3: The Permian-Triassic transition in Colorado
Published:September 07, 2016
- PDF LinkChapter PDF
James W. Hagadorn, Karen R. Whiteley, Bonita L. Lahey, Charles M. Henderson, Christopher S. Holm-Denoma, 2016. "The Permian-Triassic transition in Colorado", Unfolding the Geology of the West, Stephen M. Keller, Matthew L. Morgan
Download citation file:
The Lykins Formation and its equivalents in Colorado are a stratigraphically poorly constrained suite of redbeds and intercalated stromatolitic carbonates, which is hypothesized to span the Permian-Triassic boundary. Herein we present a preliminary detrital zircon geochronology, new fossil occurrences, and δ13C chemostratigraphy for exposures along the Front Range and in southeastern Colorado, to refine understanding of the unit’s age and depositional history.
Detrital zircons from the uppermost Lykins Formation and an overlying eolianite consist of a complex and highly diverse primary and multi-cycle grain population transported from Laurentian and Gondwanan terranes, potentially both by wind and water. Youngest concordant zircons do not rule out deposition of the uppermost Lykins Formation during a portion of Early Triassic time. Conodonts from the lower Lykins Formation require Middle Permian (Guadalupian) deposition. Conodont alteration indices of 1 indicate the unit has a shallow burial history and is amenable to paleomagnetic inquiry. Conodonts, together with other vertebrate, invertebrate, microfossil, and trace fossils, suggest a very shallow to emergent marine origin for the unit’s most substantial carbonates, and hint at a marine origin for the unit’s intercalated gypsum-anhydrite members. Chemostratigraphy corroborates field evidence of emergence and karst development capping certain units, like the Forelle Limestone Member of the Lykins Formation, where potential sequence boundaries appear to be punctuated by a short-lived meteoric signature.
Results presented here are a progress report of ongoing work in these successions. This field trip consists of a brief tour through exposures of the Lykins Formation, in which we will examine well-known localities as well as view new ones for which we seek insights.
Introduction and Historical Context
Permian-Triassic successions in Colorado and vicinity chronicle the evolution of tropical low-relief continental environments during an extinction-punctuated supercontinental cycle. Unfortunately, these local ‘redbed’ successions are challenging to correlate or interpret within the region’s depositional basins (see synthesis in Boyd and Maughan, 1973), much less integrate into the global picture of the Permian-Triassic transition. In part this is because they are poorly fossiliferous and are commonly incised by overlying Jurassic unconformities and strata. Permian-Triassic successions often lack clear sequence boundary cues and are dominated by easily weathered siltstones and evaporites that inhibit outcrop analyses and challenge surface-subsurface correlations. These successions are nomenclaturally complex, owing to the evolution of different stratigraphic schemes within the modern depositional basins into which their strata are now partitioned (Fig. 1).
Here we scaffold on the efforts of previous workers to present a progress report of our work to constrain the temporal and paleoenvironmental framework for these strata, with emphasis on exposures of units in Colorado’s Front Range and in southeastern Colorado. These units include the Lykins Formation, Day Creek Dolomite, and Taloga Formation, as well as the conformably underlying eolianites of the Lyons and Whitehorse Sandstones, and the overlying eolianites of the Jelm, Sundance, and Entrada formations. These units are also well known from the subsurface. In order to focus our efforts on strata akin to what will be seen on this field trip, we only present outcrop-based data here.
This field trip will consist of up to four stops, illustrated in Figures 2 and 3, in which we will highlight some of the features of the Lykins Formation. The trip will progress roughly from the bottom of the unit to its top, with opportunities to see its contact with the underlying Lyons Sandstone and the overlying Jelm Formation. There will be a brief detour out of stratigraphic order so that we can see a gypsum quarry that highlights the Blaine Gypsum Member of the Lykins Formation—a member rarely seen in surface outcrops. Stops 1-3 are on private property; if you are taking this trip on your own, please request permission before visiting them.
Stop 0: Trip Origin (0 mi)
The trip will begin at the east loading dock area (39.741494° N, 104.995392° W) of the Colorado Convention Center (700 14th Street, Denver, Colorado 80202), along Wel-ton Street. Vans will depart and drive to Stop 1.
Stop 1: Golden Maintenance Facility (15.1 mi) (39.777946° N, 105.239224° W)
This stop is illustrated in Figures 3A and 3B, and in our reference section F6 (Fig. 2). This stop will focus on surveying the transition from the Lyons Sandstone to the upper Forelle Member of the Lykins Formation. The first half of this stop provides a cross-sectional view of the lower part of the Lykins Formation, illustrating from bottom to top, the Greenacre Lentil, Falcon Member, and Forelle Member. Stromatolites, solution cavities, and related features are all visible in these units. At the second part of this stop, we will examine and discuss an enigmatic bed top in the upper Forelle Member, which is locally brecciated, contorted, and possibly slumped. How might these bedforms have formed? Prior to departure, there will be an opportunity to use the facilities here.
Stop 2: Park Creek Reservoir (114 mi) (40.854090° N, 105.169430° W)
This stop is illustrated in Figure 3C and in our reference section F12 (Fig. 2). It offers an opportunity to walk the section from the upper Lyons Sandstone through the lower half of the Lykins Formation. Here we will see the unit’s main carbonate members, including the basal carbonates and the Falcon, Forelle, Poudre, and Park Creek members. We will discuss their contacts, possible karst features, and possible loci of deposition. We will eat lunch at this stop.
Stop 3: Monroe Gypsum Quarry (115.9 mi) (40.875817° N, 105.166702° W)
This private quarry offers a rare opportunity to see a clean exposure of the Blaine Gypsum Member of the Lykins Formation. Here, the gypsum is mined as a cement additive, and more rarely is mined in the region for artist’s alabaster and as an agricultural soil amendment. At this locality, do the evaporites show signs of subaerial exposure or demonstrably marine features?
Stop 4: Optional Stop—Red Mountain Open Space (121.4 mi) (40.8922929° N, 105.157826° W)
This stop is illustrated in the background of Figure 3C. It offers an opportunity to hike through the ‘color transition’ of the Red Hill Shale Member (described below), and see the contact with the overlying Red Draw Member of the Jelm Formation. At this site, participants may discuss the depositional environments or diagenetic conditions that fostered the color change in the Lykins Formation, the marine versus nonmarine nature of the redbeds, and similarities between the Jelm and Lyons formations.
Stop 5: Colorado Convention Center (210.6 mi)
We will likely make a bathroom stop in Fort Collins, en route to our return to the east loading area (39.741494° N, 104.995392° W) of the Colorado Convention Center.
Nomenclature and Methods
To simplify and standardize the nomenclature used on the field trip and in this contribution, we have synthesized the region’s nomenclatural history graphically, with emphasis on units for which new data is presented and that are of utility for regional correlation. These are presented as Figs. DR1-DR21 and are grounded in earlier revisions of the region’s Permian-Trias-sic nomenclature (Hill, 1899; Lee and Knowlton, 1917; Clifton, 1925; Sanders, 1934; Maher and Collins, 1952; Shaw, 1956; Broin, 1957; Maughan, 1980; Wiggs, 1986; Merewether, 1987). Our methodology for field and laboratory analyses is described in the GSA Data Repository (Appendix DR1; see footnote 1).
Sedimentology and Stratigraphy
There has been substantial investigation of the sedimentol-ogy, diagenesis, stratigraphy, facies architecture, and accommodation history of the Lykins Formation (Hayden, 1869; Darton, 1901, 1905, 1906, 1908, 1909; Fenneman, 1905; Henderson, 1908, 1909; Tieje, 1923; Duce, 1924; Heaton, 1933; Iglehart, 1948; Maher, 1954; Oriel and Mudge, 1956; Shaw, 1956; Van Horn, 1957; Ogden, 1958; Oriel and Craig, 1960; Campbell, 1963; Story and Howell, 1963; Wells, 1967; Grose, 1972; Court-right, 1974; Berman, 1978; Wiggs, 1986), including regional studies that also integrate data from coeval units in bordering states (Lee, 1927; Reeside, 1929; Condra et al., 1940, rev. 1950; Maher, 1952; Hoyt, 1960; Maughan and Wilson, 1963; Momper, 1963; Mudge, 1967; Pearson, 1972; Rascoe and Baars, 1972; Pipiringos and O’Sullivan, 1976; Maughan, 1980; Renner, 1988) or rely heavily on subsurface well logs, cuttings, and cores (Dorell, 1940; Thompson and Kirby, 1940; Maher and Collins, 1952; Broin, 1957; Johnson and Baltz, 1960; Merriman, 1960; Greer, 1985). Rather than summarizing this work here, we highlight how we have begun to strategically build upon it, and refer participants to key syntheses where relevant.
The Lykins Formation comprises an ~300-m-thick redbed succession dominated by poorly indurated maroon siltstones, which are locally intercalated with minor very fine sandstones and rare shales. It was deposited in a tropical low-relief, near-sea level, arid epicontinental setting (see summary in Wiggs, 1986). The base of the unit interfingers with the sandy coastal eolianite facies of the Lyons Sandstone (Butters, 1913; Tieje, 1923; Struby, 1957; Willhour, 1958; Pearson, 1972, 1976; Courtright, 1974; Deacon et al., 2014; Fig. 4), and it is capped by the coastal eolianite facies of the Jelm Formation (Heaton, 1939; Pipiringos, 1968; Courtright, 1974; Picard, 1978; Heckert et al., 2012). In the upper half of the Lykins Formation, there is a prominent color change in the siliciclastics, where the unit changes from pale at its bottom to dark reddish-brown at its top, and the abundance and thickness of interbedded very fine sandstones increases. Sandstones in this darker, upper half of the Lykins Formation contain low-relief asymmetrical ripple marks and mud chips, and the siltstones have raindrop imprints and polygonal cracks, implying shallow water to emergent conditions (Broin, 1957; Fig. 5). Yet as Boyd and Loope (1984) and Boyd (1993) indicated in coeval strata in Wyoming, the lack of paleosols, caliches, and terrestrial fossils intimates that these strata do not reflect an exclusively terrestrial landscape. Yet the paucity of channels, current-and-wave ripples, dunes, and coarse lithologies suggests its depocenters were not dominated by high-energy fluvial, tidal, or wave processes, or long-lived eolian processes like its overlying and underlying units. While arid and intermittently exposed, the Red Hill Shale Member of the upper part of the Lykins Formation has been hypothesized to be influenced by a mix of distal floodplain deposition (Picard, 1967, 1978; Greer, 1985), marine flooding (Sheldon et al., 1967), influx of fine-grained windborne sediments (sensu Renner, 1988) or other factors (Broin, 1957; Mudge, 1967; Pearson, 1972; Maughan, 1980; Kauffman, 1986). The redbeds of the lower Lykins Formation also contain structures indicating subaerial exposure and shallow ponding of water, but have many more asymmetrical ripples, rare shallow channels, and a few bona fide shale intervals (Tieje, 1923; Iglehart, 1948; Van Horn, 1957; Greer, 1985; Wiggs, 1986). Moreover, the lowest third of the Lykins Formation succession is punctuated by thin but regionally extensive limestones, dolostones, and anhydrite-gypsum beds. Previous workers principally employed lithostratigraphic correlation to tie the Lykins Formation to the Phosphoria and Park City formations to the west (e.g., Lee, 1927; Maughan, 1964; Pearson, 1972; Renner, 1988), interpreting the Lykins Formation’s carbonates to represent marine incursions across a slowly subsiding low-relief continental landscape. The unit’s two widespread gypsum-anhydrite-halite beds imply regular and prolonged evaporation of marine waters in a basin characterized by well-stratified marine brines (Greer, 1985). Overall, each redbed-carbonate-evaporite package can be interpreted as a series of marine flooding events that yielded shallow subtidal to eventually supratidal deposits. Siliciclastic muds may have been deposited in shallow subtidal to mud flat settings, with sediments possibly being reworked by distal floodplain processes and/or exposed multiple times during and between flooding events. Because the carbonates are more distinct and easily correlated in well logs than the redbeds, they have been elevated to member status and been the focus of extensive study to better understand regional correlations.
The Lykins Formation outcrops poorly. Where decent outcrops are available, two depositional themes predominate: (1) overall low hydraulic energy represented by sedimentary structures and bedforms in the unit; and (2) repeated evidence of subaerial exposure and aridity in the basal portions of the unit. Some of these features will be viewed on field-trip Stops 1 and 2.
In the carbonates, common exposure features include m-scale cavities, disrupted laminations, injection of underlying muds into the unit, slumping, and solution breccias. Some of the units’ solution breccias and conglomerates are mantled by stromatolites, which further strengthens hypotheses that many of these are paleokarst (rather than modern karst) features. They are significant because, in tandem with fossil and isotopic data (presented below), exposure features may represent quasi-isochronous surfaces that can be used for regional correlation between portions of the Lykins Formation and other units in bordering states.
In the siliciclastics in the lower portion of the Lykins Formation, shorter-lived emergence or desiccation indicators are present, like hoppered and cubic pseudomorphs, which sometimes mantle polygonally cracked mudstones or interference-rippled sandstones and siltstones (Fig. 5). Elsewhere in the lower portion of the succession, shallow water indicators like ladder-back ripples or short-wavelength oscillation ripples are common. Arid depositional conditions are reflected throughout the lower Lykins Formation by cm-scale gypsum stringers and common pitting, fenestral fabrics, and evaporite mineral inclusions in carbonates. Evidence of aridity and brine formation also comes from two major bedded gypsum (anhydrite) horizons—one just above the Falcon Member (Figs. 2, 4) and another, called the Blaine Gypsum, that occurs between Greenacre Lentil and the Lykins Formation’s basal carbonates. These two evaporites are thick and well known from subsurface data, but are often thin or weathered in surface outcrops, except where quarried. We will visit one of these as Stop 3 on the field trip (Fig. 3).
Detrital zircons were analyzed in an attempt to characterize the maximum depositional age of the unit and to better understand its provenance. No tuffs, ignimbrites, or volcanic rocks were identified in the examined cores or surface exposures of the Lykins Formation. Herein we report results from our first four analyzed samples, those that come from the top of the Lykins Formation and from the base of the Red Draw Member of the overlying Jelm Formation at section F12 (Figs. 2, 6). The lowest Lykins Formation sample, DZ1, is a thinly bedded siltstone rich in very fine sand that consists of very fine sand sized, well rounded, and generally stubby to elongate zircons that in transmitted light are colorless, yellow, pink, and brown. Approximately 34 m stratigraphically higher in the Lykins Formation, sample DZ2 is a thinly bedded very fine sandstone with visible mud drapes and interbedded siltstone that contains rounded to very well rounded, stubby to prismatic very fine sand sized zircons that vary in color from colorless to brown and pink. A meter below the Lykins-Jelm contact, sample DZ3 is a very fine to fine muddy sandstone that contains very well-rounded, stubby to prismatic zircons, which vary from colorless to yellow, brown, and pink. The uppermost sample, DZ4, from the Jelm eolianite facies, is a fine-grained tabular, cross-bedded quartz arenite containing very well-rounded, stubby to prismatic zircons that are mainly colorless, with minor pink and brown grains.
In addition to being generally well to very well rounded, zircons have numerous collision marks or pitting on their surfaces, and often display collision-induced cracks. These surface features suggest a high degree of transport and/or very active abrasion during sedimentary processes experienced by the grains. Age spectra of the grains suggest that they may have been recycled by multiple sedimentary processes.
The four detrital zircon samples, which are hypothesized to span the Early Triassic to Middle Jurassic, have broadly similar age spectra. Composite peak age populations are 418 Ma, 608 Ma, and 1044 Ma with a broad peak age spectrum between ca. 1500 Ma and 2100 Ma. A minor age peak at 2700 Ma is also present (Fig. 6). Minimum concordant analysis ages for each of the samples are 258 Ma (DZ1), 254 Ma (DZ2), 295 Ma (DZ3), and 223 Ma (DZ4); however, interpreted maximum depositional ages/youngest age populations (more than 3 grains of similar age define a population) are ca. 400 Ma, ca. 400 Ma, ca. 330 Ma, and ca. 300 Ma respectively. A ca. 254 Ma ash was recently discovered in the Meade Peak Formation of eastern Idaho (Davydov et al., 2014), close in age to the youngest grains from the Lykins Formation.
The diversity of peak age populations implies they may be derived from multiple primary sources; however, the majority could potentially be derived from eastern North America and marginal terranes. North American terranes and orogenic belts that are common sources of detrital zircons (Whitmeyer and Karlstrom, 2007; Lund et al., 2015, and references therein) that overlap with those observed in this study include the Archean Superior craton (ca. 2800-2600 Ma), Penokean and Dakota/Trans-Hudson orogenies (ca. 1900-1800 Ma), Yavapai-Mazatzal/Great Plains terranes (ca. 1800-1600 Ma), Granite-Rhyolite and Llano provinces/Shawnee terrane (ca. 1550-1350 Ma), Grenville orogeny (ca. 1200-950 Ma), and Appalachian orogeny (ca. 480-265 Ma). Gondwanan (e.g., Suwanee-Wiggins terrane) and peri-Gondwanan terranes are a potential source for ca. 650-500 Ma age peaks and trans-Amazonian terranes for the ca. 2200-2000 Ma detrital zircons (Mueller et al., 2014), both of which were transferred to North America along the Appalachian-Ouachita orogenic margin in the late Paleozoic. Specifically, 610 Ma and 2080 Ma are the most prominent peaks in units of the Anti-Atlas Supergroup, which was part of the conjugate margin to North America after Pangea dispersal, suggesting an affinity with peri-Gondwanan terranes and the western Africa craton (Abati et al., 2010). The youngest, sub-285 Ma zircons in the samples could be derived from Cordil-leran arc terranes (Dickinson and Gehrels, 2009).
There are few detrital zircon studies of similarly aged units in the region. Hager (2015) analyzed a mudstone from the Lykins Formation and a very fine sandstone from the Red Draw Member of the Jelm Formation near our section S2. Four grains were analyzed from the Lykins Formation sample, which did not permit assessment of the peak age spectra; the youngest grain yielded a 206Pb/238U age of ca. 318.9 Ma, but without substantiating whether or not that analysis is concordant. A total of 248 grains were analyzed from the Jelm sample, with peak age spectra overlapping with those from this study, including robust peaks at 413 Ma, 600 Ma, 1038 Ma, a broad peak at ca. 1440-2100 Ma, and a minor Archean peak of 2700-3000 Ma. This overlap suggests that Jelm facies in southeast Colorado drew upon similar sources as those from the northern Front Range. Within this Jelm sample, the youngest grain yielded an age of 245.5 ± 5.9 Ma, but unfortunately the age is highly discordant (206Pb/238U = 245.5 ± 5.9 Ma, 207Pb/235U = 320 ± 10.7 Ma, and 207Pb/206Pb = 905.2 ± 62.6 Ma); however, a relatively young population of concordant zircons appears to exist at ca. 270 Ma or older (Hager, 2015).
Detrital zircon studies of the southern United States and Colorado Plateau for Permian-Jurassic sedimentary rocks (e.g., Dickinson and Gehrels, 2008, 2009; Gehrels et al., 2011; Gehrels and Pecha, 2014; Weislogel et al., 2015) yield largely similar results to ours with minor regional differences (Fig. 6). Dickinson and Gehrels (2009) show evidence for transcontinental transport and intraregional recycling. It is suggested that the majority of the sediment transported to central and western North America during the Triassic was from the southeast (Dickinson and Gehrels, 2008), and from the east-southeast during the Jurassic (Dickinson and Gehrels, 2009) along major paleo-rivers with several tributaries. Reworking of detritus was both due to eolian and fluvial processes because intrabasinal recycling is apparent (Dickinson and Gehrels, 2009). Interestingly, very few “local” basement sources have strong signals in the detrital zircon record, suggesting that exposure of bedrock was limited. The broad similarities of peak ages in latest Paleozoic-Middle (?) Jurassic sedimentary rocks of this study and other regional studies suggest similar distal source rocks and recycling of the distal detrital zircon components. Because the Jelm samples are from the unit’s eolian facies (the Red Draw Member), and because the sands in the upper Lykins Formation could have been influenced by a similar source (potentially even the same source as the underlying eolianites of the Lyons Sandstone), wind transport could account for the diversity of distant zircon sources comingled with fluvi-ally transported grains in the succession.
Fossils are rare in the Lykins Formation, but permit preliminary constraints on the depositional environments, age(s) of deposition, and burial history of the unit. Many of the earlier reports of fossils, beginning with the survey work of Hayden (1869), Cross (1894), and Fenneman (1905), lack sufficient stratigraphic context needed to leverage fossils into a modern geologic time scale. Float-collected or stratigraphically misidentified specimens from the Lykins Formation provide historical challenges in need of unraveling. For example, the most notable macrofossil reported from the Lykins Formation was a fish discovered near our southernmost exposures, section F1 (Schoewe, 1930). Subsequent work at this site and identification of comparable specimens elsewhere intimates that it may have originally come from nearby exposures of the fish-rich Ordovician Harding Sandstone. The U.S. Geological Survey invertebrate paleontology collection provides another example, but from near our northernmost exposure, section F13. Here, collections of bellerophont mollusks were made in the Lykins Formation in 1944 by W.O. Thompson. It turns out that these were likely collected from what today we would call the Ingleside Formation, below a “crossbedded sandstone,” which is now known to be the Lyons Formation (Butters, 1913; Deacon et al., 2014).
The most common in situ fossils in the Lykins Formation succession are biosedimentary structures, specifically stromatolites (Fig. 7), which we will have ample opportunity to examine at each field-trip stop (Fig. 3). Although originally described as crinkled, distorted, ocherous, or cherty limestones and sandstones (Marvine, 1874; Fenneman, 1905; Lee, 1927; Heaton, 1933; LeRoy, 1946; Iglehart, 1948), Walker (1957) and Rezak (in Van Horn, 1957) were some of the first to relate the Lykins Formation’s distinctive bedforms to microbial activities that produced stromatolites. Stromatolites occur in all six carbonate-dominated intervals of the Lykins Formation, including the Park Creek, Poudre, Forelle, and Falcon members, in the Greenacre Lentil, and in the unit’s unnamed basal carbonates. Stromatolites exhibit a range of sizes from mm-scale growths to bioherms tens of meters wide with potentially several meters of synoptic relief. Stromatolites nucleate atop clastic mudstones, carbonate clasts, carbonate beds, other stromatolites, and slumped beds. The most common morphotypes are linked hemispherical to domal forms and flat, wavy, and pseudocolumnar stratiforms, with rare nodular to hemispherical forms that exhibit elongate, oblong, and circular plan view geometries (Ogden, 1958; Greer, 1985; Wiggs, 1986). In comparable Proterozoic strata, such morphologies are interpreted as typical of low-energy settings, where the stromatolites develop behind barrier reefs or rimmed shelves (Walter et al., 1992). Given the absence of high-energy indicators elsewhere in the Lykins Formation, the depocenters where Lykins Formation stromatolites grew may have also been protected or characterized by low energy. Thrombolites, a common element of Phanerozoic marine carbonate-hosted micro-bialites, are absent. Stromatolites in the Lykins, Day Creek, and equivalent formations differ in their size, shape, and growth modes from Early Triassic fully marine stromatolites of Utah, Nevada, and California (e.g., Schubert and Bottjer, 1989; Mary and Woods, 2008).
Darton (1905) was the first to report body fossils in the Lykins Formation, noting the presence of a Natica or Naticop-sis-like gastropod and some “small indeterminate” bivalves at a locality in Lyons, Colorado, and bivalves akin to Bakewellia at Morrison, Colorado. R.M. Butters collected a diverse assemblage of bivalves (Alula gilberti?, A. squamulifera, Myalinaperattenu-ata, M. subquadrata, M. wyomingensis, Pleurophorus sp.), a gastropod (Murchisonia sp.), and a bellerophont mollusk (Bel-lerophon crassus) in 1910 near our sections F3, F10, and F13; these were reported by Girty (1912). Unfortunately, the fossils’ position within the Lykins Formation is poorly constrained and could have come from either the Falcon or the Forelle members. We have since relocated Butters’ original sites and recollected comparable fossils from two of these localities in the Falcon and Forelle members (Fig. 8); the third locality is now underwater in Horsetooth Reservoir. To further contextualize this early work, we examined unpublished fossil collections in the U.S. Geological Survey invertebrate paleontology collections and learned that in 1911, G.B. Richardson collected a similar fauna dominated by gastropods, bivalves (including Allerisma?) from Lykins Formation exposures near section F3, as well as coquinas from near F5 that contained the bivalves Allorisma sp., Alula squamulifera, Aviculopecten sp., Edmondia sp., and Myalina perattenuata, and the gastropod Loxonema? sp. In 1954, G.R. Scott and R. Van Horn collected bivalves possibly including Nuculana sp., Pleu-rophorus sp., and Schizodus sp. from a few feet above the Lykins-Lyons contact. We hypothesize that all of these early published and unpublished collections came from the Falcon Member of the Lykins Formation, an assignment further intimated by the presence of a comparable bivalve- (Alula squamulifera, Mya-lina wyomingensis) and gastropod-bearing coquina (Fig. 8B) in the Falcon Member at a locality midway between the sections of Butters’ and Thompson’s original localities (Willhour, 1958). This is corroborated in the “fossiliferous” horizon documented in Iglehart’s (1948) Perry Park section, which is not only at the right level to be the Falcon Member, but was actually measured by W.O. Thompson, collector of the above-mentioned fossils we examined in the U.S. Geological Survey collections, together with one of his colleagues. The only other known horizon in the Lykins Formation that has macrofossils is a horizon in the upper Forelle Member, near our section F4. Here, bivalves and high-spired gastropods occur in a thin coquina that was deposited in between stromatolite heads (Wiggs, 1986).
Despite these fossil occurrences in the lower carbonates of the Lykins Formation, the unit is primarily a siliciclastic succession, with over 90% of its strata comprising maroon-red siltstones and very fine sandstones. Fossils have not been documented previously from these redbeds. We found that sandstone-mudstone partings from the lower portions of the Lykins Formation occasionally preserve sole markings and trace fossils that represent sand casting of burrows, tracks, or trails produced in underlying muds. Traces are dominated by simple horizontal burrows preserved in convex epirelief on bed soles, including forms such as Palaeophycus and Planolites (Fig. 8). One float specimen of vertebrate tracks was collected, and may come from the Red Hill Shale at least tens of meters above the Park Creek Member.
Microfossils are also rare. Ogden (1958) was the first to report microfossils from the Lykins Formation, noting benthic foraminifera belonging to Opthalmidiidae in the Falcon Member at Garden of the Gods, Colorado. Also from the Falcon Member, but at localities nearer our sections F5 and F6, Wiggs (1986) reported unidentifiable uniserial foraminifera. Higher in the succession, in the Forelle-equivalent Day Creek Dolomite (see section S1 in Fig. 2), Ogden (1958) noted possible codiacean algae. Van Horn (1957) also discovered algae in the unit, reporting the calcareous dasycladacean alga Mizzia minuta from the top of the Forelle Member at a site in Golden, Colorado.
To augment these earlier discoveries, we systematically surveyed 420 thin sections from our reference sections for fossils, and also disaggregated samples, described below. In the Falcon Member, we identified demineralized bivalves preserved as ghost cast and mold structures or steinkerns, and a foraminiferal assemblage dominated by small attached coiled tests (Fig. 8). From nearby localities in Wyoming, Chen and Boyd (1997) described nearly identical foraminifera, along with scaphopods (Plagioglypta sp.), bivalves (Permophorus sp., Schizodus sp., Wilkingia?), high-spired and conispiral gastropods, and a bellero-phont. But these fossils appear to occur stratigraphically higher, and from our examination of the section, the fossils come from what we cautiously interpret to be the transition beds that grade upward into the lower Forelle Member (illustrated in Figs. 3A and 4E). The Wyoming foraminefera are compared to Pseu-doglomispira or possibly Glomospiroides. In the lower Forelle Member in Colorado, we identified thin-walled concave shells and hooked elements interpreted to be ostracod valves, spheroidal spinose fossils akin to calcispheres or calcified charophyte oogonia, and some unidentified algae-like fabrics. From the upper Forelle Member at a locality near our section F4, Wiggs (1986) identified bivalves and high-spired gastropods in lenses between stromatolite heads. In the Park Creek Member, we found structures akin to siliceous monaxon sponge spicules. Unfortunately, none of these fossils are markedly age or environment diagnostic; even the foraminifera represent opportunitistic groups that occur in a wide range of marine depositional environments.
Acid digestion of eight limestone and dolostone samples from the Greenacre Lentil through Park Creek members of the Lykins Formation revealed a spartan assemblage of phosphatic microfossils, including platform- and single cone-type conodont elements (Fig. 9), fish teeth, and scales. All of the conodonts come from the Falcon Member, and provide new temporal and environmental constraints on the lower Lykins Formation. Specimens have a conodont alteration index (CAI) of 1.0, indicating the Lykins Formation is thermally immature and experienced limited burial. The most notable conodonts are Hindeodus exca-vatus, which ranges from latest Kungurian to Roadian, and H. permicus, which is known elsewhere from latest Kungurian to earliest Capitanian strata (Shen et al., 2012). The apparatus of H. excavatus has been described by Clark and Ethington (1962) and Behnken (1975) from the Falcon-equivalent Minnekahta Member of the Goose Egg Formation in Wyoming. The specimens are very similar to those described by Wardlaw (2000) from the Glass Mountains of west Texas in association with Sweetina festiva, transitional Jinogondolella nankingensis, Neostreptognathodus clinei, Pseudohindeodus ramovsi, and Sweetognathus bicarinum. Collectively, the west Texas fauna is latest Kungurian. However, the species from the Lykins Formation likely range higher. For example, a specimen of Sweetina cf. triticum from the Falcon is similar to prioniodinid and lonchodinid conodonts described from the Falcon and Forelle members of Wyoming (Pearson, 1970), which were form species now attributable to Sweetina. Sweetina triticum ranges from Roadian to early Wordian (Ward-law and Collinson 1986) in the Phosphoria Formation of Wyoming, a unit that has been lithostratigraphically correlated to the Lykins Formation of Colorado. The fused carina of the Mesogon-dolella species in the Lykins Formation is comparable to that in M. phosphoriensis of the Phosphoria Formation, which is latest Roadian to Wordian. Although we did not find any conodonts in our initial sampling of the Poudre and Park Creek members, Yochelson et al. (1985) described reworked Neognathodus, Neostreptognathodus, and Hindeodus typicalis and Paull and Paull (1983) described H. typicalis which is normally found just below Isarcicella isarcica from the Early Triassic Dinwoody Formation in Wyoming, a unit thought to be correlative to the Park Creek Member (Schock, 1981).
Phosphatic actinopterygian remains include palaeoniscoid ganoid scales and three types of teeth (Fig. 10). These forms are comparable to taxa reported from the Capitanian of west Texas (Ivanov et al., 2013), but the full range of such taxa are not known. The ganoid scales, especially those in Figure 10C, are comparable to Variabilis sp., a taxon not only known from west Texas, but from the Ufimian (late Kungurian) of Russia. Thus the fish remains support a Roadian age, as indicated by the con-odonts for the Falcon Member.
Implications of Paleontological Results
These fossils help constrain the environments of deposition of the Falcon and Forelle members, with the bivalves, bellero-phonts, conodonts, foraminfera, palaeoniscoids, and actinopter-ygians indicating deposition in fully marine to restricted marine environments—rather than in alkaline lakes where stromatolites also form. For example, the conodont Sweetina is typical of shallow-marine environments and Mesogondolella is a more open marine form. The ostracods, calcispheres, gastropods, and unidentified invertebrate and vertebrate tracemakers are less diagnostic. They may also occur in freshwater to brackish settings, including hypersaline continental settings—places where stromatolites may also form. The conodont Hindeodus is typical of shallow water facies, which is consistent with the restricted marine to intertidal stromatolitic facies from which these samples were collected, and the abundant evidence for shallow water depths and intermittent exposure witnessed by both the carbonate and clastic strata of the lower Lykins Formation (e.g., Fig. 5). The specimens’ small size and low abundance are comparable to material that one of us (CMH) recovered from stromatolitic facies near the Permian-Triassic boundary at Curak Dag in Turkey. It is possible that juvenile Hindeodus conodont animals sheltered around these stromatolites. The close association of high-spired gastropods and bivalve coquinas with domal to columnar stromatolites supports this working hypothesis, and suggests inter-head facies might be fruitful areas to explore for fossils higher in the Lykins Formation succession. The Lykins Formation’s fish remains also occur in two of the units’ bivalve-dominated coquinas that abut or underlie stromatolites and gypsum-anhydrite beds that imply they lived in a restricted setting. At one of these localities, all of the fish teeth are low conical forms (Fig. 10, parts O-Q) and many show significant wear (Fig. 10, parts N-R). Given the shape and wear, it is hypothesized that these fish fed on some of the invertebrates in the coquina.
Considered together, the conodonts and fish remains provide the best lower age constraints for the Lykins Formation. They support a Roadian (i.e., early Guadalupian) age for the Falcon Member but do not exclude a Wordian to Capitanian (i.e., later Guadalupian) age. This age assignment is of interest because the Roadian spans the onset of a global Guadalupian-spanning transgression (see summary in Henderson et al., 2012). Given its interpreted marine deposition and its marine macrofossils, the Falcon Member is thought to be correlative with this Guadalupian, or basal Middle Permian transgression. Widespread bioherm development and faunas of the thick, regionally extensive overlying Forelle Member, often correlated based on lithostratigraphy to Wordian (middle Guadalupian) strata in western Wyoming, could represent the acme of this Guadalupian transgression during Wordian to Capitanian time. In light of the tectonic quiescence of the region’s Pangean landscape, and the lack of indicators of large or proximal paleotopographic highs during Lykins Formation deposition, it is possible that the thick, regionally extensive Forelle carbonates represent the late transgressive phase or highstand of the Guadalupian transgression. Breccias, slumps, and paleokarst that locally cap the Forelle Member (see Stop 1; Fig. 3) could represent a major sequence boundary reflecting lowstand conditions associated with the Middle-Upper Permian or Guadalupian-Lopingian boundary (latest Capitanian to earliest Wuchiapingian; see summary in Henderson et al., 2012).
Surprisingly, no identifiable fossils other than the stromatolites are known from the limestones of the Poudre Member, Greenacre Lentil Member, nor the Lykins Formation’s basalmost carbonates, nor were they found in thin sections of the siliciclastics. Absence of fossils in highly oxidized continental to marginal-marine clastics is not surprising, although the absence of conchostracans is odd, given that they seem to be a relatively common element of Permian-Triassic continental lacustrine redbeds (e.g., Kozur and Weems, 2010). The barren nature of these carbonates is more perplexing, especially if they represent deposition or precipitation during regional flooding events. Absence of fossils in these strata could imply poor preservation due to exposure and meteoric diagenesis, although ghost structures are known from the succession and would be expected to be visible, and phosphatic fossil elements ought to have been recovered in our preliminary acid digestions. Another possibility is that these carbonates, unlike the Falcon and Forelle members, could represent precipitation in an alkaline lake or in a suite of salinas, sabkhas, or alkaline lagoons that inhibited biomineralized organisms from thriving.
Further work is needed to resolve these issues, but unfortunately the Lykins Formation lacks many microfossil-amenable lithologies, like organic-rich early diagenetic cherts or coarse siliciclastic lags. In our field surveys, we did identify two gray-to-black shales from the base of the Falcon Member, and samples of these were digested in hydrogen flouride in the search for paly-nomorphs. Both were barren.
Chemostratigraphic profiling of the Lykins Formation’s carbonates offers opportunities to identify local or pervasive episodes of meteoric diagenesis, and to interpret secular changes in seawater chemistry that are not apparent in thin section or in trace element profiles. If a robust chemostratigraphic framework for the Lykins Formation can be constructed and matched to the global 513C curve, it might permit closer integration of Lykins strata into the global database for the Permian-Triassic transition, perhaps including direct comparison of individual units and their contained information (e.g., fossils, unconformities, pole positions) to sections elsewhere in the globe. Similarly, within the Rocky Mountain and High Plains region, chemostratigraphic profiles might be used to test the hypothesized isochroneity of individual Lykins units, or to determine if they represent susb-stantially time-transgressive facies.
At the time of this writing, our Lykins Formation 513C, 518O, and 87Sr/86Sr data are incomplete. Thus for the purposes of this field trip, we present a preliminary 513C chemostratigraphy (Fig. 11) for a portion of our studied sections, and urge caution in interpreting the results.
Across the seven sections (Fig. 11), the 513C composition varies widely between units, indicating that secular changes are recorded within the Lykins Formation. Some intervals, like the Lykins Formation’s basal carbonates, vary widely between outcrops, which together with moderate covariance of 518O with 513C (Fig. DR3 [see footnote 1]), hints at variable diagenesis between localities, and/or our inability to correlate these thin carbonate stringers to one another. In contrast, the Greenacre Lentil, a comparably thin, often silicified carbonate, displays remarkably tight clustering of values at or near +4 per mil, but exhibits the highest observed covariance with 518O (Fig. DR3). Given that none of the basal carbonates exhibits a comparable signal, and given how hard it is in the field to distinguish between the Greenacre Lentil and the basal carbonates, perhaps the Greenacre’s isoto-pic signature might be the main difference between them, even if it does potentially represent a regional event, such as meteoric diagenesis associated with exposure. Falcon Member limestones are rather negative (~-2 to -3 per mil), except in sections F5-F7, where they exhibit as much as a 6 per mil spread. Together with the lack of a systematic correlation between 513C and 513C values across the studied localities, it seems reasonable to suspect that meteoric diagenesis has acted on these localities, and/ or that there is a temporal trend within this thin carbonate, which is not distinguishable at the current sampling resolution. Perhaps the most interesting isotopic pattern occurs in the lower Forelle Member, where nearly every section exhibits a steady increase toward heavier values, from ~-3 per mil to ~3 per mil. Some sections, like F13, exhibit a nearly 8 per mil decrease in values at the top of the unit. Given the possibly short-lived nature and magnitude of this negative excursion, the meteoric diagenesis may have removed the heavier carbon, despite the lack of a significant covariance with 518O. The uniformly lighter 513C values in the upper Forelle Member are interesting given the widespread occurrence of paleokarst, solution breccias, and related exposure features at the top of the Forelle Member. Perplexingly, portions of its 518O profile are remarkably stable (Fig. DR3), akin to carbonate successions where meteoric diagenesis is minimal. Values in the Poudre Member are centered at 0 per mil but sections F11 and F12 show wide scatter, a change to light values at the top of the unit, and exhibit 513C values that covary moderately with 518O. The Park Creek Member is uniformly negative with values tightly clustered between -3 to -6 per mil, but all sites exhibit a trend toward progressively negative values at their top. It is also the only member of the Lykins Formation in which all 518O values are highly negative, suggesting that diagenesis may have been uniform throughout the region at this time, strongly influencing the carbonate values.
Despite these challenges, the consistency of Park Creek 513C values, like the consistent positive increase through the lower Forelle Member or the uniformly high positive values in the Greenacre Lentil, could prove useful as a way to distinguish between the otherwise sedimentologically and paleontologically identical Park Creek and Poudre members. This seems apropos in sections where the underlying Poudre is absent, like in the type section of the Lykins Formation at F7, where the member’s iso-topic characteristics might help fingerprint the unit better than conventional approaches.
Many of the reported values from this preliminary che-mostratigraphy are more negative than known signals from fully marine sections (e.g., Corsetti et al., 2005; Korte and Kozur, 2010; Henderson et al., 2012), and 513C and 518O covary in many samples. The aberrant values from the Lykins Formation highlight the need to identify the depth of meteoric diagenesis on a locality-by-locality basis, and to revisit field sites to couch the data with local subaerial exposure features. With such an approach, it may be possible to distinguish between altered and near-primary signals in an attempt to reconstruct a usable marine chemostratigraphic profile for the Lykins Formation. That said, it is possible that only the demonstrably marine units of the Lykins Formation (e.g., the Falcon and Forelle members) might contain marine 513C or 87Sr/86Sr signals useful for global correlation, and that other units (e.g., the basal carbonates, Greenacre Lentil, Poudre Member, and Park Creek Member) might have signatures that reflect mixed influence of marine and continental signals, or entirely diagenetic signals as hinted at above.
Synthesis and Next Steps
The Lykins Formation has long been overshadowed by its more fossiliferous, economically useful counterparts in western Wyoming, Idaho, and Utah, in part because it is challenging to interpret. The Lykins Formation’s sole economic value comes from its evaporites, which are locally mined for artist’s alabaster and as an amendment for cement and soils. Locally, the unit is unloved, because it continues to form hazardous karst as its evaporites dissolve (http://coloradogeologicalsurvey.org/geologic-hazards/subsidence-natural/case-studies/), which when combined with its paleokarst, poses major engineering challenges for road- and home-building (White, 2012), as well as dam projects in the region (Pearson, 2004). Further diminishing its stature, the unit is so fine-grained and poorly indurated that it is one of the primary local targets for landfills, reservoirs, and roadbeds (e.g., Merriman, 1960; Ellis, 2000).
Yet the Lykins Formation may contain a reasonable record of one of Earth’s most interesting pivot points. A century of solid lithostratigraphic and sedimentologic work has grounded our understanding of the Lykins Formation, and hinted at its role in understanding what happened in Colorado during the Permian-Triassic transition. The new detrital zircon, biostratigraphy, and chemostratigraphic data presented herein is incomplete, but begins to tighten our understanding of this interval a bit more. We present a summary of this information in Figure 12, as a graphical hypothesis that integrates our work with that of our predecessors. The take-away from this figure is that the lower Lykins Formation was deposited during the Early to Middle Permian, and its uppermost strata were likely deposited during the Early to Middle Triassic, leaving open the possibility that the boundary interval between these epochs may be present within the Lykins Formation. This hypothesis is imperfect, but we anticipate refining it with additional analyses.
Such work is under way, including approaches employed in this contribution as well as strontium isotope stratigraphy, magne-tostratigraphy, and basin modeling of the Lykins Formation and its equivalents. We hope some of these approaches are successful in narrowing gaps in our understanding of the Lykins Formation and equivalent units. In the meantime, please consider our data as preliminary and our interpretations as working hypotheses.
M. Dechesne, S. Egenhoff, and M. Pommer are thanked for reviewing this manuscript, although we take responsibility for any errors. We are grateful to the ranchers, homeowners, and agencies who permitted us to access exposures on their property. Many thanks to P. Plink-Björklund and I. Miller for catalyzing JWH’s initial interest in the Lykins Formation, and to L. Abbott, D. Boyd, K. Houck, V. Matthews, S. Mojzis, and B. Schumacher for fostering our understanding of key exposures. We are grateful to K.C. McKinney for facilitating our work in the U.S. Geological Survey invertebrate paleontology collections and to the U.S. Bureau of Reclamation and U.S. Geological Survey Core Research Facility teams for facilitating study of cores. Many thanks to J. Whiteley for help in the field; to L. Scott at the Colorado Geological Survey for the Colorado base map; to W. Allee, M. Deacon, M. Frasier, P. Marenco, and L. Smith for lively Permian-Triassic discussions; to S. Chakraborty and D. Dettman for assistance with isotopic analyses; to K. Honda and B. Wagner for loan of key theses; to S. Bala and T. Strong for conducting zircon mineral separation; and to F. Koether for making a heroic number of thin sections. JWH, KRW, and BLL’s work was funded by a KT Challenge grant, the Trinchera
Blanca Foundation, and the individual donors who directly support the Denver Museum of Nature & Science’s Department of Earth Sciences’ field and lab programs. Processing of micro-paleontology samples was supported by an NSERC Discovery Grant to CMH.
Figures & Tables
Unfolding the Geology of the West
Prepared in conjunction with the 2016 GSA Annual Meeting in Denver, Colorado, this volume contains sixteen guides to field trips in this rich geologic region. The four “Great Surveys” of the late 1800s ventured west to explore and document the region’s unknown natural resources and collect valuable geologic information. Many of the field guides in this volume, aptly titled Unfolding the Geology of the West, will cover the same hallowed ground as the early geologic expeditions. Organized into four sections, this volume spans some of the major subdisciplines of geology: (1) stratigraphy, sedimentology, and paleontology; (2) structure and metamorphism; (3) Quaternary landscape evolution; and (4) engineering and environmental geology.