A GIS FOR THE SIEG

Françoise Gourmelon (1) , Frédéric Bioret (1), Judd Howell (2)

(1) LETG UMR 6554 CNRS - Institut Universitaire Européen de la Mer (UBO), Technopôle Brest-Iroise, Plouzané (FR)
(2) USGS-Patuxent Wildlife Research Center, Laurel MD (US)

Introduction
In 1999, the Small Island Ecosystem Group (SIEG) was created after the international workshop " Long term monitoring of ecological processes of protected small islands " organised in 1999 by the "marine islets and submarine environment" working group of Réserves Naturelles de France, USGS and the French CNRS, to identify a network of small islands managed as nature reserves as potential sites for long term ecological monitoring methodologies and for understanding ecological processes . The goal was to develop a common protocol for interdisciplinary monitoring of marine and terrestrial environments of islets as nature sanctuaries and integral reserves for applications to research, management and education. The islets in concern are located within protected areas of Mediterranean, northern Atlantic and Pacific oceans. They are nature reserves or nature sanctuaries where the human activities are absent or linked with touristic frequentation. Long term monitoring programmes with common issues and methodology would provide a contribution to research on the effects of Global climate change. This group encompasses European and US scientists and managers. During the 1999 workshop, several recommendations were produced for ecological processes research, for long term ecological monitoring and for testing methodologies for adaptive management:

In this pluridisciplinary context, the use of GIS is recommended for their capability to provide an efficient methodology for storing, retrieving, analysis, modelling and visualising the vast amount of spatial data collected on such a complex network by different providers (Aspinall, 1995).

Case-studies
One of the french contributions to this international project concerns the terrestrial component of protected islets. In 1992, the Marine Islets Working Group of Reserves Naturelles de France has identified common ecological processes on the islets of the network: the degeneration of the vegetation cover probably due to fauna pressure such as rabbits and seagulls. According to scientific and management purposes, the terrestrial vegetation mapping of many atlantic islets (nature reserves of Iroise and Sept-Iles, in Brittany) and mediterranean islets (Iles Lavezzi and Iles Cerbicale nature reserves, in Corsica) has been performed according to a common methodology (figure 1). Based upon a phytosociological approach, the data are collected with local aerial photographs and updated every six years according to the management plan of each reserve (Bioret & Gourmelon, 1999).

Due to the small size of the islets (<63 ha), to the precision of dynamic processes and to the frequence of the monitoring, the scale of analysis is 1 : 5 000, and the phytosociological nomenclature is used to characterize the vegetation units (Bioret et al., 1995). Integrating the vegetation serie concept, this method describes vegetation cover changes taking into account environmental abiotic and biotic factors, as well as vegetation types associated to several stages of their progressive or regressive dynamics. Based on the sigmatist method, the phytosociological survey of the islets was lead in 1990 and 1995. The maps based on these surveys were realized by systematic field prospections and the report of each vegetation unit identified on aerial photographs enlarged at the scale of investigation. For further field surveys, only vegetation units and dynamic stages have been identified by photo-interpretation of aerial photographs, completed in 2000 by ULM photographs. Following photo-interpretation and field validation, the interpreted documents were transferred into a vector based database (Arc Info GIS package). Several thematic layers were obtained and used to produce maps and descriptive statistics (figure 2).

In order to study vegetation cover dynamics between two successive surveys, we used a polygon overlay process which leads to the production of two union covers. They are analyzed according to decision criteria. A spatial object with the same value for dynamic stages is considered as being stable. An entity which value of the oldest vegetation cover dynamic is above the other, is considered as being on a regeneration process (by example, 1990 value > 1996 value, 1996 value > 2000 value). An entity which value of the oldest vegetation cover dynamic is below the other, is considered as being on a degeneration process (by example, 1990 value < 1996 value, 1996 value < 2000 value). By these synthetic criteria, maps of vegetation changes are produced (figure 2). They are fairly helpful for the interpretation of the dynamic processes in relation with environmental factors such as breeding gull populations dynamics, fallow encroachment of former cultivated parcels and vegetalisation of former bare zones in relation with deflation or accretion (Gourmelon et al., 2003).
An equivalent methodology based on common typology and time-scale schedule, standardized protocols for data collection, and data analysis with a GIS provides the establishment of comparative synthesis (figure 1).

Conclusion
Protected small islands are complex systems characterized by two main interacting geographical domains. Like all coastal zones, the understanding of global ecological processes is based on a great variety of geographical data collected at many spatial and temporal scales. It also implies the use of technological tools to input, store, analyze, model, display and handle this spatial information. The issues for management is to develop robust approaches based on the estimation of ecological processes, on the prediction of implications for management or science, on the design of studies to evaluate predictions and on the formulation of optimal management decisions. GIS may contribute to long term ecological monitoring methodologies and for understanding of ecological processes on small protected islands for data inventory, environmental analysis and decision making (Gibbs et al., 1999; Gourmelon, 2002).
Concerning the long term monitoring of the vegetation, the interest of the methodology corresponds to the effort of coherence within the network of atlantic and mediterranean protected islets, that will establish diachronic and synchronic comparisons and will provide a common assessment tool for the study of long term global ecological changes. This type of approach could also be useful in order to evaluate the impact of management policies. Nevertheless, the complexity of the environmental processes will probably justify to adopt a modelling approach and the use of a simulation tool to bring a better knowledge of the interactions between natural and social dynamics. Applying to an international network organization, it appears that this kind of approach may provide the first step for an adaptive management which would contribute efficiently to the understanding of management issues (Brunner & Clark, 1997; Holling, 1978; Ringolds et al., 1996; Williams et al., 2002).

References