storm-induced redistribution of deepwater sediments in...

J. Great Lakes Res. 32:348–360Internat. Assoc. Great Lakes Res., 2006

348

Storm-induced Redistribution ofDeepwater Sediments in Lake Ontario

John D. Halfman1,*, Dawn E. Dittman2, Randall W. Owens3, and Margaret D. Etherington1

1Department of Geoscience and Finger Lakes InstituteHobart and William Smith Colleges

Geneva, New York 14456

2Tunison Laboratory of Aquatic SciencesUnited States Geological Survey

Cortland, New York 13045

3USGS, Great Lakes Science CenterLake Ontario Biological Station

17 Lake StreetOswego, New York 13126

ABSTRACT. High-resolution seismic reflection profiles, side-scan sonar profiles, and surface sedimentanalyses for grain size (% sand, silt & clay), total organic carbon content, and carbonate content alongshore-perpendicular transects offshore of Olcott and Rochester in Lake Ontario were utilized to investi-gate cm-thick sands or absence of deepwater postglacial sediments in water depths of 130 to 165 m.These deepwater sands were observed as each transect approached and occupied the “sills,” identifiedby earlier researchers, between the three deepest basins of the lake. The results reveal thin (0 to 5-cm)postglacial sediments, lake floor lineations, and sand-rich, organic, and carbonate poor sediments at thedeepwater sites (> 130 m) along both transects at depths significantly below wave base, epilimnetic cur-rents, and internal wave activity. These sediments are anomalous compared to shallower sedimentsobserved in this study and deeper sediments reported by earlier research, and are interpreted to indicatewinnowing and resuspension of the postglacial muds. We hypothesize that the mid-lake confluence of thetwo-gyre surface current system set up by strong storm events extends down to the lake floor when thelake is isothermal, and resuspends and winnows lake floor sediment at these locations. Furthermore, webelieve that sedimentation is more likely to be influenced by bottom currents at these at these sites than inthe deeper basins because these sites are located on bathymetric highs between deeper depositionalbasins of the lake, and the bathymetric constriction may intensify any bottom current activity at thesesites.

INDEX WORDS: High-resolution seismic profiles, side scan sonar records, bottom sediment mobiliza-tion, Lake Ontario.

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

INTRODUCTION

A better understanding of the sedimentaryprocesses influencing the lake floor is critical forforecasting the fate of pollutants associated withthem, defining the habitat for benthic organismswithin the lake ecosystem and other concerns. Re-search in the 1970s revealed that postglacial sedi-ment thickness typically increased from nearshore

to offshore locations with over 12 m of sediment inthe deepest subbasins of the lake (Thomas etal.1972, Kemp and Harper 1976). Notable excep-tions to this trend are at the Whitby-Olcott andScotch-Bonnet Sills (Fig. 1), the bathymetric highs(130 and 150 m depths, respectively) between theNiagara, Mississanga, and Rochester basins (maxdepths of 140, 190, and 240 m, respectively), wherepostglacial sediment accumulation is minimal eventhough the water depths are significantly belowwave base (Thomas et al. 1972). These sills are ele-

Storm-induced Sediment Redistribution in Lake Ontario 349

FIG. 1. Bathymetric maps (20 m contour interval) of Lake Ontario with the location of the basins andsills discussed in the text, and the Olcott and Rochester survey areas revealing the high-resolution seismicreflection and side scan sonar shiptrack, and Ponar grab sites.

350 Halfman et al.

vated due to the topographic relief in the underlyingglacial tills and/or bedrock (Hutchinson et al.1993). The previous authors speculated withoutelaboratation that these thin sediment accumula-tions resulted from sediment focusing around thetwo sills due to the local bathymetric relief(Thomas et al. 1972, Hutchinson et al. 1993).

Growing evidence suggests that a variety of addi-tional processes including density currents, turbid-ity currents, contour currents and othersedimentological events can influence sedimentwinnowing and redistribution below wave base(e.g., Flood and Johnson 1984, Johnson et al. 1984,Johnson and Ng’ang’a 1990, Vickman et al. 1992,Hawley et al. 1996, Scholz et al. 1993, Soreghan etal. 1999). This paper examines the sedimentologi-cal evidence for the controls on deepwater sedimen-tation offshore of Olcott and Rochester in LakeOntario using high resolution seismic reflectionprofiles, side scan sonar profiles, and sedimentanalyses.

Lake Ontario is the smallest (19,477 km2) andhydrologically farthest downstream of the Laurent-ian Great Lakes. It occupies a trough cut by fluvialand glacial processes into the northern margin ofthe Appalachian basin and much of the present daybathymetry reflects the topography of the underly-ing bedrock and glacial drift (Martini and Bowlby1991). The offshore sediment sequence, from oldestto youngest, is glacial till, glaciolacustrine claysand postglacial muds (Lewis and Sly 1971, Thomaset al. 1972, Kemp et al. 1974, Kemp and Harper1976, Hutchinson et al. 1993). The postglacial sedi-ment is dominated by silicate minerals with up to5% carbonate and 5% organic matter (Thomas et al.1972). The primary sediment sources include flu-vial inputs and shoreline erosion of exposed glacialmaterials for the allochthonous fraction and in-lakeprecipitation for the autochthonous calcite and or-ganic matter. Sediment resuspension, shoreline ero-sion and sediment transport from west to eastdominates the nearshore region along the southernmargin of the lake (e.g., Lewis and Sly 1971, Sut-ton et al. 1970, Rukavina 1976).

METHODS

High resolution seismic reflection profiles, sidescan sonar profiles, and Ponar grab sediment sam-ples were collected in 2001 and 2002 during multi-ple daylong cruises on the USGS research vessel,R/V Kaho (Fig. 1). Navigation was by DifferentialGlobal Positioning System (DGPS, 10 meter accu-

racy). The 2001 fieldwork centered on two shoreperpendicular transects from 30 to 165 m of wateroffshore of Olcott and Rochester. The transects andPonar sites were predetermined by an ongoing in-vestigation of the benthic Diporeia (arthropod) andDriessena (zebra and quagga mussels) communityin Lake Ontario (Owens and Dittman 2003,Dittman et al. 2005). The 2002 field season ex-tended the 2001 survey lakeward in a north tonorthwesterly direction to further investigate thethin, cm-thick, postglacial sands or missing post-glacial sediments recovered at the deepest-watersites in 2001 (Etherington et al. 2002, Walker et al.2002).

Approximately 150 km of high-resolution seis-mic reflection data were collected from the lake, di-vided between each field area, at a ship’s speed of ~5 km/hr (Fig. 1). Profiles were collected withEG&G’s (EdgeTech) X-Star subbottom profilingsystem, that utilizes chirp technologies, sweep fre-quencies of 2–12 kHz, and a SB-216B tow vehicle.Seismic and navigation data were recorded on mag-netic tape and plotted in real time on an EPCgraphic recorder (GSP-1086) with a time-varyinggain that increased linearly (0.4 dB/m) below thelake floor. In general, subbottom penetration withthis system varies from nil over bedrock, muds withgas, or sand-rich and coarser sediments to 50 m ormore in mud-rich, gas-free sediments with decime-ter resolution. Seismic reflectors correspond toacoustic impedance boundaries (changes in sedi-ment bulk density and sound velocity), and aremore reflective (darker in the profiles) with largercontrasts. Higher amplitude (darker) reflectors typi-cally correspond to coarser-grained sediments at thelake floor or larger changes in grain-size or watercontent at the internal reflector boundary, whereaslower amplitude reflectors correspond to finer-grained or smaller water content contrasts. TheEG&G software assumed the conventional soundvelocity of 1,500 m/s in the water column andupper sediments for the distance to two-way traveltime conversions even though the actual velocity ofsound is ~1,430 m/s or 3 to 5% slower.

Approximately 10 km of side scan sonar records,recording a 200 m swath on either side of the fish,were collected and printed in real time from eachfield area. The shiptrack was parallel to and within1 km of the 2001 sediment Ponar sites (Fig. 1). Theside scan sonar profiles were collected fromnearshore (30–40 m depth) to maximum waterdepth of ~130 m due to the length of the tow cable.The data were collected using EG&G’s AS-600 side


scan sonar system with an analog towfish set at afrequency of 100 kHz and a ship’s speed of ~ 2km/h. Higher-amplitude (darker) reflectors corre-spond to coarser-grained and/or rippled sediments,larger sediment bulk densities, and/or lake floortopographic surfaces that tilt toward the tow vehi-cle, whereas relatively lighter reflectors corre-sponding to finer-grained and/or smooth sediments,smaller sediment bulk densities, and/or topographicsurfaces that tilt away from the tow vehicle.

Lake floor sediments were collected by Ponargrabs from a total of 25 sites, recovering at leastthree grabs at each site (Table 1, Fig. 1). The Ponarrecovers a 30 × 20 × 15 cm thick block of mud atideal sites. Stiff clays and sand-rich or coarser sedi-ments hamper sediment recovery. The 2001 siteswere visited twice, once in May and again in Au-gust, collecting a total of six separate sedimentssamples from each site. The only exception was the30 m site, which was only visited in August andthus only had subsamples from three separate sedi-ment grabs. The sediments recovered by each grabwere photographed, described, and the upper 2 to 5cm of mud was subsampled onboard. Sediment sub-samples were stored in Ziplock plastic bags andkept refrigerated at 4°C until analysis in the labora-tory for grain size, total organic carbon and carbon-ate analyses. During the 2001 cruise, the upper 5 to10 cm of sediment was subsampled and homoge-nized in the baggie. During the 2002 cruise, if alake floor sand layer was present and at least 1 cmthick, then the sands (top) and the underlyingglaciolacutrine muds (bottom) were subsampledand analyzed for the same parameters separately.Once the recovered sediments were described andsubsampled, the remaining mud was carefullysieved (500 µm) onboard for enumeration of Dipor-eia, Dreissena, and other macroinvertebrates (seeDittman et al. 2005 for details).

Grain size, total organic carbon, and carbonatepercentages were determined in the laboratory. Per-cent sand (> 63 µm), silt (63 – 4 µm), and clay (< 4µm) relative to the dry sediment weight were deter-mined by wet sieving and pipette analysis (Folk1974). Total organic carbon and carbonate percent-ages were determined by weight loss on ignition at550° to burn the organics and then at 1000°C to ox-idize the carbonates relative to the dry sedimentweight (Dean 1974, Heiri et al. 2001). Each sedi-ment subsample was analyzed twice to quantify theprecision of the technique. The average and stan-dard deviation of all the analyses from each site arereported (Table 1).

RESULTS

High-resolution Seismic Reflection Profiles

The high-resolution seismic reflection profilesdifferentiate two acoustic sequences in the lake(Fig. 2). The uppermost sequence is characterizedby a low-amplitude surface reflector and a numberof low-amplitude, parallel to subparallel, internalreflectors. This sequence is 0 to 10 m thick, transi-tioning from highly-reflective nearshore sands atwater depth of ~20 m with no subbottom penetra-tion to poorly-reflective deepwater muds and deep-est penetration at water depths shallower than ~100m. The acoustic character of and sediment samplesfrom this sequence are consistent with postglacialmuds recovered in this and many other lakes withinglaciated terrains (e.g., Hutchinson et al. 1993,Halfman and Herrick 1998, Mullins and Halfman2001). The underlying sequence typically definesacoustic basement. Where this lower unit is imaged,decimeter-scale, high-amplitude internal reflectorsare either parallel to subparallel or chaotic. If bothparallel and chaotic reflectors are observed, the par-allel reflectors overlie the chaotic reflectors. Theseacoustic characteristics are typical of glaciolacus-trine muds overlying glacial outwash or till in lakesfrom glaciated terrains with the chaotic packagefilling underlying kettle holes or other depressions(e.g., Halfman and Herrick 1998, Mullins and Half-man 2001). Glaciolacustrine muds were recoveredby Ponar grabs where this lower sequence was ex-posed at or near the lake floor.

The thickness of the postglacial sequence spansfrom 0 to just over 10 m offshore of Olcott andfrom 0 to just over 5 m offshore of Rochester (Fig.3). A single, high-amplitude, lake-floor reflectorwas traced from nearshore to water depths of about25 m. The postglacial sequence then thickens lake-ward to their maximum thickness at water depthsbetween 80 and 120 m, thinning over bathymetrichighs and thinning over bathymetric lows in the un-derlying glacial sediments. The postglacial sedi-ments are absent at the Rochester 90 m scarp;presumably the lake floor is too steep for modernsediment accumulation. Farther offshore, the post-glacial sediments pinch out to a single, high-ampli-tude reflector. The single, high-amplitude reflectoris detected between water depths of ~110 m and150 m offshore of Olcott and between ~110 and 160m offshore of Rochester, except for a few isolated,decimeter-long pods that fill depressions on the un-derlying floor. The low-amplitude character of the

352 Halfman et al.

TA

BL

E1.

Sit

eav

erag

esan

dst

anda

rdde

viat

ion

ofdu

plic

ate

sedi

men

tan

alys

esof

mu

ltip

leP

onar

grab

sat

each

site

.S

ites

are

arra

nge

dfr

omn

ears

hor

eto

offs

hor

elo

cati

ons.

Bu

lk/T

op/B

otto

m:

Bu

lkse

dim

ent

sam

ples

wer

ean

alyz

edfr

omea

chP

onar

grab

un

less

sepa

rate

subs

ampl

esco

uld

isol

ate

the

lake

floo

rsa

nds

(top

)fr

omth

eu

nde

rlyi

ng

glac

iola

cust

rin

em

uds

(bot

tom

).

Loc

atio

nL

ocat

ion

(deg

rees

)S

and

(wt.

%)

Sil

t(w

t.%

)T

otal

Org

anic

Car

bon

(wt.

%)

Car

bona

te(w

t.%

)

Dep

th(m

)L

atit

ude

Lon

gitu

deB

ulk

Top

Bot

tom

Bul

kT

opB

otto

mB

ulk

Top

Bot

tom

Bul

kT

opB

otto

m

Olc

ott

01–3

0m

43.3

5833

78.7

5703

5±

1—

—64

±11

——

2.6

±0.

7—

—11

.6±

0.7

——

01–5

5m

43.3

6983

78.7

5617

18±

3—

—64

±8

——

3.9

±1.

9—

—12

.6±

1.1

——

01–7

5m

43.3

8282

78.7

5002

11±

2—

—71

±9

——

3.2

±0.

4—

—9.

2±

0.5

——

01–9

5m

43.3

9703

78.7

4610

13±

1—

—69

±3

——

3.1

±0.

2—

—6.

6±

0.6

——

01–1

10m

43.4

0915

78.7

4168

14±

1—

—67

±8

——

3.0

±0.

1—

—6.

7±

1.3

——

01–1

30m

*43

.439

0378

.762

5028

±9

74±

110

±2

62±

208

±3

50±

221.

5±

0.6

0.7

±0.

01.

3±

0.1

2.7

±0.

71.

9±

0.0

4.4

±0.

102

–130

m*

43.4

3913

78.7

6225

—79

±1

6±

1—

9±

356

±17

—0.

6±

0.0

1.3

±0.

1—

2.1

±0.

66.

0±

0.7

01–1

50m

*43

.469

6078

.722

2821

±8

76±

55

±1

70±

2112

±8

34±

211.

2±

0.4

0.6

±0.

00.

9±

0.0

4.0

±2.

51.

6±

0.1

8.5

±0.

002

–143

m*

43.4

6998

78.8

2795

—39

±12

4±

1—

59±

1987

±3

—0.

8±

0.2

1.3

±0.

1—

2.8

±2.

35.

4±

4.5

02–1

44m

*43

.475

9578

.799

60—

63±

148

±2

—24

±29

74±

14—

0.6

±0.

11.

1±

0.1

—2.

8±

2.6

5.1

±0.

302

–148

m*

43.4

8175

78.7

7753

—69

±2

33±

6—

13±

749

±21

—0.

7±

0.1

1.0

±0.

2—

2.3

±0.

14.

6±

1.5

02–1

54m

43.5

2957

78.7

8947

—6

±3

4±

1—

45±

1757

±18

—3.

4±

1.2

2.8

±0.

7—

6.9

±1.

95.

9±

0.8

Roc

hes

ter

01–3

0m

43.3

0220

77.5

6413

37±

3—

—43

±13

——

0.9

±0.

1—

—6.

4±

0.3

——

01–5

5m

43.3

3252

77.5

4940

17±

4—

—60

±20

——

2.2

±0.

9—

—7.

0±

0.5

——

01–8

0m

43.3

5993

77.5

3228

55±

10—

—24

±10

——

1.5

±0.

7—

—3.

5±

2.4

——

01–9

5m

43.3

6543

77.5

2938

30±

6—

—56

±7

——

1.1

±0.

2—

—2.

2±

0.6

——

01–9

5D43

.385

5277

.644

0755

±1

——

35±

1—

—3.

3±

0.4

——

7.4

±1.

1—

—01

–110

m43

.368

0377

.528

2827

±6

——

52±

11—

—1.

8±

0.4

——

3.4

±1.

1—

—01

–130

m43

.376

0277

.522

1311

±3

——

69±

4—

—2.

9±

0.3

——

5.2

±1.

4—

—01

–150

m*

43.3

8595

77.5

1603

13±

310

±1

8±

171

±12

63±

463

±6

3.2

±0.

84.

0±

0.9

2.9

±0.

35.

3±

0.8

7.0

±1.

05.

8±

0.3

02–1

34m

*43

.395

6777

.567

2313

±3

——

41±

5—

—2.

1±

0.4

——

4.3

±0.

7—

—02

–158

m*

43.4

6672

77.6

4832

19±

880

±4

26±

1520

±25

7±

151

±34

1.4

±0.

10.

7±

0.1

1.2

±0.

34.

0±

0.2

1.7

±0.

43.

7±

0.2

02–1

64m

*43

.500

1077

.686

7316

±6

——

28±

4—

—2.

1±

1.1

——

4.6

±0.

3—

—02

–160

m*

43.5

3417

77.7

2412

7±

2—

—52

±29

——

1.5

±0.

2—

—4.

1±

0.1

——

02–1

50m

*43

.575

3577

.774

6032

±12

——

38±

11—

—3.

9±

0.9

——

12.2

±5.

7—

—

nosa

mpl

e*w

inno

wed

site

s


isolated pods is similar to the postglacial mudssampled elsewhere.

Side Scan Sonar Profiles

Side scan sonar profiles reveal a nearshore to off-shore transition of higher to lower amplitude back-ground scatter, with occasional acoustic backscatteranomalies and lineations in water depths greater

than 90 m (Fig. 4). The high to low amplitude tran-sition from shallow water to 90 m parallels the ob-served decrease in surface sediment acousticreflectivity and the increase in postglacial sedimentthickness observed in the high-resolution seismicreflection data.

Acoustic backscatter anomalies (ABAs) were su-perimposed on the background scatter. These ABAs

FIG. 2. Representative high-resolution seismic profiles from off-shore of Olcott, NY, along a north-northeasterly shiptrack, nearlyperpendicular to the bathymetric contours. The first high-ampli-tude acoustic reflector below the lake floor is interpreted as thebase of the postglacial sequence. Top: A profile from 70 to 80 m ofwater revealing the thick accumulation of postglacial muds. Bot-tom: A profile from ~100 to 110 m of water revealing the thinningof postglacial sediments to deeper water. Depths assumed a conven-tional speed of sound of 1,500 m/s.

354 Halfman et al.

were up to 5 to 10 m across, irregular to oval inshape, and did not reveal any topographic relief atthe displayed scale and frequency of sound. Up to15 or 20 ABAs are observed over an area of 250m2, typically in random distributions. Theseacoustic backscatter anomalies (ABAs) may bepatches of ship-derived debris (coal, taconite and/orcoal clinkers) or patches of Dreissena spp. The for-mer were observed in side scan data, ROV imagesand box cores from the western portion of Lake On-tario (Ferrini and Flood 2001) under shipping lanes.However, our study areas are removed from majorshipping lanes and clinkers were not abundant in

the sediment grabs collected in this study. Ourgrabs recovered 0 to a few thousand individuals/m2

Dreissena spp. in water depths greater than 75 m(Dittman et al. 2005). The observations indicatethat the ABAs in these surveys are most likelypatchy groups of Dreissena spp. ABAs in easternLake Ontario also have been attributed to patch dis-tributions of Dreissena spp. (Charles McClennen,personal communication).

Lineations were observed in the records at waterdepths of 90 m to the deepest portion of the side-scansonar survey (~130 m). These lineations are parallelor slightly subparallel to the bathymetric contours

FIG 3. Postglacial sediment thickness maps in the survey area. Both areas reveal thin, if any, post-glacial sediments in the deepwater portions of the surveys. Left/Right: Sediment thickness data offshoreof Olcott and Rochester, respectively (2 m contour interval).


and lack discernable bathymetric relief at the dis-played scale and frequency of sound. The lineationsare more pronounced towards deeper water alongboth profiles, and are more pronounced offshore ofOlcott than Rochester at the same water depths.

Sediment Analyses

The sediments recovered by the Ponar grabsgrade from stiff sandy silts at 30 m to olive-gray,flocculent (water-rich) muds at 80 to 120 m. Thistransition parallels the absent to 5–10 m thick post-glacial sediment transition, the high to low ampli-tude surface reflection, and published trends (e.g.,Thomas et al. 1972). In contrast, the sediments re-covered at 130 m and farther offshore of Olcott(Sites 01-130 m, 01-150 m, 02-143 m, 02-144 m,02-148 m, and 02-154 m) revealed a 1 to 5 cmthick, lake floor sand layer overlying stiff glaciola-custrine silty clays. The sands were massive, lackgraded bedding, and the basal contact was sharp.The sands at Site 02-154 m, the site farthest off-shore of Olcott, contained more silt and was a fewcm thicker than the other deepwater sands. TheRochester sediment samples lakeward of Site 01-150 m also recovered minimal, if any postglacialsediment. The grabs typically recovered up to 100sand grains, dispersed on top of stiff and eroded(dimpled surface) glaciolacustrine silty clays. Sites01-150 m and 02-158 m recovered a lake floor sandlayer, ~1 cm thick, overlying the glaciolacustrinemuds.

Site averaged grain size of the surface sedimentsranged from silty sands to clayey silts with sandpercentages ranging from 3 to 80%, silt from 8 to86%, and clay from 9 to 61% (Table 1, Fig. 5). Thestandard deviation of the sand percents at each siteranged from < 1 to 15%. The average standard de-viation is 3% after excluding the standard devia-tions from the deepwater sites as these subsamplesmay have mixed different percentages of sand andunderlying glaciolacustrine clays. Among the Ol-cott sites, sediments with the largest sand concen-trations were detected in the sand layers recoveredfrom the 130 m and deeper sites, and the smallestsand concentrations were detected in the underlyingglaciolacustrine muds. Among the Rochester sites,the largest sand concentrations were detected in thesediments shallower than 110 m, at Site 02-160 m,and from the surface sand lens recovered at Site 02-158 m. The percent clay and silt data lacked spatialtrends and lacked significant correlations betweeneach other, except significantly less clay and siltwas detected in the deepwater sand lenses.

Site averaged total organic carbon concentrationsranged from 0.6 to 4.0% (Table 1, Fig. 5). The stan-dard deviation of the TOC concentrations at eachsite ranged from 0.1 to 1.3%. The largest deviationsat any one site were detected in the nearshore sites

FIG. 4. Representative side scan sonar profilesfrom offshore of Olcott, NY. Top/Bottom: Anearshore profile in a water depth of 80 m, and anoffshore profile at 120 m, respectively. The deep-water profile reveals well-developed lineationsalong the lake floor that are indicative of bottomcurrent activity.

356 Halfman et al.

FIG. 5. Average grain size (sand wt. % and silt wt. %), total organic carbon content (dry wt. %), and car-bonate content (dry wt. %) at each Ponar grab site. Left/Right: Graphs of the sediment data from offshoreof Rochester and Olcott, respectively. Bulk/Top/Bottom: Bulk sediment subsampled from each Ponar grabor separate subsamples isolating the lake floor sand (top) and underlying glaciolacustrine mud (bottom).Averages and standard deviations of the duplicate analyses of multiple sediment grabs at each site areshown. The sites, from left to right, correspond to nearshore to offshore locations at each survey area. Thewinnowed designation and, e.g., 01-150 m* asterisk, identifies those sites influenced by bottom currents.


and Site 02-154 m, and may reflect TOC patchinessat the site. Among the Olcott sites, the largest TOCconcentrations were detected in the sediments shal-lower than 100 m, and the farthest offshore site(Site 02-154 m). Among the Rochester sites, TOCincreased slightly from 30 to 150 m and then re-turned to lower values in the sites farther offshore.The exception was at Site 02-150 m, the site far-thest from the survey area with TOC concentrationssimilar to the largest values along the 30 to 150 mtransect. Each deepwater sand had less TOC thanthe underlying glaciolacustrine clays by approxi-mately 0.5%, except at Site 02-154 m and 01-150 moffshore of Olcott and Rochester, respectively.

Site averaged carbonate concentrations rangedfrom 1 to 13% (Table 1, Fig. 5). The standard devi-ation of the carbonate concentrations at each siteranged from < 0.1 to 4.5%, and lacked spatialtrends. Offshore of Olcott, the carbonate concentra-tions decreased with water depth until the deepestwater sites. Offshore of Rochester, carbonate con-centrations did not change in a systematic manner.Each deepwater sand lens, when present, had ap-proximately 3% less carbonate than the underlyingglaciolacustrine clays, except at the site farthest off-shore of Olcott (Site 02-154 m) where carbonatewas 1% larger in the sand lens, and Site 01-150 moffshore of Rochester. Carbonate content weaklyco-varied with TOC content (r2 = 0.53).

DISCUSSION

The classic lacustrine depositional model dictatesthat nearshore sediments are winnowed of fine-grained, TOC-rich materials by surface waves andepilimnetic currents leaving behind coarse, TOC-poor sediments. The winnowed material is selec-tively deposited below wave base as thick,undisturbed, accumulations of TOC-rich muds(Thomas et al. 1972, Kemp and Harper 1976, Sly1980, Rea et al. 1981, Johnson 1984). In Lake On-tario, the nearshore sediments offshore of Olcott arealso influenced by easterly transport of NiagaraRiver sediments due to the strong, storm-induced,eastward flowing currents that occasionally extenddown to water depths of 100 m when the lake isisothermal (Scrudato and DelPrete 1982, Hawley etal. 1996). The sediments from 30 to 100 m at Olcottand 30 to 130 m at Rochester follow this generalpattern, especially if the previously published datafrom the deepest basins in the lake (thickness > 12 m, silty clays, 4 to 5% TOC) are included.However, the deepwater sediments from 130 to 155

m offshore of Olcott and 140 to 160 m offshore ofRochester consisting of thin, if any, postglacial sed-iments, lake floor sediment lineations, sand lagsover glaciolacustrine clays, and small TOC concen-trations are clearly anomalous to the overall trendand the focus of this discussion.

Previous workers speculated without elaborationthat these offshore sites may be sediment starved(e.g., Thomas et al. 1972, Kemp et al. 1974, Kempand Harper 1976, Hutchinson et al. 1993). In sup-port, they noted the minimal accumulations of post-glacial muds at these sites and indicated that thesesites are located along “sills” between the Niagara,Mississauga, and Rochester basins. They proposedthat the bathymetric relief was sufficient to preventthe accumulation of postglacial sediments. How-ever, their bathymetric data revealed more reliefthan more recent data (Virden et al. 1999).

We propose that the thin (1 to 5 cm-thick), mas-sive, sand-rich postglacial sediment implies sedi-ment deposition and subsequent reworking at thesedeepwater sites. Fine-grained organic rich sedi-ments are accumulating at shallower and deeperwater depths. The accumulation of sands and podsof postglacial muds in the area suggests postglacialsediments are deposited in the area as well. It indi-cates that the fine-grained, organic rich materialsmust be removed from these sites, except from iso-lated depressions on the lake floor, to leave behindthe sands. The areas with no postglacial sedimentsoffshore of Rochester revealed dimpled glaciolacus-trine muds and a dusting of sand grains at the sur-face, indicative of active and perhaps more intenseerosion offshore of Rochester than Olcott. Finally,the lineations in the side scan sonar profiles areconsistent with sediments that are reworked andwinnowed by bottom currents. Bottom currentsleave ribbons of coarser sediments, with the lin-eations parallel to the current flow (e.g., Flood1981, Flood and Johnson 1984). More ribbons weredetected offshore of Olcott, because Olcott hadmore sand on the lake floor to be modified by bot-tom current activity. The lineations increased in in-tensity toward the deepest portion of both transects,and imply more intense reworking of the lake floortoward progressively deeper water. This interpreta-tion is consistent with the high-resolution seismicdata and recovered sandy or missing postglacialsediments, and thus extends the anomalous charac-ter of the sediments lakeward through the remain-der of both transects.

Resuspension by surface and/or internal waves orassociated epilimnetic currents brought about by

358 Halfman et al.

large storms is an unlikely means to impact sedi-mentation at these deepwater sites. The nearshoretransition between the single, hard reflector to dis-cernable postglacial sediments occurs at waterdepths of ~25 m, and suggests that surface wavesand current activity decreases its influence on thelake floor at these water depths. This water depth isconsistent with models of wave erosion based onthe effective fetch, water depth, and typical storm-induced winds in Lake Ontario (e.g., Johnson1980). The accumulation of thicker and more floc-culent sediments at 80 to 100 m suggests that sur-face and/or internal waves and other surfaceprocesses do not influence the lake floor at theseand deeper sites. Neither is ice-rafted debris norfish burrows consistent with these continuous, thin,massive, sandy sediments, and the lake floor lin-eations. No ice-rafted dropstones were observed,nor were burrow structures detected in the side scansonar and Ponar grab samples.

Evidence for post-depositional reworking of sedi-ments at water depths significantly below the sum-mer thermocline is expanding in the literature. Forexample, wavy bedforms, sediment thinning, andother features suggest that bottom currents are ac-tive in Lake Malawi despite the permanent water-column stratification in these lakes (Johnson andNg’ang’a 1990). The majority of the deepwatersands, gravels, and erosional channels on the lakefloor, however are the result of turbidity current ac-tivity (Scholz et al. 1993, Soreghan et al. 1999).Turbidites are not consistent with the Lake Ontariodeepwater sands, as the sands do not reveal gradedbedding. They lack a significant sediment source,the lake floor lacks distributary channels closer toshore, and distal turbidites are not reported from thesediments recovered from the deepest basins of thelake.

Bottom contour-current modified structures werefound along the deep lake floor (> 200 m of water)just north of the Keweenaw Peninsula in Lake Su-perior (Flood and Johnson 1984, Johnson et al.1984, Viekman et al. 1992). The authors proposedthat the strong surface currents offshore of thepeninsula generated by major storms may extenddown to the lake floor and winnow and resuspendthe sediments when the lake is isothermal. We sug-gest a similar but slightly modified scenario to in-fluence the deepwater sediments at our two surveysites. Specifically, we hypothesize that surface cur-rents generated by major storms in Lake Ontarioextend down to the lake floor at these deepwatersites, reworking and winnowing the lake floor sedi-

ments when the lake is isothermal. However, an off-shore extension of the coastal current system is anunlikely mechanism for the sediment reworking at130 and 150 m in Lake Ontario because sedimentsare meters thick and thus less influenced by bottomcurrent activity in shallower water between thesesites and the coastal current system.

Theoretically, wind stress sets into motion a two-gyre surface flow in Lake Ontario (Csanady 1978).The gyre locations are dictated by the geometry ofthe basin and bathymetry of the lake floor. In LakeOntario, both gyres flow eastward along the north-ern and southern shorelines of the lake and the re-turn limbs combine, intensify, and flow westwarddown the middle and deepest reaches of the lake.Current meter records, along with geophysical andsedimentological evidence from the nearshore re-gions, confirm the nearshore flow of these gyres(e.g., Lewis and Sly 1971, Sutton et al. 1970,Rukavina 1976, Simons et al. 1985). However, cur-rent meter measurements are lacking offshore alongthe lake floor. Models of current flow indicate thatthe location for the westward return flow followsthe deepest portions of the lake and travels over thedeepwater sites in our two study areas (Csanady1978), and are supported by mean circulation pat-terns observed in surface-water current meterrecords (Beletsky et al. 1999).

Benthic nepheloid layers and resuspension ofbottom sediments have been detected in Lake On-tario and the other Great Lakes (e.g., Mudroch andMudroch 1992, Sandilands and Mudroch 1983,Eadie et al. 1984, Rosa 1985, Hawley et al. 1996).In Lake Ontario, the height and suspended sedimentconcentration of the benthic nepheloid layer in-creased from the summer to the early wintermonths. The trend suggests that the seasonal decayof the thermocline and associated water columnstratification coupled with lower biological produc-tivity and increased frequency and intensity ofstorm activity as time progresses into the wintermonths may promote lake floor resuspension of thesediments. In fact, storm-induced bottom sedimentresuspension events were documented for westernLake Ontario due to easterly nearshore currents inwater depths down to 100 m when the lake wasisothermal (Hawley et al. 1996). Therefore, the sed-imentologic evidence suggests that westward flow-ing surface currents extend from the surface to thelake floor during larger storms when the lake isisothermal, and the events are sufficient to occa-sionally resuspend and rework sediments in thestudy area. Furthermore, the sedimentological evi-


dence suggests that these deepwater currents havethe largest influence on sediment accumulation andredistribution at the “sill” locations in this survey.Perhaps these currents intensify while constrictedover these sills between the deepest basins of LakeOntario.

CONCLUSIONS

High-resolution seismic profiles, side scan sonarprofiles, and sand-rich sediment lags or no post-glacial sediment accumulation indicate that bottomcurrents resuspend and rework the sediments at thedeepwater sills that separate the deepest bathymet-ric basins of Lake Ontario. We hypothesize that thesuspected bottom currents result from the westernconfluence of the two-gyre surface current systemin the lake that extends down to the lake floor andinfluences sedimentation processes at these deep-water sites when the lake is isothermal duringstorm-driven events.

ACKNOWLEDGMENTS

We would like to thank the captain and crew ofthe R/V Kaho, and USGS technicians, who were in-valuable in collecting the field data and invertebratedata for this study. We also thank Hobart andWilliam Smith Colleges undergraduates SandraBaldwin, Laura Calabrese, Emily LaDuca, LindseyBowser, Caterina Caiazza, Robert Stewart, and AnnWalker, who assisted in the field and/or laboratorywork. Funding was provided by the EnvironmentalProtection Agency, the USGS, The Tripp Founda-tion, The Mellon Foundation, and Hobart andWilliam Smith Colleges. Finally, we thank CharlesMcClennen and anonymous reviewers for theircomments that improved earlier drafts of this manu-script. This article is Contribution 1314 of theUSGS Great Lakes Science Center.

REFERENCESBeletsky, D., Saylor, J.H., and Schawb, D.J., 1999. Mean

circulation in the Great Lakes. J. Great Lakes Res.25:78–93.

Csanady, G.T. 1978. Water circulation and dispersalmechanisms. In Lakes, Chemistry, Geology andPhysics, A. Lerman, ed., pp. 21–64. New York, NY:Springer-Verlag.

Dean, W.E. 1974. Determination of carbonate andorganic matter in calcareous sediments and sedimen-tary rocks by loss on ignition: Comparison with othermethods. J Sed. Petrol. 44:242–248.

Dittman, D.E., Halfman, J.D., and Owens, R.W. 2005.Influence of Exotics and Physical SedimentologicalFactors on Diporeia. Final Project Report USEPA.

Eadie, B.J., Chambers, R.L., Gardner, W.S., and Bell,G.L. 1984. Sediment trap studies in Lake Michigan:Resuspension and chemical fluxes in the southernbasin. J. Great Lakes Res. 10:307–321.

Etherington, M.D., Halfman, J.D., and Dittman, D.E.2002. Influence of sedimentological factors on thedistribution of Diporeia and Dreissena species in LakeOntario. Geological Society of America NortheastRegional Annual Meeting Abstracts with Programs, v.34, p. A31.

Ferrini, V.L., and Flood, R.D. 2001. Sedimentary charac-teristics and acoustic detectability of ship-deriveddeposits in western Lake Ontario. J. Great Lakes Res.27:210–219.

Flood, R.D. 1981. Distributions, morphology, and originof sedimentary furrows in cohesive sediments,Southampton Water. Sedimentology 28:511–529.

———, and Johnson, T.C. 1984. Side-scan targets inLake Superior—Evidence for bedforms and sedimenttransport. Sedimentology 31:311–333.

Folk, R.L. 1974. Petrology of Sedimentary Rocks.Austin, Texas: Hemphill Publishing Co.

Halfman, J.D., and Herrick, D.T. 1998. Mass movementand reworking of late glacial and postglacial sedi-ments in northern Seneca Lake, New York. NE Geol.Environ. Sci. 20:227–241.

Hawley, N., Wang, X., Brownawell, B., and Flood, R.1996. Resuspension of bottom sediments in LakeOntario during the unstratified period, 1992–1993. J.Great Lakes Res. 22:707–721.

Heiri, D., Lotter, A.F., and Lemke, G. 2001. Loss onignition for estimating organic and carbonate contentin sediments: reproducibility and comparability ofresults. J. Paleolimn. 25:101–110.

Hutchinson, D.R., Lewis, C.F.M., and Hund, G.E. 1993.Regional stratigraphic framework of surficial sedi-ments and bedrock beneath Lake Ontario. Geographiephysique et Quaternaire 47:337–352.

Johnson, T.C. 1980. Sediment redistribution by waves inlakes, reservoirs and embayments. In Proc. Sym. Sur-face Water Impoundments, pp. 1307–1317. Amer.Soc. Civil Eng.

———. 1984. Sedimentation in Large Lakes. Ann. Rev.Earth Planet. Sci. 12:179–204.

———, and Ng’ang’a, P. 1990. Reflections on a rift lake.In Lacustrine Basin Exploration—Case studies andmodern analogues, B.J. Katz, ed., pp. 113–135. Tulsa,OK: AAPG Memoir 50.

———, Halfman, J.D., Busch, W.H., and Flood, R.D.1984. Effects of bottoms currents and fish on sedi-mentation in a deep-water, lacustrine environment.Bull. Geol. Soc. Am. 95:1425–1436.

360 Halfman et al.

Kemp, A.L.W., and Harper, N.S. 1976. Sedimentationrates and a sediment budget for Lake Ontario. J.Great Lakes Res. 2:324–340.

———, Anderson, T.W., Thomas, R.L., and Mudrochova,A. 1974. Sedimentation rates and recent sediment his-tory of Lakes Ontario, Erie and Huron. J. Sed. Petrol.44:207–218.

Lewis, C.F.M., and Sly, P.G. 1971. Seismic profiling andgeology of the Toronto waterfront area of LakeOntario. In Proc. 14th Conf. Great Lakes Res., pp.303–354. Internat. Assoc. Great Lakes Res.

Martini, I.P., and Bowlby, J.R. 1991. Geology of theLake Ontario Basin: A review and outlook. Can. J.Fish. Aquat. Sci. 48:1503–1516.

Mudroch, A., and Mudroch, P. 1992. Geochemical com-position of the nepheloid layer in Lake Ontario. J.Great Lakes Res. 18:132–153.

Mullins, H.T., and Halfman, J.D. 2001. High-resolutionseismic reflection evidence for middle Holocene envi-ronmental change, Owasco Lake, New York.Quatern. Res. 55:322–331.

Owens, R.W., and Dittman, D.E. 2003. Shifts in the dietsof slimy sculpin (Cottus cognatus) and lake whitefish(Coregonus clupeaformis) in Lake Ontario followingthe collapse of the burrowing amphipod Diporeia.Aquat. Ecosys. Health Manage. 6(3):311–323.

Rea, D.K., Owen, R.M., and Meyers, P.A. 1981. Sedi-mentary processes in the Great Lakes. Rev. Geophys.Space Phys. 19:635–648.

Rosa, F. 1985. Sedimentation and sediment resuspensionin Lake Ontario. J. Great Lakes Res. 11:13–25.

Rukavina, N.A. 1976. Nearshore sediments of LakeOntario and Erie. Geosci. Can. 3:185–190.

Sandilands, R.G., and Mudroch, A. 1983. Nepheloidlayer in Lake Ontario. J. Great Lakes Res. 9:190–200.

Scholz, C.A., Johnson, T.C., and McGill, J., 1993.Deltaic sedimentation in a rift-valley lake: New seis-mic reflection data from Lake Malawi, (Nyasa), EastAfrica. Geology 21:395–398

Scrudato, R.J., and DelPrete, A. 1982. Lake Ontario sed-iment-mirex relationships. J. Great Lakes Res.8:695–699.

Simons, T.J., Murthy, C.R., and Campbell, J.E. 1985.Winter circulation in Lake Ontario. J. Great LakesRes. 11:423–433.

Sly, P.G. 1980. Sedimentary processes in lakes. InLakes, chemistry, geology and physics, A. Lerman,ed., pp. 63–69. New York, NY: Springer-Verlag.

Soreghan, M.J., Scholz, C.A., and Wells, J.T. 1999.Coarse-grained deep-water sedimentation along a bor-der fault margin of Lake Malawi, Africa: SeismicStratigraphic Analysis. J. Sediment. Res. 69:832–846.

Sutton, R.G., Lewis, T.L., and Woodrow, D.L. 1970.Near-shore sediments in southern Lake Ontario, theirdispersal patterns and economic potential. In Proc.13th Conf. Great Lakes Res., pp. 308–318. Internat.Assoc. Great Lakes Res.

Thomas, R.L., Kemp, A.L.W., and Lewis, C.F.M. 1972.Distribution, composition and characteristics of thesurficial sediments of Lake Ontario. J. Sed. Petrol.42:66–84.

Viekman, B.E., Flood, R.D., Wimbush, M., Faghri, M.,Asako, Y., and Van Leer, J.C. 1992. Sedimentary fur-rows and organized flow structure: A study in LakeSuperior. Limnol. Oceanogr. 37:797–812.

Virden, W.T., Warren, J.S., Holcombe, T.L., Reid, D.F.,and Berggren, T.L. 1999. Bathymetry of LakeOntario. National Geophysical Data Center, BoulderCO. USGS MGG-15.

Walker, A.E., Halfman, J.D., Dittman, D.E. 2002. Sedi-mentological controls on the distribution of Diporeiaand Dreissena species in Lake Ontario. GeologicalSociety of America Annual Meeting Abstracts withPrograms, v. 34, p. 258.

Submitted: 28 January 2005Accepted: 26 February 2006Editorial handling: Harvey Thorleifson

storm-induced redistribution of deepwater sediments in...

Documents