Paleosols of the Upper Paleozoic Sangre De Cristo Formation, North-Central New Mexico: Record of Early Permian Palaeoclimate in Tropical Pangaea

Home , Glorieta Sandstone, Wellington Formation, Yeso Group

Accepted Manuscript

Paleosols of the Upper Paleozoic Sangre de Cristo Formation, north-central New Mexico: Record of early Permian palaeoclimate in tropical Pangaea

Lawrence H. Tanner, Spencer G. Lucas

PII: S2095-3836(16)30076-1 DOI: 10.1016/j.jop.2017.02.001 Reference: JOP 47

To appear in: Journal of Palaeogeography

Received Date: 8 November 2016 Revised Date: 30 January 2017 Accepted Date: 4 February 2017

Please cite this article as: Tanner, L.H., Lucas, S.G., Paleosols of the Upper Paleozoic Sangre de Cristo Formation, north-central New Mexico: Record of early Permian palaeoclimate in tropical Pangaea, Journal of Palaeogeography (2017), doi: 10.1016/j.jop.2017.02.001.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT

Palaeoclimatology

Paleosols of the Upper Paleozoic Sangre de Cristo Formation, north-central New Mexico: Record of early Permian palaeoclimate in tropical Pangaea

Lawrence H. Tanner a,*, Spencer G. Lucas b a Department of Biological Sciences, Le Moyne College, Syracuse, NY 13214, USA b New Mexico Museum of Natural History, 1801 Mountain Road N.W., Albuquerque, NM 87014, USA

Abstract The lower Permian (Wolfcampian) Sangre de Cristo Formation of northern New Mexico consists of silty mudstones and laterally discontinuous sandstones deposited on an aggrading alluvial plain. Locally, mudstones display a variety of pedogenic features. Common mudstone fabrics vary from platy to prismatic; some beds display prominent pedogenic slickensides. Drab-colored root traces are common throughout the section, as are calcareous nodules, which vary from small bodies with diffuse boundaries to vertically stacked, discrete, cm-scale nodules (rhizocretions), and less commonly form coalescing horizons. Vertisols occur only in the lower portion of the ca. 90-m measured section. Most of the mudstone beds contain calcretes that are immature (calcic Protosols to calcic Argillisols), but the lower to middle portion of the section also contains mature calcrete horizons (argillic Calcisols and Calcisols). Intercalated micritic limestone beds with sharp contacts containing root traces, are of laterally variable thickness and grade to nodular calcretes. These are interpreted as floodplain pond carbonates that have undergone pedogenic alteration (palustrine limestones), indicating long periods of exposure under strongly seasonal climatic conditions. The isotopic composition of the pedogenic carbonate displays a substantial range of values, but most of the range of variation in isotopic composition is accounted for by isotopically heavier carbonate (both carbon and oxygen) precipitated in shallow ponds subject to intense pedogenic reworking (palustrine carbonate). MANUSCRIPT During the early Permian, northern New Mexico was situated in a near equatorial position ( ca. 4° N). The overall character of the paleosols suggests a persistent warm, semi-humid, seasonal climate throughout most of the interval of deposition during the Wolfcampian, but with episodically increased aridity during formation of the more mature calcretes. No long-term trend of climate change is evident in the stratigraphic section examined for this study.

Keywords Sangre de Cristo Formation, Wolfcampian, Vertisol, Argillisol, Calcisol, Protosol

© 2017 China University of Petroleum (Beijing). Production and hosting by Elsevier B.V. on behalf of China University of Petroleum (Beijing). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Received 6 November 2016; accepted 19 December 2016; available online xxx

1. Introduction

ACCEPTED * Corresponding author. Email address: [email protected] (L. H. Tanner). Peer review under responsibility of China University of Petroleum (Beijing). http://dx.doi.org/xxxxxx 2095-3836/© 2017 China University of Petroleum (Beijing). Production and hosting by Elsevier B.V. on behalf of China University of Petroleum (Beijing). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

ACCEPTED MANUSCRIPT

The significance of the sedimentology of the uppermost Carboniferous through lower Permian strata of central Pangaea stems in great part from recognition of the palaeoclimatic archives contained therein. Indeed, the sedimentary records of this interval are particularly valuable for their potential to resolve major questions of the timing and mechanisms of late Paleozoic global change, such as environmental changes coincident with the end of Gondwanan glaciation, northward migration of the Pangaean continent, the interpreted late Paleozoic aridification of the Pangean interior and potential changes in atmospheric circulation ( e.g. DiMichele et al ., 2006; Dinterman et al ., 2000; Giles et al ., 2013; Kessler et al ., 2001; Mack et al ., 2003; Montañez et al ., 2007; Mountney, 2006; Parrish, 1993; Soreghan et al ., 2002; Tabor and Montañez, 2002, 2004; Tabor et al ., 2008; Zhu and Tabor, 2014). Despite computer models that indicate palaeoclimate stability across the Pennsylvanian–Permian transition, controlled by maintenance of zonal atmospheric circulation (Kutzbach and Gallimore, 1989; Kutzbach and Ziegler, 1994), numerous authors have proposed that humid to subhumid- seasonal climate conditions at tropical latitudes during the latest Carboniferous (Virgilian) were replaced by more intensely arid climate during the earliest Permian (Wolfcampian) across central Pangaea and at least as far north as the subtropical Appalachian Basin (Cecil, 2013; DiMichele et al ., 2006; Tabor and Montañez, 2002, 2004; Tabor et al ., 2008); this change is widely considered driven by the onset of the Pangaean megamonsoon (Parrish, 1993). Particularly valuable to resolving the questions of environmental change described above are the pedogenic features of the strata from this time of transition, with numerous studies citing specific pedologic morphologies and/or the stable isotopic composition of pedogenic minerals ( e.g. DiMichele et al ., 2006; Giles et al ., 2013; Kessler et al ., 2001; Mack et al ., 2003; Montañez et al ., 2007; Tabor and Montañez, 2002, 2004; Tabor et al ., 2008; Zhu and Tabor, 2014). Tabor et al . (2008), in particular, presented a regional examination of the palaeoclimatic record of the Pennsylvanian–Permian transition for tropical western Pangaea that relied exclusively on the distribution of paleosol types. Most of the studies listed above describe pedogenic features from this time interval only broadly for the purpose of supporting or critiquing palaeoclimatic models. In an attempt to add greater resolution to the previously published research, we focus here on describing in greater detail than previously the palaeopedologic features of the upper part of the Sangre de Cristo Formation in north-central New Mexico and interpret the likely palaeoclimate for Wolfcampian time.

2. Setting and lithostratigraphy

The mostly lower Permian (Wolfcampian) Sangre de Cristo Formation accumulated as an alluvial red-bed succession in the Taos Trough of northern New Mexico following uplift ofMANUSCRIPT the Ancestral Rocky Mountains (Baltz and Myers, 1999; Soegaard and Caldwell, 1990). Overall, the formation presents a fining-upward sequence of sediments sourced mainly from the basement uplifts of the Ancestral Rocky Mountains to the north; outcrops of the lower part of the formation consist dominantly of conglomerate and coarse sandstones, while most of the upper part of the formation consists of red-bed mudstone and subordinate sandstone (Soegaard and Caldwell, 1990). These are interpreted as the record of overbank and channel deposition on a low-gradient alluvial plain (Lucas et al ., 2015). Extensive outcrop exposures of the Sangre de Cristo Formation occur in the Pecos River drainage in southwestern San Miguel County, New Mexico (Fig. 1). The formation has long been known to be fossiliferous, especially for vertebrate footprints and bones (Berman, 1993; Hunt et al ., 1990; Langston, 1953; Vaughn, 1964). Details of the lithostratigraphy, sedimentology and palaeontology are well-presented in Lucas et al . (2015). For the most part, the Sangre de Cristo Formation is of Wolfcampian age (Fig. 2; Berman et al ., 2013; Krainer et al ., 2004), equivalent to the Asselian through early Artinskian, which spans the Coyotean and Seymouran land- vertebrate faunachrons of Lucas (2005, 2006). The upper part of the formation exposed in the study area contains a vertebrate fauna (temnospondyls, synapsids and captorhinomorphs) typical of the late Wolfcampian (Lucas et al ., 2015). Although Soegaard and Caldwell (1990) assign the base of the formation to the late Desmoinesian, we find no biostratigraphic evidence for an age older than Wolfcampian, but note that the lower portion of the formation is mostly lacking in useful fossils. The formation is both correlative with and lithologically similar to the Abo Formation of central and southern New Mexico (Lucas et al ., 1999, 2005, 2012, 2013a, 2013b) and correlative with the lower part of the Hueco Group in southernmost New Mexico (Krainer and Lucas, 1995; Lucas et al ., 1998, 1999). The Sangre de Cristo Formation, however, was deposited on the piedmont and upper coastal plain that surrounded the Ancestral Rocky Mountains, in contrast to the finer-grainedACCEPTED terrestrial clastics of the Abo Formation, which were deposited on a plain near the coastline of the Midland Basin. The lower Hueco Group, even farther to the south, consists of shallow-marine carbonates deposited on the margins of the Permian Basin of west Texas and southeastern New Mexico. The thickness of the Sangre de Cristo Formation is mapped as about 300 m in the study area (Baltz and Myers, 1999; Johnson, 1969; Read et al ., 1944). Notably, Baltz and Myers (1999) and Soegaard and Caldwell (1990) cursorily described calcareous paleosols and pedogenically modified limestone beds extending through almost the entire thickness of the formation. Although no complete stratigraphic sections of the formation are available for study in any single location, the upper part of the formation is well exposed in the Pecos River drainage where it is capped by coarser, more resistant Yeso

ACCEPTED MANUSCRIPT

Group strata along the flanks of Glorieta Mesa. Outcrops south of Interstate Highway 25 present the most continuous sections of the Sangre de Cristo available for direct measure, totaling approximately 160 m (Lucas et al ., 2015). Lucas et al . (2015) presented the description of an extended section of the formation ( ca . 160 m) that includes general information on paleosols in strata stratigraphically below the section described herein. Here we present a detailed analysis of the pedogenic features in a section exposed in Chamizal Arroyo that includes the uppermost 83 m of the formation and the basal strata of the Yeso Group. The uppermost portion of the Sangre de Cristo Formation is completely represented by our measured stratigraphic section (Fig. 3). As described by Lucas et al . (2015), the section is mudstone-dominated overall, but with an abundance of lenticular, multi-story sandstones that are meters thick and typically dominated by trough cross-bedding and horizontal lamination as well as thin, laterally extensive sheet sandstone bodies that display abundant climbing-ripple lamination. Carbonates, consisting of both paleosols (calcrete) and palustrine limestones less than one meter thick, are a minor component of the section unit volumetrically, amounting to less than 10% of the section. As interpreted by Lucas et al . (2015), the upper Sangre de Cristo Formation consists of red beds dominated by overbank fine-grained sediments and volumetrically subordinate coarser-grained sediments of broad, but shallow multi-storied fluvial channel systems. Thin sandstone beds and lenses resulted from extra-channel deposition of crevasse channels and crevasse splay deposits on the floodplain adjacent to major channels.

MANUSCRIPT

Fig. 1 Location of the measured section in north-central New Mexico, with location and coordinates of the base and top of the section indicated (image adapted from Google Earth®). Inset map shows major depositional basins and Late Paleozoic uplifts of the Ancestral Rocky Mountains (adapted from Lucas et al ., 2010). I-25 = Interstate Highway 25. ACCEPTED

ACCEPTED MANUSCRIPT

Fig. 2 Stratigraphy and age assignments for upper Paleozoic strata of the Taos Trough, north-central New Mexico, indicating ambiguities and questions on the nature and duration of the unconformity underlying the Sangre de Cristo Formation (adapted from Krainer et al ., 2004). MANUSCRIPT

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED

Fig.3 Section measured in Chamizal Arroyo for this study. The stratigraphic locations of the paleosols (P1 – P7) and palustrine limestones (pl1 – pl4) described in the text are indicated. 5

ACCEPTED MANUSCRIPT

3. Location and methods

The section we measured and sampled for this study is located in San Miguel County, northern New Mexico, about 5 km south of the town of Las Vegas, New Mexico. The base of the section is located in Chamizal Arroyo at N 35°21.084’, W 105°28.068’ and the top at N 35°22.022’, W 105°30.615’ (Fig. 1). Baltz and Myers (1999) indicated that the base of the Sangre de Cristo Formation in southwestern San Miguel County is a regional unconformity, scoured into marine strata of Pennsylvanian (Desmoinesian to Virgilian) age. The basal contact is not exposed in the study area, although Lucas et al . (2015) noted the Sangre de Cristo Formation resting on the Desmoinesian Provenir Formation about 10 km to the northeast of the base of the section described here. The upper contact of Sangre de Cristo Formation strata with the Lower Permian (Leonardian) Yeso Group is very well exposed on the hillside of the Glorieta Mesa to the north and west of the arroyo. Paleosols in the section were measured and described using standard Munsell colour codes. We place the upper contact of the Sangre de Cristo Formation at the lowest bed of sandy dolomite or gypsiferous sandstone of the Glorieta Sandstone. In addition to field measurement, carbonate samples were collected for petrographic examination and stable isotope analysis. Petrographic studies were conducted on standard petrographic thin sections of 30 m thickness. Carbonate samples were collected for isotopic analysis by selectively drilling micritic calcite in lapped slabs corresponding to prepared thin sections with an ultrafine engraving tool while viewing under a binocular microscope. The samples were analyzed for δ13 C and δ18 O by Isotech Laboratories, Inc., Champaign, Illinois, by reacting the carbonate with 100% H 3PO 4 at 70°C using a Finnigan GasBench II continuous flow device paired with a Thermo Delta V Plus isotope ratio mass spectrometer. Results are reported in Table 1 in parts per thousand (‰) relative to the Vienna Pee Dee Belemnite standard (VPDB). The mineralogy of the carbonate was determined by X-ray diffraction analysis of bulk powder samples with a Bruker Phaser D2 X-ray diffractometer using a Cu-anode tube operating at 30 kV and 10 mA.

4. Description of paleosols

In general, Sangre de Cristo mudstones display a variety of pedogenic features. Common mudstone fabrics vary from platy to prismatic, although crumb fabrics occur rarely. Drab-colored root traces are common, extending to a maximum vertical length of 40 cm. Some beds display prominent pedogenic slickensides. Also common are calcareous nodules, which vary from small (1 to 2 cm) bodies with diffuse boundaries to verticallyMANUSCRIPT stacked discrete cm-scale nodules (rhizocretions). Most of the mudstone beds in the ca. 90-m measured section (Fig. 3) contain calcretes that are immature (Stage I to II of Gile et al ., 1966), but the lower and middle portions of the section contain meter-scale mature calcrete beds (Stage III), with stacked nodules forming coalescing upper boundaries, in some places with a platy fabric, suggesting Stage IV calcite (Gile et al ., 1966; Machette, 1985). Where the paleosols in the section display sufficiently distinctive pedogenic features to allow classification, we apply the terminology of Mack et al . (1993) in assigning names of paleosol orders, modified by adjectives describing the most prominent subordinate characteristic. The most common paleosol type we find in the measured section is calcic Argillisol, denoting a paleosol horizon in which the defining characteristic is a clay-rich B horizon that is visibly enriched in CaCO 3 in the form of nodules, i.e. a Btk horizon. The calcic Argillisols may also display multiple B horizons, each with distinctive characteristics, such as a Bt horizon underlain by a Bk horizon. This demonstrates a higher degree of pedogenic maturity as (presumably) greater time is required for differentiation of separate B horizons, each with distinct characteristics. In any paleosol profile with multiple defining characteristics, such as illuviated clays and pedogenic carbonate, whether they occur in the same horizon or in separate horizons, a subjective judgement was made to determine which feature is dominant and which subordinate. Thus, there may be a fine distinction between a calcic Argillisol and an argillic Calcisol. In profiles where calcareous B horizons (Bk, or Btk if argillic and calcic) outnumber the argillic (Bt) horizons, we assign the designation Calcisol or argillic Calcisol, as appropriate. In addition to Argillisols and Calcisols, and variations thereon, we recognize Vertisols, paleosols with prominent vertic fractures in the B horizon. Many mudstone beds of varying thickness display pedogenic features, such as calcareous nodules and drab-colored root traces, but lack distinct horizonation. We term these units Protosols, andACCEPTED where they contain calcareous nodules we label them calcic Protosols. Most paleosol profiles in the section are truncated in that they lack a discernible eluviated upper (A) horizon and display evidence of sediment removal, although several occurrences of calcic Argillisols do appear to display a distinct epipedon. We also recognize several compound profiles in the section. These are marked by repetition of specific types of B horizons (Bt, Btk or Bk) without an intervening A horizon, suggesting interruption of sedimentation and an erosional episode (Kraus, 1999). This pattern is noted in some Calcisol and argillic Calcisol profiles. Although common, pedogenic features are not ubiquitous in the mudstones of the upper Sangre de Cristo Formation. Several mudstone units that are meters thick (up to 5 m) lack distinct pedogenic features other than fabrics, such as blocky

ACCEPTED MANUSCRIPT

or crumb, or may simply appear massive. Less common are mudstones that display horizontal lamination, indicating deposition from suspension. These occur typically in beds less than 0.5 m thick. Clearly, rates of sedimentation varied during mud deposition from low accumulation rates that favoured pedogenic modification to high rates that prevented pedogenesis.

4.1. Individual paleosol descriptions

In the following, we present descriptions of individual paleosols in the measured section where pedogenesis has produced significantly distinctive features to permit assignment to a paleosol order.

4.1.1. Paleosol 1 (calcic Argillisol) The stratigraphically lowest distinct paleosol in the measured section occurs in Unit 3 at about 3 m in the section and displays a complex pedogenic unit that is distinctive for the presence of multiple Btk horizons as well as a preserved epipedon (A horizon)(Fig. 4). The entire profile is about 2.6 m thick. The uppermost meter is massive, reddish-yellow (7.5YR 7/6), coarse sandy mudstone with some drab-colored root traces up to 40 cm long. This unit grades downwards to finer-grained sandy mudstone that is darker brown (7.5YR 4/4). This part of the profile records superposed A and Bt horizons. The underlying Btk horizon (1BtK) is capped by a discontinuous layer of coalescing nodules 10 to 20 cm thick that grades downwards to reddish-yellow (7.5YR 7/6), coarse sandy mudstone with vertically stacked, discoidal-shaped, centimeter-scale calcareous nodules that impart a crude prismatic fabric. If the exposed profile represents a complete profile, i.e. , the A horizon is not truncated, the depth to carbonate in the profile is about 1.2 m. The abundance of nodules decreases downwards in this 60-cm thick unit. The underlying 1.3 m is brown to dark brown (7.5YR 3/3) coarse mudstone with vertically stacked nodules, decreasing in abundance downwards. This lower unit constitutes a lower Btk horizon (2Btk). Therefore, we regard this sequence of horizons as a compound paleosol, specifically a calcic Argillisol. This profile forms a distinctive marker that can be traced up the arroyo along strike for a distance of several hundred meters. Locally, the density of the nodules varies so that the B horizons are best described as Bk horizons. The variation in the maturity of the calcretes is best explained as an effect of position on the floodplain relative to the stream channel; thicker, less mature calcretes formed closer to the active channel, whereas the more mature but thinner profile formed farther from the channel where sedimentation was slower. MANUSCRIPT

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIPT

Fig. 4 Compound paleosol near base of the measured section (Paleosol 1, i.e. , P1 in Fig. 3; the same follows). A– Overview of the completeACCEPTED profile, including the pale epipedon (A horizon), best seen in the background, a clay-rich (Bt horizon) and two superimposed argillic-calcic horizons (1Btk and 2Btk). Staff for scale is 1.3 m; B–Detail of the upper Btk horizon illustrating near, but not complete coalescence of nodules with vertical stacking and prismatic fabric of underlying mudstone; C–Laterally equivalent exposure of Paleosol 1 ca. 150 m along strike (up arroyo) where the calcretes are more densely coalescing (1Bk and 2Bk).

4.1.2. Paleosols 2 and 3 (Vertisol)

ACCEPTED MANUSCRIPT

At 4.5 m, Unit 5 is a 55-cm thick mudstone, containing abundant fine, drab-colored root traces, a few larger (decimeter-scale) root traces, millimeter-scale calcite nodules and prominent vertic fractures (Fig. 5A, 5B). The mudstone fabric is platy to crudely prismatic. The mudstone in the upper part of the profile displays a pale red (5R 6/2) color, darkening downwards to weak red (5R 4/3). There is no evidence of an overlying epipedon, so we regarded this profile as truncated, but composite, with multiple Bw (Bw1 and Bw2) horizons (Kraus, 1999). We regard the vertic fractures as the most prominent characteristic of the profile, so we interpret this paleosol as a Vertisol. At 10 m, Unit 11 is a 40-cm thick weak red (10R 5/4) sandy mudstone penetrated by sandstone-filled desiccation cracks, but also containing drab-colored root traces and prominent vertic fractures (Fig. 5C, 5D). This profile is truncated by overlying sandstone sheets and so lacks an epipedon. We classify this paleosol also as a simple (having a single B horizon) Vertisol.

MANUSCRIPT

Fig. 5 Vertisols (paleosols 2 and 3). A–Overview of Paleosol 2 (P2) overlain by channel-fill sandstone; B–Vertic fractures (immediately above scale and at arrow) are prominent in the mudstone underlying the channel fill; C–Multi-story channel sandstone beds overlie Paleosol 3 (P3). A large sandstone-filled scour pit at the base of the channel occurs above the hammer. Hammer length = 26 cm; D–The arrows highlight intersecting curved fractures (vertic fractures) that characterize the mudstone.

4.1.3. Paleosol 4 (calcicACCEPTED Protosol) Unit 22 at 20 m in the measured section is a mudstone unit representative of several units in the measured section by evidence of pedogenesis, but lacking the development of distinct horizonation (Fig. 6A, 6B). For this example, the 1.7 m of mudstone underlying a tabular bed of muddy sandstone, is weak red (7.5R 5/4) and exhibits a blocky to crudely prismatic fabric. Root traces are prominent, drab-colored, downward branching and 10 to 25 cm long. Calcareous nodules are 1 to 5 cm in diameter and common, but not so abundant as to form aggregates. If a distinctly separate horizon were present, either underlying or overlying, we would identify this paleosol as a calcic Argillisol. However, there is no evidence that an epipedon or other horizon was present formerly, so we suggest that a more suitable identifier is calcic Protosol that possibly

ACCEPTED MANUSCRIPT

developed during continuous sedimentation. Extensive rooting is also evident in a number of the sandstone bodies in the measured section, manifested as both drab-coloured root traces up to 70 cm long and rhizoliths (calcified root molds) 1 to 2 cm in diameter (Fig. 7A, 7B).

Fig. 6 Calcic Protosols. A–Immature paleosol (Paleosol 4) contains prominent drab-colored root traces and scattered calcrete nodules, but lacks horizonation. Hammer length = 26 cm; B–Similarly weakly developed protosol farther up section displays drab-colored root traces, nodules and blocky mudstone fabric. Staff for scale is 1.3 m.

MANUSCRIPT

Fig. 7 Rhizoliths in sandstone. A dividing ruler of 10 cms for scale. A–Fine-grained sandstone bed (Unit 31) shows evidence of significant pedoturbation, including calcite-filled root casts (rhizoliths); B–Detail of A illustrates concentric internal layering of the calcite fill.

4.1.4. Paleosol 5 (argillic Calcisol) Unit 24 at 22 m is a compound profile. The nodular calcareous mudstone with platy fabric of the upper Btk (1Btk) horizon truncates a sandy mudstone Bt (2Bt) horizon characterized by reddish-brown color (5YR 5/3), platy fabric and prominent drab-colored root traces (Fig. 8A, 8B). This horizon grades downwards to a massive calcareous pinkish-gray (5YR 6/2) mudstone withACCEPTED abundant centimeter-scale calcite nodules that locally are stacked vertically. These we interpret as rhizocretions, nodules that encased plant roots. We assign this horizon a designation of Btk, the second in the compound profile (2Btk). We describe the paleosol as a compound argillic Calcisol.

ACCEPTED MANUSCRIPT

MANUSCRIPT Fig. 8 Outcrops of Paleosol 5 (P5; argillic Calcisol). A–Truncated section showing a coalesced Bk horizon overlying a Btk horizon; B–Laterally-equivalent outcrop (more complete than A) of Paleosol 5 illustrating an upper Btk (1Btk) horizon overlying the Bk horizon.

4.1.5. Paleosols 6 and 7 (Calcisol) Unit 28 at 27 m is a 1.5-m thick profile developed in pale red (7.5R 6/4) mudstone. The uppermost 50 cm is mudstone with few calcite nodules, but where nodules do occur, they exhibit vertical stacking, suggesting association with roots. This Bt horizon transitions gradationally to a Btk horizon marked by calcareous nodules up to 8 cm in diameter, locally aggregated as botryoidal masses. The Btk horizon transitions abruptly to a 10 to 15 cm thick Bk horizon. This carbonate layer sharply overlies a second Btk horizon (2Btk) of weak red (7.5R 6/2) mudstone hosting abundant centimeter-scale nodules with prominent vertical stacking. We describe this profile as a Calcisol in which the Bk horizon constitutes a plugged layer (Stage III calcrete) that bisects the profile, separating the two Btk horizons. Unit 40 at 52 m is a 90-centimeter unit divisible into an upper calcareous ledge constituting a Bk horizon and a lower mudstone unit (Fig. 9A, 9B). The upper calcareous layer is a Bk horizon that is platy-weathering at the top, grading downward to nodular-weathering, locally penetrated by root traces. The top of the unit is sharply truncated and overlain by the mudstone Bt horizon of the superposed paleosol. The Bk horizon grades downward to yellowish-brown (10YR 5/4) mudstone of the underlying Btk horizon. The most distinctive feature of this horizon is the semi-prismatic fabric resulting from the vertically-stacked,ACCEPTED centimeter-scale nodules. We interpret this profile as an erosionally truncated Calcisol with an upper plugged Bk horizon (Stage IV calcrete). This is the uppermost distinctive and well-developed paleosol in the measured section.

ACCEPTED MANUSCRIPT

Fig. 9 Paleosol 7, truncated Calcisol. A–Outcrop view of Paleosol 7 (P7). Profile is truncated above the laminar Bk horizon. Btk horizon contains rhizocretions (vertically stacked discoidal nodules; arrow). Overlying carbonate ledges are beds of palustrine limestone. Hammer (26 cm in length) for scale in center of bed pl2; B–Closer view shows that root channels (visible immediately above the letters tk) penetrate the profile. Arrow indicates rhizocretion.

4.1.6. Other pedogenic features Pedogenic features occur in the section above Unit 40, although these consist for the most part of nodules and root traces widely dispersed in mudstone, and root traces penetrating the sandstone beds in Units 47 through 58, as described for Paleosol 6 (calcic Protosol). We also note that several of the beds in the uppermost seven meters of the formation are intraformational conglomerates consisting of reworked clasts of pedogenic carbonate. 4.2. Fossil forest MANUSCRIPT A nearby flagstone quarry exposes a one-meter thick sandstone bed, correlated to Unit 27 in the measured section, that exposes the growth-position molds of the stems of 165 coniferophytic gymnosperms (Fig. 10A–10D; Rinehart et al ., 2015). Most of the stems are partially to completely filled with grayish calcite that has a texture that is typically nodular, but less commonly is layered parallel to the walls of the mold. Calcareous nodules, 1 to 3 cm in diameter, occur in the very-fine grained sandstone that hosts the stem molds and in the sandy mudstone that overlies this bed.

ACCEPTED

ACCEPTED MANUSCRIPT

Fig. 10 Views of “fossil forest” described by Rinehart et al . (2015).MANUSCRIPT A dividing ruler of 10 cms for scale in B, C and D. A– Overview of flagstone quarry in which stem casts are preserved in a single, continuous sandstone bed (1.3 m staff at arrow for scale); B–View of closely-spaced stems with adventitious roots in quarry wall; C–Close-up of stem case filled by nodular calcrete. Some isolated calcrete nodules occur in the host sandstone (see to the right of the scale); D–Stem cast filled by calcrete with a ropy/vertically-layered texture.

4.3. Petrography

The carbonate in the pedogenic nodules consists of low-Mg calcite, as determined by X-ray diffraction (XRD) analysis. Petrographically, the carbonate in the nodules consists primarily of dense, silty micrite, but the texture is not uniform, areas of denser micrite are interspersed with lighter, less dense regions (Fig. 11A). Locally, peloids occur (Fig. 11B), and there are variations in the concentration of the silt component. The micrite fabric is disrupted by one millimeter- wide veins of sparry calcite (crystallaria) and two- to three-millimeter wide channels (rootlets) filled by a combination of microspar, peloids and silt. Larger siliciclastic grains commonly display coronas of microspar and corroded margins. Millimeter-scale concentrations of organic matter occur locally. ACCEPTED

ACCEPTED MANUSCRIPT

Fig. 11 Microscopic features of calcrete nodules. A–Lighter and darker mottling characterizes the silty micrite. Clay-lined rootlet (r) in center of the field; B–Lighter region in center contrasting with darker micrite is a root channel partially filled by peloids (p).

4.4 Pedogenic interpretation

A pedogenic origin for the features described above is indisputable given the common association of roots and calcareous nodules in these strata. Moreover, petrography reveals the presence of typical alpha-type calcrete fabrics in the nodules, such as grain coronas and corroded floating grains that indicate that these nodules represent the accumulation of carbonate in the B horizon of paleosols (Alonso-Zarza and Wright, 2010a). Thus, we are confident in our interpretation that these carbonates represent calcretes formed within soils developed on alluvial muds. The mudstone-hosted calcareous nodule horizons represent simple Stage II to III calcrete profiles (Esteban and Klappa, 1983; Gile et al ., 1966; Machette, 1985). Stage II profiles comprise isolated nodules in discrete horizons with gradational bases and tops, typical of the calcic Argillisols described above, whereas the more mature Stage III MANUSCRIPTprofiles in the Calcisols consist of coalesced horizons in which the nodules are generally vertically stacked. A platy fabric in the upper part of a Bk horizon provides evidence that the calcrete reached a stage of maturity in which the carbonate cement occluded porosity, causing development of a plugged layer, as a Stage IV calcrete. No calcretes with maturity higher than Stage IV were observed in the section.

5. Limestone beds

Carbonate ledges occur at several locations in the measured section. The lowermost of these, Unit 19, is a mottled pinkish-gray micritic limestone with a sharp base overlying blocky mudstone containing calcareous nodules (Fig. 12A). The limestone is nodular throughout its thickness and characterized by a mottled, brecciated texture comprising sharply angular domains highlighted by pinkish-gray to brown mottling. The entire thickness of the bed, which varies laterally from 30 to 50 cm, is penetrated by root channels. Two more ledges occur higher in the section at Units 42 and 44 (Fig. 12B, 12C). Each bed is 50 cm thick, has a sharp lower contact and consists of muddy limestone with a brecciated, mottled texture as described above for Unit 19. These units also are penetrated by drab-colored root traces and have nodular-weathering upper contacts. ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIPT

Fig. 12 Palustrine limestone beds. A–The stratigraphically lowest palustrine bed, unit 19, displays a strongly nodular upper surface; B–Two successive palustrine beds (units 42 and 44) located above Paleosol 7 (Fig. 8) both display very abrupt lower contacts. Hammer (26 cm in length) for scale above the letter t in Unit 42; C–Internal structure of Unit 42 (palustrine bed 2) displaysACCEPTED mottling caused by brecciation and pervasive penetration of the bed by roots (r).

5.1. Petrography

The carbonate in these limestone beds is confirmed by XRD analysis as low-Mg calcite. Petrographically, the bedded limestone carbonate consists of very dense, silty micrite. Although charophyte debris is lacking, some body fossils,

ACCEPTED MANUSCRIPT

potentially fragments of ostracod shells, occur, most prominently in Unit 19. A brecciated fabric throughout all or part of the bed is the most consistent characteristic (Fig. 13A). This fabric consists of semi-angular to angular domains of dense micrite up to 10 cm in diameter, although the domain boundaries vary from diffuse to sharp, and typically are separated by millimeter-wide to centimeter-wide veins of equant crystals of microspar to sparry calcite. The veins between the domains vary from straight to curved, have orientations ranging from horizontal to vertical, and the calcite filling is nonluminescent. Subvertical to vertical channels surrounded by drab-colored haloes form fissures up to 40 cm long and 6 cm wide (Fig. 13B); these typically extend from the upper contact of beds and generally taper downwards and bifurcate. Smaller millimeter-scale, drab-gray filaments are filled variously by fine to coarse calcite spar, as well as peloids, silt and clay.

Fig. 13 Microscopic features of palustrine limestones. A–Micrite in Unit 44 (pl3) displays incipient brecciated texture with micritic domains bounded by spar-filled circumgranular cracks (bright features); B–Root channel in Unit 44 filled by micrite, surrounded by much denser micrite. 5.2. Palustrine interpretation MANUSCRIPT Based on the brecciated texture and geometry of the beds, in particular the sharp lower contact and lateral variations in bed thickness, and the lack of a characteristic lacustrine fauna, we interpret these beds as palustrine carbonate sediments deposited in shallow ephemeral to perennial ponds that occupied low areas on the broad floodplain and were subsequently subjected to extensive pedogenesis (Alonso-Zarza, 2003; Alonso-Zarza and Wright, 2010b; Alonso-Zarza et al ., 1992; Platt and Wright, 1991; Tanner, 2000). The brecciation of the carbonate and accompanying color mottling reflect pedogenic modification of the original carbonate textures by repeated desiccation during episodes of water body evaporation, during which plant roots penetrated the sediment (Alonso-Zarza, 2003; Alonso-Zarza and Wright, 2010b; Freytet and Verrecchia, 2002; Platt, 1989, 1992; Plaziat and Freytet, 1978). Phreatic cement subsequently filled the desiccation fractures with sparry cement. Generally, the pervasiveness of the brecciation suggests pedogenesis under semi-arid conditions (Platt and Wright, 1992). Some beds exhibit nearly vertical, downward-tapering or bifurcating channels or fissures that were likely created by root penetration and solution enlargement, indicating extensive subaerial exposure (Platt, 1989).

6. Isotopic analysis

6.1. Isotopic data

Samples for stableACCEPTED isotope analysis ( δ13 C and δ18 O) included nodules from four of the paleosol units described above, as well as from the mudstones overlying the fossil forest described above and from the sandstone bed hosting the stem casts; additional analyses included five samples of the carbonate filling the stem molds and one from the limestone bed Unit 53. The isotopic analyses of the carbonate from the calcrete nodules from paleosol units in the measured section and from the fossil forest locality demonstrate consistency (Table 1). Mean δ13 C = -5.97‰ (VPDB) and mean δ18 O = -4.64‰ (VPDB), although Unit 51 is an outlier in this data set with δ13 C and δ18 O both enriched more than 1.0‰ compared to the mean for the calcrete carbonate. The composition of the carbonate from five stem casts from the fossil forest is nearly identical to that of the nodules, with mean δ13 C = -6.14‰ (VPDB) and mean δ18 O = -5.31‰ (VPDB), suggesting that the carbonate infilling

ACCEPTED MANUSCRIPT

occurred in the soil profile following the decay of the buried stems (as described by Rinehart et al ., 2015). Notably, the carbonate from Unit 53, δ13 C = -4.80‰ (VPDB) and δ18 O = -3.95‰ (VPDB), is significantly enriched compared to the calcrete carbonate. This is consistent with the interpretation of carbonate precipitation from evaporatively enriched waters (Talbot, 1990; Talbot and Kelts, 1990, Tanner, 2000, 2010; Alonso-Zarza, 2003, Alonso-Zarza and Wright, 2010b). Unit 51, which is interpreted as calcrete carbonate based on the morphological characteristics of the carbonate, yielded an isotopic composition almost identical to that of Unit 53. Potentially, the Unit 51 composition reflects wetland (palustrine) depositional influence that is not apparent due to the pedogenic overprint. Alternatively, the sample records the enriched composition of carbonate that accumulated in the plugged horizon of the profile, closer to the surface than most calcrete carbonate.

Table 1 Isotopic composition of nodular calcrete, calcrete from stem casts at “fossil forest” (East Quarry) and palustrine limestone. Sample locations are shown in Fig. 3. Sample location δ13 C‰ (VPDB) δ18 O‰ (VPDB) Facies Unit 6 -5.56 -4.81 Calcrete Unit 19 -6.25 -4.95 Calcrete Unit 51 -4.65 -3.62 Calcrete (?) Unit 53 -4.80 -3.95 Palustrine limestone Unit 60 -5.70 -4.88 Calcrete East Quarry 14A -6.58 -5.06 Calcrete East Quarry 14B -7.07 -4.53 Calcrete East Quarry 4A -6.28 -5.51 Stem mold fill East Quarry 4B -5.83 -5.82 Stem mold fill East Quarry 5 -6.40 -4.82 Stem mold fill East Quarry 6 -6.20 -5.23 Stem mold fill East Quarry 8 -6.00 -5.17 Stem mold fill 6.2. Interpretation. MANUSCRIPT The isotopic analyses presented above demonstrate substantial consistency in the isotopic composition of the carbonate in the calcrete nodules, and in the carbonate-filled stem casts, with no significant difference between the oldest and youngest samples. The oxygen isotope values presented herein (calcrete mean δ18 O = -4.64‰) are depleted relative to those presented by Tabor and Montañez (2002), who documented isotopic depletion with distance from both the coast and the equator, in addition to a temporal trend of enrichment from Virgilian to Wolfcampian time consistent with 1.5 to 3 °C warming. The authors presented a value of δ18 O = -3.2‰ for a Wolfcampian paleosol from a paleolatitude of about 5° N, which corresponds approximately to the palaeoposition of the Taos Trough. We note, however, that limestones we interpret as likely palustrine in origin are much closer to the value of Tabor and Montañez (2002). Our calcrete mean δ13 C of -5.97‰ is well within their range of δ13 C = -1.2‰ at the equator to -7.8‰ at 8° N.

7. Interpretation and discussion

7.1. Palaeoclimatic interpretation

7.1.1. Paleosols All of the Sangre de Cristo Formation paleosols described herein developed on alluvial sediments, i.e. , on the broad floodplains of the SangreACCEPTED de Cristo streams. The most common features of these paleosols are the calcrete nodules and drab- colored root traces in the reddish-brown sediments, indicative of formation in well-drained soils (Kraus and Hasiotis, 2006). In general terms, pedogenic carbonate accumulates during the dry season in regions that experience significant seasonal fluctuations in climate (Allen and Wright, 1989; Birkeland, 1999; Retallack, 2001, 2005; Schaetzl and Thompson, 2015), but this encompasses a broad range of climates, from semi-arid to sub-humid (250 mm to > 750 mm precipitation per year). Calcareous paleosols with nodular features and fabrics similar to those we observed have been described from the Miocene Siwalik of Pakistan and interpreted as produced by a climate that was tropical and subhumid, but highly seasonal (Zaleha, 1997). Seasonality of precipitation is also a requirement for the development of vertic fractures in soils, which also form

ACCEPTED MANUSCRIPT

most commonly in semi-arid to subhumid environments (500 mm to 1000 mm precipitation; Birkeland, 1999; Schaetzl and Thompson, 2015). Most of the paleosol profiles in our section are truncated, preventing a measurement of the depth to carbonate. However, we note here that Paleoso l 4 (in Unit 22) is the sole profile in which we observed an apparent epipedon. If we assume that the epipedon is complete, we can measure a depth to carbonate of 1.2 m (corrected to 1.3 m to account for compaction). Using the algorithm of Retallack (2005), we calculate a mean annual precipitation of 756 mm. Although this calculation represents a rough approximation, and the depth to carbonate calculation method remains controversial ( cf . Royer, 1999), the figure obtained is consistent with the observations presented herein that suggest mainly subhumid, but highly seasonal environmental conditions during soil formation. The palaeontological record of the Sangre de Cristo Formation also supports strong seasonality. Lucas et al . (2015), in their review, noted the presence of estivation burrowers in the tetrapod fauna, including the amphibians Dissorophus and Diplocaulus . Relatively few of the Sangre de Cristo paleosols display very high levels of maturity, however; the maximum stage of development of the calcretes in the Calcisols is Stage IV, indicating that prolonged depositional hiatuses were few and confirming that the palaeoclimate was not overly arid. More common are the less mature calcic Argillisols and the immature calcic Protosols. Collectively, these paleosol types support the interpretation of generally subhumid seasonal conditions and relatively continuous sedimentation. Additionally, the occurrences of mudstone beds that display little or no evidence of pedogenic modification demonstrates consistent and relatively rapid sediment accumulation in the Taos Trough during Wolfcampian time, likely a function of both the proximity to the basement uplift sediment sources in the Ancestral Rocky Mountains (Fig. 1) as well as a palaeoclimate conducive to sediment production and transport.

7.1.2. Palustrine limestones The formation of pedogenic fabrics in lacustrine and wetland carbonate is explained by episodic, perhaps seasonal changes in the regional water table, or hydrologic base level, which subjected the carbonate sediments to pedogenesis, i.e. , repeated episodes of wetting/drying and root penetration. The modern Las Tablas de Daimiel wetlands in Spain displays similar features, and Alonso-Zarza et al . (2006) suggested that these are an appropriate modern analog for the numerous examples of palustrine carbonates interpreted in the ancient record ( e.g. Sanz et al ., 1995; Tanner, 2000; Talbot et al ., 1994). Because the formation of the characteristic textures of palustrine carbonates is very sensitive to changes in hydrology, and consequently, to variations in precipitation or groundwater levels, there exist a variety of features that are indicative of water availability during the formation and subaerial exposure of palustrine sediments. Organic-rich lithologies, for example, have a higher preservation potential in subhumid than in semi-arid climates, while greater aridity will cause more intense desiccation, with consequent developmentMANUSCRIPT of brecciated and peloidal fabrics, microkarst surfaces and, with much greater aridity, the formation of evaporites that may be altered subsequently to chert (Alonso-Zarza, 2003; Alonso-Zarza et al ., 1992; Bustillo, 2010; Platt and Wright, 1992). Although the palustrine sediments of the Sangre de Cristo Formation lack organic-rich lithologies, evaporites and chert are similarly lacking, as are microkarst surfaces and strongly peloidal fabrics. Hence, the palustrine limestones here are not indicative of a strongly arid palaeoclimate.

7.1.3. Comparison to previous studies Consistent with our interpretation, Mack et al . (2003) described a pedogenic record from the time-equivalent Abo Formation in the nearby Orogrande Basin of south-central New Mexico that suggested a subhumid to semi-arid, seasonal climate during the Wolfcampian. The study of paleosols in northeastern New Mexico by Kessler et al . (2001) suggested cyclical fluctuations in humidity from early Wolfcampian through early Leonardian, but with an overprint of long-term aridification. Isbell et al . (2003) linked the formation of cyclothems during the Visean to Artinskian to the fluctuating size of large Pangaean ice sheets, but admitted that the periodicity was very unclear. Importantly, we find no evidence of a temporal trend in the types or maturity of the Sangre de Cristo Formation paleosols. Although the study herein focuses on just the uppermost 80+ m of the formation, the description of an extended section of the formation (ca . 160 m) presented in Lucas et al . (2015) includes information on paleosols in strata stratigraphically below our section. While lacking detail, this study also describes mostly immature calcareous paleosols, but including some calcretes of significant maturity. Similarly, Soegaard and Caldwell (1990) and Baltz and Myers (1999), in their broader studies of the Sangre de Cristo Formation, described calcareous paleosols and pedogenically modified limestone beds extending through almost the entire ca. 300-m thickness of the formation.ACCEPTED Hence, the overall palaeoclimatic record presented by the Sangre de Cristo paleosols is one of general subhumid conditions, but with shorter-term fluctuations in humidity that produced Calcisols, perhaps over time spans of 10 4 years. However, we find no evidence of periodicity or an overall long-term trend during the interval of Sangre de Cristo deposition, encompassing most of the Wolfcampian and potentially the latest Desmoinesian time.

7.2. Published models

ACCEPTED MANUSCRIPT

Mack (2003), based on the presence of vertic and calcic paleosols, xerophytic vegetation and large ectothermic vertebrate fauna, described the Wolfcampian palaeoclimate in New Mexico as warm, semi-arid to subhumid, and seasonal, with an estimated 300 to 1000 mm precipitation annually. Leonardian time, by contrast, is marked by a transition to more arid conditions, recorded by gypsic paleosols, eolian sandstones and evaporates. Tabor and Montañez (2004), however, noted a pronounced change in the character of the paleosols in the coastal plain environment of the Eastern Shelf of the Midland Basin of North-Central Texas. Tropical, humid conditions during the Late Pennsylvanian are recorded by humid- climate typical soil types as Histosols, Ultisols and Inceptisols, as well as Vertisols and Entisols. These are replaced in the Early Permian by drier, more seasonal soils types, including Vertisols, Aridisols, and Entisols. DiMichele et al . (2006) noted a similar change in paleosols occurring across the Pennsylvanian–Permian boundary on the Eastern Shelf of the Midland Basin as the humid conditions indicated by coals and Histosols are replaced by drier conditions recorded by Vertisols and evaporates. Giles et al . (2013) documented a similar transition in Oklahoma. Tabor et al . (2008) described paleosols from the Virgilian in the Taos Trough of New Mexico that range from Argillisols to Calcisols, but including also Vertisols, calcic Argillisols and argillic Calcisols. They attributed this wide range to a climate that varied between alternating conditions of stable, nonseasonal subhumid conditions to semi-arid with very strong seasonality. In this same region, paleosols of Wolfcampian age are more commonly Calcisols, and Argillisols and Vertisols become less abundant. According to Tabor et al . (2008), the palaeolatitudes a few degrees above the equator, including the Four Corners region of the Colorado Plateau (adjacent regions of Colorado, Utah, Arizona and New Mexico) were continuously semi-arid to arid from the latest Pennsylvanian (Virgilian) through the early Permian, while areas straddling the palaeoequator, including most of New Mexico, experienced a progressive shift over this same time interval. Equatorial regions experienced variably seasonally subhumid to semi-arid conditions through the Late Pennsylvanian, continuing into the early Permian (through the Wolfcampian), before a shift to greater aridity near the end of the early Permian (Leonardian). The shift in climate documented by paleosols across the Pennsylvanian–Permian boundary in many studies, or alternatively, during the early Permian, is at odds with climate models for this interval. Kutzbach and Gallimore (1989) and Kutzbach and Ziegler (1994) presented models that indicated that tropical easterly (zonal) flow maintained year-round moisture at tropical latitudes at this time. Parrish (1993), however, suggested that during the early Permian the onset of large-scale monsoonal circulation, which hypothetically might have triggered reverse (westerly) equatorial flow (Kessler et al ., 2001; Parrish, 1993; Soreghan et al ., 2002; Tabor and Montañez, 2002) and strong seasonality. Alternatively, aridification of the more northerly portions of the tropical region of western Pangaea at this time may have accompanied the northward drift of the supercontinent across narrow, unmoving climate belts (Kessler et al ., 2001; Kutzbach and Ziegler, 1994; Parrish, 1993). MANUSCRIPT As stated above, however, our data suggest that the palaeoclimate during Wolfcampian deposition of the Sangre de Cristo Formation varied from subhumid to semi-arid, but appears consistently more humid than modelled for this region by Tabor et al . (2008). Furthermore, we do not detect a temporal trend from our data, although we admit that our data temporally limited, and the changes modelled by Tabor et al . (2008) for the region including the Taos Trough (their Mid- Region North, p. 298) are both subtle and gradual, and vary by palaeolatitude.

8. Conclusions

The mudstones of the Lower Permian (Wolfcampian) Sangre de Cristo Formation host a variety of pedogenic features including pedogenic slickensides, drab-colored root traces and common calcareous nodules, which vary from small isolated bodies to vertically stacked nodule accumulations (rhizocretions), locally forming coalescing horizons and rarely plugged calcrete horizons. Paleosol orders interpreted from these features include Vertisols, Argillisols, calcic Argillisols, argillic Calcisols, Calcisols and calcic Protosols. Of these, the calcic Argillisols and calcic Protosols are the most common. The maturity of the calcrete is low (Stage I to II) in most of the soil profiles, but the lower to middle portion of the section contains mature calcrete beds (Stage III to IV), typically in compound soil profiles. Thin limestone beds interbedded with the mudstones have sharp bases and brecciated textures and were penetrated by root exhibit features consistent with carbonate deposition in ponds or wetlands, subsequently modified by pedogenesis. Although variationsACCEPTED in type and maturity of the paleosols in the section indicate episodic climatic fluctuations, the overall character of the paleosols suggests predominantly borderline subhumid, but highly seasonal climate during most of the interval of deposition. Hence, the section studied here supports previous studies that interpret a subhumid, but strongly seasonal climate during the early Permian (Wolfcampian); consistent semi-arid conditions are not in evidence for this location at this time. From the section studied in detail herein as well as data from previous work on this formation we find no evidence for a temporal shift in climate during deposition of the Sangre de Cristo Formation.

ACCEPTED MANUSCRIPT

Acknowledgements

The authors gratefully acknowledge the assistance of L. Rinehart, W. J. Nelson, S. Elrick, D. Chaney and W. DiMichele for assistance in conducting the field study, and M. Zaleha for insightful discussion. Helpful reviews and comments on the manuscript were provided by W. J. Nelson, S. Elrick and two anonymous reviewers. Additionally, we thank X. F. Hu for editorial assistance.

References

Allen, J.R.L., Wright, V.P., 1989. Paleosols in siliciclastic sequences. University of Reading (UK), Postgraduate Research Institute of Sedimentology Short Course Notes, no. 1, 97 pp. Alonso-Zarza, A.M., 2003. Palaeoenvironmental significance of palustrine carbonates and calcretes in the geological record. Earth-Science Reviews , 60(3–4), 261–298. Alonso-Zarza, A.M., Calvo, J.P., García del Cura, M.A., 1992. Palustrine sedimentation and associated features— grainification and pseudomicrokarst—in the Middle Miocene (intermediate unit) of the Madrid Basin, Spain. Sedimentary Geology , 76, 43–61. Alonso-Zarza, A.M., Dorado-Valiño, M., Valdeolmillos-Rodríguez, A., Ruiz-Zapata, M.B., 2006. A recent analogue for palustrine carbonate environments: The Quaternary deposits of Las Tablas de Daimiel wetlands, Ciudad Real, Spain. In: Alonso-Zarza, A.M., Tanner, L.H. (Eds.), Paleoenvironmental Record and Applications of Calcretes and Palustrine Carbonates . GSA Special Paper , 416, pp. 153–168. Alonso-Zarza, A.M., Wright, V.P., 2010a. Palustrine carbonates. In: Alonso-Zarza, A.M., Tanner, L.H. (Eds.), Carbonates in Continental Environments: Processes, Facies and Applications. Developments in Sedimentology 61, Elsevier, Amsterdam, pp. 103–131. Alonso-Zarza, A.M., Wright, V.P., 2010b. Calcretes. In: Alonso-Zarza, A.M., Tanner, L.H. (Eds.), Carbonates in Continental Environments: Processes, Facies and Applications.MANUSCRIPT Developments in Sedimentology 61, Elsevier, Amsterdam, pp. 226–267. Baltz, E.H., Myers, D.A., 1999. Stratigraphic Framework of Upper Paleozoic Rocks, Southeastern Sangre de Cristo Mountains, New Mexico: With a Section on Speculations and Implications for Regional Interpretation of Ancestral Rocky Mountains Paleotectonics . New Mexico Bureau of Mines and Mineral Resources, Memoir , 48, 269 pp., 5 plates. Berman, D.S., 1993. Lower Permian vertebrate localities of New Mexico and their assemblages. New Mexico Museum of Natural History and Science Bulletin , 2, 11–21. Berman, D.S., Henrici, A.C., Lucas, S.G., 2013. Ophiacodon (Synapsida, Ophiacodontidae) from the Lower Permian Sangre de Cristo Formation of New Mexico. New Mexico Museum of Natural History and Science Bulletin , 60, 36–41. Birkeland, P.W., 1999. Soils and Geomorphology, 3 rd edition. Oxford University Press, New York, 536 pp. Bustillo, M.A., 2010. Silicification of continental carbonates. In: Alonso-Zarza, A.M., Tanner, L.H. (Eds.), Carbonates in Continental Environments: Processes, Facies and Applications . Developments in Sedimentology 62, Elsevier, Amsterdam, pp. 153–179. Cecil, C.B., 2013. An overview and interpretation of autocyclic and allocyclic processes and the accumulation of strata during the Pennsylvanian–Permian transition in the central Appalachian Basin, USA. International Journal of Coal Geology , 119, 21–31. DiMichele, W.A., Tabor,ACCEPTED N.J., Chaney, D.S., Nelson, W.J., 2006. From wetlands to wet spots: Environmental tracking and the fate of Carboniferous elements in Early Permian tropical floras. In: Greb, S.F., DiMichele, W.A. (Eds.), Wetlands through Time . GSA Special Paper , 399, pp. 223–248. Dinterman, P.A., Mack, G.H., Leeder, M., Perez-Arlucea, M., 2000. Palaeoclimatic control on Early Permian fluvial to eolian transition in New Mexico. GSA Abstracts with Program , 32(7), 254. Esteban, M., Klappa, C.F., 1983. Subaerial exposure environments. In: Scholle, P.A., Bebout, D.G., Moore, C.H. (Eds.), Carbonate Depositional Environments . America Society of Petroleum Geologists Memoir, 33, pp. 1–54.

ACCEPTED MANUSCRIPT

Freytet, P., Verrecchia, E.P., 2002. Lacustrine and palustrine carbonate petrography: An overview. Journal of Paleolimnology , 27(2), 221–237. Gile, L.H., Peterson, F.F., Grossman, R.B., 1966. Morphological and genetic sequences of carbonate accumulation in desert soils. Soil Science , 101(5), 347–360. Giles, J.M., Soreghan, M.J., Benison, K.C., Soreghan, G.S., Hasiotis, S.T., 2013. Lakes, loess, and paleosols in the Permian Wellington Formation of Oklahoma, U.S.A.: Implications for paleoclimate and paleogeography of the Midcontinent. Journal of Sedimentary Research , 83(10), 825–846. Hunt, A.P., Lucas, S.G., Huber, P., 1990. Early Permian footprint fauna from the Sangre de Cristo Formation of northeastern New Mexico. New Mexico Geological Society Guidebook , 41, 291–303. Isbell, J.L., Miller, M.F., Wolfe, K.I., Lenaker, P.A., 2003. Timing of Late Paleozoic glaciation in Gondwana: Was glaciation responsible for the development of northern hemisphere cyclothems? In: Chan, M.A., Archer, A.W. (Eds.), Extreme Depositional Environments: Mega End Members in Geologic Time . GSA Special Paper , 370, pp. 5–24. Johnson, R.B., 1969. Geologic map of the Villanueva quadrangle, San Miguel County, New Mexico. U.S. Geological Survey, Geologic Quadrangle Map GQ-869, scale 1:62,500. Kessler, J.L.P., Soreghan, G.S., Wacker, H.J., 2001. Equatorial aridity in western Pangea: Lower Permian loessite and dolomitic paleosols in northeastern New Mexico, U.S.A. Journal of Sedimentary Research , 71(5), 817–832. Krainer, K., Lucas, S.G., 1995. The limestone facies of the Abo–Hueco transitional zone in the Robledo Mountains, southern New Mexico. New Mexico Museum of Natural History and Science Bulletin , 6(1), 33-38. Krainer, K., Lucas, S.G., Kues, B.S., 2004. Tectonically induced clastic-carbonate depositional sequences of the Pennsylvanian–Permian transition in the Rowe-Mora basin, northern New Mexico. New Mexico Geological Society 55 th Annual Fall Field Conference Guidebook , 55, 314-325. Kraus, M.J., 1999. Paleosols in clastic sedimentary rocks: Their geologic applications. Earth-Science Reviews , 47(1–2), 41– 70. Kraus, M.J., Hasiotis, S.T., 2006. Significance of different modes of rhizolith preservation to interpreting paleoenvironmental and paleohydrologic settings: ExamplesMANUSCRIPT from Paleogene paleosols, Bighorn Basin, Wyoming, U.S.A. Journal of Sedimentary Research , 76(4), 633–646. Kutzbach, J.E., Gallimore, R.G., 1989. Pangaean climates: Megamonsoons of the megacontinent. Journal of Geophysical Research , 94(D3), 3341–3357. Kutzbach, J.E., Ziegler, A.M., 1994. Simulation of Late Permian climate and biomes with an atmosphere-ocean model: Comparisons and observations. In: Allen, J.R.L., Hoskins, B.J., Sellwood, B.W., Spicer, R.A., Valdes, P.J. (Eds.), Palaeoclimates and Their Modelling: With Special Reference to the Mesozoic Era . Chapman and Hall, London, pp. 119–132. Langston, W., Jr., 1953. Permian amphibians from New Mexico. University of California, Publications in Geological Science , 29, 349–416. Lucas, S.G., 2005. Permian tetrapod faunachrons. New Mexico Museum of Natural History and Science Bulletin , 30, 197– 201. Lucas, S.G., 2006. Global Permian tetrapod biostratigraphy and biochronology. In: Lucas, S.G., Cassinis, G., Schneider, J.W. (Eds.), Non-marine Permian Biostratigraphy and Biochronology . Geological Society, London, Special Publication , 265, pp. 65–93. Lucas, S.G., Harris, S.K., Spielmann, J.A. Berman, D.S, Henrici, A.C., Krainer, K., Rinehart, L.F., DiMichele, W.A., Chaney, D.S., Kerp,ACCEPTED H., 2012. Lithostratigraphy, paleontology, biostratigraphy, and age of the Upper Paleozoic Abo Formation near Jemez Springs, northern New Mexico, USA. Annals of the Carnegie Museum , 80, 323–350. Lucas, S.G., Heckert, A.B., Estep, J.W., Hunt, A.P., Anderson, O.J., 1998. Stratigraphy, paleontology and depositional environments of the Lower Permian Robledo Mountains Formation of the Hueco Group, Robledo Mountains, New Mexico. New Mexico Museum of Natural History and Science Bulletin , 12, 29–41.

ACCEPTED MANUSCRIPT

Lucas, S.G., Krainer, K., Chaney, D.S., DiMichele, W.A., Voigt, S., Berman, D.S., Henrici, A.C., 2013b. The Lower Permian Abo Formation in central New Mexico. New Mexico Museum of Natural History and Science Bulletin , 59, 161–179. Lucas, S.G., Krainer, K., Colpitts, R.M., Jr., 2005. Abo-Yeso (Lower Permian) stratigraphy in central New Mexico. New Mexico Museum of Natural History and Science Bulletin , 31, 101–117. Lucas, S.G., Krainer, K., Dimichele, W.A., Voigt, S., Berman, D.S., Henrici, A.C., Tanner, L.H., Chaney, D.S., Elrick, S.D., Nelson, W.J., Rinehart, L.F., 2015. Lithostratigraphy, biostratigraphy and sedimentology of the Upper Paleozoic Sangre de Cristo Formation, southwestern San Miguel County, New Mexico. In: Lindline, J., Petronis, M., Zebrowski, J. (Eds.), Geology of the Las Vegas Area . New Mexico Geological Society 66 th Annual Fall Field Conference Guidebook , 66, pp. 211–228. Lucas, S.G., Krainer, K., Voigt, S., 2013a. The Lower Permian Yeso Group in central New Mexico. New Mexico Museum of Natural History and Science Bulletin , 59, 181–199. Lucas, S.G., Rowland, J.M., Kues, B.S., Estep, J.W., Wilde, G.L., 1999. Uppermost Pennsylvanian and Permian stratigraphy and biostratigraphy at Placitas, New Mexico. New Mexico Geological Society, 50 th Field Conference Guidebook , 281–292. Lucas, S.G., Schneider, J.W., Spielmann, J.A., 2010. Carboniferous–Permian transition in Cañon del Cobre, northern New Mexico: An overview. New Mexico Museum of Natural History and Science Bulletin , 49, 1–5. Machette, M.N., 1985. Calcic soils of the southwestern United States. In: Weide, D.L. (Ed.), Soils and Quaternary Geology of the Southwest United States . GSA Special Paper , 203, pp. 1–21. Mack, G.H., 2003. Lower Permian terrestrial palaeoclimatic indicators in New Mexico and their comparison to paleoclimate models. New Mexico Geological Society Guidebook , 54, 231–234. Mack, G.H., James, W.C., Monger, H.C., 1993. Classification of paleosols. GSA Bulletin , 105(2), 129–136. Mack, G.H., Leeder, M., Perez-Arlucea, M., Bailey, B.D.J., 2003. Early Permian silt-bed fluvial sedimentation in the Orogrande Basin of the Ancestral Rocky Mountains, New Mexico, USA. Sedimentary Geology , 160, 159–178. Montañez, I.P., Tabor, N.J., Niemeier, D., DiMichele, W.A., Frank,MANUSCRIPT T.D., Fielding, C.R., Isbell, J.L., 2007. CO 2-forced climate and vegetation instability during Late Paleozoic deglaciation. Science , 315(5808), 87–91. Mountney, N.P., 2006. Periodic accumulation and destruction of aeolian erg sequences in the Permian Cedar Mesa Sandstone, White Canyon, southern Utah, USA. Sedimentology , 53(4), 789–823. Parrish, J.T., 1993. Climate of the supercontinent Pangea. The Journal of Geology , 101(2), 215–253. Platt, N.H., 1989. Lacustrine carbonates and pedogenesis: Sedimentology and origin of palustrine deposits from the Early Cretaceous Rupelo Formation, W Cameros Basin, N Spain. Sedimentology , 36(4), 665–684. Platt, N.H., 1992. Fresh-water carbonates from the lower Freshwater Molasse (Oligocene, western Switzerland): Sedimentology and stable isotopes. Sedimentary Geology , 78, 81–99. Platt, N.H., Wright, V.P., 1991. Lacustrine carbonates: Facies models, facies distributions and hydrocarbon aspects. In: Anadón, P., Cabrera, L., Kelts, K. (Eds.), Lacustrine Facies Analysis . International Association of Sedimentologists Special Publication , 13, pp. 57–74. Platt, N.H., Wright, V.P., 1992. Palustrine carbonates and the Florida Everglades: Towards an exposure index for the freshwater environment? Journal of Sedimentary Petrology , 62(6), 1058–1071. Plaziat, J.C., Freytet, P., 1978. Le pseudo-microkarst pédologique: un aspect particulier des paléo-pédogenèses développées sur les dépôts calcaires lacustres dans le tertiaire du Languedoc. Comptes Rendues Academie Science Paris , 286, 1661– 1664. ACCEPTED Read, C.B., Wilpolt, R.H., Andrews, D.A., et al ., 1944. Geologic map and stratigraphic sections of Permian and Pennsylvanian rocks of parts of San Miguel, Santa Fe, Sandoval, Bernalillo, Torrance, and Valencia Counties, north central New Mexico. U.S. Geological Survey, Oil and Gas Investigations, Preliminary Map 21. Retallack, G.J., 2001. Soils of the Past , 2 nd edition. Blackwell, Malden, MA, 404 pp. Retallack, G.J., 2005. Pedogenic carbonate proxies for amount and seasonality of precipitation in paleosols. Geology , 33, 333–336.

ACCEPTED MANUSCRIPT

Rinehart, L.F., Lucas, S.G., Tanner, L.H., Nelson, W.J., Elrick, S.D., Chaney, D.S., DiMichele, W.A., 2015. Plant architecture and spatial structure of an Early Permian woodland buried by flood waters, Sangre de Cristo Formation, New Mexico. Palaeogeography, Palaeoecology, Palaeoclimatology , 424, 91–110. Royer, D.L., 1999. Depth to pedogenic carbonate horizon as a paleoprecipitation indicator? Geology , 27, 1123–1126. Sanz, M.E., Alonso-Zarza, A.M., Calvo, J.P., 1995. Carbonate pond deposits related to semi-arid alluvial systems: Examples from the Tertiary Madrid Basin, Spain. Sedimentology , 42, 437–452. Schaetzl, R.J., Thompson, M.L., 2015. Soils: Genesis and Geomorphology , 2 nd edition. Cambridge University Press, New York, 777 pp. Soegaard, K., Caldwell, K.R., 1990. Depositional history and tectonic significance of alluvial sedimentation in the Permo- Pennsylvanian Sangre de Cristo Formation, Taos trough, New Mexico. New Mexico Geological Society Guidebook , 41, 277–289. Soreghan, G., Elmore, R., Lewchuck, M., 2002. Sedimentologic–magnetic record of western Pangean climate in Upper Paleozoic loessite (lower Cutler beds, Utah). GSA Bulletin , 114, 1019–1035. Tabor, N.J., Montañez, I.P., 2002. Shifts in Late Paleozoic atmospheric circulation over western equatorial Pangea: Insights from pedogenic mineral δ18 O compositions. Geology , 30(12), 1127–1130. Tabor, N.J., Montañez, I.P., 2004. Permo-Pennsylvanian alluvial paleosols (north-central Texas): High-resolution proxy records of the evolution of early Pangean paleoclimate. Sedimentology , 51, 851–884. Tabor, N.J., Montañez, I.P., Scotese, C.R., Poulsen, C.J., Mack, G.H., 2008. Paleosol archives of environmental and climatic history in paleotropical western Pangea during the latest Pennsylvanian through Early Permian. GSA Special Paper , 441, 291–303. Talbot, M.R., 1990. A review of the paleohydrological interpretation of carbon and oxygen isotopic ratios in primary lacustrine carbonates. Chemical Geology , 80, 261–279. Talbot, M.R., Holm, K., Williams, M.A.J., 1994. Sedimentation in low-gradient desert margin systems: A comparison of the Late Triassic of northwest Somerset (England) and the Late Quaternary of east-central Australia. In: Rosen, M.R. (Ed.), Paleoclimate and Basin Evolution of Playa Systems . GSA SpecialMANUSCRIPT Paper , 289, pp. 97–117. Talbot, M.R., Kelts, K., 1990. Paleolimnological signatures from carbon and oxygen isotopic ratios in carbonates from organic rich lacustrine sediments. In: Katz, B.J. (Ed.), Lacustrine Exploration: Case Studies and Modern Analogues . AAPG Memoir , 50, pp. 99–112. Tanner, L.H., 2000. Palustrine/lacustrine and alluvial facies of the Norian Owl Rock Formation (Chinle Group), Four Corners region, southwestern USA: Implications for Late Triassic paleoclimate. Journal of Sedimentary Research , 70(5), 1280–1289. Tanner, L.H., 2010. Terrestrial carbonates as indicators of palaeoclimate. In: Alonso-Zarza, A.M., Tanner, L.H. (Eds.), Carbonates in Continental Settings: Geochemistry, Diagenesis and Applications . Elsevier, Developments in Sedimentology 62, pp. 179–214. Vaughn, P.P., 1964. Evidence of aestivating lungfish from the Sangre de Cristo Formation, Lower Permian of northern New Mexico. Los Angeles County Museum Contributions to Science , 80, 1–8. Zaleha, M.J., 1997. Siwalik paleosols (Miocene, northern Pakistan): Genesis and controls on their formation. Journal of Sedimentary Research , 67(5), 821–839.

Zhu, L., Tabor, N.J., 2014. Lower Permian paleosol morphologies and paleoatmospheric pCO 2 estimates from north-central Texas, USA. GSA AbstractsACCEPTED with Program , 46(6), 131. (Edited by Xiu-Fang Hu)