Previously reported biostratigraphic relationships from middle-upper Miocene sections exposed near Buff Bay, Jamaica (18°N, tropical bioprovince), differ from the subtropical North Atlantic (Sites 563 and 558). Time scales for this interval rely on correlations established at these subtropical sites, and the differences with the tropical section have implications to global correlations. Planktonic foraminiferal Zones N13 and N15 are thick at Buff Bay but are virtually absent at Sites 563 and 558; nannofossil Zone NN9 is associated with Zone N15 and uppermost Zone N14 at Buff Bay but is associated with Zone N16 at the other sites. Magnetostratigraphic data presented here further complicate the interpretation: Zone NN9 is associated with a thick normal magnetozone at Sites 563 and 558; at Buff Bay, it is associated with a thick reversed magnetozone. Although a secondary magnetization at Buff Bay makes it difficult to identify confidently Miocene normal magnetozones, the thick reversed magnetozone most likely represents the paleomagnetic field and correlates with Chron C5r. The magnetobiostratigraphic relationships require either diachrony of taxa or two mutually exclusive hiatuses in Jamaica and the North Atlantic.
We address this problem by analyzing benthic foraminiferal δ18O and δ13C from the Buff Bay section. These isotopic data allow us to evaluate three hypotheses that reconcile the magneto-, bio-, and isotopic stratigraphic data and conclude that the first and last occurrences of five taxa were diachronous by ∼0.3-0.5 m.y. between tropical and subtropical locations. This requires revised age estimates for late middle to early late Miocene biostratigraphic datum levels. We suggest that the ranges of several taxa are useful for endemic tropical or subtropical zonations, but correlations between the low and midlatitudes were affected by an increase in latitudinal thermal gradients during the late middle Miocene. However, we admit that further studies are needed before this issue is resolved.