Habitat potential modeling of Astragalus gossypinus using ecological niche factor analysis and logistic regression (Case study: summer rangelands of Baladeh, Nour)

Subject Areas : Geospatial systems development

Farhad Borna ¹ , Reza Tamartash ² , Mohammadreza Tatian ³ , Vahid Gholami ⁴

1 - MSc. Student of Rangeland Management, Sari Agricultural Sciences and Natural Resources University
2 - Assis. Prof. College of Natural Resources, Sari Agricultural Sciences and Natural Resources University
3 - Assis. Prof. College of Natural Resources, Sari Agricultural Sciences and Natural Resources University
4 - Assis. Prof. College of Natural Resources, University of Guilan

Received: 2016-02-06 Accepted : 2016-10-06 Published : 2017-02-19

Keywords: Summer rangelands, Baladeh- Nour, Logistic regression (LR), Ecological niche factor analysis (ENFA), Geographic Information System (GIS), Astragalus gossypinus,

Abstract :

This study has been done with the purpose of modeling and prediction of the habitat Astragalus gossypinus map using Logistic regression (LR) and Ecological niche factor analysis (ENFA)  in summer rangelands of Baladeh, Nour in Mazandaran province. To achieve this objective, environmental map variables were prepared with the help of ArcGIS®9.3 techniques in cell size of 10 × 10. Also, 80 site as well as the presence or absence of species was recorded by sampling classified-random. For each sampling site was recorded information about the presence or absence of species and environmental variables, and the relationship between species distribution and environmental factors was determined by using logistic regression and ecological niche factor analysis, and forecast maps the distribution of the Astragalus gossypinus was produced in the study area. According to LR results, Elevation, pH, organic carbon, average temperature of the wet season and average temperature during the dry season were the most important environmental factors influencing the distribution of the species. According to this model, variable aspect, sand Percent, TNV of soil, precipitation in the wet season and average temperature during the coldest season were used as influential environmental variables. Evaluate the correctness statistical models were performed by using the kappa coefficient and ROC area under the curve plots. Value indices, respectively 0.42 and 0.78 for the logistic regression model and 0.84 and 0.92 for the ecological niche factor analysis, which represents that profile model shows higher accuracy than the discrimination group models in the study area.