Abstract

The McClure Mountain–Iron Mountain igneous complex is an alkalic intrusive center in the northern Wet Mountains of southern Colorado. It was emplaced in early Cambrian time into gneissic/granitic 1.75–1.45 Ga Proterozoic host rocks. Numerous dikes are associated with the complex, primarily along the western side. Although the main intrusive nepheline-syenite body is well dated, the ages of the surrounding dikes are poorly known. Crosscutting relationships and poorly defined K-Ar dates suggest that the dikes are younger than the main intrusion. Paleomagnetic samples were collected from dikes associated with the McClure Mountain igneous complex. Geochronologic samples were also collected from two dikes sampled for their paleomagnetism. We obtained U-Pb zircon ages of 526 ± 8 Ma for a lamprophyric extracomplex dike and 483 ± 2 Ma for a trachytic extracomplex dike. These ages suggest either multistage or protracted dike intrusion around the ca. 524 Ma McClure Mountain complex. Our paleomagnetic data are consistent with previously published results. Dikes of the complex primarily exhibit southeast and shallow paleomagnetic directions, with variable declinations. Results from several baked contact tests indicate that the magnetizations are secondary. A steeply inclined magnetization is pervasive and was acquired over a protracted interval from late Laramide time to the present day.

INTRODUCTION

Rapid cooling and the potential for both relative and absolute dating render dikes as highly favorable targets for paleomagnetic research (Van der Voo, 1990; Halls, 2008). Integrated studies to simultaneously recover both magnetizations and emplacement ages for these intrusive bodies allow researchers to determine accurate spatiotemporal relationships throughout the entire range of geologic time. Several external factors complicate the retention of a primary magnetization. These factors include heat (Pullaiah et al., 1975), fluid flow (McCabe and Elmore, 1989; Geissman and Harlan, 2002), viscous remanent magnetizations (Kent, 1985), instability of magnetic phases (Butler, 1992), and lightning strikes (Graham, 1961).

An enigmatic collection of igneous rocks with early Paleozoic ages in southern Colorado and New Mexico intruded Paleoproterozoic to Mesoproterozoic host rocks of this area long after the development of a Neoproterozoic–Cambrian passive margin (Armbustracher and Hedge, 1982; Armbustracher, 1984; Bickford et al., 1989; McMillan and McLemore, 2004; Yonkee et al., 2014). The tectonic setting of these Paleozoic igneous intrusions—postdating the host rocks by over a billion years—is still unclear, with some workers arguing for a connection to the Southern Oklahoma aulacogen far to the east (Larson et al., 1985), while others propose a N-S–trending aulacogen setting (McMillan and McLemore, 2004). The McClure Mountain igneous complex is a relatively large, well-studied example of these early Paleozoic intrusions (Fig. 1).

We present here an integrated geochronologic and paleomagnetic investigation of dikes surrounding the McClure Mountain igneous complex in southern Colorado to better understand the temporal evolution of the complex. Previous paleomagnetic studies (Larson and Mutschler, 1971; French et al., 1977; Lynnes and Van der Voo, 1984) recovered multiple paleomagnetic directions from dikes in and around the complex but lacked reliable geochronologic constraints. Additionally, the magnetizations have not been shown to be primary. The crystallization age of the main syenite body of the McClure Mountain igneous complex is well dated, and detailed examination of its thermochronologic evolution at medium to low temperatures has been conducted (Schoene and Bowring, 2006; Samson and Alexander, 1987; Renne et al., 1998; Spell and McDougall, 2003; Anderson et al., 2017, 2018; Weisberg et al., 2018a, 2018b). However, the temporal progression of the end-stage McClure Mountain igneous complex dike intrusions is poorly known. Defining a temporal framework for the McClure Mountain igneous complex will allow us to better ascertain the regional-scale magmatic and tectonic relationships of the southern Colorado intrusive episodes.

SETTING

Regional Geology

Proterozoic basement rocks (typically 1.70–1.36 Ga migmatitic gneisses and granites) of the Yavapai terrane exposed in the uplifted Wet Mountains of southern Colorado (Taylor et al., 1975a, 1975b; Armbrustmacher and Hedge, 1982; Bickford et al., 1989) host a number of alkaline igneous complexes of early Cambrian age: McClure Mountain, Gem Park, Democrat Creek, and Iron Mountain (Figs. 1 and 2; Olson et al., 1977; Armbustracher and Hedge, 1982; Armbustracher, 1984; Bickford et al., 1989). The Wet Mountains preserve structures from regional tectonic events such as the uplift of the Ancestral Rockies in the late Paleozoic, reactivation and deformation during Laramide time, and Rio Grande rifting beginning ca. 29 Ma (Marshak et al., 2000; Lindsey, 2010). The McClure Mountain igneous complex is the largest of the late Paleozoic igneous bodies (Taylor et al., 1975a). The complex is roughly bimodal, with a mafic-ultramafic complex emplaced in the northeast followed by a larger alkaline complex (Fig. 1; Parker and Hildebrand, 1962; Shawe and Parker, 1967). The youngest igneous activity in the McClure Mountain igneous complex is represented by dikes of varied mineralogy (i.e., lamprophyric, carbonatitic, and trachytic) intruding both the main complex and Proterozoic host rocks to the west (Heinrich and Dahlem, 1969; Alexander, 1981). The area is cut by a number of faults, including the large Texas Creek fault to the west of the complex and the large Ilse fault to the east (Taylor, 1975). Weisberg et al. (2018a, 2018b) suggested that Phanerozoic normal displacement on the Ilse fault was minimal (see also Kelley and Chapin, 2004), although Noblett et al. (1987) argued for at least 16 km of strike-slip displacement. The larger estimate of strike-slip deformation, if coupled to offsets along the numerous smaller faults, might indicate significant structural rotations through the area (Taylor et al., 1975a, 1975b). The Texas Creek fault also shows little evidence for major displacement (French et al., 1977), although detailed examination of the Phanerozoic history of the fault is lacking. The numerous minor faults throughout the region exhibit minimal offsets, although they may serve as conduits for fluid flow (French et al., 1977; Heinrich and Dahlem, 1969). Our field observations of dikes along the Texas Creek fault (State Highway 65; Fig. 2) confirmed that the dikes are highly altered and sheared along their margins. Dikes exposed distal to faults are less altered.

Previous Paleomagnetic Studies

Two previous paleomagnetic studies have been conducted on intrusive rocks throughout southern Colorado (Larson and Mutschler, 1971; Larson et al., 1985), and two additional studies specifically targeted igneous complexes in the Wet Mountains (French et al., 1977; Lynnes and Van der Voo, 1984).

The first paleomagnetic study on intrusive rocks of southern Colorado was conducted by Larson and Mutschler (1971). That study sampled three localities: dikes in the Black Canyon of the Gunnison River, a gabbroic dike in the Powderhorn alkalic complex, and major rock types within the McClure Mountain igneous complex. The exact sampling locations were not documented in that study. Samples in that study were subjected to limited alternating field (AF) and thermal demagnetization and yielded a south-southeast, shallowly inclined paleomagnetic direction. Larson and Mutschler (1971) calculated separate poles for the means of the AF-treated and thermally treated samples. Unfortunately, the authors did not provide site or sample directional data in their publication. The two virtual geomagnetic poles (VGPs) are not significantly different (AF: 37°N, 122°E, A95 = 8°; thermal: 44°N, 100°E, A95 = 8°), and we therefore combined the data to compute a new paleomagnetic pole at 41°N, 112°E. The same workers examined the paleomagnetism of Black Canyon dikes in an attempt to link the southern Colorado intrusions to the Southern Oklahoma aulacogen (Larson et al., 1985). They reported a mean declination (D) = 157° and inclination (I) = +19° from nine dikes. The paleomagnetic pole calculated from the mean direction falls at 37°N, 102°E and is consistent with their results from the McClure Mountain igneous complex.

Two papers focused exclusively on the paleomagnetism of the McClure Mountain igneous complex (French et al., 1977; Lynnes and Van der Voo, 1984). French et al. (1977) sampled dikes around the complex (29 sites) along with the ultramafic-mafic part of the McClure Mountain igneous complex. Their detailed demagnetization experiments revealed two groupings of directions. The first, poorly clustered grouping combines 21 sites (I and II directions) with a mean D = 104°, I = –2° (a95 = 11°, k = 9). French et al. (1977) suggested the poor grouping resulted from temporal differences between the dikes. This magnetization appears to be carried by magnetite and yields a pole at 12°N, 156°E (A95 = 12°).

The second grouping (III directions) in that study is coincident with the results of the studies by Larson and Mutschler (1971) and Larson et al. (1985). The mean direction from seven sites is D = 159°, I = –2° (a95 = 16°, k = 15). In their discussion of this result, the workers noted that most sites with this characteristic magnetization were taken meters from the Texas Creek fault (French et al., 1977). Lynnes and Van der Voo (1984) targeted trachytic dikes along with the main syenite bodies of the complex. Although Lynnes and Van der Voo (1984) were unable to isolate stable vector end points, their great-circle analysis yielded a mean direction with D = 152°, I = +2°; that is nearly identical to the group III directions of French et al. (1977), as well as the studies of Larson and Mutschler (1971) and Larson et al. (1985). A brick-red color in the trachytes, high blocking temperature, and thin-section analysis suggested hematite as the carrier of this remanence. Lynnes and Van der Voo (1984) interpreted the direction to reflect a fluid-triggered Pennsylvanian-age remagnetization.

METHODS

Geochronologic Methods

Geochronologic hand samples were taken from two extracomplex dikes: site 1 (trachyte dike), and site 8 (lamprophyric dike; Fig. 2). Standard mineral separation techniques (crushing, sieving, heavy liquids, and magnetic methods) were used to isolate zircons from the dike samples. Samples were handpicked for zircon grains under a binocular microscope and then transferred to double-sided tape in preparation for mounting. The grains were set in epoxy and then polished to expose interiors. Zircons were examined on an scanning electron microscope (SEM) with backscattered electron (BSE) and cathodoluminescence (CL) imaging. Epoxy plugs were then washed and sonicated in an HNO3 solution prior to laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) analysis.

U-Pb isotopic analyses were carried out using the Nu-Plasma multicollector plasma source mass spectrometer at the Department of Geological Sciences at the University of Florida. The LA-ICP-MS is equipped with a custom-designed collector block for simultaneous acquisition of 204Pb (204Hg), 206Pb, and 207Pb signals on the ion detectors and 235U and 238U on the Faraday detectors (Mueller et al., 2008). Prepared zircon grains were laser ablated using a New-Wave 213 nm ultraviolet laser beam. During U-Pb analyses, the sample was decrepitated in a He stream and then mixed with Ar gas for induction into the mass spectrometer. Background measurements were performed before each analysis for blank correction and to determine contributions from 204Hg. Each sample was ablated for ∼30 s in an effort to minimize pit depth and fractionation following standard practice for zircon analyses at the University of Florida. Data calibration and drift corrections were conducted using the FC-1 Duluth Gabbro zircon standard. Data reduction and correction were conducted using in-house software and Python.

Paleomagnetic Methods

Independently oriented core samples (72) and hand samples (12) were collected from 12 dikes and other intrusions around the McClure Mountains igneous complex in the Wet Mountains area in southern Colorado (Fig. 2). Paleomagnetic core samples were collected in the field with a water-cooled, gasoline-powered drill and oriented with magnetic and sun compass. Deep erosion of dikes as compared with host rocks (Heinrich and Dalhem, 1969) rendered drilling infeasible at four sites; oriented hand samples were taken at these sites. All samples were returned to the University of Florida, where they were trimmed to a standard size. Long core samples yielded multiple specimens. Natural remanent magnetization (NRM) directions were collected using either a Molspin spinner magnetometer or 2G-77R cryogenic magnetometer. Pilot samples from all sites were demagnetized by thermal and alternating methods using either an ASC TD-48 thermal demagnetizer or DTech 2000 AF demagnetizer. Subsequent demagnetization treatment was optimized based on the results from the pilot samples. Paleomagnetic vector directions were determined via principal component analysis (Kirschvink, 1980) using IADP software (Torsvik et al., 2016). Additional analysis was carried out using PmagPy (Tauxe et al., 2016). Powdered/crushed material from several samples at each site was analyzed with a KLY-3S Kappabridge with a CS-3 furnace attachment and a vibrating sample magnetometer (VSM) in order to ascertain magnetic carriers and magnetic domain characteristics.

RESULTS

Geochronologic Results

Ordovician Trachyte

The latest stage of intrusion at the McClure Mountain igneous complex is a series of prominent red trachytic dikes. These dikes were dated (four-point Rb-Sr isochron) by Olson et al. (1977) as Cambrian–Ordovician at 495 ± 10 Ma. K-Ar whole-rock dates from Lynnes and Van der Voo (1984) were scattered, with the youngest Pennsylvanian age (ca. 300 Ma) interpreted by the authors as evidence of fluid alteration at this time.

A hand sample from the paleomagnetically sampled trachyte dike in this study yielded over 50 zircons. The paleomagnetic data from this dike are discussed below. Over 80 analyses were conducted on the zircons from the trachyte dike (GSA Data Repository Table DR11). To confidently calculate the crystallization age of this dike, we rejected any analysis that was more than 3% discordant (calculated as % discordance between 207Pb/235U and 206Pb/238U ages). This reduced the number of analyses to 18, on 13 zircons (Table 1). One zircon from the trachyte dike did have an older date comparable to the main complex body (Fig. 3A), was likely inherited, and was also removed from the calculation of the dike crystallization age. The weighted mean age (206Pb/238U, 2σ uncertainty) of the 17 analyses under 3% discordance is 483.0 ± 1.8 Ma (Figs. 3A and 3B). We regard this result as a more reliable update to the older Rb-Sr and K-Ar ages, fixing the age of intrusion of this dike as Tremadocian. Notably, this dike postdates the main body of the complex (524 Ma; Schoene and Bowring, 2006) by ∼40 m.y. Our result is consistent with the interpretation of Olson et al. (1977), who viewed the red trachyte dikes as a late intrusive stage of the McClure Mountain igneous complex.

Cambrian Lamprophyre

Lamprophyric dikes are another constituent of the late-stage McClure Mountain igneous complex magmatism. These dikes cut all major units of the complex (Heinrich and Dalhem, 1969; Alexander, 1981). Their relationship with other late-stage dikes (i.e., the red trachytes) is unclear based solely on field relationships.

A hand sample from a paleomagnetically sampled extracomplex lamprophyre dike yielded three zircons suitable for analysis. One of these zircons was inherited, with a Paleoproterozoic age (ca. 2430 Ma). The other two zircons yielded concordant, Cambrian dates (Table 1). The weighted mean age (206Pb/238U, 2σ uncertainty) of the two analyses is 525.6 ± 7.9 Ma (Fig. 3A). Although this date is based on only two zircons, it is comparable to the reliable McClure Mountain igneous complex syenite age of Schoene and Bowring (2006). We consider the age of intrusion for this dike as on the “young-uncertainty” side of our analyses (i.e., probably younger than the McClure Mountain igneous complex syenite based on crosscutting relationships elsewhere), but we allow for the possibility that lamprophyric intrusion was coeval with as well as older than the main McClure Mountain syenite intrusion.

Paleomagnetic Results

Ten of 13 sites sampled for paleomagnetic analysis yielded suitably consistent (i.e., stable and interpretable) paleomagnetic directions. Our paleomagnetic results confirm earlier data from the area (Figs. 4A and 5; Larson and Mutschler, 1971; French et al., 1977; Lynnes and Van der Voo, 1984). The main magnetic carrier is magnetite (Fig. 4C), with some exceptions discussed below. Baked contact tests were conducted to appraise the magnetic stability of paleomagnetic data (Fig. 6). Ubiquitous steep inclination magnetic directions (also noted by French et al., 1977) suggest Cenozoic to recent remagnetization around the complex (Fig. 7).

Southeast-Directed Paleomagnetic Results

Five sites in our study, including both sites with geochronologic control (see above), yielded southeast and shallow to medium paleomagnetic high-temperature/coercivity directions (Tables 2 and 3). The dikes show a consistent steep-positive (down)–directed low laboratory unblocking temperature overprint with a mean D = 52°, I = +78° (a95 = 5, k = 14; n = 68). In sites such as C161, the overprint directions fall on a path between recent field directions and the primary magnetic directions seen at the site. Other sites have more scattered overprints (i.e., site C164), but the majority of steep down-inclination overprint components are very consistent with characteristic steep down-inclination components at other sites. In order to best compare our data with those of previous studies, we separated the data into groups based on declinations (Fig. 5A; after French et al., 1977).

Two dikes (C161 and C168), both sites with geochronologic control (see above), yielded east-southeast and shallow directions with high laboratory unblocking temperatures (Table 2). The dikes show a steep-down–directed overprint with low laboratory unblocking temperatures yielding a mean D = 65°, I = +67° (a95 = 6°, k = 18; n = 29, where n is the number of specimens used in the mean calculation; Figs. 4B and 5B). The mean high laboratory unblocking temperature direction from these two dikes combined with previous data yields a mean D = 106°, I = –1° (a95 = 10°; k = 9; N = 23, where N is the number of sites used in the mean calculation). The very poor grouping of these data may reflect a variety of complications. The most obvious explanation for this disparity is the ∼40 m.y. time span separating the emplacement of the several different dikes. Due to this age disparity, we see no reason to compute a mean pole by combining our data with the group I and II directions isolated by French et al. (1977).

A baked contact test at site C168, a lamprophyric dike (dated at 525 Ma; this study) emplaced into Proterozoic granitic rocks west of the main body of the complex, is problematic (Fig. 6A). Unbaked host-rock samples yield data identical to those at the dike-granite contact and exhibit a mean D = 69°, I = +65° (a95 = 7°, k = 37; n = 12). Samples from the dike are also heavily overprinted (some completely) and yield the same steep direction; however, most dike samples showed a distinct, high laboratory unblocking temperature direction of D = 129°, I = +21° (a95 = 9°, k = 45; n = 7). Low-temperature/coercivity magnetizations overlap with recent geomagnetic field directions for southern Colorado (reference direction is D = 8°, I = +65°; International Geomagnetic Reference Field 2012 [Thebault et al., 2012]). The granite and medium laboratory unblocking temperature dike magnetic components lie along a great circle between the recent field overprint and the high-temperature/coercivity direction of the dike (Fig. 6A). The baked contact test here is therefore negative.

Three sites (2 dikes; sites C164, C165, and C1611) yielded south-southeast, shallow to moderate negative inclination paleomagnetic directions (Table 3). These characteristic directions are overprinted by a steep, positive inclination direction with a mean D = 14°, I = +84° (a95 = 6°, k = 16; n = 39; Fig. 3B). The mean direction from our sites is D = 153°, I = –23° (a95 = 33°, k = 15; N = 3). This mean direction is similar to the group III directions of French et al. (1977) and the trachyte grouping of Lynnes and Van der Voo (1984). When our results are combined with those of French et al. (1977), they yield a mean direction from 10 sites in and around the complex of D = 150°, I = –5° (a95 = 8°, k = 19), with a corresponding paleomagnetic pole at 51°S, 292°E. Because only mean directions were reported, we are unable to incorporate site mean data from Lynnes and Van der Voo (1984) or Larson and Mutschler (1971). We do note, however, that the VGP reported for red trachyte dikes (43°S, 294°E; Lynnes and Van der Voo, 1984), the alkaline complex VGP (41°S, 312°E; Larson and Mutschler, 1971), and the Black Canyon dikes pole (37°S, 282°E; Larson et al., 1985) closely resemble our paleomagnetic pole.

A single dike from this grouping (group III; French et al., 1977) yielded a reverse polarity direction (D = 338°, I = +8°; Fig. 4A). The dikes that exhibit this direction were all sampled along the Texas Creek fault zone. The paleomagnetic pole calculated from our mean direction falls at 51°S, 292°E (A95 = 10°, k = 24; N = 10). This pole coincides with the Laurentian apparent polar wander path (APWP; Torsvik et al., 2012) at ca. 300 Ma (Fig. 8) and is consistent with a K-Ar whole-rock age on a trachyte from Lynnes and Van der Voo (1984) of 300 ± 11 Ma and a less precise fission-track date from Olson et al. (1977) at 293 ± 62 Ma. Lynnes and Van der Voo (1984) argued that this remagnetization was acquired during a late Paleozoic hydrothermal event. Low-temperature thermochronologic studies on the McClure Mountain igneous complex syenite (Anderson et al., 2017; Weisberg et al., 2018a, 2018b) support the assertion that the McClure Mountain igneous complex syenite remained at low temperatures (<180 °C) during the late Paleozoic. The late Paleozoic alteration is also confirmed by regional-scale paleomagnetic studies (Geissman and Harlan, 2002). Geissman and Harlan (2002) discovered that the late Paleozoic remagnetization is carried by hematite and was acquired primarily via fluid remagnetization at low (<200 °C) temperatures, which is consistent with our observations and those of Lynnes and Van der Voo (1984).

A baked contact test at site C1611 was conducted on a small (0.75-m-wide) dike adjacent to the Texas Creek fault zone, several hundred meters north of sites C164 and C1611. All of these dikes exhibited variable alteration, quartz/calcite veining, and shear along their margins. The dike (C1611) yielded a high-temperature/coercivity magnetization with a mean D = 55°, I = +76° (a95 = 7°, k = 43; n = 10). Low-temperature/coercivity overprints yielded a mean direction of D = 357°, I = +75° (a95 = 4°, k = 128; n = 14). Granite samples taken at the contact with the dike carry a south-southeast, shallow inclination magnetization held in hematite with a mean D = 158°, I = 0° (a95 = 6°, k = 59; n = 11; Fig. 6C). These data illustrate a negative baked contact test at site C1611, but the direction isolated in the contact granite samples is consistent with the fluid remagnetization model discussed above.

Steep Paleomagnetic Data

Six sites (6 dikes) carry a dual-polarity, steep paleomagnetic direction as their characteristic magnetization component (Table 4; Fig. 7). This component is similar to the overprints on dikes that carry shallow, southeast paleomagnetic directions (Fig. 4C). These sites carry less prominent secondary magnetizations (Fig. 7), which fall close to Earth’s present-day field direction. In particular, site C169 was defined based on great-circle analysis, because of a prominent north, down-inclination overprint, whereas the great-circle intersection point was south-directed, steep-up inclination.

A baked contact test was carried out at site C1610 (Fig. 6B). This site is in a network of dike and sill gabbroic intrusions sampled near Canon City, Colorado (sites C162, C163, and C1610). Sites C162 and C163 display univectorial, steep positive inclination magnetizations. Results from C1610 are more complex. A well-exposed contact with host Proterozoic granite made a baked contact test possible. The mafic intrusion samples produce a southeast, steep to intermediate inclination magnetization, D = 145°, I = –58° (a95 = 6, k = 96; n = 6). However, close to the contact (i.e., samples 5–7), demagnetization data progress through this negative inclination along a great circle toward a steep, positive inclination at high (>580 °C) temperatures. A sample of the dike sample from directly at the contact yielded unstable results. Samples from the granite have uncomplicated, steep-down inclination magnetizations, with a mean of D = 288°, I = +88° (a95 = 13, k = 55; n = 4). All rocks (both intrusion and host) have low-temperature/low-coercivity remanence components with a mean direction of D = 8°, I = +70° (a95 = 11, k = 16; n = 11). The area around this site was affected by Cenozoic-age folding and faulting (Taylor et al., 1975b).

The mean of the north-directed, steep inclination remanence (with two sites inverted) is D = 10°, I = +73° (a95 = 13°, k = 29; N = 6). Recent magnetic field directions have slightly shallower inclinations (I = +58° GAD [Geocentric Axial Dipole] assumption, I = +65° IGRF12).

DISCUSSION

Deformation and Magnetization

A striking feature of the paleomagnetic data from early Paleozoic intrusions around the McClure Mountain igneous complex (French et al., 1977; Lynnes and Van der Voo, 1984; this study) is the directional scatter. Based on structural features, paleomagnetic data, and geochronologic data, we argue that a combination of factors, including faulting, fluid flow, and age disparities, is responsible for the large spread in directions. Large-magnitude displacements on the mapped faults in the immediate vicinity of the McClure Mountain igneous complex are unlikely, although localized deformation is possible. The Ilse fault, the trace of which is located ∼6 km east of the McClure Mountain igneous complex, accommodated ∼16 km of right-lateral strike-slip motion during Laramide deformation (Noblett et al., 1987). The Laramide-age deformation may also have generated coeval vertical-axis rotations along the intervening smaller-scale faults. Fault orientations, especially in the smaller faults, are highly variable, and the estimated offsets are relatively unconstrained. This variability suggests that any rotations are less likely to be systematic in the region, resulting in differential rotations throughout the study area. Based on the scattered paleomagnetic directions along with more recent assessments of fault movement (e.g., Noblett et al., 1987), we revise the conclusion of French et al. (1977) of “no structural disturbance” to “potentially considerable structural disturbance.” There appears to be no correlation between dike trend and declination that might hint at some systematic rotation.

Additionally, our new geochronologic data show that certain dikes (trachytes) postdate the main complex by 40 m.y. If magmatic activity was semicontinuous between 525 and 480 Ma, the paleomagnetic data may represent “snapshots” of the geomagnetic field over a 40 m.y. period, with the scatter generated as a result of continental motion. Given the difficulties associated with dating many fine-grained dikes individually, addressing this scenario is difficult. Revisiting the age constraints on other igneous centers in the area such as Gem Park and/or Democrat Creek may shed light on whether or not magmatism in the area occurred semicontinuously or in discrete phases.

Group III dikes of French et al. (1977) along with the data from trachyte dikes of Lynnes and Van der Voo (1984) were interpreted by the original authors and others (Geissman and Harlan, 2002) to be indicative of a late Paleozoic remagnetization event associated with fluid migration during the formation of the Ancestral Rockies. This remagnetization is prominent along the Texas Creek fault zone—a likely conduit for fluid movement at that time.

There is a key question pertinent to the present study, given the structural disturbance, fluid remagnetization, and scattered paleomagnetic directions: Is it even possible to separate these paleomagnetic data from early Paleozoic intrusions into different remanence signatures? That is, might the entire southeast, shallow-inclination data set be a product of differential rotations and various levels of remagnetization of the analyzed dikes? There is a natural “break” in declination between directional groupings (Fig. 5A). This “break” separates paleomagnetic data into east-southeast–directed versus south-southeast–directed groups. Additionally, the south-southeast declination paleomagnetic data originate from sites taken along the Texas Creek fault zone (French et al., 1977; this study). The data polarity values are consistent with directional separation of the paleomagnetic data. The first group of data (with geochronological ages ranging from the Cambrian to Ordovician) includes multiple sites with dual polarities, whereas the sites with late Paleozoic magnetizations are dominantly southwardly directed and shallow (with one exception), consistent with remanence acquisition during the Kiaman superchron (318–265 Ma; base from Opdyke et al., 2014; top from Belica et al., 2017). A comparison with the North American apparent polar wander path (NAM APWP; Torsvik et al., 2012) further reinforces the data break, with the mean of the east-southeast declination data falling along the late Cambrian–Ordovician segment of the path, while the south-southeast and shallow remanence falls, as expected, around the late Paleozoic part of the apparent polar wander path (Fig. 8). The paleomagnetic data are consistent with a protracted history of early magnetization acquisition, remagnetization, and deformation. Our overall assessment of these data is that, although they are broadly consistent with the NAM APWP, the likely impact of deformation and the large range of interpreted ages obtained from igneous rocks render them unsuitable for direct incorporation into the NAM APWP.

The steep directions recorded in both high-temperature/coercivity components and overprints agree well with expected Cenozoic to recent magnetic field directions. We therefore conclude that these directions were acquired during regional Cenozoic-aged igneous activity (Baldridge et al., 1991) or as viscous overprints acquired during the Neogene to present day.

Far-Field Tectonic Implications

The McClure Mountain igneous complex and other Cambrian igneous complexes have been temporally linked with the Southern Oklahoma aulacogen (Loring and Armstrong, 1980; Larson et al., 1985; Bickford et al., 1989; Hanson et al., 2013; Brueseke et al., 2016; Weisberg et al., 2018a), although such associations are controversial (Van Der Voo, 1986; Loring et al., 1987; McLemore, 1987; Purucker, 1988; McMillan and McLemore, 2004).

Some recent papers on the Southern Oklahoma aulacogen (Hanson et al., 2013; Brueseke et al., 2016) have mentioned the temporal correlation of central/southwestern diabase and alkali intrusions with the Southern Oklahoma aulacogen. Precise age constraints on the bimodal igneous rocks of the Southern Oklahoma aulacogen are also limited, but gabbros, granites, and rhyolites of the Wichita and Arbuckle Mountains give early Cambrian (539–530 Ma) ages (Wright et al., 1996; Hames et al., 1998; Hanson et al., 2009; Wall et al., 2018). Geophysical data provide support for the connection (Keller and Stephenson, 2007), with discontinuous gravity anomalies traceable from the Southern Oklahoma aulacogen into Utah. The southern Colorado–New Mexico igneous rocks are geochemically indicative of an extensional environment or mantle plume (McMillan and McLemore, 2004), similar to the Wichita igneous province suite (Hanson et al., 2013). However, the most reliable ages for Southern Colorado–New Mexico igneous rocks are appreciably younger than those of the Southern Oklahoma aulacogen (539–530 Ma): The main syenite at McClure Mountain was dated to 523.98 ± 0.12 Ma by Schoene and Bowring (2006), the Florida Mountains were dated at 503 ± 10 Ma (Clemons, 1987), and trachyte dikes of the McClure Mountain igneous complex were dated at 483 ± 2 Ma (this study). Dates obtained with the K-Ar system are more scattered, but generally fall within the same age range (McMillan and McLemore, 2004).

These data suggest that either rift-related magmatism along the lengthy aulacogen proposed by Larson et al. (1985) was severely protracted (over 40 m.y.), extending well into passive-margin development in the western United States (Dalziel, 1991; Yonkee et al., 2014), or the tectonic setting was more complex. McMillan and McLemore (2004) proposed a “New Mexico aulacogen,” analogous to the Southern Oklahoma aulacogen but differently oriented. Mantle plume impingement along the Southern Oklahoma aulacogen has been proposed (Hanson et al., 2013; Brueseke et al., 2016), which may help to explain the age disparity between the two regions.

A better definition of the upper and lower age limits of the early Paleozoic intrusions in southern Colorado and New Mexico, as well as further investigation of the Southern Oklahoma aulacogen Wichita igneous province are necessary to fully explore the tectonic setting of central-southwestern North America in the late Paleozoic. Given the current relative age disparity (exacerbated by the Early Ordovician dike age obtained in our present study), a direct relationship between the Cambrian complexes of southern Colorado and the Southern Oklahoma aulacogen remains unclear. Many putatively Paleozoic intrusions in the U.S. Southwest lack reliable age constraints (McMillan and McLemore, 2004). Obtaining accurate ages for these bodies is a challenge for the future, with the promise of untangling the apparently complex tectonic regime in the area during the transition to a passive margin.

CONCLUSIONS

Dikes around the McClure Mountain igneous complex record at least three different paleomagnetic directions. We tentatively interpret these as earliest Paleozoic (Cambrian–Ordovician), late Paleozoic (Permian–Carboniferous), and Cenozoic in age. Due to the strong likelihood of local deformation related to faults in the region, none of the paleomagnetic poles based on those data is suitable as reference results for the NAM APWP. An Early Ordovician U-Pb age of 483.0 ± 1.8 Ma on a trachytic dike near the McClure Mountain igneous complex suggests a protracted interval of magmatism in the area after emplacement of the main syenite intrusion at 524 Ma. At present, given the prolonged magmatic activity we have shown in the McClure Mountain igneous complex, we speculate that magmatism in southern Colorado was related to a different tectonic setting than that of southern Oklahoma in the early Paleozoic.

ACKNOWLEDGMENTS

Support of field work for this project was provided by grants to Pivarunas from the Colorado Scientific Society and University of Florida Department of Geological Sciences. Pivarunas was further supported by a University of Florida Graduate Student Fellowship. Rob Van der Voo encouraged the coauthors to re-explore the McClure Mountain igneous complex and provided a variety of materials that were helpful in the field. Pivarunas would also like to thank his coauthor for help in the challenging field environment of southern Colorado. We thank John Geissman and Daniel Pastor-Galán for constructive reviews that significantly improved this paper.

1GSA Data Repository Item 2019280, geochronologic data and paleomagnetic sampling site and analytical information, is available at http://www.geosociety.org/datarepository/2019, or on request from editing@geosociety.org.
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.