Bioavailable nitrogen governs ocean productivity and carbon fixation by regulating phytoplankton growth and community composition. Nitrogen input primarily results from fixation, while denitrification and ANAMMOX removes bioavailable nitrogen in oxygen-depleted conditions. Traditionally considered limited to highly suboxic (i.e., O2 < 5 µM) waters, recent studies suggest fixed-nitrogen removal processes may extend beyond, elevating global nitrogen loss estimates. This study directly quantifies fixed-nitrogen loss across oxygen gradients (from 140 to 35 µM) along the Estuary and Gulf of St. Lawrence using N cycle tracers (N2/Ar, N* and N2O). Notably, we observe significant N2 production when O2 concentrations fall below 57-52 µM, including unexpected water column fixed-nitrogen removal processes above suboxia. Benthic N2 production remains unaffected under intensifying deoxygenation from 50 down to 34 µM, but sedimentary nitrification contribution to denitrification diminishes with intensifying deoxygenation. Combined, water column and benthic fixed-nitrogen removal processes drive N* anomalies and strong \({NO_3^-}\) deficiency in bottom waters. Additionally, observed O2 concentration threshold triggers N2O production, unveiling the profound impact of ocean deoxygenation on nitrogen cycling, challenging conventional expectations even at hypoxic concentrations.

Anthropogenic impacts and global changes have profound implications for natural ecosystems and may lead to their modification, degradation or collapse. Increases in the intensity of single stressors may lead to ecological thresholds, which can create abrupt shifts in biotic responses. The effects of multiple interacting stressors may create synergistic or antagonistic interactions, leading to non-linear responses. Here we combine both concepts -- ecological thresholds and interactions between multiple stressors -- to understand the effects of multiple interacting stressors along environmental gradients, and how this can affect the occurrence of thresholds. Using an experimental approach to investigate the effect of nutrient enrichment and saltwater intrusion on mortality in the zebra mussel, Dreissena polymorpha, we show that multiple stressors can create thresholds at lower levels of an environmental gradient. Our results reveal that by treating the concepts of ecological thresholds and multiple stressors individually we risk underestimating widespread anthropogenic impacts. Our results reveal a major shortcoming in how we currently investigate these two ecological concepts, as considering them separately may be causing underestimation of thresholds and stressor-interaction impacts.