Introduction and Objective: Due to the extensive climate changes and the increase in the frequency of drought stress, the Hyrcanian forests are facing serious threats that challenge the growth and dynamics of native species of these unique ecosystems. Caucasian alder (Alnus subcordata C.A.Mey.), as one of the key species in afforestation and rehabilitation of degraded lands, is particularly affected by these climate changes due to its environmental characteristics and strong need for water resources. Since this species plays an important role in nitrogen fixation and maintaining the stability of Hyrcanian forests, it is necessary to study the effect of drought stress on its presence and distribution, both in terms of biodiversity protection and forestry management. According to the literature review, no research has been done regarding the effect of drought stress on the distribution of alder species, so the purpose of this research was to investigate the effect of drought stress on the distribution of alder species in the entire area of Hyrcanian forests.
Material and Methods: This research was carried out in the entire area covered by Hyrcanian forests. For this purpose, the data bank of forest inventory in the north of the country was used. Meteorological data, including annual precipitation and mean annual temperature, were obtained from the POWER project. In order to calculate the drought stress, the medfate software package was used in R software. In this research, environmental variables including elevation, slope, radiation, annual rainfall, mean annual temperature, distance from the valley, topographic location index, topographic curvature, topographic humidity index and drought stress were prepared for each sample plot, and were used for modeling the suitable habitat of Caucasian alder. Caucasian alder distribution in Hyrcanian forest was predicted using three models including random forest (RF), support vector machine (SVM) and generalized linear model (GLM). Modeling algorithms and relative importance of variables were implemented and calculated in the caret package in R software. The cross-validation technique was used to evaluate the performance of the models. Then, the area under the curve (AUC) statistic was used to evaluate the performance of each model.
Results: The value of the area AUC for the RF model (0.73) was higher than other models. Elevation and drought stress were respectively the most important factors for predicting the presence of Caucasian alder using the RF model. Also, based on the results of SVM and GLM, annual rainfall was the most important factors in predicting the presence of Caucasian alder. Also, the results of investigating the respond of Caucasian alder to the elevation and radiation using the RF model showed that the probability of Caucasian alder presence increases with the increase of both variables. While the probability of Caucasian the presence of alder decreases with the increase of drought stress. The results of interpolation maps showed that Caucasian alder has a higher probability of presence in the western and central parts of the Hyrcanian forest, while the probability of its presence in the eastern parts of the Hyrcanian forest is very low.
Conclusion: The results of the present study showed that drought stress is one of the most important factors limiting the distribution of Caucasian alder in Hyrcanian forests. The decrease in rainfall and increase in temperature due to climate change will increase the drought stress in the region and will directly limit the growth and regeneration of alder. In general, it is suggested to avoid afforestation with Caucasian alder in areas that are vulnerable due to drought stress, and instead use species more resistant to low water conditions in order to preserve the structure and function of forest ecosystems in the best way.