In order to constrain the onset timing of the Median Tectonic Line in SW Japan, detrital zircon geochronology is examined for pre-Miocene undated sandstones at both sides of the subhorizontal MTL in the Mikawa-Ono area, central Japan, where the 3-D configuration of the primary MTL between the Ryoke and Sanbagawa belts is observed. In addition, another Cretaceous sandstone unit (the Idaira Formation) in the Chichibu belt, ca. 10 km to the south, is analyzed for comparison. The following new results were obtained. The undated sandstone-dominant unit, previously called the Kawachi Formation or the Nanasato-Isshiki Formation as a whole is, in fact, composed of two distinct units with contrasting characteristics of distribution, rock type, and zircon age spectrum, although both are revealed to be of the Cretaceous. The one to the north of the subhorizontal MTL (in the Ryoke belt) is re-defined as the Nanasato-Isshiki Formation sensu stricto, whereas the one to the south (in the Sanbagawa belt) is newly designated as the Rokutazawa Formation. The Lower Cretaceous Idaira Formation forms the third group with a distinct signature in the zircon age spectrum. The Nanasato-Isshiki Formation (together with the associated Atera-Nanataki conglomerate) and the Idaira Formation, overlying the Ryoke granitoids and Chichibu pre-Cretaceous accretionary complex, respectively, form autochthonous units. In contrast, the Rokutazawa Formation alone represents an allochthonous units sitting structurally on the Sanbagawa schists. Age spectra of detrital zircons suggest that both the Nanasato-Isshiki Formation and Atera-Nanataki conglomerate are roughly dated as the Late Cretaceous, and also that they are correlated with the Upper Cretaceous Izumi Group in the Shikoku and Kinki regions of SW Japan. On the other hand, the sandstone of the Rokutazawa Formation demonstrates a totally different age spectrum with full of older zircon grains of the Paleozoic and even of the Neoarchean, and with the youngest grain suggesting the Early Cretaceous age of the formation. Its distribution, rock type, and age spectra of detrital zircons indicate that this unit can be correlated with the Atogura/Maana formations within the allochthonous klippe unit in the Sanbagawa belt in the Kanto Mtn. and western Shikoku, SW Japan. These units were probably transported for a long distance on a hundred-kilometer scale from the primary depositional site in the Hida belt. The correlation of the Idaira Formation and the Lower Cretaceous strata of the Chichibu belt in other areas (e.g. the Ryoseki–Monobegawa Group in Shikoku) is also confirmed by the age spectrum of detrital zircons. All of these data suggest that the low-angle faulting of the primary has started probably some time in the Oligocene, after the deposition of the Nanasato-Isshiki Formation on the hanging wall (the Maastrichtian or later), after the emplacement of the klippe with Lower Cretaceous onto the Cretaceous blueschists of the Sanbagawa belt on the foot wall (after the Eocene), and before the deposition of the neighboring Miocene strata (the Shitara Group). These results constrained for the first time the initiation timing of the MTL, not in the Cretaceous as previously imagined, but in the Paleogene, most probably in the Oligocene.
