Setting

Geography and Hydrology

The Little Sequatchie River valley is located on the eastern escarpment of the Cumberland Plateau in southern Tennessee (fig. 1). The valley spans both Grundy and Marion Counties and is near the towns of Sequatchie and Gruetli-Laager (fig. 4A). The Little Sequatchie River is the largest tributary to the Sequatchie River, by drainage area, with an area of 131.7 square miles (mi²) at the confluence with the Sequatchie River. The Little Sequatchie watershed is bounded to the north by the Collins River watershed, to the west by the Battle Creek and Pryor Cove Branch watersheds, and to the east by the Griffith Creek watershed (USGS, 2019) (fig. 4B). Vertical relief is as much as 1,730 feet (ft) from the lowest to highest point in the watershed; however, the relief is closer to 1,000 ft within the valley itself. The river has no continuous record of stage or discharge, although streamflow measurements and peak streamflow data indicate a range in flow from 0.1 to 10,600 cubic feet per second (ft³/s) (USGS, 2023b).

Pryor Cove is located directly southwest of the lower Little Sequatchie River valley, sharing an east-west boundary with the Little Sequatchie watershed (fig. 4A). The cove consists of two main valleys whose principal streams, West and North Fork Pryor Cove Branches, contribute flow to Pryor Cove Branch. The total area for the Pryor Cove watershed is 12.8 mi² at the confluence between Pryor Cove Branch and Standifer Branch. The cove is bounded on the east by the Little Sequatchie watershed, to the north by the Fiery Gizzard Cove watershed, and on the west by the Battle Creek watershed (fig. 4B) (USGS, 2019). Vertical relief of the Pryor Cove watershed is very similar to that of the Little Sequatchie watershed, with a total relief of 1,327 ft, and an elevation range from 650 to 1,977 ft above the National Geodetic Vertical Datum of 1929 (USGS, 1947).

Climate within the study area is classified as humid subtropical (Beck and others, 2018), with warm, humid summers and typically mild winters. Precipitation averages greater than 59 inches per year (U.S. Climate Data, 2023), with the winter and early spring being the seasons having the highest rainfall totals, although afternoon atmospheric convection brings additional rainfall during the summer months. The forest cover in the study area is a mixed mesophytic forest, with forest communities varying by both position along the Cumberland Plateau top and escarpment as well as by landform type (slopes versus gorges versus stream terraces) (Tennessee Division of Forestry, 2020). Pines are largely restricted to the top of the Cumberland Plateau on the acidic soils weathered from sandstones, whereas mixed oak and chestnut oak forests occupy the Cumberland Plateau escarpment and hemlocks may be found in deeper gorges and may occur next to river birch and sycamore trees alongside streams (Tennessee Division of Forestry, 2020). Forests in Tennessee have undergone dramatic changes since the initial settlement of the land because of historic logging practices and fire suppression (Baker and Van Lear, 1998, Woodbridge and Dovciak, 2022). Within the Little Sequatchie and Pryor Cove watersheds most areas have received some timber harvest, although some small areas of undisturbed forest do exist in the more remote and inaccessible portions of the study area. Land ownership within the study area is predominantly privately owned and several landowners own large parcels exceeding 1,000 acres, particularly within the Little Sequatchie watershed. State-owned lands are present in the Little Sequatchie watershed at Sequatchie Cave State Natural Area and The Chimneys State Natural Area, which total 133 and 33 acres, respectively. At least one small parcel of land in Pryor Cove is owned by a nonprofit organization, Tennessee River Gorge Trust, that preserves a rock-climbing area at Castle Rock Point (fig. 4A). Land use-land cover data were analyzed overall for each of the watersheds. Both watersheds are forest cover dominated, with over 75 percent of the watersheds classified as forest cover (Dewitz and USGS, 2021). Developed land is an overall higher percentage in the Pryor Cove watershed, because of the presence of the Town of Jasper, but is overall a much smaller area than the areas classified as developed in the Little Sequatchie watershed because the Little Sequatchie watershed is much larger. Land use-land cover data are more varied in the Little Sequatchie watershed, likely because of the much larger area (Dewitz and USGS, 2021).

Geology

The geology of the Little Sequatchie and Pryor Cove watersheds has been described and mapped at varying scales by multiple geologists and agencies (Wilson and others, 1956; Milici and Finlayson, 1979, 1983; Jones and Moore, 1986; Milici and others, 1986). The geology of the Little Sequatchie River valley and of nearby Pryor Cove (fig. 4) is similar to other portions of the Cumberland Plateau where the top of the plateau and ridgetops are capped by Lower Pennsylvanian siliciclastics although the lower slopes and valley floor are largely composed of Upper Mississippian carbonate strata (fig. 5A). Although much of the following discussion focuses on the surficial geology of the study area, the subsurface geology does play a role in the karst hydrology of the area, particularly within Pryor Cove. The units in the study area extend upward from the Mississippian Monteagle Limestone in the valley bottoms to the Pennsylvanian Rockcastle Conglomerate of the Crab Orchard Mountains Group along the top of the Cumberland Plateau. Quaternary alluvial deposits occur along the lower stream valleys for the Little Sequatchie River (fig. 5B) and Pryor Cove. The Cambrian and Ordovician units in the Sequatchie Valley east of the Sequatchie fault are not included in the study area. The following descriptions of the stratigraphic units are compiled and summarized from detailed geologic maps published at a scale of 1:24,000 and created by Milici and Finlayson (1979, 1983), Jones and Moore (1986), and Milici and others (1986).

The Mississippian Monteagle Limestone is a grayish, medium- to thick-bedded limestone, which can be cross-bedded and commonly has stylolites as well as chert-rich zones (fig. 5C). The Monteagle Limestone can be 250–290 ft thick in the study area; however, the unit crops out only in limited locations in the study area, existing below the valley floors and alluvial deposits. The Monteagle Limestone is the primary unit in which the passages of Blue Spring (known locally as “Jasper Blue spring”) are formed. Above the Monteagle Limestone, but still largely located beneath the valley floors, is the Mississippian Hartselle Formation, which is a calcareous sandstone and arenaceous limestone that can often act as an aquiclude or aquitard in similar areas. The Hartselle Formation is as much as 50 ft thick and is absent in portions of the study area where the Bangor Limestone lies unconformably on the Monteagle Limestone. Throughout the study area, the Hartselle is largely subvalley floor (fig. 5D), although some limited exposures are present in areas adjacent to the portion of the main Sequatchie Valley near the mouth of both Pryor Cove and the Little Sequatchie River valley (Milici and others, 1986). Throughout much of the study area, the valley floors and base of the Cumberland Plateau are underlain by Mississippian Bangor Limestone, a light gray to dark gray limestone with a total thickness of 200–250 ft. In the study area, the Bangor Limestone is heavily karstified, with many of the springs and caves located in the Bangor. Overlying the Bangor Limestone is the Mississippian Pennington Formation, a highly varied formation consisting of shales, siltstones, sandstones, and limestones, with a total thickness of 350–400 ft. A lower member of the Pennington is a carbonate member that can be either a dolomite or dolomitic limestone, which is often karstified; many small springs discharge from this lower unit before sinking at the contact with the Bangor Limestone.

Above the Pennington Formation are a series of Lower Pennsylvanian siliciclastic strata that range from shales to conglomeratic units. Directly above the Pennington Formation is the Gizzard Group, consisting of the Raccoon Mountain Formation and the Warren Point Sandstone (fig. 5B, C). The Raccoon Mountain Formation is a sandstone with shale interbeds and a coal unit near the base of the formation. The Raccoon Mountain Formation underlies the Warren Point Sandstone, a formation that contains a middle shale unit and is as much as 400 ft thick in the Little Sequatchie watershed. The Crab Orchard Mountains Group overlies the Gizzard Group and consists of the Sewanee Conglomerate, Whitwell Shale, and Newton Sandstone. The Sewanee Conglomerate overlies the Warren Point Sandstone and is a 70- to 140-ft-thick formation that is cross-bedded and includes quartz pebbles as much as 1.5 inches in length. Above the Sewanee Conglomerate is the Whitwell Shale, a grayish silty shale with two major coal seams (the Sewanee and Richland coal seams) found throughout the 30- to 120-ft thickness. The Newton Sandstone overlies the Whitwell Shale and can reach thicknesses of as much as 140 ft, and is generally cross-bedded but can be conglomeratic in areas. The Vandever Formation, which overlies the Newton Sandstone, consists of two to three members, depending on location, and generally consists of a middle sandstone member in between two shale members with a total thickness of 250–600 ft. In the northern portions of the watershed at the highest elevations are areas underlain by the Rockcastle Conglomerate, a medium- to coarse-grained conglomerate with quartz pebbles that is greater than 80 ft thick. The youngest geologic units in the study area are the Quaternary alluvial deposits, which are present in the valley floors, with thicknesses of 30 to 100 ft, although most deposits are near the lower end of this range.

Caves and Karst

Cave and karst features are abundant throughout the Little Sequatchie and Pryor Cove drainage areas, particularly in areas with exposures of the Pennington Formation and Bangor Limestone, which typically occur along the lower slopes and valley floors. Although limestone beds are present in the Pennington Formation, the interbeds of shale, chert, and sandstone can make cave and karst development less consistent than in the other Mississippian strata. However, in the study area, both caves and stream sinks can be found throughout the entire thickness of the Pennington Formation. The majority of the major stream sinks throughout the study area are located at the contact between the Pennington Formation and the underlying Bangor Limestone, hereafter referred to as the “Pennington-Bangor contact.” The sinkpoints can migrate upstream or downstream depending on hydrologic conditions, typically migrating upstream during drier periods and downstream during wetter periods. Downstream from these major sinkpoints, extensive reaches of dry streambed are common that only receive flow during large flood events (figs. 6 and 7). Eventually, several miles downstream from the stream sinkpoints, springs resurge creating the perennial reaches of the major streams. There are 103 documented caves in the Little Sequatchie and Pryor Cove drainage areas, ranging from 50 ft to 6.75 miles (mi) in length (Tennessee Cave Survey, unpub. data, 2023).¹ In the study area, 60 caves (58 percent) are in the Bangor Limestone, 40 (39 percent) are in the Pennington Formation, 2 (2 percent) are in the Monteagle Limestone, and 1 (<1 percent) is in the Hartselle Formation (Tennessee Cave Survey, unpub. data, 2023). The numerous caves in the study area range widely in form, from stream caves to vertical pit caves to short bluff caves. Many of the longer caves in the upper portion of both Little Sequatchie River valley and Pryor Cove are formed below the elevation of the valley floor. In the Little Sequatchie River valley upstream from Pocket Creek, many of these “subvalley floor caves” are air-filled for much of the year, but portions of the caves may flood for extended periods following larger precipitation events. The longest cave in Pryor Cove, Blue Spring, is a 2.46-mi-long underwater cave, formed below the valley floor, and is the only known cave in Pryor Cove formed in Monteagle Limestone. The caves and springs in the study area are also habitat for unique biota, with threatened and endangered species inhabiting spring runs at both Sequatchie Cave and Blue Spring (Thompson, 1977; U.S. Fish and Wildlife Service, 1994).

¹

Data either are not available or have limited availability owing to sensitivity concerns. Contact the Tennessee Cave Survey for more information.

This study initially focused on delineating recharge areas for some of the biologically significant springs; however, as fieldwork was conducted, results began to indicate that recharge areas could be delineated for several other significant springs. Each major spring that was studied and had a recharge area delineated is briefly described next. The locations of the springs and their geographic relation to one other are shown on figures 4 and 6.

Sequatchie Cave

Sequatchie Cave has several synonyms, namely “Owen” Spring, Blowing Spring, and Blowing Cave. The spring, located in the town of Sequatchie, Tennessee (fig. 4A), at the mouth of the Little Sequatchie River valley, issues from the entrance to a 1.25-mi-long stream cave (figs. 6 and 8) (Tennessee Cave Survey, unpub. data, 2023). The cave extends into the hillside and can be followed upstream travelling north before the cave turns westward approximately 1,700 ft from the entrance. The only accessible tributary to the main cave stream is located 400–500 ft in-cave from the entrance and enters the main stream from the southwest. The 1.25-mi-long spring branch, downstream from the cave, represents a significant portion of the habitat for the federally endangered Marstonia ogmorhaphe and for the Glyphopsyche sequatchie, which is designated as a Species of Greatest Conservation Need in Tennessee (Tennessee State Wildlife Action Plan, 2015).

“Keira” Spring

A feature known locally (and referred to herein) as “’Keira’ spring” is an ephemeral spring located on the west side of Little Sequatchie River valley between Sawmill Creek and Dixon Cove (fig. 6). The spring rises from a bluehole at the base of a Bangor Limestone bluff (fig. 9), ultimately issuing from a cave opening located 10–15 ft below the water surface when the spring is flowing. Spring flow can cease for extended periods during extremely dry conditions (fig. 9B). During these cessations in flow, the air-filled cave can be entered and its passage followed for 300 ft before reaching a sump pool (Tennessee Cave Survey, unpub. data, 2023).

Blue Spring

Blue Spring is a large spring located at the base of the Cumberland Plateau escarpment and near the mouth of Pryor Cove in the Town of Jasper, Tennessee (figs. 4 and 6). The spring discharges from a 2.46-mi-long underwater cave that is over 144 ft deep (Tennessee Cave Survey, unpub. data, 2023). The underwater cave is formed in Monteagle Limestone and spring flow rises from the underwater passage in a large, elongated sink (karst window) (fig. 10) before flowing beneath a short natural bridge to resurge from the spring outlet, which is the start of Town Creek (fig. 4). The spring is utilized as a drinking-water source by the Town of Jasper, and a treatment facility is located adjacent to the spring. The spring run is also habitat for the federally listed Marstonia ogmorhaphe (U.S. Fish and Wildlife Service, 1994; P.D. Johnson and others, Alabama Department of Conservation and Natural Resources, written commun. 2019).² Flows at Blue Spring fluctuate with hydrologic conditions and, based upon observations of stream width, depth and velocity, are estimated to exceed 100 ft³/s during flood conditions.

²

At the time of publication, data were not publicly available from U.S. Fish and Wildlife Service.

Dancing Fern Cave

Dancing Fern Cave is a large spring cave located on the east side of the Little Sequatchie River valley near its mouth where the valley merges with the Sequatchie Valley (fig. 6). A large spring issues from the cave entrance (fig. 11) into a 2,000-ft-long spring branch that drains into the Little Sequatchie River. The cave is a 1 mi-long stream cave characterized by long sections of passage that require swimming and by faults in the Bangor Limestone that are located toward the back of the cave (Tennessee Cave Survey, unpub. data, 2023). The cave is largely one main passage, although one short side passage is located 75 percent of the way into the cave system. Flows from Dancing Fern Cave are typically large, and flows are estimated to exceed 50–100 ft³/s are not uncommon, based upon observations of stream width, depth, and velocity.

Ship Cave

Ship Cave and its accompanying spring are located upstream from a spring group known locally (and referred to herein) as “the ‘Boils’” on the tip of the ridge between Indian Cove and the main Little Sequatchie River valley (fig. 6). The spring issues from a large pile of boulders (fig. 12) before flowing down a 0.45 mi-long spring branch to join the stream from Indian Cove and then flowing 0.4 mi farther to the confluence with the Little Sequatchie River. The spring creates the first perennial flow in the Little Sequatchie River downstream from the major stream sinks and upstream from the “Boils.” The water that discharges from Ship Cave spring can also be followed underground for approximately 0.5 mi. through large, voluminous cave passages that generally trend up the north side of Indian Cove (Tennessee Cave Survey, unpub. data, 2023). Although the spring primarily discharges from one outlet, additional overflow springs are activated during high flow that are near the main outlet along the spring branch. One additional perennial spring, known locally (and referred to herein) as “’Indian Ship’ spring,” is part of the Ship Cave hydrologic system and is also near an ephemeral overflow spring.

The “Boils”

The “Boils” (fig. 13) is a multiple-outlet distributary spring system located on the west side of the Little Sequatchie River valley, approximately 0.5 mi downstream from the mouth of Indian Cove (fig. 6). The spring group varies in the number of spring outlets depending on hydrologic conditions, with only 1 or 2 outlets during low-flow conditions and greater than 11 outlets during high-flow conditions. Although one major spring “boils” up from the streambed out of cobbles and sediment (site no. 005 on table 1 and fig. 14C), the majority of the springs within this system discharge from outlets along the spring branch banks. Springs discharge from both sides of the spring branch as well as from the streambed sediments. During base-flow periods, the “Boils” provides the bulk of the flow observed in the Little Sequatchie River downstream from the springs. The river is nearly dry upstream from the “Boils,” except for the stream flow from Ship Cave that joins the river just upstream from the “Boils” spring branch. Discharge from the spring group is estimated to exceed 100 ft³/s during high-flow periods and drop below 10 ft³/s during low-flow periods, based upon observations of stream width, depth and velocity.

Table 1.    

Monitoring sites used in this study.

[Elevation data from the 3DEP lidar explorer (U.S. Geological Survey, 2023a). Elevation data referenced to the North American Vertical Datum of 1988. no., number; ft, foot; CV, cave stream; SP, spring; ST, surface stream; Hwy, Highway]

Table 1.    Monitoring sites used in this study.

Site no.	Site name	Site type	Elevation (ft)
001	Sequatchie Cave stream, just inside dripline	CV	650
002	Owen Spring Branch	SP	650
003	Wine Cave Spring	CV	674
004	REW large spring at The Boils	SP	770
005	Main boil at The Boils	SP	770
006	REW spring above The Boils	SP	770
007	Double springs at head of channel above The Boils	SP	770
008	LEW spring downstream from The Boils at campsite	SP	770
009	REW spring under large roots downstream from The Boils	SP	770
010	REW spring downstream from 009, below broken tree	SP	770
011	Ship Cave	CV	800
012	Indian Ship Spring	SP	800
013	Indian Cove Creek	ST	764
014	Little Sequatchie River upstream from Ship Cave	ST	746
015	Keira Spring	SP	674
016	Dancing Fern Cave stream	CV	670
017	Ellis Spring	SP	647
018	Jasper Blue Spring	SP	618
019	Little Sequatchie River downstream from Hwy 28	ST	673
020	Little Sequatchie River upstream from Sequatchie Cove Farm	ST	604
021	Clear Spring	SP	651
022	Dixon Cove at Coppinger Cove Road	ST	681
023	Sawmill Creek at Coppinger Cove Road	ST	651
024	Owen Spring Branch backup	SP	650
025	Ship Cave Spring	SP	790
026	New site at top of Boils Spring Group	SP	770
027	Ship Cave Spring Branch	SP	785
028	Druin Cave Spring	SP	636
029	Pryor Cave Spring	SP	643
030	Mandys Cave	SP	948
031	The Boils low flow failsafe	SP	739
032	Grays Creek Spring	SP	982
033	Grays Creek karst window	ST	1023
034	Niagara Spring	SP	737
035	Town Creek at Highway 72	ST	610
036	Standifer Branch	ST	620
037	Glover Branch	ST	646
038	Simpson Road Spring	SP	609
039	North Pryor Cove Spring	SP	775

Previous Dye Tracing Research

Some previous unpublished groundwater tracing had been conducted in the Little Sequatchie River valley and Pryor Cove by USGS personnel, environmental consultants, and volunteer researchers (Miller and Hourigan, 2023). In 1990, a sinkhole known locally (and referred to herein) as “’Asher’ sink” (fig. 6) opened up in the Town of Jasper following some construction activities. Rhodamine WT dye was injected into the sinkhole feature and flushed with nearly 50,000 gallons (gal) of water from a nearby fire hydrant. The dye was detected at nearby Blue Spring (site no. 018 on table 1 and fig. 14A) 8 days postinjection, via water samples, and continued to be at detectable levels for an additional 4 weeks. This trace is shown later in the “Round 4, August 2022” section of this report, and the injection is shown with an injection ID of JC-MB. Later, cave divers working in Blue Spring were able to document and map where flow from this sinkhole enters the cave system (Jason Richards, written commun., 2023).

Tracing was conducted in 2015 in the lower portion of the Little Sequatchie River valley and in tributaries near the mouth of the valley by an environmental consultant (Richard D. Martin, Griggs and Maloney Incorporated, unpub. data, 2015).³ This work delineated two recharge basins for Sequatchie Cave, referred to respectively as a “conduit recharge basin” and a “percolation recharge basin.” This work consisted of four different dye injections and was conducted on behalf of Tennessee Wildlife Resources Association. The total recharge area delineated by this work was approximately 5 mi² and primarily encompassed the drainage areas of Sawmill Creek and Water Fall Cove (Richard D. Martin, Griggs and Maloney Incorporated, unpub. data, 2015).

In 2017, a dye injection was conducted by volunteers as a demonstration for a cave biology course from Berry College. The dye injection was conducted in an unnamed tributary to the Little Sequatchie River, located between Sawmill Creek and Dixon Cove. The stream sank entirely below a large waterfall and rhodamine WT was injected into the stream immediately upstream from the sinkpoint. Dye from the injection was expected to flow to nearby “Keira” spring (site no. 015 on table 1 and fig. 14A) only 0.5 mi away; however, the dye flowed to Sequatchie Cave nearly 2 mi to the south of the injection site (Miller and Hourigan, 2023). The dye arrived at Sequatchie Cave within 1 week postinjection. This trace is shown in the “Round 1, January 2022” section of this report and has an injection ID of BM-BH.

Round 1, January 2022

The first round of dye injections was conducted in January 2022, with the injections occurring on January 6, 7, and 19. These injections focused on determining the flow paths for some of the upstream swallets on the Little Sequatchie River as well as investigating if some lower valley tributaries hydrologically connect to Sequatchie Cave. The first dye injection (I-1), was conducted on January 6, using approximately 2 pounds (lb) of fluorescein injected into the main stem of the Little Sequatchie River near a major swallet feature known locally (and referred to herein) as “Party Well” near the contact of the Pennington Formation and Bangor Limestone (Pennington-Bangor contact) (fig. 15). On the day of the injection, flow in the Little Sequatchie River at the sinkpoint was estimated to be approximately 200 ft³/s. Although the main “Party Well” swallet was overtopped, all of the streamflow sank underground within 300 ft of the injection location. The second dye injection (I-2) was conducted the following day at a large sinking waterfall in Lick Branch, where a spring, discharging from the Pennington Formation, sinks immediately into a 90-ft-deep swallet in the Bangor Limestone. At this location, 2 lb of sulphorhodamine b was injected directly below a spring discharging from the Pennington Formation (fig. 16). The third and final injection for round 1 (I-3) was conducted on January 19 at a subterranean waterfall located inside a cave in Water Fall Cove (fig. 14A). One lb of eosine oj was injected immediately downstream from the waterfall and upstream from the point beyond which the stream cannot be humanly accessed.

Results from the round 1 dye injections helped to set some initial boundaries for the various recharge areas within the Little Sequatchie River valley. Injection I-1 was positively traced to all the monitored sites at the “Boils” (site nos. 004–010, figs. 14C and 17) and the subsequent sites located downstream from the “Boils” (site nos. 019 and 020, fig. 14A). Dye from this injection appears to have stayed subsurface the entire distance from the sinkpoint to the “Boils.” At a length of 7.8 mi, this trace is one of the longer documented dye traces in Tennessee, on par with major dye traces documented on the Cumberland Plateau (Crawford, 1989; Miller and others, 2019). Dye from injection I-2 was traced to “Keira” spring (site no. 015) (figs. 14A and 18) as well as a site downstream from “Keira” spring along the main Little Sequatchie River (site no. 019, fig. 14A). Dye at injection site I-3 was positively traced to Sequatchie Cave (site no. 001, figs. 14A and 18), creating a southwestern boundary for the Sequatchie Cave recharge area.

Round 2, March 2022

The second round of dye injections was conducted during March 28–29, 2022, and focused on determining flow paths for water that sinks in two major western tributaries to Little Sequatchie River, namely the main stream in Indian Cove and a side tributary to Dixon Creek, as well as two eastern tributaries, Pocket Creek and Alum Creek (fig. 6). Injection I-4 was conducted on March 28 in Indian Cove at a major stream sink located near the Pennington-Bangor contact (fig. 5A). Two lb of powdered fluorescein was injected into the main stream in Indian Cove, which sinks entirely into the open swallet at the base of a small bluff (fig. 19). The streambed from this point down to the Little Sequatchie River remains dry most of the year except following major rain events. Injection I-5 was conducted in a southern tributary to Dixon Cove with the intent to refine a possible recharge area boundary between “Keira” spring and Sequatchie Cave. In this unnamed tributary, 2 lb of sulphorhodamine b was injected into the flow of a small spring, which sank 20 ft downstream from the resurgence. Injection I-6 was conducted March 29 at a major swallet for Pocket Creek, where the entire stream sinks near the Pennington-Bangor contact (fig. 20). Following this injection, the dye was observed in the stream of a short cave formed in the Bangor Limestone approximately 200 ft downstream from the injection location. The last injection in round 2 (I-7) was conducted the same day in Alum Cove near the Pennington-Bangor contact, where 2 lb of powdered eosine oj was injected into a small sinking stream.

Results from the second round of injections delineated what would ultimately become two of the larger recharge areas in the study area and provided insight into karst resurgences with shared recharge areas in western tributaries to the Little Sequatchie River. Injection I-4 in Indian Cove was positively traced to both Ship Cave (site no. 011) and the smaller adjacent “Indian Ship” spring (site no. 012) (figs. 14B and 17). “Indian Ship” spring appears to be a perennial overflow spring for the stream in Ship Cave, and other ephemeral springs have been observed near “Indian Ship” spring following rain events. Injection I-5 in the small sinking spring system in Dixon Cove was positively traced to both “Keira” spring and Sequatchie Cave (fig. 18), arriving at “Keira” spring within 3 days postinjection and at Sequatchie Cave, via the packet in Owen Spring (site no. 002, fig. 14A), within 2 weeks postinjection. Injections I-6 and I-7, from Pocket Creek and Alum Cove respectively, were both positively traced to Dancing Fern Cave (site no. 016) (figs. 14A and 18). The trace from Pocket Creek to Dancing Fern Cave had one of the fastest traveltimes for the longer traces in the project, with dye arrival less than 48 hours postinjection. These traces establish the Dancing Fern Cave recharge area as the second largest recharge area in the Little Sequatchie River and are notable for the fact that they appear to have stayed subsurface from sink to spring, which in the case of Pocket Creek is slightly less than 7 mi.

Round 3, June 2022

The third round of dye injections was conducted in early June 2022 and focused on upper tributaries to the Little Sequatchie River while defining recharge area boundaries between Dancing Fern Cave and the “Boils.” Additionally, an injection in Dixon Cove would help delineate a recharge area for Sequatchie Cave and might help further explain the shared recharge areas between “Keira” spring and Sequatchie Cave. Injection I-8, conducted on June 1, was located in Grays Creek, approximately 1,000 ft upstream from the Pennington-Bangor contact (fig. 5A). Prior to injection, two springs were discovered (Grays Creek Spring and Grays Creek karst window, fig. 18) and added to the monitoring network (site nos. 032 and 033, fig. 14A). Four lb of eosine oj was injected into a pool, and the small stream flowing out of the pool sank entirely within 20–30 ft. Injections I-9 to I-11 were conducted the following day, June 2. Injection I-9 was conducted in the main stream in Peter Cave Cove approximately 1,500 ft upstream from the Pennington-Bangor contact and near the namesake for the cove, a large shelter cave utilized by early settlers of the cove in the late 1800s through the early 1900s (Sequachee Valley News, 1900). A total of 3 lb of rhodamine wt was injected into the small stream immediately upstream from the discrete sinkpoint (fig. 21). Injection I-10 directly followed the Peter Cave Cove injection and was located in Mill Creek on the west side of the Little Sequatchie River valley (fig. 17). At this location, approximately 1,050 ft upstream from the Pennington-Bangor contact, a total of 2 lb of powdered fluorescein was injected into the flowing stream just above the sinkpoint. The final injection for round 3, namely I-11, was conducted in Dixon Cove where the main stream in the cove sank at a site 0.5 mi upstream from the “normal” sinkpoint for the Dixon Cove stream (fig. 18). At this site, 2 lb of sulphorhodamine b was injected into the stream and sank entirely 1–200 ft downstream from the injection site (fig. 22). All of the injection sites for round 3 were located upstream from the Pennington-Bangor contact or upstream from the perennial sinkpoints for the streams. This behavior may have occurred because round 3 was conducted in June and there had been an overall lack of rain for several weeks, causing the sinkpoints to migrate upstream as groundwater levels dropped in the system.

Results from round 3 helped to further delineate a recharge area for the “Boils” and confirmed an additional shared recharge area between “Keira” spring and Sequatchie Cave. The dye injections in Grays Creek, Peter Cave Cove, and Mill Creek (I-8 through I-10) were positively traced to the “Boils” (site nos. 004–010, 026, figs. 14C and 17) and to the subsequent sites located downstream (site nos. 019 and 020, fig. 14A). This delineated a boundary between the “Boils” and the Dancing Fern Cave recharge areas, located along the drainage divide between Grays Creek and Pocket Creek. The dye injection in the main stream of Dixon Cove (I-11) was positively traced to both “Keira” spring (site no. 015) and Sequatchie Cave via the packet in Owen Spring (site no. 002) (figs. 14A and 18). Similar to the results from round 2, the dye arrived first at “Keira” spring, within 1 day for round 3, and then several days later at Sequatchie Cave, with 6 days for round 3. The faster traveltimes may be due to a short burst of heavy rainfall approximately 30 minutes after the dye injection in Dixon Cove. A light pink color was visible in the water at “Keira” spring the next morning, indicating sulphorhodamine b, as confirmed later in the charcoal packets retrieved from the site.

Round 4, August 2022

The fourth round of dye injections was conducted in August 2022 and focused on defining a southern boundary for the “Boils” recharge area, closing spatial gaps in the Sequatchie Cave and “Keira” spring recharge areas, and beginning work to delineate a recharge area for Blue Spring. Injection I-12 was conducted in Sawmill Creek, a large hollow located between Water Fall Cove, to the south, and Dixon Cove, to the north. During wetter time periods, Sawmill Creek flows overland to the Little Sequatchie River; however, in summertime, all of the flow sinks near the Pennington-Bangor contact. On August 1, a sinking stream was located in Sawmill Creek approximately 300 ft upstream from the mapped Pennington-Bangor contact and 1 lb of powdered fluorescein was injected into the pool at the sinkpoint (fig. 23). Following injection 1-12 in Sawmill Creek, injection I-13 was conducted the same day in a southern tributary to Dixon Cove between injections I-5 and I-11. Injection I-13 utilized 1 lb of sulphorhodamine b, which was injected into the sinking stream immediately above the sinkpoint for the small stream. Because of accessibility issues and flow conditions, the third and fourth injections (I-14 and I-15) were not conducted until 4 weeks later on August 29; this also allowed for dye from I-12 and I-13 to be recovered by charcoal packets and eliminated any possible masking or interference from dyes with similar wavelengths. On August 29, injection I-14 was conducted at a swallet on the western flank of the main Little Sequatchie River valley. At this site, 2 lb of rhodamine wt was injected into the small stream, which immediately sank into a small cave-like feature (fig. 24). Following this injection, the final injection (I-15) for round 4 was conducted into the eastern arm of the northern fork of Pryor Cove located to the west of the lower Little Sequatchie River valley near the Town of Jasper. Two lb of eosine oj was injected into a small flowing stream, and although it did not immediately sink at the injection site, the stream is known to sink entirely within a short distance downstream.

Round 4 results helped to extend the “Boils” recharge area, further delineated Sequatchie Cave and “Keira” spring recharge areas, and aided in the delineation of the first portions of a recharge area for Blue Spring. The dye injection in Sawmill Creek (I-12) was positively traced to Sequatchie Cave, arriving at the Owen Spring packet (site no. 002) within 1 week of the dye injection, and was not detected at “Keira” spring. The dye injection at Dixon Cove (I-13) was positively traced to both “Keira” spring (site no. 015) and Sequatchie Cave (via Owen Spring [site no. 002]), with the dye detected on packets collected 1 week postinjection (figs. 14A and 18). The dye injected into a swallet along the Little Sequatchie River upstream from Pocket Creek (I-14) was detected at the “Boils” spring group (site nos. 004–010, 026, fig. 14B) on packets collected 16 days postinjection, although the dye had not arrived at packets collected 3 days postinjection. This trace extends the recharge area for the “Boils” farther downstream and added 8.7 mi² to the total area (fig. 17). The eosine oj from the injection site in the northern fork of Pryor Cove (I-15) was positively traced to Blue Spring (site no. 018) within 9 days of the dye injections (figs. 14A and 25); however, packets collected 3 days postinjection showed no dye. Aside from the JC-MB dye injection performed in 1990 from “Asher” sink (fig. 6), discussed in the “Previous Dye Tracing Research” section, this trace would represent the first dye injection positively traced to Blue Spring. This trace indicated that the northern fork of Pryor Cove is a significant recharge source for the spring because the trace was located far up the cove, well beyond the mapped end of the water-filled cave (fig. 25).

Round 5, November 2022

The fifth round of dye injections was conducted on November 1, 2022, to continue delineating a recharge area for Blue Spring. Building off the positive trace from injection I-15, the dye injection locations included sites in western fork of Pryor Cove and in the western arm of the northern fork of Pryor Cove. The current mapped end of the underwater cave at Blue Spring is located under West Fork Pryor Cove Branch; thus, this particular stream seemed to be a likely source of recharge to the spring. Hydrologic conditions during round 5 were similar to those during round 4 because the area had not received rainfall in a few weeks, and the stream sinkpoints appear to have migrated upvalley from their normal location near the Pennington-Bangor contact. Injections I-16 and I-17 were both conducted in the upper half of the western fork of Pryor Cove. Injection I-16 was conducted in an unnamed southern tributary to West Fork Pryor Cove Branch, approximately 600 ft upstream from the mapped Pennington-Bangor contact, near a small spring located on the streambank of the small channel. One lb of sulphorhodamine b was injected into the small spring and sank immediately below the outlet (fig. 26). Injection I-17 was conducted in the main stream channel of West Fork Pryor Cove Branch, 0.8 mi upstream from the confluence with the tributary where injection I-16 was conducted. One lb of eosine oj was injected at injection site I-17, where the small stream sinks at the Pennington-Bangor contact. Injection site I-18 was located in the western arm of the northern fork of Pryor Cove, 0.5 mi upstream from the mapped Pennington-Bangor contact (fig. 25). A hillside spring discharges from the lower Pennington Formation and flows a short distance overland (~500 ft) before sinking into an open swallet on the side of the main stream channel where 1 lb of fluorescein was injected (fig. 27).

Results from round 5 helped to further refine the Blue Spring recharge area and significantly increased the size of the delineated recharge area. All three injections (I-16 to I-18) from round 5 were positively traced to Blue Spring (site no. 018, figs. 14A and 25). The dyes from the West Fork Pryor Cove Branch injections (I-16 and I-17) were detected on packets collected on November 16, indicating dye arrived between 10 and 15 days postinjection. Packets collected on November 30 indicated that dye from North Fork Pryor Cove Branch (injection I-18) had arrived between 15 and 29 days postinjection. Several rainfall events in November each exceeded 1 inch of precipitation (Community Collaborative Rain, Hail & Snow Network [Cocorahs], 2022); however, the dry conditions at the time of dye injection may have caused the traveltime (from sinkpoint to spring) to be longer than would have been during wetter time periods.

Round 6, December 2022

The sixth round was an intermediate round, consisting of a single dye injection conducted in a portion of the study area that previously had been confirmed to be hydrologically disconnected from any of the karst systems in Pryor Cove. This is important to note, as conducting traces can be problematic if all dyes have not been accounted for or sufficient precipitation has not flushed through a system, returning the system to preinjection background levels of fluorescence. The round 6 injection (I-19) was conducted at a large swallet adjacent to the “Boils” spring group. Here, a surface stream sinks into a short cave formed in the Pennington Formation before resurging immediately above a large swallet formed in the Bangor Limestone. The injection was conducted to determine whether this system, directly adjacent to the “Boils,” might resurge from a subset of the eight monitored springs. On December 12, 2022, 1 lb of fluorescein was injected upstream from the stream sink and was observed to quickly pass through the short cave segment and emerge into the large swallet (fig. 28). The USGS team then immediately made their way downslope to the “Boils” spring group to observe the dye emerging approximately 3–400 ft downstream from the “Boils” from three small springs on the western bank of the spring channel (figs. 17 and 29). The springs (site name: Niagara spring, site no. 034, table 1, fig. 14C) had eluded previous documentation because some resurgences were located under tree roots and others obscured by boulders. The time of travel was the shortest of any trace for the project, with the dye arriving at the spring resurgence less than 30 minutes postinjection. An average velocity for this trace would be approximately 1,400 feet per hour, nearly 60 percent faster than the trace from Pocket Creek to Dancing Fern Cave.

Round 7, January 2023

The seventh round of dye injections was conducted on January 9, 2023, in several unnamed tributaries to North Fork Pryor Cove Branch as well as one unnamed tributary to main stem of Pryor Cove Branch, with the purpose of further delineating a recharge area for Blue Spring. The first injection for round 7 (I-20) was conducted in an unnamed tributary to Pryor Cove Branch near Castle Rock Point on the northeast side of Pryor Cove, where the stream flows across the Pennsylvanian siliciclastic strata and cascades over a series of waterfalls before sinking into the subsurface at the top of the Pennington Formation. One lb of fluorescein was injected into the flowing stream approximately 80 ft upstream from the sinkpoint and above an 18-ft waterfall (fig. 30). Injection I-21 was conducted in a small, unnamed ravine with a flowing stream on the western side of the northern fork of Pryor Cove. At the injection site, 1 lb of eosine oj was injected into the flowing stream immediately upstream from where the small stream sank into a closed depression (sinkhole) with an open throat near the Pennington-Bangor contact. The third and final dye injection for round 7 (I-22), was located in an unnamed western tributary to North Fork Pryor Cove Branch, where another stream flowed off the Pennsylvanian siliciclastic strata and then sank at the top of the Pennington Formation, similar to the geologic conditions at injection site I-20. One lb of sulphorhodamine b was injected approximately 20 ft upslope from the sinkpoint, which was covered by sandstone talus (fig. 31).

Results from the seventh round of dye injections helped to further delineate the Blue Spring recharge area and defined part of its eastern boundary. Dye from injection I-20 was positively traced to Pryor Cave Spring (site no. 029) located near the head of Standifer Branch (figs. 14A and 25). The dye arrived at the site within 24 hours, and dye was still visible when packets were retrieved the next day, 22 hours postinjection. This trace further delineated an eastern boundary for the Blue Spring recharge area, which also shares a boundary with the recharge area for Sequatchie Cave. Dye from injection I-21 was detected at Blue Spring (site no. 018), arriving within 3 days postinjection, adding further confirmation that the western side of the northern fork of Pryor Cove provided recharge to the spring (figs. 14A and 25). The same dye from I-21 was also detected at a nearby spring monitoring site (site name: North Fork Pryor Cove spring, site no. 039), a valley floor spring downslope from the injection site (figs.14A and 25). Dye from injection I-22 was not detected at any of the monitoring sites established for the project.

Round 8, March 2023

The eighth round of dye injections was conducted on March 16, 2023, and focused on portions of the main streams of the North and West Fork Pryor Cove Branches, located in the Town of Jasper and one last unnamed tributary to North Fork Pryor Cove Branch near Highway 41. These injections aimed to show that portions of the Town of Jasper were hydrologically connected to Blue Spring and to extend the recharge for Blue Spring to the east and downstream toward the confluence of the two branches of the cove. The first injection of round 8 (I-23) was located in an eastern unnamed tributary to North Fork Pryor Cove Branch, where a stream sank into the lower Pennington Formation immediately upstream from the stream crossing at Highway 41. One lb of powdered fluorescein was injected into the small stream approximately 20 ft upstream from the sinkpoint for the stream (fig. 32). The second injection for round 8 (I-24) was conducted at the main stream channel of North Fork Pryor Cove Branch where it passes alongside North Pryor Cove Road. At this location the stream was completely dry, so a fire tanker truck provided by the Jasper Volunteer Fire Department was used to flush the 0.5 lb of sulphorhodamine b with approximately 3,500 gal of water postinjection. Following this injection, the third dye injection for round 8 (I-25) was conducted into the West Fork Pryor Cove Branch where the stream crosses Pryor Cove Road. Similar to the conditions present at I-24, the streambed was completely dry and a fire tanker truck flushed the 0.5 lb of eosine oj with 3,500 gal of water following the injection (fig. 33). Overall, smaller amounts of dye were used for injections I-24 and I-25 because of the close proximity of the dye injection locations to the mapped passages in Blue Spring.

Round 8 helped to extend the recharge area in the western fork of Pryor Cove and defined an eastern boundary for the Blue Spring recharge area. The dye from injection I-23 was detected on packets at Blue Spring (site no. 018) gathered 4 days postinjection and for several weeks thereafter (figs. 14A and 25). Eosine oj, from injection I-25, was detected on packets at Blue Spring gathered 8 days postinjection, but was not detected on the 4-day packets retrieved on March 20, 2023 (fig. 25). Sulphorhodamine b from injection I-24 was detected at Blue Spring on packets retrieved on April 10, indicating the dye took between 8 and 25 days to arrive at the spring. The reason for the delay on the arrival of the dye from injection I-24 compared to that of I-23 and I-25 is unknown and may have been due to the drier conditions at this site relative to the other two, the lack of a flowing stream, and the distance from this site to the spring conduit.

Recharge Areas Delineated from Dye Tracing

Six recharge areas were delineated within the Little Sequatchie and Pryor Cove drainage areas using the traces and results from the eight rounds of dye injections. Recharge areas were delineated for Sequatchie Cave, “Keira” spring, Dancing Fern Cave, Ship Cave, the “Boils,” and Blue Spring. The recharge areas range in size from 7.25 to over 65.2 mi² (table 4). Each of the delineated recharge areas is described in the following sections.

Sequatchie Cave Recharge Area

The recharge area for Sequatchie Cave covers approximately 11.76 mi² and is primarily located in three different drainage areas: Dixon Cove, Sawmill Creek, and Water Fall Cove (fig. 34). A total of six traces hydrologically connected various portions of the recharge area to the cave system. Several dye injections in Dixon Cove were also traced to “Keira” spring, with the dye from injections in Dixon Cove typically appearing at “Keira” spring within 1–3 days and then at Sequatchie Cave 1–2 weeks postinjection. These results indicate that Sequatchie Cave shares approximately 51 percent of its recharge area with “Keira” spring. This finding may help explain why the streamflows observed at “Keira” spring were less than those at Sequatchie Cave, even ceasing during periods of drought. Land use-land cover data indicate that forest cover dominates the recharge area (table 5), with 87.75 percent of the recharge area being either deciduous forest, evergreen forest, or mixed forest (Dewitz and USGS, 2021). Developed land in the recharge area accounts for less the 2.5 percent of the overall recharge land use, and much of this development consists of roads and smaller subdivisions along U.S. Highway 41 and on top of the Cumberland Plateau between Dixon Cove and Sawmill Creek.

Table 5.    

Land use-land cover percentages in the Sequatchie Cave recharge area.

[Data from Dewitz and U.S. Geological Survey (2021). Totals provided for selected land use-land cover categories discussed in text]

Table 5.    Land use-land cover percentages in the Sequatchie Cave recharge area.

Land use-land cover category	Percentage of recharge area	Area (square miles)
Open water	0.30	0.04
Developed, open space	2.07	0.24
Developed, low intensity	0.32	0.04
Developed, medium intensity	0.03	0.00
Deciduous forest	76.87	9.04
Evergreen forest	7.07	0.83
Mixed forest	3.81	0.45
Shrub/scrub	2.46	0.29
Herbaceous	3.81	0.45
Hay/pasture	3.22	0.38
Woody wetlands	0.01	0.00
Emergent herbaceous wetlands	0.04	0.00
Total: Developed land	2.41	0.28
Total: Forest	87.75	10.32

Blue Spring Recharge Area

The recharge area for Blue Spring covers 11.68 mi² and includes both the northern and western arms of Pryor Cove (fig. 35). Multiple dye injections conducted throughout both of the Pryor Cove drainage areas were positively traced to Blue Spring (fig. 25). Several of these traces were relatively rapid, arriving at the spring within a few days postinjection. Injections I-21 and I-23 were both detected at Blue Spring within 3–4 days postinjection. Dye injections in the Blue Spring recharge area were also conducted across a range of geologic units. Streams were commonly found to sink at the Pennington-Bangor contact; however, several sinking streams in the area sank at the top of the Pennington Formation, several hundred feet above the contact with the underlying Bangor Limestone. Several dye traces in the Sequatchie Cave-”Keira” springs recharge area helped to define an eastern boundary for the recharge area, as did the trace from injection I-20 near Castle Rock Point to Pryor Cave Spring (029). Traces in the Blue Spring recharge area indicate that sinking streams, both high on the flanks of the Cumberland Plateau and at the main valley sinkpoints, contribute recharge to this large spring. The final traces from the lower part of the Pryor Cove northern and western arms (I-24 and I-25) also show that portions of the Town of Jasper are located within the recharge area of the spring. Land use-land cover within the recharge area is similar in overall category percentages to the other spring recharge areas in the study area. Forestland covers 81.7 percent of the overall recharge area, and developed or barren land covers 3.2 percent (0.37 mi²; table 7) (Dewitz and USGS, 2021). The developed and barren land portions are largely located in the lower portions of the recharge area, in the neighborhoods along the North and West Fork Pryor Cove Branches. There are some larger pasture lands in upper portions of both coves above the rim of the Cumberland Plateau. Although the percentages of developed or barren land are relatively low, these areas are concentrated directly adjacent to the mapped spring conduit of Blue Spring; thus, these areas may have a larger influence on water quality than other land use-land cover types because of their proximity to the spring.

Table 7.    

Land use-land cover percentages in the Blue Spring recharge area.

[Data from Dewitz and U.S. Geological Survey (2021). Totals provided for selected land use-land cover categories discussed in text]

Table 7.    Land use-land cover percentages in the Blue Spring recharge area.

Land use-land cover category	Percentage of recharge area	Area (square miles)
Open water	0.38	0.04
Developed, open space	1.97	0.23
Developed, low intensity	1.06	0.12
Developed, medium intensity	0.05	0.01
Barren land	0.12	0.01
Deciduous forest	69.52	8.12
Evergreen forest	9.06	1.06
Mixed forest	3.08	0.36
Shrub/scrub	3.89	0.45
Herbaceous	2.13	0.25
Hay/pasture	8.70	1.02
Woody wetlands	0.01	0.00
Total: Developed land and barren land	3.21	0.37
Total: Forest	81.67	9.53

Dancing Fern Cave Recharge Area

The recharge area for Dancing Fern Cave was delineated from two positive traces in Pocket Creek (injection I-6) and Alum Cove (injection I-7). The Dancing Fern Cave recharge area covers 16.75 mi² and consists of portions of each cove upstream from the injection sites (fig. 36). Similar to other settings in the Little Sequatchie River valley, there are valley segments downstream from the injection sites that may contribute additional recharge to Dancing Fern Cave, although in the case of Pocket Creek, it likely would contribute recharge to the “Boils” shortly thereafter, and in Alum Cove, these downstream segments may flow directly to the Little Sequatchie River. Land use and land cover data for the recharge area (table 8) indicate that forest covers 79.9 percent of the total area, with the majority of the forest cover being deciduous. Developed land, which includes barren land as well, totals 4.3 percent (0.67 mi²) of the recharge area (Dewitz and USGS, 2021). The trace from Pocket Creek to Dancing Fern Cave (fig. 18) was noteworthy both in the length of the trace (6.74 mi) and the short period of time for the dye to arrive at the spring (48.8 hours). The trace length from Pocket Creek to Dancing Fern Cave makes this trace one of the longer dye traces in the State, comparable to those for other major karst systems on the Cumberland Plateau (Crawford, 1989; Miller and others, 2019). The field-measured groundwater velocities for this trace were 749 feet per hour, which would only occur in an aquifer with a very high hydraulic conductivity (Freeze and Cherry, 1979; White, 1988), and indicate significant portions of conduit flow between the sinkpoint in Pocket Creek cove and Dancing Fern Cave.

Table 8.    

Land use-land cover percentages in the Dancing Fern Cave recharge area.

[Data from Dewitz and U.S. Geological Survey (2021). Totals provided for selected land use-land cover categories discussed in text]

Table 8.    Land use-land cover percentages in the Dancing Fern Cave recharge area.

Land use-land cover category	Percentage of recharge area	Area (square miles)
Open water	0.07	0.01
Developed, open space	2.83	0.47
Developed, low intensity	1.13	0.19
Developed, medium intensity	0.24	0.04
Developed, high intensity	0.01	0.00
Barren land	0.09	0.02
Deciduous forest	65.96	11.05
Evergreen forest	4.74	0.79
Mixed forest	9.23	1.55
Shrub/scrub	1.02	0.17
Herbaceous	0.70	0.12
Hay/pasture	2.73	0.46
Woody wetlands	0.00	0.00
Emergent herbaceous wetlands	0.01	0.00
Total: Developed land and barren land	4.30	0.72
Total: Forest	79.93	13.39

Ship Cave Recharge Area

The Ship Cave recharge area is located entirely within the Indian Cove drainage area and extends upstream from the injection site (I-4) to encompass a total of 10.18 mi² (fig. 37). This recharge area is shared with “Indian Ship” spring, which appears to be a perennial overflow spring on the basis of trace results and observations following rainfall events. The streambed within Indian Cove is typically dry downstream from the injection site to the confluence with the Little Sequatchie River and it is likely that portions of the valley, in this reach, drain to Ship Cave and “Indian Ship” spring. However, the lack of flowing water precluded the ability to conduct further dye injections, and no additional traces were conducted in the valley to confirm any further connections. Land use within the Ship Cave recharge area (table 9) is largely forest, composing 77 percent of the recharge area, with deciduous forest being the predominant forest type. Developed or barren land constitutes 1.47 percent of the recharge area (Dewitz and USGS, 2021). Currently, the Ship Cave recharge represents the fifth largest recharge in the Little Sequatchie River valley.

Table 9.    

Land use-land cover percentages in the Ship Cave recharge area.

[Data from Dewitz and U.S. Geological Survey (2021). Totals provided for selected land use-land cover categories discussed in text]

Table 9.    Land use-land cover percentages in the Ship Cave recharge area.

Land use-land cover category	Percentage of recharge area	Area (square miles)
Developed, open space	1.25	0.13
Developed, low intensity	0.15	0.02
Developed, medium intensity	0.03	0.00
Developed, high intensity	0.01	0.00
Barren land	0.03	0.00
Deciduous forest	62.95	6.41
Evergreen forest	7.05	0.72
Mixed forest	7.01	0.71
Shrub/scrub	13.16	1.34
Herbaceous	3.56	0.36
Hay/pasture	4.77	0.49
Woody wetlands	0.05	0.01
Total: Developed land and barren land	1.47	0.15
Total: Forest	77.00	7.84

The “Boils” Recharge Area

The “Boils” recharge area is the largest recharge area in the Little Sequatchie watershed, covering 65.2 mi², and includes the upper portion of the Little Sequatchie River as well as several major tributaries, including Mill Creek, Peter Cove Creek, and Grays Creek (fig. 38). Five total traces were used in the delineation of the recharge area, and several of these traces have lengths exceeding 5 mi, including the trace from injection I-1, which is one of the longer dye traces in Tennessee, with a length of 7.65 mi. The recharge area extends to the watershed divide with both the Collins River, to the north, and the Elk River, to the west (fig. 4B). Previous dye tracing research has shown that streams sinking in the headwater drainages of the Collins River watershed (consisting of the upper portion of the Collins River valley, Big Creek valley, and Savage Gulf) eventually flow to the large feature known locally (and referred to herein) as “’Grundy Big’ spring” near Tarlton, Tennessee (fig. 4B) (Miller and others, 2019). Thus, the recharge area for the “Boils” shares a boundary with the “Grundy Big” spring recharge area. The “Boils” recharge area is one of the only recharge areas in the Little Sequatchie drainage area that encompasses any towns or communities. The communities of Coalmont, Tracy City, and Gruetli-Laager are all partially located within the northern boundaries of the “Boils” recharge area (fig. 38). Land use-land cover within the “Boils” recharge area is dominated by forest cover, with approximately 87 percent of the recharge covered by forest (table 10) (Dewitz and USGS, 2021). Developed land or barren land covers 3.5 percent or approximately 2.28 mi² of the recharge area and is the result of several communities that are, at least partially, located within the recharge area boundary.

Table 10.    

Land use-land cover percentages in the “Boils” recharge area.

[Data from Dewitz and U.S. Geological Survey (2021). Totals provided for selected land use-land cover categories discussed in text]

Table 10.    Land use-land cover percentages in the “Boils” recharge area.

Land use-land cover category	Percentage of recharge area	Area (square miles)
Open water	0.08	0.06
Developed, open space	2.45	1.60
Developed, low intensity	0.53	0.35
Developed, medium intensity	0.19	0.13
Developed, high intensity	0.03	0.02
Barren land	0.29	0.19
Deciduous forest	67.09	43.76
Evergreen forest	11.28	7.36
Mixed forest	8.41	5.49
Shrub/scrub	4.07	2.66
Herbaceous	1.78	1.16
Hay/pasture	3.73	2.43
Woody wetlands	0.03	0.02
Emergent herbaceous wetlands	0.02	0.01
Total: Developed land and barren land	3.50	2.28
Total: Forest	86.78	56.58

Land Use–Land Cover Patterns and Possible Threats to Water Quality and Quantity

Within the study area are a number of land use-land cover types and other human activities that pose potential risks to water quality and quantity in the springs, caves, and stream ecosystems. Urban development, although relatively sparse in some of the recharge areas, does pose potential threats from roads, residential development, wastewater effluent discharges, and stormwater runoff from impervious surfaces. These areas are primarily in the higher elevation portions of the recharge areas, with the exception of the Town of Jasper, which overlaps with the recharge area for Blue Spring. In addition to the risks to water quality within the urban areas themselves, roadside dumps occur in the areas surrounding some towns and are especially common in rural areas having no community trash service. Both the Little Sequatchie and Pryor Cove drainage areas have numerous roadside dumps that vary widely in scale and size. In addition, barren land and hay/pasture land-use areas may pose potential threats to water quality from sedimentation of unvegetated areas or bacterial contamination from livestock grazing on pastures.

Although land use-land cover data are a valuable tool for examining larger scale patterns, potential risks to karst aquifers exist within the study area that may not be easily identified from land use-land cover data. Timber harvesting is prevalent within the study area, and extensive tracts of land are still owned by timber companies within the Little Sequatchie watershed. Logging on steep slopes above stream channels, particularly in areas near stream sinks or karst openings, can result in sedimentation from runoff that can have detrimental impacts to karst biota that often inhabit interstitial spaces in streambeds (Holliday and others, 2020). Rock harvesting or “rock mining” has increased in frequency throughout the Little Sequatchie and Pryor Cove drainage areas. This activity involves excavating talus and sandstone “float” from the side slopes of the Cumberland Plateau, and the resulting disruption of the forest floor results in increased runoff from precipitation events, leading to gully formation and increased sedimentation. The historical remains of coal mines in both drainages can be sources of acid-mine drainage where groundwater seepage through the mine interacts with minerals to create extremely acidic, low-pH water that then discharges into streams. Although this issue may not be as severe at this location as at similar settings in Tennessee, such as the East Fork Obey River (Sasowsky and White, 1993; Webb and Sasowsky, 1994), acid mine drainage can be particularly harmful to sensitive cave biota such as cavefish and cave crayfish. Two major pipelines cross through the Little Sequatchie River valley, one containing natural gas and the other containing a refined and (or) petroleum product (non-high volatile liquid). Pipeline locations and information for Marion County and Grundy County were obtained from the National Pipeline Mapping System (NPMS, 2024). The natural gas pipeline runs through portions of the Sequatchie Cave/“Keira” spring and Ship Cave recharge areas, and the non-high-volatile-liquid pipeline extends through the Dancing Fern Cave and the “Boils” recharge areas. Although this liquid is relatively inert if the pipeline is properly operating, the primary risk to karst groundwater is from accidental rupture of the pipeline. If the rupture occurred over or directly adjacent to stream channels, the liquid would likely sink nearly immediately into the subsurface (Vandike, 1982; Rosen, 1997; Haugh and Bradley, 2001). Lastly, another potential threat to water quality that is particularly prevalent in the Little Sequatchie River valley is off-road-vehicle (ORV) usage. Although sporadic ORV usage might not be a major water-quality disturbance, the valley has seen a dramatic increase in ORV usage in recent years. Often on summer weekends, dozens to even hundreds of ORVs travel up through portions of the Little Sequatchie drainage area. This activity frequently includes driving through streams, sometimes into cave entrances, and has the potential to denude relatively extensive portions of the forest floor because of trail widening. Off-road vehicle usage has the potential to create increased sedimentation into streams from trail migration or attempts to drive on steep slopes.

Characteristics of Recharge Areas in the Little Sequatchie River Valley and Pryor Cove

The upper reaches of the delineated recharge areas in both the Little Sequatchie River valley and Pryor Cove have fairly distinct boundaries (plate 1), in most cases, because of the Pennsylvanian siliciclastic strata (the Crab Orchard Mountain Group and the Gizzard Group) that cap the top of the Cumberland Plateau (fig. 5A, B). These strata are unlikely to be karstified, because of the lack of soluble strata, and thus the topography of these upper reaches is one of the constraints on the recharge areas. Once streams flow off the insoluble strata, they sink into soluble strata and the karst groundwater is able to move more freely through the landscape independent of topography. As a result of this combination of strata and topography, many of the spring systems have recharge areas that, although sharing a boundary with another spring, show limited evidence of interbasin transfer from one recharge area to another. A precipitation event may pass over one or more recharge areas and cause a corresponding rise (both in flow and turbidity) at one spring while other adjacent springs show no indication of the precipitation event and continue to flow at a lesser volume. This behavior was observed at the Ship Cave system on February 28, 2022, when the Ship Cave system was flooding from a precipitation event from the previous evening, creating an additional overflow spring near “Indian Ship” spring. On this date, Ship Cave was the only system in flood stage. Throughout the project, similar events were later observed at both “Keira” spring and at the “Boils,” on separate days, following storm events from the previous day.

The only exception to the distinct recharge area boundaries was the 5.8-mi² shared recharge areas for Sequatchie Cave and “Keira” spring in Dixon Cove (plate 1). Multiple traces in Dixon Cove, both in the main stream and in western tributaries, were traced to both “Keira” spring and Sequatchie Cave (fig. 34). Typically dye appeared at “Keira” spring a few days prior to arriving at Sequatchie Cave. Although the documentation of shared recharge areas is not novel in karst areas (Vandike, 1992; Aley and others, 2007), this is an uncommon characteristic for karst systems on the Cumberland Plateau. In this system, it appears that “Keira” spring may function as an overflow spring for the Sequatchie Cave flow system, because “Keira” spring may dry up completely during low flow periods whereas Sequatchie Cave continues to flow. Although overflow springs are a common feature, often found in karst areas and even within the study area (for example, the “Boils,” “Indian Ship” spring), this particular overflow spring is located several miles up into the recharge area and thus is only recharged by a portion of the total recharge area for the main spring. Both “Keira” spring and Sequatchie Cave are clearly recharged, at least in part, by conduit flow systems and “Keira” spring appears to be an additional outlet for the system located along the conduit network from Dixon Cove to Sequatchie Cave.

Geologic Influences on Karst Groundwater Flow Paths and Trace Lengths

There are notable differences between the karst groundwater behavior in the upper portion of the Little Sequatchie River valley, specifically upstream from Indian Cove, compared to the behavior observed in the lower Little Sequatchie River valley and Pryor Cove (fig. 6). The differences between the two regions appear to be influenced by underlying geology and the degree to which streams have dissected into the Cumberland Plateau stratigraphy. Recharge features for the karst systems are largely focused at discrete locations where insoluble and soluble strata meet, allowing allogenic recharge to the karstic units. In both the Little Sequatchie and Pryor Cove drainage areas, these recharge features are concentrated along the Pennington-Bangor contact, although at some locations in Pryor Cove streams sank at the very top of the Pennington Formation several hundred feet above the Pennington-Bangor contact. Although stream sinks and recharge features are concentrated in the Pennington Formation and Bangor Limestone, the proximity of the valley floor to the top of the Hartselle Formation seems to affect the length of karst groundwater flow paths, the recharge areas that drain to specific springs, and whether perennial streamflow is present in the main channel. In other portions of the Cumberland Plateau, the Hartselle Formation is an aquitard that causes perching of both surface and subterranean streams (Crawford, 1987; Davis and Brook, 1993; Simpson and Florea, 2009; Blair and Vanderhoff, 2019). In the portions of the Little Sequatchie River valley that are upstream from Indian Cove, the Hartselle is considerably deeper, occurring below the valley floor. Enterable caves upstream from Indian Cove can be formed well below (as much as 120 ft beneath) the valley floor and are entirely within Bangor Limestone (Tennessee Cave Survey, unpub. data, 2023). Although these caves have active streams and periodically flood, the main flow from many of the major stream sinks is located below these caves, deeper in the Bangor Limestone. Farther down the valley, the surface streams have dissected further into the Bangor and the depth to the Hartselle below the valley floor decreases until the contact between the Bangor and Hartselle is at valley floor level, albeit covered by alluvium (Milici and Finlayson, 1979). In Pryor Cove (fig. 5A), the Hartselle Formation is beneath the valley floor for the entire drainage area and largely does not outcrop in the watershed. There is a small, documented Hartselle outcrop directly adjacent to the drainage near Pryor Cave Spring (Milici and others, 1986).

Traces in the middle to upper Little Sequatchie River valley had an overall longer distance from the injection locations to recovery sites than those in the lower Little Sequatchie River valley and Pryor Cove (figs. 17 and 18, table 11). Several traces in the upper Little Sequatchie River valley, particularly above Indian Cove, had lengths that exceeded 6 mi, and injection I-1 had a trace length that totaled 7.65 mi from the main sinkpoint on the Little Sequatchie River to the “Boils” (table 11). These longer traces may be due to conditions that allow flow from sinking streams to drop deep into the Bangor Limestone, following conduits or preferential flow paths and eventually resurging from springs. The added thickness of the Bangor Limestone below the valley floor also allows for more groundwater movement from a stream or cove on one side of the valley to traverse and eventually resurge from a spring on the opposite side of the valley. This is true for flow from stream sinks in Peter Cave Cove and Grays Creek that travels from the east side of the Little Sequatchie River valley, under the main valley, and resurges from the “Boils” on the west side of the valley (fig. 17).

Toward the lower end of the Little Sequatchie watershed, below Indian Cove, the trace lengths are shorter, overall, and karst groundwater flow paths appear to be constrained to the flanks of the main valley (fig. 18, table 11). In the lower Little Sequatchie River valley, this may be due in part to the proximity of the Hartselle Formation beneath the valley floor inhibiting downward flow of stream water and decreasing the distances that subterranean streams can traverse. In these portions of the Little Sequatchie River valley, cave passages are typically located above the valley floor and the Little Sequatchie River flows perennially downstream from the “Boils.” The close proximity of the Hartselle to the floor of the lower Little Sequatchie River valley also means the ability for karst groundwater to travel from stream sinks below the main stream is limited and thus the recharge areas for the individual springs are located on the same side of Little Sequatchie River valley as the spring outlets. The inability for recharge from sinking streams in the lower part of the Little Sequatchie River valley to travel deep below the valley floor, because of the Hartselle Formation impeding the downward movement of water, likely influences both the distances that karst groundwater in the Bangor Limestone can travel before discharging at springs as well as what areas are able to contribute recharge to springs.

In Pryor Cove (fig. 25), trace length may have been more influenced by the smaller extent of the watershed itself than in areas with a similar geologic setting in the Little Sequatchie River valley. Karst groundwater from stream sinks in Pryor Cove successfully breaches the Hartselle Formation to flow to the conduits in the underlying Monteagle Limestone. The Blue Spring conduit is formed entirely in the Monteagle Limestone; thus, all positive traces performed from locations in the Pennington Formation or at the Pennington-Bangor contact had to pass through the Hartselle Formation to recharge the spring. Despite this behavior, the Hartselle Formation does appear to have some influence on streamflow in the Pryor Cove area, as perennial flow in Standifer Branch begins at or near where the Hartselle Formation crops out on the surface. Although both dye injections in the lower portions of the Pryor Cove northern and western arms are located on Bangor Limestone, on the basis of geologic mapping and outcrops, this proximity to the Hartselle Formation may have influenced traveltimes from injections I-24 and I-25. At injection site I-24 in the lower part of the northern fork of Pryor Cove, dye took considerably more time to flow to Blue Spring compared to the dye injected at site I-25 in the western fork of Pryor Cove. Injection I-25 in western fork of Pryor Cove was conducted nearly on top of the mapped spring conduit, whereas injection I-24 was only conducted 0.5 mi away (fig. 25). It is possible that a more competent section of the Hartselle Formation at injection site I-24 caused the dye to ultimately remain in the stream channel longer, rather than immediately sinking into the subsurface to recharge the spring. This may have been why the dye from the northern fork of Pryor Cove (I-24) was not detected on packets from Blue Spring until April 10, 2023 (25 days), whereas packets collected on March 28, 2023 (12 days) indicated dye from West Fork Pryor Cove Branch (I-25) had already arrived at Blue Spring.