ZR/SR RATIOS DISTINGUISH COOL & WET FROM HOT & DRY MAGMATIC SUITES
In the absence of feldspar and zircon, Sr and Zr are incompatible and their concentrations rise during crystallization. Once feldspar saturates, invariably as a large fraction of the crystallizing assemblage, Sr behaves as a highly compatible dispersed element and its concentration in melt falls as crystallization proceeds. Because Zr is an essential structural constituent of zircon, its concentration is buffered at saturation levels once zircon appears, falling as T drops though not so precipitously as Sr. Thus, Zr/Sr invariably rises until zircon saturates (assuming that feldspar saturates first). The longer the relative delay in zircon saturation, the more Zr/Sr increases.
We find that, in rocks with high-T histories (zircon saturation Ts >850 C: Icelandic rhyolites, Grenville granitoids of the southern Appalachians, some volcanic centers in the Miocene Colorado River rift), Zr/Sr rises to >>2 at 70 wt% SiO2. In contrast, rocks that represent cooler magma systems, e.g. Mount St. Helens high-Si dacites, Paleozoic granitoids from the southern Blue Ridge, Cretaceous granitoids from SE California, have Zr/Sr <<1 at 70 % SiO2. We interpret this contrast to reflect the phase equilbria of zircon, feldspar, and melt under wet vs. dry conditions: saturation of feldspars is strongly influenced by water content of magma, whereas zircon saturation is not (Watson & Harrison 1983; Boehnke et al 2013). Simple Rhyolite-MELTS + zircon saturation modeling demonstrates that for granitic compositions the difference between 2 wt% and 4% H2O has a dramatic effect on how much feldspar crystallizes prior to zircon, and thus on how much Zr/Sr rises. At higher SiO2 Zr/Sr continues to rise in both ‘hot’ and ‘cool’ systems, but the distinction appears to be clear as long as Sr concentration remains above 10-20 ppm.