No description available.

Peatlands are globally important for carbon storage, water quality and biodiversity. However, many have been degraded by land use, and efforts to conserve or restore them are often contested by stakeholders with different objectives. Peat accumulates as a result of a complex network of interactions, which makes it challenging to predict the impact of climate and land use. Stakeholders’ knowledge may help to provide insights into these interactions and into the issues that underpin conflict. To investigate the impact of social and ecological factors on blanket peatland carbon storage, an interdisciplinary approach was used to couple cognitive and peatland development models. Blanket peatland stakeholders developed fuzzy cognitive maps based on their perceptions of peatland interactions, which they validated to agree on the structure of an aggregate network. To explore the impact of land–use objectives on carbon stocks, stakeholders proposed changes to a set of factors that controlled the state of the network. The changes identified to improve carbon storage and water quality had a positive effect on carbon stored, but those that were proposed to support local livelihoods had no effect on carbon. This was partly because some stakeholders perceived that supporting livelihoods was incompatible with measures that were likely to result in shallower water tables. However, further discussions between stakeholders suggested that land–use objectives could complement each other. To enrich the results of the network model, the DigiBog peatland model was modified to simulate blanket peat accumulation. Using two factors from the cognitive model, climate change and gully blocking, two novel modelling studies were produced. The first showed that existing peatland development models may overestimate peat accumulation because they aggregate climate variables into annual rather than weekly inputs; the second, that gully blocking is needed to arrest peat losses from oxidation in gullied systems, but that these losses would not be recovered 200 years after gully blocking. The combination of both cognitive and process–based modelling provides an example of how stakeholder knowledge can be incorporated into simulations of complex ecosystems which is likely to be applicable to other social–ecological systems where land use is contested. In this case, doing so provided holistic insights into how stakeholders’ perceptions, and the impacts of climatic forcing and restoration, affect carbon storage in blanket peatlands.

The world’s second largest tropical wetland is found in the central Congo Basin. Ambiguous grey-literature reports of peat, coupled with the large area of wetland suggest this region may be a globally significant carbon store. In this thesis I aim to establish whether this region, known as the Cuvette Centrale, harbours significant peatlands, to characterise them, compute the first estimate of peatland extent and C stocks based on ground data, to determine the factors which led to peat initiation and their maintenance today. Fieldwork within the Likouala Department, Republic of Congo, confirmed widespread peat presence. Peat-vegetation associations were recorded in the field, which combined with remotely sensed radar, optical and elevation data was used to estimate the area of peatland; at 145,529 km2 (95% CI, 134,720-154,732 km2), the Cuvette Centrale is the single most extensive tropical peatland complex in the world. The peat is shallow (maximum depth: 5.90 m) and characterised as non-domed, nutrient poor systems, occupying large interfluvial basins. Area measurements combined with those of peat depth, bulk density and C concentration, collected in the field, suggest a total peat C stock of 30.2 Pg C (90% CI, 27.8-32.7 Pg C). This increases the current global tropical peatland C stock estimate from 88.6 Pg C to 115.8 Pg C. Radiocarbon dates show peat initiated early Holocene (dated from 10555 cal yrs BP onwards), with a possible Mid- to Late-Holocene hiatus in peat accumulation, with both likely linked to changes in regional precipitation. Pressure transducers measuring the peatland water tables, rainfall estimates and water source geochemistry imply that the peatlands today are predominantly rain-fed systems. My discovery that the Congo Basin, not tropical Asia, is home to the world’s largest single peatland complex elevates the current global peatland C stock estimate from 88.6 Pg C to 115.8 Pg C and will require new regional management plans if the destructive fate of tropical Asian peatlands are to be avoided in central Africa.

In this thesis I took a novel interdisciplinary approach involving remote sensing, ecological and palaeoecological techniques to address some of the most fundamental gaps in our understanding of Peruvian peatlands. The existence of these peatlands was only recently confirmed and although they were known to store large quantities of carbon, initial assessments of carbon stocks were highly uncertain. In addition, little was known of their biodiversity or how they have developed. Firstly, I used data fusion remote sensing and extensive field data to generate a high resolution, landscape scale map of peatland ecosystems in the largest peatland complex in Amazonia. This approach confirmed that peatland ecosystems in northern Peru are the most carbon dense ecosystems in Amazonia storing up to 1391 ± 710 Mg C ha-1, and have a total carbon stock of 3.14 (0.44–8.15) Pg C, which equates to nearly 50 % of the total above-ground carbon stocks of the whole country. Secondly, I established a new network of floristic inventory plots and described the composition and diversity of peatland tree communities. I demonstrated that peatland pole forest has the lowest alpha diversity of all tree communities in lowland Amazonia. In contrast, by comparing these data with three larger plot networks from other ecosystems in the region, I also showed that they have surprisingly high beta diversity, and harbour important populations of species that were previously thought to be restricted to other habitat types such as white sand forest. Finally, pollen analysis was undertaken across eight peat cores from two sites to test the significance of historical processes in determining current patterns of composition and diversity. Both autogenic (internal biotic) and allogenic (external environmental) processes operating through time were important determinants of current floristic patterns. Demonstrating that such historical processes have an important role in determining the composition of tropical ecosystems is valuable as they are often overlooked – or in many cases impossible to study in such detail. Overall this thesis shows that peatland ecosystems in the Peruvian Amazon have high conservation value both as a carbon store and for regional ecosystem diversity. In addition, peatland ecosystems provide an exciting opportunity to investigate the importance of fundamental historical and ecological processes for determining the composition and diversity of tropical forests.

No description available.

The primary aim ofthis project was to use a range of molecular markers to estimate genetic diversity levels of a range ofcommon plant species in peatbogs, in cut and uncut areas of six differentbogs across Northern Ireland. The main species studied were the moss Polytrichum commune, the ericaceous angiosperm Calluna vulgaris, and the sedge Eriophorum vaginatum. For the moss, levels of genetic diversity were lower in fragmented populations, as predicted by population genetic theory. For the two higher plant species, no decrease in genetic diversity was found at the cut sites. Calluna vulgaris showed an increase in genetic diversity at cut sites, possibly associated with the effects of a substantial, soil seed bank, while for Eriophorum vaginatum all populations examined had a similar level of genetic diversity. The differences observ~d between the three species can be related to their different life . history traits. E. vaginatum and C. vulgaris are long lived species with the capacity for clonal growth and any detrimental effect of habitat fragmentation is likely to take many more generations to become apparent, if such an effect is to be observed. P. commune is a bryophyte where the haploid gametophyte is the dominant generation, with a short life span and a small dispersal distance compared to C. vulgaris and E. vaginatum, it is not unexpected that such a species would be the first to display the detrimental signs of habitat fragmentation. Based on the findings ofthis study, recommendations are made for the efficient conservation and management of peat bogs.

Peat deposits from three sites in south wales are examined and subjected to pollen analysis. The results show them to represent a period of over 12,000 years of peat accumulation and throw some light upon conditions in late glacial and post glacial times in south wales. The late glacial evidence points to severe climatic conditions in pollen zone 1 within the highlands of south wales but fluvial activity is present within the coastal lowlands. Favourable conditions to the north east of this escarpment enabled the early deposition of mud and clays with some establishment of vegetation from the Bolling oscillation. Highland corrie glaciation is suspected at the time. Within the warm Allerod period tree parkland spread across the country and persisted although restricted during zone 3.Periglacial activity is general and it is thought unlikely that corrie glaciers developed at this time. The vegetation colonization through post glacial times provides interesting examples of westerly influence, plant migration and the importance of ecological factors. This is well shown by comparing the vegetation succession and plant dominance at Waen Ddu and Cwmllynfell. The richer vegetation soils and more favourable easterly siting od Waen Ddu also proved more attractive to prehistoric man. Here vegetation interference is more apparent, and took place earlier than at Cwmllynfell. several different clearance methods and phases of activity are noted at the two sites and these have been tentatively correlated with archaeological periods.

No description available.

Anthropogenic emissions of the greenhouse gases carbon dioxide and methane have stimulated a rise in global surface temperature of 0.76 ºC since the turn of the 20th Century. Such climate warming has already had significant impacts on the terrestrial biosphere, raising concerns that ecosystems will feed back to future climate by altering the balance of carbon flow between the land and atmosphere. It is well established that warming can directly affect rates of photosynthesis and ecosystem respiration, which together dictate the carbon balance of most ecosystems. However, warming is also causing shifts in the productivity and composition of vegetation, and there is growing recognition that this can have indirect effects on carbon cycling via its influence over soil properties and the activity of the soil microbial community. Despite this, much uncertainty currently surrounds the effects of warming on vegetation, both between individuals of the same species and at the plant community scale. Furthermore, the consequences of vegetation change at either scale for carbon dynamics are not well understood when considered in tandem with warming. Northern peatlands are of particular relevance to ecosystem climate feedbacks, holding one third of global soil carbon in regions vulnerable to rapid temperature change. The aim of this thesis was to explore how different scales of vegetation change regulate peatland carbon cycle responses to warming. This was achieved using field experiments across Europe, including manipulations, reciprocal transplants and gradient studies, which integrated a range of approaches for detecting responses from the genetic to ecosystem scale. By measuring the growth responses of dominant plant species to warming and the presence or absence of different plant functional types (Chapter 2), I reveal interdependencies between peatland plants that will cause community change over decades to centuries. I suggest that such responses occur due to both plant-plant interactions and the influence of vegetation over microclimate, the latter of which I also show to be regulated by vegetation composition (Chapter 3). Additionally, I demonstrate that warming at different temporal scales has contrasting effects on the metabolism, photosynthesis and growth of Eriophorum vaginatum (Chapter 5), causing decreases in growth over seasons to years through phenotypic plasticity and increases in growth over centuries to millennia through natural selection. I show that such adaptation of individual plants to rising temperature has potential consequences for peatland carbon dynamics. Moreover, I reveal that changes in vegetation composition at the community level could destabilise the peatland carbon stock under warming by accelerating decomposition of ancient carbon (Chapter 4). Through these findings, I provide novel insight into the scales and mechanisms by which vegetation responses to warming impact carbon dynamics. Given the key role of northern peatlands in the global carbon cycle, I suggest that warming effects on peatland vegetation may have considerable consequences for future climate by controlling liberation of peatland carbon into the atmosphere.

This work examines the chemical composition of peatland soils and their associated fluxes of organic carbon. This thesis has demonstrated through the long-term incubations of peat material that ex-situ greenhouse gas (GHG) production from peats is variable over time, despite constant conditions during the incubation period, suggesting that the uncertainties associated with ex-situ GHG measurements of peat may have often been underestimated. In addition, the in-situ GHG monitoring carried out in this research suggests the ability of O2 to penetrate into even the deepest regions of the peat profile, potentially providing a possible explanation for the temporal variability in GHG production. This work has also demonstrated how peatland restoration may potentially affect the mobilisation of carbon from peatland sites. The simulations of runoff derived from Sphagnum vegetation and areas bare peat show significantly different chemical compositions and behaviour when treated with iron flocculating agents, suggesting that the re-vegetation of upland peats to reduce GHG emissions may have the additional benefit of making the peat runoff easier to treat. However, the assessment of the restoration campaign carried out across a lowland peat mire suggests that the restoration of the site has introduced a significant variability into the hydrology of the restored areas of the bog, which may have triggered a change in the chemical composition of the bog and increased in-situ concentrations of GHGs within the peat. In addition, the utility of chemical analysis in the differentiation of spatially and temporarily different peats and prediction of their operational behaviour has been examined. This analysis has suggested that although certain analytical methods may be inappropriate for predicting future behaviour, they may prove useful as a means of monitoring peatland restoration.