Combining geochemical data with physical data produces a powerful method for understanding sources and fluxes of waters to river systems. This study highlights this for river systems in regions of complex hydrogeology, shown here through the identification and quantification of base-flow sources to the Verde River and its tributaries within the middle Verde River watershed. Specifically, geochemical tracers (major solutes, stable and radioactive isotopes) characterize the principal aquifers (C, Redwall-Muav, and Verde Formation) and provide a conceptual understanding of the hydrologic connection between them. For the surface-water system, PCA is utilized to identify potential base-flow sources to the Verde River on a several-kilometer scale. Solute mixing diagrams then provide relative inputs of these sources, and when combined with stream discharge, allow for quantification of water sources. The results of this study provide an improved conceptual model that reveals the complexity of groundwater-surface water exchanges in this river basin.

<p>Sea turtles are long lived marine species that are currently endangered because their life history and population dynamics hinder them from withstanding modern anthropogenic threats. Worldwide, fisheries bycatch in on the major threats to the survival of sea turtles and that is also the case in the Eastern Pacific. To establish regional conservation priorities for the mitigation of bycatch, it is essential to first obtain a comprehensive picture of the regional sea turtle bycatch situation. This comprehensive analysis was lacking for the Eastern Pacific; therefore one component of this dissertation (the first chapter) is focused on delivering a regional bycatch analysis for the Eastern Pacific. A literature review was conducted to obtain numbers of turtles captured, frequencies, bycatch and mortality rates per species and country in trawl, longline, and gillnet fisheries, and to compile results of mitigation measures. Moreover, estimates for current annual capture rates in trawl fisheries were obtained and compared with population numbers. </p><p>This regional bycatch used all the information compiled and synthesized to give conservation priorities at the regional level. The review underlines the high bycatch rates in trawls for Costa Rica, Guatemala, and El Salvador and the detrimental impact that these captures could have specially for hawksbill Eretmochelys imbricata due to its reduced population numbers and for green turtle Chelonia mydas due to its highest mortality rate. It also emphasizes the continuous lack of use of TEDs as a bycatch mitigation measure. In longline fisheries, the review identifies the high bycatch rates in pelagic longline fisheries of Costa Rica, Ecuador, and Nicaragua in a global context but given that olive ridley Lepidochelys olivacea is the most common species captured in these countries, it highlights the capture of loggerhead Caretta caretta and leatherback Dermochelys coriacea off Peru and Chile due to their small population numbers. Bottom longlines have high mortality rates compared with pelagic longlines in the region and the review identifies a need for further research in this area due to the scarce information but high mortality rates. The review also noted that some mitigation measures for pelagic longlines like circle hooks and hooks with appendages could bring improvements in the mitigation of bycatch in longline fisheries in the region, there is still considerable work to be done in technology transfer, sea turtle handling, and estimates of post-release mortality rates. </p><p>For gillnet fisheries, the most important highlight is how little information exists for the region given the high rates of bycatch for sea turtles in this gear. However, the difficulties of studying bycatch in highly dynamic and artisanal fisheries are recognized as the major impediment for this situation. Nevertheless, the high bycatch rates in areas where sea turtles congregate in high numbers like in foraging grounds for loggerhead in Baja California, Mexico and for greens in Paracas and Sechura, Peru, calls for either gear modifications (which has not been that successful), change of gear, or areas closed for gillnets. </p><p>The second half of the dissertation is focused on foraging ecology of oceanic sea turtles in the Southeast Pacific Ocean. Sea turtles in the oceanic stage are the least known stage due to the difficulty of accessing these individuals. However, it is a very important stage in the life cycle and can be critical for the population dynamics of sea turtles as some population models have shown. Therefore, this dissertation is filling a gap in the life cycle of sea turtle populations in the Eastern Pacific. </p><p>To study foraging ecology, we used Stable Isotope Analysis (SIA) of turtle tissues as well as potential prey items from the oceanic realm. SIA is a great tool because it gives an integrated view, from days to weeks, of prey from a consumer tissue. SIA also can be used to link consumers to habitats when elements that have spatial trends are used. In chapter two, we investigate the foraging ecology of three species of sea turtles to compare trophic status and to observe if spatial patterns were shown in the SIA signatures of sea turtles. To our knowledge this is the first study employing SIA to research the ecology of three species of sea turtles from the same time and space. Our results show that spatial patterns in delta15N and delta13C were observed in sea turtle's tissues as correlations with latitude. We also found that loggerhead's signatures differed significantly from green and olive ridleys, especially in terms of delta15N. Loggerheads had higher values of delta15N and also a wider nitrogen trophic niche. Greens and olive ridleys were similar in isotopic nitrogen values but they were significantly different in carbon. When analyzing a smaller group of animals captured in a more restricted area, nitrogen differences were not found which suggests that latitudinal spatial patterns play an important role in the nitrogen signature. On the contrary, carbon signatures still differed among turtles in the restricted area which suggest that the inshore-offshore trend is strong and made us conclude that loggerheads are restricted to oceanic areas but that greens and olive ridleys could be using both coastal and oceanic areas.</p><p>In chapter three, we conduct a mixing model analysis using the Bayesian program SIAR to identify the most important prey items for green, olive ridley, and loggerhead off Peru. Also, we wanted to identify the contribution of longline baits in the diet of oceanic turtles. The analysis was restricted to the central zone of our study area to avoid spatial trends in nitrogen. To use as sources in the model, we collected potential prey items offshore Peru during trips on longline fishing vessels and obtained their stable isotope signatures. Results from our mixing models show that for greens and olive ridleys, crustaceans, mollusks, and coastal Ulva (indicator of coastal prey) were the only important food items. In the case of greens, crustaceans had a very high proportional contribution and due to the fact that nitrogen values of crustaceans were the lowest ones among the sources it seems that greens would be eating in a lower trophic level. The importance of coastal Ulva for greens and olive ridleys is a confirmation of our findings from chapter two where we suggest that these two species could be using oceanic as well as coastal areas. </p><p>Results for loggerheads showed cnidarians, mollusks, mackerel and squid bait as foraging items and highlights the differences among this species and the other two. The lack of importance of coastal Ulva again suggests that loggerheads remain only in oceanic areas off Peru. Moreover, the importance of mackerel and squid, the most common longline baits, suggests this species is the one interacting the most with longline fisheries and that cumulative effect of multiple interactions could have a detrimental effect in this population.</p> / Dissertation

L’érosion des sols constitue une menace majeure pour la durabilité de la production agricole. Cette étude vise à développer une méthode originale de traçage des sédiments au sein de deux bassins versants agricoles aux caractéristiques contrastées. Différents marqueurs ont été utilisés pour tracer l’origine lithologique des particules (concentrations élémentaires, rapports ⁸⁷Sr/⁸⁶Sr), déterminer de quel compartiment des sols (surface ou profondeur) elles proviennent (¹³⁷Cs) et étudier leur dynamique temporelle (⁷Be, ²¹⁰Pb_xs). Sur le bassin du Louroux (24 km², France), caractéristique des plaines agricoles du Nord-Ouest de l’Europe, les résultats ont montré un transfert accru de particules très fines (<2µm) via le réseau de drainage. Lors de l’étude de trois crues successives, les sédiments accumulés dans le chenal ont d’abord été transportés, avant l’exportation de particules directement érodées de la surface des sols. L’utilisation de modèles de mélange a montré que les particules érodées à la surface des sols contribuent majoritairement aux sédiments de l’étang situé à l’exutoire. D’autre part, les deux zones lithologiques discriminées par le rapport ⁸⁷Sr/⁸⁶Sr contribuent équitablement aux particules qui se déposent dans l’étang. Les fortes variations observées depuis les années 1950 peuvent être interprétées comme le résultat de l’aménagement du bassin. Ces résultats ont montré le potentiel de l’isotopie du strontium (⁸⁷Sr/⁸⁶Sr) pour le traçage lithologique des sources de particules. La méthode a dès lors été transposée au site de Guaporé (2000 km², Brésil), un bassin agricole au climat plus érosif, à la topographie plus accentuée et aux lithologies plus contrastées. En combinant les rapports ⁸⁷Sr/⁸⁶Sr et les éléments les plus discriminants des cinq types de sols dans les modèles de mélange, les résultats montrent une contribution très majoritaire des types de sol localisés en aval du bassin. Celle-ci reflète une sensibilité croissante des sols à l’érosion selon un transect amont-aval, avec un système d’agriculture intensive appliquant des techniques de conservation des sols à l’amont et de plus petites exploitations familiales qui pratiquent un labour traditionnel en aval. Le couplage de ces traceurs améliore la compréhension de la dynamique des particules à l’échelle du bassin versant et fournit des données spatio-temporelles essentielles pour mettre en place des mesures de lutte contre l’érosion des sols. À terme, ces données pourraient être utilisées pour valider des modèles d’érosion et de transfert particulaire. / Soil erosion is recognized as one of the main processes of land degradation in agricultural environments. This study develops an original fingerprinting method to examine sediment source contributions in two contrasted agricultural catchments. Several properties were used to trace sediment lithological sources (⁸⁷Sr/⁸⁶Sr ratios, elemental concentrations), soil surface and subsurface sources (¹³⁷Cs) and to quantify their temporal dynamics (⁷Be, ²¹⁰Pb_xs).In the Louroux catchment (24 km², France), representative of drained areas of Northwestern Europe, results showed the dominant contribution of very fine particles (<2 µm) transiting through the tile drainage system to suspended sediment. Sediment accumulated in the river channel was mainly exported during the two first floods investigated in 2013 whereas the next event was characterized by the transport of sediment eroded from the cultivated soil surface. Mixing models were used, and results indicated that surface sources contributed the majority of sediment deposited in the pond, at the outlet of the catchment. The two lithological sources, discriminated using ⁸⁷Sr/⁸⁶Sr ratios, contributed in similar proportions to downstream pond sediment. In contrast, significant variations were observed since the 1950s. These changes may be related to the progressive implementation of land consolidation schemes within the catchment. Results obtained in the Louroux catchment revealed the potential of ⁸⁷Sr/⁸⁶Sr ratios to trace sediment lithological sources. The methodology was then applied to the larger and steeper Guaporé catchment (2000 km², Brazil), exposed to a more erosive climate and characterized by contrasted laterite soils. ⁸⁷Sr/⁸⁶Sr ratios and elements that discriminated the five soil types found in the catchment were incorporated in mixing models. Results showed that the major part of sediment was supplied by soils located in lower catchment parts. This result suggests the higher sensitivity of land cultivated with conventional farming practices (downstream contribution), compared to areas cultivated under conservation agriculture (upstream contribution). Coupling several tracers (⁸⁷Sr/⁸⁶Sr ratios, elemental concentrations and radionuclides) was shown to improve the understanding of sediment sources and dynamics at the catchment scale and to provide crucial information to guide the implementation of management measures to limit soil erosion. This fingerprinting approach produced an original dataset that may be used to calibrate and validate models simulating erosion and sediment transfers.

Water security requires that sufficient quantities of water be available at critical times. This is particularly challenging for high-intensity urban and agricultural settings. In underdeveloped nations, streamflow is commonly the preferred water source, as it is readily available and delivered cost-free to users. Yet, the sources of these critical streamflows are often unknown. This issue is salient in the Northern Andes, where basic knowledge of controlling factors for the quantity, quality, and timing of runoff is lacking. High-elevation headwaters are the primary catchment areas in the Northern Andes, but the extent of water providing to municipalities in the Northern Andes is unknown. In this study, the contribution of water derived from the upper watershed to the streamflow in the Tulúa River which supplies water to 200,000 people in the city of Tulúa was quantified. The river runs 72 km through urban, agricultural, and industrial land use in the Central Cordillera of the Colombian Andes. We collected 32 and 34 water samples in August and November, respectively. The water samples were representative of high-elevation headwaters runoff, shallow groundwater discharge, and streamflow throughout the watershed. Samples were analyzed for dissolved constituents and stable isotopes. The dissolved constituents were used in mass-balance mixing models to identify the source of streamflow in the lower watershed of the Tulúa River, where it the river supports a large municipality. Results indicate that approximately 50% surface runoff largely originates as high-elevation headwater runoff, including high-elevation settings where páramos dominate the land cover. These findings underscore the need for source-water protection efforts in the upper watershed, including the páramos. This project serves as a model for other páramo derived watersheds, where source-water protection is a critical challenge.

Green sea turtles in Bermuda are overgrazing the seagrasses on which later life stages are thought to specialize. I hypothesized that larger green turtles in Bermuda would display individual diet specializations during seagrass scarcity. Stable isotope methods were used to determine the diet composition of green sea turtles from the Bermuda Platform as a function of size class and in turtles captured in successive years. Individual turtles had a wide range of diets, however, the variation in diets was driven by differences among size class rather than within the size classes of larger turtles, indicating that green turtles undergo a dietary ontogenetic shift during their residency on the Bermuda Platform and no clear specialization of diets among late-stage individuals. The apparent lack of dietary specialization of larger turtles indicates that older turtles are not diversifying their diets in response to the drastic reductions in seagrass in Bermuda.

Organic matter (OM) is often considered the “currency” for ecosystem processes, such as respiration and primary production. OM in aquatic ecosystems is derived from multiple sources, and is a complex mixture of thousands of different chemical constituents. Therefore, it is difficult to identify all the sources of OM that enter and exit aquatic ecosystems. As humans develop undisturbed land, the rate at which terrestrial OM (e.g.soil and plants) and associated nutrients (e.g.nitrogen) enters rivers has increased. Increased nutrients may lead to increased primary production from aquatic plants and algae, potentially causing eutrophication and harmful algal blooms. In this study, I identified and characterized different sources of OM in four watersheds of Northeastern Utah with multiple land covers such as cities, forests, and crops. I expected OM in watersheds with human-altered land cover would have more OM produced instream by algae and other primary producers, than OM in less disturbed watersheds, which typically have OM from terrestrial sources. I found that OM at river sites with high human impact had high amounts of OM from instream primary production, but there was also OM produced in-steam at sites with low human impact. The greatest differences in OM across watersheds was due to wastewater treatment effluent. I also measured microbial consumption rates of algal derived and terrestrially derived DOM in experimental streams to quantify how much faster algal derived OM was consumed than terrestrial OM. I found algal derived OM was consumed extremely fast, so fast that realistic measurements of its consumption in some river ecosystems may not be possible. It is important to identify and characterize sources of OM to rivers, so watershed manager scan devise effective OM reduction plans appropriate for the constituent of concern unique to that watershed or region. Constituents of concern associated with OM include pathogens affiliated with manure, toxins in harmful algal blooms, metals, and pharmaceuticals from wastewater treatment effluent. Each pollutant requires a unique mitigation strategy and therefore the first step to pollution mitigation is source identification.

Advanced monitoring of the spatio-temporal distribution of process parameters in large-scale vessels and containers such as stirred chemical or bioreactors offers a high potential for the investigation and further optimization of plants and embedded processes. This applies especially to large-scale fermentation biogas reactors where the process performance including the biological processes highly depend on mixing parameters of the complex bio-substrates. Sufficient mixing is a basic requirement for a stable operation of the process and adequate process performance. However, this condition is rarely met in agricultural biogas plants and the process efficiency is often reduced dramatically by inhomogeneities in the agitated vessels. Without a doupt, investigation and monitoring of biochemical parameters, such as the fermentation rate, pH distribution as well as O2 and CO2 concentration is of great importance. Nevertheless, also understanding of non-biological parameters, such as fluid dynamics (flow velocity profiles, circulation times), suspension mixing (homogeneity, location of dead zones and short-circuits) and heat transfer (temperature profiles), is necessary to analyze the impact of mixing on the biological system and also to improve the process efficiency. However, in most industrial scale applications the acquisition of these parameters and their spatial distributions in the large-scale vessels is hampered by the limited access to the process itself, because sensor mounting or cable connections are not feasible or desired. Therefore, state of the art instrumentation of such reactors is commonly limited to few spatial positions where it is doubtfully assumed that the measured parameters are representative for the whole reaction mixture. In this work, a concept of flow following sensor particles was developed. The sensor particles allow long-term measurement of spatially distributed process parameters in the chemically and mechanically harsh environments of agitated industrial vessels. Each sensor particle comprises of an onboard measurement electronics that logs the signals of measurement devices, namely temperature, absolute pressure (immersion depth, axial position) and 3D acceleration. The whole electronics is enclosed in a robust neutrally buoyant capsule (equivalent diameter 58.2 mm; sphericity 0.91), to allow free movement with the flow. The sensor particles were tested in pilot fermenters under comparable flow conditions of biogas fermenters. The experiments proved the applicability of the sensor particles and the robustness to resist the harsh environments of mixing processes. Moreover, the results show the capabilities of the sensor particles to monitor the internal conditions of the vessel correctly and thus deliver significant information about the flow regime. Therefore effects of liquid rheology, vessel geometry, impeller speed and axial impeller position on the macro-mixing process were properly detected. Evaluation of the impeller efficiency and the mixing processes was done based on mixing homogeneity, location of dead zones, axial velocity profiles, circulation time distributions as well as average circulation times, acceleration spectra and temperature profiles that were extracted from the measured data. Furthermore, it is shown, that parameters of mixing models such as circulation number, impeller head, PECLÉT-number and variance of suspended solid particles can be estimated from the measured data. The main achievement of this work is therefore the development and validation of instrumented flow followers for the investigation of macro-mixing effects in agitated vessels. The sensor particles show potential for employment to real applications such as biogas fermenters or large bioreactors and to monitor and improve the mixing and heating regimes.

Advanced monitoring of the spatio-temporal distribution of process parameters in large-scale vessels and containers such as stirred chemical or bioreactors offers a high potential for the investigation and further optimization of plants and embedded processes. This applies especially to large-scale fermentation biogas reactors where the process performance including the biological processes highly depend on mixing parameters of the complex bio-substrates. Sufficient mixing is a basic requirement for a stable operation of the process and adequate process performance. However, this condition is rarely met in agricultural biogas plants and the process efficiency is often reduced dramatically by inhomogeneities in the agitated vessels. Without a doupt, investigation and monitoring of biochemical parameters, such as the fermentation rate, pH distribution as well as O2 and CO2 concentration is of great importance. Nevertheless, also understanding of non-biological parameters, such as fluid dynamics (flow velocity profiles, circulation times), suspension mixing (homogeneity, location of dead zones and short-circuits) and heat transfer (temperature profiles), is necessary to analyze the impact of mixing on the biological system and also to improve the process efficiency. However, in most industrial scale applications the acquisition of these parameters and their spatial distributions in the large-scale vessels is hampered by the limited access to the process itself, because sensor mounting or cable connections are not feasible or desired. Therefore, state of the art instrumentation of such reactors is commonly limited to few spatial positions where it is doubtfully assumed that the measured parameters are representative for the whole reaction mixture. In this work, a concept of flow following sensor particles was developed. The sensor particles allow long-term measurement of spatially distributed process parameters in the chemically and mechanically harsh environments of agitated industrial vessels. Each sensor particle comprises of an onboard measurement electronics that logs the signals of measurement devices, namely temperature, absolute pressure (immersion depth, axial position) and 3D acceleration. The whole electronics is enclosed in a robust neutrally buoyant capsule (equivalent diameter 58.2 mm; sphericity 0.91), to allow free movement with the flow. The sensor particles were tested in pilot fermenters under comparable flow conditions of biogas fermenters. The experiments proved the applicability of the sensor particles and the robustness to resist the harsh environments of mixing processes. Moreover, the results show the capabilities of the sensor particles to monitor the internal conditions of the vessel correctly and thus deliver significant information about the flow regime. Therefore effects of liquid rheology, vessel geometry, impeller speed and axial impeller position on the macro-mixing process were properly detected. Evaluation of the impeller efficiency and the mixing processes was done based on mixing homogeneity, location of dead zones, axial velocity profiles, circulation time distributions as well as average circulation times, acceleration spectra and temperature profiles that were extracted from the measured data. Furthermore, it is shown, that parameters of mixing models such as circulation number, impeller head, PECLÉT-number and variance of suspended solid particles can be estimated from the measured data. The main achievement of this work is therefore the development and validation of instrumented flow followers for the investigation of macro-mixing effects in agitated vessels. The sensor particles show potential for employment to real applications such as biogas fermenters or large bioreactors and to monitor and improve the mixing and heating regimes.