Use this URL to cite or link to this record in EThOS:
Title: Forest conservation, expansion, restoration and management in a National Park : modelling ecology, suitability, biodiversity priorities, temporal and climate change using GIS and spatial data
Author: Gkaraveli, Anthi G.
ISNI:       0000 0001 3500 6662
Awarding Body: University of Wales, Bangor
Current Institution: Bangor University
Date of Award: 2002
Availability of Full Text:
Access from EThOS:
Access from Institution:
When maintaining the biodiversity and ecological integrity of forests is a goal of management, a primary requirement is to assess the status, condition, conservation value of each forest, and change in forest conditions over time. GIS procedures were used here to compare different map-based surveys and look in detail at changes in woodlands of the Snowdonia National Park, North Wales, from 1970 to 2000. The maps from the 1970s and 1980s generated by Silsoe College did not compare well with the map from the 1990s produced by the Forestry Commission and no meaningful changes could be measured. This was attributed to difficulties with comparison of different interpretations and classification schemes used by the two organisations. A series of landscape surveys using the same minimum mapping unit, classification scheme, and methodology in general is needed. The potential changes in broadleaved and scrub woodland area were modelled based on the two most extreme climate change scenarios, termed the Low and High scenarios. Temperature and rainfall models formed the basis for logistic regressions of woodland type and distribution. A declining trend in probability of presence for both woodland types from the present sites was shown under the UKCIP98 High climate change scenario. The results emphasized the conceptual difficulties in using fragments of woodland within the realised niche rather than the fundamental niche as the basis for environmental modelling of plant community distributions. GIS based models were generated to address the key question in the biodiversity action plan process of where should new woodland be created or plantations restored. Ecological criteria were developed to identify the priority areas for native woodland expansion taking into account of the requirements for successful woodland expansion from the nature point of view and specific policy aims. The results were interesting and suggested that there is ample land potentially suitable in Snowdonia for new native woodland. The models could be used to aid decision-making for new native woodland in the National Park. A further extension of GIS-based modelling was developed for the prediction of individual NVC types and BAP priority woodland types. The environmental spaces occupied by the fragments of NVC woodland types currently present in Snowdonia were defined and used as templates to produce maps of potentially suitable sites for the occurrence of each NVC type. The results were not as clear-cut as had been hoped because of overlaps in the predicted occurrences of various woodland types. Independent verification of the predictions using non-spatial data for 24 sites revealed that the model produced was very poor. This was not, however, a fault of the modelling but a reflection of the fact that some of the environmental data were at too coarse a scale and that NVC types are not solely determined by environmental factors. In spite of some weaknesses in the data, the use of GIS for modelling these scenarios proved useful. Nowadays, forest policies in Wales, Europe and elsewhere are changing rapidly to meet modified global, national, and local objectives. GIS is, and will increasingly be so, proving to be a useful and flexible tool for translating forest policy into practical application on the ground.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: Forest habitat networks