Notes: Abstract Geologic mapping on a scale of 1:10000 and detailed stratigraphic studies of lava flows and tephra deposits of the Arenal-Chato volcanic system reveal a complex and cyclic volcanic history. This cyclicity provides insight into the evolution of magma batches during the growth of the andesitic volcanic system. The Arenal and Chato volcanoes have a central zone comprised of a lava armor and a distal zone comprised of a tephra apron. During Arenal's last two eruptive periods major craters formed near intersections of regional fractures at the lava armortephra apron transition. We suggest that such intersections are potential sites for future major explosions. The earliest rocks, i.e., the Chato lava flows, range in composition from basaltic andesite to andesite. These rocks, except for the andesitic domes of Chatito and La Espina, appear to have evolved from a common parental magma. The last active period of Chato volcano occurred 3550 B. P. The earliest known activity of Arenal volcano is 2900 B. P. Arenal lava flows have 54–56 wt% SiO2 and may be subdivided into a high-alumina group (HAG, Al2O3 = 20 wt%) and a low-alumina group (LAG, Al2O3 = 19 wt%). Compared to the HAG, the LAG also has smaller amounts of incompatible elements and higher amounts of FeO and MgO. Arenal tephra deposits were emplaced by Plinian-Sub-Plinian explosions occurring at 300±150-yr intervals. These deposits are compositionally zoned and alternate between dacite and basalt. The stratigraphy reveals an apparent magmatic cycle consisting of (a) dacitic-andesitic tephra, (b) HAG lava flows, (c) LAG lava flows, and (d) andesitic-basaltic tephra. This magmatic cycle is repeated four times during Arenal's history and is interpreted to have developed by the crystal fractionation and crystal redistribution of a single magma batch. The period of this cycle, and consequently the “life” of a magma batch, is about 800 years. If the cyclic pattern continues, a basaltic explosive phase may occur in the next 250 years.

Notes: Abstract Gabbroic enclaves ejected during the current eruption phase (A-1) and during the latest prehistoric eruption phase (A-2) of Arenal Volcano show systematic variations in texture, mineralogy and composition as a function of host rock chemistry and timing of eruption. The most differentiated enclaves occur in the more differentiated A-2 lavas. Enclaves in the A-1 volcanics are consistently less evolved. Within the current A-1 eruption, the most mafic enclaves are amphibole-bearing rocks that were erupted during the first 2–3 years of activity (1968–1970). These enclaves occur in the most differentiated A-1 volcanics and are not in equilibrium with their host rocks. They crystallized from a hydrous melt that was slightly more mafic than anything erupted during the current cycle. We interpret the enclaves as sidewall crystallization products of a melt, possibly a high-alumina basalt, that was immediately parental to the A-1 lavas. Enclaves that occur in A-1 rocks erupted after 1970 and all of the A-2 enclaves are amphibole-free and less mafic than the early A-1 enclaves. Their chemistry suggests that they formed during the early to intermediate crystallization of their host lavas. None of the enclaves contain minerals that might have equilibrated with a primary, mantle-derived melt. Geothermometry is consistent with geochemistry, with amphibole-bearing A-1 enclaves yielding the highest pyroxene temperatures (ave. 1090° C) and A-2 enclaves the lowest (ave. 1030° C). Geobarometry suggests mid- to upper crustal depths for the crystallization of all enclaves. The enclaves are cognate and reflect pre-eruptive crystallization of Arenal magmas. They record evolution from a hydrous, basaltic magma to the drier basaltic andesites that characterize the current eruption. Volatiles appear to have been lost due to depressurization during the slow ascent of the magmas through the upper levels of the crust following the initial explosive eruption. Volatile loss and depressurization resulted in the destabilization and the progressive resorption of amphibole. The A-2 lavas may represent the long-term fractionation products of basaltic andesite magmas similar in composition to the A-1 lavas. Anorthitic plagioclase, commonly thought of as a phase stabilized by high Ca/Na and high water pressure, continued to crystallize in a system with relatively low Ca/Na and which had dehydrated and/or depressurized to the point at which amphibole was no longer stable. This suggests that compositional characteristics other than high Ca/Na or high water content may have stabilized the anorthite in the basaltic and basaltic andesite melts at Arenal. We speculate that the high-alumina content of the Arenal magmas may be the stabilizing factor.