Reply to Hrudey: Tracking the extent of oil sands airborne pollution.

We welcome Hrudey’s comments (1) regarding “the most noteworthy findings” of Kurek et al. (2) and have thus focused our response to discuss “several features” of the polycyclic aromatic hydrocarbon (PAH) data (and related dibenzothiophenes, DBTs) from Namur Lake. First, historic forest fires cannot adequately explain the overall increasing trends of ΣPAHs, C1–C4 alkylated PAHs, and DBTs at Namur (Fig. 1 and figure 1A in ref. 2). Retene, a characteristic marker of conifer fires (3), is very low throughout the sediment record, and there is no significant correlation between the percentage of retene and ΣPAH flux (radj = 0.06, F = 2.3, P = 0.15). In addition, PAH ratios of retene to unsubstituted and C1–C4 alkylated PAHs have declined toward modern times at Namur (figure S3 in ref. 2), which is inconsistent with forest fires being a major contributor to the overall PAH trends. Second, all of our study sites located west of the bitumen upgraders and major open-pit mining areas show variable PAH trends. This finding can be attributed to their greater distance from the upgraders and to prevailing winds, which will deliver pollutants in pulses and may contribute to the observed variation. Additionally, the quality-control data for the sedimentary PAH analysis from Namur were very good and generally met laboratory performance standards in terms of blanks, recoveries of spiked unsubstituted and alkylated PAHs, and of deuterated internal standards. Thus, we are confident that the “oscillations” are reliable measurements. Third, apart from only two intervals, Namur ΣPAH flux values are comparable to four of our five near-field study lakes, except NE20 (figure S2 in ref. 2). The average sedimentation rate at Namur (0.027 g·cm·yr) and its temporal pattern are also similar to our near-field study lakes (except NE20), and other mid-latitude boreal forest lakes (4). It is possible that winter snowpack sampling from one season may underestimate the true spatial deposition pattern of PAHs, as priority pollutants were recorded as far as 85 km from the upgraders (5). These multiple lines of evidence suggest a detectable, although variable, PAH pollution signal at Namur. Finally, Hrudey notes (1) “the striking concordance” between sedimentary PAHs and inferred chlorophyll a (VRSchla). Our study was not the first to recognize increased primary production driven by climatic warming; notable shifts in production are a hallmark of paleolimnological studies from the greater region (2). Unfortunately, we have only limited organic carbon data for Namur, although extensive data indicate that VRS-chla (available for near-field lakes) tracks primary production patterns well (6). We are also continuing to investigate historic PAH deposition patterns under the Joint Canada Alberta Oil Sands Monitoring Plan and will soon have additional data to address the issue of deposition in distant lakes. Although government funding of oil sands environmental monitoring before 1997 did provide some baseline data (e.g., condition and mercury levels in Athabasca River fish), it did not focus on temporal patterns in PAH deposition or the long-term ecological conditions in small lakes receiving emissions from nearby upgraders. Instead, the focus has been on variable river environments and downstream impacts. Reviews of previous monitoring programs have also largely been critical of their past monitoring approaches.