M.S. Thesis

This study will evaluate mechanisms of sinkhole formation, the hydrologic function that they serve, and the extent to which they are developed in the Karst Plain in the Bellevue-Castalia region of north-central Ohio.

This study falls within the realm of a Unified Source Water Protection Project being conducted in the Bellevue-Castalia Karst Plain in north-central Ohio. The carbonate rock aquifer that underlies the study area is highly susceptible to contamination from surface and near surface sources. Progressive karstification of the aquifer in the form of dolines or sinkholes provides a direct connection between surface and groundwater, greatly increasing the number of Potential Contaminant Sources in the region. Therefore, it is essential that the evolution, spatial distribution, and aerial extent of possible sinkhole development be fully understood, so that these areas might be included in a Source Water Protection Area.

Location

Preliminary boundaries of the study area will be based on Kihn’s (1988) study of the karst plain in the Bellevue-Castalia Region of north-central Ohio in portions of Seneca, Sandusky, Erie and Huron Counties (Figure 1). This area encloses just under 625 km2 with the only definite boundary being the southern shore of Sandusky Bay on Lake Erie to the north. The eastern, western, and southern boundaries will be delineated based on the presence or non-existence of karst features, namely sinkholes, throughout the region.

Figure 1. Location of the study area in north-central, Ohio (derived from Physiographic Regions of Ohio by C. Scott Brockman, 1998).

Physiography

Except for small portions of Erie, Huron, and Seneca counties, a majority of the study area lies in the Midwestern Basin and Arches Region (Shaver, 1985; Bugliosi, 1999) and is entirely within the Central Lowlands Physiographic Province (Burgess and Niple, 1967; Kihn, 1988). The area so defined is, in turn, part of two smaller physiographic sections; the Till Plains Section and the Eastern Lake Section that are included in the St. Lawrence River Basin (Shindel et al., 2001) (Figure 1). The entire karst plain has been subjected to Early, Middle, and Late stages of Wisconsinan age glaciation and possibly more episodes of glaciation during the Pleistocene Epoch, but evidence of these deposits has been removed by latest Wisconsinan ice sheets (Bugliosi, 1999). Surficial deposits left by the Laurentide ice sheet include end moraines, ground moraine (till), outwash deposits, and glacio-lacustrine sediments (Leverett, 1902).

The Till Plains section comprises the lower third of the study area and is a gently undulating region with deposits consisting of soft calcareous clays interbedded with small bedrock fragments that grade downward into a blue-gray clay containing pebbles (Shaffer, 1951). The average thickness of the till is from 0 to less than 25 ft. near Lake Erie and from 20 to 75 ft. farther southward from the lake (Stout, 1941). Overlying the till in the northern two-thirds of the study area are deposits of glacio-lacustrine clays that comprise the Eastern Lake Section in the region (Kihn, 1988). The topography is relatively flat and is separated from the Till Plains Section by a low beach ridge formed by glacial Lake Maumee that occupied this area as ice retreated out of Ohio and dammed northward drainage into the basin (Leverett, 1902). As ice continued to retreat to the north, several other beach ridges were formed and eventually the water that formed these ridges occupied Lake Erie (Forsyth, 1959). Glacio-lacustrine deposits are generally 20-30 ft. thick and the land surface gently rises 8-10 ft. /mi. moving southward away from the lake.

Stratigraphy

The bedrock underlying the study area is composed of Middle to Upper Silurian and Middle to Late Devonian limestones, dolomites, and shale (Figure 2). This sequence of rocks was referred to by drillers as the “Big Lime” and is the main water-bearing unit in the region (Owens, 1970). A great unconformity exists between the two systems and is marked by the erosion or non-deposition of Upper Silurian and Lower Devonian formations in the area (Stauffer, 1909). At nearly 700 feet thick, this entire composite of carbonates has been interpreted as being deposited in a shallow, relatively quiet, lagoon-type basin, east of the Findlay Arch (Kihn, 1988).

Middle to Late Silurian deposits related to the study area begin with the Rochester Formation which is comprised mostly of shale and serves as the basal confining unit for the overlying carbonate aquifer (Drane, 1993). This is followed by the Lockport Group which is left undifferentiated in the study area and consists of massive, somewhat porous, fine-grained dolomite (Chaffee, 1995). Both of these units are Niagaran Series deposits, although these units do not crop out in the study area. Overlying the Lockport is the Bass Islands Group, consisting of, in ascending order, the Greenfield, Tymochtee, Put-in-Bay, and Raisin River Formations. These units are part of the Cayugan Series, Salina Group and are primarily dolomites of varying thickness containing appreciable amounts of gypsum and other evaporites (Kihn, 1988). The Greenfield dolomite is in part a reef deposit, and its contacts with the underlying Lockport and overlying Tymochtee Formation range from disconformable to gradational (Norris and Fiedler, 1971). The Tymochtee Formation consists of a thinly bedded dolomite that is more shaly than the other formations of the Bass Islands Group (Carman, 1927).

The overlying Put-in-Bay and Raisin River Formations are found to be dolomites of fair purity. The Put-in-Bay Formation is drab to gray to light brown in color with an open crystalline texture and is thin to massively bedded, while the Raisin River Formation is bluish gray to brownish gray with regular bedding 2-6 inches in thickness (Stout et al., 1943). Recent studies have left these two formations undifferentiated and grouped into the Bass Islands Dolomite due to difficulties in separating the two in the subsurface (Chaffee, 1995).

The overlying formations consist of Middle Devonian Ulsterian and Erian Series dolomites and limestones and rest unconformably on the Bass Islands Dolomite. All Devonian rocks in the region are interpreted as being marine in origin with the lowest units being the Amherstburg and Lucas Formations (Stout et al., 1943). These two formations are assigned to the Detroit River Group and are unconformably overlain by the Columbus Limestone that occupies the stratigraphically highest position in this group. This is followed by the Erian Delaware Limestone and the Senecan Olentangy Shale respectively. The Olentangy Shale has been subdivided in the region into the Plum Brook Shale and the overlying Prout Limestone (Hull, 1990).

According to Stout et al. (1943), the Amherstburg is a true dolomite with a slight drab to brownish gray appearance that is massively bedded and has an open to cavernous texture due to its susceptibility to acid bearing waters and the secondary solution of gypsum that may locally produce sinkholes and underground passages leading to springs (Stout, 1941). The overlying Lucas Formation is a thinly bedded dolomite but locally can be somewhat calcareous, and depending on the degree of oxidation of the iron minerals, can vary from a light blue to a drab color (Stout et al., 1943).

The top of the Detroit River Group, the Columbus Limestone, is considered a limestone of remarkable purity in Ohio, containing 95% CaCO3 with only tiny amounts of impurities (Forsyth, 1988). This formation varies from a dolomite at the base, upwards, to a limestone at the top and ranges in color from light gray to brown (Stout, 1941). The formation is massively bedded and contains several prominent layers of chert nodules (Kihn, 1988). As a result of the purity of the Columbus Limestone, all solution formed caves in Ohio are expressed in this formation as well as a majority of the sinkholes and sinking streams in the Bellevue-Castalia region (Forsyth, 1988). Disconformably overlying the Columbus Formation is the Delaware Limestone. This unit is light to very dark bluish gray and is primarily a limestone with some interbedded shale layers (Stout, 1941). The Delaware Limestone is harder, denser, and lacks the purity of the underlying Columbus Limestone, owing to its resistance to karst forming processes (Drane, 1993). At the eastern edge of the karst plain, the Olentangy Shale disconformably overlies the Delaware Limestone (Hume, 1991), and acts as a confining unit for the Silurian-Devonian “Big Lime” carbonate sequence (Owens, 1970). In north-central Ohio, this formation has been separated into the Plum Brook Shale and the overlying Prout Limestone (Hume, 1991).

Structural Geology

The Bellevue-Castalia Karst Plain is located in the transitional region between the central Appalachian foreland basin and the interior cratonic basins of Michigan and Illinois on the eastern limb of the Cincinnati Arch in north-central Ohio (Figure 3). Locally, the structure is referred to as the Findlay Arch and stretches from the Nashville Dome in Tennessee north-northeast through Ohio to its terminus in southern Ontario where the arch becomes unrecognizable in the subsurface (Norris and Fiedler, 1971). In west-central Ohio the arch branches into a northwest trending structure called the Kankakee Arch that terminates in Wisconsin and separates the Illinois and Michigan Basins. Another major basement arch, trending north-south, exists just to the east of the karst plain and has been termed the Waverly Arch (Woodward, 1961). Although it has no surface expression, it was a prominent geographic feature in the Ordovician and Silurian Eras (Root and Onasch, 1999). Terminal closing of the Laurentian continent with the North African continent during the twilight of the Paleozoic produced the Cincinnati-Findlay arch as a result of structural loading and Appalachian basin subsidence (Root & Onasch, 1999). Bedrock units dip south eastward into the basin at 30 feet per mile and are heavily jointed and fractured (Kihn, 1988).

Hydrogeology

The area under investigation is characterized by a regional ground water flow path in a general south to north direction towards Sandusky Bay. Discharge in the region occurs through a series of springs located in the northern portion of the study area (Kihn, 1988; Forster, 1997). The aquifer in this portion of the study area is under confined conditions (confined by glaciolacustrine clays) and the potentiometric surface rises above the land surface producing a region of flowing artesian wells and springs (Forster, 1997). Ten springs have been located in the study area and classified as fracture, contact, sinkhole, and subaqueous type springs (Kihn, 1988). Springs range in size from 20-500 feet in diameter and discharge anywhere from a trace to over 6,700 gpm (Kihn, 1988; Forster, 1997). The largest of these springs, in terms of discharge, occurs at the contact between the Lucas and Columbus Formations, due to the high contrast in permeability and porosity between the two formations (Kihn, 1988). Also, several subaqueous springs have been suggested to exist on the floor of Sandusky Bay (Norrocky, 1987) and are recognized in the winter by unfrozen zones of upwelling water above the conduits and in the early spring when clear pools of water are visible in the sediment laden lake waters (Kihn, 1988). The depth of groundwater flow in the Bellevue-Castalia region is uncertain. Groundwater flow modeling has shown that the saturated thickness of the carbonate rock aquifer ranges from 260 to 600 feet (Hanover, 1994). Elevated Sulfur34 (69.1 per mil) from wells near Lake Erie indicate relatively long residence times and suggest that flow-paths may be vertical and associated with discharge from deeper portions of the aquifer system (Eberts and George, 2000). This is also indicated by the springs’ constant yearly discharge, temperature, specific conductance, total dissolved solids (SO4), hardness, and the lack of turbidity following storm events (Kihn, 1988). This indicates a regional groundwater flow-path that is diffuse in nature, although conduit flow does exist in portions of the study area (Drane, 1993). However, the overall groundwater gradient may be difficult to reconcile with this.

Photo of The Castalia Blue Hole. This is a spring that formed in the 1860's following the construction of a dam and a grist mill along Cold Creek. Photo by Dr. Ira Sasowsky.

The Dolines(Sinkholes)

Sinkholes, also termed dolines are bowl-shaped depressions ranging in size from several to more than 3000 ft. in diameter, and are anywhere from less than several feet to hundreds of feet in depth (White, 1988). In karst regions, these negative relief landforms are regarded as the fundamental expression of karst relief due to their propensity to replace the stream and river valleys of fluvial terrain (Sweeting, 1973). According to White (1988), every closed depression has three components:

• A drain—some high permeability pathway that carries the internal runoff collected by the sinkhole into the subsurface.

• The solutionally modified zone at or just below the bedrock surface.

• A cover of soil, colluvium, glacial drift or moraine, volcanic ash, or other unconsolidated material that makes up the land surface. In some landscapes, the cover may be absent.

Close to 20% of the earth’s dry land surface and about 40% of the United States east of Tulsa, Oklahoma is karst (White et al., 1995). In Ohio, 1/3 of the state is underlain by Paleozoic carbonates, but less than 2% includes karst terrain (Hull, 1999).

Rebecca Bixby, Dr. Sasowsky, & Mr. Dick Bell(owner of Seneca Caverns) standing in a depression at the "Big Sink" in Thompson Township, Seneca County, Ohio. For a view of the bedrock fracture looking down the installed drain CLICK HERE

Tintera (1980) mapped 260 closed depressions in the Bellevue-Castalia Karst Plain, ranging in size from a foot in diameter up to 270 acres in area. These features provide an intimate connection between surface and groundwaters, leading to a high susceptibility for groundwater contamination. Most of the depressions are mantled by soils, covered by trees, filled with trash, or with water, making detailed interpretation and classification difficult at best (Tintera, 1980).

However, most of the closed depressions have been classified as subsidence type dolines (Kihn, 1988). These along with collapse dolines, alluvial streamsink dolines, and solution dolines are all found on the karst plain (Figure 5) , and are expressed almost exclusively in the Columbus Limestone (Kihn, 1988).

Figure 5. Five types of Dolines: a) collapse doline, (b) solution doline, (c) subsidence doline, (d) cover collapse doline, (e) alluvial stream sink doline (after Jennings, 1985).

Procedure

Preliminary investigations into the evolution and spatial distribution of sinkholes in the Bellevue-Castalia region of north-central Ohio will involve an intensive review of published literature pertaining to subsidence, subsidence mechanisms, and karst processes. This will be followed by incorporating previous government, published, and unpublished studies in order to evaluate structural, lithologic, and hydrogeologic controls on the origin and distribution of these features on the karst plain. Previous works by Tintera (1980), Kihn (1988), Hume(1991), and Ludwikowski (1993) will serve as a preliminary template for the location of sinkholes, joints, fractures, and lineaments that have been mapped. These will be incorporated into a GIS (Geographic Information System) along with bedrock, hydrologic, and geomorphologic maps in order to thoroughly evaluate any controlling variables on the preferential spatial distribution of sinkholes in the area.

Spatial distributions will also be evaluated by examining depression densities, doline areas, and the relative spacing of nearest neighbor dolines. These will be geometrically classified based on their morphologies (after White, 1988). Also, following the inspection of selected caves, it will be determined if there is any relation between cave formation and the presence of sinkholes at the surface.

The next step in this study will focus on refining a conceptual model for sinkhole occurrence on the Bellevue-Castalia Karst Plain. It has been hypothesized that the sinkholes in the region result from the dissolution of gypsum in the subsurface followed by the progressive upward collapse to the surface (Kraus, 1905, Verber and Stansbery, 1953). Another idea is that the sinkholes form by the dissolution of the very pure Columbus Limestone (>94.0% CaCO3) (Forsyth, 1988) along joints and fractures, piping soils and dissolved bedrock into the subsurface. This will be evaluated by geochemical analysis of selected spring waters emerging from sinkhole, contact, and fracture springs in the northern portion of the study area. Existing data of the chemical character of the waters will be used and any new data will be collected as required. Spring waters will be tested for concentrations of anions and cations, including sulfate or carbonate, and compared to concentrations found in selected sinking streams and wells in the area. The geochemical program, WATEQ 4F, and possibly, NETPATH, will be used to identify the reactions occurring along the south-north flow path.

Several geophysical methods will also be employed, and run over selected dolines in order to reveal their subsurface morphology. Microgravity techniques have proved to be a useful tool in locating low gravity anomalies which indicate bedrock caves, voids in the regolith, or places where the depth to bedrock abruptly increases possibly indicating a zone where soils are being piped into bedrock fractures (Crawford et al., 1999). A quasi 3D gravity survey will be run, and, using the known densities of the bedrock, clays, and soils within the depression, will be modeled for depth and orientation of any subsurface voids in the underlying bedrock or regolith. The information gained from this survey will be used to constrain a resistivity survey that will be run over the same features. This should serve to indicate the depth of the subsurface void or drain by profiling and trenching across and around the borders of the feature indicating whether or not the dolines are forming along preferred joints or fractures that can be traced laterally outside the area of subsidence.

Expected Results

The results obtained from this investigation are expected to characterize the origin and evolution of these features in the Bellevue-Castalia region of north-central Ohio. The use of a Geograpic Information System (GIS) help will delineate any structural, lithologic, or morphologic controls on the preferred spatial distribution of sinkholes. By determining the concentration of major anions and cations in spring waters and comparing those results against those obtained from selected surfaceand well water sources, the preferential dissolution of limestone, dolomite, or gypsum may be identified as a mechanism of subsidence. Finally, geophysical surveys should serve to indicate the subsurface morphology of these features and whether or not they are spatially controlled by the presence of fractures in the bedrock. These surveys may also indicate the depth at which dissolution and void forming processes are taking place in the subsurface. And, knowing the thickness of the bedrock in the survey area this should indicate the formation that is susceptible to dissolution.

References Cited

• Bugliosi, E.F., 1999, The Midwestern basins and arches regional aquifer system in parts of Indiana, Ohio, Michigan, and Illinois—Summary: U.S. Geological Survey Professional Paper 1423-A.
• Burgess and Nipple, LTD, 1967, The northwest Ohio water development plan: Columbus, Ohio, 299 p.
• Carman, J.E., 1927, The Monroe division of rocks in Ohio: Journal of Geology, v. 35, p. 481-506.
• Chaffee, M.C., 1995, Geophysical prospecting for karst features in the Bellevue-Castalia region of north-central Ohio: Unpublished M.S. Thesis, The University of Toledo, Toledo, Ohio, 101 p
• Crawford, N.C., Lewis, M.A., Winter, S.A., and Webster, J.A., 1999, Microgravity techniques for subsurface investigations of sinkhole collapses and for detection of groundwater flow paths through karst aquifers: in Hydrogeology and Engineering Geology of Sinkholes and Karst, Beck, Pettit, and Herring (eds), Balkema, Rotterdam, p. 203-218.
• Drane III, L.A., 1993, Hydrogeology and geophysics of karst terrain in Thompson Township, Seneca County, Ohio, with investigation of the ground-water non-point source pollution from bacteria, nitrates, and pesticides: Unpublished M.S. Thesis, The University of Toledo, Toledo, Ohio, 200 p.
• Eberts, S.M. and George, L.L., 2000, Regional groundwater-water flow and Geochemistry in the Midwestern Basins and Arches aquifer system in parts of Indiana, Ohio, Michigan, and Illinois: U.S. Geological Survey Professional Paper 1423-C.
• Forster, K.M., 1997, A geochemical and flood pulse analysis through continuous electronic monitoring of springs in the Bellevue-Castalia area, north-central Ohio: Unpublished M.S. Thesis, The University of Toledo, Toledo, Ohio, 162 p.
• Forsyth, J.L., 1959, The beach ridges of northern Ohio: Ohio Department of Natural Resources, Geological Survey Information Circular No. 25.
• Forsyth, J.L., 1988, The uniqueness of the Devonian Columbus Limestone: Ohio Journal of Science, v. 88, no. 2, p. 14.
• Hanover, R.H., 1994, Analysis of ground-water flow along a regional flow path of the Midwestern Basins and Arches Regional Aquifer System in Ohio: U.S. Geological Survey Water-Resources Investigations Report 94-4105
• Hull, D.N., 1990, Generalized column of bedrock units in Ohio: Ohio Department of Natural Resources, Division of Geological Survey, 1 p.
• Hull, D.N., 1999, Mapping Ohio’s karst terrain: Ohio Geology, no. 2, 8 p.
• Hume, D.S., 1991, Structural geology and radon gas in groundwater in parts of Erie, Huron, and Seneca Counties: Unpublished Masters Thesis, University of Toledo, Toledo, Ohio, 257 p.
• Jennings, J.N., 1985, Karst Geomorphology: New York, Basil Blackwell, 293 p.
• Kihn, G.E., 1988, Hydrogeology of the Bellevue-Castalia Area, north-central Ohio, with emphasis on Seneca Caverns: Unpublished M.S. Thesis, The University of Toledo, Toledo, Ohio, 163 p.
• Kraus, E.H., 1905, On the origin of the caves of the island of Put-In-Bay, Lake Erie: American Geologist, v. 35, p. 167-171.
• Leverett, F., 1902, Glacial formations and drainage features of the Erie and Ohio basins: U.S. Geological Survey Monographs v. XLI, 802 p.
• Ludwikowski, D.L., 1993, Geomorphology in parts of Sandusky, Seneca, Erie, and Huron Counties, Ohio: Unpublished Masters Thesis, University of Toledo, Toledo, Ohio, 219 p.
• Norris, S.E., and Fiedler, R.E., 1971, Availability of ground water from limestone and dolomite aquifers in northwest Ohio and its relation to geologic structure: U.S. Geological Survey Professional Paper 750-B, p. B229-B235.
• Owens, G.L., 1970, The subsurface Silurian-Devonian “Big Lime” of Ohio: Ohio Department of Natural Resources, Geological Survey Report of Investigations No. 75, 17 p.
• Root, S., and Onasch C.M., 1999, Structure and tectonic evolution of the transitional region between the central Appalachian foreland and interior cratonic basins: Tectonophysics, v. 305, p. 205-223.
• Shaffer, P.R., 1951, Shore erosion on Sandusky Bay: Ohio Department of Natural Resources, Geological Survey Report of Investigations No. 7, 5 p.
• Shaver, R.H., 1985, Midwestern Basin and Arches Region-correlation of stratigraphic units of North America (COSUNA) project: American Association of Petroleum Geologists, 1 sheet.
• Shindel, H.L., Mangus, J.P., and Trimble, L.E., 2001, Water resources data, Ohio, water year 2000, Volume 2, St. Lawrence River Basin and statewide project data: U.S. Geological Survey Water Data Report OH-00-2.
• Stauffer, C.R., 1909, The Middle Devonian of Ohio: Ohio Geological Survey Bulletin 10, 204 p.
• Stout, W., Ver Steeg, K., and Lamb, G.F., 1943, Geology of Water in Ohio (A Basic Report): Ohio Geological Survey Bulletin 44, 694 p.
• Stout, W., 1941, Dolomites and limestones of western Ohio: Ohio Geological Survey Bulletin 42, 468 p.
• Sweeting, M.M., 1973, Karst landforms: New York, Columbia University Press, 362 p.
• Tintera, J.J., 1980, The identification and interpretation of karst features in the Bellevue-Castalia region of Ohio: Unpublished M.S. Thesis, Bowling Green State University, Bowling Green, Ohio, 62 p.
• Verber, J.L., and Stansbery, D.H., 1953, Caves in the Lake Erie Islands: Ohio Journal of Science, v. 53, p. 358-362.
• White, W.B., Culver, D.C., Herman, J.S., Kane, T.C., and Mylroie, J.E., 1995, Karst lands: American Scientist, v. 83, p. 450-459.
• White, W.B., 1988, Geomorphology and hydrology of karst terrains: New York, Oxford University Press, 464 p.
• Woodward, H.P., 1961, Preliminary subsurface study of southeastern Appalachian Interior Plateau: American Association of Petroleum Geologists Bulletin, v. 45, p. 1634-1655.