Upwelling of nutrient-rich deep waters make eastern boundary upwelling systems (EBUSs), such as the Humboldt Current system, hot spots of marine productivity. Associated settling of organic matter to depth and consecutive aerobic decomposition results in large subsurface water volumes being oxygen depleted. Under these circumstances, organic matter remineralisation can continue via denitrification, which represents a major loss pathway for bioavailable nitrogen. Additionally, anaerobic ammonium oxidation can remove significant amounts of nitrogen in these areas. Here we assess the interplay of suboxic water upwelling and nitrogen cycling in a manipulative offshore mesocosm experiment. Measured denitrification rates in incubations with water from the oxygen-depleted bottom layer of the mesocosms (via ¹⁵N label incubations) mostly ranged between 5.5 and 20 nmol N₂ L⁻¹ h⁻¹ (interquartile range), reaching up to 80 nmol N₂ L⁻¹ h⁻¹. However, actual in situ rates in the mesocosms, estimated via Michaelis–Menten kinetic scaling, did most likely not exceed 0.2–4.2 nmol N₂ L⁻¹ h⁻¹ (interquartile range) due to substrate limitation. In the surrounding Pacific, measured denitrification rates were similar, although indications of substrate limitation were detected only once. In contrast, anammox (anaerobic ammonium oxidation) made only a minor contribution to the overall nitrogen loss when encountered in both the mesocosms and the Pacific Ocean. This was potentially related to organic matter C $/$ N stoichiometry and/or process-specific oxygen and hydrogen sulfide sensitivities. Over the first 38 d of the experiment, total nitrogen loss calculated from in situ rates of denitrification and anammox was comparable to estimates from a full nitrogen budget in the mesocosms and ranged between ∼ 1 and 5.5 µmol N L⁻¹. This represents up to ∼  20 % of the initially bioavailable inorganic and organic nitrogen standing stocks. Interestingly, this loss is comparable to the total amount of particulate organic nitrogen that was exported into the sediment traps at the bottom of the mesocosms at about 20 m depth. Altogether, this suggests that a significant portion, if not the majority of nitrogen that could be exported to depth, is already lost, i.e. converted to N₂ in a relatively shallow layer of the surface ocean, provided that there are oxygen-deficient conditions like those during coastal upwelling in our study. Published data for primary productivity and nitrogen loss in all EBUSs reinforce such conclusion.

中文翻译：

---

在以反硝化作用为主的秘鲁沿海环境中，响应上升流的氮损失过程——一种中观方法

富含营养的深水上升流使东部边界上升流系统 (EBUS)，例如洪堡海流系统，成为海洋生产力的热点。与深度相关的有机物沉降和连续的有氧分解导致大量地下水体被氧气耗尽。在这种情况下，有机物再矿化可以通过反硝化作用继续进行，这是生物可利用氮的主要损失途径。此外，厌氧铵氧化可以去除这些区域的大量氮。在这里，我们评估了在操纵性近海中观实验中低氧水上涌和氮循环的相互作用。在与来自中间宇宙缺氧底层的水一起孵化时测量的反硝化率（通过¹⁵N 标记孵育）主要在 5.5 和 20 nmol N ₂  L ^-1  h ^{-1 之间}（四分位距），达到 80 nmol N ₂  L ^-1  h ^-1。然而，通过 Michaelis-Menten 动力学标度估计，中宇宙中的实际原位速率很可能不超过 0.2–4.2 nmol N ₂  L ^-1  h ^-1（四分位距）由于底物限制。在周围的太平洋，测得的反硝化率相似，尽管只检测到一次底物限制的迹象。相比之下，厌氧氨氧化（厌氧氨氧化）在中世界和太平洋中遇到时对整体氮损失的贡献很小。这可能与有机物 C 相关 $/$ N 化学计量和/或工艺特定的氧和硫化氢敏感性。在实验的前 38 天，根据反硝化和厌氧氨氧化的原位速率计算的总氮损失与中间宇宙中全氮预算的估计值相当，范围在~  1 到 5.5  µ mol N L ^{-1 之间}。这占 最初生物可利用的无机和有机氮常备库存的约 20%。有趣的是，这种损失与出口到中层世界底部约 20 m 处沉积物捕集器中的颗粒有机氮总量相当深度。总而言之，这表明，如果存在缺氧条件，例如可输出到深海的大部分氮，即使不是大部分，也已经丢失，即在相对较浅的表层海洋中转化为 N ₂在我们的研究中沿海上升流期间的那些。所有 EBUS 中初级生产力和氮损失的已发布数据证实了这一结论。