treatment dose.
We also calculated PSD accumulation factors (PAFs), i.e. ratio of PSD levels (dw basis) to TOC-normalized soil PBDE levels. PAFs were much lower than BSAFs for worms, i.e. from 0.03 to 0.2 versus 4.8 to 20, respectively (Table SlO). As observed in worms, PBDE concentrations in the PSDs generally increased with soil dose while PAFs generally decreased or were unchanged with dose. Calculated PAFs generally tracked
those of medium and high dose worm BSAFs (0.71<r2<0.97), whereas low dose PAFs
were not as well correlated (r2=0.58). PBDE partitioning into PSDs presumably occurs
via passive partitioning between substrate TOC and the surface layer of the PSD polymer
phase. Regression analysis of PSD and ADB-amended soil PBDE congener burdens
(TOC-normalized) revealed that the two quantities were well correlated in 14 d (^0.91) and 28 d (r2>0.89) treatments.
While previous lab studies have demonstrated the availability of PBDEs following their addition via solvent carrier to soil our results demonstrate that PBDEs are similarly bioavailable from soils amended with sewage sludge biosolids generated by different wastewater treatment processes. Class B biosolids (consistent with our ADB) is the most commonly land-applied in North America. "EQ" biosolids (e.g. our CB) are promoted as pathogen-free and containing fewer contaminants. EQ distribution to and use by the general public is unregulated and essentially untracked. Its recent distribution as "organic compost" has been controversial in some US cities, e.g. San Francisco (Washingtontimes.com, 2010). Data on PBDE burdens in worms collected from historically sludge-amended field sites are crucial. However, the extent of exposure is difficult to characterize. Nonetheless, BSAFs estimated from controlled exposures to well characterized biosolid-amended soils (dependent upon dose) in the lab were comparable to values reported for worms of several species obtained from biosolids-amended agricultural soils in Sweden (Matscheko et al., 2002; Seilström et al., 2005).
This is important as E. fetida is not common in agricultural soils and species-specific behavior in soils could impact contaminant accumulation. For example, Kelsey et al.
(2005) reported that DDE accumulation from natural soils by E. andrei was significantly higher compared to uptake by Lumbricus terrestris (anecie) and A. caliginosa. In contrast, however, van der Wal et al. (2004) and Jager et al. (2005) found that uptake of PCBs and organochlorine pesticides from natural soils by the epigeic E. andrei and
endogeic Apporectodea caliginosa was comparable. From our PUF bioassay results, it is clear that earthworms exposed to small amounts of PUF particles in soil are able to accumulate PBDEs therein to high levels with no apparent toxicity. The earthworm burdens may be readily transferred further into terrestrial food webs. Though Penta-BDE manufacture has ceased, emission from extant and relic PUF will likely contribute substantially to global emissions for decades to come.
Figures
Figure 1. Mean £Penta-related PBDEs in worms ^g/kg lipid) versus levels in high dose
ADB- and CB-amended soils and low dose Penta-BDE spiked AS ^g/kg TOC) after 28 d. Worm concentrations exceeded those in soils in all non-control treatments (p<0.01;two-sample t-test, assuming unequal variance). Error bars represent standard deviations
from the means.
80000
60000
Ü0
Ì
QS3 40000
Q-20000
? earthworm lipid
? substrate TOC
ADB-amended soil CB-amended soil
exposure substrate
SAS
Figure 2. PBDE congener composition in worms exposed to soil with dispersed PUF microparticles and solvent-spiked Penta-BDE for 28 d (N=3); compared to PUF, PUF-amended soil and a commercial Penta-BDE formulation, DE-7 1 . Error bars are standard deviations from the means.
50
¦ 28d PU F-soi I worm
¦ 28d spiked soil worm
¦ 28d PUF-soil substrate
40 ¦ PUF
? DE-71
^
G 30
IO
(J UJ
5 20
10
154 153
85
100 99
47
PBDE congener
Figure 3. Mean worm BSAFs (lipid/TOC normalized basis; N=3) for BDE-47, 100 and 99. Low and high dose ADB and CB and low dose SAS soil exposure results on 28 d shown. Soil XPBDE concentrations in µg/kg dw (see legend). For each congener, bars with different letters are statistically different (ANOVA with Tukey HSD testing). Pair-wise comparisons: ADB low versus high doses (BDE 47 p<0.001; BDE 100, 99 pO.Ol);
CB low versus high doses (BDE-47, 100, 99 p<0.01). High dose BSAFs for all congeners in all biosolid treatments are significantly lower than low dose BSAFs (0.001<p<0.05). For the SAS substrate, means of all congener BSAFs are significantly different from each other, i.e. BDE 47>100>99 (0.001<p<0.05). Error bars represent
standard deviations from the means.
ADB-low(72)
? CB-low(48) ADB-high(169) CB-high(196) SAS-low(676)
PBDE
Figure 4. Comparison of mean PBDE BSAFs calculated for earthworms exposed to the PUF-amended soil mixture (83.3 mg/kg dw) and the solvent-spiked artificial soil (SAS;
5.6 mg/kg dw) here and those previously reported for earthworms exposed to PBDE-spiked artificial (Asamoah, 2005; Nyholm, 2009) and natural (Liang et al., 2010) soils and biosolids-amended field soils (Seilström et al., 2005; Matscheko et al., 2002).
40
E3 14d PUF-soil worm
? 28d PUF-soil worm
? 14d spiked soil worm
35
? 28d spiked soil worm
? Nyholm, 2009
¦ SellstrometaL 2005 30
? Matscheko et al., 2002
I
? Asamoah, 2005Additional notes: For the Matscheko et al. (2002) data, we calculated means for their reported BSAFs for seven (BDE-47 and 99) and four (BDE 100) sludge-amended field sampling sites, respectively. Error bars are standard deviations from treatment means for this study and our meta-analysis of the Matscheko et al. (2002) means. Error bars for Liang et al. (2010) data are their reported standard deviations for BDE 47. In that study, means for all other congeners were reported as a range and are thus not plotted. Error bars are standard error of the mean for Seilström et al. (2005) data and one standard deviation from the mean for the Nyholm (2009) data. No replication was reported in the Asamoah (2005) study. Note that only highest dose (100 µg/kg dw) BDE-47 BSAF data were available for analysis from the Liang et al. (2010) study (all other means given as a range). Missing congeners were not reported for a given data set.
Figure 5. Mean 28 d Penta-BDE BSAFs (for low- and high-dose biosolids-exposed
worms and low-dose Penta-BDE-spiked SAS-exposed worms) versus published log Kow values (Braekevelt et al., 2003). Congeners for which data are missing were below QL.Error bars represent standard deviations from the means (N=3).
1^ = 0 71 ?ADB-low
25 r2 = 0.78 »CB-IOW
r2 = 0.74 »ADB-high r2 = 0-91 1 ACB-high r* = 0.74 XSAS-low 20
15
to CQ
X
10 85
47 100 99
154 153
?