Determination and Zoning of Suitable Planting Date of Rainfed Wheat (Triticum aestivum L.) in Golestan Province based on Different Levels of Occurrence Probability of Autumn Rainfall

Wheat has strategic importance in Iran and Golestan province is one of the major regions of wheat production in the country. Province has the third place among all provinces of the country in term of wheat cultivated lands (about 400 thousand hectares) and grain production (more than one million tons). The optimal planting date for any crop normally occur within a sowing window. It is defined as beginning and end of the planting period which guarantee the achievement of reasonable yield. In rainfed farming, farmers face a decision of whether or not to plant when sufficient rainfall accumulates to increase soil moisture content. If the average meteorological data is used to design a farming system, such as planting date, its probability will be 50 percent. Average rainfall data (50 percent probability level) are quite unreliable for cropping planning. Usually, risk levels of 60 to 80 percent are recommended depending on the sensitivity of the crop to water stress, acceptable risk by farmer and soil. Therefore, it can be stated that in rainfed cultivation, farmers choose their risk based on the selection of the planting date. Considering the importance of predicting the appropriate planting date in optimizing farm management in rainfed farming and its dependence on the occurrence of autumn rainfall on the one hand and the stochastic nature of rainfall, on the other hand, the present study was conducted to determine the sowing window of wheat in the Golestan province at different occurrence probability levels.  Initially, by reviewing the geographic distribution of meteorological stations in Golestan province and their statistical period, daily rainfall data of 58 meteorological stations in the 1991- 2016 period (26 years) were gathered. Then, suitable planting date of wheat in each station and year was determined based on the first rainfall date equal to or more than 25 mm over a period of 10 consecutive days in October. In the next step, suitable planting date for each station was calculated by the occurrence probability level 25, 50, 75, 85 and 95 percent based on statistical analysis and choose the best probability function using the software SMADA. Verification of zoning maps was done based on findings of field scale researches. Finally, in the meantime of statistical analyzing the results, zoning maps of suitable planting date of rainfed wheat were prepared using ArcGIS software by Inverse Distance Weighting (IDW).  The results showed that the suitable planting date of rainfed wheat in the southern half of the Golestan province was earlier than the northern part. The occurrence probability level had a significant effect on suitable planting date but the effect of the station was not significant. The statistical difference between all occurrence probability levels was significant, so that the occurrence probability level of 25 and 95 percent indicated the earliest and bottommost suitable planting date of rainfed wheat in Golestan province, respectively. Within different stations of the province, the suitable planting dates of rainfed wheat for 25, 50, 75, 85 and 95 percent occurrence probability were between 22 November to 1 December, 22 November to 13 December, 24 November to 3 January, 2 December to 15 January and 12 December to 20 January, respectively. Increasing the occurrence probability levels prolonged the sowing window so that the difference between suitable planting dates in different stations was 10 and 40 days for 25 and 95 probability level, respectively. According to the results, a small reduction in the risk of farming (an increase in the occurrence probability level of suitable rainfall from 25 to 50 percent) is possible with a delay in planting date for a few days, while a greater reduction in the risk (an increase in the occurrence probability level of suitable rainfall to 75, 85 and 95 percent) will be possible through a delay of the planting date for several decades. Verification and validation of the results of this study based on findings of field scale researches confirmed the accuracy of predicting the suitable planting date by autumn rainfall. The results showed that the farmers of the Golestan province with the correct understanding of the role of autumn rainfall in crop growth and yield and based on their experimental knowledge, selected the appropriate planting date of rainfed wheat with a high probability level.