Epeirogenic Controls on Canadian River Incision and Landscape
Evolution, Great Plains of Northeastern New Mexico
Paul A. Wisniewski1 and Frank J. Pazzaglia2
Utah Department of Natural Resources, Division of Oil, Gas, and Mining, Salt Lake City, Utah 84114, U.S.A.
ABSTRACT
The Canadian River has carved a deep bedrock canyon into the Great Plains of northeastern New Mexico in response
to a complex interaction of post-Laramide, epeirogenic rock-uplift processes along the Jemez lineament and downstream baselevel fall caused by late Tertiary and Quaternary evaporite dissolution. Broad, slow-bulging along the
Jemez lineament, a lithospheric-scale structure with prolific late Cenozoic volcanism along its trend, helps drive
incision in the canyon. Canadian River canyon fluvial terraces, mapped and correlated according to soils, elevation,
and rock-type criteria, reveal longitudinal profiles that appear broadly warped across the Jemez lineament and converge
to the modern channel profile both upstream and downstream of the canyon. Similarly, the modern channel long
profile is distinctly and broadly convex through its Great Plains reach, a unique feature among major rivers draining
the eastern flank of the southern Rocky Mountain front. Radiocarbon-dated charcoal from terrace alluvium and an
40
Ar/39Ar-dated basalt flow within the canyon constrain long-term average rates of bedrock incision to ∼0.06–0.07
mm/yr. River incision along the Jemez lineament has produced youthful canyons and escarpments and eroded Tertiary
rocks relative to other regions along the southern Rocky Mountain front. Whereas rock-uplift processes prevail in
the canyon reach, baselevel changes driven by a well-documented process of salt-dissolution subsidence have controlled the evolution of reaches downstream of the canyon. Removal of bedrock along the northern flank of the Jemez
lineament bulge fosters opposing dips in the valley flanks, resulting in the Canadian escarpment (northern flank) that
stands up to 200 m higher than the Caprock (southern flank). These findings suggest that post-Laramide, thermally
driven epeirogeny is in part responsible for the high-standing topography evident on the western Great Plains of
northeastern New Mexico.
Introduction
Throughout the American West, deeply incised river
canyons are a dramatic geomorphic expression of the
complex contributions of Cenozoic eustatic fall, increasingly cooler climates, and epeirogeny of the
post-Laramide Rocky Mountains (reviewed in Pazzaglia and Kelley 1998). Late Cenozoic tectonism or
epeirogeny (Epis and Chapin 1975; Huber 1981;
Steidtmann et al. 1989; Unruh 1991; Kerr 1997) has
been traditionally favored as the primary cause of
the incision of major river canyons including the
Grand Canyon of the Colorado River (Hunt 1956,
1969; Lucchitta 1984), the Black Canyon of the Gunnison River (Hansen 1965, 1967; Epis and Callender
1981), and the Royal Gorge of the Arkansas River
Manuscript received February 13, 2001; accepted October 23,
2001.
1
E-mail: [email protected].
2
Author for correspondence: Department of Earth and Environmental Sciences, Lehigh University, Bethlehem, Pennsylvania 18015, U.S.A.; e-mail: [email protected].
(Powers 1935). But the role of epeirogeny has been
questioned (Gregory and Chase 1992) or at least diminished (Chase et al. 1998; Small and Anderson
1998; Wolfe et al. 1998) due to the absence of a welldefined driving mechanism for the uplift and a growing data set of paleobotanical estimates for paleoelevation that place the entire Rocky Mountain
region high standing since the close of the Laramide
(Chase et al. 1998; Wolfe et al. 1998).
This study describes the geomorphology, rate of
river incision, and long-term landscape evolution
of the Canadian River and Canadian escarpment
of northeastern New Mexico (figs. 1, 2) in the context of post-Laramide epeirogenesis. The Canadian River canyon, its related escarpment, and a
suite of landforms and numerically dated deposits
straddle a Precambrian lithospheric suture (Karlstrom and Humphreys 1998) coincident with a
thermal bulge called the Jemez lineament. The
coincidence of the lineament, canyon, and escarp-
[The Journal of Geology, 2002, volume 110, p. 437–456] 䉷 2002 by The University of Chicago. All rights reserved. 0022-1376/2002/11004-0005$15.00
437
438
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
Figure 1. Color-shaded topography of the study area illustrating the coincidence of the Canadian escarpment and
Canadian River canyon (rectangle) with the Jemez lineament (gray-shaded overprint) and Sierra Grande arch, a
basement high illustrated by the black contours drawn on top of Precambrian bedrock (500-m contours with respect
to sea level). The inset map shows the location of the color-shaded map (rectangle) with respect to the Jemez lineament
(shaded strip) and its line of major volcanic centers: RH p Red Hills, MT p Mount Taylor, J p Jemez, O p Ocate,
and CR p Clayton-Raton. Major roads are shown as thin, white, curvilinear lines. The black rectangle is the area
mapped in detail and subsequently illustrated or referred to in figure 3; figure 4b, 4c; and figures 5–7 and 9. Towns
and other geographic locations are annotated as follows: SF p Santa Fe, LV p Las Vegas, WM p Wagon Mound,
SP p Springer, R p Roy, S p Sabinoso, T p Tucumcari, and MC p Maxson Crater, the source of the basalts that
flowed down the Mora River.
ment provides an opportunity to directly document a river’s incision history in the context of
this deep earth structure and potential mechanism
capable of generating post-Laramide rock uplift.
In doing so, this study provides a better idea
of how epeirogenic factors are partitioned in
the long-term landscape evolution of the postLaramide western Great Plains of northeastern
New Mexico. The findings of this study also force
a reexamination of the precise effects of late Cenozoic climate cooling on the general pattern of
Rocky Mountain exhumation by a fluvial incision
Journal of Geology E P E I R O G E N I C C O N T R O L S O N C A N A D I A N R I V E R I N C I S I O N
Figure 2. Cross sections A-A, B-B, and C-C. Vertical
exaggeration is #100. Profiles were extracted from the
same 3⬙ digital elevation model used to generate figure 1.
process and identify a setting in the Rocky Mountain region where high-standing topography was
likely produced by epeirogenic processes since the
close of the Laramide.
Setting
The Canadian River has carved a ∼100-km-long,
300–400-m-deep bedrock canyon into the gently
warped Tertiary and Mesozoic strata of the Las Vegas Plateau section of the Great Plains of northeastern New Mexico (figs. 1, 2; Dobrovolny et al.
1947; Baldwin and Muehlberger 1959; Dane and
Bachman 1965; Spiegle 1972; Dolliver 1984; Wisniewski 1999). Elevated 1700–2100 m above sea
level, the laterally extensive pediments, topographically inverted basalt-capped mesas, and stripped
structural surfaces of the Las Vegas Plateau slope
gradually to the southeast away from the eastern
flank of the Sangre de Cristo Mountains (O’Neill
and Mehnert 1988), which represent both the
southern Rocky Mountain front in New Mexico as
well as the eastern flank of the Rio Grande rift. The
Las Vegas Plateau terminates to the south in a
250–300-m-high, embayed line of cliffs known as
the Canadian escarpment (figs. 1, 3a). The canyon
439
is deepest (∼400 m) and widest (∼1.5 km) where it
breaches the escarpment north of Conchas Lake
near Sabinoso, New Mexico.
Numerous Quaternary terraces and an early
Pleistocene basalt flow are preserved in the lower
portion of the Canadian River canyon and the canyon of a major tributary, the Mora River (figs. 1,
3b; Roberson 1972; Dolliver 1984). The basalts,
whose origin is Maxson Crater in the Ocate volcanic field (O’Neill and Mehnert 1988), collectively
define an outstanding time line that preserves the
configuration of the ancestral Mora and Canadian
River channels and allows the rate of vertical incision since the early Pleistocene to be calculated.
The terraces and Maxson Crater basalts are inset
into a suite of older basalts locally interbedded with
the middle Miocene–early Pliocene Ogallala Group
(O’Neill and Mehnert 1988; Stroud 1997). Ogallala
Group deposits represent a suite of complexly inset
alluvial paleovalleys separated by wide, eolian interfluves. On the western Great Plains, the Quaternary has been marked by river incision, terrace
formation, and valley widening that isolated the
Ogallala Group on interfluves between Ogallala paleovalleys (Gustavson and Winkler 1988). The
bracketing ages of the Ogallala Group are a critical
piece of evidence in establishing the onset of fluvial
incision. Tephrochronology and paleontologic correlations of the Ogallala Group in western Texas
provide a regional age range of ∼12 to ∼4.5 Ma for
these sediments (Winkler 1987; Holliday 1990;
Gustavson et al. 1991). Inset relationships between
the Ogallala Group and basalts of the Ocate and
Raton-Clayton volcanic fields further constrain the
age of the Ogallala Group within the study area.
Volcanic units cap progressively higher and older
surfaces for the topographically inverted Las Vegas
Plateau. Basalts of the Ocate volcanic field (5.9 Ⳳ
0.40 Ma 40K/40Ar; O’Neill and Mehnert 1988) preserved on Las Mesas del Conjelon east of Wagon
Mound, New Mexico, are inset above the base of
the Ogallala Group where it outcrops on the north
rim of the Mora River canyon. This indicates that
the Ogallala Group is younger than 5.9 Ma. Volcanic rocks west of Wagon Mound (3.07 Ⳳ 0.34 and
3.3 Ⳳ 0.30 Ma 40K/40Ar; O’Neill and Mehnert 1988)
are inset below the base of the Ogallala Group. This
indicates an age greater than ∼3.3 Ma. On the basis
of these inset relationships, we estimate the approximate basal age of the Ogallala Group within
the field area to be ∼ 4.0 Ⳳ 0.6 Ma, approximately
midway between the local constraining ages.
The degree and age of fluvial dissection of the
western Great Plains and southern Rocky Mountain
440
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
Figure 3. a, Photograph of Canadian escarpment looking north near the mouth of the Canadian River. b, Photograph
of river terrace Qt8, which overlies the Chinle Group (Triassic) bedrock. Note the thin alluvium of the terrace deposit
and the low-relief strath.
front varies from north to south along the front. In
Wyoming, the middle Miocene–Pliocene Ogallala
Group unconformably overlies the Oligocene White
River Group. Together, these deposits extend as a
finger (the Gangplank) onto the Laramie Range west
of Cheyenne. Farther south in New Mexico, the
White River Group has been completely stripped, so
that the Ogallala Group rests unconformably on Cretaceous rocks. “Gangplanks” similar to but broader
than the Gangplank in Wyoming are preserved on
the Great Plains between the four major river valleys
draining the eastern flank of the southern Rocky
Mountain front—the Canadian, Arkansas, South
Platte, and Pecos—such that the degree of large-scale
denudation and dissection is directly correlated with
proximity to a large drainage. The greatest degree of
Great Plains stripping, however, has occurred in
northeastern New Mexico. The missing Oligocene
section, the apatite fission-track thermochronologic
data (Kelley and Chapin 1995), and the preservation
of adjacent high-standing but shallow-level intrusions such as the late Oligocene Spanish Peaks and
the Ortiz Mountains collectively attest to the great
depth of late Cenozoic denudation accomplished
presumably by the ancestral Canadian River and related drainages. The relative proximity to Gulf Coast
baselevel and position along the eastern flank of Alvarado Ridge (Eaton 1987) and the Rio Grande rift
Journal of Geology E P E I R O G E N I C C O N T R O L S O N C A N A D I A N R I V E R I N C I S I O N
(Roy et al. 1999) likely contributed to this advanced
stage of unroofing.
Downstream of the Canadian escarpment, the
Canadian River flows in a wide valley underlain by
upper Paleozoic and lower Mesozoic sedimentary
rocks, some of which, like the Permian Yeso Formation, are rich in gypsum. Dissolution of the gypsum has resulted in extensive collapse, karstification, and local baselevel fall of major river valleys
(Gustavson et al. 1982, 1991; Dolliver 1984; Gustavson 1986). Dissolution beneath the Canadian
River has played a key role in maintaining it as a
master drainage of the Great Plains.
The Canadian escarpment roughly parallels structure contours of a major northeast-trending structural anticline, the Sierra Grande arch (fig. 1). Deeply
entrenched reaches of the Canadian River system
coincide with the location of this basement high,
which represents one of a series of late Paleozoic,
intracratonic block uplifts formed during ancestral
Rocky Mountain deformation that were reactivated
during Laramide deformation (Woodward 1987). In
addition, the Sierra Grande arch is coincident with
a northeast-trending line of late Cenozoic silicic volcanism, called the Jemez lineament (fig. 1). A nearly
linear arrangement of volcanic fields stretching from
east-central Arizona to northeastern New Mexico
(Baldwin and Muehlberger 1959; Calvin 1987; Karlstrom and Daniel 1993) defines the Jemez lineament.
Long recognized as a lithosphere-penetrating (Ander
1981) Proterozoic (∼1.6–1.7 Ga) crustal suture or
weakness between 1.7-Ga Yavapai and 1.4-Ga Mazatzal crust (Cordell and Keller 1984; Karlstrom and
Daniel 1993; Karlstrom and Humphreys 1998), the
Jemez lineament is characterized by anomalously
high heat flow and is the zone of greatest volcanic
activity in the southwestern United States (Smith
and Luedke 1984; Spence and Gross 1990).
Methods
Data for this study originate from field studies that
included mapping, hand-sample rock-type identification, and soil stratigraphy and morphology. We
compliment these data with map-derived topographic metrics such as valley long profiles and the
stream gradient (SL) index of Hack (1973). See Wisniewski (1999) for detailed descriptions of methods
and sampling protocol.
Mapping was conducted on U.S. Geological Survey 7.5⬘, 20-ft contour topographic maps photo enlarged to 1 : 12,000 scale (Wisniewski 1999). We map
terraces as allostratigraphic deposits distinguished
by their relative height above the channel, sedimentology, and relative weathering characteristics. Cor-
441
relation of the deposits is accomplished by map-scale
lateral continuity, gravel rock type, and soil morphology criteria. We based gravel rock-type criteria
on the relative amounts of sedimentary, granitic and
metamorphic, and extrusive igneous (basaltic) rock
types by using a population of at least 100 randomly
selected clasts at each sampling locality. Soils data
are derived from 16 hand-dug pits covering a wide
stratigraphic range of terrace treads. Each soil pit was
excavated to an average depth of 1 m. Soils stratigraphy and morphology were described according to
the methods summarized by the Soil Survey Staff
(1975) and Birkeland (1999). We measured the calcium carbonate content of the ! 2-mm portion of
each soil horizon in the laboratory by using the Chittick device following the methods described by
Singer and Janitsky (1986). Relative degree of soil
development is based on the degree of calcic horizon
development (Birkeland 1999) and the profile development index (PDI) (Harden 1982). Wherever possible, charcoal samples were collected from the terrace deposits. Radiocarbon age determinations on
the charcoal were provided by the Beta Analytic Laboratory (Miami). The 40Ar/39Ar ages on the Maxson
Crater basalts were provided by B. Olmsted and W.
McIntosh (unpub. data).
Elevations of mapped terraces are compiled along
with the valley long profile of the Canadian River
to determine the degree of fluvial incision. All terrace and channel elevation data are orthogonally
projected to a vertical plane constructed down the
center of the valley. This convention removes problems with channel gradient associated with a variable sinuosity and allows for direct comparison of
the terrace elevation to the channel elevation in
places where the terrace may span the width of an
entire large meander. Long profile data is readily
translated into the SL index (Hack 1973). We use
adjacent topographic contours as the incremental
stream reach in the index. Geology used for the SL
index and all long profile plots were taken from a
1 : 96,000-scale bedrock map (Wanek 1962).
We quantified incised meanders and goosenecks
by using a comparison between sinuosity and upstream drainage area. We calculated channel sinuosity by subtracting the length along the valley
bottom (Lv) from the length along the channel (Lc)
divided by the valley bottom length ([Lc⫺Lv]/Lv).
Segments were chosen with 1 : 100,000 topographic maps and defined visually on the basis of
broad differences in channel sinuosity.
Results
The Channel and Longitudinal Profiles. The Canadian River channel exposes reaches of both bed-
442
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
rock and alluvium through the canyon reach. Point
bar exposures typically reveal !3 m of alluvium on
the modern channel bed strath. The channel assumes a more alluvial character both upstream and
downstream of the canyon. Longitudinal (long) profiles for most alluvial and mixed bedrock-alluvial
rivers are concave up because discharge increases
as grain size decreases downstream. Such is the
case for the four major streams draining the eastern
flank of the southern Rocky Mountain front with
the exception of the Canadian River (fig. 4). On the
Great Plains, the long profiles of the Pecos, South
Platte, and Arkansas Rivers flatten but remain concave up. The long profile of the Canadian River is
distinct because it steepens significantly and even
becomes convex for a ∼200-km reach across the
northeastern New Mexico Great Plains (fig. 4a).
This convexity is coincident with the Canadian
River canyon and straddles the projected trace of
the Jemez lineament (fig. 1). Major tributaries to
the Canadian River such as the Mora River exhibit
equally steep concave upward profiles with many
knickpoints, which are typically coincident with
harder rock types like beds of sandstone (fig. 4b,
4c). Both the Canadian River and Mora River convexities terminate at the upstream limit of their
respective canyons. Reconstructed long profiles of
the Maxson Crater basalts (fig. 4b) serve as a key
terrace correlation guide through the canyon reach.
The gradient of the basalt top matches general
trends observed in the terrace correlations as well
as the gradient of the modern channel.
The SL indices (Hack 1973) calculated along the
long profile through the canyon reach are expectedly higher than reaches outside of the canyon and
display an overall downstream increase (fig. 4c). For
the Canadian River, there is a poor correspondence
of SL index to major changes in rock type at the
map-scale resolution.
Despite its deep vertical incision, the Canadian
River continues to be a meandering stream through
the canyon reach. Channel sinuosity in the canyon
nearly doubles with respect to the average sinuosity
both up and downstream of the canyon reach.
Goosenecks are locally present. There is a strong
positive correlation between reach sinuosity and
the area of contributing tributary basins (fig. 5).
This morphometric relationship, observed for unincised alluvial rivers (Schumm 1967), is consistent
with a dynamic feedback between the ability of the
main channel to erode both vertically and horizontally and the delivery of alluvium by the tributaries.
Maxson Crater Basalts. Composed of at least five
separate flows that were channelized down the already entrenched ancestral Mora River canyon
and lower Canadian River canyon, the Maxson
Crater basalts provide a widespread, datable timestratigraphic horizon. Although the basalt contact
with the underlying bedrock is largely concealed
by thick and laterally extensive talus cones, we
estimate the thickness of the basalts to be between ∼30 and 60 m depending on paleovalley geometry. No evidence of fluvial gravel was observed
beneath the basalts at the Canadian River–Mora
River confluence, hereafter referred to as “confluence reach” (cf. O’Neill and Mehnert 1988) or elsewhere within the study area. Instead, the basalts
occupy toeslope positions in the paleovalley bottom, and assuming that the flows were fluid, the
preserved base of the basalt is roughly coincident
with the elevation of the ancestral valley bottom.
Several new samples of individual basalt flows
from the confluence reach were collected and
dated for this study. All of the samples yielded
40
Ar/39Ar ages within the error of the method, so
the flow units were treated as a single isochronous
unit for the purposes of this study. The weighted
mean age of the Maxson Crater basalts is 1.46 Ⳳ
0.02 Ma (B. Olmsted and W. McIntosh, unpub.
data).
Terraces and Terrace Correlation. We identify and
map 10 unpaired fluvial terraces (Qt1–Qt10fp) distributed throughout the canyon reach and inset
into the Las Vegas Plateau below the Ogallala
Group and Raton-Clayton and Ocate basalts (fig. 6;
table 1). Terraces are best preserved where the valley bottom widens with respect to canyon depth.
For this reason, few terraces are preserved in the
narrow reaches immediately upstream of the confluence reach, and many terraces are preserved in
the wide valleys surrounding Sabinoso, New Mexico (Wisniewski 1999). Terraces, which record the
long-term incision history of the Canadian River,
are progressively inset into the canyon walls such
that the topographically highest terrace is stratigraphically the oldest one.
Qt1, the highest and oldest terrace within the
canyon study area, is the only terrace with a base
stratigraphically above the Maxson Crater basalts
(fig. 6b). Qt1 is a prominent strath terrace. The
strath has several meters of relief, and the terrace
deposit is characterized by up to 8 m of stratified,
coarse-grained alluvium composed primarily of
well-rounded granitic and metamorphic cobbles
(fig. 7). Where present, soils developed on the gravelly alluvium of Qt1 are well developed and exhibit
stage 3 carbonate morphology. The PDI values,
however, tend to be low (∼0.2) because most soils
have been stripped and modified by surface erosion.
Inset below Qt1 are a series of strath terraces
Figure 4. a, Annotated long profiles of the Canadian, South Platte, Pecos, and Arkansas Rivers. Arrows point to
where these streams exit the mountain front. b, Long profile of the Mora River and the top of the basalt flow preserved
in the Mora valley from Maxson Crater to its confluence with the Canadian River. Note the general coincidence in
the basalt and channel gradient until about 14 km upstream of the confluence when the profiles begin to diverge
significantly. c, The stream length–gradient index (SL index) of Hack applied to the Canadian River. The plot contains
a long profile (thin black line), a curve of SL indices (gray line), and a first-order linear regression line (thick black
line) fit through the SL curve. SL indices used DL p 1 km along the channel thalweg. Long profiles were drawn from
1 : 24,000 topographic maps. Symbols are as follows: Trsr, sandstone of the Triassic Santa Rosa Formation; Trcl, the
lower shale formation of the Triassic Chinle Group; Trcm, the middle sandstone formation of the Chinle Group;
Trcu, the upper shale formation of the Chinle Group; Jbe, sandstone of the Jurassic Bells Ranch and Entrada formations.
444
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
Figure 5.
a, Schematic diagram showing Canadian
River channel segments, location of goosenecks (segments 5 and 7), together with (b) a plot of segment sinuosity as a function of contributing upstream drainage
area. Segments (reaches between the short gray horizontal lines) were chosen with 1 : 100,000 topographic maps
and defined visually based on broad differences in channel sinuosity. For example, straighter reaches were distinguished from more sinuous reaches. Map scale is approximate. Note that the X-axis of the plot is a log scale.
Unlike most of the other Canadian River tributaries that
are ephemeral streams, the Mora River is a large perennial stream with a relatively large discharge that does
not fall on the general trend of the other data. Sinuosity
was calculated by subtracting the length along the channel (Lc) from the length along the valley bottom (Lv)
divided by the valley bottom length ([Lc⫺Lv]/Lv).
(Qt2–Qt6) whose alluvium contains a greater proportion of sedimentary and extrusive igneous clasts
relative to Qt1 alluvium (figs. 6b, 7). Straths for
Qt2–Qt6 also contrast with Qt1 in that they are generally flat with !2 m of local relief. Thin and poorly
exposed, the sedimentology of alluvial deposits of
Qt2–Qt6 is only generally known to be stratified
sandy gravel that lacks fine-grained facies. Qt2 and
Qt3 alluvium is thin and discontinuous, which precludes us from establishing accurate clast counts
that might be used to determine their composition.
Intact soils are not present for Qt2 and Qt3 alluvium.
Qt4 alluvium is also largely stripped; but locally, preserved alluvium does contain boulder-sized
clasts. Intact soils developed on the tread of the Qt4
alluvium exhibit stage 2–3 carbonate morphology
and have a mean PDI of 8.2 (table 1). Qt5 is characterized by moderately thick alluvium (!5–10 m)
with well-developed soils exhibiting stage 4–5 carbonate morphology, maximum carbonate content
exceeding 30%, and a mean PDI of 15 (table 1). Qt6
is poorly preserved in the canyon reach and has only
limited exposure near Sabinoso. Its alluvium has a
greater percentage of granitic and metamorphic cobbles than Qt2–Qt5 but still less than the percentage
found in Qt1 (fig. 7). Soils on Qt6 alluvium exhibit
stage 3 carbonate morphology, have a maximum carbonate content exceeding 35%, and a mean PDI of
16.4 (table 1).
Well-preserved, ubiquitous, and aerially extensive (locally 1300 m2), Qt7 is the only true fill terrace (110 m of alluvium atop the strath) within the
canyon study area (fig. 6c). The Qt7 strath has several meters of relief, is consistently located approximately 6–20 m above the modern channel in
the medial reaches of the study area, and provides
an excellent reference surface from which to identify and correlate terraces near the modern river
channel. Qt7 alluvium is composed of well-sorted
beds of stratified sand alternating with wellrounded clast-supported gravel and cobble beds.
Gravelly beds are dominated by sedimentary clasts
(fig. 7) that are locally cemented with carbonate.
Soils are generally intact on the Qt7 tread and contain a relatively uniform distribution of pedogenic
carbonate throughout the soil profile—averaging
between 10% and 20%—and have a mean PDI of
10.3 (table 1; Wisniewski 1999). Multiple carbonate
bulges in several profiles, however, suggest buried
soils indicative of a tread subjected to both postdepositional erosion and net deposition. The sedimentology, clear presence of buried soils, and carbonate accumulation distinguish Qt7 from both
older and younger terraces. Detrital charcoal from
Figure 6. Representative cross sections oriented generally from the southwest to the northeast orthogonal to the
orientation of the Canadian River canyon showing (a) relationship of the canyon to the Ogallala Group and numerically
dated basalts of the Ocate and Raton-Clayton volcanic fields. Numeric ages west of the canyon are 40K/40Ar dates
from O’Neill and Mehnert (1988), and those east of the canyon are 40Ar/39Ar dates summarized from Stroud (1997).
Small open circles represent the Ogallala Group and gray regions are basalts. Thickness of deposits are schematic
only. b, The inset terrace stratigraphy of the Canadian River canyon. c, The details of middle-Pleistocene and Holocene
terrace stratigraphy adjacent to the river. Note locations of radiocarbon dates from terraces in (c) and the 40Ar/39Ar
date obtained from the Maxson Crater basalt in (b). Dark gray caps represent approximate thickness of terrace alluvium.
Qt1, Qt2, etc. p Quaternary terraces, Qt10fp p Quaternary terrace flood plain, Qau p Quaternary alluvium undifferentiated, and Qb p Quaternary basalt.
446
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
Table 1. Summary of Numeric Age, Approximate Age, Incision Rate, and Soil Characteristics for Canadian River
Terrace Deposits, Associated Surficial Deposits, and Maxson Crater Basalts
Terrace or
related
deposit
Qau
Qt10fp
Qfu
Qt9
Qt8
Qt7
Qt6
Qt5
Qt4
Qt3
Qt2
Qb
Qt1
To
Numeric age
Approximate age
120 Ⳳ 40 yr b.p.
600 Ⳳ 40 yr b.p.
e
22,510 Ⳳ 70 yr b.p.
26,980 Ⳳ 700 yr b.p.
150,000 yr b.p.f
122,510 Ⳳ 70 yr b.p.
1.46 Ⳳ 0.02 Mag
∼4.0 Ⳳ 0.6 Mah
Strath height
above channel
(m)a
Incision
rate
(mm/yr)b
Mean
soil
PDIc
Correlative
Pecos Valley
terracesd
Late Holocene
Holocene
Late Pleistocene
Late Pleistocene
Late Pleistocene
Middle Pleistocene
Middle Pleistocene
Middle Pleistocene
Middle Pleistocene
Middle Pleistocene
Early Pleistocene
Early Pleistocene
Early Pleistocene
Early Pliocene
2.2
3.6
2
6
12
31
38
65
83
∼90
92
∼118
∼300
(5–7)
(6–20)
(28–32)
(32–42)
(60–70)
(78–88)
.074
.128f
Qp3a/b
10.3
16.4
15
8.2
Qp2
.063
.2
a
Strath height reported as the modal height. In parentheses are ranges of strath height for a given terrace throughout the canyon.
Reported as a representative rate consistent with strath heights at the canyon mouth at Sabinoso.
c
Full soil data are reported in Wisniewski (1999); PDI calculated according to Harden 1982. Field and lab procedures from Birkeland
(1999) and Singer and Janitzky (1986).
d
Pecos Valley geomorphic and soils data are reported in Karas (1987, 1988).
e
Radiocarbon ages reported in radiocarbon years before present. Analyses by Beta Analytic with original reports recorded in Wisniewski (1999).
f
Sample was radiocarbon dead so the age is a minimum, whereas the incision rate is a maximum.
g 40
Ar/39Ar age. Value is a weighted mean of four samples of basalt collected at the confluence reach (B. Olmsted and W. McIntosh,
unpub. data).
h
Age is reported as the median age between stratigraphically limiting basalt flows in the Ocate volcanic field.
b
an alluvial fan (Qfu) that locally caps Qt7 yielded
an age of 22,510 Ⳳ 70 14C yr b.p. (fig. 6c).
Qt8 and Qt9 are inset below Qt7 and typically lie
a few meters above the modern channel (fig. 6c). Qt8
is characterized by thick (1–3.5 m) beds of imbricated, matrix-supported gravel with cross-bedded
silty sand interbeds. Qt8 is locally capped by a mottled, organic-rich soil, commonly buried by finegrained material associated with undifferentiated alluvium (Qau). Qt9 is limited in aerial extent but
clearly occupies a strath above the modern channel.
Qt9 alluvium is composed primarily of imbricated,
well-rounded clasts with a poorly developed soil on
the tread. In contrast, Qt10fp alluvium lies on a
strath coincident with the exposed channel bottom,
is composed of moderately stratified, well-sorted
sand and silt with local gravel interbeds, and has
poorly developed soils with ∼5% carbonate and a
mean PDI of 3.6. Detrital charcoal found in finegrained facies of Qt8 and Qt9 alluvium yielded radiocarbon ages of 150,000 14C yr b.p. and 26,980 Ⳳ
700 14C yr b.p., respectively (table 1; fig. 6c). Detrital
charcoal collected from fine-grained alluvium (Qau)
capping Qt8 and Qt9 yielded respective ages of
600 Ⳳ 40 and 120 Ⳳ 40 14C yr b.p. (table 1; fig. 6).
Terraces are correlated on the basis of map continuity, alluvium composition, and relative soil de-
velopment, a crude proxy for terrace age. Distinct
vertical separation between terrace straths (fig. 6b,
6c) suggests that the amount of time between terrace levels is greater than the amount of time
spanned during the formation of a given terrace.
Thus, compositional (fig. 7) and relative age criteria
(table 1) provide useful guides in terrace correlation
(Pazzaglia and Gardner 1993).
Three general trends are obvious in the terrace
alluvium compositional data: volcanic clasts, presumably from the Mora River watershed, become
more dominant in terrace alluvium in the downstream direction; sedimentary clasts become increasingly frequent in progressively younger terrace
alluvium; and the composition of alluvium in the
modern channel is distinct from the composition
of terrace alluvium (fig. 7). The first two compositional trends help distinguish Qt1 from younger
terraces, which provides an upper-bounding limit
on how the terraces can be correlated. Qt1 alluvium
is rich in quartzite, gneissic, and granitic cobbles
that have no local bedrock source, while all other
terraces contain proportionally more igneous (basaltic) or sedimentary clasts (fig. 7). Post-Qt1 terrace alluvium compositions are influenced by local
contributions from the sedimentary strata that
compose the canyon walls, whereas Qt1 clasts are
Journal of Geology E P E I R O G E N I C C O N T R O L S O N C A N A D I A N R I V E R I N C I S I O N
Figure 7. Ternary plot showing distribution of terrace
gravel rock-types. The bold numbers adjacent to or within
a circled data population indicate the relative distance
downstream a population lies from the Route 120 bridge
at the north end of the study area. See Wisniewski (1999)
for precise locations of rock-type data. There are three
general trends obvious in these data. The dashed arrow
shows volcanic clasts, from the Mora River watershed presumably, becoming more dominant in terrace alluvium in
the downstream direction. The solid arrow shows sedimentary clasts becoming increasingly frequent in progressively younger terrace alluvium. Finally, clast counts
within the modern channel do not reflect the trends discernible in the terrace alluvium, which suggests that deposition of the terrace alluvium is dominated by an upper
basin (Sangre de Cristo) provenance and/or watershed
hydrology.
likely derived primarily from the reworking of
quartzite and granitic clasts from the basal conglomerate of the Ogallala Group preserved along
the canyon rim. The fact that the modern channel
alluvium contains proportionally less crystalline
rock types than the terrace alluvium suggests that
the upper reaches of the Canadian watershed in the
Sangre de Cristo Mountains delivered proportionally more discharge and sediment during times of
terrace alluvium deposition than at present.
Correlation of younger terraces (Qt5–Qt10fp) relies more heavily on relative soil development criteria. Soils on these younger terraces become increasingly well developed, and PDI values generally
increase with vertical distance above the modern
447
channel (table 1). For example, soils developed on
the modern floodplain (Qt10fp) show uniform and
only small accumulations of pedogenic carbonate
throughout the soil profile. In contrast, stage 3 or
greater calcic horizon development is present in the
soils of Qt6 or older. The lower and more variable
PDI numbers, more variable but typically greater
depths to and percentages of carbonate, and more
complex calcic horizon morphology of older terraces (table 1) is consistent with significant postdepositional erosion or aggradation on the terrace
treads.
Intact soils developed on terraces of the Canadian
River were generally compared to a soil chronosequence constructed by Karas (1987, 1988) for terraces of the Pecos River (table 1). Although this
chronosequence is not well dated, it is the closest
local calibration of calcic soil development keyed
into the Desert Project in southern New Mexico
(Gile et al. 1966, 1981). For example, Qt7 is most
similar to Qp2 in the Pecos Valley, which is thought
to be late middle Pleistocene age. Likewise, Qt9
(numerical age is ∼27 ka) is similar to Qp3a/b (estimated age range of 10–50 ka).
Incision Rates. The rate of vertical incision of
the Canadian River can be reconstructed for any
reach where there are dated stratigraphic horizons
like terrace alluvium and the Maxson Crater basalt.
The reach with the best numeric age control is near
the town of Sabinoso where the Ogallala Group
could be projected with reasonable confidence from
the eastern rim of the canyon, the base of the Maxson Crater basalt flow is exposed, the soil in Qt7
is well expressed and likely intact, and the younger
terraces are radiocarbon dated. The reconstructed
long-term average incision rate for this reach, according to the elevations of the terrace straths,
ranges between ∼0.06 and 0.07 mm/yr (60–70 m/
m.yr.) (fig. 8). Within the number of constraints of
our dated horizons, we have no reason to suggest
that there has been any significant acceleration or
deceleration of those rates over the Plio-Pleistocene
time interval.
Discussion
Reconstruction of Paleo-Canadian River Long Profiles. Canyons are dramatic evidence that the rates
of river incision can locally far exceed the average
regional rate of denudation. In the case of the Canadian River, a basalt horizon and well-preserved
and dated suite of terraces define paleo-long profiles
of the Canadian River that not only quantify the
incision rate but also document how the incision
was accomplished (fig. 9). Both the incision rate and
448
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
Figure 8. Rate of incision of the Canadian River near the mouth of the canyon at Sabinoso, New Mexico. The inset
graph shows the long-term rate of incision (0.074 mm/yr; 74 m/m.yr.) produced by regressing through the stratigraphically constrained age of the Ogallala Group strath and Maxson Crater basalts. The large plot shows the relationship of the long-term regression line with respect to the incision rate averaged since the emplacement of the
basalt (0.063 mm/yr; 63 m/m.yr.). Note that the age of the Qt8 terrace is a minimum at 50,000 yr b.p., so the inferred
incision rate for the plotted point is a maximum.
processes are consistent with contemporary rock
uplift coincident with magmatic processes associated with the Jemez lineament.
The Canadian River apparently was already well
established with the Mora River as a major tributary at the time of Maxson Crater basalt flows ∼1.5
Ma (Dolliver 1984). The basalt and modern Mora
River long profiles are parallel from Maxson Crater
(fig. 4b; Dolliver 1984) to a knickzone approximately 20 km from the confluence. Downstream
of the knickzone, the channel gradient steepens relative to the basalts. Divergence of the basalt and
the Mora River channel downstream of the knickzone strongly suggests that incision of the Canadian River caused a relative baselevel fall for the
Mora River, resulting in the generation of a knickzone that has propagated and reclined upstream on
the Mora River (Gardner 1983).
Terraces represent the analogous paleo-long profiles in the Canadian valley that the Maxson Crater
basalt provides in the Mora valley; Qt1 and Qt7
contain sedimentologic and/or soil morphologic
criteria that makes their correlation particularly
compelling and their long profiles as upper and
lower guides for the intervening terraces (fig. 9).
Both the Qt1 and Qt7 straths are subparallel to the
modern channel, exhibiting a subtle downstream
convergence. Knickpoints in the modern channel
long profile are mirrored in the long profile of Qt7
(fig. 9), a relationship we also observe between the
profile of the Maxson Crater basalts and the Mora
River profile (fig. 4b). Alternative correlations of
Qt7 that project it upstream into a knickpoint in
the modern Canadian River channel (fig. 9a) are not
supported by map, soil, or gravel rock-type data.
The resulting correlation of all terraces produce terrace long profiles that appear to be nearly parallel
to one another over short distances (fig. 9) but that
are broadly convex or warped over larger distances
(fig. 9b).
Our finding that the terraces within the Canadian River canyon mimic or even exaggerate the
Journal of Geology E P E I R O G E N I C C O N T R O L S O N C A N A D I A N R I V E R I N C I S I O N
449
Figure 9. Proposed terrace correlation through the canyon reach for Canadian River terraces (Qt1–Qt7) and Maxson
Crater basalts (Qb). Qt8 and Qt9 are too close to the channel to plot at this scale. Filled diamonds represent fieldchecked terrace strath elevations. Open triangles denote terraces inferred from topographic maps and aerial photographs. Lines are solid where correlations are firm and dashed where correlations are less confident. Guide terraces
are those for which correlation is well supported by rock type, soils, or related data. Note the broad convexity in the
modern channel profile mimicked in an exaggerated fashion by the terrace profiles. Terrace profiles are parallel to
one another over short length scales, but downstream of the confluence reach they begin to converge toward the
modern channel, in contrast to the long profile of the Mora River and basalts in figure 4b. Note that the long profile
correlations are roughly parallel to the modern long profile and, in places, even mimic knickpoints such as the one
formed between the Bells Ranch–Entrada and Chinle formations. Terrace data do not support correlations such as
those in a, a finding that is consistent with upstream migrating knickpoints. Rather, we envision the general correlation as resembling the schematic shown in inset b. All terrace data for this plot are projected to a vertical plane
constructed down the center of the Canadian valley. Symbols for rock types exposed beneath the river are as described
in figure 4.
broad modern channel convexity is an unexpected
result in light of previous interpretations of the origin of the canyon that focus on evaporite dissolution in the eastern Great Plains of New Mexico
and Texas panhandle and the commensurate downstream baselevel fall (Gustavson 1986). The general
downstream increase in SL indices is consistent
with a relative downstream baselevel fall. However, if the canyon is a broad knickzone composed
of many separate knickpoints that all have a common downstream origin, then we would expect to
find terraces that diverge in the downstream direction and project directly to knickpoints that are
propagating upstream (fig. 9a). We do not observe
this and are left to conclude that the broad con-
vexity, at least for the time represented by the age
of the oldest terraces, is a persistent feature. Perhaps the rate of upstream propagation of the knickzone is slow such that we cannot resolve it using
terraces that are created by other mechanisms, over
shorter time spans. Different processes appear to
drive incision on the Jemez lineament-parallel
Mora and Jemez lineament-orthogonal Canadian
Rivers; and Canadian River incision is not dominated by parallel retreat of a knickzone.
The presence of terraces in the Canadian River
canyon indicate that the river experienced relatively brief periods of stasis or even aggradation
during its general history of incising and deepening
the canyon. Assuming that the long-term average
450
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
rate of incision has been relatively constant (fig. 8),
we find that the terrace strath elevations predict
times of terrace formation to be, in a general sense,
coincident with known times of alpine glaciation
in the Rocky Mountains (Rogers and Smart 1996;
table 2). These data by no means prove that the
terraces have a climatic origin but suggest that the
long-term incision of the canyon was punctuated
by periods of strath carving and aggradation atop
the strath driven by up-basin changes in discharge
and sediment yield. This interpretation is supported by the distinction in terrace alluvium composition with respect to modern channel alluvium
composition (fig. 7) as well as terrace long profiles
that do not appear to have a genetic relationship to
the channel convexity/knickzone (fig. 9).
Origin of the Canadian River Canyon and Escarpment. Incision of the Canadian River canyon is
linked to the origin and evolution of the Canadian
escarpment. Downstream of the canyon, the Canadian River flows in a broad low-standing valley
that has been strongly influenced by baselevel fall
caused by evaporite-dissolution subsidence (Dolliver 1984; Gustavson 1986). The topographic expression of the Canadian escarpment is produced
by high rates of erosion within the river valley in
conjunction with relatively low rates on valley
margins that are capped with cemented, resistant
Ogallala Group.
Although the exposed bedrock stratigraphy is
similar for the Canadian escarpment and the Caprock escarpment, the Canadian escarpment stands
100–200 m higher (fig. 1) and is much more em-
bayed than the Caprock escarpment. The drainage
is consequent to the Caprock escarpment as it is
the local drainage divide for streams flowing into
or away from the Canadian drainage. In contrast,
south-flowing streams like the Canadian River
head far north of the Canadian escarpment (fig. 1),
resulting in a greatly embayed escarpment face.
The drainage on the Canadian escarpment appears
to be antecedent or subsequent, which is consistent
with a well-established drainage network prior to
the formation of the escarpment. The lack of large
consequent drainages and the higher-standing topography of the Canadian escarpment indicate that
it is a relatively youthful landform.
A possible drainage evolution scenario would
place the ancestral Cimarron River flowing directly
east from its headwaters in the Sangre de Cristo
Mountains across the region that would become
the Raton-Clayton volcanic field (fig. 10a). The ancestral Mora and Pecos Rivers would have had similar courses, both draining the Santa Fe Range east
to the Great Plains. Eruption of the Raton-Clayton
field filled the ancestral Cimarron valley with volcanics and diverted the drainage to the south, taking advantage of a minor south-flowing tributary
of the ancestral Mora River (fig. 10b). Concurrently,
the lower valleys of the Canadian and ancestral Pecos Rivers are strongly localized by evaporite dissolution (fig. 10b) and have a competitive advantage
for elongation of their headwaters. Eventually, this
dissolution-controlled baselevel fall fosters integration of the Canadian across what is now the Ca-
Table 2. Terrace Ages and Correlation to Pleistocene Marine Oxygen Isotope Stages Based on Average Strath Height
above Channel
Terrace
Qt9b
Qt8
Qt7
Qt6
Qt5
Qt4
Qt3
Qt2
Basalte
Qt1
a
Strath height
above channel
(min-max) (m)
2
6
12
31
38
65
83
∼90
92
∼118
(5–7)
(6–20)
(28–32)
(32–42)
(60–70)
(78–88)
Age based on
.063 incision rate
(ka)
Age based on
.074 incision rate
(ka)
Inferred marine
oxygen isotope
stagea
Inferred
Rocky Mountain
glaciationa
31.7
95.2
190.5
492.1
603.2
1031.8
1317.5
1428.6
1460
1873
27
81.1
162.2
418.9
513.5
878.4
1121.6
1216.2
1243.2
1594.6
2
4–5bc
6
12
14–16d
22
?
?
n/a
?
Late Pinedale
Early Pinedale?c
Bull Lake
Pre-Illinoian B
Pre-Illinoian C-D
Pre-Illinoian G
From Richmond and Fullerton (1986).
Has a numeric age of 26,980 Ⳳ 700 14C yr b.p.
c
An early Pinedale age !79 ka is permissive using the minimum elevation of the strath. Although recognized in both the marine
oxygen isotope record and for the continental ice sheets (Richmond and Fullerton 1986), mountain glaciation of this age is only
locally recognized. It is found, for example, in the Olympic Mountains of Washington State (Thackray 2001) but not in Wyoming
(Chadwick et al. 1997).
d
Strath elevation range is permissive for deposits of late stage 16 age, which is common in New Mexico and locally contains Lava
Creek B tephra.
e
Has a numeric age of 1.46 Ⳳ 0.02 Ma.
b
Figure 10. Proposed drainage reconstructions for the Canadian and Pecos River systems from the (a) late Miocene,
to the (b) Pliocene, and (c) Pleistocene superimposed on a map with county outlines for northeastern New Mexico.
Shaded regions are uplands. The ancestral Mora, Canadian, Cimarron, and Brazos-Colorado (?) (Pecos) Rivers were
important Great Plains streams draining the southern Rocky Mountain front and depositing the Ogallala Group in
the late Miocene. By the Pliocene, Ogallala Group gravel deposition was confined to the increasingly incised river
valleys, whereas the interfluves were being buried by eolian sediments. Major evaporite dissolution beneath the
ancestral Canadian and Pecos valleys helped localize these drainages as the master streams. At the same time,
volcanism in the Raton and Ocate fields disrupted rivers in the drainage headwaters. The Canadian River was likely
integrated and diverted south across the ancestral escarpment at this time. By the early Pleistocene, the modern
Pecos River was integrated. Continued fluvial erosion and dissolution of evaporites in conjunction with epeirogenic
uplift along the Jemez lineament exacerbated the relief between the lower Canadian River valley and the Canadian
escarpment.
452
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
nadian escarpment, as the Pecos beheads the ancestral Brazos and/or Colorado Rivers in its
integration as a major north-south-flowing stream
(fig. 10c). The Plio-Pleistocene record of incision of
the Canadian River canyon cannot be entirely explained with this drainage integration story and the
associated baselevel fall, however, because terrace
profiles mimic and exaggerate the convexity of the
modern channel profile and do not diverge downstream. Therefore, we are forced to consider an additional mechanism contributing to Canadian
River canyon incision.
Jemez Lineament.
The spatial coincidence of
convex terrace profiles (fig. 9), the modern Canadian River profile convexity (fig. 4a), and the Canadian escarpment (fig. 1) with the Jemez lineament (fig. 11) suggest that Canadian River canyon
incision may be related to contemporary, aseismic
(Sanford et al. 1997) uplift centered along the axis
of the Jemez lineament (fig. 1). Broad arching across
the Jemez lineament would both exacerbate the
topographic relief across the Canadian escarpment
and focus incision through the study reach. From
the Jemez lineament reference frame, rock uplift is
the same as a downstream baselevel fall resulting
in a similar broad convexity or knickzone forming
along strike with the lineament. But unlike a
knickzone propagating through an already existing
escarpment, continued rock uplift fixes the relative
position of the convexity (Harbor 1998). The terrace
and basalt long profiles support persistence of this
knickzone over at least ∼1.5 Ma. A greater degree
of uplift near the axis of the Jemez lineament has
resulted in interfluves capped by Ogallala Group
that stand higher on the Canadian escarpment than
on the Caprock escarpment, which lies further
south on the flanks of the uplift (figs. 1, 11).
Late Cenozoic arching along the Jemez lineament may have a positive feedback on the processes
of evaporite dissolution and opposing tilts of both
flanks of the lower Canadian River valley (fig. 11).
The dissolution-controlled lower Canadian River
valley lies on the southeast flank of this hypothesized arch, so it is possible that dissolution was
focused here by the coincidence of evaporites and
fractures genetically related to the arching. The extensive removal of rock from the arch flank by both
dissolution and river incision would locally unload
this portion of the Great Plains. The valley margins
Figure 11. Cross-profile C-C of figure 2 annotated to illustrate our proposed model for epeirogenic deformation and
landscape evolution. The relative position of the Jemez lineament is ∼20 km west of this profile so the Canadian
escarpment is a feature of the eastern flank of the lineament. The lineament is shown schematically as a thermal
feature, with both deep, buoyant asthenopheric (Humphreys and Dueker 1994a, 1994b) and middle crustal intrusive
origins that have upwardly bulged the lithosphere. There are two trends to the topographic data supportive of this
interpretation. The first is a general tilt to the southeast that we interpret to be driven by the long-wavelength thermal
bulge. The second is a shorter wavelength tilt of the Canadian Valley flanks away from the valley center that we
interpret as a possible local flexural response to removal of rock by both incision and dissolution processes concentrated
here on the thermal bulge’s eastern flank.
Journal of Geology E P E I R O G E N I C C O N T R O L S O N C A N A D I A N R I V E R I N C I S I O N
Table 3.
453
Comparison of Quaternary Incision Rates
River
Rio Puerco
Jemez River
Rio Grande
Canadian River
Incision rate
(mm/yr)
Location
.24
.17
.09
.063–.074
Western rift flank and Jemez lineament (Hallet 1994)
SW flank of Jemez mountains (Formento-Trigilio and Pazzaglia 1998)
Española Basin (Detheir et al. 1988)
Western Great Plains, Jemez lineament (this study)
respond to the unloading by gently flexing upward
(fig. 11) in the same manner as the flanks of a rift
flex upward in response to crustal thinning (Brown
and Phillips 1999; Roy et al. 1999). In this way, PlioPleistocene arching beneath the Jemez lineament
both drives rock uplift and exacerbates the relative
baselevel fall by fostering evaporite dissolution
along its southeast flank. The combined effect produces an impressive south-facing escarpment (the
Canadian escarpment) and a lower-standing, northfacing escarpment (the Caprock), both of which
gently tilt away from the greatest degree of rock
removal (fig. 11).
Another important piece of geomorphic evidence
supportive of Plio-Pleistocene arching and flexure
along the Jemez lineament is the presence of deformed incised meanders in the Canadian River
canyon. Incised meanders are observed in various
parts of the Western Cordillera of North America
but are best expressed on the Colorado Plateau. Deformed incised meanders are a special subset of incised meanders in which the meander loops become so sinuous that they nearly intersect each
other, forming so-called goosenecks that are typically associated with sedimentary rocks that gently
dip upstream (Gardner 1975; Harden 1990). We observe goosenecks in the Canadian River only where
the Mesozoic sedimentary rocks dip gently upstream (fig. 5). The broad, shallow changes in bedrock dip orthogonal to the Jemez lineament may
be related to Jemez lineament arching and/or
flexure.
The driving force behind Jemez lineament uplift
is a matter of conjecture, but recent geophysical
data support the notion that the Jemez lineament
has a different crustal composition, tectonic history, and thermal structure related to its origin as
a Proterozoic suture (Karlstrom and Humphreys
1998). Thermally buoyant support of the lithosphere (Lerner-Lam et al. 1998) is consistent with
the volcanic activity localized along the Jemez lineament. Dynamic support of high-standing topography along the Jemez lineament would be similar
in process, but smaller in scale, to that which is
well documented for the Yellowstone hotspot
(Pierce and Morgan 1992; Humphreys and Dueker
1994a).
Incision Rate Comparisons.
Incision of the Canadian River may also reflect a rock uplift signal
that, unlike most major rivers in New Mexico, is
not obscured by rift tectonics. Long-term average
rates of incision determined for the Canadian River
are significantly slower than those values reported
for other New Mexico rivers like the Rio Puerco
(Hallet 1994), the Rio Jemez (Formento-Trigilio and
Pazzaglia 1998), the Rio Chama (Gonzalez and Dethier 1991), and the master stream, the Rio Grande
(Dethier et al. 1988; table 3). All of these rivers, including the Canadian River, cross the Jemez lineament, but only the Canadian River lies outside of
the influence of rift tectonics. The combined effects
of rock uplift associated with rifting and those associated with the Jemez lineament may explain why
rates of incision are greater for the Rio Grande and
its tributaries than incision rates calculated for the
Canadian River. The rate of Canadian River incision
may solely reflect a characteristic rate of dynamic
lithospheric inflation by a thermally buoyant asthenosphere.
Conclusions
Post-Laramide epeirogeny astride the Jemez lineament has played a key role in the geomorphic evolution of the Canadian River canyon of northeastern New Mexico. A broad convexity in the long
profile of the modern Canadian River, which flows
orthogonal to the northeast strike of the Jemez lineament axis, is absent from other major rivers
draining the eastern flank of the southern Rocky
Mountain front. Basalt and river terrace long profiles that mimic and exaggerate the channel long
profile convexity suggest that epeirogenic arching
has been persistent over geologically significant
time scales. The 1.46 Ⳳ 0.02 m.yr. Maxson Crater
basalts and nine major fluvial terraces (Qt1–Qt9)
elucidate the pattern and rate of fluvial incision
since the early Pleistocene. Despite the documented processes of evaporite-dissolution that
drive baselevel fall downstream of the canyon, the
reconstructed terrace long profiles within the can-
454
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
yon study area are more indicative of epeirogenic
arching than knickpoint propagation as the dominant process influencing river incision. We speculate that the epeirogenic arching is related to the
dynamic support of the lithosphere by warm, buoyant asthenosphere in a process analogous to, but
on a smaller scale than, that described for Yellowstone. In concert with the evaporite dissolution on
the eastern Great Plains of New Mexico, this epeirogenic arching has allowed the Canadian River to
etch out impressive opposing-faced escarpments on
the southeastern flank of the Jemez lineament,
whose backslope dips may be modified by local
flexural effects.
Long-term average rates of fluvial bedrock incision determined for the Canadian River are relatively steady and slow at ∼0.06–0.07 mm/yr. These
rates are slower than those measured for other,
nearby New Mexico rivers but may represent a
baseline proxy for rates of rock uplift along the Jemez lineament outside of the influence of the Rio
Grande rift. If Canadian River incision rates approximate rates of uplift for the Las Vegas Plateau,
then the Jemez lineament section of the Plateau
should have been uplifted by ∼0.6 km since the late
Miocene (∼9 Ma), the age of the oldest volcanic
units in the nearby Raton-Clayton and Ocate vol-
canic fields. The presence of the 300–400-m-deep
Canadian River canyon indicates that a significant
quantity of the uplifted rock has not yet been
stripped from the Las Vegas Plateau upland, resulting in an effective increase in mean elevations
there since the late Miocene.
ACKNOWLEDGMENTS
This research was supported by Continental Dynamics of the Rocky Mountain Project NSF-EAR9416787 (F. J. Pazzaglia), the Geological Society of
America (P. A. Wisniewski), the Colorado Scientific
Society (P. A. Wisniewski), and the Department of
Earth and Planetary Sciences, University of New
Mexico. We wish to thank J. Ashby, M. Eppes, B.
McIntosh, J. Hawley, and K. Karlstrom for logistical, financial, and moral support during this study.
We thank the people of Sabinoso, New Mexico, for
their warmth and hospitality during the field investigations. We completed this work while we
were both at the University of New Mexico, and
we wish to thank the Department of Earth and
Planetary Sciences for its support. The detailed and
insightful reviews of this manuscript by M. Blum
and G. Meyer were greatly appreciated.
REFERENCES CITED
Ander, M. E. 1981. Geophysical study of the crust and
upper mantle beneath the central Rio Grande Rift and
adjacent Great Plains and Colorado Plateau. Los Alamos Scientific Laboratory 8676-T, 235 p.
Baldwin, B., and Muehlberger, W. R. 1959. Geological
studies of Union County New Mexico. N.M. Bur.
Mines Miner. Resour. Bull. 63, 171 p.
Birkeland, P. W. 1999. Soils and geomorphology (3d ed.).
New York, Oxford University Press, 430 p.
Brown, C. D., and Phillips, R. J. 1999. Flexural rift flank
uplift at the Rio Grande Rift, New Mexico. Tectonics
18:1275–1291.
Calvin, E. 1987. A review of the volcanic history and
stratigraphy of northeastern New Mexico, the Ocate
and Raton-Clayton volcanic fields. N.M. Geol. Soc.
Guidebook 38:83–85.
Chadwick, O. A.; Hall, R. D.; and Phillips, F. M. 1997.
Chronology of Pleistocene glacial advances in the central Rocky Mountains. Geol. Soc. Am. Bull. 109:
1443–1452.
Chase, C. G.; Gregory-Wodzicki, K. M.; Parrish, J. T.; and
DeCelles, P. G. 1998. Topographic history of the Western Cordillera of North America and controls on climate. In Crowley, T. J., and Burke, K. C., eds. Tectonic
boundary conditions for climate reconstructions. Oxf.
Monogr. Geol. Geophys. 39:73–99.
Cordell, L., and Keller, G. R. 1984. Regional structural
trends inferred from gravity and aeromagnetic data in
the New Mexico–Colorado border region. N.M. Geol.
Soc. Guidebook 25:21–23.
Dane, C., and Bachman, G. O. 1965. Geologic map of
New Mexico. U.S. Geol. Surv., scale 1 : 1,000,000.
Dethier, D. P.; Harrington, C. D.; and Aldrich, M. J. 1988.
Late Cenozoic rates of erosion in the western Española
basin, New Mexico: evidence from geologic dating of
erosion surfaces. Geol. Soc. Am. Bull. 100:928–937.
Dobrovolny, E.; Summerson, C. H.; and Bates, R. L. 1947.
Geology of northwestern Quay County, New Mexico.
Sheet 1: geologic map and stratigraphic sections of
Mesozoic rocks (Dobrovolny, E., and Summerson, C.
H.). Sheet 2: Subsurface geology (Bates, R. L.). U.S.
Geol. Surv. Oil Gas Investig. Map, scale 1 : 62,500.
Dolliver, P. N. 1984. Cenozoic evolution of the Canadian
River basin. M.S. thesis. Baylor Geol. Stud. Bull. 42,
96 p.
Eaton, G. P. 1987. Topography and origin of the southern Rocky Mountains and Alvarado Ridge. In Coward, M. P.; Dewey, J. F.; and Hancock, P. L., eds. Continental extensional tectonics. Geol. Soc. Spec. Publ.
28:355–369.
Epis, M. K., and Chapin, C. E. 1975. Geomorphic and
tectonic implications of the post-Laramide, late Eocene erosion surface in the Southern Rocky Mountains. In Curtis, B. F., ed. Cenozoic history of the
Journal of Geology E P E I R O G E N I C C O N T R O L S O N C A N A D I A N R I V E R I N C I S I O N
Southern Rocky Mountains. Geol. Soc. Am. Mem.
144:45–74.
Epis, R. C., and Callender, J. F. 1981. Geologic and physiographic highlights of the Black Canyon of the Gunnison River and vicinity, Colorado. N.M. Geol. Soc.
Guidebook 32:145–154.
Formento-Trigilio, M. L., and Pazzaglia, F. J. 1998. Tectonic geomorphology of the Sierra Nacimiento: traditional and new techniques in assessing long-term
landscape evolution in the southern Rocky Mountains. J. Geol. 106:433–453.
Gardner, T. W. 1975. The history of part of the Colorado
River and its tributaries: an experimental study. Four
Corners Geol. Soc. Guidebook 8:87–95.
———. 1983. Experimental study of knickpoint and longitudinal profile evolution in cohesive, homogeneous
material. Geol. Soc. Am. Bull. 94:664–672.
Gile, L. H.; Hawley, L.; and Grossman, R. 1981. Soils and
geomorphology in the Basin and Range area of southern New Mexico. Guidebook to the desert project.
N.M. Bur. Mines Miner. Resour. Mem. 39, 222 p.
Gile, L. H.; Peterson, F. F.; and Grossman, R. B. 1966.
Morphological and genetic sequences of carbonate accumulation in desert soils. Soil Sci. 101:347–360.
Gonzalez, M. A., and Dethier, D. P. 1991. Geomorphic
and neotectonic evolution along the margin of the
Colorado Plateau and Rio Grande rift, northern New
Mexico. N.M. Bur. Mines Miner. Resour. Bull. 137:
29–45.
Gregory, K. M., and Chase, C. G. 1992. Tectonic significance of paleobotanically estimated climate and altitude of the late Eocene erosion surface, Colorado.
Geology 20:581–585.
Gustavson, T. C. 1986. Geomorphic development of the
Canadian River Valley, Texas Panhandle: an example
of regional salt dissolution and subsidence. Geol. Soc.
Am. Bull. 97:459–472.
Gustavson, T. C.; Baumgardner, R. W., Jr.; Aran, S. C.;
Holliday, V. T.; Mehnert, H. H.; O’Neill, J. M.; and
Reeves, C. C., Jr. 1991. Quaternary geology of the
southern Great Plains and an adjacent section of the
Rolling Plains. In Morrison, R. B., ed. Quaternary nonglacial geology, conterminous U.S. The geology of
North America. Geol. Soc. Am. K-2:477–501.
Gustavson, T. C.; Simpkins, W. W.; Alhades, A.; and
Hoadley, A. 1982. Evaporite dissolution and development of karst features on the rolling plains of the
Texas Panhandle. Earth Surf. Proc. Landf. 7:545–563.
Gustavson, T. C., and Winkler, D. A. 1988. Depositional
facies of the Miocene-Pliocene Ogallala Formation,
northwestern Texas and eastern New Mexico. Geology 16:203–206.
Hack, J. T. 1973. Stream-profile analysis and streamgradient index. U.S. Geol. Surv. J. Res. 1:421–429.
Hallet, R. B. 1994. Volcanic geology, paleomagnetism,
geochronology and geochemistry of the Rio Puerco
Necks, west-central New Mexico. Ph.D. dissertation,
New Mexico Institute of Mining and Technology, Socorro, 340 p.
455
Hansen, W. R. 1965. The Black Canyon of the Gunnison:
today and yesterday. U.S. Geol. Surv. Bull., 76 p.
———. 1967. The lower Black Canyon of the Gunnison.
Natl. Parks Conserv. Mag. 41:14–19.
Harbor, D. J. 1998. Dynamic equilibrium between an active uplift and the Sevier River, Utah. J. Geol. 106:
181–194.
Harden, D. R. 1990. Controlling factors in the distribution
and development of incised meanders in the central
Colorado Plateau. Geol. Soc. Am. Bull. 102:233–242.
Harden, J. W. 1982. A quantitative comparison of soil
development in four climatic regimes. Quat. Res. 20:
342–359.
Holliday, V. T. 1990. Soils and landscape evolution of
eolian plains: the southern High Plains of Texas and
New Mexico. Geomorphology 3:489–515.
Huber, N. K. 1981. Amount and timing of uplift and tilt
of the central Sierra Nevada, California: evidence from
the upper San Joaquin River basin. U.S. Geol. Surv.
Prof. Pap. 1197, 28 p.
Humphreys, E. D., and Dueker, K. G. 1994a. Physical
state of the western U.S. upper mantle. J. Geophys.
Res. 99:9635–9650.
———. 1994b. Western U.S. upper mantle structure. J.
Geophys. Res. 99:9615–9634.
Hunt, C. B. 1956. Cenozoic geology of the Colorado Plateau. U.S. Geol. Surv. Prof. Pap. 279, 28 p.
———. 1969. Cenozoic history of the Colorado River.
U.S. Geol. Surv. Prof. Pap. 669-C, p. 59–130.
Karas, P. A. 1987. Quaternary alluvial sequence of the upper Pecos River and a tributary, Glorieta Creek, northcentral New Mexico. In Menges, C. M.; Harrison, B.;
and Enzel, Y., eds. Quaternary tectonics, landform evolution, soil chronologies, and glacial deposits: northern
Rio Grande rift of New Mexico. Friends of the Pleistocene—Rocky Mountain Cell Field Trip Guidebook,
p. 159–176.
———. 1988. Quaternary alluvial sequence of the upper
Pecos River and a tributary, Glorieta Creek, northcentral New Mexico. M.S. thesis, University of New
Mexico, Albuquerque, 112 p.
Karlstrom, K. E., and Daniel, C. G. 1993. Restoration of
Laramide right-lateral strike slip in northern New
Mexico by using Proterozoic piercing points: tectonic
implications from the Proterozoic to the Cenozoic.
Geology 21:1139–1142.
Karlstrom, K. E., and Humphreys, E. D. 1998. Persistent
influence of Proterozoic accretionary boundaries in
the tectonic evolution of southwestern North America: interaction of cratonic grain and mantle modification events. Rocky Mt. Geol. 33:161–179.
Kelley, S. A., and Chapin, C. E. 1995. Apatite fissiontrack thermochronology of southern Rocky Mountain–Rio Grande Rift–western High Plains provinces.
N.M. Geol. Soc. Guidebook 46:87–95.
Kerr, R. A. 1997. Why the West stands tall. Science 275:
1564–1565.
Lerner-Lam, A. L.; Sheehan, A. F.; Grand, S.; Humphreys,
E. D.; Dueker, K. G.; Hessler, E.; Guo, H.; Lee, D. K.;
and Savage, M. 1998. Deep structure beneath the
456
P. A . W I S N I E W S K I A N D F. J . P A Z Z A G L I A
southern Rocky Mountains from the Rocky Mountain
front broadband seismic experiment. Rocky Mt. Geol.
33:199–216.
Lucchitta, I. 1984. Development of landscape in northwestern Arizona: the country of plateaus and canyons.
In Smiley, T. L.; Nations, J.; Péwé, T.; and Schafer, J.,
eds. Landscapes of Arizona. New York, University
Press of America, p. 269–302.
O’Neill, M. J., and Mehnert, H. 1988. Petrology and physiographic evolution of the Ocate volcanic field, northcentral New Mexico. U.S. Geol. Surv. Prof. Pap. 1478,
pts. A and B, 45 p.
Pazzaglia, F. J., and Gardner, T. W. 1993. Fluvial terraces
of the lower Susquehanna River. Geomorphology 8:
83–113.
Pazzaglia, F. J., and Kelley, S. A. 1998. Large-scale geomorphology and fission-track thermochronology in
topographic and exhumation reconstructions of the
southern Rocky Mountains. Rocky Mt. Geol. 33:
229–257.
Pierce, K. L., and Morgan, L. A. 1992. The track of the
Yellowstone hotspot: volcanism, faulting, and uplift.
Geol. Soc. Am. Mem. 179:1–53.
Powers, W. E. 1935. Physiographic history of the upper
Arkansas River valley and the Royal Gorge, Colorado.
J. Geol. 43:184–199.
Richmond, G. M., and Fullerton, D. S. 1986. Introduction
to Quaternary glaciations in the United States of
America. In Sibrava, V.; Bowen, D. Q.; and Richmond,
G. M., eds. Quat. Sci. Rev. 5:3–10.
Roberson, G. D. 1972. The geomorphology of the Canadian
River basin. M.S. thesis, Baylor University, Waco, Tex.,
156 p.
Rogers, J. B., and Smart, R. A. 1996. Climate influences
on Quaternary alluvial stratigraphy and terrace formation in the Jemez River valley. N.M. Geol. Soc.
Guidebook 47:347–356.
Roy, M.; Karlstrom, K. E.; Kelley, S. A.; Pazzaglia, F. J.;
and Cather, S. 1999. Topographic setting of the Rio
Grande rift, New Mexico: assessing the role of flexural
“rift-flank uplift” in the Sandia Mountains. N.M.
Geol. Soc. Guidebook 50:167–174.
Sanford, A. R.; Lin, K. W.; and Tsai, I. C. 1997. Preliminary
seismicity map for New Mexico and bordering areas.
New Mexico Institute of Mining and Technology, Geophysical Research Center, Open File Rep. 83, 2 p.
Schumm, S. A. 1967. Meander wavelength of alluvial
rivers. Science 157:1549–1550.
Singer, M., and Janitzky, P., eds. 1986. Field and labora-
tory procedures used in a soil chronosequence study.
U.S. Geol. Surv. Bull. 1648, 49 p.
Small, E. E., and Anderson, R. S. 1998. Pleistocene relief
production in Laramide mountain ranges, western
United States. Geology 26:123–126.
Smith, R. L., and Luedke, R. G. 1984. Potentially active
volcanic lineaments and loci in western coterminous
United States. In Explosive volcanism: inception, evolution, and hazards. National Academy Press, p. 47–65.
Soil Survey Staff. 1975. Soil taxonomy: a basic system of
soil classification for making and interpreting soil surveys. U.S. Dep. Agric. Handb. 436, 754 p.
Spence, W., and Gross, R. S. 1990. A tomographic glimpse
of the upper mantle source of magmas of the Jemez
lineament. J. Geophys. Res. 95:10,829–10,849.
Spiegel, Z. 1972. Cenozoic geology of the Canadian River
valley, New Mexico. N.M. Geol. Soc. Guidebook 23:
118–119.
Steidtmann, J. R.; Middleton, L. T.; and Shuster, M. W.
1989. Post-Laramide (Oligocene) uplift in the Wind
River Range, Wyoming. Geology 17:38–41.
Stroud, J. R. 1997. Geochronology of the Raton-Clayton
volcanic field, with implications for volcanic history
and landscape evolution. M.S. thesis, New Mexico Institute of Mining and Technology, Socorro, 51 p.
Thackray, G. D. 2001. Extensive early and middle Wisconsin glaciation on the western Olympic Peninsula,
Washington, and the variability of Pacific moisture
delivery to the northwestern United States. Quat. Res.
55:257–270.
Unruh, J. R. 1991. The uplift of the Sierra Nevada and
implications for late Cenozoic epeirogeny in the Western Cordillera. Geol. Soc. Am Bull. 103:1395–1404.
Wanek, A. A. 1962. Reconnaissance geologic map of parts
of Harding, San Miguel, and Mora Counties, New Mexico. U.S. Geol. Surv. Map OM-208, scale 1 : 96,000.
Winkler, D. A. 1987. Vertebrate-bearing eolian unit from
the Ogallala Group (Miocene) in northwestern Texas.
Geology 15:705–708.
Wisniewski, P. A. 1999. A fluvial record of rock uplift
along the Jemez lineament, Great Plains of northeastern New Mexico. M.S. thesis, University of New
Mexico, Albuquerque, 118 p.
Wolfe, J. A.; Forest, C. E.; and Molnar, P. 1998. Paleobotanical evidence of Eocene and Oligocene paleoaltitudes in midlatitude western North America. Geol.
Soc. Am. Bull. 110:664–678.
Woodward, L. A. 1987. Tectonic map of the Rocky Mountain region of the United States. Geol. Soc. Am., scale
1 : 2,500,000.