Differentiation Timescales and Processes for Products From Recent Eruptions of Arenal Volcano, Costa Rica (1968-1999)

Arenal Volcano in Costa Rica has continuously erupted since 1968 exhibiting a variety of eruption behaviors and evolving in a complex interplay between crystal fractionation, magma mixing, degassing, and wall-rock interaction (Reagan et al., 1987; Cigolini, 1998). We performed a pilot study of trace-element variations and U-series disequilibria in five whole rocks and, in three cases, mineral separates over the course of the eruption from 1968 to 1999. Mineral phases (pyroxene, plagioclase, magnetite, and glass fractions) from a 1970 and two freshly collected 1999 rock samples were analyzed for U-Th disequilibria by TIMS and a complete suite of trace elements by ICP-MS. Plagioclase from the 1999 samples was analyzed for ²¹⁰Pb-²¹⁰Po disequilibria. (²¹⁰Pb)/(²¹⁰Po) activity ratios at the time of collection (9/15/99) are ~2 for both 1999 plagioclase separates. Both samples produce the same calculated date of initial plagioclase crystallization (3/5/99) assuming complete ²¹⁰Po degassing. The presence of (²¹⁰Pb)/(²¹⁰Po) disequilibria in plagioclase either implies that the mm-sized plagioclase crystals grew within a very short time of eruption or that ²¹⁰Po continuously diffuses out of already formed plagioclase crystals right up until the time of eruption from a Po-degassed magma. The observed ²³⁸U-²³⁰Th disequilibria are consistent with the differentiation process operating on a timescale much less than that of ²³⁰Th decay. All measured (²³⁰Th)/(²³²Th) activity ratios are identical indicating that mineral formation in both early and late eruptions (1970, 1999) is instantaneous with respect to ²³⁰Th and that none of the minerals are older than ~10⁴ years (they are not pre-Holocene "xenocrysts" or restites). (²³⁴U)/(²³⁸U) disequilibria results generally indicate that post-eruptive alteration is not a problem (with the exception of one pyroxene sample). Despite the (²³⁰Th)/(²³²Th) activity ratios remaining constant, Th/U consistently changes from low values in the early samples to a constant, higher value from 1986 to present. This implies that mantle-derived magmas entering the Arenal magma chamber through time have changed from more ²³⁸U-enriched to less ²³⁸U-enriched, or that assimilation has changed the Th/U but not the (²³⁰Th)/(²³²Th) activity ratios. Results from our trace elements analyses show that as Th/U increases with time, LREE/MREE ratios increase slightly indicating addition of a more enriched magma, consistent with the model of Reagan et al. (1987). LREE/MREE ratios, however, have not changed after 1986. Absolute abundances do show important differences between 1986 and 1999 samples with the 1986 sample more differentiated than both 1999 samples. The more differentiated 1986 sample may reflect slowing replenishment activity in the Arenal magma chamber and domination by differentiation processes leading up to 1986. Following 1986, absolute abundances decrease which suggests increasing replenishment rates and a decreasing role for differentiation mechanisms.