HIGH-PRESSURE AMPHIBOLE IN JURASSIC ARC MAGMAS: INSIGHTS INTO DEEP CRUSTAL MAGMATIC PROCESSES

A suite of distinctive, late Middle Jurassic garnet-bearing andesitic dikes crops out in the Klamath Mountain province south of ca. 41°15’ over an area of at least 10 x 35 km. The dikes are characterized by cm-scale plagioclase phenocrysts with Finisterre habit, mm- to cm-scale Al-rich amphibole, Ca-rich almandine-pyrope garnet and apatite ± quartz ± epidote ± ilmenite. Rutile is rare: preserved as microphenocrysts and inclusions in garnet. Garnet–rutile equilibrium suggests initial crystallization at >1.0 GPa, essentially at the base of the crust. Amphibole phenocrysts are aluminous (13.7–19.0 wt%) tschermakite and ferrotschermakite. Some grains are weakly zoned, whereas others display intricate concentric color zoning from colorless to blue-green to brown. Calculated temperatures (Putirka, 2016) range from 850–990°C. Ti, Sr, and Ba vary regularly with decreasing temperature, as does Eu/Eu*. In contrast, the heavy REE slope (Gd/Lu) is essentially constant above 900°C, but steepens (higher Gd/Lu) at T < 900°C. This change in slope probably marks precipitation of garnet.

Unlike amphibole in upper-crustal plutons, which shows regular core-to-rim zoning (e.g., Barnes et al., 2016), complexly zoned amphibole in the garnet andesites lacks regular core-to-rim zoning in any element or ratio. For example, the highest T values were calculated for crystal cores and crystal terminations. Similarly, high Ti contents and high Gd/Lu ratios are found in both core zones and rims, in the same grains. These zoning patterns indicate that deep-crustal reservoirs of garnet andesite magmas were locations of repeated magma mixing. Mixing was initially at T higher than garnet stability but ultimately involved garnet-bearing and garnet-free magmas.

Preservation of the high-P assemblage indicates rapid upward magma migration, leaving the open question: would the high-P assemblage or the geochemical signature of garnet fractionation be preserved after long-term mid-crustal storage?

Barnes et al. (2016) Am. Min. 101, 328–342. Putirka, K. (2016) Am. Min. 101, 841–858.