Detrital zircon U-Pb studies of mudstone provenance are rare but may preferentially fingerprint distal zircon sources. To examine this issue, Pierre Shale and Trinidad Sandstone deposited in a Late Cretaceous deltaic environment in the Raton Basin, Colorado (USA), were measured for detrital zircon U-Pb age by laser ablation–inductively coupled plasma–mass spectrometry. Two major detrital zircon age peaks at ca. 70 and 1690 Ma are found in both Pierre Shale and Trinidad Sandstone but in inversely varying proportions: 68% and 16%, respectively, for the finest zircon fraction (~15–35 μm) in the shale, and 25% and 32%, respectively, for the coarsest zircon fraction (~60–80 μm) in the sandstone. Proximal sources in the Sangre de Cristo Mountains, directly west of the Raton Basin, contain coarse-grained, ca. 1690 Ma zircon, whereas distal sources in Laramide uplifts and basins in Colorado, New Mexico, and Arizona contain fine-grained, ca. 70 Ma zircon. The results indicate that U-Pb zircon provenance of mudstone reflects availability of volcanic and other fine-grained source rocks rather than simply distal sources. U-Pb zircon provenance studies should routinely include mudstone units because these units may identify fine-grained zircon sources more reliably than sandstones alone.

Uranium-lead (U-Pb) detrital zircon geochronology of clastic sedimentary rocks is a critical tool for sedimentary provenance, definition of paleo-drainage routes, and reconstructions of paleogeography through time (Košler et al., 2002). Most studies have focused on sandstones and quartzites, which are typically enriched in sand-sized zircon grains separated routinely from less-dense minerals by heavy-liquid concentration, mounted in epoxy, and dated in moderate to large numbers (n = 100–1000) by laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) or secondary ionization mass spectrometry (SIMS).

In contrast, mudstone and its metamorphic equivalents, which make up two-thirds of the sedimentary rock record (Schieber et al., 1998), have been dated only rarely by U-Pb detrital geochronology (e.g., Sláma and Pedersen, 2015; Leary et al., 2020). This is because mudstone commonly contains silt-sized detrital zircon grains (Totten and Hanan, 2007) that can be difficult to separate from clay mineral–rich rock matrices using heavy liquids (Hoke et al., 2014). Also, such tiny grains require very small (≤20 μm) laser or ion beam spots for U-Pb analysis by LA-ICP-MS or SIMS and, for LA-ICP-MS, short signal integration times (e.g., 10–15 s of ablation at 5 Hz; Mukherjee et al., 2019) to limit variable sample-to-sample Pb/U downhole fractionation.

More fundamentally, the significance and value of U-Pb detrital zircon geochronology of mudstone relative to sandstone from the same deposit is not clear. Sláma and Pedersen (2015) reported that U-Pb ages for fine-grained detrital zircon from metamorphosed siltstone and mudstone of the Caledonides in southern Norway were derived from far-traveled (~2000 km) distal sources and thus complement the record of coarser detrital zircon from associated sandstone, which reflects the U-Pb zircon age profile of local sources. However, Leary et al. (2020) found that only some mudstone units in the late Paleozoic Paradox and Eagle Basins, southwestern USA, contained substantial populations of fine-grained zircon derived by long-distance transport from the Grenville and Appalachian orogens.

To examine controls on distal and proximal zircon sampling of mudstone and sandstone, we compared the U-Pb detrital zircon geochronology of Pierre Shale and facies-equivalent Trinidad Sandstone in the Raton Basin, south-central Colorado, USA (Fig. 1). A Late Cretaceous fluvial system delivered clastic detritus predominantly from the west, forming the units in a deltaic facies environment (Billingsley, 1977; Cather, 2004). Contribution of Laramide-arc detritus by ash fall is also possible (Bush et al., 2016; Schwartz et al., 2021).

The study area is a railroad cut near Trinidad Lake reservoir, 3.7 km southwest of Trinidad, Colorado (37°08′40.1″N, 104°32′21.5″W) (Fig. 1), exposing the upper prodelta facies of Pierre Shale and delta-front facies of Trinidad Sandstone (Flores, 1987). A sample of each of the uppermost Pierre Shale and overlying lowermost Trinidad Sandstone was collected from a 3 m vertical section of interbedded mudstone and sandstone. The top of Pierre Shale in the Trinidad area corresponds with the lower Maastrichtian Baculites clinolobatus ammonite zone (69.59 ± 0.36 Ma) (Berry, 2018).

The Trinidad Sandstone sample is a fine-grained subarkosic sandstone composed of very fine sand (63–125 μm) and coarse silt (20–63 μm) grains, whereas the Pierre Shale sample is a silty mudstone made up by coarse silt and finer grains. Three grain mounts were made from heavy-liquid concentrates of coarser (63–125 μm) and finer (20–63 μm) sieve fractions of crushed samples of Trinidad Sandstone, referred to as TSC (C—coarse) and TSF (F—fine), respectively, and the 20–63 μm fraction of Pierre Shale, referred to as PSM (M—grain mount). Also, three polished thin sections (TS) were made from mudstone-rich layers in Pierre Shale, with U-Pb zircon analyses from all three sections pooled to a single sample referred to as PSTS. The pooled thin-section sample was used to date zircons associated with the finest fraction of shale and compare in situ shapes of zircon with those in the grain mounts. Mud-rich layers are not present in Trinidad Sandstone thin sections.

Scanning electron microscopy (SEM) with backscattered electron (BSE) imaging guided spot placement within zircon grains for laser ablation (using 20 μm and 12 μm spots in grain mounts and thin sections, respectively) and provided measurements of grain dimensions and definition of grain morphologies. Experimental procedures and metadata for SEM imaging and LA-ICP-MS U-Pb zircon ages are given in Table S1 of the Supplemental Material1.

LA-ICP-MS U-Pb age results for sample and reference zircons are listed in Tables S2–S7 and plotted on concordia diagrams in Figures S2–S4 in the Supplemental Material. Typical precision on individual 206Pb/238U dates for both 20 μm and 12 μm spots for Paleozoic zircons (Plešovice and Temora-2) is ~2%. U-Pb detrital zircon age histogram–kernel density estimation plots for Pierre Shale (PSTS, PSM) and Trinidad Sandstone (TSF, TSC) samples are shown in Figure 2A. Two major age peaks at ca. 70 Ma and 1690 Ma are found in all four samples but in inversely varying proportions (Fig. 2B). The ca. 70 Ma peak decreases progressively in abundance from 68% and 43% for Pierre Shale samples PSTS and PSM, respectively, to 37% and 25% for Trinidad Sandstone samples TSF and TSC, respectively. In contrast, the ca. 1690 Ma peak increases progressively in abundance from 16% and 25% for PSTS and PSM, respectively, to 25% and 32% for TSF and TSC, respectively. An older ca. 1740 Ma peak is present in Trinidad Sandstone (16% for TSF and 14% for TSC) but absent from Pierre Shale (Fig. 2B). Two other significant but subordinate peaks are found in all four samples but without systematic differences between Pierre and Trinidad varieties: ca. 160 Ma peak (6%–14%) and ca. 1410 Ma peak (2%–8%).

Four types of zircon grain morphologies (Fig. 3) are present in Pierre Shale and Trinidad Sandstone samples: complete, euhedral grain sections with discrete oscillatory zoning (type 1); complete or nearly complete, subhedral grain sections, commonly fractured, with diffuse oscillatory to patchy zoning (type 2); grain sections with irregular, embayed crystal faces (type 3); and crystal fragments (type 4). Additional BSE images of various types of zircons in the two dominant ca. 70 Ma and 1690 Ma age populations are given in Figure S1.

Type 2 and 4 zircons are dominant in all samples. At least for the Pierre Shale thin sections, fracturing in type 2 and grain breakage in type 4 zircons are not the result of laboratory crushing. Textures and zoning of type 1, 2, and 3 grains indicate derivation from magmatic source rocks. Euhedral shapes of type 1 grains imply early crystallization free from competing mineral growth (e.g., in rapidly quenched ash-fall tuffs). Embayed textures of type 3 grains suggest late crystallization interstitial to adjacent minerals. Type 1 and 3 grains are absent from the ca. 1690 Ma zircon age population.

The Th/U ratios of 71 ± 3 Ma zircon tend to be greater in Pierre Shale compared to Trinidad Sandstone. Mean ratios for PSTS and PSM are 1.01 ± 0.08 (n = 21) and 0.95 ± 0.13 (n = 13), respectively, with 12/34 or 35% with Th/U ≥1.2. In contrast, mean values for TSF and TSC are 0.66 ± 0.10 (n = 14) and 0.71 ± 0.08 (n = 12), respectively, with only 1/26 or 4% with Th/U ≥1.2 (Fig. 4A). Zircon with Th/U ≥1 is more typical of crystallization in basic to intermediate than acidic magmatic rocks (Wang et al., 2011; Kirkland et al., 2015). Th/U ratios of 1690 ± 15 Ma zircon tend to be low for both Pierre and Trinidad samples: 0.65 ± 0.12 (n = 14) for PSTS; 0.63 ± 0.08 (n = 22) for PSM; 0.52 ± 0.06 (n = 23) for TSF; and 0.70 ± 0.05 (n = 26) for TSC (Fig. S5).

Proximal sources of ca. 1740, 1690, 1410, and 160 Ma detrital zircons of the Pierre-Trinidad sedimentary system are exposed presently in the Sangre de Cristo Mountains along the western margin of the Raton Basin but in the Late Cretaceous may have been exposed farther west in the San Luis Uplift, a broad Laramide highland, exposed today as a valley (Cather, 2004) (Fig. 1). Jones and Connelly (2006) reported U-Pb zircon ages for 1695–1682 Ma post-orogenic plutons, ca. 1700 Ma quartzite, a suite of 1750–1730 Ma calc-alkaline intrusions, and 1434–1407 Ma granites in Sangre de Cristo crystalline basement (Fig. 2C). Siliciclastic cover units in the Sangre de Cristo Mountains could also be proximal zircon sources for Pierre Shale and Trinidad Sandstone. Bush et al. (2016) found that the Pennsylvanian–Permian Sangre de Cristo Formation has major detrital zircon U-Pb ages peaks at ca. 1720 and 1680 Ma, whereas the Cretaceous Dakota Formation has a major detrital zircon peak at ca. 160 Ma (Fig. 2C).

Sources of ca. 70 Ma detrital zircons of Pierre Shale–Trinidad Sandstone are found only in more distal terranes (Fig. 1) with Cretaceous volcanic rocks, hypabyssal intrusions, and sedimentary units exposed in basement-cored uplifts and basins related to Laramide arc magmatism (Seedorff et al., 2019). In southwestern New Mexico, felsic ash-fall tuffs with U-Pb zircon ages of 75–70 Ma are present in Love Ranch Basin (Amato et al., 2017) and Ringbone Basin (Clinkscales and Lawton, 2015) (Fig. 2D). A dacite sill and monzonite porphyry in the Burro Mountains have U-Pb zircon ages of ca. 75 Ma (Amato et al., 2017). Cretaceous Ringbone Formation contains a major U-Pb detrital zircon age peak at 73 Ma (Clinkscales and Lawton, 2015) (Fig. 2D). Volcanic and hypabyssal rocks with 75–70 Ma U-Pb zircon ages are present farther west in Arizona (Mizer, 2018; Seedorff et al., 2019) (Fig. 1).

In the Colorado Mineral Belt (Fig. 1), U-Pb zircon ages are 73 Ma for the diorite of Sleeping Ute Mountain, 70 Ma for the diorite of La Plata Mountains, 68 Ma for the granodiorite-diorite porphyry of Hermosa Peak, and 69–65 Ma for the granite-diorite porphyry of Coal Bank Pass (Gonzales, 2015) (Fig. 2D). In the San Juan Basin, the upper Campanian Kirtland Formation, a fluvial sandstone, and the overlying basal McDermott Formation, a trachyandesite debris flow (Wegert and Parker, 2011), have major U-Pb detrital zircon age peaks of 75 and 70 Ma, respectively (Pecha et al., 2018) (Fig. 2D).

Zircon from the trachyandesite of the McDermott Formation have unusually high Th/U ratios (as high as 3) compared to those of zircons measured in other Laramide sources of ca. 70 Ma detrital zircons such as the Kirtland Formation sandstone, Ringbone Formation, and felsic ash-fall tuffs of southwestern New Mexico, which are mostly <1 (Fig. 4B). Elevated Th/U ratios in ca. 70 Ma detrital zircon population of Pierre Shale are consistent with sampling a significant component of basic to intermediate sources of zircon like those in the McDermott Formation.

The ca. 1740, 1690, 1410, and 160 Ma zircon age peaks of the Pierre-Trinidad sedimentary system are also found in distal Proterozoic crystalline basement rocks and Cretaceous magmatic and sedimentary units in Laramide uplifts and basins (Fig. 2D) as well as in proximal sources described above. In the Burro Mountains of southwestern New Mexico, amphibolite and rapakivi granite have U-Pb zircon ages of 1684 Ma (Amato et al., 2011) and 1461 Ma (Rämö et al., 2003), respectively. Inherited grains in ash-fall tuff in the Little Hatchet Mountains of southwestern New Mexico have a mean U-Pb zircon age of 163 Ma (n = 6) (Clinkscales and Lawton, 2015). In the Needle Mountains of southwestern Colorado, U-Pb zircon ages are 1772–1754 Ma for the Twilight Gneiss and 1698–1695 Ma for the Bakers Bridge Granite (Gonzales and Van Schmus, 2007). Among Cretaceous sedimentary units, the Ringbone Formation contains U-Pb detrital zircon age peaks of 1700 and 165 Ma (Clinkscales and Lawton, 2015), Kirtland Formation has peaks of 1700, 1430, and 170 Ma, and basal McDermott Formation has peaks of 1690, 1410, and 165 Ma (Pecha et al., 2018). Thus, it is probable that both proximal and distal sources contributed to ca. 1740, 1690, 1410, and 160 Ma zircon age peaks of the Pierre-Trinidad delta.

Late Cretaceous (ca. 70 Ma) distal magmatic activity associated with the Laramide arc in southwestern New Mexico and Arizona and with the Colorado Mineral Belt was a major source of detrital zircon for both Pierre Shale and Trinidad Sandstone of the Raton Basin but relatively more so for the shale. The presence of substantial basic to intermediate rocks, which tend to be zircon poor relative to felsic sources (Keller et al., 2017), in the Laramide source region are more clearly identified by the elevated Th/U ratios of the Late Cretaceous detrital zircon population of Pierre Shale than Trinidad Sandstone samples.

Fluvial transport of distal, fine-grained detritus derived from Laramide volcanic and hypabyssal intrusive rocks, possibly in concert with Laramide ash-fall volcanism, enhanced eastward delivery of zircon silt to the Raton Basin and incorporation in Pierre Shale. Zircon grain size may have been reduced further by crystal breakage (morphology type 4; Fig. 3) during Laramide faulting of the source rocks. Grain fragmentation is more likely caused by deformational events at the weathering site prior to transport, which simply tends to abrade and chip grains to rounded shapes (Novák-Szabó et al., 2018). The zircon silt, mostly <45 μm, was capable of substantial long-distance transport in the suspended loads of fluvial (Milliman and Meade, 1983) and aerial volcanic systems (Stevenson et al., 2015). In contrast, older, mostly Proterozoic zircon sources for Pierre Shale–Trinidad Sandstone were typically deeper-level intrusions and gneisses (or sedimentary units derived from them) that crystallized a large fraction of coarser grains in both distal and proximal rocks and preferentially formed the detritus supply of the Trinidad Sandstone.

The broader significance and implications of our results are that U-Pb zircon provenance of mudstone reflects availability of volcanic and other fine-grained source rocks rather than simply distal sources. Mudstones will not record distal sources dominated by coarse-grained source rocks, as shown by the results of Leary et al. (2020). U-Pb zircon provenance studies should routinely include mudstone because it may identify zircon derived from fine-grained source rocks more clearly than associated sandstones. This extends earlier studies that reported biases in detrital zircon populations in sandstones as a function of grain size (e.g., Ibañez-Mejia et al., 2018) and facies (e.g., Anders et al., 2021). Future studies should explore controls on zircon populations in mudstone-sandstone pairs of depositional settings beyond the fluvially influenced delta studied here.

We thank Geology editor Marc Norman, reviewers David Chew and Jaime Toro, an anonymous journal reviewer, and U.S. Geological Survey reviewer Marieke Dechesne for helpful comments. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

1Supplemental Material Sample preparation, SEM and LA-ICP-MS analysis, (text and Table S1), detailed zircon BSE images (Figure S1), and U-Pb zircon data for zircon unknowns (Tables S2–S5, Figures S2 and S5) and reference materials (Tables S6 and S7, and Figures S3 and S4). Please visit https://doi.org/10.1130/GEOL.S.19176050 to access the supplemental material, and contact editing@geosociety.org with any questions.
Gold Open Access: This paper is published under the terms of the CC-BY license.