(Table 1) Lithology of DSDP Hole 72-516F sediments

Drilling at Site 516 on the northern shoulder of the main Rio Grande Rise has improved our understanding of the tectonic evolution and subsidence history of the Rise and of the entire Rio Grande-Walvis seamount and aseismic ridge system. Evidence from this site indicates that basalts at the bottom of Hole 516F were produced at the ridge crest and that the ridge crest was subaerial. I attribute the anomalous elevation of the Rise to an eastward ridge crest jump to the western end of the Rise, 91 Ma, and, recognizing a southward progression of such eastward jumps, I suggest a model for the Rio Grande-Walvis system involving a slow westward component of drift of the ridge crest off a hot-spot swell. This drift caused off-axis volcanism that in the Cretaceous succeeded but in the Cenozoic failed to capture the ridge crest. The probable mechanism of capture involved counteraction of the ridge push force within the lithosphere by a swell push force. This would make capture more likely in lithosphere produced by fast spreading, perhaps explaining the change of mode at the end of the Cretaceous. Within the pelagic carbonate sedimentary succession at Site 516, the partly volcaniclastic, turbiditic middle Eocene Unit 4 contains volcanic ash beds (yielding a 47.4 ± 0.7 Ma K-Ar age from fresh alkalic biotite) and a 15-m-thick basal slide. Reflection profiles show it was produced by sliding from and by subaerial erosion of a large tilted and uplifted guyot upslope from the site. Data from the site suggest that a single short off-axis event affected the entire crestal region of the Rise. Perhaps the same midplate hot spot that produced an 80-50 Ma volcanic episode in the Serra Geral of Brazil was responsible, but that was not the present Tristan hot spot. The data from Site 516 have been incorporated into a detailed model for the subsidence history of the main body of the Rise. The model uses an "oceanic" thermal isostatic model, but accounts for the effect of subaerial subsidence and incorporates the changes of the middle Eocene event. A detailed sediment compaction model is developed, and a smooth eustatic sea level correction is applied. The effects of "basement compaction" and use of local rather than regional isostatic compensation are assessed each at about 50 m. The computed paleodepth at Site 516 ranges from sea level 84.0 Ma through a Paleocene 1250 m maximum and middle Eocene 600 m minimum to 1313 m today. The "tectonic" depth curves for both Sites 516 and 357 are compared with paleoecologic depth estimates. In general, these paleoecologic estimates lie deeper, probably because of the difficulties of applying accurate subsidence and compaction corrections to the comparison sites