Volcanic and structural history of the Hohi volcanic zone,central Kyushu,Japan

More than 5000 km³ of magmatic material was erupted in Pliocene-Pleistocene times in a volcano-tectonic depression, i. e., the Hohi volcanic zone (HVZ) in central Kyushu, Japan. The eruptive deposits consist mainly of andesite lava flows and large-scale pyroclastic-flow deposits. Their eruptions were accompanied by the formation of an EW-oriented graben (70 km × 45 km) under regional NS extensional stress. Pre-Tertiary basement rocks are absent on the surface of the graben but occur at depth, having subsided up to 3 km. Radiometric ages of volcanic rocks on the surface show zoned isochrons from 5 Ma at the margin to 0.3 Ma in the center of the HVZ. The youngest center of age zonation coincides with a 30 mgal negative Bouguer gravity anomaly. Radiometric ages of rocks from drill cores are older toward the bottom of the graben, reaching a maximum of at least 4 Ma. Volcanic activity concentrated over time toward the center of the graben and buried successively erupted material. Areas of active volcanism in the HVZ became smaller and changed in style during the 5-Ma history of activity. Volcanism of the early stage (5-2 Ma) was characterized by voluminous eruptions of andesitic lava flows that formed lava plateaus and were intruded by EW-oriented feeder dikes, perhaps related to fissure eruptions. In contrast, late-stage volcanism (2-0 Ma) resulted primarily in andesitic to dacitic lava domes with features of monogenetic volcanoes produced at low eruption rates. The HVZ shows unimodal volcanism dominated by andesitic and dacitic lavas with a small amount of rhyolite and only traces of basalt; these characteristics differ from those that typify volcanism in most other extensional areas. Erupted material in the HVZ is of the calc-alkali and high-alkali tholeiite series and shows no significant chemical changes over 5 Ma, except for an increase in K₂O after 1.6 Ma. The net horizontal displacement along normal faults indicates that the HVZ widened by about 10%–20% across the graben at an average rate of 0.1 cm/yr. I interpret the HVZ to be neither a pull-apart structure of the pre-Tertiary basement nor the result of propagation of the Okinawa Trough, but rather the earliest stage of rifting when vertical subsidence caused by normal faulting is compensated by filling with volcanic material.