Investigating Management Factors affecting Weed Biodiversity Indices and yield of Wheat in Chenaran Township (Iran) Using CART Decision Tree

In order to study the effect of field management methods and environmental factors on wheat (Triticum aestivum L.) yield, weed control, weeds in two consecutive years were studied in 200 farms of 20 villages located in four directions of Chenaran Township, Iran. For this purpose, sampling was carried out using (w) method with a 0.25 m 2 quadrate. Weed species were identified and their number per square meter was determined. Then, the Shannon-Weiner Index and Simpson Equilibrium Index were calculated for biodiversity measurement. Quantitative and qualitative management factors were prepared in the form of farmers' questionnaire. For this purpose, all information on agronomic management including land area, farmers' history, seedbed preparation and weed control were recorded the forms of a questionnaire during the growing season. At the end of the growing season, the actual yield obtained by the farmers was recorded. The analysis using Classification and Regression Trees (CART) method showed that among different parameters, the agricultural experience, number of dual purpose herbicides (herbicides which control both grasses and broad leaf weeds), nitrogen, potassium, summer planting last two recent years before centrifugal wheat seed planting, farmer age, the ShannonWeiner and Simpson indices had significant changes. The most important management factors affecting wheat yield were splitting of fertilizer, the number of dual purpose herbicides and fertilizer, as well as rotation and educational levels. The results of this study showed that the appropriate amount of potassium and nitrogen fertilizer and selection of suitable alternatives are effective management strategies to improve wheat yield and increase biodiversity in Chenaran area.