Rear-arc vs. arc-front volcanoes in the Katmai reach of the Alaska Peninsula: a critical appraisal of across-arc compositional variation

Physical and compositional data and K-Ar ages are reported for 14 rear-arc volcanoes that lie 11-22 km behind the narrowly linear volcanic front defined by the Mount Katmai-to-Devils Desk chain on the Alaska Peninsula. One is a 30-km(3) stratocone (Mount Griggs; 51-63% SiO2) active intermittently from 292 ka to Holocene. The others are monogenetic cones, domes, lava flows, plugs, and maars, of which 12 were previously unnamed and unstudied; they include seven basalts (48-52% SiO2), four mafic andesites (53-55% SiO2), and three andesite-dacite units. Six erupted in the interval 500-88 ka, one historically in 1977, and five in the interval 3-2 Ma. No migration of the volcanic front is discernible since the late Miocene, so even the older units erupted well behind the front. Discussion explores the significance of the volcanic front and the processes that influence compositional overlaps and differences among mafic products of the rear-are volcanoes and of the several arc-front edifices nearby. The latter have together erupted a magma volume of about 200 km(3), at least four times that of all rear-arc products combined. Correlation of Sr-isotope ratios with indices of fractionation indicates crustal contributions in volcanic-front magmas (0.7033-0.7038), but lack of such trends among the rear-arc units (0.70298-0.70356) suggests weaker and less systematic crustal influence. Slab contributions and mantle partial-melt fractions both appear to decline behind the front, but neither trend is crisp and unambiguous. No intraplate mantle contribution is recognized nor is any systematic across-are difference in intrinsic mantlewedge source fertility discerned. Both rear-arc and arc-front basalts apparently issued from fluxing of typically fertile NMORB-source mantle beneath the Peninsular terrane, which docked here in the Mesozoic.