A summary of special scientific sessions of the 44 ^th Conference on Great Lakes Research of the International Association for Great Lakes Research

University of Wisconsin-Green Bay
Green Bay, Wisconsin
June 10-14, 2001

Prepared on behalf of IAGLR Board by
Patricia Chow-Fraser, President IAGLR 2001
for
The Council of Great Lakes Research Managers International Joint Commission

July 9, 2001

• All summaries have been prepared by students/persons attending the IAGLR 2001 Conference through a contract to IAGLR. The perspectives taken are those of the authors and do not necessarily represent the view of either the Council of Great Lakes Research Managers, or the Board of Directors, IAGLR. Any comments, questions or criticisms should be directed to the students whose names and email addresses appear immediately after the title of each summary.

Evidence and Modeling of Climate-Driven Global Change in the Great Lakes (by Piet Verburg, U of Waterloo, pverburg@scimail.uwaterloo.ca )

The fact of global warming is now generally accepted among scientists, and most scientists agree that the major cause is antropogenic, the emission of greenhouse gasses. The earth’s temperature has risen by an average 0.6 °C over the past century, and historic variability has been surpassed. For the North American Great Lakes there is no comparable temperature data set. However, the data available on ice cover, indicating much later freeze and earlier breakup dates of ice, and therefore a drastic reduced duration of ice cover of the lakes, by an average of 12 days over the past century, are convincing evidence of lake warming. Local urban warming shortened ice cover even much more, such as at the Toronto harbor (by 31 days over the past century). Increased snowfall south of Lake Superior since 1930 is indicative of increased heat storage in Lake Superior. On the other hand, a significant increase in variability in various climatic parameters complicates weather prediction and the use of climate models. Also the increase in variability may be a result of greenhouse gas-forced climate warming, with an intensification of the El Niño oscillation cycle.

A minimum global rise of 3 °C is expected over the 21 ^st century, depending on a substantial reduction of greenhouse gas production. This is of grave importance to species with a narrow temperature range. Especially fish and other aquatic species are affected. Species such as the threadfin shad will move northwards, but cold water species will find their habitat much reduced. Interactions between native and exotic species are expected to affect these changes in species ranges. Overall extinction rates will increase, and many species will disappear from the southern range of the North American continent. Even over the past century the influence of climate was apparent on small scales such as for the small mouth bass (introduced) and the lake trout populations in Lake Opeongo, Ontario. After correction for density-dependent processes, recruitment and growth of small mouth bass correlated positively with summer air temperatures. Peaks coincided with El Niño years, except when the effect of El Niño was masked by such events as the Pinatubo explosion. The latter reduced the global temperature in 1992, and growth of small mouth bass was less than expected that year. However for the abundances of 50 fish species in Lake Ontario over the past 28 years, the effects of climate forcing was much less clear, with abundances of different species apparently varying with different sets of climate and limnological indexes.

Global warming is expected to influence hydrology and biology of the North American Great Lakes. An increase in evaporation, larger than the expected increase in rainfall, will lower water levels by 1 to 3 feet towards the end of the century. Primary production in the Great Lakes will decrease by 10-20 %. The period of summer stratification will increase by several months with climate warming over the 21 ^st century, causing the early termination of spring phytoplankton blooms. Eventually this will reduce fishery production in the Great Lakes. For many species of trees the climate around the Great Lakes will be too warm during summer to survive. An estimation of such climatic effects expressed in dollars instead of in degrees and feet would be more instructive. However there was little consensus as to whether in future science should be more focussed on basic research of the climate problem or more oriented towards finding solutions to the problem.

The frequency of heavy precipitation events (during 1-7 days) in the Great lakes basin was higher over the last twenty years, by 15 %, compared with earlier in the century. Such events lead to increased flooding and property damage. Correlation with circulation patterns which were associated with heavy precipitation and future circulation patterns as predicted by climate models indicate that this trend can be expected to continue during the 21 ^st century. Largest % increases are predicted in the heaviest precipitation events. Enhanced tropical sea surface temperatures appear to be a leading cause for the global increase in the frequency and intensity of heavy precipitation events.

Rainfall and runoff to Lakes Michigan and Huron show increasing annual trends, and runoff was an increasing component of the total input over the previous century. There was also a shift in the seasonal runoff cycle over the past century, with decreased runoff in spring and increases in the fall and winter seasons. Peak flows were larger towards the end of the previous century and occurred earlier in the year. Future research will focus on the question how these findings relate to either seasonal precipitation patterns or to changes in land use practices such as irrigation. The seasonal shifts in basin inputs and resulting changes in the lake water balance are also indicated by trends in seasonal lake levels. At Lakes Erie and Ontario, towards the end of the 20 ^th century the annual lake level rises and falls occurred 1 month earlier than at the end of the 19 ^th century. These trends were probably similar for each of the Great Lakes, but analysis of past water levels in the Great lakes basin is far from complete. At present it is not possible to come up with a definitive or even reasonable answer as to why the seasonal cycle of the water balance and the level of Lake Ontario is changing. The data analysis is complicated by the regulation of the outflow at the St. Lawrence River since 1959, which leads to much more variable outflow rates. Until the onset of regulation for instance, the annual difference between lake level maxima and minima was increasing, for unknown reasons.

Much work seems still needed to understand the processes involved in climate forcing, their interactions, and effects on the Great Lakes basin water balance and how climate effects translate into lake levels. Scientists generally expect lower lake levels in the 21 ^st century due to increased evaporation. However, acid gas emissions which peaked in the mid West around 1970, increased cloud cover by aerosol formation and lowered incoming solar radiation and evaporation rates. This contributed to higher than normal Great Lake levels from 1965 to 1997. Since 1997, with lowered emissions, lake levels started to decrease. Aerosol loads are clearly an important factor in the climate forcing but to include estimates of future aerosol concentrations in climate modelling is still problematic.

The Green Bay Ecosystem: An Interdisciplinary Synthesis, New Findings, New Directions (by Sheila McNair, McMaster University, mcnairs@mcmaster.ca )

Green Bay has been subject to anthropogenic inputs including nutrients and toxics, models have been developed for the entire area and for smaller sections that attempt to predict how and where inputs move and accumulate within the bay. Most of the talks in these 3 sessions focused on abiotic elements, others related to biological organisms and some incorporated interactions of biotic and abiotic elements within a whole system context.
Some key messages:

1. Nutrient loading is high, land use is implicated. Industrial inputs are of a major concern, but agricultural land use has a big impact on total phosphorus loadings (non-point 65%) seen in trophic gradient from south to north of Green Bay. Agricultural practices need to be reconsidered.
   
2. CO2 flux across surface water interface is net heterotrophic in north of Green Bay, net autotrophic in south, overall result is a net export of CO2 to the atmosphere.
   
3. TOC moves quickly through the system, eutrophication increases the burial flux of total organic carbon in sediment. There are implications for movement of contaminants from rivers to the lake.
   
4. PCBs are known contaminants in the Fox River, along with other chemicals. The movement of contaminants from the river to the bay and within the sediment vertically is a major concern and a social/political problem. Local inputs have been stopped, but movement continues. Patterns of sedimentation and models of rates and directions predict further dispersal of contaminant load, and more difficult remediation in the future if not addressed quickly. 90% of PCBs are over an area of 20% of Green Bay, and moving, but models have limitations.
   
5. PCBs (and other contaminants) are taken up by biota. Zebra mussels show that transport is via the sediment, not the water column.
   
6. Zebra mussels have had some effects on water quality and clarity in lower Green Bay, but there is no clear signal. There appears to be decoupling of the total phosphorus-chlorophyll relationship, this may be a signal of change induced by Zebra mussels. Changing water levels (current lows) may be affecting the system as well.
   
7. Yellow Perch are declining. Recruitment, habitat, over-harvesting, water quality may be reasons. White perch have too high PCB levels for harvesting.

Overall, Green Bay is a system impacted by humans, with a clear gradient of impact high in the south, lower in the north. Contaminant and nutrient loading from historical sources needs to be addressed, management solutions need to be found and quickly. Living organisms (perch, humans, zebra mussels) are integrating the effects of loadings over the long term, and the solutions will be equally long-term. Action needs to be taken now. As one speaker said sediment contamination is a Global problem and contaminants move through the atmosphere. The important thing is to take action quickly because dispersal compounds the problem of cleanup. These issues have to be considered in the long term.

Sediment Transport (by Donald Hughes, SUNY-College of Environmental Science & Forestry, djhughes@mailbox.syr.edu )

Kirk Ziegler (Quant. Envir. Analysis) started the session with an overview of the basic requirements for modeling sediment transport. For a model to succeed, it must account for:

1. temporal and spatial variations, and
   
2. the variety of sediment types.

He classified sediments as (a) flocculating (silt, clay) or (b) coarse (sand, etc.) Resuspension is governed by bottom shear stress () and the cohesiveness of the bed. Mr. Ziegler reviewed inputs and calibration for both hydrodynamic and sediment transport models. Requirements for modeling include: a numerical grid with sufficient density; grids for hydrodynamic and sediment transport models must match. Also, the model must include at least two particle size classes--one is not adequate! For a good description of deposition, one must account for flocculent effects on settling speed, include effects of shear stress on settling speed, and include vertical stratification where it occurs. For resuspension, one must include effects of armoring, and include bed load where needed. A mass balance approach improves the understanding of the system, but one must have good data set to support it.

Joe DePinto (LimnoTech, Inc.) provided a useful comparison of three devices used to measure sediment resuspension:

1. the shaker calibrated with annular flume;
   
2. the SEDFLUME device; and the Ravens inverted flume.
   

#1 and #2 rely on measurement of an intact core, whereas #3 is conducted in situ. Despite its drawbacks, most models are designed to use data from the shaker device. #2 and 3 provide superior data, but models which use those results are still in R&D stages.

Craig Jones (U.C. Santa Barbara) presented some interesting model results for sediments in the lower Fox River. Hydrodynamics were modeled using 2-D vertically integrated transport. Sediment transport was modeled using SEDFLUME data, with sediment divided into multiple size classes. He showed patterns of deposition in which fines accumulated along the sides of the river, grading to coarser sands in the center channel. A high-flow event (June 1989) was seen to wash away fines, resulting in a broader area of sands in the river channel. However, sediments mobilized from upstream sources soon re-deposited in these areas once flow subsided. Jones said this was consistent with real world data. In conclusion, he stressed the need for accurate data, including SEDFLUME data which are more detailed and accurate than shaker experiments.

Similarly, X. Zhang (Wisconsin DNR) presented sediment modeling results for the Milwaukee Harbor. Sediments there are contaminated with up to 40 ppm PCBs at a depth of 1m. Important processes incorporated into the model included: bioturbation, sediment dynamics and diffusion of PCBs into and out of pore water. PCB water concentrations were projected for the next 30 years under a variety of scenarios. The model was found to be relatively insensitive to the influence of bioturbation. However, changes in the thickness of the active sediment layer had profound effects on the model output.

Dave Fowler (Milwaukee Metro. Sewerage District) talked about efforts to restore river channels in the Menomonee River. The river has over 5 miles of concrete walls, some of which are crumbling. Data were collected on river cross-sections and sediment bed composition; this formed the basis for hydrodynamic and sediment transport models. Findings included: existing banks and river bottoms are very stable; removal of walls will require plan for bank stabilization (soft engineering under consideration); sediments are not deposited in downstream channelized areas. MMSD is working to convince upstream municipalities that "hard" bank stabilization measures are not needed. Chris Marvin (Environment Canada) reviewed sediment contaminant data for Lake Erie over the past 30 years. Most of the data were from surface sediments collected from 70 stations throughout the lake.

Highlights of results are as follows:

1. PCBs: 3-4-fold decrease (1971-1999); there is a general downward trend from west to east, and from south to north. Congener distribution has not changed.
2. alpha-HCH: also decreased; conc. now <1 ?g/kg
3. gamma-HCH (lindane): has increased due to increased use in Canada; avg. conc. =2.5 ug/kg.
4. mercury: spatial distribution similar to PCBs; highest conc. near Detroit R. (0.7 mg/kg)
5. dieldrin: more uniformly distributed; not decreasing over time
6. endosulfan: currently in use; conc. up to 20 ug/kg
7. dioxins/furans: strong E-W gradient; Detroit R. is apparent source; range 0-35 pptr TEQ; max conc. 50-60 pptr TEQ.
8. toxaphene: Avg. conc. = 15 ug/kg; no decrease over time; minor contaminant

An overall sediment quality index was calculated, which ranged qualitatively from poor to excellent. A strong W-E gradient was found. Overall conclusions are:

1. dioxins and mercury are largest sources of toxicity;
2. Detroit R. is significant source (or vector) of contaminants to lake; and
3. this is a practical application of sediment quality guidelines.

In response to questions, Marvin noted that the overall index is subjective. Were sediment concentrations corrected for sand content? Yes, but this had little effect on patterns of contamination. Normalizing with TOC content produced greater uniformity, overall. In response to another attendee, Marvin agreed that the Maumee River is a likely source of contaminants.

G.L. Smith (U. Iowa) began the afternoon session with a study using sediment traps in S. Lake Michigan. Monthly samples were collected at monthly intervals over the period June '98 to May '99 at water depths ranging from 12 to 62 m. Zebra mussels interfered with operation of the traps at some locations. The measured sediment fluxes ranged up to 20 g/m2day; there were two storms associated with much resuspension. Both PCBs and DDT were found at levels exceeding the method detection limits; other compounds were below MDLs. Smith suspected an anaerobic source of DDT, since "fresh" (not degraded) DDT was widespread. The flux of PCBs was up to 800 ng/m2-day. He also described the variability in congener patterns among the various compartments. As expected, settling particles were enriched in the heavier congeners, compared with suspended matter, the water column and the air.

Sandra Miller (U. Iowa) talked about the effect of storms on resuspension of particles containing persistent organic pollutants in L. Michigan. The goal of this study was to find out more details about the spatial (between sites, and with depth) variability and temporal (esp. storm-related) variability. Sampling was conducted each spring in 1998, '99, and 2000. For PCBs, deep water concentrations were much lower than near shore. Also, dissolved phase PCBs dropped where settling rates of particles increased. The partitioning between dissolved (measured with XAD resin) and particle (0.7 um filter) was not found to vary over time, but did vary from location to location.

Lake Michigan Yellow Perch: On the Road to Recovery? (by Todd Morris, Erindale College, U of Toronto, tmorris@credit.utm.utoronto.ca )

D.F. Clapp identified that Lake Michigan yellow perch stocks showed drastic declines during the late 80’s and early 90’s. Commercial trawls showed the first signs of declines with creel surveys showing a 4-5 year lag. As a result of these declines the Yellow Perch Task Group, a multi-disciplinary agency, was formed to address concerns and attempt to determine the cause of these declines. This group initially set out a list of 17 possible hypotheses to explain the declines but has since narrowed the scope to 5 main areas.

These areas are:

1. egg stage and spawning stock influences,
2. physical factors,
3. prey availability,
4. predation, and
5. stock recruitment.

Research into these areas has been ongoing since 1997 and the remainder of this session focused on specific results of studies designed to examine each of these areas.

Proper management of these stocks requires knowledge of their genetic makeup and whether they can be managed as a single stock representing L. Michigan or whether they should be managed on a smaller scale. L. M. Miller summarized genetic studies using microsatellite DNA loci which revealed much greater genetic variability than previously thought (heterozygosity ranging from 0.21 to 0.86). The Green Bay stock is clearly distinct from the main Lake Michigan stock showing as much variability as inland lake stocks. The Main lake Michigan stock should also be managed as 2 separate stocks northern and southern.

Limitation by prey availability was investigated by A. S. McNaught et al. as well as A. Mahan et al.. Large declines in zooplankton abundance have been seen to coincide with the decline in Yellow Perch recruitment (from 50 70 /L prior to the crash to 5 30/L after). A.S. MCNaught reported that larval perch abundance appeared temporally out of sync with the zooplankton abundance with the larval peak coming in early June and the zooplankton peaking in late June. A Mahan suggested that larval yellow perch ( <10mm) might be gape limited in their prey selection. Both studies found larval preference for copepod nauplii. C. J. Heyer took the findings of decreased zooplankton abundance and spatial overlap with prey items a step further and addressed the issue of larval condition using RNA-DNA ratios as an indicator of protein synthesis and condition. Results indicated that these ratios did not differ between the main body and the Green Bay stocks suggesting that larvae are not starving and that prey availability is not the cause of current failures.

Physical factors, in particular wind-aided larval transport, were investigated by J.A. Janssen and D. Jude as well as Dettmers et al.. J. A. Janssen reported that larval perch abundance was closely linked to wind direction with higher abundances observed at downwind sites. Janssen reported that major lake currents might cause larvae to be moved from rocky nest sites through areas where large Diporeia losses have been seen. Dettmers also found higher larval abundance offshore where zooplankton abundance is higher and size larger which may favour increased growth in these regions however spatial overlap with predators is complicated by this movement and may be higher or lower. Hydroacoustic studies by M.B. Balge and Mason show high spatial overlap with larval perch and alewife in these areas.

D.G. Fitzgerald et al. investigated winter mortality as a possible contributing factor to declines and found that a minimum size of about 75mm was required for overwinter survival. As age-0 growth has not shown a change since before the decline, changes in winter survival or predation may have contributed to a change in overwinter survival.

Lake Michigan Yellow Perch: On the Road to Recovery? (continued by Rajat Chakraborti, SUNY at Buffalo, rkc@acsu.buffalo.edu )

Various studies pointed out that the yellow perch ( Perca flavescens ) population in the Great Lakes and in the surrounding lakes has decreased at an alarming rate, resulting in various changes in the ecosystem, and affecting the commercial and sport fishing industry. The state of the art research goal was to evaluate size- and age-specific dynamics of growth rate (due to net effect of growth and mortality) as a function of hydrodynamic and limnological conditions, prey density, and predator size.

Various modeling approaches, aided by field measurements, have been presented where simulation models were used as tools to address such contemporary intrinsic issues on yellow perch population dynamics. From all the presentations in this session, it is evident that the decline in population of yellow perch is due to high abundance of alewife ( Alosa pseudoharengus ) that may be out competing yellow perch for food and space, decrease in phosphorus loading, increase in piscivorous cormorants, decline in Diporeia populations, and high levels of harvesting. In addition, research findings pointed out that the decline may be due to an increase in water transparency caused by the zebra mussel ( Dreissena polymorpha ) invasion, which may have altered patterns of growth and mortality of yellow perch. Moreover, the significance of compensatory growth, mortality, and reproduction may have been influenced by changes in walleye forage (mayfly, white perch, gizzard shad) that controls the yellow perch population dynamics. Field data on cormorant predation on perch population has been cited as another major factor on their decline. Interestingly, adapting to piscivory at a comparatively younger age of yellow perch has been a concern of recent changes in the lake’s ecosystem.

The findings from this research suggest that a reduction in zooplankton production and biomass, increase in benthic production and biomass, and an increase in macrophytes might have impacted the perch recruitment. As a predictive tool for ecosystem changes, models can be very useful for assessing compensatory growth of perch and walleye, since these are coupled in an aquatic system. Good quality field data are critical for model accuracy. Water level fluctuations might have impacted yellow perch habitat, which is an important emerging issue that needs more research.

The other emerging issues are:

1. Can yellow perch growth compensate mortality at low abundance? and,
   
2. Should we try to manage for a more optimal population size and density?

It is required to develop a stock-recruitment model of yellow perch population by identifying growth-limiting factors and addressing recruitment problems resulting from over harvesting of the single year classes, and possible changes in growth, fecundity and spawning habits. This may lead to a control of commercial and sport fishing both in the Great Lakes and other lakes as a part of the lakewide management program, by imposing limits on the quantity of sport catch above certain size, and by restricting the catching season. There is a greater need to harmonize basin-wide harvest regulations.

Evaluating Ecological Integrity of Benthic Invertebrate Communities in The Great Lakes (by Alison Fraser, Trent University, afraser@trentu.ca )

This first part of this session focused on the assessment of the integrity of benthic communities, including an evaluation of the methods currently used to do so, as well as improvements/extensions to these methods. One limitation often experienced when conducting such an assessment is the unavailability of pre-disturbance data. This was addressed by L.C. Grapentine who suggests using a reference condition approach which involves the use of predictive, multiple regression type models. These types of models can then be extended in order to assess the change in biota in response to degradation or rehabilitation as presented by R.H. Norris. This approach is extremely useful from a managerial perspective and it was found that the most useful predictor variables include total nitrogen and conductivity. A suggestion of incorporating a ‘time for recovery component’ into the method was offered by the audience. Another dilemma often encountered in bioassessments is the inclusion of rare species. T.B. Reynoldson addressed this through the examination of a sequential elimination of rare species from such assessments. It was concluded that rare species are not important for community analyses if multivariate methods are used and each assemblage must be examined to determine the appropriate degree of species removal. However, they must be included if univariate methods are used. As well, rare species are critical if concerns relate to conservation or biodiversity. An additional dilemma encountered in the sampling of benthic invertebrates is the difficulty associated with the collection of these organisms in stony littoral zones of Lakes. M. Moertl addressed this issue and presented research concerning the development of two types of underwater suction samplers:

1. a modified surber sampler which can be used in shallow areas ( < 1m) and
2. an underwater suction sampler which can be used in greater water depths with the use of scuba divers.

Research concerning the composition of specific benthic invertebrate communities in various aquatic systems was also discussed in this session. T.G. Horvath presented results from an assessment of the benthic invertebrate community from the nearshore sands of southern Lake Michigan. It was found that this community was dominated by oligochaetes and nematodes, and there was an occurrence of 2 new harpaticoids with invasive species dominating the harpaticoid community. D.E. Dittman presented results on work conducted to quantify and compare the spatial and temporal composition of macroinvertebrates in Southern Lake Ontario. There have been considerable declines in amphipods, clams and oligochaetes, however dreissenid mussels have greatly increased in numbers (~100 fold) since 1992. M.A. Quigly examined the abundance and life-stage composition of Diporeia spp. in Lake Michigan. Of particular interest is the fact that despite investing energy into transforming to the adult phase, these organisms tend to move into the water column, where they are subject to greater predation.

L.D. Corkum presented work conducted on using abiotic factors to predict the mass emergence of Hexagenia adults in Lake Erie. This work is of importance as although these organisms are important prey items for fish and birds, they can become a great nuisance as their peak densities can reach 24 740 adults/m2. It was concluded that using wind chill allows for the prediction of the emergence of both adult males and females.

The last section of this session focused on the presence of zebra mussels in aquatic systems. A.L. Fettes presented work on the patchy distribution of zebra mussels in lower Green Bay and it was hypothesized that substrate is a limiting factor for the presence of these organisms. Further to this, S.J. Wielgus discussed the greater abundance and diversity of benthic macroinvertebrates within zebra mussel aggregations (druses) compared to the surrounding sediment in lower Green Bay. As well there was a greater abundance of benthic macroinvertebrates in sediments containing zebra mussels. The druses contained high numbers of amphipods, and it was suggested by the audience that echinogammarus be examined in the study. The ecological impact of zebra mussels was then addressed by M.T. Bur who examined the effects on the diets of both yellow and white perch in Lake Erie. The zebra mussel component of the diet of these fish has increased, however, it is also limited by the difficulty involved in ingesting the hard shells of these organisms.

Lake Huron fish community and environmental objectives (by Brian Ellrott, University of Vermont, bellrott@zoo.uvm.edu )

The Lake Huron Initiative (LHI) was established by local, state, provincial, and federal agencies to restore and maintain the chemical, physical, and biological integrity of the waters, tributaries, and nearshore terrestrial and aquatic ecosystems of Lake Huron.

To accomplish this goal efforts are directed toward two key issues:

1. critical pollutants/use impairments and
2. fish and wildlife habitat and biodiversity.

J.H. Bredin stated that although chemical contaminant concentrations are a problem as indicated by fish consumption advisories, loss of habitat is the primary ecological concern on Lake Huron. Near-shore and tributary habitats are being lost and/or degraded by the effects of eutrophication, harmful exotic species, persistent chemical inputs, and impassable dams.

D.G. Fielder et al. reported that walleye populations are low in some locations in part due to dams and other impediments to riverine spawning habitat. Similarly, L.C. Mohr and J.R. McClain stated that lake sturgeon ( Acipenser fulvescens ) numbers are extremely low and that successful recovery depends on the improvement of access to traditional spawning and juvenile habitats. Another concern is the incidental mortality of young lake sturgeon when streams are treated with TFM for sea lamprey ( Petromyzon marinus ) control.

T.F. Nalepa et al. examined the recent trends in benthic macroinvertebrates in Lake Huron. Similar to other large lakes the benthic amphipod Diporeia population has declined as zebra mussel colonization has expanded. At a deep site in Saginaw Bay, Diporeia began to decline one year after zebra mussels became abundant in shallow areas. The cause of this apparent relationship is unknown.

High levels of mortality of native lake trout ( Salvelinus namaycush ) beginning in the 1940s drove populations close to extinction in the main basin by 1966. Over-harvest and sea lamprey parasitism were the causes of this population crash and continue to be major impediments of lake trout restoration efforts today. Statistical-catch-at-age (SCAA) modeling of lake trout mortality by A.P. Woldt et al. showed that commercial fishing and lamprey account for most mortality in northern Lake Huron while recreational fishing and lamprey are the major sources in the southern section.

T.J. Morse et al. presented a review of sea lamprey control in Lake Huron. The sea lamprey control program was initiated on a lake wide basis in 1970 and was successful in reducing the original population of sea lampreys by almost 85%. However, the uncontrolled St. Marys River population allowed sea lamprey populations in the lake to increase in the 1990s. D.M. Reid and J.E. Johnson indicated that progress toward lake trout rehabilitation has been greatly limited by the abundance of parasitic sea lamprey entering the lake from the St. Marys River. Treatment of the St. Marys River in 1998 and 1999 resulted in a 45% percent reduction of the larval sea lamprey population and is expected to significantly reduce lamprey induced lake trout mortality.

M.P. Ebener presented a review of commercial fishing in Lake Huron and indicated that fishing power has changed in the last 20 years. Improvements in tracking ability, communication, and gear have increased fishing efficiency. The use of taller gill nets has increased net catch-ability for targeted lake whitefish ( Coregonus clupeaformis ) and also incidental lake trout. M.P. Ebener and L. Mohr reported that the coregonine community was healthy and most objectives were being met.

Lake Huron fish community and environmental objectives (CONTINUED by Norine Dobiesz, Michigan State University, dobieszn@pilot.msu.edu )

Papers in this session addressed the status of the Lake Huron fish community and ecosystem by bringing together new information on the lake and the processes involved in its dynamics. The focus of the session was to compare the current status with objectives established by the Great Lakes Fishery Commission's Lake Huron Committee (in 1992) and the Lake Huron Initiative (recently published). The afternoon session focused on trends in the chinook salmon population, the status of the prey fish community, estimates of consumption by the key predators, and an evaluation of the Fish Community Objectives (FCO) for Lake Huron. The FCO’s set guidelines for individual species. Given the loss of habitat and invasions by exotic species there is concern that the current fish community can not be rehabilitated to sustain the level of harvest in the FCO’s (J.R. Schrouder et al.).

Although the average chinook salmon catch is still below the FCO guidelines, the rate has increased during the last half of the 1990’s. This increase was not been driven by increases in stocking but is attributed to the recent hatchery use of pen culture, shown to increase the survival of planted fish. There are no estimates of wild production of chinook salmon although a study is now underway to help estimate this. Chinook salmon condition, as represented by the weight-length relationship, was higher in the 1970s and lower in the 1990s. Such data suggests that chinook salmon were stressed during the 1990s raising concerns of a bacterial kidney disease (BKD) outbreak as seen in Lake Michigan. However, both catch rates and weir examined fish show no evidence of BKD (J. Johnson et al.)

The prey fish community consists of a diverse array of species however alewife, rainbow smelt, and bloater dominate the biomass. Bloater biomass has also shown cyclic tendencies. Since 1990 alewife and rainbow smelt biomass has declined while prey size structure has shown a trend toward smaller fish. This may indicate strong predatory demand and points to the need for continued monitoring of the prey fish community (G.L. Curtis et al.).

Estimated consumption by the key predators (chinook salmon, lake trout, walleye, and burbot) in the open waters of Lake Huron has been increasing in recent years. Nearly 77% of the estimated consumption in the main basin is attributed to chinook salmon and lake trout. The primary forage fish consumed are alewife and rainbow smelt, comprising 81% of the consumption by the key predators. Consumption by lake trout in Georgian Bay and the North Channel rivals consumption by lake trout in the main basin. Double-crested cormorants also play a significant predatory role in Georgian Bay and the North Channel. Further study is needed to reduce areas of uncertainty in the models used to estimate consumption (N.E. Dobiesz and J.R. Bence).

GIS Applications for Great Lakes Coastal Management (by Jagjit Kaur, SUNY at Buffalo, jkaur@acsu.buffalo.edu )
In this session, the applications of GIS in the Great Lakes’ watershed were discussed. There were altogether three talks, which were completely different from each other with regard to the use of GIS in dealing with various contemporary environmental problems.

They were on the use of GIS for:

1. tracking the contaminant plumes in surface waters;
2. developing a spatially explicit habitat-based population model; and
3. analyzing watershed characteristics for determining metal speciation.

In dealing with the problem of contamination in waters of the Buffalo and Niagara River basins, GIS was used to visualize the plumes of dissolved oxygen, temperature, and turbidity. The study suggested that the method could be used to develop a model of contaminant transport out of the Buffalo River and into adjacent receiving waters. The study results reported that Buffalo River is a source of high temperature and turbidity, but the impact is still in question. Future research needs to focus on other watershed problems such as the role of Scajaquada Creek as a source of pollutants. The developed technique could also be used for tracking contaminant plumes in streams and rivers with similar watershed characteristics.

Research priorities include:

1. tracking and assessing the contamination in the Buffalo and Niagara Rivers (the Areas of Concern), and
2. evaluating the impact of temperature, dissolved oxygen, and turbidity on aquatic life.

A spatially explicit habitat-based population model developed by using remote sensing, field data, and GIS coverages on disturbance, macrophyte cover, and substrate type and depth, was discussed. The model incorporated the physical and biological factors that limit space and resources. Priority activities include the simulation and testing of habitat-based modeling framework under different scenarios such as: climate change and habitat loss/modifications, which are recent important issues needing attention. This modeling work could be used in fish conservation programs, such as in assessing the emerging issue of decline in yellow perch population in the Great Lakes. The information could be used for providing guidance on protecting and restoring habitats for fish populations, which is a priority issue in the Great Lakes. Another study focused on the emerging issue of metal speciation in aquatic systems. The idea of employing dissolved fraction of metals in Aquatic Life Criteria (ALC) instead of using total recoverable metal is still in discussion. However, the study suggested that the permits must still be expressed as total recoverable metals. The capabilities of GIS were explored to determine the metal speciation as influenced by watershed characteristics. The study assessed whether watershed characteristics can be used to predict chemical controls of trace metal levels and speciation, levels of total, filterable and truly dissolved metals, and metal-particle partitioning. The use of multivariate models and their effectiveness at various spatial scales were examined. The associated priorities include the determination of metal speciation in aquatics, which is important in setting permits and other regulations.

Can we stop them? Preventing introductions and controlling the spread of nonindigenous species in the Great Lakes (by Todd Morris, Erindale College, U of Toronto, tmorris@credit.utm.utoronto.ca )

Johengen et al reported on the role of ballast water in the transport of nonindigenous species. Since 1970 75% of introductions have resulted from shipping practices. Although ballast water exchange programs have been initiated requiring a salinity of 30ppt most vessels in the Great Lakes are NOBOB (NO Ballast On Board) vessels and are unregulated. Most vessels still contain residual ballast water and sediment in the tanks, which may be discharged when these ships take on ballast in the Great Lakes. 33% of Ballast on Board (BOB) vessels still had live animals even after ballast exchange and large numbers of viable resting eggs have been found in sediment samples from ballast tanks.

A new study being initiated to:

1. characterize the communities in ballast tanks on NOBOB vessels,
2. determine if biota are released from these vessels and
3. test effectiveness of open-ocean exchange.

Chakraborti et al. and Drake et al. presented results of modeling attempts to predict and describe invasion success. Chakraboti et al. used a GIS based multivariate approach to examine likelihood of successful invasion of zebra mussels in Saginaw Bay. They modeled growth potential instead of the usual density variable and found that chlorophyll a was the most important predictive variable. Their model showed good agreement with actual mussel growth in the bay. Recognizing that most invasions are initiated by only a few colonizers, Drake et al. examined the role of stochastic events in determining invasion success (using Ruffe as a case study). Their results suggest that a high likelihood of success can be obtained with even a very small number of initial invaders ( <5). The inclusion of Allee effects results in substantial reduction in success.

The remaining presenters looked at population dynamics of pre-existing invaders. T. P. Diggins et al. proposed that interference competition for space and niche partitioning of settling substrata might be combining to allow expansion of quagga mussels and exclusion of zebra mussels from all regions of the lower Great Lakes. They reported that zebra mussels have been found attached to macrophytes in high proportions (> 70% zebra mussels) in areas where benthic numbers are 100% quagga mussels. They propose macrophytes as a potential refuge for zebra mussels. R. K. Wolfe reported on nest selection in Round Gobies. Through a presentation of artificial nest structures he determined that round gobies prefer large nests with single openings.

K. O. Rothhaupt reminded us that invasions are not a North American phenomenon through a brief history of invasions in Lake Constance where at least 10 invasions (6 invertebrates, 4 fish) have occurred in the 20 ^th century. Invasion of Cyclops vicinus in the course of eutrophication has caused the extinction of a native copepod, zebra mussels (introduced by boat traffic) have caused a 10-fold increase in waterfowl, rainbow trout are outcompeting native trout and Ruffe are adversely affecting native perch and whitefish.

G. Zimmerman took us out of the water and onto the land reporting on biocontrol of Purple Loosestrife with Galerucella beetles. Initial control showed a time lag of about 2 years after which time dramatic declines in Purple Loostrife were seen. Open spaces created by the loss of Purple Loosestrife were recolonized by a mixture of native plants and Purple Loosestrife.

Air pollution and policy (by Donald Hughes, SUNY-Environmental Science and Forestry, djhughes@mailbox.syr.edu )

Seung-Muk Yi (Clarkson U.) discussed deposition of trace metals in two Korean reservoirs. The goal of her research was to quantify dry deposition fluxes and identify sources. A major source was "yellow sand events", i.e. times when high winds transported fine sands from arid regions of China. During these events deposition rates increased dramatically. These sands also brought in high levels of Cr, Pb, V with fluxes far exceeding those seen in the Great Lakes. To identify other sources, PSCF (potential source contribution function) methods were used.

Tom Holson (Clarkson U. too) also applied PSCF methods to identify sources of air pollution, in this case PCBs in Chicago. The PSCF pointed to the northwest and southwest of the city. Possible sources were identified: sludge drying beds, a transformer storage yard, and a landfill. Estimated loading rates from these sources ranged up to 0.2 kg/day; however one would expect 2-20 kg/day to explain the observed PCB concentrations in Chicago's air.

A. Peck (U. of Iowa) used an experimental chamber (30 m3) to measure plant uptake of anthracene. He observed air concentrations of 4-9 ng/m3 which matched model predictions. However, there were diurnal variations which could not be explained--anthracene concentrations in the air fell during both light and dark periods, but were expected to rise during dark periods.

Ron Hites (Indiana U.) posed the question: Why have PCB concentrations in the air not decreased as they have for other environmental compartments? Data from Bloomington, IN, for example, showed little long-term trend. PCB concentrations in Chicago range from 1000-2000 pg/m3, and from 70-200 pg/m3 in other locations around the Great Lakes (IADN data). He demonstrated that temperature can explain 50% of the variability in the data. Even with temperature correction, no significant trends emerge from the IADN data. But if older data 1970-1995) is included a significant downward trend is found, indicating a PCB half-life of about 7.5 years for Bloomington. A similar result was seen with Sleeping Bear Dunes data (half-life ~5-6 years).

Steve Eisenreich (Rutgers U.) an atmospheric deposition monitoring network comprising 9 stations in New Jersey (NJADN). Like the IADN, samples are collected every 12 days and analyzed for a number of organic compounds and trace metals. Eisenreich focussed on chlordane, a mixture of > 100 compounds that were banned in the US in 1988; and in Canada in 1990. Back-trajectory models showed clearly that chlordane was mostly emanating from the southern US. Fingerprinting of cis:trans ratios corroborated this conclusion. In contrast, chlordane in Chicago seemed to be stemming from local sources.

Contaminants in Wildlife (by Jeanette Jaskula, U of Wisconsin at Green Bay, jaskjm25@uwgb.edu )

As the human population continues to grow, pollution threatens the integrity of our waterways and the creatures that depend on a healthy environment. This session focused on assessing the impacts of effluent from refineries and wastewater treatment plants on rainbow trout and determining if eagles are an effective indicator of persistent toxicity in the Great Lakes.

The presence of the egg protein, vitellogenin (Vg), is accepted as an indicator of exposure to environmental estrogens. Exposure of this kind has resulted in the feminization of male fish and reduced reproductive success. To determine if effluent is responsible for Vg induction, rainbow trout and their eggs were caged close to the point of discharge from two refineries, a wastewater treatment plant, and an upstream reference site in Ontario. After fifteen days the trout were assessed for the presence of Vg. There was no evidence of Vg induction in these samples; however, the researchers indicated that analytical tool inadequacies might have produced a false negative result. Research continues in Canada to determine if water pollution actually causes Vg induction resulting from estrogen. If, in fact, there is a link, it is impossible to say at this point how this could affect entire ecosystems and humanity.

Because the bald eagle is a tertiary predator in aquatic ecosystems, it is possible they can be used as indicators of persistent toxicity in the Great Lakes Basin. Blood and feather samples were taken from eagles in six areas around the Great Lakes between 1999 and 2000. Study sites in Michigan included the Upper and Lower Peninsulas, Lake Michigan, Lake Heron, and Lake Superior. The last site was located in Voyageurs National Park in Minnesota. Concentrations of mercury and PCBs were compared to a dataset from 1985-1989. There was no significant difference in mercury concentrations between sites in 1999-2000. Mercury concentrations were significantly lower in only Lake Michigan and Lake Huron compared to samples from the 1985-1989 dataset. Concentrations of PCBs have declined among most subpopulations. Bald eagles appear to be good sentinel species to monitor concentrations of toxins within the Great Lakes Basin ecosystem.

Food Webs and Fisheries in the Great Lakes (by Rajat Chakraborti, SUNY at Buffalo, rkc@acsu.buffalo.edu )

This session synthesized the dynamism in the lower food web, including the impacts of exotic species. The emerging issues covered a wide range of topics, including phosphorus loading and its impacts, phytoplankton and zooplankton production, zebra mussel effects (Dreissena polymorpha) on the food web, and contaminant transport in fish and terrestrial species. Primarily, the focus of the research presented was on the use of modeling frameworks and field measurements to evaluate lower food web dynamics, including phytoplankton, zooplankton, and zebra mussel species.

In this session, particular attention was drawn to the studies of zooplankton community. Recent decline in Diporeia spp., along with an increase in alewife production, both in the Great Lakes and other lakes have raised a vast interest towards zooplankton specific research. It has been found that the introduction of zebra mussels has impacted negatively on zooplankton productivity and biomass, by causing a decline in food availability and creating a competition for food. The overwhelming feeding capacity of dreissinid mussels may have been a major factor in the reduction of Diporeia populations. Mussels are out-competing many pelagic and benthic species by drawing algae that is essential for their survival. The impacts of zebra mussels on the fate and transport of PCBs have been studied in a coupled ecosystem level modeling framework, which has demonstrated a transfer of food and energy from the pelagic food web to the benthic food web. That research revealed the need to conduct field studies of PCBs in lower food web species, particularly in phytoplankton and zooplankton, after the zebra mussel invasion. It also emphasized a need for conducting studies on the impact of zebra mussels on contaminant transport in higher trophic level species, by including interactions and feedback between pelagic and benthic food chains.

In most of the studies, zebra mussels have been indicated as a major source of various ecological changes. This session pointed out the following future research priorities, and provided some suggestions to enhance the understanding of the interactions among species in the food web:

• Need for improved methods for assessing Dreissena distribution and biomass,
• Need for long term monitoring of the Great Lakes benthic community,
• Need for better understanding of trophic food web interactions from diet studies (all trophic levels),
• Conduct isotope experiment or field observation, and
• Conduct bioenergetics measurements.

The focus of this session can be broadly divided into three different categories of environmental problems:

1. decline in amphipod Diporeia population, and the associated changes in the Great Lakes’ ecosystem;
   
2. impact of round goby in the food web dynamics; and
   
3. the effect of prey’s (Bythotrephes) body structure (e.g. spine) in predator’s (Rainbow smelt) body.

The first group of studies indicated that the decline in Diporeia species might be due to competition for food caused by the zebra mussel (Dreissena polymorpha) invasion in the Great Lakes and the surrounding shallow lakes. Other major factors for this decline include possible pathogens such as a disease, fungus and bacteria. The decline in Diporeia populations has resulted in the change of diet choice of forage fish (bloater, alewife, rainbow smelt, slimy sculpin, deepwater sculpin) towards Mysis relicta, other zooplankton, and chironomids. Field and laboratory studies addressed this problem from various perspectives. In one interesting study, feeding mechanisms, success or failure in foraging, and the energetic costs of diet shifts from Diporeia to Mysis have been documented.

The focus of another set of presentations was on the diet choice of an exotic species, Round goby (Neogobius melanostomus) and its impact on the food web. The laboratory and field studies suggest the variability in predation by round gobies under variable limnological conditions. In a mixed diet, the effect of the presence of zebra mussels on the feeding habits of round gobies has been studied to evaluate the switching of diet under various scenarios. Preliminary results on a study of predation by higher trophic level species on round gobies showed the future research direction focused on contaminant transport paths through food webs, and bioaccumulation.

Apart from Diporeia and the zebra mussel, the research findings addressed the effect of Bythotrephes in the diet of rainbow smelt. This study pointed towards the effect and appearance of Bythotrephes spines in the predator’s body from the viewpoint of the structure and morphology of spines.

Future research needs to address the question of whether the decline in Diporeia population and abundance is correlated with the zebra mussel introduction. This puzzle is cited as one of the most important contemporary issues that will help in understanding many ecological problems. The suggestion from the researchers was to manage zebra mussels in a way so that Diporeia density in the Great Lakes increases.

There is a need for studies determining contaminant transport in food webs including exotic species such as zebra mussels and round gobies. Particular research priorities include the study of contaminant transport by round gobies by incorporating multiple trophic level species.

A spatially explicit, individual-based simulation model was used to investigate the impacts of habitat changes on steelhead parr in the Manistee River, Michigan. Results suggest that density dependence and changes in discharges had little effect on parr growth and survivorship. However, decreased temperatures lead to decreased mortality and increases in length. Possible management actions include controlling discharges. (J.A. Tyler et al.)

Bioenergetics models are used to define an ecosystem in terms of energy. Parameters representing adult physiology have been applied to all alewife life stages. R.A. Klumb et al. derived empirical response data for larval and young-of-year (YOY) alewife then compared these to the adult responses. Results indicate that the adult parameter values estimated higher respiration rates for YOY and lower respiration rates for juvenile alewife. Since bioenergetics models are used to study many aspects of individual species energy allocation as well as food web connectivity, this study suggests that applying adult parameter values to other life stages may be problematic.

In Lake Opeongo, Ontario it was found that gillnet catchability in the spring is greater than in summer. Transmissivity did not differ between spring and summer and so changes in visibility could not account for differences in catchability. However, seasonal changes in forage behavior resulted in decreased catchability in the summer. Findings suggest that behavior of the target organism is very important to gillnet catchability (T.M. Janoscik et al.).

Lake trout restoration in Lake Champlain was investigated looking at spawning activity across 7 areas that included natural and artificial reefs. Spawning activity was evident at multiple sites throughout the lake. Hatching occurred at several sites where fry were detected, however, fry were more often found at the artificial sites. There was no evidence of EMS detected and recruitment to age 3+ appears to be on the rise. Artificial reefs play an important role in this lake (B.J.Ellrott and J.E. Marsden). However, shallow nearshore areas, where most of the artificial reefs are located, are poor habitat for spawning. Using an ROV to search for lake trout eggs, suitable habitat was found at 10-20m depths. Further sampling of the deep spawning substrate and calibration of the ROV process is needed. Additionally, the lake trout strain in Lake Champlain may have always been a shallow water species, accounting for their success on the generally shallow artificial reefs (J.E. Marsen and B.J.Ellrott).

Advances in the Lake Michigan Mass Balance Study (By Aaron Peck, University of Iowa, apeck@blue.weeg.uiowa.edu )

New field and modeling results from the Lake Michigan Mass Balance (LMMB) Study were presented. Topics ranged from foodweb dynamics to PCB source identification.

A lake-wide toxics model has been developed for Lake Michigan. Three versions of the model are available:

• Level 1 contains 6 surface segments and 9 water segments;
• Level 2 contains 10 surface segments and 41 water segments;
• Level 3 contains 2,318 surface segments and 44,042 water segments.

The necessary inputs for the model include transport, loading, boundary conditions, and initial concentrations (Richardson, W.L., et. al.). A lake-wide model was developed to examine the fate and transport of atrazine in the lake. By examining tributary load ratios over time it was evident that significant runoff of atrazine occurs immediately after application. The greatest loading were from tributaries in southwest Michigan. From the Level 1 MICHTOX model, it appears that atrazine is decaying at one percent or less per year (Rygwelski, K.R., et. al.). The uncalibrated Level 2 Lake Michigan toxics model estimates the organic carbon sorbent dynamics very well (Zhang, X., et. al.).

One important aspect of the LMMB was the identification of contaminant sources. A mass balance model on the Milwaukee Harbor Estuary was developed in order to estimate PCB loadings to Lake Michigan for the Lake Michigan Mass Balance Study. Between 1990 and 1995, the model predicted a total PCB loading to Lake Michigan from the Milwaukee Harbor Estuary of 138 kg (Zhang, X). Chicago is an important atmospheric source of PCBs to Lake Michigan contributing about 330 kg to the total atmospheric load to Lake Michigan (Hornbuckle, K.C., et. al.). A comparison of congener profiles for simultaneous samples taken upwind and downwind from a potential source was used to identify atmospheric PCB sources in Chicago. Atmospheric PCB sources included wastewater treatment facilities, landfills, and a large transformer storage yard. Additionally, K-modeling based on horizontal wind fluctuations for lateral dispersion and vertical dispersion was used to estimate the fluxes from these sources. The total source from Chicago was estimated as 28 kg/day (Hsu, Y.K., et. al.).

Atmospheric sources of PCBs and PAHs have been found to be important sources of these contaminants to Lake Michigan. The atmosphere plays an important role the fate of these contaminants in the lake. The atmospheric data and model used in the Lake Michigan Mass Balance Study indicate that deposition of PCBs to Lake Michigan is highly variable over space and time (Hornbuckle, K.C., et. al.) The annual atmospheric loading of PCBs to Lake Michigan is about 3,200 kg. The air-water exchange fluxes of PCBs estimated for Green Bay (1989-1990) and for southern Lake Michigan (1994-1995) have been previously reported. Because improved Henry’s Law constants are available, the air-water exchange in these areas has been revisited. In addition to using new Henry’s Law constants, the new calculations were based on air temperatures instead of water temperatures. The new net flux estimation for Green Bay is a loss of 1100 ng/m2 day. For Lake Michigan, the net loss was estimated as 27 ng/m2 day. Both of these fluxes are larger than originally reported (Totten, L.A., et. al.). Air-water exchange fluxes of PAHs were estimated for Green Bay and southern Lake Michigan using improved Henry’s Law constants. For Green Bay (1988-1989) variable air concentrations dominated air-water exchange. For southern Lake Michigan (1994-1995) wind direction was important in flux estimates. For south and southwest winds a net flux into the lake was estimated. For north and northeast winds, a net flux out of the lake was estimated. Net depositional fluxes to southern Lake Michigan ranged from 5, 000 20, 000 ng/m2 day (Gigliotti, C.L., et. al.)

The role of foodweb dynamics is important in the fate of toxics Lake Michigan. Dissolved phase, phytoplankton, bulk zooplankton, Mysis, and Diporeia samples were collected from eleven sites over two sampling years for the Lake Michigan Mass Balance Study. The dissolved phase was highest in tri-chlorinated PCBs. Phytoplankton were highest in tetra-chlorinated PCBs. Zooplankton, Mysis, and Diporeia were highest in penta-chlorinated PCBs. The half-life for PCBs in the lake was calculated as 6 years. The bioaccumulation factors for PCBs varied considerably over space and time. Slopes for log BAF vs. log Kow plots were 0.27, 0.36, and 0.6 for phytoplankton, zooplankton, and Mysis, respectively. This trend indicates that organisms higher in the food chain are closer to equilibrium than those lower in the food chain Normalizing PCB concentrations to lipid content does not help resolve spatial or temporal variability. Changing lipid concentrations preclude lipid normalization (Trowbridge, A.G., et. al.).

PCB concentrations were 45 ng/g, 169 ng/g, 253 ng/g, and 417 ng/g dry weight in phytoplankton, zooplankton, Mysis, and Dyporeia, respectively. PCB concentrations in these organisms were not at equilibrium or steady-state with respect to dissolved phase concentrations. Foodweb dynamics were studied by comparing biomagnification factors (BMF) between trophic levels. The Mysis-zooplankton BMF was the lowest and the Diporeia-phytoplankton BMF factor was the highest (Trowbridge, A.G., et. al.).

Most foodweb models use a single spatial compartment and, consequently, a single exposure temperature. Because of the temperature dependence of PCB concentrations in fish it is important for models to differentiate exposure temperatures. Model results show that a multicompartment model can better accommodate migratory species (Zhang, X., et. al.).

Total mercury concentrations in Lake Michigan surficial sediments was 140 ? 38 ng/g. Methyl-mercury concentrations were 0.57 ? 0.34 ng/g. There was more methyl-mercury found in the northern part of the lake than in the southern part and more found in the eastern part than the western part. There were no spatial trends found for total mercury. Additionally, the methyl-mercury concentrations were not correlated with total mercury concentrations (Rossman, R., et. al.).

The original Lake Michigan Mass Balance Study dataset did not include PCB concentrations. Due to numerous requests EPA GLNPO has funded the calculation and addition of PCB values to the dataset. Because of differences in calculation procedures for PCB in different phases, care must be taken when using these numbers (Kuehl, M.A., et. al.).

Remote Sensing and spatial data applications for the Great Lakes (by Christina Clark, Environment Canada, christina.clark@ec.gc.ca )

Gunatilaka et al. (2000) presented some of the automated continuous monitoring systems that are currently available for waterways, and are summarized by Scully et al. and Cowell et al. The Daphnia movement system, for example, sets off an alarm when movement is below a threshold since the number alive indicates the health of the river system. This 24 hour monitoring system can also watch for ‘midnight dumping’. Systems are currently in use in Northern Europe and Asia. After ten years of operation the quality of river water has greatly increased in some areas.

The Great Lakes Coast Watch program announced that data from NOAA’s GOES (Geostationary Operational Environmental Satellites) can now be viewed in an online geographical information system ( http://coastwatch.glerl.noaa.gov/ ). Java GIS is the technology that allows some analysis/interpretation of data online.

Using optical infrared sensors, passive microwave sensors, synthetic aperture radars and scatterometers, Nghiem and Leshkevich are developing methods to monitor ice coverage and its impacts. A unique data set has been collected to ground truth areas along Lake Superior. QuikSCAT launched June 1999 offers 25km resolution, inner and outer swath, 25x35km. They can process to 7km resolution due to Doppler technology. There is a SeaWINDS/ ENVISAT/ AIRSAR partnership to validate lake ice mapping products.

Lesht et al. showed that satellite observations contribute to the quantitative analysis of the Great Lakes chlorophyll levels. Analysis of an unexpected bloom revealed that a four-hour 10m/s+ wind event deepened the thermocline on June 2, 1998 which triggered the early summer phytoplankton bloom.

Stroud et al. presented numerical aspects of combing satellite images, and hydrogeological data to model events backwards or forwards in time. Currently a 5km grid is used, though a 1.1km grid is possible. The model is dynamic, not just an interpolation between known images since it uses satellite images and hydrogeological data.

Plankton in the Great Lakes (by Sarah Wielgus, University of Wisconsin at Green Bay, wielsj12@uwgb.edu )

The impact of ultraviolet radiation (UVR) is of concern in the Great Lakes. UVR penetration is highly impacted by dissolved organic matter (DOM) which has varying optical qualities within a lake system. Consequently, this variation causes the UVR exposure in the water column to be highly variable. In Lake Erie, bacteria appear most susceptible to damage from UVR exposure while phytoplankton and phosphorus cycling are less affected.

Nutrient controls are important and generally result in a decline in the biomass of phytoplankton. However, in Western Lake Erie the establishment of the zebra mussel caused the biomass of phytoplankton to further decline even though phosphorus loading continued to increase. The zebra mussels were causing the biomass to decrease but the number of different species within each type of algae assemblage was increasing. It seems that the phytoplankton standing crop was not limited by nutrients but by zebra mussel grazing in this system. Likewise, in the Severn Sound phytoplankton biomass was lower than predicted by a reduction in nutrient loading. This is also a result of zebra mussel grazing. In addition, the proportion of Bosmina, a representative eutrophic water zooplankton, has noticeably decreased since zebra mussel establishment and reduced phosphorus loading.

Plankton in the Great Lakes ( Continued by Todd Morris, Erindale College, University of Toronto, tmorris@credit.utm.utoronto.ca )

K. E. Seifried reported on changes in macrophytes in the Bay of Quinte area of Lake Ontario. With the recent invasion of zebra mussels into this area and the associated increase in water clarity macrophytes have shown drastic increases. Biomass has increased by up to 300% and the macrophyte beds have extended in to deeper water than has been seen in the last 20 years. Regression analysis (parametric and non-parametric) suggests that light on bottom and current velocity on bottom (both associated with the zebra mussel invasion) are the 2 most important factors in this increase.

S. W. Wilhelm et al (presented by M. R. Twiss) presented evidence to suggest that trace metals are important regulators of photosynthesis in fresh water systems such as Lake Erie. Field experiments revealed that additions of Cd, Zn and Co all caused increases in photosynthetic activity in the picoplankton. These metals may serve as limiting agents in the pelagic epilimnetic waters of Lake Erie.

In recent years temporary episodes of foul smelling drinking water have been reported along the Canadian shores of Lake Ontario. T. Howell et al. have determined that Geosmin is the agent responsible. Field studies in 2000 along with an analysis of historic records indicate that rises in Geosmin levels are associated with nearshore downwelling events and are likely caused by cyanobacteria in the offshore surface water which become entrained during these downwelling events.

T.J. Morris et al presented a new technique describing spatial structure in zooplankton. The technique, a combination of Principal Coordinates analysis and multiple regression, allows analysis of spatial structure at all scales simultaneously and eliminates the need for preprocessing of samples. Using this technique they showed that zooplankton populations in Lake Erie are structured at multiple spatial scales (>1.5 km, <10m) and that this structure varies by season with less predictable structure in the Spring and Fall. Spatial structure also varies by size of organism with large animals showing structure at more scales. Zooplankton patterns overlap with phytoplankton distributions at some scales but not all.

Silica flux is critical for structuring the growth, abundance and composition of diatoms and H. A. Bootsma et al. examined this in Lake Malawi. They found that silica flux was dominated either by river flows or by internal cycling depending on the time of year with maximal upwelling occurring from December to May.

E.A. Blukacz et al. investigated Bythotrepehes population dynamics in two smaller inland lakes (Harp Lake and Lake Simcoe) using an individual based model originally designed for use in Harp Lake in 1995. They found that the model could easily be re-parameterized to replicate seasonal population characteristics in Harp Lake and Lake Simcoe in 1998. However, in order to produce abundance estimates that were realistic more than life history parameters needed to be changed. An adjustment in predation pressure allowing for a period of intense planktivory in the mid-summer period replicated the bimodal dynamics seen in Lake Simcoe.

Advances in Great Lakes Coastal Processes (by Sarah Wielgus, University of Wisconsin at Green Bay, wielsj12@uwgb.edu )

Documenting changes in substrate mobility is important for predicting future movement. Sidescan sonar can be an effective tool for doing this especially when considering large areas of substrate. When sidescan sonar was used to track littoral sediment movement for selected sites in Lake Michigan and Lake Erie it was discovered that even though the substrate area changed by more than 20% per year at two of the four sites the sand bodies were relatively stable within the nearshore zone. It seems that sizeable movement of the sand bodies is restricted to significant storm events. Sidescan sonar proved effective in mapping the bottom substrate and documenting the spatial and temporal variability of the bottom sediment.

Reconstructing the history of dune growth and migration can be used to help better understand the processes involved in dune development and the temporal scale over which these processes tend to occur. Carbon dating has proved effective in predicting the initiation of dune growth at which point the burial of the basal soil occurred. Further dune growth is marked by the formation of a lower inceptisol series. The development of a spodosol indicates a period of general dune stability. If an upper inceptisol forms this is an indication of a period of dune remobilization. However, there appears no correlation between dune growth and the rising or lowering of lake levels during dune development along the southeastern shore of Lake Michigan.

Predicting future lake shore erosion in the Great Lakes basin may be the first step in developing strategies for reducing lake shore erosion, which consequently reduces potential damages attributed to this erosion over time. Extreme spatial and temporal variability in lake shore erosion for Lake Michigan required current methods of calculating average annual recession rates to be standardized. A standard method was recommended for calculating annual erosion rates with an emphasis placed on the physical processes that are ultimately causing the variability in the calculated erosion rates. It was recommended to consider a minimum temporal scale of 50 years, utilize 3-D analysis, calculate erosion rates for both the top and bottom of the bluff slope, and take measurements along multiple transects. Likewise it was found that there was a strong correlation between wave impact heights and shoreline recession rates along two shoreline reaches of Lake Michigan. However, no correlation was observed between recession rates and precipitation amounts or storm activity at the time scales considered.

Previous river shoreline stabilization practices strongly utilized hard infrastructure, including concrete and steel, to protect developments along a water way from erosion. More recently soft engineering practices have been used to achieve the same purpose as hard engineering structures. However, soft engineering structures utilize rocks, vegetation, and other materials for shoreline stability at a much lower cost. Waterfront redevelopment projects could use soft engineering to stabilize shorelines. Several demonstration projects have shown that soft engineering is as effective as hard engineering and it gives the waterway much more aesthetic appeal. A similar approach was taken at Presque Isle State Park along the shoreline of Lake Erie. Shoreline erosion within the back bay area of the park was in close proximity to the park’s heavily used multi-purpose trail. Natural vegetation, downed trees, sand dredged from a local sandbar, and riprap were used to stabilize the shoreline and save the trail at a minimal cost.

Food webs and fisheries in the Great Lakes (by Norine Dobiesz, Michigan State University, dobieszn@pilot.msu.edu )

A spatially explicit, individual-based simulation model was used to investigate the impacts of habitat changes on steelhead parr in the Manistee River, Michigan. Results suggest that density dependence and changes in discharges had little effect on parr growth and survivorship. However, decreased temperatures lead to decreased mortality and increases in length. Possible management actions include controlling discharges. (J.A. Tyler et al.)

Bioenergetics models are used to define an ecosystem in terms of energy. Parameters representing adult physiology have been applied to all alewife life stages. R.A. Klumb et al. derived empirical response data for larval and young-of-year (YOY) alewife then compared these to the adult responses. Results indicate that the adult parameter values estimated higher respiration rates for YOY and lower respiration rates for juvenile alewife. Since bioenergetics models are used to study many aspects of individual species energy allocation as well as food web connectivity, this study suggests that applying adult parameter values to other life stages may be problematic.

In Lake Opeongo, Ontario it was found that gillnet catchability in the spring is greater than in summer. Transmissivity did not differ between spring and summer and so changes in visibility could not account for differences in catchability. However, seasonal changes in forage behavior resulted in decreased catchability in the summer. Findings suggest that behavior of the target organism is very important to gillnet catchability (T.M. Janoscik et al.).

Lake trout restoration in Lake Champlain was investigated looking at spawning activity across 7 areas that included natural and artificial reefs. Spawning activity was evident at multiple sites throughout the lake. Hatching occurred at several sites where fry were detected, however, fry were more often found at the article sites. There was no evidence of EMS detected and recruitment to age 3+ appears to be on the rise. Artificial reefs play an important role in this lake (B.J.Ellrott and J.E. Marsden). However, shallow nearshore areas, where most of the artificial reefs are located, are poor habitat for spawning. Using an ROV to search for lake trout eggs, suitable habitat was found at 10-20m depths. Further sampling of the deep spawning substrate and calibration of the ROV process is needed. Additionally, the lake trout strain in Lake Champlain may have always been a shallow water species, accounting for their success on the generally shallow artificial reefs (J.E. Marsen and B.J.Ellrott).

Remediation of Contaminated Sediments: Evaluation of Remediation Methods, Effectiveness and Transport to the Environment (Alison Fraser, Trent University, afraser@trentu.ca )

In a brief pre-session discussion, a general overview of sediment remediation was given. P. Keillor discussed remediation options which are available including

1. natural recovery,
2. in place capping,
3. in place treatment,
4. relocation, and
5. ex situ treatment.

Furthermore, J.J. Steuer emphasized that there is a need to consider the mass balance approach to remediation efforts which shifts the focus from concentration to mass in order to get a full perspective on remediation activities. However, D.B. McLaughlin, underscored the importance of understanding and including concentration of contaminants in environmental compartments from a human health risk assessment perspective. Therefore, there is a need for not only an understanding of the mass but also the concentration in remedial activities. Some debate ensued concerning the mass balance versus concentration approach.

Polychlorinated biphenyls (PCBs) are of major concern with respect to sediment remediation and advancements in this area were presented. T.E. Myers discussed recent developments in the bioremediation of PCBs in dredged material. This work is of major importance as confined disposal facilities (CDF) are filling up. To date, research has been conducted at 3 treatment scales with 3 different sediments using 2 biotechnologies and PCB disappearance ranged from 0-70%. Another approach to treating PCB contamination was given by J.E. Duffy who described a technique using wet air oxidation. It is hoped that the method can be modified to extract heavy metal contaminants.

A section of the session focused on the contaminated sediment of the Lower Fox River. To begin with, D.B. McLaughlin discussed remediation strategies for sediments in large rivers, with the Lower Fox River as an example. These areas are problematic in that they often contain large amounts of sediment and remediation strategies occur over a longer timescale. It was stressed that risk is a function of exposure and because exposure is related to contaminant concentration in environmental compartments, there is a need to fully understand changes in the contaminant concentrations. As well, J.J. Steuer applied the mass balance approach to the Fox River and concluded that this approach is scientifically defensible. Research concerning the estimates of PCB mass removal due to remediation efforts from the Lower Fox River was presented by S.L. Laszewski. Pre-project PCB in-situ sediment mass estimates were compared to post-project PCB mass derived from analytical measurements from the on-shore sediment dewatering site and the two estimates were not in total agreement. On-shore measurements were considered superior due to the intensive on-shore sampling measurements. The concern for river turbidity was outlined by J.B. Hutchinson for the Fox River. Dredging may cause considerable sediment resuspension followed by downstream transport. Using turbidity curtains, it was concluded that it is possible to minimize sediment loss and river turbidity monitors provide real time data which can be used to statistically determine measured effects.

Another system which has been studied with respect to sediment contamination is the Minnesota Slip as presented by J.L. Crane. Using invertebrate sediment toxicity tests for a suite of chemicals including PAHs and mercury, the degree of sediment contamination in the area was assessed. Not only can invertebrates be used to assess sediment contamination via toxicity tests, they can also be used as biomonitors for habitat recovery as discussed by L.E.J. Lee. More specifically, gastropods were used to assess the long-term impacts at a controlled oil spill and the effectiveness of bioremediation treatments.

The use of contaminant transport and fate models in assessing the remediation of sediments was presented by H.P. Homberg. Models can be used to

1. integrate and synthesize multimedia data,
2. eliminate the “black box” in order to achieve a better system understanding,
3. identify data gaps or inconsistencies in data,
4. test hypotheses, and
5. combine effects

of natural processes and contaminant sources to reproduce observable trends.

Another issue that was raised concerning sediment contamination is the importance of the organic carbon content of the media. M. Tomczak compared two methods (TOC and LOI) for determining the organic carbon content using sediments from the Detroit River. It was concluded that LOI is unbiased compared to the TOC method and can serve as a TOC estimate since it is both simpler and cheaper.

The State of Health in Lake Michigan: Ecosystem Health Today and Tomorrow (by Vicki Medland, University of Wisconsin at Green Bay, MedlandV@uwgb.edu )

Madejian stated that chinook and rainbow trout have substantial natural reproduction (~25% but needs to be confirmed) and will probably be a permanent part of the fish community. Yellow perch populations started decreasing in the north basin and that has continued in the south. Population decline is due to a suite of factors including low spring warming rates, fisheries selecting large females, zooplankton abundance decline in early summers. Rutherford added that lake circulation moves fry east into Diporeia-poor areas where they starve.

Fitzsimons et al. evaluated spawning activity by lake trout in Northeastern Lake Michigan. Data suggest that gobies may be offshore by the time trout spawn since they were not caught in nets and so might not be important egg predators, but that rusty crayfish and sculpins are present and might be.

Jude and Janssen reported that round gobies are expanding to deeper water (30 m) and that they have just arrived in Green Bay. That they move around so well because they eat zebra mussels, small ones crushed, large eaten whole and egested. Some concern public might think to use gobies as a zebra mussel control and start introducing them in other places. Their best predators are bass, bowfin, which don’t occur in L. Michigan.

Lavis et al. indicated that St. Marys River is a “hot zone” for sea lamprey introduction into Huron/Michigan and needs to be the focus for lamprey reduction.

Warren and Horvatin: Lake Michigan Mass Balance study indicated that Lake Michigan is losing mercury by volatilization and burial. Used to think of the lake as a source, but now (see other sessions on toxics) many think of it as a sink, but this is being debated.

Jude and Janssen also mentioned there are now 2 types of herniations found on zooplankton, one that has already been described as an Ellobiobsidae (sp?) parasite, but also a new not yet described type.

Palmer of Warren and Palmer: A recent history of Lake Michigan nutrients show that chloride levels are rising more than they had predicted in Lake Michigan, probably from road salts.

Fitness of Fish Populations in Large Lakes of the World (by Piet Verburg, University of Waterloo, pverburg@scimail.uwaterloo.ca )

Factors such as environmental pressure, climate change and species introductions affect the evolutionary fitness of fish species in Large Lakes around the world. Fitness is initially determined by the adaptive and reproductive success of ancestors under past conditions. Fitness of fish populations is a concept of interest to the management of fish populations in lakes and past and extrapolated prospective change in fitness merits examination. Fish in lakes generally evolved from species in large river systems, which are much less ephemeral on geological time scales than the lakes which they form. However, Great Lakes supplied a vastly different environment and ecological conditions for fish species to adapt to and exploit, compared with the riverine habitats of their ancestors. Though there are large trophic similarities between pelagic fish communities in large tropical and large northern lakes, they evolved from different taxa and have developed different biological properties in response to their environment. Those in northern lakes typically have seasonally varying growth rates, relatively high fecundity, and varying age and size at maturity. In the warm tropical lakes seasonality is near absent, fecundity lower, but reproduction continuous, and age and size at maturity is relatively constant.

Upon formation of the lakes, adaptation of founding fish species to their new lacustrine environment transpired through natural selection. After having to evolve and adapt in a relatively short time, and restricted to taxa available in the catchment before the formation of the lake, the relative fitness of lacustrine species may often be suboptimal even under low environmental stress. Now, with modern environmental conditions changing rapidly in many lake systems, the natural selection process is affected, and native fish species become less well adapted to their lacustrine environment. The loss of competitive edge is often too rapid for evolutionary adaptation and leads to major shifts in the fauna composition of several Great lakes. Introduced species can compete for resources and habitat and alter fitness of native species. Zebra mussels for instance in the North American Great Lakes direct energy flow from the pelagic to the littoral, with complex consequences for the native fish species.

The East African Great Lakes are each isolated systems, each containing generally hundreds of endemic fish species. Especially cichlids adapted well to the lacustrine habitat in East Africa, but species flocks also originated in other fish families. Evolution and radiation in fish species was explosive in these lakes, with many specialists filling the new trophic possibilities. The relatively young Laurentian Great Lakes are interconnected by navigable waterways. The number of endemics in the East African Great Lakes is often over 80 %, while in the Laurentian Great Lakes, with a lower total number of fish species (159), endemicity is only 3 %. Species introductions have been relatively numerous in the Laurentian Great Lakes, with 23 exotic fish species, compared with the African lakes. Like the African Great lakes, also the old and isolated lake Baikal in Siberia, where the sculpins radiated into a species flock, has a high degree of endemicity (60 %) and 6 introduced fish species. Though introduced species numbers are low in the African lakes, the introduction of the Nile perch in Lake Victoria has had a dramatic influence. The Nile perch became very abundant in little more than a decade and decimated native fish species. Several hundred fish species were lost. Also eutrophication in Lake Victoria has contributed to loss of fish species through reduced light penetration causing loss of chromatic volume, reduced benthic algal depth distribution, and deep water hypoxia. Foodwebs in Lake Victoria were dramatically altered in as little time as a decade. In the temperate lakes pollution is more of a concern. In Lake Erie, the presence of sediments contaminated with PCB’s and other compounds, is linked with high rates of abnormalities in fish, up to 75%.

In the relatively undisturbed Lake Tanganyika, the main environmental changes over the past century are the result of high fishing pressure since the early sixties on the few pelagic species and the effects of climate warming. The primary production in the offshore foodweb is maintained by seasonal and local inputs of phosphorus, nitrogen and silica by upwelling and recycling in this deep meromictic lake. Annual differences in upwelling, and a general decrease of upwelling intensity and mixing due to climate warming, have a major effect on the production and may force changes in the foodweb and in the fisheries. The major pelagic planktivores in the relative simple offshore foodweb of Lake Tanganyika are clupeids, while those in Lakes Victoria and Malawi are cyprinids. Clupeids are of marine and pelagic origin, and have sometimes been considered more efficient by adaptation to the offshore environment of Large Lakes than cyprinids which as a group are predominantly riverine. A comparison of the fitness of these groups by study of growth parameters revealed no substantial differences in their performances. There is therefore no argument for the proposed introduction of clupeids to Lakes Victoria or Malawi.

Fitness predicts success for progeny in an environment similar to which their ancestors were adapted to. In changed environments there is a need to monitor the ecological fitness of key fish populations. Climate warming decreases the fitness in Pacific salmon. Migration and reproduction are affected by high temperatures, increasing energy use during migration and spawning and slowing growth rates. The salmon are affected in each stage of their life history by increased water temperatures, resulting in smaller fry, greater predation risk, a development out of phase with the spring plankton bloom and decreased energy reserves during migration. With an expected temperature increase of 3 °C this century, the expected weight loss in Pacific salmon is c. 10%. Extirpation at their southern range is expected, with repercussions for the riverine food webs dependent on the salmon.

Concerns for salmon and trout stocks in the Baltic (which in view of its low salinity can be considered a lake) are of a completely different nature. Between 90 and 75% of all salmon in the Baltic are of hatchery origin. The hatchery fish have a less well adapted antipredator behavior, lower variance, lower survival rates, and generally a lower fitness, than fish of wild origin. The abundance of hatchery reared brook trout in tributaries to Lake Superior may also pose a problem for the coaster brook trout population, which are indigenous to the lake but spawn upriver.

With the introduction of many non-native fish species in the North American Great Lakes, the predation pressure for invertivores has shifted from immature to more mature age classes. This has caused an onset of reproduction several years earlier in the native invertivore fish. Also hatchery reared lake trout, a piscivore, reproduces several years earlier than wild lake trout. The trade off between reproduction and growth in the upper levels of the food webs of the Great Lakes has important implications for the fitness of the populations and fishery management.

An especially poignant case of loss of fitness by change of fitness determinants is illustrated by the developments in Lake Victoria. Eutrophication occurred over the last few decades in Lake Victoria through increases of nutrient loads by human impacts. Transparency decreased to such levels that numerous fish species, which depend on vision for mate recognition, are threatened in their existence. The decreased transparency by particulate and dissolved matter not only lowers ambient light levels but also narrows the bandwidth of the visible spectrum of light wave lengths by wave length specific absorption. Red and blue are the main colors for males in many cichlids species. Wave lengths of blue and red light, at both ends of the visible spectrum, are both more scattered or absorbed by organic matter than light with colors of intermediate wavelengths. Red and blue colors are therefore often not or less discernable to their potential mates. In Lake Victoria where fish species have evolved only since 12 ka BP, and where female mate recognition and sexual selection have been a main driving force of evolution, many species have not yet evolved beyond reproductive barriers. Sexual selection is slowed down, or in the worst case, incipient sympatric species tend to merge and disappear. The decrease of the visual field of fish may also lead to a decrease in prey encounter rate and affect species coexistence and survival by dietary changes, generally towards less specialization and dietary overlap.

In general the increased eutrophication in Lake Victoria will lead to a further loss of diversity by several mechanisms, in this already severely impacted lake. Lake Victoria has the largest total fish-catch of all lakes in the world. With eutrophication large shifts in nutrient cycles occurred, with phosphorus increasing, while silica fell to much lower levels. This precipitated a large shift in phytoplankton taxa, from diatoms to predominantly blue-greens. Important shifts in the composition of zooplankton and benthic invertibrates followed. The environmental developments favored generalist species and impacted especially the numerous cichlid species, most of which are highly specialized. The pelagic planktivore cyprinids were relative unaffected. However the introduced Nile perch started feeding on cyprinids after the collapse of cichlids. Monitoring of these developments is of paramount importance to a proper management of the fish stocks and will provide insight into dynamic lake ecosystems functioning under stress. High stability tends to promote narrowing of specializations of fish species and a high fitness. In a dynamic, changing environment fitness decreases and more diverse diets are more of advantage. Only those species with the genetic capability to adjust to changing variability will survive. Adaptability may be related to life history traits or to physiological characteristics. Limited as they may be in their adaptative possibilities, freshwater pelagic and deep water species may be vulnerable to exotic invasions in disturbed environments. The concept of fitness can be of pragmatic use in fisheries management monitoring population dynamics of key species.

Biological, chemical and physical process studies in Lake Superior (by Christina Clark, Env. Canada, christina.clark@ec.gc.ca )

Keough and Hicks found that archeal nucleic acids account for 1-10% of the total nucleic acids in picoplankton from the Great Lakes.

Elenbaas et al.’s Lake Superior study of bacterioplankton communities used a new BIOLOG microplate technology that contains 95 carbon sources to identify metabolic ‘fingerprints’ for each community. This identification technology allowed them to discover that there are temporal changes in nearshore waters but there is no temporal change in offshore waters throughout the summer.

Studies of phytoplankton populations in Lake Superior revealed that horizontal gradients do exist and are seasonally influenced (strongest when density gradients present due to thermal bars). Bub et al. attribute vertical variation in populations to the density difference in turnover that trap the phytoplankton in the DCM.

Smutka et al. concluded that metals likely contribute to determining phytoplankton growth.Lake Superior algae growth responded most to added phosphorus, even more when both phosphorus and iron were added. Smaller growth effects were observed when both zinc and iron were added.

Early and late in the season, light limitation was observed in Lake Superior phytoplankton production. During the middle of the season phosphorus effects can be seen. Schulz et al. determined that the ratio of light and nutrients are a reasonable predictor of the Seston C:P even within a single lake. Low growth (temp) and light limitation can result in low C:P despite low nutrient levels. Nutrients limits were only a factor during stratification.

Initially Sterner and Schulz thought that Lake Superior was conservative and followed the Redfield Seston ratio. They found that Lake Superior went from the Redfield ratio in 1997 to the highest values ever recorded in 1999-2000. 1998 data are missing, but the variation may be due to climate changes, similar to decadel ENSO cycles in the Pacific ocean.

Heinen determined that Duluth basin in Lake Superior is very important in Si burial since it contains a large fraction of the total Si burial for the lake. The basin is also disproportionately important in phosphorus burial since it buries 9-15% of the total phosphorus.

Biological, chemical and physical process studies in Lake Superior (CONTINUED by Norine Dobiesz, dobieszn@pilot.msu.edu )

Water levels in Lake Superior have varied widely and current management strategies are being revisited. Using pre-regulation St Marys River discharge and a time series of net basin supplies the natural, and upper and lower bounds of the lake’s water levels were determined (A.H. Clites and F.H. Quinn). A new approach to studying turbidity uses SeaDAS processing to observe the southwestern arm of Lake Superior. Southeast winds did not affect turbidity but other wind directions have much large impacts on turbidity (A. Vodacek and K. Knobelspiesse).

Elevated levels of contaminants are a major concern in Lake Superior but the cycles of many contaminants are not well understood. The concentrations of many trace metals were found to change spatially and temporally while others were uniform across space and time (J.T Overdier and M.M Shafer). Contributions of mercury (Hg) to Lake Superior are thought to come from tributaries and atmospheric deposition, however sediment concentrations were found to rival deposition by the atmosphere (K.R. Rolfhus et al.). H.E. Sakamoto et al. detected a decreasing west to east Hg concentration gradient. River sampling within the watershed suggests that the Hg phase distribution depends on particle size. Flow rates also affected the phase distribution. (C.L. Babiarz et al.). Elevated copper (Cu) concentrations in coastal waters is related to historical copper mining. Continued dissolution of Cu may result in higher nearshore concentrations (J. Jeong and M.R. Urban). Bio-uptake and regeneration does not play a significant role in the regulation of Cu in Lake Superior (J. Jeong and M.R. Urban).

Bioaccumulation of contaminants posses a problem for the Lake Superior food web. Mysis relicata were 1.5 to 2 times higher in methyl mercury than various copepods (K.R. Rolfhus et al.). In lab experiments, Cu concentrations above ambient levels caused reductions in photosynthesis and biomass of phytoplankton (M.R. Twiss and R.M.L. McKay). Total PCBs in lake trout do not vary seasonally while smelt and lake herring contained higher concentrations in May than in October. Lake trout appear to metabolize some PCB congeners and uptake other PCB congeners (C.S. Wong et al.). Persistently high levels of toxaphene have been detected in Lake Superior waters and fish. Some congeners bioaccumulate in the food web and others do not. Seasonal diet shifts may account for the higher toxaphone concentrations in lake herring and dioperia in May than October (H. Karlsson et al.).

Economic, Social, and Education Issues in the Great Lakes (by Sheila McNair, McMaster University, mcnairs@mcmaster.ca )

Speakers addressed economic, social and education issues often within a political or regulatory context. The first part of the session was related more directly to economic and social issues, but often with an educational component while the latter session was more specifically about education with economic and social implications.

This whole area is an ‘emerging issue’ that of moving the scientific research into the more public sphere, and is also the identified theme for IAGLR 2001. The ‘emerging issues’ from these sessions are:

1. It is difficult to assign economic values to environmental elements such as wetlands, exotic species, contaminated sediments, and there are a variety of approaches that can be used.

2. Even when it is very costly (in $$) to respond to environmental damages (e.g. keeping zebra mussels off of water intake pipes) and the problem is shared throughout the basin, the magnitude of the basin-wide impact (in $$) is not known. A basin-wide perspective would inform a basin-wide response. In general, we do not have a very good idea of the costs of human-mediated impacts to the Great Lakes.

3. Pollution does not respect political boundaries. Coalitions and partnerships can more effectively mobilize cleanups and political will.

4. Education is a very effective way of addressing the scientific, social and economic issues in the Great Lakes. Both with adults and children. Innovative people are generating good programs and taking the long view toward stewardship of the lakes. There was a comment that there are social/political pressures being exerted to prevent ‘environmental brainwashing’ (i.e. promoting an environmental ‘agenda’) in Wisconsin.

Beyond PCBs: Emerging Organic Chemicals of Concern in the Great Lakes (by Aaron Peck, University of Iowa, apeck@blue.weeg.uiowa.edu )

Polychlorinated biphenyls (PCB) have been and continue to be important contaminants of concern in the Great Lakes. A variety of other compounds, however, are beginning to be identified and studied in the Lakes. Like PCBs, the compounds discussed here were all chlorinated organic compounds (except polybrominated biphenyls which are the brominated analog to PCBs)

Polybrominated diphenyl ethers (PBDE) were discussed in several presentations. Concern about these compounds is becoming important worldwide. Surface water concentrations of polybrominated diphenyl ethers (PBDE) in Lake Ontario were reported. PBDE concentrations in surface water ranged from 4-6 ng/L and about 80% was in the dissolved phase (20% associated with particulates). There was no spatial concentration or homolog pattern within the lake (Luckey, F.J., et. al.). Lake Michigan salmon (16 coho and 5 chinook) were analyzed for PBDEs. PBDE was detected in all 21 fish. PBDE #47 was the most abundant congener found. The average PBDE concentration was 80,1 ng/g wet weight (2240 ng/g lipid). These concentrations are higher than those reported elsewhere. A very good correlation between PBDE and PCB concentrations was found (Manchester-Neesvig, J.B., et. al.). Lake trout from the Great Lakes were analyzed for PBDEs and polybrominated biphenyls (PBBs). PBDE concentrations were about forty times higher than PBB concentrations. PBDE congener #47 was the most abundant congener found (55% of total PBDE mass). PBDE #99 (15% of total PBDE mass) and PBDE #100 (10% of total PBDE mass) were also commonly found. Lake Ontario fish had the highest PBDE concentrations. Lake Superior and Lake Huron fish had similar concentrations. Lake Erie fish had the lowest concentrations. It was suggested that this trend may show an atmospheric signal (Luross, J.M., et. al.). PBDEs were measured in municipal sewage sludge at eleven sites in Wisconsin. PBDE #47 and PBDE #99 were detected at all 11 wastewater treatment sites (Degenhardt, D.P., et. al.). A search for a PBDE source in a new laboratory due to high blank values resulted in the removal and replacement of the ventilation system. PBDEs are common in building products and care must be taken to insure a clean laboratory environment when analyzing these compounds (Luckey, F.J., et. al.).

Polybrominated biphenyls (PBB) concentrations in Lake Huron fish were much higher than in other lakes. PBB concentrations increased significantly from 1978-1998 and concentrations doubled between 1993 and 1998 (Luross, J.M., et. al.).

Polychlorinated naphthalenes (PCNs) associated with suspended sediments in the Detroit River were determined. The homolog profile for the sediment samples was very similar to that found in a common commercial product, Halowax 1014. PCN concentrations in suspended sediment ranged from 1.25 to 400 ng/g. The TEQs associated with these PCN concentrations were higher than those from dioxins and PCBs combined (Marvin, C.H., et. al.).

Short-chain chlorinated paraffins are polychlorinated alkanes (C10-C13). GC-high resolution ECNIMS was used to analyze sediments, municipal waste treatment plant effluents, lake water, air, lake trout, and lake carp samples taken from western Lake Ontario. The concentrations found in sediments does not follow the production profile for the Great Lakes basin. Fugacity ratios between water and air favor the deposition of these compounds from the air to the water (Muir, D., et. al.).

Nineteen in-use pesticides were detected from 1994 to 2000 in precipitation and surface water at various Great Lakes locations. Atrazine (1,039 ng/L maximum) metolachlor (736 ng/L maximum) surface water concentrations show declining trends (L’Italien, S., et. al.).

Toxaphene was measured in fish from Lake Superior. An enantioselective column was used so that enantiomeric ratios could be determined for each fish community. Smelt and lake herring had enantiomeric ratios less racemic than was found for lake trout. Changes in enantiomeric ratios correlated with a decline in the abundance of smelt in the Lake Superior fish community (Whittle, D.M., et. al.).

Evidence that the deepwater sculpin, Myoxocphalus thomsoni, can metabolize PCBs to methylsulfonyl PCBs was presented. Methylsulfonyl PCB concentrations ranged from 4.2 – 9.1 ng/g lipid in sculpin liver tissue from Grand Travers Bay (Stapleton, H.M., et. al.).

Passive in-situ concentration extraction samplers (PISCES) were used to analyze surface waters of several lakes in New York. With PISCES, 2-3 L/d can be analyzed. High MTBE concentrations were found in lakes where motorboats were allowed (Hughes, D.J., et. al.).

Brauer reported modeled increases in PCB from <40kg in the early 1990s to 1536kg in the late 1990s and finally a load of 3200kg in 2000. Certain regulations and guidelines allow waste with less than 50 ppm to be dumped in special landfills that do not have to be monitored afterwards. New permit conditions now exist for air PCBs.

Aquatic mercury sampling techniques have improved so that the equipment does not accumulate mercury. Now these new methods report more accurate values, which show a decrease since the early 1980s. Knauer has found that mining wastes were a source of mercury inputs into Lake Michigan. In Lake Superior they did not observe methyl mercury in oxic bottom waters, possibly due to bacteria mediating the process.

Comment from M. Burrows -- These two paragraphs don't appear to follow from the preceding page. Is this section supposed to go with another paper?

Advances in the biogeochemistry of sediment pore waters (by Donald Hughes, SUNY-Environmental Science and Forestry, djhughes@mailbox.syr.edu )

T.Mayer (NWRI, Burlington, Ontario) presented a paper on the effects of road salts on porewaters in a small pond located in the Rouge River valley. This pond is adjacent to a major highway (the 401). The influence of road salts was seen in the increasing conductivity in waters leading to the pond. Sediment porewaters contained up to 2000 mg/L of chloride, and elevated levels of ammonia. Toxicity testing showed a dose-response relationship, with no survival in 100% porewater, and increasing survival with increasing dilution.

The release of mercury from porewaters was discussed by Jim Hurley (U. Wisconsin). His research employed a unique "close interval sampler" --a device implanted in the sediments which enabled sampling of water just above the sediment-water interface. Porewaters were collected using special all-Teflon units. In the METALICUS project in western Canada, stable-isotope Hg was added to a lake and subsequently monitored. Despite efforts to measure concentration profiles of total and methyl-mercury at the sediment-water interface, modeling had to be used instead. This suggested that both Hg-T and Hg-Me peaked at the interface.

The next speaker, Gerry Matisoff (Case Western Reserve U.), gave two talks. The first of these examined the succession of benthic organisms and their effects on porewater chemistry in Lake Erie. Sediments are first colonized by chironomids, which make U-shaped burrows extending 10 or more cm beneath the sediment surface, allowing for greater exchange of porewater with the overlying water column. They are succeeded by tubificid oligochaetes, which stay in the top several cm of sediments. When chironomids dominate, the concentrations of dissolved silica, iron, phosphorus, and alkalinity are all decreased in porewater. The flux of silica out of the sediment is enhanced by chironomids. However, iron and phosphorus precipitate at the surface (Fe reacts with oxygen; P sorbs to Fe). The flux of ammonia is high in either case.

In his second talk, Matisoff presented a model for sediment-water interactions based on work done on marine systems. The starting point is the "G-model" developed by Berner (1974), where "G" represents potentially degradable organic carbon. Modern refinements on this model allow for differing types of carbon, the re-oxidation of reduced substances, and bioturbation. A modern carbon-nitrogen diagenesis model (CANDI) was applied to freshwater systems. The major difficulty here is that the reaction rate constants in freshwater are largely unknown. Nonetheless, with salinity, T, and P as inputs, concentration profiles (vs. sediment depth) of major species were estimated. Problems need to be worked out yet: pH is over-estimated, and organic matter inputs do not reach steady-state.

Louis Thibodeaux (L.S.U.) gave a fascinating talk about the release of soluble PCBs from contaminated sediments, particularly those in the Hudson River. He noted that traditionally, contaminant transport models have focussed on particles, but releases of soluble components represent a significant mode a transport (30% in Fox R.; 60+% in Hudson R.). In his river model, soluble losses from sediment deposits were either transported downstream or lost via evaporation. PCB transport out of the sediments included several steps: enhanced movement of particles via bioturbation; equilibrium desorption at the sed,-water interface; and movement across the benthic boundary layer. He modeled the release of PCBs with 3-10 chlorines from the Thompson Island pool in the upper Hudson River. The model results showed a strong dependence on both flow and temperature. Increasing the bioturbation term tended to lessen the temperature dependence and strengthen flow dependence. Overall, he reported that the measured coefficient is the same magnitude as the calculated values, using no adjustable parameters. He also noted that dissolved organic matter isn the most important vector for soluble PCB release. In response to questions, Thibodeaux said that there is substantial release of PCBs even without resuspension of sediments.

Finally, Carrie Turner (LimnoTech, Inc.) presented a follow-up talk which incorporated Thibodeaux's two-layer mechanistic model describing release of soluble PCBs from Hudson River sediments. The Thompson Is. pool, which holds 50% of the PCB mass, was modeled in three longitudinal sections. Data under low-flow conditions ( <10,000 cfs) were used to determine soluble PCB losses, since resuspension was minimal. A wide range of PCB congeners were modeled using a two-layer resistance model: heavy congeners were limited by water-side resistance, whereas the light congeners were sediment controlled. Overall, the model accounted for the measured release of different congeners quite well.

Riparian zones and watersheds (by Vicki Medland, University of Wisconsin at Green Bay, MedlandV@uwgb.edu )

Inland lakes cores were used by Simpson et al. to develop a better understanding of chemical loading in Great Lakes watersheds. Lakes appear to be more sensitive to land-use changes.

Birds, Habitat, and Landscapes in the Great Lakes Basin (by Sarah Wielgus, University of Wisconsin at Green Bay, wielsj12@uwgb.edu )

There seems to be some indication that breeding and migratory birds do exhibit a “Great Lakes effect.” When examining the species richness of land birds in woodlots near the shore of Lake Michigan and farther inland at the same latitude it was noticed that the species richness of these land birds was greatest around the shore area. This “lake effect” seems to be restricted to sites closer to the shore ( <10 km). Overall land bird abundance and diversity was greatest within 5 km of the lake shore. It is thought that there are multiple causes responsible for this “lake effect.” Two important differences between near shore and inland sites are microhabitat and microclimate. The tree species composition and average temperature are noticeably different near the lake shore compared to farther inland even at the same latitude.

Developing models that predict the abundance of breeding birds as a result of changing habitat patterns is important for estimating future bird populations. Bird counts and aerial photos of the local vegetation in the Seney National Wildlife Refuge were used to develop models that would predict the abundance of individual bird species. It was discovered that patch size in 38 of the 40 models was the most important predictor of bird abundance. Patch size was far more important than patch characteristics in predicting bird abundance. Therefore, bird species are likely declining due to habitat loss and fragmentation. A similar modeling study conducted in the Nicolet National Forest showed somewhat different results. Habitat choices of various bird species were based on both landscape and site-specific habitat characteristics. For some birds, small-scale factors were more important.

Forest managers could benefit from bird surveys that use certain bird species as indicators of habitat fragmentation and overall ecosystem health. Owls could be key indicators of habitat degradation but little is known about their habitat requirements in many areas. An owl survey in the Nicolet National Forest showed that barred and saw-whet owls are the most well distributed owl species in the forest. Saw-whet owls showed low site fidelity and generally selected upland spruce and lowland conifer for habitat. In most cases great-horned owls are used as indicators of forest degradation but this owl species is rare in the Nicolet National Forest.

The altering of a natural phenomenon in a landscape can have significant effects on the surrounding bird populations. Sharp-tailed grouse populations in northeast Minnesota are showing this trend. Landscape composition was surveyed around active and inactive lek sites as well as at random points. Active bird sites had higher proportions of native sedge grass whereas inactive sites were dominated by upland forest and brush cover types. A model was developed to predict the probability of grouse occurrence in a specified landscape cover type. Only 8% of the land area had a greater than 80 percent chance of an active lek site present. This indicates that there is a low proportion of the landscape with suitable habitat for lek sites. Sharp-tailed grouse declines are likely linked to changes in landscape cover types. Suppression of fires in these areas has led to succession within the landscape changing it from a grass-dominated habitat which favors grouse lek sites to an upland forest. Therefore, future management of the sharp-tailed grouse populations in this region must be done at a landscape level in which fire is no longer suppressed.

Staging, Wintering, and Breeding Waterfowl on the Great Lakes (by Jeanette Jaskula, University of Wisconsin at Green Bay, jaskjm25uwgb.edu )

Ducks Unlimited is currently developing a GIS-based decision support system used to model mallard duck ‘hotspots’ in the Great Lakes. The model will incorporate an engineering perspective to correlate land use factors and wetland types with mallard counts. This model will also assist Ducks Unlimited in predicting which areas in the Great Lakes need wetland restorations based on habitat deficiency information collected in the field. It will further determine how costly and feasible a wetland restoration is to create duck habitat. This appears to be a wonderful support system not only aimed at conserving ducks, but restoring the valuable services wetland provide to animals and humans alike.

Twelve tundra swans were tracked from Long Point, Lake Erie using radio transmitters from 1998-2000. Tundra swans spent considerable time in farm fields acquiring the necessary fat needed for the long migration to breeding sites. It is interesting to note that tundra swans are less dependent on high quality wetlands as they are on agricultural fields.

The historical distribution of Phragmites australis was studied at Long Point, Lake Erie and a prediction was made on its future spread in the area. In 1964, Phragmites distribution dropped on Long Point most likely due to rising water levels. However, by 1995 it had expanded probably because of lower lake water levels, nitrate levels, and perhaps a new genotype. Continuing studies will address the extent to which this plant could invade further areas based on water levels and soil type.

Feral mute swans are non-migratory birds that displace native birds and severely reduce aquatic plant populations. They have few natural predators and the recent warming trend has caused a sharp increase in their reproductive output and overwinter survival. Aerial survey data indicate that their population will increase 10-30 % each year. An increase of 10% results in the population doubling every 7 years while a 30% increase results in the population doubling less than every 3 years! It was suggested that these birds be removed from the list of protected species under the Migratory Bird Act in Canada and lose protection in the U.S. as well. However, as discussed at the conclusion of the session, one can foresee a huge public relations problem when we open hunting season on a bird long epitomized as an eternal symbol of grace and elegance!

How did deepwater Great Lakes fishes survive glaciation: geological and biological approaches (by Brian Ellrott, University of Vermont, bellrott@zoo.uvm.edu )

C.S. Brockman presented evidence that large long-lived deepwater lakes did not exist along the late Wisconsinan maximum ice margin. Since the maximum advance was well south of the St. Lawrence-Mississippi divide lakes that occurred there were temporary and drained to the Mississippi River. Deepwater refugia adjacent to advancing and retreating ice margins were more likely available during the middle Pleistocene.

D. M. Mickelson stated from about 23,000BP to 14,000BP glacier ice filled the upper Great Lakes drainage basin, eliminating the possibility of fish survival in the basin. Then around 14,000BP ice retreated into the southern part of the Lake Michigan basin. Since 13,000BP deep lake conditions existed as the ice retreated further and connected Huron and Michigan.

C.F.M. Lewis and S.M. Blasco presented evidence that retreating glaciers may have caused catastrophic melt-water outbursts that removed biota from the lower Great Lakes basins. Glaciofluvial flows at minimum velocities of 5-10m/s significantly eroded bedrock and glacial drift in the basins. When catastrophic melt-water events ceased lake levels were likely well below outlet elevations in the Ontario, Erie, and Georgian Bay basins. The Huron-Georgian Bay biota would have been isolated for at least two successive periods spanning 300-600 years.

A macrophysical climate model developed at the University of Wisconsin-Madison was presented by R.A. Bryson. The model can be applied as a tool to understand climates of the proglacial region in the late Pleistocene and early Holocene. Information on the climate of this period will provide a better understanding of the environmental conditions that allowed deepwater fishes to survive glaciation.

C.C. Wilson presented genetic evidence that strongly suggests that pre-Wisconsinan glacial events caused the observed evolutionary divergences among the lake trout (Salvelinus namaycush) that survived in at least six peripheral refugia. Genetic data indicate that refugial lake trout populations were size constrained and isolated leading to genetic divergence within the species.

R.L. Eshenroder and J. A. Fuller present evidence that blue pike (Stizostedion vitreum glaucum) evolved from walleye during the Holocene. Rapid refilling of Lake Erie’s western basin around 5,000BP may have created conditions for reproductive isolation and then differentiation of blue pike from walleye. Prior to refilling one or two waterways flowed across the Pelee Moraine into the central basin. As lake levels rose habitat on top of the submerged moraine may have been suitable for walleye egg incubation isolating this population from those spawning in the upland rivers. Then selection favored a more pelagic morphotype (blue pike).

Research and Monitoring: enhancing linkages and communication (by Christina Clark, Environment Canada, mailto:christina.clark@ec.gc.ca )

SOLEC (State of the Lakes Ecosystem Conference) objectives are to assess the state of the Great Lakes, strengthen decisions and management, inform local decision makers, and provide a forum for networking among stakeholders. An estimate of ~ 4 million dollars to implement the entire suite of 80 indicators developed to evaluate the state of the Great Lakes was presented. Data gaps still exist for some of the indicators, so outside assistance is welcome. Visit the website: http://www.on.ec.gc.ca/solec/intro.html for more details.

The Lake Erie Monitoring Network (LEMNet) has provided workshops for data sampling for a number of years at the Stone laboratories, but has not maintained the data to report or monitor conditions in the basin. The data will now be available online for their summer program and for the year-round monitoring sonde at the website: http://www.sg.ohio-state.edu/ .

Meadows et al. have developed a real-time environmental monitoring program for Lake St. Clair that uses both remote sensing data and in-situ marine observation stations. Flow and chemical concentrations are used to trace and predict pollutant movement. Hydraulic flow and west wind influence are also incorporated in the model since the east end of Lake St. Clair has poor circulation unless winds are present. Currently working on forecasting pollution transportation.

Analytical Tools Applicable to Ecosystem-based Decision-making in the Great Lakes (by Alison Fraser, Trent University, afraser@trentu.ca )

The main objective of this session was to evaluate various tools that are being used to assist in the decision-making processes dealing with ecological concerns in the Great Lakes. These tools include models, indices of biotic integrity, and probabilistic risk assessment individually or in combination. To begin with, S.W. Effler discussed the development of a dynamic, two-dimensional, mass balance river quality model, which can be used for both dissolved oxygen and chlorophyll. The model is unique in that it also accounts for the oxygen demand and filtering effects of zebra mussels. The use of ecological models was discussed by H. Zhang, who presented work on modeling Daphnia in Lake Erie. It is hoped that this trophodynamic model will be combined with a two-dimensional hydrodynamic model for this system. M.J. Boote has extended conventional food web bioaccumulation models to take into consideration the habitat preference of each organism. This becomes important for organisms such as fish due to their mobility and preference for certain habitats. Using the model, it is possible to further direct remedial efforts in contaminated areas, however, the availability of site specific habitat use data is limited.

As with models, Indices of Biotic Integrity (IBI) are often used in assessing the state of an ecosystem. D.D. Kane presented research on the development of an IBI for planktonic communities (P-IBI) which will allow for improved water quality monitoring for offshore areas as well as assist in decision-making. Some concerns with the P-IBI are

1. the effects of lower taxonomic resolution,
2. the unavailability of reference conditions,
3. natural cycles and variability, and
4. whether the samples are representative on a spatial and temporal scale.

Probabilistic ecological risk assessments are also valuable tools in assessing ecological concerns in the Great Lakes. D.R. Passino-Reader presented such an assessment for Lake Erie, using the Crystal Ball software package, to identify contaminants posing a potential risk to the biota of this aquatic system. The results from the assessment can in turn be used to quantify the ecological cost-effectiveness of sediment remediation.

T.W. Chapman and J.E. Scholl described a flood management strategy which has been developed and applied to the Menomonee River watershed and includes the use of a variety of analytical tools in combination. This area has been substantially developed and flood events have resulted in major economic losses. Tools used to in this scenario include hydrologic and hydraulic models, sediment transport investigations, environmental analyses, public involvement activities, corridor field monitoring , and preliminary design evaluations. As with the Menomonee watershed, a wide variety of analytical tools were used in a project by L.J. Smith in Door County. The main objective of the study was to generate future scenarios based on a variety of growth and development policies as well as related decisions in the area. It was stressed that public involvement in this type of analysis is of extreme importance and should be encouraged by all parties participating in the process. R.B. Wenger discussed work conducted on the Upper Fox River basin which combines analytical tools including Ecological Risk Assessment (ERA), Ecosystem Stressor Indicators (ESI), and Mediated Modelling (MM). This will allow for the identification of critical stressors in a multiply stressed ecosystem and help stakeholders prioritize the stressors and direct management efforts. It is hoped that this study will address the following questions:

1. what is the best combination of tools with respect to ecosystem management and
2. what is the most effective way to develop these tools and the subsequent results.