Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
An age-depth model and revised stratigraphy of vertebrate-bearing
units in Natural Trap Cave, Wyoming
David M. Lovelacea*, Cory M. Redmanb, Thomas A. Minckleyc, Blaine W. Schubertd, Shannon
Mahane, John R. Woodf, Jenny L. McGuireg, Juan Ladenh, Kathleen Bittermani; Holly Heinigerj,
Lindsey Fendersonj,k, Alan Cooperl, Kieren J. Mitchellj, Julie A. Meacheni
aUniversity of Wisconsin-Madison, Dept. of Geoscience, UW Geology Museum, 1215 W. Dayton Street, Madison,
WI 53706, USA
bGrand Rapids Public Museum, 272 Pearl St NW, Grand Rapids, MI 49504, USA
cUniversity of Wyoming, Department of Geology and Geophysics, 1000 E. University Ave., Laramie, WY 82071,
USA
dCenter of Excellence in Paleontology and Department of Geosciences, East Tennessee State University, Johnson
City, TN 37614, USA
eU.S. Geological Survey, Denver Federal Center, Bldg 95 Luminescence Geochronology Lab, Denver, CO 80225,
USA
fGeologic Resources Division, National Park Service, Lakewood, CO 80235, USA
gGeorgia Institute of Technology, School of Biological Sciences, School of Earth and Atmospheric Sciences,
Interdisciplinary Graduate Program in Quantitative Biosciences, 311 Ferst Dr., Atlanta, GA 30332, USA
hLaden affiliation
iDes Moines University, Dept of Anatomy, 3200 Grand Ave, Des Moines, IA 50312, USA
jAustralian Centre for Ancient DNA, School of Biological Sciences, North Terrace Campus, University of Adelaide,
Adelaide, Australia, SA 5005, AU
kDepartment of Natural Resources and the Environment, University of New Hampshire, Durham, NH 03824, USA
lCooper affiliation
*Corresponding author.
E-mail address: dlovelace@wisc.edu (D.M. Lovelace)
1
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
ABSTRACT
Almost a half-century ago excavations at Natural Trap Cave (NTC) began to yield evidence of the
steppe paleoecology along the western slope of the Bighorn Mountains in north central Wyoming. The
first decade of fieldwork led to the discovery of a diverse fauna that existed at the end of the Last Glacial
Maximum. Stratigraphic deposits below the entrance of the cave were studied soon after excavations
began, but never formally published. Although stratigraphy, taphonomy, and depositional circumstances
were briefly discussed over the following years, little has been done to correlate the numerous
stratigraphic schemes used by various authors. In this study, four stratigraphic sections were measured
and analysed to establish an easily modifiable lithostratigraphic system of nomenclature. We provide the
first correlations of all stratigraphic nomenclature used throughout excavations at NTC to facilitate
comparisons with current and previous collections and publications. By leveraging more than 100
radioisotopic dates we developed an age-depth model and chronostratigraphic framework to further
interrogate spatiotemporal relationships between strata, paleoenvironmental proxies, and fossil
assemblages. Deposition is shown to be discontinuous; sediment accumulation in the study area is
restricted to the buildup through peak penultimate and Last Glacial maxima. More recent (<10 ka)
Holocene deposits unconformably cover the eroded surface of underlying Pleistocene strata. There is
active reworking of sediments with transport and deposition of reactivated sediments within the Lower
Chamber. We note that the two hiatuses coincide with interglacial periods and may reflect changing
depositional circumstances within the cave such as extended periods of non-deposition, erosion, or bypass
(possibly leading to deposition in the Lower Chamber). Contrary to previous reports, we demonstrate that
it is unlikely a prominent snow cone existed or contributed to the pattern of sediment and fossil
distribution within the study area, furthermore, we do not observe a continuous Pleistocene-Holocene
transition in the study area. Further stratigraphic work will be needed to better understand the
2
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
interrelationship between Main and Lower chamber deposits and the evolution of sediment accumulation
in NTC.
KEYWORDS:
Stratigraphy; Age-Depth Model; Last Glacial Maximum; Penultimate Glaciation, Sedimentology; Cave
Survey; Geochronology; Natural Trap Cave
1. Introduction
Upper Pleistocene sediments in the Rocky Mountain West record environmental and
ecological histories during the last major advance and retreat of continental and alpine glacial ice
(Illinoian–Wisconsin and Bull Lake–Pinedale glaciations; Nelson et al., 1979, Pierce, 2003;
Benson et al., 2005; Pierce et al., 2018). Changing climate across this time period altered
ecozones and shifted ranges for flora and fauna, leading to extirpation or extinction of the
majority of vertebrate megafauna and establishment of non-analog floras and faunas (Graham
and Grimm, 1990; Stafford et al., 1999; Koch and Barnosky, 2006; Stewart, 2009; Kohn and
McKay, 2012). Studying these evolutionary events is often hindered by taphonomic biases (e.g.,
hydrodynamic sorting, variable deposition, time-averaging, and rates of accumulation) that make
community level sampling more difficult (Olsen, 1980). High-resolution stratigraphic and
temporal control can limit (but not eliminate) biases, although novel circumstances such as traps
and ashfalls can minimize these biases and more faithfully capture signals of local ecology
(Rogers and Kidwell, 2007). Many Late Pleistocene fossil localities in North American preserve
diverse vertebrate assemblages (Kurtén and Anderson, 1980), and some sites are particularly
well known for dense accumulations and/or diversity, for example: Rancho La Brea, CA; the
3
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Mammoth Site, SD; and the Ziegler Reservoir fossil site, CO (e.g., Agenbroad and Mead, 1994;
Akersten et al., 1983; Sertich et al., 2014). These fossil deposits provide an extraordinary
opportunity to answer far ranging questions about the evolution and extinction of local flora and
fauna.
Natural Trap Cave (NTC) is another significant fossil locality, known for its
exceptionally well preserved Late Pleistocene vertebrate fossils (e.g., Martin and Gilbert, 1978a
& 1978b; Gilbert and Martin, 1984; Wang, 1988; Chorn, 1988). The cave contains fossils that
span from late Illinoian to recent (i.e., ca. 150,000 ka - present); however, the vast majority of
pre-Holocene specimens coincided with the last Pinedale glacial maximum (~23,400 to 29,200
cal yr B.P.; Pierce, 2003), which terminated around ~17,000 ka (Benson et al., 2005).
The fossiliferous deposits of NTC were first collected in 1969 (Rushin, 1973), subsequent
excavations occurred throughout the 1970s and into the early 1980s (Martin and Gilbert, 1978a;
Gilbert and Martin, 1984; Wang and Martin, 1993; Meachen et al., this issue); NTC was heavily
sampled during this 11-year period (Wang and Martin, 1993; Meachen et al., this issue). Over the
course of three years (1976-1978), geologist John Albanese developed seven cross sections, a
description of sediments (including compositional analyses), and a working stratigraphic
nomenclature for the cave (Albanese. 1976 Unpublished Report).
Although subsequent authors have drawn extensively from Albanese’s unpublished data
(with some modifications in nomenclature including a reversal of the original numbering
scheme; e.g., Martin and Gilbert, 1978a; Gilbert and Martin, 1984; Wang and Martin, 1993;
Williams, 1993), no detailed stratigraphic work of the cave has been published to date.
Furthermore, stratigraphic nomenclature was not consistently applied throughout the course of
4
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
study at NTC, leading to confusion and difficulties in reconstructing previous excavations and
placing collected material into a coherent stratigraphic and temporal context. Given the success
of earlier excavations and the importance of a well-preserved repository of Late Pleistocene
paleoenvironmental and paleoecological signals, a re-evaluation of stratigraphic and temporal
relationships of deposits within NTC is warranted. After a 30-year hiatus, fieldwork at NTC
resumed from 2014 to 2018 (Meachen et al., this issue), leading to new insights into
paleobiogeography, depositional history, palynology, phylogeny, and an updated geochronology.
This paper aims to correct any confusion by establishing a correlation between early
published and unpublished accounts of sedimentary deposits within NTC, as well as reconcile
historic and recently acquired 14C determinations in order to construct an age-depth model. Our
age-depth model provides a more comprehensive understanding of chronostratigraphic
relationships and creates a more stable foundation upon which paleoecological inferences can be
built, and previous hypotheses can be tested.
2. Geographic and historical context
Natural Trap Cave (NTC) is a karst pit cave located on the western slope of the northern
Bighorn Mountains northeast of Lovell, Wyoming (Fig. 1) on a plateau (aka Little Mountain) at
an elevation of 1,512 m (4960 ft). The cave entrance is situated on a narrow ridge that is sloping
westward and flanked by deeply incised ephemeral surface drainages. It should be noted that the
geographical depression containing the cave entrance represents a very limited catchment (< ca.
40,000 m2). In 1973, the Bureau of Land Management placed a steel gate across the 6 x 8.5 m
horizontal entrance to protect cave resources and prevent accidental falls. A 3.5 m vertical drop
5
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
below the gated entrance ends at a 2 m wide ledge that runs along the southwestern edge before
transitioning to a final vertical drop into the main chamber where the current chamber floor is
24.5 m below the grate. The roughly bell-shaped main chamber is ~43 m in diameter with a
perimeter that is subcircular and uneven. The cave was originally mapped between 1969 and
1972 (Hill et al., 1976, p.86-87). In 2016 the cave was re-surveyed with improved techniques
(see section 3.1 below).
The bedrock geology of Little Mountain plateau largely consists of the clastic-dominated
Amsden Formation (Early Pennsylvanian; May et al., 2013) that overlies the Early-Middle
Mississippian Madison Limestone Formation (Lageson et al., 1979; Sonnenfeld, 1996). The
geologic setting of the cave, is an important component to the stories for the stratigraphy and the
latest efforts of geochronology, and we refer the reader to Mahann et al., (this issue) for a more
thorough examination of the importance of Amsden and Madison contributions to in-cave
sediment.
Geographically, the cave is positioned south of a gap that developed between the
Laurentide and Cordilleran ice sheets in central North America (Fig. 1) at the end of the Last
Glacial Maximum (LGM) ~15-14,500 cal ka BP (Dyke, 2004). This corridor was open prior to
the coalescence of the two ice sheets >~23 ka ago (Heintzman et al., 2016), and some taxa that
spanned this passage would have been isolated north and south of the joined ice sheets for nearly
10,000 years. Subtle faunal interchanges between Beringia and North America are known to
have occurred when this ‘ice-free corridor’ re-emerged uniting taxa divided by the coalesced ice
sheets; NTC is ideally situated to capture that signal (Shapiro et al., 2004; Meachen et al., 2016;
Pedersen et al., 2016; Heintzman et al., 2017; Wang et. al, 2021). Although the entrance to the
6
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
cave was certainly known to the earliest people in the region (Crow Elder Grant Bulltail, Pers.
Comm to JAM), only one known artifact has been found: a red-jasper biface that was surface
collected immediately below the cave entrance. A putative atlatl shaft was reportedly found in a
packrat midden near the surface (see Martin and Gilbert, 1978a), but this artifact was later
determined to more likely be a coup stick and a relatively recent addition to NTC ( Meachen et
al., this issue).
2.1 Land acknowledgment
As researchers studying the geologic history of NTC, we must acknowledge that we are
visiting the ancestral lands of the Apsaalooké (Crow), Očhéthi Šakówiŋ (Great Sioux Nation),
and Tséstho’e (Cheyenne), lands that were forcibly taken from them by white colonizers during a
period of extensive genocide. NTC is situated between Medicine Mountain, and the Pryor
Mountains, a region that remains culturally significant to many Indigenous people. We recognize
the sovereignty of Indigenous people throughout the region whose history stretches millennia
into the past and who are an active part of the present. Although we may employ western science
to frame stories hidden in rock, we must not overlook Native ways of knowing the history of the
region as well. There is much to share and much to learn.
3. Methods
Original fieldwork was used in conjunction with unpublished data such as field notes,
collection data, field and analytical reports (i.e., Albanese. 1976 Unpublished Report), and other
communications. These data were obtained from a variety of sources including the University of
7
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Kansas Biodiversity Institute Division of Vertebrate Paleontology, Bureau of Land Management,
and personal collections. These data can be found at http://dx.doi.org/10.17632/f8frrskzpz.1, an
open-source online data repository hosted at Mendeley Data ([dataset] Lovelace et al., 2021).
3.1 Cave surveys
The Vedauwoo Student Grotto initially surveyed NTC between 1968 and 1972 using the
steel tape and Brunton compass method. Their finished cave map was published in a compilation
of Wyoming caves (Hill et. al., 1976, p. 86-87). Members of the 2016 field party were granted
permission by the BLM to re-survey the cave using more accurate mapping methods, including
DistoX and LiDAR technology. The DistoX survey used a modified Leica laser distance meter to
measure distance, azimuth, and vertical angle. The instrument was calibrated for local
declination prior to the survey. The LiDAR survey used a Zeb REVO to scan the cave; scanning
and post processing were conducted by H2H Geoscience Engineering, Troy, New York. The
principal goal was to determine if alternate entrances into the main chamber existed, produce a
volumetric reconstruction of the cave, and ensure all observed passages were mapped in more
detail.
3.2 Stratigraphy
Two unpublished stratigraphic schemes were devised (and used extensively during early
excavations) by S.A. Chomko/K.A. Lippencott and J. Albanese, an alphanumeric ‘zone’ and
‘unit/sub-unit’ based nomenclature, respectively (Appendix Figure 1). Martin and Gilbert
(1978a) were the first to publish a simplistic profile outlining seven discrete stratigraphic levels.
Their descriptions of each ‘stratum’ are minimal, and no correlation to previous unit
8
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
nomenclature(s) used during excavation was provided. Gilbert and Martin (1984) published
another schematic, but no unit designations or descriptions were included. Wang and Martin
(1993) proposed several hypotheses related to the origin and taphonomy of fossiliferous cave
deposits, but the stratigraphic nomenclature used was that of Albanese (1976 Unpublished
Report). The complexity of the deposits below the entrance of NTC required us to correlate new
observations with previous stratigraphic work to avoid confusion between studies; something
that had not been attempted before this paper.
During the 2016 field season four stratigraphic sections were measured with some of the
units traceable between each section. Fresh sediment was exposed in vertical profiles at each
section in order to observe a non-weathered surface. Grain-size, texture, color, thickness,
sedimentary features and lateral relationships were recorded. Photography was used to record the
lateral continuity of units across several meters and along the transition between north-south and
east-west trending walls. Work was conducted under Bureau of Land Management permit
PA-13-WY-207.
3.3 Geochronology
3.3.1 Tephrochronology
At least five discrete volcanic ash accumulations have been observed within NTC
deposits over the course of study (1972-2017); however, the lateral and temporal relationship of
some ash deposits remain poorly constrained. Most ashes appear as primary deposition and can
be stratigraphically traced through the deposits, however there is evidence of some
remobilization of ash within the cave. Albanese (1976 Unpublished Report) was the first to note
9
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
the presence of an ash (unmapped) along the southwall of the 1976 NTC excavation and
remarked that little had been done with the ash up to that point; that still remains the case and
should be a focus for future work within the cave. As excavations continued between 1977 and
1983, three new ash deposits described as a white volcanic ash layer overlying two grey volcanic
ash layers were reported (Gilbert [unpublished NSF report] see [dataset] Lovelace et al., 2021;
Gilbert et al., 1980; Gilbert and Martin, 1984) from units discovered below the lowermost
deposits of Martin and Gilbert (1978a). In an abstract, Gilbert and others (1980) provide a ca.
110 ka BP fission-track date for the stratigraphically youngest of those three ashes (Mahan et al.,
this issue). Vertebrate remains, including Ovis,Equus,Lepus, along with microvertebrate
material (Gilbert, 1980 BLM Report [unpublished] see [dataset] Lovelace et al., 2021; Gilbert
and Martin, 1984; Wang and Martin, 1993) were found below the undated lowermost ash.
During the course of recent fieldwork two ashes were encountered low in section, rather
than three as noted by Gilbert and Martin (1984). The ashes were grey in appearance, and we
make the assumption, in the absence of direct correlation with previous observations, that these
represent the lower two grey ashes of Gilbert and Martin (1984). Two additional ashes were
observed, one on the southwall at the surface of the cave deposits that is likely the ash referenced
by Albanese (1976 Unpublished Report), and a 20 cm thick, but laterally limited (<1 m) ash
accumulation on the north wall just below the Upper Red layer (described below).
The middle of the ‘lower three ashes’ has a reported 40Ar/39Ar date of 129.6 ±17.4 ka
(Mahan et al., this issue), sanadine was absent from the lowermost ash sample but zircons were
present in a grab sample taken for palynology. A 238U/206Pb age of 140 ± 7 ka was determined for
the lowermost ash (Clementz et al., this issue) indicating fossil accumulations at NTC extend
10
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
into the Bull Lake Glaciation (Pierce, 2003). These two dates are used to help temporally
constrain the lowermost sediments observed in the excavation.
3.3.2 Radiocarbon
Clear stratification is present in the cave, and several successive units contain datable
vertebrate fossils (Martin and Gilbert, 1978a; Gilbert and Martin, 1984). This has provided the
foundation upon which larger scale interpretations of timing of floral and faunal changes have
been made (e.g., Gilbert and Martin, 1984; Wang, 1988; Wang and Martin, 1993). Specimens
used for radiocarbon analyses were from one of two institutions: UW: University of Wyoming
Geological Museum; KUVP: University of Kansas Biodiversity Institute Division of Vertebrate
Paleontology.
DICARB (and 2 unknown) Determinations 1973-1980: Eight 14C dates were
determined from NTC specimens between 1973 and 1980. The first 14C determinations from
bone found within NTC were analysed in 1973, and later published as 10,920 ±300 yr BP and
12,770 ±900 yr BP (Martin et al., 1977). No information other than the ages were published and
no original data pertaining to the analytical lab, specimens, methods or results have been found;
these dates were later perpetuated (Martin and Gilbert, 1978a) or cited as ‘Loendorf, personal
communication’ (Gilbert and Martin, 1984; Williams, 2009). Considering the lack of supporting
information, analytical lab, or methods, we consider these two ages suspect. Martin and Gilbert
(1978a) report a ‘17,620 yr BP’ and ‘20,170 yr BP’ age, also with no error or associated data.
The ‘20,170’ age remains unverified, however, the ‘17,620’ age is likely analytical number
DICARB 690, published with an associated error ±1490-1820 yr BP (Martin et al., 1979) but is
later misidentified as DICARB 1690 (Gilbert and Martin, 1984). Another typographical error
11
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
occurred again when the age determination 14,670 ±670/730 yr BP (DICARB 689; Martin et al.,
1979) was misidentified as DICARB 1689 (Gilbert and Martin, 1984). The datasheets for
DICARB 689 and 690 have not been found. Three additional radiocarbon ages were reported in
Gilbert and Martin (1984): 17,870 ±230 yr BP (DICARB 1686), 20,250 ±275 yr BP (DICARB
1687), and 21,370 ±830/920 yr BP, the correct determination for DICARB 1689; [dataset]
Lovelace et al. (2021).
Of the eight historical radiocarbon dates listed above only the last three dates have
preserved datasheets with specimen information, however both DICARB 1689, and 1687 contain
more than one specimen combined “at the sender's request” to produce enough datable material
([dataset] Lovelace et al., 2021). We do not consider the mixing of two specimens a reliable
method to obtain a radiocarbon date and we consider the ages of these two ‘mixed sample’ dates
to be suspect. As such, DICARB 1687 (20,250 ±275 yr BP) is the only radiocarbon age from
NTC that has been verified (e.g., a known datasheet exists) and is based on a single specimen
with stratigraphic control prior to 2001. DICARB age determinations have been questioned
(Reuther and Gerlach, 2005), though their incorporation into larger datasets used in Bayesian
models may still have utility (Hamilton and Krus, 2018).
OxA Determinations 2001-2005: Forty-three AMS 14C ages were determined from NTC
specimens at the Oxford Radiocarbon Accelerator Unit (ORAU) between 2001 and 2019.
Twenty-two of those ages have been referred to in publications (Shapiro, 2004; McKay, 2008;
Williams, 2009; Salis et al., 2020; Perri et al., 2021; Salis et al., 2021; Table S1). LLNL
Determinations 2015-2018: Fifty-eight (including two inter- and two intralab replicate samples)
AMS 14C ages were determined from NTC specimens at the Lawrence Livermore National
12
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Laboratory CAMS facility (LLNL; Table S1). KCCAMS Determinations 2015-2018: Four
AMS 14C ages were determined from NTC specimens at the Keck Carbon Cycle AMS Facility
University of California-Irvine (KCCAMS; Table S1). See Appendix for methods used for OxA,
LLNL, and KCCAMS determinations.
3.4 Age-Depth Model
Stratigraphic units within the cave represent a complex history varying between periods
of relatively continuous deposition, subtle to significant erosion, or relatively stable surfaces with
low sediment accumulation. The contacts between stratigraphic units range from gradational to
unconformable. Although clear boundaries exist and can be traced across many meters of
excavated surfaces, the units themselves are non-uniform, varying in thickness with lateral
changes in lithology and contact type. For this reason, we employ an age-depth model using R (R
Core Team, 2017) with Bacon (Blaauw and Christen, 2011; R package ‘rbacon’ v. 2.5.3) based
on the IntCal20 calibration curve (Reimer et al., 2020) and a simple procedure for correcting
age-depth relationships between sections with high variability in unit parameters such as depth
and thickness (Pelton et al., 2017).
Dated materials have been collected from NTC since the earliest excavations, but not all
have had their stratigraphic position recorded in a way that allows for simple age-depth
relationship determinations. The first 14C dates and tephrochronology from NTC (~1977-1980)
gave some indication that the cave’s depositional history extends at least to 110 cal ka BP (early
Sangamonian) and it was proposed that there was relatively continuous deposition to the present
(Martin and Gilbert, 1978a; Gilbert and Martin, 1984). This has provided the foundation upon
which larger scale interpretations of timing of floral and faunal changes have been made (e.g.,
13
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Gilbert and Martin, 1984; Wang, 1988; Wang and Martin, 1993). Of those early 14C
determinations, only one of the eight dates has verified analytical results and is based on a single
specimen. To test the hypothesis that NTC preserves a continuous record of deposition that spans
from (at least) 110 cal ka BP to present, we leverage a series of 14C determinations (n=105;
including 4 replicate and 79 previously unpublished determinations; Table S1) measured over the
last two decades, as well as recent 40Ar/39Ar and 238U/206Pb dates for material older than 45,000
years to constrain boundaries within an age-depth model. Most of the dated material does not
have strict stratigraphic control, but all dates can be constrained to one of 11 lithostratigraphic
units defined below.
The model presented below (section 4.3) is based on 22 radiometric dates with strict
stratigraphic control; 21 dates are from the same analytical lab. Two significant hiatuses are
interpreted from the strata and are included in the model, the youngest hiatus defines the
boundary between the Upper Beds and Primary Beds; it is a clear erosional surface that incises
through the Primary Beds along the South Wall. The second hiatus was placed at an abrupt
change in lithology that we infer to be the boundary between the relatively continuous deposition
of the fine-grained dominated Lower Pebbly Grey bed and the overlying boulder-sized
breakdown of the Maroon Blocks (269 cm depth, Reference Section). Through an iterative set of
modelling runs, we determined that the age-depth models should not assume conformal
sedimentation. Despite the depth of dated material collected from the cave, we screened dated
materials for positional accuracy and provenience, allowing for extrapolation to younger and
older unconformable boundaries (Blaauw and Christen, 2011). Based on our selection criteria,
we chose 22 of the >100 radiometric dates as appropriate for age-model determinations. Those
14
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
22 samples have the strongest stratigraphic control, the remainder are limited to unit-level (e.g.,
their position within the unit is not well constrained). In order to make sure we were not
significantly altering model results by excluding 81 dates (a suggestion outlined by Hamilton and
Krus, 2018), we repeated the study with the inclusion of all NTC 14C determinations from the last
two decades, regardless of lab, as well as DICARB 1687. Given the lack of stratigraphic control
beyond unit-level for the majority of specimens, 14C determinations without strict stratigraphic
control were randomly distributed within their respective units (rather than stacking the dates
chronologically within each unit, which would reach beyond the information at hand). We
independently modeled two discrete sections (i.e., Upper Beds through Primary and Rainbow
Beds [with a hiatus at 40 cm], and Lower Beds [all strata below 269 depth in Reference
Section]). Age-depth relationships were estimated using 7,000-8,000 iteration model runs using
R (R Core Team, 2017) with Bacon (Blaauw and Christen, 2011; R package ‘rbacon’ v. 2.5.3)
based on the IntCal20 calibration curve (Reimer et al., 2020). Our assumptions were that
sedimentation is relatively constant over time, and that contacts between the stratigraphic units
were of similar age regardless of horizontal position on the cave floor.
4. Results
4.1 Survey
The results of the DistoX survey conducted in 2016 demonstrated a strong similarity to
earlier mapping done by the Vedauwoo Student Grotto (Hill et al, 1976, p. 86-87), though the
recent survey produced a more detailed map with a number of cross sections and map profile
views (Fig. 2; for complete map see Fig. S2). A LiDAR survey was also conducted to provide a
15
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
volumetric model of the cave system ([dataset] Lovelace et al., 2021). Although several domes in
the ceiling of the main chamber were explored, no additional entrances to the cave were
discovered; however, further exploration is warranted to more fully understand the evolution of
NTC.
During the 1985 excavation, an entrance into what is now called the Lower Chamber was
breached in the deepest part of the main excavation (Meachen et al, this issue). Wang and Martin
(1993) stated that the Lower Chamber contains Pleistocene fossils and it undoubtedly received
sediment inputs from the overlying room where excavations have been focused. This was
corroborated by brief forays into the Lower Chamber by members of the 2014 and 2016 field
crews. Fifty-two vertebrate fossil specimens were collected from the surface of the Lower
Chamber (2014: n=50; 2016: n=2; Appendix Table 1). An un-altered impression of a large bone
was also observed. It is possible that this is the mammoth femur that was removed in 1985 along
with a complete equid skull (Wang and Martin, 1993). Although the DistoX survey team did not
survey this chamber, only part of the Lower Chamber was included in the LiDAR survey: further
work is warranted to complete the survey of this portion of the cave system. The LiDAR survey
of the Lower Chamber provides the only quantifiable information on the length, volume, and
overall shape of this room and its spatial relationship with the overlying main chamber (Fig. 3).
4.2 Stratigraphy
There is abundant evidence that water was the primary agent in the deposition of the
various stratigraphic units in NTC. The presence of banding, gravel lenses, pronounced erosional
surfaces, and the distribution of disarticulated bones all point to the action of running water. Silt
16
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
and clay constitute the bulk of the cave sediments. Albanese (1976, Unpublished Report) also
noted that the majority of sediment (excluding obvious breakdown) is dominated by fine grained
clastic material exhibiting sedimentary structures consistent with deposition by moving water.
Sediment flux into the cave from the surrounding surface is likely sourced from the
erosion of proximal outcrops of Amsden Formation and soils on the plateau, as well as more
broadly sourced aeolian contributions (Stock et al., 2006). Additional sources include breakdown
and weathered host rock and paleokarst infill. Multiple surficial processes are likely providing
material that is derived from or subsequently blown into the catchment where it can be
introduced to the cave by surface water. For instance, aeolian transport of local fine-grained
sediment into the immediate catchment can be stored and then mobilized by surface runoff where
it is entrained and redeposited below the cave entrance.
Four measured stratigraphic sections demonstrate a relatively consistent pattern of
deposition, which can be divided into four fundamental beds that we label the Upper Beds,
Primary Beds, Rainbow Beds, and the Lower Beds (listed youngest to oldest). The four beds
were further differentiated into 11 identifiable strata that were assigned lithostratigraphic names
based on characteristic features (Table 1). An important criterion for determining stratum
boundaries was the traceability of a boundary across exposed surfaces.
The Upper Beds and Rainbow Beds are traceable from the South Wall section (Fig. 4A)
to the Reference Section (Fig. 4B), which is used to tie all sections together. The Primary Beds
are absent in the South Wall section due to erosion prior to the deposition of the Upper Beds that
unconformably overlie the Rainbow Beds. This unconformable contact is also traceable to the
Reference Section where the Upper Beds overlie the Cobbly Grey. The erosive surface can
17
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
further be traced to the Saddle Area (Fig. 4C) section where the Grey Silt and Blocky Brown
gradually appear as the degree of truncation decreases. This section terminates near the base of
the Primary Beds (lowest point of excavation at time of observation). The West Wall (Fig. 4D)
section encompasses the Lower Beds through the majority of the Rainbow Beds where it is
truncated within the Mottled Silt due to pre-1985 excavations that removed Primary Beds and
younger strata.
4.2.1 Unit descriptions
The Upper Beds are composed of two intervals, the Upper Cobbly and the Upper Red.
Both the Upper Cobbly and Upper Red beds preserve vertebrate remains, but only the Upper
Cobbly exhibits abundant pebble to cobble sized breakdown. The Upper Red is a conspicuous
red to pale reddish-brown clay-rich silt unit that varies from 2–20 cm in thickness, whose base
contains highly-weathered pebble to medium gravel sized breakdown blocks with a chalky
appearance (Fig. 5A-C). The lower contact of Upper Red is sharp in the western part of the
quarry and irregular but quite apparent along the south wall. This boundary exhibits a notable
erosive surface whose incision depth (≥ 1.5 m) increases from the Saddle Area section to the
South Wall representing a significant disconformity in both the Reference Section and the South
Wall stratigraphy (Fig. 4). Below the boundary, the Blocky Brown is the youngest unit of the
Primary Beds, and features well developed fine blocky peds that increase in size from mm to cm
scale over ca. 20 cm of depth. Argilans are not observed, though the diffuse and gradational
contact with the underlying clay-rich Grey Silt suggests that some degree of illuviation occurred
during pedegenisis. Numerous molds of fibrous plant remains were also encountered in both the
Blocky Brown and the Grey Silt, though no apparent root traces were documented in either.
18
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
There is a gradual transition between the Grey Silt and the Cobbly Grey units as breakdown
content increases rapidly in both size and abundance. This transition and the upper two units of
the Primary Beds are only visible in the Saddle Area, which is effectively the southern edge of
the pre-Meachen excavations on the west side of the main quarry. The Upper Beds and Primary
Beds make up the vast majority of the earlier excavation; Albanese (1976, Unpublished Report)
demonstrated a similar relationship between his Unit 4 (=Upper Beds) and Units 3 and 2 (=
Primary Beds).
The otherwise silt-rich Cobbly Grey is quite apparent because of the high concentration
of fine to very coarse gravel, with rare boulder-sized breakdown clasts. It is also noteworthy for
the high concentration of megafauna fossils - considerably more than any other unit in the cave
to date. The breakdown blocks are angular and show little sign of weathering, sorting, or
preferred orientation leading to a massive appearance. The base of the Cobbly Grey is marked by
a rather abrupt loss of the gravely component, although the contact is diffuse it does appear to be
relatively conformable with the underlying Rainbow Beds (as seen along the southern west wall
near the Reference Section, e.g. Fig. 5D).
Relative to the overlying Upper Beds, the Rainbow Beds do not contain nearly as many
vertebrate remains (Martin and Gilbert, 1978a; Gilbert and Martin, 1984). Gilbert and Martin
(1984) denote the presence of Panthera and ‘wolverine’ in the unit below the main bone bearing
unit (our Cobbly Grey), but little else. One notable Panthera specimen (KU4409) from
Albanese’s sub-unit 4 (our Mottled Silt) was dated to 27.8–28.7 cal ka BP (OxA 10078;
Williams, 2009). An additional 14C determination was recently acquired just prior to submission
19
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
yielding a 27.6–28.3 cal ka BP (CAMS 186413) date from a specimen of Vulpes (UW 54558)
collected in the uppermost Rainbow Beds (upper Mottled Silt).
The Rainbow Beds represent four relatively conformable units, three of which display
little lateral change in thickness; each unit maintains distinct and mappable characteristics. These
four units are traceable across the west wall, and all but the lowermost unit is present on the
south wall (Fig. 6). The youngest unit of the Rainbow Beds is the Mottled Silt, a clay-rich silt
with apparent, but diffuse, redox features consistent with fluctuations in saturation. There are few
randomly distributed breakdown blocks in the pebble to medium gravel size range within this
otherwise fine unit. The base appears to be conformable with the underlying Upper Pebbly Grey.
The top of the Upper Pebbly Grey unit is defined by the abrupt presence of thin (5–20
cm) laterally extensive medium to coarse gravel lenses that show some degree of sorting and
lateral accretion suggesting transportation by moving water. These lenses are interbedded with
finer grained clay-rich silt, often with mm scale laminations or very low angle cross beds.
Distinct color banding is seen in the fine grained intervals throughout this unit. The base of the
unit is defined as the top of a paired thin (2–5 cm) but persistent oxidized ‘red’ (10R 4/6) and
‘olive yellow’ (2.5Y 6/6) clay-rich silt layers that abruptly transition to another clay-rich silty
unit with abundant pebble to medium gravel breakdown clasts with little to no preferred
orientation or sorting giving the Pebbly Brown unit a more massive appearance (Fig. 6A-B). The
Pebbly Brown is seen deposited between the upper gaps of boulder-sized breakdown of the
underlying Maroon Blocks. The boundary between the Pebbly Brown and Maroon Blocks (Fig.
6B) has only been observed along the west wall; excavations have not extended deep enough to
encounter it elsewhere.
20
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
The Maroon Blocks are unique within NTC deposits. This unit is most striking in the
Reference Section where large (boulder-sized) breakdown with a notable red color, poor sorting,
and no preferred orientation are stacked ca. 70 cm thick. Immediately above the Reference
Section on the roof of the cave, near the gated entrance, is a portion of ceiling that exhibits the
same red blocks that compose (what appears to be) part of brecciated paleokarst infilling (Fig.
6C; Mahan et al., this issue). These blocks become rapidly less common north of the West Wall
stratigraphic section where the Maroon Blocks is only a few cm thick and composed of clay-rich
silt with the same red color as the isolated large breakdown. This unit, and the overlying units of
the Rainbow Beds dips 25-30° N30E below the excavated horizon (e.g., the slump, runoff, and
backfilled area of previous excavations where two rungs of an aluminum ladder are exposed, the
rest being buried by recent sedimentation) along the northern end of the west wall (Fig. 7). A
larger scale unconformity is hypothesized to exist at the base of the Maroon Blocks; regardless,
this is the boundary between the Rainbow Beds and Lower Beds.
The top of the Lower Pebbly Grey (Lower Beds) sees a return to deposits similar to those
observed in the Upper Pebbly Grey, with discrete gravely lenses interbedded with clay-rich silt
layers. A 2–5 cm thick charcoal-rich silty layer is a prominent feature of the Lower Pebbly Grey
(Fig. 6B,D). The Lower Beds are partially exposed at the base of the West Wall section and they
are exposed to bedrock in the Reference Section; however, due to temporary BLM permitting
restrictions the Lower Beds were not accessible for more detailed observations at the Reference
Section beyond a visual inspection from a distance. The base of the Lower Pebbly Grey is
21
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
constrained by the occurrence of volcanic ash ca. 40 cm below the laterally extensive
charcoal-rich layer.
Two volcanic ashes are present at the base of the Reference Section (Lower Ashes; Fig.
8A-B); these were sampled for geochronology along with an isolated ash (Fig. 8C) on the north
wall (Mahan et al., this issue; and Minckley et al., this issue). The ashes are greatly diminished
and diffuse at the base of the West Wall section. However, it is apparent that the ash was
ultimately deposited by moving water and not via airfall. Sedimentary features demonstrate
small-scale low angle bedding (Fig. 8B). Given the limited exposures and lack of access at the
time of study, the Lower Beds are the least well understood part of the NTC depositional
sequence.
4.2.2 Correlations
The correlation between the Saddle Area, Reference section, and South Wall sections exhibits a
significant erosive surface between the Upper Beds and underlying Primary Beds and Rainbow Beds (Fig.
4). The depth of the erosive surface increases with distance from west to east (Saddle Area towards the
South wall) along the southern portion of the excavation; this surface can be seen in the reference section
where the Blocky Brown and Grey Silt have been completely removed (Fig. 5D; Fig. 9). On the southern
end of the east wall the surface dips ~25°N, where its greatest observed depth extends into the Pebbly
Brown (Fig. 9). The remainder of the east side of the excavation (~495 W line) was covered with
excessive backfill or slumped material and fresh outcrop was not encountered for ~3 m until the 500N
line of the excavation (for excavation coordinate system see Meachen et al., this issue) where an
unconformity between the Upper Beds and strata that is temporally equivalent to the Primary Beds is seen
around 1477.1 m (~2.5 m below the same contact on the South Wall); this unit was not traceable across
22
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
the north or east portions of the excavation. Further work is needed to better understand the relationship
between this unit and those outlined in this study, but strata that are temporally equivalent to the Primary
Beds are present in the northeast corner of the current excavation (500N-515N and 490W-505W) that
unconformably underlie Upper Beds that contain a mixture of Holocene and reworked Late Pleistocene
vertebrates; this is supported by 14C dates.
The Reference section is ~2.5 m south of the West Wall section and units are easily traceable
along the exposed surface of the excavation. The strata are northward dipping at ~30° for the majority of
the exposed west wall (Fig. 7). Units of the Lower Beds maintain a relatively uniform thickness and
bedding tends to follow the contours of strata below. This is disrupted by the presence of the chaotic
assemblage of boulders in the Maroon Blocks, just above the entrance to the Lower Chamber. The
Maroon Blocks demonstrate a rapid decrease in thickness north of the entrance to the Lower Chamber.
This is linked to a decrease in the size and concentration of breakdown. Excavations from the 1970’s-80’s
were backfilled and can be clearly seen in cross section (Fig. 7).
4.3 Age-depth model
4.3.1 Radiocarbon ages
Radiocarbon has been the favored approach to NTC geochronology primarily due to the
plethora of well-preserved bones in the sediments. Over the course of the last two decades, three
different labs have analyzed more than 100 collagen samples extracted from NTC vertebrates to
determine the 14C age of those specimens (Table S1). Two samples (UW-51516 [OxA 37895;
CAMS 170343]; UW-51520 [OxA 37991; CAMS 170433]) were each analyzed at the Oxford
Radiocarbon Accelerator Unit (ORAU) and Lawrence Livermore National Laboratory (LLNL)
for an interlab comparison. Two other samples (OxA: 14903, 14904 & 14907, 14908) were each
23
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
replicated at ORAU and another pair (CAMS: 176891, 176897 & 176893, 176898) were each
replicated at LLNL for intralab comparisons. Interlab comparisons demonstrate a slightly
younger age reported for each LLNL sample relative to the ORAU results. Considering subtle
differences in lab protocol and instrumentation, this is not unexpected. Each of the interlab
replicates were very consistent with their respective laboratories. Intra- or interlab comparisons
with four samples analyzed at Keck Carbon Cycle AMS Facility University of California-Irvine
(KCCAMS) were not conducted.
We compiled the results of our analyses and a tabulation of the calibrated ages shows a
bimodal distribution (Fig. 10), one that contains post-Pleistocene ages from specimens collected
within the Upper Beds, and one that contains the vast majority of total analyzed dates ranging
from ca. 23–26 cal ka BP from the Cobbly Grey unit of the Primary Beds. Due to limited
exposure of the upper Primary Beds vertebrate material from the Blocky Brown and Grey Silt
were inadvertently excluded from 14C analyses. The Grey Silt is fully represented in the Saddle
Area though it rapidly becomes truncated towards the east and south, while the overlying Blocky
Brown is truncated by an erosional surface. We chose one specimen from each unit, both with
strong stratigraphic control to further constrain the maximum depositional age of Blocky Brown.
Although the specimen from the Blocky Brown did not yield enough collagen for analysis, the
Grey Silt specimen (UW 54757) yielded a 16.6–18.6 cal ka BP (CAMS 186411) date, the
youngest verifiable Pleistocene 14C date for NTC.
Three specimens from the Upper Beds have dates that are typical of Cobbly Grey
specimens (ca. 23.5, 23.8, & 24.5 cal ka BP [CAM 179248; 179245; 176891]) . We suggest that
24
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Pleistocene aged specimens collected in the Upper Beds demonstrate reworking of older deposits
and are co-mingled with syndepositional post-Pleistocene specimens; therefore, these
significantly older specimens are not representative of the age of the deposit. The concentration
of vertebrate fossils below the Cobbly Grey appears to be greatly reduced; however, recent
excavations have not yet been extended below the Cobbly Grey (e.g., in the Saddle Area, or on
the northeast corner), except in specific locations such as the South Wall where Primary Beds are
missing. At the South Wall location, the Pebbly Brown is currently the most prolific producer of
Pleistocene-aged vertebrate fossils, and these range from ca. 34–39 cal ka BP.
4.3.2 Boundary ages and sedimentation rate
Of the 101 14C age determinations (Table S1), 20 have depth-controlled stratigraphic
information, that is to say, their exact position relative to the top and bottom of a particular unit is
constrained. The remainder of the 14C ages are restricted to unit-level, but not necessarily
positionally within their respective units, as the motivation for sampling was for the
distributional ages of the faunal remains rather than focused stratigraphic age-depth calculations.
The 20 depth-controlled determinations (Table 2) are used as anchors for the age-depth model for
the Pebbly Brown and younger units. Two radiometric ages (determined from the lower two
ashes) are used to constrain strata below the Pebbly Brown. The median extrapolated age for the
base of the Reference Section in NTC is 154.5 cal ka BP, demonstrating that the deposits extend
(with recognizable breaks in deposition) from the penultimate glacial period (Bull Lake Glacial)
to the Holocene. Three distinct intervals of sedimentation are noted: an upper interval (Upper
Beds) that records Holocene accumulation (10.5 cal ka BP to present); a middle interval
25
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
(Primary Beds through Rainbow Beds) bound by an upper and lower hiatal surface records
deposition (53.0 to 17.2 cal ka BP) throughout the Pinedale Glaciation; and a lower interval
(154.5 to 132.0 cal ka BP).
Unlike the clear erosional surface of the upper hiatus (between the Upper Beds and
Primary Beds), the lower hiatus (between the Rainbow Beds and Lower Beds) is much less clear.
We hypothesize that the base of the Maroon Blocks, which exhibits the most dramatic change in
lithology, represents an event that initiates a new interval of deposition after a ca. 80 ka hiatus.
We used the basal depth of the Maroon Blocks (~269 cm) in the Reference Section as the
boundary for the second hiatus in the model. Although our model results suggest sedimentation
likely ceased somewhere around 132 ka BP, it should be noted that we did not include the 110 ka
BP age for an ash that was dated using fission track ([abstract] Gilbert et. al., 1980; Mahan et al.,
this issue); we have no stratigraphic control for that ash nor was it encountered during the course
of our study. Until further dates help to anchor the Maroon Blocks and the Lower Pebbly Grey,
we rely on extrapolated ages for this interval. Undoubtedly, this will be refined as new data
become available, though we feel the model honors the available data.
In the Reference Section sedimentation initiated around 153 ka with an accumulation rate
of around 115 yr/cm. Sediment accumulation ceases around 132 ka at the start of the lower
hiatus. The hiatus is terminated with the onset Maroon Block deposition ca. 53 ka. The rate of
sedimentation gradually increases from ca. 200 yr/cm to 110 yr/cm throughout the Rainbow
Beds. The transition to Primary Beds deposition is marked by higher concentrations of vertebrate
remains and medium to coarse gravel and exhibits a much more rapid rate of accumulation (35
26
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
yr/cm). There is an approximately 10 fold decrease in accumulation rate (325 yr/cm) between the
top of the Cobbly Grey and the truncated surface of the Blocky Brown. There is only one
confirmed 14C date (16.6–18.6 cal ka BP; CAMS 186411) from a specimen near the top of the
Grey Silt. Further sampling is needed in this interval to better understand the duration of the
hiatus. Sedimentation resumes above an erosive contact and exhibits a 250 yr/cm rate of
deposition, the second slowest in observed NTC deposits.
The above sedimentation rates and median extrapolated ages for major stratigraphic
boundaries are based on 20 well constrained 14C ages and two other radiometric ages. Hamilton
and Krus (2018) suggest the inclusion of all radiometric dates tied to a study area unless there are
reasons for their exclusion (i.e., poor sample prep/yield/lack of historical documentation), and
suggest the inclusion of DICARB dates even in light of their inherent issues (sensu Reuther and
Gerlach, 2005). After randomizing ages constrained to particular units (bound depth ranges), we
repeated the earlier analyses with an additional 70 14C ages (Table S1) including DICARB 1687.
The results were quite similar, with the exception of the 40 cm above the upper hiatus where the
larger dataset hypothesizes two intervals of rapid followed by slow rates of accumulation rather
than relatively stable rate of accumulation (Fig. 11). Tighter stratigraphic control of Holocene
samples will help resolve this issue with additional dating in this part of the section.
5. Discussion
Reconstructing the temporal relationships between sediment (and bone) accumulation,
paleoenvironment, and paleoecological circumstance is critical to understanding the context and
history of paleontological localities such as NTC. The fossil-rich deposits have proven to be a
27
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
treasure trove of vertebrate stories ranging from the preservation of genetic material, trophic
associations, paleoecology, paleogeography (e.g., Mitchell et al., this issue; Minkley et al, this
issue; Higgins et al., this issue; Redman et al., this issue). However, outside of the unpublished
work by John Albanese (1976), little has been done to further our understanding of stratigraphic
relationships of excavated sediments. Moreover, confusion has arisen as various authors have
created their own nomenclature to apply to NTC deposits without the requisite correlations
needed to maintain continuity across studies (e.g., Martin and Gilbert, 1978a; Gilbert and Martin,
1984; Wang and Martin, 1993; Williams, 2009).
5.1 Stratigraphic Nomenclature
A hand-drawn note (Appendix, Fig. S1) found amongst NTC fieldnotes and files curated
at the University of Kansas Biodiversity Institute Division of Vertebrate Paleontology is the only
known correlation of the ‘zone’ and ‘unit’ systems of stratigraphic nomenclature created in
1976/1977: the ‘zone’ system was first devised by SAC (S.A. Chomko - 1976) and modified by
SAC/KAL (S.A. Chomko & K.A. Lippencott - 1977) and Albanese’s ‘unit’ system of
nomenclature (Albanese, 1976 Unpublished Report). Unfortunately the long-term utility of the
latter was complicated by the inverted nature of the original unit designations. In his 1976 report,
Albanese described 11 discrete stratigraphic intervals (numerically descending with stratigraphic
depth). He later (1977) amended his original report by adding hand-written notes assigning each
of his 11 units to one of four primary ‘Units’ ([dataset] Lovelace et al., 2021). In other words,
Unit 1: sub-unit 1 was the oldest stratigraphic unit known at the time of his description, and Unit
4: sub-unit 11 was stratigraphically youngest. The ‘zone’ nomenclature along with Albanese’s
28
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
‘unit’ nomenclature were used extensively during the 1974-1985 excavations to record the
stratigraphic context of fossil specimens. Issues arose when excavations began to encounter
stratigraphic layers underlying Albanese’s Unit 1: sub-unit 1.
It is unclear when a shift in nomenclature occurred. The first example is found in Martin
and Gilbert’s (1978a, fig 5) schematic stratigraphy figured with fauna typical of each unit along
with an abbreviated description of seven stratigraphic units; they labeled their units Stratum 1–7
(youngest to oldest, respectively). While there is utility in this nomenclatural shift (e.g., it is
easier to add new units as they are discovered below the last labeled unit), the new stratum
intervals do not correlate 1:1 with the ‘zone’ or ‘unit’ system previously established, and the
published strata lacked detailed descriptions or correlations with Albanese’s prior descriptions. In
addition, descending numerical stratigraphic nomenclatures can potentially complicate issues as
new strata may be discovered between existing units. Six years later, Gilbert and Martin (1984,
fig. 6.3) published a simplified stratigraphic diagram (which appears to be modified after
Albanese’s A-A’ cross section; Albanese, 1976 Unpublished Report) to demonstrate the relative
position of three ashes (found in 1979) interbedded with cave sediments found well below
Stratum 7 of Martin and Gilbert (1978a). However, none of the stratigraphic intervals were
described or labeled and the schematic is inconsistent with their previous publication making a
direct comparison between the two difficult.
At least two units have commonality between all stratigraphic schemes (based on unit
descriptions/figures), the laterally continuous clay-rich silt with weathered (chalky appearance)
pebble to gravel size limestone clasts at the base (our Upper Red), and the main bone bearing
unit (our Cobbly Grey) with abundant fine to very coarse gravel breakdown in a grey/green silt
29
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
matrix (Albanese, 1976 Unpublished Report; Martin and Gilbert, 1978a). Descriptions of an
associated Panthera specimen (skull = KU 31417; Martin and Gilbert, 1978b) from the just
below the contact with the breakdown-rich ‘main bone layer’, (i.e., Unit 2 -sub-unit 5 and Unit 1
-sub-unit 4), allows correlation with Stratum 4-5 of Martin and Gilbert (1978a) and our Mottled
Silt unit. Albanese’s Unit 1 (sub-unit 1) is his lowest described stratum and is noted as a clayey
light tan silt with abundant angular limestone fragments (fine to very coarse gravel) that occur
scattered throughout, and as lenses (Albanese, 1976 Unpublished Report). This is comparable to
the Martin and Gilbert (1978a) Stratum 7 and our observations of the Pebbly Brown (see section
4.2.1).
Wang and Martin (1993) used Albanese’s unpublished cross sections and sub-unit
designations (not cited), but misidentified the ‘main bone bearing layer’ as Unit 4. They clearly
were discussing sub-unit 5, which is the main bone-bearing unit of Albanese, and is even drawn
and labeled correctly in their reconstructed fence diagram (Wang and Martin, 1993 fig. 5). We
have collated all significant contributions to NTC stratigraphic nomenclature and correlated them
with our proposed lithostratigraphic nomenclature, including our extrapolated ages for each unit
boundary (Fig. 12).
5.2 The ‘snow cone’ hypothesis and the ‘Lower Chamber’
While we cannot fully reconstruct the evolution of deposits within the cave, specific observations
do provide insight and allow for the support or rejection of other hypotheses. The possible presence of a
year-round snow cone during the Pleistocene deposition was used to explain the displacement and
concentration of fossils several meters lateral to the footprint of the cave entrance. The snow cone
hypothesis was also invoked to explain unusual topography of the cave floor, namely that the most
30
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
fossiliferous strata were found not only lateral to (max = <10 m), but also elevated above (≤2 m) the
footprint of the cave entrance (Martin and Gilbert, 1978a; Wang and Martin, 1993).
The current average temperature is reported to be around 5° C (Martin and Gilbert, 1978a; Wang
and Martin, 1993; Meachen et al., 2016). While modern average cave temperature alone does not preclude
the possibility of year round snow during peak glacial periods, it demonstrates that under recent (i.e.,
interglacial) conditions a permanent snow cone does not exist. It should be noted that permanent snow
does occur in several caves in the Cottonwood Creek area south of NTC, and northwest of the cave in the
Pryor Mountains; however, those caves are greater than 800 m higher in elevation and are located in more
forested areas (Elliot, 1963; Hill, 1976; Higham and Palmer, 2018). Further observations of these caves
could be used to compare the resultant sediment accumulation and distribution in the waning presence of
a permanent or semi-perminant snow cone. In addition, no periglacial features have been observed in
NTC (Albanese, 1976 Unpublished Report). The topographic depression found immediately below the
cave entrance was proposed to represent the location of the snow cone, and as a result sediment, bones,
and animals, that fell into the cave were proposed to have effectively been shunted laterally to build up
around the margin of the cone causing a high concentration of vertebrate remains (Wang and Martin,
1993).
If, as it has long been presumed, vertebrate remains are sourced from animals that accidentally
fell into the cave, it must be at least noted that there is the distinct possibility that those animals were in
motion (walking or running across the landscape). Under those circumstances, moving animals will
maintain the velocity at which they lose contact with the ground for the duration of the fall.
Conservatively the vertical descent is 24 m, thus a free fall of this distance would take ~2.2 s (impact
velocity ~22 m/s). Objects entering free fall with a horizontal velocity less than 1 m/s would fall within 2
m of the footprint of the entrance. Megafauna trotting or running down the ridgeline and entering the cave
at higher velocities (i.e., 1.5 to 4 m/s; Higgins et al., this issue) would impact somewhere between 3–9 m
31
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
from the entrance footprint; momentum could potentially carry their carcasses even farther after impact. If
the animal did not rebound off of the opposite side of the cave entrance, it is possible to fall via projectile
motion > 20 m horizontal to the fall line. The higher accumulation of vertebrate remains west of and
lateral to the cave entrance might well be explained with projectile motion rather than sliding down a
snow cone.
We support an alternate explanation for the depression underneath the entrance: that erosion was
the major cause of surface deflation observed during earlier investigations. Erosion as a potential
mechanism was originally proposed and simultaneously dismissed as unlikely because of a presumed lack
of slope capable of allowing material to move laterally (Martin and Gilbert, 1978a) and it was thought
that small scale cut and fill structures suggested erosion wasn’t a prominent surface process (Wang and
Martin, 1993). However, active redistribution of sediment by moving water has been documented
throughout the history of excavation. For instance, curated BLM reports discuss means of controlling
erosion between field seasons with plywood structures ([dataset] Lovelace et al., 2021), and observed
stratigraphic relationships show clear evidence of erosion as a significant form of surficial modification.
Our stratigraphic sections demonstrate that the Rainbow Beds are at their maximum elevation just
below the cave entrance (Fig. 4). Given the discovery of in situ vertebrate remains that yielded 14C dates
from the Mottled Silt (27.8 to 28.7 cal ka BP; OxA-10078) and Pebbly Brown (37.6 to 39.2 cal ka BP;
CAMS 216524) under the entrance to the cave during the buildup and into the peak of the Last Glacial
Maximum, 21,000 cal yr BP (Pinedale Glaciation; Pierce, 2003; Pierce et al., 2018), the presence of a
permanent snow cone is rejected. Late Pleistocene strata equivalent to the Primary Beds most likely
existed below the entrance, dipping to the north and west where it was possible for material to be moved
laterally downslope. We suggest early Holocene erosion prior to deposition of the Upper Beds is the most
parsimonious solution to the missing strata.
32
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Although we propose that Primary Beds equivalent strata under the cave entrance were removed
by erosion, it raises the question: where could eroded sediment/bones have gone? There is support for
reworking of vertebrate material derived from Cobbly Grey sediments in the Upper Beds (see section
4.2.2), and it is possible that the eroded material was transported down the main talus slope mixing with
Holocene vertebrate remains; this may be the case for several bones attributed to Pleistocene-aged
mammals found on the surface of the slope in early reconnaissance (Rushin, 1973). Numerous 14C age
determinations with stratigraphic control will be required to further interrogate this hypothesis in future
excavations. The presence of the Lower Chamber is another possible route of sediment transport and
deposition of reworked material.
The entrance to the Lower Chamber was not encountered prior to the 1985 excavation indicating
that the entrance was covered by breakdown prior to that date. An initial investigation of the Lower
Chamber in 1985 ended with the collection of two Pleistocene aged taxa, including a mammoth femur
(KUVP 85000) and an equid skull (KUVP 84947; Wang and Martin, 1993). This demonstrates
communication with the Upper Chamber at some point with an opening big enough to accept
boulder-sized Pleistocene material. Outside of the LiDAR survey of the physical dimensions and limited
surface collection of vertebrate remains (see section 4.1) in 2014/2016, little is known about the Lower
Chamber and its relationship with the main chamber. It is proposed that the Lower Chamber, at least
periodically, acted as a repository for material derived from deposits within the Main Chamber. The
complexities of sedimentation in the Lower Chamber may be exacerbated by intermittent sediment supply
and will likely exhibit significant temporal mixing of material (including reverse stacking with older on
younger material). Understanding the relationship between the Lower and Main chambers is critical to
piecing together the history of sedimentation within the cave; this should be a focus of future work in the
cave.
33
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
5.3 Constraining depositional history
Considering the concentrated distribution of vertebrate remains within NTC deposits, we
employed the age-depth model to help frame fossil-dense strata (i.e., Cobbly Grey) within a spatial and
temporal context. Although there are three hiatus bound sedimentary packages with evidence of truncated
or missing stratigraphic units, sedimentation is rather consistent between comparable units (i.e., the
Mottled Silt/Upper Pebbly Grey and the Lower Beds) with an ~110–150 yr/cm accumulation rate; the
exceptions are the Cobbly Grey with a much higher rate of sedimentation (~25 yr/cm) and the upper
Primary Beds (>300 yr/cm). In addition, units that are bound above and below hiatal surfaces have
comparable lithology, breakdown distribution, clast size, and intra-unit thicknesses (i.e., Upper Pebbly
Grey and Lower Pebbly Grey), suggesting similar depositional circumstances. The taphonomic signal is
also suggestive of a consistent depositional setting (Redman et al., this issue), though it is unclear why
there is an increase not only in depositional rates, but also the dramatic increase in vertebrate remains in
the Cobbly Grey.
By demonstrating that stratigraphic boundaries can be traced between sections we suggest that
extrapolated boundary ages can also be translated across these surfaces making temporal correlations
possible. Furthermore, the extrapolated boundary ages allows us to place fossiliferous intervals into a
chronostratigraphic framework (Fig. 13). When compared to local and regional glacial history (e.g.,
Salisbury and Blackwelder, 1903; Nelson, 1977; Pierce et al., 2018; Minckley et al., 2021), the vast
majority of preserved sediments were deposited during the buildup of both the penultimate and Last
Glacial maxima. Interglacial deposits were not encountered between the two glacial maxima, though
erosion, bypass, and deposition are clearly active processes in the Holocene.
The base of the Lower Beds in the Reference Section represents the start of sediment
accumulation in NTC which coincides with buildup of the penultimate glacial period (Bull Lake
34
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Glaciation ca. 150–130 ka; Pierce et al., 2018). The upper surface of the Lower Beds marks the base of
the lower hiatal boundary. This surface does not exhibit significant erosion supporting a period of bypass
rather than downcutting, at least in the exposures of the Reference Section. The hiatus is terminated with
the onset of deposition at the base of the Maroon Blocks which has an extrapolated age of 53 cal ka BP.
The Maroon Blocks represent a significant event composed of large red-colored breakdown blocks. The
thickest accumulation is observed at the base of the Reference section and on both the east and west side
of the entrance to the Lower Chamber. A conspicuous portion of the cave roof immediately over the
Reference Section appears to be dark red paleokarst infill that may have been the source of the breakdown
(Fig. 6C). We tentatively hypothesize that this event resulted in a physical barrier that altered the
topography of the floor, or plugged the entrance to the Lower Chamber, reinitiating deposition in the main
excavation area throughout the Pinedale Glaciation.
Sometime near the end of the Pleistocene, erosion likely began to reshape previous deposits
resulting in the mixing of Holocene and Pleistocene aged faunas in the Upper Beds. The record of Early
Holocene through present is more consistent with deposition of eroded Pleistocene material along with
more contemporaneous specimens and sediment. Presently erosion and bypass are dominant processes
within the cave as material is transported to the lowest portions of the excavation (which is currently
bowl-shaped) and overflow into the Lower Chamber. Given the active surficial processes seen in the cave
today, and depositional history throughout the last ca. 150 ka, it seems unlikely that sediment input from
the surface would be halted for >70 ka. At minimum, aeolian contributions would be expected. It is more
likely that warmer and wetter interglacials were (are) relative periods of erosion and bypass rather than
deposition in the area of the main excavation.
A final point to note: the Saddle Area is our only measured section that contains the Grey Silt and
Blocky Brown, which correlate to Albanese’s Unit 3 (sub-units 6–8). Given the increasing presence of the
Blocky Brown to the east as the depth of the erosive unconformity lessens in that direction, it is possible
35
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
that the Blocky Brown west and north of the Saddle Area (i.e., Martin-Gilbert excavations) were thicker
than in the section we observed. This would effectively reduce the amount of time missing between the
Upper Red and underlying Blocky Brown. For this study, we have discounted most of the early DiCARB
data because of the inclusion of multiple specimens (to get enough collagen to analyse at that time) and/or
the original analytical data are not available. However, three of the DiCARB dates fall within our modeled
upper hiatus (10.9, 12.8, 14.7 uncalibrated 14C ka BP). Our final 14C analysis (in 2021) yields one date
from the upper Grey Silt at the older end of this range (i.e., 14.5 uncalibrated 14C ka BP; CAMS 186411).
It is possible that the DiCARB dates reflect reasonable radiocarbon ages, but were not reproduced in our
studies because the strata they derived from were absent (eroded) from the sections we studied. A
systematic study to determine 14C dates from specimens collected during Martin-Gilbert excavations
within a single square with known depth and stratigraphic position spanning sub-units 5 through 11 would
help establish if the erosional surface extends to this portion of the excavation, or if the
Pleistocene-Holocene boundary is preserved as hypothesized by previous workers (Martin and Gilbert,
1978a; Gilbert and Martin, 1984). If the Pleistocene-Holocene boundary is found to exist in NTC, further
study or continued excavation in the western and north of the deposits is warranted considering the
paucity of sites that record this important transitional period.
6. Conclusions
●Using lithostratigraphic terminology we have established an easily modifiable system to
incorporate new units as they are discovered, regardless of their position in the profile relative to
existing units. This new system is correlated with previous stratigraphic nomenclatures to
establish a consistent framework for future excavations and to allow the comparison of current
and future work with previously published results that form the backbone of our understanding of
NTC.
36
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
●We used more than 100 radioisotopic dates to develop an age-depth model that enables us
to assign an extrapolated age to each stratigraphic boundary. These boundaries are
traceable between sections allowing us to place deposits throughout NTC excavations
into a chronostratigraphic framework making broader correlations and comparisons
possible.
●Counter to the original hypothesis that NTC represents continuous deposition over the
last 110 ka, we demonstrate that deposition likely began around 155 ka and that there is
an ~75 ka period of non-deposition in the study area between the penultimate and Last
Glacial maxima.
●There appears to be a relationship between sediment accumulation in the main
excavation area and the buildup through peak glacial periods. Interglacial periods are not
represented in the stratigraphic assemblage and may reflect a period of erosion, bypass,
or nondeposition. Warmer and wetter conditions of the interglacials would likely see
more influx of water, and sediment may have been transported further downslope or into
the Lower Chamber (e.g., bypassing the main excavation area).
●Latest Pleistocene to earliest Holocene strata are not present in the four stratigraphic
sections measured for this study. It is possible that strata of this age were removed during
early Holocene erosion prior to onset of deposition that preserves a mixture of Holocene
and reworked Late Pleistocene vertebrate remains. Additionally, it may be possible that
the erosional surface is restricted to the southern and eastern portions of the excavation
(our study sections), allowing for the potential preservation of the Pleistocene-Holocene
37
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
boundary in historically excavated (and yet to be excavated) areas in the northwestern
areas of NTC deposits.
Author contributions
Conceptualization, DML, CMR, TAM; Methodology, DML, TAM; Investigation, DML, CMR,
BWS, JL; Formal analysis, DML, TAM; Data Curation, DML, CMR, KB; Resources (14C data),
KJM, AC, LF, HH; Resources (cave safety), JL; Visualization and Writing (Original draft),
DML; Writing (Review and editing) all authors; Funding acquisition, JLM, JAM.
Declaration of competing interests
The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
Data availability
Datasets related to this article can be found at http://dx.doi.org/10.17632/f8frrskzpz.1, an
open-source online data repository hosted at Mendeley Data (Lovelace et al., 2021).
Acknowledgements
We would like to thank the many volunteers, students, conservators and researchers who
contributed to the excavation, preparation, curation, and study of NTC over the last four and a
half decades. Our sincere appreciation to our colleagues xx for supportive dialog and guidance.
We thank Jason Ballensky, Edmund Tucker, Ian Chechet, Ken Stahley, and Hans Bodenhamer
for their efforts in surveying NTC. Thanks to Ryan Gardner for assistance in measuring sections
and thoughtful discussion of observations. We would also like to thank Megan Sims (University
38
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
of Kansas’ Vertebrate Paleontology Collec) for her help accessing information in the NTC
database, and Emma Hendersen for helping collate archived data at the Oxford Radiocarbon
Accelerator Unit. This project was made possible by the Wyoming Bureau of Land Management
permit number PA-13-WY-207 to JAM, and we would like to thank B. Breithaupt and G. Hurley
of the BLM for their assistance. This project was funded by Des Moines University (JAM),
National Geographic Society Grant NGS 9479-14 to JAM, NSF EAR/SGP 1425059 to JAM,
NSF EAR/SGP 1945013 to JLM, a Cave Conservancy grant to JAM and CR. Any use of trade,
firm, or product names is for descriptive purposes only and does not imply endorsement by the
U.S. Government.
References
Agenbroad, L. D., & Mead, J. I., 1994, The Hot Springs Mammoth Site: A decade of field and laboratory
research in paleontology, geology, and paleoecology: The Mammoth Site of South Dakota. Hot Springs,
South Dakota, 457.
Akersten, W. A., Shaw, C. A., Jefferson, G. T., & Page, G. C., 1983, Rancho La Brea: status and future.
Paleobiology, 9(3), 211-217.
Albanese, J. P., 1976, Preliminary Report on the Geology of Natural Trap Cave, Big Horn County,
Wyoming. [Unpublished Report].
http://dx.doi.org/10.17632/f8frrskzpz.1#file-50d270a4-a9e6-4bf1-840c-ae4165598e2c
Batchelor, C. L., Margold, M., Krapp, M., Murton, D. K., Dalton, A. S., Gibbard, P. L., Stokes, C. R.,
Murton, J. B., & Manica, A., 2019, The configuration of Northern Hemisphere ice sheets through the
Quaternary. Nature Communications, 10(1), 3713. https://doi.org/10.1038/s41467-019-11601-2
Benson, L., Madole, R., Landis, G., & Gosse, J., 2005, New data for Late Pleistocene Pinedale alpine
glaciation from southwestern Colorado. Quaternary Science Reviews, 24(1–2), 49–65.
https://doi.org/10.1016/j.quascirev.2004.07.018
Blaauw, M., & Christen, J. A., 2011, Flexible paleoclimate age-depth models using an autoregressive
gamma process. Bayesian Analysis, 6(3), 457–474. https://doi.org/10.1214/11-BA618
Clementz, M.T., Schmitt, A.K., & Chamberlain, K.R., this issue. TITLE. Quaternary International.
Chorn, J., Frase, B. A., & Frailey, C.D., 1988, Late Pleistocene Pronghorn, Antilocapra Americana, from
Natural Trap Cave, Wyoming. Transactions of the Nebraska Academy of Sciences and Affiliated
Societies. XVI. 127-139.
Dyke, A. S., 2004, An outline of North American deglaciation with emphasis on central and northern
Canada. In Developments in Quaternary Sciences, 2. Elsevier: 373–424.
https://doi.org/10.1016/S1571-0866(04)80209-4
39
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Elliot, J. K., 1963, Cave Occurences in the Mississippian Madison Limestone of the Pryor Mountains,
Montana. Billings Geological Society Papers, 12 Annual Field Conference, 15 p.
Gilbert, B.M., and Martin, L.D., 1984, Late Pleistocene fossils of Natural Trap Cave, Wyoming, and the
climatic model of extinction. In P. S. Martin and R. G. Klein (eds.), Quaternary Extinctions: A Prehistoric
Revolution. Tucson, University of Arizona Press: 138-147.
Gilbert, B. M., Pearsall D., & Boellstorff J., 1980, Post -Sangamon record of volcanism and climatic
change at Natural Trap Cave, Wyoming. Abstracts and Program, sixth biennial meeting, American
Quaternary Assoc., 26.
Graham, R. W., & Grimm, E. C., 1990, Effects of global climate change on the patterns of terrestrial
biological communities. Trends in Ecology & Evolution, 5(9), 289–292.
https://doi.org/10.1016/0169-5347(90)90083-P
Hamilton, W. D., & Krus, A. M., 2018, The Myths and Realities of Bayesian Chronological Modeling
Revealed. American Antiquity, 83(2), 187–203. https://doi.org/10.1017/aaq.2017.57
Heintzman, P. D., Froese, D., Ives, J. W., Soares, A. E. R., Zazula, G. D., Letts, B., Andrews, T. D.,
Driver, J. C., Hall, E., Hare, P. G., Jass, C. N., MacKay, G., Southon, J. R., Stiller, M., Woywitka, R.,
Suchard, M. A., & Shapiro, B., 2016, Bison phylogeography constrains dispersal and viability of the Ice
Free Corridor in western Canada. Proceedings of the National Academy of Sciences, 113(29), 8057–8063.
https://doi.org/10.1073/pnas.1601077113
Heintzman P. D., Zazula G. D., MacPhee R. D., Scott E., Cahill J. A., McHorse B. K., Kapp J. D., Stiller
M., Wooller M. J. & Orlando L., 2017, A new genus of horse from Pleistocene North America. E-Life, 6,
e29944.
Higham, S. R., & Palmer, A. N., 2018, Ice Caves in the USA. In Ice Caves, Aurel Perşoiu, Stein-Erik
Lauritzen, eds., 705–716. https://doi.org/10.1016/B978-0-12-811739-2.00031-0
Higgins, P. Meachen, J.A., Lovelace, D.M., this issue, Were Pronghorns (Antilocapra) Primary Prey for
North American Cheetahs (Miracinonyx)? Quaternary International.
Hill, Chris, Sutherland, W., and Tierney, L., 1976, Caves of Wyoming: Geological Survey of Wyoming
[Wyoming State Geological Survey] Bulletin 59, 230 p., 37 pls., 4 sheets.
Jicha, B.R., Singer, B.S., Sobol, P., 2016, Re-evaluation of the ages of 40Ar/39Ar sanidine standards and
supereruptions in the western U.S. using a Noblesse multi-collector mass spectrometer, Chem. Geol. 431,
54-66, doi:10.1016/j.chemgeo.2016.03.024.
Koch, PL, Barnosky, TD., 2006, Late Quaternary Extinctions: State of the Debate. Annu. Rev. Ecol. Evol.
Syst, 37:215–50
Kohn, M. J., & McKay, M. P., 2012, Paleoecology of late Pleistocene–Holocene faunas of eastern and
central Wyoming, USA, with implications for LGM climate models. Palaeogeography, Palaeoclimatology,
Palaeoecology, 326–328, 42–53. doi.org/10.1016/j.palaeo.2012.01.037
Kurtén, B., & Anderson, E., 1980, Pleistocene mammals of North America. New York (N.Y.): Columbia
university press. 442 p.
40
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Lageson, D., Maughan, E., & Sando, W. J., 1979, The Mississippian and Pennsylvanian (Carboniferous)
Systems in the United States—Wyoming. USGS Professional Paper No. 1110-U. USGS. 41 p.
Lee, J.-Y., Marti, K., Severinghaus, J.P., Kawamura, K., Yoo, H.-S., Lee, J.B., Kim, J.S., 2006, A
redetermination of the isotopic abundance of atmospheric Ar. Geochim. Cosmochim. Acta 70,
4507–4512. doi:10.1016/j.gca.2006.06.1563
[dataset] Lovelace, D.M.; Redman, C.M.; Sims, M., 2021, Natural Trap Cave: Historical Documents,
photos, and related datasets. Mendeley Data, v1. http://dx.doi.org/10.17632/f8frrskzpz.1
Mahan, S.A., Hanson, P.R., Mead, J., Holen, S., and Wilkins, J., 2016, The Mammoth Site at Hot Springs,
South Dakota: use of OSL dating to calibrate the sinkhole time machine. Geological Society of America
Abstracts with Programs, 48(7). doi:10.1130/abs/2016AM-281798
Mahan S., Wood J., Lovelace D., Laden J., Mcguire J., Meachen J. this issue. Illuminating a dark and
dangerous hole: Interpreting historical and recently acquired geochronology, while making sense of the
divergence of stratigraphy and ages at Natural Trap Cave, Wyoming. Quaternary International.
Martin, L. D., & Gilbert, M. B., 1978, Excavations at Natural Trap Cave. Transactions of the Nebraska
Academy of Sciences and Affiliated Societies, 336, 107–116.
Martin, L. D., & Gilbert, M. B., 1978, An American Lion, Panthera Atrox, from Natural Trap Cave, North
Central Wyoming. Contributions to Geology, 16(2), 95–101.
Martin, L. D., Gilbert, B. M., & Adams, D. B., 1977, A Cheetah-Like Cat in the North American
Pleistocene. Science, 195(4282), 981–982.
Martin, L. D., Gilbert, B. M., & Chomko, S. A., 1979, Dicrostonyx (Rodentia) from the Late Pleistocene
of Northern Wyoming. Journal of Mammalogy, 60(1), 193–195. https://doi.org/10.2307/1379770
May, S.R., Gray, G.G., Summa, L.L., Stewart, N.R., Gehrels, G.E., & Pecha, M.E., 2013, Detrital zircon
geochronology from the Bighorn Basin, Wyoming, USA: Implications for tectonostratigraphic evolution
and paleogeography. Geological Society of America Bulletin, 125(9–10), 1403–1422.
https://doi.org/10.1130/B30824.1
McKay, M.E.P., 2008, Paleoecologies of the mammalian fossil faunas of Natural Trap Cave and Little
Box Elder Cave,Wyoming [Masters Thesis]. University of South Carolina. 143 p.
Meachen, J.A., Brannick, A.L., & Fry, T.J., 2016, Extinct Beringian wolf morphotype found in the
continental U.S. has implications for wolf migration and evolution. Ecology and Evolution, 6(10),
3430–3438. https://doi.org/10.1002/ece3.2141
Meachen, J.A., Redman, C.M., Gilbert, B.M., Reppen, R., Chomko, S., Lippincott, K., Breithaupt, B.H.,
Lovelace, D.M., & Laden, J., this issue, A History of Paleontological Excavations at Natural Trap Cave
(1969 to present). Quaternary International.
41
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Minckley, T.A., Clementz, M., Kornfeld, M., Larson, M.L., & Finley, J.B., 2021, Late Pleistocene
environments of the Bighorn Basin, Wyoming-Montana, USA. Quaternary Research, 99, 128–141.
https://doi.org/10.1017/qua.2020.78
Minckley, T.A., Clementz, M.T., Lovelace, D.M., Chamberlain, K.R., & Schmitt, A.K., this issue,
Late-Quaternary Vegetation and Isoscapes History of Natural Trap Cave. Quaternary International.
Mitchell, K.J., Bover, P., Mudge, C., Sais, A.T., Heiniger, H., Weyrich, L.S., Meachen, J.A., & Cooper,
A., this issue, Patterns of mitochondrial DNA diversity in extinct horses and camels from Natural Trap
Cave. Quaternary International.
Nelson, R.S., 1977, Glacial Deposits of the Lake Angeline Quadrangle Bighorn Mountains, Wyoming.
The Wyoming Geological Association Earth Science Bulletin, 10(1), 7.
Nelson, A.R., Millington, A.C., Andrews, J.T. & Nichols, H., 1979, Radiocarbon-dated upper Pleistocene
glacial sequence, Fraser Valley, Colorado Front Range. Geology, 7, 410–414.
Olson, E. C., Behrensmeyer, A. K., & Hill, A. P. (1980). Taphonomy: its history and role in community
evolution. Fossils in the Making: Vertebrate taphonomy and Paleoecology, 5-19.
Pedersen, M. W., Ruter, A., Schweger, C., Friebe, H., Staff, R. A., Kjeldsen, K. K., Mendoza, M. L. Z.,
Beaudoin, A. B., Zutter, C., Larsen, N. K., Potter, B. A., Nielsen, R., Rainville, R. A., Orlando, L.,
Meltzer, D. J., Kjær, K. H., & Willerslev, E., 2016, Postglacial viability and colonization in North
America’s ice-free corridor. Nature, 537(7618), 45–49. https://doi.org/10.1038/nature19085
Pelton, S. R., Kornfeld, M., Lou Larson, M., & Minckley, T., 2017, Component age estimates for the Hell
Gap Paleoindian site and methods for chronological modeling of stratified open sites. Quaternary
Research, 88(2), 234–247. https://doi.org/10.1017/qua.2017.41
Perri, A.R., Mitchell, K.J., Mouton, A., Álvarez-Carretero, S., Hulme-Beaman, A., Haile, J., Jamieson,
A., Meachen, J., Lin, A.T., Schubert, B.W., Ameen, C., Antipina, E.E., Bover, P., Brace, S., Carmagnini,
A., Carøe, C., Samaniego Castruita, J.A., Chatters, J.C., Dobney, K., … Frantz, L.A.F., 2020, Dire wolves
were the last of an ancient New World canid lineage. Nature, 591, 87–91
Pierce, K. L., 2003, Pleistocene glaciations of the Rocky Mountains. In Developments in Quaternary
Sciences, 1, 63–76. doi.org/10.1016/S1571-0866(03)01004-2
Pierce, K. L., Licciardi, Joseph. M., Good, J. M., & Jaworowski, C., 2018, Pleistocene Glaciation of the
Jackson Hole Area, Wyoming. USGS Professional Paper 1835, 1–55.
R Core Team (2014). R: A language and environment for statistical computing. R Foundation for
Statistical Computing, Vienna, Austria. URL http://www.R-project.org/
Redman, C.M., Moore, J.R., Lovelace, D.M., & Meachen, J.A., this issue, The rank abundance structure
of large-bodied vertebrates from Natural Trap Cave, Wyoming. Quaternary International.
Reuther, J. D., & Gerlach, S. C., 2005, Testing the “Dicarb Problem”: A Case Study from North Alaska.
Radiocarbon, 47(3), 359–366. https://doi.org/10.1017/S003382220003513X
42
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Rushin, C. J., 1973, Interpretive and Paleontologic values of Natural Trap Cave, Bighorn Mountains,
Wyoming [Masters Thesis]. University of Montana. 116 p.
Reimer, P.J., Austin, W.E.N., Bard, E., Bayliss, A., Blackwell, P.G., Bronk Ramsey, C., Butzin, M.,
Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg,
A.G., Hughen, K.A., Kromer, B., Manning, S.W., Muscheler, R., … Talamo, S., 2020, The IntCal20
Northern Hemisphere Radiocarbon Age Calibration Curve (0–55 cal kBP). Radiocarbon, 62(4), 725–757.
doi.org/10.1017/RDC.2020.41
Rogers, R. R., & Kidwell, S. M., 2007, A conceptual framework for the genesis and analysis of vertebrate
skeletal concentrations. In Bonebeds: Genesis, analysis, and paleobiological significance. 1-63.
Salis, A.T., Bray, D., Lee, M.S.Y., Heiniger, H., Barnett, R., Burns, J.A., Doronichev, V., Fedje, D.,
Golovanova, L., Harington, C.R., Hockett, B., Kosintsev, P., Lai, X., Mackie, Q., Vasil’ev, S., Weinstock,
J., Yamaguchi, N., Meachen, J., Cooper, A., and Mitchell, K.J., 2020, Lions and brown bears colonized
North America in multiple synchronous waves of dispersal across the Bering Land Bridge. bioRxiv
(https://doi.org/10.1101/2020.09.03.279117)
Salis, A.T., Gower, G., Schubert, B.W., Soibelzon, L.H., Heiniger, H., Prieto, A., Prevosti, F., Meachen,
J., Cooper, A., and Mitchell, K.J., 2021, Ancient genomes reveal hybridisation between extinct
short-faced bears and the extant spectacled bear (Tremarctos ornatus). bioRxiv
(https://doi.org/10.1101/2021.02.05.429853)
Salisbury, R. D., & Blackwelder, E., 1903, Glaciation in the Bighorn Mountains. The Journal of Geology,
11(2), 216–223.
Sertich, J. J., Stucky, R. K., McDonald, H. G., Newton, C., Fisher, D. C., Scott, E., ... & Davis, E. B.
(2014). High-elevation late Pleistocene (MIS 6–5) vertebrate faunas from the Ziegler Reservoir fossil site,
Snowmass Village, Colorado. Quaternary Research, 82(3), 504-517.
Shapiro, B., Drummond, A. J., Rambaut, A., Wilson, M. C., Matheus, P. E., Sher, A. V., Pybus, O. G.,
Gilbert, M. T. P., Barnes, I., Binladen, J., Willerslev, E., Hansen, A. J., Baryshnikov, G. F., Burns, J. A.,
Davydov, S. P., Driver, J. C., Froese, D., Harington, C. R., Keddie, G., … Cooper, A. (2004). Rise and
Fall of the Beringian Steppe Bison. Science, 306(5701), 1561–1565.
https://doi.org/10.1126/science.1101074
Sonnenfeld, M. D., 1996, Sequence Evolution and Hierarchy within the Lower Mississippian Madison
Limestone of Wyoming. In M. W. Longman and M. D. Sonnenfeld, eds., Paleozoic Systems of the Rocky
Mountain Region. Rocky Mountain Section, SEPM (Society for Sedimentary Geology), 165-192.
Stafford Jr, T. W., Semken Jr, H. A., Graham, R. W., Klippel, W. F., Markova, A., Smirnov, N. G., &
Southon, J., 1999, First accelerator mass spectrometry 14C dates documenting contemporaneity of
nonanalog species in late Pleistocene mammal communities. Geology, 27(10), 903-906.
Stewart, J. R., 2009, The evolutionary consequence of the individualistic response to climate change.
Journal of Evolutionary Biology, 22(12), 2363–2375. https://doi.org/10.1111/j.1420-9101.2009.01859.x
43
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Stock, Greg M., Riihimaki, Catherine A., Anderson Robert S., 2006. Age constraints on cave
development and landscape evolution in the Bighorn Basin of Wyoming, USA. Journal of Cave and Karst
Studies, 68(2), 76–84.
Wang, X., 1988, Systematics and population ecology of Late Pleistocene bighorn sheep (Ovis canadensis)
of Natural Trap Cave, Wyoming. Transactions of the Nebraska Academy of Sciences and Affiliated
Societies, 16, 173–183.
Wang, Xiaoming and Martin, Larry D., 1993, Natural Trap Cave. National Geographic Research &
Exploration, 9(4), 422-435.
Wang, Y., Widga, C., Graham, R. W., McGuire, J. L., Porter, W., Wårlind, D., & Williams, J. W., 2021,
Caught in a bottleneck: Habitat loss for woolly mammoths in central North America and the ice-free
corridor during the last deglaciation. Global Ecology and Biogeography, 30(2), 527–542.
https://doi.org/10.1111/geb.13238
Williams, D., 2009, Small mammal faunal stasis in Natural Trap Cave (Pleistocene–Holocene), Bighorn
Mountains, Wyoming [Dissertation]. University of Kansas. 252 p.
Figure Captions.
Color Figures: 1-2, 4-9, 12-13. Black and White Figures: 3, 10-11.
Figure 1. Geographic position of Cordilleran and Laurentide ice sheets (CIS, LIS, respectively),
Yellowstone ice cap (YIC), Natural Trap Cave (asterix), and glacially influenced giant lakes (blue). Ice
sheet reconstruction approximates ice boundaries as the Beringian corridor began to coalesce prior to ca.
23 ka ago (Heintzman et al., 2016; Batchelor et al., 2019)
Figure 2: Plan (A;D) and profile (B;E) view of cave passages from DistoX (A-B) and LiDAR (D-E)
surveys, with an aerial view (Google Earth) containing a superimposed cave map (C). Note: A-B scale in
feet; modified after Hans Bodenhamer et al., 2016 (Fig. S2). D-E scale in meters ([dataset] Lovelace et
al., 2021). Profile view (B) looking N20W.
44
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Figure 3. Cross section of main and Lower Chamber (A) ca. 5 m northwest of and parallel to A - A’ of
Fig.MAP. Four cross sections of the Lower Chamber (a-d) show the constriction of the chamber
downslope (B). 3D rendering of Lower chamber in map (C) and profile (D) view. Depth in feet.
Figure 4. Stratigraphic profiles of four measured sections and their lateral correlations. Inset map (lower
left) shows column locations (letters in circles correspond to stratigraphic profiles). Entrance to Lower
Chamber (E); North Wall ash (F).
Figure 5. Selected features of the Upper Beds and Primary Beds as seen in the Saddle Area (A-C) and the
Reference Section (D). UR = Upper Red; BB = Blocky Brown; GS = Grey Silt; CG = Cobbly Grey; and
RB = Rainbow Beds. Red dotted line = contact of UR and CG, note angular incision from right to left.
Bright green dotted line = contact between CG and the underlying Rainbow Beds.
Figure 6. Selected features of the Rainbow Beds and Lower Beds as seen in the Reference Section (A)
and West Wall section (B). Dotted green line = Upper Beds/Rainbow Beds contact. Dotted black line =
Rainbow Beds/Lower Beds contact. In the upper ¼ of the Lower Beds (B), a laterally continuous
charcoal-rich layer is present. C = infilled paleokarst in ceiling lateral to cave entrance. D = charcoal-rich
layer (hand sample). MS = Mottled Silt; UPG = Upper Pebbly Grey; PB = Pebbly Brown; MB = Maroon
Blocks. RB = Rainbow Beds; LB = Lower Beds.
Figure 7. Profile of the west wall (top) with labeled interpretive illustration (bottom). The West Wall
section is located vertically below the scale bar (bottom). Note the tip of the ladder buried in sediment
accumulated in the ~2 m deep excavation into the Lower Chamber (ca. 1985; bottom right).
Figure 8. Two grey ash layers (A) were encountered above the entrance to the Lower Chamber (bottom
center). The lower ash in (A) exhibits evidence of transport and sorting (B, blowup box in A). Another
ash is visible on the north wall (C).
Figure 9. Erosive contact between the Upper Beds and the Rainbow Beds in the southeast corner of the
excavation. A = southern East Wall. B = eastern South Wall. Note, in A the contact extends into the
Pebbly Brown.
45
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Figure 10. Distribution of median calibrated ages (ka BP) at 95% probability for 97 14C determinations
(Table S1); one sample (not plotted) extended beyond the range of the calibration curve.
Figure 11. Results of the age-depth model and rates of sedimentation (yr/cm) for the first 50 ka. Analyses
with stratigraphically well constrained samples (n=20; black line; data points are plotted with error) and
all 14C determinations (n=101; orange line) are plotted for comparison. No samples were analyzed from
the Blocky Brown (BB). GS = Grey Silt; UPG = Upper Pebbly Grey; PB = Pebbly Brown.
Figure 12. Correlation of historical nomenclature including unpublished and published accounts. The
Upper Red (highlighted with light red correlation) and the Cobbly Grey (highlighted with light blue
correlation) share commonalities across multiple publications. *Note that Wang and Martin (1993)
misidentify Unit 4 as equivalent to ‘zone’ VII and referred to Unit 4 as the main bone-bearing unit; their
description and fence diagram (their Fig. 5) is based on sub-unit 5 of Albanese (1976 Unpublished
Report). **Inferred position of ashes from Gilbert and Martin (1984).
Figure 13. Chronostratigraphic representation of Natural Trap Cave sediment accumulation compared to
regional glaciations. Note the lack of preserved stratigraphy deposited during the Sangamonian
Interglacial in the area of the main excavation(s). Shaded areas on the left of the timeline represent
interglacials (white), buildup/decline (grey), and peak (black) glacial periods. Red = Upper Beds; light
blue = Cobbly Grey; dark red = Maroon Blocks; dark blue = lower two ashes.
Table 1. Stratigraphic description
Table 2. Modeled 14C determinations
46
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Table 1. Stratigraphic description
Bed Stratum Munsell Description
Upper Beds
Upper Cobbly 10YR 4/4 dark
yellowish brown
Silt with weak calcareous cement. Abundant pebble to cobble size limestone breakdown, rare boulder size
breakdown present. Abrupt lower contact.
Upper Red 10R 5/8 red
Prominent silt-rich bed that can be traced laterally across much of the excavated area. Medium to coarse
gravel sized limestone fragments with a strongly weathered, chalky appearance. Abrupt unconformable erosive
contact with significant incision into underlying units.
Primary
Beds
Blocky Brown
7.5YR 4/4 brown (top)
10YR 3/2 very dark
greyish brown (base)
Slightly clayey silt with blocky peds that increase in size from fine to medium with increased depth. Clay
content increases with depth, and clay skins are present. Fine to medium gravel, diffuse mottles (10R 4/8), and
plant fiber molds with oxidized halos are present towards the base. The lower boundary is diffuse.
Grey Silt 2.5YR 5/3 reddish
brown Slightly gravelly clayey silt with weak calcareous cement. Gradational contact with Cobbly Grey (10–20 cm).
Cobbly Grey 2.5YR 6/1 reddish
grey
Clayey silt with abundant (~20–30% of unit) medium to coarse gravel sized breakdown; boulder size
breakdown is rare. Gravel and bone clasts are randomly oriented and poorly sorted throughout the unit. Some
diffuse mottling (5YR 5/6; yellowish red) is present. The lower boundary is marked by a distinct abrupt change to
a near absence of breakdown.
Rainbow
Beds
Mottled Silt 10YR 6/3 pale brown
Clayey silt with diffuse mottling (5YR 5/8; yellowish red) and rare (<5%) matrix supported medium to coarse
gravel. The lower contact is wavy and abrupt.
Upper Pebbly
Grey 7.5YR 4/2 brown
Clayey silt with abundant lenses and accretion sets composed of moderately sorted fine to very coarse
gravel at the top of the unit, discrete lenses are more widely distributed within the clayey silt matrix. Lower
contact is distinct.
Pebbly Brown 7.5YR 4/4 brown
A distinct 'purple' (10R 4/6; red) clayey silt layer (2–5 cm thick) marks the upper contact. This is followed by a
distinct 'ochre' (2.5Y 6/6 olive yellow) clayey silt layer (2–5 cm). Underlying the 'ochre' layer is a clast rich clayey
silt matrix with abundant fine to coarse gravel distributed in shallow lenses. The contact is irregular when
following boulder sized clasts of the Maroon Blocks, otherwise abrupt.
* Maroon Blocks 10R 4/6 red
Boulder sized breakdown (20–60 cm) is randomly oriented, though it rapidly pinches out to medium to coarse
cobbles as the bed dips northward where exposed on the west wall. Silt matrix with randomly oriented fine to
coarse gravel size breakdown fills between boulders. Abrupt lower contact.
Lower Beds
* Lower Pebbly
Grey 7.5YR 4/2 brown
Clayey silt matrix with broadly distributed discrete thin (2–5 cm) lenses of fine to medium gravel. A laterally
traceable thin (2–4 cm) charcoal rich layer with fine to coarse gravel sized limestone fragments with a strongly
weathered, chalky appearance is present ca. halfway through this interval.
** Lower Ashes 5YR 6/2 light reddish
brown
Clayey silt matrix. Two discrete layers composed of volcaniclastic material are present. The lower ash
exhibits sorting and low angle cross bedding. Otherwise similar to Lower Pebbly Grey.
* = exposure limited to the west wall
** = exposure limited to entrance to Lower Chamber laterally for ~1 m.
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Table 2. Modeled 14C determinations
Specimen
Center
Center ID
Date
Error
to:
from:
%
Stratigraphic Interval
Depth
UW 53755 LLNL CAMS 176896 2,550 60 2375 2764 95.45 Upper Beds 15
UW 51659 LLNL CAMS 179250 5,795 30 6499 6666 95.45 Upper Beds 25
UW 53754 LLNL CAMS 176895 8,150 35 9003 9265 95.45 Upper Beds 35
UW 54757 LLNL CAMS 186411 14450 360 16644 18630 95.4 Grey Silt 44
UW 52702 LLNL CAMS 172884 19,480 100 23164 23775 95.45 Cobbly Grey 60
UW 51656 LLNL CAMS 179248 19,480 90 23185 23770 95.45 Upper Red 60
UW 53734 LLNL CAMS 179241 20,000 100 23815 24238 95.45 Cobbly Grey 70
UW 51516 LLNL CAMS 170433 20,140 60 23896 24412 95.45 Cobbly Grey 88
UW 52730 LLNL CAMS 176886 20,270 80 24120 24625 95.45 Cobbly Grey 88
UW 52713 LLNL CAMS 172888 20,850 120 24791 25581 95.45 Cobbly Grey 92
UW 52732 LLNL CAMS 176887 21,060 90 25184 25662 95.45 Cobbly Grey 130
UW 52733 LLNL CAMS 176888 21,510 90 25703 25968 95.45 Cobbly Grey 130
UW54558 LLNL CAMS 186413 23720 140 27641 28286 95.4 Mottled Silt 140
UW54521 LLNL CAMS 186412 27710 220 31178 32112 95.4 Pebbly Brown 168
UW 52717 LLNL CAMS 176876 29,410 1,710 30862 39246 95.45 Pebbly Brown 175
UW 52721 LLNL CAMS 176877 30,010 260 34065 35129 95.45 Pebbly Brown 175
UW 52701 LLNL CAMS 172883 31,030 400 34626 36192 95.45 Pebbly Brown 181
UW 52724 LLNL CAMS 176881 33,500 1,250 35849 41278 95.45 Pebbly Brown 189
UW 53731 KCCAMS CAMS 216524 33,560 230 37597 39220 95.45 Pebbly Brown 200
UW 52722 LLNL CAMS 176880 >29200 N/A N/A N/A N/A Pebbly Brown 200
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Supplementary Information:
Appendix
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Figure 1. Hand-written correlations of early ‘zone’ and modified ‘zone’ nomenclature with that
of Albanese’s (1976) nomenclature, Albanese’s 1977 written modifications, and a final partial
system (RR) was not included in this study. SAC = Stephen A. Chomko; KAL = Kerry A.
Lippencott; JA = John Albanese; RR = Randi Reppen.
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
Collection Year
UW_Specimen #
Taxon
ID element
2014
50603
Aves
coracoid
2014
50604
Aves
coracoid
2014
50605
Aves
coracoid
2014
50606
Aves
humerus
2014
50607
Aves
humerus
2014
50608
Aves
humerus
2014
50609
Aves
humerus
2014
50610
Aves
ulna
2014
50611
Aves
ulna
2014
50612
Aves
ulna
2014
50613
Aves
femur
2014
50614
Aves
femur
2014
50615
Aves
femur
2014
50616
Aves
femur
2014
50617
Aves
femur
2014
50618
Aves
femur
2014
50619
Aves
tibiotarsus
2014
50620
Aves
tibiotarsus
2014
50628
cf. Bison sp.
atlas
2014
50629
Bison sp.
lunar
2014
50630
Bison sp.
proximal phalanx
2014
50631
Bison sp.
middle phalanx
2014
50632
Bison sp.
distal phalanx
2014
50633
Bison sp.
distal phalanx
2014
50694
?Bison sp.
cervical vertebra
2014
50695
?Bison sp.
sacral vertebra
2014
50696
?Bison sp.
thoracic vertebra
2014
50635
Ovis canadensis
metatarsal
2014
50636
Ovis canadensis
atlas
2014
50637
Ovis canadensis
ulna
2014
50638
Ovis canadensis
astragalus
2014
50639
Ovis canadensis
lunar
2014
50640
?Bovidae
dentary
2014
50641
?Bovidae
radius
2014
50644
cf. Canis lupus
proximal phalanx
2014
50697
cf. Canis lupus
cervical vertebra
2014
50645
Vulpes sp.
tibia
2014
50648
Mustela sp.
dentary
2014
50650
Lepus
humerus
2014
50652
Sylvilagus
humerus
2014
50653
Sylvilagus
radius
2014
50660
Lagomorpha
innominate
2014
50676
Equus sp.
femur
2014
50698
cf. Equus sp.
thoracic vertebra
2014
50681
Neotoma sp.
cranium
2014
50699
Mammalia
cervical vertebra
2014
50700
Mammalia
caudal vertebra
2014
50701
Mammalia
rib
2014
50702
Mammalia
rib
2014
50703
Mammalia
rib
2016
53920
Bison sp.
tooth
2016
53949
Canis sp.
metatarsal IV
Table S1. A List of specimens collected in the Lower Chamber in 2014 and 2016.
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
2.0 14C analytical methods
OxA Determinations 2001-2005 Methods: AMS radiocarbon preparation procedures
generally followed Brock et al. (2010). Bone or tooth fragments were sampled by using a
tungsten carbide drill to generate coarse powder, which then underwent an acid-base-acid
pretreatment of 0.5 M hydrochloric acid, followed by 0.1 M sodium hydroxide, and then another
pretreatment of 0.5 M hydrochloric acid - with thorough rinsing using ultrapure water between
each reagent. Collagen was then gelatinized at 75 C for 20 hr. Residual solids were filtered out
prior to being ultrafiltered to remove the >30-kDa fraction. This fraction was freeze-dried and
graphitized for AMS analysis.
LLNL Determinations 2015-2018 Methods: AMS radiocarbon preparation procedures
generally followed Brown et al. (1988) and Bronk Ramsey et al. (2004). Decalcification with 0.25
N HCl was used to obtain collage that was then gelatinized at 58 °C for 16 h. Residual solids
were filtered out prior to being ultrafiltered to remove the >30-kDa fraction. This fraction was
freeze-dried and graphitized for AMS analysis. Dates were calibrated using OxCal v.4.4.3
(Bronk Ramsey, 2021), using the IntCal20 calibration curve (Reimer et al., 2020).
KCCAMS Determinations 2015-2018 Methods: AMS radiocarbon sample preparation
largely follows Brown et al. (1988) and Bronk Ramsey et al. (2004). Samples were dissolved in
0.5 N HCl over 2 weeks (acid changed 3 times); bone chunks remained after dissolution. Filter
pre-cleaning follows (Bronk Ramsey, 2004). Residual acid was removed and collagen was
rinsed (Mili-Q water) prior to adding 10 mL of NaOH (removing humic acid). The NaOH was
removed and collagen was rinsed (Mili-Q water) prior to adding 0.1 N HCl. Samples were
gelatinized at 58 °C for 16 h. Unwound collagen was pre-filtered to remove residual solids, and
then ultrafiltered to obtain >30-kDa fraction. The retantate was then freeze-dried and graphitized
for AMS analysis. Dates were calibrated using OxCal v.4.4.3 (Bronk Ramsey, 2009), using the
IntCal20 calibration curve (Reimer et al., 2020).
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
References:
Brock F, Higham T, Ditchfield P, Bronk Ramsey C., 2010, Current Pretreatment Methods for
AMS Radiocarbon Dating at the Oxford Radiocarbon Accelerator Unit (ORAU). Radiocarbon,
52(1), 103-112. doi:10.1017/S0033822200045069
Bronk Ramsey, C., Higham, T., Bowles, A., & Hedges, R. (2004). Improvements to the
Pretreatment of Bone at Oxford. Radiocarbon, 46(1), 155–163.
https://doi.org/10.1017/S0033822200039473
Bronk Ramsey, C., 2009, Bayesian analysis of radiocarbon dates. Radiocarbon,51(1), 337-360.
Brown, T. A., Nelson, D. E., Vogel, J. S., & Southon, J. R., 1988, Improved Collagen Extraction
by Modified Longin Method. Radiocarbon, 30(2), 171–177.
https://doi.org/10.1017/S0033822200044118
Reimer, P. J., Austin, W. E. N., Bard, E., Bayliss, A., Blackwell, P. G., Bronk Ramsey, C., Butzin,
M., Cheng, H., Edwards, R. L., Friedrich, M., Grootes, P. M., Guilderson, T. P., Hajdas, I.,
Heaton, T. J., Hogg, A. G., Hughen, K. A., Kromer, B., Manning, S. W., Muscheler, R., …
Talamo, S. (2020). The IntCal20 Northern Hemisphere Radiocarbon Age Calibration Curve
(0–55 cal kBP). Radiocarbon, 62(4), 725–757. https://doi.org/10.1017/RDC.2020.41
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 1, 2021. ; https://doi.org/10.1101/2021.07.30.454390doi: bioRxiv preprint
