Geochronological Constraints on the Evolution and Petrogenesis of the Malwa Plateau Subprovince of the Deccan Traps

TL;DR

New 40Ar/39Ar plagioclase ages and geochemistry for 25 Malwa Plateau lavas show Deccan volcanism there began by 66.834 ± 0.075 Ma, at least ~400 ka before the oldest dated Western Ghats lavas, peaked at 66.4–66.3 Ma, and continued to 65.6 Ma. Malwa lavas are geochemically distinct from Western Ghats chemotypes, and REE modeling suggests lower-degree, lower-temperature, shallower melting consistent with an earlier, cooler stage of the Deccan plume.

Key findings

• 40Ar/39Ar dating of 21 lavas spanning the exposed stratigraphic extent shows eruptions began prior to 66.8 Ma; the oldest sample, KI18-1 of the Mandleshwar Formation, gives 66.834 ± 0.075 Ma (1σ analytical), at least 400 ka older than the oldest high-precision Western Ghats age of 66.413 ± 0.067 Ma (Sprain et al., 2019). Exposed southern Malwa lavas span ~700 m elevation and roughly 1 million years, with quasi-continuous eruptions for ~350,000 years leading up to the KPB; eruption rates peaked from 66.4 to 66.3 Ma and then diminished until 65.6 Ma, coinciding with the Late Maastrichtian Warming event.
• All but one analyzed sample are tholeiitic basalts or basaltic andesites with SiO2 49.3–54.2 wt% on a volatile-free basis; except topmost JNP-813 (TiO2 = 0.99 wt%, FeO = 10.8 wt%, MgO = 6.8 wt%), samples are high in TiO2 (2.3–3.3 wt%) and FeO (10.7–15.7 wt%) and low in MgO (3.8–6.9 wt%). One alkaline trachybasalt, MAL19-2 (66.51 ± 0.28 Ma), has strong LREE/HREE enrichment (La/Lu = 109.6).
• Discriminant-function analysis returns some Malwa correlations to Western Ghats chemotypes, but many samples do not reliably match any chemotype and several matches are out of stratigraphic order; Khandala chemotypes erupted earlier in Malwa (66.8–66.2 Ma) than in the Western Ghats (66.2–66.1 Ma), while Bushe-like lavas erupted slightly later in Malwa (65.9–65.8 Ma) than in the Western Ghats (66.1–66.0 Ma).
• REE modeling suggests Malwa lavas were derived by slightly lower degrees of partial melting, at lower mantle temperatures and depths, than Western Ghats lavas; Malwa thus appears to record an earlier, cooler stage of the Deccan plume’s evolution and continued erupting through a large part of the main Deccan province’s lifetime.

Methods

• Collected 25 lava samples from Malwa Plateau outcrops, predominantly in the southern region; 21 were selected for 40Ar/39Ar dating, and four were unsuitable due to plagioclase inclusions or alteration. Seven plagioclase megacrystic basalts had geochemistry previously presented by Marzoli et al. (2022).
• 40Ar/39Ar dating used plagioclase separates co-irradiated with Fish Canyon sanidine in four separate 7 hr irradiations in the CLICIT facility of the Oregon State University TRIGA reactor; 4–8 aliquots per sample were analyzed in 12–15 step incremental CO2 laser heating experiments at the Berkeley Geochronology Center using a MAP 215C mass spectrometer. Weighted mean plateau ages are based on the Renne et al. (2011) calibration, and age models were produced with Chron.jl using sample elevations and 2σ weighted mean plateau ages including systematic uncertainties.
• Whole-rock major and trace element data were obtained on all 25 samples by XRF and ICP-MS at the Washington State University GeoAnalytical Lab; data were compared with Western Ghats chemotypes via discriminant-function analysis in SPSS v.27 following Peng et al. (1998) and Vanderkluysen et al. (2011), and REE patterns were modeled with REEBOX PRO and EC-AFC.

Limitations

• Stated: the duration of the waxing and waning phases of eruption rate is poorly constrained because of unknown erosion and the undated lava section below the lowest dated flow at the Narmada River; two dated samples (KOL20-1 and SK20-1) cannot be confidently placed in the age model.
• Stated: current age resolution permits only coarse climate modeling via geochemical box models, and there remains disagreement between pulsed versus quasi-continuous eruptive tempo; intercalibration uncertainties between 40Ar/39Ar and U-Pb dating persist, with offsets depending on whether the Renne et al. (2011) or Kuiper et al. (2008) calibration is used.
• Evident: sampling is concentrated in the southern Malwa Plateau where sections are least faulted and expose the maximum elevation range; geochemical correlations rely on liberal DFA criteria (Mahalanobis distance <20), and the paper notes that differences in Ti, Zr/Y, and Nb/La cast doubt on almost every correlation. The study also does not directly bear on the pulsed versus quasi-continuous eruptive tempo of the LIP as a whole.