One hypothesis for the transcontinental and intra-Great Lakes basin transfer of round gobies (Neogobius melanostomus) has been that round gobies were pumped into the ballast water of ships. During June 2005 in Lake Erie, we obtained evidence of a vertical migration of round goby larvae, when we collected 167 round goby larvae in surface ichthyoplankton net tows at night and zero during day. These results complemented similar findings from the Muskegon River estuary of Lake Michigan during 2003 and 2004, documenting diel vertical migration for the first time in larval round gobies. We suggest vertical migration behavior may have allowed larval round gobies to be transported to and within the Great Lakes via ballast water and dispersed in the Great Lakes via advection of 6.5–8.5-mm long larvae at the surface. Based on our results, if ballast water was only taken on near the surface during daylight hours from May through September when larval round gobies were present, it would have mitigated the spread of round gobies throughout the Great Lakes.

Great Lakes coastal wetlands may be more resistant to invasion by certain nonindigenous species and thus serve as refuge habitats for native species. As a first step in testing this hypothesis, we investigated the distribution of round goby (Apollonia melanostomus, formerly Neogobius melanostomus) in the lower reaches of several Lake Michigan tributary systems that contain both wet-land and lake habitats near their confluences with Lake Michigan. Using fyke nets, we sampled round gobies in lake and adjacent wetland habitats in four systems in 2004 and six systems in 2005. In each macrohabitat (lake or wetland), we sampled three microhabitats (mono-dominant stands of Nuphar, beds of submersed aquatic vegetation, and bare sediment). We found that round goby catch was generally lower in wetland macrohabitats than adjacent lake macrohabitats and that round gobies appeared to prefer beds of submersed aquatic vegetation in lakes among the three microhabitats. The majority of round gobies in all habitats were relatively small (< 7 cm standard length). We also found a significant negative correlation between round goby catch and distance of sampling points from the Lake Michigan shoreline in 2005, suggesting that 1) Lake Michigan nearshore waters (including the connecting navigation channels and pier areas) may be serving as round goby spawning and nursery habitats with subsequent dispersal into the tributary lake/wetland complexes, and 2) round gobies may still be invading these systems from Lake Michigan. Our results provide evidence that coastal wetland habitats are more resistant to invasion by round gobies than adjacent lake habitats.

Patterns in abundance, growth, and condition of lake whitefish (Coregonus clupeaformis) from Lake Erie were compared with those from Lake Ontario. Discontinuous assessment data were available from 1972 to 2003 for each lake to describe abundance, growth, and condition, while a seasonally intensive field program was undertaken in 2003 to describe diet, energy density, and gonadosomatic index (GSI). Through time, abundance declined more in Lake Ontario than in Lake Erie. Length-at-age (growth) and condition both declined significantly in Lake Ontario but did not change in Lake Erie. Diet analysis revealed chironomids, dreissenid mussels and sphaeriids made up the bulk of lake whitefish diet in Lake Erie. Diet in Lake Ontario exhibited more seasonal variability with amphipods and gastropods comprising the bulk of the spring and fall diets, and dreissenid mussels dominating summer diets. Lake whitefish energy density (J/g wet mass) was significantly higher in Lake Erie than in Lake Ontario, increasing with body mass and strongly correlated with water content. Female gonadosomatic index was also significantly greater in Lake Erie than in Lake Ontario. Biological attributes of lake whitefish from Lake Erie did not change greatly from the late 1980s to 2003 while fish from Lake Ontario exhibited decreased size-at-age and condition likely due to decreased energy content of diets compared to pre-dreissenid mussel conditions, and possibly lower availability of benthic prey compared to Lake Erie.

Potential negative ecological interactions between ruffe Gymnocephalus cernuus and round goby Apollonia melanostoma (formerly Neogobius melanostomus) might affect the colonization dynamics of these invasive species where they are sympatric in the Great Lakes. In order to determine the potential for ecological interactions between these species, we examined the activity, aggression, and habitat use of round gobies and ruffe in single species and mixed species laboratory experiments. Trials included conditions in which food was concentrated (in light or darkness) or scattered. Results showed that ruffe were more active than gobies, particularly when food was scattered. Activity of both species was significantly lower during darkness. Round gobies were significantly more aggressive than ruffe, and total aggression was lower in mixed species trials. Habitat use by ruffe and round gobies overlapped considerably, but we observed significant differences between species in their use of specific habitats that depended on experimental conditions. Overall, ruffe used open habitats more often than did round gobies, primarily when food was scattered. Round gobies used rocks significantly more frequently than did ruffe, but their use of rock habitat decreased during dark conditions. Ruffe were found more often in plant habitats and less often near the wall of the pool in trials during daylight with concentrated food. Activity and habitat use of ruffe and round goby did not significantly differ between single and mixed species trials. Overall, we found little evidence for negative ecological interactions between ruffe and round goby in these laboratory experiments.

At some coastal beaches, concentrations of fecal-indicator bacteria can differ substantially between multiple points at the same beach at the same time. Because of this spatial variability, the recreational water quality at beaches is sometimes determined by stratifying a beach into several areas and collecting a sample from each area to analyze for the concentration of fecal-indicator bacteria. The average concentration of bacteria from those points is often used to compare to the recreational standard for advisory postings. Alternatively, if funds are limited, a single sample is collected to represent the beach. Compositing the samples collected from each section of the beach may yield equally accurate data as averaging concentrations from multiple points, at a reduced cost. In the study described herein, water samples were collected at multiple points from three Lake Erie beaches and analyzed for Escherichia coli on modified mTEC agar (EPA Method 1603). From the multiple-point samples, a composite sample (n =116) was formed at each beach by combining equal aliquots of well-mixed water from each point. Results from this study indicate that E. coli concentrations from the arithmetic average of multiple-point samples and from composited samples are not significantly different (t = 1.59, p = 0.1139) and yield similar measures of recreational water quality; additionally, composite samples could result in a significant cost savings.

Epiphytic macroinvertebrate communities of four coastal wetlands of Green Bay, Lake Michigan were compared by taxonomic composition, feeding group composition, and environmental influences using Bray-Curtis ordination. Ordination scores from the most sheltered oligotrophic site, Portage Marsh, were distinct from the eutrophic, exposed sites located in middle and lower Green Bay—Seagull Bar, Little Tail Point, and Dead Horse Bay. Epiphyton chlorophyll a, phytoplankton chlorophyll a, and specific conductance strongly correlated to the ordination axes, indicating the trophic gradient within Green Bay was a primary environmental influence. The feeding group compositions at the sites were consistent with the type and abundance of food available. Portage Marsh is a scraper-shredder system, with macroinvertebrates feeding mainly on epiphyton and coarse particulate detritus. Dead Horse Bay and Little Tail Point are collector systems, sustained by phytoplankton and fine particulate organic matter. Seagull Bar is intermediate in trophic position along the ordination axes, but more closely resembles the latter two sites. The type and abundance of food resources available to these invertebrate communities are influenced by wave exposure, light attenuation, nutrient levels, and algae levels of the littoral and pelagial waters. Macroinvertebrate communities were sensitive to shifts in food resources, which generated shifts in trophic structure.

Natural hybridization between pink salmon (Oncorhynchus gorbuscha) and Chinook salmon (O. tshawytscha) has been observed only in the Great Lakes, where both species have been introduced. The direction of hybridization between these species is poorly understood, thus the present study analyzed mitochondrial DNA (mtDNA) to determine whether the maternal parents were pink or Chinook salmon. During annual salmon population surveys on the St. Marys River (1998–2002), fifty putative hybrids of pink salmon and Chinook salmon were identified from meristic, morphometric and color characters. We designed primers to amplify a fragment of the mitochondrial D-loop control region that included both a variable and a control restriction site (BstNI). Polymerase chain reaction, restriction endonuclease digestion and capillary electrophoresis of mtDNA were used in order to identify the maternal parent of each hybrid salmon. In addition, the amplified fragments from the three fish species were sequenced to further verify the results. All hybrid salmon specimens were found to possess Chinook salmon mtDNA, indicating that hybridization between Chinook salmon and pink salmon is asymmetrical and likely unidirectional, occurring only between male pink salmon and female Chinook salmon. Influences contributing to the hybridization of these salmonid species could include limited spatial and temporal segregation of spawners, sexual selection, a limiting number of Chinook males, and/or physiological factors.

Water-level change is integral to the structure and function of Great Lakes coastal wetlands, and many studies document predictable relationships between vegetation and water level. However, anthropogenic stressors, such as invasive species, land-use change, and water-level stabilization, interact to shift the historical cycle (of native vegetation migration up- and down-slope) toward dominance by invasive Typha species. Knowing from earlier studies that water-level stabilization alters the historical vegetation cycle, we asked if similar shifts can occur where water levels are not stabilized. Using historical aerial photographs of three coastal wetlands (in Lake Michigan's Green Bay, Wisconsin), we determined that habitat dominated by Typha species has expanded to eliminate wet meadow habitat. Between 1974 and 1992, linear regressions showed strong, significant relationships of both meadow area (R² ≥ 0.894; p < 0.02) and marsh area (R² ≥ 0.784; p < 0.05) to water level in all three wetlands. In 2000, meadow area was below that predicted by the historical pattern due to the landward advance of marsh habitat during a year of decreasing water levels. In the same period, land use in the wetland watersheds converted from agriculture to urban. Urbanization and the replacement of native Typha latifolia by the invasive hybrid Typha xglauca may have overwhelmed the beneficial impact of water-level fluctuation. The documentation of vegetation shifts, as herein, is an essential step in the process of preserving and restoring ecological integrity.

Total phosphorus (TP) inputs to Lake Simcoe have led to hypolimnetic dissolved oxygen (DO) depletion and loss of cold water fish habitat. Since 1990, efforts have been made to reduce the total TP input to the lake below a defined target of 75 t/year, which was predicted to lead to reductions in spring TP concentration and improvements in end-of-summer hypolimnetic DO concentrations. The total TP load to the lake during the most recent period of record (1998/99–2003/04) ranged from 53 to 76 t/yr and averaged 67 t/yr, compared to an average of 114 t/yr estimated between 1990/91 and 1997/98 (range 85–157 t/yr). Reductions in TP loads from the catchment via tributary discharge (~26 t) accounted for the majority of the decrease in total load between the two time periods. Total P concentrations decreased significantly in four out of six long-term monitored tributaries; however, concentrations in all six tributaries remain above the level recommended to avoid nuisance plant growth (30 μg/L). Although TP loads to the lake are currently below the target 75 t/yr, excessive growths of filamentous algae and macrophytes continue to be a problem in the nearshore zone. End-of-summer minimum hypolimnetic DO concentrations (average 4.3 mg/L, 1998/99–2003/04) remain substantially below the level (7 mg/L) that is considered protective of lake trout. Efforts to reduce TP loads to the lake therefore need to continue.

Historic reports imply that the lower Detroit River was once a prolific spawning area for lake whitefish (Coregonus clupeaformis) prior to the construction of the Livingstone shipping channel in 1911. Large numbers of lake whitefish migrated into the river in fall where they spawned on expansive limestone bedrock and gravel bars. Lake whitefish were harvested in the river during this time by commercial fisheries and for fish culture operations. The last reported landing of lake whitefish from the Detroit River was in 1925. Loss of suitable spawning habitat during the construction of the shipping channels as well as the effects of over-fishing, sea lamprey (Petromyzon marinus) predation, loss of riparian wetlands, and other perturbations to riverine habitat are associated with the disappearance of lake whitefish spawning runs. Because lake whitefish are recovering in Lake Erie with substantial spawning occurring in the western basin, we suspected they may once again be using the Detroit River to spawn. We sampled in the Detroit River for lake whitefish adults and eggs in late fall of 2005 and for lake whitefish eggs and fish larvae in 2006 to assess the extent of reproduction in the river. A spawning-ready male lake whitefish was collected in gillnets and several dozen viable lake whitefish eggs were collected with a pump in the Detroit River in November and December 2005. No lake whitefish eggs were found at lower river sites in March of 2006, but viable lake whitefish eggs were found at Belle Isle in the upper river in early April. Several hundred lake whitefish larvae were collected in the river during March through early May 2006. Peak larval densities (30 fish/1,000 m³ of water) were observed during the week of 3 April. Because high numbers of lake whitefish larvae were collected from mid- and downstream sample sites in the river, we believe that production of lake whitefish in the Detroit River may be a substantial contribution to the lake whitefish population in Lake Erie.

Selective grazing by zebra mussels has altered phytoplankton communities in many North American lakes, but the specific changes are not the same in each ecosystem. Because of this variation in response, we investigated the impacts of zebra mussels on the plankton community of Lake Champlain with two objectives: first to determine whether zebra mussels increased the dominance of potentially toxic cyanobacteria in the phytoplankton, and second to explore the impact of zebra mussels on protozoans, rotifers, copepod nauplii, and other microzooplankton in the lower food web. Experiments were conducted in 200-L mesocosms filled with Lake Champlain water filtered through a 150-μm sieve to remove macrozooplankton. Zebra mussels were added to half of the mesocosms while the others were maintained as controls. Over a 96-hour experimental period, we tracked nitrogen and phosphorus concentration, chlorophyll α, microcystin concentration, and both phytoplankton and microzooplankton composition and abundance. We found an increase in SRP and total nitrogen concentration and a decrease in the ratio of TN:TP in the zebra mussel treatments over time. Microcystin was undetectable throughout the experiment using the ELISA assay. Phytoplankton biovolume, including cyanobacteria biovolume, declined significantly in the zebra mussel treatments, as did rotifer, protozoan and nauplii abundance. By both direct (consumption) and indirect (altered nutrient availabilities and increased competition) means, zebra mussels clearly seem capable of strongly influencing the lower planktonic foodweb in the many shallow water habitats of Lake Champlain.

Surveys of the benthic macroinvertebrate community were conducted in the main basin of Lake Huron in 2000 and 2003, and in Georgian Bay and North Channel in 2002. Results were compared to surveys conducted in the 1960s and early 1970s. Although data of earlier surveys were inconsistent, our best estimates suggest that total density of the four major benthic taxa (Diporeia spp., Oligochaeta, Sphaeriidae, and Chironomidae) in the main basin declined dramatically between the early 1970s and 2000. Populations of all major taxa continued to decline between 2000 and 2003, particularly Diporeia and Sphaeriidae. Diporeia was rare or absent in the southern end of the lake and in some nearshore areas in 2000, and by 2003 was not found at depths < 50 m except in the far northeastern end of the lake. Densities of the major taxa in Georgian Bay and North Channel in 2002 were not different from densities in 1973 despite differences in survey methods. A limited study in southern Georgian Bay, however, found that densities of both Diporeia and Sphaeriidae declined to zero at most sites between 2000 and 2004. The population of Dreissena polymorpha was stable in all lake areas, but Dreissena bugensis increased, particularly at the 31–50 m depth interval in the main basin. Since there were no extensive surveys in Lake Huron in the period between nutrient abatement (late 1970s) and the establishment of Dreissena (early 1990s), it is difficult to determine relative roles of these events on observed declines. However, since phosphorus loads have been stable since the early 1980s, declines between 2000 and 2003 can likely be attributed to Dreissena.

The linkage between land use in a catchment basin and downstream aquatic ecosystems, especially effects on algae attached to substrata or loosely aggregated in the littoral zone, represents a void in our understanding of lake systems. The occurrence of beds of metaphyton at some stream mouths and not others in Conesus Lake, NY (USA) provided an opportunity to consider the relationship between land use and phosphorus and nitrogen losses on the development of shoreline metaphyton blooms. Experiments were performed in the littoral zone of a large temperate lake to test the hypothesis that effluent high in phosphorus and nitrate from tributaries draining agricultural watersheds had a stimulatory effect on the growth of littoral metaphyton, while effluent from a forested watershed did not. The study encompassed six watersheds of varying agricultural use (60–80%) and a forested watershed (12% agriculture). For each experiment, two quadruplicate sets of plexiglass incubation chambers (height = 50 cm, interior diameter = 9.5 cm) containing native assemblages of metaphyton received lake or tributary water continuously over a 3-day lake incubation period. Growth of metaphyton incubated in lake water and in tributary water was compared and differences appeared to be related to nutrient concentrations. A statistically significant stimulatory effect was measured for the six tributaries draining agricultural watersheds but not for the forested watershed. Tributary loadings appear to stimulate metaphyton at sites where the hydrology and hydrodynamincs are suitable. A significant positive linear relationship existed between percent metaphyton cover in the littoral zone and the percent of land use in agriculture. Metaphyton abundance is impacted by land use practices and subsequent loss of nutrients from the catchment.

Lake Malawi, in south-eastern Africa, is subject to increasing loading of suspended solids caused by land use pressure in its watershed. Whether this load is transported into the lake as overflow, interflow or deep underflow determines to a large extent its effect on the lake ecosystem. In this paper, vertical distributions of suspended solids in the Linthipe River delta region of the lake are described from multiple surveys during two rainy seasons. These data are supplemented by data from a single survey near four northern rivers also tributary to the lake. Profiles of temperature, conductance, and suspended solids concentrations (SSC, estimated from optical backscatter and beam transmission) are used to identify fluvial intrusions into the water column. Most inflow plunged to the seasonal metalimnion where it spread along high density gradients as interflow. While SSC in surface plumes rarely exceeded 10 g m⁻³, and in intrusions in the lower metalimnion was rarely greater than 1 g m⁻³, concentrations up to 420 g m⁻³ were recorded in interflow near the thermocline. Although storm runoff density often exceeded 100 m depth-equivalence in the lake, underflow density was reduced to metalimnion-equivalence (30–50 m depth) within a few 100s of meters of the river mouth. We attribute bottom-attached turbid layers, and the few unattached turbid layers in the lower metalimnion, all with positive conductance anomalies, to sediment resuspension and not to runoff. We conclude that the upper metalimnion is the prevailing pathway carrying watershed runoff horizontally throughout Lake Malawi.

Increasing land use pressure has resulted in increasing suspended solids loading to many African lakes in recent decades, causing concern that littoral habitat may be degraded by sedimentation and light limitation. In a companion paper in this issue, we showed that much runoff into Lake Malawi plunges below the euphotic zone and spreads through the lake via the upper metalimnion, thus mitigating nearshore and upper water column impacts. In this study we use data from multiple surveys through two rainy seasons to describe spatiotemporal distributions of suspended solids concentration (SSC) and transparency caused by the flow of runoff through the Linthipe River delta region of Lake Malawi. SSC in interflow at 30–50 m depth was typically an order of magnitude greater than in surface plumes. Seasonal median transparency near the river mouth was inversely proportional to the suspended load in surface-buoyant inflow, but independent of total loading. This is because storm runoff was more dense (cooler and with higher SSC) than runoff during low-flow periods, so that it was more likely to plunge. The effect was to moderate inter-annual variability in light attenuation near the river mouth because transparency was lower (due to more surface-buoyant inflow of turbid water) in the year when solar irradiance was higher (less cloud cover). The effect was, however, local. Beyond a few kilometers of the river mouth upward mixing of suspended particles carried in interflow reasserted the more intuitive inverse dependence of transparency on total loading.

Spectral slope (S), describing the exponential decrease of the absorption spectrum over a given wavelength range, is an important parameter in the study of of chromophoric dissolved organic matter (CDOM) dynamics, and also an essential input parameter in remote sensing models. Furthermore, S is often used as a proxy for CDOM composition, including the ratio of fulvic to humic acids and molecular weight. The relative broad range in S values reported in the literature can be explained by the different spectral ranges and fitting methods used. A single exponential model is used to fit the S values for 17 investigations involving 458 samples in Lake Taihu from January to October in 2004. The average S value was 15.18 ± 1.39 μm⁻¹ for the range of 280–500 nm, which fell within the range reported in the literature. The frequency distribution of S value basically obeyed a normal distribution. Significant differences in S values between summer and other seasons showed that phytoplankton degradation was one of the important sources of CDOM in summer, whereas CDOM mainly came from the river input in other seasons. Furthermore, the estimated S value decreased with increasing wavelength range used in regression. The maximum and minimum values derived from the regression were 17.89 ± 1.25 μm⁻¹ and 13.62 ± 2.11 μm⁻¹ for the wavelength ranges of 280–380 nm and 400–500 nm, respectively, a decrease of 23.9%. S values significantly decreased with the increase of CDOM absorption coefficients. CDOM absorption coefficients could be more appropriately estimated from exponential model introducing the variation of S with absorption coefficients, making them useful for a remote sensing bio-optical model of Lake Taihu. DOC-specific absorption coefficient a*(λ) and the parameter M describing molecular size of the humic molecules could also be used as a proxy for the sources and types of CDOM. A general relationship was found between S and a*(λ), and M values. S increased with the decrease of DOC-specific absorption coefficient and the increase of M corresponding to the decrease of molecular weight.

An established population of Cabomba caroliniana now covers extensive littoral areas in the shallow waters of Kasshabog Lake (Ontario). This is the first known naturalized population of this non-native aquatic macrophyte, commonly called fanwort, on the Canadian side of the Great Lakes basin, despite the fact that it was first reported in the 1930s. High dispersal potentials combined with the ability to adapt and grow in a range of environmental conditions have made C. caroliniana a nuisance species in Australia, Japan, and parts of the United States. However, little is known about the broader ecological implications of its introduction and establishment. Using a survey approach, we conducted a preliminary assessment of water chemistry, macrophyte, epiphytic algae, and macroinvertebrate communities found in C. caroliniana beds and compared them with native macrophyte beds in Kasshabog Lake. Light penetration was significantly reduced in the C. caroliniana beds and was the only sampled physiochemical parameter that differed between bed types. We also found several notable differences in the structure and composition of biological communities within macrophyte beds. While native macrophytes were present in dense C. caroliniana beds, abundance was considerably low and unevenly distributed. Significantly more epiphytic algae was present on C. caroliniana plants, however community composition was comparable with epiphytic algae found on native macrophytes. The taxonomic composition of macroinvertebrates was similar between C. caroliniana and native beds, while abundance was substantially higher in C. caroliniana beds, owing to high densities of coenagrionids and chironomids. These differences suggest that C. caroliniana is changing macrophyte community composition in this lake, having an impact on epiphytic algae, and creating a new habitat for some macroinvertebrates. Further studies are required to determine the extent of these ecological impacts.

To understand the functional-physiological influences on the depth distribution of deep-water cisco (bloater), Coregonus hoyi, in the Laurentian Great Lakes, we compared their gas bladder anatomy with the putative ancestor, shallowwater cisco, C. artedi, from a shallow inland lake. There were no general differences in the gas bladder anatomy—both coregonines possessed simple micro retia (counter-current gas multipliers) and an open pneumatic duct connection between the bladder and esophagus. Bloater possessed a significantly smaller duct diameter relative to body size in comparison with shallowwater cisco, yet the estimated resistance to air flow through the duct of bloater was ~7.2 times lower than that of cisco. In light of these findings, we discuss evidence for two modes of gas bladder inflation in bloater: 1) gas deposition via micro retia at depth, and 2) gas acquisition via aquatic surface gulping (ASG). Because the hypolimnetic distribution of adult bloater precludes them from using ASG, we infer that these fish utilize micro retia to deposit gas into their bladders. We hypothesize that bloater compensate for relatively modest rates of gas deposition by not voiding excess gas. This “gas bladder conservation (GBC) hypothesis” for bloater is supported by four pieces of evidence: 1) deepwater habitation with no access to the surface to perform ASG, 2) a lack of extensive retia and modest gas deposition rates in coregonines, 3) a tendency to bloat when brought from depth to the surface, and 4) previous findings of a confined depth range, centered around the depth of neutral buoyancy.

The importance of fish spawning habitat in channels connecting the Great Lakes to fishery productivity in those lakes is poorly understood and has not been adequately documented. The Detroit River is a reputed spawning and nursery area for many fish, including walleye (Sander vitreus) that migrate between adjacent Lakes Erie and St. Clair. During April–May 2004, near the head of the Detroit River, we collected 136 fish eggs from the bottom of the river on egg mats. We incubated the eggs at the Great Lakes Science Center until they hatched. All eleven larvae that hatched from the eggs were identified as walleye. These eggs and larvae are the first credible scientific evidence that walleye spawn in the Detroit River. Their origin might be a stock of river-spawning walleye. Such a stock of walleye could potentially add resilience to production by walleye stocks that spawn and are harvested in adjacent waters.