Bluffer’s Park Beach in the Toronto and Region Area of Concern had a history of beach postings often exceeding 80% of the beach season since the 1980s. A study applied expanded E. coli surveillance and microbial source tracking techniques in 2005–2007 to identify fecal pollution sources contributing to beach postings. Expanded surveillance in the beach vicinity identified significant E. coli hotspots in the foreshore beach sand (pore water max E. coli = 255,000 CFU 100 ml−1) and associated with a marsh inland of the beach. During rain events, streams from the marsh (max E. coli = 173,000 CFU 100 ml−1) and runoff from the parking lot (max E. coli = 4100 CFU 100 ml−1) were observed to overflow across the beach to contaminate beach waters. Microbial source tracking using library-dependent (antibiotic resistance and rep-PCR DNA fingerprinting of E. coli isolates) and library-independent (human HF183 bacterial DNA marker) methods indicated the prevalence of animal fecal pollution sources at the beach rather than human sewage. These results were consistent with sanitary survey information, observations of wildlife in the marsh area, and Gulls and Canada Geese on the beach. In 2006, a bird management program was initiated, and remedial actions continued in advance of the 2008 bathing season to engineer a berm to prevent marsh runoff into beach water and re-direct parking lot drainage into the marsh. Since these remediation actions, Bluffer’s Park Beach has been posted less than 20% of each beach season, and it was awarded a Blue Flag accreditation in 2011.
Beaches are important water quality symbols for the public. If a beach is frequently posted and the public is afraid to swim, perceptions of poor water quality conditions can extend far beyond the beach. For this reason, reducing beach postings and associated Beneficial Use Impairments (BUIs) are an integral part of reducing health risks and demonstrating improved water quality conditions in many Great Lakes Areas of Concern (AOC).
Bluffer’s Park Beach is located on Lake Ontario in the Toronto and Region Area of Concern. Characterized by a beautiful backdrop of the Scarborough Bluffs, an extensive marsh and wooded area occurs behind the beach and extends part way up the Bluffs (Figure 1). However, Bluffer’s Park Beach had a history of beach postings often exceeding 80% of the beach season since the 1980s (Environmental Defence, 2004; City of Toronto, 2009). This study applied expanded E. coli surveillance, groundwater investigations, and microbial source tracking techniques (US EPA, 2005; Edge and Schaefer, 2006) to better understand the sources of E. coli causing beach postings at Bluffer’s Park Beach. We summarize water quality research conducted from 2005 to 2007 that helped guide remedial actions by the City of Toronto and the Toronto and Region Conservation Authority to turn one of Toronto’s worst beaches into one of its best.
Surface water samples were collected weekly at Bluffer’s Park Beach by wading out from the shoreline during bathing seasons from 2005 to 2007. All water samples were collected in sterile polypropylene bottles and returned on ice for lab analysis within several hours. Samples were collected at ankle and chest depths along two transects perpendicular to the shoreline that were at the City’s sampling locations 47E and 51E. Based upon sanitary survey information, surface water samples were also collected from the marsh and associated streams behind the beach, rain-event runoff from a parking lot near the beach, and from a stormwater outfall and stormwater retention pond on the other side of the park west of the public boat launch area. Groundwater samples were collected from bore holes as part of groundwater investigations below the beach on two sampling occasions, in September 2006 and April 2007, as outlined in Edge et al. (2007a). Sand samples were obtained from the wet foreshore sand within a meter of the lake, using a sterile plastic corer (diameter = 2.5 cm). Additional samples of sand pore water were collected by digging a hole in the wet foreshore sand about one meter inland from the lake, and collecting the pore water seeping into the hole. General observations of the number of animals were recorded during sampling visits each week, and any fecal droppings on the wet foreshore sand within two meters of the waterline were enumerated. Fecal source sampling was conducted simultaneously to collect E. coli isolates and DNA extracts for microbial source tracking analyses. Municipal wastewater samples were collected from pumping stations near Bluffer’s Park and from the Ashbridges Bay Sewage Treatment Plant. Fecal samples from birds (gulls, Canada Geese, ducks, cormorants, swans) were obtained from fresh fecal droppings on the beach near Bluffer’s Park Beach and from the Toronto area.
Water and wastewater samples were analyzed by membrane filtration and E. coli enumeration was expressed as colony forming units per 100 ml of water (CFU 100 ml−1). Serial dilutions of water samples were performed and membrane filters were placed on the chromogenic differential coliform (DC) agar media supplemented with cefsulodin (Oxoid Inc.) for 18 h incubation at 44.5 °C. Reference water samples were routinely filtered as negative controls. Sand pore water samples were analyzed following the same filtration method as surface water samples, while sand samples were analyzed by a blender-based method (Edge et al., 2007a) and E. coli enumeration was expressed as colony forming units per gram of dry sand.
For library-based microbial source tracking methods used in 2006, antibiotic resistance analysis and rep-PCR DNA fingerprinting techniques were used to compare the similarity of water and sand E. coli isolates to those from known fecal sources (Edge and Hill, 2007; Edge et al., 2010). If E. coli isolates could not be reliably classified as more similar to bird or wastewater isolates, they were classified as unknown source. The growth of E. coli isolates on agar plates with antibiotics was compared to their growth on a control plate without antibiotics to develop a resistance profile. E. coli isolates were also lysed for DNA fingerprinting using a BOX-PCR primer approach using a Mastercycler Gradient PCR system (Eppendorf, Hamburg, Germany). For the library-independent method in 2007, an endpoint polymerase chain reaction (PCR) assay was used to analyze water samples for the presence of a DNA sequence unique to Bacteriodales bacteria in the human gut (Edge et al., 2010). Up to 300 ml of water samples was filtered on 0.45 µm membrane filters, which were then frozen at -80 °C for subsequent DNA extraction using a Powersoil DNA isolation kit (Mo BIO Laboratories, Inc.). A 1 µl DNA extract was used as template in an endpoint polymerase chain reaction (PCR) assay using primer HF183F to amplify the human Bacteroidales DNA sequences and BAC32 to amplify universal Bacteroidales sequences if they were present in the sample. A human fecal DNA extract was run as a positive control for each set of reactions, along with sterile water as a negative control.
Results and discussion
Sanitary surveys and animal observations
Sanitary survey results in the vicinity of Bluffer’s Park Beach did not provide any clear indication of the presence of a significant human sewage source. Dye and smoke testing of local sewage infrastructure, including a public restroom at the beach and sanitary sewers on top of the bluffs, did not reveal any signs of sewage leakage (City of Toronto, 2007). During our field observations, a used diaper was observed on one occasion in the stream draining the marsh area near the parking lot at Bluffer’s Beach. There was also a dumpster located in the northeast corner of this parking lot that could have received such wastes. It is possible that these types of human fecal sources could have presented some sporadic human fecal contamination in the beach area. The marina at Bluffer’s Park has a pumping station to handle fecal wastes from boaters; however there was no evidence to indicate boaters were a significant source of E. coli loading into the Beach area. A stormwater outfall and the outlet of the Dunkers stormwater storage facility occur on the far western side of Bluffer’s Park. These outfalls are small and likely too far away to have an impact on Bluffer’s Park Beach.
An extensive marsh and wooded area occurs behind Bluffer’s Park Beach and abuts up against the bluffs. Streams were observed to drain from the marsh across the beach toward the lake during rain events. The most prominent stream over the study period was at the western end of the beach near the parking lot. During dry weather, this stream only reached about half to three-quarters of the way across the beach toward the lake before drying up. During rain events, this stream, and several others, flowed directly across the beach into the lake. During rain events, runoff from the parking lot also flowed into the marsh stream and entered the lake. A variety of urban wildlife species were known to occur in the marsh area behind Bluffer’s Beach. On sampling visits we observed deer, coyotes, foxes and raccoons. Raccoons were also observed around the parking lot and the dumpster. The only animals observed regularly on Bluffer’s Beach from 2005 to 2007 were birds; notably Ring-billed Gulls (Larus delawarensis) and Canada Geese (Branta canadensis). Gulls commonly numbered between 50 to 200 birds, although over 500 Gulls were counted in the beach area on one occasion. The numbers of Geese were usually less than 25 birds, although over 150 Geese were counted on one occasion. Swans, ducks, and dogs were occasionally seen. Gull fecal droppings were commonly observed within two meters of the lake (usually more than 50), although numbers along the beach exceeded 500 droppings on one occasion. Canada Geese droppings were fewer (usually less than 50), although numbers exceeded 400 on one occasion. These Gull and Geese droppings on the foreshore sand were subject to wave action in the swash zone, and were observed to be a source of E. coli into adjacent beach waters.
E. coli surveillance
The highest E. coli concentrations measured in this study were usually in the foreshore beach sand, and results from 2006 are presented in Figure 2. E. coli concentrations in sand pore water were as high as 55,000 CFU 100 ml−1 in 2006 (10 July) and 255,000 CFU 100 ml−1 in 2007 (23 July). The highest E. coli concentration recorded in sand samples was 10 887 CFU g−1 dry sand on 24 July 2006, although its concentration was only 2 CFU g−1 dry sand on 12 December 2005. The finding of relatively high concentrations of E. coli in beach sand is a common phenomenon at beaches around the Great Lakes (Whitman and Nevers, 2003; Alm et al., 2003; Edge and Hill, 2007; Staley and Edge, 2016). Beach sand can serve as a reservoir for E. coli, and a secondary source of E. coli entry into adjacent beach water (Edge and Hill, 2007; Vogel et al., 2016). The E. coli in Bluffer’s Park beach sand would be subject to wave resuspension and transport into Bluffer’s Park beach waters.
E. coli concentration gradients were found across Bluffer’s Park Beach depth zones (Figure 2). E. coli concentrations were higher in ankle depth water than chest depth water in each year. Since Bluffer’s Park Beach is a very shallow beach, water samplers typically needed to wade out 50 meters or so offshore to reach chest depth. It is possible that sampling in more shallow water associated with transient sand bars could have contributed to some higher E. coli counts in the past.
Expanded E. coli surveillance found several E. coli hotspots in the immediate vicinity of the beach associated with the marsh and parking lot (Figure 2). High concentrations of E. coli were found in the ephemeral stream draining the marsh near the parking lot, with concentrations as high as 7167 CFU 100 ml−1 in 2005 (28 June), 173,000 CFU 100 ml−1 in 2006 (10 July) and 21,300 CFU 100 ml−1 in 2007 (27 May). There was a seasonal pattern for E. coli concentrations in the stream runoff with a significant increase in June and July followed by a decline in August down to lower levels in the fall. Similar E. coli results were reported by Lake Ontario Waterkeepers in 2005. Higher E. coli concentrations in the stream were associated with rain events and large increases in water turbidity from clay-like particles in the stream, likely originating from erosion of the bluffs. Parking lot runoff was also observed during rain events, where it flowed onto the beach to join the flow of a stream draining the marsh. Concentrations of E. coli in parking lot runoff ranged from 700 to 4100 CFU 100 ml−1 in 2006, with one sample from 2007 at 2100 CFU 100 ml−1. Gulls were the most common animal observed in the parking lot area, although other birds (e.g. sparrows and starlings) and raccoons were occasionally observed there on sampling days. The stormwater outfall on the other side of Bluffer’s Park had E. coli concentrations reaching as high as 15,600 CFU 100 ml−1 in 2006 and 23,000 CFU 100 ml−1 in 2007.
Groundwater studies were conducted in September 2006 and April 2007 and were described in Edge et al. (2007a). These studies showed that the depth to the water table varied from approximately 0 m (water table intersects the ground surface) at the lake-beach interface to approximately 1.0–2.0 m at the landward edge of the beach where the sand dunes and vegetation are present. The depth to the water table increased with increasing elevation of the surface of the beach away from the lake, with a total rise in ground surface elevation of approximately 2.1 m. The elevation of the water table below the beach during September 2006 declined across the beach towards the lake with a slope or hydraulic gradient of 0.00675. This hydraulic gradient is a bit lower than the hydraulic gradients measured at other beaches of the Great Lakes (0.010 to 0.024). Given that the hydraulic conductivity of the beach sand is approximately 3.6 cm hr−1 and the porosity is approximately 0.35, the groundwater velocity beneath the beach was estimated to be approximately 1.5 cm d−1 or 6 m yr−1 towards the lake. E. coli analyses from September 2006 bore hole sampling showed that there was no E. coli present in the groundwater below Bluffer’s Park Beach up to 50 m towards the shore line from the marsh (60 m to 100 m from the shoreline). Although groundwater was flowing from the marsh towards the lake, there was no evidence that E. coli from the marsh was being transported via groundwater flow to the shoreline.
Microbial source tracking
An E. coli library (decloned) of 1133 bird fecal isolates and 976 wastewater isolates from the Toronto area was used for library-dependent microbial source tracking analyses as outlined in Edge and Hill (2007) and Edge et al. (2010). About 2500 E. coli isolates collected over the bathing season in 2006 from the marsh stream, beach sand, and beach water at Bluffer’s Park Beach were more often similar to bird E. coli (generally 35 to 45% of isolates) than wastewater E. coli (generally 20 to 30% of isolates) by both antibiotic resistance analysis and rep-PCR DNA fingerprinting. In contrast, about 400 E. coli isolates from the stormwater outfall on the other side of Bluffer’s Park were more often similar to wastewater E. coli (about 35% of isolates) than bird E. coli (about 30% of isolates) by both antibiotic resistance analysis and rep-PCR DNA fingerprinting. However, many sand and water E. coli (often 30 to 40% of isolates) could not be reliably classified to either source, and so, these library-based source tracking results were treated more qualitatively than quantitatively. It is possible these unknown isolates were from urban wildlife sources not well represented in our library of fecal isolates.
Analyses for Bacteroidales DNA markers in water samples found that the generic BAC32 DNA marker was detected in about 98% of water samples analyzed across Toronto in 2007 (n = 434), suggesting little concern about inhibition of PCR assays. Edge et al. (2010) summarize the excellent host-specificity of the HF183 DNA marker assay, and more recent results in our lab have continued to validate its host-specificity, with only very rare detections in animal fecal samples. In 2007, the human HF183 DNA marker was only detected in two beach water samples (2%; n = 102) and one marsh stream sample (6%; n = 17) at Bluffer’s Park Beach. In contrast, the human HF183 DNA marker was detected in sixteen (70%; n = 23) of the stormwater outfall samples from the other side of Bluffer’s Park in 2007. These results suggest this outfall has a human sewage cross-connection problem, similar to what has been reported for other stormwater systems around the Great Lakes (Sauer et al. 2011; Staley et al. 2016). However, this outfall was considered unlikely to impact Bluffer’s Park Beach, similar to conclusions about the impacts of effluent from Dunkers stormwater storage facility also on the other side of the Park from the beach (TRCA, 2005).
Sanitary survey information and the results from human Bacteroidales DNA assays and E. coli library-based microbial source tracking assays all suggested human fecal contamination was not significantly impacting Bluffer’s Park Beach. The limited detection of human fecal impacts could have been the result of sporadic sources like diaper disposal or perhaps bather shedding. In contrast, bird fecal droppings were frequently observed to be direct sources of E. coli entry into Bluffer’s Park beach waters. Numerous studies have documented the significance of bird fecal droppings as fecal pollution sources at beaches around the Great Lakes (Edge and Hill, 2007; Edge et al. 2007b; Lu et al. 2011; Converse et al. 2012). In addition, streams draining from the marsh with E. coli concentrations exceeding 100,000 CFU 100 ml−1, and parking lot runoff were also observed to be direct sources of E. coli entry into Bluffer’s Park beach waters. Similar results pointing towards the importance of localized E. coli sources such as parking lot runoff have been identified at other Great Lakes beaches (McLellan and Salmore, 2003). E. coli library-based microbial source tracking suggested that E. coli from the marsh and beach waters were both more likely from non-human sources, consistent with numerous waterfowl and wildlife in the beach area. Beach observations, E. coli surveillance, and microbial source tracking results were all consistent in suggesting the importance of better managing bird impacts and runoff from the parking lot and marsh behind the beach.
In 2006, the City of Toronto began to implement a bird management program at Bluffer’s Park Beach (City of Toronto, 2006, 2007). This program was enhanced in 2007 and included use of trained dogs to deter waterfowl, expanding the Canadian Wildlife Service waterfowl transfer and egg oiling program, a public education and communication campaign advising park and beach goers against feeding birds, enhanced waste and recycling pickup and beachcombing operations, and enhanced beach grooming/tilling to remove bird droppings, food, and algae that might attract birds. Implementation of bird deterrence programs have been shown to lead to significant water quality improvements at Great Lakes beaches (Converse et al., 2012).
From the fall of 2007 into the spring of 2008, Toronto Water and the Toronto and Region Conservation Authority worked to re-engineer the wetland and dune system behind Bluffer’s Park Beach as part of a wetland improvement project (TRCA, 2016). The project deepened the existing marsh behind the beach to increase its water retention capacity, built a small dyke at the base of the bluffs to slow and control release of runoff, redirected parking lot runoff away from the beach into the marsh, directed water flows in the marsh towards the east end of the beach away from the swimming area, and created a dune system at the marsh edge to serve as a barrier to prevent water runoff onto the beach and direct stormwater overflows into infiltration basins. Community involvement assisted with planting dune and wetland areas and resulted in two hectares of restored wetland and 0.3 hectares of restored beach dunes within a three hectare wetland complex.
Through these remedial actions, water quality improved significantly at Bluffer’s Park Beach (Figure 3), and the beach was awarded an internationally recognized Blue Flag designation in 2011. Similar beach remedial actions to reduce impacts from birds and stormwater runoff in Racine, Wisconsin successfully increased beach use by residents and tourists, and expanded the role of the beachfront in the local economy (Kinzelman and McLellan, 2009). Increasingly, remediation efforts to reduce beach Beneficial Use Impairments in Areas of Concern around the Great Lakes need to address local impacts from waterbirds/waterfowl and stormwater runoff in addition to any legacy human sewage concerns related to combined sewer overflows and discharges from wastewater treatment plants.
Bluffer’s Park Beach in the Toronto and Region Area of Concern had a history of beach postings often exceeding 80% of the beach season since the 1980s. In 2005–2007, field studies found the highest E. coli concentrations near the beach in foreshore beach sand, and rain event runoff from a parking lot and marsh behind the beach. Microbial source tracking results were consistent with sanitary survey and field observations in identifying animal fecal contamination as a more significant source of E. coli contamination at the beach rather than human sewage contamination. It was concluded that the most important sources of E. coli at the beach were localized bird fecal droppings on the foreshore sand and runoff from the marsh and parking lot, which were observed to directly transport E. coli into beach waters. These results helped guide an adaptive management approach including a new bird and beach grooming management program and a marsh and dune restoration project by the City of Toronto and the Toronto and Region Conservation Authority in 2006–2008. These remedial actions significantly improved water quality at Bluffer’s Park Beach, and the Beach received an internationally recognized Blue Flag designation in 2011.
We would like to thank many people for their help with field and laboratory work including, Chad Boyko, Charlotte Curtis, Bailey Davis, Jason Demelo, Ian Droppo, Christine Jaskot, Alyssa Loughborough, Lauren McShane, Greg Meek, John Minor, Kinga Smolen, Gary Stinson, and Flora Suen. Assistance for the groundwater field work was provided by Environment and Climate Change Canada’s Technical Operations personnel. We thank the City of Toronto for their support of this research.