This dissertation investigates the climate system response during periods of relative warmth over the last 10 million years (Myr) at different temporal scales: thousand-year (i.e., orbital) timescales during the warmer glacial cycles of the early Pleistocene and million-year timescales during the warm mean-state of the late Miocene. The relatively warmer and less intense glacial cycles of the early Pleistocene are an ideal period to investigate what drives high latitude climate during the ice ages, which remains not entirely understood. The warm mean-state of the late Miocene is a period not currently well characterized especially in the western equatorial Pacific (WEP) therefore, we investigate how WEP climate responds as the mean-state shifts from the warm late Miocene into the Pleistocene ice ages. This dissertation is presented in three parts. Chapter 2 provides an efficient method to characterize changes in high-resolution multi-sensor track data (i.e., Gamma Ray Attenuation (GRA) bulk density) as a function of sediment composition and/or grain size. This sedimentological study focuses on the diatom-rich sediments of the high latitude Bering Sea. An interpretation of the GRA bulk density data is developed based on laser particle size and smear slide analysis. Results show that down-core variability in GRA reflects changes in the sediment packing due to changings in diatom valve abundance and fragmentation. The method presented in Chapter 2 is adapted and applied in Chapter 3, which studies the cyclicity of the Bering Sea cryosphere to test theories of what causes ice ages. We focus on the warmer early Pleistocene glacial cycles because they are shorter (40 thousand years (kyr)), less intense, and are thought to be a more straightforward response to insolation forcing than the most recent ice ages. To measure local high latitude ice variability Chapter 3 presents an orbitally resolved Ice Rafted Debris (IRD) flux record that is interpreted with help from the sedimentological method of Chapter 2, modified to interpret Natural Gamma Radiation (NGR) instead of GRA. Results suggest the record dominantly represents transport of IRD by sea ice. While spectral analysis reveals there are no 40 kyr cycles, there is significant sub-orbital variability in the record, which suggests the Bering Sea cryosphere is sensitive to local seasonal forcing. Chapters 4 evaluates the long-term (~10 Myr) evolution of sea surface temperatures (SST) in the WEP warm pool since the late Miocene. There is little data from the WEP warm pool, but over this ~10 Myr period many other regions show a long-term cooling trend. We also assess whether the tropical Pacific El Padre mean-state, characteristic of the early Pliocene, is present in the late Miocene. The SST record presented in Chapters 4 use the Mg/Ca paleothermometer by measuring trace metal ratios of planktic foraminifer Trilobatus trilobus. Results show there is long-term stability in the SSTs of the WEP warm pool, and the El Padre mean-state is present during the late Miocene. Overall, this dissertation broadly investigates the climate system response during past periods of relative warmth. Chapters 2 and 3 focus on the high-latitude Bering Sea at orbital timescales and show that Bering Sea cryosphere is sensitive to local seasonal forcing during the early Pleistocene. Chapter 4 focuses on the low-latitude WEP over million-year timescales and finds long-term stability in the WEP warm pool SSTs since the late Miocene. Synthesized, this work finds strong regional responses, since the 40 kyr cycles were not found in Chapter 3, and the long-term cooling trend was not found in Chapter 4.

Climate dynamics are primarily forced by physical parameters such as insolation and ocean circulation; however, purely physical models fail to replicate the abrupt changes seen in climate records, implicating biogeochemical internal feedback mechanisms as important factors in the global climate system. Marine nutrient supply is a primary avenue that can potentially propagate climate signals to disparate parts of the globe; for example, a rapid response of the biological pump can both amplify signals and force further climate changes. Our current knowledge of paleo-biogeochemistry is limited by scant evidence and, often, low temporal resolution. This dissertation uses high-resolution marine sediments from the Subarctic and Subantarctic Pacific to reconstruct relationships between climate change, nutrient supply, and the biological pump.The Subarctic Pacific experienced a brief interval of extremely high primary productivity during the global transition from the glacial to the interglacial climate regimes. The cause of this high productivity has been debated, with both iron fertilization and reorganizations in Pacific circulation proposed: iron fertilization would suggest a strengthened biological pump, while circulation changes may indicate a weakened biological pump. Here, I reconstruct the diatom community response in terms of silicic acid utilization (single-genus silicon isotopes) and species composition, finding that silicic acid utilization was not enhanced, and that low-iron-adapted species were the primary responders to the high-productivity. These results suggest iron was the limiting nutrient during the high-productivity interval, consistent with the major reorganizations in Pacific circulation causing increased macronutrient supply and briefly releasing large amounts of CO2 to the atmosphere. Nutrient supply during the deglaciation is further explored via amino-acid-bound nitrogen isotopes from the same sediments from the Subarctic Pacific. This investigation finds that the high-productivity intervals carry an elevated signal of source δ15N (i.e., the δ15N of primary producers), corroborating a change in circulation. We also find high-productivity intervals have the lowest community Trophic Positions, indicative of shorter and more direct trophic chains, consistent with high macronutrient supply. Finally, I examine changes in sediment color in the Subantarctic Pacific across the glacial-interglacial transitions of marine isotope stages 7/8 and 17/18 and find that sediment color is strongly related to sediment composition. Interglacial intervals contain high weight percent calcium carbonate and glacial periods contain high weight percent nitrogen. This suggests the position of the Subantarctic Front was strongly linked to changes in climate, with cold periods characterized by northward expansion of high-silicate waters, favoring diatom production over coccolithophores, and strengthening the local biological pump.

Due to processes that are poorly understood, the arctic/subarctic region is currently warming at approximately twice the rate of the rest of the globe, and these regional changes may in turn have important consequences on global climate. The Bering Sea is an ideal location to study the relationship between sea ice distribution, ocean circulation, and biological productivity, which are all thought to be components of important interactions that both respond to and influence climate changes on regional (North Pacific) to global scales. Geologic records are essential for understanding and predicting future oceanic and climatic changes, but most previous studies of Bering Sea and North Pacific climate and oceanic conditions were limited by short records that only spanned one glacial-interglacial climate cycle (~100 kyr). This dissertation presents new geochemical and sedimentological records of oceanic and climatic conditions over the past 1.2 Myr using sediment cores from Integrated Ocean Drilling Program Site U1342 (54.83N, 176.92E, 818 m water depth) in the Bering Sea, providing an unprecedented opportunity to examine trends over multiple glacial-interglacial cycles. Long benthic foraminiferal stable isotope records from Site U1342 provide the first evidence of an important recurring relationship in which extremely cold glacial conditions lead to enhanced Bering Sea ice extent and enhanced local formation of North Pacific Intermediate Water (NPIW), the water mass primarily associated with large-scale North Pacific circulation and heat transport. These results contradict model predictions for weaker NPIW formation during extreme glacial conditions and show that North Pacific and North Atlantic circulation acted in phase on glacial-interglacial timescales. Furthermore, new records of nitrate utilization based on nitrogen isotopes demonstrate that glacial climates were consistently associated with enhanced physical stratification in the Bering Sea, which resulted in increased utilization of nutrients by phytoplankton and a reduction in the carbon dioxide released from the ocean to the atmosphere. Finally, multiple new records that demonstrate high productivity during intervals of laminated sediments, which represent brief anoxic events, indicate that interglacial climates precondition the Bering Sea for rapid oceanic and biological change.