Annually, 0.5-1 % of the OM productivity escapes mineralization to be preserved in the geosphere, the largest reservoir of organic matter (OM) on Earth. This preservation of OM has a direct impact on carbon bioavailability, on the carbon cycle and also on the formation of fossil fuels. This thesis aims to provide insights into the processes that lead to the preservation of OM on Earth by examining recalcitrant OM at early stages of diagenesis and, to a lesser extent, its stability with time. This is achieved by examining the p311itioning of prokaryotic biomarkers between solvent-extractable and insoluble OM fractions from a Sphagnum peat bog, geothermally-derived silica sinters and SE continental slope marine sediments. The insoluble organic matter (IOM) was examined using stepwise chemical degradation involving base and acid hydrolysis, and biomarkers that occur in the IOM fractions are expected to have an enhanced preservation potential. Iso- and anteiso- Cl5 and C 17 fatty acids (F As), hopanoic acid and bishomohopanol - diagnostic for· the Bacteria domain - and archaeol - diagnostic for Archaea - were quantified via GCIMS, and glycerol dialkyl glycerol tetraethers (GDGTs), including both branched and isoprenoid form, were examined using HPLC/MS. Results clearly reveal that a large proportion of prokaryotic biomarkers - up to 80 % - occur in insoluble fractions at early diagenesis. Different factors seem to dete1111ine the partitioning of biomarkers between extractable and insoluble fractions. First, intact polar (IP) head groups, whose phosphate or glycoside groups contain motifs likely promote covalent or sorptive interactions that enhance lipid insolubility. This factor seems particularly important in peat. Second, esterification reactions with time. Such mechanisms for prokaryotic lipid insolubilization seems particularly imp0l1ant in the marine sediments analysed, with larger proportions of base and acid hydrolysis-released biomarkers at deeper sections of the core. Third, physical encapsulation or sorptive interactions. These mechanisms are unlikely in peat, but account for at least 40 % of the marine branched F As, most likely via interactions with CaC03. Likewise, it is suggested that sorptive interactions of prokaryotic biomarkers with the silica matrix of the geothermal sinters promote their insolubilization. However, the specific distribution of IP head groups in such settings might additionally bias the partitioning of prokaryotic lipids, with archaeol also present in IOM pools as opposed to marine and peat sediments. An additional novel experiment was performed on the Sphagnum peat bog. A pulse of 13CH4 was introduced, targeting methanotrophic bacteria. Subsequently, I3C-Iabelled bacterial F As were traced among the extractable and insoluble fractions over 8 months and under different conditions, anaerobic vs aerobic and UV-Vis vs dark. This allowed to examine first, whether prokaryotic lipids are inherently resistant to solvent-extraction, and second, whether IOM is further increased in situ during the experiment. Up to 5 % of I3C_ labelled lipids were insoluble initially after the I3C-pulse; thus, indicating a ce11ain inherent recalcitrance of prokaryotic cells. Moreover, no in situ transfer from extractable to insoluble fractions was observed during the course of the experiment, which is likely due to the short timescale. However, direct evidence is provided of the preferential protection of insoluble 13C-labelled FAs against oxidative mineralization. Collectively, these results show that putatively labile compounds are entrained into insoluble forms essentially immediately; with mechanisms that range from the inherent recalcitrance of cells to covalent condensation reactions and sorptive interactions, all of which strongly depends on the depositional environment. Insolubility affords some degree of protection to these biomolecules, but there is some evidence that this IOM is dynamic and lipids can also be released from the IOM during early diagenesis, challenging their stability over geological time.

This practice-based thesis aims to inform new kinds of creative art and design practice that are engaged with the production and critique of new technological artefacts and systems. The core focus of this research is the role of materiality in making-processes as part of an ecology of technological experience, in other words an 'experiential ecology'. Building on an existing, ontological argument that meaning is an a priori property of being, a proposition for materiality's intrinsic historicity is developed. It is argued that material things have meaning, anterior to interpretation, and that consequently meaning is connected to objects' pasts. It is then described how a principal implication of this understanding in the context of making with contemporary technologies is the adoption of an 'archaeological' approach to studying and producing artefacts which engages with such pre-existing meanings. It is suggested that a possible consequence of this archaeological engagement, for practitioners, is that they view their making activities as situated within ecologies of existing agencies and interactions of the material world into which practice intervenes. Building on Dewey's articulation of the connection of aesthetics to the circumstances of art's production aesthetic experience within such ecologies integrates the activity of making to the material histories of artefacts. To inform making-practices thus conceived, a framework of five separate facets of materiality is developed: performative, distributed, spatio, temporal, fragile, and future-oriented. This framework is applied through a case study describing the making-process of a new artwork entitled 'Neurotic Armageddon Indicator'. The results of this application are examined to assess their contribution to the earlier introduced experiential ecology. This practice-based research makes the following contributions: (i) a design framework that defines five facets of materiality and suggested applications and benefits, (ii) a series of experimental artworks that articulate and develop that framework and (iii) an articulation of experiential ecologies that offers ways for designers and artists to contextualise their production as part of a vibrant material world.

Mathematical models are a theoretical tool used to understand ecological processes. In this thesis we create mathematical frameworks to describe and evaluate four ecological systems. In the first case study we extend a host-pathogen framework to include a maternal effect which increases the disease resistance of offspring when the maternal environment is poor. Maternal effects impacting life-history traits have been shown to increase the propensity for population cycles. Our contrasting results show maternal effects acting on disease resistance stabilise host-pathogen systems. The second case study examines the impact infection may have on population estimates using Capture-Mark-Recapture (CMR) studies. We show that the estimates using the statistical Program Capture are accurate when capture rates are infection dependent. The final two case studies use spatial, individual-based, stochastic models to simulate disease spread and the colonisation of the Eurasian red squirrel (Sciurus vul- garis) on real-life landscapes. Using novel techniques we highlight the role habitat connectivity has on the dispersal routes which influence the spread of disease and re-population dynamics. Moreover the inclusion of seasonality shows that squirrel population dynamics are driven by the multi-year signal of resources.

Many ecosystems in Kenya are undergoing severe alteration due to landuse changes and associated human impacts. In one such large ecosystem, the Athi-Kapiti plains to the south of the city of Nairobi (and including Nairobi National Park), relatively healthy populations and migrations of large herbivores have been maintained despite increasing concentrations of domestic stock, settled agriculture amongst former pastoralists and increasing industrialization. This study measured the factors that control the vegetation structure and the dynamics of the large herbivore communities. The effects of rainfall on vegetation quality, quantity and species composition was measured inside and outside exclosures, so that the effects of large herbivore grazing could be isolated. The distribution and movement patterns were established by aerial, ground and dung counts. Interactions between xi le herbivores and vegetation were studied by measuring patterns of vegetation use, species composition, chemical analyses for vegetation quality and dung quality. The results show that the quality of vegetation and its biomass reflect the gradient of total rainfall across the Athi-Kapiti plains. Biomass accumulation rates in the exclosures were directly proportional to rainfall. Vegetation biomass and quality changed seasonally and with this, the distribution and utilization patterns by large herbivores. In general, vegetation quality and biomass were highest during the wet months and lowest during the dry regardless of grazing intensity. However, during both seasons, the heavily grazed areas had the highest quality vegetation. The plains in the south experienced heavy grazing in the wet season compared to Nairobi National Park. The impact of heavy grazing in the wet season was to increase plant diversity and produce shorter, more prostrate growth forms of plants. Wildebeest and zebra migrated down the rainfall gradient from Nairobi National Park to the plains in the wet season and up the gradient on their return to the park in the dry season. Small ungulates tended to remain on the plains during the dry season while the large bulk feeders (e.g buffalo) remained in the park. The vegetation selection patterns shown by these herbivores were distinct: large grazers were negatively correlated with quality in contrast to small herbivores, which were strongly positively correlated. Medium sized grazers showed a positive relationship with vegetation quality but in the wet season only. Variation in dung protein between the various species also illustrated the differences in the quality of diet selected by the large herbivores in the ecosystem. Browser dung had the highest crude-protein content followed by small, medium and large grazers in that order. The data revealed significant declines in dung protein with increasing body weight. Zebra, the only non-ruminant in the group, was the only exception to the quality body weight trends. The conclusions drawn from the study show that despite being a highly altered ecosystem the large herbivore movements and diet selectivity patterns still follow rainfall and food quality pulses. This information can be used to make recommendations for the conservation of parts of the ecosystem which are critical to the migrations and the continued survival of this important wildlife population.

This thesis is mainly concerned with the study of solvents of relevance for carbon capture and storage (CCS). The current industrial approach relies on amines to capture CO2, however this thesis describes a range of alternative non-amine based solvents, particularly phenols, saturated and unsaturated aliphatic acids, long chain fatty acids, aromatic acids, and β-dicarbonyl compounds (mainly ketones and esters). The CO2 capture capacities of these substrates have been compared with the industrial model substrate (monoethanolamine, MEA). From this study, phenols show 49-90% CO2 capture capacity whereas aromatic phenolic acids such as gallic acid show up to 100% theoretical CO2 capture capacity. Acetylacetone, a β-diketone shows 88% and the rest of the substrates from this and other groups show 50-60% CO2 capture capacity. However, some of the substrates, particularly simple mono and polycarboxylic acids showed negligible CO2 capture capacity, which can be understood on the basis of pKa. Particular attention has been paid to understanding the formation of bicarbonate salts, which would be expected under aqueous conditions. Clear evidence for their formation is provided by 13C NMR studies. The measured CO2 capture capacities of the new substrates have been correlated with pKa values. Those with pKa values between 9-13 have excellent to good CO2 capture capacity whereas others having pKa < 7, have insignificant CO2 capture capacity. The substrates with low pKa capture less volume of CO2 compared to those having high pKa, but release it more easily on heating. The second body of work is a study on the chemistry of MEA and its oxidised derivatives. MEA is currently the industry standard for CO2 capture. Given that power station flue gases contain large amounts of oxygen, and trace metals which may act as oxidation catalysts, understanding the chemistry of oxidised MEA derivatives is of increasing importance. This can have a significant effect on solvent activity and lifetime, which are important aspects of the economic profile of CCS. MEA was oxidised under a variety of conditions, and a complex mixture of products was formed. Many of these have been unambiguously identified by comparison with commercial or synthetic samples. The main oxidation products were then heated at 100°C for prolonged periods of time to mimic conditions in a commercial CCS system. Thermal degradation of MEA oxidation products was clearly observed, and in some cases, could be rationalised on the basis of established organic reactivity.

The dearth of archaeomagnetic intensity data from the southern hemisphere is a limiting factor in evaluating models of global geomagnetic field evolution during the Holocene. High quality microwave archaeointensity data were obtained from 106 individual ceramic fragments (33 archaeological contexts) from Fiji, Vanuatu, New Caledonia, Bismarck Archipelago and Solomon Islands in the SW Pacific. Complementary Thelliertype experiments, corrected for anisotropy give good agreement with the microwave results. Magnetic mineralogical investigations were performed using a Variable Field Translation Balance (VFTB). Single Curie temperatures between 460°C to 580°C were typically found for most of the island regions indicating an impure to pure magnetite phase. The remaining samples revealed a second TC around 350°C, which was primarily found in samples from the Island of Efate. SEM and EDX investigations have been carried out on selected samples and identified some large titanomagnetite grains (> 50 µm) which were probably introduced by adding beach or river sand to the clay matrix during the manufacturing of the pottery. The newly established archaeointensity curve covers a time interval from 1000 BC to 1750 AD and suggests a more distinctive geomagnetic field strength variation during the last three millennia than those predicted by current global models. The majority of the new data prior to 250 AD exhibit significantly lower intensity than predicted by current global field models (CALS3k.3 and ARCH3k) for the region, with an apparent intensity minimum at 250 BC reaching as low as 50% of the present-day field strength. Between 400 AD and 1500 AD, the data are broadly consistent with the global field models but with a 20% higher field between 1200 and 1400 AD. Two significant increases of the geomagnetic field strength between 200 AD and 400 AD as well as 1200 AD and 1400 AD correlate with intensity peaks found in French data sets and hint at an occurrence of archaeomagnetic jerks in the SW Pacific. Strong climatic variations around 1300 AD in the SW Pacific seem to fit to the second intensity peak and could support the hypothesis that archaeomagnetic jerks influence the climate. Nevertheless more data are needed for an unambiguous identification of such geomagnetic events and their correlation with the climate. For the first time in the SW Pacific this study has attempted to provide chronological information using archaeomagnetic dating. Three poorly dated archaeological sites on the Santa Cruz Islands, which are an important link between Near and Remote Oceania, were investigated. Similar archaeointensity results found from sites RF-2 and RF-6 on Reef Island contrast with a significantly higher value obtained from Nendö (SZ-8) potsherds suggesting a different age for this site. The SW Pacific archaeointensity results highlighted the potential of a new and independent dating tool that can be used by archaeologists to solve problems that cannot be addressed using conventional dating methods, such as the timing of the advance of the Lapita peoples across Oceania.

The shelf seas account for only 9% of the surface area of the global ocean, yet are estimated to be responsible for 16% of global primary productivity and 47% of global annual carbon export. While the importance of these seas in the marine carbon cycle is well-recognised, the relationship between the physical environment and the magnitude and fate of biological production is still largely undetermined. Using a range of approaches, this project examined the relationship between physical processes and biological carbon flow over various time and space scales in a temperate shelf sea. Autonomous high-resolution sensor surveys in Liverpool Bay and the Irish Sea reveal that this region is net autotrophic on an annual scale with estimates of regional net community production rates based on oxygen mass balance ranging from 1.3 to 4.2 mol C m-2 y-1. The highest NCP rates were measured in a Region Of Freshwater Influence (Liverpool Bay). Ship-based sampling in the Celtic Sea illustrated the influence of water column structure on rates of primary and bacterial production and community structure. The highest rates of primary production were observed in the fully mixed water column (55.1 mmol C m-2 d-1). However, bacterial production rates were lowest at the mixed site (6.8 mmol C m-2 d-1). In contrast, primary production and bacterial production at the seasonally stratified sites and shelf break were comparable (34.8 mmol C m-2 d-1 and 37.2 mmol C m-2 d-1 respectively, and 12.9 and 12.9 mmol C m-2 d-1 respectively). The phytoplankton community was dominated by diatoms at the mixed site and dinoflagellates at the seasonally stratified sites, with both diatoms and dinoflagellates occurring at the shelf edge. Bacterial carbon demand at each site was estimated using published empirical relationships. However, at three out of the four sites examined, the estimated daily rate of total primary production was insufficient to meet the daily bacterial carbon demands, implying that either an external source of carbon was required or that the assumptions used during estimates of bacterial processing of carbon were incorrect. Finally, the definition of export production in shelf seas is examined in relation to the potential pathways of carbon flow on daily and seasonal timescales in shelf seas.

In recent years, seabird tracking studies have become ever more popular as a means of informing and evaluating the effectiveness of marine protected areas and offshore energy developments as well as in understanding the ecology and behaviour of seabirds. This study uses tracking data collected from the European shag, Black-legged kittiwake, Northern gannet and Brown booby to identify important foraging areas around four seabird colonies located in the UK, Channel Islands and Anguilla and examines the impact of offshore renewable energy developments in the Channel Islands. As well as providing examples of how seabird tracking data can be useful in informing marine spatial planning, this study also considers the impact that sample size and the sample composition may have on the foraging areas predicted for any colony. Small sample sizes are a common feature of tracking studies, often due to logistical and financial constraints, meaning that seabirds are often tracked over short spatial and temporal scales which may not fully represent the important foraging areas and behaviours of the colony or individual. This study therefore provides recommendations to improve the predictions of area use and foraging strategy for future tracking studies to ensure the most representative and useful data is collected and used to inform marine spatial planning issues.

Planetary dynamos, resulting from fluid flow in electrically conductive parts of their interior, are thought to be highly time dependent. Currently, our understanding of temporal variation of these fields is limited because we only have observations for one example, the Earth. To overcome this, data acquired by 6 NASA space missions between 1973-2003 are used to investigate possible time variation (secular variation) of Jupiter's magnetic field. Previous attempts to constrain jovimagnetic secular variation have been inconclusive or ineffective for various reasons. We attempt to resolve these issues in a number of ways. All data available within 12 Jovian radii are considered and modelling of the external field takes place for each individual orbit. Whilst we find that non-uniqueness limits resolution of Jupiter's magnetodisk configuration, it does not prevent the resulting field from being constrained: we find this field does not vary greatly with time and conclude that solar activity is not a strong control on its generation. Of particular significance is our regularised minimum norm approach to modelling the planetary field. This approach allows construction of numerically stable models with small-scale (high spherical harmonic degree) structure that directly fit the observations. Two models of Jupiter's magnetic field are presented: the first time-averaged over the whole dataset, whilst the second allows for linear time variation of the field. With the inclusion of secular variation, we find an improved fit with fewer additional parameters, suggesting that changes to the field can be resolved. Our favoured solution indicates a 0.012%/year increase in Jupiter's dipole magnetic moment over the investigated time period; this value is roughly a factor of four less then that currently observed at Earth. Relating field changes to internal dynamics, we use our models of secular variation to infer the motion of material at the top of the dynamo source. Velocities on the order of 200km/year are found, approximately an order of magnitude faster then at Earth, but with similarities in flow configuration. Further analysis shows that our optimal solution may be too conservative and that we are able to approximate a polar configuration previously only attained by models employing the additional constraint of auroral footprint position. In doing so, the models favour reversed polar flux, which has important dynamical implications for the interior. The longitude of data used in these models are defined relative to the System III 1965.0 rotation period, itself defined by the magnetic field. Thus, some of the secular variation could result from inaccuracies in the determination of this reference frame. We find that such an affect cannot explain all the observed secular variation. The constraint of our magnetic models on changes to planetary rotation rate, allows a bound to be placed on angular momentum transfer between the atmosphere and deep interior, analogous with variations in Earth's observed length-of-day. This provides strong observational evidence against models directly linking surface winds to deep Jovian convection.

Furthering our understanding of the Southern Ocean as a critical component of the global climate system and its variability in both space and time is the focus of many investigations. This thesis aims to add to that effort by addressing two important questions. The first chapter will set the role of the Southern Ocean into greater context, clarifying the framework and background for its examination. The second chapter will examine a simple 1-dimensional model of the modification of the Lower Circumpolar Deep Water as it moves southwards across the ACC and into the Weddell Gyre. This leads into the third chapter which will address the first question of how variable are the characteristics of the Lower Circumpolar Deep Water? The variability of the salinity maximum associated with the Lower Circumpolar Deep Water is examined in order to address its spatial variability – from entry via the South Atlantic, through its circumpolar transit, ending with a terminal repository in the Weddell Gyre – by making use of freely available hydrographic data. It also examines temporal variability where repeat data allows, and provides new estimates of deep ocean mixing rates. The fourth chapter will seek to address the second question of how variable was the volume transport during the Lazarev Sea Krill Study (LaKriS) cruises? The LaKriS cruises provide a rare set of semi-repeat grids of hydrographic measurements near the Greenwich Meridian. This provides the opportunity to attempt to assess seasonal and inter-annual variability by making use of an inverse modelling technique. The fifth and final chapter will set the knowledge gained from addressing these two questions against the wider knowledge of the Southern Ocean system and consider the implications for future oceanic sampling and research.