FROM SUBDUCTION TO EXTENSION/TRANSTENSION; A CASE STUDY IN TRANSITIONAL GEOCHEMISTRY FROM SONORA, MEXICO

Plate tectonic reconstructions indicate Miocene volcanic rocks in the coastal province of Sonora, Mexico were erupted both before and after the ~12.5 Ma transition from subduction to transtension/rifting in the proto-Gulf of California. We examine major and trace element geochemistry of well-dated volcanic sections from this province in order to compare them to accepted petrotectonic models. Syn-subduction volcanic rocks (24 - 12.5 Ma) range from calc-alkalic basaltic andesite to dacite and have typical arc affinities, including a depletion in the HFSE with the characteristic Nb-Ta troughs. Post-subduction volcanic rocks (12.34 – 8.25 Ma) also range from calc-alkalic basalt to rhyolite (including abundant andesite-dacite) and the mafic end-members have a typical intraplate affinity with a residual arc depletion in the HFSE. Previous studies in the Gulf Extensional Province have suggested there was a sudden transition from syn-subduction calc-alkaline volcanism to syn-rifting, tholeiitic or bimodal volcanism, a conclusion our study does not support. Instead, our study documents a progressive change in chemistry that is still ongoing when volcanism shuts off in coastal Sonora several million years after subduction ends. In addition, this change is only evident in the trace element compositions.

The characteristics of geochemical change in Sonora in response to a change in tectonic setting highlight the need for caution in future interpretations of tectonic environments from classic petrotectonic models. Sonora appears to fall into a long list of post-subduction extensional provinces (the Rio Grande Rift, the Mexican Borderland rift, the southwestern Mexican Volcanic Belt, and the SCORBA region of SW North America) that do not exhibit typical extensional bimodal tholeiitic and rhyolitic volcanism from the first onset of extension. Instead, intermediate composition volcanism persists until the hydrous, metasomatized lithospheric mantle is exhausted. This further reinforces the need to address the geochemistry of a region as the result of both: 1) the contemporary processes causing melting and 2) the processes operating in the lithosphere over the previous several m.y.