying particle-bound semi-volatile organic
Ž
chemicals e.g. polycyclic aromatic hydrocarbons, polychlorinated dibenzo-p-dioxins and dibenzofurans, polychlori-
.
nated biphenyls to vegetation. Retention of compounds on particles brought to the above-ground plant surfaces is
discussed. In the absence of definitive fieldrexperimental studies, calculations are made drawing on the literature
data to estimate the contributions of atmospheric and soil particle-bound organic contaminants to the plant
concentration. These show that depending on the site-specific, species-specific and compound-specific scenarios
considered, particulate-bound inputs may be negligible or may dominate the supply of organic contaminants to the
above-ground portion of plants. However, fieldrexperimental studies and direct measurements are needed to provide
reliable quantitative data on this topic. 䊚2000 Elsevier Science B.V. All rights reserved.
Keywords: Semi-volatile organic chemicals; Vegetation; Atmospheric particles; Soil particles
UCorresponding author. Tel.: q44-1524-593972; fax: q44-1524-593985.
Ž.
E-mail address: k.c.jones@lancaster.ac.uk K.C. Jones
0048-9697r00r$ - see front matter 䊚2000 Elsevier Science B.V. All rights reserved.
Ž.
PII: S 0 0 4 8 - 9 6 9 7 9 9 00459-3

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236208
1. Introduction
Vegetation plays a major role in the movement
Ž.
of semi-volatile organic chemicals SOCs through
the environment through its ability to retain them,
to modify air᎐soil exchange processes and to in-
troduce them into foodchains through grazing
Ž
animals Lorber et al., 1994; Simonich and Hites,
1994, 1995; Welsch-Pausch et al., 1995; McLach-
lan, 1996, 1997; Douben et al., 1997; Duarte-
.
Davidson et al., 1997; Wagrowski and Hites, 1997 .
The generic term SOCs includes a number of well
studied organic contaminants, such as polycyclic
Ž.
aromatic hydrocarbons PAHs , dibenzo-p-di-
Ž.
oxinsrdibenzofurans PCDDrFs and polychlori-
Ž.
nated biphenyls PCBs , as well as a variety of
other chemical classes such as the chlorinated
napthalenes, various brominated species, organo-
chlorine pesticides and so on.
There are a number of different pathways by
which SOCs may enter vegetation, summarised in
Fig. 1. These include uptake by roots and subse-
Ž
quent translocation in the xylem McCrady et al.,
1990; O’Connor et al., 1990; Wild and Jones,
1991, 1992; Ye et al., 1992; Schroll and Scheunert,
.
1993; Schroll et al., 1994; Nakamura et al., 1995 ,
volatilisation from the soil followed directly by
Ž
foliar uptake from the gaseous phase Trapp and
.
Matthies, 1997 or directly from the surrounding
atmosphere via wet or dry deposition. There is
also the possibility that some SOCs may be syn-
Ž
thesised naturally by the vegetation Edwards,
.
1983 . In addition, re-suspension of soil particles
by varied mechanisms such as the actions of wind
and rain may result in their subsequent capture
by vegetation. Should such re-suspended particles
be contaminated by SOCs, this may provide a
further route of plant contamination, which may
assume particular importance in those areas
where the soils are more heavily contaminated.
This process is of course not exclusive from the
atmospheric deposition described below. Soil par-
ticles may be re-suspended by wind into the atmo-
sphere only to later be removed by one of the
depositional processes and be retained by plant
surfaces.
Atmospheric deposition can therefore, poten-
tially be a route of contamination and hence it is
important to understand how the atmosphere and
vegetation interact. SOCs can exist in the air
either in the gas phase or associated with parti-
cles. Most SOC research on the relationships
Fig. 1. The uptake pathways of SOCs by vegetation.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 209
between the atmosphere and plant to date has
focused on modelling dry gaseous phase-transfer
Ž
processes Bacci et al., 1990a,b; McCrady and
Maggard, 1993; McCrady, 1994; Tolls and
McLachlan, 1994; Smith et al., 1995; Trapp and
.
Matthies, 1995 with little emphasis on the poten-
Ž
tial influence of dry Lorber, 1995; Chrostowski
.
and Foster, 1996; McLachlan, 1997 or wet parti-
cle-bound deposition.
An extensive literature on radioisotopes and
metals can be drawn on which suggests that atmo-
spheric particulate deposition can indeed be im-
portant in supplying particle associated contami-
nants to vegetation and that soil splashr
re-entrainment may also be important in certain
circumstances. The objectives of this review are
Ž.
therefore to: i briefly comment on the factors
and properties controlling the distribution of
SOCs between the gas and particle phases in the
Ž.
air; ii review the processes whereby particles
Ž
transfer to vegetation dry particle deposition, wet
.
particle deposition and direct soil contamination ;
Ž.
iii consider the influence of vegetation proper-
ties and type on the entrapment and washoff of
Ž.
particles; and iv relate the above information to
organic contaminants associated with particles in
order to investigate their potential to contami-
nate vegetation of various types.
2. Semi-volatile organic chemicals in the
atmosphere
SOCs can exist in the atmosphere either associ-
ated with the aerosol fraction or in the gas phase
Ž.
Fig. 2 . Generally, SOCs with subcooled liquid
vapour pressures lying between approximately
10y6and 10y2Pa will be distributed between
both the gas and particle phases at ambient tem-
Ž
peratures Eisenreich et al., 1981; Bidleman,
.
1988 . The proportion associated with atmo-
spheric particulate matter will depend on the
compound’s vapour pressure, the ambient air
temperature, the relative humidity, and the rate
Ž.
Fig. 2. Broad comparison of vapour pressures Pa nominally @ 25⬚C for a range of organic contaminants together with comments
on their typical atmospheric gasrparticle distributions.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236210
of exchange with the sorbing surface which will in
turn be influenced by the nature and availability
Ž
of the atmospheric aerosol Yamasaki et al., 1982;
Bidleman et al., 1986; Pankow, 1987; Bidleman,
1988; Cotham and Bidleman, 1992; Pankow et al.,
.
1993 . Some SOCs associated with airborne par-
Ž.
ticulates e.g. PAHs can exist as non-exchangea-
ble and exchangeable fractions. It is the amount
in this latter fraction which is in equilibrium with
Ž.
the vapour phase concentration Bidleman, 1988 .
Ž.
Other SOCs e.g. PCBs appear to display com-
pletely reversible temperature-controlled parti-
tioning between the gas and particle phases
Ž.
Simcik et al., 1998 .
The aerosol or total suspended particulate
Ž.
TSP loading of the atmosphere varies between
locations, with urban areas generally having
higher levels than rural ones. Typical values range
y3Ž
from 15 to 272 ␮g m for urban settings Pierce
and Katz, 1975; Katz and Chan, 1980; Yamasaki
et al., 1982; Ligocki et al., 1985b; Bidleman et al.,
1986; Shah et al., 1986; Ligocki and Pankow,
1989; Gardner et al., 1992, 1995; Koester and
.
Hites, 1992; Harner and Bidleman, 1998 and
19᎐64 ␮gm
y3for ruralrsemi-urban locations
ŽShah et al., 1986; Gardner et al., 1992, 1995;
.
Kaupp and McLachlan, 1999 . The TSP loading
will influence the proportion of SOC found in the
particulate phase and consequently the particle
associated contribution to the total plant con-
Ž.
tamination Chrostowski and Foster, 1996 .
The distribution of SOCs between the gas and
particle phases controls their atmospheric re-
moval and deposition to various natural surfaces.
Gas phase compounds are subject to sorption to
the plant surface andror uptake via the stomata.
Particle associated compounds may be retained
by the vegetation surface and if small enough also
taken up via the stomata. However, the processes
of interception and retention by the plant are
different for each phase. The link between the
SOC concentrations in these two phases in the
atmosphere and that measured in the vegetation
is difficult to study in practice using conventional
techniques for measuring wet, dry, and bulk depo-
sition because the interception and retention be-
haviour of gases and particles by the surrogate
surface may be very different from that of the
Ž.
plant surface Hicks, 1986 . It is therefore, neces-
sary to perform experiments with the vegetation
itself acting as the deposition collector in order to
gain insights into the processes of interception
and retention.
3. Atmospheric deposition
3.1. Atmospheric aerosols
Atmospheric aerosols occur in a range of sizes,
and can be operationally defined as having a
Ž
trimodal distribution Willeke and Whitby, 1975;
Whitby, 1978; Bidleman, 1988; Kao and
.
Venkataraman, 1995 .
Ž.
The smallest particles Aitken nuclei with di-
Ž. Ž .
ameters d-0.08 ␮m median df0.01 ␮m are
formed during combustion when hot vapour con-
denses to form small primary particles which can
then coagulate to form chain aggregates. These
particles are removed from the atmosphere by
interception because of their rapid Brownian mo-
tion. Although this mode contains most of the
number of particles, it contributes little to the
overall mass of a typical urban aerosol. Particles
grow out of this range mainly by coagulation into
larger aggregates. The atmospheric lifetimes of
Aitken nuclei are short, being of the order of 1 h,
because of this rapid coagulation.
The second size range is the ‘accumulation
mode’, consisting of particles 0.08 -d-2␮m
Ž.
median df0.3 ␮m . These are also formed by
gas to particle conversions and by coagulation of
Aitken nuclei. Once particles in this range reach
diameters of approximately 0.1 ␮m, removal by
diffusion is negligible and they continue to grow
by coagulation until they exceed 2.0 ␮m, when
removal by impaction and sedimentation become
significant. Those particles at the upper end of
the accumulation mode size range will deposit to
the ground before growing much larger. The
lifetime in the atmosphere of accumulation mode
particles may be up to several days. This mode
comprises most of the surface area and up to
approximately half of the total mass of the typical
urban aerosol. In terms of SOCs it is this mode
which has the highest concentrations when com-

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 211
Ž
pared to the coarse particle mode see for exam-
ple Kaupp et al., 1994; Schnelle et al., 1995,
.
1996 . The higher concentration of contaminant
on this particular particle size range does not
necessarily imply a greater contribution to the
depositional flux, because although larger parti-
cles may have lower concentrations they have
larger deposition velocities and are thus de-
posited faster.
The coarse mode consists of large particles
Ž.
with d)2.0᎐2.5 ␮m median df8␮m which
are produced mainly by mechanical means such
as the aeolian weathering of soils, sea-spray and
release from plants of pollen and spores. Coarse
mode particles settle out by sedimentation and
have lifetimes in the atmosphere ranging from a
few to many hours. As a result they tend to
deposit locally, close to their source, rather than
undergoing regional scale transport. SOCs associ-
ated with re-entrained soil particles would be
present in this fraction of the aerosol. There is
little exchange of mass between this mode and
the accumulation mode, with the result that the
two modes are chemically different. An interest-
Ž.
ing paper Kao and Venkataraman, 1995 investi-
gated the importance of soil re-entrained into the
atmosphere in contributing to atmospheric depo-
sition. The authors estimated that re-entrained
soil dust contributes approximately 4% to the
ambient PCDDrF concentrations. However, since
Ž.
soil dust particles d)2.5 ␮m have higher depo-
sition velocities than those arising from, e.g. com-
Ž.
bustion d-2.5 ␮m , they argued that this ‘old’
dioxin may account for up to 70᎐90% and 20᎐40%
of the total dioxin deposition in urban and rural
areas, respectively.
The major constituents of urban particulate
matter include organic and elemental carbon, sul-
phate, nitrate, ammonium, silicates, alkali and
alkaline earth metals, iron, and lead. The total
carbon content of the aerosols from urban and
rural air ranges from 12 to 30% of the total TSP
Ž
mass Ligocki et al., 1985b; Lioy and Daisey,
1986; Shah et al., 1986; Ligocki and Pankow,
.
1989 . This is of importance in terms of the
adsorptive capacity of the atmospheric aerosols
Ž
for hydrophobic organic chemicals Finizio et al.,
.
1997 . In addition, urban aerosols may be coated
or contain oil and grease which can influence
compound partitioning and particle coagulation
Ž.
behaviour Harner and Bidleman, 1998 .
3.2. Depositional processes
SOCs can be transferred from air to plant
surfaces by a number of different processes: dry
particle deposition, vapour exchange and the in-
terception of rain, snow or fog containing vapours
and particulates. Therefore, dry deposition in-
cludes both dry particle deposition and vapour
exchange, while wet deposition includes the re-
moval of vapours and particles by rain, snow or
fog.
3.2.1. Dry particle deposition
The processes of transport and deposition of
gases or particles to vegetation can be considered
Ž.
in four stages Chamberlain, 1986 :
䢇
Transport by wind in the free air.
䢇
Transport by eddy diffusion and sedimenta-
tion across the boundary layer to the canopy.
䢇
Transport by eddy and molecular diffusion,
interception, impaction and sedimentation to
the surfaces of the vegetation and soil.
䢇
Physical adherenceradsorption to the surface.
Therefore, the dry deposition of gases and parti-
cles involves four mechanisms: diffusion, inter-
ception, impaction and sedimentation. Where
there is the development of a boundary layer
above a vegetation canopy, the downward dry
depositional flux of both gases and particles can
be described mathematically by an equation of
the form:
Ž.
DsV⭈C1
dda
Žy2
where Dis the deposition rate or flux g cm
d
y1.
s,Vis the deposition velocity of the gas or
dŽy1.
particle cm s and Cis the air concentration
a
Žy3.
of the organic chemical g cm .
Vis an empirical parameter analogous to the
d
gravitational settling velocity of particles and

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236212
should strictly be defined relative to the height
above the surface at which the concentration is
measured, by convention set at approximately
1᎐1.5 m for land surfaces. Therefore, the parame-
ter Vprovides a useful indication of the poten-
dŽ
tial magnitude of particle flux to a surface Mc-
Mahon and Denison, 1979; Sehmel, 1980; Cham-
.
berlain, 1986 .
Submicron particles behave similarly to gases,
conforming almost perfectly to the streamlines of
flow in the atmosphere and even following small
eddies, with their diffusivity being independent of
density and dependent only on the size and shape
of the particle. As the diameter of a submicron
particle decreases its Brownian diffusion coeffi-
cient starts to approach that of a gas. However,
for such particles the deposition velocity is still
considerably lower than that of a gas, because
mass transfer at the interface involves mainly
diffusion. In addition, the wet deposition and
retention behaviours will be different for the gas
and particle phases. For such submicron particles
the nature of the receiving surface is not so
important. For aerosols in the 1᎐30-␮m size
range, the depositional velocity is made up of two
components: the velocity of deposition because of
impaction and the sedimentation velocity influ-
enced by gravity acting on the particle density and
diameter. For these particles the detailed struc-
ture of the surface roughness is all important
since it will influence the efficiency of impaction.
The deposition velocity of a particle will be
equal to or greater than the gravitational settling
velocity, which increases in proportion to the par-
ticle density and the square of the particle diame-
ter. Dry deposition is therefore, a function of
micrometeorological variables such as the atmo-
spheric turbulence and wind speed profile above
Ž
the surface of the vegetation deposition is ap-
.
proximately a linear function of the wind speed ;
the size, shape and other physical ᎐chemical
properties of the particle which will influence the
atmospheric friction velocity; and the features of
the receiving surface. The values for Vare mini-
d
mal for particles in the 0.1᎐1-␮m size range, i.e.
accumulation mode particles. For smaller parti-
cles, despite gravitational settling forces being
small, values of Vare higher because of Brow-
d
nian motion; while for larger particles they are
higher because of gravitational settling and turbu-
lent impaction. This is of potential significance
for SOCs since they are predominantly associated
Ž
with the 0.1᎐1-␮m size range this is further
.
discussed below . For a multimodal aerosol the
numerical value of Vis highly dependent on the
d
characteristics of the size distribution.
Because of the large number of variables af-
fecting the value of Vit is notoriously difficult
d,
to make generalizations about the value for a
particular particle size being deposited in differ-
ent situations, and the results quoted in the liter-
ature pertain to a specific set of circumstances.
Values obtained for particle Vtherefore range
d
over three orders of magnitude. For example, the
values of Vas measured specifically over grass-
d
land range from 0.02 to 3.5 cm sy1depending on
the conditions present in the different studies
ŽLittle and Wiffen, 1977; Wesley et al., 1977;
McMahon and Denison, 1979; Sehmel, 1980; Gar-
.
land and Cox, 1982 . Table 1 shows the variability
of this parameter for a range of different sized
aerosols deposited to various vegetation surfaces.
The average dry depositional velocity for
PCDDrF homologues to flat platesrfrisbees was
y1Ž
calculated to be 0.2 cm s Koester and Hites,
.
1992 , while the depositional velocity for individ-
ual congeners and homologue groups to bulk
Ž.
collectors wet and dry deposition at a regional
y1Ž
background site averaged 0.27 cm s Jones and
.
Duarte-Davidson, 1997 . The depositional veloci-
ties for particle-bound PAH compounds to a wa-
ter surface at a ruralrsemi-urban location were
y1Ž.
calculated to be 0.2 cm s Gardner et al., 1992 .
Using data from a study on the particle size
distribution of PCDDrFs and PAHs, calculations
were made to identify those particle size fractions
Ž
dominating dry particle-bound deposition Kaupp
.
and McLachlan, 1999 . For both compound classes
the dry deposition velocities for the particle-bound
substances were calculated at 0.05 cm sy1.De-
spite the low SOC loadings on particles of aerody-
Ž.
namic diameter d)1.35 ␮m, these larger par-
ae
ticles contributed 60᎐70% of the dry particle flux
Ž
in both summer and winter for the less volatile
PAHs, in summer these larger particles only con-
.
tributed approx. 20% to the flux . However, dry

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 213
Table 1
a
Selected particle dry deposition velocities to various vegetation surfaces
Vegetation surface Depositing material Particle diameter Deposition velocity
y1
Ž. Ž .
␮mcms
Grass Pb aerosol 0.03 0.13
Grass Au colloid particles 0.05᎐0.1 0.1᎐1.1
Grass Atmospheric aerosols 0.05᎐0.1 0.1᎐1.1
Grass Atmospheric aerosols 0.05᎐0.2 -0.1
Grass Pb aerosol 0.2 0.02
Grass ᎐0.1 0.03
Grass ᎐1 0.03
Grass ᎐2 0.1
Grass ᎐3 0.75
Grass ᎐4 1.1
Grass ᎐5 0.8
Grass CuSO 3᎐7 0.015᎐0.15
4
Grass Atmospheric aerosol 1 ᎐10 0.8
Grass ᎐16 0.5᎐2.1
Grass ᎐30 3.4
Grass Lycopodium spores 32 0.7᎐3.5
Sagebrush Zns 2.5 0.5
Sagebrush Zns 5 1.5᎐3.4
Nettle ᎐2.75 0.5
Nettle ᎐5 0.9
Nettle ᎐8.5 1.5
Beech ᎐2.75 0.04
Beech ᎐5 0.1
Beech ᎐8.5 0.3
White poplar ᎐2.75 0.3
White poplar ᎐5 0.3
White poplar ᎐8.5 0.8
Clover CuSO 3᎐7 0.018᎐0.15
4
Bean leaves ᎐0.8 0.004
Pine and oak shoots Uranine 2 0.003᎐10
aŽ. Ž. Ž. Ž.
Adapted from Little and Wiffen 1977 , Wesley et al. 1977 , McMahon and Denison 1979 , Sehmel 1980 and Garland and
Ž.
Cox 1982 .
particle deposition was found to contribute less
than 25% of the combined wet and dry particle
depositional fluxes. Interestingly the size distribu-
tions of PCDDrFs and PAHs were similar in
both the atmospheric particles and bulk particle
deposition. The calculated average gas and parti-
cle PCB dry depositional flux to smooth plates
y1Ž
was measured at 0.5 cm s range 0.4᎐0.6 cm
y1.Ž .
s Holsen et al., 1991 . The authors suggested
that particle associated PCBs contribute signifi-
cantly to the observed fluxes, despite the majority
of the compounds being in the gas phase. Again
with these studies it serves to emphasise the
problems of attempting to extrapolate the values
of Vto artificial surfaces as obtained here to
d
those actually occurring over vegetation since, for

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236214
example, the plant surface may be more or less
efficient at intercepting particulates and their as-
sociated SOCs.
3.2.2. Wet deposition
Wet deposition includes a number of processes,
e.g. the dissolution of gases and particles in cloud
or rain water, nucleation where an aerosol parti-
cle serves as a condensation nucleus for water,
Brownian capture occurring in clouds where ki-
netic motion brings the contaminant into contact
with cloud droplets, and impaction resulting from
the collision of rain drops and pollutant particles
Ž.
Schroeder and Lane, 1988 .
In the case of vapour dissolution in cloud or
rain water, the extent of dissolution can be de-
termined by the compound’s Henry’s law con-
Ž3y1.
stant, Hatm m mol . Vapour scavenging is
favoured by low Hvalues. For example, in studies
on PAHs it was found that those with two to four
rings were washed out as vapours whereas those
with greater than five rings were washed out as
Ž.
particles Ligocki et al., 1985a,b . However, if H
is sufficiently high, vapour dissolution in droplets
is negligible and only the particulate fraction is
removed by wet deposition. In the case of PCBs,
despite the fact that they occur mainly in the
vapour phase, wet deposition mainly removes the
particulate fraction as a result of their low aque-
Ž.
ous solubility Duinker and Bouchertall, 1989 .
Calculations using PCDDrF and PAH particle
size distributions indicated that approximately
80᎐90% of the particle-bound concentrations
observed in rain resulted from the removal of
Ž
particles of d-1.35 ␮m Kaupp and McLach-
ae
.
lan, 1999 .
The overall pollutant removal by wet deposi-
tion can be expressed empirically by the scaveng-
ing or washout ratio, W.
Žy3.Ž
y3.Ž.
Ws␮g m rain r␮g m air 2
Wis therefore composed of two components:
gas scavenging which is the equilibrium partition-
ing between the vapour and aqueous phases, and
particle scavenging which is the removal of parti-
cles from the atmosphere. A large Wtherefore,
results from a high water solubility, the efficient
scavenging of particles with a high sorbed concen-
tration, or both. However, there are considerable
uncertainties involved in using scavenging ratios
to represent the removal of pollutants by rain
both spatially and temporally, as it is assumed
there is a linear relationship between the air
concentration and that in the precipitation. This
Ž.
is probably not normally the case Barrie, 1992 .
The washout ratio for particles will depend on
Ž
the particle size and rainfall intensity Gatz and
Dingle, 1971; McMahon and Denison, 1979;
Lindberg, 1982; Eitzer and Hites, 1989; Koester
.
and Hites, 1992 . The concentrations of particle-
bound species scavenged by rain are dependent
on the rainfall amount, since most particles con-
taining pollutants are washed out of the atmo-
Table 2
Ž.
Scavenging ratios wfor SOCs
Compound Scavenging ratio Comments Reference
Ž.
PCDDrF homologue groups 9300᎐90 000 Gas and particle Eitzer and Hites 1989
PCDDrF homologue groups 12 000᎐72 000 Particle
Ž.
PCDDrF homologue groups 15 000᎐150 000 Gas and particle Koester and Hites 1992
PCDDrF homologue groups 16 000᎐100 000 Particle
Ž.
PAH compounds 1600᎐7000 Gas and particle Ligocki et al. 1985b
PAH compounds 1300᎐1700 Particle
Ž.
Individual PCB congeners 100᎐650 000 Gas and particle Duinker and Bouchertall 1989

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 215
sphere at the beginning of the rain event. Thus,
smaller precipitation amounts will be expected to
have higher particle-bound concentrations. Parti-
cle size is another important factor influencing
scavenging efficiency, with minimum scavenging
Ž.
around the 1 ␮m size range Radke et al., 1980 .
In terms of SOCs, this is significant in view of
their association predominantly with this size of
particles. Particle scavenging is difficult to predict
theoretically since it is a complex process depend-
ing on the meteorological conditions in the cloud
along with the chemical and physical properties
of the aerosol. The scavenging ratios for a range
of SOCs are given in Table 2.
It can be seen therefore, that the wet depositio-
nal flux will depend to some extent on the season
via variations in the amount of precipitation.
However, there is little knowledge of the interac-
tions between dissolved gaseous or particulate
SOCs in wet deposition and plant surfaces, and
how this is related to the different scavenging
ratios for SOCs. This pathway is seldom con-
sidered when looking at the contamination of
vegetation.
3.3. SOC particle size distribution
In order to understand the deposition and sub-
sequent behaviour of SOCs associated with atmo-
spheric aerosols it is clearly important from the
discussion above to have a knowledge of the
particle size distribution of SOCs in the atmo-
sphere.
The PCDDrF particle size distribution was
studied during summer in a rural atmosphere
Ž.
Kaupp et al., 1994 and ;90% of the airborne
PCDDrF was found to be associated with parti-
cles of d-1.35 ␮m, i.e. accumulation mode
ae
particles. The highest concentrations of
Žy1.
ÝPCDDrF;64 ng g occurred on particles
d-0.45 ␮m. Comparison of these concentra-
ae
tions with those of uncontaminated bulk soil
shows that they are between a factor of 20 and
100 times higher. These very fine particles will
also have longer residence times on the plant
surface. The PCDDrF homologue group patterns
were similar on all of the particle size classes
during any particular sampling run, implying that
the particulate depositional processes will be
w
similar for all of the PCDDrFs studied indeed
Ž.
according to Koester and Hites 1992 , particle-
bound PCDDrFs all appear to be removed from
the atmosphere with similar efficiencies, suggest-
ing that they are all bound to similar particle size
xŽ
classes . A further study Kaupp and McLachlan,
.
1999 also indicated that )80% of PCDDrFs
were associated with particles d-1.35 ␮min
ae
winter and summer. Again the particle size dis-
tributions were similar for each of the individual
compounds, with an increase in the particle asso-
ciated ÝPCDDrF concentrations during the win-
ter. The distribution of PCDDrFs on various
sizes of airborne particles in rural, urban and
industrial atmospheres was investigated by
Ž.
Kurokawa et al. 1996 . Between 50 and 60% and
70᎐80% of ÝPCDDrF were associated with par-
ticles d-1.1 and-2␮m, respectively. In con-
ae
trast to the findings above, most of the hexa- to
octa-CDDrFs existed on small particles, whereas
the less chlorinated PCDDrFs were widely dis-
tributed on all particle sizes. The larger particles
therefore had a pattern similar to that of gas
phase PCDDrFs.
Ž
One of the older studies Pierce and Katz,
.
1975 on 10 PAH compounds associated with the
aerosol fraction, suggested that 85᎐90% of the
ÝPAHs are associated with particles -5.0 ␮m
diameter in winter, with 70᎐85% in summer. A
study on the particle size distribution of three
PAHs in city and suburban atmospheres indicated
that they are associated with particles of mass
Ž. Ž
median diameter MMD ;0.5 ␮m Butler and
.
Crossley, 1979 . Both sites had similar aerosol
characteristics, with a MMD of total particulates
of ;1␮m. More than 50% of the PAHs were
associated with particles -1␮m, while 70᎐90%
were associated with particles -3␮m. Katz and
Ž.
Chan 1980 indicated that 72᎐79% of PAHs were
associated with particles in the size range -
1.3᎐3.3 ␮m, while 88 ᎐95% were associated with
particles -1.1᎐7.0 ␮m. These authors went on to
conclude that although particles )10 ␮m con-
tribute significantly to the aerosol mass they are
unimportant to the ÝPAH burden. A study by
Ž.
Schnelle et al. 1995 indicated that 50᎐60% of

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236216
the ÝPAHs are associated with particles -0.49
␮m and that )85% are associated with particles
Ž
-1.5᎐1.72 ␮m the particle size cut-off point
.
depending on the sampling system . The lower
Ž.
molecular weight MPAHs increased in rela-
w
tive concentration with increasing particle size,
while some of the higher Mones decreased in
w
relative concentration with increasing particle
Ž
size. The rest of the PAHs mainly the highest
.
Mcompounds showed no dependence of rela-
w
tive concentration on particle size. In another
study the particle size dependence of 11 PAHs in
the outskirts of Munich were investigated
Ž.
Schnelle et al., 1996 . Different sampling meth-
ods indicated that between 50 and 60% of the
ÝPAH was associated with particles of d-
Ž
0.5᎐0.66 ␮m again depending on the sampling
.
method employed . There was a unimodal dis-
tribution with a mean diameter of ÝPAH concen-
tration in the range 0.26᎐0.42 ␮m. A decrease in
the mean diameter of ÝPAH concentration was
observed with an increase in the compound’s boil-
ing point. The maximum PAH concentration on
particles occurred in the range 0.13᎐0.50 ␮m. A
Ž.
study by Kaupp and McLachlan 1999 found
)80% of PAHs on particles with d-1.35 ␮m,
ae
with particle associated concentrations increasing
in winter. There was a similar particle size dis-
tribution for the individual PAHs during winter
and summer, except for phenanthrene, fluoran-
thene and pyrene which were relatively enriched
on the larger particle sizes during summer.
For an urban and industrial site, the particle
size distributions of ÝPCB and TSP were both
Ž.
bimodal Chen et al., 1996 . The urban site had
Žy3.
the highest ÝPCB concentrations ng m in the
size ranges 0.32᎐0.56 ␮m and 5.6᎐10 ␮m, the
major peak in concentration occurring in the
larger of the two size ranges. The mass median
diameters of the TSP and ÝPCB concentration
were 3.99 and 5.39 ␮m, respectively. Thirty-five
percent of PCBs were associated with particles of
d-2.5 ␮m and 80% with particles of d-10.0
␮m. The industrial site had the highest ÝPCB
Žy3.
concentrations ng m in the size ranges
0.056᎐0.1 ␮m and 1.0᎐1.8 ␮m, with the major
peak in concentration occurring in the smaller of
the two size ranges. The mass median diameters
of the TSP and ÝPCB concentration were 1.06
and 1.07 ␮m, respectively. Seventy-four percent
of PCBs were associated with particles of d-2.5
␮m and 91% with particles of d-10.0 ␮m.
Therefore, compared to the industrial site, PCBs
in the urban site were increasingly associated with
the coarse aerosol. A study measuring the dry
depositional flux of PCBs in an urban setting
found them to be associated with all sizes of
Ž.
atmospheric aerosols Holsen et al., 1991 . The
PCB concentration normalised to particle mass
decreased with increasing particle size as a result
of the corresponding decrease in surface to
volume ratio. This concentration decreased from
50 ␮gg
y1for particles -1␮m to approximately
30 ␮gg
y1for particles )25 ␮m. Comparison of
these concentrations with those of semi-rural
wy1
agricultural bulk soils 0.011᎐0.054 ␮gg ;Al-
Ž.x
cock et al. 1993 , shows them to be approxi-
mately three orders of magnitude higher.
It therefore appears that particulate phase
PCDDrFs and PAHs are primarily associated
with particles of approximately 1 ␮m in diameter.
Consequently, it is primarily the interception and
retention behaviour of this size range of particles
on plant surfaces that needs to be considered.
This is indeed fortunate, since a similar size range
of particles has been found to be associated with
the plutonium contamination of vegetation via
Ž
atmospheric deposition Gay and Watts, 1981;
.
Pinder et al., 1990 . Reference to these studies
might therefore allow inferences to be made re-
garding SOC behaviour. Caution is needed how-
ever, as there may be differences in both the
chemical nature of the particles which SOCs and
radionuclides associate with and also the mecha-
nisms by which the SOCs and radioactive activity
are bound to the particles. Interestingly however,
in studies using a range of radioactive aerosol
species the physical or chemical form of the activ-
ity was found to have little influence on the
interception characteristics of the grass sward
Ž.
Chamberlain, 1970, 1986 . In contrast, PCBs ap-
pear to be associated with a wider range of parti-
cles than the combustion derived PAHs and
PCDDrFs. This might result in a greater varia-
tion in their depositional and interception be-
haviour.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 217
4. Plutonium as a tracer for particle movement
Ž.
Plutonium Pu can exist in the environment in
a number of forms: as large fragments when
associated with soil, submicron particulates in
aerosols or in aqueous solution. The Pu contami-
nation of vegetation can result from root uptake
and translocation, the interception of atmo-
spheric deposition or the impaction of re-
suspended soil particles arising from wind action,
raindrop splash, mechanical disturbance or ani-
mal activity. Pu can be used as an effective tracer
of atmospheric deposition and soil movement as
it is poorly absorbed by plant roots, which ensures
that the contamination of the vegetation is surfi-
cial because of physical transport processes
Ž.
Romney et al., 1981; Adriano et al., 1986 . Pu is
relatively insoluble and therefore not readily
leached from transported particles and it can be
measured at low concentrations. This allows for
the detection of low concentrations of deposi-
Ž
tionrsoil on the plant biomass McLeod et al.,
1980; Adriano et al., 1982; Pinder and McLeod,
.
1988, 1989 . However, a precise knowledge of the
nature of Pu particles in both atmospheric fallout
and the soil is needed before general conclusions
about the behaviour of SOCs associated with
their own specific particle types can be made.
Fallout Pu enters the terrestrial environment
as oxides of 238 Pu and 239r240 Pu which are rela-
Ž.
tively stable and insoluble Hakonson, 1975 . As
discussed above, the size range of Pu particles in
the atmosphere and that with which SOCs are
associated are similar, although questions remain
about differences in the chemical nature of the
particles which Pu and SOCs associate with.
Understanding the contamination of vegetation
by Pu present in the soil requires a knowledge of
the association with different types and sizes of
soil particles. A study on four soils separated into
Ž
different particle size fractions Livens and Bax-
.
ter, 1988 showed the radionuclides investigated
to be concentrated in the finer size fractions. The
Ž.
clay sized particles -2␮m had Pu levels up to
40 times higher than those of the bulk samples.
Since the physical transport processes operate
preferentially on these smaller particles, those
re-suspended by wind or rain will have a higher
specific activity than the material from which they
are derived. This will have an influence on the
calculation of soil mass loadings, a fact that needs
to be taken into consideration when comparing
studies. This is similar to the situations for PAHs
and PCBs, where the majority of the SOC burden
was found to be associated with the finer particle
Ž
size ranges Guggenberger et al., 1996; Wilcke
.
and Zech, 1998 . The case for other SOCs, e.g.
PCDDrFs, is unknown.
A further advantage of using Pu is that the
change over time in the isotopic ratio of 23 8 Pu
and 239r240 Pu in atmospheric deposition has al-
lowed for the different processes of atmospheric
particulate deposition and soil re-suspension to
Ž
vegetation to be investigated see for example
.
Pinder and McLeod, 1989; Pinder et al., 1990 .
These data provide an insight into the two
processes of interception of atmospheric deposi-
tion and soil re-suspension, and how they change
relative to one another with the progression of
the growing season for a range of different plant
species. This may be particularly relevant for crop
species which change their canopy structure and
have a significant increase in biomass as the
growing season progresses. Initially, soil re-sus-
pension from the exposed soil may be the more
important route of contamination but as the
Ž
canopy biomass increases both increasing the
interception of atmospheric deposition and pro-
tecting the soil from the erosive effects of wind
.
and rain , atmospheric deposition may increase in
relative importance. For grazed pasture there is a
more constant biomass throughout the growing
season, and the relative contribution of these two
processes may be expected to remain more or less
constant.
5. Particulate contamination of vegetation
The contribution of particle bound SOCs to the
overall vegetation SOC burden will be a product
of the particle deposition velocity, the substance
concentration on the particle, the residence time
of the particle on the vegetation surface and the
transfer of the substance from the particle to the
Ž
leaf during this period Umlauf and McLachlan,

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236218
.
1994 . The rate of transfer will be determined by
the relative absorptive strengths of the particle
and the vegetative surface for the SOC, as well as
the contact surface area. These parameters will
be much influenced by the presence of surface
films. The concentration of contaminant associ-
ated with the particle, or more specifically the
concentrations of the exchangeablernon-ex-
changeable fractions and the distribution of the
chemical within the particle, may be useful for
determining the diffusion gradient and therefore,
the rate of transfer. There is as yet no informa-
tion available concerning the transfer of SOCs
from particles to the leaf. However, even con-
taminated particles that are only surficially at-
tached to the leaf surface will contribute to the
contamination. In addition, it appears that some
particles trapped by vegetation remain perma-
nently associated with the leaf surface and are
Ž
not removed by weathering this is discussed fur-
.
ther below .
The extent of particulate contamination of veg-
etation is a complex process involving different
types of particles in atmospheric deposition and
soil re-suspension, varying climatic conditions and
variations in the nature of the receiving surface
we.g. area of surface exposed, external morphology
of the above-ground parts, developmental season
Ž.x
of the plant Sawidis, 1988 . Grasses appear to
have higher concentrations of particulates on their
surfaces compared to herbs, shrubs and trees
Ž.
Hakonson, 1975; Brabec et al., 1981 . This is
perhaps because they have a significant propor-
tion of their standing biomass from previous years’
dead growth and therefore, a significant propor-
tion of deposition and re-suspension from the
previous year. It may also be the case that this
dead material has a higher retention. There is
therefore considerable uncertainty about many of
the factors involved in the particulate contamina-
tion of vegetation.
In the case of atmospheric contamination it is
assumed that when the particulate deposition of a
pollutant is constant, the concentration in the
plant will increase and gradually level off as an
equilibrium is established between interception of
Ž.
deposition and removal Chamberlain, 1986 .
Ž.
However, a study Brabec et al., 1981 on three
Ž
vegetation stands maize, barley and a semi-natu-
.
ral grassland made during the entire growing
season found that instead of an equilibrium being
reached between deposit interception and re-
moval, there was a seasonal accumulation of de-
posited particles which was only partly explained
by the increase in biomass resulting in increased
interception. The authors suggest that initially the
particles are distributed uniformly on the plant
surface by the processes of sedimentation and
impaction where some, especially the larger sizes
and lighter organic ones, are easily eroded by the
action of wind and rain. For some there occurs a
rain induced translocation to various cavities, e.g.
leaf axils where the particle deposition is resistant
to wet and dry removal.
Vegetation is selective in the size of particles
that are intercepted and retained, with particles
of sizes larger than approximately 100 ␮m not
Ž
usually found on the surface. A study Romney et
.
al., 1963 investigating the fallout particles lodg-
ing on different types of plant foliage found them
to be predominantly of d-100 ␮m. The plant
foliage was selective in trapping andror retaining
particles of d-44 ␮m. The action of wind re-
moved ;23% of the foliar contamination from
alfalfa over a 2-day period, with all particles of
d)44 ␮m being removed. Foliage which was not
exposed to wind had ;5% of the fallout on the
surface as particles of d)44 ␮m. In addition, the
nature of the plant surface had an influence on
the retention behaviour, with washing removing
particles more easily from smooth leaf surfaces
than sticky, resinous, or hairy surfaces. Particle
size spectra for three types of herbaceous stands
Ž.
alfalfa, maize and a grass stand were investi-
Ž.
gated by Kovar et al. 1987 . There was found to
´
be a remarkable degree of variation in the range
of particle sizes both between and within stands,
caused by a number of mainly meteorological
factors influencing the conditions within the stand.
The plant material tested for had particles largely
in the supermicron size category, with particles
just )1␮m being most frequent in incidence. A
study investigating material )1.4 ␮m accumulat-
ing on tree foliage indicated that a large propor-
tion of the particles were organic in origin, e.g.
spores, bacteria, pollen grains, seeds, etc. Those

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 219
inorganic particles contained Si and Al, suggest-
ing that they were soil derived. Particles -10 ␮m
made up the majority of the particle numbers,
with amounts varying between 57 and 89%. Mean
Ž
sizes ranged from 5.8 to 12.7 ␮m Freer-Smith et
.
al., 1997 .
5.1. Interception of particulate deposition
The loading of atmospherically deposited parti-
cles on vegetation at any given time is a function
of the initial interception and subsequent loss;
and will be influenced by factors such as the
particle size, wind speed, and the morphology and
physiology of the vegetation. These in turn will
Ž
depend on the location affecting the meteorolog-
.Ž
ical conditions and season affecting vegetation
.
growth and dying-back . Grazing and weather
conditions will also affect the rates of field loss
ŽChamberlain, 1970; Chamberlain and Little,
.
1981 .
The proportion of the particulate deposition
initially retained by the vegetation, p, is assumed
w2Ž
to be related to an uptake coefficient ␮mkg
.y1x
DW by a relation of the form:
Ž. Ž.
ps1yexpy␮w3
wŽ.
y2x
where wis the herbage density kg DW m . It
can be seen that the interception fraction will
increase as the plant grows. For grassland, experi-
mental values of ␮lie in the range 2.3᎐3.3 m2kg
DWy1. In the case of radioactive aerosol deposi-
tion the value of ␮appears to be independent of
the physical or chemical form of the deposited
radioactive activity. This has the advantage of
allowing some estimate of a probable value of ␮
to be made when the nature of the activity is
Ž.
variable or uncertain Chamberlain, 1970 . The
use of interception fractions for vegetation re-
quires that the total amount of deposition to a
specified area of ground has been determined.
This may prove to be a particularly difficult
parameter to measure. For example, should the
deposition velocity to the grass above a specified
area of ground be measured and not the total
deposition onto that ground area, then it follows
that the value of pused should be set to 1.
A comprehensive review of the interception
fractions of wet and dry deposited radioactive
aerosols by different vegetation types has been
Ž.
compiled by Miller 1980 . The interception frac-
tion is related to the density of the vegetation w
as described by the equation above, with an in-
crease in the biomass of the vegetation resulting
in increased interception. For forage grasses the
pvalues were found to be approximately nor-
mally distributed with mean s0.47, range s
0.02᎐0.82 and S.D.s0.30. The prwratio had an
approximately log-normal distribution. Normalis-
ing the interception fraction to biomass gives a
mass interception factor equal to the fraction of
the total deposit intercepted and initially retained
per unit mass of vegetation. This value decreases
with increasing biomass as a result of tissue growth
dilution. For non-grassland types of vegetation it
was not possible to examine the distribution of
the pvalues because of the limited size of the
data set.
A study using 20 3 Pb labelled aerosols showed
that rough or hairy leaf surfaces were up to
approximately eight times more efficient at col-
lecting the aerosols than the smooth leaf surfaces
Ž.
Little and Wiffen, 1977 . The authors attributed
this to the increased surface area for eddy diffu-
sivity and the projection of roughness through the
boundary layer. For particles in the 1.0᎐5.0 ␮m
size range, deposition by impaction is inefficient
and the presence of fine hairs on the vegetation
may be of importance. For submicron particles,
eddy diffusivity is the main mechanism of transfer
and the nature of the surface is perhaps not as
important as in the case of larger particles. The
leaf laminae, petioles and stems all had different
collecting efficiencies as a result of their different
size, shape and surface texture. There were dif-
ferent collecting efficiencies between non-senes-
cent and senescent leaves. For some species
senescence resulted in an increased collection
efficiency while for others the reverse was true.
Therefore, the collection efficiency of a particular
vegetation type may change during the growing
season. The presence of grass increased the total

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236220
Ž.
deposition of Pb MMD 0.03᎐0.2 ␮m per unit
area of ground by three to four times as com-
pared to bare soil, with the grass collecting
70᎐90% of the total Pb intercepted.
The deposition velocities for 14 1 Ce and 13 4 Cs
aerosols were investigated for Artemisia tridentata
Ž. Ž
Big sagebrush and Elymus elymoides Squirrel-
.Ž .
tail bottlebrush Fraley et al., 1993 . Here for
example, since the deposition velocities were
measured directly to the vegetation surfaces, the
value of pshould be set to 1. The aerosols of the
two radionuclides were both of approximately 1
␮m activity median aerodynamic diameter. There
were no differences between the aerosols in their
velocities of deposition or retention half lives, but
species differences were significant. A.tridentata
had a higher Vvalue as a result of its larger and
d
pubescent leaf surfaces, and also the greater
height of the plant resulting in increased atmo-
spheric turbulence.
Simulated rain containing 3, 9 and 25-␮m
polystyrene microspheres resulted in grass inter-
Ž
ception fractions calculated for total vegetation
.
)5 cm above the soil surface ranging from 0.1
Ž.Ž
to 0.6 geometric mean ;0.3 Hoffman et al.,
.
1992 . There was an increased dependence of this
interception fraction on biomass rather than on
the rain amount. In the case of pasture, biomass
is an indicator of the leaf area and canopy cover.
The influence of the leaf texture and structure
was also found to have less of an effect. The
authors postulate that there is rapid settling out
of the particles from the rain droplets followed by
adsorption to the plant surface.
5.2. Soil particle mo¨ement
The soil contamination of vegetation can be
expressed in two different forms: as concentra-
Žy1.Ž
tions g soil g plant tissue or as inventories g
y2.
soil m land . The first term is particularly in-
fluenced by the change in biomass of the vegeta-
tion as the growing season progresses, which
makes it difficult to make observations on the
processes of interception and retention. In addi-
tion, the soil contamination of vegetation occurs
primarily below a height of 0.5 m and conse-
quently the soil concentration will be diluted by
any mass of vegetation above 0.5 m which will be
relatively free of soil. The use of inventories
allows the comparison of different crop species
without different crop masses and canopy struc-
tures complicating the issue. However, in terms of
exposure assessment models, the use of plant
concentrations may be more appropriate.
Soil may move to plant surfaces by the actions
of raindrop splash or wind erosion. There are also
the effects of mechanical disturbance and live-
stock activity. For crops there is also the specific
scenario of soil contamination during harvest.
Ž.
Large particles 53᎐105 ␮m may remain on the
surface for several days while the smaller parti-
Ž.
cles -53 ␮m may remain for over a week
Ž.
Dreicer et al., 1984 . The importance of soil
retention with respect to contamination is in-
creased in situations where the contaminant is
preferentially sorbed to smaller and more readily
transported soil particles, and also preferentially
concentrated in the surface soils. Unfortunately,
estimates of mass loadings are not available for
most crops, soil types, or climatic regions. How-
ever, the soil type appears to have little influence
on the amount of soil found on the vegetation
Ž.
Pinder and McLeod, 1989 . Most data estimating
mass loadings come from the US and lie in the
region of ;1to)200 mg gy1depending on the
Ž.
vegetation type Table 3 . These variations result
from differences in plant morphologies, with the
largest values on those broad leaved species
Ž.
growing close to the ground Pinder et al., 1991 .
Pasture crops appear to have increased mass
loading levels relative to row crops. One might
expect the litter cover found in pastures to reduce
the effects of wind and rain. However, this may
be offset by the fact that the vegetation is growing
close to the ground surface, and mass loadings
have been found to decrease with increasing plant
height. There is also the potential influence of the
trampling activities of grazing animals affecting
the level of soil contamination. Rapid fluctuations
in mass loadings might also be expected because
of rain and wind events. It may be that the
differences in mass loadings between the various
vegetation types merely reflect differences in cli-
mate between regions. Few of the studies quanti-
fying soil contamination identify the mechanisms

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 221
Table 3
a
Mass loadings on different vegetation types
Vegetation type Concentration Inventory Comments
y1y2
ŽŽ
mg soil g g soil m
..
plant DW land
US meadow 18
UK pasture 50᎐200 Cattle present
Herbage 70
Live bahia grass 9 "1.7 2.9 "0.6 Inventory and concentration higher in dead leaf layer
Winter wheat 4.3 1.3 Average during growing season
Maize 1.7 1.2 Average during growing season
Soybean 8.4 1.4 Average during growing season
Sunflower 2.6"0.8 0.79"0.29
Bush beans 30᎐60 Whole plant
Broccoli 10"8.1 0.19"0.16 At maturity
Tomatoes 17 1.3 Lower 0.4 m. Repeated rainfall.
Turnip greens 32"11 1.1"0.4 At maturity
Cabbage 1.1"1.1 0.03"0.03 At maturity
Lettuce 260"100 5.7"2.6 At maturity
Alfalfa 1.12
Tobacco 2.1"0.6 0.26"0.08
aŽ. Ž. Ž. Ž. Ž .
Adapted from McLeod et al. 1980 , White et al. 1981 , Adriano et al. 1982 , Dreicer et al. 1984 , McLeod et al. 1984a,b ,
Ž. Ž. Ž. Ž .
Summerling et al. 1984 , Kovar and Mejstrik 1987 , Green and Dodd 1988 , Pinder and McLeod 1988, 1989 and Pinder et al.
´
Ž.
1991 .
responsible for the differences in mass loadings
between the crops. In summary therefore, plants
growing close to the surface, those with growth
forms that are effective at intercepting soil parti-
cles, or those which are contaminated by the
trampling activities of grazing animals will have
the highest concentrations of soil.
There is a paucity of information on SOCs
associated with different types and sizes of soil
particles. The nature of the soil particles involved,
in particular the size, will have a profound impact
on such processes as re-entrainment into the at-
mosphere by wind, the contamination of vegeta-
tion by raindrop splash, and once intercepted by
the plant the degree of retention. In terms of
re-suspension of soil, it will be the surface of the
soil with its characteristic particle size distribu-
tion and associated SOC concentrations that will
influence its relative importance to the contami-
nation of vegetation. Additionally, the precise na-
ture of the association of the SOC with different
types of soil particles may result in varying ex-
changeable and non-exchangeable fractions,
which will in turn result in different contamina-
tion behaviours. For example, the exchangeable
fraction may partition to the plant cuticle, while
the non-exchangeable fraction may remain associ-
ated with the particle. In the latter instance the
contamination of the plant surface will be tem-
porary as the particles are subjected to normal
weathering processes. There is also little informa-
tion on the behaviour in soil of atmospherically
derived particle associated SOCs with their char-
acteristically higher concentrations. The down-
ward migration of such particles, together with
their possible association with other particles in
the soil, will affect their susceptibility to re-sus-
pension and therefore their potential to subse-
quently contaminate vegetation. The lack of such
fundamental information makes it difficult to un-
derstand and evaluate the extent of SOC soil
particle contamination of vegetation.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236222
A study investigating the behaviour of PAHs in
the mineral horizons of different forest soils
Ž.
Guggenberger et al., 1996 found 80% of the
Ž.
PAH burden to be in the silt 2᎐20 ␮m and
Ž.
coarse clay 0.2᎐2␮m fractions, irrespective of
the soil texture. In terms of the direct soil con-
tamination of plants it is these particles which are
preferentially transported to and retained by the
vegetation. Silt generally showed the highest con-
centrations of organic carbon related PAH con-
centrations, and there was an increase in the
proportion of the high MPAHs as the size of
w
the separates decreased. The PCB concentrations
of soil in a rural area were investigated by Wilcke
Ž.
and Zech 1998 . PCB concentrations generally
increased as the particle size decreased, although
when normalised to organic carbon the fine sand
Ž.
fraction 20᎐250 ␮m had the highest concentra-
tions. There was a slight increase in the relative
proportion of the lower chlorinated PCBs as the
particle size decreased. This situation differs from
that of the PAHs described above in that the
PCB distribution was not as homogenous and
they were associated with a larger size range of
particles. Should this trend of increasing concen-
tration with decreasing soil particle size also be
the case for other SOCs, then the use of the bulk
soil concentration to calculate the impact of soil
in contaminating the vegetation will result in an
underestimation of its contribution, since this
concentration will be lower than that of the finer
particles that are preferentially transported to
and retained by the plants.
5.2.1. Wind
Wind driven soil movement occurs by surface
creep, saltation and suspension, and is affected by
Ž.
soil moisture and texture Bisal and Hsieh, 1966 .
Particles moving by surface creep can be large,
with diameters of up to 2 mm. Saltation is the
most important process initiating and sustaining
soil movement and acts mainly on particles in the
diameter range 50᎐500 ␮m. Those particles mov-
ing by suspension are small, with diameters -100
␮m, and can be transported significant distances.
The greatest mass of eroding soil is moved by
saltation or surface creep. Therefore, depending
on the soil particle size and wind velocity, wind
may either suspend particles or cause them to
bounce along the soil surface by surface creep or
Ž.
saltation Anspaugh et al., 1975 . Such particles
Ž
can then settle onto plant surfaces Pinder and
.
McLeod, 1988 .
5.2.2. Rain
One of the processes by which soil particles can
be transported to the surfaces of plants is via the
action of rain. Here the particle size distribution
on the surface of the soil is all important. Soil loss
is related to the total rainfall energy and intensity
Ž.
Wischmeier and Smith, 1958 . Particle detach-
ment will also be influenced by the particle size,
type and density; the type of aggregate; or the
Ž
presence or absence of overland flow Mazurak
.
and Mosher, 1968, 1970; Farmer, 1973 .
Ž.
Dreicer et al. 1984 investigated the soil partic-
ulate contamination of tomato plants by rain-
splash. No particles of d)105 ␮m were found on
the plants. It is possible that larger particles were
trapped by the vegetation but washed off more
rapidly and therefore not detected so readily.
Most of the soil was not splashed higher than 40
cm in height. Approximately 1.3 g my2of soil was
Ž
observed on the lower 0.4 m of the plants Pinder
.
and McLeod, 1988 . A linear relationship was
found between the concentration of particles d-
53 ␮m and certain rainfall characteristics ᎏthe
storm intensity and average kinetic energy. After
a single storm event, the differences in soil found
on the tomato plants between the experimental
and control plots were only significant for the
Ž.
silt-clay size fraction d-53 ␮m . The concentra-
tions of silt-clay were related to height and storm
characteristics, but not to the plant surface area
or canopy cover. A single storm resulted in soil
concentrations of up to 1.6 mg gy1plant WW. In
Ž.
this instance larger particles d)53 ␮m were
either not re-suspended or not retained. After a
number of storms contamination was by particles
of up to 105 ␮m size, with up to 2.5 mg gy1plant
WW of both the silty-clay and fine sand fractions
being detected. It is of course the concentration
of SOC in each of these soil particle size classes
that will determine the importance of their con-
tribution to the total contamination. Scanning
electron microscope analysis showed )99% of

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 223
the particles to be of d-55 ␮m, and to be
associated with the lower vegetation strata. Since
soil particles can only be transported locally by
raindrop impact, this route of contamination will
be of particular importance for low growing
species.
5.3. Retention
Removal of particle-bound SOCs from the
above-ground parts of plants may occur by leach-
ing; removal of particles by wind, rain or other
disturbance; and the loss of plant parts such as
the shedding of cuticle and the die-back of leaves.
Removal will therefore, be influenced by the
physical and chemical form of the contaminant,
vegetation type and growth form, climate, and
Ž
season Chamberlain, 1970, 1986; Miller and
.
Hoffman, 1983 . Of particular relevance for SOCs
there is also the possibility of photolytic degrada-
Ž.
tion e.g. McCrady and Maggard, 1993 or the
revolatilisation into the surrounding air of SOCs
sorbed to the particles.
Ž.Ž
The environmental half-life Ti.e. the
w
amount of time for one half of the deposited
particles on the vegetation to be removed by
.
environmental processes provides an indication
of the degree of retention. A knowledge of Tw
values for different types and sizes of particles on
various vegetation types is essential. Of impor-
tance when comparing the values obtained from
different studies is the experimental procedure,
and whether the contamination is of an acute or
chronic nature. In an acute deposition situation,
one might expect the particle concentrations on
the plant to decrease with time as a result of the
various weathering processes until a residual level
is reached. However, in cases of continuous depo-
sition, there may result an equilibrium situation
with interception being balanced by removal. The
contribution of specific factors to the removal of
particles from vegetation is difficult to determine
in a manner that will permit the derivation of Tw
as a function of specific removal mechanisms.
Therefore, Tis determined empirically via the
w
regression of the concentration on the vegetation
with time. The effect of growth dilution is implic-
itly included when values of Tare based on a
w
per unit mass basis and will result in lower esti-
mates. This effect is excluded if calculations are
made on a per unit ground basis. On the other
hand, the influence of grazing will have more of
an effect on the loss rate calculation when based
on a per unit ground basis. The estimates of
particle residence times commonly used in the
literature are obtained from studies of the deposi-
tion of radionuclides to vegetation. It remains to
be seen whether these values are also appropriate
for SOCs, as perhaps different types of particles
are involved in their transport and deposition.
Particles of a given aerosol do not attach them-
selves to a rough surface uniformly, but exhibit a
range of adhesive forces which reflect the range
of adhesive sites available on the surface. The
distribution of adhesive forces is typically log-nor-
mal. Under dry conditions the energy required to
overcome the force binding a particle to a surface
is derived from the turbulent kinetic energy of
the surrounding air. Therefore, a more tightly
bound particle will require a greater total amount
of energy to be liberated. The rate of re-entrain-
ment of particles will vary approximately inversely
with time, so that a decreasing proportion of the
residual deposit will be lost with successive time
intervals. However, the plant canopy is exceed-
ingly complex and its structure will change over
time. Therefore, there is not only a wide range of
adhesive sites but also considerable spatial and
temporal variation in the intensity and frequency
of turbulent kinetic energy. In addition, there are
further complicating factors such as the presence
or absence of rainfall along with biological
processes such as cuticular wax shedding and leaf
loss. Thus, while time dependency might be ex-
pected in the rate at which contamination is re-
moved, this might not necessarily conform to the
simple situation described above. Furthermore,
there may be sites within the canopy for which
there would be insufficient turbulence to re-en-
train more tightly bound particles and therefore,
particles in such sites would remain on the plant
indefinitely, or else be lost by some other process
such as cuticular wax shedding or defoliation
Ž.
Kinnersley et al., 1996 .
It has been suggested that the removal of parti-
cles is more influenced by the physical process of

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236224
cuticle removal than by the chemical or physical
form of the contaminant or the frequency of
Ž
rainfall Chamberlain, 1970; Chamberlain and
.
Little, 1981 . Rain may influence the removal
process indirectly by increasing the rate of wax
removal from the cuticle. Loss has been shown to
occur from plants in the absence of rain, and the
rate of loss is greatest when the vegetation is
growing most rapidly and producing the most new
cuticle. There also seems to be longer particle
retention times in slow growing species as com-
pared to rapidly growing ones, again perhaps be-
cause of differences in cuticle production and
shedding. Thas been found to be lower for
w
growing than dormant vegetation, which might
explain the higher particulate levels found in lit-
Ž
ter as compared to growing grass Little et al.,
.
1980 . In studies on the influence of leaf structure
on the retention of particles under the influence
Ž.
of rain Barthlott and Neinhuis, 1997 , it was
found that on water repellent leaves the particles
were removed completely by water droplets, inde-
pendent of the chemical nature or size of the
particles. This appears to contradict the discus-
sion above on the limited influence of rain, but
may just be a reflection of the range of plant
types to be found. Leaves that were water repel-
lent had distinctively convex to papillose epider-
mal cells and a dense layer of epicuticular waxes.
Therefore, although particles are captured more
effectively by rough leaf surfaces, it appears that
such plants have a very effective self-cleaning
mechanism. The epicuticular waxes are fragile
structures easily damaged by mechanical abrasion
and therefore particles may be retained in such
altered areas. Wettable leaves which had smooth
surfaces without any prominent surface sculptur-
ing retained 40᎐80% of particles, with more par-
ticles being removed in heavier rains. A study on
the distribution of such water repellent leaves
found them to be common in herbaceous plants,
especially the grasses, which in view of their im-
portance in agricultural foodchains could be sig-
Ž.
nificant Neinhuis and Barthlott, 1997 . Pu parti-
cles of MMD-1␮m aerosolised and deposited
on the foliage of Phaseolus ¨ulgaris were resistant
Ž.
to removal by wind Cataldo et al., 1981 . Under
the influence of simulated rainfall foliar retention
ranged from 20 to 92% depending on the particle
size, chemical form of the Pu and environmental
conditions, such as the relative humidity and rain-
fall characteristics with mists being more effective
than droplets in removing the deposits. In each
case the first leaching event accounted for the
major loss of deposit from the leaves. The reten-
tion curves therefore exhibited two components, a
relatively rapid loss in response to the first leach-
Ž
ing event this may represent the loss of larger
.
particles andror readily solubilised components
followed by a much reduced rate of loss. A study
203 Ž.
using Pb labelled aerosols MMD 0.03᎐0.2 ␮m
found that light rainfall removed 30% of the
deposited Pb from the grass while heavy or re-
Ž
peated rainfall removed up to 50% Little and
.
Wiffen, 1977 . Again these studies would seem to
indicate that in contrast to the assertion above
about the dominance of cuticle removal as a loss
mechanism, rain can indeed play a significant
role.
A default value of Ts14 days is usually used
w
for radiological risk assessment, with the variabil-
ity of Tgenerally being less than that of other
w
parameters used in such models. A comprehen-
sive summary of literature values of Thas been
w
Ž.
compiled by Miller and Hoffman 1983 , see Table
4. The values of Tfor a range of plant types and
w
particle size ranges from later studies has been
Ž
summarised in Table 5 these studies all refer to
cases where the particulate contamination was of
an acute nature and the values of Twere calcu-
w
lated in such a way as to negate the effects of
growth dilution in reducing the contaminant lev-
.
els .
5.4. Normalised Specific Acti¨ity
Ž.
The Normalised Specific Activity NSA is a
concept originally developed for the atmospheric
deposition of contaminants to grassland. It is de-
fined as the ratio of contaminant per unit mass of
vegetation to the daily rate of deposition onto the
Ž
ground Chamberlain, 1970, 1986; Simmonds and
.
Linsley, 1982 . The expression of the relationship
between continuous deposition and the contami-
nation of different vegetation types as the NSA,
allows an assessment of the contamination levels

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 225
Table 4
a
Ž.
Summary of environmental half-lives T of particles on vegetation
w
Vegetation type Geometric mean Range Geometric S.D.
Ž. Ž.
days days
b
Herbaceous vegetation 17 9᎐34 1.6c
12 9᎐19 1.5d
20 10᎐24 1.5
b
Woody vegetation 20 12᎐28 1.5c
20 12᎐28 1.5 d
NA. NA. NA.
b
Dormant vegetation 32 22 ᎐49 1.4
aŽ.
Adapted from Miller and Hoffman 1983 .
bSummary of all measurements.
cTdetermined from measurements of concentration per unit mass.
w
dTdetermined from measurements of concentration per unit ground area.
w
on vegetation and also provides a useful compar-
ison which can help in elucidating the mecha-
nisms of retention and field loss of pollutants.
Larger NSA values are therefore indicative of
greater interception andror retention. As the
length of the growing season increases, the NSA
tends towards a constant value as field losses
balance the continued deposition. In poor growing
conditions and in winter the NSA values are
higher since there is less growth dilution, slowly
growing foliage will have been exposed for longer,
and the rate of field loss is possibly correlated
with the growth rate. The use of the NSA implic-
itly takes account of processes affecting foliar
contamination such as growth dilution and con-
tamination as a result of soil resuspension.
Field measurements on grassland have allowed
empirical interception and retention models to be
developed. The majority use the following
Ž.
parameters: the interception fraction pand the
Ž.
retention half-life T. A number of studies on
w
herbage have indicated pto be in the range
0.2᎐0.28 and Tto be in the range 13᎐19 days
w
Ž
during summer during winter when there is little
.Ž
growth the retention is longer Chamberlain,
.
1970 . In generic radionuclide modelling studies,
values of ps0.25 and Ts14 days are com-
wŽ
monly used for all vegetation types Chamberlain,
.
1970; Simmonds and Linsley, 1982 . Provided the
Ž
density of the herbage with its important influ-
.
ence on the interception fraction is taken into
account, empirical models of this type predict
fairly reliably the contamination of herbage due
to deposition. At a given location the density of
pasture can be considered constant during the
growing season and therefore so will the intercep-
tion fraction be. Therefore, the model will predict
Ž
the average concentration Simmonds and Lins-
.Ž.
ley, 1982 . Chamberlain 1970 has derived a the-
oretical relationship between the NSA for herbage
and experimentally measurable quantities: the
Ž
uptake coefficient ␮used to calculate pas de-
Ž..
scribed above in Eq. 3 , the retention half-life
T, and the period of growth between successive
w
cuttings. This theoretical approach provided good
agreement with NSA values from experimental
studies. Herbage experiments have indicated that
the NSA is similar for a number of contaminants
throughout the growing season and independent
Ž
of the physical form of the contaminants Cham-
.
berlain, 1970, 1986 .
For crops on the other hand, the concentration
is required at a specific time, i.e. at harvest. The
concentration will be influenced by a number of
factors, such as changes in the physical plant size
during growth, which will influence the intercep-
tion fraction and also result in the dilution of
activity. The extent of external contamination of
plants following a single deposition event depends
on when the event occurs relative to the growth
cycle of the plant. The timing will affect both the

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236226
Table 5
Ž.
Environmental half-lives Tof particles on vegetation
w
Ž.
Particle type Size Vegetation type Tdays Comments Reference
w
238 Ž.
Pu Median ds0.8 ␮m Maize 12 Pinder and Doswell 1985
Ž. Ž.
Silica spheres MMAD 1.9 ᎐8.4 ␮m Exposed wheat 1᎐24᎐8% residue Rapid initial loss Kinnersley et al. 1996
with residual level
determined by the
turbulence of the
weather.
Ž.
Partially sheltered wheat 3᎐422᎐52% residue
Ž.
Greenhouse grown broad beans 0.5᎐1.5 22᎐26% residue
141 134 Ž. Ž.
Ce and Cs MMAD ;1␮mArtemisia tridentata 1᎐2 initial loss Fraley et al. 1993
Ž.
and Elymus elymoides Several weeks subsequent loss

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 227
Ž
fraction of the deposit intercepted because of
.
changing biomass and the extent of dilution and
weathering. In such instances it remains to be
seen whether the simple relationship between the
NSA and the uptake coefficient, retention half
life and period of growth still holds.
Ž.
A study Simmonds and Linsley, 1982 de-
termined the NSA values for a range of grains
and leafy vegetables, and found the values for Pu
to be lower than those for the other radionuclides
investigated. The authors suggested that the lower
NSA values for Pu observed were the result of
the different retention behaviour of this radionu-
clide. The reasons for this were unclear, and
could have resulted from the insolubility of its
oxides preventing translocation away from the
leaf surface thus allowing for increased weather-
ing. For each individual radionuclide the NSA
values were similar for the range of Brassicas
tested, despite differences in their growth struc-
tures. According to this study a smaller fraction
of the daily deposited activity was present per
unit mass of the leafy vegetables than on herba-
ceous vegetation because of morphological dif-
ferences and a smaller surface area to mass ratio.
Ž.
However, a study by Pinder et al. 1985 investi-
238 Ž
gating the Pu NSA values for crops wheat,
.
soybean, corn and tobacco and leafy vegetables
Ž.
lettuce, turnip, broccoli and cabbage , found in
contrast to the findings above that the NSA val-
ues for Pu were similar to those of the other
radionuclides investigated on comparable crops.
These NSA values were also found to be similar
to those for herbage. This would indicate similar
interception andror retention efficiencies both
between the different radionuclides and the vari-
ous vegetation types. Much higher values were
found for lettuce, which the authors attributed to
the type of lettuce being a semi-head variety
allowing for increased efficiency in trapping depo-
sition and soil resuspension.
6. Contribution of particle-bound SOCs to plant
contamination
The evidence for the importance of particle-
bound SOC contamination of vegetation is mixed.
Given the large number of factors discussed above
which affect the interception and retention of
particles by a range of plants under different
scenarios, it is perhaps not surprising that gener-
alisations cannot readily be made about the rela-
tive contributions of each of the various contami-
nation pathways. The experimental and theoreti-
cal evidence for a quantitative contribution of
atmospheric particulate SOC deposition is re-
viewed in the first section below. The second
section attempts to quantify the proportion of the
overall SOC contamination of vegetation ac-
counted for by soil present on the plant surface.
6.1. E¨idence for particulate-deri¨ed SOC
contamination of ¨egetation
6.1.1. PCDDrFs
The uptake and subsequent translocation of
PCDDrFs from the soil by vegetation is generally
Ž
a relatively minor pathway of contamination Mc-
Crady et al., 1990; Schroll and Scheunert, 1993;
.
Schroll et al., 1994; Nakamura et al., 1995 . How-
ever for zucchini and pumpkins, both members of
Ž.
the cucumber family Cucurbitaceae , it has been
reported that contamination by root uptake and
Ž.
translocation is important Hulster et al., 1994 . A
¨
further study by the same authors goes on to
Ž.
suggest that in zucchini Cucurbita pepo L. this
phenomenon results from exudates from the root
complexing with PCDDrFs in the soil, thus facili-
Ž
tating root uptake and translocation Hulster and
¨
.
Marschner, 1995 . Therefore, in some cases path-
ways which are normally discounted as being mi-
nor in contributing to plant contamination may in
fact be important. According to a model devel-
Ž.
oped by Trapp and Matthies 1997 , volatilisation
from the soil followed by foliar uptake would
appear to be a minor pathway for PCDDrFs,
except in those cases where the soils are heavily
contaminated and the plants are growing close to
the ground.
Ž.
Chrostowski and Foster 1996 developed a
model to explain PCDDrF uptake by vegetation
from the atmosphere. They predicted comparable
contributions to plant contamination from the
vapour phase and direct particulate deposition for
TCDDrF, one of the more volatile of the homo-

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236228
logue groups. The importance of particulate phase
transfer was predicted to increase with the more
chlorinated homologues which occur primarily in
Ž
the particulate phase of the atmosphere see Fig.
.
2 . Other models developed for PCDDrF uptake
Ž.
to grass Lorber, 1995; McLachlan, 1997 con-
sider vapour phase transfer to be the dominant
process, although both authors point out the large
uncertainties associated with modelling the par-
ticulate deposition because of a lack of knowledge
in this area. In addition, both models did not
include potential inputs from wet deposition. An
evaluation of different modelling approaches to
the transfer of airborne PCDDrFs to grass was
Ž.
considered by Douben et al. 1997 . The scaveng-
ing model originally proposed by McLachlan
Ž.
1995 gave good predictions when compared to
an experimental data set. One of the implications
of this approach was that airborne PCDDrFs
present in the atmosphere in the vapour or par-
ticulate phases were transferred with similar ef-
ficiencies to the grass and remained associated
with it after interception. Again the fate of parti-
cles on plants was one of the key areas mentioned
in which there was a poor understanding.
Ž.
Welsch-Pausch et al. 1995 conducted a study
on the uptake of PCDDrFs by Welsh Rye grass
in controlled growth chambers. Their work indi-
cated that for the tetra- to hexa-chlorinated con-
geners dry gaseous deposition is the main uptake
pathway. They also concluded that the contribu-
Ž.
tion of small particles d-2.9 ␮m to the con-
tamination of the grass was negligible for all
compounds. However, their experimental set-up
was such that grass exposure to particles of d)2.9
␮m was minor, and it may be important to study
the transfer of PCDDrFs to vegetation under
field-based experimental conditions where the full
range of aerosol particle sizes and wet deposition
could potentially impact on the vegetation con-
tamination. Indeed a further study by the same
authors on the uptake of PCDDrFs by native
grassland containing herbaceous species indicated
that, although dry gaseous deposition dominated
the accumulation of the lower chlorinated homo-
logues, particle-bound deposition was important
for homologues with six or more chlorine atoms
Ž.
Welsch-Pausch and McLachlan, 1996 . The im-
portance of this pathway varied with different
plant species and will be related to leaf orienta-
tion, exposure, and leaf surface characteristics.
Ž.
As part of a study by Prinz et al. 1991 , the
deposition of PCDDrFs in particulate matter was
compared with PCDDrF uptake by standardised
grass cultures grown in uncontaminated condi-
tions and protected from the effects of soil con-
tamination. There was a good correlation between
the two parameters of PCDDrF uptake and
PCDDrF particulate deposition. Schuler et al.
Ž.
1995 investigated the transfer factors of
PCDDrFs into dairy milk and concluded that the
soil contamination of the fodder was not impor-
tant, but that atmospheric particulate deposition
was. Their study site was situated close to a
modern municipal waste incinerator and there-
fore subject to higher deposition levels than aver-
age, while the soil concentrations were found to
be only slightly elevated compared to those of
remote areas. They found deposition onto the
pasture was not congener-specific and estimated
Ž
that during the period of investigation 55 days
.
over summer , 40% of the deposited PCDDrF
was retained on the grass with the remainder
perhaps being washed off by rain. This fraction
was arrived at by comparison of the deposition as
measured by the Bergerhoff method and the
amount found in the grass. A study investigating
the transfers of airborne PCDDrFs to bulk depo-
sition collectors and herbage found the mixture of
2, 3, 7, 8 substituted congeners to be similar in
the air, bulk deposition and grass over a given
Ž
sampling period and site Jones and Duarte-
.
Davidson, 1997 . This suggested that PCDDrFs
Ž
of different levels of chlorination and therefore
.
different gasrparticle partitioning were transfer-
ring with similar efficiencies from the air to the
grass. The authors suggested that the similar
transfer efficiencies could result from wet deposi-
tional inputs supplying additional particulate and
dissolved PCDDrF to the herbage or because
ultra-fine aerosols carrying PCDDrFs and gas
phase PCDDrFs have similar deposition veloci-
ties. For the background site, the ratio of the flux
Žy2y1.
removed by the grass per day pg m day to

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 229
Žy2y1.
the bulk deposition flux pg m day ranged
Ž.
from 0.63 to 3.0 average 1.1 . Values greater
than 1 would suggest that the bulk deposition
collectors are underestimating deposition to the
grass itself. In the two studies above, in order to
fully understand the processes of interception and
retention of the gas and particle phase com-
pounds the nature of the bulk deposition to the
grass sward needs to be understood. A study by
Ž.
Hulster and Marschner 1993 on the transfer of
¨
PCDDrF from contaminated soils to lettuce,
potato and hay concluded that the contamination
of crops by soil particles can be important under
certain circumstances. Although the contamina-
tion of lettuce and potato shoots resulted mainly
from atmospheric deposition, soil particle accu-
mulation assumed importance for hay.
6.1.2. PAHs
The direct uptake of PAHs from the soil via
root uptake or adsorption to the root surface
appears to be minimal in comparison to the other
Ž
uptake pathways from the atmosphere Wild and
.Ž.
Jones, 1991, 1992 . Nakajima et al. 1995 related
the concentrations of three PAHs; pyrene,
wx
benzo apyrene and perylene, in azalea leaves to
wx
the atmospheric concentrations. Benzo apyrene
and perylene are found mainly in the particulate
phase in the atmosphere, whereas pyrene is found
Ž.
in both the gas and particulate phases see Fig. 2 .
The concentrations of all of the compounds in the
leaves and the TSP decreased in summer and
wx
increased in winter. Indeed the benzo apyrene
and perylene leaf concentrations were found to
be approximately proportional to those in the
TSP. In other words, for these two compounds
the patterns of seasonal change were similar for
the leaf and TSP, with the highest concentrations
occurring in winter. The authors suggested that
contamination resulted from the attachment of
the TSP to the leaf surface, followed by
permeation of the PAHs into the cuticle. How-
ever, the situation with pyrene was more complex
because the pattern of seasonal change in the
leaves did not match that of the particulate, the
vapour, or the sum of both phases. The pyrene
vapourrparticulate distribution varied strongly
with temperature, with )80% in the vapour
phase in summer and 40᎐70% in the vapour
phase in winter. Pyrene contamination of leaves
was thought to be primarily influenced by the
temperature-dependant partitioning of vapour to
Ž.
the leaf surface. Nakajima et al. 1995 concluded
that in their study approximately 75% of its up-
take was from the gas phase, with the remainder
attributable to the particulate phase.
6.1.3. PCBs
A theoretical approach to investigate the depo-
sition of PCBs to spruce needles considered the
pathways of dry gaseous deposition, particle-
bound deposition and wet deposition and con-
cluded that each of the different pathways con-
tributes significantly to the concentrations found
Ž.
in the needles Umlauf and McLachlan, 1994 . In
an attempt to further investigate this, a green-
house system was set up to study the various
Ž.
pathways in isolation Umlauf et al., 1994 . The
authors concluded that the air to plant transfer of
PCBs was dominated by dry gaseous deposition,
with dry particulaterwet deposition or the washoff
of the particle scavengedrdissolved substances
not affecting the atmospheric input.
There is still very little information on the wet
and dry particulate deposition of SOCs to vegeta-
tion. Where it is considered, it is usually under
experimental conditions which are perhaps not
representative of those present in the field. In
those studies where the bulk deposition to vegeta-
tion has been studied, there is much uncertainty
as to the exact nature of the particulate deposi-
tion reaching the plant surface and its subsequent
fate. It is perhaps the first of these two unknowns
that needs most urgent definition, since without a
knowledge of the deposition reaching the plant
the subsequent fate processes cannot be accu-
rately investigated. Therefore, no attempt is made
here to quantify the importance of the SOC par-
ticulate deposition contamination pathway, de-
spite there being in the literature a considerable
body of information on the physical processes of
interception and retention of a range of particle
sizes by different vegetation types which could be
legitimately used for SOCs.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236230
6.2. Illustrati¨e calculations of the soil associated
SOC contamination of ¨egetation
In order to quantify the contribution of soil
associated SOCs to the overall level of plant
contamination, information from various sources
in the literature has been used. Only data sets
where the SOC concentrations were measured in
the vegetation and soil at the same site and time
have been considered. This limited the calcula-
tions to PCDDrFs, PAHs, and PCBs for grass
and maize. For grass, the soil contamination sce-
narios considered included a conservative back-
ground estimate of 20 mg soil gy1grass DW, and
a worst case situation of 100 mg soil gy1grass
Ž.
DW see Table 3 . For maize a soil mass loading
y1Ž.
of 2 mg soil g plant DW was used see Table 3 .
This was taken to be typical of those levels result-
ing from standard agricultural practices. The bulk
soil SOC concentration has been used. This will
result in an underestimation of the PAH and
PCB soil associated contamination since these
compounds are concentrated in the finer soil par-
ticle size fractions. For PCDDrFs the situation
with respect to soil particle size is unknown.
Table 6 shows the percentage contribution of
SOCs associated with the soil to the total plant
SOC concentration under scenarios of back-
ground and worst case soil mass loadings. There
appears to be systematic differences between the
different compound groups. For grass, given a
typical background soil mass loading of 20 mg
gy1, soil particle contamination was significant for
PCDDrFs and PAHs. This route of contamina-
tion accounted for approximately 30% of the total
SOCs found in the grass. For PCBs however, the
contribution was much smaller, accounting for
only approximately 3᎐6%. With a worse case situ-
ation of 100 mg soil gy1plant DW, soil associated
SOCs would account for the majority of the
Ž
PCDDrF and PAH grass contamination percent
Table 6
Percentage contribution of SOCs associated with soil to the total plant SOC concentration
a
Vegetation type Soil mass SOC Bulk soil Plant Percentage Reference
loading concentration concentration contribution of
y1y1y1
ŽŽ.Ž.
mg soil g pg g dry wt. pg g dry wt. soil associated
.
plant dry wt. SOCs to the
total plant SOC
concentration
Grass 20 ÝPCDDrF 467 31 30.1 a
100 )100 b
Grass 20 ÝPCDDrF 37 2.5 29.6 b
100 )100
Grass 20 ÝPAH 2 697 000 153 000 35.3 c
100 )100 b
Grass 20 ÝPCB 1098 725 3.0 b
100 15.1
Grass 20 ÝPCB 2464 852 5.8 d
100 28.9
Maize 2 ÝPCDDrF 467 3.3 28.3 a
b
Maize 2 ÝPCDDrF 37 28.7 0.3 b
b
Maize 2 ÝPCB 1098 3730 0.1 b
aŽ. Ž. Ž. Ž
References: a, Schuler et al. 1995 ; b, McLachlan 1996 ; c, Meharg et al. 1998 ; d, Dr G.O. Thomas personal communica-
.
tion .
bFor b a 20% fraction for dry mass was assumed for the grass and maize measurements.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 231
contributions greater than 100% were obtained,
this probably resulting from using an unrealisti-
cally high soil mass loading for samples represent-
.
ing background situations . However, for PCBs,
although significant with soil associated PCBs ac-
counting for 15᎐30% of the total plant contami-
nation, it appears that other contamination path-
ways still dominate.
The differences between species were not so
clear. In one of the studies soil associated
PCDDrFs accounted for 28% of the maize con-
tamination, while in the other the contribution
was negligible. The reasons for this are unknown,
and may genuinely reflect the differences between
the studies in soil associated SOC contamination
levels, or perhaps are the result of different sam-
Table 7
Percentage contribution of individual SOC homologues or compounds associated with soil to the total plant SOC concentration
a
Vegetation Soil mass SOC Bulk soil Plant Percentage Reference
type loading concentration concentration contribution of
y1y1y1
ŽŽ.Ž.
mg soil g pg g dry wt. pg g dry wt. soil associated
.
plant dry wt. SOCs to the
total plant SOC
concentration
Grass 20 ÝCL4 DD 0.05 0.0095 11
Grass 20 ÝCL5 DD 0.18 0.018 20 b
Grass 20 ÝCL6 DD 0.56 0.075 15 b
Grass 20 ÝCL7 DD 5.00 0.35 29
Grass 20 ÝCL8 DD 16.00 0.95 34
Grass 20 ÝCL4 DF 0.48 0.115 8
Grass 20 ÝCL5 DF 1.35 0.145 19 b
Grass 20 ÝCL6 DF 2.31 0.185 25 b
Grass 20 ÝCL7 DF 7.40 0.5 30
Grass 20 ÝCL8 DF 3.70 0.145 51
Grass 20 ÝCL4 DD 0.36 0.08 9
Grass 20 ÝCL5 DD 2.40 0.16 30
Grass 20 ÝCL6 DD 16.30 1.43 23 a
Grass 20 ÝCL7 DD 90.00 5.1 35
Grass 20 ÝCL8 DD 266.00 16 33
Grass 20 ÝCL4 DF 1.80 0.58 6
Grass 20 ÝCL5 DF 2.82 0.32 18
Grass 20 ÝCL6 DF 22.42 3.3 14 a
Grass 20 ÝCL7 DF 41.70 2.57 32
Grass 20 ÝCL8 DF 22.00 1.5 29
Grass 20 PCB 28 29.00 47.1 1
Grass 20 PCB52 58.60 36 3
Grass 20 PCB101 106.20 49.1 4
Grass 20 PCB118 139.90 57.1 5 d
Grass 20 PCB153 251.80 69.9 7
Grass 20 PCB138 239.00 63.2 8
Grass 20 PCB180 93.80 29.3 6
aŽ. Ž. Ž. Ž .
References: a, Schuler et al. 1995 ; b, McLachlan 1996 ; c, Meharg et al. 1998 ; d, Dr G. Thomas personal communication .
bFor b a 20% fraction for dry mass was assumed for the grass and maize measurements.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236232
Ž.
pling protocols. In McLachlan 1996 the soil had
a much lower PCDDrF concentration and only
the maize leaves were sampled. The position of
the sampled leaves in the canopy may have an
effect on the leaf PCDDrF concentrations. Alter-
natively, the lower contribution of PCDDrFs as-
sociated with soil may result from the lower soil
mass loadings on maize compared to grass, be-
cause of differences in canopy structure. Again
PCBs associated with soil made a negligible con-
tribution to the maize concentration.
In an attempt to investigate systematic differ-
ences within the different SOC classes, individual
homologues or compounds were examined for
their contribution to the plant contamination
Ž.
when associated with soil Table 7 . This was only
done for PCDDrFs and PCBs on grass because
of the limitations of the data-sets available. For
both PCDDrFs and PCBs the calculations sug-
gest that soil contamination becomes increasingly
important with increasing level of chlorination.
7. Concluding remarks
There is clearly the necessity to verify the dif-
ferent plant SOC contamination pathways by the
use of experimental measurements made in real-
world field situations, and to use this new infor-
mation to develop and adapt existing models.
Such studies should of course involve a range of
relevant plant types, and be expanded to include
a full range of SOCs with widely differing
physico-chemical properties.
Acknowledgements
We are grateful to the Food Contaminants
Division of the UK Ministry of Agriculture, Fish-
eries and Food and the Natural Environment
Research Council’s Environmental Diagnostics
Programme for funding research on the air᎐plant
transfer of SOCs at Lancaster.
References
Adriano DC, Pinder JE, McLeod KW, Corey JC, Boni AL.
Plutonium contents and fluxes in a soybean crop ecosystem
near a nuclear fuel chemical separations facility. J Environ
Qual 1982;11:506᎐511.
Adriano DC, McLeod KW, Ciravolo TG. Long-term availabil-
ity of Cm and Pu to crop plants. Health Phys
1986;50:647᎐651.
Alcock RE, Johnston AE, McGrath SP, Berrow ML, Jones
KC. Long-term changes in the polychlorinated biphenyl
content of United Kingdom soils. Environ Sci Technol
1993;27:1918᎐1923.
Anspaugh LR, Shinn JH, Phelps PL, Kennedy NC. Resuspen-
sion and redistribution of plutonium in soils. Health Phys
1975;29:571᎐582.
Bacci E, Cerejeira MJ, Gaggi C, Chemello G, Calamari D,
Vighi M. Bioconcentration of organic chemical vapours in
plant leaves: the azalea model. Chemosphere 1990a;21:
525᎐535.
Bacci E, Calamari D, Gaggi C, Vichi M. Bioconcentration of
organic chemical vapors in plant leaves: experimental mea-
surements and correlation. Environ Sci Technol 1990b;
24:885᎐889.
Barrie LA. Scavenging ratios: black magic or a useful scientific
tool. In: Schwartz SE, Slinn WGN, editors. Precipitation
scavenging and atmosphere ᎏsurface exchange processes,
vol. 1 ᎏThe Georgii volume: precipitation scavenging
processes. Washington: Hemisphere Pub. Corp, 1992:
403᎐419.
Barthlott W, Neinhuis C. Purity of the sacred lotus, or escape
from contamination in biological surfaces. Planta 1997;
202:1᎐8.
Bidleman TF. Atmospheric processes. Environ Sci Technol
1988;22:361᎐367.
Bidleman TF, Billings WN, Foreman WT. Vapor-particle par-
titioning of semivolatile organic compounds: estimates from
field collections. Environ Sci Technol 1986;20:1038᎐1043.
Bisal F, Hsieh J. Influence of moisture on erodibility of soil by
wind. Soil Sci 1966;102:143᎐146.
Brabec E, Kovar P, Drabkova A. Particle deposition in three
´´´
vegetation stands: a seasonal change. Atmos Environ
1981;15:583᎐587.
Butler JD, Crossley P. An appraisal of relative airborne sub-
urban concentrations of polycyclic aromatic hydrocarbons
monitored indoors and outdoors. Sci Total Environ
1979;11:53᎐58.
Cataldo DA, Garland TR, Wildung RE. Foliar retention and
leachability of submicron plutonium and americium parti-
cles. J Environ Qual 1981;10:31᎐37.
Chamberlain AC. Interception and retention of radioactive
aerosols by vegetation. Atmos Environ 1970;4:57 ᎐78.
Chamberlain AC. Deposition of gases and particles on vegeta-
tion and soils. In: Legge AM, Krupa SV, editors. Air
pollutants and their effects on the terrestrial ecosystem.
New York: Wiley International, 1986:189 ᎐209.
Chamberlain AC, Little P. Transport and capture of particles
by vegetation. In: Grace J, Ford PG, Jarvis PG, editors.
Plants and their atmospheric environment. Oxford: Black-
well Scientific Publications, 1981:147᎐173.
Chen S, Hsieh L, Hwang P. Concentration, phase distribution,

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 233
and size distribution of atmospheric polychlorinated
biphenyls measured in southern Taiwan. Environ Int
1996;22:411᎐423.
Chrostowski PC, Foster SA. A methodology for assessing
congener-specific partitioning and plant uptake of dioxins
and dioxin-like compounds. Chemosphere 1996;32:
2285᎐2304.
Cotham WE, Bidleman TF. Laboratory investigations of the
partitioning of organochlorine compounds between the gas
phase and atmospheric aerosols on glass fibre filters. Envi-
ron Sci Technol 1992;26:469᎐478.
Douben PET, Alcock RE, Jones KC. Congener specific trans-
fer of PCDDrFs from air to cow’s milk: an evaluation of
current modelling approaches. Environ Pollut 1997;95:
333᎐344.
Dreicer M, Hakonson TE, White GC, Whicker FW. Rain-
splash as a mechanism for soil contamination of plant
surfaces. Health Phys 1984;46:177᎐187.
Duarte-Davidson R, Sewart A, Alcock RE, Cousins IT, Jones
KC. Exploring the balance between sources, deposition,
and the environmental burden of PCDDrFs in the U.K.
terrestrial environment: an aid to identifying uncertainties
and research needs. Environ Sci Technol 1997;31:1᎐11.
Duinker JC, Bouchertall F. On the distribution of atmo-
spheric polychlorinated biphenyl congeners between vapor
phase, aerosols, and rain. Environ Sci Technol
1989;23:57᎐62.
Ž.
Edwards NT. Polycyclic aromatic hydrocarbons PAH’s in the
terrestrial environment ᎏa review. J Environ Qual
1983;12:427᎐441.
Eisenreich SJ, Looney BB, Thornton JD. Airborne organic
contaminants in the Great Lakes ecosystem. Environ Sci
Technol 1981;15:30᎐38.
Eitzer BD, Hites RA. Atmospheric transport and deposition
of polychlorinated dibenzo-p-dioxins and dibenzofurans.
Environ Sci Technol 1989;23:1396᎐1401.
Farmer EE. Relative detachability of soil particles by simu-
lated rainfall. Soil Sci Soc Am Proc 1973;37:629 ᎐633.
Finizio A, Mackay D, Bidleman T, Harner T. Octanol ᎐air
partition coefficient as a predictor of partitioning of semi-
volatile organic chemicals to aerosols. Atmos Environ
1997;31:2289᎐2296.
Fraley L, Chavez G, Markham OD. Seasonal variations in
deposition and retention of cerium-141 and cesium-134 in
cool desert vegetation. J Environ Radioact 1993;21:203᎐212.
Freer-Smith PH, Holloway S, Goodman A. The uptake of
particulates by an urban woodland: site description and
particulate composition. Environ Pollut 1997;95:27 ᎐35.
Gardner B, Hewitt CN, Jones KC, Smith D. Deposition of
polycyclic aromatic hydrocarbons to natural water surfaces
in the United Kingdom. In: Shwartz SE, Slinn WGN,
editors. Precipitation scavenging and atmosphere-surface
exchange processes, vol. 2 ᎏThe Semonin volume: atmo-
sphere-surface exchange processes. Washington: Hemi-
sphere Pub. Corp, 1992:649 ᎐660.
Gardner B, Hewitt CN, Jones KC. PAHs in air adjacent to
two inland water bodies. Environ Sci Technol 1995;
29:2405᎐2413.
Garland JA, Cox LC. Deposition of small particles to grass.
Atmos Environ 1982;16:2699᎐2702.
Gatz DF, Dingle AN. Trace substances in rain water: concen-
tration variations during convective rains, and their inter-
pretation. Tellus 1971;23:14᎐27.
Gay DD, Watts JR. Particle size distribution of airborne
plutonium near a chemical separations facility. Aiken, S.C:
E.I. du Pont de Nemours and Co.; Report Dp-1610, 1981.
Green N, Dodd NJ. The uptake of radionuclides from inad-
vertent consumption of soil by grazing animals. Sci Total
Environ 1988;69:367᎐377.
Guggenberger G, Pichler M, Hartmann R, Zech W. Polycyclic
aromatic hydrocarbons in different forest soils: mineral
horizons. Z Pflanzenernahr Bodenk 1996;159:565᎐573.
¨
Hakonson TE. Environmental pathways of plutonium into
terrestrial plants and animals. Health Phys 1975;29:
583᎐588.
Harner T, Bidleman TF. Octanol᎐air partitioning coefficient
for describing particlergas partitioning of aromatic com-
pounds in urban air. Environ Sci Technol 1998;32:
1494᎐1502.
Hicks BB. Measuring dry deposition: a reassessment of the
state of the art. Water Air Soil Pollut 1986;30:75᎐90.
Hoffman FO, Thiessen KM, Frank ML, Blaylock BG. Quanti-
fication of the interception and initial retention of radioac-
tive contaminants deposited on pasture grass by simulated
rain. Atmos Environ 1992;26A:3313᎐3321.
Holsen TM, Noll KE, Liu S, Lee W. Dry deposition of
polychlorinated biphenyls in urban areas. Environ Sci
Technol 1991;25:1075᎐1081.
Hulster A, Marschner H. Transfer of PCDDrPCDF from
¨
contaminated soils to food and fodder crop plants. Chemo-
sphere 1993;27:439᎐446.
Hulster A, Marschner H. PCDDrF-complexing compounds in
¨
zucchini. Organohalogen Compd 1995;24:493᎐496.
Hulster A, Muller JF, Marschner H. Soil᎐plant transfer of
¨¨
polychlorinated dibenzo-p-dioxins and dibenzofurans to
Ž.
vegetables of the cucumber family Cucurbitaceae . Environ
Sci Technol 1994;28:1110᎐1115.
Jones KC, Duarte-Davidson R. Transfers of airborne
PCDDrFs to bulk deposition collectors and herbage. Envi-
ron Sci Technol 1997;31:2937᎐2943.
Kao AS, Venkataraman C. Estimating the contribution of
reentrainment to the atmospheric deposition of dioxin.
Chemosphere 1995;31:4317᎐4331.
Katz M, Chan C. Comparative distribution of eight polycyclic
aromatic hydrocarbons in airborne particulates collected by
conventional high-volume sampling and by size fractiona-
tion. Environ Sci Technol 1980;14:838᎐843.
Kaupp H, McLachlan MS. Atmospheric particle size distribu-
tions of polychlorinated dibenzo-p-dioxins and dibenzofu-
Ž.
rans PCDDrFs and polycyclic aromatic hydrocarbons
Ž.
PAHs and their implications for wet and dry deposition.
Atmos Environ 1999;33:85᎐95.

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236234
Kaupp H, Towara J, McLachlan MS. Distribution of polychlo-
rinated dibenzo-p-dioxins and dibenzofurans in atmo-
spheric particulate matter with respect to particle size.
Atmos Environ 1994;28:585᎐593.
Kinnersley RP, Shaw G, Bell JNP, Minski MJ, Goddard AJH.
Loss of particulate contaminants from plant canopies un-
der wet and dry conditions. Environ Pollut 1996;91:227᎐235.
Koester CJ, Hites RA. Wet and dry deposition of chlorinated
dioxins and furans. Environ Sci Technol 1992;26:1375᎐1382.
Kovar P, Mejstrik V. Interception and retention of atmo-
´
spheric particles by stands of alfalfa and maize. Sci Total
Environ 1987;65:215᎐225.
Kovar P, Brabec E, Riegerova A. Size spectra of particulate
´´
matter in herbaceous stands. Sci Total Environ
1987;65:241᎐246.
Kurokawa Y, Takahiko M, Matayoshi N, Satoshi T, Kazumi F.
Distribution of polychlorinated dibenzo-p-dioxins and
dibenzofurans in various sizes of airborne particles.
Organohalogen Compd 1996;28:232᎐236.
Ligocki MP, Pankow JF. Measurements of the gasrparticle
distributions of atmospheric organic compounds. Environ
Sci Technol 1989;25:75᎐83.
Ligocki MP, Leuenberger C, Pankow JF. Trace organic com-
pounds in rain-II. Gas scavenging of neutral organic com-
pounds. Atmos Environ 1985a;19:1609᎐1617.
Ligocki MP, Leuenberger C, Pankow JF. Trace organic com-
pounds in rain-III. Particle scavenging of neutral organic
compounds. Atmos Environ 1985b;19:1619᎐1626.
Lindberg SE. Factors influencing trace metal, sulfate and
hydrogen ion concentrations in rain. Atmos Environ
1982;16:1701᎐1709.
Lioy PJ, Daisey JM. Airborne toxic elements and organic
substances. Environ Sci Technol 1986;20:8᎐14.
Little P, Wiffen RD. Emission and deposition of petrol engine
exhaust Pb-I. Deposition of exhaust Pb to plant and soil
surfaces. Atmos Environ 1977;11:437᎐447.
Little CA, Whicker FW, Winsor TF. Plutonium in a grassland
ecosystem at Rocky Flats. J Environ Qual 1980;9:350 ᎐354.
Livens FR, Baxter MS. Particle size and radionuclide levels in
some West Cumbrian soils. Sci Total Environ 1988;70:1᎐17.
Lorber M. Development of an air-to-leaf vapor phase transfer
factor for dioxins and furans. Organohalogen Compd
1995;24:179᎐186.
Lorber M, Cleverly D, Schaum J, Phillips L, Schweer G,
Leighton T. Development and validation of an air-to-beef
food chain model for dioxin-like compounds. Sci Total
Environ 1994;156:39᎐65.
Mazurak AP, Mosher PN. Detachment of soil particles in
simulated rainfall. Soil Sci Soc Am Proc 1968;32:716 ᎐719.
Mazurak AP, Mosher PN. Detachment of soil aggregates by
simulated rainfall. Soil Sci Soc Am Proc 1970;34:798 ᎐800.
McCrady JK. Vapor-phase 2, 3, 7, 8-TCDD sorption to plant
foliage ᎏa species comparison. Chemosphere 1994;
28:207᎐216.
McCrady JK, Maggard SP. Uptake and photodegradation of 2,
3, 7, 8-Tetrachlorodibenzo-p-dioxin sorbed to grass foliage.
Environ Sci Technol 1993;27:343᎐350.
McCrady JK, McFarlane C, Gander LK. The transport and
fate of 2, 3, 7, 8-TCDD in soybean and corn. Chemosphere
1990;21:359᎐376.
McLachlan MS. Accumulation of PCDDrF in an agricultural
foodchain. Organohalogen Compd 1995;26:105 ᎐108.
McLachlan MS. Bioaccumulation of hydrophobic chemicals in
agricultural food chains. Environ Sci Technol 1996;
30:252᎐259.
McLachlan MS. A simple model to predict the accumulation
of PCDDrFs in an agricultural food chain. Chemosphere
1997;34:1263᎐1276.
McLeod KW, Adriano DC, Boni AL et al. Influence of a
nuclear fuel chemical separations facility on the plutonium
contents of a wheat crop. J Environ Qual 1980;9:306᎐315.
McLeod KW, Alberts JJ, Adriano DC, Pinder JE. Plutonium
contents of broadleaf vegetable crops grown near a nuclear
fuel chemical separations facility. Health Phys 1984a;
46:261᎐267.
McLeod KW, Pinder III JE, Watts JR. Contribution of a
nuclear fuel chemical separations facility to the plutonium
content of a tobacco crop. Health Phys 1984b;46:1205 ᎐1211.
McMahon TA, Denison PJ. Empirical atmospheric deposition
parameters ᎏa survey. Atmos Environ 1979;13:571᎐585.
Meharg AA, Wright J, Dyke H, Osborn D. Polycyclic aromatic
Ž.
hydrocarbon PAH dispersion and deposition to vegetation
and soil following a large scale chemical fire. Environ
Pollut 1998;99:29᎐36.
Miller CW. An analysis of measured values for the fraction of
a radioactive aerosol intercepted by vegetation. Health
Phys 1980;38:705᎐712.
Miller CW, Hoffman FO. An examination of the environmen-
tal half-time for radionuclides deposited on vegetation.
Health Phys 1983;45:731᎐744.
Nakajima D, Yoshida Y, Suzuki J, Suzuki S. Seasonal changes
in the concentration of polycyclic aromatic hydrocarbons in
azalea leaves and relationship to atmospheric concentra-
tion. Chemosphere 1995;30:409᎐418.
Nakamura M, Suda R, Matsueda T, Kurokawa Y, Takada S,
Fukamachi K. Uptake of PCDDs and PCDFs by radish
plant. Organohalogen Compd 1995;24:497᎐500.
Neinhuis C, Barthlott W. Characterization and distribution of
water-repellent, self-cleaning plant surfaces. Ann Bot
1997;79:667᎐677.
O’Connor GA, Kiehl D, Eiceman GA, Ryan JA. Plant uptake
of sludge-borne PCBs. J Environ Qual 1990;19:113᎐118.
Pankow JF. Review and comparative analysis of the theories
on partitioning between the gas and aerosol particulate
phases in the atmosphere. Atmos Environ 1987;21:
2275᎐2283.
Pankow JF, Storey JME, Yamasaki H. Effects of relative
humidity on gasrparticle partitioning of semivolatile or-
ganic compounds to urban particulate matter. Environ Sci
Technol 1993;27:2220᎐2226.
Pierce RC, Katz M. Dependancy of polynuclear aromatic

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236 235
hydrocarbon content on size distribution of atmospheric
aerosols. Environ Sci Technol 1975;9:347᎐353.
Pinder III JE, Doswell AC. Retention of 238 Pu-bearing parti-
cles by corn plants. Health Phys 1985;49:771᎐776.
Pinder III JE, McLeod KW. Contaminant transport in
agroecosystems through retention of soil particles on plant
surfaces. J Environ Qual 1988;17:602᎐607.
Pinder III JE, McLeod KW. Mass loading of soil particles on
plant surfaces. Health Phys 1989;57:935᎐942.
Pinder III JE, McLeod KW, Simmonds JR, Linsley GS. Nor-
malized specific activities for Pu deposition onto foliage.
Health Phys 1985;49:1280᎐1283.
Pinder III JE, McLeod KW, Adriano DC, Corey JC, Boni AL.
Atmospheric deposition, resuspension, and root uptake of
Pu in corn and other grain-producing agroecosystems near
a nuclear fuel facility. Health Phys 1990;59:853᎐867.
Pinder III JE, McLeod KW, Lide RF, Sherrod KC. Mass
loading of soil particles on a pasture grass. J Environ
Radioact 1991;13:341᎐354.
Prinz B, Krause GHM, Radermacher L. Criteria for the
evaluation of dioxins in the environment. Chemosphere
1991;23:1743᎐1761.
Radke LF, Hobbs PV, Eltgroth MW. Scavenging of aerosol
particles by precipitation. J Appl Meteorol 1980;19:715᎐722.
Romney EM, Lindberg RG, Hawthorne HA, Bystrom BG,
Larson KH. Contamination of plant foliage with radioac-
tive fallout. Ecology 1963;44:343᎐349.
Romney EM, Wallace A, Schulz RK, Kinnear J, Wood RA.
Plant uptake of 23 7 Np, 239,240 Pu, 241Am, and 244 Cm from
soils representing major food production areas of the
United States. Soil Sci 1981;132:40᎐59.
Sawidis T. Uptake of radionuclides by plants after the Cher-
nobyl accident. Environ Pollut 1988;50:317᎐324.
Schnelle J, Jansch T, Wolf K, Gebefugi I, Kettrup A. Particle
¨¨
size dependent concentrations of polycyclic aromatic hy-
Ž.
drocarbons PAH in the outdoor air. Chemosphere
1995;31:3119᎐3127.
Schnelle J, Wolf K, Frank G, Hietel B, Gebefugi I, Kettrup A.
¨
Particle size-dependent concentrations of polycyclic aro-
matic hydrocarbons. Analyst 1996;121:1301᎐1304.
Schroeder WH, Lane DA. The fate of toxic airborne pollu-
tants. Environ Sci Technol 1988;22:240᎐246.
Schroll R, Scheunert I. Uptake pathways of octa-
chlorodibenzo-p-dioxin from soil by carrots. Chemosphere
1993;26:1631᎐1640.
Schroll R, Bierling B, Cao G et al. Uptake pathways of
organic chemicals from soil by agricultural plants. Chemo-
sphere 1994;28:297᎐303.
Schuler F, Schmid P, Schlatter C. Transfer factors of
PCDDrFs from deposition into dairy milk. Organohalogen
Compd 1995;26:59᎐62.
Sehmel GA. Particle and gas dry deposition: a review. Atmos
Environ 1980;14:983᎐1011.
Shah JJ, Johnson RL, Heyerdahl EK, Huntzicker JJ. Carbona-
ceous aerosol at urban and rural sites in the United States.
J Air Pollut Control Assoc 1986;36:254᎐257.
Simcik MF, Franz TP, Zhang H, Eisenreich SJ. Gas-particle
partitioning of PCBs and PAHs in the Chicago urban and
adjacent coastal atmosphere: states of equilibrium. Environ
Sci Technol 1998;32:251᎐257.
Simmonds JR, Linsley GS. Parameters for modelling the in-
terception and retention of deposits from atmosphere by
grain and leafy vegetables. Health Phys 1982;43:679 ᎐691.
Simonich SL, Hites RA. Importance of vegetation in removing
polycyclic aromatic hydrocarbons from the atmosphere.
Nature 1994;370:49᎐51.
Simonich SL, Hites RA. Organic pollutant accumulation in
vegetation. Environ Sci Technol 1995;29:2905᎐2914.
Smith DG, Chaudhuri IS, Heinold D, Ruffle B. An alternative
approach for estimating plant uptake of dioxin vapors.
Organohalogen Compd 1995;24:187᎐193.
Summerling TJ, Dodd NJ, Green N. The transfer of Stron-
tium-90 and Caesium-137 to milk in a dairy herd grazing
near a major nuclear installation. Sci Total Environ
1984;34:57᎐72.
Tolls J, McLachlan MS. Partitioning of semivolatile organic
Ž
compounds between air and Lolium multiflorum Welsh
.
Rye grass . Environ Sci Technol 1994;28:159᎐166.
Trapp S, Matthies M. Generic one-compartment model for
uptake of organic chemicals by foliar vegetation. Environ
Sci Technol 1995;29:2333᎐2338.
Trapp S, Matthies M. Modeling volatilization of PCDDrF
from soil and uptake into vegetation. Environ Sci Technol
1997;31:71᎐74.
Umlauf G, McLachlan M. Deposition of semivolatile organic
compounds to spruce needles I. Calculation of dry and wet
fluxes. Environ Sci Pollut Res 1994;1:146᎐150.
Umlauf G, Hauk H, Reissinger M. Deposition of semivolatile
organic compounds to spruce needles II. Experimental
evaluation of the relative importance of different pathways.
Environ Sci Pollut Res 1994;1:209᎐222.
Wagrowski DM, Hites RA. Polycyclic aromatic hydrocarbon
accumulation in urban, suburban, and rural vegetation.
Environ Sci Technol 1997;31:279᎐282.
Welsch-Pausch K, McLachlan MS. Pathways of atmospheric
PCDDrF to a native grassland culture: the importance of
particle bound deposition. Organohalogen Compd
1996;28:72᎐75.
Welsch-Pausch K, McLachlan MS, Umlauf G. Determination
of the principal pathways of polychlorinated dibenzo-p-di-
Ž
oxins and dibenzofurans to Lolium multiflorum Welsh Rye
.
grass . Environ Sci Technol 1995;29:1090᎐1098.
Wesley ML, Hicks BB, Dannevik WP, Frisella S, Husar RB.
An eddy-correlation measurement of particulate deposition
from the atmosphere. Atmos Environ 1977;11:561 ᎐563.
Whitby KT. The physical characteristics of sulfur aerosols.
Atmos Environ 1978;12:135᎐159.
White GC, Hakonson TE, Ahlquist AJ. Factors affecting
radionuclide availability to vegetables grown at Los Ala-
mos. J Environ Qual 1981;10:294 ᎐299.
Wilcke W, Zech W. Polychlorinated biphenyls PCBs in bulk
soil and particle size separates of soils in a rural commu-
nity. Z Pflanzenernahr Bodenk 1998;161:289 ᎐295.
¨
Wild SR, Jones KC. Studies on the polynuclear aromatic

()
K.E.C. Smith, K.C. Jones rThe Science of the Total En¨ironment 246 2000 207᎐236236
Ž.
hydrocarbon content of carrots Daucus carota . Chemo-
sphere 1991;23:243᎐251.
Wild SR, Jones KC. Polynuclear aromatic hydrocarbon uptake
by carrots grown in sludge-amended soil. J Environ Qual
1992;21:217᎐225.
Willeke K, Whitby KT. Atmospheric aerosols: size distribution
interpretation. J Air Pollut Control Assoc 1975;25:529 ᎐534.
Wischmeier WH, Smith DD. Rainfall energy and its relation-
ship to soil loss. Trans Am Geophys Union 1958;39:
285᎐291.
Yamasaki H, Kuwata K, Miyamoto H. Effects of ambient
temperature on aspects of airborne polycyclic aromatic
hydrocarbons. Environ Sci Technol 1982;16:189᎐194.
Ye Q, Puri RK, Kapila S, Yanders AF. Studies on the transport
and transformation of PCBs in plants. Chemosphere
1992;25:1475᎐1479.