A unique science and management strategy has been developed for the Laurentian Great Lakes due to their enormous size, geographic-ecological diversity, political and economic importance. This article is a documentary of more than 40 years of research conducted at the base of the foodweb by Fisheries and Oceans Canada, which has contributed significantly to the management of the Great Lakes. In the 1960s, the governments of Canada and the United States responded to the threat of cultural eutrophication which eventually resulted in the signing of the binational Great Lakes Water Quality Agreement. Dr. R. A. Vollenweider and Dr. J. R. Vallentyne were instrumental in developing a phosphorus abatement program, as well as the adoption of the “ecosystem approach” resulting in an holistic and integrated protocol for managing multiple environmental stressors. By showcasing some selected examples (Lake Ontario, Bay of Quinte, current research activities), an attempt is made to chronicle the evolution of phytoplankton, primary productivity and microbial foodweb research in the Great Lakes. Some of the research programs, techniques, models, policies and international cooperation are highlighted, in addition to the strong European influences on Great Lakes research. The lessons learned from the long-term Great Lakes research experience could be extrapolated and applied to enhance understanding of the ecology and management of other large lake ecosystems throughout the world.
Concerns over the health of aquatic ecosystems reached a crescendo in the 1950s and 1960s as the public became aware of, and alarmed by, some of the more obvious signs of cultural eutrophication including algal blooms and declining fish catches. These problems were observed all over the world and such was the extent of the problem that the United Nations Organization for Economic Development and Cooperation (OECD) commissioned a study on cultural eutrophication in 1966. The final report (Vollenweider, 1968) provided the scientific impetus for governments, particularly in North America and Western Europe, to implement full scale research and monitoring programs in threatened watersheds. What follows is a narrative of two scientists – Dr. Richard A. Vollenweider and Dr. John R. Vallentyne – who were instrumental in dealing with issues of cultural eutrophication and adoption of the ecosystem approach in the Laurentian Great Lakes. Their scientific passion, expertise, dedication and personal interactions set in place a paradigm for managing aquatic environments that is still relevant today. This chronicle of the past 50 years of lower trophic level research in the Great Lakes begins with the impact of these two scientists who left their fingerprints in the past, present, and future of Great Lakes research. While this article will largely focus on their contributions to the Great Lakes, we will also highlight the lessons that can be applied globally.
Dr. Vollenweider, born and educated in Switzerland, had already begun to establish himself as one of the foremost authorities on nutrient dynamics of phytoplankton and lake trophic state when he was hired by the OECD in Paris to produce the report, Scientific Fundamentals of the eutrophication of lakes and flowing waters, with particular reference to nitrogen and phosphorus as factors in eutrophication (Vollenweider, 1968), that would become the catalyst for global action on eutrophication. During the same period, Dr. Vallentyne, who left academia in 1966 to become head of the Eutrophication section of the Fisheries Research Board of Canada (FRBC) at the Freshwater Institute in Winnipeg, was at the forefront of crafting the Canadian government’s response to cultural eutrophication. He believed that strong science should inform management and part of his efforts included establishing the Experimental Lakes Area in northeastern Ontario. He also helped to establish a research program for the Great Lakes based at Burlington, Ontario, now the Canada Centre for Inland Waters. Along with the help of colleagues such as Dr. W. Johnson (FRBC) and Mr. J. Bruce (Environment Canada), Dr. Vallentyne was able to recruit Dr. Vollenweider to take up residence at the new Great Lakes facility in Burlington beginning in 1968 (Bruce, 2011a). The synergistic interactions of Vallentyne and Vollenweider resulted in two fundamental outcomes for ecosystem health and management:
The North American Great Lakes, especially Lakes Erie and Ontario, would become the proving ground for efforts to alleviate the impacts of cultural eutrophication, and
Two sovereign countries, Canada and the United States, would agree by treaty to protect rather than exploit a shared resource (Great Lakes Water Quality Agreement (GLWQA), 2012).
The scientific and political influence of both these outstanding scientists cannot be overstated. Vollenweider’s eutrophication models – relating phosphorus loadings to primary production and algal standing crop – would be used to classify all of the Great Lakes according to trophic state (Vollenweider et al., 1974; Dave and Munawar, 2014). These were predictive models that could be used to estimate the magnitude of phosphorus load reductions needed to alleviate cultural eutrophication. Prior to their publication in 1974, Vollenweider’s models had already played a key role in efforts to clean up the Great Lakes as they formed the underpinnings of the phosphorus management provisions of the Great Lakes Water Quality Agreement (GLWQA). Signed in 1972 by the governments of Canada and the United States, the GLWQA committed both countries to reducing phosphorus discharges into the lakes.
In the public hearings leading up to the GLWQA, Dr. Vallentyne took on the role of advocating for phosphorus (P) abatement and defending the emerging scientific consensus that excess phosphates were responsible for algal blooms, against opposition from the detergent industry (Hamilton, 2011; Bruce, 2011b). He also advocated for adopting an “ecosystem approach” to managing environmental stressors, a term that would become enshrined in the revised GLWQA of 1978. The crux of the ecosystem approach was the recognition that humans both affected and were affected by the health of the biosphere in addition to recognizing that any system had an inherent carrying capacity (Great Lakes Research Advisory Board, 1978; Leach, 1978; Vallentyne, 1993).
The legacy of Vollenweider and Vallentyne was to provide a robust infrastructure for environmental protection that includes, above all: sound science, extensive research and monitoring programs, and an international policy framework. A special memorial issue of Aquatic Ecosystem Health and Management (Vol. 14, No. 2) was published in 2011 to honour and recognize the outstanding and long-lasting contributions of both these scientists. The current article will highlight two case studies of long-term research programs in the Great Lakes that show the evolution of this legacy from the early emphasis on eutrophication and algal blooms to a broader understanding of foodweb dynamics, perturbations and remediation. We will use these case studies to highlight the broader lessons that can be applied to the recovery of stressed ecosystems throughout the world.
The North American Great Lakes (Superior, Michigan, Huron, Erie and Ontario) cover approximately 250,000 km2 and account for 20% of the world’s surface freshwater (Table 1). Lake Superior is the largest (82,100 km2) with a maximum depth of 407 m and flushing time of 170 years. Home to 37 million people, the lakes are shared by Canada and the United States except for Lake Michigan, which falls exclusively under US jurisdiction. The Great Lakes were formed at the end of the last glacial period approximately 12,000 years ago. The southern portion of the watershed – including parts of Lakes Michigan and Huron in addition to Lakes Erie and Ontario as well as their connecting channels – is home to several large US and Canadian metropolitan areas including: Chicago, IL; Detroit, MI; Toledo, OH; Cleveland, OH; Erie, PA; Buffalo, NY; Hamilton, ON; Toronto, ON; and Rochester, NY (Environment Canada and United States Environmental Protection Agency, 1995). Consequently, the southern portion of the watershed has also been the most heavily impacted by multiple anthropogenic stressors such as: overexploitation of fisheries and habitat alterations; industrial pollution and toxic contamination; cultural eutrophication from municipal and agricultural sources; invasive species; and climate change. The upper Great Lakes (Lakes Superior, northern parts of Lakes Michigan and Huron) have been less impacted, with most stressors confined to nearshore areas; however long-term warming of the surface waters associated with climate change have been reported in Lake Superior (Austin and Colman, 2007). A general overview of major ecological events and perturbation available from the archives (since 1930) is given by Munawar et al. (2005). These records clearly demonstrate that the Great Lakes have been subjected to multiple anthropogenic stresses, as well as biological invasions, which drastically affected their health and integrity over the past several decades.
|.||Superior .||Michigan .||Huron .||Erie .||Ontario .|
|Geological Origin .||Glacial scour and tectonic .||Glacial .||Glacial .||Glacial .||Glacial .|
|Latitude||47°33′ N||44°00′ N||45°00′ N||42°09′ N||43°39′ N|
|Longitude||87°47′ W||87°00′ W||82°15′ W||81°15′ W||77°47′ W|
|Mean depth (m)||149||85||59||19||86|
|Maximum depth (m)||407||282||229||64||245|
|Flushing time (yr)||170||100||21||2.5||8|
|.||Superior .||Michigan .||Huron .||Erie .||Ontario .|
|Geological Origin .||Glacial scour and tectonic .||Glacial .||Glacial .||Glacial .||Glacial .|
|Latitude||47°33′ N||44°00′ N||45°00′ N||42°09′ N||43°39′ N|
|Longitude||87°47′ W||87°00′ W||82°15′ W||81°15′ W||77°47′ W|
|Mean depth (m)||149||85||59||19||86|
|Maximum depth (m)||407||282||229||64||245|
|Flushing time (yr)||170||100||21||2.5||8|
Limnological surveys of Lakes Erie and Ontario were organized in 1962 and 1965, respectively, as the Government of Canada began to assess the threat of eutrophication. The first comprehensive lake-wide plankton surveys were initiated and organized by Dr. Vollenweider during 1970 and included: primary productivity, phytoplankton, zooplankton, and benthos (Figure 1); in addition to a standard suite of physical – chemical parameters (nutrients, Secchi depths, chlorophyll a, etc.). These lake-wide surveys were conducted monthly, typically from April – December, although the first Lake Ontario cruise was year round (January – December). The lower Great Lakes, Ontario and Erie, were sampled during 1970, while the Upper Great Lakes surveys were conducted during 1971 (Lake Huron), 1973 (Lake Superior) and 1974 (Georgian Bay-North Channel). The results of these surveys would be incorporated into Vollenweider’s eutrophication models (Vollenweider et al., 1974). The enormous amount of new data generated regarding phytoplankton biomass, its composition, and primary productivity for all the North American Great Lakes was later published in the form of two peer-reviewed books (Munawar and Munawar, 1996, 2000).
The expanded scope of the Great Lakes surveillance program also reflected the beginning of a broader ecosystem approach. To wit, while Dr. Vollenweider was busy overseeing the first comprehensive bio-limnological surveys of the Great Lakes, Dr. Vallentyne was in turn establishing a policy for managing eutrophication and working closely with the Science Advisory Board of the International Joint Commission for developing a comprehensive management plan for the Great Lakes (Vallentyne and Munawar, 1993; Dave and Munawar, 2014).
Some of the highlights of the status of the lower lakes (Erie and Ontario) including management actions and plankton surveys are given below by date. The list is not exhaustive.
1960s: increasing eutrophication in Lake Ontario and Erie; nutrient surveillance begins.
1970s: continued eutrophication; Vollenweider’s lake-wide biological surveys commence; GLWQA signed; launching of the intensive Bay of Quinte monitoring program; P-abatement programs implemented.
1980s: Revised GLWQA implemented and ecosystem approach adopted; designation of Areas of Concern, P-abatement continues; Dreissenid Mussels appear in Lake Erie.
1990s: Remedial Action Plans (RAPs) implemented for Areas of Concern; Dreissena invasions continue; Lake Ontario and Lake Erie Trophic Transfer lake-wide surveys; increased water clarity and lower foodweb alterations observed in both lakes with invasive mussels and P-abatement identified as contributing factors.
2000s: Oligotrophication of Lake Ontario resulting in increased water clarity; Round Gobies (Neogobius melanostomus) and predatory Waterfleas (Cercopagis) invasions and disappearance of amphipods (Diporeia) and diatoms, large abundance of heterotrophic nanoflagellates; towards the end of the decade Lake Ontario is becoming mesotrophic with elevated chlorophyll a and phytoplankton biomass during summer; Harmful Algal Blooms re-emerge in Lake Erie related to ecosystem changes brought on by Dreissenid Mussels.
2010s: Prolonged algal bloom in Lake Erie (2011) reignites concerns over eutrophication; algal blooms continue to be observed in nearshore areas of Lake Ontario such as the Bay of Quinte and Hamilton Harbour.
The early phytoplankton surveys
The phytoplankton communities of all of the Great Lakes under Canadian jurisdiction were sampled from 1970 – 1974. Detailed observations of phytoplankton dynamics from this period are available in Munawar and Munawar (1996, 2000, 2001). Figure 2 depicts the seasonality of phytoplankton and its taxonomic composition and here we offer a brief summary for each ecosystem.
The lake did not show any seasonal trends and contained very low biomass dominated by phytoflagellates (Chrysophyceae, Cryptophyceae and Dinophyceae) as well as diatoms indicating ultra-oligotrophic conditions.
North Channel, Georgian Bay and Lake Huron
Low biomass with a single peak was observed during the summer in the North Channel which was dominated by diatoms. Georgian Bay exhibited a relatively higher biomass with peaks apparent in both the spring and late summer/fall respectively and diatoms were typically the most prevalent. Lake Huron had higher biomass concentrations including a pronounced spring peak and a pronounced early fall peak composed largely of diatoms.
The seasonal fluctuations of phytoplankton biomass in Lake Erie (lake-wide means) indicated a high concentrations during the spring composed of diatoms, dinoflagellates and chlorophytes. Maximum biomass was observed in the late summer/fall and included diatoms, dinoflagellates, chlorophytes (greens) and cyanophytes (blue-greens).
Only one clearly pronounced peak was observed in late summer and two smaller pulses were observed during the late winter and spring. Diatoms were abundant during the winter and spring periods. But more diversity was apparent from June onwards when phytoflagellates, greens and blue-greens contributed significantly to total biomass throughout the summer and fall periods.
These early plankton surveys also included size fractionated primary productivity experiments for all of the Great Lakes (except Michigan) being performed for the first time (Figure 3). These experiments were conducted during a summer cruise in 1973. Smaller sized plankton (<20 μm) made an overwhelming contribution to the photosynthetic rates in all the lakes. The size fractionation technique has since been refined considerably with improvements in filter design. Currently the following size classes are used: picoplankton (<2 μm), nanoplankton (2–20 μm) and net plankton (>20 μm). Extensive picoplankton surveys were conducted in the late 1980s which demonstrated the significance of small sized organisms (Munawar et al., 1987; Munawar and Weisse, 1989) in both the community structure and photosynthetic rates of the Great Lakes.
Long-term trends and multiple stressors
The lower Great Lakes, especially the western and central basins of Lake Erie in addition to the coastal areas of Lake Ontario, were the most heavily impacted by eutrophication. From the 1970s until the late 1990s, the reduction in the concentration of phytoplankton biomass and chlorophyll a in western and central Lake Erie has been attributed to both the impact of P-abatement and the filtering activities of invasive Zebra Mussels (Leach, 1993; Munawar and Munawar, 1999; Munawar et al., 1999). An alternate hypothesis proposed by Charlton et al. (1999) suggested that most of the water quality improvements took place before the advent of mussels. Figure 4a shows the long term changes in the mean concentration of chlorophyll a for the west basin which showed a significant decrease in 1992 (post P-abatement, post mussel invasion) compared to 1975 (pre P-abatement, pre mussel). From our data, it is impossible to separate the cumulative impacts of P-abatement and Dreissenid Mussel grazing. While the data from this period indicates that the lake was becoming less eutrophic, more recent information indicates that Lake Erie is once again becoming eutrophic (Kane et al., 2014).
Lake Ontario followed a similar trajectory to that of Lake Erie although the change in trophic state was more pronounced (from eutrophic to oligotrophic) due to the integrated impact of P-abatement and Dreissena filtering. Figure 4b depicts the long term lake-wide means of spring chlorophyll a concentrations in the nearshore region (<30 m) of Lake Ontario. High chlorophyll a values coincided with high TP concentrations (>20 μg l−1) in the 1970s (Figure 5) prior to P-abatement. The pigment decreased significantly indicating high water quality following the invasion of mussels in 1990 and at a time when TP concentrations were <10 μg l−1. Very low chlorophyll a concentrations were also recorded during summer and fall seasons (Munawar and Munawar, 2003). However, there is evidence that increased phosphorous loadings from tributaries are leading to increased chlorophyll a concentrations in localized nearshore areas of Lake Ontario (Makarewicz et al., 2012; Pavlac et al., 2012).
Lake Ontario case study
Long term research and monitoring of Lake Ontario: Spatially intensive surveillance
With a volume of 1651 km3, Lake Ontario is the world’s 11th largest lake and the second smallest of the Laurentian Great Lakes (Table 1). It is also the final lake in the watershed which drains into the Atlantic Ocean via the St. Lawrence River. The first lake-wide survey of the limnology of Lake Ontario was organized in 1965 as the governments of Canada and the United States began to recognize the threat of cultural eutrophication and the need for baseline measurements of nutrients, transparency (Secchi) and chlorophyll a. As discussed above, the first biological surveys of the lower trophic levels, including phytoplankton and zooplankton, were organized by Dr. Vollenweider in 1970. Phytoplankton biomass was observed to be quite high in the mid-summer period (≈9 g m−3) indicating persistent algal bloom conditions (Munawar and Naurwerck, 1971; Munawar and Munawar, 1982). Overall, the lake was determined to be mesotrophic according to Vollenweider’s eutrophication models (Vollenweider et al., 1974) with phosphorus loadings of 1 g m−2 y−1, primary production of 220 g m−2 y−1 and mean chlorophyll a of 5 μg l−1. These are the same models used by the governments of Canada and the United States to develop the P-abatement strategy for the Great Lakes formally recognized as Annex 7 of the Great Lakes Water Quality Agreement.
One of the legacies of Vollenweider’s tenure with the FRBC, and later Environment Canada, was the establishment of a regular monitoring program on the Great Lakes, although comprehensive biological studies were less frequent. The program was designed to be spatially intensive (i.e. lake-wide coverage) but limited in scope temporally (usually spring and summer cruises). The program was initially intended to assess the response of the lake to P-abatement. Two examples from the monitoring program include spring TP concentrations and spring chlorophyll a concentrations. Following the implementation of the P-abatement program in the Great Lakes, spring TP in Lake Ontario showed an immediate decline from 23.5 μg l−1 in 1977 to 16.8 μg l−1 in 1978 and has shown steady declines since then, levelling out at <8 μg l−1 beginning in 1998 (Figure 5; Dove, 2009). As noted in the previous section (Figure 4b), spring chlorophyll a concentrations were more variable over this time, but were observed to be as high as 12 μg l−1 in 1974 compared to 1 μg l−1 in 1996 and 1998 in response to the cumulative stresses of P-abatement and Dreissenid Mussel filtration (Mills et al., 2003; Munawar and Munawar, 2003). While much has been written about the impacts of these and other stressors on Lake Ontario, the broader point to be made is that the long-term surveillance program has provided researchers with the capacity to do so.
The surveillance program has created opportunities to intensively study the phytoplankton community of Lake Ontario. Reduced phosphorus loads were expected to have the greatest impact on the algal standing crop during the peak of the growing season in summer. Long term changes in summer phytoplankton biomass and composition are depicted in Table 2 at comparable stations distributed evenly across the lake. The highest observed biomass was in 1970 (6.2 g m−3) for which Chlorophyta (green algae) represented about half of the biomass. Following the implementation of nutrient controls in 1978, phytoplankton biomass declined to 1.1 g m−3 and contained a mixture of Cryptophyceae, Diatomeae, Chrysophyceae and Dinophyceae. Biomass declined even further between 1978 and 2003 – a period that included not only the continued impacts of P-abatement but also the establishment of Dreissenid Mussels – to an historic low of 0.2 g m−3and contained Cryptophyceae and Dinophyceae. During the lake-wide survey of 2008, biomass had increased to 3.0 g m−3 indicating a return to mesotrophic conditions and contained a mixture of diatoms, dinoflagellates and chrysophytes.
|.||1970 .||1978 .||2003 .||2008 .|
|.||1970 .||1978 .||2003 .||2008 .|
Taking a more holistic approach to the study of lower trophic levels, the first intensive survey of the microbial – planktonic foodweb of Lake Ontario was undertaken by Fisheries & Oceans Canada in collaboration with researchers from the University of Toronto in 1990 (Munawar and Weisse, 1989; Munawar et al., 2003). In addition to phytoplankton, these studies also included microbial loop (bacteria, autotrophic picoplankton and heterotrophic nanoflagellates [HNF]) and ciliate communities. This approach would be repeated for lake-wide biological surveys conducted in 2003, 2008 and 2013 and has allowed for the assessment of the organic carbon resources. Using the summer surveys as an example, we (Munawar et al., 2015) estimated the size of the organic carbon pool to be, on average, 520 mg C m−3 in 1990, 265 mg C m−3 in 2003 and 850 mg C m−3 in 2008 (the 2013 survey data are not yet available). In addition to fluctuations in the size of the organic carbon pool, the results show a remarkable shift in the structure from being primarily autotrophic (dominated by nanoplankton) in 1990 to being mainly heterotrophic in 2003 (dominated by nanoflagellates) and again autotrophic in 2008. With respect to the 2003 data set, we previously concluded that the lake was in poor health since HNF were sequestering autochthonous carbon and as a consequence higher trophic levels, which depended on this resource (e.g. zooplankton, planktivorous fish), were imperiled (Munawar et al., 2010). Other researchers have also found, at different times, the structure of the lower foodweb of Lake Ontario to be both heterotrophic (Pick and Caron, 1987) and autotrophic (Fahnenstiel et al., 1998). The longer term view would be that the lake appears to fluctuate between periods of heterotrophy and autotrophy and this has significant implications for ecosystem health assessments based primarily on short term synoptic surveys.
Regular monitoring efforts in Lake Ontario, and throughout the Great Lakes, have focused on basic physical and chemical measurements rather than intensive biological surveys. The result has been relatively infrequent assessments of the lower trophic levels. One major attempt at addressing this deficiency has been the Coordinated Science and Monitoring Initiative (CSMI) to better integrate Canadian and American research efforts. Under CSMI, each lake has been sampled intensively once every 5 years beginning in 2001. The 2003, 2008 and 2013 surveys of Lake Ontario fell under this banner. While this is a distinct improvement, our results show spring and summer surveys at 5 year intervals are not sufficient to capture the variability of the lower trophic levels (Munawar et al., 2014) and the dramatic shifts in structure of the organic carbon pool. More intensive research and monitoring of Lake Ontario at all trophic levels are needed in order to understand this complex and dynamic ecosystem.
The Bay of Quinte Case Study
Four decades of temporally intensive research and monitoring
One of the deficiencies of the Great Lakes Surveillance program has been that, due to the enormous size of the lakes, logistics prohibit frequent sampling. The trade-off is to undertake more frequent sampling over a much smaller geographic area, often a single index station, in order to gain a better understanding of the temporal dynamics (Millard et al., 1996). One such example of a temporally intensive research and monitoring program was developed at the Bay of Quinte, on the northeastern shore of Lake Ontario. The program, dubbed Project Quinte, began in 1972 and was aimed at addressing issues related to cultural eutrophication (Johnson and Hurley, 1986; Minns et al., 2011). The bay itself is a large ‘z’-shaped riverine system with 5 major tributaries that spans 254 km2 with a mean depth of 5 m although it is ≈30 m deep at the interface with Lake Ontario. The original program included regular monitoring of the lower trophic levels during the growing season (May – October) which meant that phytoplankton and zooplankton populations were assessed in addition to nutrient and chlorophyll a concentrations. Sampling was conducted on a weekly or bi-weekly basis beginning in 1972. The aims of the original program were first to understand the dynamics of a culturally eutrophic system and second to measure the response of the system to P-abatement. However these objectives were broadened over the years in response to an increasing awareness on the part of scientists, managers and the general public of the impacts of other anthropogenic stressors, leading to a changing policy framework.
Among these policy changes was a commitment by the governments of Canada and the United States to clean up severely degraded coastal areas in the Great Lakes basin under the terms of the revised Great Lakes Water Quality Agreement in 1987. In an effort to focus remediation activities, a total of 42 Areas of Concern (AoCs) were identified on both sides of the border for having at least 1 of 14 possible Beneficial Use Impairments (BUIs). Potential BUIs included “degradation of fish and wildlife populations,” “fish tumours or other deformities,” “eutrophication or undesirable algae” and “degradation of phytoplankton and zooplankton populations.” Despite being symptomatic in nature, the 14 BUIs are representative of environmental degradation at all trophic levels and are consistent with the principles of the “ecosystem approach” (Vallentyne and Beeton, 1988) adopted in the revised GLWQA. The Bay of Quinte was deemed to have 10 BUIs and thus designated an AoC which provided a new impetus for Project Quinte (Bay of Quinte Remedial Action Plan, 1993).
Cultural eutrophication was the original focus of Project Quinte and the intent was to study the response of phytoplankton and zooplankton communities to declining phosphorus loads. At the time of the bay’s designation as an AoC, it was deemed to have the related BUIs of “eutrophication or undesirable algae” and “degradation of phytoplankton and zooplankton populations” which had the effect of making Project Quinte a key aspect of the subsequent remediation efforts (Minns et al., 2011). In addition, the long term nature of the program meant that the cumulative impacts of other stressors, such as the colonization of the bay by Dreissenid Mussels in the mid-1990s, could also be studied. To our knowledge, this is the longest continuous running research and monitoring program in the Great Lakes and what we will show here is that the underlying structure of the program has proven robust enough to adapt to changing environmental stressors and management priorities.
Long term results from the Bay of Quinte monitoring program dating back to 1972 are summarized in Figure 6. Point source phosphorus loadings in the Bay of Quinte exceeded 200 kg d−1 in the early 1970s. At that time, total phosphorus concentrations averaged, for the May – October period, 70 μg l−1, chlorophyll a averaged 40 μg l−1 and total phytoplankton biomass averaged 12 g m−3. Beginning in 1978, point source phosphorus loadings were reduced to <70 kg d−1 as a result of improved sewage treatment and the current recommended cap is 15 kg d−1 (Minns et al., 1986; Minns and Moore, 2004). As part of remediation activities, targets were established for total phosphorus concentrations (30 μg l−1), phytoplankton biomass (4–6 g m−3), and chlorophyll a (12–15 μg l−1) which were based on mean concentrations over the growing season (i.e. May – October) (Bay of Quinte Remedial Action Plan, 1993). While these targets have been met intermittently since about 1996, our own assessment of the targets using a battery of tests approach suggests that these would not be sufficient to alleviate eutrophic conditions (Munawar et al., 2012). We have also observed that annual primary production in the Bay of Quinte exceeds 300 g C m−2 (Munawar et al., 2012) which is indicative of eutrophic conditions based on Vollenweider’s eutrophication models (Vollenweider et al., 1974). Results from the long term monitoring program has shown that despite the incredible progress made in reducing point source phosphorous loadings, non-point sources also contribute significantly to eutrophication in the bay (see Munawar et al., 2012, for a detailed discussion).
The scope of Project Quinte was expanded in 2000 to include the microbial loop so that the entire microbial – planktonic foodweb was studied. In addition to regular measurements of phytoplankton and zooplankton, bacteria, autotrophic picoplankton, heterotrophic nanoflagellates and ciliates were added to provide a more holistic assessment of the structure of the foodweb. This change was in keeping with the principles of the ‘ecosystem approach’, by responding to the growing awareness of the role of microbes in recycling nutrients and transferring energy to higher trophic levels (Pomeroy, 1974; Sherr et al., 1986). In eutrophic environments, there was also evidence that elevated inputs of phosphorus and allochthonous matter could enhance heterotrophic microbial processes (Pace, 1993; Porter, 1996). Over the course of 8 years (2000 – 2007), we examined the structure of the microbial foodweb, bi-weekly from May to October, following standard microscopic techniques (Munawar et al., 2011). On the basis of seasonal weighted means, we found that phytoplankton biomass ranged from 2.4 – 6.0 g m−3 typically 2–5 times greater than the zooplankton biomass (0.5 – 2.1 g m−3) when expressed as fresh weight. The microbial loop biomass ranged from 2.2 – 8.7 g m−3 and was up to 7 times greater than zooplankton biomass. Overwhelmingly composed of HNF, the microbial loop was equal to or exceeded the combined phytoplankton and zooplankton biomass in 3 of those years.
The relative importance of the microbial loop is even more pronounced in the organic carbon pool. Since organic carbon is essentially the currency of energy exchange, the size of the pool represents the amount of energy that is available for transfer to higher trophic levels and the structure of the pool affects how autochthonous energy is transferred. In the Bay of Quinte, we observed that the organic carbon pool was evenly split between autotrophs and heterotrophs during the first 3 years of our study; heterotrophs (primarily HNF) dominated for the next 2 years followed by an autotroph-dominated foodweb for the final 2 years of the study (Figure 7). The change in structure from heterotrophy to autotrophy was related to the appearance of algal blooms in the mid- and late summer. These results confirm that questions of trophic state are more than just the sum of autotrophic processes (Dodds and Cole, 2007) and that heterotrophic nanoflagellates have a significant, albeit not well understood, role in the transfer of energy from lower to higher trophic levels. The consistently high proportion of HNF to zooplankton observed in our study suggests that HNF may be outcompeting zooplankton for food and therefore limiting the food available to planktivorous and piscivorous fish (see Munawar et al., 2011, for a more detailed discussion).
With respect to the Bay of Quinte’s status as an Area of Concern, our microbial foodweb assessment does suggest an impaired foodweb consistent with the BUIs of “Eutrophication or undesirable algae” and “degradation of phytoplankton and zooplankton communities.” Our assessment of the Remedial Action Plan targets using a battery of tests strategy also tells a similar story. In the broader terms of ecosystem health, these results show that taking an ‘ecosystem approach’ is useful for both the diagnosis and remediation of impaired ecosystems around the world.
It would be useful and appropriate to review the lessons from past research in the Great Lakes. This could then be followed by considering the current and emerging research activities designed to enhance intensity, rapidity and precision of sampling of these large ecosystems.
European-North American collaboration
It had indeed been interesting and thought provoking to overview the development of Great Lakes science and some of the people who made it happen. Science does not develop in a vacuum and here we provide an account of the European - North American connections that have influenced Great Lakes science. The awareness of, and fight against, nutrient enrichment was undertaken more or less simultaneously in the Swedish large lakes (Willén, 2001) and the North American Great Lakes (Vollenweider et al., 1974) with P-abatement programs being implemented in the mid-1970s. Dave and Munawar (this issue) provide a detailed account of the trophic state of both ecosystems and their response to P-abatement and remediation programs. A similar situation existed for toxic chemicals such as metals, pulp and paper effluents, and other organic pollutants in both the Swedish and the North American Great Lakes. Two complimentary papers included in the special issue (Minns, 2014; Nalepa, 2014) provide detailed information about the other components of the foodweb such as fisheries and exotic species.
The European – North American connection is rather exciting with respect to large lakes research and eutrophication abatement (Minns, 2011; Leach and Mills, 2011). It probably originates with Professor Einer Steemann Nielsen, the Danish phytoplankton taxonomist turned experimental biologist who developed the 14Carbon method for measuring aquatic primary production that would be used for the first time on the Galathea expedition of 1950–1952 (Steeman Nielsen, 1951; Søndergaard, 2002). During the same period in Sweden, Dr. Wilhelm Rodhe had an active limnological research program on Lake Erken, near Uppsala. He was keen to undertake primary production experiments using the 14Carbon technique in Lake Erken and compare the results with various European lakes (AEHMS, 2011; Nauwerck, 2011). While conducting this research, he met Dr. Vollenweider, who was then working at the Instituto Italiano di Idrobiologia in Pallanza, Italy. Together they developed a research program fueled by their similar interests in the nutrient requirements of planktonic algae (Rodhe et al., 1958). This research would be instrumental for Vollenweider’s landmark report on cultural eutrophication to the Organization for Economic Cooperation and Development (Vollenweider, 1968). It would also bring him to the attention of Dr. Vallentyne, the Canadian scientist charged with crafting his government’s response to the issue of cultural eutrophication in the North American Great Lakes. Dr. Vollenweider deployed Steemann Nielsen’s 14Carbon technique in the Great Lakes to develop his eutrophication models that became the cornerstone of the Great Lakes Water Quality Agreement. Meanwhile, Dr. Vallentyne would succeed in having the same treaty adopt the ‘ecosystem approach’ for the management of environmental stressors.
Development of ‘Great Lakes’ science
Due to their enormous size, large watersheds and unique ecological diversity as well as political, economic and strategic importance, one can easily argue that the Great Lakes research community has developed its own ways and means to science. The large number of research institutions, representing government agencies and universities from both Canada and the United States, has given birth to a unique collection of tools, techniques, and approaches. Much of the credit resides with the foresight of Vollenweider and Vallentyne. Dr. Vollenweider designed and implemented a classical and extensive research and monitoring program that would result in the development of the mathematical models for implementing the P-abatement policy in the Great Lakes. Simultaneously, Dr. Vallentyne envisioned a broader paradigm called ‘the ecosystem approach’ that allowed two countries to manage environmental stressors in an integrated fashion that would be fair and equitable; protecting rather than exploiting a shared precious resource. Over four decades of research and monitoring has resulted in the systematic growth of a characteristic and formidable brand of science deserving to be coined ‘Great Lakes science’. Another excellent example of unique scientific restoration and remediation comes from the Bay of Quinte experience. From humble beginnings of preliminary summer-time nearshore sampling to year-round lake-wide pelagic surveys and temporally intensive long-term assessments, Great Lakes science has grown robust. The hypotheses, approaches, technologies and models deployed can also be applied globally to other large aquatic ecosystems. Some highlights from the Great Lakes experience are summarized below:
Lake-wide plankton and primary productivity surveys of the Great Lakes.
Long-term lake-wide surveillance of major nutrients and chlorophyll a.
P-abatement models and loading targets enshrined in law by two countries (Great Lakes Water Quality Agreement).
Adoption of the “Ecosystem Approach” to environmental management.
Identification of impaired Areas of Concern (AoCs), Beneficial Use Impairments (BUIs), and development of management plans for remediation and restoration.
Use of temporally-intensive long-term monitoring and research at selected index stations (e.g. Bay of Quinte).
Research into microbial foodweb and linkages between lower and higher trophic levels.
Invasive species research and management.
Current five-year cycle of lake-wide surveys under Coordinated Science and Monitoring Initiative (CSMI) which includes a cross section of researchers and managers from all jurisdictions.
Updating of Great Lakes Water Quality Agreement to emphasize sound science for dealing with emerging stressors, threats and risks, especially invasive species and climate change.
Deployment of new tools and techniques, especially in situ probes and sensors, and rigorous assessment to ensure consistency with long-term data.
Current research, issues and approaches
Resurgence of harmful algal blooms
Concerns that eutrophication of coastal areas and harmful algal blooms are increasing in frequency throughout the world have been raised by the scientific community (Rabalais et al., 2009; Paerl et al., 2011). While algal blooms have been a problem in the Great Lakes since at least the 1950s, the P-abatement strategies implemented in the 1970s were felt to have brought the problem under control (Burns, 1985) despite continued recurrences in specific nearshore areas and embayments (Dyble et al., 2008; Munawar et al., 2012). However, the resurgence of algal blooms over large portions of Lake Erie and embayments of Lake Ontario (Bay of Quinte and Hamilton Harbour) and the presence of microcystin have re-ignited concerns regarding eutrophication and ecosystem health (Kane et al., 2014; Michalak et al., 2013). Toxin producing algae can pose an immediate risk to public health and it is imperative that tools be available to rapidly assess this risk (de Figueiredo et al., 2004). However, more conservative techniques are also needed including comprehensive taxonomic analyses of the phytoplankton community in order to identify which species may produce toxins and what factors contribute to their growth. A true ‘ecosystem approach’ involves developing an understanding of how all components of the phytoplankton community interact before, during and after bloom formation and what factors lead to the production and release of algal toxins. We have begun conducting detailed taxonomic assessments of algal blooms as part of our ongoing research efforts in eutrophic areas, notably the Bay of Quinte and also Hamilton Harbour in Lake Ontario. Our lab has always placed great emphasis on phytoplankton taxonomy as a necessary and critical tool for assessing ecosystem health in the Great Lakes (Munawar and Munawar, 1996; 2000). In a recent paper (Munawar et al., 2014) we called for a major capacity building effort to expand the pool of qualified phytoplankton taxonomists. The increasing frequency of harmful algal blooms in the Great Lakes and throughout the world makes that call more urgent.
Emerging tools and techniques
Recent improvements in technologies used to measure limnological parameters in situ have allowed for an increased number of measures over space and time. Sensors with automated data loggers can be deployed in a lake for months at a time to capture diurnal and seasonal trends in parameters such as oxygen, temperature, depth, pH, conductivity, and light among others. The increasing prevalence of optical technologies in these sensors means that calibrations can be maintained for long periods of time. Similarly, profilers equipped with these new generation sensors can also be used to obtain data in situ rather than collecting water samples for analyses in vitro. In 2013, DFO was part of the binational Lake Ontario 2013 Coordinated Science and Monitoring Initiative that employed some of these new and developing technologies in an attempt to sample the lake ecosystem intensively, and ultimately generate a synthesis for the entire lake.
One objective of the project was to estimate primary productivity on a lake-wide scale. Traditional methods incubate samples in vitro using radioactive carbon as a tracer. Out of necessity, these samples are collected at a specific place and time and results are extrapolated over much wider temporal and spatial scales. This can result in a high level of uncertainty when estimating annual lakewide primary production. The time and effort necessary to collect the samples in vitro and conduct the incubation restricts the number of samples that can feasibly be run. In the 2013 study, in vitro samples were collected biweekly at only 3 stations and twice across the lake at 15 stations (in April and August). In an attempt to reduce the level of uncertainty in primary production estimates researchers from DFO, Ontario Ministry of Natural Resources and United States Geological Survey deployed in situ oxygen-temperature loggers at 13 stations across Lake Ontario at a number of depths to model community production using the “diel” or “free” oxygen method (Staehr et al., 2010). Solomon et al. (2013) uses changes in oxygen levels at a scale of minutes to estimate daily net production, driven by light attenuation, incident light levels, wind speeds and thermocline depth. A comparison of radioisotope and free oxygen techniques is a primary data synthesis objective for the Lake Ontario CSMI group.
Another technology employed in the 2013 Lake Ontario study was the use of a towed array of instruments to create 2D transects of the water column. An active towbody was outfitted with sensors for temperature, conductivity, pH, oxygen, and fluorometers for chlorophyll a, phycocyanin and CDOM coupled to a Laser Optical Plankton Counter which measures the size and density of particles in the water. These particles can be matched with zooplankton taxa based on size and comparison with calibration samples which have been identified using microscopic means. The towbody can be deployed at a set depth to investigate a specific depth layer or towed in a sinusoidal curve to give a ‘tow-yo’ of a specific transect. Mapping mid-summer results indicated a fully developed thermocline at 11 m with a deep chlorophyll layer observed around 15 m. Particle sizes were differentiated by depth, indicating some depth preferences by specific groups of zooplankton and intensified patchiness as particle size increased.
Discrete depth water samples were collected to determine phytoplankton biomass microscopically. Our results (Figure 8) showed that the highest phytoplankton biomass did not correspond with the highest levels of chlorophyll a, as seen by the towed array. A large diatom assemblage was recorded below the metalimnion which was underestimated by fluorometry. Since diatoms are considered an important food for zooplankton, the use of a fluorometer in foodweb studies may be misleading if the data is not verified with standard microscopic data. It is important to consider the application of new technologies in order to improve the efficiency and effectiveness of sampling programs and it is equally important to thoroughly assess the accuracy of their output.
In situ fluorometry versus microscopy
The need for rapid assessment of the algal standing crop has resulted in the development of a variety of in situ fluorometers including a multi-spectra device capable of identiflying pigment-based algal classes in addition to chlorophyll a (Beutler et al., 2002). This fluorometer has been deployed extensively in Europe and North America (Gregor and Maršálek, 2004; Gregor et al., 2005; Ghadouani and Smith, 2005; Twiss and Macleod, 2008). However, there is a great need to compare results with more traditional taxonomic assessments (microscopy).
There are two fundamental problems inherent in such a comparison. The first is that, as an indicator of algal standing crop, chlorophyll a is not equivalent to estimates of phytoplankton biomass (or biovolume) that a taxonomist would generate. The second is that the fluorometer is limited by design to identifying 4 algal classes by pigment, whereas 6 taxonomic groups (containing hundreds of species) have been commonly identified in the Great Lakes (Munawar and Munawar, 1996, 2000). Given these limitations, the only reasonable approach is to compare results in the broadest possible context. Our expectation is that the spectral group identified as “Greens” should correspond to Chlorophyta; “Bluegreens” should represent Cyanophyta; “Browns” would include Chrysophyceae, Diatomeae and Dinophyceae, and “Mixed” be indicative of Cryptophyceae based on the descriptions provided by Beutler et al. (2002).
In this example, we compared fluorometric and microscopic techniques in water sampled from oligotrophic Lake Superior during August, 2011 at depths of 3 m (epilimnion), 8 m (metalimnion), 13 m (deep chlorophyll layer) and 35 m (hypolimnion). The fluorometer revealed that the water column of Lake Superior is dominated by “Greens” which accounted for 55–70% of the algal standing crop (chlorophyll a) and the largest proportions were observed in the upper strata. The fluorometer also found that “Browns” form a significant proportion of the standing crop, 30–45%, with no other pigment groups being detected. The microscopic assessment revealed a different story. The combination of Chrysophyceae, Diatomeae, and Dinophyceae together (i.e. “Browns”), accounted for 20–50% of the total phytoplankton biomass and was reasonably close to the fluorometric readings. However, Chlorophyta (“Greens”) represented only 15–35% of the biomass. Cyanophyta composed 10–50% of the biomass yet the fluorometer did not detect any “Bluegreens.” Similarly, Cryptophyceae (“Mixed”) composed 10–15% of the biomass in sub-epilimnetic strata but were also not detected by the fluorometer.
Our preliminary comparison of fluorometry and taxonomic evaluation indicated that pigment based classifications can be misleading and not representative of the actual composition of the phytoplankton community. We recognize that many species and taxonomic groups contain similar pigments and this helps explain the disparity in results. We would also point out that this is precisely the problem. Rapid in situ assessments of the algal standing crop are of no value if they are not accurate. Microscopic techniques can be used to provide thorough information on phytoplankton biomass, size and species abundance and most importantly taxonomic composition but require a highly trained taxonomist. A commitment to improving taxonomic training is the only way to ensure holistic, accurate assessments of the phytoplankton biodiversity and the status of ecosystem health (Wilson, 2013; Munawar et al., 2014).
The need for rapid assessment of the algal standing crop has resulted in the development of a variety of in situ fluorometers. One such device, described by Beutler et al. (2002), is based on in vivo fluorescence of photosynthetic pigments. This fluorometer has been the focus of many studies comparing the in vivo results to the results of in vitro chlorophyll a extraction techniques, in Europe (Beutler et al., 2002; Gregor and Maršálek, 2004; Gregor et al., 2005) and North America (Ghadouani and Smith, 2005; Twiss and Macleod, 2008). However, a complete assessment of the algal standing crop requires microscopic assessments of taxonomy and biomass by a trained taxonomist and there is a need to compare the results of the in situ fluorometer to those obtained through standard microscopy.
Chlorophyll a is one pigment, among many, that composes only a very small proportion of the biovolume of an individual cell and that proportion is highly variable among cells, species and environments. We would generally not expect to find any meaningful correlation between the two despite the widespread assumption that such a relationship exists.
This article commenced by showcasing the contributions of Vollenweider and Vallentyne to Great Lakes science. And here we reiterate that the influence of these two scientists extends far beyond eutrophication models and the ecosystem approach. What they helped to create was an institutional and policy framework that led to robust research and monitoring programs capable of responding on a large scale to emerging risks and threats while dealing with existing stressors. The fact that the policy framework includes sovereign countries agreeing by international treaty to protect a shared resource cannot be understated. At the same time, Vollenweider and Vallentyne were part of a global confluence of events that included: widespread public concern over eutrophication; a new methodology to accurately measure algal growth rates; recognition that phosphorus was a limiting nutrient for algae; recognition that other anthropogenic stressors were simultaneously affecting the health of large lakes, and that a concerted public policy effort, based on sound science, was needed to address these threats. The legacy left behind by them still echoes, not only in all corners of the Laurentian Great Lakes, but across the world.
Looking in to the future, Great Lakes science presents an excellent example of international collaboration between Canada, United States and Europe in developing programs for the conservation of ecosystem health and policies for the management of precious aquatic resources. The future is promising since emerging tools, techniques and models appear not only to complement the traditional limnology/oceanography but also enhance the intensity, rapidity and precision of field and laboratory research, which will certainly improve our understanding of the Great Lakes.
This review is dedicated to the memories of Dr. Richard Vollenweider and Dr. John R. (Jack) Vallentyne.
The long term inventory of Great Lakes research resulted from the contributions of numerous colleagues, students and contractors from both Canada and the United States during a period of more than four decades. We thank all these people who contributed to various aspects of microbial-plankton research. The Munawars are grateful to their teachers: R. Vollenweider, J. Vallentyne, J. Verduin, M. R. Suxena and A. R. Zafar, and to senior scientists: H. Regier, J. Talling, C. Reynolds, K. Minns, N. Mandrak, E. Mills, J. Leach, R. Heath, T. Nalepa, G. Dave, M. Van der Knaap for their guidance, advice and friendship. The support of the Munawars’ daughters (Saberina and Nabila) and son (Ahmad), who often pleasantly tolerated Dr. Munawar’s long absences from home due to field work or official travel, is gratefully acknowledged.