The Sedimentology of the Ordovician Harding formation.

The Sedimentology of the Ordovician Harding Formation, in the Northern Sangre de Cristo Range.

Ralph Falsetto, Bruce Bartleson and Robert Fillmore

Bushnell Lake 1The Harding Formation was named by C. D. Walcott in 1892. The type locality is one mile northwest of the State Penitentiary at Canon City, Colorado (S 1/2 , NW ¼ , Sec. 31, T 18 S., R.70 W.). The Harding crops out in several locations within central Colorado (Sweet 1954), and correlates with the Simpson Formation in the southeastern Colorado, and the Winnipeg Sandstone, in north central Wyoming and central Montana (Foster, 1972). The Harding represents the beginning of the Tippecanoe transgression, a craton wide sea level rise that began in the early to middle Ordovician (Sloss, 1963).

Study Area

The study area is located in the northern Sangre De Cristo range in the San Isabel National Forest of south central Colorado, and extends from the town of Wellsville south to Hayden Pass, and is bounded by State Highways 50 to the east and 285 to the west (Figure 1). The field work consists of three measured sections: Wellsville (SW. ¼, Sec. 13, R. 9 E., T. 49 N.); Bushnell Lake (SW. ¼ , Sec. 5, R. 10 E., T. 48 N.); Galena Peak (SW. ¼ , Sec. 21, R. 10 E., T. 47 N.). Generalized sections were simplified from the detailed measured sections for correlation purposes (Figure 2).

Bushnell Lake 2Stratigraphy

The Ordovician strata in the study area contains three formations: the lower Ordovician Manitou, Formation Middle Ordovician Harding Formation and Upper Ordovician Fremont Formation (Figure 3). The Manitou Formation consists of brown to dark gray dolomitic limestone, with chert nodules and stringers. The Manitou lies unconformably on Precambrian strata in the study area (Graffin, 1992; Rold, 1961; Litsey, 1958) and likely correlates to the mid-continent Arbuckle Formation (Rold, 1961). The Harding Formation rests unconformably on the erosion surface at top of the Manitou (Rold, 1961). The Harding, which correlates with the Eureka Quartzite, Simpson Group, St. Peter Sandstone, and Winnipeg Formation, marks the beginning of the Tippecanoe transgression (Sloss ,1963). The overlying Fremont Formation consists of two dolomite members, a massive dark gray lower member and a thin bedded upper member. The Fremont is probably equivalent to the mid-continent Viola and Bighorn dolomite in Wyoming (Rold, 1961).


The purpose of this study was to create a detailed sedimentological record of the Harding Formation in the study area, correlate the sections and interpret the facies. A more specific goal was to investigate the Bushnell Lakes locality that has been described as a conformable Ordovician sequence with a fluvial unit in the upper part of the Harding Formation (Graffin, 1992).

Previous Work

Rold (1961) studied the Lower and Middle Paleozoic strata in the Wellsville area where he measured sections and mapped the geology. He found that the " Manitou sea transgressed over a broad gentle Precambrian surface." The fine grained, even bedded sandstones of the Harding Formation were deposited during an influx of source sand and not the result of a regressing sea. The upper part of the Fremont is thinner bedded and contains interbedded limestones which possibly represent minor regressions or fluctuations of the Ordovician sea. Litsey (1958) studied the structure and stratigraphy in the northern Sangre de Cristo range, which included a measured section near Galena Peak. The basal unit of the Harding Formation consists of easily weathered beds and is only exposed in a few areas. Litsey describes the Harding/Manitou contact as a disconformity. His evidence includes the varying thickness of the Manitou and the abrupt change in lithology at the contact, to support this conclusion. The contact between the Harding and Fremont is covered in most places. However, near Galena Peak the contact is undulatory. Graffin (1992) studied the Bushnell Lakes locality and interpreted the Ordovician system as a shallowing upward carbonate sequence and suggests the Harding Formation represents an influx of clastic sediments that halted the limestone deposition for a period of time. He describes the contact between the Manitou, Harding and Fremont formations as conformable, and interprets the upper unit in the Harding as fluvial.

Paleoenvironment during Harding deposition

Ross (1976) reconstructed the Ordovician of the North American craton. Weathering of the Canadian shield provided the source sediments, during late Canadian time. The Harding sediment was swept in by northern trade winds that moved clastic sediments southwest along the Transcontinental Arch. The clastic sediments temporarily halted carbonate production while the Harding Formation was deposited. Canadian Shield erosion halted, due to inundation during Cincinnatian time and carbonate deposition renewed and dominated, after Harding deposition (Figure 3).

This Study


Trace fossils are present in all three measured sections of this study and some units are completely bioturbated. Vertical Skolithos-like burrows are the most common and while horizontal Thalassinoids-like borrows dominate the surfaces of some units although, they are less common. Diplocraterion burrows were observed rarely. The trace fossil assemblages fall within the Skolithos ichnofacies, which indicate a shallow marine environment (Frey, 1978). Since the trace fossils fall within a single ichnofacies their utility is limited separating the shallow marine sub-facies in this area. Fish plates are abundant in the upper sandstone units of all three sections. The fish plates are fossilized Ostracoderm (Astraspis and Eriptychius) remains (Spjeldnaes 1979, Graffin 1992). Although the fish plates have been the subject of considerable study, they are not useful in the facies analysis.


The Harding Formation in the study area mainly consists of quartz arenite sandstone and quartzite. The grains generally are well rounded and well sorted in the center part of the sections. Grains in the upper and lower portions commonly exhibit a bimodal grain size distribution and coarser grains are typically subangular to rounded. The Harding contains varying amounts of shale and siltstone, which appear as partings between the sandstone beds and several inch thick individual beds.

Galena Peak Section

The Galena Peak section is the most weathered and has more cover than the other two sections. Litsey (1958) measured a section near Galena peak and recorded its thickness to be 111'. The measured section in this study includes a lower 75'; exposed part with an upper 35' covered interval. The contact with the overlying Manitou is covered. The section is approximately 15' thicker and has slightly more shale than the Bushnell Lake section. Individual bed thickness ranges from 1" to 46". Sedimentary structures include planar and ripple parted beds (Figure 2).

Wellsville Section

The Wellsville section is well exposed and mildly metamorphosed. The contact with the Manitou is covered. The contact with the Fremont Formation is sharp, exhibiting an abrupt lithologic change from tan quartz sandstone to sandy dolomite. The Wellsville section contains less shale and is approximately 30' thinner than the Bushnell Lakes section. Bed thickness range from 1" to 108" and are almost exclusively planar parted. Preserved sedimentary structures are rare, unit 10 is the only one observed to have internal bedding structures (Figure 2).

Bushnell Lakes Section

The sedimentology and facies analysis of the Bushnell Lakes locality is the focus of this study. It is well exposed with little cover. This section is described in detail below with the facies analysis (Figure 2).

Manitou Formation

The top of the formation consists of sandy dolomite with interspersed rounded to well rounded fined grained quartz sand.

18 " cover

Harding Formation

Basal Sandstone: units 31-34, thickness 114"

The basal unit consists of 8 partially covered beds of weathered bimodal quartz sandstone, that coarsens upward overall. The base is light green and mottled by vertical burrows. It consists of medium to coarse rounded to well rounded grains and small scale trough cross beds are present in a few undisturbed areas. The texture of the mid-part is similar to the base, but is light gray with no sign of bioturbation. The upper part is light green with a mottled appearance. It is coarser grained , consisting of very coarse grains and clay clasts up to 4mm. The top is a dark gray sandstone composed of medium to very coarse grains and contains clay clasts up to 4mm. The bedding surface contains horizontal and vertical burrows. Below it is a coarse grained tan sandstone, with a thin shale bed beneath (Figure 2).


The lower part of the Harding Sandstone is thought to represent the beginning of the Tippecanoe transgression. The slight bioturbation and coarse nature of this unit suggest a near shore line deposition (Reinson, 1984). The reverse grading of this unit corresponds with a foreshore deposition (Galloway and Hobday, 1983). The unit which is consistent with and could represent the beginning of a transgression. The overlying shale bed suggests the unit is probably a near shore sand deposit behind a barrier bar.

14" cover

Lower Shale: unit 30 Thickness 17"

The lower shale is light gray, weathered and wavy laminated.


Due to a position below a coarse cross bedded sand which is interpreted a barrier bar deposit and above another coarse foreshore sand this shale probably represents lagoonal muds.

Lower Sandstone: units 27-29 Thickness 48 "

The base of the lower sandstone consists of light tan medium lower to very coarse upper, subangular to well rounded quartz grains. It has low angle planar and herringbone cross beds and the top is slightly bioturbated with vertical burrows (Cross bed photos). Above this is 102" unit consisting of several 7" - 32" beds which are light to dark gray well sorted fine grained sandstone. The beds are parted by shale which are slightly thicker toward the base of the unit. The upper and lower partings are planar and the middle part of the unit show wavy partings. The bedding surfaces are horizontally and vertically burrowed. Above is a 130" unit consisting of several 7 " - 24" beds of planar parted mottled light to dark gray very fine grained sandstone. The partings overall are thinner in the middle of the unit. Bedding surfaces in the lower part show intense horizontal burrowing, the middle surfaces show slight horizontal burrowing and the upper shows abundant vertical burrows.


The base consists of very coarse grained low angle and herringbone cross beds, which is consistent with upper to middle shoreface deposition. The features in this unit could also be interpreted as bar crest deposits (Davis, 1983). As the sea transgressed the coarse grained cross bedded barrier bar sands, migrated land ward over the lagoonal muds. This unit abruptly covers a back barrier lagoon and is in turn covered by deeper water sediments which is typical of some transgressing barrier bar deposits (Galloway and Hobday, 1983). This interpretation explains the underlying shale. The fine grained beds above have planar and ripple partings, with horizontal and vertical burrows. Its finer grained texture and greater bioturbation indicate slightly deeper water conditions (Reinson, 1984). This unit is likely middle to lower shoreface. The very fine grained upper beds are planar parted with overall dominant horizontal burrows at the parting surfaces, which are characteristics of lower shoreface deposition.

Middle Shale: unit 26 Thickness 36"

The middle shale consists of sandy shale with a 2" very fine grained sand stone lens approximately 15' long that pinches out laterally. The shale below the lens is slightly wavy and above the lens is planer laminated (Figure 2).


This unit probably represents a storm influenced deposit, where the wave base was temporarily lowered and wavy laminated shale with a sandstone lens were deposited. This was covered by planar laminated shale indicating below normal wave base deposition (Harms et al. , 1982).

Lower Middle Sandstone: units 22-25 Thickness 86"

The middle lower sandstone consists of two units. The beds consist of alternating light and dark quartz sandstone layers separated by shaly partings. The units show slight vertical burrows and the light layers are thicker than the dark. The first unit is 13" of fine grained quartz sandstone with 5mm to 120mm thick white layers and 10 mm to 35mm dark layers. The layers are thinner at the top of this unit. The units are separated by fine grained intensely bioturbated sandstone below and above the first unit (Figure 2) . The second unit is 41" thick consisting of several 1" to 5" fine grained sandstone beds with wavy shale partings. The alternating varve like layers are 1mm to 15mm thick.


The unit is very fine grained with generally planar parted beds, which suggests lower shoreface deposition (Reinson, 1984).

Middle Sandstone: units 15-21 Thickness 252"

The middle sandstone is composed of light gray very fine to fine grained sandstone, that is bioturbated through out. There are several beds 9" to 42" thick with planar and wavy partings. The basal 171' of intensely vertically burrowed (Skolithos-like) fine grained sandstone and most of the partings are planar . The upper beds are very fine grained with wavy partings. The burrowing varies, the lower part shows intense vertical and sparse horizontal (Thalassinoids-like) burrows and the middle shows sparse horizontal burrows and the top is intensely vertically burrowed.


The base of the unit is fine grained, planar parted, with dominant vertical burrows, features consistent with lower to middle shoreface deposition (Elliott , 1979). The upper beds are very fine grained, generally ripple parted and have some. This unit is probably middle to lower shoreface.

Upper Middle Sandstone: units 8-14 Thickness 136 "

The upper middle sandstone and consists of sixteen fine grained sandstone beds. The bed thickness ranges from 2" to 22" and they are separated by dominantly shaley planar partings. Bedding surfaces are slightly bioturbated by vertical burrows. The 18" bed near the middle is slightly coarser grained (fine lower) and the grains are subrounded to rounded. The upper beds are less bioturbated, finer grained (very fine lower) with horizontal (Thalassinoids-like) and vertical (Skolithos-like) burrows at the top (Figure 2).


The bed partings are dominantly planar, slightly coarser grained and slightly burrowed which probably indicates middle to possibly upper shoreface deposition (Reinson, 1984).

The Upper Sandstone: units 1-7 Thickness 195"

The upper sandstone consists of two units separated by 16" of cover. The basal unit consists of two beds that are very coarse sandstone at the base and fine upwards to fine grained sandstone. The lower part contains seven planar shale parted beds and shows slight to moderate vertical burrows. The upper is a massive sandstone with sparse horizontal burrows at the top of the unit (Figure 2). The second 77" unit consists of six fine to very coarse sandstone beds separated by wavy shale parting. The shale partings become thicker up to 2" at the middle of the unit. The beds are slightly to moderately bioturbated by vertical burrows. The upper beds contain abundant fish plates.


Very coarse grained sandstone with planar partings and slight to moderate bioturbation in unit 1 suggests an upper shoreface deposit. Very coarse grained sandstone with wavy partings and slight to moderate bioturbation in unit 2 suggesting a transition from upper shoreface to foreshore deposition. The lack of burrows near the top likewise suggests a shallower, higher energy setting (Elliot , 1979).

Fremont Formation

There is a sharp planner contact between the top of the Harding/Fremont boundary, that can be traced laterally for tens of meters. The base of the Fremont Formation consists of sandy dolomite, with approximately 30% rounded to subangular, medium to 3mm quartz grains.

Depositional Environment

The Bushnell Lakes section is interpreted to represent a transgressing barrier bar sequence from the basal sandstone to the middle shale. A regressive sequence begins at the middle sandstone and continues to the top of the section. The basal near shore sandstone is overlain by thin lagoonal shale. The overlying lower sandstone begins with a barrier bar sandstone. The deeper water facies trend continues upsection. The facies change from middle to lower shoreface deposits and finally the below wave base at the middle shale (Figure 5A). The middle shale marks the transitional middle point from a transgressive to regressive cycle. The units show a shoreward transition upsection from the middle shale. The middle sandstone changes from lower to middle, and then upper shoreface facies. Likewise, the upper sandstone changes from upper shoreface to foreshore facies at the top of the section (Figure 5B). The transgressive to regressive sequence is evident in the Wellsville section. Although, the Galena Peak section is missing the upper 35' it also reflects the transgressive sequence found in the lower units of the other two sections (Figure 2).

This interpretation is similar to a work done by Thyer (1991) at the Canon City Embayment. He interpreted the Harding in that area as a prograding, wave-dominated barrier bar system. He described a series of depositional cycles where the lower units show a transgression, the middle shale was deposited at the height of the transgression and the upper units record a regression. Several small scale cycles, due to small fluctuations in relative sea level, are shown in a depositional sequence (Figure 6). The onlapping transgressive track resembles the lower units of the measured sections. The middle shale would correspond to the maximum landward point of the transgression. The upper units are analogous to downlapping regressing sea trend toward the top of the diagram. The coastal plain sediments are absent in the Harding Formation of the study area, and may have been eroded prior to deposition of the Harding.

Possible fluvial upper unit

The upper part of the Harding has been interpreted as a fluvial deposit (Graffin 1992). The festoon and epsilon cross beds used to support the fluvial interpretation were not observed in Bushnell Lakes section in this study. Additionally, the vertical (Skolithos-like) burrows found in the upper sandstone indicate a marine environment (Chamberlain, 1978).

Nature of the upper and lower contacts in the study area

The lower contact is covered in all 3 measured sections of this study. Hence the assertion that the contact is gradational and conformable cannot be addressed directly. Instead a brief summary of past work concerning the Manitou/Harding contact is presented. The Harding, which correlates with the Eureka Quartzite, Simpson Group, St. Peter Sandstone, and Winnipeg Formation, marks the beginning of the Tippecanoe transgression (Sloss ,1963) The Tippecanoe is one of six unconformity bounded sequences of the continental interior of North America. Ross and Tweto (1980) estimate an approximately 35 my. gap between the Manitou/Harding boundary. Even though the quartz sand to carbonate ratio decreases from the upper unit of the Manitou to basal unit of the Harding. Graffin (1992), it unlikely that the contact is conformable given the regional unconformity.
The erosional surface at the top of the Harding would be consistent with another transgression where the Fremont was deposited on this surface. The Harding/Fremont boundary exhibits a sharp lithologic change from quartz arenite to dolomite in the Wellsville and Bushnell lakes sections, which is interpreted as paraconformity in this study.
An erosion surface was recognized by Litsey (1958) in the Galena Peak area. He described 4' of dolomitic sandstone at the base of the Fremont Formation that may have been deposited during reworking of the top of the Harding. Rold (1961) described the base of the Fremont as a result of a dramatic change in sediments "rather than uplift and long periods of erosion.". Ross and Tweto (1980) estimated up to a 10 my. gap between the Harding/Fremont contact in a regional scope. Ross and Tweto(1980) state "Nowhere has the Fremont been found without the Harding Sandstone beneath it". This occurrence suggests that the two formations occur as a depositional package, and an unconformity between the two would likely be a minor one.


The Harding Formation at Bushnell Lakes locality probably represents a transgressive sequence. The findings in this study are consistent with transgressive barrier bar sequence, where the facies show a progressively seaward stepping depositional environment, from the base of the Formation to the middle shale. The middle shale represents the maximum height of the transgression and the transition from a transgressing to regressing sea. The regressive cycle begins at the middle shale, where the facies indicate progressively landward stepping deposition environments, upsection to the top of the Formation. Erosion of the Canadian Shield provided the sediments for Harding deposition. The clastic sediments were swept southwest along the Transcontinental Arch by trade winds, during Harding deposition. The carbonate deposition was stopped by the influx of clastic material. The Canadian Shield was inundated, halting erosion and clastic input into the system, allowing carbonate deposition to resume.
The base of the Harding corresponds with the beginning of the early Middle Ordovician Tippecanoe transgression in the study area. The Tippecanoe is one of six unconformity bounded sequences of the continental interior of North America. The upper and lower boundaries of the Harding Formation are most likely unconformities. The bulk of the evidence, including an interregional unconformity at this juncture, indicates an unconformity at the Manitou/Harding boundary. The nature of the Harding/Fremont boundary is less clear. The contact is interpreted as a paraconformity due to the abrupt change in lithology and sharp planar contact that can be traced for tens of meters. The supporting evidence from other workers includes, when the Fremont is present it overlays the Harding and the estimated gap between the Formations is less than 10 my.

References Sited

This research was funded by a grant from the Barrett Resources Fund for Excellence in Geology.