A test of whether contrasting magma supply rates (hotter vs. cooler lower crust) influences the differentiation of arc magmas in the Mexican arc

Numerical modeling studies (e.g., Dufek and Bergantz, 2005; Annen et al., 2006) have proposed that the extent of differentiation of arc magmas in the lower crust will depend on its thermal state. In these models, mantle-derived hydrous basalts are emplaced into the lower crust as a series of sills and/or dikes. Differentiation to form evolved melts can occur both by partial crystallization of the basalts and by partial melting of surrounding mafic crust. The extent of basalt crystallization and/or wall rock melting, and therefore the compositions of evolved melts, is predicted to be controlled by the temperature of the lower crust, with higher temperatures promoting less evolved melts (e.g., basaltic andesite) and relatively lower temperatures favoring more evolved melts (e.g., dacite). There are three main factors proposed to control the relative temperature of the lower crust, and therefore the composition of the most abundant differentiated melts: (1) initial geotherm, (2) crustal thickness (a deeper lower crust is hotter for a similar geotherm), and (3) supply rate of mantle-derived basalt into the lower crust. The Mexican volcanic arc provides an opportunity to test the predictions of these numerical models on the basis of recent work that has documented magma eruption rates along different segments of the arc. In this study, we compare results from the Tancítaro volcanic field (4000 km2), associated with subduction of the Cocos plate (Ownby et al., 2009) with those from the Mascota volcanic field (2000 km2), associated with subduction of the Rivera plate (Ownby et al., 2008). By the critieria listed above, the lower crust beneath the Tancítaro vs. Mascota volcanic field is expected to be hotter because: (1) the crust is inferred to be thicker beneath the Tancítaro vs. Mascota volcanic fields by ~5-10 km based on gravity studies, and (2) the magma eruption rate is nearly two orders of magnitude higher in the Tancítaro vs. Mascota volcanic fields. Despite these differences, the calc-alkaline differentiation trend at both volcanic fields is nearly identical, with erupted lavas ranging continuously from 52-63 wt% SiO2 and 9-2 wt% MgO. No dacite or rhyolite of any significant volume has erupted in either volcanic field over the last 2 Myr. Moreover, the proportions of erupted calc-alkaline lava types are strikingly similar, with twice as much andesite as basaltic andesite erupted in both volcanic fields. The evidence from the two volcanic fields indicates that there is no difference in the degree of calc-alkaline differentiation despite expectations that the lower crust is hotter beneath the Tancitaro vs. Mascota volcanic fields. As a consequence, thermal models of the lower crust may need to be re-evaluated, and it is proposed here that the transport of evolved liquids out of the lower crust is a major mechanism of heat transfer that prevents the lower crust from becoming hotter on a regional scale and on a long-term temporal scale, and that elevated temperatures associated with the injection of basaltic sills occur over shorter length scales and time scales in the lower arc crust than previously assumed.