USGS - science for a changing world

U.S. Geological Survey - Great Lakes Science Center

GLSC Reports

This is a compilation of six reports presented at the annual lake committee meetings.

This is a compilation of seven reports presented at the annual lake committee meetings.

This is a compilation of seven reports, each of which is individually numbered on the bottom of the page. As such, numbering for the document as a whole is located on the upper right corner of each page.
 

The assessment of prey fish stocks in the Great Lakes have been conducted annually with bottom trawls since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. These stock assessments provide data on the status and trends of prey fish that are consumed by important commercial and recreational fishes. Although all these annual surveys are conducted using bottom trawls, they differ among the lakes in the proportion of the lake covered, seasonal timing, bottom trawl gear used, and the manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspects, direct comparison of prey fish catches among lakes is not straightforward. However, all of the assessments produce indices of abundance or biomass that can be standardized to facilitate comparisons of status and trends across all the Great Lakes. In this report, population indices were standardized to the highest value for a time series within each lake for the following principal prey species: cisco (Coregonus artedi), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). Indices were also provided for round goby (Neogobius melanostomus), an invasive fish that has proliferated throughout the basin over the past 18 years. These standardized indices represent the best available long-term indices of relative abundance for these fishes across all of the Great Lakes. In this report, standardized indices are presented in graphical form along with synopses to provide a short, informal cross-basin summary of the status and trends of principal prey fishes. In keeping with this intent, tables, references, and a detailed discussion were omitted.

The Great Lakes Science Center has conducted daytime nearshore bottom trawl surveys of Lake Superior (15-80 m bathymetric depth zone) each spring since 1978 and an offshore survey (>80 m) since 2011 to provide long-term trends of relative abundance and biomass of the fish community. In 2012, 72 nearshore and 34 offshore stations were sampled with a 12-m Yankee bottom trawl.

The 2012 estimate of lake-wide nearshore fish community biomass was 1.14 kg/ha, second lowest in the 35-year survey history, down from 3.63 kg/ha observed in the 2011 survey. Dominant species in the catch, in order of relative biomass, were bloater, rainbow smelt, lake whitefish, pygmy whitefish, and shortjaw cisco. Compared to 2011 levels, biomass of all species decreased. Year-class strengths for the 2011 cisco and bloater cohorts were well below average and ranked as the second weakest year-classes in the past 35 years. Year-class strength of rainbow smelt was the weakest in the survey record, continuing a decline that began in 2008. As in 2011, densities of hatchery lake trout remained near zero in 2012, while densities of wild (lean) lake trout and siscowet lake trout decreased. Proportions of total lake trout density in 2012 that were hatchery, wild, and siscowet were 5, 74, and 21%, respectively.

The 2012 estimate of lake-wide offshore fish community biomass was 6.9 kg/ha, down from 9.0 kg/ha in 2011. Deepwater sculpin, kiyi, and siscowet lake trout represented 98% of the fish caught in terms of both density and biomass. Community composition, number of species collected and densities and biomass for most species were similar to that observed in 2011.

Due to ship mechanical failures, nearshore sampling was delayed from mid-May to mid-June to mid-June to late August. The shift to summer sampling when the lake was stratified may have affected our estimates, thus our estimates of status and trends for the nearshore fish community in 2012 are tentative, pending results of future surveys. However, the results of the 2012 survey are comparable with those during 2009 and 2010 when lake-wide fish biomass declined to < 1.40 kg/ha. Declines in prey fish biomass since the late 1990s can be attributed to a combination of increased predation by recovered lake trout populations and infrequent and weak recruitment by the principal prey fishes, cisco and bloater. In turn declines in lake trout biomass since the mid-2000s are likely linked to declines in prey fish biomass. If lean and siscowet lake trout populations in nearshore waters continue to remain at current levels, predation mortality will likely maintain the relatively low prey fish biomass observed in recent years. Alternatively, if lake trout populations show a substantial decline in abundance in upcoming years, prey fish populations may rebound in a fashion reminiscent to what occurred in the late 1970s to mid-1980s. However, this scenario depends on substantial increases in harvest of lake trout, which seems unlikely given that levels of lake trout harvest have been flat or declining in many regions of Lake Superior since 2000.

The USGS Great Lakes Science Center (GLSC) conducted acoustic/midwater trawl surveys of Lake Huron during 1997 and annually during 2004-2012. The 2012 survey was conducted during September and October, and included transects in Lake Huron’s Main Basin, Georgian Bay, and North Channel. Pelagic fish density (638 fish/ha) was lower in 2012 compared to 2011, with density in 2012 only 34% of 2011. Total biomass in 2012 was 74% of the 2011 value. Alewife Alosa pseudoharengus remained nearly absent, and only one cisco Coregonus artedi was captured. Rainbow smelt Osmerus mordax density was only 31% of the 2011 density. Bloater Coregonus hoyi density was less than half the 2011 density, mostly as a result of lower density of small bloater. Density and biomass of large bloater in 2012 were similar to 2011 levels. During 2012 we observed significantly higher fish biomass in North Channel than in the Main Basin or Georgian Bay. Prey availability during 2013 will likely be similar to 2012. Lake Huron now has pelagic fish biomass similar to that observed in recent lakewide acoustic surveys of Lake Michigan and Lake Superior, but species composition differs in the three lakes. There is an increasing diversity and prevalence of native species gradient from Lake Michigan to Lake Superior, with Lake Huron being intermediate in the prevalence of native fish species like coregonines and emerald shiner Notropis atherinoides.

The U.S. Geological Survey Great Lakes Science Center has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. The resulting data on relative abundance, size and age structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the survey were completed in 2012. The survey provides relative abundance and biomass estimates between the 5-m and 114-m depth contours of the lake (herein, lake-wide) for prey fish populations, as well as burbot, yellow perch, and the introduced dreissenid mussels. Lake-wide biomass of alewives in 2012 was estimated at 9 kilotonnes (kt, 1 kt = 1000 metric tonnes), which continues the trend of unusually low alewife biomass since 2004 but represented a 20% increase from the 2011 estimate. The age distribution of alewives larger than 100 mm was dominated (i.e., 84%) by age-2. Record low biomass was observed for several species, including bloater (0.4 kt), rainbow smelt (0.1 kt), deepwater sculpin (1.5 kt), and ninespine stickleback (0.01 kt). Slimy sculpin lake-wide biomass was 0.73 kt in 2012, which was the third consecutive year revealing a decline. Estimated biomass of round goby increased by 79% to 3 kt. Burbot lake-wide biomass (0.5 kt in 2012) has remained below 3 kt since 2001. Numeric density of age-0 yellow perch (i.e., < 100 mm) was only 2 fish per ha, which is indicative of a relatively poor year-class. Lake-wide biomass estimates of dreissenid mussels have continued to increase from 2010, from 12 to 95 kt in 2012. Overall, the total lake-wide prey fish biomass estimate (sum of alewife, bloater, rainbow smelt, deepwater sculpin, slimy sculpin, round goby, and ninespine stickleback) in 2012 was 15 kt, which represented the lowest total biomass of the time series.

The USGS Great Lakes Science Center has conducted trawl surveys to assess annual changes in the offshore demersal fish community of Lake Huron since 1973. Sample sites include five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2012 fall bottom trawl survey was carried out between 20 October – 5 November 2012 and included all U.S. ports as well as Goderich, ON. The 2012 main basin prey fish biomass estimate for Lake Huron was 97 kilotonnes, higher than the estimate in 2011 (63.2 Kt), approximately one third of the maximum estimate in the time series, and nearly 6 times higher than the minimum estimate in 2009. The biomass estimates for adult alewife in 2012 were higher than 2011, but remained much lower than observed before the crash in 2004, and populations were dominated by small fish. Estimated biomass of rainbow smelt also increased and was the highest observed since 2005. Estimated adult bloater biomass in Lake Huron has been increasing in recent years, and the 2012 biomass estimate was the third highest ever observed in the survey. Biomass estimates for trout-perch and ninespine stickleback were higher than in 2011 but still remained low compared to historic estimates. The estimated biomass of deepwater and slimy sculpins increased over 2011, and slimy sculpin in particular seem to be increasing in abundance. The 2012 biomass estimate for round goby was similar to that in 2011 and was the highest observed in the survey. Substantial numbers of wild juvenile lake trout were captured again in 2012, suggesting that natural reproduction by lake trout continues to occur. The 2012 Lake Huron bottom trawl survey results suggest that several species of offshore demersal fish are beginning to increase in abundance.

Acoustic surveys were conducted in late summer/early fall during the years 1992-1996 and 2001-2012 to estimate pelagic prey fish biomass in Lake Michigan. Midwater trawling during the surveys as well as target strength provided a measure of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. The 2012 survey consisted of 26 acoustic transects (576 km total) and 31 midwater tows. Mean total prey fish biomass was 6.4 kg/ha (relative standard error, RSE = 15%) or 31 kilotonnes (kt = 1,000 metric tons), which was 1.5 times the estimate for 2011 and 22% of the long-term mean. The increase from 2011 resulted from increased biomass of age-0 alewife, age-1 or older alewife, and large bloater. The abundance of the 2012 alewife year class was similar to the average, and this year-class contributed 35% of total alewife biomass (4.9 kg/ha, RSE = 17%), while the 2010 alewife year-class contributed 58%. The 2010 year class made up 89% of age-1 or older alewife biomass. In 2012, alewife comprised 77% of total prey fish biomass, while rainbow smelt and bloater were 4 and 19% of total biomass, respectively. Rainbow smelt biomass in 2012 (0.25 kg/ha, RSE = 17%) was 40% of the rainbow smelt biomass in 2011 and 5% of the long term mean. Bloater biomass was much lower (1.2 kg/ha, RSE = 12%) than in the 1990s, and mean density of small bloater in 2012 (191 fish/ha, RSE = 24%) was lower than peak values observed in 2007-2009. In 2012, pelagic prey fish biomass in Lake Michigan was similar to Lake Superior and Lake Huron. Prey fish biomass remained well below the Fish Community Objectives target of 500-800 kt, and key native species remain absent or rare.

The assessment of prey fish stocks in the Great Lakes have been conducted annually with bottom trawls since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. These stock assessments provide data on the status and trends of prey fish that are consumed by important commercial and recreational fishes. Although all these annual surveys are conducted using bottom trawls, they differ among the lakes in the proportion of the lake covered, seasonal timing, bottom trawl gear used, and the manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspects, direct comparison of prey fish catches among lakes is not straightforward. However, all of the assessments produce indices of abundance or biomass that can be standardized to facilitate comparisons of status and trends across all the Great Lakes. In this report, population indices were standardized to the highest value for a time series within each lake for the following important prey species in the Great Lakes: cisco (Coregonus artedi), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). Indices were also provided for round goby (Neogobius melanostomus), an invasive fish presently becoming established throughout the basin. These standardized indices represent the best available long-term indices of relative abundance for these fishes across all of the Great Lakes. In this report, these standardized indices are presented in graphical form along with synopses to provide a short, informal cross-basin summary of the status and trends of prey fishes. In keeping with this intent, tables, references, and a detailed discussion were omitted.

For each lake, standardized relative indices for biomass of age-1 and older fishes and numeric density of recruits were calculated as the observed value divided by the maximum value observed in the times series. Recruitment indices of year-class strength reliably reflect the magnitude of the cohort recruited at subsequent ages. Differences in the timing of surveys across lakes and differences in methodology to distinguish age classes resulted in adopting different fish age-classes to index year-class strength for each species. Year-class strengths were based on aged cisco, bloater, and smelt in Lake Superior and alewife in Lake Michigan. For other species and lakes, age-classes were assigned based on fish length cut-offs. Depending on the lake and species, year class strengths were assessed from densities of age-0, age-1, or age-3 fish.

The Kendall coefficient of concordance (W) was calculated to determine if the time series of relative abundances for a given species was statistically “concordant” across lakes. W can range from 0 (complete discordance or disagreement) to 1 (complete concordance or agreement). The P-value for W provides the probability of agreement across lakes.

When making statistical comparisons of trends among lakes, data were restricted to years when all or a group of lakes were sampled. For all lakes, data from 1992, 1993, 1998, and 2000 were omitted from statistical comparisons because missing or atypical data were collected in one or more lakes. Comparisons with Lake Erie were restricted to 1990-2010, years when surveys with a consistent sample design were conducted. Beginning with our 2010 report, a complete series of catch data from Lake Huron was made available for comparison with other lakes because fishing power corrections to the Huron data were developed to account for the use of a larger bottom trawl to conduct surveys during 1992-2011. Assessment of cross-basin trends for round gobies begins with 1994, the first year that these fish were detected in bottom trawl surveys in the Great Lakes.

The USGS Great Lakes Science Center conducted acoustic/midwater trawl surveys of Lake Huron during 1997 and annually during 2004-2011. The 2011 survey was conducted during September and October, and included transects in Lake Huron’s Main Basin, Georgian Bay, and North Channel. Main Basin estimates of pelagic fish density and biomass were higher in 2011 compared to 2010. Bloater Coregonus hoyi densities and biomass did not change between 2010 and 2011, but we observed increases in rainbow smelt Osmerus mordax (1.8x), and emerald shiner Notropis atherinoides (625x) biomass. Alewife Alosa pseudoharengus remained nearly absent, but ciscoes Coregonus artedi were captured in all four trawls on both North Channel transects. During 2011 we observed no significant differences in fish density or biomass among North Channel, Georgian Bay, or the main basin. That spatial pattern differed from patterns we found during 2004-2007 when biomass in the sub-basins was higher. Prey availability during 2012 will likely be higher than 2011 due to increases in rainbow smelt and emerald shiner. Lake Huron now has almost two times greater pelagic biomass than Lake Michigan, but species composition differed. Alewife predominated in Lake Michigan, while pelagic biomass in Lake Huron was comprised of rainbow smelt, bloater, and to a lesser extent, emerald shiner.

The Great Lakes Science Center (GLSC) has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. The resulting data on relative abundance, size structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the survey were completed in 2011. The survey provides relative abundance and biomass estimates between the 5-m and 114-m depth contours of the lake (herein, lake-wide) for prey fish populations, as well as burbot, yellow perch, and the introduced dreissenid mussels. Lake-wide biomass of alewives in 2011 was estimated at 7.64 kilotonnes (kt) (1 kt = 1000 metric tons), which was 19% higher than the 2010 estimate. Adult alewife abundance has remained low during 2004-2011 and the age distribution of the adult population was truncated during 2009-2011 compared with previous years. Lake-wide biomass of bloater in 2011 was estimated at 3.70 kt, which was 52% lower than the 2010 estimate. Rainbow smelt lake-wide biomass equaled 0.47 kt in 2011, which was 38% lower than the 2010 estimate. Deepwater sculpin lake-wide biomass equaled 1.86 kt, which was 29% lower than the 2010 estimate. Slimy sculpin lake-wide biomass remained relatively high in 2011 (1.93 kt), but was still 30% lower than the 2010 level. Ninespine stickleback lake-wide biomass equaled 0.04 kt in 2011, which was the lowest value in the time series. Estimated biomass of round goby decreased by 78% between 2010 and 2011, from 8.51 to 1.83 kt. Burbot lake-wide biomass (2.00 kt in 2011) has shown a recent increase. Numeric density of age-0 yellow perch (i.e., < 100 mm) equaled only 1 fish per ha, which is indicative of a relatively poor year-class. Lake-wide biomass estimates of dreissenid mussels increased by more than fourfold from 12.18 kt in 2010 to 52.70 kt in 2011. Overall, the total lake-wide prey fish biomass estimate (sum of alewife, bloater, rainbow smelt, deepwater sculpin, slimy sculpin, round goby, and ninespine stickleback) in 2011 was 17.47 kt, which represented the lowest total biomass of the time series.

The USGS Great Lakes Science Center has conducted trawl surveys to assess annual changes in the offshore demersal fish community of Lake Huron since 1973. Sample sites include five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2011 fall bottom trawl survey was carried out between 22 October – 5 November 2011 and included all U.S. ports as well as Goderich, ON. The 2011 main basin prey fish biomass estimate for Lake Huron was 63.5 kilotonnes, more than double the estimate in 2010 (29.1 Kt) and approximately 17 percent of the maximum estimate in the time series. Biomass estimates for adult alewife and rainbow smelt in 2011 were slightly lower than 2010, and remained near the lowest observed in the time series, and populations were dominated by small fish. Estimated adult bloater biomass in Lake Huron has been increasing in recent years, and the 2011 biomass estimate was the highest observed since 1995. Biomass estimates for trout-perch and ninespine stickleback were higher than in 2010 but still remained low compared to historic estimates. The biomass of slimy and deepwater sculpins was higher than estimated in 2010, but remain near the lowest levels observed. The 2011 biomass estimate for round goby was much higher than in 2010 and was the highest observed in the survey. Substantial numbers of wild juvenile lake trout were captured again in 2011, suggesting that natural reproduction by lake trout continues to occur. Although bloater abundance and biomass are increasing, the overall low prey fish abundance in Lake Huron may have continuing negative implications for Chinook salmon in the lake.

The Great Lakes Science Center has conducted daytime nearshore bottom trawl surveys of Lake Superior (15-80 m bathymetric depth zone) each spring since 1978 to provide long-term trends of relative abundance and biomass of the fish community. Between 19 May and 20 June 2011, 82 stations distributed around the perimeter of the lake were sampled with a 12-m Yankee bottom trawl towed cross-contour. The 2011 estimate of fish community biomass was 3.63 kg/ha, seventh lowest in the 34-year survey history, but up from 1.37 kg/ha observed in the 2010 survey. The distribution of biomass across jurisdictions was uneven; mean biomass in Canada East, Canada West, Michigan, Minnesota and Wisconsin waters were 2.23, 3.64, 2.07, 0.01, and 7.73 kg/ha, respectively. Dominant species in the catch, in order of relative biomass, were lake whitefish, rainbow smelt, bloater, cisco, and longnose sucker. Compared to 2010 levels, biomass of cisco, bloater, shortjaw cisco, lake whitefish, rainbow smelt, and lean and siscowet lake trout increased. Year-class strengths for the 2010 cisco and bloater cohorts were well below average and ranked as the ninth and twelfth weakest year-classes, respectively, in the past 34 years. The 2011 cisco age structure was dominated by age-2 fish (2009 year-class), which accounted for 91% of the ciscoes captured. Remaining ciscoes captured were composed mostly of adults from the 2005, 2003 and 1998 year-classes. Year-class strength of rainbow smelt was the fifth weakest in the survey record, continuing a decline that began in 2008.

As in 2010, densities of small, intermediate and large hatchery lake trout remained near zero in 2011. Densities of small and intermediate wild (lean) lake trout increased in 2011 while density of large wild lake trout decreased. Density of all sizes of siscowet lake trout declined in 2011. Proportions of total lake trout density in 2011 that were hatchery, wild, and siscowet were 2, 66, and 32%, respectively.

Declines in prey fish biomass since the late 1990s can be attributed to predation by recovered lake trout populations. In turn, recent declines in lake trout biomass are likely linked to declines in prey fish biomass. If lean and siscowet lake trout populations in nearshore waters continue to remain at current levels, predation mortality will likely maintain the relatively low prey fish biomass observed in recent years. Alternatively, if lake trout populations show a substantial decline in abundance in upcoming years, prey fish populations may rebound in a fashion reminiscent to what occurred in the late 1970s to mid-1980s, however, this scenario is unlikely given present management strategies.

The U.S. Geological Survey (USGS) and New York State Department of Environmental Conservation (NYSDEC) have cooperatively assessed Lake Ontario prey fishes each year since 1978. Bottom trawling has been conducted during spring to assess alewife Alosa pseudoharengus, summer to assess rainbow smelt Osmerus mordax, and autumn to assess benthic prey fish (slimy sculpin Cottus cognatus and deepwater sculpin, Myoxocephalus thompsonii). Timing of these surveys was originally selected based on the season in which bottom trawl catches of the target species peaked during trawling conducted in May to October 1972 (Owens et al. 2003). Twelve transects were established at roughly 25-km (15.5- mile) intervals along the U.S. shoreline (Figure 1). Bottom trawling was generally conducted at all transects to assess alewife, at all transects except Fair Haven to assess rainbow smelt, and at six transects to assess benthic prey fish (Figure 1). Although each survey targets specific fish species, catches of non-target fishes are also tracked and provide information on ecologically important changes in the fish community, such as the resurgence of once abundant native species (e.g., deepwater sculpin Myoxocephalus thompsonii, Lantry et al. 2007) or increasing abundance of recently introduced invasive species (e.g., round goby Neogobius melanostomus (Walsh et al. 2007).

The underlying principle of our original sampling plan was to concentrate sampling effort in the depth zone where the target species was most abundant by using our knowledge of each fish species’ unique bathymetric distribution. At each transect, we typically made trawl hauls at 10-m depth intervals through the range of depths occupied by the target species. In 1997, however, for the alewife and rainbow smelt assessments we modified the number and range of depths fished at each transect, as well as the trawling gear, in response to the invasion of zebra and quagga mussels (Dreissena polymorpha and D. bugensis, respectively, hereafter referred to collectively as dreissenids). Dreissenids contributed to changes in fish distribution in the early 1990s (O’Gorman et al. 2000) and affected net efficiency by clogging bottom trawls.

Fixed-station sampling designs, such as ours, are commonly used for assessing fish populations in the Great Lakes and in northern Europe (ICES 2004). The underlying assumption is that changes in relative abundance at the fixed stations are representative of changes in the whole population. Mean abundance from fixed-station surveys will not be biased if the fish are randomly distributed. We have always assumed that the fish are randomly distributed in the geographic area in which a transect is located, and because we have numerous transects spaced at regular intervals around the shore, that our abundance indices are unbiased. Acoustic sampling conducted during the 2004-2006 alewife bottom trawl assessments confirmed our assumption of random distribution within geographic areas. Furthermore, differences among geographic areas in densities of alewifestrength targets measured with acoustics were reflected in the densities of alewife measured by bottom trawl. However, there is no assurance that this has always been true given the large-scale shifts in fish distribution since dreissenids proliferated. Although random sampling is preferable for estimating precision, the systematic, fixed-station sampling that we employ in Lake Ontario will often be optimal for generating the most precise estimate of relative abundance even though the variance of the estimated relative abundance will be biased (ICES 2004).

Two vessels participated in prey fish surveys during 1978 - 1982, the 19.8-m (65-ft), steel hull R/V Kaho (USGS) and the 12.8-m (42-ft), fiberglass hull R/V Seth Green (NYSDEC). During 1983 - 1985, all assessment trawling was conducted by the Kaho (the fiberglass Seth Green was permanently retired in fall 1982). In 1985, the NYSDEC accepted delivery of a new R/V Seth Green and this 14-m (46-ft), steel hull vessel participated with the Kaho in prey fish surveys during 1986-2002 and in 2004-2008.

Acoustic surveys were conducted in late summer/early fall during the years 1992-1996 and 2001-2011 to estimate pelagic prey fish biomass in Lake Michigan. Midwater trawling during the surveys as well as target strength provided a measure of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. The 2011 survey provided data from 24 acoustic transects (442 km) and 26 midwater tows. Mean total prey fish biomass was 4.8 kg/ha (relative standard error, RSE = 23%) or ≈25.6 kilotonnes (kt = 1,000 metric tons), which was only 24% of the estimate for 2010 and 16% of the long-term mean. The decrease from 2010 was the largest single-year decrease in the time series and resulted from decreased biomass of age-1 and older alewife as well as a weak 2011 year class. The weak 2011 year class was the result of very low spawner densities rather than early mortality. The 2011 alewife year-class contributed <1% of total alewife biomass (3.5 kg/ha, RSE = 25.0%), while the 2010 alewife year-class contributed ≈68%. In 2011, alewife comprised 72% of total prey fish biomass, while rainbow smelt and bloater were 16 and 12% of total biomass, respectively. Rainbow smelt biomass in 2011 (0.75 kg/ha, RSE = 38%) was similar to biomass in 2010 (0.6 kg/ha). Bloater biomass was much lower (0.6 kg/ha, RSE = 31%) than in the 1990s, and mean density of small bloater in 2011 (4 fish/ha, RSE = 23%) was the lowest observed in any acoustic survey on record. Previous high densities of small bloater (2007-2009) appear to have resulted in minimal recruitment to larger sizes. In 2011, pelagic prey fish biomass in Lake Michigan was much lower than in Lake Huron for the first time in the eight years in which acoustic surveys were done in both lakes. Prey fish biomass remained well below the Fish Community Objectives target of 500-800 kt, and key native species remain absent or rare.

Assessment of prey fish stocks in the Great Lakes using bottom trawls have been conducted annually since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. Prey fish stock assessments differ among lakes in the proportion of a lake covered, seasonal timing, bottom trawl gear used, sampling design, and the manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspects, a direct comparison of prey fish catches among lakes is problematic. All of the assessments, however, produce indices of abundance or biomass that can be standardized to facilitate comparisons of trends among lakes and to illustrate present status of the populations. We present population indices for important prey fishes in the Great Lakes standardized to the highest value for a time series within each lake: cisco (Coregonus artedi), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). We also provide indices for round goby (Neogobius melanostomus), an invasive fish presently spreading throughout the basin. Our intent is to provide a short, informal report emphasizing data presentation rather than synthesis; for this reason we avoid use of tables and the need to cite references.

Gear bias is always a concern when conducting fishery assessments. Bottom trawls are reliable tools for measuring relative abundance and biomass of fish living near the lake bottom. Inter-annual variation in the proportion of a fish population collected near the lake bottom will, of course, result in some measurement error. The types, sizes, and numbers of fish caught by the trawl are influenced by many variables, such as dimensions of the net and speed at which it is towed. Nonetheless, we believe that the information presented in this report is the best available long-term index of relative abundance for these fish in the Great Lakes. Our approach of indexing relative abundance estimates from bottom trawl catches to the maximum observed estimate in the time series is the only practical means of comparing relative trends across the Great Lakes and is relatively insensitive to differences in sampling design.

To make statistical comparisons of trends among lakes, we restricted data to years when all or a group of lakes were sampled. For all lakes, data from 1992, 1993, 1998, and 2000 were omitted from comparisons because of missing or atypical data were collected in one or more lakes. Specifically, 1992 and 1993 were omitted because Lake Huron was sampled 1 month earlier than the rest of the time series, 1998 was omitted because the vessel speed was too fast in lakes Michigan and Huron, and 2000 was omitted because Lake Huron was not sampled. Comparisons with Lake Erie were restricted to 1990-2010, years when surveys with a consistent sample design were conducted. Beginning with our 2010 report, a complete series of catch data from Lake Huron was made available for comparison with other lakes because fishing power corrections to the Huron data were developed to account for the use of a larger bottom trawl to conduct surveys during 1992-2010. For each lake, standardized relative indices for biomass of age-1 and older fishes and numeric density of recruits were calculated as the observed value divided by the maximum value observed in the times series. To determine whether time series for a given species were statistically “concordant” across the basin, we calculated the Kendall coefficient of concordance (W), which can range from 0 (complete discordance or disagreement among trends) to 1 (complete concordance or agreement among trends). The P-value for W provides the probability of agreement across the lakes. First, we present trends in relative biomass of age-1 and older prey fishes to show changes in populations within each lake. Then, we present standardized indices of numerical abundance of a single age-class to show changes in relative year-class strength within each lake. Indices of year-class strength reliably reflect the magnitude of the cohort size at subsequent ages. However, differences in survey timing across lakes results in adopting different methods for selecting the fish age-class that is used to index year-class strength for each species. For Lake Superior cisco, bloater, smelt, and Lake Michigan alewife, assessment of year- class strengths are based on aged fish, and for all other samples age-classes were assigned based on fish length cut-offs. Depending on the lake and species, year class strengths were assessed from density of age-0, age-1 fish, or age-3 fish, which necessitated using years 1977-2010 to include all species. Our intent with this report is to provide a cross-lakes view of population trends but not to propose reasons for those trends.

NOTE: The Lake Michigan Committee indicated that annual forage fish survey reports were not needed in 2011, given the effort spent towards State of the Lake. Nonetheless, we offered to still develop short summaries of our surveys given the historical interest in this report.

The Great Lakes Science Center has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. All seven established index transects were completed in 2010. Below, we created Table 1 to summarize the 2010 lake-wide biomass results, and place them in context with the more recent (2005-2010) and long-term (1973-2004) averages.

The Great Lakes Science Center has conducted daytime nearshore bottom trawl surveys of Lake Superior (15-80 m bathymetric depth zone) each spring since 1978 to provide long-term trends of relative abundance and biomass of the fish community. Between 27 April and 23 June 2010, 62 stations distributed around the perimeter of the lake were sampled with 12-m Yankee bottom trawls towed cross-contour. The 2010 estimate of fish community biomass was 1.37 kg/ha, the second lowest in the 33-year survey history. The low biomass estimate continues a trend of declining fish community biomass since 2005. The distribution of biomass across jurisdictions was uneven; levels in Canada East, Canada West, Michigan, Minnesota and Wisconsin waters were 2.65, 0.51, 0.42, 0.03, and 3.94 kg/ha, respectively. Dominant species in the catch, in order of relative abundance, were cisco, rainbow smelt, lake whitefish, bloater, shortjaw cisco, and siscowet lake trout. This is the first year that shortjaw cisco ranked in the top tier of prey species. Compared to 2009 levels, cisco, bloater, lake whitefish, and shortjaw cisco increased while rainbow smelt, lean lake trout and siscowet biomass decreased. Year-class strengths for the 2009 cisco and bloater cohorts were below average but ranked as the tenth strongest year class in the past 33 years. The 2010 cisco age structure was dominated by yearlings, which accounted for 98% of the ciscoes captured. Remaining ciscoes captured were composed of adults mostly from 2003 and 1998 year classes. Year class strength or rainbow smelt was the weakest in the survey record, continuing a decline that began in 2008.

Densities of small- ( 400 mm TL) lean lake trout declined to the lowest level since 1983, prior to the period of recovery. Siscowet lake trout has shown a stable pattern of variable density since 2000. For 2010, densities of small- and intermediate-size siscowet decreased slightly while densities of large siscowet increased slightly. In the 2010 survey, proportions of total lake trout density that were hatchery, lean and siscowet were 5, 33, and 62%, respectively.

The USGS Great Lakes Science Center has conducted trawl surveys to assess annual changes in the offshore demersal fish community of Lake Huron since 1973. Sample sites include five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2010 fall bottom trawl survey was carried out between 22 October – 12 November 2010. The 2010 main basin prey fish biomass estimate for Lake Huron was 29.09 kilotonnes, the second lowest estimate in the time series, and less than five percent of the maximum biomass estimated in 1987. The estimated biomass of adult alewife and rainbow smelt in 2010 were higher than 2009, but remained near the lowest observed in the time series, and populations were dominated by small fish. Estimated adult bloater biomass in Lake Huron has been increasing in recent years, and the 2010 biomass estimate was the highest observed since 1997. Biomass estimates for trout-perch and ninespine stickleback were the lowest observed in the time series. Deepwater sculpin biomass was higher than estimated in 2009, but remains near the lowest levels observed; slimy sculpins have not been captured since 2006. The 2010 biomass estimate for round goby was higher than in 2009 but remains relatively low. Wild juvenile lake trout were captured again in 2010, suggesting that low levels of natural reproduction by lake trout may be occurring. High variability in the abundance and biomass of several species may indicate that the offshore demersal fish community in Lake Huron is in an unstable state. Low prey fish abundance in Lake Huron may have continuing negative implications for populations of lake trout and Chinook salmon in the lake.

Assessments of prey fishes in the Great Lakes using bottom trawls have been conducted annually since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. Prey fish assessments differ among lakes in the proportion of a lake covered, seasonal timing, bottom trawl gear used, sampling design, and the manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspects, a direct comparison of prey fish catches among lakes is problematic. All of the assessments, however, produce indices o f abundance or biomass that can be standardized to facilitate comparisons of trends among lakes and to illustrate present status of the populations. We present indices for import ant prey fishes in the Great Lakes standardized to the highest value for a time series within each lake: cisco (Coregonus artedi), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). We also provide indices for round goby (Neogobius melanostomus), an invasive fish presently spreading throughout the basin. Our intent is to provide a short, informal report emphasizing data presentation rather than synthesis; for this reason we avoid use of tables and the need to cite references.

Although gear bias is always a concern when conducting fishery assessments, bottom trawls used to collect these data are reliable tools for measuring relative fish abundance near the lake bottom. Inter-annual variation in the proportion of a fish population collected near the lake bottom will, of course, result in some measurement error. The types, sizes, and numbers of fish caught by the trawl are influenced by many variables, such as dimensions of the net and speed at which it is towed. Nonetheless, we believe that he information presented in this report is the best available, long-term index of relative abundance for these selected fish in the Great Lakes. Our approach of indexing relative abundance estimates from bottom trawl catches to the maximum observed estimate in the time series allows comparison of relative trends across the Great Lakes and is relatively insensitive to differences in sampling design.

To make statistical comparisons of trends among lakes, we restricted the data to years when all or a group of lakes were sampled. For all lakes, data from 1992, 1993, 1998, and 2000 were omitted from comparisons because of missing data in one or more lakes. Specifically, 1992 and 1993 were omitted because Lake Huron was sampled 1 month earlier than the rest of the time series, 1998 was omitted because the vessel speed was too fast in lakes Michigan and Huron, and 2000 was omitted because Lake Huron was not sampled. Comparisons with Lake Erie were restricted to 1990-2009, years when surveys were conducted. Available for the first time this year is a complete data series for Lake Huron, the result of fishing power corrections to make the data comparable across two different bottom trawls. For each lake, standardized relative indices for biomass of age-1 and older fishes and numeric density of recruits were calculated as the observed value divided by the maximum value observed in the times series. To determine whether time series for a given species were statistically “concordant” across the basin, we calculated the Kendall coefficient of concordance (W), which can range from 0 (complete discordance or disagreement among trends) to 1 (complete concordance or agreement among trends). The P-value for W provides the probability of agreement across the lakes. First, we present trends in relative biomass of age-1 and older prey fishes t o show changes in populations within each lake. Then, we present standardized in dices of numerical abundance of a single age-class to show changes in relative y ear-class strength within each lake. Indices of year-class strength reliably reflect the magnitude of the cohort size at subsequent ages. However, because of differences in survey timing across lakes, the age class that is used for each species to index year-class strength varies across lakes and, just as surveys differ among lakes, methods for determining fish age-class differ also. For Lake Superior cisco, bloater, smelt, and Lake Michigan alewife, assessment of year- class strengths are based on aged fish, and for all other samples age-classes were assigned based on fish length cut-offs. Depending on the lake and species, year class strengths were assessed from density of age-0, age-1 fish, or age-3 fish, which necessitated using years 1977-2009 to include all species. Our intent with this report is to provide a cross-lakes view of population trends but not to propose reasons for those trends.

We sampled with acoustics (AC) and midwater trawls (MT) to determine cisco abundance in Lake Superior’s Thunder and Black bays during 8-14 November, 2009. Total abundance of spawning-size (≥ 250 mm total length) ciscoes was estimated at 6.25 million in Thunder Bay and 1.12 million in Black Bay. Exploitation fractions of market-size (≥ age 6) females from Thunder and Black bays for 2009 were estimated at 7.1% and 11.3%, respectively; below the recommended maximum annual harvest of 15% recently adopted by Lake Superior fisheries managers. Given Thunder Bay spawner densities are on a downward trajectory, and recruitment since the 2003 year-class has been low, it is likely the exploitation fractions will increase in the future. After 2010, the Ontario Ministry of Natural Resources (OMNR) will carry on the AC program as a management activity. It is likely suspended experimental gill net (GN) samples will be used to ground truth future AC samples. In 2009, we characterized the length and age structure of Thunder Bay ciscoes using both MT samples and GN samples. Females represented 49% of the MT catch, but only 39% in GN samples. Catching a smaller proportion of females in GN samples resulted in a lower female population estimate and a higher estimated exploitation fraction (10.4%) compared to MT samples (7.1%). Experimental gill net effort was limited to 10-11.8 m water column depths where midwater trawl samples also caught roughly 40% females. Ciscoes ≥ age 17 (≥ 1992 year class) were common in Black Bay, but rare in Thunder Bay suggesting:

  1. the stocks may be distinct; and
  2. total mortality of ciscoes returning to spawn in Black Bay in recent years has been lower than ciscoes returning to Thunder Bay.

Our mid-November 2009 effort to assess the Black Bay stock by sampling outside of the bay in the lake proper was deemed successful, but this should be confirmed by sampling the Black Bay region during both mid- and late-November 2010.

The Great Lakes Science Center (GLSC) has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. The resulting data on relative abundance, size structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the survey were completed in 2009. The survey provides relative abundance and biomass estimates between the 5-m and 114-m depth contours of the lake (herein, lake-wide) for prey fish populations, as well as burbot, yellow perch, and the introduced dreissenid mussels. Lake-wide biomass of alewives in 2009 was estimated at 13.03 kilotonnes (kt) (1 kt = 1000 metric tons), which was more than double the 2008 estimate. Lake-wide biomass of bloater in 2009 was estimated at 6.98 kt, which was nearly three times higher than the 2008 estimate. Rainbow smelt lake-wide biomass equaled 1.26 kt in 2009, which was nearly double the 2008 estimate. Deepwater sculpin lake-wide biomass equaled 3.73 kt, which was only 4% lower than the 2008 estimate. Nevertheless, the 2009 estimate was the lowest value in the deepwater sculpin time series. Slimy sculpin lake-wide biomass remained relatively high in 2009 (3.59 kt), increasing 72% over the 2008 level. Ninespine stickleback lake-wide biomass equaled 0.39 kt in 2008, which was nearly identical to the 2008 estimate. The final prey fish, exotic round goby, decreased by 83% between 2008 and 2009, from 3.76 to 0.63 kt. Burbot lake-wide biomass (0.90 kt in 2009) has remained fairly constant since 2002. Numeric density of age-0 yellow perch (i.e., < 100 mm) equaled 38 fish per ha, which is indicative of a relatively strong year-class. Lake-wide biomass estimates of dreissenid mussels increased by more than fivefold from 7.57 kt in 2008 to 40.79 kt in 2009. Overall, the total lake-wide prey fish biomass estimate (sum of alewife, bloater, rainbow smelt, deepwater sculpin, slimy sculpin, round goby, and ninespine stickleback) in 2009 was 29.62 kt, which represented a 52% increase over the 2008 estimate.

The Great Lakes Science Center has conducted annual daytime bottom trawl surveys of the Lake Superior nearshore (15-80 m bathymetric depth zone) each spring since 1978 to provide long-term trends of relative abundance and biomass of the fish community. Between 27April and 19 June 2009, 63 stations distributed around the perimeter of the lake were sampled with 12-m Yankee bottom trawls towed cross-contour. The lakewide mean relative biomass estimate for the entire fish community was 1.22 kg/ha which is the lowest in the 32-year survey history. Biomass across jurisdictions was relatively even; levels in Canada East, Canada West, Michigan, Minnesota and Wisconsin waters were 1.99, 1.29, 1.12, 1.12, and 0.62 kg/ha, respectively. Dominant species in the catch, in order of relative abundance, were rainbow smelt, lean lake trout, siscowet lake trout, bloater, and lake whitefish. Compared to 2008 levels, rainbow smelt, lake whitefish, bloater and cisco biomass decreased while lean lake trout biomass increased. Year-class strengths for the 2008 cisco and bloater cohorts were well below average. A decline in smelt year class strength reversed a trend of increasing strength from 2003-2008. The 2008 cisco age structure was dominated by age-6 and older fish, which accounted for 79% of the ciscoes captured.

Densities of all sizes of hatchery lake trout continued a pattern of decline observed since 1993-1996. Densities of small- ( 400 mm TL) lean lake trout has been relatively stable since 1986. Siscowet have shown a pattern of variable but increasing density since 1980. For 2009, densities of small- and intermediate-size siscowet decreased while densities of large siscowet remained unchanged. In the 2009 survey, proportions of total lake trout density that were hatchery, lean and siscowet were 5, 61, and 34%, respectively.

The U SGS Great Lakes Science Center has conducted trawl surveys to assess annual changes in the offshore demersal fish community of Lake Huron since 1973. Sample sites include five ports in U.S. waters with additional sampling near Goderich, Ontario, since 1998. This is the first time that the entire time series of abundance and biomass estimates for Lake Huron have been presented since the sampling gear was changed in 1992. The 2009 fall bottom trawl survey was carried out between 18 October – 4 November 2009. The 2009 main basin prey fish biomass estimate for Lake Huron was 16.53 kt, the lowest estimate in the time series, and less than five percent of the maximum biomass estimated in 1987. The estimated biomass of adult alewife and rainbow smelt in 2009 were the lowest observed in the time series, and populations were dominated by small fish. Adult bloater densities in Lake Huron have been increasing in recent years, but the 2009 biomass estimate was less than that estimated by the last full survey in 2007. Biomass estimates for trout-perch, ninespine sticklebacks, and slimy and deepwater sculpins in 2009 were the lowest observed in the time series. The 2009 biomass estimate for round goby was the lowest since 1998, the year after the species was first captured in the survey. Lake whitefish abundance and biomass remain depressed compared to the 1990s, but biomass appears to be increasing. No wild juvenile lake trout were captured in the 2009 survey for the first time since 2004. Relatively large numbers of juvenile walleye were captured near Au Sable Point in 2009, the first time that walleye in this size range have been captured since the inception of the survey.

The U.S. Geological Survey (USGS) and New York State Department of Environmental Conservation (NYSDEC) have cooperatively assessed Lake Ontario prey fishes each year since 1978. Bottom trawling has been conducted during spring to assess alewife Alosa pseudoharengus, during summer to assess rainbow smelt Osmerus mordax, and during autumn to assess slimy sculpin Cottus cognatus. Timing of the surveys was selected to correspond with the season in which bottom trawl catches of the target species peaked during trawling conducted May through October 1972 (Owens et al. 2003). Twelve transects were established at roughly 25-km (15.5-mile) intervals along the U.S. shoreline (Figure 1). Bottom trawling was generally conducted at all transects to assess alewife, at all transects except Fair Haven to assess rainbow smelt, and at six transects to assess slimy sculpin (Figure 1). Although each survey targets one species of fish, catches of non-target fishes are also tracked and provide information on ecologically important changes in the fish community, such as the resurgence of once abundant native species like deepwater sculpin Myoxocephalus thompsonii (Lantry et al. 2007) or increasing abundance of recently introduced invasive species like round goby Neogobius melanostomus (Walsh et al. 2007a).

The underlying principle of our original sampling plan was to concentrate sampling effort in the depth zone where the target species was most abundant by using our knowledge of each fish species’ unique bathymetric distribution. At each transect, we typically made trawl hauls at 10-m depth intervals through the range of depths occupied by the target species. In 1997, we modified the number and range of depths fished at each transect, as well as the trawling gear, in response to the invasion of zebra and quagga mussels (Dreissena polymorpha and D. bugensis, respectively, hereafter referred to collectively as dreissenids), which changed fish distribution in the early 1990s (O’Gorman et al. 2000) and resulted in bottom trawls clogged with shells. Fixed station sampling designs, such as ours, are commonly used for assessing fish populations in the Great Lakes and in northern Europe (ICES 2004). The underlying assumption is that changes in relative abundance at the fixed stations are representative of changes in the whole population. Mean abundance from fixed station surveys will not be biased if the fish are randomly distributed. We have always assumed that the fish are randomly distributed in the geographic area in which a transect is located, and because we have numerous transects spaced at regular intervals around the shore, that our abundance indices are unbiased. Acoustic sampling conducted during the 2004 - 2006 alewife bottom trawl assessments confirmed our assumption of random distribution within geographic areas. Furthermore, differences among geographic areas in densities of alewife strength targets measured with acoustics were reflected in the densities of alewife measured by bottom trawl data. However, there is no assurance that this has always been true given the large scale shifts in fish distribution since dreissenids proliferated. Although random sampling is preferable for estimating precision, the systematic, fixed-station sampling that we employ in Lake Ontario will often be optimal for generating the most precise estimate of relative abundance even though the variance of the estimated relative abundance will be biased (ICES 2004).

Acoustic surveys were conducted in late summer/early fall during the years 1992-1996 and 2001-2009 to estimate pelagic prey fish biomass in Lake Michigan. Midwater trawling during the surveys provided a measure of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. In 2005, we began sampling Mysis diluviana during the survey. The 2009 survey provided data from 22 acoustic transects (433 km), 27 midwater tows, and 14 mysid tows. Mean total prey fish biomass was 21.3 kg/ha (relative standard error, RSE = 27.5%) or ≈114 kilotonnes (kt, = 1,000 metric tons), which was 1.4 times higher than the estimate for 2008 and 1.2 times the long-term mean. The increase from 2008 was because of increased biomass of age-2 and older alewife, of which the 2005 year-class showed the largest increase. The 2009 alewife year-class contributed ≈2% of total alewife biomass (18.6 kg/ha, RSE = 29.0%), while the 2005 alewife year-classes contributed ≈34%. In 2009, alewife comprised 87% of total prey fish biomass, while rainbow smelt and bloater were 8 and 5% of total biomass, respectively. Rainbow smelt biomass in 2009 (0.95 kg/ha, RSE = 39%) was lower than biomass in 2008 (1.6 kg/ha). Bloater biomass was much lower (1.75 kg/ha, RSE = 17.3%) than in the 1990s, but mean density of small bloater in 2009 (574 fish/ha, RSE = 22.9 %) was the highest observed in any acoustic survey on record. Additionally, this was the third consecutive year of increased small bloater density. Prey fish biomass remained well below the Fish Community Objectives target of 500-800 kt and only alewife and small bloater are above or near long-term mean abundance levels. Mean density of Mysis diluviana has remained relatively constant over time with an observed range from 185 ind./m2 (RSE = 6.8%) in 2005 to 112 ind./m2 (RSE = 5.1%) in 2007, and no significant difference in mean density among years. In 2009, mean density of M. diluviana was 117 ind./m2 (RSE = 16 %).

Assessments of prey fishes in the Great Lakes have been conducted annually since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. Prey fish assessments differ among lakes in the proportion of a lake covered, seasonal timing, bottom trawl gear used, sampling design, and the manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspects, a direct comparison of prey fish catches among lakes is problematic. All of the assessments, however, produce indices of abundance or biomass that can be standardized to facilitate comparisons of trends among lakes and to illustrate present status of the populations. We present indices of abundance for important prey fishes in the Great Lakes standardized to the highest value for a time series within each lake: cisco (Coregonus artedi), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). We also provide indices for round goby (Neogobius melanostomus), an invasive fish presently spreading throughout the basin. Our intent is to provide a short, informal report emphasizing data presentation rather than synthesis; for this reason we intentionally avoid use of tables and cited references.

For each lake, standardized relative indices for annual biomass and density estimates of important prey fishes were calculated as the fraction relative to the largest value observed in the times series. To determine whether basin-wide trends were apparent for each species, we first ranked standardized index values within each lake. When comparing ranked index values from three or more lakes, we calculated the Kendall coefficient of concordance (W), which can range from 0 (complete discordance or disagreement among trends) to 1 (complete concordance or agreement among trends). The P-value for W provides the probability of agreement across the lakes. When comparing ranked index values from two lakes, we calculated the Kendall correlation coefficient (τ), which ranges from -1 (inverse association, perfect disagreement) to 1 (direct association, perfect agreement). Here, the P-value for τ provides the probability of either inverse or direct association between the lakes. First, we present trends in relative biomass of age-1 and older prey fishes to show changes in populations within each lake. Then, we present standardized indices of numerical abundance of a single age class to show changes in relative year-class strength within each lake. Indices of year-class strength reliably reflect the magnitude of the cohort size at subsequent ages. However, because of differences in survey timing across lakes, the age class that is used for each species to index year-class strength varies across lakes and, just as surveys differ among lakes, methods for determining fish age-class differ also. For Lake Superior cisco, bloater, smelt, and Lake Michigan alewife, year- class strengths are based on aged fish and age-length keys, and for all other combinations of lakes and species, age-classes are assigned based on fish length cut-offs. Our intent with this report is to provide a cross-lakes view of population trends but not to propose reasons for those trends.

The Great Lakes Science Center of the United States Geological Survey (USGS) is working cooperatively with the Ontario Ministry of Natural Resources (OMNR) on a threeyear study to develop standard procedures for acoustic and midwater trawl (AC-MT) assessments of spawning cisco Coregonus artedi that the OMNR can carry forward as a management activity. In year two (2008), we conducted an AC-MT survey of the northern shore from Nipigon Bay to Thunder Bay. Spawning-cisco (> 250 mm total length) densities were lowest near Nipigon Bay (<10/ha), moderate in and around Black Bay (15- 30/ha), and highest in Thunder Bay (118/ha). Rainbow smelt Osmerus mordax densities were highest in Nipigon (2,179/ha) and Black (3,219/ha) bays, and lowest in Thunder Bay (961/ha). We combined our AC-MT survey results with commercial catch records to estimate exploitation fractions of female cisco in Thunder Bay during the 2008 fishery at 4% for ages 1-5, 8.7% for ages 6-12, and 4.4% for ages ≥ 13. Lake Superior fishery managers recently recommended that annual exploitation of adult female lake cisco be kept below 10-15%. Recruitment of cisco since 2003 has been low and there is a strong probability the Thunder Bay stock will decline into the future. Using a simple population dynamics approach we estimated that if the current total allowable catch (TAC) quota is held constant, exploitation fractions could exceed 10% by 2010 and 15% by 2011. Our 2008 collections suggested the survey of Black Bay was likely conducted before all spawners had returned there to spawn. Our data also suggested that cisco collected in Black Bay and east of this site in mid-November may be from the same stock. During November 2009 we will attempt to get better definition of the area occupied by cisco around Black Bay and also determine when surveys should be conducted at this location.

The USGS Great Lakes Science Center conducted acoustic/midwater trawl surveys of Lake Huron during 2004-2008. The 2008 survey was conducted during September and October, and included transects in Lake Huron’s Main Basin, Georgian Bay, and North Channel. Main Basin estimates of pelagic fish density and biomass were higher in 2008 compared to surveys in 2004-2007 because of increases in both age-0 and adult bloater. Native species now comprise the majority of the Main Basin biomass. We also observed substantial increase in the abundance of threespine and ninespine sticklebacks, although they contributed little to total community biomass increase due to small size. Rainbow smelt densities and biomass appeared similar to other years, and both alewife and emerald shiner were scarce. Also notably absent was cisco which historically were an important pelagic prey fish in Lake Huron. Unlike previous surveys, we did not observe differences in fish density or biomass among Lake Huron’s basins; during 2008 both density and biomass in the North Channel, Georgian Bay, and Main Basin were similar. This appeared to be a result of increases in the Main Basin and not declines in other areas. Main Basin prey availability for salmonids will depend largely on the extent of their predation on bloater which now comprise the majority of the prey biomass there. The Georgian Bay prey biomass had almost equal proportions of bloaters and rainbow smelt, while the North Channel pelagic biomass remained dominated by rainbow smelt. The present situation in Lake Huron where bloater is relatively abundant but alewife and other prey are scarce may result in dependence on bloater as the primary prey for salmonids.

The Great Lakes Science Center (GLSC) has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. The resulting data on relative abundance, size structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the survey were completed in 2008. The survey provides relative abundance and biomass estimates between the 5-m and 114-m depth contours of the lake (herein, lake-wide) for prey fish populations, as well as burbot, yellow perch, and the introduced dreissenid mussels. Lake-wide biomass of alewives in 2008 was estimated at 8.27 kilotonnes (kt) (1 kt = 1000 metric tons), which was the smallest biomass estimate in the entire time series and 29% lower than the 2007 estimate. Lake-wide biomass of bloater in 2008 was estimated at 3.33 kt, which was the lowest estimate since 1977 and 38% lower than the 2007 estimate. Rainbow smelt lake-wide biomass equaled 0.89 kt, which was only 0.01 kt higher than 2007, which is the lowest estimate in the time series. Deepwater sculpin lake-wide biomass equaled 5.23 kt, which is the fourth straight year of declining biomass. The 2008 estimate is the second smallest in the time series, and 39% lower than the 2007 estimate. Slimy sculpin lake-wide biomass remained relatively high in 2008 (2.75 kt), increasing 25% over 2007. Ninespine stickleback lake-wide biomass equaled only 0.50 kt in 2008, which was 79% lower than the 2007 estimate. The final prey fish, exotic round goby, increased two orders of magnitude between 2007 and 2008, from 0.02 to 4.65 kt. Round gobies now represent 18% of the prey fish biomass. Burbot lake-wide biomass (0.91 kt in 2008) has remained fairly constant since 2002. Numeric density of age-0 yellow perch (i.e., < 100 mm) equaled 0.7 fish per ha, which is indicative of a relatively poor year-class. Lake-wide biomass of dreissenid mussels dropped precipitously in 2008, down to 9.47 kt, and a 96% decline from the 2007 biomass estimate. Overall, the total lake-wide prey fish biomass estimate (sum of alewife, bloater, rainbow smelt, deepwater sculpin, slimy sculpin, round goby, and ninespine stickleback) in 2008 was 25.62 kt, which was the lowest observed since the survey began in 1973.

The Great Lakes Science Center has conducted annual daytime bottom trawl surveys of the Lake Superior nearshore (15-80 m bathymetric depth zone) every spring since 1978 to provide a long-term index of relative abundance and biomass of the fish community. Between May 5 and June 14, 2008, 58 stations were sampled around the perimeter of the lake with 12-m wide bottom trawls. Trawls were deployed cross-contour at median start and end depths of 17 and 55 m, respectively. The lakewide mean relative biomass estimate for the entire fish community was 4.61 kg/ha which was similar to that measured in 2007, 4.81 kg/ha. Dominant species in the catch were lake whitefish, rainbow smelt, longnose sucker and cisco, which represented 49, 18, 11, and 7 % of the total community biomass, respectively. Compared to 2007 levels, lake whitefish and cisco biomass increased 35% and 55%, respectively, while bloater and rainbow smelt biomass declined 69% and 41%, respectively. Increased biomass of lake whitefish and decreased biomass in bloater represent trends observed since 2007; however, reversed trends in biomass were observed for cisco and rainbow smelt. Year-class strength for the 2007 cisco cohort (0.20 fish/ha) was below the long-term (1977-2007) average (73.31 fish/ha), as was year-class strength for the 2007 bloater cohort (0.33 fish/ha) compared to the long-term average (11.11 fish/ha). Smelt year class strength (226.26 fish/ha) continues a trend of increasing strength from a 31-year low of 56.75 fish/ha in 2001 and was above the long-term average of 193.81 fish/ha. The 2008 cisco age structure was dominated by age 5 and older fish, which accounted for 82% of the mean relative density. Wisconsin waters continue to be the most productive (mean total community biomass of 17.09 kg/ha), followed by western Ontario (5.40 kg/ha), eastern Ontario (3.08 kg/ha), Michigan (2.82 kg/ha), and Minnesota (0.89 kg/ha).

Densities of small (400 mm) hatchery lake trout continued a pattern of decline observed since 1993-1996 to 0.04, 0.03 and 0.01 fish/ha in 2008, respectively. Densities of small and large wild (lean) lake trout continued a decreasing trend observed since 1996-1998. From 2007 to 2008, density of small lean lake trout declined from 0.29 to 0.15 fish/ha, the lowest value since 1978. Density of large lean lake trout has been relatively stable since 1986 but more recently density declined from 0.43 fish/ha in 2006 to 0.10 fish/ha in 2008. Density of intermediate size lean lake trout showed a small increase from 0.31 in 2007 to 0.41 fish/ha in 2008. Siscowet lake trout have shown a pattern of variable but increasing density since 1980. Since 2006, densities of small and intermediate size siscowet lake trout have increased from 0.10 to 0.12 and 0.08 to 0.15 fish/ha, respectively. Densities of large siscowet lake trout have fluctuated between 0.10 and 0.07 fish/ha since 2000. In 2008 the proportions of total lake trout density that were hatchery, lean and siscowet were 8, 60, and 32%, respectively.

The U.S.Geological Survey Great Lakes Science Center has conducted trawl surveys to assess annual changes in the deepwater demersal fish community of Lake Huron since 1973. Since 1992, surveys have been carried out using a 21 m wing trawl towed on-contour at depths ranging from 9 to 110 m on fixed transects. Sample sites include five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2008 fall bottom trawl survey was carried out between October 24 and November 20, 2008 and sampled only the three northern U.S. ports at DeTour, Hammond Bay, and Alpena due to mechanical problems with the research vessel and prolonged periods of bad weather. Therefore, all data presented for 2008 are based on samples collected from these ports. Compared to previous years, alewife populations in Lake Huron remain at low levels after collapsing in 2004. Age-0 alewife density and biomass appears to have increased slightly but overall levels remain near the nadir observed in 2004. Density and biomass of adult and juvenile rainbow smelt showed a decrease from 2007 despite record-high abundance of juveniles observed in 2005, suggesting recruitment was low. Numbers of adult and juvenile bloater were low despite recent high year-classes. Abundances for most other prey species were similar to the low levels observed in 2005 - 2007. We captured one wild juvenile lake trout in 2008 representing the fifth consecutive year that wild lake trout were captured in the survey. Based on pairwise graphical comparisons and nonparametric correlation analyses, dynamics of prey abundance at the three northern ports followed lakewide trends since 1992. Density of benthic macroinvertebrates was at an all-time low in 2008 since sampling began in 2001. The decline in abundance was due to decreases in all taxonomic groups and a large reduction in recruitment of quagga mussels. Density of Diporeia at northern ports in 2008 was the lowest observed. Diporeia were found only at 73-m sites of three ports sampled in northern Lake Huron. While no lakewide estimate of prey biomass was calculated due to the limited spatial scope of the 2008 survey, existing data suggest prey biomass remains depressed. Prey available to salmonids during 2009 will likely be small alewives, small rainbow smelt and small bloaters. Predators in Lake Huron will continue to face potential prey shortages.

The U.S. Geological Survey (USGS) and New York State Department of Environmental Conservation (NYSDEC) have cooperatively assessed Lake Ontario prey fishes each year since 1978. Bottom trawling has been conducted during spring to assess alewife Alosa pseudoharengus, summer to assess rainbow smelt Osmerus mordax, and autumn to assess slimy sculpin Cottus cognatus. Timing of the surveys was selected to correspond with the season in which bottom trawl catches of the target species peaked during May to October trawling conducted in 1972 (Owens et al. 2003). Twelve transects were established at roughly 25-km (15.5 mile) intervals along the U.S. shoreline (Figure 1). Bottom trawling was generally conducted at all transects to assess alewife, at all transects except Fair Haven to assess rainbow smelt, and at six transects to assess slimy sculpin. Although each survey targets one species of fish, catches of non-target fishes are also tracked and they provide information on ecologically important changes in the fish community such as the resurgence of once abundant native species like deepwater sculpin Myoxocephalus thompsonii (Lantry et al. 2007) or increasing abundance of recently introduced invasive species like round goby Neogobius melanostomus (Walsh et al. 2007a).

The underlying principle of our original sampling plan was to concentrate sampling effort in the depth zone where the target species was most abundant by using our knowledge of each fish species’ unique bathymetric distribution. At each transect, we typically made trawl hauls at 10-m depth intervals through the range of depths occupied by the target species. In 1997, however, we modified the number and range of depths fished at each transect, as well as the trawling gear, in response to the invasion of zebra and quagga mussels (Dreissena polymorpha and D. bugensis, respectively, hereafter referred to collectively as dreissenids) which changed fish distribution in the early 1990s (O’Gorman et al. 2000) and resulted in bottom trawls clogged with shells. Fixed station sampling designs, such as ours, are commonly used for assessing fish populations in the Great Lakes and in northern Europe (ICES 2004). The underlying assumption is that changes in relative abundance at the fixed stations are representative of changes in the whole population. Mean abundance from fixed station surveys will not be biased if the fish are randomly distributed. We have always assumed that the fish are randomly distributed in the geographic area in which a transect is located, and because we have numerous transects spaced at regular intervals around the shore, that our abundance indices are unbiased. Acoustic sampling conducted during the 2004-2006 alewife bottom trawl assessments confirmed our assumption of random distribution within geographic areas. Furthermore, differences among geographic areas in densities of alewife-strength targets measured with acoustics were reflected in the densities of alewife measured by bottom trawl. However, there is no assurance that this has always been true given the large scale shifts in fish distribution since dreissenids proliferated. Although random sampling is preferable for estimating precision, the systematic, fixed-station sampling that we employ in Lake Ontario will often be optimal for generating the most precise estimate of relative abundance even though the variance of the estimated relative abundance will be biased (ICES 2004).

Acoustic surveys were conducted in late summer/early fall during the years 1992-1996 and 2001-2008 to estimate pelagic prey fish biomass in Lake Michigan. Midwater trawling during the surveys provided a measure of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. In 2005, we began sampling Mysis diluviana during the survey. The 2008 survey provided data from 24 acoustic transects (734 km), 33 midwater tows, and 39 mysid tows. Mean total prey fish biomass was 15.3 kg/ha (relative standard error, RSE = 7.6%) or ~82 kilotonnes (kt, 1,000 metric tons), which was 1.9 times higher than the estimate for 2007 but 78% lower than the long-term mean. The increase from 2007 was because of increased biomass of age-1 and age-3 alewife. The 2008 alewife year-class contributed ~12% of total alewife biomass (11.0 kg/ha, RSE = 9.0%), while the 2007 and 2005 alewife year-classes contributed ~33% and 35%, respectively. In 2008, alewife comprised 72% of total biomass, while rainbow smelt and bloater were 11 and 17% of total biomass, respectively. Rainbow smelt biomass in 2008 (1.6 kg/ha, RSE = 10.6%) was identical to the biomass in 2007 (1.6 kg/ha). Bloater biomass was again much lower (2.6 kg/ha, RSE = 15.2%) than in the 1990s, but mean density of small bloater in 2008 (534 fish/ha, RSE = 10.9) was the highest observed in any acoustic survey on record. Prey fish biomass remained well below the Fish Community Objectives target of 500-800 kt and only alewife and small bloater are above or near long-term mean biomass levels. Mysis diluviana remains relatively abundant. Mean density ranged from 185 ind./m2 (RSE = 6.8%) in 2005 to 112 ind./m2 (RSE = 5.1%) in 2007, but there was not a statistically significant difference among years.

The Lake Erie Biological Station has conducted bottom trawl assessments of fish populations in western Lake Erie near East Harbor State Park, Ohio each summer and autumn since 1961. The catches of most age-0 forage fishes in 2008 were less than their 15-year means. Mean densities for five species exceeded the long term mean. Mean density of age-0 white perch Morone americana was more than double last year’s mean and the long-term mean. Densities of both yellow perch Perca flavescens and walleye Sander vitreus were at or above their long term mean. Catches of round goby Neogobius melanostomus (all ages combined) exceeded the previous five years, but were substantially below the species’ 13-year (1996-2008) mean. For five species, mean total lengths of age-0 individuals captured in 2008 were greater than their respective 20-year means. During 1988-2007, the frequency of occurrence for yellow perch diet items was highest for benthic invertebrates and zooplankton during the summer and benthic invertebrates in autumn. During summer zooplankton and benthic invertebrates were frequently consumed by white perch. Benthic prey for white perch occurred more frequently in the last 5-10 years.

Assessments of prey fishes in the Great Lakes have been conducted annually since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. Prey fish assessments differ among lakes in the proportion of a lake covered, seasonal timing, bottom trawl gear used, sampling design, and manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspect, a direct comparison of prey fish catches among lakes is problematic. All of the assessments, however, produce indices of abundance or biomass that can be standardized to facilitate comparisons of trends among lakes and to illustrate present status of the populations. Herein we present indices of abundance, standardized to the highest value for a time series within each lake, for important prey fishes in the various Great Lakes: lake herring (Coregonus artedi), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). We also provide indices for round goby (Neogobius melanostomus), a new invasive fish presently spreading throughout the basin.

Cisco Coregonus artedi form pre-spawning aggregations in Lake Superior during November with the bulk of spawning occurring during late November through early December (Dryer and Beil 1964). Eggs are broadcast into open water (Smith 1956) with fertilized eggs settling to the lakebed (Dryer and Beil 1964). Peak hatching occurs the following May (United States Geological Survey – Great Lakes Science Center, GLSC, unpublished data). Interannual variability in year class strength is high, but tends to be synchronous across different regions of Lake Superior (Bronte et al. 2003). November 2005 sampling of Thunder Bay showed 14 year-classes were present with the oldest fish being from the 1984 year-class (Yule et al. 2008). The ciscoes sampled were predominantly from five year classes that hatched during 1988, 1989, 1990, 1998, and 2003. These same strong year-classes were found in the western arm of Lake Superior during November 2006 (GLSC, unpublished data). Growth is rapid in the first few years of life with minimal growth after age-8 (Yule et al. 2008). Ciscoes exceeding 250 mm total length (TL) are typically sexually mature (Yule et al. 2006b, 2008). Thunder Bay ciscoes have high annual survival with rates for females and males averaging 0.80 and 0.75, respectively; females have higher rates of fishing-induced mortality compared to males but lower rates of natural mortality (Yule et al. 2008). Some Lake Superior stocks are currently commercially fished with the bulk of harvest occurring during November when fishers target females for their roe. The bulk of fish are harvested from Thunder Bay using suspended gillnets with mesh sizes ranging from 79-89 mm stretch measure. Ciscoes younger then age-5 make up a very small proportion (<0.1%) of the harvest (Yule, et al. 2008).

The USGS Great Lakes Science Center conducted annual acoustic/midwater trawl surveys of Lake Huron from 2004-2007. The 2007 survey was conducted during September and October, and included transects in Lake Huron’s main basin, Georgian Bay, and North Channel. Main basin estimates of pelagic fish density were similar during all four years. Biomass estimates were significantly lower during 2006 compared to other years, but all annual estimates ranged between 3.5 and 8.8 kg/ha. During 2004-2007 alewife were virtually absent, and rainbow smelt and bloaters comprised the majority of pelagic biomass. However, emerald shiners Notropis atherinoides were prominent in 2006, while cisco Coregonous artedi were prominent in 2007. Ciscoes were absent in trawl catches in 2004-2006, but comprised about 30% of main basin pelagic fish biomass in 2007 due to their large size compared to other species. As with previous surveys during 2004 and 2005, pelagic fish density and biomass in the North Channel was significantly higher than Georgian Bay and the main basin. Native species now comprise the majority of the main basin biomass.

The Great Lakes Science Center (GLSC) has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. The resulting data on relative abundance, size structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the survey were completed in 2007. The survey provides relative abundance and biomass estimates between the 5-m and 114-m depth contours of the lake (herein, lake-wide) for prey fish populations, as well as burbot, yellow perch, and the introduced dreissenid mussels and round gobies. Lake-wide biomass of alewives in 2007 was estimated at 11.67 kilotonnes (kt) (1 kt = 1000 metric tons), which was 18% higher than in 2006. Lake-wide biomass estimates of bloater (5.39 kt) and rainbow smelt (0.88 kt) in 2007 were 59% and 63%, respectively, lower than in 2006. Bloater biomass has declined drastically since 1989, and the 2007 estimate was the lowest since 1978. The 2007 rainbow smelt lake-wide biomass estimate was the lowest biomass estimate for rainbow smelt on record. Deepwater sculpin lake-wide biomass had shown neither an increasing nor decreasing trend during 1990-2006, but then decreased from 22.86 kt in 2006 to 8.53 kt in 2007. Slimy sculpin lake-wide biomass had been steadily increasing since 2001, but then decreased from 8.16 kt in 2006 to 2.20 kt in 2007. Ninespine stickleback lake-wide biomass remained relatively high in 2007 (2.37 kt), as the species has generally increased in abundance from 1996-present compared to 1973-1995. Burbot lake-wide biomass (1.91 kt in 2007) has remained fairly constant since 2002. After a record-high 2005 year-class, numeric density of age-0 yellow perch (i.e., < 100 mm) remained relatively high (4.7 fish per ha) in 2007 compared with the 1996-2004 period. Lake-wide biomass of dreissenid mussels appeared to be leveling off in 2007 at 245.51 kt, after increasing exponentially during 2003-2006. Round goby abundance decreased from 27.7 fish per ha in 2006 to 1.0 fish per ha in 2007. Overall, the total lake-wide prey fish biomass estimate (sum of alewife, bloater, rainbow smelt, deepwater sculpin, slimy sculpin, and ninespine stickleback) in 2007 was 31.04 kt, which was the lowest observed since the survey began in 1973.

The Great Lakes Science Center has conducted annual daytime bottom trawl surveys of the Lake Superior nearshore (15-80 m depth zone) every spring since 1978 to provide a long-term index of relative abundance and biomass of the fish community. Between April 24 and June 19, 2007, 54 stations were sampled around the perimeter of the lake with 12-m wide bottom trawls. Trawls were deployed cross-contour at median start and end depths of 17 and 56 m, respectively. The lakewide mean relative biomass estimate for the entire fish community declined 29% from 6.80 kg/ha in 2006 to 4.81 kg/ha in 2007. Most of this decline was a result of decreased biomass estimates for cisco (89%) and bloater (55%). However, rainbow smelt and lake whitefish biomass increased by 61% and 13% from 2005 to 2006, respectively. Lake whitefish represented the highest percentage of biomass for the entire community (31%), followed by rainbow smelt (27%), bloater (13%), and cisco (4%). Year-class strength for the 2006 cisco cohort (0.3 fish/ha) was well below the long-term (1977- 2006) average (76 fish/ha), as was year-class strength for the 2006 bloater cohort (0.3 fish/ha) compared to the long-term average (12 fish/ha). The 2007 cisco age structure was dominated by the strong 2003 year class, which accounted for 42% of the mean relative density. Wisconsin waters continue to be the most productive (mean total community biomass of 12.60 kg/ha), followed by western Ontario (6.60 kg/ha), eastern Ontario (2.88 kg/ha), Michigan (1.98 kg/ha), and Minnesota (0.06 kg/ha).

For the first time, we examined lake trout densities from the bottom trawl series by size classes to detect population trends. Our analysis showed a strong recovery of adult lake trout in the mid-1980s preceded by increasing densities of small and intermediate size classes. Periods of increased recruitment of small ( 400 mm) adult wild lake trout have remained relatively steady since 1986, indicative of a recovered adult spawning population for 20+ years.

As a follow-up to fall 2005 cisco spawning surveys and spring 2006 larval cisco surveys in Thunder and Black bays, we conducted additional day bottom trawls in these bays during May 2007 to characterize age-1 densities and thereby assess survival from egg to larval to age-1 life stages. Based on these surveys, we estimated survival of eggs to larvae at 0.29% and 0.46%, and larvae to age-1 at 0.16% and 0.52% in Thunder and Black bays, respectively.

In May 2007 we conducted a study of diel variation in depth distribution of the principal prey fishes. We observed that prey species have discrete habitat distributions that shift from demersal or lower pelagic during day to pelagic at night (rainbow smelt, cisco, bloater, kiyi), or shift from deeper to shallower demersal habitat from day to night (lake whitefish). These patterns reveal the potential for Lake Superior prey fishes to couple resource use in nearshore and offshore zones by movement across habitats.

The U.S.G.S. Great Lakes Science Center has conducted trawl surveys to assess annual changes in the fish community of Lake Huron since 1973. Since 1992, surveys have been carried out using a 21 m wing trawl towed on-contour at depths ranging from 9 to 110 m on fixed transects. Sample sites include five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2007 fall prey fish survey was carried out between October 15 and November 4, 2007 and sampled all five U.S. ports and Goderich, Ontario. The alewife population remained at low levels after collapsing in 2004. Adult and age-0 alewife density and biomass declined from 2006 levels and remain near the historical low observed in 2004. Density and biomass of adult and juvenile rainbow smelt showed only a slight increase from 2006 levels despite record-high abundance of juveniles observed in 2005 suggesting recruitment was limited. Juvenile rainbow smelt abundance in 2007 remained slightly higher than the long-term mean for the time series since 1992. Numbers of adult bloater increased in 2007 approaching the high levels observed in the early 1990s showing evidence of recruitment, but biomass remains low due the small size distribution of the new recruits. Juvenile bloaters remained ubiquitous with abundance surpassing the previous record high observed in 2005. Abundances for most other prey species were similar to the low levels observed in 2005 and 2006. We captured 11 wild juvenile lake trout in 2007; this represents the fourth consecutive year that substantial numbers of wild lake trout were captured in the survey. Benthic macroinvertebrates have been sampled during the fall survey since 2001. Density of benthic macroinvertebrates showed an increase from 2005-2006 values due largely to increases in oligochaete and quagga mussel densities. Prey biomass available to the bottom trawl increased slightly from 2006 levels due to increases in bloater and rainbow smelt biomass. However, total prey biomass remains near the all-time low observed in 2006. The prey available to salmonids during 2008 will be small rainbow smelt and small bloaters. Predators in Lake Huron will continue to face potential prey shortages.

The U.S. Geological Survey (USGS) and New York State Department of Environmental Conservation (NYSDEC) have cooperatively assessed Lake Ontario prey fishes each year since 1978. Bottom trawling has been conducted during spring to assess alewife Alosa pseudoharengus, summer to assess rainbow smelt Osmerus mordax, and autumn to assess slimy sculpin Cottus cognatus. Timing of the surveys was selected to correspond with the season in which bottom trawl catches of the target species peaked during May to October trawling conducted in 1972 (Owens et al. 2003). Twelve transects were established at roughly 25-km (15.5 mile) intervals along the U.S. shoreline (Figure 1). Bottom trawling was generally conducted at all transects to assess alewife, at all transects except Fair Haven to assess rainbow smelt, and at six transects to assess slimy sculpin. Although each of the three surveys targets one species of fish, catches of non-target fishes are also tracked and they provide information on ecologically important changes in the fish community such as resurgence of once abundant native species like deepwater sculpin Myoxocephalus thompsonii (Lantry et al. 2007) or increasing abundance of recently introduced invasive species like round goby Neogobius melanostomus (Walsh et al. 2007a).

The underlying principle of our original sampling plan was to concentrate sampling effort in the depth zone where the target species was most abundant by using our knowledge of each fish species’ unique bathymetric distribution. At each transect, we typically made trawl hauls at 10-m depth intervals through the range of depths occupied by the target species. In 1997, however, we modified the number and range of depths fished at each transect, as well as the trawling gear in response to the invasion of zebra and quagga mussels (Dreissena polymorpha and D. bugensis, respectively, hereafter referred to collectively as dreissenids) which changed fish distribution in the early 1990s (O’Gorman et al. 2000) and resulted in bottom trawls clogged with shells. Fixed station sampling designs, such as ours, are commonly used for assessing fish populations in the Great Lakes and in northern Europe (ICES 2004). The underlying assumption is that changes in relative abundance at the fixed stations are representative of changes in the whole population. Mean abundance from fixed station surveys will not be biased if the fish are randomly distributed. We have always assumed that the fish are randomly distributed in the geographic area in which a transect is located and, because we have numerous transects spaced at regular intervals around the shore, that our abundance indices are unbiased. Acoustic sampling conducted during the 2004-2006 alewife assessments confirmed our assumption of random distribution within geographic areas and differences among geographic areas in acoustically measured density of alewife-strength targets were reflected in the densities of alewife measured by bottom trawl. However, there is no assurance that this has always been true given the large scale shifts in fish distribution since dreissenids proliferated. Although random sampling is preferable for estimating precision, the systematic, fixed-station sampling that we employ in Lake Ontario will often be optimal for getting the most precise estimate of relative abundance even though the variance of the estimated relative abundance will be biased (ICES 2004).

Acoustic surveys were conducted in late summer/early fall during the years 1992-1996 and 2001-2007 to estimate pelagic prey fish biomass in Lake Michigan. Midwater trawling during the surveys provided a measure of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. The 2007 survey provided data from 21 acoustic transects and 41 midwater tows. Mean total prey biomass was 8.2 kg/ha (RSE = 6.4%) or ~40 kilotonnes (kt) (or 1,000 metric tons), which was 32% lower than the estimate for 2006 and, largely a result of decreased rainbow smelt biomass. Total biomass in 2007 was 47% lower than in 2005, largely a result of decreased biomass of alewife and rainbow smelt. The 2007 alewife year-class contributed ~33% of total alewife biomass (4.7 kg/ha, RSE = 10.7%), while the 2005 alewife year-class contributed ~47% of alewife biomass. The remaining 20% of alewife biomass consisted of the 2000, 2002, 2003, 2004, and 2006 year-classes. In 2007, alewife comprised 57% of total biomass, while rainbow smelt and bloater were 19 and 23% of total biomass, respectively. Rainbow smelt biomass exhibited an increasing trend from 2002-2006 with the population consisting primarily of larger fish (≥ 90 mm), but rainbow smelt biomass was much lower in 2007 (1.6 kg/ha, RSE = 15.8%) than in 2006. Bloater biomass was again much lower (1.9 kg/ha, RSE = 24.0%) than in the 1990s, but mean density of small bloater in 2007 (320 fish/ha, RSE = 14.2) was the highest observed in any acoustic survey on record. Although acoustic and midwater trawl data suggest that the preyfish community is somewhat more diverse than in the previous five years, preyfish biomass remained well below the Fish Community Objectives target of 500-800 kt.

Each summer and autumn since 1961, the Lake Erie Biological Station has conducted assessments of fish populations in western Lake Erie near East Harbor State Park, Ohio, and more recently has included assessments of fish diets, zooplankton, and benthic macroinvertebrates. The catches of all major age-0 forage fishes in autumn 2007 were less than their 15-year (1993-2007) means. Catches of round goby Neogobius melanostomus (all ages combined) were similar to the previous four years, but substantially below the species’ 12-year (1996-2007) mean. For five species examined, mean total lengths for age-0 individuals in 2007 were greater than their respective 20-year means; for four species the reverse was true. Yellow perch Perca flavescens and white perch Morone americana adults (age-2 and older) consumed predominately benthic macroinvertebrates in summer and fall. Although the autumn diet of white perch was dominated by benthic macroinvertebrates, fish accounted for a relatively large proportion (41.3% by wet weight). The proportion of benthic macroinvertebrates in the summer diet of yellow perch was the highest since 2004, and dominated the autumn diet. The proportion of benthic macroinvertebrates in the summer diet of white perch has increased since 2004. Dreissena spp. dominated the benthic macroinvertebrates collected in 2007, followed by Chironomidae. Veligers of Dreissena spp. was the most abundant taxon in the summer zooplankton samples.

The two largest remaining cisco Coregonus artedi commercial fisheries on Lake Superior are supported by the Thunder and Black Bay, Ontario, stocks. The sustainability of these fisheries relies on con rolling the harvest in relation to the size of the populations. During mid-November 2005 we conducted acoustic and midwater trawl surveys of both bays to assess abundances of various year-classes at-large which we compared to commercial catches to estimate exploitation levels. At present, the cisco population in Thunder Bay is dominated by the 2003, 1998, 1990, 1989 and 1988 yearclasses. These five strong year-classes were caught previously during an annual USGS spring bottom trawl survey as age-1 fish, while the remaining year-classes were largely absent from trawl catches as yearlings. Using acoustic methods, we estimate the numbers of large (> 250 mm) ciscoes in Thunder and Black bays at 5.2 million (95% CI = 4.3 – 6.2 million) and 244,000 (95% CI = 32,000 – 456,000), respectively. We estimate that commercial fishers harvested 2.49% of males older than age-6 and 8.46% of age-6 and older females from Thunder Bay. Plots of catch-per-unit-effort (CPUE) of commercial nets versus landing date showed that our survey of Thunder Bay occurred when CPUE was high, indicating most spawners were present. The CPUE of ciscoes in Black Bay increased rapidly after our survey, suggesting our acoustic abundance estimate is conservative. Given the 1998 and 2003 year-classes are both strong and well established, we conclude that present commercial harvest quotas will not jeopardize the persistence of the Thunder Bay stock into the foreseeable future (i.e., next 5-10 years). Given that our abundance estimate from Black Bay is conservative and the estimate suffers from poorprecision, we can not comment on the sustainability of the current Black Bay harvest quota.

Assessments of prey fishes in the Great Lakes have been conducted annually since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. Prey fish assessments differ among lakes in the proportion of a lake covered, seasonal timing, bottom trawl gear used, sampling design, and manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspect, a direct comparison of prey fish catches among lakes is problematic. All of the assessments, however, produce indices of abundance or biomass that can be standardized to facilitate comparisons of trends among lakes and to illustrate present status of the populations. Herein we present indices of abundance, standardized to the highest value for a time series within each lake, for important prey fishes in the various Great Lakes: lake herring (Coregonus artedii), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). We also provide indices for round goby (Neogobius melanostomus), a new invasive fish presently spreading throughout the basin.

The USGS Great Lakes Science Center conducted acoustic and midwater trawl surveys of Lake Huron in 1997, 2004, 2005, and 2006. The 2006 survey was conducted during August and was comprised of Lake Huron’s main basin. Main basin estimates of pelagic fish density were higher in 2006, primarily due to increased age-0 rainbow smelt Osmerus mordax abundance. However, biomass estimates were lower than in 2004 and 2005 due to lower abundance of adult rainbow smelt and adult bloaters Coregonus hoyi. Alewife Alosa pseudoharengus density and biomass remained low; alewife densities were 56 fish·ha-1 and all alewives captured in trawls were age-0’s. Emerald shiner Notropis atherinoides density and biomass increased substantially throughout the main basin in 2006, and they comprised about 25% of pelagic biomass. Increased emerald shiner biomass did not offset decreases in other species, and total pelagic fish biomass estimates during 2006 were 69% lower compared with 2004. Results of the acoustic survey were consistent with observations of low prey abundance in bottom trawl surveys, and suggest that overall prey availability in the main basin may be at an all time low since inception of annual surveys in 1973.

The Great Lakes Science Center (GLSC) has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. The resulting data on relative abundance, size structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the survey were completed in 2006. The survey provides relative abundance and biomass estimates between the 5-m and 114-m depth contours of the lake (herein, lake-wide) for prey fish populations, as well as burbot, yellow perch, and the introduced dreissenid mussels and round gobies. Lake-wide biomass of alewives in 2006 was estimated at 9.86 kilotonnes (kt) (1 kt = 1000 metric tons), which was 26% lower than 2005. Lake-wide biomass estimates of bloater (13.30 kt) and rainbow smelt (2.39 kt) in 2006 were 46% and 69%, respectively, lower than in 2005. Bloater biomass has declined drastically since 1989 and the 2006 estimate is the lowest since 1978. Abundance of age-0 bloaters (< 120 mm) in 2004-2006, however, has been higher than in the previous 10-year period, perhaps signaling a bloater recovery. The 2006 rainbow smelt lake-wide biomass estimate was similar to the previous 10-year period; the 2005 estimate was the highest since 1993. Deepwater sculpin lake-wide biomass (22.86 kt in 2006) has not shown a pronounced temporal trend during 1990-2006. Slimy sculpin lake-wide biomass has been increasing since 2001, and biomass in 2006 (8.16 kt) was the highest in the overall time series. Ninespine stickleback lake-wide biomass remained relatively high in 2006 (4.05 kt), as the species has generally increased in abundance from 1996-present compared to 1973-1995. Burbot lake-wide biomass (2.05 kt in 2006) has remained fairly constant since 2002. After a record-high 2005 year-class, numeric density of age-0 yellow perch (i.e., < 100 mm) remained relatively high (5.2 fish per ha) compared to the 1996-2004 period. Lake-wide biomass of dreissenid mussels has been increasing since 2003, and the 2006 estimate (212.27 kt) was a nearly 3-fold increase over the 2005 estimate. Round goby were first captured in 2003 and have since been increasing in abundance. Round goby abundance in 2006 (27.7 fish per ha) was a 16-fold increase over the 2005 estimate. Overall, the total lake-wide prey fish biomass estimate (sum of alewife, bloater, rainbow smelt, deepwater sculpin, slimy sculpin, and ninespine stickleback) was 60.62 kt, which was the lowest observed since the survey began in 1973.

The Great Lakes Science Center conducts an annual daytime bottom trawl survey of the Lake Superior fish community every spring to provide a long-term index of relative abundance and biomass. The survey began in 1978 for U.S. waters and was expanded in 1989 to include Canadian waters. Currently, 86 fixed stations are distributed around the perimeter of Lake Superior. In 2005, a total of 51 stations were sampled with 12-m bottom trawls between April 27 and June 15. Trawls were deployed cross-contour at median start and end depths of 21 and 55 m, respectively. The lakewide mean relative biomass estimate for all species combined increased from 6.29 kg/ha in 2004 to 9.13 kg/ha in 2005. Most of this increase was a result of increased biomass estimates for lake whitefish, rainbow smelt, and lake trout. Lake whitefish made up the highest percent of the total mean biomass for any species (34%), followed by lake herring (21%), bloater (14%), and rainbow smelt (12%). Lake herring and bloater biomass remained at similar levels from 2004 to 2005. We predict that biomass estimates for lake herring in 2006 will drop as the moderately strong 2003 cohort matures and becomes less susceptible to day bottom trawling. Biomass of siscowets decreased in 2005, while wild and hatchery lake trout biomass increased by 5x and 42x, respectively. The increases were a result of increased catch of large (> 500 mm) lake trout (wild and hatchery) and higher frequency of non-zero catches in the trawls (wild lake trout). Year-class strengths for the 2004 lake herring (1 fish/ha) and bloater (< 1 fish/ha) cohorts were below their long-term (1977-2004) averages (76 and 11 fish/ha, respectively). Wisconsin waters continue to be the most productive (mean total biomass of 26.78 kg/ha), followed by western Ontario (10.96 kg/ha), eastern Ontario (2.29 kg/ha), Michigan (1.61 kg/ha), and Minnesota (0.02 kg/ha). Estimates of mean relative density from the spring survey indicate the biological impossibility of increased densities of age-2 lake herring in 2005 compared to age-1 lake herring in 2004. Age-2 rainbow smelt density in 2005 was also greater than age-1 density in 2004. These results suggest improvements to our sampling strategy and survey design should be considered. Additionally, qualitative comparisons of relative biomass estimates based on all stations versus a subset of stations reveal similar trends in biomass dynamics at the community and species levels. These results suggest reduced effort could provide similar information. Any savings in reduced effort could be directed at improving assessment strategies and/or adding new ecological investigations.

The U.S.G.S. Great Lakes Science Center has conducted trawl surveys to assess annual changes in the fish community of Lake Huron since 1973. Since 1992, surveys have been carried out using a 21 m wing trawl towed on-contour at depths ranging from 9 to 110 m on fixed transects. Sample sites include five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2006 fall prey fish survey was carried out between October 17 and November 7, 2006 and sampled all five U.S. ports and Goderich, Ontario. The alewife population remained at low levels after collapsing in 2004 due to four consecutive years of poor recruitment. Adult and age-0 alewife density and biomass declined from 2005 levels and remain near the historical low observed in 2004. Density and biomass of adult rainbow smelt continued to decline from 2004 levels and juvenile rainbow smelt declined from the record-high abundance observed in 2005. Juvenile rainbow smelt abundance in 2006 is slightly higher than the long-term mean for the time series since 1992. Adult bloater abundance increased slightly showing some recruitment, but juvenile bloaters remained ubiquitous with abundance levels second to the record high for the time series observed in 2005. Abundances for most other prey species were similar to the low levels observed in 2005. We captured 15 wild age-0 lake trout in 2006; this represents the third consecutive year that substantial numbers of wild fish were captured in the survey. Benthic macroinvertebrates have been sampled during the fall survey since 2001 and the six-year time series shows a general declining trend. Density of benthic macroinvertebrates showed a slight increase from 2005 values due largely to increases in oligochaete and quagga mussel densities. However, macroinvertebrate densities remain about half of that observed in 2002. Prey biomass available to the bottom trawl decreased slightly from 2005 levels due to declines in alewife and rainbow smelt biomass. Total prey biomass remains near the all-time low observed in 2004. The prey available to salmonids during 2007 will be small rainbow smelt and small bloaters. Predators in Lake Huron will continue to face potential prey shortages.

Acoustic surveys were conducted in late summer/fall during the years 1992-1996 and 2001-2006 to estimate prey fish biomass in Lake Michigan. Midwater trawling during the surveys provided a measure of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. The 2006 survey provided data from 13 acoustic transects and 23 midwater tows. Mean total prey biomass was 10.1 kg/ha (RSE=8.9%), similar to the estimate for 2005 (12.7 kg/ha, RSE=9.7%). The 2006 alewife year-class contributed ~10% of alewife biomass, while the 2005 alewife year-class contributed ~46% of alewife biomass. Another 36% of alewife biomass consisted of similar contributions (11-13%) of the 2002-2004 year-classes. The 1998 year-class was still present and made up 1% of biomass. In 2006, alewife and rainbow smelt biomass were similar (4.6 and 4.2 kg/ha, respectively), whereas in the previous five years rainbow smelt biomass was only 2-36% of alewife biomass. Rainbow smelt biomass has exhibited an increasing trend since 2002 and in 2006 consisted primarily of larger fish (=90 mm, 92% of rainbow smelt biomass). Bloater biomass was again much lower (1.4 kg/ha, RSE=7.1%) than in the 1990s, but mean density of small bloater in 2006 (50.0 fish/ha, RSE=16.7) was the fourth highest observed since 1992. Although acoustic and midwater trawl data suggest that the preyfish community is somewhat more diverse than in the previous five years, preyfish biomass remained well below the Fish Community Objectives target of 500-800 kt. Without recovery of the bloater population to levels in the 1980s and early 1990s, it seems likely that preyfish biomass will remain below the target level.

Each summer and autumn since 1961, the Lake Erie Biological Station has conducted assessments of fish populations in western Lake Erie near East Harbor State Park, Ohio, and more recently has included assessments of fish diets, zooplankton, and benthic macroinvertebrates. The catches of nearly all major age-0 forage fishes in autumn 2006 were considerably less than their 15-year (1992-2006) means. Rainbow smelt (Osmerus mordax), was an exception, as catches were higher than the 15-year mean by nearly 50%. Catches of all age-0 spiny-rayed fishes caught in 2006 were less than 25% of their 15-year means. In particular, no age-0 walleye (Sander vitreus) were captured 2006. Catches of round goby (Neogobius melanostomus) (all ages combined) were similar to the previous three years and substantially below the species’ 11-year (1996-2006) mean. For most species examined, mean total lengths for age-0 individuals in 2006 were similar to their respective 15-year means. Yellow perch (Perca flavescens) and white perch (Morone americana) adults (age-2 and older) consumed predominately zooplankton and benthic macroinvertebrates in summer. The proportions of zooplankton in the diet of yellow perch and of benthic macroinvertebrates in the diet of white perch have increased since 2004. Although the autumn diet of white perch was dominated by benthic macroinvertebrates, fish composed a relatively large proportion (25.3% by wet weight). Dreissena spp. dominated the benthic macroinvertebrates collected in 2006, followed by Chironomidae. Chironomid density was the highest recorded on this survey since the first benthic samples were collected in 1997. Cyclopoid copepods were the most abundant zooplankton taxon detected during summer sampling, and Bosminidae was the most abundant zooplankton taxon detected in the autumn.

Assessments of prey fishes in the Great Lakes have been conducted annually since the 1970s by the Great Lakes Science Center, sometimes assisted by partner agencies. Prey fish assessments differ among lakes in the proportion of a lake covered, seasonal timing, bottom trawl gear used, sampling design, and manner in which the trawl is towed (across or along bottom contours). Because each assessment is unique in one or more important aspect(s), a direct comparison of prey fish catches among lakes is problematic. All of the assessments, however, produce indices of abundance or biomass that can be standardized to facilitate comparisons of trends among lakes and to illustrate present status of the populations. Herein we present indices of abundance, standardized to the highest value for a time series within each lake, for important prey fishes in the various Great Lakes: lake herring (Coregonus artedi), bloater (C. hoyi), rainbow smelt (Osmerus mordax), and alewife (Alosa pseudoharengus). We also provide indices for round goby (Neogobius melanostomus), a new invasive fish presently spreading throughout the basin. To determine whether basin-wide trends were apparent for each species, we first ranked standardized index values within each lake. When comparing indices from three or more lakes, we calculated the Kendall coefficient of concordance (W), which can range from 0 (complete discordance or disagreement among trends) to 1 (complete concordance or agreement among trends). The P-value for W provides the probability of agreement across the lakes. When comparing indices from two lakes, we calculated the Kendall correlation coefficient (τ), which ranges from -1 (inverse association, perfect disagreement) to 1 (direct association, perfect agreement). Here, the P-value for τ provides the probability of either inverse or direct association between the lakes. First, we present trends in relative biomass of age-1 and older prey fishes to show changes in populations within each lake. Then, we present standardized indices of numerical abundance of a single age class to show changes in relative year-class strength within each lake. Indices of year-class strength are intended to reliably reflect the magnitude of the cohort size at subsequent ages. However, because of differences in survey timing across lakes, the age class that is used for each species to index year-class strength varies across lakes.

Over the last century humans have harvested tens of thousands of metric tons of lake herring Coregonus artedi from Lake Superior. The commercial harvest of lake herring in Wisconsin waters of Lake Superior peaked in the 1940s at 2,450 metric tons annually and harvest remained high through the early 1960’s when operators targeted herring flesh (Selgeby 1982) mostly during the spawning season (MacCallum and Selgeby 1987). Through the 1960s, annual harvest dropped markedly as stocks began to plummet. Lawrie and Rahrer (1972) suggested the lake herring decline likely resulted from successive overfishing whereby effort was first directed to spawning grounds near ports and successively to more remote grounds. By the early 1960s many stocks had likely collapsed, but the decline in readily-accessible stocks was not apparent from harvest statistics that remained relatively stable as more remote stocks were fished. Selgeby (1982) later analyzed commercial catch data from Wisconsin waters and concurred with the Lawrie and Rahrer’s (1972) sequential overfishing hypothesis. Following the collapse, annual harvest of lake herring in Wisconsin waters generally remained low (roughly 100 metric tons), showed a modest increase in the early 1990’s (roughly 230 metric tons), but has since steadily declined back to a level of 100 metric tons.

The USGS Great Lakes Science Center conducted acoustic/midwater trawl surveys of Lake Huron in 1997, 2004, and 2005. The 2005 survey was conducted during September-October, and included transects in Lake Huron’s main basin, Georgian Bay, and the North Channel. Estimates of pelagic fish density were higher in 2005 than 2004, primarily due to increased rainbow smelt abundance in the North Channel and a lake-wide increase in age-0 bloater abundance. However, biomass estimates did not differ significantly between the two years, and were lower than estimates from the initial survey in 1997. Alewife density and biomass remained low; alewife density in 2005 was about 1 fish per hectare, compared with a density of 580 fish per hectare in 1997. Slight increases in the density of sticklebacks and emerald shiners during 2005 did not contribute substantially to pelagic fish biomass. Between 1997 and 2004-05, Lake Huron’s pelagic fish biomass decreased from about 150 kg per hectare to 10-20 kg hectare. This was due to almost complete disappearance of alewife and a change in age and size structure in the bloater population from dominance by adults to increased prevalence of smaller or younger bloaters. Rainbow smelt density varied among regions and North Channel density and biomass may have been higher than overall lake-wide densities observed during 1997. Lake Huron appears to have lost a substantial amount of pelagic fish biomass between 1997 and 2004-05. Results of the acoustic survey support observations of lower prey abundance observed in bottom trawl surveys, but suggest that overall prey availability remains higher in Georgian Bay and the North Channel than in the main basin.

The Great Lakes Science Center (GLSC) has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven index transects. The resulting data on relative abundance, size structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the lake-wide survey were completed in 2005. Lake-wide biomass of alewives in 2005 was estimated at 13.401 kilotonnes (kt) (1 kt = 1000 metric tons), which was very similar to the lake-wide biomass estimate of 13.721 kt for alewives in 2004. Catch of adult alewives in 2005 was dominated by the 2002 year-class. Lake-wide biomasses of deepwater sculpin, bloater, rainbow smelt, and slimy sculpin in 2005 were estimated at 32.030 kt, 24.546 kt, 7.816 kt, and 5.474 kt, respectively. Bloater biomass drastically declined between 1989 and 2005. Abundance of juvenile bloaters steadily increased during 2003-2005, perhaps signaling the start of a bloater recovery. Rainbow smelt biomass increased substantially from 1.854 kt in 2004 to 7.816 kt in 2005. Deepwater sculpin biomass has not shown a pronounced temporal trend during 1990-2005. Burbot abundance showed a slight decrease during the early 2000s. Slimy sculpin abundance showed an increasing trend during 1985-2005. Yellow perch year-class strength in 2005 was the highest on record dating back to 1973. The lake-wide biomass estimate of dreissenid mussels was relatively high in 2005. Total catch of round gobies in 2005 was 37 fish, compared with 26 fish in 2004.

The Great Lakes Science Center conducts an annual daytime bottom trawl survey of the Lake Superior fish community every spring to provide a long-term index of relative abundance and biomass. The survey began in 1978 for U.S. waters and was expanded in 1989 to include Canadian waters. Currently, 86 fixed stations are distributed around the perimeter of Lake Superior. In 2005, a total of 51 stations were sampled with 12-m bottom trawls between April 27 and June 15. Trawls were deployed cross-contour at median start and end depths of 21 and 55 m, respectively. The lakewide mean relative biomass estimate for all species combined increased from 6.29 kg/ha in 2004 to 9.13 kg/ha in 2005. Most of this increase was a result of increased biomass estimates for lake whitefish, rainbow smelt, and lake trout. Lake whitefish made up the highest percent of the total mean biomass for any species (34%), followed by lake herring (21%), bloater (14%), and rainbow smelt (12%). Lake herring and bloater biomass remained at similar levels from 2004 to 2005. We predict that biomass estimates for lake herring in 2006 will drop as the moderately strong 2003 cohort matures and becomes less susceptible to day bottom trawling. Biomass of siscowets decreased in 2005, while wild and hatchery lake trout biomass increased by 5x and 42x, respectively. The increases were a result of increased catch of large (> 500 mm) lake trout (wild and hatchery) and higher frequency of non-zero catches in the trawls (wild lake trout). Year-class strengths for the 2004 lake herring (1 fish/ha) and bloater (< 1 fish/ha) cohorts were below their long-term (1977-2004) averages (76 and 11 fish/ha, respectively). Wisconsin waters continue to be the most productive (mean total biomass of 26.78 kg/ha), followed by western Ontario (10.96 kg/ha), eastern Ontario (2.29 kg/ha), Michigan (1.61 kg/ha), and Minnesota (0.02 kg/ha). Estimates of mean relative density from the spring survey indicate the biological impossibility of increased densities of age-2 lake herring in 2005 compared to age-1 lake herring in 2004. Age-2 rainbow smelt density in 2005 was also greater than age-1 density in 2004. These results suggest improvements to our sampling strategy and survey design should be considered. Additionally, qualitative comparisons of relative biomass estimates based on all stations versus a subset of stations reveal similar trends in biomass dynamics at the community and species levels. These results suggest reduced effort could provide similar information. Any savings in reduced effort could be directed at improving assessment strategies and/or adding new ecological investigations.

The Great Lakes Science Center has conducted trawl surveys to assess annual changes in the fish community of Lake Huron since 1973. Since 1992, surveys have been carried out using a 21 m wing trawl towed on-contour at depths ranging from 9 to 110 m on fixed transects. Sample sites include five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2005 fall prey fish survey was carried out between October 18 and November 7, 2005 and sampled all five US ports and Goderich, Ontario. The alewife population remained at low levels after collapsing in 2004 due to three consecutive years of poor recruitment. Adult alewife density and biomass remained near the historical low observed in 2004. Age-0 alewife showed a slight increase in abundance over 2004 estimates, but remain below the long-term mean for the time series. Density and biomass of adult rainbow smelt decreased from 2004 levels but juvenile rainbow smelt were at a record-high abundance in 2005. Adult bloater abundance decreased slightly, but juvenile bloaters remained ubiquitous with record high abundance for the time series. Abundances for most other prey species were lower than 2004. We captured 11 wild age-0 lake trout in 2005; this represents the second consecutive year that substantial numbers of wild fish were captured in the survey. Density of benthic macroinvertebrates was the lowest observed since collections began in 2001 due to declines in abundance of oligochaetes, Diporeia spp., and quagga mussels while sphaerid clams and chironomids increased slightly over 2004 levels. Prey biomass available to the trawl increased slightly over 2004 levels due to increases in age-0 alewife and rainbow smelt biomass. However, total prey biomass remains near the all-time low observed in 2004. The prey available to salmonids during 2006 will be small rainbow smelt and small bloaters. Predators in Lake Huron face potential prey shortages; estimates of predatory demand are now similar to estimates of prey biomass, and nearly all the remaining prey species are smaller than the adult alewives traditionally consumed.

The U.S. Geological Survey (USGS) and New York State Department of Environmental Conservation (NYSDEC) have cooperatively assessed Lake Ontario prey fishes each year since 1978. Bottom trawling has been conducted during spring to assess alewife Alosa pseudoharengus, summer to assess rainbow smelt Osmerus mordax, and autumn to assess slimy sculpin Cottus cognatus. Timing of the surveys was selected to correspond with the season when bottom trawl catches of the target species peaked during May to October trawling conducted in 1972 (Owens et al. 2003). Twelve transects were established at roughly 25-km (15.5 mile) intervals along the U.S. shoreline (Figure 1). Bottom trawling was generally conducted at all transects to assess alewife, at all transects except Fair Haven to assess rainbow smelt, and at 6 transects to assess slimy sculpin. Although each of the three surveys targets one species of fish, catches of non-target fishes are also tracked and they provide information on ecologically important changes in the fish community such as resurgence of once abundant native species (e.g. deepwater sculpin Myoxocephalus thompsoni) or increasing abundance of invasive species (e.g. round goby Neogobius melanostomus).

At each transect, trawl hauls were usually made at 10-m depth intervals through the range of depths occupied by the target species. Fixed station sampling designs, such as ours, are commonly used for assessing fish populations in the Great Lakes and in northern Europe (ICES 2004). The underlying assumption is that changes in relative abundance at the fixed stations are representative of changes in the whole population. Mean abundance from fixed station surveys will not be biased if the fish are randomly distributed. We have always assumed that the fish are randomly distributed in the geographic area in which a transect is located and, because we have numerous transects spaced at regular intervals around the shore, that our abundance indices are unbiased. However, we did not initiate acoustic sampling to test the assumption of random distribution within geographic areas until 2004 when we began an acoustic evaluation of fish distribution during the alewife assessment (see Status of Alewife, below). If the fish are not randomly distributed within geographic areas, mean abundance will be biased, although if the non-random pattern of fish distribution persists through time, the differences in mean abundance between years will be unbiased (Warren in ICES 1992). Although random sampling is preferable for estimating precision, the systematic, fixedstation sampling that we employ in Lake Ontario will often be optimal for getting the most precise estimate of relative abundance even though the variance of the estimated relative abundance will be biased (ICES 2004).

Two vessels participated in prey fish surveys during 1978-1982, the 19.8-m (65 ft), steel hull R/V Kaho (USGS) and the 12.8-m (42 ft), fiberglass hull R/V Seth Green (NYSDEC). During 1983-1985, all assessment trawling was conducted by the Kaho (the fiberglass Seth Green was permanently retired in fall 1982). In 1985, the NYSDEC accepted delivery of a new R/V Seth Green and this 14-m (46 ft), steel hull vessel participated with the Kaho in prey fish surveys during 1986-2002 and in 2004.

Acoustic surveys were conducted in the fall during the years 1992-1996 and 2001-2005 to estimate prey fish biomass in Lake Michigan. Midwater trawling during the surveys provided a measure of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. The 2005 survey included 31 transects and 62 midwater tows. Alewives were the dominant species in the 2005 trawl catch by mass, followed by rainbow smelt then bloater. Alewife, rainbow smelt, bloater, and yellow perch all produced abundant young in 2005, with YOY alewife density highest since 1995. Bloater and yellow perch YOY were more abundant in 2005 than in any other year of the survey. Numeric density of YOY alewife in 2001-2005 was positively correlated with May-August surface water temperature (r2=0.95). Total prey biomass (alewife, rainbow smelt, bloater, sticklebacks, and yellow perch) was 2.2x higher in 2005 (69 kt) than in 2004, but was only ~40% of 2001 biomass. Of the lakewide prey fish total of 69 kt, 49% was alewife. In 2005, there was evidence of spatial structure at multiple scales, with small-scale autocorrelation in density occurring up to ~5 km accompanied by large-scale (regional) differences in distribution. The regional differences consisted of spatial segregation among species and between size groups within species in 2004-2005. Numeric fish density was again highest in offshore regions (north and south offshore), but biomass density was highest in nearshore areas (depths <100 m).

Each summer and autumn since 1961, the Lake Erie Biological Station (LEBS) has conducted assessments of fish populations in western Lake Erie near East Harbor State Park, Ohio, and more recently has included assessments of fish diets, zooplankton, and benthic macroinvertebrates. The catches of major age-0 forage fishes (gizzard shad, [Dorosoma cepedianum], alewife [Alosa pseudoharengus] and emerald shiner [Notropis atherinoides]) fell below their 15-year (1991-2005) means. In contrast, the catches of spottail shiner (Notropis hudsonius) and trout-perch (Percopsis omiscomaycus), were higher than the long-term means. Catches of all major age-0 spiny-rayed fishes (yellow perch [Perca flavescens], walleye [Sander vitreus], white perch [Morone americana], white bass [Morone chrysops], and freshwater drum [Aplodinotus grunniens]) in 2005 were lower than the 15-year means. Catches of round goby (Neogobius melanostomus) (all ages combined) were lower than in any year since the species was first captured (1996). Mean lengths of emerald shiner, spottail shiner, yellow perch, walleye, white perch, and white bass were greater than their respective long-term means. In summer 2005, yellow perch diets were dominated by crustacean zooplankton and benthic macroinvertebrates, whereas white perch diets were dominated by zooplankton. Autumn 2005 diets for yellow perch consisted mostly of benthic macroinvertebrates and fish, while white perch consumed higher proportions of fish and zooplankton. Hexagenia sp. nymphs were the dominant taxonomic group in the fall diet of yellow perch. Zooplankton have contributed a higher mean percent weight of the diet in the past three years for both yellow perch and white perch. Benthic macroinvertebrates made up the largest proportion of the autumn diets for yellow perch during 2001-2004 and for white perch during 2003-2004. Dreissena sp. made up the largest portion of the benthic community near East Harbor State Park (mean = 62%), followed by Gastropoda. Calanoid copepods were the most abundant zooplankton taxon detected during both summer and fall sampling. The spring and autumn samples of yellow perch from commercial trap nets in 2005 were dominated by individuals from the 2001 year class, similar to last year.

In 2004, LEBS began an assessment of fish populations in the Ontario and Michigan waters of the western basin. Species diversity in 2005 was greater in autumn than in spring. Autumn densities of most age-0 forage fishes were higher in 2005 than in 2004. Both yellow perch and white perch consumed mainly zooplankton in spring and benthic macroinvertebrates in fall.

The Great Lakes Science Center (GLSC) has conducted lake-wide surveys of the fish community in Lake Michigan each fall since 1973 using standard 12-m bottom trawls towed along contour at depths of 9 to 110 m at each of seven to nine index transects. The resulting data on relative abundance, size structure, and condition of individual fishes are used to estimate various population parameters that are in turn used by state and tribal agencies in managing Lake Michigan fish stocks. All seven established index transects of the lake-wide survey were completed in 2004. Alewife abundance decreased substantially between 2003 and 2004, as estimated lake-wide biomass dropped from 43.234 kilotonnes (kt) (1 kt = 1000 metric tons) in 2003 to 13.721 kt in 2004. Catch of adult alewives in 2004 was dominated by the 1998 yearclass. Lake-wide biomasses of bloater, deepwater sculpin, slimy sculpin, and rainbow smelt in 2004 were estimated at 34.321 kt, 26.519 kt, 3.658 kt, and 1.854 kt, respectively. Bloater biomass drastically declined between 1989 and 2004. Abundance of juvenile bloaters in 2004 was the highest juvenile bloater abundance observed since 1991. Rainbow smelt biomass declined during 1992-1997, and has remained low since 1997. Deepwater sculpin biomass has shown neither an increasing nor decreasing trend from 1990 to 2004. Burbot abundance decreased during 2001-2003, but increased in 2004. Slimy sculpin abundance appeared to be leveling off during 1999-2004, following an increase during the 1990s. Yellow perch year-class strength was poor in 2004. Lake-wide biomass estimates of dreissenid mussels trended neither upward nor downward during 1999-2004. Total catch of round gobies in 2004 (26 fish) was similar to that in 2003 (23 fish), and catches were again limited to the Manistique and Saugatuck transects.

The Great Lakes Science Center conducts an annual daytime bottom trawl survey of the Lake Superior fish community every spring to provide a long-term index of relative abundance and biomass. The survey began in 1978 for U.S. waters and was expanded in 1989 to include Canadian waters. Currently, 87 fixed stations are distributed around the perimeter of Lake Superior. In 2004, a total of 75 stations were sampled with 12-m bottom trawls between May 10 and June 19. Trawls were deployed cross-contour at median start and end depths of 15 and 65 m, respectively. Acoustic data were also collected concurrently at 70 of the 75 stations to test the assumption that fish are primarily demersal during the day and thus susceptible to bottom trawls. The mean lakewide biomass estimate for all species combined increased from 4.71 kg/ha in 2003 to 6.29 kg/ha in 2004, halting the downward trend that began in the early 1990s. Most of this increase was a result of recruitment of the 2003 lake herring year-class, which was the seventh highest during the 27-year survey. Bloater, burbot, and longnose suckers also made minor contributions to the increase in biomass. Lake whitefish made up the highest percent of the total mean biomass for any species (30%), followed by lake herring (29%), bloater (18%), longnose sucker (7%), and rainbow smelt (5%). Lake whitefish and rainbow smelt biomass remained at similar levels from 2003 to 2004, in spite of a moderate 2003 year-class of rainbow smelt. Biomass of lean lake trout and siscowets decreased in 2004, with siscowet biomass exceeding wild lake trout biomass for the first time and hatchery lake trout reaching their lowest biomass over the time series. The 2002 and 2003 lake herring yearclasses were the largest consecutive cohorts since the 1988-1990 cohorts, though much smaller in magnitude. Bloater year-class strength increased for a second consecutive year, but still remains below the 1978-2004 average. Wisconsin waters continue to be the most productive (mean total biomass of 19.75 kg/ha), followed by western Ontario (5.33 kg/ha), Michigan (3.01 kg/ha), eastern Ontario (1.91 kg/ha), and Minnesota (0.75 kg/ha). Comparisons between acoustic techniques and bottom trawls indicate that, at a lakewide scale, mean fish biomass in the water column above the trawl path was 70% of the mean estimate derived from the bottom trawl data. Comparing estimates from both gears at western Lake Superior stations sampled concurrently in 2001, 2003, and 2004 shows high year-to-year and site-to-site variability in pelagic fish biomass during the day. The assumption that fish are primarily demersal during the day in the spring appears to be incorrect. These results suggest that fish behavior can have an impact on biomass estimates when daytime bottom trawling is the sole gear, and that relative differences in our spring survey indices may result from changes in fish behavior as well as actual changes in abundance. If absolute fish biomass estimates are a long-term goal for Lake Superior management agencies, consideration of a comprehensive lakewide survey that integrates multiple gears will be necessary.

The Great Lakes Science Center has conducted trawl surveys to assess annual changes in the fish community of Lake Huron since 1973. Since 1992, surveys have been carried out using a 21 m wing trawl towed on-contour at depths ranging from 9 to 110 m on fixed transects at five ports in U.S. waters with less frequent sampling near Goderich, Ontario. The 2004 fall prey fish survey was carried out during October and sampled all five US ports and Goderich, Ontario. The alewife population collapsed during 2004, due to three consecutive years of poor recruitment. Both adult and age-0 alewife density and biomass were at historical lows for the time series. Density and biomass of adult rainbow smelt abundance increased from 2003 levels due to strong year classes in both 2003 and 2004. Adult bloater abundance decreased slightly, but juvenile bloaters remained ubiquitous. Density of small bloaters increased during 2004 due to continued abundance of the exceptionally strong 2003 year class combined with substantial numbers of age-0 fish produced during 2004. Abundances for most other prey species were lower than 2003. We captured 22 wild age-0 lake trout in 2004; this represents the first time that substantial numbers of wild fish were captured in the survey. Prey biomass available to the trawl decreased by 65 % during 2001-2004; nearly all the decrease was due to reduced alewife biomass, and no other species has increased in abundance enough to compensate for their loss. The primary source of food for salmonids during 2005 will be small rainbow smelt. Predators in Lake Huron face potential prey shortages; estimates of predatory demand are now similar to estimates of prey biomass, and nearly all the remaining prey species are smaller than the adult alewives consumed traditionally.

We began a comprehensive re-analysis of our bottom trawl assessments, conducted annually since 1978, with a re-evaluation of the alewife Alosa pseudoharengus assessment. Although, the re-evaluation resulted in numerous changes to the calculation of alewife abundance, the new indices showed the same trends as the historical indices (Spearman rank correlation, P Osmerus mordax in 2004 were markedly higher than the record lows recorded in 2003. The increase was due entirely to a strong 2003 year class and not to a decrease in mortality rates of adult rainbow smelt -- age-2 and older smelt remain scarce. We have lost the ability to track abundance of slimy sculpin Cottus cognatus along the south shore -- dreissenid numbers now preclude towing the trawl gear historically used to assess sculpins, and trawling with other gear produced inconsistent results. One deepwater sculpin Myoxocephalus thompsonii was collected in spring 2004.

To provide annual estimates of prey fish biomass in Lake Huron, acoustic surveys were conducted in the fall during the years 1997 and 2004. The survey conducted in 1997 was limited to areas Coregonus hoyi) were the dominant component of acoustic biomass in both years, but rainbow smelt (Osmerus mordax) were numerically dominant in both years. Alewives (Alosa pseudoharengus) made up ~5% of biomass in 1997 and less than 0.05% in 2004 Acoustic estimates of total biomass, rainbow smelt biomass, and bloater biomass were significantly higher in 1997 than in 2004. This difference was not the result of limited spatial coverage. Total biomass estimated from the acoustic survey in 2004 was similar to the USGS bottom trawl survey estimate. The decline in biomass between 1997 and 2004 was of similar magnitude for both surveys. The Lake Huron preyfish community continues to be dominated by the native bloater. Density and biomass did not vary significantly among regions in 1997. In 2004, significant regional variation in rainbow smelt and YOY bloater density and biomass was observed; decreased abundance may have contributed to greater patchiness. It is unclear how biomass estimates presented here compare to recent salmonine consumption rates, but it is evident that preferred prey (alewife) is not abundant enough to support predator demand.

Acoustic surveys were conducted in the fall during the years 2001-2004 to estimate prey fish biomass in Lake Michigan. Surveys conducted in 2001-2003 were limited in coverage, but regions included in the 2004 lakewide survey represented ~94% of the lake area. Midwater trawling during the surveys provided measures of species and size composition of the fish community for use in scaling acoustic data and providing species-specific abundance estimates. Alewives (Alosa pseudoharengus) were the largest proportion of the catch in all years but 2003. No trends were observed in the proportion (by weight) of any species in the trawl catch, but acoustic estimates of total biomass and alewife biomass exhibited a decreasing trend across years. The decrease in total biomass was driven primarily by a decline in alewife biomass. Alewife biomass and variance estimates from the acoustic surveys were similar to estimates derived from the USGS fall bottom trawl survey. The decline in alewife biomass over time was of similar magnitude for both the acoustic and bottom trawl surveys, suggesting that the decrease in the acoustic estimates of biomass was not an artifact of the limited spatial coverage in 2001- 2003. Two alewife year classes (2002 and 2003) made up a large portion of the midwater catch of yearling-and-older alewives at age one. Even though the highest young-of-the-year alewife biomass occurred in 2002, only the 2003 year class made up a significant portion of the total catch beyond age 0, suggesting the 2003 year class may dominate the alewife population for the next several years. The 95% confidence interval for lakewide biomass of alewives in 2004 was 18-32 kt, with a mean of 25 kt.

Each summer and autumn since 1961, the Lake Erie Biological Station (LEBS) has conducted assessments of fish populations in western Lake Erie near East Harbor State Park, Ohio, and more recently has included assessments of fish diets, zooplankton, and benthic macroinvertebrates. The catches of major age-0 forage fishes (gizzard shad, [Dorosoma cepedianum], alewife [Alosa pseudoharengus] and emerald shiner [Notropis atherinoides]) fell below their 15-year (1991-2005) means. In contrast, the catches of spottail shiner (Notropis hudsonius) and trout-perch (Percopsis omiscomaycus), were higher than the long-term means. Catches of all major age-0 spiny-rayed fishes (yellow perch [Perca flavescens], walleye [Sander vitreus], white perch [Morone americana], white bass [Morone chrysops], and freshwater drum [Aplodinotus grunniens]) in 2005 were lower than the 15-year means. Catches of round goby (Neogobius melanostomus) (all ages combined) were lower than in any year since the species was first captured (1996). Mean lengths of emerald shiner, spottail shiner, yellow perch, walleye, white perch, and white bass were greater than their respective long-term means. In summer 2005, yellow perch diets were dominated by crustacean zooplankton and benthic macroinvertebrates, whereas white perch diets were dominated by zooplankton. Autumn 2005 diets for yellow perch consisted mostly of benthic macroinvertebrates and fish, while white perch consumed higher proportions of fish and zooplankton. Hexagenia sp. nymphs were the dominant taxonomic group in the fall diet of yellow perch. Zooplankton have contributed a higher mean percent weight of the diet in the past three years for both yellow perch and white perch. Benthic macroinvertebrates made up the largest proportion of the autumn diets for yellow perch during 2001-2004 and for white perch during 2003-2004. Dreissena sp. made up the largest portion of the benthic community near East Harbor State Park (mean = 62%), followed by Gastropoda. Calanoid copepods were the most abundant zooplankton taxon detected during both summer and fall sampling. The spring and autumn samples of yellow perch from commercial trap nets in 2005 were dominated by individuals from the 2001 year class, similar to last year.

In 2004, LEBS began an assessment of fish populations in the Ontario and Michigan waters of the western basin. Species diversity in 2005 was greater in autumn than in spring. Autumn densities of most age-0 forage fishes were higher in 2005 than in 2004. Both yellow perch and white perch consumed mainly zooplankton in spring and benthic macroinvertebrates in fall.

The objective of this study was to inventory reference locations at selected lakes in the National Parks Services Great Lakes Cluster Parks and to install uniform, scientifically defendable biomonitoring programs that could detect long-term ecological trends.

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo USA.gov logo U.S. Department of the Interior | U.S. Geological Survey
URL: https://www.glsc.usgs.gov/reports
Page Contact Information: GLSC Webmaster
Page Last Modified: Thursday December 15, 2016