Fixed nitrogen is the key limiting nutrient for primary production in much of the world ocean. While the concentration of dissolved inorganic nitrogen is below detection in open ocean subtropical gyres, micro-molar concentrations of dissolved organic nitrogen (DON) persist throughout the surface ocean. Similar to the broader dissolved organic matter (DOM) pool, the origin and molecular structure of this refractory material are largely unknown. The “microbial carbon pump” has emerged as a proposed framework for understanding microbial production of refractory DOM (R-DOM), but whether the same mechanisms also form R-DON remains unknown. Spectroscopy, radiocarbon (Δ14C), and concentration depth profiles of common amide-containing biomolecules together suggest that R-DON originates from proteinaceous material. Therefore, if an analogous microbial N pump exists, it most likely operates through microbial transformations of proteinaceous DON. Amino acids provide multiple potential tracers for this process, as the likely monomers contributing to amide R-DON material while their D-enantiomers (D-AAs) are unambiguous biomarkers for prokaryotic organisms.

Here, for the first time, we couple 14C ages, %D-AAs, and AA molar percent abundance (Mol%) data within DOM size-reactivity fractions, in order to assess the importance of bacterial sources in the formation of R-DON. We analyzed DOM isolates representing opposite ends of the size-reactivity continuum, from surface to deep ocean in the North Pacific Subtropical Gyre. Results from this first comparison of prokaryotic contributions to DON in isolated size-reactivity fractions strongly support the idea of a microbial N pump, with higher %D in R-DOM for almost all AAs. Quantification based on mass spectral data also allowed the unambiguous identification of three new D-AAs which had not previously been definitively identified in DOM, with oceanographically consistent depth profiles and relationships with Mol% and Δ14C, expanding the suite of environmentally relevant D-AAs from four to seven. Our results also suggest that several novel D-AA subgroupings could be developed as tracers: Mol% and %D trends with Δ14C for Alanine (Ala) suggest that D-Ala has independent cycling mechanisms from other D-AAs, and may be strongly linked to bacterial peptidoglycan sources and cycling, while D-Leucine and D-Valine have a particularly strong relationships with Δ14C, with a %D maximum in the twilight zone suggesting a link to sinking particles. Interestingly, traditional degradation indices based on Mol% changes appear to indicate bacterial inputs rather than age, though we show that consistent changes in Mol% are observed with age in both DOM size-reactivity fractions as well. Overall, our results are largely consistent with the dominant production of R-DON by prokaryotic organisms, and support a greater mechanistic understanding of degradation and recalcitrance through distinct relationships between %D, Mol%, and Δ14C for our expanded suite of environmentally relevant chiral AAs.

Female sperm whales (Physeter macrocephalus) are top predators in mesopelagic ecosystems, integrating chemical information about ecosystem structure through their diet. By studying proxies for diet and habitat use, we may be able to learn about how sperm whales' foraging and environment change through time. We measured stable isotopes of carbon (δ13C) and nitrogen (δ15N) from individual growth layer bands from the teeth of 10 female sperm whales, to track changes in diet and habitat use from ca. 1930 to 1960, and to investigate relationships with major environmental events. While El Niño events can strongly affect food webs, particularly in the Eastern Tropical Pacific, we found no clear linkage between bulk δ13C and δ15N records and El Niño records, possibly due to high variability among dentinal isotopic records or weak effects of El Niño events during the 20th century. However, we found that bulk δ13C and δ15N records fall into three temporal patterns, suggesting distinct groupings of whales with clearly differentiated life-long foraging strategies. Average bulk δ13C and δ15N values for each tooth were positively correlated, and we found individual whales generally separated in isotopic space according to temporal pattern groupings.

To determine if whales from each temporal pattern foraged in different regions with distinct isotopic baselines, we measured δ13C and δ15N values from individual amino acids (AAs) in a subset of samples. Amino acid isotope results clearly indicate that the bulk isotopic trend is due to baseline differences, as opposed to differences in diet or ecosystem structure. Specifically, our results indicate that whales from each of our identified groupings used different geographic regions, but had similar trophic positions, because essential- and source-AA isotope values correlated with bulk isotopic values, while both non-essential- and trophic-AAs had no relationship to bulk δ13C and δ15N values, respectively. Considering the bulk isotopic records together with CSIA data, we suggest that female sperm whales inhabiting the eastern Tropical Pacific likely had three different life-long foraging strategies under similar large-scale environmental constraints. Together, these results provide novel insight into social bonds among female sperm whales, since each social group shared the same habitat and diet over their life-time, but had separate trophic niches between adjacent social groups possibly due to environmental gradients.

Compound specific isotope analysis of individual amino acids (CSI-AA) is a powerful tool for tracing nitrogen (N) source and transformation in biogeochemical cycles. Specifically, the d15N value of phenylalanine (d15NPhe) represents an increasingly used proxy for baseline d15N values, with particular promise for paleoceanographic applications. Further, multiple studies have shown that d15N values of Phe and glutamic acid (Glu) can be coupled to provide precise estimates of trophic position (TP), while simultaneously decoupling baseline d15N values from the effects of trophic transfer. However, current derivatization/gas chromatography (GC) methods require expensive and relatively uncommon instrumentation, and have relatively low precision, posing significant issues for many potential applications. Here we present a new approach to CSI-AA based on a high-pressure liquid chromatography (HPLC) purification of underivatized amino acids, followed by offline elemental analysis-isotope ratio mass spectrometry (EA-IRMS). The utility of this approach is demonstrated in two different applications using natural materials: 1) the construction of a high precision paleoceanographic record from the d15NPhe in proteinaceous coral tissue and 2) the TP estimation of five marine organisms using the coupled d15N values of Glu and Phe. Both demonstrations were compared to values measured by an established GC based method. In all cases both methods produced equivalent values, however, values generated by the new HPLC/EA-IRMS approach had higher average precision in all cases. These results demonstrate that this approach represents a viable alternative to traditional GC based methods for d15NAA measurement and may therefore allow significant expansion of CSI-AA applications, requiring only commonly available instrumentation to produce high precision d15NAA values.

Throughout most of the world’s oceans, the bioavailability of marine dissolved organic nitrogen (DON) acts to limit marine primary production and thus is a key control on marine biogeochemical cycles and carbon sequestration. Understanding the processes that control marine DON cycling and availability, and by extension long-term carbon storage in the ocean, are thus of vital importance. Nevertheless, despite significant research, the mechanisms that lead to the accumulation of refractory DON (RDON) largely remain an enigma. In this thesis, I address this knowledge gap by investigating the source and degradation processes of traditionally studied younger, high molecular weight (HMW) DON and, for the first time, directly isolated older, low molecular weight (LMW) solid phase extracted (SPE) DON. I used a range of novel proxies and analytical approaches, including a new suite of D-amino acids and the first compound-specific specific amino acid measurements across the DON age/size spectrum. These powerful molecular level proxies for proteinaceous marine DON source and cycling allowed me to posit a new theory for bacterial control of RDON production. Finally, I pair these analyses with broader advanced solid state NMR techniques to present new information regarding overall marine DON chemical composition. Together, I use these data to propose a novel theory regarding production of the most refractory nitrogen containing molecules in the ocean, suggesting a fundamental paradigm shift in our understanding of the marine DON pool.

This thesis explored the use of bulk stable nitrogen (δ15N) and carbon (δ13C) isotope records from central Pacific proteinaceous deep-sea corals to create novel multi-millennial scale, sub-decadal resolution records of export production in this ocean region. I used live collected and subfossil coral specimens from both the central North Pacific Subtropical Gyre and Line Islands, to explore how coral isotope records can reconstruct climate forced changes in oceanography in very different areas of the Pacific, as well as to explore the limits of isotope records and coral skeleton preservation over time.

My thesis is organized in three parts. In the first chapter, I used three Kulamanamana haumeaae specimens from the main Hawaiian Islands to create ~5000 year record of North Pacific Subtropical Gyre (NPSG) export production δ15N and δ13C history, greatly expanding previous coral records (by ~4000 years) into the Mid-Holocene. These bulk isotope records revealed a very different oceanic regime during the mid-Holocene, with higher δ15N values indicative enhanced flow of source waters bearing signatures of denitrification from the eastern Pacific. Further, these new δ15N data confirmed that the extremely low present day δ15N values recorded by deep sea corals (∼8‰) are unprecedented for the NPSG, at least within the past five millennia. My data also suggested that the coupling or decoupling of δ15N and δ13C values may be an important new tool for understanding the ocean’s response to climate shifts. During most biogeochemical regimes shifts in this period, δ15N and δ13C trends were synchronous (similar to recent coral records), consistent with nutrient concentration and isotopic values acting together with plankton community shifts. However, several distinctly decoupled regimes also appear, indicating unusual mechanisms must have been at work. Phytoplankton species composition and nutrient source changes are hypothesized as the dominant mechanisms which controlled the coupling and decoupling of δ15N and δ13C values, likely primarily influenced by changing oceanographic conditions (e.g., stratification versus entrainment).

Chapter 2 was a methodological study, aimed at understanding the extent in time over which proteinaceous coral records from this region can be extended. Specifically, this study examined the relative changes due to degradation in subfossil corals, asking to what degree bulk versus compound-specific isotope values remain viable over multi-millennial time scales. This chapter examined one of the oldest deep sea K. haumeaae corals ever collected and the specifics of proteinaceous skeleton preservation over ~10 kyrs on the seafloor. I compared amino acid (AA) molar composition and isotope pattern changes, together with bulk δ13C and δ15N values in a subfossil (~9.6-11.6 kyrs BP) coral skeleton with a live collected specimen from the same region (Cross Seamount, Hawaii). I aimed to understand the effects of long-duration benthic oxic exposure on AAs and organic preservation in a subfossil coral. This study shows that while bulk isotope values are dramatically altered in a clear diagenetic outer zone, likely due to carbonate deposition and protein-structural degradation, most of the AA isotopic data remains almost entirely unaltered. Some notable results included nearly ~50% of glycine loss over ~10 kyrs, likely due to abiotic water based hydrolysis of the protein matrix, which nevertheless does not appreciably shift glycine’s isotope values. Despite changes to molecular composition, compound-specific isotope analysis of amino acids (CSIA-AA) data patterns remained almost entirely unchanged in inner coral layers. This indicates that in contrast to bulk isotope data, CSIA-AA data can reconstruct paleo-oceanographic biogeochemical and ecosystem information in subfossil corals beyond the clear outer diagenetic horizon, which is easily identifiable from an evaluation of C/N ratios and the ΣV degradation proxy.

My last chapter, switched back to paleo-oceanographic application for a very different region in the equatorial Pacific (Kingman Reef, Line Islands), one for which no prior proteinaceous coral records exist for the most recent two millennia. I examined the bulk δ13C and δ15N values and CSIA-AA δ15N values of both a live-collected and subfossil specimen to determine how export production isotope values have changed in response to climate shifts through the last ~2000 years. I proposed shifts in both δ13C and δ15N values are fundamentally driven by changing current dynamics, with an alternation between dominance by two water sources: western Pacific waters in the North Equatorial Countercurrent, and east Pacific waters in the North Equatorial Current with distinctly different nutrient concentrations and isotopic signatures. I utilized modern day El Niño Southern Oscillation patterns as a model to interpret the overall data, e.g. lower δ13C and δ15N values are indicative of more El Niño-like (warmer/wetter) conditions, with smaller-celled prokaryotic phytoplankton communities, and advection from the west Pacific. These changes in current dynamics recorded in my new records and are likely related to shifts in the Intertropical Convergence Zone, which is the major source of ocean stratification due to controls on precipitation and climate in the region.

Overall, by combining both new methodological work utilizing CSIA-AA to determine preservation state, and a novel set of coral archive records, my thesis expanded the utility of and potential application of deep sea coral archives, while specifically greatly expanding the biogeochemical history of oceanic change in two separate central Pacific regions. My thesis examined time periods and regions for which no such prior records exist, showing that both modern and subfossil specimens can provide high resolution information of Pacific Ocean variability through the Holocene.

Novel molecular isotope proxies measured in bivalve tissues represent a wide range of approaches for tracing ecological and climatological change through time and space. In comparison to other oceanographic data used to reconstruct modern and past biogeochemical cycles (e.g., marine sediment cores), chemical signals preserved in sessile, filter-feeding mollusks record climate and ecological signals directly in the context of a local environment. As global planetary warming accelerates climate change, marine ecosystems are exhibiting vastly different responses in terms of nutrient availability, phytoplankton community composition, and fundamental baseline biogeochemical cycles. Responses are particularly dynamic in highly variable coastal regions, underscoring the need for creating detailed, local historical and geologic records to understand how specific ecosystems and regions have responded to past climatic regime shifts. In my thesis, I have developed a set of novel approaches that couple bulk and compound-specific isotopes of amino acids (CSI-AA) in multiple bivalve tissues (shell and soft tissue) to reconstruct the integrated biogeochemical histories of environment in which bivalves grow. These new integrative isotope techniques are particularly valuable for environments with extreme variability (e.g., nearshore margins), those that are very difficult to sample (e.g., the deep-sea), or as new paleo reconstruction approaches which can be used with any reasonably preserved shell sample from either sediment cores or archaeological sites. Key findings include: the development of a novel suite of geochemical proxies for tracing chemosynthetic production in spatially heterogenous, deep-sea methane cold seeps, including for the first time quantifying the amount of nitrogen obtained through heterotrophic filter-feeding vs. chemoautotrophy in a chemosymbiotic bivalve (Chapter 1). These new geochemical proxies can now be applied to any remnant bivalve shell to reconstruct the biogeochemical histories of often times transient methane seep systems. Next, I tested and calibrated CSI-AA proxy approaches in shell matrix protein (Chapter 2) to develop preserved bivalve shell as a bioarchive across multiple environments. I compared isotope patterns in three different bivalve species from two coastal ecosystems (littoral and estuary) and show that the well-established ecological isotope proxies (niche width, baseline δ13C and δ15N, trophic level, and resource contribution) calibrated in bivalve soft tissue are directly transferred and preserved to shell protein matrix. However, in this chapter I also show that past CSI-AA trophic level calculations are fundamentally inaccurate and propose a new mollusk-specific trophic level equation required for both soft tissue and shell, linked to consistently compressed trophic discrimination factors and shell isotope routing. Finally, to determine the fidelity of isotope signals in ancient bivalve shell, I tested the preservation of bulk and CSI-AA isotope values and patterns in a suite of archaeological shells spanning a wide range of time periods between the middle to late Holocene and depositional preservation environments (Chapter 3). I show that even in subfossil shells whose bulk isotopes are strongly diagenetically altered, CSI-AA proxies are intact across almost 6 kyr of preservation. This work quantitively demonstrates CSI-AA preservation in the insoluble shell matrix protein of bivalves and sets the stage for using bivalve shell as bioarchives for CSI-AA parameters, to reconstruct paleoecological histories of nearshore systems with far more detail and precision that has ever been possible.

Marine dissolved organic matter (DOM) is the largest pool of actively cycling organic carbon and nitrogen in the ocean. Despite the importance of DOM in biogeochemical cycles, understanding of its composition and cycling is limited, in part due to difficulty in isolating representative material. Here we present a new approach to DOM isolation via a sequential combination of ultrafiltration and solid phase extraction, which allowed the selective isolation of two fractions with different compositions and reactivities. By limiting the influence of reactivity mixtures, we were able to directly investigate the composition and cycling of both labile and refractory DOM. Measurements of bulk isotopic (Δ14C, δ13C, δ15N) and compositional (C/N ratio, 13C and 15N NMR) parameters confirmed the uniqueness of the two fractions. The ultrafiltered high molecular weight (HMW) fraction was younger and dominated by carbohydrate-like and protein-like molecules. In contrast, the low molecular weight (LMW) fraction was older and dominated by alicyclic carbon and heterocyclic nitrogen containing molecules.

Spectroscopic studies have shown that the nitrogen-containing fraction of DOM (DON) is dominated by labile proteinaceous material. Despite this, other evidence suggests that a large fraction of DON is resistant to degradation. However, the paradigm of DON composition is based on measurements of HMW DON alone. Measurements of isotopic (δ15N, AA-Δ14C) and compositional (15N NMR, AA-D/L ratios) parameters of LMW DON provided a new perspective on the composition and cycling of DON in the ocean. Amino acids in LMW DON had both old radiocarbon ages and high D/L ratios, suggesting a bacterial mechanism capable of preserving proteinaceous material on millennial timescales. NMR analysis and nitrogen isotopic ratios demonstrated that LMW DON is dominated by heterocyclic nitrogen-containing molecules likely with intrinsically refractory molecular structures. These observations present a new paradigm for marine DON composition and suggest a mechanism for its long-term persistence.

Paleoceanographic studies typically employ stable nitrogen isotopic measurements of total combustible N (δ15NTN) from marine sediments to infer temporal changes in surface ocean marine N dynamics. However, δ15NTN is a highly non-specific measurement, susceptible to confounding issues including non-marine N sources, and offers little, if any, information regarding degradation and/or alteration of primary δ15N signals. Recent development of novel biomarker and microfossil based proxies provide new analytical capabilities to circumvent issues with δ15NTN.

Here, I explore the utility of compound specific nitrogen isotopic analysis of individual amino acids measurements (δ15NAA) of proteinaceous material isolated from marine sediments and deep-sea coral. With these data I infer past changes in nitrate δ15N fueling primary production at the base of the food web, ecosystem structure, and the extent of heterotrophic resynthesis influencing sedimentary and coral-skeletal δ15NTN records. Within sediments and corals, I demonstrate that δ15NAA based parameters reflect that of sinking particles and surface collected planktonic biomass and thus serve as a robust, albeit complex proxy to infer ecosystem and nutrient dynamic changes for at least the last few hundred years on the California Margin. In sediments I find an unexpected δ15N offset in major operational fractions and molecular compound classes. This result (1) challenges a major assumption organic nitrogen in sediments is dominated by amide functionality (i.e., proteinaceous N) and (2) highlights possible issues with standard protein hydrolysis techniques.

Lastly, I analyze the δ15N composition of foraminifera-bound N (FB-δ15N) to assess the Plio-Pleistocene history of whole ocean nitrogen dynamics from the oligotrophic North Pacific for the last 5 million years. This region contains calcareous nannofossil rich sediments, ideal for preserved FB-δ15N records. These new FB-δ15N records support an expansion of suboxic regions, consistent with previous evidence, but also indicate that the whole ocean δ15N budget was stable throughout the Plio-Pleistocene. This latter result implies that kinetic isotopic fractionations of marine N transformations and the ratio of pelagic to benthic denitrification, must not have changed during the Plio-Pleistocene transition, despite drastic global scale changes in sea level, shelf area, and volume of the major oxygen deficient zones at this time.

Compound-specific nitrogen (N) isotopic analysis (δ15N) of amino acids (AA), or CSI-AA, is a novel approach to understand N cycling. We expand upon initial observations at a productive, hypoxic margin to provide insight into source and transformation of sedimentary organic nitrogen from varied depositional conditions in a complex N cycling zone, the eastern tropical Pacific. δ15NAA patterns are generally well preserved, matching original mixed plankton inputs with some evidence for microbial degradation. There is appreciable δ15N offset between total N (bulk) and total AA N, where AA N is enriched. Source AA, previously understood to record changing baseline δ15N, are fundamentally different in marine sediments. We use modeling to rigorously test whether the total AA δ15N instead tracks changing baseline δ15N. Lastly we employ this new understanding of δ15NAA in sediments to compare with published whole sediment (δ15Nbulk) records, demonstrating the utility and wealth of information provided by CSI-AA.