Probing Paleoceanographic Proxies

Abstract

This thesis provides additional mechanistic understanding of both thermal and non-thermal influences on several marine temperature proxies: TEX86, based on the number of cyclopentane moieties in isoprenoid glycerol dibiphytanyl glycerol tetraethers (GDGTs) made by ammonia oxidizing archaea; UK’37, based on the number of double bonds in long-chain alkenones made by coccolithophores (haptophyte algae); the clumped isotope composition (Δ47), which relates the number of carbonate ions in which the rarer, heavier isotopes of both carbon and oxygen (13C and 18O) occur in calcitic coccoliths made by the same haptophyte algae; and the ratio of magnesium to calcium (Mg/Ca) and oxygen isotope composition (δ18O) in the calcitic tests of planktonic foraminifera. To this end, I used a long-term (10-year) sediment trap record in the eastern Mediterranean Sea, where proxy values typically deviate from expected values based on global core tops, to examine proxy variability captured in sinking particles at three depths in the water column (~500, 1500, and 2500 m) with ~2-week resolution. This temporal resolution allows for a robust understanding of seasonal variability, while the depth profile provides insight into possible effects of degradation or dissolution in the water column. Sediment trap results were compared with real-time physical and biogeochemical models to help in disentangling thermal and non-thermal influences on proxy values. I additionally use computational methods to address targeted questions on possible proxy biases.These approaches led to several important findings: Langrangian particle tracking simulations demonstrate that the impact of ocean currents laterally transporting proxy carriers has little impact on proxy temperatures in the Mediterranean Sea. A global compilation of GDGTs in sediment trap material demonstrates that, rather than subseasonal variations in temperature, subseasonal variations in the depth of GDGT export likely control the variations in TEX86. Using biogeochemical model output, the changing depth of GDGT export is linked to changes in the depth or quantity of net primary production. Using evidence from C37 and C38 alkenones, coccolith Δ47, and the hydrogen isotope composition of alkenones, thermal and non-thermal influences on alkenone-based paleothermometry were disentangled in the Mediterranean Sea sediment trap, leading to an inferred dwelling depth of ~40-50m in summer and the inference two alkenone-producing strains which are dominant in different seasons.Little seasonal variability in Mg/Ca and δ18O in the planktonic foraminifer G. ruber occurs in the Mediterranean Sea sediment trap. This lack of variability could be a result of habitat tracking, where the individuals vary the depth of their habitat to maintain an invariant temperature, or may be due to longer-lived individuals (~4 months) than is typically assumed (~1 month), where individuals calcify under optimal conditions in summer, and persist into winter and early spring.Finally, the temperature dependence of GDGTs was re-examined using both surface and ancient sediments. Rather than an index-based approach, the temperature dependence of each GDGT is described as a Gaussian curve, resulting in an idealized GDGT distribution for a given temperature. Using the fit to ancient sediments to extrapolate beyond the modern temperature range should improve temperature reconstruction in warm climates.