Two forms of volcanism: Implications for mantle flow and off‐axis crustal production on the west flank of the southern East Pacific Rise

Abstract

SeaMarcII side‐scan sonar and bathymetric data on the west flank of the East Pacific Rise (16°–18°S) reveal a large number of seamounts, organized primarily in chains. The easternmost ends of the chains near the ridge axis appear to be active as indicated by fresh lava flows. In addition, areas of unusually high reflectivity representing recent lava flows were found as far as 60 to 80 km from the spreading axis. The flows are sometimes dammed by abyssal hill scarps formed near the ridge axis that have maximum relief of the order of 100 m. In other cases, the scarps appear to be buried by flows. The disappearance of scarps can be used to estimate the extent of old lava flows or volcanic debris that are not detected acoustically in the sonar image. Under this criterion, lava flows and volcanic debris surrounding the seamounts cover 40–50% of the survey area of about 18,000 km2. With an estimated average thickness of about 100 m, the volume of lava flows and volcanic debris is about 0.8±0.3% of that of the total crust. The volume of the seamounts is estimated by isolating the seamounts from the background topography with an anisotropic, median‐filtering technique. The volume of seamount edifices is 1.05±0.05% of the volume of the crust. Thus the total volume of volcanic material extruded off‐axis is about 1.5 to 2.2% of the volume of the crust. The decreasing volcanic activity away from the ridge axis suggests that near‐axis seamounts may be associated with the upwelling system beneath the ridge, perhaps by preferential melting of embedded, passive heterogeneities. However, the decrease in volcanism with increasing distance from the axis also could be caused by a decrease in lithospheric vulnerability to penetration by magma. Since no apparent east‐west faults and no simultaneous volcanic activity along the length of a seamount chain are observed, north‐south extension and ridge‐perpendicular convection are unlikely to be major causes of seamount formation in the study area. The existence of closely spaced linear chains and the tendency for fresh flows to be found at near‐ridge ends of chains suggest that there are discrete sources in the upper mantle active for extended periods (over 1.8 m.y.). Passive heterogeneities embedded in the upwelling mantle would have to be highly elongated along streamlines to produce persistent, nearly stationary melting anomalies. A variety of mechanisms may be responsible for the off‐axis volcanism, but the single hypothesis most consistent with all observations is a mini‐hotspot origin. Although fresh lavas 80 km off‐axis are consistent with broad mantle upwelling beneath the ridge, the possibility of narrow upwelling cannot be ruled out, because the seamounts could be caused by independent buoyant upwelling of mini‐plumes that can penetrate thin lithosphere more easily near the axis.