Can distribution modeling inform rare and endangered species monitoring in Mediterranean islands?

Abstract

Phlomis cypria ssp. occidentalis is one of the Red Data Book species of the island of Cyprus with restricted distribution, whose conservation status must be periodically assessed under Article 17 of the Habitats Directive. We used the known species occurrence records (120 geo-referenced points) and six environmental variables (bioclimatic and biophysical) within the maximum entropy distribution modeling, to predict the potential suitable habitat for the species. We constructed three models, one using three biophysical variables, the second using three bioclimatic variables and a third model where we pulled all six variables together. We also calculated the species extent of occurrence (EOO) using a Minimum Convex Polygon and its area of occupancy (AOO) using Range Tool. We compared the results between the three Species distribution models (SDMs) with those obtained from EOO and AOO mapping. Out of the three SDMs the smallest predicted area was by Model III (all variables) while the largest by Model II (bioclimatic variables only) which also predicts the largest area of actual AOO (66%). Although Model III predicts only 28% compared to the actual AOO area, 84% of its prediction is spatially located within the AOO. The methods employed herein, when used in combination, have a twofold importance for Mediterranean islands which have a high percentage of rare and endemics species since a) they may significantly improve our knowledge on distributional patterns and b) they can guide future monitoring activities. In addition, and since they require moderate computer literacy, they can be readily employed by conservation agencies.