We welcome Clemens and Finger’s (2012) comments on our paper regarding the metasedimentary source for the Jiuyishan high δ18O fayalite-bearing A-type granite in southern China (Huang et al., 2011). They concur with our classification of the Jiuyishan granite as A-type, but question our interpretation of a metasedimentary source for these rocks. Instead, they propose an old, felsic, meta-igneous protolith for the granite.

Clemens and Finger state that evolved Sr and Nd isotopes are the chief reasons for us proposing a sedimentary source; alternatively, they propose that fluid-absent melting of older granitoids can also account for the crustal isotope feature, as well as the high-temperature, A-type character of the Jiuyishan granite. In fact, the most compelling evidence for a metasedimentary source for the Jiuyishan granite is its high δ18Ozircon of 8.0‰–9.8‰ (Huang et al., 2011), rather than the evolved Sr and Nd isotopes. Melting of older granitoids cannot account for the 18O-enriched feature of the Jiuyishan granite, unless the protolith itself has a sedimentary origin. Besides, melting of older granitoids does not always produce high-temperature, A-type rocks, and when it does, the melts are always highly felsic (e.g., Patiño Douce, 1997), incompatible with a low-SiO2 (64 wt%) magma parental to the Jiuyishan A-type granite. In addition, low-volume melt proportions derived from a granitoid protolith probably would not produce granitic plutons as large as the Jiuyishan granite (>1000 km2 outcrop).

Clemens and Finger propose that “Fa90 would be stable at ∼1.4 times the fO2 of FMQ” and suggest that the relatively reduced condition of the Jiuyishan granite would be satisfied by melting of an ilmenite-series granitoid; therefore, relatively low fO2 does not unambiguously support a metasedimentary source. Indeed, low fO2 equivalent to that of the Jiuyishan fayalite-bearing granite could be met by many sources, including the ilmenite-series granitoid noted here and metasedimentary rocks exemplified by Clemens et al. (2011). However, as noted above, melts of any granitoid are incomparable with the low SiO2 nature of the parental magma for the Jiuyishan A-type granite.

On the basis of Sr, Nd, Hf, and O isotopes, Clemens and Finger suggest an I-type connection for Mesozoic granitoids in the South China Block, including the Jiuyishan granite. They propose that these rocks could have been formed from re-melting of large-volume, crustal-derived Early Paleozoic felsic igneous rocks. We emphasize that radiogenic isotopes cannot be used to unambiguously distinguish an I-type source, whereas O isotope ratios of the Jiuyishan granite support a sedimentary source. First, zircon δ18O results for the Jiuyishan granite overlap entirely with those of the S-type Hawkins dacite-Cootralantra granodiorite (7.5‰–10‰) and Violet Town ignimbrite-Strathbogie granite (7.1‰–12.9‰) in southeastern Australia (Kemp et al., 2008). Second, the estimated whole-rock δ18O values of 9.5‰–11.6‰ for the Jiuyishan granite are comparable with S-type granites worldwide (>10‰), but higher than I-types (5.3‰–10‰) (e.g., Eiler, 2001). It is noted that the whole-rock δ18O of 10.8‰ for a magnetite-series granite sample cited by Anderson and Morrison (2005) is much higher than the rest of the group (<9.6‰). We interpret that this could be a result of alteration and/or input of sedimentary components. Melting of the Early Paleozoic felsic igneous rocks cannot reproduce the low-SiO2 Jiuyishan granite. The fact that the Jiuyishan granite outcrops at the same level as the Early Paleozoic granites casts further doubt on a melt-source connection.

Clemens and Finger propose that “felsic magmas derived from metasedimentary sources are strongly peraluminous.” Therefore, the metaluminous to weakly peraluminous Jiuyishan rocks are similar to many I-types, but different from the strongly peraluminous high-temperature, metasediment-derived volcanic rocks referred to by Clemens et al. (2011). However, an examination of Clemens et al.’s data reveals that many relatively low-SiO2 samples of the metasediment-derived fayalite-bearing Toombullup ignimbrite are weakly peraluminous, with a lowest ASI value of 1.0, comparable with the Jiuyishan rocks in the same SiO2 range.

To conclude, we contend that the Jiuyishan high δ18O fayalite-bearing A-type granite most likely originated from melting of a granulitic metasedimentary source (Huang et al., 2011), rather than an old, felsic, meta-igneous protolith.