Direct and Indirect Atmospheric Deposition of PCBs to the Delaware

To develop the total maximum daily load (TMDL) for polychlorinated biphenyls (PCBs) for the tidal Delaware River (water-quality Zones 2−5), estimate...
17 downloads 8 Views 251KB Size
Download PDF
Environ. Sci. Technol. 2006, 40, 2171-2176

Direct and Indirect Atmospheric Deposition of PCBs to the Delaware River Watershed L I S A A . T O T T E N , * ,† M A Y A P A N A N G A D A N , † S T E V E N J . E I S E N R E I C H , †,‡ GREGORY J. CAVALLO,§ AND THOMAS J. FIKSLIN§ Department of Environmental Sciences, Rutgers University, 14 College Farm, Road, New Brunswick, New Jersey 08901 and Delaware River Basin Commission, 25 State Police Drive, West Trenton, New Jersey 08628

Atmospheric deposition can be an important source of PCBs to aquatic ecosystems. To develop the total maximum daily load (TMDL) for polychlorinated biphenyls (PCBs) for the tidal Delaware River (water-quality Zones 2-5), estimates of the loading of PCBs to the river from atmospheric deposition were generated from seven air-monitoring sites along the river. This paper presents the atmospheric PCB data from these sites, estimates direct atmospheric deposition fluxes, and assesses the importance of atmospheric deposition relative to other sources of PCBs to the river. Also, the relationship between indirect atmospheric deposition and PCB loads from minor tributaries to the Delaware River is discussed. Data from these sites revealed high atmospheric PCB concentrations in the Philadelphia/Camden urban area and lower regional background concentrations in the more remote areas. Wet, dry particle, and gaseous absorption deposition are estimated to contribute about 0.6, 1.8, and 6.5 kg year-1 ΣPCBs to the River, respectively, exceeding the TMDL of 0.139 kg year-1 by more than an order of magnitude. Penta-PCB watershed fluxes were obtained by dividing the tributary loads by the watershed area. The lowest of these watershed fluxes are less than ∼1 ng m-2 day-1 for penta-PCB and probably indicates pristine watersheds in which PCB loads are dominated by atmospheric deposition. In these watersheds, the passthrough efficiency of PCBs is estimated to be on the order of 1%.

Introduction The Delaware River Basin Commission (DRBC) recently developed a total maximum daily load (TMDL) of 380 mg day-1 for the sum of all polychlorinated biphenyls (PCBs) in Zones 2-5 of the Delaware River (Figure 1) (1). This TMDL, which was formally established by the USEPA on December 15, 2003, was a phased effort with additional modeling and refinement of the TMDL planned. Development of the waterquality model and the TMDL required estimates of all PCB loadings to the river, including atmospheric deposition. The * Corresponding author phone: (732) 932-9588; fax: (732) 9328644; e-mail: [email protected]. † Rutgers University. ‡ Current Affiliation: European Chemicals Bureau, European Commission, Joint Research Centre, TP 290, I-21020 Ispra (VA), Italy. § Delaware River Basin Commission. 10.1021/es052149m CCC: $33.50 Published on Web 03/04/2006

 2006 American Chemical Society

initial water-quality model was developed for pentachlorinated PCBs only. The calculated TMDL for penta-PCB was converted to ΣPCBs by multiplying by a conversion factor of 4, which represents the average ratio between ΣPCBs and penta-PCB in ambient water samples (1). To evaluate atmospheric deposition of PCBs to the river, the DRBC commissioned Rutgers University to operate four new atmospheric deposition monitoring sites in New Jersey, Pennsylvania, and Delaware. These sites represent an extension of the New Jersey Atmospheric Deposition Network (NJADN), which had previously consisted of nine sites in New Jersey where atmospheric concentrations of PCBs, polycyclic aromatic hydrocarbons (PAHs), organochlorine pesticides, nutrients, and trace metals were measured in atmospheric samples (2). PCBs and organochlorine pesticides were measured in the gas and particle phases at the new sites. The data were used to generate atmospheric input functions for the Delaware River. PCBs are a problem in the Delaware River because their ambient concentrations exceeded the DRBC and adjoining state water-quality criteria for PCBs of approximately 44 pg L-1 for the protection of human health from carcinogenic effects. Typical PCB concentrations in the river range from 2 to 10 ng L-1 (1). Fish from the Delaware River have been shown to contain elevated levels of PCBs, resulting in fish consumption advisories for the region. The relative magnitude of the sources of PCBs to the Delaware River was not obvious at the beginning of the TMDL process. The loading estimates constructed for the TMDL model (1) indicate that nonpoint sources, mostly urban runoff, constitute the largest single input of PCBs to the system, but the estimate of the nonpoint source load is highly uncertain. Contaminated sites appear to constitute a significant loading category, but again, these estimates are associated with a great deal of uncertainty owing to the difficulties of identifying the contaminated sites (some 30 have been identified) and in estimating the export of PCBs from these sites to the River. The absence of a single clearly identifiable major source of PCBs sets the Delaware apart from most rivers where PCB contamination has been extensively studied and modeled, for example, the Lower Fox River (Wisconsin) (3, 4) and the Hudson River (New York) (5). The types of PCB sources important in the Delaware River are probably typical of most urbanized rivers in the United States. The new monitoring sites were established at Lum’s Pond on the Chesapeake and Delaware Canal, at Swarthmore College south of Philadelphia, at the Northeast Philadelphia Airport, and at Alloways Creek in New Jersey (Figure 1). In addition, previous NJADN sites on the Delaware River included the Delaware Bay site (near Cape May Courthouse), Washington Crossing, and Camden, bringing the total number of sites used to estimate atmospheric deposition to the river to seven. The concentrations of PCBs in the gas, aerosol, and precipitation phases from these seven sites were used to estimate atmospheric deposition fluxes (Fatm) to the river. The purpose of this paper is to present the atmospheric PCB data from these seven sites, use that data to estimate atmospheric deposition fluxes to the River, and assess the importance of atmospheric deposition relative to other sources of PCBs to the river. Additional data collected by the DRBC on tributary loads to the Delaware River allow an investigation of the relationship between direct and indirect atmospheric deposition, which is defined at deposition to land surfaces which can be remobilized to enter water bodies. Thus, a second purpose of this investigation was to examine VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2171

FIGURE 1. Map showing the water-quality zones, monitoring sites, and subwatersheds of the Delaware River. the yield of PCBs from watersheds feeding into the Delaware River and attempt to determine the fraction of tributary inputs that are attributable to indirect atmospheric deposition and the efficiency with which watersheds trap atmospherically deposited PCBs.

Materials and Methods Site Characterization. The Delaware River flows south from its headwaters in New York State to form the natural boundary between New Jersey and Pennsylvania/Delaware. The tidal portion of the river extends ∼133 mi from Trenton, NJ, south to the mouth of Delaware Bay (Figure 1). Seven air-monitoring locations were operated within the watershed of the Delaware River: Camden, Swarthmore, and Northeast Philadelphia Airport (hereafter referred to as “Northeast”), Alloways Creek, Lum’s Pond, Washington Crossing, and Delaware Bay. The Camden, Delaware Bay, and Washington Crossing sites have been described previously (6). The Northeast, Swarthmore, and Lum’s Pond locations were operated from December 2001 to December 2002. These locations were chosen because they were already in use as air-quality-monitoring sites for various regional authorities. Northeast Philadelphia Airport is located within the city limits of Philadelphia about 10 mi to the northeast of the city center. The sampler was located on top of an air sampling trailer, which is located away from most airport operations but close to one rarely used runway. The Swarthmore site was on the roof of Hicks Hall, a building on the campus of Swarthmore College. This building was built ca. 1888 but extensively renovated in 1974. The roof of Hicks Hall was replaced in 1991. The Lum’s Pond site was located inside Lum’s Pond State Park near the Chesapeake and Delaware Canal. Here the sampler was again placed on top of an air-monitoring trailer. At Alloways Creek the sampler was located in a fenced enclosure in an undeveloped area. 2172

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006

Sample Collection and Analysis. Details of sample collection, preparation, extraction, and analysis can be found elsewhere (7, 8). Air samples (24 h) were typically collected at 12-day frequencies using a modified high-volume air sampler (Tisch Environmental, Village of Cleves, OH) with a calibrated airflow of ∼0.5 m3 min-1. Quartz fiber filters (QFFs; Whatman) were used to capture the particulate phase, and polyurethane foam plugs (PUFs) were used to capture the gaseous phase. The resulting samples were extracted using a Soxhlet apparatus, cleaned up using 3% deactivated alumina, and analyzed for PCBs by electron capture detection (ECD). Sixty peaks representing 93 PCB congeners were quantified and summed to yield “ΣPCBs”. Quality Assurance. Field blanks and matrix spikes were used for quality-control purposes. Field blanks were not subtracted from samples but were used to calculate the method detection limit (MDL) for each congener by taking the mean of the mass detected in all field blanks plus three times the standard deviation about the mean. No significant differences were observed between masses of PCBs measured in field blanks collected at the different sampling sites. Thus, one detection limit was calculated which applies to all sites. ΣPCB method detection limits were 19 and 18 ng for QFF and PUF samples, respectively. These can be converted to concentration units by dividing by the exact sample volume (typically ∼600 m3). The relative percent difference between two side-by-side samples for ΣPCBs was 2% for QFFs and 9% for PUFs. Surrogate recoveries were used to judge the recovery of all PCBs in the sample. Samples with recoveries below 50% were discarded from the data set. For the samples that remained, surrogate recoveries averaged more than 78% for PCBs 23, 65, and 166 in all matrixes and were used to correct individual compound concentrations for surrogate recoveries. PCB concentrations measured in the particle phase were

TABLE 1: Parameters for the Seven Monitoring Sites on the Delaware River (arranged from north to south) site a b R2

Wash. Crossing

Northeast

-3393 15 0.36

-7705 31 0.77

Camden

Swarthmore

Lum’s Pond

penta-PCB temperature dependence -6922 -11745 -3556 30 47 16 0.66 0.95 0.42

Alloways Creek

Delaware Bay

-3392 15 0.17

-3778 17 0.45

average concentrations (pg m-3) penta-PCB gas particle ΣPCBs gas particle

32 1.4

140 5.1

820 36

1400 12

31 1.4

35 1.0

33 2.8

180 7.1

700 19

3300 176

3300 32

160 6.1

150 4.6

160 14

fluxes (ng m-2 day-1)1 penta-PCB gas dry wet ΣPCBs gas dry wet

1.1 0.61 0.15

6.1 2.2 0.56

41 15 3.9

80 5.0 1.3

1.5 0.62 0.16

1.5 0.44 0.11

1.4 1.2 0.30

15 3.1 0.78

49 8.0 2.0

270 76 19

230 14 3.5

14 2.6 0.67

12 2.0 0.50

18 5.9 1.5

also corrected for laboratory blank masses as described in VanRy et al. (9). Gas-phase PCB concentrations were not corrected for laboratory blanks because masses in the blanks were in all cases less than 5% of the mass in the samples. Several PUFs were cut in half before deployment in the field in order to quantify gas-phase breakthrough. The bottom half PUF contained on average (n ) 3) 13% of the total mass of PCBs and on average less than 10% of each individual congener (n ) 3), except for the tri-chloro PCBs, for which a maximum of 31% was found in the bottom half PUF.

Results Ambient Air Concentrations. Gas-phase PCB concentrations were highest at Camden and Swarthmore, where ΣPCBs averaged about 3300 pg m-3 (Table 1). Despite the similar concentrations measured at both sites, the congener patterns were very different, as described below. ΣPCBs at Northeast averaged 700 pg m-3. At the other sites (Washington Crossing, Lum’s Pond, Alloways Creek, and Delaware Bay) gas-phase ΣPCB concentrations were essentially the same (averaging 150-220 pg m-3), consistent with the background PCB concentration for the region (6). The temperature dependence of atmospheric PCB concentrations is well documented (10-16) and expressed via the Clausius-Clapeyron equation

ln C )

a +b T

(1)

where a and b are the constants used in the TMDL model, C is the concentration of the individual congener in pg m-3, and T is the average temperature during the sampling period in Kelvin. Because the initial water-quality model was developed for penta-PCBs, only the results of the ClausiusClapeyron regressions for the penta homologue are presented in Table 1. Despite a high level of uncertainty in these slopes, it is immediately obvious that both the slope and the intercept at Swarthmore are significantly higher than at the other sites. Average gas-phase congener patterns at each site were regressed against each other to determine whether the sources of PCBs were similar at all sites. The results of these regressions revealed that most sites were similar (R2 > 0.6) but that the congener pattern at Swarthmore was significantly different from the other sites, with the R2 less than 0.35 for all sites except Camden (R2 ) 0.61). The difference is due to a shift toward heavier congeners at Swarthmore. The average

molecular weight (MW) of the PCB profile at Swarthmore is 315 g mol-1, while at the other NJADN sites the average ((stdev) MW is 290 ((4) g mol-1. The average MW at Camden is 298 g mol-1. The high concentrations, very high correlation with temperature, and distinct congener pattern lead us to conclude that the PCB signal measured at Swarthmore was due to volatilization of PCBs from construction materials used in Hicks Hall during the 1974 renovation. Field blanks from this site revealed no obvious contamination; however, data from Swarthmore were not used to calculate deposition fluxes and loads in this report. Particle-phase PCB concentrations followed roughly the same spatial trend as in the gas phase. The highest particlephase concentrations were observed at Camden (averaging 180 pg m-3) and Swarthmore (30 pg m-3) followed by Northeast (19 pg m-3), Delaware Bay (13 pg m-3), and Washington Crossing, Lum’s Pond, and Alloways Creek (all averaging about 6-8 pg m-3). Deposition to the Delaware River. The Stage 1 TMDL model divided the River into 105 junctions and 111 channels (17). Data from the closest of the seven monitoring sites was applied to each model segment to estimate wet and dry particle deposition of penta-PCB to the river during the model calibration period (September 2001 through March 2003 (1)). These loads were summed and are presented here as composites for Zones 2-5 (which are covered by the TMDL) and Zone 6. Although Zone 6 was not included in the TMDL, deposition to this zone was included in the model because PCB inputs there can impact Zones 2-5 via tidal mixing. Fatm consists of dry particle (Fdry), wet (Fwet), and gaseous (Fgas) deposition fluxes (all in units of ng m-2 day-1), which are calculated by applying the appropriate mass transfer coefficients (v in units of m day-1)

Fatm ) Fdry + Fwet + Fgas ) Cpvd + CVWMvp + Cgvg (2) where vd is the dry particle deposition velocity, vg is the gaseous deposition velocity or air-water exchange mass transfer coefficient (sometimes also referred to as KOL), and vp is the precipitation intensity (meters of rain per day). Cp is the particle-phase PCB concentration, Cg is the gas-phase PCB concentration, and CVWM is the volume-weighted mean concentration of PCBs in rainwater or snowmelt. Dry deposition (Table 1) was calculated by assuming a constant modeled deposition velocity (vd) of 0.5 cm s-1. The VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2173

chosen deposition velocity is similar to that used in the Chesapeake Bay model (0.49 cm s-1) (18). It is significantly higher, however, than the deposition velocity of 0.2 cm s-1 used to model deposition to the Great Lakes (19). The constant dry deposition flux described above was applied to the river at all time steps of the water-quality model, resulting in dry deposition loads of penta-PCB of ∼0.25 kg year-1 to Zones 2-5 and ∼0.36 kg year-1 to Zone 6. The dry deposition loads of ΣPCBs are ∼1.3 kg year-1 in Zones 2-5 and ∼1.8 kg year-1 in Zone 6. The loads to Zone 6 are higher than the loads to Zones 2-5 despite the lower concentrations assumed for Zone 6 due to its large surface area (1700 vs 240 km2 for Zones 2-5). Wet deposition (Fwet) was modeled by applying a volumeweighted mean concentration (CVWM) in rain whenever the meteorological data indicated that rain fell. Because rain samplers were not installed at all of the deposition sites, data from the other NJADN sites (Camden, New Brunswick, Chester, Sandy Hook, Jersey City, Tuckerton, and Pinelands) was regressed against annual average particle-phase PCB concentrations (Cp in pg m-3) and annual CVWM (pg m-3) (6). The intercept was not statistically significant and thus was assumed to be zero. Therefore, the slope (1.0 ((0.1) × 105, where the error represents the 95% confidence limit) is equivalent to the particle-phase scavenging ratio (Wp) and was found to be similar to values reported in the literature (ref 9 and references therein). Wet deposition loads were for penta-PCB 0.25 kg year-1 in Zones 2-5 and 0.12 kg year-1 in Zone 6 and for ΣPCBs 1.1 kg year-1 in Zones 2-5 and 0.61 kg year-1 in Zone 6. Because of the constant dry deposition velocity and scavenging ratio used and because rainfall is nearly constant across the study area, the ratio of the dry deposition flux to the wet deposition flux at each site is fixed at 3.9 (Table 1). Gaseous deposition is thought to be the dominant mechanism by which atmospheric PCBs are delivered to surface waters (6). For purposes of comparison with the wet and dry deposition fluxes, here we calculate the gaseous absorption of PCBs to the River by assuming a constant annual average vg for each homologue (20). By our simplified calculation, gaseous deposition of penta-PCB delivers ∼1 kg year-1 to Zones 2-5 and ∼1 kg year-1 to Zone 6. Gaseous deposition of ΣPCBs delivers ∼5 kg year-1 to Zones 2-5 and ∼7 kg year-1 to Zone 6. The water-quality model used a more complicated approach to calculating gaseous deposition. While dry and wet deposition are irreversible, gaseous deposition is reversible and therefore calculated in the water-quality model as a net flux, which is the balance between gaseous absorption/ deposition (Fgas) and volatilization (Fvol)

(

Fnet ) Fvol - Fgas ) vg Cd -

)

Cg Kaw

(3)

where Cd is the dissolved phase PCB concentration in water and Kaw is the dimensionless Henry’s Law constant (21, 22). The water-quality model calculated Cd and separately, used the temperature dependence of the gas-phase penta-PCB concentrations at each monitoring station (Table 1) to predict Cg in each model segment. At each time step, the model calculated vg based on temperature, wind speed, and river flow velocity (see ref 1 for details). The water concentrations (1) suggest, however, that the reverse process of volatilization of dissolved penta-PCB and ΣPCBs from the river to the atmosphere exceeds gaseous deposition in all river Zones. The water-quality model predicts that air-water exchange results in net losses of penta-PCB of 17 kg year-1 in Zones 2-5 and 26 kg year-1 in Zone 6. The yearly volatilization loss of ΣPCBs has not been calculated by the water-quality model, but if it is assumed that penta-PCB represents 25% of the 2174

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006

TABLE 2. Tributaries to the Delaware River and Their Penta-PCB Loads (1)

watershed

annual load,a g year-1

surface area, km2

watershed yield (Fws), ng m-2 day-1

E, %

vws, cm s-1

Brandywine Neshaminy Christina Rancocas Delawareb Salem Pennypack Raccoon Crosswicks Chester Alloways Schuylkill Poquessing Big Timber Frankford Pennsauken Mantua Cooper Darby

30 30 40 150 3330 70 40 40 110 70 80 2530 30 90 130 130 220 190 350

831 606 420 891 17560 303 145 127 339 174 176 4903 54 155 104 96 130 109 199

0.10 0.14 0.26 0.46 0.52 0.63 0.76 0.86 0.89 1.1 1.3 1.4 1.5 1.6 3.4 3.7 4.6 4.8 4.8

0.5% 0.7% 1% 2% 3% 3% 4% 4% 4% 6% 6% 7% 8% 8% 17% 19% 23% 24% 24%

0.003 0.004 0.008 0.01 0.02 0.02 0.02 0.02 0.03 0.03 0.04 0.04 0.04 0.05 0.1 0.1 0.1 0.1 0.1

a On the basis of normalizing the 577-day load given in ref 1 to 365 days. b Load from the Delaware River at Trenton (i.e., head of tide load).

ΣPCB flux, then the total loss of ΣPCBs is on the order of 170 kg year-1 from Zones 2-6. Thus, even though, at present levels, atmospheric deposition exceeds the TMDL, the atmosphere acts as a net sink for PCBs in Zones 2-6 of the Delaware River. The estimated wet + dry deposition of ΣPCBs exceeds the TMDL of 0.139 kg year-1 by more than an order of magnitude. Even so, wet and dry atmospheric deposition are thought to be responsible for less than 6% of the total load of PCBs to Zones 2-5 (1). The major loading categories are thought to be nonpoint sources (primarily urban runoff) accounting for about 26% of the total and inputs from point sources (∼17%). The loading from contaminated sites is estimated to be ∼10%. The various tributaries are thought to contribute ∼34% of the total load. Tributary loads are dominated by the Delaware River at Trenton and the Schuylkill River, which together contribute about 70% of the water in the Delaware River (1). Indirect Atmospheric Deposition. Some of the PCBs in the tributaries and the urban runoff come from indirect atmospheric deposition, i.e., atmospheric deposition to surfaces that is remobilized by rainwater to eventually enter the river. The DRBC constructed head-of-tide loadings for penta-PCBs in the tributaries listed in Table 2 (1). Examination of these tributary loads to the Delaware provides clues about the extent of indirect atmospheric deposition in this region. Also, these data can be used to develop a method for estimating tributary loads that arise from atmospheric deposition alone. This is a useful parameter because it represents a minimum load for each tributary. Sampling by DRBC indicates that local sources of PCBs, other than atmospheric deposition, exist in many of these watersheds. For example, both the Delaware River at Trenton and the Schuylkill Rivers are contaminated with PCBs from a variety of point and nonpoint sources (23). To attempt to identify which tributaries receive PCB inputs primarily from the atmosphere, the penta-PCB load was divided by the watershed surface area. This ratio represents a watershed yield (Fws) or flux of penta-PCB per unit area of the watershed. To compare it with atmospheric deposition fluxes, it is expressed in Table 2 in units of ng m-2 day-1, and the watersheds are arranged in order of increasing yield. This

analysis helps to prioritize watersheds to track down point sources of PCBs because watersheds at the bottom of the table, such as Darby Creek and Cooper River, probably harbor significant local PCB sources. In other words, atmospheric deposition is not the only source of PCBs in these watersheds, and efforts to track down other sources of PCBs in these watersheds are justified. The ranking of watersheds in Table 2 holds few surprises. Darby Creek runs through southern Philadelphia and harbors a PCB-contaminated landfill and several other industrial facilities. The Cooper River runs through Camden, NJ. Mantua Creek is home to a PCBcontaminated site under investigation by the state of New Jersey. This analysis also suggests that significant sources of PCBs exist in the Schuylkill River, a finding that is in accord with the industrialized nature of that river and it’s listing as impaired by the Commonwealth of Pennsylvania for this pollutant. While the Delaware River at Trenton has a relatively low PCB yield for its watershed area, it is known that PCB sources exist on one of its major tributaries (the Lehigh River). Watersheds at the top of the table are relatively unimpacted by PCB contamination. It is possible that atmospheric deposition is the only source of PCBs in these watersheds, and track down efforts in these watersheds are not likely to be successful. Watersheds with penta-PCB yields between 0.4 and 1.6 ng m-2 day-1 (except the Schuylkill River) display a significant (p < 0.05) correlation between watershed area and the size of the tributary load, as expected for watersheds in which atmospheric deposition is the major source of PCBs. This relationship is significant with or without the Delaware River at Trenton and Rancocas Creek, which may dominate the regression due to their large areas. A relationship of this type can be used to estimate loads to the river for tributaries that were not sampled for PCBs. The intercept is significant (p < 0.05), suggesting that although the atmosphere may be the dominant source of PCBs in these watersheds, it is not the only source. For watersheds in which atmospheric deposition is thought to be the dominant source of PCBs, it should be possible to estimate their loads from the atmospheric concentrations of PCBs by first calculating the atmospheric deposition flux (Fatm) and then estimating the pass-through efficiency (E) of PCBs in the watershed

E)

Fws Fatm

(4)

In practice, this is difficult because most of the models developed to convert atmospheric concentrations to deposition fluxes were developed to describe deposition to water surfaces not terrestrial ecosystems. Wet deposition can be assumed to be constant regardless of the surface to which the rain falls, but dry particle deposition and gaseous absorption are likely to be very different in terrestrial ecosystems. In fact, Horstmann and McLachlan (24) demonstrated that both dry particle and gaseous deposition velocities depend heavily on the type of terrestrial ecosystem. They calculated dry particle deposition velocities (via impaction and diffusion but not sedimentation of particles) for PAHs (and applied these values to PCBs) that were nearly 15 times greater in the deciduous vs coniferous canopies (0.73 and 0.050 cm s-1, respectively). Similarly, they calculated gaseous deposition velocities for PCBs 84, 90, and 101 (penta congeners) that were nearly 10 times greater in deciduous forest canopies than in coniferous canopies (1.15 and 0.13 cm s-1, respectively). These gaseous deposition velocities did, to some extent, take into account the reversible nature of the gaseous absorption process. Bohme et al. (25) suggest that gaseous absorption of PCBs to a variety of crops reaches equilibrium but that when it is kinetically limited a deposition velocity of 0.14 cm s-1 is applicable.

To obtain an estimate of the atmospheric deposition flux, we applied an average deposition velocity of about 0.6 cm s-1 to the entire land surface and assumed an uncertainty of 80%, which encompasses the highest and lowest deposition velocities cited above. This results in an atmospheric deposition flux (wet + dry + gas) of penta-PCBs of about 20 ng m-2 day-1 and results in the estimates of the pass-through efficiency (E) given in Table 2. The calculated values range from 0.5% to 24%, although the high calculated efficiencies for the more contaminated watersheds at the bottom of the table are not realistic. In this case, the lowest efficiencies of about 1% are most likely to be correct since they result from the “cleanest” watersheds, in which atmospheric deposition is most likely to drive PCB loads. Assuming 20% uncertainty in the atmospheric concentrations and watershed loads and 80% uncertainty in the deposition velocities, the propagated error in E is about 85%. As part of the phase 2 TMDL, the DRBC will estimate tributary loads of most PCB homologues which will be used to further examine the pass-through efficiency of PCBs in these watersheds. In particular, it will be useful to examine the heavier homologues for which gaseous absorption is relatively unimportant. Given the levels of uncertainty associated with the passthrough efficiency (E) and the atmospheric deposition velocities (vd) it may be more useful to relate the watershed yield (Fws) directly to the atmospheric PCB concentrations because both of these parameters are relatively easy to measure. This is accomplished by defining a new mass transfer coefficient (vws) for the process by which the watershed transfers PCBs from the atmosphere downstream

Fws ) vwsCatm = vwsCg

(5)

Since atmospheric deposition of penta-PCBs in these systems appears to be dominated by gas absorption, Catm can be approximated as Cg. vws is a useful parameter because it can be used to generate estimates of tributary loads to the river by multiplying by Cg and the watershed area (Aws)

Load ) vwsCgAws

(6)

Using a regional average gas-phase penta-PCB concentration of 40 pg m-3, the penta-PCB watershed delivery rates (vws) for watersheds with Fws < 1.3 ng m-2 day-1 range from 0.003 to 0.05 cm s-1 (Table 2). Again, the lowest values of about 0.005 cm s-1 are most likely to be correct since they arise from the most pristine watersheds. It should be noted that because of the many simplifying assumptions used to calculate E and vws, the values in Table 2 are related by a constant factor (E ) 1.7vws). If the uncertainties in Cg and the watershed loads are assumed to be 20%, then the propagated error in vws is about 28%. This relatively low degree of uncertainty is misleading, however, because the accuracy of vws relies on the underlying assumption that atmospheric deposition dominates the inputs of PCBs to these watersheds. This assumption is difficult, perhaps impossible, to validate. The deposition velocities of Horstmann and McLachlan (24) can also be used to estimate that indirect atmospheric deposition at regional background concentrations delivers 17-470 ng m-2 day-1 ΣPCBs to coniferous and deciduous forests, respectively. The Delaware River (including the portion above Trenton) drains a watershed area of about 35 000 km2. Although this area is not 100% forested, these fluxes can be multiplied by watershed surface area to give a rough estimate of the atmospheric deposition of PCBs to the watershed of between 200 and 6000 kg year-1 ΣPCBs. For comparison, the TMDL model used present-day inputs of penta-PCB to the tidal Delaware River of 49 kg during the 577-day model calibration period or about 120 kg year-1 of ΣPCBs. Thus, atmospheric deposition to the watershed is VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2175

large compared to the amount of PCBs actually reaching the river. This is logical considering that the organic matter in soil should strongly bind PCBs entering the watershed from the atmosphere. A fraction of the indirect atmospheric deposition may come from recycling of PCBs at regional background concentrations, but NJADN data has demonstrated that the urban area of Philadelphia/Camden has elevated atmospheric PCB concentrations and is acting as a source of PCBs to the regional atmosphere. Thus, the net export of atmospheric PCBs from this metropolis is probably supplying hundreds and perhaps thousands of kilograms of PCBs to the regional atmosphere each year, some fraction of which is re-deposited to the land surface.

Acknowledgments This study was funded by the DRBC and the NJ Department of Environmental Protection (project officer Michael Aucott). We particularly thank the staff of the Modeling & Monitoring Branch at DRBC and Joe Martini and Rick Greene at DNREC for help in collecting samples.

Literature Cited (1) Fikslin, T. J.; Suk, N. Total Maximum Daily Loads For Polychlorinated Biphenyls (PCBs) For Zones 2-5 Of The Tidal Delaware River, Report to the USEPA regions II and III, 2003; Dec 15, 2003. (2) Reinfelder, J. R.; Totten, L. A.; Eisenreich, S. J. The New Jersey Atmospheric Deposition Network (NJADN), New Jersey Department of Environmental Protection, Final Report, 2004. (3) Wisconsin Department of Natural Resources; Record of Decision, Operable Units 3, 4, and 5, Lower Fox River and Green Bay, Wisconsin, USEPA: 2003. (4) Wisconsin Department of Natural Resources; Record of Decision, Operable Unit 1 and Operable Unit 2, Lower Fox River and Green Bay, Wisconsin, USEPA: 2002. (5) Hudson River PCBs Site New York Record of Decision, USEPA: 2001. (6) Totten, L. A.; Gigliotti, C. L.; VanRy, D. A.; Offenberg, J. H.; Nelson, E. D.; Dachs, J.; Reinfelder, J. R.; Eisenreich, S. J. Atmospheric Concentrations and Deposition of PCBs to the Hudson River Estuary. Environ. Sci. Technol. 2004, 38, 2568-2573. (7) Brunciak, P. C.; Dachs, J.; Gigliotti, C. L.; Nelson, E. D.; Eisenreich, S. J. Atmospheric polychlorinated biphenyl concentrations and apparent degradation in coastal New Jersey. Atmos. Environ. 2001, 35, 3325-3339. (8) Totten, L. A.; Brunciak, P. A.; Gigliotti, C. L.; Dachs, J.; IV, G. T. R.; Nelson, E. D.; Eisenreich, S. J. Dynamic Air-Water Exchange of Polychlorinated Biphenyls in the NY-NJ Harbor Estuary. Environ. Sci. Technol. 2001, 35, 3834-3840. (9) VanRy, D. A.; Gigliotti, C. L.; T. R. Glenn, I.; Nelson, E. D.; Totten, L. A.; Eisenreich, S. J. Wet deposition of polychlorinated biphenyls in urban and background areas of the Mid-Atlantic states. Environ. Sci. Technol. 2002, 36, 3201-3209. (10) Hoff, R. M.; Muir, D. C. G.; Grift, N. P. Annual cycle of polychlorinated biphenyls and organohalogen pesticides in air in southern Ontario. Air concentration data. Environ. Sci. Technol. 1992, 26, 266-275.

2176

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006

(11) Wania, F.; Haugen, J.-E.; Lei, Y. D.; Mackay, D. Temperature dependence of atmospheric concentrations of semivolatile organic compounds. Environ. Sci. Technol. 1998, 32, 10131021. (12) Hornbuckle, K. C.; Eisenreich, S. J. Dynamics of semivolatile organic compounds in a terrestrial ecosystem-Effects of diurnal and seasonal climate variations. Atmos. Environ. 1996, 30, 39353945. (13) Hillery, B. R.; Basu, I.; Sweet, C. W.; Hites, R. A. Temporal and spatial trends in a long-term study of gas-phase PCB concentrations near the Great Lakes. Environ. Sci. Technol. 1997, 31, 1811-1816. (14) Hoff, R. M.; Brice, K. A.; Halsall, C. J. Nonlinearity in the Slopes of Clausius-Clapeyron Plots for SVOCs. Environ. Sci. Technol. 1998, 32, 1793-1798. (15) Simcik, M. F.; Basu, I.; Sweet, C. W.; Hites, R. A. Temperature dependence and temporal trends of Polychlorinated Biphenyl Congeners in the Great Lakes atmosphere. Environ. Sci. Technol. 1999, 33, 1991-1995. (16) Carlson, D. L.; Hites, R. A. Temperature Dependence of Atmospheric PCB Concentrations. Environ. Sci. Technol. 2005, 39, 740-747. (17) Suk, N. S. DYNHYD5 Hydrodynamic Model (Version 2.0) and Chloride Water Quality Model for the Delaware Estuary; Delaware River Basin Commission: 2003. (18) Leister, D. L.; Baker, J. E. Atmospheric deposition of organic contaminants to the Chesapeake Bay. Atmos. Environ. 1994, 28, 1499-1520. (19) Hoff, R. M.; Strachan, W. M. J.; Sweet, C. W.; Chan, C. H.; Shackleton, M.; Bidleman, T. F.; Brice, K. A.; Burniston, D. A.; Cussion, S.; Gatz, D. F.; Harlin, K.; Schroeder, W. H. Atmospheric deposition of toxic chemicals to the Great Lakes: A review of data through 1994. Atmos. Environ. 1996, 30, 3505-3527. (20) Totten, L. A. Present-Day Sources and Sinks for Polychlorinated Biphenyls (PCBs) in the Lower Hudson River Estuary. In Pollution Prevention And Management Strategies For Polychlorinated Biphenyls In The New York/New Jersey Harbor; Panero, M., et al., Eds.; New York Academy of Sciences: New York, 2005; p 18. (21) Bamford, H. A.; Poster, D. L.; Baker, J. E. Henry’s Law constants of polychlorinated biphenyl congeners and their variation with temperature. J. Chem. Eng. Data 2000, 45, 1069-1074. (22) Bamford, H. A.; Poster, D. L.; Huie, R.; Baker, J. E. Extrathermodynamic relationships and estimation of the temperature dependence Henry’s Law constants for PCBs. Environ. Sci. Technol. 2002, 36, 4395-4402. (23) A Report on the State of the Schuylkill River Watershed; The Conservation Fund: 2002. (24) Horstmann, M.; McLachlan, M. S. Atmospheric deposition of semivolatile organic compounds to two forest canopies. Atmos. Environ. 1998, 32, 1799-1809. (25) Bohme, F.; Welsch-Pausch, K.; McLachlan, M. S. Uptake of Airborne Semivolatile Organic Compounds in Agricultural Plants: Field Measurements of Interspecies Variability. Environ. Sci. Technol. 1999, 33, 1805-1813.

Received for review October 27, 2005. Revised manuscript received January 30, 2006. Accepted February 7, 2006. ES052149M