The Bodrum Igneous Complex is the largest remnant of an post-collisional Late Miocene to Early Pliocene magmatic province of subalkaline and potassic character in the SE Agean and SW Anatolian region. Magmatic products at Bodrum, showing the typical high-field strenght element (HFSE) anomaly of lavas at convergent margins, span a wide compositional range including also little modified mantle magmas, unusual for the province. Volcanic rocks vary from high-Mg ultrapotassic varieties to trachybasalts, trachyandesites, alkali trachytes, trachydacites and rhyolites, intrusive bodies from monzogabbros to monzodiorites, monzonites, and quartz-monzonites. Intermediate rocks are not only by far the most voluminous, they also appear recurrently throughout the lifespan of this complex between, at least, 12 and 7.5 Ma whereas silica-rich felsic rocks seem to be restricted to the interval from 12 to 11 Ma and more alkali-rich felsic rocks from 10 to 9 Ma.
Major element variations among volcanic rocks indicate two increasingly divergent trends towards higher evolution degree of the magma, starting from a trachybasaltic parent. In SiO2 vs. total alkali space, one trend evolves from trachybasalt to trachyte (T-trend) near the joint of trachytic and phonolitic rocks; the other trend evolves near the alkalic/subalkalic transition from trachybasalt to trachydacite and finally to rhyolite (R-trend). Only within the most evolved compositional spectrum, the two trends are clearly separated from one another; before, compositions form wide arrays between the two, more and less alkalic, envelopes. With the exception of amphibole and accessory titanite restricted to the evolved R-trend, mineralogical evolutions involving mainly cpx, plagioclase and biotite are very similar and indicate a relationship between the two trends. Isotopic (Sr, Nd, O) compositions also show only slight and non-systematic differences between the R- and T-trend, especially for Nd with the lowest values being those of
rhyolites as well as mantle-derived ultrapotassic rocks.
HFSE among others trace elements can differentiate between plausible and unlikely processes yielding compositions of the T-trend, of the R-trend, and in between. Enrichment of incompatible elements along the T-trend, starting with primitive trachybasalt, follows typical Rayleigh fractionation trends suggesting an origin through such process. On the contrary, throughout the R-trend, predominantly incompatible trace element concentrations like HFSE are often lower or equal to that of the parental magma requiring unlikely steady high bulk partition coefficients near unity or above to validate a pure xls-fractionation scenario. Thus, Bodrum rhyolites are either partial crustal melts and yield intermediate compositions by mixing or,
at least, crustal assimilation played a vital role during xls-fractionation along the R-trend. In any case, the importance is that the T-trend is likely to trace a nearly pure fractionation process and that crustal contributions are significant enough to produce the R-trend and compositions in between either by pure mixing or by AFC processes.
The proposed model provides means to link alkalic and more subalkalic trends to one common parental magma. Consequently, in igneous systems with similar, alkaline versus more subalkaline, compositions but with obscured mafic lineages, our model is an alternative to several mafic parental magmas commonly used to explain the range in alkalinity.