Inland waters receive significant inputs of organic and inorganic carbon (OC and IC) from terrestrial ecosystems. This water-borne carbon (C) is subsequently stored, processed, outgassed, or exported downstream depending a suite of biogeochemical controls. These processes are increasingly well-constrained in temperate systems, but our global models are hindered by a lack of quantitative and mechanistic understanding of the tropics. Within the tropics, there is no larger knowledge-gap than the Congo Basin. Although the Congo is still mostly pristine, increasing rates of deforestation threaten to mobilize soil organic carbon (SOC) to rivers, albeit with unknown fate. Here I present seasonal data from a pristine montane forest system in the Congo Basin that highlights the onset of the wet season as a key period for C export. Results from this pristine system show a flushing of boilable and heterogeneous dissolved organic matter (DOM) during the first seasonal rains. Ultimately, this novel dataset provides a baseline against which to assess future change. To examine the effect of deforestation on stream C biogeochemistry in the Congo, I employed a paired-watershed approach in which catchments with varying degrees of forest loss were compared to pristine, primary forest endmembers in both lowland and montane forest ecosystems. Carbon from deforested catchments was old, aliphatic, and biolabile, exhibiting a composition similar to that of microbial biomass and SOC from deep horizons. Deforested streams were also warmer, lower in dissolved oxygen, and more supersaturated in carbon dioxide, potentially reflecting higher rates of in-situ OC respiration relative to forested catchments. Together, these results suggest that destabilized SOC may be respired and vented through the aquatic pathway following deforestation and land-use conversion to agriculture. Lastly, to uncover the source of condensed aromatics present in streams draining both lowland and montane pristine forest catchments, the seasonal composition of wet and dry deposition in a lowland forest was assessed. The deposition of nitrogen and condensed aromatic compounds was associated with the seasonal burning of savanna-woodland biomass, indicating the widespread effect of slash-and-burn agriculture on the interior biogeochemistry of the Congo basin. The results presented in this thesis provide new insight into the effects of seasonality, deforestation, and fire on the carbon cycle of a major and understudied tropical watershed. / A Dissertation submitted to the Department of Earth, Ocean and Atmospheric Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2018. / December 7, 2018. / Includes bibliographical references. / Robert G. M. Spencer, Professor Directing Dissertation; Tingting Zhao, University Representative; Jeffrey P. Chanton, Committee Member; Mariana Fuentes, Committee Member; Johan Six, Committee Member.

Graduation date: 1980 / Best scan available. Original appears to be a copy of a copy. The text is light on many of the pages.

During the GEOSECS cruise of the R/V KNORR, July 1972-April 1973, a very complete and high quality nutrient data set was acquired for the Atlantic Ocean. One hundred and twenty-one hydrographic stations were occupied throughout the Atlantic providing an internally consistent picture of the nutrient dynamics for this ocean. The dynamic and biological controls on the nutrient distribution were viewed by means of horizontal distribution patterns, vertical profiling, and statistical modeling of relationships between oxygen, potential temperature, salinity, and nutrients. The general conclusions are summarized as follows: 1. The nutrient concentrations in the Atlantic exhibit the interplay at all depths of nutrient rich waters of South Atlantic origin with nutrient poor waters of the North Atlantic. This interrelationship of the two water sources manifests itself in numerous extrema (maxima and minima) in the water column. 2. For intermediate and deep waters, the strong predominance of lateral transport over processes of vertical dissipation are apparent in the Atlantic. Identifiable water types with only small variations of potential temperature (θ), salinity (S), and preformed nutrients can be characterized thousands of miles from their region of origin. 3. Silicate distribution in the Atlantic exhibits very marked gradients between waters of South and North Atlantic origin. Variations of up to 100 μm/kg occur where salinity differences are less than 0.3‰. Great potential exists for the use of silicate as a water mass tracer for Antarctic Intermediate Water (AAIW), North Atlantic Deep Water (NADW), and Antarctic Bottom Water (AABW). 4. The deep and bottom water nutrient distribution can be explained purely from hydrodynamic considerations. Nutrients, dissolved oxygen (O₂), and apparent oxygen utilization (AOU) behave like conservative parameters. The rates of oxidation in deep water are slow relative to the physical processes of mixing and advection, 5. The total organic carbon (TOG) is relatively invariant below a few hundred meters. Significant variation at the cores of NADW, AAIW, and at the ocean bottom is indistinguishable at the present analytical capability. This supports the observation of very low rates of oxidation in the abyssal waters of the Atlantic. 6. The use of statistical models of O₂ as a function of θ or S and a nutrient are consistent with θ-S diagrams in distinguishing the influence of various water types. In addition, a subsurface water type is seen in temperate and equatorial regions which is due to biochemical activity. This water type corresponds to the portion of the water column where rapid oxidation of organic carbon ceases. It is characterized by a low preformed nutrient concentration but a relatively high oxidative nutrient portion. 7. Statistical modeling for a series of stations in the Drake Passage shows the extent of biological depletion across the Passage and points out the influence of an oxygen rich bottom water in the southern reaches of the Drake Passage. This is bottom water from the South Scotia Sea observed by other authors. 8. An apparent breakdown of Redfield's ratio for the Δ O₂: Δ PO₄ and the Δ O₂:Δ NO₃ in the bottom waters of the Atlantic is seen. My analysis indicates that the variation is due not to an inconsistency in the Redfield ratio but to the very low rates of oxidation at great depths. Nearly all the variation in the oxygen content of the deep water at an equatorial station and a station in the Drake Passage can be explained by the use of a conservative variable such as θ or S. Significant oxidation larger than the analytical errors of the GEOSECS methods cannot be seen for the stations considered at present. / Graduation date: 1974

An experimental investigation of the thermohaline, diffusive interface between convecting layers, with heat fluxes more similar to natural fluxes than in previous studies, shows that the formula suggested by Huppert (1971) for the dependence of heat flux on interface stability cannot be extrapolated to stability numbers higher than seven and a new formula is proposed. The non-dimensional ratio of salt to heat flux is observed to increase from the value 0.15, found by Turner (1965), as the heat flux is lowered through almost three orders of magnitude. Migration of the interface is found even in experiments with anti-symmetric temperature boundary conditions; Huppert's (1971) analysis of the stability of a pair of diffusive interfaces was based on the assumption of stationary interfaces. For oceanic values of the heat flux, the thickness of the interface was in the range observed for the layered system of microstructure in the Arctic Ocean. / Graduation date: 1974

Thesis (Ph. D.)--University of Rhode Island, 2004. / Typescript. Includes bibliographical references (leaves 190-195).

<p> An imbalance in the flux of oxidants and reductants to the suboxic zone (which lies between layers containing oxygen (O₂) and hydrogen sulfide (H₂S)) has been observed in many anoxic basins. These basins also have high levels of chemoautotrophy and elemental sulfur in the suboxic zone, which is defined here as the region where oxygen levels are &lt; 2&micro;M and sulfide levels are &lt; 2&micro;M. It has been hypothesized that there is a &lsquo;cryptic sulfur cycle&rsquo; in the suboxic zone that is mediated by sulfate (SO4^2&ndash;) reducers and sulfide oxidizers. This study examines elemental sulfur in the Cariaco Basin and a meromictic lake using HPLC and Raman microspectroscopy techniques to evaluate if the elemental sulfur is produced by biological or chemical oxidation. Concentrations of particulate sulfur (> 0.2&micro;m) and total zerovalent sulfur (TZVS made up of particulate sulfur, colloidal sulfur, and polysulfides) were determined in the suboxic zone of the Cariaco Basin in November 2014 and 2015. Analysis of samples from November 2015 with Raman microspectroscopy found sulfur inside cells at two depths in the suboxic zone. HPLC and Raman analyses of samples from Fayetteville Green Lake found little elemental sulfur, perhaps indicating that the layer previously observed to contain high concentrations of elemental sulfur and TZVS by Zerkle et al (2010) may have been missed, and that samples need to be collected with higher depth resolution than used during the present study. Raman microspectroscopy was effective for finding sulfur associated with cells. Future work could include examining samples collected with high depth resolution from Fayetteville Green Lake, examining previous Cariaco samples with Raman microspectroscopy, and exploring density data from the Cariaco Basin to see if high levels of elemental sulfur are associated with intrusions of oxygenated water.</p><p>

Northern high-latitude regions are undergoing rapid changes as the Arctic warms at about twice the rate of mid-latitudes. Climate change is causing permafrost thaw, vegetation and hydrologic shifts, and the increased incidence of wildfire, all of which have major implications for regional and global carbon (C) cycling. In this study, I evaluate dissolved organic matter (DOM) composition across temporal and spatial gradients using chromophoric DOM (CDOM), the biomarker lignin phenol, and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). The goal of this dissertation was to improve spatial and temporal understanding of DOM composition and cycling across aquatic gradients by improving spatial (Chapter 1) and temporal (Chapter 2) coverage of DOM composition, and using space for time gradients to understand the seasonal and landscape scale controls on DOM composition in lakes and rivers (Chapters 3 and 4, respectively) and how they may change into the future. Finally, an overarching theme of these studies were the utilization of optical measurements to estimate dissolved organic carbon (DOC) concentration and DOM composition for future applications for in situ and remote sensing technology. By including an understudied, mid-sized watershed in pan-Arctic flux estimates as a model for the unsampled portion of the pan-Arctic watershed (i.e. not encompassed in the major six Arctic rivers from which historic estimates are extrapolated) DOC flux estimates were increased from 27 Tg C to 34 Tg C annually to the Arctic Ocean. Additionally, the residence time of lignin and thus terrestrial DOM was further constrained from previous studies to 0.5 to 1.8 years. This refinement of the pan-Arctic flux estimate and terrestrial DOM residence time is important for the accurate assessment of land-ocean C fluxes and their implications for future change. Temporal DOM dynamics were also evaluated in both rivers and lakes. Diel lake sampling revealed that seasonal variability accounted for the greatest changes in DOM composition and underscored the need to sample lakes seasonally while regular diel trends were not observed. Finally, using space for time transitions in both lakes and rivers the seasonal and landscape drivers of DOM composition were evaluated to allow future projections in a changing Arctic. In lakes that were relatively hydrologically disconnected there was lower CDOM compared to hydrologically connected lakes and a strong decoupling of DOC from CDOM. Further evidence showed that DOM in these lakes was driven by autochthony and that a future drier climate is unlikely to cause browning in these lakes as suggested by recent research. In rivers, where hydrologic connectivity is high, the watershed relief, soils and vegetation played an important role in determining DOM composition. Lower DOC yields and more aliphatic DOM were mobilized in watersheds underlain by continuous permafrost compared to discontinuous permafrost. Together these studies show the utility of space for time gradients to project future change and the use of CDOM parameters as a proxy for DOC concentration and DOM composition. Taken together these results allow us to make projections for boreal and arctic change, as well as allow future studies to improve spatial and temporal resolution via CDOM parameters. / A Dissertation submitted to the Department of Earth, Ocean and Atmospheric Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester 2019. / April 15, 2019. / Arctic, Boreal, Carbon Cycling, Chromophoric Dissolved Organic Matter, Dissolved Organic Matter, Lignin / Includes bibliographical references. / Robert G. M. Spencer, Professor Directing Dissertation; Alan Marshall, University Representative; Jeffrey P. Chanton, Committee Member; Markus Huettel, Committee Member; Angela Knapp, Committee Member.

In the spring of 2010, the MC 252 Deepwater Horizon well blow out lead to nearly five million barrels of Gulf of Mexico light sweet crude to be released into the northern Gulf at a depth of 1500 meters. Dispersant injected into the plume at the wellhead helped little to keep the oil below the surface. This dispersant inefficiency, coupled with the limited effectiveness of sea surface mitigation, allowed an estimated 150,000 barrels to impact the shores of the Northern Gulf of Mexico, from East Texas to the Western Florida Panhandle. Nearly half of the impacted coastline is comprised of permeable sandy beaches. The surface oil took several months to reach the shore, and over that time it was degraded by heat, UV light, oxygen, and microbes. The weathered oil final reached the shores of Pensacola Beach, Florida on June 22, 2010. In the surf zone, the weathered oil was mixed with sand to form Sediment-Oil-Aggregates (SOA) that sank in the swash zone between the beach and longshore bar. This SOA material was transported with longshore currents, and repeatedly buried and exhumed. The weathered oil also came ashore at the same time as Tropical Storm Lee, whose increased wave action cause some of the SOA material to be deposited and buried into the dry beach sediment above the high-water line. The goal of this dissertation is to investigate the fate of MC252 crude oil from the Deepwater Horizon accident on Northern Gulf of Mexico permeable sandy beaches. Sampling trips to Santa Rosa Island, Florida were performed monthly from July 2010 to July 2011, where sediment cores from the dry beach above the high-water line were taken. These cores were sectioned and incubated to measure microbial activity in response to the buried oil, using oxygen as a proxy, over the year after the oil came ashore. On these trips, SOA material on the beach and in the surf was collected, homogenized, and use for lab incubations to investigate the role of microbes, temperature, and mechanical stress due to wave action, on the degradation of SOA material in the surf zone. Column experiments were also performed to investigate the aerobic decomposition of SOA material in the coastal water column and permeable sediments. The time series incubations showed that clearly oiled sections of sediment had significantly higher oxygen consumption rates, compared to sections that were visibly clean. In October of 2010, beach cleaning crews used heavy machinery to exhume the top meter of beach, sieve out the large SOAs, and in the process homogenized the smaller oil particles throughout the top meter of beach sand, increasing the surface area available to microbes for degradation. By April of 2011, a clearly oiled layer in the beach was no longer visible. Along with decreasing visible oil in the dry sediment, SOA material in the swash zone also decreased during the year. In laboratory incubations, it was found that microbes play a large part in the degradation of the SOA material, with microbes accounting for 80% of the oxygen consumption in SOA incubations. Higher temperatures increased the rate of oxygen consumption, with warmer summer temperatures causing a 4-fold increase in oxygen consumption rates over winter temperatures. The mechanical stress of wave action also causes the SOA material to quickly fall apart. In incubations, SOA material was rotated at 0.5rpm, and SOAs were disintegrated within 24 hours. In column experiments, it was found that increased fluid front velocity increased the oxygen consumption rates of sediment with artificially weathered crude oil. In columns amended with SOA material, there was no difference in oxygen consumption compared to sediment with no SOA material. There was also very little DOC release in SOA columns where the water was amended with Corexit 9500®, suggesting that the small surface area to volume ratio of larger, intact SOAs buried in the sediment develop a tough crust of highly degraded hydrocarbons, protecting the more labile inside from microbial degradation. This research shows the importance of microbial activity, wave action, and temperature on the degradability of the Deepwater Horizon oil. The wave energy of the environment, coupled with the permeable sandy sediments and warm temperature of the Florida summer, all contributed to the rapid degradation of the oil. / A Dissertation submitted to the Department of Earth, Ocean and Atmospheric Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester 2019. / April 12, 2019. / Beach, Corexit, Deepwater Horizon, Oil, SOA / Includes bibliographical references. / Markus Huettel, Professor Directing Dissertation; Thomas E. (Tom) Miller, University Representative; Jeffrey P. Chanton, Committee Member; William Dewar, Committee Member; Olivia Mason, Committee Member.

The Deepwater Horizon (DWH) blowout of 2010 released an estimated 4.5-4.9 million barrels of oil and 500,000t of methane into the Gulf of Mexico (GOM). Some of this oil rose to the surface, forming oil slicks, while an estimated 30% of the smaller particles and gasses formed a deep-water hydrocarbon plume between 1000-1200m (Ryerson et al., 2013; Valentine et al., 2010). The oil slicks at the surface covered a total area of 149,000km2 (MacDonald et al., 2015), but less than 0.01% of the gaseous hydrocarbons reached the surface (Kessler et al., 2011; Yvon-Lewis et al., 2011). After capping the wellhead and following natural and human remediation efforts, an estimated 11-30% of the oil was left unaccounted (Lehr et al., 2010). Studies of δ13C and Δ14C tracers in particulate organic carbon (POCsusp) in the water column and in sediments have shown the accumulation of fossil carbon in these pools. This dissertation explores the POC and sedimentary organic carbon pools using δ13C and Δ14C to characterize and track the recovery of these carbon pools following the DWH blowout. Due to the small particle size, residence time, and sensitivity to inputs, POCsusp provides a link between microbial processes in dissolved organic carbon and larger particles that pass carbon up the food web. Through this link we can evaluate the incorporation of hydrocarbons using δ13C and Δ14C. POCsusp was collected over 6 years from 43 sites across the Northern GOM. At the time of collection these sites were classified as seep or non-seep. We observed a wide range of natural variability in both δ13C (-17.8 to -35.4‰) and Δ14C (+71 to -755‰) throughout the water column. We found that deep-water POCsusp of the GOM was always more depleted than POCsusp from the euphotic zone. POCsusp collected from seeps was more depleted in Δ14C than non-seep sites. Endmember modeling indicated that in these particles, as much as 73% of the carbon was incorporated from oil. Four years following the blowout, we observed recovery in the Δ14C of deep-water POCsusp settling at a baseline of Δ14C=-164.4±18.9‰. We found the δ13C of POCsusp from the euphotic zone became more depleted over time, potentially due to the continuous incorporation of hydrocarbons. The deposition of oil in the sediments of the GOM has been estimated to be up to 14% of the total oil released (Valentine et al., 2014; Chanton et al, 2015), with marine oil snow as the primary mode of deposition. We employed inverse distance weighted interpolation to the surface sediment δ13C and Δ14C data. From these interpolations, we calculated the area affected by petrocarbon and followed its quantity through time. The area affected by petrocarbon decreased each year at a rate of -2x108 g/yr. Our maps indicated an east-west trend in depletion of both δ13C and Δ14C likely caused by the increasing importance of output from natural seeps and the Mississippi River. We also found significant differences between the sediment of the northern and southern GOM, with the north being much more depleted in δ13C and Δ14C than the south. Ramped pyrox paired with δ13C and Δ14C was used in previous studies of oil contaminated marsh sediments, showing the evolution of the thermostability and isotope signatures as the oil was transformed and the system recovered. We used ramped pyrox paired with δ13C and Δ14C measurement of the evolved fractions to explore the recovery of two DWH affected time-series sites, GIP07 and GIP17, and one site that had high PAH levels in 2010. We found differences in the thermographs and δ13C and Δ14C of the evolved CO2 between crude oil and the seep and control sediment. We observed shifts in the CO2 evolution over time from lower to higher-temperature at GIP17 (~16km from the wellhead), followed then by a loss of higher temperature peaks at GIP07 (~90km). At both sites we observed recovery going from bulk Δ14C=-491‰ in 2010 to almost background by 2015, Δ14C=-264‰ at GIP17. Our study supports ideas from Bagby et al. (2016) and Stout and Payne (2016) indicating a relationship between degradation rate and distance travelled in the water column. The further the hydrocarbons traveled in the water column, the faster they degraded before being sedimented. Following sedimentation, degradation rates were much slower than while the oil was in the water column. The level of contamination also affected the degradation rate, with high contamination recovering at slower rates (20‰ y-1) than sites with lower contamination (46‰ y-1). / A Dissertation submitted to the Department of Earth, Ocean and Atmospheric Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester 2018. / April 30, 2018. / carbon isotopes, Deepwater Horizon Blowout, Particulate organic carbon, ramped pyrolysis, sediment / Includes bibliographical references. / Jeffrey Chanton, Professor Directing Dissertation; Tingting Zhao, University Representative; Olivia Mason, Committee Member; Joseph Montoya, Committee Member; Yang Wang, Committee Member.

<p>Cosmogenic ³²Si (t_1/2 &sim; 140 yrs) was used in a novel way to constrain the quantity of reactive Si storage and early diagenetic reactions of Si in the highly mobile deltaic sediments along the coast of French Guiana, representative of deposits along the ~1600 km Amazon&ndash;Guianas coastline downdrift of the Amazon delta. A sequential leach was developed to extract and purify SiO₂ from different operational pools in large samples of surface sediments (0&ndash;10cm). This methodology, a hot 1% Na₂CO₃ leach followed by a hot 4M NaOH leach, was adapted from the existing leaches widely used to estimate biogenic silica (bSi) content in marine sediments, and ultimately to constrain the global oceanic Si budget. ³²Si activity was determined in each pool via its daughter product ³²P. Results from several sites in coastal mudbanks near Kourou and Sinnamary indicate no detectable ³²Si activity in the bSi fraction, whereas ³²Si was detected in the Si-NaOH fraction after removal of bSi. The lack of detectable activity in the 1% Na₂CO₃ leach and its detection in the NaOH fraction (0.4&ndash;2.5 dpm) indicate that the method widely used to determine bSi content recovers only a minor fraction of the originally deposited reactive bSi in these deposits. The results are consistent with rapid alteration of biogenic silica and clay authigenesis or reverse weathering. They also demonstrate that the current estimate of biogenic silica storage in tropical deltaic sediments is significantly underestimated. Assuming an initial diatom specific activity range of &sim;5&ndash;40 dpm/kg SiO₂, the ³²Si activity in the NaOH fraction corresponds to a reactive Si storage of &sim;150&ndash;18,000 &micro;mol Si/g sediment. This magnitude is more consistent with estimates of reactive Si (&Sigma;Si_hr) storage in the Amazon delta based on modified operational leach techniques that target poorly crystalline clays and with diagenetic modeling of pore water K⁺, F^&minus;, and Si(OH)₄, though these modified leaches also appear to underestimate the amount of reactive Si stored along this system. To directly confirm whether these modified operational extractions underestimate reactive Si storage, a sequential extraction methodology was also developed to first isolate ³²Si activity in the &Sigma;Si_hr fraction (0.1N HCl followed by 1% Na₂CO₃) and then extract any remaining ³²Si from the residual fraction using 4M NaOH. </p><p> Sediment from 2 stations in the Gulf of Papua, Papua New Guinea, 1 station in the northern Gulf of Mexico near the Southwest Pass, and 1 station in Long Island Sound (Smithtown Bay) were also extracted for ³²Si in the bSi fraction as well as the residual fraction after removal of bSi. Bulk ³²Si activities in the residual fractions in the Gulf of Papua (0.5&ndash;0.7 dpm/kg sediment) were used to extrapolate Si storage in the outer topset and forset of the clinoform delta. </p><p> ³²Si activity was detected in the both the bSi (0.21 &plusmn; 0.04 dpm/kg sediment) and the residual fraction (0.44 &plusmn; 0.08 dpm/kg sediment) from the site in the Gulf of Mexico. A Si burial rate using the ³²Si activity in the bSi fraction (assuming an activity of 15dpm/kg in starting Si materials) of 0.004Tmol/y was calculated over approximately 5000 km2 of the delta, whereas the burial rate calculated using the Si content in this same fraction from a classic bSi leach, was &sim;0.006Tmol/y. Adding the Si burial rate using the ³²Si activity in the residual fraction (0.008Tmol/y) yielded a total storage per year of 0.012Tmol Si, &sim;10% of the total Si inputs (dissolved and amorphous Si) from the Mississippi-Atchafalaya river system. ³²Si activity was also detected in the residual fraction (0.53 &plusmn; 0.08 dpm/kg sediment) after removal of &Sigma;Si_hr and using this activity yielded similar calculated rates of Si burial (&sim;0.01 Tmol/y). </p><p> In Smithtown Bay, Long Island Sound, ³²Si activity was also detected in both the bSi (0.15 &plusmn; 0.05 dpm/kg sediment) and the residual (0.4 &plusmn; 0.2 dpm/kg sediment) fractions from the site in Smithtown Bay, Long Island Sound, yielding a total Si storage estimate (assuming an activity of 15 dpm/kg in starting Si materials) of 1.6 &times; 10^&minus;3 Tmol/y over the entire Sound, comparable to estimates of Si storage calculated using the Si content in the classic bSi (1.1 &times; 10^&minus;3 Tmol/y) and the classic &Sigma;Si_hr (2.2 &times; 10^&minus;3 Tmol/y) leaches. It appears that reverse weathering is an important sink of Si in these deposits and that classic bSi or &Sigma;Si_hr leaches can underestimate Si storage in these system by two to four-fold. (Abstract shortened by UMI.)</p>