Nitrogen enrichment in sensitive habitats has become a matter of concern in recent years, and has led to the need to develop a bio-monitoring scheme that could be used by non-specialists to undertake site evaluation across the UK. Epiphytes are ideal candidates for such a project due to their high sensitivity to atmospheric pollutants. Indicator species analysis identified a comprehensive UK-specific list of epiphytic indicators for NH3 pollution in the UK that for the first time also included microlichens. A shortlist of easily identifiable, widely distributed indicator species was produced from these and used to trial the effectiveness of a frequency based scoring system. A simple, un-weighted frequency based scoring system founded on the Lichen Acidophyte Nitrophyte index was found to correlate strongly with NH3 concentrations in air. Assessment of other atmospheric chemistry and climate variables with the scoring system showed that NO2 concentrations in air and bark pH were confounding factors. The influence of pH was further validated by observations in a field experimentation plot where different N forms were applied. In order to address this, a regression equation was formulated that incorporated NH3, NO2 and bark pH to produce a predictive model that could potentially be used to evaluate site condition with respect to atmospheric N pollution, as defined by the combined effect of NH3 and NO2. Investigations into the biochemistry of an indicator species identified that rates of phosphomonoesterase (PME) activity were higher in N-sensitive species than N-tolerant ones. N-sensitive lichen thallus pH measurements for were also consistently more acidic than those of N-tolerant species, and closer to the observed pH optima of PME activity.

Using spatial patterns to infer biotic and abiotic processes underlying plant population dynamics is an important technique in contemporary ecology, with particular utility when investigating and shrub population dynamics, for which experimental and observational methodologies are rarely feasible. Using a novel one-class classification technique, the locations of over 17,000 Spartocytisus supranubius individuals were mapped from aerial imagery generating a spatially extensive (162 ha), yet accurate, dataset. The recent rapid increase in studies using pattern-process inference has not been accompanied by a rigorous assessment of the behaviour of these techniques, nor an appraisal of their utility in addressing ecological research questions. The first part of the thesis addresses these concerns, investigating whether current methodologies are adequate to test hypotheses concerning spatial interactions. A literature review reveals a preponderance of studies of small, little-replicated plots. The results of the research raise concerns about the utility of spatial point pattern analyses as currently applied in the literature. To avoid inaccurate description of fine-scale spatial structures it is recommended that researchers increase plot replication. Furthermore, studies of spatial structure and population dynamics should account for spatial environmental gradients, whatever plot size is used. The second part of the thesis presents a rigorous investigation, incorporating a priori inference and the application of fine-scale spatial statistical and modelling techniques, of the biotic and abiotic mechanisms underlying the spatial structure and population dynamics of S. supranubius, a leguminous shrub species endemic to the Canary Islands. The spatial structure of S. supranubius populations is consistent with the operation of clonal reproduction and intra-specific competition. However, the results indicate that spatial environmental heterogeneity (from small to broad scales), in particular topography, can interact with biotic processes to generate quantitatively different S. Supranubius patterns in different locations. Future research into the spatial and temporal dynamics of interactions between abiotic and biotic processes is recommended.

Knowledge about the distribution of species is limited, with extensive gaps in our knowledge, particularly in tropical areas and in arid environments. Species distribution models offer a potentially very powerful tool for filling these gaps in our knowledge. They relate a set of recorded occurrences of a species to environmental variables thought to be important in determining the distributions of species, in order to predict where species will be found throughout an area of interest. In this thesis, I explore the development, potential applications and possible limitations of distribution models using species from various taxonomic groups in two regions of the world: butterflies, mammals, reptiles and amphibians in Egypt, and butterflies, hoverflies and birds in Great Britain. Specifically I test: 1) which modelling methods produce the best models; 2) which variables correlate best with the distributions of species, and in particular whether interactions among species can explain observed distributions; 3) whether the distributions of some species correlate better with environmental variables than others and whether this variation can be explained by ecological characteristics of the species; 4) whether the same environmental variables that explain species’ occurrence can also explain species richness, and whether distribution models can be combined to produce an accurate model of species richness; 5) whether the apparent accuracy of distribution models is supported by ground-truthing; and 6) whether the models can predict the impact of climate change on the distribution of species. Overall the use of distribution models is supported; my models for species in both Egypt and Britain explained observed occurrence very well. My results shed some light on factors that may be important in determining the distributions of species, particularly on the importance of interactions among species. As they currently stand, distribution models appear unable to predict accurately the impacts of climate change.

Mathematical models are extensively used to support decision-making in many disciplines. Nevertheless there are not clear standard guidelines to assess models performance. This significantly affects model selection processes, which aim to determine the "best model", among several possible candidates. Model performance is often measured by the accuracy with which models predictions fit independent observations. However this test assesses only a single aspect of a model. A model selection process should establish the similarities between the constructed and the conceptual model. Therefore it should be based on a comprehensive assessment of the models capabilities, which is the objective of the multi-aspect comparison approach proposed in this work. The innovative aspect of this approach is to create a relationship among four conventional tests, i.e. uncertainty and sensitivity analysis, goodness-of-fit prediction-observations, model complexity and level of details, in order to provide a reliable estimation of the differences between the constructed and conceptual models. Although, model complexity is quantified using a standard approach, a novel methodology is proposed in this thesis, intended to be an intuitive and illustrative approach in creating a linkage between model complexity and level of detail. Five radioecological models have been considered: SAVE rural model, TEMAS rural model, SAVE semi-natural model, FORM and RIFE1. The results show that there is a limited resemblance between these models and the respective conceptual models. This is due to low prediction accuracy (RIFE1 and FORM); high level of uncertainty (SAVE rural); sensitivity to parameters which is not consistent with the current understanding of radiocaesium behaviour in the environment (TEMAS and SAVE rural). The SAVE rural model has been revisited in order to increase the similarity between the constructed and conceptual model. The resulting model prediction shows lower degree of uncertainty and there is a significant agreement between the model sensitivity results and the general understanding of the processes affecting Cs soil-to-plant transfer. Nonetheless the revised model does not show higher prediction accuracy than the original model. It is concluded that a reliable methodology for model selection should be based on a comprehensive investigation of each considered model aspect and that there is not a single best approach. The methodology proposed in this work has been successful in the case of the five radioecological models studied.

Lake Ohrid is a large, deep, ancient lake located on the Balkan Peninsula and has an extraordinary degree of endemic biodiversity. A deep drilling campaign was carried out in 2013 as part of the Scientific Collaboration on Past Speciation Conditions in Lake Ohrid (SCOPSCO) project to better understand the influence of climate and environmental change on evolutionary patterns and endemism. Initial results from a 569 m sediment succession drilled in the centre of the basin indicate continuous lacustrine conditions over the past ca. 1.2 Ma. This thesis presents new stable isotope data from carbonate within the upper 248 m of the composite profile, covering the last ca. 640 ka at a millennial-scale resolution (≈ 0.5 ka). Isotope data are used to provide a long-term palaeoenvironmental reconstruction, which is supported by a quantitative model of modern lake water isotope composition and a detailed multi-proxy investigation of climate evolution since the last glacial period. Water balance calculations confirm Ohrid to be an evaporative system with a complex hydrology. Variations in the isotope composition of modern lake water are suggested to represent long-term fluctuations in water balance. A trend from wetter to drier conditions through the Holocene is consistent with regional and hemispheric processes related to changes in insolation. Over the last ca. 640 ka, endogenic calcite is precipitated in abundance during warm stages, however carbonate is negligible during glacial periods with the exception of discrete horizons comprising early diagenetic authigenic siderite. The oxygen isotope composition of lake water, calculated using calcite and siderite, indicates more evaporated conditions during warm stages and fresher lake water in glacial phases. Low frequency variability shown by calcite isotope data suggests relatively stable conditions before ca. 450 ka, a transition to a wetter climate between ca. 400-250 ka, and a trend to drier climate conditions after ca. 250 ka. Higher frequency millennial-scale oscillations and abrupt climate events observed during warm stages are likely associated with regional climate change as a function of orbital forcing. This study emphasises the potential of Lake Ohrid as a valuable archive of climate change in the central Mediterranean region and demonstrates the efficacy of isotope data for palaeoenvironmental reconstruction at Lake Ohrid.

The hoverflies (Diptera:Syrphidae) represent an apparently paradoxical visual Batesian mimicry complex, with what appear to be "poor" Mimics outnumbering their more accomplished counterparts. The purpose of this thesis is to determine how far conventional mimicry theory is capable of explaining the apparent paradoxes of mimicry in the hoverflies. It becomes obvious that determining the mimetic status of the supposedly poor Mimics is not a trivial task. Conventional experimental tests of mimicry, using captive predators, seem incapable of predicting the degree of protection enjoyed by a Mimic in the field. The research therefore concentrates on developing some novel empirical approaches to the study of mimicry. This includes developing a method of image analysis which yields an objective, single-value measure of the similarity between Model and Mimic patterns. This index of similarity is used to produce unique descriptions of the structure of mimetic communities in terms of Mimic frequency and similarity to the supposed Model. These profiles indicate that there is an objective basis to the perceived paradox, and suggest that there is not a simple relationship between the actual and perceived similarity of two patterns. The perceived similarity of Model and Mimic will be a key determinant of mimetic success. The index of similarity is also used as a basis for direct comparison of the supposedly mimetic hoverflies with a more established example of mimicry in the butterflies. This exercise demonstrates that an index of pattern similarity enables a unique comparative analysis of mimicry. It is proposed that an index of similarity also provides a unique opportunity to test our theoretical understanding of mimicry, if it is used in conjunction with a mathematical model that possesses some specific attributes. A suitable prototype model is developed and demonstrated. The thesis concludes with an indication that the novel empirical approaches developed here, have been adopted elsewhere. This latter work indicates that those hoverfly species which are apparently "poor" Mimics, may be exploiting some constraint in predator perceptual and cognitive systems to achieve mimetic protection, despite a relatively low degree of actual similarity to the Model species.

The chemical composition of snow, terricolous lichens and top-soil along with abundance and diversity of lichen communities were assessed in the Pechora and Usa basins, North-Eastern European Russia. Transects were established through the principal industrial towns of Vorkuta, Inta and Usinsk to assess the spatial extent of acid or alkaline and metal deposition. A further eight sites were selected to assess local impacts of oil and gas operations. In the Usa basin decreases of nitrogen concentration in the lichen Cladonia stellaris and winter deposition of non-sea salt sulphate moving northward were attributed to long range transport of oxides of nitrogen and sulphur from lower latitudes. Increased ionic content and pH of snow, along with elevated nitrogen concentrations and modified cation ratios in lichens (Cladonia arbuscula and Flavocetraria cucullata) within 25-40 km of Vorkuta and Inta were attributed to local deposition of alkaline coal ash. Nitrate concentration in snow did not vary with proximity to perceived pollution sources. Trace metal composition of winter snowpack, snow-melt filter residues and top-soils indicated elevated concentrations of elements associated with alkaline combustion ash around coal mining operations in Vorkuta and Inta, adding significantly to the soil metal loading as a result of ash fallout. Around the petrochemical industry near Usinsk there was little evidence of trace metal deposition. Acid deposition was associated with pristine areas, whereas alkaline combustion ash near to emission sources more than compensated for the acidity due to S02 and NOx. There were limited perturbations in the chemical signals in lichens, top-soils and lichen diversity close to an oil and gas industrial complex on the Kolva river. Here, there were elevations of lead and nitrogen concentrations in lichen apices and in the apical : basal nitrogen ratio in Flavocetraria cucullata, with lower lichen diversity of epigeal and epiphytic lichens. Elevated concentrations of Ba and Ca were found in soil-ash, probably as a result of local emissions from construction activity and gas flaring, rather than from long-range transport. Virtually all other sites remained unmodified and reflected background concentrations.The ecological impacts of the measured pollution loads are discussed.

Hoverflies (Diptera: Syrphidae) vary widely in their mimetic associations, comprising wasp-mimetic, bee-mimetic and non-mimetic species. Social wasp mimics are dominated by 'imperfect mimics' which outnumber their supposed models (Hymenoptera: Vespidae) by large factors. The purpose of this thesis is to determine to what degree Batesian mimicry can account for these paradoxes, and to test alternative hypotheses for the evolution of the yellow-and-black patterns. There is little evidence of an effect of wasp abundance on 'imperfect mimic' abundance across 23 years of trapping data, as predicted if mimics are protected from predators through their resemblance to wasps. The seasonal asynchrony and high abundance of 'imperfect mimics' relative to their models is also notable, as well as the possible significance of wasp predation on hoverflies. Predictions concerning the function of the colour patterns of 'imperfect mimics' are tested using the association between similarity to the model and flight agility (indirectly measured assuming a trade-off between reproductive potential and flight agility). There is no strong indication that mimetic protection is the primary function of the colour patterns, but the evidence concurs with an aposematic function, signalling to predators the unprofitability of attempting capture. These conclusions are tentatively supported by direct measures of flight agility, though the small differences among species are difficult to pick up. The data on reproductive morphology of hoverflies show considerable variation across species, especially in males. The existence of giant testes in some species suggests that methods of dealing with sperm competition in hoverflies are diverse and deserve further study. The high ratio of 'imperfect mimics' to both models and good wasp mimics is also partly explained by habitat disturbance; undisturbed habitats show significantly less 'imperfect mimics' as a proportion of the hoverfly population. Current relative abundance in the UK may therefore be very different to when the colour patterns evolved.

Lowland tropical peatlands can act as sinks and sources of carbon, interchanging greenhouse gases (GHG) with the atmosphere. Despite the importance of lowland tropical peatlands in the past, present and future global carbon cycle, uncertainties exists about the controls regulating the processes of carbon turnover. Therefore, this study examined different controls of carbon turnover, including abiotic, biotic and anthropogenic. For this purpose, six peatlands with different dominant vegetation were selected in the north western region of the Republic of Panama (9° 4' 16.06" N; 82° 6' 28.98" W). Two phasic communities were used as experimental models; Raphia taedigera palm swamps and mixed forest swamps with Campnosperma panamensis. A combination of in situ and ex situ experiments were performed between March 2010 and November 2012: i) ex situ respirometric assays were used to quantify differences in carbon turnover through the peat profile under different redox regimes, ii) litter bags experiments were used to investigate the effect of plant materials with distinct botanical origins on peat accumulation, iii) the effect of vegetation on greenhouse gases emissions was assessed with in situ and ex situ experiments and iv) land use change (LUC) was monitored to evaluate its consequences on the short term carbon turnover processes. Ex situ respirometric assays suggested that organic matter composition of peat plays a major role in controlling the potential CO2 and CH4 production. Under anaerobic conditions, the potential CO2 and CH4 production decreased with depth. The potential CO2 and CH4 productions in the surface peat layers of the anaerobic assays were 7 and 120 fold higher than those in deeper layers of the peat profile respectively. The change in redox regime affected the carbon turnover; the CO2 potential production in the surface layers (< 50 cm depth) increased 20 fold when exposed to aerobic conditions, whilst the deeper layers (> 50 cm depth) increased 47 fold. In contrast, CH4 production was reduced two orders of magnitude under aerobic conditions. Tissue types of R. taedigera and C. panamensis showed different in situ decomposition rates. Decomposition was significantly slower belowground than at the surface, reflecting the importance of the redox regime on the litter decomposition. Roots presented the lowest in situ decomposition rates among tissues both at the surface (R. taedigera: 0.59 ± 0.04 y-1; C. panamensis: 0.45 ± 0.01 y-1) and belowground (R. taedigera: 0.13 ± 0.01 y-1; C. panamensis: 0.17 ± 0.005 y-1). Macromolecular analyses revealed that roots and stems have similar composition to the peat material accumulated in deeper layers. Vegetation exerted a direct control on GHG fluxes from lowland tropical peatlands. In both ex situ and in situ measurements, fluxes of CO2 and CH4 varied with vegetation activity. In terms of CO2eq (Addition of mass flow of GHG, converted with the global warming potential of each gas), the agricultural LUC increased CO2eq emissions from the R. taedigera swamp at Cricamola by ca. 20 t CO2eq ha-1 y-1. At the pristine site, CO2, CH4 and N2O contributed with ca. 90, 9 and 1 % of the TCO2eq respectively. In contrast, in the anthropogenically impacted plot, CO2, CH4 and N2O contributed with ca. 29, 69 and 2 % of the TCO2eq respectively. Water table strongly influenced the carbon turnover. Under flooded conditions (water table at or above the surface; 0 to 0.15 m), the CH4 emissions were ca. 4 times higher in comparison with those where the water table was below the surface (−0.01 to −0.4 m). In contrast, CO2 emissions were ca. 1.5 times higher when the water table was below the surface. It was concluded that the interdependence of hydrology, peat composition and vegetation activity are the main factors controlling carbon turnover in the lowland peatlands of the north western region of Panama. This thesis has shown that fine scale alterations of these three factors can have large scale consequences, demonstrating sensitivity to perturbations and ease shift of lowland tropical peatlands from carbon sinks to carbon sources.

Nutrient dynamics of two sub-types of peat swamp forest, mixed swamp forest and low pole forest, in the upper catchment of the Sebangau River in Central Kalimantan, Indonesia were studied. Three permanent study plots, 50 x 50 m, were established in each forest sub-type to facilitate collection of throughfall, stemflow, litterfall, decomposition, above ground and below ground biomass, peat and water samples. Graphical presentation, Wilm's method, and analysis of variance were carried out for both sub-types of forest in order to analyse data to detect any significant differences. Rainfall is slightly acid (pH 5.96+0.35) with a predominance of NH4-N, Ca and K. Throughfall and stemflow are enriched in most elements analysed compared to rainfall and the pH values are lower. Throughfall pH is 4.76±0.33 in mixed swamp forest and 4.37±0.33 in low pole forest. Stemflow pH is 4.03±0.19 in mixed swamp forest and 3.57±0.11 in low pole forest. Greater litter production was obtained in mixed swamp forest (8,411 kg ha-1 yr-1) than in low pole forest (6,534 kg ha-1 yr-1). Dry weight of the different fractions of litterfall (leaves, branches, reproductive parts and other debris) for MSF and LPF were 6216,1246, 460 and 489 kg ha-1 and 4864,1251,169 and 251 kg ha-1, respectively. Decomposition rates (k) in the MSF and LPF are 0.396 yr-1and 0.285 yr-1 respectively. Above ground biomass in MSF and LPF are 313,899 and 252,547 kg ha-1 respectively, while below ground (root biomass) is 26,533 and 14,382 kg ha respectively. Nitrogen is the predominant nutrient in peat soil at 50 cm depth in both MSF and LPF, while manganese is the lowest. Calcium is the element in greatest amount in water run off in MSF and LPF at 8,15 and 7.15 kg ha-1 yr-1 respectively, while manganese was the lowest at 0.01 and 0.02 kg ha-1 yr-1, respectively. Nutrient inputs were higher than nutrient losses during the 1-year study period with the greatest nutrient gain for calcium while manganese was the lowest in both sub-types of forest. Moreover, the results of this study highlight that nutrient concentrations in peat soils are low and the substrates are acidic. These factors are likely to be strongly limiting to agricultural development, including plantations of estate crops and trees. Under such conditions the maintenance of intact forest for natural ecosystem services (e. g. carbon storage, watershed, biodiversity maintenance, timber production in certain time period) is likely to be a far wiser land use from a long-term perspective.