EVALUATING MAFIC HOLOCENE TEPHRA AS PARENTS OF EVOLVED ROCKS ON UNALASKA ISLAND, ALEUTIAN ISLANDS, AK

New major and trace element data provide additional constraints on the composition of parental basalts and the petrogenesis of andesitic lavas from Unalaska Island - located within the southwestern end of the Cold Bay volcanic-tectonic segment of the eastern Aleutian arc. Makushin volcano is an active stratovolcano comprised of basaltic to dacitic lavas that occupies the northwest extension of Unalaska Island. Holocene mafic tephra similar in composition to lavas previously proposed to be “parental” (Gust and Perfit, 1987) were sampled from Pakushin, a satellite cone approximately 7 Km southwest of Makushin. The scoria contain abundant phenocrysts of olivine and clinopyroxene and minor spinel, but only microlites of plagioclase within a vitric matrix. Whole-rock analyses yield anomalously mafic compositions with MgO ranging from 7.25 to 14.8 wt. %, Ni and Cr contents up to 214ppm and 1167ppm respectively and (La/Sm)_N from 1.7 to 1.9. Makushin samples come from two caldera-forming eruptions (8050 BP and ~8790 yr BP), an andesite scoria eruption (~9000 BP), and the Driftwood Pumice eruption (~ 8000 BP). Samples range from basaltic andesite to andesite with MgO ranging from 1.37 to 2.65 wt. %, SiO₂ from 55.6 to 61.2 wt. % and Ni and Cr less than 17ppm.

MELTS calculations suggest the andesites could created by ~ 50 to 68% fractional crystallization of high MgO melts with compositions like the Pakushin tephra at 3 to 1 kb in the order Olivine - Clinopyroxene - Plagioclase - Magnetite with initial water contents set less than 1 wt. % and temperatures ranging from 1355 to 1060 C°. However, in the more evolved Makushin samples linear major and trace element trends indicate the basaltic andesites may represent mixes between high-SiO₂ andesites and the mafic basalts. We propose a petrogenetic model whereby mixing of hot, primitive basaltic melt into resident andesitic magmas initiate explosive eruptions – a scenario similar to that proposed by Lerner et al. (2018) in Driftwood pumice. We also surmise that early high-T fractionation resulted in accumulation of olivine in some melts resulting in the most mafic Pakushin compositions. Phase chemical analyses are underway in order to evaluate MELTS and mixing models and to constrain possible links between each volcanic center.