Modeling the habitat distributions of endemic abies species in Turkiye under climate change scenarios using machine learning techniques

One of the most determining ecological factors on ecosystems and species is climate. Climate change, defined as one of the most important problems of our time, poses a major threat to forest ecosystems. It is estimated that the distribution, composition and functions of forest ecosystems, which are important in terms of biodiversity and carbon sequestration, will change significantly under the pressure of climate change. The aim of this study is to model the current and future potential distribution of fir taxa (Abies nordmanniana subsp. equi-trojani, Abies cilicica subsp. cilicica, Abies cilicica subsp. isaurica, Abies nordmanniana), which are endemic in Turkiye and stand out with their ecological contribution to the forest ecosystem, with species presence data and environmental variables. In this direction, species distribution was carried out according to the climate data of 1970-2000 for today and the SSP2 4.5 and SSP5 8.5 scenarios for the future, covering the time intervals of 2081-2100. In order to perform these analyses, machine learning methods, which are frequently used today in modeling species distributions, were used with the help of Geographic Information Systems software. The machine learning methods used in this thesis are Maximum Entropy (MaxEnt), Multiple Logistic Regression, Multi Layer Perceptron and Linear Mixture Model. The Receiver Operating Characteristic (ROC) curve, which expresses the estimation accuracy of the model results, was used in the accuracy analysis of the study and the MaxEnt model was determined as the most accurate method with an Area Under Curve (AUC) value of 0.999 for both species. The MaxEnt model was applied to future scenarios to determine the future status of the species and change analyses. In addition, anthropogenic effects (population density and access to road networks) that could directly affect the future planning decisions of the species were also analyzed. It was determined that the findings obtained in this study are important in terms of using them to determine strong strategies for the sustainability and protection of fir taxa in the future and contributing to the studies to be carried out to determine species distributions.