O2

oxidative damage

Figure 2. Photochemical processes involved in photoactivated toxicity. Generation of excited states and pathways for their decay are illustrated in (A). Possible mechanisms underlying toxic activity are illustrated in (B).

excited-state orbitals. The excess energy is then dissipated in radiationless transitions from excited to ground rotational and vibrational electron levels, radiative transitions from excited-state orbitals to ground-state orbitals (fluorescence or phosphorescence, depending on whether triplet state intermediates occur via intersystem crossing), or direct energy transfer from the sensitizer to other molecules present in the biological matrix. Biologically damaging events occur during these sensitizer transitions from excited to ground state. Although not generally considered to be a major component of environmental aquatic phototoxicity, there exists the potential for solar radiation activated modification of non-toxic parent compounds to more toxic photodegradation products (e.g., quinones and other oxygenation products). In this chapter, unless otherwise noted, discussion will be limited to photosensitization reactions, rather than photomodification reactions.

The photosensitization process follows two possible pathways, conventionally referred to as Type I and Type II reactions following the terminology suggested by Gollnick [96] and refined by Foote [97,98], and illustrated in Figure 2B. The two pathways are differentiated by whether the excited-state sensitizer molecule transfers energy directly to molecular oxygen (Type II) or to another molecule within the biological matrix (Type I). As illustrated in Figure 2B, singlet oxygen is the primary damaging intermediate in Type II reactions, and may also contribute to damage resulting from Type I reactions as well. Type I reactions produce molecular radicals and reactive oxygen species (superoxide radicals, peroxides, hydroxyl radicals) formed during interaction of the excited-state sensitizer and other constituents of the biological matrix (referred to as solvent or substrate in non-biological photochemical systems). These are frequently competing reactions, the predominance of one or the other depending upon the specific sensitizer and its concentrations, the extent of oxygen saturation, the nature of the biological matrix, and the wavelengths of excitation radiation present in the system [74,98]. For example, given sufficient quantities of oxygen (which favors a Type II pathway), a matrix of greater lipid content relative to higher water content is likely to produce higher substrate oxidation rates because singlet oxygen lifetimes are significantly longer in lipid-rich reaction systems. Additionally, experimental evidence indicates that singlet oxygen quantum yields are greater for reactions occurring within membranes [99].

The probability that sensitizer excited states will be generated during UV irradiation is determined by the energy difference between electrons in the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) in the sensitizer molecule [95]. This energy difference, often termed the HOMO-LUMO gap, is illustrated in Figure 2A by the dark lines in the ground and excited-state diagrams. In PAHs, this gap is sufficiently narrow so that the relatively low energy present in UV photons is sufficient to promote electrons from occupied, bonding n orbitals to unoccupied, antibonding n orbitals (n to Ti* transitions), thus initiating the photoactivation process. This relatively narrow HOMO-LUMO gap is characteristic of aromatic systems owing to the extensive it conjugation extent in the benzene constituents. In many heterocyclic and substituted PAHs, electron transitions from non-bonding 2p orbital to a* or n* orbitals may also contribute to their potential phototoxicity. The HOMO-LUMO gap has been used in several QSAR (quantitative structure-activity relationship) analyses (discussed later) to predict whether specific PAHs are potentially photoactivated or non-photoactivated.

There are several aspects of photoactivated toxicity yet to be thoroughly elucidated. For most compounds, it has not been established clearly whether initial singlet excited states decay to excited triplet states prior to energy transfer to oxygen or substrate. This is not a trivial question, as the probability of singlet oxygen production increases by orders of magnitude when the sensitizer passes through a triplet excited state. This increased probability is due to the large difference in lifetimes for singlet (ns) and triplet (/is to ms) excited states. For most compounds, it also has yet to be established whether photoreactions occur primarily via Type I or Type II reactions.

Regardless of the exact excited-state processes involved, photoactivated damage to biological systems appears to proceed primarily via oxidative damage. Studies at the molecular level, as well as empirical whole-organism or whole-tissue studies, indicate that exclusion of oxygen from the experimental system or the presence of singlet oxygen quenchers (e.g., /^-carotene, a-tocopherol, etc.) greatly, or entirely, eliminates photoactivated damage. As well as singlet excited-state oxygen, other reactive species may also be produced via Type I pathways where sufficient oxygen is present. Choi and Oris [104] recently demonstrated very clearly that simultaneous exposure of fish liver microsomes to PAH and UVR resulted in oxidative stress, specifically lipid peroxidation resulting from formation of superoxide anion. Other highly toxic species include oxygen-free radical, hydroxyl radicals, and peroxides, all of which have been demonstrated to disrupt cellular membranes, amino acids, DNA, and other cellular and tissue components [101].

7.5 Predicting phototoxicity

As was stated earlier, PAHs are among the more problematic contaminants, relative to potential environmental phototoxicity, because they generally occur in contaminated aquatic systems as complex mixtures. Thousands of unsub-stituted and variously substituted PAHs have been identified as contributing to environmental contamination [22,102], As it is untenable to test each of these possible compounds to determine the extent of their photoactivated toxicity, a significant effort has been made by various researchers to develop predictive models to address this issue. An interesting aspect of phototoxic chemicals is the clear relationship between their phototoxic potency and their carcinogenic potential. Some of the earliest efforts to predict phototoxicity potential were based on these relationships, although much effort was directed at accomplishing the opposite, that is predicting carcinogenicity based on phototoxic potential. The elucidation of the carcinogenicity-phototoxicity relationship facilitated development of QSAR models. Based on this relationship, it was considered parsimonious (and ultimately correct) to use the molecular parameters proven to be significant in carcinogenicity QSARs when developing phototoxicity QSARs.

The earliest studies suggesting a carcinogenicity-phototoxicity relationship were conducted by Mottram and Doniach [47,48], Doniach [49], and Calcutt [50] and suggested that many chemicals capable of causing tumors were also phototoxic. It should be noted though that these authors all suggested that, for some of these compounds (e.g., the arsenicals), the toxicity in the presence of light was mechanistically different than for the classic photosensitizers such as ben-zo[a]pyrene. This is an important distinction for the development of QSARs, as phototoxicity resulting from photodynamic action is likely to have a very different molecular basis relative to other modes of action (e.g., the photo-dermatitis associated with porphyria mentioned earlier).

Epstein et al. [56] tested the carcinogenicity and phototoxicity of 157 compounds spanning a broad structural range and including substituted and unsub-stituted linear and cyclic compounds. The relationship between carcinogenicity and phototoxicity was apparent. Most notable was the fact that a larger portion of compounds that exhibited high phototoxic potency were carcinogens, compared to non-carcinogens. Rather than being diagnostic, these results suggested that potential for carcinogenicity and phototoxicity may have similar molecular foundations. Santamaria [103] also compared carcinogenicity and phototoxicity potentials of 36 PAHs in isolated mitochondria, and reported a significant correlation. Epstein [56] suggested, tested, and rejected the hypothesis that formation of charge-transfer complexes was a common molecular activity responsible for both carcinogenicity and for phototoxicity, a finding that caused other researchers to investigate alternative molecular explanations for both.

Morgan et al. [57] and Morgan and Warshawsky [102] linked several molecular parameters to carcinogenicity, and, based on the relationship between carcinogenicity and phototoxicity, suggested that these same parameters would be predictive of phototoxic potential. They determined energy levels for lowest singlet and triplets states, singlet-triplet splitting energy, and phosphorescence lifetimes for 18 carcinogens and 31 non-carcinogens. Of the parameters examined, excited singlet-state energy was highly significantly correlated, and singlet-triplet splitting energies significantly correlated with carcinogenicity, but triplet-state energy and phosphorescence lifetime were not. Compounds with singlet energies within the range of 297 to 310 kJ mol-1 were 22.8 times more likely to be carcinogenic. When compounds were plotted against singlet-state energy and singlet-triplet-state energy, carcinogens were clustered into a clearly defined ellipse, indicating that these two parameters could successfully discriminate between the carcinogenic and non-carcinogenic compounds examined.

Morgan and Warshawsky [57] related carcinogenicity, and indirectly the parameters used to predict it, to phototoxic potential by completing phototoxicity assays with brine shrimp (Artemia salina) nauplii. The greatest contribution of this work was the incorporation of quantum yields to accessing phototoxic potential, or potency. The authors accomplished this by testing each compound with essentially identical organisms and light levels. The data were then used to develop relative photodynamic activity (RPA) value, which could then be compared to carcinogenicity potential for these same compounds. RPA was cal culated by first characterizing the relationship between the average number of nauplii immobilized (ANI) and several exposure parameters:

where I0 = light intensity, / = nauplii path-length, a = proportion of PAH absorbed by nauplii, fa = quantum yield for immobilization, e = molar absorptivity of the PAH, C = exposure PAH concentration, t = duration of light exposure, and B = an integration constant. The RPA was then determined for each PAH tested:

RpA = Va0I = *fa where fa and (j>\ are quantum yields for the each PAH and the reference PAH (benz[c]acridine), respectively, and a and ar are molar absorptivities for each PAH and the reference PAH, respectively. A graphical representation of RPA estimation is shown in Figure 3. One limitation of this approach is that variability in the bioconcentration factors (BCF) for the examined compounds could significantly affect RPA, but was not accounted for. Thus Morgan and War-shawsky's [57] estimation of quantum efficiency was based on whole organism response, rather than a photochemical property specific to the compound itself. Regardless of this deficiency, their results were adequate to demonstrate a strong relationship between carcinogenic potential and phototoxicity, and to suggest

Figure 3. (A) Average number of nauplii immobilized (ANI) as a function of time. (B) Average number of nauplii immobilized as a function of (2.3034 e Ct). Benz[c]acridine, □—, 85.7 nM; benzo[a]pyrene, O—, 22 nM. [Data from [57], with permission of the publisher]

Figure 3. (A) Average number of nauplii immobilized (ANI) as a function of time. (B) Average number of nauplii immobilized as a function of (2.3034 e Ct). Benz[c]acridine, □—, 85.7 nM; benzo[a]pyrene, O—, 22 nM. [Data from [57], with permission of the publisher]

that the molecular parameters related to the former would also be related to the latter.

The predictive and QSAR models that were developed to predict phototoxicity have been based on the molecular characteristics that describe the probability that electrons will reach excited states when they interact with photons, and the pathway of decay to the ground state condition that is most likely to occur. As described previously, this latter question addresses the disposition of excess energy, and whether the decay processes are of sufficient duration to allow for energy, or electron, transfer among substrate or oxygen molecules. The characteristics examined in these modeling efforts include energy of lowest singlet and triplet states, HOMO-LUMO gap, energy of singlet-triplet interconversion, molecular connectivity, phosphorescence lifetime, and various parameters that describe molecular conformation and stability.

Newsted and Giesy [103] extended the work of Morgan and Warshawsky [102] by comparing the results of toxicity tests (Daphnia magna LT50) of 20 PAHs with several of their molecular parameters to determine which would best predict phototoxicity. These parameters included lowest-energy singlet and triplet states, singlet-triplet splitting energy, phosphorescence lifetime, and first-and second-order connectivity. These parameters, except for molecular connectivity, have been discussed in the section describing photosensitization mechanisms. Molecular connectivity describes molecular structure based on physical three-dimensional placement of skeletal atoms and their valence electrons, and molecular density, and has been correlated with bioconcentration potential and toxicity [106]. Following the assumption accepted by most phototoxicity researchers, that PAHs in organism tissues are responsible for toxicity (as opposed to those in the water column), Newsted and Giesy [103] attempted to conduct assays at equivalent molar tissue concentrations for each PAH. Where these attempts failed, they adjusted estimated lethal times based on tissue concentrations, resulting in estimates of potency they termed the median adjusted lethal time. Their modeling effort involved estimation of potency (<f) for each PAH, quantification of the photon energy reaching the tissue PAHs, the molar absorptivity of each PAH, and values for each of the molecular parameters. Ultimately, an RPA, analogous to that of Morgan and Warshawsky [102], was calculated for each PAH as follows: First, a potency value for each PAH was calculated:

where t = time and /a = the rate of UVR absorbtion (quanta per minute) expressing /a in quantifiable units. This equation was then rearranged:

d[%mortality]/di n d(%mortality)

d t n where, X is the integrated waveband (e.g., UV-A 315 to 336 nm), = the intensity of the waveband, T^ b, and Ca = optical transmittance, path length, and molar

PAH concentration in the organism, A = average quanta absorbed, n = number of X summed. Integration of this equation yields:

which expresses the linear relationship of % mortality and exposure duration, with B = y-intercept, and slope = A(f>t. Finally, RPA was calculated as:

where <p' and (j> are the potency for each PAH and for the reference PAH, benzo[i>] anthracene, respectively.

Using linear regression analysis of the relationship between RPA and the various molecular parameters, the authors found that phosphorescence lifetime explained the greatest proportion of the variation, and that all other parameters, when added to the model, increased the regression R2, but did not reduce residual variance. Curve-linear modeling produced a parabolic relationship with triplet-state energy providing the best fit (Figure 4). Finally, the authors successfully used principal-component analysis to cluster the 20 PAHs into three groups defined as very toxic, moderately toxic, and nontoxic. Discriminant analysis of these results indicated that phosphorescence lifetime and lowest triplet state energy were the parameters best able to reliably achieve these groupings.

Mekenyan et al. [107] and Newsted and Giesy [103] reanalyzed these data in an effort to develop a QSAR model that would predict phototoxicity using calculated ground-state molecular parameters, rather than excited-state parameters which are more difficult to compute. This approach was intended to predict toxicity where empirical data, e.g., measured triplet-state energy, were not available. The basis of Mekenyan et al.'s [107] modeling effort was the observation that the parabolic relationship revealed in Newsted and Giesy's [105] analysis might indicate multiple, competing processes, including internal factors (molecular parameters) and external factors (chemical and UV dosimetry). The molecular, internal parameters chosen for consideration included HOMO-LUMO gap and molecular stability.

Mekenyan et al.'s [108] results indicated that the HOMO-LUMO gap was a suitable ground state predictor of PAH phototoxicity and accurately placed the PAHs examined (the 20 studied by Newsted and Giesy [103]) into toxic or non-toxic groups (Figure 5). Adding to the significance of this work was the consistent relationship between the calculated HOMO-LUMO gap and the singlet and triplet state energies of the PAHs examined. This is significant, as the HOMO-LUMO gap is directly related to the probability that electrons will achieve the excited states necessary to initiate toxicity, but is not functionally related to the fate of that excess energy (a direct determinant of ultimate phototoxic potency), as were the parameters used by Newsted and Giesy [103], As pointed out by Mekenyan et al. [108], the consistent relationship between these two parameters enables their QSAR approach to be fully consistent with

1600

1400

1200

c 1000 2

E 800

O io

tZ3 600

400 200

100 120 140 160 180 200 220 240 260 280 Triplet Energy(ET) in kJ/Mole

1600 1500 1400 1300 1200 1100 c 1000 c 900

8 800

H 700

< 600 500 400 300 200 100

Figure 4. Top: Median lethal time (LT50) as a function of lowest triplet energy (ET). Bottom: Adjusted median lethal time (ALT50) as a function of lowest triplet energy. The polycyclic aromatic hydrocarbons are identified by number. [Data from [103], with permission of the publisher]

100 120 140 160 180 200 220 240 260 280 Triplet Energy(ET) in kJ/Mole

ALT50 = 1.53E -20(E T)? "+6.52E-12(E T J"470

ALT50 = 1.53E -20(E T)? "+6.52E-12(E T J"470

100 120 140 160 180 200 220 240 260 280 Triplet Energy(ET) in kJoules/Moles

100 120 140 160 180 200 220 240 260 280 Triplet Energy(ET) in kJoules/Moles

HOMO-LUMO Gap [eV]

Figure 5. Variation of toxicity (log(l/ALT)) with HOMO-LUMO gap. The solid triangles denote the predicted toxicity values for some as yet untested PAHs. [Data from [108], with permission of the publisher]

HOMO-LUMO Gap [eV]

Figure 5. Variation of toxicity (log(l/ALT)) with HOMO-LUMO gap. The solid triangles denote the predicted toxicity values for some as yet untested PAHs. [Data from [108], with permission of the publisher]

Newsted and Giesy [103], and yields a predictive tool that can be used in the absence of empirical data.

Veith et al. [109,110] extended the work of Mekenyan et al. [107,108] by demonstrating that HOMO-LUMO gap energies are also excellent predictors of phototoxicity of various a-terthienyls. Even though these toxicity data did not allow for potency estimates to be corrected for tissue concentrations (as were the data used in the previously described QSAR studies), the compounds studied were accurately predicted to be toxic or non-toxic. This is notable because, although of relatively high phototoxicity potency, these a-terthienyls are chemically significantly different (e.g., they contain linked cyclopentane with sulfur substitutions rather than fused cyclohexane) than the PAHs used in Mekenyan et al.'s [107,108] initial work. Veith et al. [109,110], in a separate analysis, calculated HOMO-LUMO values for pyrene and anthracene having methyl, tert- and n-butyl, ethylene, propylene, nitro, hydroxy, and chloro substitutions. The systematic selection of these substitutions allowed the authors to make broad conclusions regarding the nature and extent of the shift in HOMO-LUMO values associated with each. In general, alkyl and hydroxyl substitution did not significantly shift the HOMO-LUMO values, whereas nitro, alkene, and chloro substitutions did. The authors point out that other factors, such as bioaccumulation and environmental half-life, will also change with substitution, and will complicate predictions of phototoxicity for these compounds in natural waters. A specific test of these model results has yet to be reported.

7.6 Photomodified toxicity

The majority of phototoxicity research in aquatic systems has been undertaken with the assumption that the primary mechanism of concern in natural systems is photosensitized photodynamic toxicity, rather than photomodified toxicity [16]. However, the work of Huang et al. [23] and Ren et al. [Ill] suggests that photomodification of PAHs may increase their toxicity to some aquatic plants. The approach in these studies was to photomodify PAHs in solution by treating them with UVR, and then to test the toxicity of the photomodified solutions by adding them to plant growth media (Hutner's medium). Typically, PAH dissolved in 0.1% solutions of DMSO in water were irradiated at 25 to 40 ¿rniol m~2 UV-B for time periods ranging from 6 h to 96 h, and then diluted to concentrations ranging from 0 to 2 mg 1_1 in the growth medium. Based on a typical terrestrial solar spectrum, these UY-B exposure values correspond to an approximate range of 950 to 1500 piW cm-2; a range approximately 3 to 5 times greater than typical in terrestrial radiation. These estimates are from three separate sources: data from mid-day summer spectral scans made in coastal California [88], similar spectral data available for Daytona Beach (source: [106]), and solar spectra generated using the SBDART [113] model discussed in the UV dose section. Acute growth effects levels (50%) for photomodified PAHs ranged from 500 to 2000 pig l-1 for anthracene, benzo[a]pyrene, fluoranthene, naphthalene, phenanthrene, and pyrene. Thresholds for effect ranged from 50 to 500 pig l-1. Huang et al. produced similar results for five PAHs irradiated in natural radiation. In this study, IC50 estimates for PAHs irradiated for 7 or 20 days ranged from 0.2 to 2.8 mg 1_1, and were consistently lower for the longer irradiation durations. These natural solar radiation exposures demonstrate that reciprocity between exposure intensity (relatively high in Huang et al.'s [23] former work) and duration of irradiation (longer in this study) must be considered when evaluating effect levels.

The investigation of toxicity of photoproducts was extended by Marwood et al. [114] who exposed Lake Erie phytoplankton to anthracene and one of its primary photoproducts, 1,2-dihydroxyanthraquinone. When exposures were conducted in solar radiation, 200 pig anthracene l-1 caused a 50% inhibition in photosynthesis. At concentrations of 2000 pig 1_1, the photoproduct 1,2-dihyd-roxyanthraquinone reduced photosynthesis slightly when exposures were conducted in the dark, and by 50% when exposures where conducted in solar radiation. While this research was designed primarily to evaluate the usefulness of specific chlorophyll fluorescence parameters as indicators of toxic effects in phytoplankton, the results suggest that anthracene photoproducts are toxic to natural phytoplankton assemblages only at concentrations orders of magnitude greater than those occurring in surface waters. Although sensitized toxicity of anthracene was apparent in these tests, it too occurred at very high concentrations.

The potential for photoinduced toxicity (combined photosensitization and photomodification processes) in plants has been QSAR modeled by Krylov et al.

[115]. Their model, which is far too complex to present in this chapter, includes parameters for rates of photomodification of 16 PAHs, their relative photomodi-fied toxicity, rates of uptake into leaf tissues, solar flux, triplet-state formation of intact PAHs, rates of production of modified in-situ biomacromolecules via both sensitization reactions and direct interaction of PAH photoproducts. Unlike the QSAR models discussed previously [115], this model incorporated kinetic parameters for mechanism-specific rates of deactivation of a model photosynthetic molecule (G) via type I and II photoreactions of intact PAH, as well as by PAH photoproducts. This component of the model was developed to incorporate the growthendpoints reported by Huang et al. [23,29] and Ren et al. [111]. Krylov et al. [115] confirmed the consistency of this modeling approach by completing toxicity tests of 16 intact and photomodified PAHs on Lemna gibba. The results indicate that, for the relatively high-concentration exposures required for toxicity of PAHs in duckweed, toxicity is best predicted by an additive model that combines both photomodified and photosensitized mechanisms of action. Additional research into photomodified PAH toxicity in Lemna gibba, closely related to the studies described above, includes development of a QSAR model that incorporates shape parameters [116] and identification of specific anthracene photoproducts and their toxicity [117].

Additional evidence for photomodified toxicity in plants is provided by Wieg-man et al.'s azaarene phototoxicity work with the diatom species, Phaeodactylum tricornutum [25,93,118]. Wiegman et al. irradiated azaarenes with environmentally-realistic intensities of UVR. The EC50 values (for reduced photosynthesis) for quinoline, isoquinoline, acridine, and phenanthridine were reduced when exposure solutions were irradiated prior to the introduction of diatoms. The reduction of EC50 concentrations ranged from a factor of three to a factor of 300. Effect concentrations for these azaarenes ranged from approximately 230 fig 1 ~1 to 1 mg 1_1.

Although the concentrations of PAH in water or growth media were relatively high in most of these studies, irradiance doses were not. Even where irradiance levels were several times higher than natural radiation, the duration of irradiation was relatively short compared to natural settings where PAH contaminated sediments are exposed continuously to solar UVR during daylight hours. These studies were also completed using simple, single-chemical exposures, a condition rarely, if ever, encountered in contaminated systems. It is reasonable to assume that the complex mixtures of PAH (and other compounds) present in most contaminated sediments consist of some compounds that have the potential for photomodified toxicity. The presence of natural and anthropogenic organic material in contaminated sediments also constitutes a chromophore-rich environment where a variety of photosensitized reactions could produce toxic photomodified products at rates and concentrations similar to those used in the studies just discussed. These studies elucidate the hazard represented by photomodified toxicity, and although they do not clearly demonstrate risk, they do indicate that the potential for photomodification in contaminated locations warrants further research.

7.7 UVR exposure

The first law of photodynamics [94,95] (only absorbed wavelengths have the potential to activate photochemical processes) suggests that PAH phototoxicity will not occur in the absence of UVR, specifically wavelengths from 280 nm to about 400 nm. Of this wavelength range, the UV-A portion (315 to 400 nm) is of greatest concern because shorter wavelengths (UV-B; 280 to 315 nm), while biologically very harmful, make up only about 8% of the total UVR present, and are filtered from the water column much more effectively than longer UV wavelengths ([119] and see Chapter 3). In addition, PAHs generally absorb radiation more affectivity in the longer, UV-A wavelength range. Phototoxic potency is ultimately a function of dose - the intensity of UVR integrated over duration of exposure. In most systems, the law of reciprocity suggests that equal damage will be caused by equivalent photon doses, regardless of the rate at which they enter the system (within some reasonable bounds).

In natural aquatic systems, UVR is attenuated in the water column at a rate that generally corresponds to the concentration of DOC. Log-transformed, UV intensity values are linearly related to depth, and the slope of a best-fit line (examples are shown in Figure 6A) represents the rate at which radiation is attenuated [120-123]. Concentration and makeup of DOC can vary significantly among water bodies and over time, and the effect on subsurface UV-A can be dramatic [124-126]. For example, the absorption coefficient for a near-shore area in Lake Superior (near Duluth, MN, USA) was estimated to be —0.355 m-1, whereas in a St. Louis Harbor (Duluth, MN, USA) PAH-contaminated site, the absorption slope was estimated to be —0.001 m-1 (unpublished data, Figure 6A). The depth at which 50% of the above-surfaces UV-A intensity would

Depth (m) Wavelength <jnm)

Figure 6. Absorption plots for UV-A radiation (plot A) and spectra at 50% UV penetration depth (plot B) for PAH-contaminated site in St. Louis Bay, Duluth, MN and a near-shore location in Lake Superior. The 50% penetration depths shown are 10 cm and 80 cm for the St. Louis Harbor and Lake Superior sites, respectively.

Depth (m) Wavelength <jnm)

Figure 6. Absorption plots for UV-A radiation (plot A) and spectra at 50% UV penetration depth (plot B) for PAH-contaminated site in St. Louis Bay, Duluth, MN and a near-shore location in Lake Superior. The 50% penetration depths shown are 10 cm and 80 cm for the St. Louis Harbor and Lake Superior sites, respectively.

be absorbed was estimated to be 0.8 m for Lake Superior, and 0.1 m for the St. Louis Harbor location, an 8-fold difference. Given equivalent tissue PAH concentrations, and assuming that 50% of surface irradiance is sufficient to photo-activate PAH toxicity, organisms would be at risk in nearly 1 m of the Lake Superior water column, versus 10 cm in the St. Louis location water column.

These estimates of broad-spectrum UV penetration do not reflect potential differences in the spectrum of radiation that would reach PAH-exposed organisms. The differences in UV spectrum at the 50% UV penetration depth for the St. Louis Harbor and Lake Superior sites just discussed are shown in Figure 6B. These differences arise because DOC attenuates shorter UV wavelengths more efficiently than longer wavelengths and also because the makeup of DOC in different locations can vary dramatically, based on its sources in the landscape. The importance of such spectral variability has been demonstrated by Diamond et al. [90] in exposures of brine shrimp (Artemia salina) nauplii to three PAHs in combination with different UV spectra. In these assays, the overlap of UV-A radiation with absorbance spectra of pyrene, fluoranthene, and anthracene was manipulated using various filters. Where the radiation spectra overlapped significant potions of PAH absorbance spectra, toxicity did not differ. Where radiation and absorption spectra differed in the extent of their overlap, toxicity differed significantly. Most importantly, the variation in the spectra of UV-A used was consistent with variation possible in natural aquatic systems. The interaction of UV-A spectra and PAH absorbance spectra is summarized the equation for the phototoxicity weighting function (PWF):

where: PWF = photoactivated toxicity weighting function, e^ = wavelength-specific molar absorptivity, and Ix = wavelength-specific irradiance. This component of UV dose was incorporated into the Morgan and Worshawsky [102] and Newsted and Giesy [103] models discussed previously.

While this approach is a logical first approximation of actual PAH-photo-activation potential, its accuracy is limited somewhat by our knowledge of mechanisms of action involved. If, for example, mortality during PAH/UV-A exposure derives from accumulated external tissue damage, then this approach to dosimetry is likely to be acceptably accurate. However, if other mechanisms such as disruption of DNA or other macromolecule function are involved, then the spectrum of light reaching these target sites is likely to vary considerably depending on specific overlying tissue types, species, and lifestages. In this case, comparisons of toxicity between sites would be consistent if the same species was considered, but would be questionable among different species. As well as altering UV spectra, these biological components are also likely to alter the photochemistry of the toxic mechanisms occurring (e.g., by quenching PAH excited states or free radical processors).

To some extent, solar flux can be predicted in aquatic systems. However, exposure to UVR entering the water column will be greatly influenced by the life history and behavior of different species. Attenuation of UYR in the water column, as well as physical shading, creates a highly heterogeneous UY environment in which exposure is largely determined by the moment to moment location of potentially-exposed organisms. Species that reside in sediments, vegetation, other highly shaded microhabitats, or deep water during daylight will receive little UV exposure. For most motile organisms (e.g., larval fish and plankton), daily accumulated UV exposure will be a complex summation of high and low exposure periods. Except in cases where behavior is well understood and quantifiable, the risk of phototoxicity can be characterized best by setting bounding conditions for possible exposure, or by describing UVR dose in limited areas in the aquatic habitat rather than as specific estimates of expected affects.

Diamond et al. [127] have estimated UVR doses in wetlands using this approach. Typical UVR doses were estimated by first generating maximal solar radiation doses for each day using a radiative transfer model, SBDART [113]. The model produced values for the full spectrum of solar radiation, from 280 to 700 nm, for cloudless conditions. These maximal values were then modified based on cloud cover effect estimates from 30 yr of historical solar radiation data (National Renewable Energy Laboratory, Department of Energy; http://rredc.nrel.gov/solar/). The values derived in this procedure were estimated daily terrestrial, spectral (2 nm increments from 280 to 700 nm) solar radiation doses. Water column doses were then derived from absorption coefficients and spectral attenuation data described by Peterson et al. [128]. Although the focus of this effort was to characterize risk of UV-B radiation effects in amphibians, the procedure is directly applicable to phototoxicity, and the resulting UV-A radiation and spectral doses could be directly incorporated into calculation of possible effects.

Because of its importance to phototoxicity, the interaction of UVR with photosensitizer has received considerable focus, and has been quantitatively incorporated into all of the QSAR and PAH absorbance/UV spectra work described above, as well as discussions of the ecological risk of PAH phototoxicity [84,86,87,89,90,104,107-110,129]. In these QSAR studies, adherence to the law of reciprocity was assumed, rather than tested specifically. Ankley et al. [130] (see Figure 7) explicitly tested the consistency of the Law of Reciprocity at predicting PAH phototoxicity by conducting assays using Lumbriculus variegatus exposed to several combinations of fluoranthene concentration (0,3.7, 7.5, 15, 30, 60, 120 ng l"1) and UVR intensity (16.6, 33.5, and 75.2 /¿W cm-2). Making simplifying assumptions, that tissue concentrations are constant over the exposure duration (a 96 h uptake period preceded initiation of UV exposure) and that damage repair is negligible, Ankley et al. [130] predicted that toxicity would be described by the equation:

where / = lethality, /c3 = rate of damage accrual, DL = the critical level of damage, R0 = initial tissue residue, I = radiation intensity, iD = time to death.

Light Intensity Initial Residue (pW-/igfcm -g)

Figure 7. Time-dependent mortality of Lumbriculus variegates (expressed as LT50 values) versus the product of light intensity and initial tissue concentration of fluoranthene.

[Data from [130], with permission of the publisher]

Light Intensity Initial Residue (pW-/igfcm -g)

Figure 7. Time-dependent mortality of Lumbriculus variegates (expressed as LT50 values) versus the product of light intensity and initial tissue concentration of fluoranthene.

[Data from [130], with permission of the publisher]

The results of their exposures were consistent with this prediction, as shown in Figure 7, with slight deviation at the highest UV exposure levels. This deviation is expected, as these low-PAH-high-UV exposures would be more strongly affected by the delay in the onset of mortality commonly observed in phototoxicity assays. Ankley et al. [131] extended and confirmed these results in subsequent studies by repeating exposures with pyrene, anthracene, and fluorene. As in the previous study, these PAHs, except for fluorene, which was not phototoxic, adhered to predictions based on the Law of Reciprocity. These results were in accord with earlier work of Oris and Giesy [84] in which bluegill sunfish were exposed to various concentrations of anthracene and three levels of UV-B radiation. The fit of the mortality data to a model assuming reciprocity was less definitive than in the work of Ankley et al. [130,131], possibly because of differences between the two compounds, the relatively limited data set of Oris and Giesy [84], and differences among tested populations of sunfish. These relationships were also demonstrated by Erickson et al.'s [132] mixture work, in which binary combinations of anthracene, fluoranthene, and pyrene were tested for interactive phototoxicity. The toxicity of these compounds was found to be additive (as opposed to antagonistic or synergistic), but, more importantly for this discussion, all single and mixture exposure responses adhered to the law of reciprocity.

7.8 Risk assessment for PAH phototoxicity

Ecological toxicologists characterize the probability of harmful effects occurring in natural settings by conducting risk assessments. These assessments, briefly, incorporate potency, lethality, or "hazard" of specific contaminants, and probabilities of exposure of various species in natural systems. This already complex process is confounded for phototoxicity risk assessment by the need to quantify both PAH and UVR exposure. Many of the confounding factors have been alluded to throughout this chapter, and have been discussed briefly by Diamond and Mount [89], and by Ankley et al. [133]. The UVR component of phototoxicity risk assessment has been addressed previously in this chapter, and also by Diamond et al. [127]

If the well-supported assumption that PAH phototoxicity is primarily a photosensitization process, then the critical measure of PAH exposure is tissue concentration. Because PAHs are hydrophobic they tend to accumulate in tissues to concentrations 100 to 100000 times higher (depending on the specific PAHs) than their environmental water concentrations. These bioaccumulation factors (BAFs) are affected by lipid concentrations and metabolic processes in organisms, by fugacity process among sediment-bound PAHs, the water column, and organic suspended and dissolved material [134,135], The pathway for accumulation of most PAHs, and other lipophyllic, potentially photoactivated compounds, is from sediment (where they tend to accumulate because of high organic content) to water, and then to aquatic organisms. Additional uptake may occur via sediment ingestion, and via the food chain.

The difficulty of estimating PAH tissue concentrations is complicated by the fact that most PAH contamination occurs as mixtures of hundreds of PAH compounds. Each of these compounds has a unique Kow (organic-water partitioning coefficient), which is a reliable indicator of its tendency to remain in sediments or dissolved in water. Each compound also has a unique phototoxic potency, resistance to environmental degradation or modification, and metabolism by exposed organisms. At high concentrations, complex mixtures can also affect the solubility of their constituents, adding to the uncertainty of fugacity estimations (e.g., [136]). The phototoxicity potential of the complex mixtures typical of contaminated sites has been demonstrated in several ways, including field collection and subsequent UV treatment of PAH-exposed organisms, controlled, in situ UV exposure, and bioassays using field-collected sediments. However, the applicability of these results to broader risk assessment is limited by the unique mixture of PAHs present at these sites, and by the myriad differences in environmental factors, including penetration of UVR, temperature, carbon concentrations, etc.

The occurrence of phototoxicity in PAH-contaminated sites is nearly impossible to observe. Most highly contaminated sites are biologically depauperate, making direct observation of the toxic processes for species that would normally reside there untenable. Hence, although the potential for phototoxicity, its mechanisms and key components, has been thoroughly demonstrated, its importance in potentially affected aquatic systems has yet to be fully characterized or quantified. Thus, there is a definite need to continue studying the phenomenon due to the following factors:

(1) the large number of sites contaminated with high concentrations of PAHs,

(2) the slow degradation of most PAHs by natural processes,

(3) the continued release of PAHs via terrestrial runolf and aerial deposition, and

(4) the great potential for increased UVR exposure in aquatic systems due to environmental changes associated with global climate change.

References

1. M.L. Mitchell, D. Karentz (1993). The induction and repair of DNA photodamage in the environment. In: A.R. Young, L.O. Björn, J. Moan, W. Nultsch (Eds), Environmental U V Photobiology (pp. 345-377). Plenum Press, New York.

2. W.H. Jeifrey, D.L. Mitchell (1997). Mechanisms of UV-induced DNA damage and response in marine microorganisms. Photochem. Photobiol., 65, 260-263.

3. R.B. Setlow (1974). The wavelengths of sunlight effective in causing skin cancer: a theoretical analysis. Proc. Natl. Acad. Sei. U.S.A., 71, 3363-3366.

4. U. Strahle, S. Jesuthasan (1993). Ultraviolet irradiation impairs epiboly in zebrafish embryos: evidence for a microtubule-dependent mechanism of epiboly. Development, 119,909-919.

5. R.P. Sinha, H.D. Kumar, A. Kumar, D.-P. Häder (1995). Effects of UV-B irradiation on growth, survival, pigmentation and nitrogen metabolism enzymes in cyanobac-teria. Acta Protozool., 34,187-192.

6. P. Morliere, A. Moysan, I. Tirache (1995). Action spectrum for UV-induced lipid peroxidation in cultured human skin fibroblasts. Free Radical Biol. Med., 19, 365-371.

7. A.A. van Zeeland, A.T. Natarajan, E.A.M. Verdegaal-Immerzeel, A.R. Filon (1980). Photoreactivation of UV-induced cell killing, chromosome aberrations, sister chromatid exchanges, mutations and pyrimidine dimers in Xenopus laevis fibroblasts. Mol. Gen. Gen., 180,495-500.

8. W.E. Daniell, H.L. Stockbridge, R.F. Labbe, J.S. Woods, K.E. Anderson, D.M. Bissel, J.D. Bloomer, R.D. Ellefson, M.R. Moore, C.A. Pierach, W.E. Schreiber, A. Tefferi, G.M. Franklin (1997). Environmental chemical exposures and disturbances of heme synthesis. Environ. Health Perspect. Suppi, 105, 37-53.

9. D.-P. Häder, H.D. Kumar, R.C. Smith, R.C. Worrest (1998). Effects on aquatic ecosystems. J. Photochem. Photobiol. B: Biol., 46, 53-68.

10. J.J. Cullen, P.J. Neale (1997). Effect of UV on short-term photosynthesis of natural phytoplankton. Photochem. Photobiol, 65,264-266.

11. F.R. de Gruijl, J.C. Van der Leun (1994). Influence of ozone depletion on the incidence of skin cancer: quantitative prediction. In: A.R. Young, L.O. Björn, J. Moan, W. Nultsch (Eds), Environmental UV Photobiology (pp. 89-109). Plenum Press, New York.

12. S. Zigman (1993). Ocular damage by environmental radiant energy and its prevention. In: A.R. Young, L.O. Björn, J. Moan, W. Nultsch (Eds), Environmental UV Photobiology (pp. 149-181). Plenum Press, New York.

13. O. Holm-Hansen, D. Lubin, E.W. Helbling (1993). Ultraviolet radiation and its effect on organisms in aquatic environments. In: A.R. Young, L.O Björn, J. Moan, W. Nultsch (Eds), Environmental UV Photobiology (pp. 379-418). Plenum Press, New York.

14. D.M. Leech, C.E. Williamson (2001). In situ exposure to ultraviolet radiation alters the depth distribution of Daphnia. Limnol. Oceanogr., 46,416-420.

15. B.E. Brown, R.P. Dunne, T.P. Scoffin, M.D.A. Le Tissier (1994). Solar damage in intertidal corals. Mar. Ecol. Prog. Ser., 105, 219-230.

16. D.P. Arfsten, D.J. SchaefTer, D.C. Mulveny (1996). The effects of near ultraviolet radiation on the toxic effects of polycyclic aromatic hydrocarbons in animals and plants: a review. Ecotoxicol. Environ. Saf, 33, 1-24.

17. R.A. Larson, M.R. Berenbaum (1988). Environmental phototoxicity: solar ultraviolet radiation affects the toxicity of natural and man-made chemicals. Environ. Sci. Techno!., 22, 354-360.

18. L. Santamaria, G. Prino (1964). The photodynamic substances and their mechanisms of action. Res. Prog. Org.-Biol. Med. Chem., 1, 259-336.

19. M.R. Berenbaum (1987). Charge of the light brigade: phototoxicity as defense against insects. In: J.R. Heitz, K.R. Downum (Eds), Light-activated Pesticides (pp. 206-216). American Chemical Society, Washington, DC.

20. E.A.J. Bleeker, S. Wiegman, P. de Voogt, M. Kraak, H.A. Leslie, E. de Haas, W. Admiraal (2002). Toxicity of azaarenes. Rev. Environ. Contam. Toxicol, 173, 39-83.

21. B. David, P. Boule (1993). Phototransformation of hydrophobic pollutants in aqueous medium I - PAHs adsorbed on silica. Chemosphere, 9, 1716-1630.

22. R. Dabestani, I.N. Ivanov (1999). A compilation of physical, spectroscopic and photophysical properties of polycyclic aromatic hydrocarbons. Photochem. Photo-bioi, 70,10-34.

23. X.D. Huang, D.G. Dixon, B.M. Greenberg (1993). Impacts of ultraviolet radiation and photomodification of the toxicity of polycyclic aromatic hydrocarbons to the higher plant Lemna gibba L. (duckweed). Environ. Toxicol. Chem., 12, 1067-1077.

24. K. Nikolaou, P. Masclet, G. Mouvier (1984). Sources and chemical reactivity of polynuclear aromatic hydrocarbons in the atmosphere - a critical review. Sci. Total Environ., 32, 103-132.

25. S. Wiegman, P.L.A. van Ylaardingen, W.J.G.M. Peijnenburg, S.A.M. van Beusekom, M.H.S. Kraak, W. Admiraal (1999). Photokinetics of azaarenes and toxicity of phototransformation products to the marine diatom Phaeodactylum tricornutum. Environ. Sci. Technol., 33, pp. 4256-4262.

26. J.R. Heitz, K.R. Downum (Eds) (1987). Light-activated Pesticides. American Chemical Society, Washington, DC.

27. L. Santamaria, G. Prino (1964). The photodynamic substances and their mechanism of action. Res. Prog. Org.-Biol. Med. Chem., 1, 259-336.

28. B.K. Sinha, C.F. Chignell (1983). Binding of anthracene to cellular macromolecules in the presence of light. Photochem. Photobiol., 37, 33-37.

29. X.-D. Huang, G.D. Dixon, B.M. Greenberg (1995). Increased polycyclic aromatic hydrocarbon toxicity following their photomodification in natural sunlight: impacts on the duckweed Lemna gibba L. G-3. Ecotoxicol Environ. Saf., 32, 194-200.

30. L. Ren, L.F. Zeiler, D.G. Dixon, B.M. Greenberg (1996). Photoinduced effects of polycyclic aromatic hydrocarbons on Brassica napus (canola) during germination and early seedling development. Ecotoxicol Environ. Saf, 33, 73-80.

31. M.A. Pathak (1986). Phytophotodermatitis. Clin. Dermatol, 4, 102-121.

32. M.A. Pathak, T.B. Fitzpatrick (1992). The evolution of photochemotherapy with psoralens and UVA (PUVA): 2000 BC to 1992 AD. J. Photochem. Photobiol, 14 , 3-22.

33. I.R. Fahmy, H. Abu-Shady (1947). Ammi majus Linn: pharmacological study and isolation of a crystalline constituent, ammoidin. Q. J. Pharm. Pharmacol, 20,281-291.

34. H.N. Nigg, H.E. Nordby, R.C. Beier, A. Dillman, C. Macias, R.C. Hansen (1993).

Phototoxic coumarins in limes. Food Chem. Toxicol., 31, 331-335.

35. G. Kavli, G. Volden (1984). Phytophotodermatitis. Photodermatology, 1, 65-75.

36. B. Ljunggren (1990). Severe phototoxic burn following celery ingestion. Arch. Dermatol., 126,1334-1336.

37. M.A. Pathak, F. Daniels, T.B. Fitzpatrick (1962). The presently known distribution of furocoumarins (psoralens) in plants. J. Invest. Dermatol., 39, 225-239.

38. G.W. Ivie (1978). Toxicological significance of plant furocoumarins. In: R. Keeler, K. Van Kampen, L. James (Eds), Effects of Poisonous Plants on Livestock (pp. 475-485). Academic Press, New York.

39. J.R. Heitz (1987). Development of photoactivated compounds as pesticides. In: J.R. Heitz, K.R. Downum (Eds), Light-activated Pesticides (pp. 1-21). American Chemical Society, Washington, DC.

40. D.J. Birmingham, M.M Key, G.E. Tublich (1961). Phototoxic bullae among celery harvesters. Arch. Dermatol., 83, 73-85.

41. C.R. Lovell (1993). Plants and the Skin. Blackwell Scientific, Oxford.

42. L.D. Scheel, V.B. Perone, R.L. Larkin, R.E. Kupel (1963). The isolation and characterization of two phototoxic furanocoumarins (psoralens) from diseased celery. Biochemistry, 2,1127-1131.

43. S.K. Hann, Y.-K. Park, S. Im, S.W. Byun (1991). Angelica-induced Photodermatitis. Photodermatol. Photoimmunol. Photomed., 8, 84-85.

44. G.D. Pimprikar, M.J. Coign (1987). Multiple mechanisms of dye-induced toxicity in insects. In: J.R. Heitz, K.R. Downum (Eds), Light-activated pesticides (pp. 134-155). American Chemical Society, Washington, DC.

45. J. Kagan, W.J. Bennett, E.D. Kagan, J.L. Maas, S.A. Sweeney, I.A. Kagan, E. Seigneurie, V. Bindokas (1987). a-Terthienyl as a photoactive insecticide: toxic effects on nontarget organisms. In: J.R. Heitz, K.R. Downum (Eds), Light-activated pesticides (pp. 176-191). American Chemical Society, Washington, DC.

46. A. Jodlbauer, H. Tappeiner (1905). Die Beteiligung des Sauerstoffs bei der Wirkung ffuorezierender Stoffe. Dtsch. Arch. Klin. Med., 82, 520-546.

47. J.J. Mottram, I. Doniach (1937). Photodynamic action of carcinogenic agents. Nature, 140,933-934.

48. J.J. Mottram, I. Doniach (1938). The photodynamic action of carcinogenic agents. Lancet, 1, 1156-1159.

49. I. Doniach (1939). A comparison of the photodynamic activity of some carcinogenic and non-carcinogenic compounds. Br. J. Exp. Pathol., 20, 227-235.

50. G. Calcutt (1954). The photosensitizing action of chemical carcinogens. Br. J. Cancer, 8,177-180.

51. S.S. Epstein, M. Burroughs (1962). Some factors influencing the photodynamic response of Paramecium caudatum to 3,4-benzopyrene. Nature, 193, 337-338.

52. S.S. Epstein (1963). The photodynamic activity of polycyclic hydrocarbons carcinogens. Acta Unio Int. Cancrum, 19, 599-601.

53. S.S. Epstein, M. Burroughs, M. Small (1963). The photodynamic effect of the carcinogen 3,4,-benzopyrene on Paramecium caudatum. Cancer Res., 23, 35-44.

54. S.S. Epstein, M. Small, J. Koplan, N. Mantel, S.H. Hunter (1963). A photodynamic bioassay of benzo[a]pyrene with Paramecium caudatum. J. Natl. Cancer Inst., 31, 163-168.

55. S.S. Epstein, M. Small, J. Koplan, N. Mantel, H.L. Falk, E. Sawicki (1963). Photodynamic bioassay of polycyclic air pollutants. Arch. Environ. Health, 7, 531-537.

56. S.S. Epstein, M. Small, H.L. Faulk, N. Mantel (1964). On the association between photodynamic and carcinogenic activities in polycyclic compounds. Cancer Res., 24,

855-861.

57. D.D. Morgan, D. Warshawsky, T. Atkinson (1977). The relationship between carcinogenic activities of polycyclic aromatic hydrocarbons and their singlet, triplet, and singlet-triplet splitting energies and phosphorescence lifetimes. Photochem. Photo-bioi, 25,31-38.

58. J.W. Bowling, G.J. Leversee, P.F. Landrum, J.P. Giesy (1983). Acute mortality of anthracene-contaminated sunfish exposed to sunlight. Aquat. Toxicol., 3, 79-90.

59. J. Kagan, P.A. Kagan, H.E. Buhse, Jr. (1984). Light-dependent toxicity of a-terthienyl and anthracene toward late embryonic stages of Rana pipiens. J. Chem. Ecol., 10, 1115-1122.

60. J. Kagan, E.D. Kagan, E. Seigneurie (1986). Alpha-terthienyl, a powerful fish poison with light-dependent activity. Chemosphere, 15, 49-57.

61. J. Kagan, K. Tadema-Weilandt, G. Chan, S.D. Dhawan, I.P. Jawarsky, K.A. Suder-shan (1984). Oxygen requirement for near-UV mediated cytotoxicity of phenylheptat-riyne to Escherichia coli. Photochem. Photobiol., 39, 465-467.

62. R. Davenport, L.R. Johnson, D.J. Schaeffer, H. Balbach (1994). Light-enhanced toxicity of TNT and some related compounds to Daphnia magna and Lytechinus variagatus embryos. Ecotoxicol. Environ. Saf., 27, 14-22.

63. L.R. Johnson, R. Davenport, H. Balbach, D.J. Schaeffer (1994). Near-ultraviolet light-enhancement of Microtox assays of trinitrotoluene and aminodinitrotoluenes. Ecotoxicol. Environ. Saf., 27, 23-33.

64. L.R. Johnson, R. Davenport, H. Balbach, D.J. Schaeffer (1994). Comparative toxicity of trinitrotoluene and aminodinitrotoluenes to Daphnia magna, Dugesia dorotocephala, and sheep erythrocytes. Ecotoxicol. Environ. Saf, 27, 34-49.

65. A. Zaga, E.A. Little, C.F. Rabeni, M.R. Ellersieck (1998). Photoenhanced toxicity of a carbamate insecticide to early life stage anuran amphibians. Environ. Toxicol. Chem., 17, 2543-2553.

66. A. S. Wernersson, S. Dave (1997). Phototoxicity identification by solid phase extraction and photoinduced toxicty to Daphnia magna. Arch. Environ. Contamin. Toxicol., 32,268-273.

67. L. Cleveland, E.E. Little, R.D. Calfee, M.G. Barron (2000). Photoenhanced toxicity of weathered oil to Mysidopsis bahia. Environ. Toxicol. Chem., 49, 63-76.

68. E.E. Little, L. Cleveland, R. Calfee, M.G. Barron (2000). Assessment of the photoenhanced toxicity of weathered oil to the tidewater silverside. Environ. Toxicol. Chem., 19,926-932.

69. M.C. Pelletier, R.M. Burgess, K.T. Ho, A. Kuhn, R.A. McKinney, S.A. Ryba (1997). Phototoxicity of individual polycyclic aromatic hydrocarbons and petroleum to marine invertebrate larvae and juveniles. Environ. Toxicol. Chem., 16, 2190-2199.

70. R. Davenport, A. Spacie (1991). Acute phototoxicity of harbor and tributary sediments from lower Lake Michigan. J. Great Lakes Res., 17, 51-56.

71. G.T. Ankley, S.A. Collyard, P.D. Monson, P.A. Kosian (1994). Influence of ultraviolet light on the toxicity of sediments contaminated with polycyclic aromatic hydrocarbons. Environ. Toxicol. Chem., 13, 1791-1796.

72. P.D. Monson, G.T. Ankley, P.A. Kosian (1995). Phototoxic response of Lumbriculus variegatus to sediment contaminated by polycyclic aromatic hydrocarbons, short communication. Environ. Toxicol. Chem., 14, 891-894.

73. D.S. Ireland, G.A. Burton, Jr., G.G. Hess (1996). In situ toxicity evaluations of turbidity and photoinduction of polycyclic aromatic hydrocarbons. Environ. Toxicol. Chem., 15, 574-581.

74. T. Ito (1978). Oxygen requirement for near-UY mediated cytotoxicity of phenyl-

heptatriyne to Escherichia coli. Photochem. Photobiol., 28, 493-508.

75. J.T. McCloskey, J.T. Oris (1991). Effect of water temperature and dissolved oxygen concentration on the photo-induced toxicity of anthracene to juvenile bluegill sunfish (Lepomis macrochirus). Aquat. Toxicol., 21, 145-156.

76. J. Kagan, A. Stokes, H.-H. Gong, R.W. Tuveson (1987). Light-dependent cytotoxicity of fluoranthene: oxygen-dependent membrane damage. Chemosphere, 16,2417-2422.

77. D.R. Kearns, A.U. Khan (1969). Sensitized photooxygenation reactions and the role of singlet oxygen. Photochem. Photobiol., 10, 193-210.

78. B.W. Gensemer, D.G. Dixon, B.W. Greenberg (1998). Amelioration of the photo-induced toxicity of polycyclic aromatic hydrocarbons by a commercial humic acid. Ecotoxicol. Environ. Saf., 39, 57-64.

79. A. Nikkilae, S. Penttinen, J.V. Kukkonen (1999). UV-B-induced acute toxicity of pyrene to the waterflea Daphnia magna in natural freshwaters. Ecotoxicol. Environ. Saf., 44, 271-279.

80. W. Bennett, J.L. Maas, S.A. Sweeney, J Kagan (1986). Phototoxicity in aquatic organisms: the protecting effect of beta-carotene. Chemosphere, 6, 781-786.

81. A.T. Hall, J.T. Oris (1991). Anthracene reduces reproductive potential and is maternally transferred during long-term exposure in fathead minnows. Aquat. Toxicol., 19,249-264.

82. J.E. Weinstein, K.D. Polk. Phototoxicity of anthracene and pyrene to glochidia of the freshwater mussel, Utterbackia imbecillis. Environ. Toxicol. Chem., in press.

83. J.E. Weinstein (2001). Characterization of the acute toxicity if photoactivated fluoranthene to glochidia of the freshwater mussel, Utterbackia imbecillis. Environ. Toxicol. Chem., 20,412-419.

84. J.T. Oris, J.P. Giesy (1985). The photoenhanced toxicity of anthracene to juvenile sunfish (Lepomis spp.). Aquat. Toxicol., 6,133-146.

85. J.T. Oris, J.P. Giesy, Jr. (1986). Photoinduced toxicity of anthracene to juvenile bluegill sunfish (Lepomis machrochirus Rafinesque): photoperiod effects and predictive hazard evaluation. Environ. Toxicol. Chem., 5, 761-768.

86. J.T. Oris, J.P. Giesy, P.M. Allred, D.F. Grant, P.F. Landrum (1984). Photoinduced toxicity of anthracene in aquatic organisms: an environmental perspective. In: T.N. Vezirogiu (Ed.), The Biosphere: Problems and Solution (pp. 639-658). Elsevier Science, Amsterdam.

87. P.F. Landrum, J.P. Giesy, J.T. Oris, P.M. Allred (1985). The photoinduced toxicity of polycyclic aromatic hydrocarbons to aquatic organisms. In: J.H. Vandermeulen, S. Hrudey (Eds), Oil in Freshwater: Chemistry, Biology, and Technology (pp. 304-318). Pergamon, New York.

88. M.G. Barron, E.E. Little, R. Calfee, S. Diamond (2000). Quantifying solar spectral irradiance in aquatic habitats for the assessment of photoenhanced toxicity. Environ. Toxicol. Chem., 19, 920-925.

89. S.A. Diamond, D.R. Mount (1998). Evaluating the role of photo-activated PAH toxicity in risk assessment. SETAC News, Learned Discourses, 18, 17-18.

90. S.A. Diamond, D.R. Mount, L.P. Burkhard, G.T. Ankley, E.A. Makynen, E.N. Leonard (2000). Effect of irradiance spectra on the photoinduced toxicity of three polycyclic aromatic hydrocarbons. Environ. Toxicol. Chem., 19,1389-1396.

91. L.L. Hoist, J.P. Giesy (1989). Chronic effects of photoenhanced toxicity of anthracene on Daphnia magna reproduction. Environ. Toxicol. Chem., 8, 933-942.

92. S.A. Diamond, J.T. Oris, S.I. Guttman (1995). Adaptation to fluoranthene exposure in a laboratory population of fathead minnows. Environ. Toxicol. Chem., 14,1393-1400.

93. S. Wiegman (2002). Photoenhanced Toxicity of Azaarenes to Marine Phytoplankton.

Ph.D. Thesis (pp. 73-90). University of Amsterdam, The Netherlands.

94. A. Gilbert, J. Baggott (1991). Essentials of Molecular Photochemistry. Blackwell Science, Paris.

95. C.H.J. Wells (1972). Introduction to Molecular Photochemistry. Chapman and Hall, London (pp. 1-13).

96. K. Gollnick (1968). Type II photooxygenation reactions in solution. Adv. Photo-chem., 6, 1-122.

97. C.S. Foote (1991). Definition of Type I and Type II photosensitized oxidation. Photochem. Photobiol., 54, 659.

98. C.S. Foote (1987). Type I and Type II mechanisms of photodynamic action. In: J.R. Heitz, K.R. Downum (Eds), Light-activated Pesticides (pp. 22-38). American Chemical Society, Washington, DC.

99. D.P. Valenzeno (1987). Photomodification and singlet oxygen generation in membranes. In: J.R. Heitz, K.R. Downum (Eds), Light-activated Pesticides (pp. 39-57). American Chemical Society, Washington, DC.

100. M.L. Lee, D.L. Vassilaros, D.W. Later (1982). Capillary column gas chromatography of environmental polycyclic aromatic compounds. Int. J. Environ. Anal. Chem., 11,251-262.

101. L. Santamaria, G.G. Giordano (1969). Effects of long-wave ultraviolet radiation on polycyclic hydrocarbon carcinogenesis. In: M.D. Urbach (Ed.), The Biological Effects of Ultraviolet Radiation (with Emphasis on the Skin) (pp. 569-580). Pergamon Press, New York.

102. D.D. Morgan, D. Warshawsky (1977). The photodynamic immobilization of Ar-temia salina nauplii by polycyclic aromatic hydrocarbons and its relationship to carcinogenic activity. Photochem. Photobiol, 25, 39-46.

103. J.L. Newsted, J.P. Giesy (1987). Predictive models for photoinduced acute toxicity of polycyclic aromatic hydrocarbons to Daphnia magna, Strauss (Cladocera, Crustacea). Environ. Toxicol Chem., 6, 445-461.

104. J. Choi, J.T. Oris (2000). Evidence of oxidative stress in bluegill sunfish (Lepomis macrochirus) liver microsomes simultaneously exposed to solar ultraviolet radiation and anthracene. Environ. Toxicol Chem., 19, pp. 1795-1799.

105. D.W. Nebert, D.D. Petersen, A.J. Fornace (1990). Cellular responses to oxidative stress: The [AH] gene battery as a paradigm. Environ. Health Perspect., 88, 99, 13-25.

106. R. Koch (1983). Molecular connectivity index for assessing ecotoxicological behavior of organic compounds. Toxicol Appl. Environ. Chem., 6, 87-96.

107. O.G. Mekenyan, G.T. Ankley, G.D. Veith, D.J. Call (1994). QSARs for photoinduced toxicity: I. Acute lethality of polycyclic aromatic hydrocarbons to Daphnia magn., Chemosphere, 28, 567-582.

108. O.G. Mekenyan, G.T. Ankley, G.D. Veith, D.J. Call (1994). QSAR estimates of excited states and photoinduced acute toxicity of polycyclic aromatic hydrocarbons. SAR QSAR Environ. Res., 2, 237-247.

109. G.D. Veith, O.G. Mekenyan, G.T. Ankley, D.J. Call (1995). QSAR evaluation of a-terth

0 0

Post a comment