Tracking crystal-melt segregation and magma recharge using zircon trace element data

The Cretaceous Yunshan caldera complex in SE China consists of an unusual coexisting assemblage of peraluminous and peralkaline rhyolites and a resurgent intra-caldera porphyritic quartz monzonite. In this study, we use zircon trace element data to study the compositional differences of zircons from cogenetic magmas and to track the evolution of the entire magmatic system. Our results indicate that the zircons from the peraluminous and peralkaline rhyolites formed from highly evolved compositions with high Hf concentrations and low Ti contents, and low Th/U and Zr/Hf ratios, which are distinct from those of the intrusive porphyritic quartz monzonite. Zircons from the peraluminous and peralkaline rhyolites display overlapping Zr/Hf and Hf, but the zircons from the peralkaline rhyolites have extremely low Eu/Eu* ratios (<0.1) and Ti contents (2.26–13.3 ppm). The lack of overlapping zircon trace element compositions between the volcanic and intrusive caldera units is interpreted to represent crystal–melt segregation processes. In addition, zircon grains from the porphyritic quartz monzonite and a few zircon grains from the peraluminous rhyolite display distinctly bright rims and whole grains in cathodoluminescence imaging, which have high Ti, Zr/Hf, and Eu/Eu*, and are similar to those of zircons from the mafic microgranular enclave within the porphyritic quartz monzonite. We interpret these signatures to reflect crystallization from a relatively hot and less evolved magma indicating a magma chamber recharge event. We further developed a model in which the magmas of the peraluminous and peralkaline rhyolites were successively extracted from a primitive crystal mush by crystal–melt segregation with a rejuvenation of the crystal mush after the extraction of the peraluminous rhyolitic melt, leaving behind residual mushes solidified as porphyritic quartz monzonite. Our study shows that trace element analyses of zircons can effectively be used to pinpoint multiple crystal–melt segregation and melt extraction events as well as magma recharge processes in silicic magmatic systems.