Download Can highly hydrophobic organic substances cause aquatic baseline

Hearing Order OH-001-2014
Tsawout IR #2, Document X
Environmental Toxicology and Chemistry, Vol. 25, No. 10, pp. 2639–2644, 2006
䉷 2006 SETAC
Printed in the USA
0730-7268/06 $12.00 ⫹ .00
CAN HIGHLY HYDROPHOBIC ORGANIC SUBSTANCES CAUSE AQUATIC BASELINE
TOXICITY AND CAN THEY CONTRIBUTE TO MIXTURE TOXICITY?
PHILIPP MAYER* and FREDRIK REICHENBERG
Department of Environmental Chemistry and Microbiology, National Environmental Research Institute, P.O. Box 358,
4000 Roskilde, Denmark
( Received 24 March 2006; Accepted 3 May 2006)
Abstract—Effect concentrations for aquatic baseline toxicity generally decrease with increasing log octanol–water partition coefficient (Kow) values of up to 5 to 6, whereas less is known about the baseline toxicity of organic chemicals with log Kow values
above 6. A physicochemical analysis of the dissolution process for organic chemicals was combined with reported baseline toxicity
data, leading to the following conclusions. First, no absolute hydrophobicity cutoff exists for baseline toxicity at a log Kow value
of 6, because aquatic baseline toxicity for fish and algae was observed for chemicals with log Kow values greater than 6.5 and with
effect concentrations less than 10 ␮g/L. Second, the baseline toxicity of hydrophobic organic substances was exerted at a relatively
constant chemical activity of 0.01 to 0.1. Finally, organic chemicals with high melting points cannot provide sufficient chemical
activity to exert baseline toxicity when considered as individual, pure chemicals. However, such substances are still expected to
contribute to baseline toxicity when part of a complex mixture.
Keywords—Hydrophobic organic chemicals
titative structure–activity relationships
Effective chemical activity
Baseline toxicity
Quan-
pure water is proportional to its concentration (Cwater), and it
expresses the potential for diffusive uptake and partitioning
into, for instance, lipid membranes. Chemical activity can be
determined using either ␥water or SL [5,6]:
INTRODUCTION
The hydrophobicity of organic chemicals is a key factor
for their aquatic baseline toxicity, because it can drive their
partitioning from water and into lipid membrane and storage
lipids of aquatic organisms. Many toxicity studies suggest that
up to a certain cutoff level, aquatic baseline toxicity increases
with increasing hydrophobicity, whereas highly hydrophobic
chemicals with log octanol–water partition coefficient (Kow)
value above, for instance, 6.0 do not exert aquatic toxicity
(see, e.g., [1]). The present paper challenges this view, and it
emphasizes the melting point as a key parameter for singlesubstance toxicity, with major implications for the baseline
toxicity of complex mixtures.
a ⫽ ␥water · Vwater · Cwater 艑
Cwater
SL
(2)
Escaping from water
The second implication of the hydrophobic property is the
tendency to escape from the dissolved aqueous form, and this
makes hydrophobic organic substances susceptible to partitioning from water and into any other phase with a higher
solubility than that in water. Octanol is such a solvent that
displays a generally high solubility for a wide range of hydrophobic organic substances [2,7,8]. This is the basis for the
use of Kow as a surrogate measure for hydrophobicity and for
empirical relations, such as Kow 艑 1,795/SL [8].
Dissolution and aqueous solubility
The transfer of a solid substance into the dissolved aqueous
form requires two energy-dependent physicochemical transformations. First, the compound must be liberated from the
crystal lattice (i.e., melt) and then be dissolved from the liquid
state into the aqueous form [2,3]. Consequently, the water
solubility (S) of a solid substance depends on both its melting
enthalpy and its hydrophobicity. The melting enthalpy also is
the main factor determining the melting point (Tm; in Kelvin)
of a substance, and this is the basis for the empirical relation
by Yalkowsky et al. [4]:
S ⫽ exp[(6.8 ⫾ 1.0)(1 ⫺ Tm/T)]SL
Toxicity cutoff
Partitioning into lipid membranes and baseline toxicity
Lipids also display a high solubility for many hydrophobic
organic chemicals that consequently are enriched into, for instance, lipid membranes, where they can cause nonspecific
effects on membranes. This phenomenon is called baseline
toxicity or (nonpolar) narcosis, and it occurs at a critical membrane concentration of approximately 40 to 160 mmol/kg lipid,
which for many aquatic species is equivalent to 2 to 8 mmol/
kg wet weight [9,10]. Baseline toxicity is controlled mainly
by the membrane–water partitioning coefficient of a chemical
[9–11], which increases with increasing hydrophobicity [12].
This is why median effective concentration (EC50) values for
baseline toxicity generally decrease with increasing log Kow
values up to 5 to 6. Substances with log Kow values exceeding
6.0 are, to the contrary, often reported not to cause aquatic
toxicity even at their solubility limit, and this gives rise to the
(1)
where SL is the subcooled liquid solubility and T is the ambient
temperature.
The SL of a substance is a ‘‘pure measure’’ for its hydrophobicity and is directly related to its aqueous activity coefficient (␥water; SL 艑 1/(␥water·Vwater), where Vwater is the molar
volume of water). The chemical activity (a) of a substance in
* To whom correspondence may be addressed ([email protected]).
2639
2640
Environ. Toxicol. Chem. 25, 2006
P. Mayer and F. Reichenberg
important question of whether highly hydrophobic organic
substances with log Kow values above 6.0 can cause or contribute to aquatic baseline toxicity.
The subsequent analysis of this problem is based on the
classical fish toxicity study by Könemann [1] and our two
working hypotheses: First, baseline toxicity requires a chemical activity of approximately 0.01 to 0.1 [6,13], and solid
substances with a high melting point are not able to provide
such high chemical activities [4,6]. Second, the duration of
many aquatic toxicity tests often is insufficient to reach critical
internal exposure concentrations of highly hydrophobic organic chemicals in aquatic macro-organisms.
SELECTION, TREATMENT, AND PRESENTATION OF
TOXICITY DATA
The primary toxicity data set was taken from Könemann
[1]. It includes median lethal concentration (LC50) values for
50 industrial chemicals derived from 14-d toxicity tests with
guppies. These data are plotted against log Kow in Figure 1A,
which also includes linear regressions for LC50 values and
aqueous solubility. Figure 1B presents the same toxicity data
together with LC50 values for naphthalene, phenanthrene, pyrene, and the two polychlorinated biphenyls (PCBs), congeners
31 and 105, for zebra fish larvae from Petersen [14]. The latter
data were obtained in toxicity tests with a sufficiently long
test duration to achieve equilibrium between the exposure medium and the organism and with measured lethal body residues
that indicate baseline toxicity. A regression for the SL from
Mackay et al. [8] also is included in Figure 1B. The LC50
values from Figure 1B were divided with the respective SL
values to derive median lethal chemical activities (LA50 ⫽
LC50/SL), which are plotted in Figure 1C.
The small size and the large surface to volume ratio of
unicellular algae make them very suited for the toxicity testing
of hydrophobic chemicals under equilibrium partitioning conditions. Therefore, EC50 values of five chlorobenzenes and
three PCBs from algal growth inhibition tests were selected
from three studies [15–17]. All these EC50 values are based
on the toxicity endpoint ‘‘inhibition of the exponential growth
rate’’ and on analytically confirmed exposure concentrations.
These toxicity data are plotted together with the respective
solubility and SL data in Figure 2.
DISCUSSION
Könemann [1] derived an empirical toxicity cutoff in his
classical fish study of 50 industrial pollutants, and he based it
on two observations: Absence of toxicity for two tetrachlorinated biphenyls in a 14-d toxicity test with guppies, and the
intersection of the solubility regression and the LC50 regression at a log Kow value of between 5 and 6 (Fig. 1A). Other
studies also suggest that very hydrophobic substances are unable to exert aquatic toxicity, and explanations include limited
solubility in the membrane [18], insufficient aqueous solubility
[19], insufficient time to reach steady state [20,21], and limited
ability to enter the membrane at a critical molecular size [22].
The subsequent discussion is directed at identifying the causes
for the apparent toxicity cutoff in the study by Könemann [1]
and at questioning the sense of defining a toxicity cutoff based
on Kow.
Kinetic aspects
A chemical needs to be taken up before it reaches a critical
internal concentration that can lead to toxicity, and this sup-
Fig. 1. Median lethal concentration (log LC50) as a function of log
octanol–water partition coefficient Kow. The LC50 values for guppy
were taken from Könemann [1]. (A) Regressions for LC50 and aqueous solubility from the same study. (B) Regression for the subcooled
liquid solubility (SL) from Mackay et al. [8], and LC50 values for
zebra fish larvae from Petersen [14]. (C) The same toxicity data expressed as median lethal chemical activities (LA50 ⫽ LC50/SL).
Aquatic baseline toxicity of hydrophobic organic substances
Environ. Toxicol. Chem. 25, 2006
2641
posedly explains the absence of toxicity for the two tetrachlorinated biphenyls. Establishing steady-state concentrations
of tetrachlorinated biphenyls in guppies requires more than 40 d
[3]; thus, internal steady-state concentrations were not reached
within the 14-d toxicity test of Könemann [1]. The time to
steady state decreases with decreasing size of the organism
[23,24], and toxicity will thus be expressed in smaller organisms within a shorter test duration. Indeed, Petersen [14] observed lethal effects of PCB 31 (trichlorinated) and PCB 105
(pentachlorinated) in 6-d toxicity tests with zebra fish larvae,
and Mayer et al. [15] reported growth inhibition of PCB 31,
PCB 48 (tetrachlorinated), and PCB 105 in 2-d tests with unicellular microalgae. The exposure times of the fish larvae test
and the algal tests were sufficiently long to reach (near) steady
state [14,25], and lethal body burdens and toxic cell concentrations were determined to be near baseline toxicity levels
[14,15]. Consequently, the absence of toxicity in guppies and
the presence of toxicity in fish larvae and algae can be explained by differences in uptake kinetics and test duration.
Thus, toxicity often will not be expressed in large organisms
that are exposed to highly hydrophobic substances, because
this combination leads to long steady-state times. The toxicity
data for fish larvae and algae were obtained at equilibrium,
and they were thus included in Figures 1B and C and 2A and
B.
Hydrophobicity and melting point
Fig. 2. (A) Median effect concentrations (log EC50) for algal growth
inhibition of three polychlorinated biphenyls (PCBs) and chlorobenzenes together with the subcooled liquid solubility (log SL) plotted
against log octanol–water partition coefficient (Kow). (B) Solubility
(log S) and log subcooled liquid solubility (SL) plotted against log
Kow. The vertical lines represent the reduction in solubility from the
melting process. (C) Toxicity data of A expressed as median effective
chemical activities (EA50 ⫽ EC50/SL). The toxicity data were taken
from Mayer et al. [15] for the three PCB congeners and from Wong
et al. [16], Calamari et al. [17], and Mayer et al. [27] for the chlorobenzenes. log S for chlorobenzenes were taken from Schwarzenbach
et al. [37] and Mackay et al. [38] and were calculated for the PCBs
by the SRC-WSKOW software (Syracuse Research, Syracuse, NY,
USA). log Kow values were taken from Bruijn et al. [39], and log SL
were calculated using Equation 1. HCB ⫽ hexachlorobenzene.
The other observation that led to the solubility cutoff hypothesis of Könemann [1] is the intersecting regressions for
LC50 and aqueous solubility at a log Kow value of between 5
and 6. Indeed, Figure 1A suggests that aqueous solubilities for
substances with a log Kow value of greater than 6 generally
are less than the predicted LC50 values, and this would then
point to a solubility cutoff. However, Kow is not a sufficient
descriptor for the aqueous solubility (S) of a solid substance,
which not only depends on its hydrophobicity but also on its
melting enthalpy [2,4].
The solubility regression should be refined by separately
accounting for the effect of hydrophobicity (i.e., the SL) and
for the effect of the melting cost (i.e., the ratio between S and
SL) (Eqn. 1). The effect of hydrophobicity is highlighted in
Figure 1B by a regression for the SL from Mackay et al. [8],
and this regression remains approximately one order of magnitude greater than all LC50 values from Könemann [1] and
Petersen [14]. This is in good agreement with the study by
Bobra et al. [13], who suggested the acute toxicity of hydrocarbons and chlorinated hydrocarbons to occur at a relatively
constant fraction of SL, which is equivalent to toxicity at a
relatively constant chemical activity [6]. No hydrophobicity
cutoff is apparent in Figures 1B and 2A, and indeed, toxicity
is observed to be exerted at a relatively constant chemical
activity, as can be seen in Figures 1C and 2C.
Presence of a melting point cutoff
The intersecting regressions for LC50 and aqueous solubility of Figure 1A can, instead, be explained by the impact
of the melting cost on the aqueous solubility in combination
with a correlation between the log Kow and the melting temperature for the chemicals that were included in the study by
Könemann [1]. Such a melting point cutoff has been suggested
by Lipnick [26] to explain observations from the classic Overton [18] study: Even though the two isomers anthracene and
phenanthrene have similar partitioning properties, phenan-
2642
Environ. Toxicol. Chem. 25, 2006
Fig. 3. The chemical activity of a substance at solubility as a function
of its melting point temperature (Eqn. 1) and the reported chemical
activity range for baseline toxicity.
threne (Tm ⫽ 99.5⬚C) produces narcosis, whereas anthracene
(Tm ⫽ 217.5⬚C) does not.
The melting point cutoff also can explain why the highly
hydrophobic PCBs are toxic to algae [15], whereas the similarly hydrophobic hexachlorobenzene is not [27] (Fig. 2). The
main difference appears to be the melting temperature, which
is between 67 and 105⬚C for the three PCB congeners and
230⬚C for hexachlorobenzene. Figure 2A shows EC50 values
of some chlorinated benzenes and PCBs together with their SL
values. The EC50 regression is 1 to 1.5 log units below the
regression for SL, suggesting that substances with a solubility
(S) of more than 1.5 log units below SL are not able to exert
baseline toxicity as individual substances. The solubility S of
hexachlorobenzene is more than two orders of magnitude below SL (Fig. 2B), meaning that even the maximum chemical
activity of the pure substance is less than that required for
baseline toxicity (Fig. 2C). Indeed, no algal toxicity of hexachlorobenzene was observed. Finally, it should be noted that
the melting cost not only reduces the solubility of a solid in
water but also reduces its solubility in lipid membranes. Thus,
the melting point cutoff also can be looked at as a lipid solubility cutoff.
Figure 3 shows the decrease of the S to SL ratio as a function
of the melting temperature of a chemical (Eqn. 1). A melting
point cutoff in toxicity applies to those high-melting-point
chemicals for which the S to SL ratio remains under the level
required for baseline toxicity. The existence of such a melting
point cutoff and of a kinetic cutoff for time needed to reach
near steady-state conditions has important implications for the
toxicity testing of individual high-melting-point chemicals, for
mixture toxicity, and for the eventual effect assessment of
high-melting-point chemicals.
Aquatic baseline toxicity of individual chemicals
The maximum chemical activity of many hydrophobic
chemicals in aquatic toxicity tests often will be just below, at,
or just above the baseline toxicity limit. Such tests may then
give two very different results. A substance will either appear
to be very toxic, with EC50 values in the ␮g/L range, or they
will appear to be nontoxic even at saturation. The chemical
activity framework and the melting point cutoff might help us
to understand this apparent discontinuity in toxicity that can
P. Mayer and F. Reichenberg
be observed between very similar substances, such as anthracene and phenanthrene. The thermodynamic stability of the
crystalline state of the pure substance defines the maximum
chemical activity (Eqn. 1), which then can be below, at, or
above the level required for baseline toxicity. The discontinuity
in such cases relates to the thermodynamic stability of the
crystal structure that defines a maximum chemical activity
rather than to properties that determine membrane permeability
or toxic mode of action. Highly symmetric molecules form
more stable crystal structures compared to less symmetric but
otherwise similar molecules. The maximum chemical activity
thus is much lower for the highly symmetric molecule anthracene than for the less symmetric molecule phenanthrene,
and it also is lower for 1,4-dichlorobenzene than for 1,2-dichlorobenzene.
Baseline toxicity has been suggested to occur within a relatively narrow range of chemical activities, between 0.01 to
0.1 [6,13], and this is in good agreement with the lethal chemical activities shown in Figure 1C (LA50 ⫽ LC50/SL) and the
median effective chemical activities (EA50 ⫽ EC50/SL) shown
in Figure 2C. Such chemical activities of organic substances
can be calculated from concentration measurements, or they
can be measured directly with equilibrium sampling devices
[28]. The effective chemical activity framework might challenge our perception of the toxicity of highly hydrophobic
organic chemicals. They are not particularly toxic in terms of
effective chemical activity (Figs. 1C and 2C), even though
they might appear to be highly toxic when expressed as effective aqueous concentrations. This is in good agreement with
the critical body residue approach that suggests baseline toxicity to occur at a relatively constant body burden [9].
Regulatory guidelines and standard methods for aquatic
toxicity tests were developed for essentially pure substances
that are soluble and stable in water, whereas they are less suited
for difficult to test substances, such as hydrophobic organics
[29]. It is particularly difficult to establish aqueous solutions
of hydrophobic organic substances and to maintain constant
and freely dissolved concentrations throughout the test [30].
Another difficulty arises from the fact that toxic responses
often are related to nominal concentrations or to total measured
concentrations, both of which can be substantially higher than
actual dissolved concentrations [31,32]. Most of these experimental errors are easily made when testing highly hydrophobic organic chemicals and lead to a reduced exposure. Thus,
it can be rather difficult to generate and test at the maximum
chemical activity, which might lead to false-negative results
for toxicity.
Effect assessments typically are based on the testing of
aquatic organisms in short-term toxicity tests [33]. The duration of the acute fish test especially is often insufficient to
establish internal steady-state concentrations or equilibrium
concentrations of highly hydrophobic organics because of their
slow uptake kinetics. A thorough assessment of the baseline
toxic potential of highly hydrophobic organics thus requires a
sufficiently long test duration and, preferably, also confirmation of stable internal concentrations. In the case of slow uptake
kinetics, metabolization within the test organism can reduce
the internal exposure and, thus, the toxicity—at least with
regards to the parent compound. Therefore, the persistence of
a hydrophobic organic chemical within the test organism at
the time scale of its bioconcentration kinetics will be a prerequisite for its baseline toxicity. The absence of aquatic toxicity for many hydrocarbons might result from metabolism of
Aquatic baseline toxicity of hydrophobic organic substances
the compound so that they do not reach a critical concentration
in the target membranes.
Baseline toxicity of hydrophobic organic chemicals in a
mixture
The described melting point cutoff in baseline toxicity does
not apply to solid substances that are part of a mixture, because
baseline toxicity is concentration additive [34] and substances
that do not exert toxicity as individual compounds can still
contribute to the toxicity of a mixture [9,35,36]. Hexachlorobenzene can, for example, be expected to contribute to the
baseline toxicity of chlorobenzene mixtures. Another example
would be anthracene and phenanthrene, which at less than
aqueous solubility can be expected to have very similar contributions to baseline toxicity.
A strategy to avoid the difficulties related to the aquatic
effect assessment of hydrophobic organics could be the introduction of a chemical activity threshold level within the environmental exposure assessment. The contribution of a chemical to the baseline toxicity of a mixture can be kept low when
keeping its chemical activity less than a certain threshold level.
Such a threshold level could, for instance, be set to 0.0001, if
the contribution of a single substance to baseline toxicity
should be kept to less than 1%. This would protect the aquatic
environment against baseline toxicity. Toxicity testing efforts
could then be directed at testing for those effects that are
exerted at chemical activities well below 0.01, which indicates
a more specific kind of toxicity.
CONCLUSIONS
No general hydrophobicity cutoff exists for aquatic baseline
toxicity in the region of log Kow values between 6 and 7. Some
highly hydrophobic substances can exert baseline toxicity at
very low aqueous concentrations but at relatively high chemical activities. Many highly hydrophobic substances can be
expected to contribute to the baseline toxicity of complex mixtures.
Slow uptake kinetics explains the absence of toxicity for
two tetrachlorinated biphenyls in the study by Könemann [1],
which suggests that hydrophobic substances require longer
exposure times to express their toxicity. Furthermore, a melting
point cutoff explains the intersecting regressions of solubility
and LC50 in the study by Könemann [1], and it also explains
the reported nontoxicity of hexachlorobenzene and anthracene.
The melting point cutoff suggests that high-melting-point substances can hardly exert baseline toxicity as single substances,
simply because they cannot reach a sufficiently high chemical
activity. These substances, however, are still expected to contribute to the toxicity of a complex mixture. The absence of
toxicity of many hydrophobic substances should be generalized with caution, because some highly hydrophobic substances can be toxic in the nmol/L range [14,15] and organisms in
the environment are exposed to mixtures rather than to single
substances.
Acknowledgement—The authors thank Joop Hermens, Lars Toräng,
and Henrik Tyle for helpful comments on the manuscript. This research has been funded by the Danish Research Academy, the Danish
Technical Research Council, and the European Commission within
the FP 6 Integrated Project ‘‘Novel Methods for Integrated Risk Assessment of Cumulative Stressors in the Environment’’ (GOCE003956).
Environ. Toxicol. Chem. 25, 2006
2643
REFERENCES
1. Könemann H. 1981. Quantitative structure–activity relationships
in fish toxicity studies. Part 1: Relationship for 50 industrial pollutants. Toxicology 19:209–221.
2. Schwarzenbach RP, Gschwend PM, Imboden DM. 2003. Environmental Organic Chemistry. John Wiley, New York, NY, USA.
3. Sijm DTHM. 1995. Solubility of difficult organic chemicals. Proceedings, Performing Aquatic Toxicity Tests with Poorly Soluble
Substances, RIVM, Bilthoven, The Netherlands, March 17, pp
21–32.
4. Yalkowsky SH, Orr RJ, Valvani SC. 1979. Solubility and partitioning. 3. The solubility of halobenzenes in water. Ind Eng Chem
Fundam 18:351–353.
5. Di Toro DM, McGrath JA, Hansen DJ. 2000. Technical basis for
narcotic chemicals and polycyclic aromatic hydrocarbon criteria.
I. Water and tissue. Environ Toxicol Chem 19:1951–1970.
6. Reichenberg F, Mayer P. 2006. Two complementary sides of bioavailability: Accessibility and chemical activity of organic contaminants. Environ Toxicol Chem 25:1239–1245.
7. Banerjee S. 1984. Solubility of organic mixtures in water. Environ
Sci Technol 18:587–591.
8. Mackay D, Bobra A, Shiu WY, Yalkowsky SH. 1980. Relationships between aqueous solubility and octanol–water partition coefficients. Chemosphere 9:701–711.
9. McCarty LS, Mackay D. 1993. Enhancing ecotoxicological modeling and assessment. Environ Sci Technol 27:1719–1727.
10. van Wezel AP, Opperhuizen A. 1995. Narcosis due to environmental pollutants in aquatic organisms: Residue-based toxicity,
mechanism, and membrane burdens. Crit Rev Toxicol 25:255–
279.
11. Vaes WHJ, Ramos EU, Verhaar HJM, Hermens JLM. 1998. Acute
toxicity of nonpolar versus polar narcosis: Is there a difference?
Environ Toxicol Chem 17:1380–1384.
12. Gobas FAPC, Lahittete JM, Garofalo G, Shiu WY. 1988. A novel
method for measuring membrane–water partition coefficients of
hydrophobic organic chemicals: Comparison with 1-octanol–water partitioning. J Pharm Sci 77:265–272.
13. Bobra AM, Shiu WY, Mackay D. 1983. A predictive correlation
for the acute toxicity of hydrocarbons and chlorinated hydrocarbons to the water flea (Daphnia magna). Chemosphere 12:1121–
1129.
14. Petersen G. 1997. Comparative studies of toxic effects and bioaccumulation of lipophilic substances in fish early life stages.
PhD thesis. University of Copenhagen, Copenhagen, Denmark.
15. Mayer P, Halling-Sorensen B, Sijm DTHM, Nyholm N. 1998.
Toxic cell concentrations of three polychlorinated biphenyl congeners in the green alga Selenastrum capricornutum. Environ
Toxicol Chem 17:1848–1851.
16. Wong PTS, Chau YK, Rhamey JS, Docker M. 1984. Relationship
between water solubility of chlorobenzenes and their effects on
a freshwater green alga. Chemosphere 9:991–996.
17. Calamari D, Galassi S, Setti F, Vighi M. 1983. Toxicity of selected
chlorobenzenes to aquatic organisms. Chemosphere 12:253–262.
18. Overton E. 1901. Studien über die Narkose zugleich ein Beitrag
zur Allgemeinen Pharmakologie. Gustav Fisher, Jena, Germany.
19. Lipnick RL. 1989. A quantitative structure–activity relationship
study of Overton’s data on the narcosis and toxicity of organic
compounds to the tadpole, Rana temporaria. In Suter GW II,
Lewis MA, eds, Aquatic Toxicology and Environmental Fate.
American Society for Testing and Materials, Philadelphia, PA, pp
468–489.
20. Hawker DW, Connell DW. 1985. Relationships between partition
coefficient, uptake rate constant, clearance rate constant, and time
to equilibrium for bioaccumulation. Chemosphere 14:1205–1219.
21. Veith GD, Call DJ, Brooke LT. 1983. Structure–toxicity relationships for the fathead minnow, Pimephales promelas: Narcotic
industrial chemicals. Can J Fish Aquat Sci 40:743–748.
22. Opperhuizen A, van de Velde EW, Gobas FAPC, Liem DAK, van
de Steen JMD. 1985. Relationship between bioconcentration in
fish and steric factors of hydrophobic chemicals. Chemosphere
14:1871–1886.
23. Gobas FAPC, Mackay D. 1987. Dynamics of hydrophobic organic
chemical bioconcentration in fish. Environ Toxicol Chem 6:495–
504.
24. Sijm DTHM, Linde AVD. 1995. Size-dependent bioconcentration
kinetics of hydrophobic organic chemicals in fish based on dif-
2644
25.
26.
27.
28.
29.
30.
31.
Environ. Toxicol. Chem. 25, 2006
fusive mass transfer and allometric relationships. Environ Sci
Technol 29:2769–2777.
Halling-Soerensen B, Nyholm N, Kusk O, Jacobsson E. 2000.
Influence of nitrogen status on the bioconcentration of hydrophobic organic compounds to Selenastrum capricornutum. Ecotoxicol Environ Saf 45:33–42.
Lipnick RL. 1990. Narcosis: Fundamental and baseline toxicity
mechanism for nonelectrolyte organic chemicals. In Karcher W,
Devillers J, eds, Practical Applications of Quantitative Structure–Activity Relationships (QSAR) in Environmental Chemistry
and Toxicology. Kluwer, Dordrecht, The Netherlands, pp 129–
144.
Mayer P, Wernsing J, Tolls J, de Maagd PGJ, Sijm DTHM. 1999.
Establishing and controlling dissolved concentrations of hydrophobic organics by partitioning from a solid phase. Environ Sci
Technol 33:2284–2290.
Mayer P, Tolls J, Hermens JLM, Mackay D. 2003. Equilibrium
sampling devices. Environ Sci Technol 37:184A–191A.
Organization for Economic Co-operation and Development. 2000.
Guidance document on aquatic toxicity testing of difficult substances and mixtures. Guideline 23. Paris, France.
European Centre for Ecotoxicology and Toxicology of Chemicals.
1996. Aquatic Toxicity Testing of Sparingly Soluble, Volatile and
Unstable Substances. Monograph 26. Brussels, Belgium.
Urrestarazu-Ramos E, Meijer SN, Vaes WHJ, Verhaar HJM, Hermens JLM. 1998. Using solid phase microextraction (SPME) to
determine partition coefficients to humic acids and bioavailable
concentrations of hydrophobic chemicals. Environ Sci Technol
32:3430–3435.
P. Mayer and F. Reichenberg
32. Landrum PF, Nihart SR, Eadle BJ, Gardner WS. 1984. Reversephase separation method for determining pollutant binding to
Aldrich humic acid and dissolved organic carbon of natural waters. Environ Sci Technol 18:187–192.
33. Robertson S. 1995. The use of aquatic toxicity test data in framework of substance classification and risk assessment. Proceedings, Performing Aquatic Toxicity Tests with Poorly Soluble Substances, RIVM, Bilthoven, The Netherlands, March 17, pp 17–
20.
34. Deneer JW, Sinnige TL, Seinen W, Hermens JLM. 1988. The
joint toxicity to Daphnia magna of industrial organic chemicals
at low concentrations. Aquat Toxicol 12:33–38.
35. Hermens JLM, Leeuwangh P. 1982. Joint toxicity of mixtures of
8 and 24 chemicals to the guppy (Poecilia reticulata). Ecotoxicol
Environ Saf 6:302–310.
36. Landrum PF, Lotufo GR, Gossiaux DC, Gedeon ML, Lee JH.
2003. Bioaccumulation and critical body residue of PAHs in the
amphipod, Diporeia spp.: Additional evidence to support toxicity
additivity for PAH mixtures. Chemosphere 51:481–489.
37. Schwarzenbach RP, Gschwend PM, Imboden DM. 1993. Environmental Organic Chemistry. John Wiley, New York, NY, USA.
38. Mackay D, Ying Shiu W, Ching Ma K. 1992. Illustrated Handbook of Physical–Chemical Properties and Environmental Fate
for Organic Chemicals. Lewis, Boca Raton, FL, USA.
39. Bruijn JD, Busser F, Seinen W, Hermens J. 1989. Determination
of octanol/water partition coefficients for hydrophobic organic
chemicals with the ‘‘slow-stirring’’ method. Environ Toxicol
Chem 8:499–512.