Amino acids (AA) represent the most abundant identifiable biomolecule class in marine dissolved organic nitrogen (DON) and provide powerful proxies for DON degradation state. In particular, the D-enantiomers of AA (D-AA) are known to be derived mainly from bacteria, making them ideal direct tracers for bacterially derived N. However, despite the widespread use of D-AA tracers, it remains unclear if the observed accumulation of different D-AA species in the ocean indicates that most DON arises from direct bacterial sources or from continual bacterial alteration of eukaryotic algal material. This difference has major implications for our understanding of sources and cycling mechanisms of DON in the ocean. Here, we present the most extensive D-AA suite ever reported in younger, high molecular weight (HMW) DON contrasted with older, low molecular weight (LMW) solid phase extracted (SPE) DON from the central Atlantic and Pacific Oceans. We evaluate D-AA in these two contrasted MW fractions in the context of multiple common AA-based proxies and bulk DOM radiocarbon (Δ¹⁴C) data. Specifically, we assess if D-AA in HMW and LMW SPE-DON are most consistent with 1) preformed bacterial source signals, 2) progressive bacterial degradation/alteration of eukaryotic algal sources, or 3) gradual, continued resynthesis/addition of new bacterial biomolecules during ocean circulation. Our results suggest that AA-containing molecules in HMW and LMW SPE-DON fractions are almost entirely distinct, with independent bacterial sources and degradation mechanisms. In HMW DON, all measured indices support a surface-produced, semi-labile component which is progressively altered in the mesopelagic with increasing radiocarbon age. In contrast, for LMW SPE-DON, AA-based proxies showed conflicting results. Some proxies (D/L ratios of most D-AA, non-protein AA, mol% Gly, and %C-AA) indicated LMW SPE-DON was less labile and more degraded than HMW DON. However other proxies (D/L-Ala, the N isotope based ΣV parameter, and the commonly used DI index) indicated equivalent or even less degradation and resynthesis in the isolated LMW material compared to HMW material, suggesting a disconnect in the underlying mechanisms reflected by different proxies. Finally, AA composition and degradation state in the subsurface samples of both HMW and LMW DON varied little with increasing radiocarbon age, suggesting that HMW and LMW pools may cycle independently. Together, our results suggest that AA DON sources, while almost entirely bacterial, may be more diverse than previously believed, and that much of the hydrolysable AA pool in the ocean may not be derived from proteinaceous material. Overall, these observations support the microbial nitrogen pump idea, but with compositionally unique refractory components in both HMW and LMW material which resist degradation over millennial timescales.

氨基酸 (AA) 代表了海洋溶解有机氮 (DON) 中最丰富的可识别生物分子类别，并为 DON 降解状态提供了强大的代理。特别是，已知 AA 的 D-对映异构体 (D-AA) 主要来源于细菌，使其成为细菌衍生 N 的理想直接示踪剂。然而，尽管广泛使用 D-AA 示踪剂，但仍不清楚是否观察到海洋中不同 D-AA 物种的积累表明，大多数 DON 来自直接细菌来源或来自真核藻类物质的持续细菌改变。这种差异对我们了解海洋中 DON 的来源和循环机制具有重要意义。在这里，我们展示了有史以来最广泛的 D-AA 套件，与较旧的相比，更年轻的高分子量 (HMW) DON，来自中大西洋和太平洋的低分子量 (LMW) 固相萃取 (SPE) DON。我们在多个常见的基于 AA 的代理和散装 DOM 放射性碳（Δ¹⁴三）数据。具体来说，我们评估 HMW 和 LMW SPE-DON 中的 D-AA 是否与 1) 预先形成的细菌源信号最一致，2) 真核藻源的渐进性细菌降解/改变，或 3) 逐渐、持续的再合成/添加新细菌海洋环流过程中的生物分子。我们的结果表明，HMW 和 LMW SPE-DON 馏分中的含 AA 分子几乎完全不同，具有独立的细菌来源和降解机制。在 HMW DON 中，所有测量的指数都支持表面产生的半不稳定成分，随着放射性碳年龄的增加，该成分在中层海洋中逐渐改变。相比之下，对于 LMW SPE-DON，基于 AA 的代理显示出相互矛盾的结果。一些代理（大多数 D-AA、非蛋白质 AA、mol% Gly 的 D/L 比率，和 %C-AA) 表明 LMW SPE-DON 比 HMW DON 更不稳定且更易降解。然而，其他代理（D/L-Ala、基于 N 同位素的 ΣV 参数和常用的 DI 指数）表明，与 HMW 材料相比，孤立的 LMW 材料中的降解和再合成相当或什至更少，这表明所反映的潜在机制存在脱节通过不同的代理。最后，随着放射性碳年龄的增加，HMW 和 LMW DON 的地下样品中的 AA 组成和降解状态变化不大，这表明 HMW 和 LMW 池可能独立循环。总之，我们的研究结果表明，AA DON 来源虽然几乎完全是细菌，但可能比以前认为的更加多样化，并且海洋中大部分可水解的 AA 池可能不是来自蛋白质物质。全面的，