Causal mechanisms of the deep chlorophyll maximum in Lake Superior: A numerical modeling investigation

The deep chlorophyll maximum (DCM) is a near ubiquitous feature in Lake Superior during the summer stratified season. Previous studies have elucidated observable characteristics of the DCM in Lake Superior but the physical and biological mechanisms controlling the creation and maintenance of the DCM remained unclear. We use a three-dimensional hydrodynamic model of Lake Superior coupled to an ecological model to perform sensitivity runs exploring the influence of photoadaptation, photoinhibition, zooplankton grazing, and phytoplankton sinking on the vertical distribution of chlorophyll in the water column. The role of a nutricline in determining the presence and nature of the DCM is also explored. The presence of the DCM is dependent upon the presence of thermal stratification in the model. The sensitivity runs reveal that photoadaptation plays a primary role in determining the depth of the DCM in the model while zooplankton grazing and phytoplankton sinking affected the magnitude but not the presence or depth of the DCM. Photoinhibition showed negligible effects on chlorophyll concentration distribution. The presence of a nutricline in the model is also a necessary condition for the formation of the DCM and it influences both the depth and magnitude of the DCM.     

Seasonality and physical drivers of deep chlorophyll layers in Lake Superior,with implications for a rapidly warming lake

A deep chlorophyll layer (DCL) is a common feature of many deep, oligotrophic lakes including Lake Superior. Mechanisms generating and maintaining DCLs are variable across lakes, and seasonal patterns and relationships of DCL structure to physical variables are not well described. Using vertical profile data for physical and biological variables from western and central Lake Superior, we described seasonal patterns in DCL structure and other physical and biological parameters and applied linear mixed-effects models to determine how different physical factors (surface temperature, thermocline depth, and 1% photosynthetically active radiation (PAR) depth) affect the depth, thickness, maximum concentration, and integrated chlorophyll of the DCL. We observed clear seasonal patterns in the development and degradation of the DCL that coincide with seasonal changes in light and temperature. Modeling analysis using linear mixed-effects models showed that the DCL thickness was best predicted by surface temperature (R² = 0.51) followed by thermocline depth (R² = 0.36), and the deep chlorophyll maximum (DCM) concentration was best predicted by surface temperature (R² = 0.26). The 1% PAR depth was not implicated as an important predictor, but observations from seasonal data suggest that it plays a role in the depth of the DCM. While no relationship was found between surface temperature and DCL-integrated chlorophyll, DCL thickness decreased and DCM concentration increased with increasing surface temperature, which could have implications for productivity in the DCL as the lake continues to warm.     

Recent change in summer chlorophyll a dynamics of southeastern Lake Michigan

Six offshore stations in southeastern Lake Michigan were sampled during a pre quagga mussel Dreissena rostriformis bugensis period (1995–2000) and a post quagga mussel period (2007–2011). Chlorophyll a fluorescence profiles were used to characterize chlorophyll a concentrations during early (June–July) and late (August–September) summer stratification. During the early summer period the average whole water column chlorophyll a, the deep chlorophyll maximum, and the size of deep chlorophyll layer decreased 50%, 55%, and 92%, respectively, between 1995–2000 and 2007–2011. By contrast, in late summer there were no changes in these metrics between periods. Surface mixed layer chlorophyll a in early and late summer did not differ between time periods. On the other hand, chlorophyll a in the near bottom zone (bottom 20 m) declined 63% and 54% between 1995–2000 and 2007–2011 in early and late summer respectively. Changes in total phosphorus between 1995–2000 and 2007–2011 were less dramatic, with declines of 22–27% in early summer and 11–30% in late summer. Changes in the chlorophyll a conditions were attributed to dreissenid mussels which reduced material available from the spring bloom and disrupted the horizontal transport of nutrients to the offshore. Although light availability increased (i.e., increased secchi depths), reduced nutrient availability and spring diatom abundance resulted in a much smaller deep chlorophyll layer in 2007–2011.     

Spatial,seasonal, and historical variation of phytoplankton production in Lake Michigan

There have been few direct measurements of phytoplankton production made in Lake Michigan since invasive dreissenid mussels became established in the lake. Here we report the results of 64 measurements of phytoplankton primary production made in Lake Michigan during 2016 and 2017. We conducted two lake-wide surveys, one in the spring 2016 isothermal period and one after summer stratification in 2017 and examined seasonal production with bi-weekly sampling between May and November 2017 at an offshore station in the southwestern part of the lake. We assessed nearshore-offshore gradients by sampling at three transect locations on three occasions in 2017. Spring 2016 production and production:biomass (P:B) ratios (reflective of growth rates) were similar across the lake and were higher than those reported before dreissenid mussels became established, suggesting that despite decreases in phytoplankton biomass, growth rates remain high. Summer 2017 production and growth rates increased from south to north. Areal production in 2017 peaked in late summer. Mean 2017 summer production (499 ± 129 mg C m^?2 day^?1) was lower than values reported prior to the mussel invasion, and the fraction of total production occurring in the deep chlorophyll layer was about half that measured pre-mussels. At the offshore site picoplankton accounted for almost 50 % of the chlorophyll. As spring P:B ratios have increased and summer P:B and seston carbon:phosphorus ratios have not changed, we conclude that the decrease in phytoplankton production in Lake Michigan is due primarily to grazing by mussels rather than to stronger nutrient limitation.     

Community structure,abundance variation and population trends of waterbirds in relation to water level fluctuation in Poyang Lake

Poyang Lake is China's largest freshwater lake, and it has been an internationally important wintering ground for migratory waterbirds. Based on waterbird censuses from 2001 to 2016, community structure and abundance trends of waterbirds were analyzed, as well as the potential correlations with the water level of Poyang Lake. The results showed that the annual average number of waterbirds in Poyang Lake was 426,707 ± 150,170, with 111 species from 17 families. Waterfowl was the most abundant family accounting for 74.4 ± 8.8% of all individuals, followed by shorebirds (14.8 ± 8.5%), wading birds (6.5 ± 1.8%) and open-water/waterbirds (4.4 ± 1.9%). Although waterbird abundance fluctuated dramatically, there were no significant trends in the abundance of most guilds or in total waterbird abundance; only geese significantly increased among the eight groups. Analysis of trends of 37 relatively abundant or regularly occurring species indicated that population trends appeared to be species-specific. As for the correlations between water level and waterbird abundance, only shorebirds showed significant correlations with average July water level, average water level, and high water level duration in wet season among four guilds, i.e. waterfowl, wading bird, shorebird and, open-water and waterbird. At the group level, abundances of swans, geese, and ducks were significantly correlated with monthly average water level during the wet season, and wader abundance was significantly correlated with average water levels and high water level duration during the wet season. Correlations between population abundance and monthly water level also exhibited species-specific patterns.     

First record of the non-indigenous parasitic copepod Neoergasilus japonicus (Harada, 1930) in the Lake Ontario Watershed: Oneida Lake,New York

Four specimens of the Asiatic parasitic copepod Neoergasilus japonicus (Harada, 1930) were collected from Oneida Lake, New York in September 2018; one specimen was from a white sucker Catostomus commersonii, another from a green sunfish Lepomis cyanellus, and two from a bluegill Lepomis macrochirus. The four adult female specimens were found attached to the base of the gills of their respective hosts along with other ergasilid species. The average total length of the adult female N. japonicus specimens we found was 0.609 mm. These detections represent the first known occurrence of this non-native species in the state of New York, extends the easternmost distribution of this parasite over 400 miles, and now includes the Lake Ontario watershed for the first time. It is commonly believed that the international aquaculture industry and aquarium fish trade are the most likely vectors of dispersal for N. japonicus. Monitoring the spread of non-indigenous aquatic species is an important step towards the development of management plans and mitigation efforts with regards to the anthropogenic causes of dispersal, and fish parasites are no exception.     

Modeling historical trends in Lake Superior total nitrogen concentrations

Nitrate concentrations in Lake Superior increased fivefold between 1900 and 1980, and have remained nearly constant since that time. Such rapid changes in concentration in a lake with a long hydraulic residence time (~ 190 years) are surprising. We developed a model to better understand the causes of the historical changes and to predict future changes in nitrate concentrations. Historical loadings were reconstructed based on average national NO_x emissions estimates, recent (past ~ 30 years) atmospheric N deposition data, recent tributary concentration data, and basin-wide runoff estimates. Increases in atmospheric N deposition alone were insufficient to have resulted in the observed trends. However, model runs combining increased atmospheric deposition with increased tributary N loading and/or decreased burial + denitrification mid-century reproduced the observed accumulation of N. Because internal N fluxes are an order of magnitude greater than external fluxes, relatively small changes in the lake's internal N cycle may produce relatively large changes in total N concentrations. Land-use changes in the watershed, particularly increases in logging activity, may have altered riverine N inputs. Regardless of the historical mechanisms leading to the rise in nitrate concentrations, it appears as though the system is currently at or is approaching peak N content.     

Implications of increasing pollution levels on commercially important fishes in Lake Victoria

Lake Victoria receives huge quantities of effluent from domestic, agricultural and industrial sources. We used fish condition factor (K), vitellogenin (VTG) production and liver lesions as biomarkers to assess pollution levels in the lake. We tested the hypothesis that pollution levels do not affect the selected biomarkers. Beach seine and cast nets were used to collect Oreochromis niloticus (n = 230), Lates niloticus (n = 99) and Protopterus aethiopicus (n = 37) in areas presumed to be less or more polluted, both inshore and offshore. K was lower in more polluted compared to less polluted areas of the lake. VTG production was high in both less and more polluted areas for O. niloticus (0.77 ± 0.08 µg/L), L. niloticus (0.73 ± 0.09 µg/L) and P. aethiopicus (0.55 ± 0.06 µg/L). Liver tissue showed lesions such as vacuolations, cellular degeneration, sinusoidal dilation, focal necrosis, increased Küpffer cells and congestion of sinusoids. The prevalence of liver tissue alteration showed normal lesion (19.9%, n = 73), slight (8.2%, n = 30), moderate (41.5%, n = 152), severe (18.6%, n = 68) alterations and irreparable damage (11.8%, n = 43). Severe liver alterations in O. niloticus, L. niloticus and P. aethiopicus were higher in more polluted compared to less polluted areas. Chemical contamination of Lake Victoria caused liver lesions and other changes in fishes, possibly leading to adverse effects on the lake’s fisheries resources. Overtime, such chemical contamination could lead to negative impacts on the consumers of fish if actions are not taken to mitigate the risks.     

Fate of phosphorus from a point source in the Lake Michigan nearshore zone

Nutrient loading into Lake Michigan can produce algal blooms which in turn can lead to hypoxia, beach closures, clogging of water intakes, and reduced water quality. The Great Lakes Water Quality Agreement targets for Lake Michigan are 5600 MT annually for total phosphorus (TP) loading, 7 μg L^?1 lake-wide mean TP concentration, and a chlorophyll-a concentration of 1.8 μg L^?1. However, in light of the recent resurgence of nuisance algal (Cladophora sp.) growth in the nearshore zone, the validity of these targets is now uncertain. The occurrence and abundance of Cladophora in the nearshore area depends primarily on the availability of dissolved phosphorus, light, and temperature. The availability of dissolved phosphorus is a potentially useful indicator of nearshore areas susceptible to excessive Cladophora growth and impaired water quality. Regulating agencies are looking for guidance in determining phosphorus loading rates that minimize local exceedance of the lake target concentration. In this study, the lake assimilative capacity was quantified by applying a biophysical model to estimate the area required for mixing and diluting wastewater treatment plant outfall TP loadings to the level of the lake target concentration during the Cladophora growing season. Model results compared well with empirical measurements of particulate and dissolved phosphorus as well as Cladophora biomass and phosphorus content. The model was applied to test scenarios of wastewater treatment plant phosphorus loading in two different years, in order to help establish phosphorus discharge limits for the plant.     

Dietary differences among commercially important fishes in Lake Tanganyika assessed using stable isotope analysis

The pelagic fishery of Lake Tanganyika, which is largely made up of the three species Lates stappersii, Limnothrissa miodon, and Stolothrissa tanganicae, has been in decline, and there is no clear understanding of the primary underlying causes. It has been suggested that climate change has altered the primary productivity of the system, but detailed knowledge of the system's food web is required to elucidate the effect on higher trophic levels. The aim of this study is to determine the diet of the three commercially important fish species. Muscle tissue samples for stable isotope analysis were taken from February through April 2017, supplemented with stomach samples from L. stappersii for use in stomach content analysis. The stomach analysis showed an ontogenetic change in the diet composition of L. stappersii, shifting from copepods to fish larvae, supplemented with shrimp, to whole fish prey as the fish grew larger. Stable carbon and nitrogen isotope values from fish muscle were consistent with this observation, and also seemed to indicate that fish of similar sizes had similar diets, irrespective of species. This suggests that the diet of the pelagic fish species might be better explained by size, rather than species. The isotope data revealed a short range of δ¹⁵N values from primary consumers to fish, which may help explain the high fisheries productivity of Lake Tanganyika, and the strong impact of primary productivity changes on fisheries yield.     

A spectral space partition guided ensemble method for retrieving chlorophyll-a concentration in inland waters from Sentinel-2A satellite imagery

This research proposes an ensemble method for synergistically combining multiple empirical algorithms to better estimate chlorophyll-a (Chl-a) concentration. In previous studies, different empirical algorithms have been employed separately and a single algorithm was often identified as the most suitable predictor for Chl-a retrieval. Our ensemble method combines different individual algorithms to form an ensemble predictor that exploits advantages of each individual algorithm to maximize the overall estimation accuracy. We evaluated two ensemble predictors: the optimally weighted ensemble predictor and the spectral space partition guided ensemble predictor. The ensemble method has been successfully applied to a Sentinel-2A multispectral image acquired over Harsha Lake, Ohio in 2016. Based on in situ water reference data and satellite imagery, we constructed two ensemble predictors that consist of three individual empirical algorithms/estimators, including 2BDA (two-band algorithm), 3BDA (three-band algorithm), and NDCI (Normalized Difference Chlorophyll Index). For the optimally weighted ensemble predictor, the weights for individual algorithms are computed by solving an overdetermined linear system with the pseudoinverse technique. For the spectral space partition guided ensemble predictor, the rules for partitioning spectral space into spectral regions were established as a decision-tree using the CART method. The optimal Chl-a estimate for a pixel is obtained by selectively using the empirical algorithm in the ensemble that has the highest expected accuracy in the spectral region where the pixel is located. Our assessments suggest that the spectral space partition guided ensemble method performs significantly better than three individual empirical algorithms and also better than the optimally weighted ensemble method.     

Phosphorus and nitrogen loading to Lake Huron from septic systems at Grand Bend,ON

Groundwater nutrient loading to L. Huron was assessed along a 1.7 km section of beach at Grand Bend, ON, Canada, where septic systems are used for wastewater disposal. The artificial sweetener acesulfame (ACE) was detected in all groundwater samples (7–842 ng/L, n = 78), revealing that the entire surficial aquifer was impacted by septic system wastewater. Nitrate concentrations (3.5 ± 1.4 mg/L, n = 78) were correlated with ACE (r² = 0.54), indicating that septic systems contribute to nitrate loading in the aquifer. Chloride was also elevated (37 ± 11 mg/L, n = 78), but was not correlated with ACE (r² = 0.008), indicating a non wastewater source was dominant, likely road salt. Soluble reactive phosphorus (SRP) values were low (5.3 ± 9.3 μg/L, n = 77) and were not correlated with ACE (r² = 0.006). Sediment profiling below two of the septic system drain-fields, showed that the sand grains had distinct secondary coatings containing P, indicating that mineral precipitation reactions played a role in limiting P concentrations present in the aquifer. Groundwater nutrient loading to the lake was estimated at 13,000 kg N/year from NO₃^? and 1.9 kg P/year from SRP. These amounts are insignificant compared to nutrient loading from a stream that drains an agricultural catchment and discharges to the lake at the north end of the study site (Parkhill Creek). This calls into question, in some cases, the rationale of decommissioning properly functioning septic systems as a mitigation measure for reducing nutrient loading to nearby water courses.     

Genetic estimates of jurisdictional and strain contributions to the northeastern Lake Michigan brown trout sportfishing harvest

The Lake Michigan brown trout (Salmo trutta) fishery is sustained by the stocking of five hatchery strains by four state natural resource agencies. In the absence of exhaustive marking programs, strain-specific measures of stocking success are lacking for brown trout in Lake Michigan. We used microsatellite-based genetic assignment testing and genetic stock identification (GSI) to determine the strain of 122 angler-caught brown trout from four northeastern Lake Michigan ports. We compared strain composition estimates for sportfishing harvest to expected proportions of each brown trout strain in Lake Michigan at the time of harvest using stocking records corrected for age-specific mortality rates. Reassignment rates of individuals from baseline strains averaged 92.1% (range: 84.1–98.0%). Assignment testing and GSI analyses consistently found Wild Rose strain brown trout represented approximately 89% of the northeastern Lake Michigan sportfishing harvest, while only comprising 43.8% of the expected stock. Of the Michigan angler harvest of Wild Rose strain brown trout, approximately half were estimated to have originated from Wisconsin hatcheries, demonstrating a propensity for lake-wide movements. Continued assessments will improve understanding of strain relative contributions to angler harvests that can direct future stocking efforts.     

Invasive cattail reduces fish diversity and abundance in the emergent marsh of a Great Lakes coastal wetland

Great Lakes coastal wetlands (GLCWs) provide critical fish habitat. The invasion of GLCWs by hybrid and narrow-leaved cattail, Typha × glauca and Typha angustifolia (hereafter Typha), homogenizes wetlands by out-competing native plant species and producing copious litter. However, the effect of this invasion on fish communities is little known. To measure the effect of Typha on fishes, we established plots in Typha invaded and native wetland emergent zones in a northern Lake Michigan coastal wetland, and measured environmental variables, plants, and fishes in each zone over two summers. Dissolved oxygen and water temperature were significantly lower in invaded compared to native plots. Invaded plots were dominated by Typha and its litter; whereas. sedges (Carex spp.) were the most abundant species in native plots. Fish abundance and species richness were significantly lower in Typha compared to native wetland plots. The Typha fish community was dominated by hypoxia tolerant mudminnow whereas other small, schooling, fusiform species such as cyprinids and fundulids were absent. These results illustrate the negative impact of a dominant invasive plant on Great Lakes fishes that is expected to be found in Typha invasions in other GLCWs.     

Telemetry reveals limited exchange of walleye between Lake Erie and Lake Huron: Movement of two populations through the Huron-Erie corridor

Immigration and emigration of individuals among populations influence population dynamics and are important considerations for managing exploited populations. Lake Huron and Lake Erie walleye (Sander vitreus) populations are managed separately although the interconnecting Huron-Erie Corridor provides an unimpeded passageway. Acoustic telemetry was used to estimate inter-lake exchange and movement within St. Clair River and Detroit River. Of 492 adult walleyes tagged and released during 2011 and 2012, one fish from Tittabawassee River (Lake Huron; 1 of 259, 0.39%) and one individual from Maumee River (Lake Erie; 1 of 233, 0.43%) exchanged lakes during 2011–2014. However, both fish returned to the lake where tagged prior to the next spawning season. The one walleye from Maumee River that moved to Lake Huron made repeated round-trips between Lake Erie and Lake Huron during three consecutive years. Of twelve fish tagged in the Tittabawassee River detected in the Huron-Erie Corridor, few (n = 3) moved south of Lake St. Clair to the Detroit River. Ten walleye tagged in the Maumee River entered the Huron-Erie Corridor, and five were detected in the St. Clair River. Our hypothesis that walleye spawning in Maumee River, Lake Erie, served as a source population to Lake Huron (“sink population”) was not supported by our results. Emigration of walleye to Lake Huron from other populations than the Maumee River, such as those that spawn on in-lake reefs, or from Lake St. Clair may contribute to Lake Huron walleye populations.     

The Composition of Particulate Organic Matter in the Euphotic Zone of Lake Superior

The composition of particulate organic matter (POM) as living and dead material in the euphotic zone of Lake Superior, May to November 1973, was derived by a regression method. The regression method reported in the literature was modified because it had a conceptual bias and overestimated detritus. POM composition expressed as chlorophyll, particulate organic carbon, particulate organic nitrogen, and particulate organic phosphorus was also analyzed. This analysis showed that living seston was the major component of POM during the 7 months. Detritus was minimal during the year and increased to 39% of POM by fall. Most detritus was autochthonous. The lack of correlation between the particulate organic phosphorus and the other components of POM shows that the available phosphorus is rapidly recycled in an efficient manner and not significantly bound in detrital material. The coupling of statistical techniques and data collected during the intensive study of Lake Superior has provided a good insight into the interrelationships and seasonal patterns of nutrients and chlorophyll.     

Spatial Distribution of Fishes in Hydropeaking Tributaries of Lake Superior

Deviation from a river's natural flow regime is considered to be one of the most serious and continuing threats to lotic ecosystems. Peaking hydroelectric facilities, which are designed to adjust the level of power generation in accordance with hourly energy demand, can dramatically alter flows and temperatures and ultimately lead to changes in the quantity and quality of habitat available to fish. In this study, we examine the spatial distribution of river fishes, benthic invertebrates and organic matter along lateral and longitudinal gradients in two hydropeaking and eight natural Lake Superior tributaries in Ontario, Canada. This study demonstrates that (i) hourly variation in flow, caused by hydropeaking, results in a varial zone that supports significantly fewer fish than the adjacent permanently wetted channel and (ii) strong longitudinal gradients in fish biomass, particularly for sedentary species such as slimy sculpin (Cottus cognatus), exist in regulated rivers, and fish biomass is up to four times greater at sites directly below the peaking dams than at sites further downstream or in nearby natural rivers. Gradients in the spatial distribution of fishes closely follow changes in food resources such as benthic organic matter and invertebrates, suggesting that these gradients are driven by spatial shifts in food availability and are ultimately caused by gradients in abiotic habitat variables. Monitoring and assessment efforts should take into account that lateral and longitudinal gradients exist in regulated rivers, and this understanding must be incorporated into sampling programmes. Failing to do so could alter the interpretation of river productivity, integrity and health. Copyright © 2013 John Wiley & Sons, Ltd.     

Ecology and timing of black bass spawning in Lake Ontario and the St. Lawrence River: Potential interactions with the angling season

The timing of spawning for largemouth bass and smallmouth bass in Lake Ontario and the St. Lawrence River was examined over a 3-year period. Temperatures were warmer in the preferred spawning habitat of largemouth bass, and the majority of nests had offspring that had reached the swim-up stage at the opening of the bass angling season in 2 of the 3 years examined. In contrast, the proportion of smallmouth bass nests that had reached the swim-up stage when the bass angling season opened ranged from 4 to 13% during these years. Using models created from nest observations and temperature data, we provide estimates of the additional time required to allow higher percentages (20%, 50% and 80%) of smallmouth bass nests to reach the swim-up stage. Invasive round goby, which are an important nest predator, were more abundant in the spawning habitat preferred by smallmouth bass. A simulated angling experiment showed that round gobies invaded about half of the nests when the guarding male was briefly angled and released. In view of these results, it may be important to re-evaluate the dates for the closed season in order to maintain the quality of the bass fisheries in Lake Ontario and the St. Lawrence River.     

Lake Ontario ice coverage: Past,present and future

Lake Ontario ice conditions are statistically linked to regional temperatures recorded in Toronto, during the most recent climate normal (1980/81–2009/10). A metric was developed to capture the net melting effect of average winter temperatures to characterize lake ice conditions, referred to as Net Melting-Degree Days (NMDD). This metric was able to account for 78% of lake ice interannual variability (R² = 0.783, P < 0.001). Based on NMDD parameters, current lake ice conditions were characterized in four ways: heavy, moderate, light and very light. Lake Ontario ice conditions were reconstructed to create a hindcast for the span of the instrumental temperature record (1840/41–1979/80). Based on a decadal analysis, heavy ice seasons decreased significantly (R² = 0.658, P < 0.001) from the 1840s to the 2000s, declining from an average of 6 heavy ice seasons per decade during the most distant climate normal (1840s to 1960s) to an average of only 1 heavy ice season per decade during the most recent climate normal (1980s to 2000s). Finally, lake ice conditions are projected to the end of the 21st century, using an optimal ensemble of Global Climate Model outputs for two different climate change scenarios (RCP4.5, RCP8.5). Heavy ice seasons no longer occur as early as the 2050s under both RCP4.5 and RCP8.5. Whereas, very light ice seasons go from being an extreme in the baseline period (10%), to the dominant characterization of Lake Ontario ice conditions by the 2080s, for both RCP4.5 (73%) and RCP8.5 (100%).