Introduction
Warren Upham recognized the significance of the proglacial lake and marine waterbodies in the Champlain and Hudson lowlands when he wrote;
“The Hudson—Champlain area, made classic in glacial geology by the works of C.H. Hitchcock, Baldwin, Baron de Geer, Gilbert, Merrill, Peet, Woodworth, and others, which through the writings of Hitchcock and Dana gave the name Champlain to the closing epoch of the Ice Age, deserves yet further work of detailed surveys, with exact leveling for determination of the relations of all its lacustrine and marine shore lines. No other area of our continent promises more important information concerning the Glacial and Recent periods” (Upham, Reference Upham1905)
One hundred and twenty years later, high-resolution LiDAR elevation models provide an important new tool to conduct the kind of detailed analysis Upham envisioned. This study reevaluates the deglacial lake and marine chronology of the eastern Ontario, western St. Lawrence, and Champlain Lowlands (Fig. 1) in the context of newly interpreted strandline data, principally beach ridges, compiled from LIDAR-derived digital elevation models (DEMs). The deglacial chronology of this region has important implications for ice recessional history, meltwater routing to the North Atlantic, and global climate (Teller, Reference Teller and Gadd1988; Broecker et al., Reference Broecker, Kennett, Flower, Teller, Trumbore, Bonani and Woelfli1989; Rahmstorf, Reference Rahmstorf1995; Donnelly et al., Reference Donnelly, Driscoll, Uchupi, Keigwin, Schwab, Thieler and Swift2005; Tarasov and Peltier, Reference Tarasov and Peltier2005; Broecker, Reference Broecker2006). Proglacial lakes and marine estuaries, described later, also overlapped with Paleoindian occupation of New York and could have factored into the migration, regional subsistence, and land-use practices of the earliest people in Hudson and St. Lawrence Valleys (Lothrop and Bradley, Reference Lothrop, Bradley and Chapdelaine2012).
Figure 1. Shaded relief map of the Ontario, St. Lawrence, and Champlain Lowlands. The study area is indicated by the red shading adjacent to the northern and eastern Adirondack Upland.
Many proglacial lake phases in the published literature were defined from small sample populations of comparatively low-resolution data that included unreliable strandline proxies, inaccurate elevation estimates, or a limited understanding of the direction and magnitude of regional isostatic rebound and, in most cases, were presumed to have stable outlet thresholds. Furthermore, shorelines of large proglacial lakes are diachronous, so elevation data used for shoreline reconstruction may not be the same age everywhere in the basin (Lewis et al., Reference Lewis, Breckenridge and Teller2022). These limitations contributed to the proliferation of names given to presumed proglacial lake stages in different areas. Published references to proglacial lake names have become a matter of individual discretion, leading to the development of an overlapping and complex nomenclature for different parts of the region (Table 1). From our spatial analysis of strandline deposits and landforms, principally beach ridges, we revise and redefine proglacial lake and marine phases in a way that reflects the interaction between ice recession and changes in outlet elevation or location, glacial isostatic adjustment, and relative sea-level change.
Table 1. Nomenclature for proglacial lake and marine waterbodies in the Hudson–Champlain, eastern Ontario, and western St. Lawrence Lowlands.a
a Principal sources: Woodworth, Reference Woodworth1905a; Chapman, Reference Chapman1937; MacClintock and Stewart, Reference MacClintock and Stewart1965; Wagner, Reference Wagner, Doolan and Stanley1972; Denny, Reference Denny1974; Muller and Prest, Reference Muller, Prest, Karrow and Calkin1985; Parent and Occhietti, Reference Parent and Occhietti1988, Reference Parent and Occhietti1999;
Pair and Rodrigues, Reference Pair and Rodrigues1993; Wall and Lafleur, Reference Wall, Lafleur, Garver and Smith1995; Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007; Cronin et al., Reference Cronin, Manley, Manley, Brachfeld, Manley, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008; DeSimone et al., Reference DeSimone, Wall, Miller, Rayburn, Kozlowski, Dineen and Hanson2008; Stanford, Reference Stanford2009; Rayburn et al., Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011; Lewis and Todd, Reference Lewis and Todd2019; the present study.
Proglacial lake and marine nomenclature
The sequence of proglacial lakes and marine waterbodies in the Ontario, St. Lawrence, and Champlain Lowlands has been studied for more than a century (Table 1). Thoughtful summaries of early work can be found in MacClintock and Stewart (Reference MacClintock and Stewart1965), Muller and Prest (Reference Muller, Prest, Karrow and Calkin1985), Rodrigues (Reference Rodrigues1992), Lewis and Todd (Reference Lewis and Todd2019), and Lewis et al. (Reference Lewis, Breckenridge and Teller2022) for the Ontario and western St. Lawrence Lowlands, and in Chapman (Reference Chapman1937), Connally and Sirkin (Reference Connally, Sirkin, Black, Goldthwait and Willman1973), Muller et al. (Reference Muller, Franzi, Ridge and Cadwell1986), Wall and Lafleur (Reference Wall, Lafleur, Garver and Smith1995), Rayburn et al. (2005, Reference Rayburn, DeSimone, Frappier, Grover and Mango2018), and DeSimone et al. (Reference DeSimone, Wall, Miller, Rayburn, Kozlowski, Dineen and Hanson2008) for the Hudson–Mohawk and Champlain Lowlands. Deglacial thinning of the LIS over upland regions (Franzi et al., Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016; Barth et al., Reference Barth, Marcott, Licciardi and Shakun2019) and drawdown of ice lobes into peripheral lowlands impounded extensive proglacial lakes along the receding margin of the LIS (Fig. 2). These lakes expanded northward with ice recession and evolved through a series of falling lake phases in response to differential glacial isostatic uplift, relative sea-level change, outlet incision, and shifting positions of their drainage outlets. We assign the names Lake Abenaki, Lake Akwesasne, endorheic Lake Ontario, Champlain Estuary, and Hochelagan Estuary in the following summary (Table 1). These names are explained in the “Discussion.”
Figure 2. Proglacial lake succession in the western Ontario, western St. Lawrence, and Hudson–Champlain Lowlands. Details given in the text. Arrows indicate outflow routes from glacial lakes. Principal sources: Muller and Prest (Reference Muller, Prest, Karrow and Calkin1985); Parent and Occhietti (Reference Parent and Occhietti1988); Pair and Rodrigues (Reference Pair and Rodrigues1993); Lewis and Todd (Reference Lewis and Todd2019), Lewis et al. (Reference Lewis, Breckenridge and Teller2022), and Franzi et al. (Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016). Modified from figures in Franzi et al. (Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016) and prepared in collaboration with the Adirondack Journal of Environmental Studies. The names Lake Abenaki, Lake Akwesasne, Endorheic Lake Ontario, Champlain Estuary, and Hochelagan Estuary (Table 1) are explained in the “Discussion.” (A) Early Lake Iroquois with outflow via the Glaciomohawk River to Lake Albany; (B) Lake Iroquois with outflow via the Iromohawk River to Lake Abenaki. The lakes are depicted near their maximum extents immediately before the Lake Iroquois breakout at The Gulf near Covey Hill; (C) Lake Akwesasne with outflow across the bedrock threshold at Fort Ann; (D) the Champlain Estuary at the upper marine limit; (E) Endorheic Lake Ontario and the Hochelagan Estuary.
Lake Iroquois formed in the Ontario basin following ice recession from the northern flank of the Appalachian Plateau ∼14.0 cal ka BP (Fairchild, Reference Fairchild1909; Coleman, Reference Coleman1937; Muller and Calkin, Reference Muller and Calkin1993; Lewis and Todd, Reference Lewis and Todd2019; Fig. 2A). Early high-level lake phases remained confined to the southern and eastern portions of the Ontario Lowland and drained eastward across a threshold near Little Falls, New York (Pair and Rodrigues, Reference Pair and Rodrigues1993; Bird and Kozlowski, Reference Bird and Kozlowski2016; Porreca et al., Reference Porecca, Briner and Kozlowski2018; Fig. 1), through the Mohawk Valley as the Glaciomohawk River (Fairchild, Reference Fairchild1912) into Lake Albany in the upper Hudson Lowland (Figs. 1 and 2A). Lake Iroquois expanded northward with ice recession and became connected to lake outflow from proglacial lakes in the upper Great Lakes basin (Fig. 2B, Table 1). Flow through the Mohawk Lowland continued as the larger Iromohawk River (Fairchild, Reference Fairchild1912). Falling water level and northward expansion of Lake Albany accompanied recession of the ice margin in the upper Hudson and Champlain Lowlands (DeSimone et al., Reference DeSimone, Wall, Miller, Rayburn, Kozlowski, Dineen and Hanson2008; Stanford, Reference Stanford2009). As Lake Albany water level fell, a low point on the Hudson–Champlain drainage divide emerged about 6 km south of Fort Ann, New York, separating Lake Abenaki in the Champlain Lowland from Lake Albany in the upper Hudson Lowland (Fig. 2B, Table 1).
Outflow from Lake Iroquois shifted northward to the Champlain Lowland about 13 cal ka BP (Lewis and Todd, Reference Lewis and Todd2019; Lewis and Anderson, Reference Lewis and Anderson2020; Lewis et al., Reference Lewis, Breckenridge and Teller2022; Fig. 2B) as lower outlets emerged on the St. Lawrence–Champlain drainage divide near The Gulf Unique Area (The Gulf) in Clinton County, New York, about 2.5 km southwest of Covey Hill, Quebec (Woodworth, Reference Woodworth1905a; Fairchild, Reference Fairchild1912, Reference Fairchild1919; Denny, Reference Denny1974; Pair et al., Reference Pair, Karrow, Clark and Gadd1988; Pair and Rodrigues, Reference Pair and Rodrigues1993; Franzi et al., Reference Franzi, Rayburn, Yansa and Knuepfer2002, Reference Franzi, Rayburn, Knuepfer and Cronin2007, Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Fig. 1). The sudden outflow scoured the surficial cover material to form bedrock pavements, or Flat Rocks, in the northwestern Champlain Lowland (Woodworth, Reference Woodworth1905a, Reference Woodworth1905b; Denny, Reference Denny1974; Franzi et al., Reference Franzi, Rayburn, Yansa and Knuepfer2002, Reference Franzi, Rayburn, Knuepfer and Cronin2007, Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005) and flow through the Mohawk Lowland as the Iromohawk River ended.
Lake Iroquois briefly stabilized at The Gulf bedrock threshold (elevation 305 m) near Covey Hill (Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007, Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005, Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011). The stable interval correlates with the Lake Frontenac phase of Pair and Rodrigues (Reference Pair and Rodrigues1993) and may have lasted ∼50 yr based upon stratigraphic evidence in varve records from the Champlain Lowland (Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007; Rayburn et al., Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011). Final drainage of Lake Iroquois occurred as ice receded and floodwater discharged eastward along the northern flank of Covey Hill. Lake Iroquois fell until proglacial lakes in the Ontario, St. Lawrence, and Champlain Lowlands merged to form Lake Akwesasne (Pair and Rodrigues, Reference Pair and Rodrigues1993; Rayburn, Reference Rayburn2004; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007; Fig. 2C, Table 1).
Post-Iroquois proglacial lakes drained ∼12.8 cal ka BP (Lewis et al., Reference Lewis, Breckenridge and Teller2022) when ice recession in the Lake Chaudière–Etchemin area, Quebec, opened a connection to the Goldthwait Sea at the mouth of the St. Lawrence River and marine water inundated the isostatically depressed St. Lawrence and Champlain Lowlands, creating the Champlain Sea (Occhietti et al., Reference Occhietti, Parent, Lajeunesse, Robert, Govare, Ehlers, Gibbard and Hughes2011). Freshwater outflow across the Frontenac Arch, a broad structural high in eastern Ontario and northern New York, restricted circulation between the Champlain Sea and coeval Early Lake Ontario (ELO) (Pair and Rodrigues, Reference Pair and Rodrigues1993; Lewis and Todd, Reference Lewis and Todd2019; Lewis and Anderson, Reference Lewis and Anderson2020; Lewis et al., Reference Lewis, Breckenridge and Teller2022; Fig. 2D, Table 1). We refer to these contiguous waterbodies as the Champlain Estuary (Fig. 2D).
Falling water level, controlled by differential isostatic uplift and associated shift in lake outlet, relative sea-level change, and warming late Pleistocene climate, isolated water in the Ontario Basin from the upper glacial Great Lakes and the Champlain Estuary (Lewis, Reference Lewis2016; Lewis and Todd, Reference Lewis and Todd2019; Table 1). Endorheic (internally drained or closed basin) Lake Ontario occupied the lowest portion of the Ontario Basin and was separated from the Hochelagan Estuary by the Frontenac Arch (Lewis, Reference Lewis2016; Lewis and Todd, Reference Lewis and Todd2019; Fig. 2E, Table 1).
Shoaling, eastward regression of the Champlain Estuary in the western St. Lawrence Lowland, and freshwater inflow from adjacent uplands freshened water in the Hochelagan Estuary (Lake Lampsilis of Elson and Elson [Reference Elson and Elson1959], Elson [Reference Elson and Clark1962, Reference Elson and Gadd1988], Richard [Reference Richard1978], and Parent and Occhietti [Reference Parent and Occhietti1988]; Fig. 2E, Table 1). The Hochelagan Estuary occupied the lowland until ∼8.3 cal ka BP, when Lake Ontario reestablished connection to the western St. Lawrence Lowland and the Great Lakes to create the modern Lake Ontario–St. Lawrence River drainage system (Lewis, Reference Lewis2016; Lewis and Todd, Reference Lewis and Todd2019).
Strandline data
Strandline deposits and landforms such as deltas, shorezone platforms, and beach ridges mark the shores of proglacial lake and marine waterbodies in the study area. Many strandline features are diachronous or yield ambiguous water level estimates, especially on regressive shorelines. Elevations of deltaic topset–foreset contacts are commonly used proxies for paleo-water levels in northern New York and New England (Pair et al., Reference Pair, Karrow, Clark and Gadd1988; Koteff and Larsen, Reference Koteff and Larsen1989; Pair and Rodrigues, Reference Pair and Rodrigues1993; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Hooke and Ridge, Reference Hooke and Ridge2016), but their use requires identification of internal stratigraphy. Delta surfaces slope basinward on regressive shorelines and topset bed thickness is often unknown, leading to uncertainty regarding paleo-water level. Erosional shorezone platforms are horizontal to subhorizontal bedrock surfaces that form between the highest elevation of spring tide or storm surge erosion and the breaking depth of incoming waves (Kennedy, Reference Kennedy2015; Rovere et al., Reference Rovere, Raymo, Vacchi, Lorscheid, Stocchi, Gómez-Pujol, Harris, Casella, O’Leary and Hearty2016). The platform–sea cliff junction at the inner edge of a platform develops within a narrower range of elevation (∼1–4 m) between mean water level and the upper limit of spring tide or storm surge erosion (Kelsey, Reference Kelsey, Shennan, Long and Horton2015; Rovere et al., Reference Rovere, Raymo, Vacchi, Lorscheid, Stocchi, Gómez-Pujol, Harris, Casella, O’Leary and Hearty2016). Shore platform morphology in the northwestern Champlain Lowland and on the northern flank of Covey Hill may be confused with the eroded edges of gently dipping lower Paleozoic sedimentary strata, especially where beach deposits lap onto horizontal or gently dipping bedrock exposures, making them less accurate as water-level indicators.
The term “beach ridge” applies to relict, shore-parallel, wave- or swash-built, gravel and sand ridges, or erosional shoreface scarps (Otvos, Reference Otvos2000). Beach ridges commonly occur as clusters of low-relief, multiple, subparallel ridges in basin-sloping, progradational strandplains (Otvos, Reference Otvos2000; Tamura, Reference Tamura2012; Lindhorst and Schutter, Reference Lindhorst and Schutter2014; Kelsey, Reference Kelsey, Shennan, Long and Horton2015; Rovere et al., Reference Rovere, Raymo, Vacchi, Lorscheid, Stocchi, Gómez-Pujol, Harris, Casella, O’Leary and Hearty2016). Lindhorst and Schutter (Reference Lindhorst and Schutter2014) found that most beach ridges on King George Island, Antarctica, formed by storm waves, but strandline progradation primarily occurred by swash deposition during calmer intervals. Individual ridges form between mean water level and the highest storm-surge or spring-tide levels (Kelsey, Reference Kelsey, Shennan, Long and Horton2015). Successive offlapping sequences of ridges form as basinward regression disconnects earlier-formed beach deposits from active beach processes. Principal controls on shoreline regression in the study area include outlet incision, glacial isostatic adjustments, and relative sea-level change. Beach ridges are the principal focus of our investigation, because they are the most conspicuous, widespread, and traceable strandline features on surface elevation models in the study area (Fig. 3).
Figure 3. Examples of Champlain Estuary beach ridges illustrating ridge-crest segments and midpoint location in the Champlain Lowland near Altona, New York.
Materials and methods
DEM data
We downloaded bare-earth DEM data for New York, Vermont, Ontario, and Quebec from publicly available data resources (Table 2) with spatial (x,y) resolution of 2 m or less and vertical accuracy of <0.2 m. DEM data were merged into a single mosaic image and resampled at 1 m resolution using ArcMap 10.8.2 (Release 10, Environmental Systems Research Institute, Redlands, CA). The resampling did not improve the resolution of the 2 m data but provided a seamless base upon which to map. Derivative surface models such as hill shade, slope, and contour lines were generated using 3D Analyst raster tools in the ArcMap 10.8.2 ArcToolbox.
Table 2. Sources of digital elevation model (DEM) data.a
a Notes: USGS – United States Geological Survey; FEMA – Federal Emergency Management Agency; NYS – New York State; OMNRF – Ontario Ministry of Natural Resources and Forests; DAENR – Department of Agriculture, Energy and Natural Resources.
DEM resolution varied throughout the study area, so data were merged with 0.75 arcsec Canadian digital elevation models and 1 m DEM data from northwestern Massachusetts and resampled at 10 m spatial resolution to produce a seamless elevation model for the purpose of plotting summary maps of beach ridge interpretation. All DEM models use Universal Transverse Mercator (UTM) Zone 18N spatial coordinates and North American Datum 1983 (NAD83) vertical datum.
Individual ridges are subtle features that may be difficult to observe in the field due to dense vegetation and low relief, but they are easily identified on DEM images. We hand-digitized beach ridge crests as polyline features at a scale of 1:1000 or larger across >7800 km2 within the study area. Beach ridge examples are shown in Figure 3 for an area near Altona, New York. We also mapped beaches on the flanks of the Monteregian Hills in southern Quebec and in the eastern Champlain Lowland of Vermont, but mapping in these areas was limited to beaches near the upper marine limit (UML) of the Champlain Estuary.
Spatial coordinates and elevations were extracted from the endpoints and midpoint of each polyline segment using the ArcMap Data Management tools Add Surface Information and Add Geometry Attributes (Fig. 4). Ridge crest segments with >3 m relief along their length or >1 km in length were checked individually for continuity and accuracy. A subset of ridges, mostly in northeastern New York, was field verified.
Figure 4. Locations of beach ridges mapped in this study. Also shown is the line used to divide the data into St. Lawrence (western) and Champlain (eastern) subsets (UTM 594790 m E) and the lines of cross section for Figs. 4 and 5. Most beaches formed in the Champlain Sea (88.5%), followed by Lake Fort Ann (10.1%), Lake Iroquois (1.3%), and Lake Coveville (0.1%).
Only beach ridges mapped on DEM images within the study area (Fig. 1) were used in statistical analyses. Beach ridges mapped outside the study area as well as strandline data from Rayburn (Reference Rayburn2004) and Lewis et al. (Reference Lewis, Breckenridge and Teller2022) are presented for comparative purposes. Rayburn (Reference Rayburn2004) compiled strandline data from beaches and other features such as deltas, bars, terraces, and outlet elevations, for proglacial Lakes Iroquois, Coveville, and Fort Ann and the Champlain Sea in the Champlain and western St. Lawrence Lowlands (Table 1). Lewis et al. (Reference Lewis, Breckenridge and Teller2022) compiled spatial data for proglacial Lake Iroquois and the Champlain Sea. These authors selected data from previous investigations, including Rayburn (Reference Rayburn2004), collected by traditional surveying or differential global positioning methods to avoid elevation inaccuracies inherent with less-reliable aneroid barometer measurements and estimates made from topographic maps. Geographic coordinates of latitude and longitude reported by Lewis et al. (Reference Lewis, Breckenridge and Teller2022) were converted to UTM Zone 18N coordinates using the online National Geodetic Survey Coordinate Conversion and Transformation Tool (www.ngs.noaa.gov/NCAT/).
Lake Iroquois and Lake Abenaki are unique to the St. Lawrence and Champlain Lowlands, respectively, so beach ridge data, including previously published strandline data, were divided into St. Lawrence and Champlain subsets east and west of a north–south line that passes through Covey Hill at Easting 594790m (Fig. 4). The midpoint of each beach ridge segment was projected onto graphs of UTM Northing (abscissa) versus elevation (ordinate) along N–S elevation profiles: in the western St. Lawrence Lowland along Easting 522320m through Cornwall, Ontario, and another in the Champlain Lowland along Easting 626108m through Plattsburgh, New York (Fig. 4). The profile lines approximately bisect the St. Lawrence and Champlain datasets.
Previously reported isobase azimuths for proglacial lakes in the western St. Lawrence and Champlain Lowlands trend 90°–112° (Parent and Occhietti, Reference Parent and Occhietti1999; Rayburn, Reference Rayburn2004; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Lewis et al., Reference Lewis, Breckenridge and Teller2022), roughly perpendicular to the N–S elevation profiles. The vertical distribution of beaches facilitates correlation to known lake or marine phases (Fig. 4). All graphics and regression analyses for the profiles were performed using Microsoft Excel.
Isobase trend surfaces
A trend surface is a three-dimensional polynomial surface produced by least-squares regression that predicts the value of an observation, in this case beach ridge elevation, as a function of geographic coordinates (Davis, Reference Davis2002). We used beach data within ∼2 m of well-defined water planes from the N-S profiles. First- and second-order trend surfaces were fit to beach ridge data to assess the magnitude and direction of postglacial uplift. The analyses were conducted using trend surface tools in R Studio v. 4.2.2.
Results
Beach ridge morphology and distribution
The morphology and distribution of beach ridges are complex functions of atmospheric variables (wind speed, wind direction, and fetch); shoreline physiography (slope, aspect, and paleobathymetry); thickness, texture, and composition of surface materials; sediment supply; and rate of shoreline regression. Beach ridges in the study area occur individually or as offlapping groups of closely spaced, subparallel, curvilinear ridges. More than 19,500 beach ridge segments were identified in the study area, across the St. Lawrence and Champlain Lowlands in northeastern New York and adjacent Quebec south of the St. Lawrence River and UTM 18N Northing 5000000 m (Fig. 4, Supplementary Data). The region north of this latitude lies below the UML where, except for the flanks of the Monteregian Hills, the St. Lawrence Lowland is covered by marine or lacustrine mud or younger eolian dune sand.
Beach ridge segments range in length from a few tens of meters to ∼1 km (average ∼250 m) and in width from a few meters to ∼20 m. Relief along individual ridge crests averages about 1.5 m. Published field descriptions report that most ridges are composed of pebble to cobble gravel, sandy gravel, and gravelly sand (Woodworth, Reference Woodworth1905a, Reference Woodworth1905b; Fairchild, Reference Fairchild1919; Denny, Reference Denny1967, Reference Denny1970; LaSalle, Reference LaSalle1985). Occasionally, beach deposits occur as thin <1-m-thick boulder to cobble lag underlain by till (Pair et al., Reference Pair, Karrow, Clark and Gadd1988).
In map view, beach ridges occur in distinct bands separated by areas with few strandline features along the southern flank of the St. Lawrence Lowland and the western flank of the Champlain Lowland in New York (Fig. 4). A similar distribution is evident in N–S elevation profiles, which show beaches distributed in three groups of closely spaced points separated by vertical gaps with few points (Figs. 5 and 6). Beach ridge groups correlate with known elevations of proglacial lake and marine waterbodies in the published literature (Woodworth, Reference Woodworth1905a, Reference Woodworth1905b; Fairchild, Reference Fairchild1919; Chapman, Reference Chapman1937; MacClintock and Stewart, Reference MacClintock and Stewart1965; Wagner, Reference Wagner, Doolan and Stanley1972; Clark and Karrow, Reference Clark and Karrow1984; Muller and Prest, Reference Muller, Prest, Karrow and Calkin1985; LaSalle, Reference LaSalle1985; Parent and Occhietti, Reference Parent and Occhietti1988, Reference Parent and Occhietti1999; Pair et al., Reference Pair, Karrow, Clark and Gadd1988; Pair and Rodrigues, Reference Pair and Rodrigues1993; Rayburn, Reference Rayburn2004; Rayburn et al., Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011; Stanford, Reference Stanford2009; Bird and Kozlowski, Reference Bird and Kozlowski2016; Lewis et al., Reference Lewis, Breckenridge and Teller2022).
Figure 5. Plot of beach ridge midpoint elevations vs. UTM Northing for beach deposits and landforms in the St. Lawrence Valley. Beaches delineated in this study are represented by circles, triangles represent strandline data from Lewis et al. (Reference Lewis, Breckenridge and Teller2022) for the upper marine limit of the Champlain Estuary and Lake Iroquois, and squares represent upper and lower Lake Fort Ann (Akwesasne) strandlines from Rayburn (Reference Rayburn2004). Elevations of presumed outlet thresholds include Rome (Kozlowski, A., and Backhaus, K., personal communication, 2023), Little Falls (Bird and Kozlowski, Reference Bird and Kozlowski2016; Porecca et al., Reference Porecca, Briner and Kozlowski2018), The Gulf near Covey Hill (this study, from digital elevation model [DEM]), upper Lake Fort Ann (Rayburn, Reference Rayburn2004), lower Fort Ann (Rayburn, Reference Rayburn2004), and the Cove at Coveville (this study, from DEM).
Figure 6. Plot of beach ridge midpoint elevations vs. UTM Northing for beach deposits and landforms in the Champlain Lowland in New York. Beaches delineated in this study are represented by circles, triangles represent strandline data from Lewis et al. (Reference Lewis, Breckenridge and Teller2022) for the upper marine limit, and squares represent Lake Coveville and upper and lower Lake Fort Ann strandlines from Rayburn (Reference Rayburn2004). Elevations of presumed outlet thresholds are the same as (Fig. 5).
The highest beach ridges in the St. Lawrence Lowland, depicted by red triangles in Figures 4 and 5, correlate with proglacial Lakes Iroquois and Frontenac (Table 1). The distribution of beach ridge elevations within this group does not provide evidence for differentiation into separate lake phases (Fig. 5). The well-defined lower boundary slopes southward at 0.91 m/km and projects northward to the elevation of The Gulf bedrock threshold southwest of Covey Hill (Fig. 5). The highest beaches in the Champlain Lowland, depicted by purple triangles in Figures 4 and 6, correlate with Lake Abenaki (Lake Coveville of Chapman [Reference Chapman1937], Denny [Reference Denny1967, Reference Denny1970, Reference Denny1974], Rayburn [Reference Rayburn2004], Rayburn et al. [Reference Rayburn, Knuepfer and Franzi2005], and Franzi et al. [Reference Franzi, Rayburn, Knuepfer and Cronin2007], and the unstable phase of Lake Albany of Stanford [Reference Stanford2009]; Table 1). Lakes Iroquois and Abenaki were separated by the LIS where it straddled the St. Lawrence–Champlain drainage divide and blocked The Gulf outlet threshold (Franzi et al., Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016). The scarcity of beaches in Lakes Iroquois and Abenaki may reflect proximity to the ice sheet, which limited fetch, and the short time interval before ice recession exhumed The Gulf threshold and water levels in both lowlands dropped to the Lake Akwesasne level (Table 1).
The intermediate group of beach ridges, depicted by green squares in Figures 4–6, represents Lake Akwesasne, a lake phase that flooded the Ontario, St. Lawrence, and Champlain Lowlands. These beaches correlate with proglacial lakes variously named Vermont (Fort Ann Phase), Fort Ann, Candona, St. Lawrence, Sidney, Belleville, and Trenton from the published literature (Table 1). Water level fell rapidly, >70 m, during the Iroquois breakout to the Champlain Lowland, as evidenced by a lack of preserved beaches between the lowest Lake Iroquois and uppermost Lake Akwesasne beaches (Fig. 5). In the Champlain Lowland, beaches span the transition between Lake Abenaki and Lake Akwesasne (Fig. 6, Table 1), indicating a continuous, steady water-level decline unaffected by inflow of water from Lake Iroquois.
Gradual incision through newly emergent glacial and lacustrine deposits or bedrock on the Hudson–Champlain drainage divide south of Fort Ann, New York, and differential isostatic rebound may account for the steady water-level decline for both Lakes Abenaki and Akwesasne in the Champlain Lowland (Fig. 1, Table 1). The lower boundary of Lake Akwesasne projects southward between 0.89 and 0.92 m/km to the village of Fort Ann in both the St. Lawrence and Champlain Lowland profiles (Figs. 1, 5, and 6). This boundary corresponds to a relatively stable phase of Lake Akwesasne (Lake Fort Ann of Stanford [Reference Stanford2009] and Rayburn et al. [Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011]), when outlet incision slowed where it exhumed a bedrock threshold at Fort Ann (Fig. 1, Table 1).
The lowest set of beaches, depicted by blue circles in Figures 4–6, correlates with the Champlain Estuary in the St. Lawrence and Champlain Lowlands. The well-defined UML slopes southward between 0.89 and 0.95 m/km in the St. Lawrence and Champlain Lowlands (Figs. 5 and 6, Table 1). Water level fell rapidly, ∼45 m, when Lake Akwesasne drained to the UML, as evidenced by the gap in beach preservation (Figs. 1, 5, and 6) between the lowest Akwesasne beaches and the UML. The UML projects northward in the St. Lawrence Lowland to the highest beaches on Mt. Rigaud in Quebec (Figs. 1 and 5) and projects above the highest beaches on Mt. St. Hilaire, Mount Yamaska, and Rougemont in the Monteregian Hills north of the Champlain Lowland (Figs. 1 and 6).
The abundance and close spacing of beaches below the UML (Figs. 5 and 6) record the forced regression of marine water in the St. Lawrence Lowland caused by the net effects of differential postglacial isostatic uplift and relative sea-level change. The Frontenac Arch (Figs. 1 and 2) emerged above sea level ∼12.6 cal ka BP, and endorheic Lake Ontario became isolated in the Ontario Lowland (Lewis, Reference Lewis2016; Lewis and Todd, Reference Lewis and Todd2019) from the Hochelagan Estuary in the St. Lawrence Lowland (Parent and Occhietti, Reference Parent and Occhietti1988; Lewis and Todd, Reference Lewis and Todd2019; Fig. 2E).
Regional isobases and glacial isostatic uplift
First- and second-order trend surfaces were fit to three water levels in N–S plots of beach midpoint elevations (Fig. 7). These levels represent Lake Iroquois, when outflow stabilized across The Gulf bedrock threshold near Covey Hill; the lowest phase of Lake Akwesasne, when the bedrock threshold near Fort Ann controlled outflow; and the UML of the Champlain Estuary. First-order trend surfaces adequately define the trends as indicated by their adjusted r 2 and the low values of the second-order regression coefficients (Table 3). Second-order trend surfaces nearly matched first-order trends and are not further described.
Figure 7. Isobases derived from first-order trend surfaces of beaches on (A) Lake Iroquois (The Gulf threshold), (B) Akwesasne (Fort Ann threshold), and (C) the upper marine limit of the Champlain Estuary in the St. Lawrence and Champlain Valleys.
Table 3. Regression coefficients and goodness-of-fit statistics for trend surfaces on stable Lake Iroquois (The Gulf threshold near Covey Hill), lower Lake Akwesasne (Fort Ann threshold), and Champlain Sea (upper marine limit), in the Ontario, western St. Lawrence, and Champlain Lowlands.
Isobases derived from the first-order trend surface for Lake Iroquois strike ESE (∼119°) with 0.70 m/km gradient to the SSW (Fig. 7A). This trend differs >25° clockwise from Lake Akwesasne and UML isobases and may reflect regional or temporal patterns of postglacial rebound or limitations associated with a small sample size (n = 10) and limited spatial distribution of Lake Iroquois beaches. The Lake Iroquois isobase gradient is also lower than the gradient estimated from the N–S elevation profile for the western St. Lawrence Lowland (0.91 m/km; Fig. 5).
Isobases derived from the first-order trend surfaces for Lake Akwesasne (LA) and UML water levels are nearly identical in strike azimuth (LA = 90.0°; UML = 88.0°), with a gradient of ∼0.92 m/km to the south (Fig. 7B and C, Table 3). These surfaces fit the observed data well (Table 3), reflecting large sample populations (LA = 196; UML n = 348) and widespread spatial distribution of beaches. The results compare well with previous studies of isostatic uplift in the region (Koteff and Larsen, Reference Koteff and Larsen1989; Parent and Occhietti, Reference Parent and Occhietti1999; Rayburn, Reference Rayburn2004; Rayburn et al., Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011; Bird and Kozlowski, Reference Bird and Kozlowski2016; Hooke and Ridge, Reference Hooke and Ridge2016; Lewis et al., Reference Lewis, Breckenridge and Teller2022). The similarity of isobase azimuth and slope for the Akwesasne and UML surfaces probably reflects a short duration for drainage to occur and serves as a quality indicator for the beach ridge datasets.
Discussion
Beach ridges in the study area are most abundant in the northwestern Champlain Lowland and the southern flank of the St. Lawrence Lowland near Covey Hill (Fig. 4). These areas would have been exposed to the greatest fetch from prevailing northeasterly winds across the lake and marine water surfaces, generating high wave energy consistent with anticyclonic atmospheric conditions in the late Pleistocene and Early Holocene. Similar conditions existed on the flanks of the Monteregian Hills. Postglacial dunes found in the St. Lawrence Lowland northeast of Montreal, Quebec (Filion, Reference Filion1987; David, Reference David and Gadd1988) and northern New York (MacClintock and Stewart, Reference MacClintock and Stewart1965; Cadwell and Pair, Reference Cadwell and Pair1991; Carlisle et al., Reference Carlisle2015) probably formed under similar wind conditions soon after the Hochelagan Estuary regressed (Filion, Reference Filion1987; David, Reference David and Gadd1988; Borracci et al., Reference Borracci, Carl, Gontz, Pantoja-Flores and Penfield2023; Carl et al., Reference Carl, Gontz, Borracci, Penfield, Pantoja-Flores and Franzi2023; Table 1). The lack of widespread beaches in Vermont and the southern Champlain Valley reflects the wind shadow created by the high relief of the Green Mountain Upland to the east and narrowing of the Champlain Lowland to the south.
Beach ridges in the St. Lawrence and Champlain Lowlands occur in distinct groups separated spatially and vertically by areas with few strandline features (Figs. 4–6). Closely spaced points within a group represent times of steady, gradual net shoreline regression. Preservation of offlapping sequences of beach deposits occurred when shoreline regression disconnected older beach deposits from the upper limit of active shoreline processes. Abandonment at any given location probably occurred over a period of decades, given the short time frame for deglaciation in the region (Rayburn et al., Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011). Gaps between groups represent times of rapid water-level fall during breakout floods when there was insufficient time for beach ridges to form. The rate of water-level decline decelerated during the waning stages of a breakout as water levels in the source and receiving reservoirs equilibrated, allowing sufficient time for beach ridges to form.
No evidence for shoreline transgression, as might occur if ice readvanced over a lake outlet or sea-level rose, was found. Strandline deposits probably formed and reformed repeatedly along the shore during periods of steady water-level decline, preserving only regressive deposits as isostatic-rebound raised beaches above the level of active shorezone processes.
Proposed proglacial lake and marine nomenclature
Beach ridges record steady, gradual shoreline regression controlled by outlet incision and the net effects of differential isostatic uplift and postglacial relative sea-level change. Spatial analyses of beach deposits and landforms interpreted from DEM data demonstrate continuous progression of steadily falling water levels was interrupted by rapid water-level drops caused by breakout floods from Lake Iroquois, across the St. Lawrence–Champlain divide in northeastern New York (Figs. 8 and 9), and Lake Akwesasne in the Lake Chaudière–Etchemin area, Quebec (Occhietti et al., Reference Occhietti, Parent, Lajeunesse, Robert, Govare, Ehlers, Gibbard and Hughes2011). The revised nomenclature for the proglacial lake sequences presented here reflects new data and interpretations of strandline data (Fig. 10).
Figure 8. Surficial features of the Covey Hill area related to Lake Iroquois outflow in the Champlain Lowland and the shorelines of lakes Abenaki, Iroquois, and Akwesasne and the upper marine limit of the Champlain Estuary. The map is adapted from LaSalle (Reference LaSalle1985), Denny (Reference Denny1970, Reference Denny1974), and Trevail (Reference Trevail2006). The Altona Flat Rock sandstone pavement is labeled AFR.
Figure 9. East–west topographic profile between the St. Lawrence River and Lake Champlain through The Gulf threshold near Covey Hill showing late Pleistocene proglacial lake and marine water levels. The transect follows UTM Northing 4984486 m N (UTM Z18N). Vertical exaggeration ∼110×.
Figure 10. New proglacial lake nomenclature for the eastern Ontario, western St. Lawrence, and Hudson–Mohawk–Champlain Lowlands. Abbreviations: ML, Mohawk Lowland; EI, Erie Interstade; SBD, Shed Brook Discontinuity; G–M, Glaciomohawk; I–M Iromohawk River. Principal sources: Muller and Prest (Reference Muller, Prest, Karrow and Calkin1985), Muller and Calkin (Reference Muller and Calkin1993), Parent and Occhietti (Reference Parent and Occhietti1988), Occhietti et al. (Reference Occhietti, Parent, Lajeunesse, Robert, Govare, Ehlers, Gibbard and Hughes2011), Ridge et al. (Reference Ridge, Franzi and Muller1991, Reference Ridge, Balco, Bayless, Beck, Carter, Dean, Voytek and Wei2012), Ridge (Reference Ridge1997, Reference Ridge2003, Reference Ridge, Ehlers, Gibbard and Part2004), Ridge and Franzi (Reference Ridge, Franzi and April1992), Pair and Rodrigues (Reference Pair and Rodrigues1993), Stanford (Reference Stanford2009), Franzi et al. (Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016), Lewis (Reference Lewis2016), Lewis and Todd (Reference Lewis and Todd2019), and Lewis et al. (Reference Lewis, Breckenridge and Teller2022).
Lake Albany
We retain the name Lake Albany to include the stable and unstable lake phases largely confined to the Hudson Lowland (Stanford, Reference Stanford2009). The extent to which Lake Albany may have expanded into the Champlain Lowland remains uncertain. Fairchild (Reference Fairchild1919) considered the shoreline features presently assigned to Lakes Abenaki and Akwesasne to be part of an estuarine system open to the Atlantic Ocean in the lower Hudson Lowland. Stanford (Reference Stanford2009) suggested that Lake Albany expanded northward to Altona, New York (Fig. 1) in the Champlain Lowland. However, Wall (Reference Wall, DeSimone, Wall, Miller, Rayburn, Kozlowski, Dineen and Hanson2008) determined that Lake Albany drained to near the present grade of the Hudson River at the time the Iromohawk River began to cut the gorge below Cohoes Falls (Figs. 1 and 2). Proglacial lakes confined to the Champlain Lowland and controlled by outflow across the Hudson–Champlain divide must predate the Iroquois breakout by at least the time required to cut the Cohoes gorge. DeSimone and LaFleur (Reference DeSimone, LaFleur and Selleck2008) and DeSimone et al. (Reference DeSimone, Wall, Miller, Rayburn, Kozlowski, Dineen and Hanson2008) indicated that the ice margin may have stood at the Forest Dale delta in Vermont when Lake Albany fell sufficiently to expose older glacial and lacustrine deposits, or bedrock on the Hudson–Champlain divide near Fort Ann (Fig. 1), separating lakes in the upper Hudson and Champlain Lowlands. We consider it unlikely that Lake Albany expanded northward beyond the southern Champlain Lowland.
Lake Iroquois
Lake Iroquois drained through the Mohawk Lowland via the Iromohawk River (Fig. 2B) until ∼13 cal ka BP (Lewis and Todd, Reference Lewis and Todd2019; Lewis and Anderson, Reference Lewis and Anderson2020; Lewis et al., Reference Lewis, Breckenridge and Teller2022), when lower outlets emerged on the St. Lawrence–Champlain drainage divide near The Gulf, about 2.5 km southwest of Covey Hill, Quebec (Figs. 8 and 9). The highest bedrock channels on the northeastern flank of the Champlain Lowland (Fairchild, Reference Fairchild1919; Denny, Reference Denny1974) may have served as ephemeral outlets during the initial stages (Fig. 8). Outflow discharged southeastward along the ice margin to Lake Abenaki in the Champlain Lowland across deglaciated upland slopes southeast of Altona. Rapid discharge scoured surficial cover and created the Altona Flat Rock (AFR in Fig. 8) sandstone pavement (Woodworth, Reference Woodworth1905a, Reference Woodworth1905b; Denny, Reference Denny1974; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007).
Beach ridge formation kept pace with water-level declines in the Champlain Lowland (Fig. 6), despite widespread geomorphic and sedimentological evidence for a high volume of water discharge during initial outflow of Lake Iroquois to the Champlain Lowland (Woodworth, Reference Woodworth1905a, Reference Woodworth1905b; Chapman, Reference Chapman1937; Denny, Reference Denny1974; Franzi et al., Reference Franzi, Rayburn, Yansa and Knuepfer2002, Reference Franzi, Rayburn, Knuepfer and Cronin2007, Reference Franzi, Ridge, Pair, Desimone, Rayburn and Barclay2016; Rayburn, Reference Rayburn2004; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005, Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011). Sandstone pavements between Covey Hill and Altona were mostly ice-covered during the initial flooding (Denny, Reference Denny1974). Lake Iroquois beaches in the western St. Lawrence Lowland provide evidence for steady water-level decline during the initial event, with no distinction between a Lake Iroquois “Main Phase” and “Frontenac Phase” proposed by Pair and Rodrigues (Reference Pair and Rodrigues1993; Fig. 5, Table 1). Steady water-level decline further indicates that the transition from the southern outlets of Lake Iroquois in the western Mohawk Lowland to northern outlets near Covey Hill did not involve an abrupt change in water level in the St. Lawrence Lowland as previously proposed (Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007).
Lake Iroquois stabilized for several decades (Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007; Rayburn et al., Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011) at The Gulf bedrock threshold until further ice recession on the north slope of Covey Hill caused catastrophic breakout of the remaining impounded water to the Champlain Lowland. Outflow around the northern flank of Covey Hill created large sandstone pavements across nearly flat-lying sedimentary rocks north and southeast of Covey Hill (Denny, Reference Denny1974; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Franzi et al., Reference Franzi, Rayburn, Knuepfer and Cronin2007; Fig. 8). Proglacial lake level fell >70 m during the event (Figs. 5 and 9). The absence of beaches between the lowest Lake Iroquois and highest Lake Akwesasne level in the western St. Lawrence Lowland (Fig. 5) indicates a rapid drop in water level. As water levels equilibrated, beaches began forming along the shore of Lake Akwesasne, a lake that spanned the Ontario, western St. Lawrence, and Champlain Lowlands.
Lake Abenaki
Proglacial Lake Abenaki (Abenaki: “dawn-land people”; Snow, Reference Snow2012) formed in the Champlain Lowland after the Hudson–Champlain divide emerged near Coveville, New York (Figs. 1, 2B, and 10). The lake is named for the indigenous Algonquin-speaking people who lived in southern Quebec and northern New England. The name Abenaki replaces the names Lake Vermont (Coveville Stage), Lake Coveville, and the northern portion of the unstable phase of Lake Albany (ULA) (Stanford, Reference Stanford2009; Table 1) in the Champlain Lowlands. Lake Abenaki differs from Lake Vermont (Woodworth, Reference Woodworth1905a; Chapman, Reference Chapman1937) in that the lake was confined to the Champlain Lowland for its entire existence and does not include the Fort Ann Stage of Woodworth (Reference Woodworth1905a) and Chapman (Reference Chapman1937), which extended into the western St. Lawrence Lowland, nor does it include unstable Lake Albany phases that may have straddled the Hudson–Champlain divide in the southern Champlain Lowland (Stanford, Reference Stanford2009). Beaches formed on the shore of Lake Abenaki in the Champlain Lowland were previously attributed to Lake Coveville (Woodworth, Reference Woodworth1905a, Reference Woodworth1905b; Chapman, Reference Chapman1937; Denny, Reference Denny1967, Reference Denny1970, Reference Denny1974; Franzi et al., Reference Franzi, Rayburn, Yansa and Knuepfer2002, Reference Franzi, Rayburn, Knuepfer and Cronin2007; Rayburn, Reference Rayburn2004; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005).
Lake Abenaki drained southward across emergent glacial and lacustrine deposits or bedrock on the Hudson–Champlain divide. The close spacing of Abenaki beaches (Fig. 6) indicates shorelines adjusted continuously in response to outlet incision and differential isostatic rebound. The lake expanded northward ∼105 km to the latitude of Altona, New York (Figs. 1 and 8), at which point the lake received discharge from Lake Iroquois. Bouldery inwash deposits on Cobblestone Hill, located at the SE margin of Altona Flat Rock, accumulated where the Iroquois floodwater entered Lake Abenaki (Franzi et al., Reference Franzi, Rayburn, Yansa and Knuepfer2002, Reference Franzi, Rayburn, Knuepfer and Cronin2007; Rayburn, Reference Rayburn2004; Rayburn et al., Reference Rayburn, Knuepfer and Franzi2005; Fig. 8). The inwash deposits range in present-day elevation between 230 and 205 m, indicating that Lake Abenaki water level fell to near the level of Lake Akwesasne while the ice margin stood at Cobblestone Hill. Lake Abenaki ended when ice receded from the northern flank of Covey Hill and water in the Ontario, western St. Lawrence, and Champlain Lowlands merged to form Lake Akwesasne (Figs. 2C and 8–10).
Franzi et al. (Reference Franzi, Rayburn, Knuepfer and Cronin2007) reported ice recession rates in the northern Champlain Lowland range between 0.19 and 0.44 km/yr (mean = 0.32 km/yr). Connally and Sirkin (Reference Connally, Sirkin, Black, Goldthwait and Willman1973), Ridge et al. (Reference Ridge, Balco, Bayless, Beck, Carter, Dean, Voytek and Wei2012), and Wright et al. (Reference Wright, Dunn, Springston and Grigg2024) similarly report recession rates of ∼0.3 km/yr in the northern Champlain Lowland, Lake Hitchcock in the Connecticut Valley, and Lake Winooski in western Vermont, respectively. We infer that Lake Abenaki existed in the Champlain Lowland for ∼340 yr, assuming ice recession of 105 km at an average rate of 0.31 km/yr.
Lake Akwesasne
Proglacial Lake Akwesasne (“land where the grouse drum”; Klemp, Reference Klemp2022) occupied the Ontario, western St. Lawrence, and Champlain Lowlands following the Lake Iroquois breakout. The lake is named after the St. Regis Mohawk tribal land in northern New York, southeastern Ontario, and southwestern Quebec that straddles the U.S.–Canada international border. Lake Akwesasne replaces Lake St. Lawrence, Lake Candona, the Frontenac, Sydney, Belleville, and Trenton phases of Lake Iroquois, and Lake Fort Ann (Fig. 10, Table 1).
Lake Akwesasne drained southward across emergent glacial and lacustrine deposits, and bedrock near the present Hudson–Champlain drainage divide near Fort Ann, New York. Lake Akwesasne water level fell gradually ∼21–24 m before stabilizing briefly when the bedrock threshold near Fort Ann was exhumed. The lake existed for at least 216 yr (Rayburn et al., Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011), at which point ice recession in the Lake Chaudière–Etchemin area, Quebec (Occhietti et al., Reference Occhietti, Parent, Lajeunesse, Robert, Govare, Ehlers, Gibbard and Hughes2011) opened a connection to the Goldthwait Sea in the Gulf of St. Lawrence and the level of Lake Akwesasne dropped abruptly ∼45 m to the UML of the Champlain Estuary (Figs. 5, 6, 9, and 10).
Champlain and Hochelagan Estuaries
The Champlain Estuary inundated the Ontario, western St. Lawrence, and Champlain Lowlands and opened northeastward to the Atlantic Ocean via the Goldthwait Sea (Chapman, Reference Chapman1937; MacClintock and Stewart, Reference MacClintock and Stewart1965; Muller and Prest Reference Muller, Prest, Karrow and Calkin1985; Occhietti et al., Reference Occhietti, Parent, Lajeunesse, Robert, Govare, Ehlers, Gibbard and Hughes2011; Figs. 2D, 9, and 10, Table 1). The highest beaches formed at the UML during the latter stages of the Akwesasne breakout. Widespread beach development in the western St. Lawrence and Champlain Lowlands below the UML records the forced regression of marine water. The Champlain Estuary phase ended when shoaling from isostatic rebound and relative sea-level change raised the Frontenac Arch and separated endorheic Lake Ontario (an internally drained lake phase) from the Hochelagan Estuary (Figs. 2E and 10).
Continued shoaling from isostatic rebound and freshwater inflow progressively decreased salinity in Hochelagan Estuary, freshening the western portion, previously known as Lake Lampsilis (Elson and Elson, Reference Elson and Elson1959; Elson, Reference Elson and Clark1962, Reference Elson and Gadd1988; Richard, Reference Richard1978; Parent and Occhietti, Reference Parent and Occhietti1988; Table 1). Lake Lampsilis was named for a genus of freshwater mussels found in sediments overlying marine deposits in the western St. Lawrence Lowland.
The Hochelagan Estuary is named for the Iroquois village Hochelaga, visited by Samuel de Champlain in 1535, at the confluence of the Ottawa and St. Lawrence Rivers near Montreal. The name is derived from the Iroquoian terms “beaver path” or “big rapids” (Gagné, Reference Gagné2013). Use of the name dates to Woodworth (Reference Woodworth1905a, p. 220), who proposed the name “Hochelagan formation” for the deposits of the marine incursion and the subepoch or stage during which they were deposited. Its use here is restricted to the estuarine conditions that were confined to the St. Lawrence and Champlain Lowlands following the separation of endorheic Lake Ontario and led to deposition of freshwater sediments in the Montreal area (Fig. 2E). The Hochelagan Estuary phase ended ∼8.3 cal ka BP (Lewis and Todd, Reference Lewis and Todd2019) with the emergence of the western St. Lawrence Lowland and hydrologic reconnection to the Ontario Basin, establishing the modern drainage system (Fig. 10). Present-day Lake Champlain formed as postglacial isostatic rebound raised ground above sea level near St. Jean, forming a threshold that separated Lake Champlain from the Hochelagan Estuary.
Conclusion
The deglacial paradigm presented here underpins a new nomenclature for proglacial lake and marine succession in the Ontario, St. Lawrence, and Champlain Lowlands (Fig. 10). The new chronology and nomenclature clarify the spatial and temporal extents of deglacial waterbodies in the region. Shoreline regression in late Pleistocene to Early Holocene proglacial lakes and the Champlain and Hochelagan estuaries in Ontario, St. Lawrence, and Champlain Lowlands created offlapping strandline deposits and landforms. Beach ridges record steady, gradual shoreline regression interrupted by breakout floods from lakes Iroquois and Akwesasne. The names for proglacial lakes Iroquois and Albany are retained but redefined to conform to new data presented in this study. The names Lake Abenaki and Lake Akwesasne replace multiple names given to proglacial lakes in the Ontario, St. Lawrence, and Champlain Lowlands (Table 1). Contiguous Early Lake Ontario and Champlain Sea and later contiguous Lake Lampsilis and Champlain Sea are assigned the names Champlain Estuary and Hochelagan Estuary. The Lake Ontario Lowstand (Lewis and Todd, Reference Lewis and Todd2019) is renamed as endorheic Lake Ontario to reflect the internally drained basin the lake occupied (Fig. 10). The new proglacial lake designations provide a simplified, uniform nomenclature that considers abundant new beach ridge data, eliminates multiple and overlapping names by removing local naming conventions, and recognizes the indigenous peoples whose ancestral lands once lay beneath these glacial waters.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/qua.2025.15.
Acknowledgments
We thank Charles Gordon, a SUNY Plattsburgh graduate, who compiled much of the spatial database in the Champlain Lowland as part of his undergraduate research project. Our colleagues Byron Stone, Michel Lamothe, Scott Stanford, Michael C.F.M. Lewis, David Barclay, David DeSimone, John Rayburn, Andy Kozlowski, Karl Backhaus, Don Pair, and Phil Wagner generously shared their insights and information during the preparation of this article. Jack Ridge and Stephen Wright read an early draft of this report and provided comments that greatly improved it. Ed Romanowicz and Robert Reams provided advice on the trend surface analyses and Eileen Allen and Mark Baran helped with GIS analyses. We are also grateful for the comments provided by Tom Lowell, Jason Briner, and an anonymous reviewer who provided peer review for the journal. We appreciate the time and effort they dedicated to improving this article. We extend special thanks to Darren Bonaparte, director of the Tribal Historic Preservation Office for the St. Regis Mohawk Tribe, and Joanne Crawford, interim chief of the Abenaki Nation of the Mississquoi, for their advice on the use of indigenous names for the proglacial lakes proposed in this study.
Data availability statement
The beach data are available in the Supplementary Data File. GIS shapefiles are available upon request from the corresponding author.
Funding statement
This research benefited from support from SUNY Plattsburgh and SUNY Potsdam.
Competing interests statement
The authors declare that they have no known competing financial interests or personal relationships that influenced the research presented here.
Open access
