Response of deep-ocean ostracodes to climate extrema of the Paleogene: Ecological, morphological, and geochemical data from the Eocene-Oligocene transition and late Paleocene thermal maximum

The Eocene-Oligocene transition is marked by a pronounced ∼1.5‰ decrease in foraminiferal δ18O termed Oi-1, reflecting ice-volume expansion and bottom-water cooling. To assess temperature change at ODP Site 744A (Kerguelen Plateau), the temperature-dependent Mg/Ca ratios of Krithe ostracodes were determined by ICP-MS. Krithe Mg/Ca shows no net trend through Oi-1, indicating that the δ 18O shift largely reflects major Antarctic ice-sheet development.; However, major deep-ocean ostracode faunal abundance, richness, and evenness changes coincide with Oi-1 at ODP Sites 744A and 689B (Maud Rise). Following Oi-1, Site 744A faunal composition shifted slightly back towards initial compositions, whereas Site 689B faunas remained relatively distinct. The Oi-1-coincident faunal changes include both uniform (e.g., disappearance of Cytherella and Bairdioppilata at both sites) and reversed (e.g., Site 689B disappearance and Site 744A appearance of Trachyleberis and Actinocythereis ). Together, these faunal patterns imply a common response to increased Oi-1 surface productivity and a disparate response to either subtle productivity intensity differences or circulation changes. At Site 744A, morphometric examination of seven Algulhasina quadrata instar stages show a strong positive correlation with biogenic opal, indicating that increased organic carbon delivery after Oi-1 allowed increased instar size through sequestering of greater metabolic reserves.; The faunal response of deep-ocean ostracodes to the Late Paleocene Thermal Maximum (LPTM) was also examined at ODP Sites 690B (Maud Rise) and 738C (Kerguelen Plateau). Abundances drop precipitously at the major δ13C negative excursion and then recovers to roughly two-thirds to fully pre-event levels. The Site 690B fauna was overall taxonomically richer than the Site 738C fauna. Site 738C richness recovered after the LPTM, whereas Site 690B richness remained low, but less variable overall, perhaps reflect lingering deleterious conditions. Both sites contain three distinct faunas bounded by the carbon isotope spike and later recovery to relatively stable isotope values. Prolonged LPTM bottom-water dysoxia is not supported by the stable proportion of dysoxia-tolerant filter-feeding ostracode taxa. The restriction of specific taxa to particular stratigraphic intervals within the event provides provisional paleoecological preferences for testing at other LPTM sites and other intervals.