@article { author = {Ajili Lahiji, A. and Mohammadi Torkashvand, A. and Mehnatkesh, A. and Navidi, M.N.}, title = {Determination of the Most Important Factors Affecting Yield of Olive (Olea europaea L.) Orchards in the North of Iran}, journal = {Journal Of Horticultural Science}, volume = {33}, number = {4}, pages = {743-755}, year = {2020}, publisher = {Ferdowsi University of Mashhad}, issn = {2008-4730}, eissn = {2423-3986}, doi = {10.22067/jhorts4.v33i4.82069}, abstract = {Introduction In order to land classification for agricultural and natural resources, the most important criterion and factor is the production or yield of the lands. The best way to evaluate a method is to interact between the yield and potential of the product with the land specification involved in the production. One of the new methods in the land evaluation for different uses is the modeling or simulation of the intended use. Artificial neural networks are one of the new tools that are used today to simulate yield and determine the effective factors in the production of agricultural crops. New land evaluation methods include different modeling techniques, which these new methods, including simulations, are used in land valuation methods to test the ability of models for determining the relationship mentioned. An artificial neural network is one of the techniques that can do nonlinear analysis. It is important to recognize the most important input parameters to the predictive models of olive yield, which are also considered to be effective variables in production. Through the process of sensitivity analysis, valuable information about the sensitivity of the model to its input variables can be provided to the designer and model architect. In north of Iran, in the southern areas of Gilan Province, the most important olive gardens of the country are located in the southern part of the country. Different soil and water, topographical and climatic factors affect the yield of olive orchards. Materials and Methods Climatic data, soil and water characteristics, topographic characteristics and leaf nutrition elements as input variables and olive yield were used as output models. Twenty-eight factors were selected as the most important factors or variables affecting the yield of olive orchards. These input variables included soil properties: EC, (TNV), organic carbon percentage, available phosphorus, available potassium, clay percentage and silt percentage; irrigation water characteristics including: EC and content of irrigation water; topographic characteristics including: altitude and the slope; the concentration of nutrients in leaves included: nitrogen, phosphorus, potassium, iron and zinc; climatic factors including: sunny hours, evaporation rate, average temperature and olive yield were considered as output of artificial neural network model. Results and Discussion Using the MATLAB software by artificial neural network, the best structure of this network was obtained for the component of the yield of olive. The trained structure had 28 input nodes in 5 groups and one output node. The number of hidden nodes, 38 nodes and the most appropriate number of repetitive learning based on the test and error, 20 were determined for olive product yield. After determining the best structure of the neural network with a R-test of 80%, using Hill's sensitivity analysis, the model's response to each of the input variables was studied and the most important factors influencing the yield of olive oil were obtained. Based on the results of sensitivity analysis, the most important parameters affecting the yield of olive are content of leaf nitrogen, soil phosphorus, winter evaporation, summer evaporation, average autumn temperature, summer sunshine, leaf potassium, salinity, salinity, and slope, respectively. In general, this study showed that the amount of nitrogen and phosphorous is the most important factor for the production of olives, and then variables that are related to the amount of water in the soil, (summer evaporation, average autumn temperature, summer sunshine), were identified as important. The concentration of potassium, iron and zinc in the leaf as a nutrient element and water and soil salinity from stressors for plants and slope are important topographical factors that affected the amount of soil depth and soil water content as the most important factors in olive crop production. Conclusion In brief, the role of micro and macro nutrients and the factors affecting the maintenance of water in the soil and providing moisture for the plant. In addition to nutrients, gardeners should consider soil and water salinity, which is a stressor for plants, as well as slope, which is important topography factor that effective on soil depth and water available and nutrients for plants as the most important factors in olive production.}, keywords = {Artificial neural network,Gilan,Olive (Olea europaea L.),Modeling,sensitivity analysis}, title_fa = {تعیین مهمترین عوامل موثر بر عملکرد باغات زیتون شمال ایران}, abstract_fa = {یکی از روش­های نوین در ارزیابی اراضی نسبت به کاربری­های مختلف، مدل­سازی یا شبیه­سازی کاربری مورد نظر است. با توجه به اهمیت عوامل موثر بر عملکرد باغات زیتون کشور، این ارزیابی در شمال ایران در بیش از 80 باغ زیتون شهرستان رودبار استان گیلان که مهمترین باغات زیتون کشور واقع شده است با استفاده از مدل شبکه عصبی مصنوعی انجام شد. عملکرد محصول باغات زیتون تحت تاثیر عوامل مختلف خاکی، آبی، توپوگرافی و اقلیمی می­باشند، که در مجموع بیست و هشت عامل موثر بر عملکرد زیتون انتخاب و به عنوان متغیرهای ورودی مدل انتخاب شدند. این متغیرها عبارت بودند از خصوصیات خاک شامل EC، میزان مواد خنثی شونده (TNV)، درصد کربن آلی، فسفر قابل جذب، پتاسیم قابل جذب، درصد رس و درصد سیلت، خصوصیات آب آبیاری شامل EC و میزان آب آبیاری، خصوصیات توپوگرافی شامل ارتفاع و شیب، غلظت عناصر غذایی در برگ شامل نیتروژن، فسفر، پتاسیم، آهن و روی، عوامل اقلیمی شامل ساعات آفتابی، میزان تبخیر، میزان متوسط دما به عنوان متغیرهای ورودی و عملکرد محصول به عنوان خروجی مدل­­های شبکه عصبی مصنوعی در نظر گرفته شد. نمونه­­های خاک، آب و برگ از باغات به طور همزمان در مرداد ماه جمع آوری شده و عوامل توپوگرافی از طریق میان­یابی در محیط GIS و با استفاده از نرم­افزار  ArcGisبدست آمد و اطلاعات مربوط به عوامل اقلیمی نیز از ایستگاه‌های هواشناسی منطقه و روش میان­یابی بدست آمد. پس از تعیین بهترین ساختار شبکه عصبی با میزان R آزمون 80 درصد، به کمک آنالیز حساسیت به روش Hill، عکس­العمل مدل به هر یک از متغیرهای ورودی بررسی و مهمترین فاکتورهای تأثیرگذار بر عملکرد محصول زیتون به دست آمد. بر اساس نتایج آنالیز حساسیت، مهم­ترین پارامترهای مؤثر در عملکرد محصول زیتون، به ترتیب نیتروژن برگ، فسفر خاک، تبخیر زمستان، تبخیر تابستان، میانگین دمای پاییز، ساعات آفتابی تابستان، میزان پتاسیم برگ، شوری خاک، شوری آب و شیب می­باشند. که به طور خلاصه می­توان به الویت تامین عناصر غذایی ماکرو المنت مانند نیتروژن و فسفر و تامین رطوبت مورد نیاز با توجه به تبخیر منطقه حتی در فصل زمستان برای باغداران تاکید نمود.}, keywords_fa = {آنالیز حساسیت,زیتون,شبکه عصبی مصنوعی,گیلان,مدل سازی}, url = {https://jhs.um.ac.ir/article_37163.html}, eprint = {https://jhs.um.ac.ir/article_37163_8beff9c87cc539180263514eb697724a.pdf} }