Handbook of ECOTOXICOLOGY - Section 4 doc

© 2003 by CRC Press LLC
SECTION IV
Methods for Making Estimates, Predictability,
and Risk Assessment in Ecotoxicology
31 Global Disposition of Contaminants Roy M. Harrison, Stuart Harrad, and

Jamie Lead
32 Bioaccumulation and Bioconcentration in Aquatic Organisms Mace G. Barron
33 Structure Activity Relationships for Predicting Ecological Effects of Chemicals
John D. Walker and T. Wayne Schultz
34 Predictive Ecotoxicology John Cairns, Jr. and B. R. Niederlehner
35 Population Modeling John R. Sauer and Grey W. Pendleton
36 Ecological Risk Assessment: U.S. EPA’s Current Guidelines and Future Directions
Susan B. Norton, William H. van der Schalie, Anne Sergeant, Lynn Blake-Hedges,

Randall Wentsel, Victor B. Serveiss, Suzanne M. Marcy, Patricia A. Cirone,

Donald J. Rodier, Richard L. Orr, and Steven Wharton
37 Ecological Risk Assessment Example: Waterfowl and Shorebirds Feeding in
Ephemeral Pools at Kesterson Reservoir, California Earl R. Byron, Harry M. Ohlendorf,
Gary M. Santolo, Sally M. Benson, Peter T. Zawislanski, Tetsu K. Tokunaga, and

Michael Delamore
38 Restoration Ecology and Ecotoxicology John Cairns, Jr
© 2003 by CRC Press LLC
CHAPTER 31
Global Disposition of Contaminants
Roy M. Harrison, Stuart Harrad, and Jamie Lead
CONTENTS
31.1 Introduction
31.2 Environmental Transport Mechanisms

31.2.1 Atmospheric Transport
31.2.2 Freshwaters
31.2.3 Marine Transport
31.2.4 Soils
31.3 Transfer Mechanisms and Fluxes between Environmental Compartments
31.3.1 Atmosphere–Land Surface Exchange
31.3.2 Air–Plant Exchange
31.3.3 Air–Sea Exchange
31.3.4 Sediment–Water Exchange
31.3.5 Solid–Solution Exchange
31.4 Chemical and Microbiological Breakdown
31.4.1 Rate Expressions
31.4.2 Environmental Lifetimes
31.5 Spatial Distribution of Contaminants
31.5.1 Microscale
31.5.2 National and Regional Scales
31.5.3 Hemispheric and Global Scales
31.6 Temporal Trends in Contaminant Concentrations
31.7 Summary
References
31.1 INTRODUCTION
In most instances, pollutant sources are relatively easy to identify. Point sources especially
present few problems of quantification, while diffuse sources (e.g., runoff from agricultural land)
are more difficult to determine with certainty. The source is, however, only the first part of the
picture, and the period that exists between emission/discharge of a pollutant and contact with the
receptor may contain many varied and interesting processes. It is the aim of this chapter to describe
© 2003 by CRC Press LLC
some of the more important processes involved in pollutant transport and removal from the envi-
ronment and to demonstrate how such processes influence the distribution of pollutants within the
environment. Of particular interest are processes leading to the transfer of chemical substances

between environmental compartments, i.e., water to air, air to soil, etc.
Environmental cycles of chemical elements and compounds are generally termed “biogeochem-
ical cycles.” Ideally, such cycles include quantitative estimates (however uncertain) of the fluxes
between compartments and the total inventory of substance within a given compartment. Such
quantification is difficult even for chemical elements, especially in relation to the flux component.
For chemical compounds subject in some cases to rather rapid chemical change, estimation of
fluxes is even more problematic.
In this chapter, the transport mechanisms responsible for pollutant transfer within and between
environmental compartments are first considered. Mathematical treatments allowing calculation of
transfer fluxes and lifetimes are then described. Some examples of pollutant distributions are given,
indicating where possible the processes responsible. Finally, some examples are provided of the
use of present-day environmental measurements to infer historical concentrations of pollutants.
31.2 ENVIRONMENTAL TRANSPORT MECHANISMS
31.2.1 Atmospheric Transport
Atmospheric motions occur on a number of spatial scales, most notably:
1. Global
2. Synoptic, or large-scale (thousands of kilometers)
3. Mesoscale, or intermediate (tens and hundreds of kilometers)
4. Microscale (ten kilometers and less)
In addition to horizontal transfer processes, movements in the vertical are important, especially
for substances with long atmospheric lifetimes. The atmosphere, viewed in the vertical (Fig
-
ure 31.1), divides readily into discrete regions. The lowermost part, known as the troposphere, is
characterized by decreasing temperature with increasing altitude. This region is the most accessible
to us and consequently is the part most thoroughly observed in scientific terms. Above the tropo
-
sphere lies the stratosphere, a region within which temperature increases with altitude. This is the
region in which ozone mixing ratios peak, as discussed later. The atmospheric regions above the
stratosphere are of little concern in relation to pollution phenomena.
The troposphere is typically also thermally stratified (see Figure 31.1). The main regions are

the boundary layer, typically about 1 km in depth during the daytime but often reducing to only
around 100 m at night, and the free troposphere, which lies between the boundary layer and the
tropopause. These regions are separated by a temperature inversion, which severely limits exchange
between them.
The extent to which a pollutant is subject to either vertical or horizontal movement in the
atmosphere is a function of its atmospheric lifetime, def ined in section 31.4. For substantial
movement of a substance between compartments, the approximate minimum lifetimes τ are indi
-
cated below:
Boundary layer to free troposphere τ > 5 days
Entire tropospheric hemisphere τ > 1 month
Global troposphere τ > 2 years
Troposphere to stratosphere τ > 10 years
© 2003 by CRC Press LLC
To cite some examples, nitrogen dioxide has a chemical lifetime of about a day, and little of
that emitted in the tropospheric boundary layer transfers to the free troposphere. Aerosol emitted
from Chernobyl, which had a lifetime of around 1 month, led to contamination of much of the
northern hemisphere. Methane, with an atmospheric lifetime of about 9 years, is rather well mixed
between northern and southern hemispheres and penetrates in modest amounts into the lower
stratosphere. Chlorofluorocarbons and nitrous oxide, with lifetimes in excess of 100 years, have
no significant tropospheric sinks and mix appreciably into the stratosphere.
Horizontal atmospheric motions are driven by gradients in temperature and pressure that lead
to general circulation, as described in Figure 31.2.
Wind speeds in the boundary layer vary greatly
Figure 31.1 Vertical structure of the atmosphere indicating approximate temperatures and the main regions.
Figure 31.2 The general circulation of the atmosphere.
50
40
30
20

10
8
6
4
2
0
STRATOSPHERE
TROPOSPHERE
Stratopause
Tropopause
High-level cloud
(Stratus)
High-level cloud
(Cirrus)
Storm
clouds
(Cumulonimbus)
Intercontinental
airliner
1
10
10
2
2
3
4x10
10
Boundary layer
200
220 240 260

280
Temperature (K)
Altitude (km)
Pressure (mb)
Polar fronts
Polar fronts
Polar easterlies
Westerlies
Westerlies
Horse latitudes
Trade winds
Trade winds
Equatorial doldrums 0
°
30°
Polar easterlies
60°
Subpolar
lows
Subpolar
lows
© 2003 by CRC Press LLC
from place to place but are typically of the order of 5 m/sec, implying transport distances of around
400 km/day. It is thus possible for a pollutant to travel around the globe at a given latitude in a
matter of days, but more substantial north–south mixing takes months. Atmospheric circulation at
low latitude is dominated by the Hadley circulation, which involves updrafts of air in equatorial
regions, with subsequent movement both north and south at high tropospheric altitudes and sub
-
sidence to form the subtropical high–pressure regions. The existence of two such circulatory cells
on either side of the equator ensures inefficient mixing between northern and southern hemispheres,

and only substances with lifetimes measured in years mix appreciably between the hemispheres.
This is particularly important in limiting movement of contaminants from the heavily populated
northern hemisphere into the far cleaner southern hemisphere.
Mesoscale circulations involve such processes as land–sea breeze circulations that can pro-
foundly influence pollutant movements in coastal regions. Mountain upslope and downslope winds
driven by thermal convective motions may also act as an important transfer route. Processes
occurring on the micro- and mesoscale are important in influencing concentrations of locally
generated pollutants but play only a minor role in transferring pollutants on a larger scale.
31.2.2 Freshwaters
Pollutants entering freshwater may be in the dissolved (< 1nm), colloidal (1 nm – 1 µm), or
particulate (> 1 µm) form. The exact nature of the pollutant species has important implications for
both bioavailability and environmental transport. For instance, biological impact of trace metals is
often related directly to the dissolved (free) metal concentration.
1
In addition, dissolved and colloidal
forms of the pollutant will tend to remain in the water column due to mixing processes, while
particulate forms will tend to settle out of the water column and be incorporated in the sediments.
2
Once in a river or lake, they may remain in their original form or repartition between these different
forms in response to a changed matrix. Important parameters affecting the distribution of pollutants
include microbiological activity, concentration, and nature of organic matter, pH, ionic strength,
redox potential, and so on. These changes can be illustrated by the following examples:
• Rain water (pH ~ 5) falling on lake water (pH ~ 7–8). As the pH increases, metals in the dissolved
form will tend to bind to solid phases. Over this pH range many metals change from 0 to 100%
bound to solid phases.
• Freshwater mixing with seawater in estuaries. Increased ionic strength will cause colloids, and
subsequently sediment, to aggregate out of the water column. The fate of associated pollutants
will also be affected.
• Oxic–anoxic boundary in lakes. The oxidized form of iron (Fe [III]) exists as the solid phase in
oxygenated surface waters. The particles sediment from the water column (with any associated

pollutants) and come into contact with the deoxygenated bottom waters. At this point the iron is
reduced, forming Fe (II), which exists in the dissolved phase. Particulate-phase pollutants may
therefore pass into the dissolved phase.
In reality, the processes occurring are much more complex than indicated, with many competing
reactions occurring on different spatial and temporal scales. For instance, in estuaries, both pH and
chloride concentration as well as ionic strength increase dramatically, with many consequences for
pollutant behavior. In the case of lake waters, as the iron is reductively solubilized across the
oxic–anoxic interface, released trace pollutants may bind to other solid phases and not be released
into the dissolved phase.
Processes such as sorption, precipitation, microbiological activity, and others may affect both
organic and inorganic pollutants. Therefore, the assumption that a substance entering a river in an
effluent will maintain its original physicochemical form will often prove to be incorrect.
Metals are important contaminants that may undergo a wide range of physical changes such as
those outlined above and, in some instances, may also undergo changes in oxidation state leading
© 2003 by CRC Press LLC
to a complete change in physical form. The relevance of such changes to environmental transport
is that they affect the size association and hence the mode of movement in the freshwater system.
Figure 31.3
exemplifies the possible forms of trace metals and their size association.
3
Dissolved contaminants will move freely with flowing water and, although subject to diffusive
movements caused by turbulence, are predominantly influenced by advective transport (i.e., they
are mainly carried along by the water, rather than diffused within it). The behavior of particulate
material is more complex, as it may deposit, entering the bottom sediment. This process is controlled
by two processes. First, the tendency to deposit is a function of the size and settling velocity of
the particles with which the contaminant is associated. However, this tendency is counteracted by
turbulence forces that keep particles in suspension and make them available for movement with
the water. Thus, in relatively static waters (lakes and ponds), there is a strong tendency for particles
to be incorporated into bottom sediments. In fast-moving rivers and estuaries, the turbulent energy
of the water keeps the particles in motion and may even lead to resuspension of bottom sediment.

In such circumstances, contaminant transport may occur by three main mechanisms:
• As truly suspended particles
• By saltation, or “bouncing” of particles along the sediment surface
• As bed load, or bulk motion, of surface sediment
Changing flow conditions within a river may lead to rapidly altering transport properties. An
example is given in Figure 31.4,
which shows measurements of total (suspended plus dissolved)
concentrations of chromium in the River Thames (U.K.) as a function of discharge (flow). As flow
increases from very low values, concentrations of contaminant fall due to dilution of dissolved
material in a greater volume of water. They reach a minimum and then begin to increase for flows
of greater than about 70 m
3 s–1
. This flow is known (from other measurements) to correspond to
that at which riverbed sediments begin to become resuspended. Concentrations of chromium
continue to increase with increasing flow as more bed sediment enters the water.
In slow-moving waters, where much of the suspended sediment enters the bottom sediment, a
historical record of inputs to the lake may be preserved in the bottom sediment. For instance, the
concentration of lead, zinc, and copper in recently deposited sediments in Lake Erie have increased
by several-fold in comparison to sediments deposited ca. 1900.
4
Such
records often reflect changes
in atmospheric deposition to the lake surface, and for substances such as heavy metals
5
and
poly-
nuclear aromatic hydrocarbons, they have been used to reconstruct a record of air quality in the past.
Figure 31.3 Typical physicochemical forms of trace metals in aquatic systems, as related to size association.
(From de Mora, S. J. and Harrison, R. M., Water Res., 17, 723, 1983. With permission.)
Metals

Species
Free
Metal
Ions
Inorganic Ion Pairs
Inorganic Complexes
Low Molecular-
Weight Organic
Complexes
High
Molecular-
Weight
Organic
Complexes
Metal Species
Adsorbed onto
Inorganic Colloids
Metals Associated
with Detritus
Metals Absorbed
into Live Cells
Metals Adsorbed
onto or
Incorporated into
Mineral Solids and
Precipitates
Examples
Mn
2+
Cd

2+
NiCl
+
HgCl
4
2-
Zn-fulvates
Pb-humates Co-MnO
2
Pb-Fe(OH)
3
Cu-clays
PbCO
3(s)
Soluble Colloidal Particulate
© 2003 by CRC Press LLC
The behavior shown in Figure 31.4 for chromium in the River Thames is not common to all
contaminants. In the same study, measurements were made of nitrate and soluble reactive phosphate
(SRP). For nitrate, concentrations were almost independent of flow, suggesting the dominance of
inputs in waters making up the major part of the river flow (surface runoff and groundwater). In
the case of SRP, concentrations of this dissolved species showed a monotonic decrease with flow
rate reflective of dilution of effluent inputs (from sewage treatment works) and no input from
resuspended sedimentary materials.
31.2.3 Marine Transport
While rivers generally take only hours or days to flow to the sea, water has a very long residence
time in the ocean. Consequently, transport and transformation processes control the distribution of
contaminants. The oceanic surface layers to a depth of around 100 m are driven largely by the
wind, the water motion being modified by the Coriolis force that arises from the rotation of the
earth. Oceanic circulations in large part consist of gyres constrained by continental boundaries.
Faster currents occur along western margins, leading to pronounced circulatory features such as

the Gulf Stream, Brazil Current, and Kuroshio Current.
At greater depths within the ocean, circulations are determined by the chemistry and temperature
of the water. Dense waters (either cold or more saline) form in cold polar regions, where the sea
ice formation leads to an increase in salinity. Antarctic bottom water flows north from the Weddell
Sea into the south Atlantic, while North Atlantic deep water moves on a timescale of around 1000
years from the Norwegian Sea through the Indian Ocean into the Pacific.
6
Surface ocean waters are relatively well mixed and show little variation of temperature with
depth. In the layer beneath, termed the “thermocline,” temperatures decline rapidly with depth,
with the most dramatic temperature changes occurring down to about 1 km in equatorial and
temperate latitudes. The deep layers beneath the thermocline show little change in temperature
with depth. The thermocline presents a rather sharp boundary and a considerable barrier against
mixing of the surface and deep layers.
Owing to the relatively slow time scales of oceanic water movement, even in the surface layers,
highest concentrations of contaminants occur in the coastal regions, where inputs are greatest —
from rivers, direct coastal discharges, and atmospheric deposition. Much of this contaminant load
may deposit to the sediments before mixing processes carry it far from coastal waters.
Figure 31.4 Concentration of chromium (suspended plus dissolved) in the waters of the River Thames as a
function of flow rate.
© 2003 by CRC Press LLC
31.2.4 Soils
Soils are by their nature rather immobile. If undisturbed, they may retain a record of contaminant
inputs over a very long period. Many pollutants bind strongly to soils and, if there is input from
the atmosphere, show a very strong surface enrichment. Mechanical mixing by plowing or other
agricultural practices, or bioturbation by burrowing organisms, can cause some vertical and lateral
mixing of pollutants in soils.
Following deposition from the atmosphere, some organic pollutants, particularly polychlori-
nated biphenyls (PCBs), readily undergo volatilization from soil to the extent that for many such
compounds, volatilization represents the principal loss mechanism from soil.
7

This has important
implications for understanding the origins of the continuing atmospheric presence of PCBs (the
manufacture of which was ceased in most western countries in the late 1970s). Specifically, although
accidental releases from PCBs remaining in use continue, such volatilization of previously deposited
material is widely considered to represent the main contemporary source of PCBs in the atmosphere.
31.3 TRANSFER MECHANISMS AND FLUXES BETWEEN ENVIRONMENTAL
COMPARTMENTS
Every time it rains, small amounts of highly persistent compounds, such as chlorinated pesti-
cides, are deposited to land and sea. Dry deposition of particles and gases also contributes to inputs.
This occurs even for some compounds no longer in large-scale production or use. Residues in soils
and waters in parts of the world where heavy usage has occurred are still evaporating into the
atmosphere and are carried to locations remote from their sources, where deposition takes place.
Exchange between environmental compartments is a major route of transfer for some substances
and can lead to unexpected instances of pollution. One such phenomenon occurred in West Cumbria,
U.K., where abnormally high levels of plutonium were found in coastal soils. Detailed investigation
showed that plutonium discharged from the Sellafield reprocessing plant to the Irish Sea was
incorporated into sea spray with an efficiency greater than that expected from its abundance in
seawater and carried back to land in marine aerosol.
5,8
Some of the major processes involved in intercompartmental transfer will now be considered.
31.3.1 Atmosphere–Land Surface Exchange
Transfer from air to land can occur by two major routes:
1. Rainfall (termed “wet deposition”)
2. As dry particles or gas, without the intervention of rain (termed “dry deposition”)
A third pathway, involving “occult” deposition of fog and/or cloudwater droplets, may also be
of localized importance.
Concepts of dry deposition were originally developed to describe the transfer of radioactive
gases and particles from the atmosphere into terrestrial systems. The process of deposition is first
order; that is, the flux to the surface depends linearly upon the atmospheric concentration. The
constant of proportionality is termed the “deposition velocity,” v

g
, defined as:
v
g
flux to surface
atmospheric concentration at 1 m
=
© 2003 by CRC Press LLC
For gases, deposition velocities to soils and vegetation vary greatly according to the affinity of
the gas for the individual surface. Thus, for chlorofluorocarbons, deposition velocities are essentially
zero, while for highly reactive and water-soluble nitric acid vapor, v
g
has a typical value of 2 to 3
cm s
–1
. The deposition velocity is not constant for a given gas or a given surface. While one can
make approximate statements about the magnitude of v
g (
such as
that above for nitric acid), in
reality v
g
v
aries with the surface characteristics and atmospheric properties at the time of deposition.
This may be expressed as follows for the total resistance to deposition, R.
(31.1)
where r
a
is the aerodynamic resistance, or the resistance to transfer downward through the atmo-
sphere to within 1 to 2 mm of the surface; r

b
is the boundary layer resistance, which is the resistance
to transfer through a laminar layer of air of about 1 mm thick over the individual roughness elements
of a surface; and r
c
is the canop
y, or surface resistance, which describes the resistance of the surface
itself to take-up of the gas. For a sticky, reactive molecule like nitric acid, r
c
i
s essentially zero,
while for chlorofluorocarbons, it is almost infinite and accounts for the differing depositional
behavior of these gases. For ozone, r
c
over vegetation typically varies with time of day according
to the opening of stomatal apertures of the vegetation, necessary for rapid ozone deposition.
Deposition velocities are not as strong a function of chemical properties in airborne particles
as in gases. The main determinant of deposition velocity is the particle size (see Figure 31.5), with
high values of v
g
appl
icable to large particles due to their inertial properties and very small particles
due to their high diffusivities, which cause them to behave more like gases.
9
Between the two lies
a region of low dry depositional efficiency. For particles in this size range, around 0.1 to 1 µm in
diameter, wet scavenging is also inefficient, and atmospheric lifetimes are long — of the order of
7 to 40 days. Chemical composition may affect dry deposition, as hygroscopic particles are liable
to grow adjacent to a humid surface, leading to enhanced deposition.
Some gases are released by soils or vegetation. If their concentration immediately adjacent to

the surface exceeds that in the atmosphere above, the concentration gradient will cause them to
diffuse upward, and the net flux will be from surface to atmosphere. In this instance, the concept
of deposition
velocity is of little value. A gas that normally exhibits upward fluxes from soil is
Figure 31.5 Typical variation of aerosol particle deposition velocity with particle size over land.
R
1
v
g
r
a
r
b
r
c
++==
© 2003 by CRC Press LLC
nitrous oxide, N
2
O. Ammonia, NH
3
, shows ambivalent behavior; net fluxes from fertilized soils
are commonly upward into the atmosphere, while over unfertilized soils, within which concentra
-
tions of ammonium (and by inference ammonia) are very low, deposition normally occurs.
31.3.2 Air–Plant Exchange
Air-to-plant transfer of persistent organic pollutants (POPs), such as dioxins, PCBs, and poly-
cyclic aromatic hydrocarbons (PAH), occurs readily, primarily as a result of vapor-phase deposition
onto the leaf surface, either as an equilibrium partitioning process or as a kinetically-controlled
process, although particle-bound deposition can be influential for some compounds.

10
Given the
facile nature of air–plant transfer of POPs, it is important to evaluate the quantitative role that
vegetation plays within the overall biogeochemical cycling of such pollutants. Several authors have
speculated that because of the vast surface area covered by vegetation, this role could be significant.
Indeed, there is ample evidence that vegetation, particularly forests, remove a substantial quantity
of airborne POPs, with consequent reductions in their atmospheric residence times. A steady-state
mass-balance model determined that vegetation removed 4% of PAHs from the atmosphere in the
northeastern region of the United States. It is evident that vegetation plays a significant role in the
biogeochemical cycling of PAHs and other POPs possessing similar propensity for atmosphere-to-
foliage transfer. The implications of this phenomenon in terms of human and ecotoxicological
impacts are at present unclear, and this research topic is likely to receive much attention in the future.
31.3.3 Air–Sea Exchange
For many large bodies of water, the atmosphere is a major route of contaminant input. For
example, for the North Sea, which is heavily influenced by adjacent industrialized countries,
atmospheric deposition is the major source of some trace metal inputs.
Particles deposit to water surfaces in much the same manner as to land. There are clearly
differences arising from the resistance terms in Equation 31.1 that are not identical for the two
surfaces or the air above them. In general, similar patterns of behavior are observed, although there
is stronger wind-speed dependence in the case of the sea, since surface characteristics change
appreciably as high winds break up the sea surface into waves and spray.
Aerosol produced from the sea surface itself can be enriched in trace elements that are trans-
ported over land. The extent of enrichment is controversial, with some research indicating a major
impact on aerosol and rainwater composition and other papers suggesting little effect in the case
of trace metals. One clear example of enrichment is plutonium, cited above. Plutonium is enriched
in aerosol and deposition some 30- to 500-fold relative to its abundance in bulk seawater. Fortu
-
nately, marine aerosol is of relatively large particle size and deposits within a rather narrow coastal
band of land; the majority of deposition occurs within 5 km of the coast.
Incorporation of contaminants in rainwater leads to wet deposition to both land and sea surfaces.

Scavenging of contaminants may occur either within the cloud (termed “rainout”) or by falling
raindrops (termed “washout”). The overall efficiency of the process is often described by the
scavenging ratio, W, also known historically as the Washout Factor:
Values of W for particulate species are typically around 200 to 1000, implying crudely that
substances present in air at microgram-per-cubic-meter concentrations will occur in rain at milli
-
gram-per-liter-levels, as a cubic meter of air weighs 1.2 kg at 25ºC and 1 atm pressure. An alternative
W
concentration in rain mg kg
1–
()
concentration in air mg kg
1–
()
=
© 2003 by CRC Press LLC
descriptor of wet removal used by numerical modelers is the washout coefficient, λ. This is a first-
order rate constant in the equation
where dc/dt is the rate of change of airborne concentration with time, t. For aerosol species, λ
typically takes values of around 10
–4 s–1
.
The sea may act as a source of atmospheric trace gases. A notable example is that of dimethyl
sulfide released by marine phytoplankton. It may also act as a sink for reactive gases such as nitric
acid and other soluble species.
The direction of flux between air and sea
11 i
s determined by Henry’s Law, which describes the
equilibrium between dissolved and gaseous forms of a substance. At equilibrium:
where C

w and Ca
are the concentrations in water and air, respectively, C
w
(eq) is the concentration
in water at equilibrium with the atmosphere, and H is Henry’s Law constant.
If, however, C
w
≠ C
w
(eq), then a net flux of material to air, or from water to air, will occur. If
C
a
H
–1 > C
w
, net transfer into the water occurs; and if C
a
H
–1 < C
w
, then the net flux is to the
atmosphere.
Suppose that:
∆C = C
a
H
–1 – C
w
The rate of gas transfer is:
F = K

(T)w
∆C
where K
(T)w
is termed the “transfer velocity.” By analogy with Equation 31.1, one can write:
(31.2)
where k
a
and k
w
are indi
vidual transfer velocities for chemically nonreactive gases in air and water
phases, respectively, and α (= k
reactive
/k
inert
) is a factor that quantifies any enhancement of gas transfer
in the water due to chemical reaction. The resistance r
a and
r
w
are directly
analogous to those in
Equation 31.1; and for chemically reactive gases, usually r
a >>
r
w
, and atmospheric transfer limits
the overall flux. For less reactive gases, r
w i

s more difficult to predict. It is highly wind-speed
dependent, as exemplified by Figure 31.6, which shows the overall transfer velocity for carbon
dioxide (for which k
w
i
s dominant) as a function of wind speed at 10 m.
13
31.3.4 Sediment–Water Exchange
The distribution of trace elements in aquatic systems is affected by a number of processes,
which are exemplified in Figure 31.7. One important process in the water column is the distribution
between the solid-solution interface (see next section). Incorporation of contaminants into sus
-
pended solids that subsequently aggregate and deposit to form bottom sediments is a major removal
mechanism of the pollutants. Sedimentation rates, as expressed by the rate of increase at the bottom
sediment surface, vary greatly from millimeters per 100 years in the deep ocean to centimeters per
dc
dt
λt–=
C
w
eq() C
a
H
1–
=
1
K
T()w

1

αk
w

1
Hk
a
+ r
w
r
a
+==
© 2003 by CRC Press LLC
year in some lakes and coastal environments. Other important controls on the transport of contam-
inants are the resuspension of particles, the entrapment of water in pore spaces, the upward flow
of pore water due to hydrostatic pressure gradients, and diffusional fluxes of dissolved species.
2
Recent work with microelectrodes and other microscale techniques
14
has shown that sediments are
extremely heterogeneous over ranges of hundreds of microns and that dissolved concentrations of
major and minor species will fluctuate not only vertically and temporally but also laterally, i.e., in
three dimensions. These heterogeneities often relate to the effects of biogenic processes.
15
Figure 31.6 Air–sea transfer velocities for carbon dioxide at 20°C as a function of wind speed at 10 m (ms
–1 or
Beaufort Scale). The graph combines experimental data (points) with a theoretical line. (From Watson,
A.J., Upstill-Goddard, R.C., and Liss, P.S., Nature [London], 349, 145, 1991. With permission.)
Figure 31.7 Processes affecting the distribution of trace elements in aquatic systems. (From Tessier, A. et al.,
in Chemical and Biological Regulation of Aquatic Systems, CRC Press, Boca Raton, FL, 1994.)
© 2003 by CRC Press LLC

The net flux of a substance across the water-sediment interface, F
z = 0
, is expressed by
(31.3)
where
φ= porosity (volume fraction occupied by water)
D = diffusion coefficient
C = concentration in solution
C
s
= concentration in solids
U = rate pore water advection
U
s
= sedimentation rate
z = vertical distance
Thus, of the three terms in the right-hand side, the first term represents diffusive transfer, the
second term advection in pore water, and the third term transfer in solids.
In a uniform, noncompacting sediment, U = U
s
, and from above, C
s
= K
d
C. Substitution yields:
(31.4)
The relative importance of the two terms, representing diffusion and sedimentation, depends
upon the sedimentation rate and the magnitude of K
d
. If either of these is large, it is likely that

sedimentation processes will dominate the flux. If both are small, diffusive transfer may be dom
-
inant, as often occurs in the deep ocean.
A fuller consideration of solute transport would also consider biological processes (bioturbation
and bioirrigation by macroinvertebrates) and physical processes (waves, currents, and tides).
12
Although bioturbation appears to be negligible, the other processes can significantly enhance
movement. The role of the biofilms may also be important in pollutant dispersion.
Interesting processes can arise in redox–active sediment components. Examples are with iron
and manganese. Both elements are present in oxidizing waters in highly insoluble oxidized forms
such as Fe
2
O
3
an
d MnO
2
, which become incorporated in surface sediments. Progressively, the
sediment becomes buried with freshly sedimenting solids and becomes anoxic as microorganisms
consume oxygen more rapidly than it can be supplied from the overlying water. In the anoxic
region, chemical reduction takes place, and Fe(III) and Mn(IV) are converted to, respectively, Fe(II)
and Mn(II). These are both liable to be more soluble than the more oxidized forms (especially
Mn), and Mn(II) diffuses upward through the interstitial water, from where it may either enter the
overlying water or reprecipitate in the oxic layer.
16
Figure 31.8
shows typical schematic profiles
of solid-associated and dissolved manganese. The subsurface maximum in dissolved manganese
occurs where the outward flux and reprecipitation balance the upward supply from greater depth.
Solid manganese below this depth is mostly present as MnCO

3
. Analysis of a core of sediment for
manganese is liable to reveal a near-surface maximum for the above reasons. This should not be
confused with near-surface enrichment of other sediment components not subject to the same redox
chemistry, which in more recent years has often been seen as a result of enhanced rates of
contaminant deposition (see Section 31.6). Post-depositional changes in sediments are common
and varied and are grouped under the term diagenesis.
Diagenetic processes in sediments can lead to the release of contaminants to the overlying
water. Thus, for example, microbial decay of organic matter and reduction of iron and manganese
in sedimentary solids can lead to rerelease of adsorbed or complexed material that, if not reabsorbed,
F
z=0
φ D
dC
dz

– UC U
s
C
s
++


=
F
z=0
φ D–
dC
dz

UK
d
1+()C+
z=0
=
© 2003 by CRC Press LLC
may be released into the overlying water. This is most likely to occur to an appreciable degree if
the bottom water of a lake becomes anoxic for an appreciable period, causing even the surface
sediment to be anaerobic.
31.3.5 Solid–Solution Exchange
Although these exchanges occur within — rather than between — environmental compartments,
e.g., in surface waters, groundwaters, or sediment pore waters, the processes occurring at the
solid–solution interface within the water body are essential to understanding the transport of
pollutants. Controlling processes, such as biological activity and sedimentation, are highly depen
-
dent on reactions occurring at the solid–solution interface. The chemical weathering of minerals
that, in large part, determines the composition of natural waters and plays a part in controlling
atmospheric carbon dioxide levels through the action of carbonic acid (see below, dissolution of
silicates and calcium carbonate) is also dependent on these solid-solution reactions.
CaSiO
3
+ 2CO
2
+ 3H
2
O = Ca
2+
+ 2HCO
3
–

+ H
4
SiO
4
CaCO
3
+ 2H
+
= Ca
2+
+ H
2
O + CO
2
The solid–solution interface is often dominant in regulating the concentrations of dissolved
trace elements and can be modelled through the use of a distribution coefficient given by the
equation,
Although the distribution coefficient is useful in modelling processes such as the sorption of
organic pollutants on organically coated particles, it is also limited in that it provides a conditional
constant, which will change with changing solution conditions. More recently, understanding the
partitioning of metals to oxide or natural organic surfaces has been performed in terms of surface
complexation theory.
17,18
However, full understanding of these processes will come only in the
Figure 31.8 Schematic profiles of manganese in bottom sediment dissolved in the interstitial waters (broken
lines) and associated with the solid sedimentary components (solid line).
K
d
concentration in solid phase material mg kg
1–

()
concentration in dissolved phase mg kg
1–
()
=
© 2003 by CRC Press LLC
distant future. The fate of organic pollutants such as PCBs is modified as the “hydrophobic effect.”
This results in a tendency of organic chemicals to avoid contact with water and results in sorption,
particularly on natural organic matter (primarily humic substances). The humic substances them
-
selves tend to sorb to the surface of mineral and microbiological particles, with significant conse-
quences for the fate of the pollutants.
31.4 CHEMICAL AND MICROBIOLOGICAL BREAKDOWN
The atmospheric lifetime of many chemical species is dictated by chemical reactions rather
than depositional processes. This is also true for some more reactive species in aquatic systems.
A simplistic description of relevant processes is given here.
31.4.1 Rate Expressions
If contaminant X is removed by reaction with reagent Y in a one-step chemical process,
X + Y → products
the rate of loss of X is given by:
where k is the rate constant for the reaction. If Y is a species whose concentration is not greatly
variable with time, either because it is present in great excess or it is continually replaced after
consumption, such that [Y] ≅ constant, then
where k′ = k[Y].
This rate expression has a first-order form and is termed “pseudo first-order.” The implication
of this form of equation, which prevails frequently in the environment, is that a contaminant that
ceases to be emitted diminishes in concentration according to a negative exponential form, as with
the decay of a radioactive source.
The second-order equation, if applicable, may be used to estimate the rate of removal of a
substance by a specific chemical reaction. For example, the main sink for nitrogen dioxide is

conversion to nitric acid by reaction with the hydroxyl radical, ·OH. For this reaction,
k = 1.1 × 10
–11
cm
3
molecule
–1
s
–1
and
and
dX[]–
dt
kX[]Y[]=
dX[]–
dt
k′ X[]=
d
dt
– NO
2
[]1.1 10
11–
×()110
6
×()NO
2
[]=
d
dt

NO
2
[]–
NO
2
[]
1.1 10
5–
s
1–
× 0.04 h
1–
==
© 2003 by CRC Press LLC
Thus, loss of NO
2
occurs at a rate of approximately 4% per hour at the above concentration of
OH radical. At twice that OH radical concentration, the loss is 8% per hour.
In aquatic chemistry, many of the reaction processes are not simple one-step reactions, and
complex rate expressions result in noninteger reaction orders.
In soils and sediments, many decay and transformations affecting chemicals are microbiolog-
ically mediated. For example, in oxidizing sediments, organic materials are oxidized, with con-
sumption of molecular oxygen, O
2
. When the O
2
is depleted, the system becomes anoxic, and
anaerobic organisms take over. Initially, nitrate-reducing bacteria are favored. Nitrate is used as an
oxygen source and ammonium is produced as product. Subsequently, sulfate is used as an oxygen

source and converted to sulfide in the process. Finally, carbohydrate material itself is used as a
source of oxygen by methanogenic bacteria, and methane and carbon dioxide are the products.
31.4.2 Environmental Lifetimes
In discussing environmental lifetimes, it is useful to think of the analogy of a bath, with a
source of input water and removal via an overflow. At equilibrium, the level of water is constant,
and the inflow is exactly equaled by the outflow through the overflow. Then, the faster the input,
the more rapid the outflow and the shorter the residence time (or lifetime) of water in the bath.
This concept is easily related to a mixed reservoir, such as the atmosphere or ocean, but becomes
more nebulous when considering solid-earth processes.
To a crude approximation, many environmental contaminants are in some form of equilibrium,
and sources (S) equal removal rates (termed “sinks”
®
):
S = R
If R and S refer to a given environmental compartment, and A is the total quantity of the
contaminant in that compartment, the lifetime (or residence time), τ, is defined as:
In practice, the lifetime defined in this way is the time taken for the concentration of the species
to diminish to
1
/e of its initial concentration if all source processes ceased.
Alternatively, dividing top and bottom by the volume of the environmental compartment gives:
or ratio of concentration to flux, S′.
If removal is via a first-order or pseudo first-order process, as described above, then:
and
τ
Akg()
S kg s
1–
()

A
R
==
τ
A[]kg m
3–
()
S′ kg m
3–
s
1–
()
=
S′
dA[]
dt
kA[]==
τ
A[]
kA[]

1
k
==
© 2003 by CRC Press LLC
where k is the first-order constant for the loss of component A. This is a very useful result and
can be used in the context of removal by chemical or depositional pathways. Taking the example
of nitrogen dioxide above,
k′ = k[·OH]
= 0.04 h

–1
and
τ = 25 h
For many atmospheric trace gases, removal is by reaction with the hydroxyl radical, and
For methane,
k = 5.6 × 10
–15 cm3 mo
lecule
–1 s–1
and
τ = 5.7 years for [·OH] = 1 × 10
6 molecule cm–3
Concentrations of hydroxyl radical do in fact vary diurnally and with latitude and season. The
use of long-term average values (as above) for lifetime estimation is therefore justified for the
longer-lived substances (e.g., methane); but for short-lived species like nitrogen dioxide, more
precise values for the location and time are appropriate.
As mentioned in Section 31.3, washout coefficients are first-order rate constants for wet
deposition. Thus, for λ = 10
–4
sec
–1
, the lifetime with respect to removal by this pathway is 10
4
sec, or 2.8 h. Such a value could obviously apply only during periods of rain and would not be
relevant to dry spells.
Lifetimes can also be calculated in relation to dry deposition processes. In this case, the rate
constant is given by
k = v
g
/H

where v
g
is the deposition velocity and H the mixing depth, or depth of the lowest mixed layer of
the atmosphere, normally the boundary layer. Thus, for sulphur dioxide, for which v
g
≅ 1 cm s
–1
and a typical mixing depth of 1000 m,
τ = H/v
g
= 28 h
Within the ocean, dissolved materials may be scavenged by adsorption onto particles, thereby
accelerating their deposition. Using the same form of mathematical analysis, deep-water scavenging
residence times vary from 0.4 years for particles, to 10 to 50 years for metals such as tin, iron,
cobalt, and lead, to over 10
4
years for cadmium.
τ
1
k·OH[]
=
© 2003 by CRC Press LLC
31.5 SPATIAL DISTRIBUTION OF CONTAMINANTS
As discussed above, spatial distributions of contaminants result from a number of factors,
including source locations, dispersal processes, and sink processes. The spatial extent of the
distribution from a single source will depend critically on the lifetime of the contaminant, as those
of shorter lifetimes will not survive over any great distance.
31.5.1 Microscale
Good examples of microscale distributions of contaminants arise from dispersion from point
sources. Effective numerical models exist for describing the three-dimensional concentration

distribution of pollutants emitted from elevated point sources into the atmosphere. The instanta
-
neous distribution of concentrations is described adequately for most purposes by the Gaussian
plume model.
19,20
Emissions to atmosphere of pollutants such as heavy metals from point sources can lead to
distributions of pollutants in local surface soils directly analogous to those for atmospheric con
-
centrations. Many surveys have been carried out of metal concentrations in soils and vegetation
around smelters, exhibiting just this kind of distribution, except where low-level, fugitive sources
of coarse dusts, such as stockpiles, lead to highly enhanced deposition close to the works. Frequently,
this has led to severe contamination, which persists unless subject to cleanup.
A well-known phenomenon of past decades has been the contamination of roadside soils with
lead from vehicle exhausts. A typical distribution involves highly elevated concentrations at curb
-
side, falling exponentially to background levels within about 100 m. Such soils typically show a
strong surface enrichment of lead, with concentrations declining rapidly in lower layers, reaching
local background concentrations by about 15 cm deep. Highly elevated concentrations of lead in
soils may also be found in areas of lead mineralization and historical lead mining, where lead-rich
soils have contaminated local soils.
Both comparable methods based on Fickian diffusion and systems procedures are available to
estimate by numerical means the dispersion of contaminants in rivers.
21
31.5.2 National and Regional Scales
Pollutants with major localized sources may give substantial and significant variations in
concentration on the micro- or mesoscale but contribute in only a minor way to national or regional
patterns of concentration. Secondary atmospheric pollutants often yield more meaningful patterns
on the latter scale, as they are produced more slowly and are less influenced by individual sources;
an example is hydrogen ion in rainwater. Strong acidity is generated only rather slowly from
oxidation of sulphur dioxide (ca. 1 to 2% per hour) and nitrogen dioxide (ca. 5% per hour).

Figure 31.9
shows the distribution of rainwater pH over North America at the height of the
acidification problem, which is now less severe. Distinct spatial gradients relate to sources, mete-
orology, and local geology, since airborne carbonate particles can neutralize acids. Another example
of a secondary pollutant showing meaningful spatial patterns in concentration is ozone. Analysis
of measurement data in Europe indicates clear north–south gradients in episodic peak hourly ozone,
both nationally and across western Europe as a whole.
31.5.3 Hemispheric and Global Scales
As mentioned above, the atmospheric circulation provides a strong barrier to tropospheric
mixing between northern and southern hemispheres; crossing of the equatorial regions occurs
appreciably only for pollutants with lifetimes of a year or more. Ship-borne measurements of air
© 2003 by CRC Press LLC
composition have revealed north–south gradients of many atmospheric constituents. In some
instances, these show sharp discontinuities at the equator, which is consistent with a predominantly
northern hemisphere source and little transfer into the southern hemisphere.
As previously mentioned, persistent organic pollutants (POPs) like PCBs are capable of vola-
tilizing from soils and water. Once airborne, they readily undergo long-range atmospheric transport
either in the vapor phase or bound to aerosol. It is via such atmospheric transport, which occurs
in a repeated cycle sometimes referred to as “the grasshopper effect,” that these chemicals distribute
in a ubiquitous fashion. This universal distribution is illustrated by the presence of POPs in polar
regions, which is a matter of concern, for although inputs to such locations are small, the low
temperatures minimize volatilization losses, leading to a steady accumulation of the overall pollutant
burden. This accumulation is compounded by the reduced biomass-to-surface-area ratio of polar
regions, with the result that the pollutant burden is distributed among a much smaller biomass than
in industrial areas. These factors account for the observations of PCB concentrations in the breast
milk of Canadian Inuit mothers, which exceed those found in women from urban Canada.
31.6 TEMPORAL TRENDS IN CONTAMINANT CONCENTRATIONS
High-quality environmental measurements of trace contaminants have been made only in
relatively recent years. There are, therefore, relatively few examples of clear temporal trends
discernible from contemporaneous environmental measurements. One very clear example is that

of tropospheric carbon dioxide measured at Mauna Loa, Hawaii by Keeling and coworkers
22
since
the late 1950s. High-quality data show a very clear upward trend modulated by a seasonal cycle
due to varying exchange fluxes with terrestrial biota. Since CO
2
has a relatively long tropospheric
lifetime, similar general trends are seen at other sites within both northern and southern
hemispheres.
It is more difficult to determine trends that predate contemporaneous measurements. However,
two methods have been particularly successful. The first involves analysis of ice cores. Gases
Figure 31.9 Calculated ground-level concentration distributions (µg m
–3
) of NO
x
around an elevated point source
within a fertilizer works. + denotes center of works; • denotes sampling sites used to measure
ground-level concentrations. (From Boutron, C.F. et al., Nature, 353, 153, 1991. With permission.)
© 2003 by CRC Press LLC
become trapped in ice during snowfall, and careful retrospective analysis of dated ice core samples
has revealed valuable information. Examples are carbon dioxide and methane, determined in
Antarctic ice
23
u
p to 1.6 × 10
5
y
ears old. Metallic pollutants are also incorporated into snow and
ice, and Figure 31.10 shows data for lead in Greenland ice, determined from several studies,
showing a massive increase in atmospheric lead deposition, commencing with the Industrial Rev

-
olution and accelerating rapidly after 1950 as use of leaded gasoline increased. More recently,
deposited ice shows a strong decline from around 1970, when leaded gasoline use was curtailed,
first in North America and subsequently in many other developed countries.
Lake sediment cores can also reveal interesting patterns of historical inputs, as the sediments
are built up of layers laid down sequentially, the most recent at the water interface. Concentrations
of trace contaminants in such sediment cores reflect input processes and subsequent diagenetic
changes. In some instances, however, profiles of pollutants have been shown to reflect clearly
historical inputs of pollutants such as trace metals from the atmosphere to the water body and hence
provide a relative measure of trends in airborne concentrations.
The environmental persistence of POPs means that correctly stored archived samples as well
as sediment and snow-pack cores provide an invaluable resource for elucidating past changes in
contaminant concentrations.
25
Kno
wledge of such past trends can be correlated with contempo-
raneous trends in source activities and used to build further understanding of the likely impact
of emission-control strategies on future concentrations. Figure 31.11
shows the temporal varia-
Figure 31.10 Changes in lead concentrations in Greenland ice and snow from 5500 BP to present. (From
Boutron, C.F. et al., Nature, 353, 153, 1991. With permission.)
Figure 31.11 Temporal trends in concentrations of dioxins (PCDD/Fs) in soil measured in the southern United
Kingdom.
19901970195019301910189018701850
20
40
60
80
100
Year

PCDD/F (ng/kg)
© 2003 by CRC Press LLC
tion of dioxin concentrations in archived soil taken in south England between 1850 and 1990.
26
This study shows that a low, essentially constant “background” of dioxins existed in the soil
throughout the latter part of the 19
th
century; since then, dioxin concentrations have risen
dramatically. The inference from this is that, while some natural sources do exist, notably forest
fires (which may arguably be of significance in some countries), the environmental burden of
these compounds has its source in human activity, in particular, the manufacture and use of
organochlorine chemicals (such as chlorophenols) and combustion processes (such as waste
incineration and coal combustion).
31.7 SUMMARY
Distributions of many pollutants are measured on spatial scales varying from the local micros-
cale to the global scale. These may generally be understood in terms of:
1. Distribution of sources
2. Transport mechanisms and intercompartmental transfer
3. Environmental lifetimes
4. Distribution of sinks
While short-lived contaminants are most readily identified close to their source, the more
persistent substances, such as heavy metals and PCBs, may achieve a truly global distribution due
to atmospheric transport and deposition to soils and surface waters.
Good mathematical treatments are available from many transfer processes and sink mechanisms,
although considerable uncertainties still exist in relation to the use of appropriate values of param
-
eters such as diffusion coefficients and transfer velocities. The analysis of pollutant deposits
accumulated historically in media such as sediments and polar ice may be used to construct a
picture of past pollutant concentrations.
REFERENCES

1. Campbell, P. G. C., Interactions between trace metals and aquatic organisms: A critique of the free-
ion activity model, in Metal Speciation and Bioavailability, Tessier, A. and Turner, D. R., Eds., John
Wiley and Sons, Chichester, 1995.
2. Stumm, W. and Morgan, J. J., Aquatic Chemistry, 3rd ed., John Wiley and Sons, New York, 1996.
3. de Mora, S. J. and Harrison, R. M., The use of physical separation techniques in trace metal speciation
studies, Water Res., 27, 723, 1983.
4. Sigg, L., The regulation of trace elements in lakes: The role of sedimentation, in Chemical and
Biological Regulation of Aquatic Systems, Buffle, J. and de Vitre, R. R., Eds., Lewis Publishers, Boca
Raton, FL, 1994.
5. Harrison, R. M., Integrative aspects of pollutant cycling, in Understanding Our Environment: An
Introduction to Environmental Chemistry and Pollution, Harrison, R. M., Ed., Royal Society of
Chemistry, London, 1992.
6. Stowe, K. S., Ocean Science, John Wiley and Sons, New York, 1979.
7. Ayris, S. and Harrad, S., The fate and persistence of polychlorinated biphenyls in soil, J. Environ.
Monit., 1, 395, 1999.
8. Cawse, P. A., Studies of Environmental Radioactivity in Cumbria, Part 4. Cesium-137 and Plutonium
in Soils of Cumbria and the Isle of Man, United Kingdom Atomic Energy Authority, UKAEA Report
No. AERE-9851, 1980.
9. Davidson, C. I. and Wu, Y. L., Dry deposition of trace elements, in Control and Fate of Atmospheric
Trace Metals, Pacyna, J. M. and Ottar, B., Eds., Kluwer Academic, Dordrecht, The Netherlands, 1989.
© 2003 by CRC Press LLC
10. Currado, G. M. and Harrad, S., Transfer of POPs in vegetation: Implications and mechanisms, in
Persistent Organic Pollutants: Environmental Behaviour and Pathways of Human Exposure, Harrad,
S., Ed., Kluwer Academic, Dordrecht, The Netherlands, 2001.
11. Liss, P. S. and Merlivat, L., Air-sea exchange rates: Introduction and synthesis, in The Role of Air-
Sea Exchange in Geochemical Cycling, in Buat-Menard, P., Ed., Reidel, Dordrecht, The Netherlands,
1986.
12. Tessier, A., Carignan, R., and Belzille, N., Processes occurring near the sediment-water interface:
Emphasis on trace elements, in Chemical and Biological Regulation of Aquatic Systems, Buffle, J. and
de Vitre, R. R., Eds., Lewis Publishers, Boca Raton, FL, 1994.

13. Watson, A. J., Upstill-Goddard, R. C., and Liss, P. S., Air-sea gas exchange in rough and stormy seas
measured by a dual-tracer technique, Nature(London), 349, 145, 1991.
14. Brendel, P. J. and Luther, G. W., Development of a gold amalgam voltammetric microelectrode for
the determination of dissolved Fe, Mn, O
2
, and S (-II) in porewaters of marine and freshwater
sediments, Environ. Sci. Technol., 29, 751–761, 1995.
15. Shuttleworth, S. M., Davison, W., and Hamilton-Taylor, J., Two-dimensional and fine structure in the
concentration of iron and manganese in sediment pore-waters, Environ. Sci. Technol., 33, 4169–4175,
1999.
16. Harrison, R. M., de Mora, S. J., Rapsomanikis, S., and Johnston, W. R., Introductory Chemistry for
the Environmental Sciences, Cambridge University Press, New York, 1991.
17. Dzomabak, D. A. and Morel, F. M. M., Surface Complexation Modelling — Hydrous Ferric Oxide,
John Wiley and Sons, New York, 1990.
18. Turner, D. R., Problems in trace metal speciation modeling, in Metal Speciation and Bioavailability,
Tessier, A. and Turner, D. R., Eds., John Wiley and Sons, Chichester, U.K., 1995.
19. Harrison, R. M. and McCartney, H. A., A comparison of the predictions of a simple Gaussian plume
dispersion model with measurements of pollutant concentration at ground-level and aloft, Atmos.
Environ., 14, 489, 1980.
20. Williams, M. L., Atmospheric dispersal of pollutants and the modelling of air pollution, in Pollution:
Causes, Effects and Control, 4
th
ed., Harrison, R. M., Ed., Royal Society of Chemistry, London, 2001.
21. Young, P. C., Quantitative systems methods in evaluation of environmental pollution problems, in
Pollution: Causes, Effects and Control, 2
nd
ed., Harrison, R. M., Ed., Royal Society of Chemistry,
London, 1990.
22. Keeling, C. D. et al., A three-dimensional model of atmospheric CO
2

transport based on observed
wind, 4. Mean annual gradients and interannual variations, Geophys. Monogr., 55, 305, 1989.
23. Inter-Governmental Panel on Climate Change, Climate Change, Cambridge University Press, New
York, 1991.
24. Boutron, C. F., Gorlack, U., Candelone, J P., Bolshov, M. A., and Delmas, R. J., Decrease in
anthropogenic lead, cadmium and zinc in Greenland snows since the late 1960s, Nature, 353, 153,
1991.
25. Sanders, G., Temporal trends in environmental contamination, in Persistent Organic Pollutants:
Environmental Behaviour and Pathways of Human Exposure, Harrad, S., Ed., Kluwer Academic
Publishers, Dordrecht, The Netherlands, 2001.
26. Kjeller, L., Jones, K. C., Johnston, A. E., and Rappe, C., Increases in the polychlorinated dibenzo-p-
dioxin and -furan content of soils and vegetation since the 1840s, Environ. Sci. Technol., 25, 1619,
1991.
© 2003 by CRC Press LLC
CHAPTER 32
Bioaccumulation and Bioconcentration
in Aquatic Organisms
Mace G. Barron
CONTENTS
32.1 Introduction
32.2 Bioaccumulation from Sediments
32.2.1 Determinants of Bioaccumulation
32.2.2 Bioaccumulation Factors
32.2.3 Bioaccumulation of Metals
32.2.4 Bioaccumulation of Organic Contaminants
32.2.4.1 Bioavailability
32.2.4.2 Equilibrium Partitioning Theory
32.3 Bioconcentration from Water
32.3.1 Bioconcentration Processes
32.3.1.1 Uptake

32.3.1.2 Distribution and Lipid Partitioning
32.3.1.3 Elimination
32.3.2 Estimation of Bioconcentration
32.3.3 Bioavailability of Waterborne Contaminants
32.3.4 Bioconcentration of Metals
32.3.5 Bioconcentration of Organic Contaminants
32.4 Biomagnification and Trophic Transfer
32.4.1 Overview
32.4.2 Dietary Absorption
32.4.3 Dietary Bioavailability
32.4.4 Aquatic-Based Food Webs
32.4.4.1 Biomagnification
32.4.4.2 Food-Web Models
32.4.4.3 Maternal Transfer
32.5 Summary
References
© 2003 by CRC Press LLC
32.1 INTRODUCTION
Concern for the bioaccumulation of contaminants arose in the 1960s because of incidents such
as toxicity from methylmercury residues in shellfish and avian reproductive failure due to chlorinated
pesticide residues in aquatic species. Bioaccumulation models were developed in the 1970s to account
for processes such as the partitioning of hydrophobic chemicals from water to aquatic organisms.
1
The study of contaminant accumulation and predictive model development expanded in the 1980s
to include bioaccumulation from sediments, food chain biomagnification, and carcinogenesis in feral
species. Today, determining contaminant bioaccumulation in aquatic organisms is an essential com
-
ponent of assessing risks to piscivorous wildlife. Additionally, current research is evaluating the
association between contaminant tissue residues and adverse effects in aquatic organisms.
2

This chapter presents an overview of the principles and determinants governing bioaccumula-
tion from sediments and water and biomagnification in aquatic-based food webs. Organic and
metal contaminants are discussed, with an emphasis on hydrophobic organics. The objective of
this chapter is to elucidate concepts relating to bioaccumulation rather than present an exhaustive
review of the literature.
32.2 BIOACCUMULATION FROM SEDIMENTS
32.2.1 Determinants of Bioaccumulation
Aquatic sediments are formed from the deposition of particles and colloids and can act as both
a sink and a source of pollutants. Long-term contaminant input leads to sediment concentrations
that can exceed the water concentration by several orders of magnitude because of partitioning of
chemicals onto sediment-binding sites. Determinants of the bioavailability of sediment-associated
contaminants to aquatic organisms include physical/chemical interactions between the chemical
and sediment constituents and biological processes determining exposure and uptake. Binding of
hydrophobic organics and metals to sediment appears to be controlled by different mechanisms.
The majority of information relating to bioaccumulation from sediments relates to hydrophobic
organics, with more limited information on metal or organic anion and cation bioaccumulation.
Bioaccumulation of sediment-associated contaminants is highly species-dependent because of
the diversity of feeding ecology and living habits of benthic organisms.
3–5
Exposure pathways from
sediment-associated contaminants include exposure to sediment pore water (interstitial water),
ingestion of sediment particles and dissolved organic matter, direct contact of sediment with body
surfaces, and exposure to the boundary layer of water overlying the sediment.
6
Utilization of
sediments as a food source by benthic organisms is common.
5
Particle ingestion via filtration of
the water column or direct consumption may be a significant route of exposure because benthic
organisms may preferentially ingest this fraction of the sediments.

7
Mode of feeding can significantly influence contaminant bioaccumulation from sediment.
6
For
example, deposit-feeding invertebrates may accumulate contaminants to a greater extent than filter
feeders. Digestion of sediment by an organism may be required for contaminant absorption because
of slow desorption of chemicals bound to sediment.
8
Sediment reprocessing by benthic invertebrates
may introduce nonequilibrium conditions in the sediment by removal of rapidly desorbable con
-
taminants.
9
Benthic organisms may also preferentially select higher organic carbon sediments,
influencing their exposure sediment-associated contaminants.
9
Types of organic carbon food sources
included in sediment can include encrusted mineral grains, organic–mineral aggregates, fecal
pellets, plant fragments, microalgae, and fungi.
5
A fraction of organic carbon is digestible, but some
organic carbon is present as refractory humic polymers that are not readily digestible.
5
Lee
9
described additional factors that may influence contaminant bioaccumulation. (1) Burrowing organ-
© 2003 by CRC Press LLC
isms (e.g., amphipods) may ventilate pore waters, whereas surface deposit feeders may not, and
(2) the formation of tubes and burrows may affect the spatial distribution of contaminants. Ulti
-

mately, contaminant bioaccumulation in benthic organisms will be influenced by the heterogenous
nature of sediment, which can vary in composition in scales of millimeters.
5
32.2.2 Bioaccumulation Factors
A term used to quantify bioaccumulation from sediments at equilibrium is the bioaccumulation
factor (BAF), defined as the chemical concentration in the organism (ng chemical/g tissue) divided
by the concentration in sediment (ng/g sediment). Either the wet or dry weight may be used; use
of dry weight may eliminate variability due to tissue or sediment hydration. The BAF is a unitless
number between 0 and infinity. A biota to sediment accumulation factor (BSAF) can also be
calculated from the lipid-normalized chemical concentration (ng chemical/g animal lipid) divided
by the organic-carbon-normalized sediment concentration (ng/g sediment organic carbon). The
BSAF provides a unitless number between 0 and infinity, with typical values reported for benthic
invertebrates ranging from 1 to 10.
4
BAFs and BSAFs are frequently used in ecological risk
assessments to estimate contaminant concentrations in benthic organisms.
10
BSAFs for polychlo-
rinated biphenyl (PCB) congeners determined in four species of benthic invertebrates collected
from Lake Erie, ranged from 0.1 to 10 and generally increased with increasing log K
ow
of the
congener.
3
BSAFs for polycyclic aromatic hydrocarbons (PAHs) ranged from 0.04 to 8 and generally
decreased with log K
ow
, suggesting biotransformation of the larger PAHs.
3
Mean BAFs for eight

metals in benthic invertebrates ranged from 0.1 to 3.5 and from 9 to 37 for PCBs.
10
32.2.3 Bioaccumulation of Metals
Concentrations of heavy metals in sediments can exceed those of the overlying water by three
to five orders of magnitude.
11
The bioavailability of sediment-associated heavy metals is related to
the presence of metal binding sites on the sediment. Increasing the concentration of iron oxides or
organic materials in the sediment appeared to reduce metal bioavailability by increasing the number
of metal binding sites.
11
Metals can also form insoluble sulfides, and the sulfide fraction of sediments
may represent a pool of adsorption sites for metal binding.
12
The toxicity of cadmium has been
shown to be related to the acid-volatile sulfide (AVS) content of sediments,
12,13
whereas the binding
of cadmium added to oxic sediments was explained by the humic acid content of the sediment.
14
Fu et al.
14
concluded that in oxic sediments, humic substances were the major absorbent for
cadmium and possibly other metals. Other factors that may influence metal bioaccumulation from
sediments include metal speciation, transformation (e.g., methylation to form hydrophobic alkyl
metals), inhibitory interactions of different metals, sediment chemistry (salinity, redox or pH), and
binding to dissolved organic matter (DOM).
11
DOM is an important energy source for microbially
based aquatic food webs, and the presence of DOM can affect both the distribution of a metal

between water and sediments and the bioavailability of pore-water contaminants.
9,15
Benthic organ-
isms may directly affect DOM concentrations through their feeding behaviors and excretory pro-
cesses.
9
32.2.4 Bioaccumulation of Organic Contaminants
32.2.4.1 Bioavailability
Binding of organic contaminants to sediments has been attributed to the organic carbon content,
clay type and content, cation-exchange capacity, pH, and particle surface area of the sediment.
6
The organic carbon content of sediment can be predictive of contaminant bioavailability because