Gravity fields over the crests of slow spreading mid-ocean ridges are characterized by small amplitude free air gravity anomaly lows (30--70 mGal) with wavelengths of 40--60 km, flanked by smaller gravity highs. In general, the amplitude and wavelength of these gravity lows decrease with increasing spreading rate until, at fast spreading ridges (>5 cm/yr), free air gravity highs (10--20 mGal) are observed. Previous explanations of these anomalies, which involve uncompensated ridges, thermally expanded ridges and elastic plate compensation models, are generally not consistent with the geological evidence regarding the nature of the crust and upper mantle beneath the ridge crest.

In particular, the postulated presence of wide zones of partial melt in the uppermost mantle is inconsistent with petrologic data from mid-ocean ridge basalts, which indicates that primary melts are in equilibrium with an extracted from residual mantle at depths of 30 km or more. In addition, similar gravity lows observed over extinct spreading ridges suggest that the anomalies are not mainly due to any dynamic aspect of the accretion process taking place at active spreading centers. Geometrical models of mid-ocean ridges derived from detailed structural and petrologic studies of the Bay of Islands ophiolite complex have been used to generate gravity anomalies over ridges with various spreading rates. Modeling results show that four main sources serve to define the amplitude and wavelength of the gravity anomaly: the ridge topography, the crustal level magma chamber, the configuration of the isotherms near the ridge crest, and a low-density, partially gabbroic cumulate root extending vertically into the mantle beneath the ridge.

Systematic variations in the contributions from these sources with increasing spreading rate control the changes in character of the gravity anomaly from slow to fast spreading centers. Major contributions are found to come from the topography and from the low-density root, which broadens with slower spreading rates, resulting in both increasing amplitude and increasing wavelength with decreasing spreading rate. At extinct spreading centers, contributions from the magma chamber, isotherms, and topography gradually diminish with time once the ridge ceases to spread, whereas the low-density root remains a permanent part of the oceanic lithosphere and continues its contribution to the gravity field. Seismic evidence from slow spreading centers appears to support the existence of the low-density gabbroic root zone inferred from the study of the Bay of Islands ophiolite and the gravity analysis presented here.