10044

9274

كارشناسي ارشد

مرتعداري

اصفهان: دانشگاه صنعتي اصفهان، دانشكده منابع طبيعي

1393

هفده، 124ص.: مصور، جدول، نمودار

ص.ع. به فارسي و انگليسي

محمدرضا وهابي، مصطفي تركش اصفهاني

سعيد پورمنافي، آندريوتا ونسندپترسون

94/2/27

رضا جعفري، مصطفي سعيدفر

1396/09/27

كتابنامه

منابع طبيعي

مهندسي منابع طبيعي

ID9274

Determining Potential Habitat of Rheum ribes L Using Maxent and GARP Models in Isfahan Province Ali Khalighifar a khalighifar@na iut ac ir Date of Submission 01 07 2015 Department of Natural Resources Engineering Isfahan University of Technology Isfahan 84156 83111 IranDegree M Sc Language FarsiSupervisor Mohammad Reza Vahabi vahabi@cc iut ac irSupervisor Mostafa Tarkesh Esfahani m tarkesh@cc iut ac irAbstractSpecies distribution modeling is based on the relationship between the occurrence data of a givenspecies and the environmental variables Nowadays in plant ecology producing the speciesdistribution maps is being greatly paid attention to due to the expansion of statistical methods andgeographical information systems In order to model the potential habitat of Rheum ribes in IsfahanProvince two modeling methods of Maximum Entropy and Genetic Algorithm were used Thelocation of species occurrence data 74 present sites were recorded during the field surveys using GPSfrom 6 sub provinces Meymeh Fereydounshahr Fereydan Khansar Golpayegan and Chadegan Climatic data 19 variables remote sensing data 20 principal components resulted from the PCA onthe NDVI and topographic data slope aspect and altitude were entered in the modeling process asenvironmental layers Using these variables seven approaches were chosen which led to the productionof 16 scenarios using two modeling algorithms MAXENT and GARP After analyzing the results ofthe 16 scenarios and evaluating the performance of the models two scenarios of PCA Topo Data three topographic data and 20 principal components of PCA and PCA Bio Data 19 climaticvariables and 20 principal components of PCA analyses were selected as the best scenarios forpredicting the potential habitat of Rheum ribes The results showed that in PCA Topo Data scenario the most effective factors on Rheum ribes distribution were elevation and the second component of thePCA while in PCA Bio Data the most important factors were temperature seasonality Bio 4 meantemperature of warmest quarter Bio 10 and the annual precipitation Bio 12 A collection ofindependent data was used to evaluate model using area under curve of ROC plot The performances ofboth models were classified as suitable and acceptable According to AUC Index Maxentoutperformed GARP Values of AUC were 0 92 and 0 89 for Maxent and GARP respectively in PCA Topo Data scenario and 0 94 and 0 88 in PCA Bio Data scenario respectively Annualprecipitation 200 250 mm elevation 2000 3000 m and a slope between 12 to 25 percent are theoptimum and suitable conditions for occurrence of Rheum ribes Also temperature seasonality about 8 C and average temperature about 13 20 C are considered suitable for this species Key words Habitat Modeling Rheum ribes L Maximum Entropy Maxent GeneticAlgorithm GARP Isfahan Province

محمدرضا وهابي، مصطفي تركش اصفهاني

سعيد پورمنافي، آندريوتا ونسندپترسون

رضا جعفري، مصطفي سعيدفر