Swollen Volcanic Ridges in Pre-Oceanic, Oceanic and Back-Arc Environments

Upwelling of the upper mantle below mid-ocean ridges is considered within the theory of Plate Tectonics to be normally a mostly passive process, induced by plate separation along accretionary boundaries. At any given mantle and/or temperature/composition, melt production and crustal thickness depend on rate of upwelling, that is, on spreading rate, resulting in a correlation along ridges of crustal thickness and topographic height with spreading rate (Shen and Forsyth, 1992). Crustal thickness and topographic level vary within a small range in much of the mid-ocean ridge system (Morgan and Chen, 1993). This means that temperature and composition of the upwelling mantle are relatively constant below ridges, except for "hot spots" and "cold spots" regions. We call attention to a set of anomalous volcanic ridges from different geotectonic settings: an anomalous mid-ocean ridge segment (Spiess Ridge, Southwest Indian Ridge); a back-arc basin ridge (Marsili Ridge, Tyrrhenian Sea); and a "pre-oceanic" rift (the Erta Ale Ridge in the Southern Red Sea rift system). These ridges are more elevated than even those mid-ocean ridge segments that are influenced by "hot spots". A possible qualitative explanation is that these ridges, that we call "swollen ridges", are underlain by upper mantle thermal and/or compositional anomalies producing an unusually high quantity of melt that cannot be accommodated by plate separation. The result is a thicker than normal basaltic crust and an anomalous high topographic level. Another result of the mantle anomaly that underlies swollen ridges is an unusually shallow subaxial magma chamber, documented particularly for the Marsili and the Erta Ale Ridges. We updated Purdy et al. (1992) curve showing an inverse correlation of subridge magma chamber depth versus spreading rate, confirming the inverse trend between spreading rate and depth of the axial magma chamber. However, our three swollen ridges plot outside the trend, their magma chamber being too shallow for their spreading rate. A qualitative explanation of these observations might be that swollen ridges are created by non-passive processes, whereby the quantity of melt produced by an upper mantle subridge thermal and/or compositional anomaly exceeds the quantity that can be accommodated by plate separation. As a result the thickness of the basaltic crust and the ridge topographic level are higher than normal.