Asymmetry in the slow-spreading Mariana back-arc basin

The Mariana Trough is an active back-arc basin opening behind the Mariana subduction zone. Its spreading history is fairly complicated, due to ridge segments propagations and trenchward jumps. The basin is asymmetric, with its spreading center located closer to the active arc than next to the West Mariana remnant arc. At latitude 18°N, the spreading axis is located 185 km from the remnant West Mariana Ridge, and 110 km from the active arc to the east. In the central part of the basin, the rift valley has a morphology that is typical of slow-spreading ridges. Its shape varies between a fairly symmetric graben and an asymmetric half-graben with a single large fault scarp on one side of the valley, generally on its eastern side. This asymmetry is well expressed at latitudes 17°20'N and 18°35'N for example. Between 17°40'N and 18°25'N, the neovolcanic zone is expressed as an unfaulted region within the valley floor, which covers an area wide of 6 km in average. The series of hummocky to linear axial volcanic ridges (AVRs) mark the locus of the most active recent accretion. The AVRs ranges in size up to 1 km high, 1-5 km wide, and tens of kilometers long. Along this segment, spreading occurs following a N75E direction at a half rate of 1.4 cm/yr. One striking feature is the location of the Brunhes anomaly, which is not centered on the AVRs. At 18°N for example, the eastern and western sides of the central Brunhes anomaly have markedly different widths, and are 15km and 10km, respectively, from the AVR, suggesting different spreading rates on both sides of the valley. We observe similar pattern more to the north, between latitudes 19°20'N and 19°50'N. We hypothesize that the asymmetry of the Mariana basin is not only due to eastward ridge jumps, but also to asymmetric magmatic accretion. It is probable that the magmatic accretion is one sided with new crust mainly added to the western flank. Along the eastern flank, spreading is mainly accommodated by significant displacement on a single large fault, leading to the formation of an asymmetric graben. This pattern has already been observed along the Mid-Atlantic Ridge at the end of segments. In our case, we suspect that this process has a greater extent since it is observed almost along the whole length (i.e. both at segment centers and ends) of the spreading ridge.