Dynamic feedbacks between landform, landscape processes and vegetation patterns : a modelling framework to predict the distribution of plant species in Lefka Ori, Crete, Greece

Research output: ThesisPhD Thesis - Research external, graduation external


The aim of the thesis has been to describe and model landforms and active landscape processes and include model feedbacks and climatic parameters into plant species distribution models. The study area has been an ecologically important mountainous area, the Lefka Ori of Crete. The starting point has been a number of previous studies underlying that climatic and geoenvironmental factors are among the most significant controlling the spatial distribution of rare and endemic species. The limitations of past studies that this study tried to address were (a) some of these factors are complex (e.g. landscape instability, mass movement), (b) they do not refer to the landscape scale (e.g. erosion-deposition, water redistribution) or (c) data are not available (i.e. climatic). To achieve these aims: (i) Object-Based Image Analysis was used to create a geological map of the area, (ii) a set of land surface parameters were created from elevation data, (iii) a semi-automated method was developed to map the landforms of potential ecological Significance, (iv) available climatic data were reviewed and satellite imagery was utilized to address the climatic component through snow cover persistence patterns, (v) a field experiment was set up for the description of active landscape processes, (vi) dynamic landscape processes were modelled for water redistribution, erosion & sedimentation, landscape (in)stability and mass movement patterns. In the course of the study, 75 vegetation plots were sampled and merged with another 80 from past studies leading to the formulation of statistical models (GLMs) for 59 species. The results show that geology, elevation and snow cover persistence patterns explain most of the variability for most species thus were more frequently included in the predictive models. This signifies the climatic and geological control over species distribution. For four species of conservation importance, model results were presented in detail leading to probabilistic distribution maps. ; EThOS - Electronic Theses Online Service ; GB ; United Kingdom
Original languageEnglish
Publication statusPublished - 2012
Externally publishedYes


  • ITC-CV

Cite this