Temporally variable crustal contributions to primitive mantle-derived Columbia River Basalt Group magmas

Abstract

The Steens Formation is one of the earliest and most primitive (>7 wt. MgO) eruptive products of the Columbia River Basalt Group (CRBG) and the CRBG-Yellowstone-Snake River large igneous province. New major-, trace-, and highly siderophile-element abundance and 87Sr/86Sr, 143Nd/144Nd and 187Os/188Os data are reported for the lower and upper Steens Formation to examine likely mantle sources and the nature of magmatic differentiation and crustal contamination acting on lavas. Examined Steens Formation basalts are relatively mafic (7–9 wt% MgO), incompatible trace element enriched, and have weaker Nb and Ta anomalies compared to other CRBG lavas. The most primitive basalts have isotopic compositions at the time of crystallization consistent with originating from a mantle source that was relatively depleted (87Sr/86Sr = ~0.7033; εNdi = ~ + 6.5; γOsi = ~ + 1). Primary magma compositions for the Steens Formation do not provide compelling evidence for a subducted slab component, instead suggesting derivation from primitive mantle sources more similar to those of other Mesozoic continental flood basalts (CFB; e.g., Deccan, North Atlantic Igneous Province). Onset of sulfide saturation in the CRBG occured at lower MgO (<7 wt%) than in other CFB (~ 8 wt%) leading to the high Os contents in the Steens Formation. Collectively, the Steens Formation exhibits decreasing Os contents, εNdi values and increasing 187Os/188Os with decreasing MgO. These geochemical signatures are consistent with increasing crustal contamination to parent melts with time, a feature that is also shared for the CRBG as a whole. Calculations based on Os and Nd isotopes of likely mantle and crust components to different formations of the CRBG indicates a progressive increase in the quantities of crustal contamination from ~1 to 2% from Steens Formation magmatism to more than 6% during Grande Ronde and Wanapum eruptions. These results would indicate increasing crustal contamination and enhanced potential cryptic degassing of CO2 in the later, more voluminous stages of CRBG magmatism, after ~16.5 Ma. Unless mantle-derived melts can produce sufficient greenhouse gas emission, there is likely an offset between the inception of the mid-Miocene Climatic Optimum at 17 Ma and maximum CO2 release, indicating that CRBG eruption was a contributing factor to climate change at that time, but was not the trigger for it.