slider slider slider slider slider slider
Integration of categorical- and gradient-based approaches in landscape fragmentation and connectivity modelling using GIS&T

Landscape structure must be identified and quantified in a meaningful way before interactions between landscape patterns and ecological processes can be properly understood. Therefore it is crucial to have appropriate representation of a landscape structure and methodology allowing to link landscape spatial pattern and ecological processes (e.g. landscape/habitat fragmentation and/or connectivity). Most quantitative methods which are used in modern landscape ecology to quantify landscape spatial pattern and to assess fragmentation and connectivity, are based on categorical (discrete) representation of landscape heterogeneity (mainly patch-based models). However, categorization (which is often subjective) obscures all internal heterogeneity of discrete landscape structures what could lead to losses of important ecological information. Therefore, currently an emerging issue in landscape and ecological studies is an alternative representation of landscape structure - landscape gradient models. They are assumed to accurately represent continuous spatial heterogeneity, in this way helping for better understanding of pattern-process relationships. The integration of these two different approaches into one coherent model could even more increase accuracy of landscape spatial pattern representation. Appropriate methods are needed for the implementation of such approaches, as well as demonstration of their utility in landscape and ecological studies.

In our project we propose to create a unique set of methods allowing for more accurate representation and description of landscape structure. This issue is extremely important in the ecological and biodiversity studies which attempt to link landscape spatial patterns with ecological processes. The uncertainty of landscape structure representation affects an accuracy of past and future analyses and models in ecological studies, e.g. habitat distribution models. Our new methodology will merge discrete and continuous approaches to landscape structure representation. In particular we will test and adapt methods from mathematics and physics (e.g. surface metrology, texture analysis, lacunarity analysis or graph theory). In this process the significance of Geographic Information Science and Technology (GISc&T) will be emphasized.

Knowledge and tools allowing for spatial analysis and modelling of landscape structure could help e.g. planners and ecologists to identify new directions in environment conservation plans that include identification of biodiversity hotspots and prevention of biodiversity loss, location of reintroduction sites for endangered species or delimitation of ecological corridors.

Project supported by a grant from the National Science Centre


National Science Centre