Peatlands are important ecosystems in terms of biodiversity, nature conservation and the hydrological cycle. They also represent the most significant long-term carbon store in the terrestrial biosphere. As there is much concern regarding the effect of climate change on peatlands, it is vital that their ecohydrological responses to future climate scenarios and human influence are understood. Studying past peatland responses to environmental change in a holistic fashion is a key part of this process. Testate amoebae are increasingly being used as indicator organisms m both palaeoenvironmental and contemporary applications. The ecology of testate amoebae and their responses to environmental factors was explored. It was found that the taxon Hyalosphenia subf1ava can thrive in peatlands severely affected by wildfire. An extensive testate amoebae training set, which includes a number of previously unquantified modem analogues, has been used to construct a new, robust transfer function for Northern England (NE). This transfer function was comprehensively tested using a novel spatially-independent cross-validation approach against two other existing models. A high-resolution, multiple-proxy palaeoenvironmental reconstruction spanning the last -2400 years from Malham Tam Moss, Northern England is presented. Peatland water tables were reconstructed using the NE transfer function together with plant macrofossil and peat humification methods. Local and regional vegetation changes were inferred through pollen and plant macrofossil analyses, changes in geochemistry were examined through X-ray fluorescence and burning events identified through charcoal concentrations. Peat and carbon accumulation were calculated through bulk density, loss-on-ignition and CIN ratio. A high-precision chronology was achieved through Bayesian modelling techniques applied to a sequence of AMS 14C dates and spheroidal carbonaceous particles.

The combined effects of ditchwater management regime and sub-irrigation spacing on water table fluctuation have been investigated for two low lying agricultural peatlands in England, West Sedgemoor in the Somerset Moors and Methwold Fen in the Norfolk Fenlands. The consequence of the resulting soil moisture regimes for microbially mediated mineralisation of soil organic matter has been examined on peat samples collected from the upper metre of peat profile from these two test sites. It is shown that sub-surface tile spacing has a strong influence on the transference of ditchwater regime to the mid-tile point in the field. Where sub-irrigation spacing is greater than 40 m the mid-point water table falls to similar levels experienced without any form of sub-irrigation intervention. Where sub-irrigation is at 10 m intervals the mid-point water table was found to be close to the water regime maintained in the ditches. Differences in field water-table level can lead to considerable variation in the matric potential experienced at different depths in the peat profile. As a consequence, peats at different stages of degradation (linked to depth) and under different land uses can exhibit variable physical and hydraulic properties. The von Post scale, which describes the degradation status of peats, has been linked to these physical properties but no simple model has been found between these properties and the von Post score. A good relationship has been found between saturated hydraulic conductivity and the van Genuchten alpha value which itself was related to the air entry value for all peats except the amorphous (unstructured) peat from Methwold fen. The water management regime, in conjunction with variations in physical and hydraulic properties of different peat types, influences the peat microbial community structure. At West Sedgemoor those peats that are wetter have predominantly anaerobic species, whilst those in drier environments have a greater proportion of aerobic species. At Methwold Fen the variable nature of the water management strategy appears to have homogenised the microbial community throughout the entire peat profile, resulting in more aerobic microbes in the deeper peat deposits. The type of microbial community and the degree of peat aeration dictate the efficiency with which soil organic matter is mineralised. Over the period October 2004 - July 2005 the rate of mineralisation in Methwold Fen peat samples averaged 0.40 g CO2-C m-2 hr-1 in saturated samples whilst in drier peat it averaged 0.72 g CO2-C m-2 hr-1. This clearly demonstrates that a wetter peat profile minimises the rate of microbially mediated organic matter mineralisation. Land use exerts an equally strong influence on microbial activity and can mask the true extent of soil organic matter mineralisation. Root exudates may offer an alternative source of organic carbon for microbial metabolic processes. Where the water table was maintained at 0.3 m below the soil surface respiration rates on grass covered West Sedgemoor peat samples was, at maximum, 1.46 g CO2-C m-2 hr-1 whilst on bare Methwold Fen peat samples it was less, at 1.06 g CO2-C m-2 hr-1. After removal of all surface vegetation the average rate of respiration switched, with Methwold Fen peats exhibiting a greater rate of organic matter mineralisation (7.27 µg CO2-C g soil-1 hr-1) than West Sedgemoor peats (3.8 µg CO2-C g soil-1 hr-1). Sub-irrigation modelling, using a drainage theory based water table model, can adequately simulate the soil water balance. Coupling the output of a comparable hydrological model (SWAP) with a process based model of nutrient dynamics(ANIMO) demonstrates that under future climate scenarios closely spaced subirrigation could reduce the mineralisation of soil organic matter to the atmosphere and reduce subsidence by up to 2mm year-1, thus reducing agricultural peatland contributions to greenhouse gas emissions and improving peatland sustainability1. Even partial aeration of a moist soil profile can lead to high rates of mineralisation. However, a combination of ditchwater management and sub-irrigation can, improve the sustainability of low lying peatlands if the management regime maximises the period of complete peatland inundation. 1 Sustainability being defined as maintenance and/or improvement of peat soil resource quality and/or longevity through the reduction of present day rates of subsidence and mineralisation.

Peatland and mineral soils are important reservoirs in the global carbon cycle and there is much interest in characterising the response of their carbon stores to temperature perturbations. Peatlands are notable also because of their dual function as carbon capacitors and amplifiers in the Earth System, sequestering CO2 into soil biomass, but also producing CH4, a more potent greenhouse gas. Cycling of H2 in soil is poorly characterised despite accounting for 80% of the annual atmospheric H2 sink. This study adapted a novel heating and cooling instrument, typically used for process chemistry, as a new form of soil incubator. Its ability to generate cyclical and ramped temperature regimes was exploited to simulate diurnal soil temperature variations while concentrations of CH4, CO2 and H2 were measured. These data were used to calculate turnover rates, F-ratios, Q10 factors, activation energy and Gibbs free energy for specific microbial and enzymatic processes. Determinations were compared to data from conventional stepped temperature incubations presently used in models. Lower rates of methanogenesis occurred in cyclical versus constant temperature incubations; however, the effect existed only in high trophic status peat. Temperature sensitivity of anaerobic CO2 production also was linked positively to nutrient status. Rates of H2 production were unaffected by temperature regime but were highly sensitive to temperature. Rates of aerobic CO2 production in mineral soil were more sensitive to temperature in organic matter rich soil while H2 cycling in the same soil occasionally exhibited unusual behaviour that suggested H2 (and possibly CH4) was produced in anoxic micro-sites. Methanogenesis functioned near its thermodynamic limit and F-ratios were consistently lower than values used in most peatland CH4 models. Living plants were excluded from these experiments and thus findings carry the caveat that a major source of labile carbon was absent. The importance of root biomass to carbon mineralization processes was established during characterisation of in situ conditions at the sites from which peat was obtained for the incubations.

Around 9% of the UK land area is blanket peat, found mostly in the uplands. Peat is composed of partially decomposed dead plants with high organic matter content up to 95%. Intact blanket peat provides a range of ecosystem functions including filtering atmospherically-deposited pollutants such as lead, removing carbon dioxide from the atmosphere by plant photosynthesis and water storage.

Peatlands occupy a mere 3% of the world's land mass, but store up to one third of terrestrial carbon stocks. Peatlands are widely regarded as carbon sinks owing to their ability to sequester more carbon than is released. Carbon cycling in peatlands is driven by environmental conditions e.g. water table levels, temperature and pH; substrate quality i.e. the ease with which microbes can synthesise the carbon; nutrient availability and the composition of the microbial community. Peatlands are valued not only for their ability to sequester carbon, but also for the range of ecosystem services which they provide including the provision of food, recreation and leisure, a source of income for rural communities, water supply and as habitats for a range of flora and fauna. As a result, management of peatlands is widespread, with the four most common methods of management of upland blanket bogs being afforestation, drainage, grazing and burning. To date, little work has been carried out on the effects of such management practices on carbon losses or drivers of the carbon cycle. The aim of this research was to identify how these management practices influenced losses of carbon from peatlands as well as the chemical and physical drivers of the peatland carbon cycle. A combination of field and laboratory work was carried out on managed peats with an unmanaged site at the Moor House National Nature Reserve in Cumbria. Field monitoring involved measurement of dissolved organic carbon (DOC) in the peat solution, water table levels and carbon dioxide gains and losses. Laboratory analysis was carried out on cores of peat to examine nutrient concentrations, the structure of the peat in terms of porosity and density; carbon stocks and the quality of the carbon. The results of this research demonstrated that all sites including the unmanaged site acted as carbon sources. Greatest losses occurred from the afforested site, where losses of DOC were significantly higher than all other sites and some of the highest losses of carbon dioxide were found. In contrast, the site that was burnt on a 10 year rotation was found to be a very slight carbon sink, held the most carbon within the peat and lost the least amount of DOC. Few significant differences in the chemical composition of the peat were observed between the sites, however, lignin, the most recalcitrant fraction was found to be significantly lower in the burnt (every 10 years) site, which had the highest carbon content. Lignin was identified as the dominant constituent of the peat for all the sites, with highest concentrations present in the afforested site. The high lignin content of the peats from all the sites indicated that the peats are in the latter stages of decomposition, and are thus fairly recalcitrant. The higher lignin content in the afforested site, coupled with the highest losses of DOC, some of the highest CO2 losses through ER (ecosystem respiration), however, suggest that the chemical composition of the peat is not a strong a driver of the peatland carbon cycle. Temperature was found to be the dominant driver of ER, accounting for between 54 and 92 % of variation in the data. The afforested site was the only treatment where a significant relationship between temperature and ER was not identified. Rates of primary productivity were highest in the burnt and grazed sites indicating that regeneration of the vegetation through management is of key importance in terms of sequestering carbon. The lowest primary productivity was identified at the drained site, where concentrations of nitrogen were also lowest. In terms of the structure of the peat, the air filled porosity of the burnt and grazed (every 20 years) site was greatest, however no linkages were established between the structure of the peat and gaseous carbon losses. This thesis has provided a unique insight into the effects of land management on the drivers of the peatland carbon cycle, carbon dioxide gains and losses, and DOC production. Further work should focus on examining the effects of the intensity of land management practices on peatland carbon budget for example, comparing low and high temperature burns, or closely spaced drains with drains that are located far apart. The results of this thesis suggest that future management needs to focus on encouraging increased PP by managing water table levels and promoting growth of peat forming species of vegetation such as Sphagnum. Light burning was also found to increase water table levels and peat solution acidity, thus reducing losses of DOC into the peat solution. The results demonstrated that temperature is the most important control on ER, and under climate change losses are likely to increase, therefore, the need twas found to be strongly linked to water table levels, pH and the carbon quality, with higher concentrations of holocellulose resulting in reduced losses of DOC.

Vegetation of a particularly diverse area of undrained flood-plain mire is described. Factors and processes important in determining the distribution of community-types within the study area have been investigated. Investigation of the alluvial stratigraphy has revealed the presence of a complex pattern of peat cuttings. Succession within the peat cuttings, with appropriate vegetation management, has led to the formation of Cladium mariscus-, Phragmites communis- and Juncus subnodulosus-dominated fen vegetation and, in some areas, poor-fen communities with much Sphagnum. In areas not cut for peat, management his also sustained various rich-fen herbaceous communities(different to those of the cuttings); in its absence, fen carr develops. The differences in successional development in peat cutting areas and those not cut for peat is mainly due to difference in hydrological status. Long term experiments established to examine effects of different management techniques are described. Studies of peat and peat water chemistry in a representative selection of community-types has demonstrated local areas of high salinity, caused by incursions of brackish water due to exceptionally high tides and, particularly, the influence of underlying estuarine deposits. Although large areas are flooded by river water there is little evidence for eutrophication, except very locally. Indeed, 'seral oligotrophication' is occurring in isolated areas. This may preceed, but is not a pre-requisite for, Sphagnum invasion. The most species-rich communities are developed in non-saline areas with a fairly stable water level; they may, however, be dependent upon flooding by river water for maintenance of their base status.

Plant macrofossil analyses of 5 peat cores obtained from undisturbed (i.e. not cut for peat) flood-plain mires situated in the Ant Valley in the Norfolk Broads have indicated the successional development of the vegetation over approximately the last 2000 years. These have been supplemented by analysis of diatom and foraminiferal content of some of the deposits together with the application of radiocarbon dating to give an approximate chronological framework within which to place macrofossil zones. Macrofossil assemblages have been grouped into 5 major zones. These are interpreted as representing: A: Carr woodland communities (pre-Roman) B: Salt marsh communities formed during a marine transgressive phase (Romano-British). C: Fen tussock/carr communities indicative of drier conditions (Early Medieval). D: Aquatic communities indicative of wetter conditions (late/post Medieval). E: Communities suggestive of present day vegetation influenced by human management (post Medieval-present day). The zones have been interpreted largely in terms of the response of the vegetation to changes in sea-level, climate and management over the last two millenia. Macrofossil analyses were also carried out on samples collected from a variety of former peat cuttings in the Catfield and Irstead Fens. Successional changes were deduced and compared with previous investigations. Historical studies of archival documentary material has suggested that the post-Medieval use of peat as a fuel in Norfolk was largely a feature of the eighteenth and nineteenth centuries and that it was a commodity of major importance amongst the poor during this time. Study of archive material specifically relevant to the Catfield and Irstead Fens has suggested that at least some of the former turf cuttings may have been dug in the first half of the nineteenth century but many may also date from the second half.

Evidence has shown that changes in land-use can affect the potential of the soil to act as a methane sink. The Lakenheath site consists of drained fenland peat, which is being re-converted to wetland. At the time of this study the site consisted of four land-use types, an intensively cultivated plot, unmanaged grassland, woodland, and wetland. Peat cores were taken from the four plots and compared for their ability to act as a methane sink. CH4 uptake was measured throughout the depth profiles using gas chromatography. Clear differences in methane oxidation rates were recorded with depth and land-use. The woodland soil showed the highest capacity for atmospheric methane oxidation, and the wetland soil was only profile that had a distinct peak of methane oxidation activity (just above the water table). Despite the change in land-use, all four soils had the capacity to oxidise both high and low concentrations of methane and so acted as a methane sink. The only exception to this was the wetland soil after persistent rainfall. The uppermost layers were water saturated and all soil sections failed to oxidise methane. Methanotroph diversity in the four soils was compared using molecular biological and enrichment techniques. Total DNA was extracted from depth profiles of the four soils and PCR amplified with 16S rRNA methanotroph group-specific primers and primers specific to subunits of the pMMO and AMO (pmoA and amoA), sMMO (mmoX) and MDH (methanol dehydrogenase, mxaF). In addition, DNA was extracted from the top 5 cm of the cultivated (drained) and flooded soil and PCR amplified with primers specific to subunits of the pMMO and AMO. These PCR products were cloned and gene libraries constructed for each soil. No significant differences were observed in retrieved methanotroph sequences from these two soils, suggesting that the methanotroph population had not altered after flooding. The sequences obtained in the molecular study were predominantly amoA sequences from nitrifiers and pmoA sequences from type II methanotrophs. No type I pmoA sequences were retrieved. Type I methanotrophs, however, were isolated directly from the peat soil in the enrichment study.

Les tourbières représentent 2 à 3% des terres émergées et stockent entre 10 et 25% du carbone des sols. Les tourbières sont soumises à des contraintes anthropiques et climatiques importantes qui posent la question de la pérennité de leur fonctionnement en puits de C et de leur stock. Une meilleure compréhension de ces écosystèmes est nécessaire pour déterminer les facteurs et les effets et interactions de ces facteurs sur les émissions de gaz à effet de serre (GES). Ce travail a consisté à suivre les émissions de GES et les facteurs contrôlant dans La tourbière de La Guette (Sologne) pour établir son bilan de C. En parallèle des expérimentations sur l’effet de l’hydrologie sur les flux ont été menées, enfin un suivi sur 4 sites a été réalisé pour étudier la variabilité à l’échelle journalière. Les résultats de ces travaux montrent que la tourbière de La Guette a fonctionné en source de C (-220 ± 33 gC m-2 an-1) et ce malgré un niveau de nappe élevé. Ils montrent également l’importance de la variabilité spatiale des flux estimés à l’échelle d’un site. Les expérimentations confirment l’importance de l’hydrologie et suggèrent à haut niveau de nappe d’eau des phénomènes liés au transport des gaz. Enfin l’étude de la variabilité journalière montre que la sensibilité de la respiration à la température peut être différente le jour et la nuit et que la synchronisation entre les températures du sol et la respiration peuvent améliorer la représentation de cette dernière. / Peatlands cover only 2 to 3% of the land area but store between 10 and 25% of the soil carbon. The outcome of the anthropic and climatic pressure on these ecosystems is uncertain regarding their functions and storage. A better understanding of these ecosystems is needed to determine the factors and their interactions on greenhouse gas (GHG) emission. This work consist in monitoring GHG emissions and controlling factors in a Sphagnum peatland to estimate its carbon balance. Experimentation on mesocosms were carried out to explore the effect of hydrology on the fluxes and a monitoring on 4 sites was made to study the daily variability. Results show that La Guette peatland was a carbon source (-220 ± 33 gC m-2 an-1) in spite of the high water table level. The importance of the spatial variability measured in the site was also demonstrate. The hydrology effect was confirmed by the mesocosms experiments and high water table level shows that gas transport might have an effect. Finally the study of the daily variability show that the temperature sensitivity of the respiration might be different between day and night and that synchronizing soil temperatures and respiration can improve the respiration representation.