جستجوی مقالات مرتبط با کلیدواژه "artemisia" در نشریات گروه "جنگلداری"

Determining the state of distribution of species and the habitats occupied by them is very important in species protection and management programs. The use of modeling to predict the distribution of species has increased in recent years. For this purpose, a wide range of modeling techniques has been developed. In this regard, the present research was conducted with the aim of evaluating the capability of logistic regression and the maximum entropy method in preparing a prediction map of the habitat expansion of Artemisia chamaemelifolia species and determining the factors affecting its distribution in Ardabil province.

 In the rangelands of Ardabil province, 449 study sites, including 102 sites of presence and 347 sites of non-presence of the species, were recorded from 2018 to 2021. Two categories of environmental factors, including bioclimatic variables (19 cases), three primary topographic indices (elevation, slope, and aspect), and five secondary topographic indices (topographic wetness index, topographic position index, topographic roughness index, stream power index, and plan curvature index), were investigated in relation to the presence of the species. Bioclimatic variables with a spatial resolution of 2.5 minutes, equivalent to 5 km², were downloaded from worldclim.org. The maps of topographical variables were produced using the maps of the digital elevation model in a geographic information system environment. Maps of all environmental factors were prepared and overlapped with 70% of the data. Environmental information was extracted for sample points. Collinearity between independent variables was checked. In the next step, the regression relationship between species presence and independent variables was extracted in SPSS software. Then, by combining the maps of factors affecting species distribution and applying the relevant regression relationship, the distribution map was predicted through the logistic regression method. To prepare the prediction map of A. chamaemelifolia habitat through the maximum entropy method in the MaxEnt software environment, the environmental layers were converted to ASCII format, and the species presence points were converted to CSV format. The maps prepared in both models were classified based on the optimal threshold of species presence into two classes: species presence and non-presence. The Kappa index was used to check the accuracy of prepared maps and compare their performance.

 The results of the modeling showed that elevation was the most effective environmental factor in the distribution of the species. The altitude range of presence of A. chamaemelifolia in Ardabil province was found to be from 2100 to 2900 meters. The Kappa coefficient obtained from the comparison of the predicted and real maps for the logistic regression model was 0.962, and for the maximum entropy model, it was 0.871, which are at a very good to excellent level. The logistic regression model identified slope percentage, precipitation of the coldest season, precipitation of the driest month, and the average range of daily temperature as influencing variables in the occurrence of the species in Ardabil province. Based on the jackknife test resulting from the implementation of maximum entropy in the MaxEnt environment, the most important variables affecting the suitability of A. chamaemelifolia habitat were seasonal precipitation and elevation.

 The results of the present study have provided key and important information about the range of tolerance of the A. chamaemelifolia species to the influencing environmental variables. Relying on the results of the current study, arrangements can be made to protect and restore habitats with current or potential distribution of A. chamaemelifolia species.