ABSTRACT Background. Sydney Harbour (SH), Nova Scotia has long been subject to effluent and atmospheric inputs of metals, polycyclic aromatic hydrocarbons (PAHs), and polychlorinated biphenyls (PCBs) from a large coking operation and steel plant that operated in Sydney for nearly a century until closure in 1988 (Fig 1). Contaminated effluents from the industrial site resulted in the creation of the Sydney Tar Ponds (STPs), one of Canadas largest contaminated sites (Fig 2). Since its closure, there have been several attempts to remediate this former industrial site and finally, in 2004, the governments of Canada and Nova Scotia committed to remediate the site to reduce potential ecological and human health risks to the environment. The STPs and Coke Ovens cleanup project has become the most prominent remediation project in Canada today. Approach. As an integral part of remediation of the site (i.e., which consisted of solidification/stabilization and associated capping of the STPs), an extensive multiple media environmental effects monitoring (EEM) program was implemented to assess what effects remediation had on the surrounding environment, and, in particular, harbour sediments. Additionally, longer-term natural sediment recovery rates of select contaminants predicted for the harbour sediments were compared to current conditions. During remediation, potential contributions to sediment quality, in addition to remedial efforts, were evaluated which included a significant harbour dredging project, propeller wash from harbour traffic, storm events, adjacent loading/unloading of coal and municipal wastewater treatment discharges. Two sediment sampling methodologies, sediment grab and gravity corer, were also compared to evaluate the detection of subtle changes in sediment quality. Results. Results indicated that overall spatial distribution pattern of historical contaminants remains unchanged, although at much lower concentrations than previously reported, due to natural recovery. Measurements of sediment indicator parameter concentrations confirmed that natural recovery rates of SH sediments were in broad agreement with predicted concentrations, in spite of ongoing remediation activities. Overall, most measured parameters in sediments showed little temporal variability even when using different sampling methodologies, during three years of remediation compared to baseline, except for the detection of significant increases in total PAH concentrations noted during year one of remediation monitoring. The data confirmed the effectiveness of mitigation measures implemented during construction relative to harbour sediment quality and of monitored natural recovery (MNR) as a management tool, despite other anthropogenic activities (e.g., dredging) and the dynamic nature of the harbour. METHODS Four assessment areas and 11 monitoring stations (~10 m deep), were sampled in SH by boat to monitor sediment and other variables throughout this study during 2009 baseline (pre- remediation), and three years of remediation (i.e., year , year and year ) (Fig 3). Assessment areas comprised of area 1–near-field; area 2–mid-field; area 3–far-field, and area 4– Sydney River estuary. Annual surface sediments (0-1 cm) were sampled as part of an EEM program and simultaneously to assess MNR. Sediments were analyzed for PAHs, PCBs, TOC, metals, and grain size. Two sediment sampling devices (an Ekman grab [a] and gravity corer [b]) were tested and compared. Sediment traps were deployed in SH to determine sediment deposition rates and to assess chemical composition for PAHs, PCBs, TOC and metals (c). Monthly water sampling was also conducted for organic and inorganic chemical parameters (d). Biota sampling for analysis of Rock crab hepatopancreas for PAHs, PCBs, metals and lipid content (e) was another component of the study. RESULTS & DISCUSSION The annual monitoring of contaminants in surface sediments and occasional monitoring of sediments collected in sediment traps in SH has not detected evidence of substantial contaminant releases into the marine environment as a result of remediation activities at the STP site. This monitoring was also used to assess MNR as a management tool. Intra-station comparison of sediment contaminant concentrations using the Ekman grab and the gravity corer device did not reveal significant differences in contaminant concentrations, therefore, the Ekman grab was used throughout this remediation monitoring study. Measured total PAH concentrations in sediments during remediation monitoring were higher than baseline and generally the highest concentration were observed in near-field stations (Fig 4 top), but despite the initial increases observed during start-up of remediation activities (i.e., Year 1), there has been a continued decrease with each subsequent monitoring year supporting the prediction of natural sediment recovery made by Smith et al. (2009) (Fig 4 bottom and inset). Results of the Year 3 remediation monitoring were lower than Year 1 and Year 2 remediation monitoring and overall variation across harbour stations was lower. These recent PAH concentrations reported here are at least an order of magnitude lower than peak concentrations measured in the 1980s. This in part may be due to the positive effects of recent harbour dredging activities associated with the proposed Sydport container terminal which may have anthropogenically enhanced natural harbour recovery predicted by Smith et al. (2009). PCB and metal concentrations (Pb shown) measured in sediments during remediation monitoring were comparable to baseline, and did not indicate substantive releases of these contaminants to the SH marine ecosystem during remedial activities (Fig 5 and Fig 6). Furthermore, Pb, Hg and Zn contaminants may be slightly decreasing during remediation monitoring, but this pattern may not be conclusive for all metals as some metal concentrations are so close to detection levels which often create noise when interpreting the data relative to baseline conditions. Whilst Pb may be slightly lower than compared to baseline, these patterns may not be statistically significant as there are too few monitoring events (i.e., 4 years) for robust statistical comparison. Therefore, we recommend long-term monitoring of these sediment contaminants in SH to confirm these MNR predictions. Sediment grain size variation observed during this study at some mid-field stations were not stable in SH (data not shown), indicating that sediments within SH are dynamic and maybe influenced by storm resuspension. Other physical sediment properties, such as, TOC have remained stable throughout the remediation monitoring study. Sediment chemical quality measured in sediment traps during recent dredging activities indicated lower contaminant concentrations compared to baseline and sediment deposition rates across SH were low (i.e. <1 cm yr -1 ) (Fig 7). This suggests that sediment trap data support the positive impacts on SH sediments from recent large scale dredging activities, where measured deposition of less contaminated sediment were between 7-35 cm in just 3 months (Fig 8). Generally water indicator parameters measured throughout the program were below detection levels with only occasional detections (Dillon, 2012a, b; data not shown). PAH concentrations were not detected in Rock crabs and PCB concentrations indicated a decreasing trend (Fig 9; Walker et al., submitted a). This coupled with associated mitigative measures implemented during remediation activities at the STP site and natural sediment recovery, as predicted by Smith et al. (2009), are having positive impacts on SH sediments. CONCLUSIONS The annual monitoring of most marine sediment parameters in SH has not detected evidence of substantial contaminant releases into the marine environment as a result of ongoing remediation activities at the STP site. However, total PAH concentrations in sediments have increased significantly during remediation compared to baseline monitoring, but absolute concentrations are low compared to historical levels and this slight increase represents only a short term interruption in the overall natural recovery of sediments in SH (Walker et al., submitted a, b). These recent ground truth results of sediment chemistry support the modeled predictions made by Smith et al. (2009) of natural containment of the main inventory of contaminants deposited during the s by layers of cleaner surficial sediments. The continuing trend of natural sediment containment may have enhanced as a result of positive impacts of recent dredging activities in SH, but will require further monitoring and assessment to confirm these trends using MNR. Not withstanding, results presented here confirm modeled predictions of natural sediment containment, despite the ongoing land based remediation activities at the STP site. Therefore, ongoing monitoring of these harbour sediments and other ecological indicators becomes increasingly important, whether the aim is to assess impacts of industrial activities (e.g., remediation of a former industrial sites or harbour dredging activities) or to gauge the rates of natural sediment recovery using MNR. ACKNOWLEDGEMENTS This study was supported by the STPA. Valuable input and improvements to this paper have been made by representatives of Public Works and Government Services Canada and the Environmental Management Committee. Particularly we would like to thank John Smith, Michael Parsons, Brent Law and Timothy Milligan from the Bedford Institute of Oceanography for their insightful comments. FIGURES Assessing Sydney Tar Ponds Remediation and Natural Sediment Recovery in Nova Scotia, Canada Tony R. Walker a, Andrew Thalheimer a, Devin MacAskill a, Peter Weaver b a Dillon Consulting Limited, Nova Scotia, Canada; b Sydney Tar Ponds Agency, Sydney (STPA), Nova Scotia, Canada REFERENCES 1.Dillon (2012a) Marine Report for Year 2 Construction. Submitted to the Sydney Tar Ponds Agency (STPA). 2.Dillon (2012b) Sydney Harbour, Nova Scotia – Water Quality Monitoring Program Final Report. Submitted to Public Works and Government Services Canada (PWGSC). 3.Smith, J.N., Lee, K., Gobeil, C., MacDonald, R.W. (2009) Natural rates of sediment containment of PAH, PCB and metal inventories in Sydney Harbour, Nova Scotia. Science of the Total Environment, 407, Walker, T.R., MacAskill, D., Weaver, P. (submitted a) Assessment of contaminants in Rock crabs in Sydney Harbour during remediation of the Sydney Tar Ponds, Nova Scotia, Canada. Journal of Environmental Monitoring and Assessment. 5.Walker, T.R., MacAskill, D., Rushton, T., Weaver, P. (submitted b) Assessing Sydney Tar Ponds remediation and natural sediment recovery in Nova Scotia, Canada. Journal of Environmental Monitoring and Assessment. Fig 4. (Top) Histogram of total PAH concentrations in surface sediments during baseline and 3 years of remediation representing a single composite sample from three grab samples per station. Far-field stations (*) are indicated. (Bottom) Box plot of total PAH concentrations showing temporal variation. Solid horizontal lines indicate ER-M (44.8 μg g -1 ) and dashed lines indicate the ER-L (4.02 μg g -1 ). (Inset) Predicted ranges (PR) for PAH concentrations (μg g - 1 ) for 2020 reported by Smith et al. (2009). Fig 7. Mean sediment deposition rates for each assessment area calculated from pooled sediment traps individual stations during baseline. Fig 8. Trailing-suction hopper dredger Oranje. Over 4.2 million m 3 of sediment was removed from a 9 km channel which was used for the construction of a confined disposal facility. Representation of the dredged harbour channel and new port facilities at Sydport, SH (top left). Suspended sediment plume following discharge of sediment for the constructed containment area (top right). Fig 9. Box plot of PCB concentrations (μg g -1 ) in Rock crab hepatopancreas tissue showing temporal variation during baseline and 3 years of remediation. Detection limit (DL) (0.05 µg g -1 ) and Canadian Food Inspection Agency (CFIA) limit (2 µg g -1 ). Fig 1. Sydney steel and coking plant Fig 2. North and South Tar Ponds in relation to SH c b a d e Fig 3 Marine Stations Fig 5. (Top) Histogram of PCB concentrations in surface sediments during baseline and 3 years of remediation representing a single composite sample from three grabs per station (RDL = 0.01 μg g -1 ). Far-field stations (*). (Bottom) Box plot of PCB concentrations showing temporal variation. Solid horizontal lines indicate ER-M (0.18 μg g -1 ) and dashed lines indicate ER-L (0.023 μg g -1 ). (Inset) Predicted ranges (PR) for PCB concentrations (μg g -1 ) for 2020 reported by Smith et al. (2009). Fig 6. (Top) Histogram of Pb concentrations in surface sediments during baseline and 3 years of remediation representing a single composite sample from three grabs per station. Far-field stations (*). (Bottom) Box plot of Pb concentrations showing temporal variation. Solid horizontal lines indicate CCME PEL and dashed lines indicate CCME ISQG values. (Inset) Predicted ranges (PR) for Pb concentrations (μg g -1 ) for 2020 reported by Smith et al. (2009).