Cumulative Habitat Impacts of Nearshore Engineeringdavem/abstracts/05-12.pdf · Cumulative Habitat...

Cumulative Habitat Impacts of Nearshore Engineering

Guy A. Meadows1,*, Scudder D. Mackey2, Reuben R. Goforth3, David M. Mickelson4,Tuncer B. Edil5, Jonathan Fuller6, Donald E. Guy, Jr.6, Lorelle A. Meadows1,

Elizabeth Brown3, Stephanie M. Carman7, and Dale L. Liebenthal6

1Department of Naval Architecture and Marine EngineeringUniversity of Michigan

Ann Arbor, Michigan 48109

2Habitat Solutions37045 N. Ganster Road

Beach Park, Illinois 60087

3Michigan Natural Features InventoryMichigan State University Extension

Stevens T. Mason Building, PO Box 30444Lansing, Michigan 48909

4Department of Geology and GeophysicsUniversity of Wisconsin

Madison, Wisconsin 53706

5Department of Civil and Environmental EngineeringUniversity of Wisconsin

Madison, Wisconsin 53706

6Ohio Department of Natural ResourcesDivision of Geological Survey

Lake Erie Geology Group1634 Sycamore Line

Sandusky, Ohio 44870

7Conservation Services DivisionNew Mexico Department of Game and Fish

PO Box 25112Santa Fe, New Mexico 87504

ABSTRACT. A multi-disciplinary, multi-institutional research team evaluated a broad range of physi-cal and biological characteristics at six Great Lakes nearshore sites in order to develop and test a con-ceptual modeling framework to assess linkages between bluff erosion, sediment supply, coastal processes,and biological utilization of nearshore and coastal habitats. The sites were chosen to represent a broadrange of hydrogeomorphic conditions, with the objective of assessing the response of these nearshore sys-tems to anthropogenic modifications and coastal change. As a result of this 2-year field effort, new meth-ods and integrated approaches were developed to characterize, map, and assess the dynamic nature ofthe nearshore zone (area generally less than 10 m water depth). Thus, these data provide an initial quan-titative assessment of nearshore change. In addition, our data indicate that shoreline modifications haveled to cumulative impacts that have irreversibly modified Great Lakes nearshore coastal habitats and the

J. Great Lakes Res. 31 (Supplement 1):90–112Internat. Assoc. Great Lakes Res., 2005

*Corresponding author. E-mail: [email protected]

90

Cumulative Nearshore Impacts 91

INTRODUCTION

Great Lakes coastlines have been subject to in-tensive coastal development (Christie et al. 1987,Steedman and Regier 1987, Edsall 1996, Edsall andCharlton 1997). In response to this coastal develop-ment, Great Lakes nearshore areas have been al-tered to maintain commercial navigation andprotect property threatened by coastal erosion.These alterations include the construction of largestructures to protect harbors and adjacent commer-cial infrastructure, dredging of channels to maintaincommercial and recreational navigation, and theemplacement of erosion-control structures to pro-tect both private and public property. These struc-tures typically reduce bluff recession, but over thelong term, may lead to: the reduction or eliminationof beaches and barrier systems, the loss ofnearshore sand substrates, and an increase inlakebed down cutting and water depths in nearshoreareas (e.g., Shabica and Pranschke 1994, Nairn andWillis 2002). The net effect of these “improve-ments” has been to alter the natural coastalprocesses that create and maintain drive erosion andsediment transport, and therefore the nature and ex-tent of nearshore habitats of Great Lakes shorelines.

These changes directly threaten the Great Lakesecosystem by impacting coastal marshes and wet-lands, reducing Great Lakes water quality, alteringhabitat heterogeneity, and impacting fish spawningand nursery habitats (Regier and Hartman 1973,Steedman and Regier 1987, Leslie and Timmins1993, Kelso et al. 1996, Brazner and Beals 1997).The destruction of barrier systems results in the lossof adjacent wetlands through direct erosion due towave attack. Restoration or protection of these wet-lands may involve armoring of the barrier system,which alters the connectivity, hydrology, and func-tion of the wetland. The resulting loss of connectiv-ity between coastal wetlands and nearshore areascan have serious implications for the reproductiveand recruitment success of many Great Lakes fishes(Brazner et al. 2001). Loss of sand and gravel sub-

strates can reduce potential spawning, nursery, resi-dential, and migratory fish habitats in nearshoreareas (Goodyear et al. 1982), and changes in ba-thymetry alter water circulation patterns, whichmay in turn affect water temperature, turbidity,available prey, and distributions of larval and juve-nile fish.

RATIONALE

The goal of this research was to implement amulti-disciplinary pilot study to develop and test aconceptual framework to identify potential cumula-tive physical and ecological impacts of anthro-pogenic shoreline modifications along Great Lakescoasts. Such a framework would assist Great Lakescoastal planning and resource management agenciesby establishing a set of criteria and pathways neces-sary to identify critical nearshore areas, functions,or processes need of protection, rehabilitation, andrestoration. Moreover, understanding the long-terminteractions between physical and biologicalprocesses will allow resource management agenciesto evaluate potential long-term effects of currentpolicies and resource management decisions onnearshore coastal processes, habitat, and the biolog-ical organisms and communities that utilize thosehabitats.

Conceptual Framework Model

Development of a conceptual framework modelto assess potential cumulative impacts of shorelinemodifications on nearshore coastal habitats requiresan understanding of four fundamental system com-ponents: 1) the energy of the system; 2) the under-lying geology of the system; 3) the hydrology of thesystem; and 4) the biology and ecology of the sys-tem. As shoreline modifications typically involvealteration of the physical characteristics of the sys-tem, a major component of this work has been todevelop tools and methods to measure critical vari-ables that will allow us to develop models to predict

processes that create and maintain them. Of special note is our observation that altered nearshore sub-strate dynamics resulting from shoreline modifications may enhance the colonization success oflithophilic aquatic invasive species in nearshore areas of the Great Lakes. Continued development of theshoreline may exacerbate changes in Great Lakes nearshore food-web structures and ecosystem services.Further study and monitoring of these phenomena are needed, and our work suggests that a holistic, mul-tidisciplinary approach is necessary to develop effective management strategies to address these andother issues affecting nearshore areas of the Great Lakes.

INDEX WORDS: Cumulative impacts, coastal erosion, nearshore habitat, sediment supply, sidescansonar, substrate mapping, aquatic invasive species.

92 Meadows et al.

how nearshore habitats, and the biological organ-isms and communities that utilize those habitats,will respond to specific abiotic stressors resultingfrom modifications to the shoreline. A flow dia-gram illustrating the relationships between abioticand biotic components of this framework model ispresented in Figure 1. The model is designed toidentify the fundamental processes, pathways, andlinkages between energy, geology, hydrology, andbiology that create and maintain nearshore coastalhabitats.

Specific variables evaluated in the site selectionprocess were, in part, determined by the types ofmethods and tools applied during this pilot study.For example, bluff erosion process models requireknowledge of the bluff height above the water, bluffstratigraphy and composition; bluff cross-sectionalprofiles; groundwater elevation; strength of materi-als; mechanisms of failure; and historical recessioninformation. Understanding nearshore erosion andsediment transport processes require knowledge oflocal bathymetry; wind and wave climatology; sub-strate composition and grain size; available sedi-ment supply; and distribution and thickness of

sediments. The ecological model requires knowl-edge of local fish, benthic, and zooplankton com-munities as a function of life-stage and season;water mass characteristics (e.g., turbidity, tempera-ture, water chemistry); local bathymetry; energydissipation by the bottom; substrate composition,grain size, stability, and distribution; and habitatavailability and connectivity. Finally, and perhapsmost importantly, we require knowledge of howthese interactions change with time. Hence, greatcare was exercised in selecting sites representativeof typical Great Lakes ecosystems.

Site Selection

The climatological parameters that drivenearshore coastal processes include wind, waves,and water levels (Meadows et al. 1997). However,it is the interaction of these parameters with physi-cal characteristics of the shoreline (i.e., geomor-phology, surficial materials and bedrock geology,sediment composition and texture) and anthro-pogenic modifications to the shoreline that struc-ture, create, and maintain nearshore coastalhabitats. For the purpose of this study, sites werefirst selected based primarily on hydrogeomorphiccharacteristics of the coastline, as it is those charac-teristics and the coastal processes that act on themthat are most severely impacted by anthropogenicmodifications to the shoreline.

To provide a broad basis for this study, three siteswith similar hydrogeomorphic and geological char-acteristics were identified. A summary of site char-acteristics is presented in Table 1 for all sixlocations. In addition Figure 2 provides the geo-graphical distribution of the sites within the basinand Figure 3 provides aerial views of eachnearshore region. These “similar sites” includePainesville, Ohio; St. Joseph, Michigan; and TwoRivers, Wisconsin (Fig. 1 and Table 1). These“mid-bluff” sites have cohesive bluffs that rangefrom 10 to 30 m in height, have similar geomorpho-logical and geological characteristics, and have anopen coast exposure. Sediments at these sites typi-cally consist of well-indurated cohesive clays—pri-marily lacustrine silts and clays, glacial till, and/orbedrock, overlain by thin mobile sand and graveldeposits. The physical characteristics of these sitesare typical of many Great Lakes coastlines.

A second set of three sites was chosen based onthe unique nature of their physical and biologicalcharacteristics. These “unique” site locations in-cluded Sheldon Marsh, Ohio; Ludington, Michigan;

FIG. 1. Flow diagram illustrating relationshipsbetween fundamental abiotic and biotic compo-nents of nearshore aquatic habitats.


and Port Washington, Wisconsin. The specificphysical and/or hydrogeomorphic characteristicsthat make these sites unique are listed in Table 1.

Note that for all of these sites, the presence or ab-sence of shore protection was also a factor in theselection process. We attempted to select sites witha range of different types of shore protection: fromgroin fields at Ludington; sheet-pile seawalls, revet-ments, and groins at St. Joseph; armored headlandsand revetments at Painesville and Sheldon Marsh,and relatively unprotected sites at Two Rivers andPort Washington, Wisconsin. Brief site descriptionsare provided that characterize the wave climate andgeological and geomorphological characteristicspresent at each of these sites. Representative pho-tographs are presented in Figure 3. Sites that havesimilar hydrogeomorphic characteristics aregrouped together and described first (“mid-bluff”sites). Followed by descriptions of the three“unique sites” (Table 1).

Painesville, OH

Painesville, Ohio, is located in the mid-longituderegion of the southern coast of Lake Erie. The

shoreline orientation is northeast to southwest andthe site is facing the northwest with the longestfetches to the northeast and west. During the mostcommon wind events, the wind blows from the westover an expansive fetch creating high-energywaves. Consequently, the highest energy waves arealso the most common and this site exhibits thelargest average wave height of all of the surveysites (Table 2). The next most significant wave ap-proach direction is from the northeast. These wavesare also characterized by large mean and maximumwave heights but not to the magnitude or frequencyof the westerly waves. Therefore, high energywaves approach along either side of the shoreline,predominantly from the west.

The ~16 m high cohesive bluff consists mostly ofcompact, Ashtabula till, water-lain diamicton, andthinly-bedded lakebed sediments. The base of thebluff is severely undercut and is undergoing activeerosion. Much of the bluff in this reach is sparselyvegetated. Shallow slides and slumps dominate thepresent failure mode, although large slumps haveoccurred in the recent past and there may be move-ment on these failure surfaces as well. This appearsto be the most rapidly eroding bluff in the study.

TABLE 1. Hydrogeomorphic basis for selecting similar (upper portion of table) vs. unique sites (lowerportion of table).

DominantBluff Nearshore

Great Geomorphic Height Surficial PhysicalSite Location Lake Setting Exposure (m) Materials Protection Characteristics

Painesville, OH Erie Open Coast NW 16 Thin Sand/ Armored Similar(SLE) (Open Coast) Cohesive Clay Headlands

St. Joseph, MI Michigan Open Coast WNW 30 Thin Sand/ Armored Similar(ELM) (Open Coast) Cohesive Clay

Two Rivers, WI Michigan Open Coast E 9 Thin Sand, Unprotected Similar(WLM) (Open Coast) Cohesive Clay

Sheldon Marsh, OH Erie Embayed NNE < 1 Thin Sand/ Armored Organic Peat(SLE) Barrier/ Lacustrine Clay, Headlands and Embayed

Wetland Organic Peat Wetland Complex Complex

Ludington, MI Michigan Low-relief WSW < 1 Sand Sheet Unprotected Sand Sheet(ELM) Dune Complex and Low-relief

Dune Complex

Port Washington, WI Michigan Open Coast ESE 43 Thin Sand/ Unprotected Rapidly Eroding(WLM) Cohesive Clay High Bluff

94 Meadows et al.

Discontinuities in the glacial till influenced theblock-type failure mode of the Painesville bluffs.Painesville has steel and rock revetments to thewest with several rock groins nearby and rubblestrewn throughout the nearshore zone of the west-ern half of the site. Several examples of “low-costshore protection” can also be found in this reach.The eastern half of the site is unprotected.

St. Joseph, MI

The site is near St. Joseph in the southwest cor-ner of Michigan near the Indiana border on thesouthwest edge of Lake Michigan. Since this sitefaces the direction of the long northwest fetch, thelargest waves approach nearly perpendicular to theshore under the influence of the prevailing westerlywinds. The most frequent wave approach is from

the west. Waves approaching directly from thenorth are relatively infrequent.

The bluffs at St. Joseph have historically suffereda great deal of bluff recession. The ~30 m high co-hesive bluff consists of mostly till in the upper partof the bluff with sand below. In the southern part ofthe reach the bluff is entirely sand. The bluff ispartly vegetated with mature trees. In the past, un-dercutting by waves has removed sand and silt fromlow on the bluff and the upper cohesive diamictonhas collapsed. Most of the site, however, is pro-tected by rock revetments or other shore-protectionstructures (e.g., rock, steel, old cars, tires, concreterubble) in an attempt to protect the fragile bluffslope. There is a groin field on the north end of thesite. The site is bounded to the north by the largeharbor structures at St. Joseph/Benton Harbor. Thisstructure acts as a sediment trap with significant

FIG. 2. Site location diagram showing “similar” and “unique” sites located in Ohio, Michigan,and Wisconsin.


FIG. 3. Photographs of the six survey sites chosen for this investigation; A) Two Rivers, WI(Lake Michigan), B) Sheldon Marsh, OH (Lake Erie), C) Port Washington, WI (Lake Michigan),D) Painesville, OH (Lake Erie), E) Ludington, MI (Lake Michigan), F) St. Joseph, MI (LakeMichigan).

96 Meadows et al.

volumes of sediment accreted on the updrift (north)side of the structure. The navigation channel is peri-odically dredged by the U.S. Army Corps of Engi-neers (USACE) and a portion of the dredgedmaterials is placed in a feeder beach on the southside of the structure.

Two Rivers, WI

The survey site at Two Rivers faces east, and it islocated south of the Door Peninsula. The shorelineat this site is in the lee of the prevailing westerlywind direction. Waves produced by westerly windsare directed offshore and are commonly high in en-ergy. This site is located in the mid-latitude regionof Lake Michigan and, therefore, is affected by thelong fetches to the northeast and the southeast.Hence, the larger waves approach from eithernortheast or southeast along the shoreline. Wavesapproaching perpendicular to shore have lowerwave heights than the oblique waves. The longfetch to the southeast coupled with the south-northshore orientation produces frequent waves (~30%)with maximum wave heights from the southeast.

The ~9 m high cohesive bluff has receded fairlyrapidly over the past 20 years. The bluff fails byshallow slides and flows, and small shallow slumps.It appears to retreat in a parallel manner. The bluffis partially vegetated with grasses and a few mature

woodlots located on the bluff crest. Clayey TwoRivers till is present along all of the upper third ofthe bluff. The middle third of the bluff is composedof lakebed sediments primarily interbedded silt andclay. The lower third of the bluff consists of theclay-rich Haven till, which is commonly coveredwith slump material, especially during periods oflower water levels. The Two Rivers site is centeredon a boat launch ramp that is protected by rockrevetments to the north and south. The remnants ofa wooden pier are evident in the nearshore zonenorth of the ramp. There are large rock revetmentslocated in front of the nuclear power plant locatedimmediately south of the site.

Sheldon Marsh, OH

The shoreline of this reach faces northeast and islocated southeast of the entrance to Sandusky Bay.At one time this area was protected by a barrier-is-land complex separating the bay from Lake Erie.This site has the most unique wave climate of all ofthe sites. Due to very limited fetch length, it is no-tably the calmest site, exhibiting the lowest maxi-mum wave height (3.6 m) among sites. Althoughthe maximum wave height is lower than other sites,the average wave height of 0.7 m is similar to theother sites when compared to the overall study av-erage of 0.75 m. The most frequent wind directions

TABLE 2. Summary of hindcast wave data from the United States Army Corps of Engineers Wave Infor-mation Study (Driver et al. 1991, Hubertz et al. 1991) for each of the survey sites.

Mean Mean Max MaxSignificant Wave Significant wave Dominant

WIS Station Wave Period Wave Period LongshoreSite (lat/long) Height (m) (sec) Height (m) (sec) Transport

Painesville, OH E13 0.9 4.1 4.6 10.0 From west(41.87N/81.13W)

St. Joseph, MI M58 0.8 4.0 6.1 10.0 From north(42.08N/86.58W)

Two Rivers, WI M18 0.6 3.8 5.6 9.0 Mixed(44.27N/87.43W)

Sheldon Marsh, OH E06 0.7 3.6 3.6 9.0 From east(41.43N/82.50W)

Ludington, MI M46 0.7 4.1 7.0 10.0 Mixed(43.95N/86.65W)

Port Washington, WI M11 0.8 3.9 6.3 9.0 From north(43.27N/87.68W)


are from the south and southwest, which are di-rected offshore.

Sheldon Marsh is bounded to the east and to thewest by rock revetments. The eastern side protects agolf resort and the NASA pumping station. Thewestern revetment protects a condominium com-plex and individual homes on the Cedar PointChaussee. This revetment extends all the way to theCedar Point Causeway. An extensive marsh lies be-hind the barrier beach and there is no bluff.

Ludington, MI

The shoreline at Ludington is oriented west-southwest, facing the prevailing wind direction. It isalso in the mid-latitude region of Lake Michiganwith long north and south fetches. Since the shore-line is oriented slightly west of north, the shorelineis shielded from the northern fetch resulting in adominant fetch to the south. Accordingly, both themaximum wave height and most frequent occur-rence come from the south-southwest.

Ludington has a major harbor structure to thesouth. Sediment transport at this site is mixed, assediment is transported approximately equally tothe north and to the south. Sections of the shorelineare protected by a vertical revetment where homesand/or public facilities are constructed near theshore. Groin fields are constructed along a portionof the site to protect the Ludington State Park ac-cess road. Many of these groins are in poor condi-tion. The northern part of the reach is immediatelysouth of the Big Sable Point Lighthouse. The light-house is fronted by a combination of rock revet-ment and sheet-pile seawall. A low relief dune fieldlies above the beach, and slopes on some dunes areactive. There is no bluff, and under present lowwater conditions, waves do not impinge on the dunecomplex.

Port Washington, WI

The survey site at Port Washington is also in themid-latitude region of Lake Michigan. This sitefaces east-southeast and is also in the lee of the pre-vailing westerly wind direction. Waves produced bywesterly winds are directed offshore and are com-monly high in energy. The greatest fetches are tothe northeast and southeast, and high-energy wavesapproach from either the northeast or southeastalong the shoreline. Waves approaching perpendic-ular to shore have lower wave heights than theoblique waves. Similar to Two Rivers, the most fre-

quent wave approach direction is from the south-east. However, the waves are smaller than at TwoRivers due to a shorter southern fetch. Approxi-mately 16 percent of the waves come from thesouth compared to 30 percent at Two Rivers.

The rapidly eroding high bluffs (~ 43 m) at PortWashington are the highest and most dramatic ofthe survey sites. Large scale, deep-seated slumps,which are then modified by shallow slides andslumps dominate the bluff face at this site. Largeblocks of bluff material typically sprawl across thenarrow sand beach and are rapidly eroded and re-moved by ensuing wave action. The upper part ofthe bluff consists of clayey diamicton that overliesthinly-bedded lakebed silts and clays. The lowerbluff is made up of clayey diamicton, but is typi-cally covered by slumped material. Portions of thePort Washington shoreline are protected by smallrock revetments at both the north and south of theends of the study area. A shore-protection structuresouth of the study site was under construction dur-ing the 1999 field season and was completed by theend of the 2000 field season.

Water Levels

The latter half of the 20th century can be charac-terized as a period of rising water levels on theGreat Lakes, with record highs in 1974 and 1986.Water levels for Lake Michigan and Lake Erie areshown graphically in Figure 4. During this sametime period, there was also rapid development ofindividual properties along the Great Lakes shore-lines. A study by the U.S. Army Corps of Engineersfrom 1972 to 1976 showed that $170 million wasspent on shoreline protection structures (Meadowset al. 1997). Many of these structures were origi-nally constructed during the record high water yearsof the early 1970s and either rebuilt or modifiedduring the record highs of the mid-1980s and late1990s.

During this study, water levels dropped by morethan 0.2 m between the 1999 and 2000 data collec-tion periods at all of the sites except at Painesville,Ohio where water levels dropped only 0.14 m(Table 3). The smaller drop in lake level atPainesville is due to a short-term fluctuation inLake Erie water levels due to a wind-driven seicheevent over the period of time that data were col-lected at the site.

98 Meadows et al.

DATA COLLECTION

Cooperative joint field operations were con-ducted during three separate weeks for the sum-mers of 1999 and 2000. Each summer, 1 week wasdevoted to each state (two sites each) and includedthe collection of bluff, beach, nearshore profile ele-vation, and sidescan sonar data lakeward to thedepth of sediment closure (~10 m water depth).Sediment and textural data were obtained fromeach of the profile sites and integrated with sedi-ment distributions interpreted from bottom samples

and sidescan sonar data. Biological communitydata were collected separately (methods describedin Goforth and Carman 2005). NonparametricKruskal-Wallis (K-W) tests were conducted to de-termine whether density and CPUE data for eachtaxonomic and functional group were different be-tween shoreline types (unique vs. mid-bluff) andamong lake areas—Southern Lake Erie (SLE),Eastern Lake Michigan (ELM), and Western LakeMichigan (WLM)). Nearshore density and CPUEmeasures were also used to determine whether

FIG. 4. Historical changes in water levels for Lake Michigan and Lake Erie showing theextended period of above-average water levels spanning the past three decades and therecent decline in water levels starting in 1998, when the project was initiated (shaded bar). (http://www.lre.usace.army.mil/greatlakes/hh/greatlakeswaterlevels/historicdata/greatlakeshydrographs/)


these measures were different between the sub-strate stability regimes. An alpha level of 0.05 wasused for all statistical tests.

RESULTS

Bluff Recession and NearshoreSediment Contributions

Sediment flux from the bluff is dynamic; deliv-ery of sediment is highly variable and varies sea-sonally, annually, and with long-term changes inclimate and water level. An analysis of the relation-ship between recession rates and water levelchanges demonstrates that there is a strong correla-tion between water level and recession rate for low-bluff shorelines (Brown 2000, Brown et al. 2005).Lower water levels generally mean lower recessionrates for shorelines with low-relief bluffs. Forshorelines with high bluffs, there are often longtime lags between removal of material at the toe ofthe bluff and an erosion event at the top. The sedi-ment flux from high bluffs is episodic, and relatedto wave impact height and the number and fre-quency of large bluff failure events along a reach ofcoastline. For example, at the Port Washington site,large slumps may occur with a periodicity on theorder of 50 to 100 years. For high-bluff shorelines,recession rates are not directly influenced byshorter-term changes in water level (Mickelson andEdil 1998).

Sediment volumes contributed to the nearshorefor a given reach of coastline can be calculated bycombining changes in water level and historic re-cession rates. Since the stratigraphy and grain sizedistribution of bluff sediments are known, reces-sion rate (either historical or predicted) can be

translated into sediment volume delivered to thebeach under certain wave impact height conditions.The proportions of sand and gravel measured ineach stratigraphic unit were multiplied by thethickness of the stratigraphic unit (from Mickelsonet al. 1977, Chapman et al. 1997, Mackey 1995),to calculate the total volume of sand and gravel en-tering the beach from the bluff for a unit distanceof recession. Assuming parallel bluff-face retreatand using the long-term bluff recession rates deter-mined from aerial photographs, a determination ofthe average total amount of sand and gravel perlinear meter of shoreline per year contributed bybluff erosion can be made. Sediment contributionsdue to erosion of the beach and nearshore areas canbe made by assuming a constant geometry andshifting the bluff-beach-nearshore profile land-ward. This is, of course, constrained by nature ofthe materials present on the lakebed and based onbeach/nearshore measurements and observations,we can then estimate the amount of sand andgravel currently in temporary storage on the beachand in the nearshore

For example, annual contributions of sediment(sand and gravel) from the bluff and thebeach/nearshore were estimated for two reacheswith parallel and relatively constant bluff retreat,one on the Lake Michigan shoreline in Wisconsin(Two Rivers), and one on Lake Erie in Ohio(Painesville) (Mickelson et al. 2002). Both siteshave low-moderate height bluffs above the beach.The Two Rivers reach has no significant shore pro-tection and the Painesville reach is unprotected, butadjacent updrift and downdrift reaches are heavilyprotected. The results are given in Table 4 for bothsites. Although more sand is produced by erosion atPainesville, there is not a great deal more sand instorage. This suggests that the Painesville site issediment starved compared to Two Rivers. A likelycause is that much of the bluff along that part of theOhio shoreline is protected from erosion therebysignificantly reducing the available supply of sedi-ment necessary to create and maintain sandbeaches. Reductions in available sediment supplyare a direct result of shoreline modification. A re-duction in sediment supply will alter the nearshoresubstrate distributions and affect nearshore aquatichabitats. Similar calculations are not possible forSheldon Marsh and Ludington as they are low-re-lief sites and actively eroding bluffs are not presentwithin the survey area.

TABLE 3. Water levels are derived from theCenter for Operational Oceanographic Productsand Services water level observations. Water levelswere determined by averaging the water level datafrom a gouge station over the course of eachnearshore hydrographic survey period.

Site 1999 (m) 2000 (m) Diff (m)

Painesville 174.34 174.20 –0.14St. Joseph 176.40 176.17 –0.23Two Rivers 176.41 176.13 –0.28

Sheldon Marsh 174.36 174.13 –0.23Ludington 176.41 176.21 –0.20Port Washington 176.40 176.16 –0.24

100 Meadows et al.

Nearshore Coastal Processes andSediment Distribution

Precision hydrographic surveying, shore-normalnearshore profiles, and sidescan sonar were used toassess changes in beach width, slope, bathymetry,and nearshore sediment distribution. Sediment sam-ples were collected using an Ekman dredge and anunderwater video camera was used to assess sedi-ment type, composition, surface texture, and struc-ture along individual shore-normal nearshoreprofiles and from selected sites based on acousticcharacteristics of the sidescan sonar data. Sedimentsamples were described in detail and sieved forgrain size analyses. The results of these analysesare reported elsewhere (Fuller et al. 2002).

A between-year comparison of 1999 and 2000hydrographic surveys shows changes in profile ba-thymetry indicative of nearshore erosion and sedi-ment transport (Fig. 5). Previous studies along theLake Michigan shoreline have demonstrated thatduring lower water levels, substantial deepening ofthe offshore profile occurs due to increased erosionby larger waves (i.e., increased wave power) and alower wave base. This steepening of the offshoreprofile was shown to be a reliable indicator of im-pending nearshore and beach erosion events (Mead-ows et al. 1999). Table 5 contains calculated

changes in nearshore slope during the period offalling water levels encompassed by this study. Thenearshore areas of profiles with steeper offshoreslopes will likely be subject to increased erosion inshallow nearshore, beach, and bluff toe areas withthe onset of rising water levels. This occurs as thenearshore system attempts to develop a shore-nor-mal profile that is in equilibrium with incidentwave energy.

When water levels drop and the beach slope re-mains constant, there is a predictable increase inbeach width. If the predicted increase in beachwidth is greater than the actual increase in width,then one can extrapolate that erosion has occurred.However, if the expected increase is less than theactual increase, than accretion has occurred. This isa useful qualitative tool for beach analysis. Be-tween-year comparisons for 1999 and 2000 beachwidths along with calculated beach widths are illus-trated in Figure 6.

Data from the hydrographic and sidescan sonarsurveys were used to map nearshore sediment dis-tributions at all of the survey sites. In 1999, shore-parallel sidescan sonar surveys were run at the sixstudy sites. At each site, the surveyed area extendedfarther alongshore than the modeling site and ex-tended offshore to the lakeward edge of continuoussand cover, or to the lakeward end of the shore-nor-

TABLE 4. Estimated bluff sediment contribution to the beach and nearshore region per unit distance ofrecession.

Sediment SedimentLong- Sediment Production Storage

Term Bluff Produced Beach & Beach &Recession from Bluff Nearshore Nearshore

Rate (m3/yr/m (m3/yr/m (m3/yr/mSite Stratigraphic Units & Their Properties (m/yr) shoreline) shoreline) shoreline)

Unit % Dry UnitThickness Sand & Weight

Unit (m) Gravel (kN/m3)

Two Rivers, WI Two Rivers Till 3 31 19 0.76 1.6 0.43 70.0

Lacustrine Sediment 3 20 17

Haven Till 3 19 18.6

Painesville, OH Ashtabula Till 2.4 20 19

1.6 8.2 1.9 88.8Sand &Gravel 14 100 18


mal bathymetric profiles. In 2000, shore-parallelsidescan sonar surveys were run at five of the sites;poor weather prevented data acquisition at the Lud-ington site.

Additional details about the sidescan sonar tooland associated data processing, interpretation, andanalyses are described elsewhere (Mackey and

Liebenthal 2005). At all six survey sites, areas areclassified and grouped together based on theiracoustic response. Areas exhibiting unique or simi-lar acoustic characteristics were sampled multipletimes and/or observed using the underwater videocamera to identify grain size, composition, surfacetexture, and structure of the bottom materials gener-

FIG. 5. Representative shore-normal hydrographic profiles from 1999and 2000 from the Two Rivers, WI (Lake Michigan) and Painesville, OH(Lake Erie) sites.

102 Meadows et al.

ating the acoustic response. Once identified, similartypes and patterns of acoustic response were associ-ated with specific sediment or substrate types andused to create substrate distribution and changemaps. Examples of sidescan sonar mosaics, inter-preted data, and substrate distribution and changemaps are presented in Figure 7. Data on substrateareas, volumes, and the results of between-yearcomparisons for sites where data are available arealso summarized in Table 6.

The amount of sand cover (by area) at the studysites ranged from nearly 100% at Ludington to only17% at Painesville (1999 data). At Ludington(100% sand cover) and St. Joseph (71% sandcover), the abundance of sand is due to the rela-tively high proportion of sand in eroding bluffs andnearshore areas updrift from the study sites. At St.

Joseph, the abundance of sand may also be associ-ated with beach nourishment activities associatedwith the St. Joseph/Benton Harbor navigation struc-ture. At Sheldon Marsh (72% sand cover), retreat ofthe protective sand barrier in the mid-1970s createdan embayment that effectively traps and retains lit-toral sand. Moreover, this reach of the coastline canbe considered to be a depocenter for the littoral cellthat extends from Huron, Ohio west to the CedarPoint Chaussee. The amount of sand cover in 1999at Port Washington (44%), Two Rivers (38%), andPainesville (20%) is considerably less due to therelatively low proportion of sand in adjacent bluffsand abundance of cohesive clay exposed in adjacentnearshore areas. Also, at Painesville, littoral trans-port is interrupted by a large harbor complex at

TABLE 5. Percent of survey lines exhibiting an increase or decrease in nearshore beach slope for eachsite. Increases in slope generally indicate removal of sediment and higher incident wave energy condi-tions.

Beach Slope Nearshore Slope Offshore Slope

Survey Site Increase Decrease NC Increase Decrease NC Increase Decrease NC

Painesville, OH 28% 43% 28% 14% 14% 14% 14% 28% 28%St. Joseph, MI — 17% — 17% — — 83% —Two Rivers, WI 60% 40% — 40% 60% — — 40% 60%Sheldon Marsh, OH — 44% 11% 11% 44% — 11% 55% —Ludington, MI — 100% — 60% 33% — 66% 33% —Port Washington, WI 60% 40% — 60% 40% — — 40% 40%

FIG. 6. Evaluation of measured average beach width at each site during 1999 and 2000 surveys;(A) Exposed beach width; (B) Changes not attributable to water level change (inundation).


FIG. 7. Sidescan sonar mosaics and interpretations for all of the Lake Michigan sites. Sandover cohesive clay and lag gravel are present at both the Two Rivers site (A) and the Port Wash-ington site (B) while the Ludington site (C) is covered by a broad sand sheet. Figure D is anexample of an area change and centroid analysis at the St. Joseph site. Details are presented ina companion paper by Mackey and Liebenthal (2005).

104 Meadows et al.

Fairport Harbor located updrift (to the west) of thesurvey site.

Thickness of sand deposits was approximatedusing both the unprocessed paper sidescan sonarrecords and the bathymetric data collected by theUniversity of Michigan. Thickness was calculatedby extrapolating a sloping, hard surface (cohesive

clay) underneath the mobile sand substrate usingshore-normal bathymetric profiles, the distributionof sediment from bottom grab samples, and patternsof acoustic backscatter on the sidescan sonarrecords. The depth over non-sand areas minus thedepth over adjacent sand areas provides an approxi-mation of the sand thickness. The vertical thickness

TABLE 6. A comparison of sediment area and volume data derived from the 1999 and 2000 sidescansonar and bathymetric surveys at five Great Lakes sites. Sand sediment volume estimates are based on anaverage thickness of 0.5 meter (see text for discussion).

St. Two Sheldon PortAttribute Painesville Joseph Rivers Marsh Washington

Dimensions of area surveyed (alongshore × offshore) (m) 2,700 × 700 4,700 × 650 4,000 × 250 4,000 × 800 4,300 × 550

Total area surveyed (m2) 2,092,000 3,168,000 1,113,000 3,246,000 2,391,000

Sand area 1999 (m2) 421,000 2,257,000 419,000 2,347,000 1,059,000Sand area 2000 (m2) 349,000 2,142,000 218,000 2,351,000 932,000

Area with stable sand in both years (m2) 266,000 1,809,000 144,000 2,048,000 890,000Area of sand lost between 1999 and 2000 (m2) 154,000 449,000 275,000 300,000 169,000Area of sand gained between 1999 and 2000 (m2) 83,000 333,000 74,000 304,000 42,000

Total Change Area (AT) (m2) 237,000 782,000 349,000 604,000 211,000

Area of sand substrate 1999 (% of site) 20 71 38 72 44Area of sand substrate 2000 (% of site) 17 67 20 72 39

Area of stable sand substrate (% of site) 13 57 13 63 37Area of sand lost between 1999 and 2000 (%of site) 3 4 18 0 5Area that changed substrate between 1999 and 2000

(%of site) 12 25 31 19 9

Area of stable sand substrate (% of 1999 sand area) 63 80 34 87 84Area of sand lost between 1999 and 2000 (% of 1999

sand area) 37 20 66 13 16Area of sand gained between 1999 and 2000 (% of

1999 sand area) 20 15 18 13 4

Area Change Ratio (ACR) 0.56 0.35 0.83 0.26 0.20Centroid Distance of Movement—

Total Survey Area (m) 105 72 215 n/a 108Centroid Direction of Movement—

Total Survey Area (deg) 252 (WSW) 207 (SSW) 178 (S) n/a 188 (S)

Sand volume 1999 (m3) 210,500 1,128,500 209,500 1,173,500 529,500Sand volume 2000 (m3) 174,500 1,071,000 109,000 1,175,500 466,000

Volume of sand stable 1999 and 2000 (m3) 133,000 904,500 72,000 1,024,000 445,000Loss in sand volume between 1999 and 2000 (m3) 77,000 224,500 137,500 150,000 84,500Gain in sand volume between 2000 and 1999 (m3) 41,500 166,500 37,000 152,000 21,000


of the mobile sand substrate was then integratedalong the shore-normal bathymetric surveys to cal-culate the average thickness (methodology similarto that described by Nairn and Willis 2002). The re-sulting calculated sediment volumes are summa-rized in Table 6. These volumes provide aqualitative measure of the potential for lakebeddowncutting in the nearshore zone. In the case ofcohesive shorelines, the thinning or loss of sandcover in the nearshore may expose underlying co-hesive clays to erosion. This process is calledlakebed downcutting. Experimental sand bed mobi-lization studies demonstrated that there is a criticalthreshold value for sand cover (~200 m3/m) wherethe underlying cohesive clays are protected fromwave erosion by the overlying sand sheet (Nairn1992). When the volume of sand is reduced belowthis critical value, the sand sheet may be mobilizedand rapid erosion of the underlying cohesive claysubstrate may occur.

Substrate-change maps were used to infer a gen-eral direction of sand transport between 1999 and2000 for five of the survey sites (Mackey andLiebenthal 2005). At several sites, similarly shapedsand polygons were identified on mosaics for bothyears. For these, the change of shape and locationof the sand polygon was used to infer direction ofsand transport. At study sites where sand occurredas a broad sheet, direction of sand transport was in-ferred by noting the shift in position of polygonwindows within the sand sheet that exposed the un-derlying hard substrate. Sand migration was alsoquantified using area change and centroid analysisas described in a paper by Mackey and Liebenthal(2005). Results of these analyses are also presentedin Table 6. Area Change Ratios (ACR) were calcu-lated for each of the five sites. ACR values nearzero represent areas of substrate stability. Valuesbetween 0.2 and 0.5 represent areas of moderatesubstrate stability, and values of 0.5 or greater typ-ify areas with unstable substrates, i.e., highly mo-bile.

Decrease in sand cover between 1999 and 2000,expressed in percent of total area surveyed, rangedfrom 0 to 5% at Sheldon Marsh, Painesville, St.Joseph, and Port Washington but was 18% at TwoRivers. Although the decrease in sand area differedfrom site to site, the ranking of the sites based onpercent of sand cover did not change between thesurveys in 1999 and 2000. Similar patterns appearwhen comparing the area that changed substrate be-tween 1999 and 2000. Values ranged from 9 to 31%(Table 6).

The decrease in sand cover at Two Rivers (18%of area) was nearly four times greater than at theother sites. Comparison of conditions at Two Riversand Port Washington, which experience similarwater levels and wave climates, suggests that sandthickness is the important variable. The sand at TwoRivers is much thinner than at Port Washington;thus, erosion of some sand from the nearshore atPort Washington causes less change in substratethan a similar event at Two Rivers.

Water depth and wave energy determine theamount of energy available to move sand in thenearshore (USACE 1973). Typically, sand in deepwater is less mobile than sand in shallow water. As-suming that substrates are utilized by biologicalcommunities, annual changes in both the extent andlocation of a sand body would affect nearshorehabitat distribution. A more detailed discussion ofsubstrate stability and biological utilization of habi-tat is given in Mackey and Liebenthal (2005).

Nearshore Biological Communitiesand Ecology

Wave energy, alongshore currents, and sedimentdynamics of Great Lakes nearshore zones exposefreshwater aquatic biota to unique environmentalconditions characteristic of few other freshwaterson Earth. These forces are much more akin to phys-ical properties of oceans than inland lakes, provid-ing habitat for taxa that are generally atypical oflake ecosystems (Janssen et al. 2004, Dettmers etal. in press). Notable examples of such taxa thatwere observed as part of this study include mayflynymphs of the Family Heptageniidae, longnosedace (Rhinichthyes cataractae), and mottledsculpins (Cottus bairdi), all of which more typicallyreside in headwater streams and small rivers(Janssen et al. 2004). Other taxa that are more com-monly associated with lakes generally do not occurwithin Great Lakes nearshore zones and were alsoabsent in this study. Examples of such taxa includesunfishes (Lepomis spp.), dragonfly and damselflynymphs (Odonata), and water boatmen (Corixidae).Hence, the biological communities of nearshorezones are unique and difficult to sample using tradi-tional lake sampling techniques. It is then notablethat our methods, while consistent with those ofother Great Lakes researchers, were likely inade-quate to address the full range of physical processesinfluencing biota in the nearshore zones sampled,especially with regard to wave energy and along-shore currents. However, given that the overall pro-

106 Meadows et al.

ject effort was concerned with cumulative impactsof shoreline anthropogenic alterations as they relateto sediment dynamics and biota, we deemed our ap-proach as suitable for exploring the general hypoth-esis that nearshore biological community patternscan be explained based solely on shoreline condi-tions.

Of the many fish (shallow water and nearshore),benthic, and zooplankton density measures (bothoverall for each group and individual taxa densitieswithin groups) used in statistical analyses to deter-mine whether these nearshore communities respond

to varied shoreline features, only zooplankton andshallow water fish measures exhibited significantdifferences between shoreline types (Table 7). Mostanalyses were hampered by high variability withinthe classes that primarily resulted from regionallake effects, low levels of replication, and consider-able heterogeneity of nearshore habitats encoun-tered during the surveys (see Goforth and Carman2005). These interactions made the results of thestatistical analyses difficult to interpret, particularlybecause the reverse pattern of response to shorelinetype occurred consistently in one of the three lake

TABLE 7. Catch per unit effort (shallow water and nearshore fish) and density (native benthic inverte-brates and zooplankton) measures based on samples collected at three sites classified by mid-bluff shore-lines and three sites classified as having unique shorelines. Further data provided in Goforth and Carman2005.

Shoreline Type

Community Component Group Mid-Bluff Unique

Piscivore 0.01 ± 0.01 0.11 ± 0.06Planktivore 0.36 ± 0.29 1.18 ± 0.56

Shallow Water Fish ≤1.0 Benthivore 0.24 ± 0.11 0.28 ± 0.17m depth (No. Indiv./Beach Seine m) Native 0.47 ± 0.18 0.85 ± 0.24

Non-Native Overall 0.14 ± 0.10 0.71 ± 0.35

CPUE 0.65 ± 0.19 2.17 ± 0.52

Piscivores 2.4 ± 0.9 4.1 ± 2.0Planktivores 1.5 ± 0.9 1.1 ± 0.5

Nearshore Fish Benthivores 1.7 ± 0.7 0.8 ± 0.33.0–6.0 m depth Native 3.6 ± 1.5 3.2 ± 1.4(No. Indiv./Gill Net ft/hr) Non-Native 2.0 ± 0.9 3.3 ± 1.5

Overall CPUE 5.6 ± 1.6 6.5 ± 2.9

Cladocerans 189.7 ± 50.9 490.9 ± 93.0Daphnia sp. 13.2 ± 2.9 318.1 ± 58.9Calanoids 29.2 ± 4.7 222.3 ± 36.7

Zooplankton Cyclopoids 76.9 ± 16.4 163.4 ± 26.8(No. Individuals/m3) Harpacticoids 0.12 ± 0.07 0.38 ± 0.27

Nauplii 32.0 ± 5.0 405.2 ± 66.2Native 5.6 ± 1.6 6.5 ± 2.9Non-native 23.2 ± 5.4 1.3 ± 0.5

Overall Density 364.2 ± 61.4 1,601.7 ± 255.5

Aquatic Insect Larvae 251.4 ± 52.0 210.6 ± 36.6Oligochaetes 79.3 ± 31.5 108.2 ± 28.6

Native Benthic Invertebrates Amphipods/Isopods 211.1 ± 67.9 102.7 ± 47.4(No. Indiv./m2) Gastropods 7.9 ± 5.4 5.9 ± 2.2

Sphaeriid Clams 0.7 ± 0.7 17.4 ± 8.5

Overall Density 588.1 ± 97.2 575.6 ± 71.5


areas compared to the other two (i.e., southern LakeErie compared to eastern Lake Michigan and west-ern Lake Michigan). Also, the magnitudes of the re-sponses were also often quite different among thelake areas, with measures differing by one or twoorders of magnitude for sites within the same shore-line class but from different lakes areas. The levelof replication for this study was necessarily low toaccommodate a study design that was appropriatefor disparate scientific disciplines. Therefore, in-creased replication within and among lake areas foreach of the shoreline types of interest, or for a sub-set of these shoreline types, would greatly enhancethe statistical power in future studies. Given thesestatistical constraints, the conclusions presentedherein must be considered as preliminary andshould be used as the basis for developing hypothe-ses to be tested in future research efforts seeking toexplore potential relationships between nearshorebiological communities and shoreline engineering.

The lower zooplankton densities, both overalland with respect to most component taxonomicgroups, at the mid-bluff vs. unique sites may be evi-dence of a community response to changing foodweb dynamics in the nearshore zone linked toshoreline engineering (Table 7). The mechanism forthis response appears to be related to the greatercolonization success of zebra mussels, Dreissenapolymorpha, at mid-bluff vs. unique sites. Althoughdreissenid densities were not explicitly measured aspart of the study, it was very obvious that they dom-inated substrates at two of the three mid-bluff sites,whereas they were rare at unique sites where suit-able substrates for colonization were sparse or ab-sent. As described previously, shoreline engineeringcan cause nearshore areas to become sand starved,exposing larger, harder substrates that can facilitatecolonization by dreissenids. This enhanced localcolonization success can result in the high dreis-senid mussel densities anecdotally observed at mid-bluff sites in this study. Dreissenids have beenreported to sequester available energy and to reallo-cate energy to benthic algae, both of which con-tribute to subsequent declines in zooplanktonpopulations (Dermott and Kerec 1997, Vanderploeget al. 2002, Dettmers et al. 2003). Thus, shoreline-mediated changes in sediment dynamics and sub-strate availability may facilitate and evenexacerbate food web alterations in nearshore areasof the Great Lakes by providing greater habitatavailability for dreissenids that can elicit a bottom-up response by native zooplankters. This effect mayalso extend to offshore (i.e., ≤ 70m water depth)

benthic invertebrates, such as Diporeia, via the ef-fects of dreissenid feeding on phytoplankton fromupwelled waters and the subsequent return (i.e.,downwelling) of these phytoplankton-stripped wa-ters to offshore areas (Janssen et al. 2004).

Although there were no statistically significantpatterns in overall benthic community and individ-ual morphospecies densities relative to shorelinestatus, there are several characteristics of thePainesville site, arguably the most highly modifiedshoreline site, that suggested community responsesto shoreline/nearshore change at this site comparedto all of the other sites. The Painesville site was ex-tensively colonized by both dreissenid mussels andround gobies (Neogobius melanostomus). While thecolonization of hard substrates by dreissenids hasbeen observed to increase local habitat complexity,creating additional habitat for local non-dreissenidbenthos (Dermott et al. 1993, Stewart and Haynes1994, Ricciardi et al. 1997, Botts et al. 1996, Stew-art et al. 1998, Haynes et al. 1999, Kuhns and Berg1999), non-dreissenid invertebrates were highly un-derrepresented (by an order of magnitude) at thePainesville site (Table 7) (see Goforth and Carmen2005). These low benthic densities may have beenthe result of the large numbers of round gobies alsooccurring at the site. Manipulative studies havedemonstrated that benthic invertebrates decline sig-nificantly in the presence of round gobies (Kuhnsand Berg 1999). This non-native benthivore hasbeen reported to rely heavily on native benthic taxaas food sources, primarily during the juvenilestages (Jude et al. 1995), despite its primary re-liance upon dreissenids as a food source. Roundgoby densities were estimated to be ≈ 16 individu-als/m2 based on SCUBA observations at depths of 3m, although comparatively few were captured inseine hauls due to the locally high substrate hetero-geneity that lowered the effectiveness of seines insampling shallow water fish. These densities wereextremely high compared to observations of roundgobies using both SCUBA observations and beachseines at other sites. Thus, the high round goby den-sities at the Painesville site likely contributedgreatly to the sparse non-dreissenid benthic com-munities observed.

The result of the species interactions and commu-nity changes at the Painesville site appears to haveshifted local productivity to favor benthic,lithophilic non-native species (i.e., round gobiesand dreissenid mussels), reflecting both a top-downresponse by native benthos and a bottom-up re-sponse by native zooplankters. The Painesville

108 Meadows et al.

shoreline has been heavily manipulated, and thenearshore areas were generally sand starved com-pared to historical times, with extensive exposedhard-pack clays and glacially deposited hard sub-strates (e.g., cobbles, boulders, and bedrock) domi-nating in the nearshore zone. These large, hardsubstrates are ideal habitat for both round gobiesand dreissenid mussels, so shoreline changes andassociated changes in nearshore substrates (com-pared to historic times) appear to have facilitatedthe dominance of non-native benthos at thePainesville site. Based on SCUBA observations in2000, the Two Rivers site appears to be following asimilar pattern (see Goforth and Carmen 2005). Al-though no dreissenid mussels were observed at theTwo Rivers site in 1999 (perhaps due to ice scourthe preceding winter), all hard substrates present atthe site were heavily colonized by small dreissenidsin 2000. The third mid-bluff site, St. Joseph, had noround gobies present during the 2000 surveys, al-though small individuals of this species were ob-served associated with groins and revetments at thesite in 2003. Open beach areas surveyed in thevicinity of the groin/revetment sample sites in 2003yielded no round gobies in beach seine hauls, pro-viding additional, although anecdotal, evidence tosuggest that the shoreline structures may facilitateinvasion of local areas by providing suitable habitatwithin a matrix of largely unsuitable habitat, in thiscase, sand. A future change in benthic communityproperties may be expected at the St. Joseph andTwo Rivers sites based on observations atPainesville. However, there will likely be a lag timebetween the dreissenid mussel colonization andround goby invasion during which benthic commu-nities will remain relatively intact, contributing tothe non-significant statistical tests conducted forthis study.

Overall shallow water (< 1.0 m depth) fish catchper unit effort (CPUE) was higher for unique shore-lines (Table 7). This appeared to be largely due tothe high productivity at the Sheldon Marsh site andthe comparatively species rich community at theLudington site. However, differences in CPUE be-tween unique and mid-bluff sites may have alsobeen due to greater seining success in the sandyshallow water substrates generally associated withthe unique sites (except Port Washington, wheresubstrates were more variable and estimates werecomparably lower than other unique sites). Thiswas, in fact, supported by the contradictory obser-vations of round goby densities between seining ef-forts (low densities) and SCUBA reconnaissance

(very high densities) conducted at the Painesvillesite as described previously. The high variability incatch rates among seine hauls further suggested thatshallow water fish were either patchily distributedor that variable substrate and/or wave conditions in-fluenced sampling efforts both within and amongstudy sites. These obvious limitations in samplingtechniques made it difficult to conclude thatnearshore habitat types associated with unique vs.mid-bluff shorelines were truly more or less pro-ductive with respect to shallow water fish. It wasnonetheless clear that sand-based nearshore areaswere characterized by sufficient shallow water fishCPUE and species richness to suggest that these areimportant habitats within the context of the GreatLakes Basin and not simply “wet deserts” as theyare often considered. Further, these sand-based sys-tems, while characterized by homogeneous habitatsat the site scale, appeared to be faunally distinctcompared to rocky nearshore areas that were moreheterogeneous with respect to substrate composi-tion, and therefore habitat, locally. While shorelinemediated habitat transformations from sandy torocky substrates in nearshore zones may increaselocal habitat heterogeneity and thus provide newand/or different foraging opportunities for predatorslocally (e.g. Wells 1977), there are likely to beother consequences resulting from these transfor-mations that are not fully understood. From thisperspective, the loss of sand-based nearshore sys-tems resulting from shoreline engineering is unde-sirable and may have consequences for losses ofbiodiversity and ecosystem services at lake and/orbasin scales.

CUMULATIVE IMPACTS

For the purposes of this discussion, cumulativeimpacts are induced by the combination of individ-ually minor effects (or impacts) of multiple naturalor anthropogenic shoreline modifications over time.As an example, shore-perpendicular navigationstructures associated with commercial and recre-ational harbors may produce far more reaching cu-mulative impacts than the sum of local impactsfrom each individual structure. The science of un-derstanding cumulative impacts is in its infancy;however, it is recognized that these cumulative im-pacts extend far beyond pure physical influence.They include modifications of the geological,chemical, and biological systems in operationwithin the nearshore region as well as changes tothe physical setting.


Numerous studies have shown that hard engi-neering structures, such as jetties, breakwalls,groins, revetments, and seawalls produce a measur-able impact on the shoreline that extends for manytimes their length (e.g., Berek and Dean 1982,Carter et al. 1986, Dean and Work 1993, Kraus1988, Stauble and Kraus 1993, Komar 1976,O’Brien and Johnson 1980, Shabica and Pranschke1994, Nairn and Parson 1995, Parson et al. 1996).Large navigational structures may extend more than400 meters into the lake from shore. The measur-able impact of these structures may extend up to 6to 10 times the overall length of the structure alongthe shoreline. The same relationship has beenshown to hold true for smaller, individual, privateshore protection structures. These structures alternatural coastal processes and interrupt the long-shore transport of littoral sediment. Littoral sedi-ments accumulate updrift of the structure therebyeffectively eliminating them from the active littoralsystem. The downdrift reduction in available sedi-ment supply results in a loss of protective sandcover, accelerates nearshore lakebed downcutting,and increases incident wave energy impinging onthe shoreline. Protective beaches become thinnerand narrower, and bluff-recession rates increase asprotective beaches become thinner and narrower(e.g., Shabica and Pranschke 1994, Nairn and Par-son 1995, Nairn and Willis 2002). These effects areinitially local, but long-term permanent reductionsin littoral sediment supplies will directly impact theentire downdrift shoreline reach.

For example, a series of man-made harbor struc-tures have been constructed along the Michigan andWisconsin shorelines of Lake Michigan, each pro-ducing its own localized set of impacts that may ex-tend many times its length laterally along theshoreline. Each of these structures captures a por-tion of the available littoral sediment supply, andmay divert those sediments into deeper offshorewaters. Depending on where these structures arelocated within what was once a natural littoral cell,each successive harbor structure may trap and re-move additional sediment from the littoral system.The net (or cumulative) effect of these anthro-pogenic modifications is to artificially subdividenatural littoral cells into discrete shoreline segments(or sub-cells), each of which becomes progressivelymore sediment-starved with increasing downdriftdistance. Under natural conditions, the downdriftportions of littoral cells are typically depocenters(i.e., areas where sediments are deposited and accu-mulate). As a result, it would appear that the poten-

tial cumulative impacts of these structures onnearshore and coastal habitats are much more sig-nificant in the downdrift portions of what were oncenatural littoral cells.

Data collected during this study show that a lossof sand cover will typically expose thin lag depositsof coarse sand, gravel, and cobble-size materialover an indurated cohesive clay or bedrock sub-strate. Others have observed this phenomenon aswell (e.g., Shabica and Pranschke 1994, Nairn andWillis 2002). While environmental responses to thisphenomenon have been relatively well understoodfor some time, we now know the nearshore ecologywill change in response to increasing habitat hetero-geneity created by the loss of sand cover and expo-sure of these rocky substrates. Of course naturallyoccurring rock-dominated substrates and associatedcommunities are important nearshore ecologicalfeatures in many areas of the Great Lakes basin(e.g., Janssen et al. 2004). Our work suggests thatshoreline alterations that result in nearshore sandstarvation facilitate habitat transformations thatmay alter the distribution and species compositionof multi-taxonomic communities, and alter trophicstructures characteristic of sand-based nearshoreecosystems. Furthermore, it appears that thesetransformations may also facilitate wider coloniza-tion of nearshore areas by lithophilic aquatic nui-sance species, such as dreissenid mussels and roundgobies, which more readily replace native benthictaxa as coarse-grained substrates become exposed.Widespread alteration of nearshore habitats mayhave significant implications for trophic dynamicsand productivity in the Great Lakes by shifting en-ergy flow from predominantly pelagic communitiesto benthic communities in nearshore areas, and po-tentially affecting upwelling/downwelling cycles inoffshore areas (MacIsaac 1996, Dermott and Kerec1997, Haynes et al. 1999, Janssen et al. 2004).This, in turn, may have considerable effects onGreat Lakes fisheries and other economically sig-nificant ecosystem services provided by the basin.

CONCLUSION

The results presented here are the result of amulti-disciplinary pilot effort to describe cumula-tive Great Lakes coastal impacts based on simulta-neous assessments of shoreline and nearshorephysical, geological, and biological attributes.Clearly, additional work is needed to more explic-itly describe the stressor-response relationships thatexist between shoreline development and Great

110 Meadows et al.

Lakes biological communities and ecologicalprocesses. What we have been able to demonstrateis that shoreline modifications may enhance habitattransformations and colonization success of aquaticnuisance species via altered nearshore substrate dy-namics that make suitable substrates more availablefor colonization. The implication of this is that ef-forts to control coastal erosion may, in fact, be fa-cilitating much larger scale changes in biologicalcommunity composition, trophic structure, ecosys-tem function, and fisheries production within theGreat Lakes basin.

ACKNOWLEDGMENTS

This research effort was funded by a grant fromthe Great Lakes Protection Fund to the Universityof Michigan through the Cooperative Institute forLimnology and Ecosystems Research (CILER). Wewould like to acknowledge the assistance of theOhio Geological Survey interns Bruce Gerke, Jen-nifer Vagen, and Pete Sokoloski and to express ourthanks to Tom Berg, State Geologist of Ohio andDivision Chief, and Constance Livchak, Lake ErieGeology Group Supervisor, for their support.

REFERENCESBerek, E.P., and Dean, R.G. 1982. Field Investigation of

Longshore Transport Distribution. In Proc. 18thCoastal Engineering Conference, pp. 1620–1638.American Society Civil Engineers.

Botts, P.S., Patterson, B.A., and Schloesser, D.W. 1996.Zebra mussel effects on benthic invertebrates: physi-cal or biotic? J. North Amer. Ben. Soc. 15:179–184.

Brazner, J.C., and Beals, E.W. 1997. Patterns in fishassemblages from coastal wetland and beach habitatsn Green Bay, Lake Michigan: a multivariate analysisof abiotic and biotic forcing factors. Can. J. Fish.Aquatic Sci. 54:1743–1761.

———, Tanner, D.K., and Morrice, J.A. 2001. Fish-medi-ated nutrient and energy exchange between a LakeSuperior coastal wetland and its adjacent bay. J.Great Lakes Res. 27:98–111.

Brown, E.A. 2000. Influence of water level, wave cli-mate, and weather on coastal recession along a GreatLakes shoreline. M.S. thesis, Department of Geologi-cal Sciences, University of Wisconsin.

———, Wu, C., Mickelson, D.M., and Edil, T.B. 2005.Factors Controlling Rates of Bluff Recession at TwoSites on Lake Michigan. J. Great Lakes Res.31:306–21.

Carter, C.H., Monroe, C.B., and Guy, D.E., Jr. 1986.Lake Erie shore erosion: the effect of beach width andshore protection structures: J. Coastal Res.2(1):17–23.

Chapman, J.A., Edil, T.B., and Mickelson, D.M. 1997.Lake Michigan shoreline recession and bluff stabilityin northeastern Wisconsin: 1996. Bay-Lake RegionalPlanning Commission Technical Report.

Christie, W.J., Collins, J.J., Eck, G.W., Goddard, C.I.,Hoenig, J.M., Holey, M., Jacobson, L.D., MacCallum,W., Nepszy, S.J., O’Gorman, R., and Selgeby, J.1987. Meeting future information needs for GreatLakes fisheries management. Can. J. Fish. Aquat. Sci.44 (Suppl. 2):439–447.

Dean, R.G., and Work, P.A. 1993. Interaction of naviga-tional entrances with adjacent shorelines. J. CoastalRes. 18:91–110.

Dermott, R., and Kerec, D. 1997. Changes to the deep-water benthos of eastern Lake Erie since the invasionof Dreissena: 1979 to 1993. Can. J. Fish. Aquat. Sci.54:922–930.

———, Mitchell, J., Murray, I., and Fear, E. 1993. Bio-mass and production of zebra mussels (Dreissenapolymorpha) in shallow waters of northeastern LakeErie. In Zebra Mussels: Biology, Impacts, andControl, eds. T.F. Nalepa and D.W. Schloesser, pp.399–413. Boca Raton, FL: Lewis Publishers.

Dettmers, J.M., Raffenberg, M.J., Matthew, J., and Weis,A.K. 2003. Exploring zooplankton changes in southernLake Michigan: implications for yellow perch recruit-ment. J. Great Lakes Res. 29:355–364.

———, J .M., Janssen, J., Pientka, B., Fulford, R., andJude, D.J. In Press. Evidence across multiple scalesfor offshore transport of yellow perch larvae. Can. J.Fish. Aquat. Sci.

Driver, D.B., Reinhard, R.D., and Hubertz, J.M. 1991.Hindcast Wave Information for the Great Lakes: LakeErie. WIS Report 22, U.S. Army Corps of Engineers,Washingtn, D.C.

Edsall, T.A. 1996. Great Lakes and Midwest region. InStatus and trends of regional resources, eds. M.J.Mac, G.S. Farris, P.A. Opler, P.D. Doran, and C.E.Puckett. National Biological Service, Washington,DC.

———, and Charlton, M.N. 1997. Nearshore waters ofthe Great Lakes. Environment Canada and U.S. Envi-ronmental Protection Agency State of the LakesEcosystem Conference. SOLEC ’96 BackgroundPaper. EPA 905-R-97-015a

Fuller, J.A., Liebenthal, D.L., Mackey, S.D., and Guy,D.E., Jr. 2002. Mapping nearshore substrates with sidescan sonar at six Great Lake sites. In A ComprehensiveStudy of Great Lakes Shorelines, Appendix II, GreatLakes Protection Fund Grant #470 Final Report.

Goforth, R.R., and Carman, S.M. 2005. Nearshore com-munity characteristics related to shoreline propertiesin the Great Lakes. J. Great Lakes Res. 31(Suppl. 1):113–128.

Goodyear, C.D., Edsall, T.A., Ormsby-Dempsey, D.M.,Moss, G.D., and Polanski, P.E. 1982. Atlas of spawn-ing and nursery areas of Great Lakes fishes. USFWS,


Report FWS/OBS-82/52, Volumes 1–14, Washington,DC.

Haynes, J.M., Stewart, T.W., and Cook, G.E. 1999. Ben-thic macroinvertebrates communities in southwesternLake Ontario following invasion of Dreissena: contin-uing change. J. Great Lakes Res. 25:828–838.

Hubertz, J.M., Driver, D.B., and Reinhard, R.D. 1991.Hindcast Wave Information for the Great Lakes: LakeMichigan. WIS Report 24, U.S. Army Corps of Engi-neers, Washington, D.C.

Janssen, J., Berg, M.B., and Lozano, S.J. 2004. Sub-merged terra incognita: Lake Michigan’s abundant butunknown rocky zones. In The State of Lake Michigan:Ecology, Health, and Management, eds. T. Edsall andM. Munawar. Ecovision World Management Series,Aquatic Ecosystem Health and Management Society.

Jude, D.J., Janssen, J., and Crawford, G. 1995. Ecology,distribution, and impact of newly introduced roundand tubnose gobies on the biota of the St. Clair andDetroit Rivers. In The Lake Huron Ecosystem: Ecol-ogy, Fisheries, and Management, eds. M. Munawar,T. Edsall, and J. Leach, pp. 447–460. SPB AcademicPublishing, Amsterdam, The Netherlands.

Kelso, J.R.M., Steedman, R.J., and Stoddart, S. 1996.Historical causes of change in Great Lakes fish stocksand the implications for ecosystem rehabilitation.Can. J. Fish. Aquat. Sci. 53 (Suppl. 1):10–19.

Kraus, N.C. 1988. The effects of seawalls on the beach:Extended literature review. J. Coastal Res. (SpecialIssue No. 4):1–28.

Komar, P.D. 1976. Beach Processes and Sedimentation.Englewood Cliffs, N.J.: Prentice-Hall.

Kuhns, L.A., and Berg, M.B. 1999. Benthic invertebratecommunity responses to round goby (Neogobiusmelanostomus) and zebra mussel (Dreissena polymor-pha) invasion in southern Lake Michigan. J. GreatLakes Res. 25:910–917.

Leslie, J.K., and Timmins, C.A. 1993. Distribution, den-sity, and growth of young-of-the-year fishes inMitchell Bay, Lake St. Clair. Can. J. Zool.71:1153–1160.

Mackey, S.D. 1995. Lake Erie Sediment Budget. In LakeErie Coastal Erosion Study Workshop—April 1995,ed. D.W. Folger, pp. 34–37. USGS Open-File Report95-224.

———, and Liebenthal, D.L. 2005. Mapping changes inGreat Lakes nearshore substrate distributions. J.Great Lakes Res. 31 (Suppl. 1):75–89.

MacIsaac, H.J. 1996. Potential abiotic and biotic impactsof zebra mussels on the inland waters of North Amer-ica. Amer. Zool. 36:287–299.

Meadows, G.A., Meadows, L.A., Wood, W.L., Hubertz,J.M., and Perlin, M. 1997. The relationship betweenGreat Lakes water levels, wave energies, and shore-line damage. Bull. Amer. Met. Soc. 78(4):675–683.

———, Bennett, T., Meadows, L.A., Caufield, B., andVanSumeren, H. 1999. Nearshore Profile Change and

its Impact on Rates of Shoreline Recession. In Proc.Coastal Sediments ’99. American Society of CivilEngineers, Long Island, NY, June 1999.

Mickelson, D.M., and Edil, T.B. 1998, Evolution of formand process on high bluffs subject to large-scaleslumping along Wisconsin’s Lake Michigan shore-line. Geol. Soc. America Abstracts with Programs30:251.

———, Acomb, L., Brouwer, N., Edil, T.B., Fricke, C.,Haas, B., Hadley, D., Hess, C., Klauk, R., Lasca, N.,and Scheider, A.F. 1977. Shoreline erosion and bluffstability along Lake Michigan and Lake Superiorshorelines of Wisconsin. Shore Erosion Study Techni-cal Report, Coastal Management Program, State Plan-ning Office.

———, Brown, E.A., Edil, T.B., Meadows, G.A.,Mackey, S.D., Liebenthal, D.L., and Fuller, J.A. 2002.Comparison of sediment budgets of bluff/beach/nearshore environments near Two Rivers, Wisconsin,on Lake Michigan, and at Painesville, Ohio, on LakeErie. Geol. Soc. America Abstracts with Programs34(2):A-12.

Nairn, R.B. 1992. Designing for Cohesive Shores, InProc. Coastal Engineering in Canada, ed. J. W. Kam-phuis. Department of Civil Engineering, Queen’s Uni-versity, Kingston, Canada.

———, and Parson, L.E. 1995. Coastal Evolution Down-drift of St. Joseph Harbor on Lake Michigan. In Proc.Coastal Dynamics ’95, pp. 903–914. American Soci-ety Civil Engineers.

———, and Willis, D. 2002. Erosion, Transport, andDeposition of Cohesive Sediments. In Coastal Engi-neering Manual, Part III, Coastal SedimentProcesses, ed. T. Walton, Chapter III-5, EngineerManual 1110-2-1100. U.S. Army Corps of Engineers,Washington, DC.

O’Brien, M.P., and Johnson, J.W. 1980. Structures andsandy beaches. In Proc. Coastal Zone ’80 , pp.2718–2740, Vol. IV. Hollywood Beach, Florida:American Society Civil Engineers.

Parson, L.E., Morang, A., and Nairn, R.B. 1996. Geo-logic Effects on Behavior of Beach Fill and ShorelineStability for Southeast Lake Michigan., U.S. ArmyEngineer Waterways Experiment Station, Vicksburg,MS. Technical Report CERC-96-10.

Regier, H.A., and Hartman, W.L. 1973. Lake Erie’s fishcommunity: 150 years of cultural stresses. Science180:248–1255.

Ricciardi, A., Whoriskey, F.G., and Rasmussen, J.B.1997. The role of the zebra mussel (Dreissena poly-morpha) in structuring macroinvertebrate communi-ties on hard substrata. Can. J. Fish. Aquat. Sci.54:2596–2608.

Shabica, C., and Pranschke, F. 1994. Survey of littoraldrift sand deposits along the Illinois and Indiana shoresof Lake Michigan, J. Great Lakes Res. 20:61–72.

Stauble, D.K., and Kraus, N.C. 1993. Beach Nourish-

112 Meadows et al.

ment Engineering and Management Considerations.Coastlines of the World Series, American Society ofCivil Engineers, New York, NY.

Steedman, R.J., and Regier, H.A. 1987. Ecosystem sci-ence for the Great Lakes: perspectives on degradativeand rehabilitative transformations. Can. J. Fish.Aquat. Sci. 44 (Suppl. 2):95–103.

Stewart, T.W., and Haynes, J.M. 1994. Benthic macroin-vertebrate communities of southwestern Lake Ontariofollowing invasion of Dreissena. J. Great Lakes Res.19:479–493.

———, Miner, J.G., and Lowe, R.L. 1998. Macroinverte-brate communities on hard substrates in western LakeErie: structuring effects of Dreissena. J. Great LakesRes. 24:868–879.

U.S. Army Corps of Engineers. 1973. Shore ProtectionManual, Volumes 1–3. U.S. Government PrintingOffice, Washington, D.C.

Vanderploeg, H.A., Nalepa, T.F., Jude, D.J., Mills, E.L.,Holeck, K.T., Leibig, J.R., Grigorovich, I.A., andOjaveer, H. 2002. Dispersal and emerging impacts ofPonto-Caspian species in the Laurentian Great Lakes.Can. J. Fish. Aquat. Sci. 59:1209–1228.

Wells, L. 1977. Changes in yellow perch (Percaflavescens) populations in Lake Michigan,1954–1975. J. Fish. Res. Board Can. 34:1821–1829.

Submitted: 9 May 2004Accepted: 11 September 2005Editorial handling: John Janssen

Cumulative Habitat Impacts of Nearshore Engineeringdavem/abstracts/05-12.pdf · Cumulative Habitat...

Documents

Transcript of Cumulative Habitat Impacts of Nearshore Engineeringdavem/abstracts/05-12.pdf · Cumulative Habitat...