The use of gas chromatography isotope ratio mass spectrometry (GC-IRMS) for compound specific stable isotope analysis is an underutilized technique because of the complexity of the instrumentation and high analytical costs. However stable isotopic data, when coupled with concentration measurements, can provide additional information on a compounds production, transformation, loss, and cycling within the biosphere and atmosphere. A GC-IRMS system was developed to accurately and precisely measure δ13C values for numerous oxygenated volatile organic compounds (OVOCs) having natural and anthropogenic sources. The OVOCs include methanol, ethanol, acetone, methyl ethyl ketone, 2-pentanone, and 3-pentanone. Guided by the requirements for analysis of trace components in air, the GC-IRMS system was developed with the goals of increasing sensitivity, reducing dead-volume and peak band broadening, optimizing combustion and water removal, and decreasing the split ratio to the IRMS. The technique relied on a two-stage preconcentration system, a low-volume capillary reactor and water trap, and a balanced reference gas delivery system. Measurements were performed on samples collected from two distinct sources (i.e. biogenic and vehicle emissions) and ambient air collected from downtown Miami and Everglades National Park. However, the instrumentation and the method have the capability to analyze a variety of source and ambient samples. The measured isotopic signatures that were obtained from source and ambient samples provide a new isotopic constraint for atmospheric chemists and can serve as a new way to evaluate their models and budgets for many OVOCs. In almost all cases, OVOCs emitted from fuel combustion were enriched in 13C when compared to the natural emissions of plants. This was particularly true for ethanol gas emitted in vehicle exhaust, which was observed to have a uniquely enriched isotopic signature that was attributed to ethanol’s corn origin and use as an alternative fuel or fuel additive. Results from this effort show that ethanol’s unique isotopic signature can be incorporated into air chemistry models for fingerprinting and source apportionment purposes and can be used as a stable isotopic tracer for biofuel inputs to the atmosphere on local to regional scales.

<p>The rationale of this study is to apply light stable isotope chemistry towards investigations that require temporally high-resolution data. High-resolution (or high sampling frequency) data sets, are critical for testing environmental and/or paleoenvironmental hypotheses that seek to explain processes occurring over rapid or short time intervals. The investigation of climate variation (e.g., seasonality, El Niño, deglaciation), animal migration and physiology, and disturbance ecology (e.g., fire, flooding) benefits from the recovery of proxy information at decadal to subannual resolutions. The type of material used also dictates a spatial scale. Herein are presented four studies that utilize high-resolution light stable isotope profiles with contrasting temporal and spatial scales. The first study employs advances in three-dimensional computer-controlled micromilling to recover ~daily to weekly deposited carbonate from small (~1 cm) mollusc shells. Stable oxygen isotope values from freshwater mollusc shells are predictably related to the local environment of growth using previously published temperature-fractionation relationships, providing a paleoclimate record of temperature and precipitation. The second study investigates variation in stable carbon isotope values from Aplodinotus grunniens otoliths, for which high-resolution patterns were critical in assessing metabolic rate as the governing control. The third study employs high-resolution stable oxygen and carbon isotope values to determine chinook salmon (Oncorhynchus tshawytscha) seasonal and ontogenetic migration in Lake Ontario and its tributaries. Lastly, high-resolution stable hydrogen and carbon isotope values of chitin derived from Mexican free-tailed bat (Tadarida brasiliensis) guano are presented, providing a record of abrupt climate change. Thus, this thesis reports on promising new research avenues for paleoclimatology, paleoecology, and modern ecology.

This thesis describes a method for calculating lake evaporation as a proportion of water inputs (E/I) for large surface water bodies, using stable isotope ratios of oxygen (18O/16O) and hydrogen (2H/1H) in water. Evaporation as a proportion of inflow (E/I) is calculated for each Laurentian Great Lake using a new dataset of 516 analyses of δ18O and δ2H in waters sampled from 75 offshore stations during spring and summer of 2007. This work builds on previous approaches by accounting for lake effects on the overlying atmosphere and assuming conservation of both mass and isotopes (18O and 2H) to better constrain evaporation outputs. Results show that E/I ratios are greatest for headwater Lakes Superior and Michigan and lowest for Lakes Erie and Ontario, controlled largely by the magnitude of hydrologic inputs from upstream chain lakes. For Lake Superior, stable isotopes incorporate evaporation over the past century, providing long-term insights to the lake’s hydrology that may be compared to potential changes under a future – expectedly warmer – climate. Uncertainties in isotopically derived E/I are comparable to conventional energy and mass balance uncertainties. Isotope-derived E/I values are lower than conventional energy and mass balance estimates for Lakes Superior and Michigan. The difference between conventional and isotope estimates may be explained by moisture recycling effects. The isotope-based estimates include only evaporated moisture that is also advected from the lake surface, thereby discounting moisture that evaporates and subsequently reprecipitates on the lake surface downwind as recycled precipitation. This shows an advantage of applying an isotope approach in conjunction with conventional evaporation estimates to quantify both moisture recycling and net losses by evaporation. Depth profiles of 18O/16O and 2H/1H in the Great Lakes show a lack of isotopic stratification in summer months despite an established thermocline. These results are indicative of very low over-lake evaporation during warm summer months, with the bulk of evaporation occurring during the fall and winter. This seasonality in evaporation losses is supported by energy balance studies. For Lakes Michigan and Huron, the isotope mass balance approach provides a new perspective into water exchange and evaporation from these lakes. This isotope investigation shows that Lake Michigan and Lake Huron waters are distinct, despite sharing a common lake level. This finding advocates for the separate consideration of Lake Michigan and Lake Huron in future hydrologic studies.

<p>The rationale of this study is to apply light stable isotope chemistry towards investigations that require temporally high-resolution data. High-resolution (or high sampling frequency) data sets, are critical for testing environmental and/or paleoenvironmental hypotheses that seek to explain processes occurring over rapid or short time intervals. The investigation of climate variation (e.g., seasonality, El Niño, deglaciation), animal migration and physiology, and disturbance ecology (e.g., fire, flooding) benefits from the recovery of proxy information at decadal to subannual resolutions. The type of material used also dictates a spatial scale. Herein are presented four studies that utilize high-resolution light stable isotope profiles with contrasting temporal and spatial scales. The first study employs advances in three-dimensional computer-controlled micromilling to recover ~daily to weekly deposited carbonate from small (~1 cm) mollusc shells. Stable oxygen isotope values from freshwater mollusc shells are predictably related to the local environment of growth using previously published temperature-fractionation relationships, providing a paleoclimate record of temperature and precipitation. The second study investigates variation in stable carbon isotope values from Aplodinotus grunniens otoliths, for which high-resolution patterns were critical in assessing metabolic rate as the governing control. The third study employs high-resolution stable oxygen and carbon isotope values to determine chinook salmon (Oncorhynchus tshawytscha) seasonal and ontogenetic migration in Lake Ontario and its tributaries. Lastly, high-resolution stable hydrogen and carbon isotope values of chitin derived from Mexican free-tailed bat (Tadarida brasiliensis) guano are presented, providing a record of abrupt climate change. Thus, this thesis reports on promising new research avenues for paleoclimatology, paleoecology, and modern ecology.

Moose populations in southern Manitoba have declined. Although the primary cause of this decline is unclear, wolf predation is considered a main contributor because wolves have limited moose populations elsewhere. To help understand the role of wolf predation in limiting moose populations we reconstructed the proportional diet of wolves using stable isotope analysis and a Bayesian stable isotope mixing model (MixSIAR). We tested the assumptions of MixSIAR by running sensitivity analyses on trophic discrimination and prior information. We observed differences in wolf diet in areas where moose and deer coexist and are spatially separated, with changes both seasonally and annually. Our results suggest white-tailed deer were preyed on during winter, even in areas where deer are locally rare, suggesting prey catchability is more important than local prey density. Seasonal changes in prey accessibility strongly affect wolf predation rates, and manipulating alternative prey populations could mitigate predation impacts on moose. / October 2015

Underlying all applications of sulfur isotope analyses is our understanding of isotope systematics. This dissertation tests some fundamental assumptions and assertions, drawn from equilibrium theory and a diverse body of empirical work on biochemical kinetics, as applied to the multiple sulfur isotope systematics of microbial sulfate reduction. I take a reductionist approach, both in the questions addressed and experimental approaches employed. This allows for a mechanistic, physically consistent interpretation of geological and biological sulfur isotope records. The goal of my work here is to allow interpreters a more biologically, chemically and physically parsimonious framework to decipher the signals coded in modern and ancient sulfur isotope records. / Earth and Planetary Sciences

This thesis describes a method for calculating lake evaporation as a proportion of water inputs (E/I) for large surface water bodies, using stable isotope ratios of oxygen (18O/16O) and hydrogen (2H/1H) in water. Evaporation as a proportion of inflow (E/I) is calculated for each Laurentian Great Lake using a new dataset of 516 analyses of δ18O and δ2H in waters sampled from 75 offshore stations during spring and summer of 2007. This work builds on previous approaches by accounting for lake effects on the overlying atmosphere and assuming conservation of both mass and isotopes (18O and 2H) to better constrain evaporation outputs. Results show that E/I ratios are greatest for headwater Lakes Superior and Michigan and lowest for Lakes Erie and Ontario, controlled largely by the magnitude of hydrologic inputs from upstream chain lakes. For Lake Superior, stable isotopes incorporate evaporation over the past century, providing long-term insights to the lake’s hydrology that may be compared to potential changes under a future – expectedly warmer – climate. Uncertainties in isotopically derived E/I are comparable to conventional energy and mass balance uncertainties. Isotope-derived E/I values are lower than conventional energy and mass balance estimates for Lakes Superior and Michigan. The difference between conventional and isotope estimates may be explained by moisture recycling effects. The isotope-based estimates include only evaporated moisture that is also advected from the lake surface, thereby discounting moisture that evaporates and subsequently reprecipitates on the lake surface downwind as recycled precipitation. This shows an advantage of applying an isotope approach in conjunction with conventional evaporation estimates to quantify both moisture recycling and net losses by evaporation. Depth profiles of 18O/16O and 2H/1H in the Great Lakes show a lack of isotopic stratification in summer months despite an established thermocline. These results are indicative of very low over-lake evaporation during warm summer months, with the bulk of evaporation occurring during the fall and winter. This seasonality in evaporation losses is supported by energy balance studies. For Lakes Michigan and Huron, the isotope mass balance approach provides a new perspective into water exchange and evaporation from these lakes. This isotope investigation shows that Lake Michigan and Lake Huron waters are distinct, despite sharing a common lake level. This finding advocates for the separate consideration of Lake Michigan and Lake Huron in future hydrologic studies.

Ce mémoire décrit les corps dolomitiques des zones de Ranero et El Moro (vallée de la Karantza, zone Cantabrique, NO de l’Espagne) et précise les variations temporelles et latérales de leurs attributs pétrographiques et géochimiques. Les corps dolomitiques sont portés par les calcaires Albiens, déposés sur la marge du bassin Basque-Cantabrique en période d’intense subsidence. Les dolomies sont formées par replacement et par cémentation, et précédées et suivies par divers types de ciment calcitique. L’étude pétrographique, minéralogique et géochimique (XRD, ICP, XRF, isotopes stables et Sr) est conduite le long de sections transverses sur les corps dolomitiques et permet de comparer les caractéristiques de plusieurs stades de circulations hydrothermales. Deux épisodes contrastés de dolomitisation sont identifiés. Les dolomies précoces sont ferreuses, très localement associées à une minéralisation de type MVT, appauvries en δ18O (de -14 à -10‰ V-PDB) et remplacent largement les calcaires massifs en générant des zebras. Les dolomies tardives sont non-ferreuses, plus sévèrement appauvries en δ18O (-19 to -15‰ V-PDB), et ne semblent pas remplacer le calcaire mais, au contraire, les dolomies précoces. Toutes ces dolomies sont pratiquement stœchiométriques (49.76 à 51.59 mole% CaCO3). Leurs inclusions fluides ont piégé des saumures de haute température (Th de 120 à 200°C). Leur contenu en Sr, radiogénique, suggère que les fluides responsables de la dolomitisation ont préalablement circulé à travers des roches silicoclastiques. La texture comme les propriétés pétrophysiques des dolomies sont largement affectées par les déformations cataclastiques et un épisode tardif de dédolomitisation (météorique).Le premier épisode de dolomitisation résulte probablement de l’expulsion des fluides issus de la compaction du bassin adjacent et de leur migration le long des fractures affectant la marge de la plateforme carbonatée Albienne. Ces fluides précoces étaient riches en Mg, Fe et peut-être légèrement acides pour pouvoir remplacer les calcaires. Les fluides responsables du deuxième épisode de dolomitisation sont pauvres en Fe, paraissent plus chauds et en relation avec une anomalie thermique. / This study documents the temporal and lateral variation in petrographic and geochemical signatures of fault-related dolomite bodies in the Ranero and El-Moro areas (Karrantza valley, Cantabrian mountains; NW Spain). These dolomite bodies are hosted in Albian carbonates, which were deposited at the margin of the Basque-Cantabrian Basin during an intense rift-related subsidence. Fluid circulations generated replacive and cementing dolomites, paragenetically predated and followed by various calcite cements. Petrography, mineralogical and geochemical systematics (XRD, ICP, XRF, stable and Sr isotopes) along sections cutting the dolostone bodies document successive hydrothermal stages. Two contrasting dolomite formation events are evidenced. Early dolomites are ferroan, locally associated with MVT mineralisation, δ18O depleted (-14 to -10‰ V-PDB) and mostly replace limestone producing abundant zebra lithotypes. Later dolomites are non-ferroan, severely δ18O depleted (-19 to -15‰ V-PDB), and do not replace limestones but rather previous dolomites. Dolomites are generally stoichiometric (49.76 to 51.59 M% CaCO3). Fluid inclusions record high temperature brines (Th 120 to 200°C). Sr isotope data suggest that the dolomitising fluids interacted upstream with siliciclastic lithologies. The dolomite fabric and its petrophysical properties are variably altered through cataclastic deformation and late (meteoric) dedolomitisation.The first episode of pervasive ferroan dolomitisation probably resulted from compactional dewatering of basinal fluids from the nearby Basque trough and hydrodynamic fluid flow along the fractures in the Albian carbonate platform. These early fluids must have been Mg, Fe-rich and slightly acidic (limestone-replacive). The second episode of very hot and localized dolomitisation may be related to a thermal anomaly and/or convective flow of Fe-poor fluids.

De nombreux sols forestiers français sont acides et pauvres en éléments nutritifs et les forêts françaises font face aujourd'hui à des pressions extérieures croissantes. Ces nouvelles contraintes sont d'ordre nutritionnel et/ou sylvicole et/ou climatique et risquent d'impacter la durabilité des écosystèmes forestiers à faible fertilité minérale. Il est donc très important de comprendre et quantifier les processus qui régissent les cycles des éléments nutritifs dans ces écosystèmes afin de prédire au mieux leurs évolutions possibles face à de telles contraintes et ainsi aider les gestionnaires forestiers à prendre des décisions réfléchies pour assurer leur durabilité. L'objectif principal de cette thèse était de comprendre le cycle du calcium et du magnésium dans un écosystème forestier productif bien que croissant sur un sol très faiblement pourvu en Ca et Mg. Pour cela, nous avons établi dans un peuplement de hêtre âgé de 35 ans des bilans de flux « entrées-sorties » basés sur des mesures classiques de pools et flux entre 2003 et 2008. Puis une expérience de multi-traçage isotopique (2H, 15N, 26Mg et 44Ca) a ensuite été conduite: les traceurs ont été apportés à la surface du sol en avril 2010 sous forme soluble et leur progression a été suivie dans l'écosystème jusqu'en mars 2012. Les résultats ont mis en évidence le rôle primordial que joue la matière organique dans les cycles du Mg et du Ca, en leur conférant notamment un caractère très conservatif. Le 26Mg et le 44Ca apportés ont été captés dans la fine couche de litière au sol puis lentement libérés. Dans les horizons minéraux, la matière organique du sol, qui semble contribuer à la plus grande part de CEC du sol, a fortement contribué à ralentir la migration du Mg et du Ca à travers le sol. Par rapport au Mg, le calcium apporté a été plus intensément retenu dans la litière, et moins intensément lixivié à travers le sol en raison de sa plus grande affinité pour la matière organique. Malgré les flux préférentiels d'eau mis en évidence par le traçage au 2H, aucune perte par drainage de 26Mg ou 44Ca n'a été mesurée. Deux années après l'apport, il restait dans la litière 8% du 26Mg et 33% du 44Ca apportés tandis que, respectivement, 52% et 46% ont été retrouvés dans les 10 premiers cm du sol. Les traceurs, rapidement absorbés par les racines fines ont migré dans l'arbre à des vitesses très différentes: les feuilles étaient enrichies en 15N dès le mois suivant l'apport des traceurs, tandis que seuls de très faibles enrichissements en 26Mg (mais pas de 44Ca) ont été mesurés dans la canopée en 2012. Le bilan en 2012 montre les arbres ont absorbé 27% de 26Mg et 20% de 44Ca et qu'ils sont majoritairement concentrés dans les tissus de la base du tronc. L'utilisation de traceurs isotopiques a permis de mettre en évidence la lenteur de la redistribution de Mg et encore plus de Ca dans l'écosystème, d'une part de la litière vers les sols et d'autre part des racines vers les canopées. Ce comportement est attribué à l'affinité de ces ions pour les charges carboxyliques portées par la matière organique. Enfin, la dilution isotopique mesurée dans les différents compartiments de l'écosystème à l'issue de deux années a permis de calculer des bilans et ainsi de valider les pertes de Mg calculées par l'approche classique de « bilan de fertilité minérale ». En revanche, les pertes de Ca calculées par cette même approche ont été mises en défaut, démontrant la contribution d'une source externe (absorption par les feuilles, altération…) ou interne (mobilisation du stock des tissus ligneux), non prise en compte dans l'approche classique. L'ensemble de ces travaux souligne l'intérêt de l'approche « traçage isotopique » en complément des approches conventionnelles afin d'étudier la fertilité minérale des écosystèmes forestiers. / Many forest soils in France are acid and nutrient poor. French forests endure today increased external constraints which may be nutritional and/or silvicultural and/or climatic. These constraints may strongly impact the sustainability of nutrient-poor ecosystems. It is thus important to understand and quantify the processes which govern nutrient cycling in these ecosystems to better predict soil fertility changes and help forest managers make sound decisions to sustain these ecosystems.The main objective of this thesis was to study magnesium and calcium cycles in a productive forest ecosystem with low available magnesium and calcium pools in the soils. For this purpose, in a 35-year-old beech stand, we computed input-output nutrient budgets between 2003 and 2008 using conventional approaches to measure nutrient pools and fluxes. A multi-isotopic (2H, 15N, 26Mg et 44Ca) tracing experiment was then carried out in the same plot: tracers were sprayed on the forest floor in April 2010 as soluble salts, tracer transfer was then monitored in the whole ecosystem until March 2012 (litter-layer, exchangeable pools, soil solutions, above-ground biomass…).The results evidenced the predominant role of organic matter in the very conservative cycling of Mg and Ca in this nutrient-poor soil. Indeed, the applied 26Mg and 44Ca were very rapidly retained in the fine litter-layer by ion-exchange processes. In the soil profile, most of the CEC was probably due to soil organic matter which played an important role in the retention of Mg and Ca in the soil. Compared to magnesium, calcium was more strongly retained in the litter layer and transferred more slowly in the soil profile because of the higher affinity of calcium for organic CEC. Despite preferential water flow evidenced by the water tracing experiment (2H), no leaching of 26Mg or 44Ca was observed during the study period. Two years after the application of tracers, 8% of applied 26Mg and 33% of applied 44Ca was still retained in the litter layer while, respectively, 52% and 46% of applied tracers were found in the first 10cm of the soil profile. Tracers were rapidly absorbed by the fine roots but were transferred from the roots to the canopy at very different velocities: leaves were enriched in 15N one month after the application of tracers while only very small 26Mg enrichments and not 44Ca were observed in the canopy in 2012. Tracer budgets in 2012 showed that trees absorbed 27% of 26Mg and 20% of 44Caand that tracers were mainly concentrated at the base of the bole. The use of isotope tracers evidenced the slow distribution of magnesium and the slower distribution of calcium in the ecosystem, on the one hand from the litter pool to the soil profile and on the other hand from the roots to the canopy. This behavior was attributed to the affinity of the cations for organic matter functional groups such as carboxyl groups.Finally the isotopic dilution technique of the applied tracers measured in the different ecosystem compartments after two years of monitoring enabled us to validate the Mg losses computed by nutrient budgets. However, Ca depletion suggested by nutrient budgets were proven to be nil thus evidencing the contribution of an external (foliar absorption, weathering) and/or internal (mobilization of nutrients stored in ligneous tissues) sources which are currently not taken into account in conventional approaches.Overall the results of the in situ ecosystem scale multi-isotopic tracing experiment show the added value of the “isotopic tracing” approach to study nutrient pools and fluxes in forest ecosystems in addition to conventional approaches.

The late Cambrian-early Ordovician transition contains several trilobite extinctions. The first of these extinctions (the end-Marjuman) is thought to coincide with the Steptoean Positive Carbon Isotope Excursion or SPICE, a large and rapid excursion in the marine carbon isotope record. This excursion, which is expressed in sedimentary successions globally, is thought to represent a large perturbation to the carbon cycle during this time. Additionally, a limited amount of carbon isotope data from the Deadwood Formation in the Black Hills of South Dakota suggests the possibility of a small negative ẟ¹³C excursion near the extinction and preceding the SPICE. Previous high-resolution biostratigraphy has identified an expanded record of extinction event within the Nolichucky Formation of the Southern Appalachians making it an excellent candidate for the study of the precise relationship between the extinction and changes in the carbon cycle. This investigation confirms the onset of the SPICE occurs at the extinction boundary however no negative ẟ¹³C excursion occurs at the extinction boundary. Further there is no systematic relationship between local facies changes and ẟ¹³C or the extinction interval across the basin, which suggests that global environmental changes were responsible for both the ẟ¹³C record and the extinction event. / Master of Science