Abstract

Crystallization ages of igneous xenoliths entrained in a salt diapir in La Popa basin, northeastern Mexico, indicate the timing of magmatism in onshore Mexico salt basins and elucidate igneous rock-salt relations. Three equigranular phaneritic mafic to intermediate xenoliths have U-Pb zircon ages from 158 to 154 Ma (Oxfordian–Kimmeridgian). Phaneritic textures, hydrothermal alteration, zircon zonation, and previously published 40Ar/39Ar cooling ages from a nearby diapir, which are younger than Upper Jurassic strata overlying the salt, combine to suggest that these samples were intruded into salt and exhumed during diapirism. A porphyritic mafic rock with a U-Pb zircon age of 150 Ma (Tithonian) has a crystallization age coeval with marine strata overlying the salt. It is interpreted as a shallow intrusion into salt emplaced after the onset of salt diapirism. Magmatism thus took place from 158 to 150 Ma. The Oxfordian–Kimmeridgian ages of the three older xenoliths overlap with published biostratigraphic ages in carbonate strata intercalated with evaporites of the Minas Viejas Formation in the Sierra Madre Oriental south of Monterrey, Mexico. The entire time interval represented by the xenolith ages is coeval with mafic flows and pillow basalts in Border rift basins in northern Mexico and the southwestern United States. Comparison of the new ages with 40Ar/39Ar ages of igneous xenoliths from coastal Louisiana diapirs implies that rift magmatism began later in northeastern Mexico than in the Gulf of Mexico. The magmatic time interval indicated by the diapiric xenoliths is younger than accepted ages for Jurassic arc magmatism in northern and central Mexico and supports the inference that the Late Jurassic magmatism was not a result of subduction. The U-Pb ages and biostratigraphy are consistent in indicating a late Oxfordian–early Kimmeridgian salt age in northeastern Mexico, younger than the Louann Salt of the Gulf of Mexico. The contrasting salt ages indicate incursion of water into La Popa basin and adjacent onshore salt basins of northeastern Mexico from a fully marine Gulf of Mexico.

INTRODUCTION

The relative timing of salt deposition and synrift magmatism is important to understanding the history and mechanisms of basin formation during rifting. Nevertheless, magmatic history and salt age in the Gulf of Mexico, one of the most extensively studied rift basins in the world, and onshore Mexico salt basins remain poorly understood. Late Jurassic (160–159 Ma) 40Ar/39Ar dates on allochthonous volcanic xenoliths in salt diapirs were previously reported from coastal Louisiana (Stern et al., 2011), but the stratigraphic relationship of the xenoliths to the salt is uncertain. Interpretation of salt age in the Gulf of Mexico basin relies principally on the biostratigraphic age of overlying strata (e.g., Salvador, 1987), whereas the age of onshore Mexico salt is suggested by biostratigraphic age of carbonate blocks in diapirs of La Popa basin (Vega and Lawton, 2011) and carbonate strata interbedded or structurally interleaved in exposed evaporite in the Sierra Madre Oriental (Cross, 2012; Zell et al., 2015). Because the stratigraphic relations of evaporite and fossiliferous carbonate strata are somewhat equivocal, independent evidence of salt age is desirable for establishing tectonic models of salt deposition in the Gulf region and elsewhere.

The mechanisms of Jurassic magmatism in Mexico remain as yet unresolved. On the basis of the geochemical attributes of igneous xenoliths in salt diapirs, magmatism in La Popa basin of northeastern Mexico has been attributed to asthenospheric melting attendant upon continental extension (Garrison and McMillan, 1999). In contrast, an approximately linear trend of Early to Middle Jurassic calc-alkaline flows and ignimbrites interbedded with red beds in northern and eastern Mexico has been interpreted as a magmatic arc created by subduction of a pre-Farallon or proto-Farallon plate beneath western Mexico (Sedlock et al., 1993; Dickinson and Lawton, 2001; Bartolini et al., 2003; Barboza-Gudiño et al., 2008). In this context, the Gulf of Mexico basin has been interpreted as a backarc basin to the Jurassic arc (Stern and Dickinson, 2010). Improved age control, in addition to igneous geochemistry, will permit improved understanding of Jurassic magmatic systems of Mexico and the Gulf of Mexico.

In order to provide insight into salt-magmatism relations and timing of magmatism in northeastern Mexico, we present U-Pb ages from four igneous blocks entrained in a salt diapir in La Popa basin, one of several onshore salt basins that flank the Gulf of Mexico (Fig. 1). From sample ages and textural characteristics, we infer the original stratigraphic context of the igneous rocks and the salt. The ages of the xenoliths, which we interpret as having originated as intrusions into the salt, range in age from late Oxfordian to early Tithonian. The ages suggest that magmatism was a synchronous Late Jurassic event in the salt basins of northeastern Mexico and the basins of the Border rift in northern Mexico and southwestern United States but was largely younger than magmatism in the Gulf of Mexico basin. Moreover, the magmatism in La Popa basin was younger than arc magmatism in northern and central Mexico, thus permitting interpretation of different mechanisms for the two magmatic events.

GEOLOGIC SETTING AND STRATIGRAPHY OF SALT BASINS

La Popa, Monterrey, and Sabinas basins, in the states of Nuevo León and Coahuila, comprise the onshore salt basins of northeastern Mexico (Hudec et al., 2013). These basins are separated from the Gulf of Mexico by NW-trending basement blocks termed the Burro and Tamaulipas arches (Fig. 1). In the Sabinas, La Popa, and Monterrey salt basins, the evaporitic section is represented by the Minas Viejas Formation, formerly termed the Olvido Formation (Heim, 1940; Salvador, 1987; Kroeger and Stinnesbeck, 2003; Vega and Lawton, 2011; Zell et al., 2015). The Minas Viejas evaporites overlie a thick red-bed succession in the Monterrey salt basin where the section is well exposed in the Sierra Madre Oriental (e.g., Kroeger and Stinnesbeck, 2003) and in the Sabinas basin (Eguiluz de Antuñano, 2001), but the subsalt section is currently unknown in La Popa basin. In the Sierra Madre Oriental, conglomerate of La Joya Formation, at the base of the Minas Viejas Formation, unconformably overlies the red beds of La Boca Formation, which has a maximum depositional age of 163 ± 3 Ma on the basis of young detrital-zircon ages (Fig. 2; Callovian–early Oxfordian; Rubio-Cisneros and Lawton, 2011; Time Scale of Walker et al., 2012).

The Minas Viejas Formation of the Sierra Madre Oriental consists of structurally complex exposures that contain locally intact stratigraphic successions of interstratified gypsum, limestone, and siliciclastic strata, as well as alkalic mafic igneous rocks and at least one ignimbrite (Fig. 2; Kroeger and Stinnesbeck, 2003; Cross, 2012). Cross (2012) reported a composite stratigraphic thickness of ∼920 m, which he divided into five carbonate-dominated members and six evaporite-dominated members. Gypsum, shale, and limestone alike contain abundant characteristics of peritidal to supratidal depositional conditions, including chicken-wire structure, small oscillation ripples, desiccation polygons, birds-eye structure, peloids, thin cryptobiotic lamination, and locally abundant untransported oyster accumulations (Kroeger and Stinnesbeck, 2003; Cross, 2012). Ammonite and bivalve fossils, including the gryphaeid oyster Nanogyra virgula, recovered from carbonate strata in the upper ∼550 m of the Minas Viejas Formation indicate a late Oxfordian–early Kimmeridgian age for the carbonates (Cross, 2012; Zell et al., 2015). The Minas Viejas section in the Sierra Madre Oriental is gradationally overlain by sandy limestone that has long been considered as pertaining to the Zuloaga Limestone (e.g., Kroeger and Stinnesbeck, 2003), which is late Oxfordian in exposures to the west where salt is absent (Imlay, 1943). Such a correlation is thus invalidated by ammonites recovered from intercalated Minas Viejas carbonate strata; these ammonites indicate ages younger than the Zuloaga Limestone (e.g., Cross, 2012). Late early Kimmeridgian ammonites occur near the base of the suprajacent La Caja Formation (Cross, 2012).

La Popa basin contains a thick succession of post-salt Lower Cretaceous–Eocene strata exposed in mini-basins formed by salt migration (Fig. 3; Lawton et al., 2001). In La Popa basin, three diapirs—La Popa, El Gordo, and El Papalote—consist of exposed insoluble caprock of gypsum and uncommon anhydrite. The evaporite section consists of halite and gypsum in the subsurface, as indicated by a nearby petroleum exploration well (Lopez-Ramos, 1982). The caprock of the three diapirs contains exotic blocks with exposed dimensions ranging from 3 to 350 m. The blocks in the caprock include fossiliferous limestone and dolostone, stratigraphically intact, interbedded laminated carbonate mudstone and gypsum with chicken-wire structure, and blocks of igneous rocks. As illustrated by block distribution at El Gordo diapir (Fig. 4), the igneous blocks, which are somewhat more abundant near the margins of the diapirs, are termed xenoliths in this paper. Despite their occurrence in salt rather than an igneous host rock, this term is nonetheless consistent with the literal translation “foreign rock” and follows the precedent of Stern et al. (2011), who used the term xenolith to refer to similar inclusions within southern Louisiana diapirs. The limestone blocks contain an oyster, Nanogyra virgula, which is locally abundant and untransported, and is considered of early to middle Kimmeridgian age (Vega and Lawton, 2011). Fossil content in the transported blocks thus suggests a direct correlation of the diapiric evaporite with the Minas Viejas Formation of the Monterrey salt basin. Although the blocks of limestone in El Papalote diapir (Fig. 3) have been identified as Zuloaga Limestone (Laudon, 1984; Lawton et al., 2001), we use the term “Nanogyra limestone” (Figs. 2, 4) to indicate that they are younger than the Zuloaga Limestone.

Angular to rounded clasts of carbonate and igneous rocks derived from the diapir are present in Lower Cretaceous–Eocene strata flanking the diapirs and exposed along the length of an elongate welded former salt wall termed La Popa weld (Fig. 2; e.g., Rowan et al., 2012). The diapir-derived fragments indicate that the carbonate and igneous xenoliths were exposed at the crests of emergent diapirs throughout the Cretaceous–Paleogene depositional history of the basin and that they were eroded from the diapirs at least since the Aptian, the age of the oldest exposed strata flanking the salt structures. The evidence for long-term entrainment suggests that the xenoliths have been extensively mobilized and shuffled within the salt of the diapirs. The diapirs contain no clasts from of the flanking beds, nor do they contain fragments of red beds or basement rocks demonstrably older than the evaporite. The absence of pre-salt or post-salt rocks in the diapirs led Vega and Lawton (2011) to infer that the Kimmeridgian limestone blocks represented strata interbedded with the youngest evaporite, a conclusion consistent with geologic relations of the exposed Minas Viejas Formation in the Monterrey salt basin of the Sierra Madre Oriental.

In Jurassic stratigraphic sections west of the salt basins, the Upper Oxfordian Zuloaga Limestone or equivalent La Gloria Formation overlies red-bed successions with intercalated volcanic rocks termed the Nazas Formation. The red beds of the Nazas Formation, as young as 169 ± 1 Ma (Fig. 2; Lawton and Molina-Garza, 2014), have been interpreted as the record of a Jurassic magmatic arc in central Mexico (Bartolini, 1998; Bartolini et al., 2003). La Gloria Formation contains a basal conglomerate containing volcanic and sandstone clasts of the Nazas Formation. The Zuloaga Limestone and La Gloria Formation, which contain late Oxfordian ammonites, are overlain by middle Kimmeridgian shale of La Casita Formation, indicating the presence of an early Kimmeridgian hiatus (Imlay, 1943). Salt is absent in these western sections, and thus its stratigraphic relation to these sections remains unknown.

In the northern Gulf of Mexico and the onshore salt basins of the United States, the Louann Salt lacks biostratigraphic indicators; therefore, bracketing stratigraphic relations provide the basis for salt age (Fig. 2). The salt overlies Mesozoic continental deposits of the Eagle Mills Formation, of uncertain upper age, and underlies thin continental and eolian strata of the Norphlet Formation and the upper Oxfordian Smackover Formation, a Zuloaga equivalent, which provides the upper limit on salt age (e.g., Imlay, 1943; Hudec et al., 2013; Olson et al., 2015). On the basis of plate reconstructions that limit the area of the Gulf of Mexico basin available for salt deposition in the Louann and Campeche basins prior to Oxfordian time, it is likely that salt age extends into the early Oxfordian (Pindell, 1985; Pindell and Kennan, 2009). Salt age in the Louann and Campeche salt basins is therefore regarded as late Middle Jurassic (Callovian; Salvador, 1987) to early Late Jurassic (Oxfordian; Pindell and Kennan, 2009). Stratigraphic relations and biostratigraphy thus indicate that the youngest salt, probably early Kimmeridgian, in onshore Mexico is younger than the youngest Louann salt, but stratigraphic relations are equivocal regarding the age of basal salt, which could be as old as Callovian (e.g., Vega and Lawton, 2011). Bromide concentrations measured in subsurface salt samples from both the Gulf of Mexico and onshore Mexico salt basins are consistent with precipitation of both halite and anhydrite from marine water (Land et al., 1988).

Extending northwest as far as the southwestern United States from the onshore Mexico salt basins, the Border rift constitutes a series of Late Jurassic marine basins (Fig. 1; Lawton and McMillan, 1999). Although these basins lack Jurassic salt, several of them contain mafic intrusions, lava flows, and pillow lavas, intercalated with Kimmeridgian marine shale, that are compositionally similar to the igneous rocks described here (Fig. 2; Lawton and Olmstead, 1995; Lawton and McMillan, 1999; Olmstead and Young, 2000; Villaseñor et al., 2005). Although well exposed in southeastern Arizona, flows and sill-like intrusions are difficult to differentiate because the interior parts of thick flow intervals, distinguished by pillowed bases and tops, commonly have phaneritic diabasic textures. Samples of mafic lavas and intrusions in Jurassic strata of southeastern Arizona have high Nb and TiO2, and low Zr, and also have εNd values near +4 at 150 Ma, consistent with an asthenospheric source for the magmas, and thus with magmatism during continental rifting (Lawton and McMillan, 1999; McClain et al., 2001; McMillan et al., 2001). The Border rift has been interpreted as an aulacogen developed during Gulf of Mexico rifting (Mickus et al., 2009; Stern and Dickinson, 2010); nevertheless, the Late Jurassic age of the Border rift postdates the age of rifting in the Gulf of Mexico and is instead time equivalent with seafloor spreading in the Gulf of Mexico basin (Pindell and Kennan, 2009).

METHODS

Zircons were analyzed for U-Pb ages using sensitive high-resolution ion microprobe (SHRIMP) at the Stanford–U.S. Geological Survey facility with the methods detailed in Williams (1998) and Wang et al. (2014). Isoplot (Ludwig, 2008) and SQUID 2 (Ludwig, 2005) programs were used for reduction and plotting of U-Pb data. Cathodoluminescence (CL) images and rare-earth element (REE) concentrations of the zircons were obtained for the analyzed grains. Additional analyses were obtained with single-collector laser ablation–inductively coupled plasma mass spectrometry (LA-ICPMS) at the University of Arizona LaserChron Center, using the methods of Gehrels (2012). Ages are reported in the text at the 2σ level, and the mean square of weighted deviates (MSWD) is included as a measure of the uniformity of the population.

Whole-rock geochemical analyses were performed using X-ray fluorescence (XRF) spectroscopy at New Mexico State University, using a Rigaku ZSX wavelength-dispersive spectrograph equipped with an end-window Rh target X-ray tube. Reference materials (BHVO-1 and BHVO-1P) were measured before and after all unknown analyses. Accuracy on major-element concentrations is generally within 1.1%.

Strontium and Nd isotopic compositions of two samples were obtained by isotope-dilution thermal ionization mass spectrometry at the University of Colorado at Boulder. Rock powder was dissolved in HF and perchloric acid in Teflon beakers. Rubidium, Sr, Sm, and Nd were separated and their concentrations and isotopic compositions were determined using the techniques described in Farmer et al. (1991). Isotope-dilution determinations are accurate to 0.5% for Sm and Nd. The total procedural blank averaged 100 pg for Nd during the study period. Eleven measurements of the La Jolla standard yielded 143Nd/144Nd = 0.511829 ± 8 (2σ mean). All errors in measured isotopic ratios are at the 95% confidence limit, corrected for mass fractionation by normalizing to 146Nd/144Nd = 0.72190.

Some mineral identifications were made using X-ray diffraction of powdered bulk sample at the New Mexico Bureau of Geology and Mineral Resources.

SAMPLE DESCRIPTION, GEOCHEMISTRY, AND GEOCHRONOLOGY RESULTS

Igneous xenoliths at La Popa and El Gordo diapirs are present throughout the exposures of the evaporite bodies, but they are more common along the margins (Figs. 3 and 4). Xenoliths are typically extensively fractured, lack chilled margins or pillow structures, and where exposed, their margins consist of a mixture of evaporite and brecciated igneous rock (Garrison and McMillan, 1999). Xenolith size ranges from extensively comminuted small blocks tens of cm long to giant blocks with long dimensions of ∼140 m (Fig. 4). Textures of greenish-gray mafic xenoliths are aphanitic, porphyritic, and phaneritic; vesicles are present in some aphanitic and porphyritic xenoliths. Altered marginal zones several cm thick of the xenoliths at El Papalote diapir were termed reaction aureoles and contain decreasing concentrations of SiO2, Al2O3, Fe2O3, TiO2, Na2O, MnO, and MgO but increasing CaO, with distance into the block (Garrison and McMillan, 1999). The reaction aureoles are continuous along the block margins and were interpreted as a result of interaction with diapir evaporite following xenolith entrainment (Garrison and McMillan, 1999). Unambiguous original block-evaporite contacts were not observed and are evidently rare or absent.

Four altered igneous xenoliths from El Gordo diapir were analyzed by SHRIMP and/or LA-ICPMS methods (Figs. 5 and 6; Tables 13). Internal zonation of zircon crystals is complex (Fig. 7). For most samples, the high-U rims were avoided. Sample PPP-142 is a fine-grained phaneritic rock with plagioclase, tremolite, muscovite, and chlorite. It has miarolitic cavities lined with epidote. It has SiO2 of 49.4 wt% and 7.7% MgO (Table 4). This is classified as a diorite based on its mafic composition, presence of amphibole, light color, and indeterminate original plagioclase composition. Rocks of this composition and texture in El Papalote diapir were likewise considered diorites (Garrison and McMillan, 1999). Sample PPP-142 has an initial 87Sr/86Sr of 0.7058 and εNd of 1.5 (Table 5). SHRIMP analysis of ten zircon crystals (Table 2) yielded a weighted-mean 238U/206Pb age of 158.6 ± 3.0 Ma (n = 8; two grains at 145.6 Ma and 168.8 Ma were not included in the mean age calculation). Zircon crystals were also analyzed by LA-ICPMS (Table 3). Extremely high U (up to 6000 ppm) was correlated with older ages (Fig. 8). The weighted-mean 238U/206Pb age of zircon grains with <2650 ppm U is 158.0 ± 1.7 Ma (n = 21); this is the preferred age for the sample.

Sample PPP-146 is an altered mafic rock with phaneritic texture and a dioritic composition. It has plagioclase, hornblende, chlorite, and clinozoisite lining miarolitic cavities. The SiO2 is 49.8 wt%, and MgO is 7.7% (Table 4). SHRIMP dating yielded a weighted-mean 238U/206Pb age of 153.9 ± 1.7 Ma (n = 10; Table 2).

Sample PPP-148 is a phaneritic tonalite with plagioclase, quartz, minor microcline, hornblende, chlorite, and epidote grown interstitially and in miarolitic cavities. It is fresher than the other two samples, and the most silicic sample analyzed, with SiO2 at 58.0 wt% and MgO at 1.6% (Table 4). It has an initial 87Sr/86Sr ratio of 0.7067 and εNd of 2.4 (Table 5). SHRIMP analyses (n = 10) yielded a weighted-mean 238U/206Pb age of 157.6 ± 1.8 Ma (Table 2). As with sample PPP-142, the LA-ICPMS data indicate correlation of high-U content and older ages (Fig. 8; Table 3). The lower-U zircon grains yielded a weighted-mean 238U/206Pb age of 154.7 ± 1.6 Ma (n = 24), our preferred age.

Sample PPP-147 is an altered porphyritic igneous rock with plagioclase phenocrysts in a matrix of plagioclase and unidentifiable fine-grained crystalline material. Minor epidote and chlorite are present. It has SiO2 of 52.4 wt% and moderate alkalis, classifying it as a basaltic andesite (Table 3). LA-ICPMS analysis yielded two ages of 301 Ma and 166 Ma that are older than the main population and not included in the mean age (Table 3). The remaining 13 analyses have a weighted-mean 238U/206Pb age of 149.5 ± 2.4 Ma. Unlike the other samples, the zircon crystals have low-U concentrations (Fig. 8).

DISCUSSION

Rift Magmatism and Comparison with the Border Rift

Xenolith lithology and chemistry are consistent with emplacement of the rocks in a continental rift (e.g., Garrison and McMillan, 1999). The diorite and basaltic andesite have a limited range of mafic compositions from 49–52 wt% SiO2 with relatively high MgO and Cr. The tonalite is more silicic. The Sr isotopic compositions may have been affected by postcrystallization processes such as Rb mobilization. Samarium and Nd are typically less mobile; thus, the positive εNd values of 1.5–2.4 and the mafic-intermediate compositions indicate that the samples were not derived exclusively from remelted continental crust. The lithology and geochemistry therefore indicate that at least some of the magma was derived from the mantle, likely generated in a continental rift zone consistent with extension in the Gulf of Mexico region and the Border rift (e.g., Garrison and McMillan, 1999; Lawton and McMillan, 1999; Pindell and Kennan, 2009; Hudec et al., 2013).

The U-Pb zircon data from the phaneritic samples are complex, mainly because of unusually high U concentrations in zircon. High U could have resulted from late-stage crystallization in a low-pressure hydrothermal environment, consistent with epidote growth in miarolitic cavities in the diorite samples (e.g., Candela, 1997), pervasive alteration, and complex zircon growth patterns such as high-U rims, sector zonation, or growth in fractures (Fig. 7). The zonation (e.g., Corfu et al., 2003), together with the other observations, provides evidence that the magma may have been intruded into the salt directly, creating a hydrothermal environment. Typically, high-U zircon experiences excessive radiation damage leading to Pb loss (e.g., Mezger and Krogstad, 1997), resulting in a negative correlation between U concentration and age. However, the ages in these samples, particularly the ones with the highest U concentrations (PPP-142; Fig. 8), are positively correlated with U. This is consistent with the findings of White and Ireland (2012) that SHRIMP analysis of higher-U zircon yielded erroneous ages related to instrumental fractionation or matrix effects. Alternatively, the positive correlation could be explained by the addition of Pb to the zircon (e.g., Williams et al., 1984) or removal of U (less likely, given the high U concentrations); either of these could result in the same relationship. Regardless of the actual cause, there is no geologic explanation that would result in ages, correlated to U concentrations, ranging semi-continuously up to 40 m.y. older than the main population in the same sample. Thus, considering the ages from zircons with <2650 ppm U, the preferred weighted-mean ages of the three plutonic samples are 158–154 Ma, and the single porphyritic sample is 150 Ma.

The U-Pb ages reported here indicate a duration of magmatism from 158 to 150 Ma, or 8 m.y. (±∼2–4 m.y.). Confidence that this age range represents the time span of magmatism is enhanced by long-term, entrainment-related block mixing implied by the occurrence of igneous clasts in diapir-flanking strata that span the Lower Cretaceous–Eocene section. Even with ages on only four different samples, extensive mixing of xenoliths in the diapir over ∼70 m.y. does not predict a block assemblage with a limited age span unless there were no other ages in the xenolith assemblage. The U-Pb ages are 4–12 m.y. older than previously reported 40Ar/39Ar biotite dates of ca. 147–146 Ma from two diapir xenoliths in El Papalote diapir reported by Garrison and McMillan (1999), who regarded the dates as metamorphic ages established when the samples cooled below 300 °C, probably due to diapiric exhumation following magma emplacement. The U-Pb ages are also younger than the 40Ar/39Ar dates of 160–159 Ma reported by Stern et al. (2011). If a similar time lag exists between magma crystallization and the 40Ar/39Ar ages reported by Stern et al. (2011), the age difference between the Louisiana and La Popa igneous xenoliths would be even greater and consistent with earlier magmatism in Louisiana than in the onshore Mexico basins. Indeed, equivalence of the Louisiana block ages to the biostratigraphic late Oxfordian age of post-salt Smackover strata (Fig. 2) suggests that either (1) those 40Ar/39Ar ages are cooling ages, or (2) the magma was intruded into the salt and crystallized during deposition of the overlying Smackover Formation.

Jurassic magmatism in Mexico and the Gulf of Mexico region likely resulted from more than one plate tectonic process and consequently remains a topic of debate. In addition to the generally mafic magmatic rocks described above from the Gulf of Mexico region, the belt of Lower to Middle Jurassic calc-alkaline magmatic rocks of northern and eastern Mexico is generally associated in time with development of the Gulf of Mexico basin (e.g., Barboza-Gudiño et al., 1998, 2008; Bartolini, 1998; Stern and Dickinson, 2010). The magmatic trend has been termed the Nazas arc by some previous workers (Dickinson and Lawton, 2001; Bartolini et al., 2003) and more recently interpreted as a transtensional backarc domain (Martini and Ortega-Gutiérrez, 2016). The time span of Nazas magmatism, as evaluated from direct dating of ignimbrites and andesite flows, as well as the age range of detrital zircons in associated sandstones, ranged from ca. 200 to ca. 165 Ma (Earliest Jurassic–middle Callovian; Barboza-Gudiño et al., 2008; Godínez-Urban et al., 2011; Rubio-Cisneros and Lawton, 2011; Lawton and Molina-Garza, 2014). The diapiric xenoliths described here, in addition to having a different geographic distribution than the Nazas rocks, are younger than volcanic rocks ascribed to the Nazas arc and have different chemical composition than Nazas igneous rocks (e.g., Garrison and McMillan, 1999). Therefore, the subduction-related magmatic rocks of the Early and Middle Jurassic are for the most part older than, and tectonically unrelated to, the continental rift magmatic rocks described here.

Implications for Age of Salt Deposition

The original stratigraphic relation of igneous xenoliths to salt in diapirs of the Gulf of Mexico region, whether plucked subsalt basement, intercalated flows, or intrusions into salt (e.g., Garrison and McMillan, 1999), has never been satisfactorily established. On the basis of equivalence of U-Pb ages with previously posited biostratigraphic ages of the evaporites, as well as the igneous rock textures themselves, we interpret the phaneritic rocks as emplaced directly into the salt during or after its deposition. Pristine contacts of the igneous rocks and evaporite are not preserved, precluding direct observation of primary textural evidence of magma-evaporite interaction. Nevertheless, emplacement of shallow intrusions even after diapir initiation is required by the post-salt crystallization age of the porphyritic xenolith (Fig. 2). Intrusion of the youngest porphyritic sample into mobile salt is suggested by its crystallization age equivalent to part of a thick succession of post-salt Tithonian marine strata (Fig. 2), which likely provided the stratigraphic load that initiated diapirism. The age of that sample (PPP-147) indicates that magmatism occurred at least until 150 Ma.

We infer that salt deposition began before 158 Ma, in order to host the intrusions, and ended at approximately the onset of diapirism near 150 Ma. On the strength of the diapiric mixing mechanism described above, we predict that older crystallization ages will not be encountered by future studies. A post-Callovian age for the base of salt is consistent with the earliest Oxfordian maximum depositional age reported from the upper part of underlying La Boca Formation in the Sierra Madre Oriental (Rubio-Cisneros and Lawton, 2011). The absence of pre-Oxfordian igneous ages can be interpreted to indicate either that (1) magmatism postdated older (Callovian–early Oxfordian?) salt deposition, or (2) salt had not yet been deposited prior to late Oxfordian time to host the intrusions, in which case the entire thickness of the onshore Mexico salt is younger than the Louann Salt of the Gulf of Mexico basin.

We interpret the U-Pb crystallization ages of the La Popa basin xenoliths, together with the stratigraphic context of the onshore salt deposits, to indicate that the salt age of the onshore Mexican salt basins is younger than the Louann Salt of the northern Gulf of Mexico and onshore U.S. salt basins, as inferred by some previous workers (e.g., Longoria, 1984; Salvador, 1987). Continued salt deposition into the Kimmeridgian in the onshore Mexico basins further indicates that the limestone strata overlying the Minas Viejas Formation cannot represent the Smackover-equivalent, upper Oxfordian Zuloaga Limestone, as previously inferred (Kroeger and Stinnesbeck, 2003). The age of the Smackover Formation requires the Louann Salt to be older than middle Oxfordian, whereas the late Oxfordian–early Kimmeridgian upper age of salt in La Popa basin temporally overlaps Kimmeridgian marine shales in the Border rift that contain mafic volcanic rocks (Figs. 1 and 2). The salt of the onshore Mexican salt basins records the initial influx of marine water, the source of the salt (e.g., Land et al., 1988), into a rift or pull-apart basin system formerly isolated from the Gulf of Mexico by NW-trending basement blocks (Fig. 1). A Gulf of Mexico, as opposed to Panthalassan, or Pacific, source for the marine water that fed the onshore salt basins is indicated by a dominant Gulf of Mexico affinity of the mollusks in the Minas Viejas carbonate strata (Zell et al., 2015). The salt of the onshore Mexican salt basins thus bears a relationship to the Louann Salt analogous to that of Holocene salt pans with intercalated mafic volcanic rocks within the Afar Depression of East Africa to Miocene salt of the Red Sea rift basin, which is separated from the depression by a basement arch (e.g., Bosworth et al., 2005).

CONCLUSIONS

Four igneous blocks with phaneritic and porphyritic textures, entrained in a salt diapir in La Popa basin, yielded six U-Pb zircon ages that range from 158 to 150 Ma (late Oxfordian–Tithonian), which we infer to be the age range of rift magmatism in northeastern Mexico. This age range correlates with mantle-derived mafic magmatism in the Border rift basins of northern Mexico and the southwestern United States but is likely younger than magmatism in the greater Gulf of Mexico on the basis of previously reported 40Ar/39Ar ages from igneous xenoliths in Louisiana salt domes. The Louisiana ages are equivalent to the biostratigraphic age of the post-salt Smackover Formation and could represent either crystallization ages of intrusions into the salt or cooling ages, analogous to 40Ar/39Ar ages previously reported from igneous blocks from a different La Popa basin diapir.

The igneous blocks originated as shallow intrusions emplaced into the salt during and shortly following deposition of the salt, possibly concomitant with early salt diapirism initiated by the load of suprajacent Upper Jurassic marine strata. No obvious pre-salt xenoliths have been dated. The new U-Pb ages corroborate previous inferences from fossil content of associated carbonate strata that salt deposits of the Minas Viejas Formation in onshore salt basins of northeastern Mexico are of late Oxfordian–early Kimmeridgian age. The salt of the onshore Mexico salt basins is therefore at least in part younger than the Louann Salt, which is older than early Oxfordian on the basis of overlying stratal ages. The salt of northeastern Mexico formed by spilling of marine water from the adjacent Gulf of Mexico over or through a basement barrier after fully marine conditions were established in the Gulf.

The mafic to intermediate magmatism recorded by the diapir xenoliths is spatially and geochemically distinct from Early to Middle Jurassic calc-alkaline subduction-related magmatism in northern and eastern Mexico, which ended ca. 165 Ma. The ages presented here thus indicate that the magmatism associated with the onshore salt basins of northeastern Mexico is also younger than the subduction-related magmatism.

ACKNOWLEDGMENTS

We thank Jennifer Garrison, Gary Gray, Kate Giles, Nancy McMillan, and Mark Rowan for discussions of diapir block origins, and Jim Pindell for insights regarding salt structure in the Gulf of Mexico. M. Coble, E. Gottlieb, J. Wooden, L. Farmer, S. Bothern, and V. Lueth helped acquire data. Arizona LaserChron Center is supported by National Science Foundation grantEAR-1338583. Partial financial support was provided by UNAM PAPIIT grant IN105714. Reviews by Michael Hudec, James Pindell, and Robert Stern improved manuscript focus and clarity.