Glacial/interglacial control of terrigenous and biogenic fluxes in the deep ocean off a high input, collisional margin: A 139 kyr-record from New Zealand

Sediment input to the deep ocean, off the New Zealand collisional plate boundary, is large and variable as revealed by a 139 kyr-old record from the 34.92 m-long core, MD97 2121. Mass accumulation rates (MARs) were derived for terrigenous and biogenic (carbonate and silica) sedimentary components, with temporal control from a stable isotope age model verified by tephra and radiocarbon ages. Terrigenous MARs changed in phase with glacio-eustatic fluctuations of sea level. Highest rates (≥ 30 g/cm2/kyr) coincided with the late regressive-lowstand-early transgressive phase of marine isotope stage (MIS) 2 when rivers discharged at or near the shelf edge. Lesser, but still high terrigenous rates (20–30 g/cm2/kyr) characterized peak warm phases of early MIS 5 and 1, when a strengthened Subtropical Inflow probably introduced sediment from fluvial and seabed sources to the north of MD97 2121. However, during prolonged highstands the flux declined as more sediment was retained on a tectonically subsiding shelf with a shore-parallel current. Paradoxically, the lowest MARs (10–5 g/cm2/kyr) occurred in MIS 4; either sea level was too high to allow much sediment to escape off-shelf, or the fluvial input was modest, or both these factors.Despite the terrigenous dominance, biogenic MARs are high, and are 2–4 times larger than SW Pacific Ocean rates. MARs were lowest during MIS 4 and 3, possibly due to reduced marine production under a lowered input of fluvial micronutrients. In contrast, MARs increased irregularly though glacial maxima to peak at 5–6 g/cm2/kyr carbonate and 2–2.5 g/cm2/kyr silica in early-mid MIS 5 and 1. Such rates reflected the interaction of macronutrient-rich subantarctic waters, sourced from the south, with local, micronutrient-rich subtropical waters. Production also responded to warmer temperatures, elevated nutrient runoff and a more stratified surface ocean. Later in MIS 5 and 1, fluxes reduced as less productive subtropical waters prevailed. That biogenic and aeolian MAR profiles are out of phase, suggests Fe-fertilization by aerosolic dust was not a major influence.