Abstract
The plants’ geographical distribution is affected by natural or
human-induced climate change. Numerous studies at both the global and
regional level currently focus on the potential changes in plant
distribution areas. Ecological niche modeling can help predict the
likely distribution of species according to environmental variables
under different climate scenarios. In this study, we predicted the
potential geographic distributions of Quercus ilex L. (holm oak),
a keystone species of the Mediterranean ecosystem, for the Last
Interglacial period (LIG: ~120 Ka), the Last Glacial
Maximum (LGM: ~22 Ka), mid-Holocene (MH:
~6 Ka), and future climate scenarios (Representative
Concentration Pathway (RCP) 4.5 and 8.5 scenarios) for years 2050-2070
obtained from CCSM4 and MIROC-ESM global climate scenarios respectively.
The models were produced with algorithms from the R-package “biomod2”
and assessed by AUC of the Receiver Operating Characteristic plot and
True Skill Statistics. Apart from BIOCLIM (SRE), all model algorithms
performed similarly and produced projections that are supported by good
evaluation scores, although Random Forest (RF) slightly outperformed all
the others. Additionally, distribution maps generated for the past
period were validated through a comparison with pollen data acquired
from the Neotoma Pollen Database. The results revealed that southern
areas of the Mediterranean Basin, particularly coastal regions, served
as long-term refugia for Q. ilex, which was supported by fossil
pollen data. Furthermore, the models suggest long term refugia role for
Anatolia and we argue that Anatolia may have served as a founding
population for the species. Future climate scenarios indicated that
Q. ilex distribution varied by region, with some areas
experiencing range contractions and others range expands. This study
provides significant insights into the vulnerability of the Q.
ilex to future climate change in the Mediterranean ecosystem and
highlights the crucial role of Anatolia in the species’ historical
distribution.