Unit:
Κατεύθυνση ΒιοπληροφορικήLibrary of the School of Science
Author:
Somponis Georgios
Supervisors info:
Αναστάσιος Λεγάκις, Αναπληρωτής Καθηγητής, Τμήμα Βιολογίας, Ε.Κ.Π.Α.
Άρτεμις Νικολαΐδου, Καθηγήτρια, Τμήμα Βιολογίας, Ε.Κ.Π.Α.
Αριστείδης Παρμακέλης, Επίκουρος Καθηγητής, Τμήμα Βιολογίας, Ε.Κ.Π.Α.
Original Title:
Εκτίμηση του οικολογικού θώκου των ελαφιών Cervus elaphus στην Πάρνηθα με τη χρήση νευρωνικών δικτύων
Translated title:
Estimation of the ecological niche of Cervus elaphus on Mt. Parnitha using neural networks
Summary:
The National Park of Parnitha is an area with rich flora and fauna which is a habitat for many species of mammals and birds. The most important species is the red deer (Cervus elaphus). The red deer in the past were distributed across most of mainland Greece. However, in a few decades the populations shrunk to such an extent that the species characterized as endangered. The population of Mt. Parnitha is currently the most thriving in Greece and it is proven that reserves the genes of native deer. The destruction of a significant part of its habitat from the fire in 2007 and the increase in population due to the previously lack of natural predators may have negative effects on Mt. Parnitha’s forest because of the pressure from the grazing deer. Therefore, a movement of deer to new habitats it is indicated for reducing pressure on the forest and the distribution of the species in Greece.
In order to increase the chances of survival of red deer in a new area the habitat suitability should be estimated first. For this reason, the biotic and abiotic parameters of the areas that the deer is present or absent in the National Park of Parnitha were recorded. Afterwards, a SOM neural network was implemented and trained with these data and assessed its effectiveness on the proper evaluation of unknown areas.
The results indicate that the performance of SOM neural network is more than satisfactory and the application developed based on this can estimate directly and quickly the suitability of an area for hosting the red deer of Mt. Parnitha. However, the dataset anoints improvement so as to include more parameters and cover larger part of the National Park of Parnitha.
The distribution of the red deer in areas where it has disappeared it will assist in its conservation, with what advantages deriving from it, and in reducing the pressure it exerts on Mt. Parnitha’s forest. Nevertheless, the reintroduction of red deer in suitable habitats does not ensure its survival if the poaching is not eliminated first. For this reason, the environmental awareness is essential for the survival of new subpopulations.
Main subject category:
Science
Keywords:
Red deer, Cervus elaphus, Mt. Parnitha, habitat suitability, SOM neural networks
