Fossil Limpets and Fossil Otoliths as Indicators of the Extent of the Wet Prairies Around Castalia in Margaretta Township, Erie County and Townsend and Riley Townships in Sandusky County, Ohio, U.S.A.

Michael James Norrocky

162 Duchess St.

Vickery, OH 43464 U.S.A.



Limpets are aquatic gastropods in the family Ancylidae, most with a shell that resembles a clam or mussels (patelliform). Ferrissia, a common genus with numerous species inhabiting many types of habitats, was found during this study.

Otoliths (ear stones) are calcium structures found in pairs (3 pairs) in the skull of bony fishes. There are three types; sagitta, astericus, and lapillus. The lapilli are found about midway up the skull and are associated with balance. Each asteriscus is nested on the lateral surface of a sagitta and the pair are found near the bottom of the skull. They are associated with hearing (Campana 2004). The presence of limpets and otoliths in the fields around Castalia, Ohio and to the west suggests that portions of the land was covered with water to some extent all year much like portions of the Black Swamp of northwest Ohio.

As the Wisconsin glacier began to retreat about 14,000 years ago, the melt water produced a series of 11 lakes, Lake Erie being the last. The first and highest of these has been named Lake Maumee I and it covered the land around the Lake Erie basin to a level of about 800 feet above mean sea level (msl). Subsequent ice invasions and melt water lakes deposited moraines and beach ridges which, along with rock outcrops, were the majority of the topographic relief in northwest and north-central Ohio (Forsyth 1959).

This low, wet land made early settlement slow and this was particularly the case in northwest Ohio. This 30-40 mile wide, low ground extended up the Maumee River to near Fort Wayne, Indiana and covered some 1500 square miles (Kaatz 1955, Verduin 1969). This area became known as the Black Swamp and extended east to just beyond the northeastern edge of Sandusky County.

During the 1800's the Black Swamp was settled and eventually drained by excavating ditches which drained into the Portage, Maumee and Sandusky Rivers and Sandusky Bay. Prior to the draining, fish such as Muskellunge (Esox mosquinongy), Northern Pike (Esox lucius) and Cisco (Coregonus artedii ssp.) entered the swamp in the spawning season and the area was important to their survival in Lake Erie (Langlois 1954, Verduin 1969). These species' numbers have been declining since around 1885, the period when the Black Swamp was almost completely drained (Trautman 1981).

The purpose of this paper is to point out the extent of the wet areas in Townsend and Riley Townships in northeast Sandusky County and Margaretta Twp. in northwest Erie County based on the fossil limpet shells and fish otoliths found in fields downstream of springs suggesting that this area was similar to the Black Swamp to the west.


The village of Castalia (41° 23' 59" N 82° 48' 15" W) lies near the middle of Margaretta Township, in northwestern Erie County, and about 6 miles southwest of Sandusky. It is near the east side of the study area and lies at the north edge of the Columbus Cuesta, an outcrop of the Findlay Arch which runs generally NE and SW and dips 20' per mile to the southeast (Forsyth 1983). The land north of the cuesta slopes to the Sandusky Bay at about 14 feet per mile. At least 18 springs are located in the approximate 40 mi2 area from Castalia west to Riley Township, Sandusky County (Fig. 1). State Route 101, which runs southwest on the Lake Warren beach ridge at 650 ft. above mean sea level (msl), forms the south boundary of the spring area. From Wade Spring (41° 20' 53" N 82° 53' 20.2" W), about 6 mi from Castalia, the line runs northwest for 6.75 mi to Blue Heron Reserve in the Sandusky County Park system (41° 24' 14.5" N 82° 57' 22.6" W). It then runs northeast 9 mi to Bay View, passing White's Landing (41° 26' 05" N 82° 54' 30" W) at 3.7 mi, where 1 spring can be found on land and at least 5 under the Sandusky Bay. From Bay View, it runs 4.9 mi south-southeast to Castalia.

Area of Study

The springs flow year around and the volume of some is substantial. They provided a constant source of water flowing over the land. It is estimated that the Castalia Blue Hole (Figs. 2 & 3) discharges between 2690 and 3053 gallons per minute (gpm) and Rockwell Trout Club Spring (Fig. 3) discharges 3175 to 3575 gpm (Kihn 1988). Ver Steeg and Yunck (1935) described the area around Castalia as, "…the prairie to the north and west of Castalia was too wet for trees." This provided the habitat in which the limpets and snails lived and the fish invaded, probably in the spring, to spawn. Even the Castalia Blue Hole, which was a popular tourist attraction from early in the 20th century until its closing in the 1970's was "… inaccessible except by boat." in the 1880's (Ver Steeg & Yunck 1932).

Area of Study

Maxfield Ludlow, surveying the west boundary of the Connecticut Western Reserve (west boundary Erie County), "Commenced the 29th mile north in fallen timbers & swamp." He notes 50 chains later, "The Prairy [sic] appears to be 5 miles E & W very wet". At the corner post he concludes that the area is, "… not worth a farthing." This location was northwest from Castalia.

Today these springs flow in channels, primarily man made. In the past they flowed in natural channels or onto the surrounding land producing the large wet prairies estimated to cover 200,000 acres (Monroe 1974) on their way to Sandusky Bay (Fig. 1). There are hundreds of wells in the study area, drilled for domestic use. Most of these flow constantly to some degree and undoubtedly reduced the flow from the springs, thus reducing the extant they flooded the land.

Area of Study

Besides the springs near Castalia, three springs are known to occur around the village of Green Springs, 15 mi (24 km) to the southwest of Castalia. It is not known if they are part of the same karst found at Castalia, but they have an extremely strong smell of sulfur, more than those near Castalia. These springs flow into Green Creek which flows NNE to Sandusky Bay. The well at Robert Wickert's home site at the corner of Sandusky Township roads T221 and TR 198 (on the SE edge of Fremont) smelled of sulfur in the spring following heavy rains, suggesting that the Green Springs aquifer was flowing NNE, flooding into this well. (Personal Communication). Seven and one half miles north-northeast of Green Springs at the junction of SR 412 and TR 220 in Riley Twp., a spring rises and 3.5 mi further NNE, the Barr artesian well flows from a casing, also with a strong sulfur smell. Water testing would help clarify this issue.

Green Creek enters Sandusky River near the mouth to the Sandusky River, in Riley Township (Fig. 1). One and eighty six hundredths of a mile to the east of the mouth of Green Creek, Pickerel Creek flows into Sandusky Bay carrying water from Mound Spring, Crane Spring, Thorbahn Spring, and an un-named spring along TR 260, west of Vickery (Figs. 1 & 3).

In response to a question about when the last of the post glacial lakes covered the study area, Dr. Ed Herdendorf replied in an email, "The last glacial lake to cover the Castalia area was Lake Lundy about 12,500 years ago at an elevation of about 615-620 feet. As soon as the ice melted back from the Niagara sill, about 12,000 years ago, the level of the lake dropped to about 490 feet initiating the post glacial (Holocene) lakes in the Erie Basin. As isostatic rebound raised the Niagara outlet, the lake rose in stages to its present level, but probably never greatly exceeded the modern level and certainly didn't flood the Castalia area in the past 8,000 years."

Paul Sears (1966) studied the pollen from herbaceous plants and trees in the 3000 acre Castalia prairie area. He recognized two different habitats, one he refers to as muck soils (highly organic) and the others as "Andropogon" prairies (Fig. 4). Presumably the latter was drier than the former. The Martin, Harkness, and Lippert springs (Fig. 3) all rise in the Andropogon area and fossil limpets were found around the Martin Spring suggesting it flooded the surrounding area in the vicinity of the present community of Crystal Rock.

The post glacial lakes deposited 54 ft. of fine clay in northern Ohio. The clay was covered by 6 ft. to 7 ft. of organic muck in the 3000 acre area (Sears 1967). The depth to the clay was variable and occurred at the surface in many places thus inhibiting water infiltration and contributing to the wet conditions.

Methods and Materials

This study began when a farmer, Leon Balduff, had a small spring on his property tiled in late 2010 (41° 24' 57" N 82° 54' 02" W)(Fig. 5, Fig. 3 - Site #1). The spring is located in a field NE of the corner of Sandusky County Road 290 and State Route 6. In January 2011, I collected approximately 2 liters of the material that had lain below the topsoil (Fig. 6). The material was washed through 2 mm and 0.8 mm screens (Fig. 7). The separate samples were dried and examined with a dissecting microscope. Specimens were separated into groups according to type: limpet, snail, etc., photographed, and stored. The only variable was the sample size which varied somewhere between 1 and 2 quarts. The terms marl and tufa, as used in this paper, are defined as: 1) tufa (travertine) is a rock that contains impressions or inclusions that indicate the rock formed from deposition of CaCO3 in an aquatic environment, around emergent vascular plants and the algae Chara, and 2) marl is black organic soil, with tufa and both terrestrial and aquatic snails.

Muck Soil

Balduf Spring

Snails Tufa


Most of the sample sites were marl, indicating that they had been covered by spring water with emergent plants sometime in the past. Clay sites had none of the marl characteristics such as snails and tufa and had no limpets. For instance, site #21 was clay had no limpets although ctenoid and cycloid fish scales were found in the field.


List of collection sites and specimens found (see Fig. 3):

1) Balduff - Home spring - fossil limpets, snails, & Percidae fish scale.
2) Balduff - Schultz field - fossil limpets, snails, Percidae fish scale.
3) Bill Warner field - fossil limpet, & Drum otoliths.
4) Meggitt spring field - fossil limpets, snails, ostracods & cycloid fish scale.
5) Balduff - Miller field - fossil limpets, snails, and Sphaeridae.
6) Oxbo Rd. field – fossil limpet shells, snails, & Pike otolith.
7) Thicket Rd. field - fossil limpets, snails, & Sphaeridae.
8) Martin Spring - fossil limpets & snails, ostracods, Sphaeridae
9) TR 280 - fossil limpets, snails, Sphaeridae, & Drum & Yellow. Perch otolith.
10) Miller Blue Hole access field - fossil limpets, snails, ostracods & Pike otolith.
11) TR 247 field - fossil limpets & snails.
12) CR 312 (NW Rd.) field - fossil limpets & snails.
13) Blue Heron spring field – fossil limpets, snails.
14) McCartney Rd. east field - fossil limpets, snails, & Pumpkinseed otolith.
15) Keller - Bailey field - fossil limpets & snails, ostracods, Sphaeridae.
16) Altvater field - fossil limpets & snails.
17) Heywood Rd. field - fossil limpets & snails, piece of broken otolith.
18) Yetter spring field - fossil limpets, snails, Drum otoliths.
19) Downing TR 263 - fossil limpets (many), Sphaeridae, snails, and ostracods.
20) Resthaven clay pit woods - fossil limpets & snails.
21) TR 260 - ctenoid & cycloid scales.
22) Downing TR 306 field - fossil limpets, otoliths, snails, ostracods.
23) TR 298 field - fossil limpets.
24) Warner TR 255 – fossil limpets, Drum otoliths & snails.
25) Yetter farm field - fossil limpets, Pike otolith, Drum otoliths, & snails.

Figure 8 is a piece of tufa from the Balduff site (#1). Water from area springs contains a high concentration of calcium (Table 1) which deposited on plants, fossilizing them as tufa rock. Tubes that formed around plant stems and a leaf impression can be seen in the photo. Also in the Balduff sample were many encrusted stems of Chara (Fig. 9), snails (Fig. 10), one Percidae scale (Fig. 11), and 5 whole and one broken limpets which appeared to be small clams.

Plant tubes

The small "clams shells" (Fig. 12), fish scale and Chara, suggested that the area had been under water all year at some time in the past. Since fish are part of the life history of clams during the juvenile glochidia stage, it was assumed that the "clam shells" were indeed naiads. I emailed pictures to Dr. Tom Watters at the museum at Ohio State University and following a telephone call and resending the pictures he agreed that the "clam shells" were indeed very young naiads. The fish scale was identified by Dr. Ted Cavender from the Ohio State University museum as being a member of the Perch family, Percidae.

My next collection was over one mile from the Balduff Spring in a field in which the soil was marl (Fig. 3 - Site #2). It was similar to the soil at the Balduff site. I found 18 of the "clams". As I was lining them up for a photograph, I noticed that all the "shells" would have been from the same side of a clam. At this point I realized that the "clams" were not clams at all.



Fish scales

Plant tubes

At the Oxbo Rd. (Fig. 3, Site # 6) site, a shell was found that did not look like the "clams" I had been finding, but reminded me of a Keyhole limpet I had seen in Florida. When I searched on the internet for freshwater limpets, pictures of the last specimen and the "clams" were displayed. I had been finding freshwater fossil limpets! Searching on the internet, I located Dr. Robert Dillon at the University of Charleston, an expert on gastropods. He verified that the "clams" were indeed fossil limpets, probably of the genus Ferrissia.

Keyhole limpits

Water analysis

Freshwater drum

Including the first collection at Balduff's, I have collected 25 sites, 24 of which have yielded over 200 fossil limpets. Site #21 was clay and had no marl ground, tufa rock, or snails. Besides the limpets, 21 Freshwater Drum (Aplodinotus grunniens) otoliths (Fig. 14) from 0.25 in up to 1.25 in were found in five fields (Fig. 3, Sites # 2, 7, 10, 24, 25). Also a few mussels fragments (Unionidae), fingernail clams (Sphaeridae), and Ostracods (Ostracoda) were found (see list above). The habitat must have been ideal for snails. Sterki (1920) reported that a quart sample from near Castalia, contained over 100,000 small snails. Figure 15 is a typical field that contained limpets and otoliths as well as tufa and snails.

Native American artifacts were found on the surface in the fields that yielded Drum otoliths. These fields were marl. Fields that were clay and/or sandy, also contained Native American artifacts but no otoliths. The Native Americans may have hunted the marl fields with the emergent vegetation and lost artifacts doing so. The lack of otoliths on the higher ground suggests that: 1) the Native Americans on the high ground did not consume Drum, 2) they put their debris in middens which were not found, 3) the otoliths were eroded away, or 4) the Drum lived in the lower ground when it was flooded by the spring waters. The findings of this study suggest the latter.

Large otoliths (sagittae) from fish other than Drum were not found but; 1) they are not as massive as those from the Drum and could have broken and not noticed; 2) because they were less massive they eroded away; or 3) just not been discovered or collected in the samples.

Otolith shapes are very species specific (Campana 2007, pg. vii). Otoliths and fragments ranging from 0.03 in. to 0.05 in were found in 7 fields. They were compared to photos of otoliths in a collection made by the author, now in the collections of the University of Michigan Museum of Zoology. Two of the otoliths resembled otoliths from Northern Pike, family Ecocidae (Figs. 16 & 17), one a Yellow Perch, family Percidae (Fig. 18), one a Pumpkinseed, family Centrachidae (Fig. 19). The fragments were not large enough to be recognized (Figs. 20 & 21).

Typical farmland

Oxbo Road otolith

Blue Hole otolith

Yellow perch


TR 306

Heywood Road


Kaatz (1955) wrote that, "In many places the swamp boundary is, indeed, difficult to determine with great accuracy". He drew the east boundary following the lower part of the Sandusky River from about Fremont. James Mollenkamp wrote that the swamp extended to Sandusky, Ohio on the east. Robert Gordon (1955) composed a map of the vegetation of Ohio based on the notes of early surveyors. His swamp forest extended into Margaretta Township, Erie County Ohio (Fig. 22).

Townsend and Riley Townships are east of Fremont in the Black Swamp (Kaatz 1955). Margaretta Township is in Erie Co. east of Townsend Township. Margaretta Township was surveyed by two surveyors in 1808. Maxfield Ludlow surveyed the western line of the Connecticut Western Reserve which is the west line of Margaretta Township. Almon Ruggles surveyed the balance of Erie County. In this survey the township was divided into 5 mile squares, therefore the survey notes did not provide much information about the vegetation. Sears (1967), Sterki (1920), Ver Steeg (1932) and Gordon (1966) provided much useful information.

The constant flow from the springs produced the environment that allowed aquatic plants to grow and limpets and snails to thrive. Limpets require water year around (Dr. Robert Dillon, Personal Communication). Victor Sterki (1920) collected the gastropods in the area around Castalia, primarily northwest. He noted that west of Castalia, "… is a marl deposit extending over several miles,… ". He also noted that, "Some parts are still marshy with luxuriant marsh vegetation and in places even covered with open water". The area was a rich environment for snails. At one site Sterki (1920) scooped up about one quart of small snails that he estimated to contain about 100,000 individuals.

Around the springs, below the 6' (1.8 m) layer of marl that formed in the spring water, clay and silts exist to 54' (16.4 m) level, followed by 8.5' (2.6 m) of till resting on bedrock. The clays, 1) Miami, Crosby, Brookston and Clyde silty loams, 2) Brookston clay, 3) Toledo silty clay is very fine and does not transport water well. Hence, water from any source lays on the surface of the clay for an extended period contributing to the condition resembling a swamp.

The Drum otoliths and the mussel shells found in samples could have been debris from Native Americans' meals. The smaller otoliths came from fish too small to have been practical as food. These were young fish living in the area in schools or singly. This would seem to be the habitat Sterki was describing. The adult members of the Percidae, Esocidae, and Centrarchidae families would have utilized these marshy and open water areas for spawning. Northern Pike still move up ditches in this area in the spring and fall, supposedly to spawn.

Drum, Perch, and Pumpkinseed consume snails (Pennak 1953). The snails living amongst the vegetation would have provided adequate food for the spawning fish. The fish would have supplied food for the Pike. The plants in the relative shallow water would have provided a place for all the species to spawn and places to hide. The small Pike, Pumpkinseed, and Perch would have found insects and the many snails Sterki noted as food. These areas were probably what would be termed "wet prairies" or swamps and numbers of these were noted by the surveyors in 1819 & 1820.

The plant impressions in the tufa rock result from the high calcium concentration of calcium in the spring water being deposited on the plant parts and offer proof of the conditions. The algae, Chara sp, found in the marl secrete calcium on its surface, fossilizing those plants. The tufa rock was found at each site that yielded limpets suggesting that the limpets lived on the plants and the high spring water kept the areas wet and formed the tufa rock. Today most of these areas are productive farm fields (Fig. 15).

Approximately 1800 acres that were prairie are in the State owned Resthaven Wildlife Area. The remainder is in private ownership and extends almost continually in patches to Blue Heron Spring in Riley Township. West of Blue Heron the land is primarily clay with sand in a few fields near Fremont.

Natural Vegetation

Robert Gordon (1966) concluded that the study area vegetation was Elm-Ash Swamp Forest type (Fig. 22). Jane Forsyth (1970), comparing the edaphic influences of geology on the vegetation and found a close relationship between the original vegetation of Gordon and the underlying geology. Her map, (Fig. 23), of the swamp forest outlines essentially the same part of the study area as Gordon (Fig. 22). On Gordon's map note the prairie grassland type around Castalia. Limpets were collected in the north part of Section 5 of Figure 23, indicating it was a wet prairie or swamp.

Area Geology

Marshy Bayview Peninsula


I would like to thank Dr. Robert Dillon for identifying snails and limpets, Dr. Jack Burch helped with limpet identifications. I thank all the people that allowed me to collect a sample from their property.


Black, Bryan A. and Marc D. Abrams. 2001. Influences of Native Americans and Surveyor Biases on Metes and Bounds Witness-Tree Distributions. Ecology 82(9): 2574-2586.

Campana, S. E. 2004. Photographic Atlas of Fish Otoliths of the Northwest Atlantic Ocean. NRC Research Press, Ottawa, Ontario. 284 pp.

Crist, Dennis R. 2006. Potentiometric Surface of the Consolidated Aquifers in Sandusky County. Ohio Dept. of Nat. Res., Div. of Water 1 pg.

Dillon, Robert T. 2000 The Ecology of Freshwater Molluscs. Cambridge University Press 499 pp.

Forsyth, Jane L. 1959. The beach ridges of Northern Ohio. Ohio Geol. Surv. Inf. Circ. 25: 1 – 10.

______ 1970. A Geologist Looks at the Natural Vegetation Map of Ohio. 70(3): 180 – 191.

Forsyth, Jane L., Kahle, Charles F., 1983. Cruising the Columbus Cuesta. Ohio Acad. Sci. Mimeo Field Guide 22 pp.

Kaatz, Martin R. 1955. The Black Swamp: A Study in Historical Geography. Annals of the Assoc. of Am. Geographers 45(1): 1-36.

Kihn, Gary E. 1988 Hydrogeology of the Bellevue-Castalia area, North-Central Ohio, With An Emphasis on Seneca Caverns. [Master's Thesis]: University of Toledo, Ohio, 163 p.

Ludlow, M. 1808 Western boundary of the Firelands. Firelands Pioneer 19: 1868-1870, 1915, pgs. 250 – 360.

Monroe, Tom. 1974 Long-range wildlife management plan for Resthaven Wildlife Area. Mimeo report, Ohio Dept. Nat. Res., Div. of Wildlife. 10 p.

Pennak, Robert W. 1953. Fresh-Water Invertebrates of the United States. Ronald Press, New York 769 pgs.

Sears, Paul B. 1966. The Castalia Prairie. The Ohio Journal of Science 67(2): 79-88.

______ 1926. The Natural Vegetation of Ohio, II. The Prairies The Ohio Journal of Science 26(3): 128-146.

Sterki, Victor 1920. Marl Deposits in Ohio and Their Fossil Mollusca. Ohio Jour. Sci. 20(6): 173-184.

Survey of Sandusky County, 1819 – 1820. Field notes for Riley Township and Townsend Township. Available from State of Ohio, Columbus, OH.

Trautman, Milton, 1981. The Fishes of Ohio. Ohio State Univ. Press 782 pp.

Verduin, Jacob 1969. Man's Influence on Lake Erie. Ohio Jour. Sci. 69(2): 65- 70.

Ver Steeg, Karl, George Yunck, 1932 The Blue Hole of Castalia. Ohio Jour. Sci. 32: 425 – 435.

Download PDF file.

Don't have Adobe® Reader to open PDF files? Free download here.Adobe Reader Icon