ALBERT

Description: Tourmaline occurs in peraluminous granites from the Central Iberian Zone associated with two main AFM mineral assemblages: (1) muscovite + biotite ± cordierite ± andalusite in the Araya-type granites; (2) muscovite ± biotite ± garnet in leucogranites from the Alamo complex. When tourmaline is dominant, biotite is an accessory or absent, and vice versa. We present field and petrographic relations, mineral chemistry, and geochemical data for tourmaline-bearing and tourmaline-free granitic rocks from various localities in the Central Iberian Zone. Compositional phase diagrams are used to evaluate the factors controlling the occurrence of tourmaline relative to biotite in granitic rocks, with particular emphasis on the relationships between mineral assemblage and whole-rock chemistry and its petrological implications. Although tourmaline stability in felsic magmas depends on the interplay between rates of changing environmental conditions such as bulk composition, T, aH 2 O, and fO 2 , the principal factor dictating tourmaline formation is the B content of the melt, judging from phase relations. In short, regardless of other variables, granitic melts have to surpass a critical boron threshold to achieve tourmaline saturation. Experimental constraints, combined with petrographic and geochemical data, suggest minimum boron contents in the range of ~500–3000 µg g –1 (depending on temperature) to saturate melt in tourmaline. Acting in concert with boron content, other variables such as Al 2 O 3 , mafic components, T, fO 2 , and so on, control not only the formation of tourmaline during melt crystallization, but also the magnitude of boron loss from the magma to the surrounding rocks. The analysis of phase relations suggests that tourmaline granites usually form units distinct from biotite granites because common granitic melts have restricted accessibility to the three-phase Tur–Bt–Ms field.