FIRE AND ICE: MANTLE MELTING BENEATH THE HIGH ARCTIC MID-OCEAN RIDGES

Abstract

Mid-ocean ridges generally are thought of as simple and well-defined geological structures. Straight ridge segments of greater or lesser extent whose faults trend perpendicular to the direction of spreading are offset by strike slip transform faults trending parallel to the direction of spreading. In a few places along ultraslow spreading mid-ocean ridges such as the Southwest Indian Ridge (15-18 mm/yr full rate), these rules seem to be violated, as the distinction between ridge segment and transform becomes blurred (as at the oblique segment of the SW Indian Ridge). This is characteristic of ultraslow spreading ridges. The Arctic Ridges violate the “rules” of mid-ocean ridge behavior almost from the moment the N. Atlantic ridge crosses the Arctic Circle in Iceland. There, north of the Jan Mayen fracture zone, the ridges deepen, become oblique and lose their stairstep geometry. The Mohns ridge, beginning about 73 degrees N, is about 15-20 degrees oblique to the spreading direction, then swings around through 0 degrees obliquity to become the Knipovich ridge, which is almost 30 degrees oblique in the opposite direction. Both Knipovich and Mohn’s ridges are devoid of fracture zones. Little or no peridotite outcrop has been reported from Mohn’s or Knipovich ridge. Basalts from these ridges tend to be enriched relative to basalts from the North Atlantic. The northern end of Knipovitch ridge is delineated by one of the few arctic fracture zones, the Molloy Fracture Zone. Immediately to the north, Molloy Deep forms an oblique, deep rift structure, from which numerous recorded dredge hauls have reported only peridotite. Molloy deep is the deepest point on the global mid-ocean ridge system (6000m), and is the first known example of amagmatic oceanic rifting. The northern part of Molloy deep contains a basement ridge structure which also consists of mantle peridotite, mostly plagioclase-bearing. These peridotites have as yet been little studied. The Spitzbergen Fracture Zone connects Molloy Deep with the Southern end of Lena Trough. Though many authors have speculated as to the nature and configuration of Lena Trough, recent work has shown that Lena Trough is a single amagmatic oblique spreading segment, devoid of fracture zones. Instead of a classical segmentation, the rift is divided into a relay zone of steep-sided basement blocks consisting entirely of mantle peridotite. The only basalts are from a single location and show a highly enriched, K-rich composition which bears little chemical similarity to other MORB, and shows a strong garnet trace element signature in their source. The peridotites of Lena Trough are highly variable, and in the basalt location also with trace element signatures suggesting garnet residue during partial melting. The minimum chrome numbers of residual peridotites are low, 13-14 in nearly every dredge haul, and are strongly variable, up to Cr# 40-50 in some stations. There appears to be a magmatic segmentation represented by the occurrence of magmatic veins and plagioclase peridotites, which are more common in the center of Lena Trough than at the ends. Lena Trough veers abruptly East at its northern end, becoming the orthogonal-spreading Gakkel Ridge. The chemistry of basalts from western Gakkel Ridge more closely approximates normal MORB, though some sites near Lena Trough show evidence of the high-K component. Between the termination of Lena Trough and a marked discontinuity at 3E, there are no deviations of axial linearity, no fracture zones, but there is evidence of primary magmatic segmentation because of vertical undulations along the ridge. At 3E another amagmatic zone begins, somewhat shorter than Lena Trough. Peridotites from the western end of this region are uniformly fertile (Cr# 12-14), but become more variable to the east. Two dredge hauls from this region contained ultra fresh peridotites, with no visible serpentinization in many samples. A reference “fresh abyssal peridotite” sample set is being prepared so that bulk and microanalysis of these critical samples can proceed in an organized fashion. The eastern edge of the 3E zone marks a dramatic change both in tectonics and in the geochemistry of peridotites and basalts. A narrow zone of basalt enrichement is flanked by a broader zone of peridotite major element depletion. Minimum Cr#’s here reach 45-50 in several spatially associated dredge hauls. One of these, HLY D40 also contains some fresh material, as small rounded clasts in a carbonate-hasted breccia, but the amount is quite small, only a few kg total. East of this zone, as the spreading rate decreases to around 10mm/yr full rate, fertile peridotites return, and basalt chemistry becomes similar to that observed along other ultraslow spreading ridges. Overall the arctic ridges are a surprise because of what they don’t show. Low degrees of partial melting would imply that basalt have a chemistry dominated by early-melting components (often referred to as “mantle veins”), the basalt geochemical database to date for the arctic rules this out except for one location in Lena Trough. Second, thin to missing crust would imply a uniformly fertile mantle, perhaps corresponding to completely unmelted mantle. Except for one small region at the western end of the 3E amagmatic zone this is not the case. Instead, nearly all locations show a dramatic variablility in chemical compositions of mantle rocks. Some of these variations are too extreme to have plausibly originated by melting beneath Gakkel Ridge, and must instead be the product of older events. Finally, the basaltic crust turned out to be surprisingly robust. Though there are several large occurrences of bare mantle outcrop along the high arctic ridges, the basaltic crust in between is surprisingly thick. Possible reasons for this include a long-lived zone of enhanced melting beneath the western Gakkel Ridge, an increase in melt extraction efficiency when highly focused, or the fact that heavy sedimentation in the eastern Gakkel Ridge covers up peridotite exposures that we would otherwise be able to sample between the basaltic edifices. The future of exploration of the high Arctic ridges is likely to include dredging of the eastern end of Gakkel Ridge, as drilling in the sediment covered basins in the East, and drilling investigation of the causes of the abrupt change in crustal thickness at 3E