Poppe, L.J.¹, Lewis, R.S.²,Denny, J.F.¹,Parolski, K.F.¹, and DiGiacomo-Cohen, M.L.²

1 U.S. Geological Survey Woods Hole, MA 02543
² Long Island Sound Resource Center, Connecticut Geological and Natural History Survey,
Department of Environmental Protection, 1084 Shennecossett Road, Avery Point, Groton, CT 06340

The U.S. Geological Survey has produced detailed geologic maps of the sea floor in Long Island Sound, a major East Coast estuary surrounded by the most densely populated region of the United States. These studies have built upon cooperative research with the State of Connecticut that was initiated in 1982. During the beginning phase of this program, geologic framework studies in Long Island Sound were completed and results published (Needell and Lewis, 1984; Lewis and Needell, 1987; Needell and others, 1987; Lewis and Stone, 1991). The second and current phase of the program emphasizes studies of sediment distributions, processes that control these distributions, nearshore environmental concerns, and the relation of benthic community structures to the sea-floor geology.

Anthropogenic wastes, toxic chemicals, and changes in land-use patterns resulting from residential, commercial, and recreational development have stressed the environment of the Sound, causing degradation and potential loss of benthic habitats (Long Island Sound Study, 1994). Detailed maps of the sea floor are needed to help evaluate the extent of adverse impacts and to help manage resources wisely in the future. Therefore, in a continuing effort to better understand Long Island Sound, we have constructed sidescan sonar mosaics (complete-coverage acoustic images) of the sea floor within areas of special interest (fig. 1). The mosaic presented herein covers approximately 17.7 km² of the sea floor in Fishers Island Sound and northwesternmost Block Island Sound (fig. 5). The mosaic enables us to define the sea floor's geological variability, which is one of the primary controls of benthic habitat diversity. It also improves our understanding of the processes that control the distribution of bottom sediments, benthic habitats, and associated infaunal community structures, and it provides a detailed framework for future research, monitoring, and management activities. The sidescan sonar mosaics produced during this program also serve as base maps for subsequent sedimentological, geochemical, and biological observations because precise information on environmental setting is important to the selection of sampling sites and to the accurate interpretation of point measurements.

Fishers Island Sound, which lies between 71^o53'W and 72^o03'W., is about 14 km long and 4.6 km wide (fig. 1). The sound is bordered by Fishers Island, N.Y., on the south, southeastern Connecticut on the north, and southwestern Rhode Island on the northeast. Fishers Island Sound opens into eastern Long Island Sound on the west and into Block Island Sound to the southeast. Water depths in Fishers Island Sound are irregular, but generally increase toward the west where the maximum water depth exceeds 28 m (fig. 2).

Crystalline rocks of pre-Mesozoic age dominate the bedrock lithology beneath Fishers Island Sound (Needell and Lewis, 1984; Lewis and Needell, 1987). Onshore in southeastern Connecticut, dark-gray, fine- to medium-grained, Precambrian (Avalonian terrane; Rodgers, 1985) rocks of the Rope Ferry Gneiss and (to a lesser extent) rocks of the Potter Hill Granitic Gneiss and the Hope Vally Alaskite Gneiss cropout in the New London, Mystic, and Old Mystic 7.5-minute Quadrangles (Goldsmith, 1967, 1985; Rodgers, 1985). The Avalonian anticlinorium is locally intruded by the Westerly Granite (Permian) now noted by Hermes and others (1994) as the fine-grained component of the Narragansett Pier plutonic suite. In southwestern Rhode Island, pinkish-gray, Permian, magmatic rocks of the Narragansett Pier Granite cropout in the Watch Hill Quadrangle and underlie Napatree Point (Moore, 1967).

The bedrock in Fishers Island Sound is unconformably overlain by glacial deposits that are composed mainly of glaciolacustrine silt and sand laid down as deltaic fans and of glacial drift laid down as part of a series of recessional end moraines across southeastern Connecticut (Goldsmith, 1962; Upson, 1971; Lewis and Stone, 1991; Stone and others, 1992). The Harbor Hill - Roanoke Point - Charlestown moraine belt, which is of late Wisconsinan age, lies across northern Long Island and also makes up Fishers Island (fig. 1). A less prominent end moraine (the Clumps moraine) controls much of the bathymetric relief in the central and southern part of Fishers Island Sound. Salt marsh and swamp deposits overlie bedrock and some thinner glacial deposits in environments of low wave energy along the northern shoreline of the sound (Stone and others, 1992).

The surficial bottom sediments of the sound are composed predominantly of materials that have been reworked from the glacial deposits during Holocene time by strong tidal currents that range up to 2.0 knots (Eldridge Tide and Pilot Book, 1994). The irregular bottom topography, well-developed bedforms, and extensive lag deposits of boulders that occur in Fishers Island Sound reflect this scour, transport, and reworking of the glacial and early postglacial deposits (Lewis and Stone, 1991).

Sidescan sonar imagery, bathymetry, 3.5-kHz high-resolution seismic-reflection profiles, and navigational data were collected along tracks spaced 150 m apart aboard the RV Asterias during June 10-11, 1995 (figs. 2, 3, and 5). The bathymetric data were collected by means of a 200-kHz echo sounder, and the data were logged digitally. The subbottom data (fig. 3) were collected in analog form using an Ocean Research Equipment 3.5-kHz profiler transmitting at a 0.25-s repetition rate. The sidescan sonar data were collected using a 100-kHz Klein sidescan sonar system set to sweep 100 m to either side of the ship's track. These data were logged digitally to 8-mm tape on a QMIPS data acquisition system (Danforth and others, 1991). Ship position was determined with a differential Global Positioning System (GPS) was logged digitally at 10-s intervals. The bathymetric data were corrected for the approximately 0.82-m tidal range in the study area by adjusting the measured depth values to the predicted mean sea level for New London, Conn., approximately 3 km west of the study area.

The sidescan sonar data were processed according to procedures summarized by Danforth and others (1991) and Paskevich (1992a, b). Briefly, the sonar data were multiplexed and filtered to convert the data into a processing format and to remove speckle noise and corrected for slant-range distortions. Additional corrections were applied to compensate for signal attenuation and dropped lines in the sonar data. A contrast enhancement based on the dynamic range of the data was applied, ship navigation was merged with the sonar data, and the data were geographically oriented and displayed on a Mercator grid.

After this preliminary processing the data were used to make the composite digital mosaic. Alternating strips of sonar data were placed in their proper geographic location at the appropriate scale and projection (Paskevich,1992c). Adjacent sonar images were matched for tone and images were progressively combined into a composite digital image; unwanted data where two images overlapped were digitally. Registration between the shiptrack navigation and the strips of sidescan sonar data are generally good throughout the study area.

The final image is a digital mosaic with a 1-m pixel size (a pixel represents one meter on the sea floor).  The 47Mbyte digital mosaic, with its associated Arcview geo-reference world file, is available for viewing as a graphic in "tif" format.  This image was also output to film, photographically enlarged to a scale of 1:10,000, and interpreted (figs. 5 and 6).

Surficial sediment sampling and bottom photography were attempted at 80 locations within the boundaries of the present study area during June 1992 aboard the RV Asterias using a Van Veen grab sampler equipped with a video camera system. The map of surficial sediment distribution shown in figure 4 is based on data from these sediment sampling and bottom photography stations, supplemented by textural data from earlier studies (Schlee, 1973; Akpati, 1974). It is also based on tonal changes in backscatter on the sidescan sonar image, and on the bathymetry. The photographic system was used to appraise intra-station bottom variability and to observe boulder fields and bedrock outcrops where sediment samples could not be collected. Surficial sediments (sediments from the sediment-water interface down to 2 cm below the interface) were collected from the grab sampler; these subsamples were frozen and stored for later analysis.

In the laboratory, the sediment samples were disaggregated and wet sieved to separate the coarse and fine fractions. The fine fraction (less than 62 microns) was analyzed by Coulter Counter (Shideler, 1976); the coarse fraction was analyzed by sieving (gravel) and by rapid sediment analyzer (sand; Schlee, 1966). The data were corrected for salt content of the samples.

Because biogenic carbonates commonly form in situ, they usually are not considered to be sedimentologically representative of the depositional environment from a textural standpoint. Therefore, bivalve shells and other biogenic carbonate debris were manually removed from the gravel fraction. Size classifications are based on the method proposed by Wentworth (1929). The verbal equivalents were calculated using the inclusive graphics statistical method (Folk, 1974) and are based on the nomenclature proposed by Shepard (1954). A detailed discussion of the laboratory methods employed is given in Poppe and others (1985); the field methods, station navigation, raw grain-size data, associated statistics, and detailed descriptions of the bottom photography are reported in Poppe and others (1993).

Bathymetry

The corrected bathymetry was contoured at 2-m intervals to facilitate an understanding of the complex geology and irregular benthic character of Fishers Island Sound (fig. 2). Bathymetric highs in the central and southern part of the sound are somewhat isolated, but primarily occur in a discontinuous east-northeast-trending band that includes West Clump, Middle Clump, East Clump, and Latimer Reef. This band of bathymetric highs defines the submerged part of a minor recessional end moraine (Clumps moraine) that occurs north of the more prominent Harbor Hill - Roanoke Point - Charlestown moraine (Lewis and Stone, 1991; fig.1), which comprises Fishers Island. Bathymetric highs in the northern part of the sound are oriented more north-south and may reflect the underlying bedrock surface.

Bathymetric lows occur between the recessional moraine segments and at the eastern and western entrances to the sound. Maximum water depths exceed 28 m in two isolated depressions that occur north of Wicopesset Island and between Intrepid Rock and West Clump. Generally, water depths shoal northward and around Fishers Island, but outlying areas of relief greater than 2 m are common.

Near-Surface Seismic Character

The 3.5-kHz seismic-reflection profiles provide information on the subsurface geology and its relation to surficial processes and sediment distribution. Acoustic facies recorded in the profiles include a prolonged facies, a relatively transparent facies, and an opaque facies. Surficial sedimentary features, such as sand waves and depressions related to current scour, were also identified on the profiles.

The acoustically prolonged facies, which typically represents acoustic basement, has an irregular, highly reflective upper surface and is interpreted to be predominantly glacial drift (fig. 3), the oldest unit mapped as part of this study. The drift is primarily till that is associated with the Clumps moraine, the minor recessional end-moraine that underlies parts of Fishers Island Sound. However, some of this facies in the northwestern part of the study area may also represent bedrock. The acoustically prolonged facies crops out on most bathymetric highs and along the central, or long, axis of Fishers Island Sound, where the tidal currents are strongest.

The relatively transparent acoustic facies overlies and is younger than the prolonged acoustic facies. This unit is thin throughout much of the study area, but can be thicker where it ponds within depressions in the prolonged acoustic facies (fig. 3B, D). It also thickens leeward of bathymetric highs (fig. 3C, D) and in protected environments, such as the area south and southeast of East Clump (fig. 3B). The relatively transparent acoustic facies is variably layered and is interpreted to be Holocene sediment. Internal layering within this unit appears to decrease toward the axis of Fishers Island Sound, presumably because of the higher energy depositional environment along the axis.

The acoustically opaque facies occurs within the relatively transparent acoustic facies and is interpreted to result from gas-charged sediment (figs. 2 and 3B). The distribution of this facies is primarily limited to areas where the subbottom depth to acoustic basement is relatively great and where the depositional environment is one of lower energy, such as the area between Ram Reef and Eel Grass Ground and the area south and southeast of East Clump. The top of this facies occurs 3-6 m beneath the sediment-water interface. Gas-charged sediment, where detected, obscures most of the underlying units and acoustic basement.

Sand waves occur in scattered patches throughout the study area, but are best developed just west of Latimer Reef (figs. 3D and 6) and southeast of Napatree Point (fig. 6). The average crest to trough height of these features is less than 0.5 m, but occasionally exceeds 2 m. Asymmetry of the sand waves suggests that they migrate. Although sediment transport around Latimer Reef is toward the east, most sediment transport within the sound appears to be westward. A similar transport regime is suggested by the distribution of iron-stained quartz (Akpati, 1974).

Depressions in the Holocene sediment cover are commonly present immediately adjacent to outcrops of the prolonged acoustic facies (fig. 3A). The location of these depressions suggests scour by currents that are deflected around the more resistant tills that typically compose the bathymetric highs.

Sidescan Sonar Mosaic

The 100-kHz sidescan sonar mosaic (fig. 5) portrays an acoustic image of the sea floor that, when combined with the 3.5 kHz subbottom and bathymetric data, can be used to interpret the surficial geology (fig. 6). Distinctive acoustic patterns revealed by this mosaic include the following: complex patches of high and low backscatter with high-backscatter targets (objects), areas of higher backscatter (light tones), areas of lower backscatter (dark tones), and elongate patches of alternating high and low backscatter. Anthropogenic phenomena, such as boat wakes and a shipwreck, are also present.

The complex patches of high and low backscatter with individual high-backscatter targets occur over areas of the sea floor that are characterized by concentrations of boulders and gravel. The distribution of this acoustic pattern coincides with areas where the prolonged facies crops out and where higher energy environments prevail, such as on the tops of bathymetric highs that are associated with the discontinuous northeast-trending Clumps moraine (figs. 5 and 6). Presumably, the strong tidal and storm currents in these high energy environments have prevented the deposition of Holocene sediments and winnowed the finer grained fraction from the glacial drift, leaving lag deposits of boulders and gravel.

The areas of higher backscatter on the sidescan sonar image are characterized by coarser-grained sediments and reflect environments of more moderate energy (figs. 5 and 6). This acoustic pattern is prevalent along the central axis of the sound, around isolated bathymetric highs, and surrounding the patches of boulders and gravel.

The areas of lower backscatter are interpreted to represent finer grained Holocene marine sediments. This acoustic pattern tends to occur in lower energy environments, such as a protected area along the north shore of Fishers Island, in isolated bathymetric lows, and on the leeward side of bathymetric highs.

Areas dominated by elongate patches of alternating high and low backscatter are scattered throughout the study area but are most extensive southwest of Ram Reef, west of Latimer Reef, north of Wicopesset Island, west of Napatree Point ledge, and southeast of Napatree Point (figs. 5 and 6). This acoustic pattern is produced by sand waves that range in wavelength up to approximately 50 m. The alternation in backscatter that creates this pattern results from a combination of topographic changes affecting the angle of incidence of the sidescan sonar and the differences in sediment texture typically present between the crests and troughs of sand waves (Reineck and Singh, 1980).

Surficial Sediment Distribution

The distribution of surficial sediment and the locations of the sediment sampling and bottom photography stations and are shown in figure 4. The irregular boundaries and patchy nature of the sediments reveal that the benthic environments vary spatially over short distances; this is also evident on the sidescan sonar image and interpretation (figs. 5 and 6). A comparison with an earlier sediment-distribution map based solely on random stations (Poppe and others, 1994) clearly demonstrates that highly variable sea floors, such as the one present in Fishers Island Sound, are too heterogeneous and complex to be mapped by measurements made solely at discrete sample locations with conventional spacings. Boundaries on this earlier map are smoother, and most of the extreme interstation variability was not detected.

Boulders, gravel, and gravelly sand dominate the surficial sediments along the central axis of Fishers Island Sound. Boulders and gravel are common at the eastern and western entrances to the sound, especially over the submerged part of the Harbor Hill - Roanoke Point - Charlestown moraine (i.e. between the northeast end of Fishers Island and Napatree Point) where the tidal flow is constricted and the currents are the strongest. Boulders and gravel are also prevalent on bathymetric highs associated with the Clumps moraine (for example Middle Clump and Latimer Reef; Lewis and Stone, 1991) and directly offshore from promontories where episodic storm currents combine with the tidal currents (Youngs Rock for example). The boulders in these areas are commonly covered by kelp and other unidentified seaweed and are variably inhabited by starfish, anemones, and mussels. No bedrock outcrops were observed at any of the sediment sampling stations.

Aprons of gravelly sand (less than 50, but greater than 10 percent gravel) commonly flank the areas characterized by boulders and gravel. Examination of the grab samples revealed that the gravel portion of the gravelly sediments often occurs in thin surficial layers, which are underlain by finer grained sediments. This stratification is caused by strong tidal currents that erode the underlying glacial tills and winnow away the finer grained materials. Subsequently, thin lag deposits of gravelly sediment remain to armor the underlying silts and sands. The grain-size distributions of the gravelly sediments tend to be poorly to very poorly sorted and bimodal. Three explanations are offered for the bimodality. First, till, which by definition is poorly sorted, underlies and is exposed in the study area (Lewis and Stone, 1991). Second, the grab sampler penetrated the lag deposits described above, collecting both the gravel armor and the underlying finer grained sediments. Third, seaweed-encrusted coarse gravel was occasionally observed being "rafted" by currents across the study area during bottom photography. When the seaweed dies, the gravel is stranded in finer grained, hydraulically unequivalent sediments and the sediment becomes a gravelly sand. Sponges, rock crabs, lobsters, starfish, welks, and mussels are commonly observed in the bottom photography of the gravelly sediments.

Sand dominates in the more protected areas, such as in the area between Middle Clump and Youngs Rock. Bedforms are common in the sandy areas and include current and oscillation ripples, and, in places along the gravelly central axis, sand waves. Earlier work (Akpati, 1974; Poppe and others, 1993) has shown that the sands in the eastern part of Fishers Island Sound and adjacent to the gravelly areas tend to have moderately to moderately well sorted, unimodal grain-size distributions. Sorting in these sands decreases both inshore toward coves, such as the southeast of East Clump, and toward the western part of the sound. Shell hash, hermit crabs, gastropods, skates, flounder, and burrows were observed in the bottom photography from these sandy areas.

Finer-grained sands and silty sands are restricted to lower energy environments, such as a small area about 1 km southwest of Ram Island and in coves along the north shore of Fishers Island. These muddy sediments are typically heavily burrowed by amphipods, polychetes, and bivalves. Crabs (spider and cancer) were commonly observed during the bottom photography of the areas characterized by finer grained sediments.

The detailed bathymetric, 3.5-kHz subbottom, sidescan sonar, sampling, and bottom photographic data presented herein reveal the extreme variability of the sedimentary environments and benthic habitats in Fishers Island Sound. Because of their irregular, patchy nature, these environments are clearly too complex to be accurately mapped by measurements made solely at reasonably-spaced random sample locations.

The present distribution of surficial sediments in Fishers Island Sound is shaped from the deposits left by the last glaciation and reflects the cumulative effects of reworking, transport, and deposition by tidal and (to a lesser degree) storm currents during and since the Holocene eustatic rise in sea level. The predominantly coarse-grained nature of the sediments and the abundance of bedforms are a direct result of these strong currents. The presence of finer grained sediments only in the westernmost part of Fishers Island Sound and in coves along the north shore of Fishers Island suggests that more tranquil conditions, which can permit muddy sediments to accumulate, are rarely present along the axis of the sound.

This report was prepared in cooperation with the State of Connecticut, Department of Environmental Protection, Geological and Natural History Survey and was originally published as US Geological Survey Geologic Investigations Map I-2640. We thank Dave Olmsted, who captained the Woods Hole Oceanographic Institution vessel RV Asterias; the University of Connecticut, for providing logistical support; and Kathy Scanlon-Catanach and Marilyn ten Brink (USGS), who critically reviewed this map. Mean sea-level tidal-height measurements used to correct the bathymetric data were computed by Michael McCray at the Tidal Analysis Branch of the National Oceanographic and Atmospheric Administration, Silver Spring, Md. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government or the State of Connecticut.

FIGURES

Figure 1.--Map showing location of study area (diagonal stripe-filled polygon). Also shown, as open polygons, are the sites of other surveys completed to date that are part of or preparatory to this series: Norwalk--Twichell and others (1997); Milford--Twichell and others (1998); New Haven Harbor and New Haven Dumping Ground--Poppe and others (1998c); Roanoke Point--Poppe and others (1998a); Falkner Island--Poppe and others (1997a); Hammonasset--Poppe and others (1997b); Niantic Bay--Poppe and others (1998b); New London--Poppe and others (1992), Moffett and others (1994), Zajac and others (1995). Major onshore moraines are indicated.

Figure 2.--Bathymetric map of the study area. Depths are in meters, have been corrected for tides, and are adjusted to mean sea level. Fine dashed lines represent tracks along which the geophysical data were collected. The distribution of an opaque subbottom acoustic facies interpreted to be gas-charged sediment and the locations of profiles A, B, C, and D, shown in figure 3, are indicated.

Figure 3.--Subbottom profiles (3.5-kHz) showing different echo- character types. (A) Profile shows the acoustically prolonged facies, interpreted to be glacial drift. Bathymetric depressions around the drift on this profile suggest scour by the strong tidal currents. (B) Profile shows the prolonged acoustic character of the glacial drift, sediment ponds that occur within depressions in the drift, the relatively transparent acoustic facies interpreted to be Holocene sediment, and the opaque acoustic facies intrepreted to result from gas-charged sediments. (C) Profile shows prolonged acoustic nature of the drift beneath Napatree Point Ledge and sand waves on the wedge of Holocene sediment (transparent acoustic facies) to the west of the ledge. (D) Profile shows the sand waves on the relatively thick wedge of Holocene sand (transparent acoustic facies) west of Latimer Reef. Locations of the profiles are shown in figure 2. Water depths are not corrected for the tidal range; distance scales are averaged and not corrected for changing ship speed.

Figure 4. Map showing distribution of surficial sediments and locations of the sediment sampling and (or) bottom photography stations (square--Schlee, 1973; triangles--Akpati, 1974; solid circles--Poppe and others, 1994). Interpretation of the sediment distribution is based on data from these stations, on tonal changes in backscatter on the sidescan sonar image (fig. 5), and on the correlation of the textural and backscatter data with the bathymetry (fig. 2).

Figure 5. Sidescan sonar mosaic of the sea floor in the study area. Light tones on the image represent areas of high backscatter (typically coarser grained sediment); dark tones represent areas of low backscatter (typically finer grained sediment).

Figure 6. Interpretation of the sidescan sonar mosaic. Map shows the distribution of boulders and gravel, sand waves, areas characterized by high backscatter, and areas marked by low backscatter. The different patterns are discussed in the text.

SIDESCAN-SONAR MOSAIC

The sidescan-sonar mosaic is a digital image with a 1-m pixel size (a pixel represents one meter on the sea floor). The 47M-byte digital image (fisher.tif) is available for viewing as a graphic in "tif" format with its associated Arcview geo-reference file (fisher.tifw).