E-Locus - Institutional Repository of the University of Crete - Χωρική ανάλυση των χερσαίων σαλιγκαριών της Κρήτης

Home Χωρική ανάλυση των χερσαίων σαλιγκαριών της Κρήτης

Results - Details

[Add to Basket]

Identifier

000376023

Title

Χωρική ανάλυση των χερσαίων σαλιγκαριών της Κρήτης

Alternative Title

Spatial analysis of the terrestrial mollusks of Crete

Author

Γεωργοπούλου, Ελισάβετ Νικ.

Thesis advisor

Μυλωνάς, Μωυσής

Reviewer

Σφενδουράκης, Σ.
Γιαννουλάκη, Μ.

Abstract

The current work is a first approach of the potential spatial distribution of the Cretan lands nails. The maximum entropy (MaxEnt) analysis was used to model the potential distributions of 56 land snail species. MaxEnt is a machine learning method that predicts the potential distribution of a target group over a geographical region from a sample of presence localities and a set of environmental data. The effect of the environmental parameters that were used to model the potential distribution of the species is explored. Results show that temperature contributes to the explanation of the potential distributions of 35 species, while two other important factors seem to be rainfall and land cover. Moreover, species with restricted distribution range like Albinaria eburnea, have higher AUC values (higher than 0.8), while species with a broader range like Cantareus aspersus have lower AUC values (lower than 0.7). The use of absences increases the AUC values (53 out of 56 species have values higher than 0.75). The potential distributions of most species like Poiretia dilatata και Trochoidea sp., are quite precise. Nevertheless, for some species like Metafruticicola dictaeus, MaxEnt does not take into account the historical reasons like the paleogeography of the study area, which are likely to affect the spatial distribution of the species.

Language

Greek

Subject

AUC

Environmental variables

MaxEnt

Potential distribution

Spatial patterns

Παρουσίες ειδών

Περιβαλλοντικές μεταβλητές

Πιθανή κατανομή

Χωρικά πρότυπα