Onset and tempo of ignimbrite flare-up volcanism in the eastern and central Mogollon-Datil volcanic field, southern New Mexico, USA

Abstract The Cenozoic ignimbrite flare-up (40–18 Ma) generated multiple volcanic fields in the southwestern United States and northern Mexico resulting from asthenospheric mantle upwelling after removal of the Farallon slab. The correlation of tuffs to one another and to source calderas within these volcanic fields is essential for determining spatiotemporal patterns in volcanism and magma geochemistry, which have been used to deduce migration of the Farallon slab at depth and associated mantle melting. However, the correlation of Eocene–Oligocene tuffs in the southwestern U.S. is difficult because of post-emplacement erosion and faulting. This study focuses on spatiotemporal patterns of the initial episode of ignimbrite flare-up activity (ca. 36.5–33.8 Ma) in the Mogollon-Datil volcanic field in south-central New Mexico, USA. We show that alkali feldspar major and trace element geochemistry is an effective tool for correlating tuffs when combined with high-precision, single-crystal 40Ar/39Ar geochronology and bulk-rock geochemistry. Using these data, we correlate several tuff units and differentiate other tuffs that have the same eruption age but very different geochemistry, and we conclude that there was a broadly northwestward migration in volcanism over time. The new tuff correlations are used to investigate spatiotemporal variations in magma geochemistry, erupted volumes, and recurrence intervals during the initial episode of Mogollon-Datil volcanic field volcanism. Early-erupted tuffs restricted to the eastern Mogollon-Datil volcanic field share similarities with western U.S. topaz rhyolites, which suggests that the silicic magmas were generated by partial melting of mafic lower crustal rocks. We also find differences in the compositions, crystallinities, and mineral assemblages between the early- and late-erupted tuffs. The early-erupted tuffs tend to have single-feldspar mineralogies, lower feldspar Or contents, large negative Eu anomalies, and low-whole–rock Ba concentrations. Conversely, late-erupted tuffs have two feldspar plus quartz assemblages, lesser Eu anomalies, higher whole-rock Ba concentrations, and feldspars have higher Or contents. Thus, we suggest that for some of the early eruptions, after magmas underwent crystal fractionation in the crust, the silicic melt largely separated from the crystalline mush prior to eruption, whereas late-erupted tuff magmas underwent crystal fractionation at near the eutectic minimum and were remobilized and erupted with a larger proportion of their crystalline mush. Using our new ages, correlations, and previously published data, we find that the initial phase of Mogollon-Datil volcanic field volcanism produced at least 15 eruptions between 36.5 Ma and 33.8 Ma, with a minimum total erupted volume of ~1350 km3 and an average recurrence interval of 170 k.y. However, eruptions were generally smaller in volume (most <15 km3) than in other coeval fields, and most eruptions (n = 11) occurred in the first 1.2 m.y. (ca. 36.5–35.3 Ma) of activity. Altogether, our work sheds new light on variations in the composition, timing, and migration of volcanism during the initial phase of Mogollon-Datil volcanic field activity and highlights the utility of feldspar geochemistry in both “fingerprinting” tuffs and elucidating magma evolution.