AbstractBiodiversity is declining globally primarily because of climatic change and anthropogenic impacts. In Europe, protected sites (e.g. Natura 2000) have been established to prevent further loss of biodiversity and protect key habitats from further deterioration, including from previous detrimental management practices. In the UK and particularly Wales, even protected habitats lack a coherent management approach leaving conservation bodies to work with communities to protect their biodiversity. Few systems exist for consistently, routinely and spatially quantifying biodiversity, assessing impacts of past management and guiding and predicting the consequences of future actions, including those relating to policy. Most current monitoring systems still rely solely on field observations and field surveys with little input from very high resolution (VHR; ≤ 2m) airborne and space data to provide baseline data suitable for monitoring the extent and condition of habitats over time.
In order to investigate the use of VHR optical imagery for monitoring, the potential for WV2 spectral indices data to be used as indicators of condition for protected bog sites was evaluated and a method developed which could be used within a protected site monitoring system for monitoring both flora and fauna distributions. In developing an approach this study focuses primarily on the protected lowland raised bog at Cors Fochno where encroachment by grass species (primarily Molinia caerulea and Phragmites australis) over decades, coupled with management actions focusing on conserving the integrity of the bog system, has led or is anticipated to lead to transitions in state with consequences on biodiversity, particularly species richness, distributions and abundance.
Field spectroradiometer data was used to develop a method for identifying optimum months of the year and best indices for species discrimination on Cors Fochno protected site that can later be used within a system for accurate classification of species. Methods developed, and ecological information from field data, were used with WV2 data acquired over 3 time periods (March, July and November) to produce a multiscale classification (LCCS to species level) of Cors Fochno SSSI and it’s surrounding landscape suitable for use within a monitoring system. Within the SSSI WV2 satellite derived parameters (spectral indices and classified dominant plant species) that relate to environmental variables relevant to invertebrate habitat suitability and invertebrate plant food sources were used to test associations with selected invertebrate species and assemblages assessed and map habitat suitability.
An ANOVA F-ratio method provided a successful method to assess the distinctive phenological differences between spectral data of the key lowland raised bog plant species. It demonstrated the importance of phenological data for selecting the best months and the best spectral indices for species discrimination and was used with success to select optimum WV2 imagery acquisitions for a more successful species classification. Classification of varying scales, from landscape to species level, was carried out effectively with good overall accuracies (over 81%) by using LCCS and an adapted EODHaM system classification method. This enabled the use of indices values as input using the best indices and image acquisition times for species discrimination, along with ecological information gathered in field surveys. WV2 satellite derived spectral indices, and WV2 satellite derived dominant plant species classified using the developed system which were indicative of the ecological gradients of the lowland raised bog, were used to show associations between invertebrate data (Araneae, Coleoptera and Diptera) sampled across the bog and were shown to be useful as invertebrate diversity and distribution indicators. This research demonstrates that spectral indices derived from VHR optical imagery can be used as indicators of the composition, structure and functional diversity of a lowland raised bog.
A key component was the development of a framework for sampling the SSSI. From these observations and models, a conceptual framework for creating a baseline which can be used and developed for quantifying losses and gains in biodiversity as a function of satellite-observed changes in habitat extent and condition was proposed. Through this approach, a better understanding of the drivers and impacts (both actual and potential) of change is anticipated with direct benefits for conservation, sustainable use of the environment and environmental and agricultural policy.
|Date of Award
|31 May 2016
|Knowledge Economy Skills Scholarships
|Pete Bunting (Supervisor), Peter Dennis (Supervisor) & George Petropoulos (Supervisor)