پهنه‌بندی اقلیمی- احتمالاتی اراضی کشاورزی استان گلستان برای کشت گندم دیم (Triticum aestivum L.)

Introduction
Stochastical analysis can increase decision-making reliability, meanwhile increasing the flexibility of decisions. The critical role of climatic factors in land zoning on the one hand and the stochastical nature of climatic variables, on the other hand, causes the uncertainty of land suitability grading for plant cultivation, especially in dryland conditions. Golestan province has the third rank among the Iran provinces with about 400,000 hectares of cultivated land and production of more than one million tons of wheat, of which about 57 and 50 percent are allocated to rainfed cultivation, respectively. The present study was conducted to examine changes in various climate variables affecting wheat growth under the influence of different occurrence probabilities and zoning agricultural land in Golestan province for the rainfed wheat cultivation under these conditions.
 
Materials and Methods
 First, dates of planting and occurrence of each phenological stage of rainfed wheat were determined by daily data of precipitation and minimum and maximum temperatures for 33 weather stations in Golestan province in a 26-year typical statistical period. Then, by fitting different probability distribution functions and selecting the best distribution, the values of these variables were estimated at five occurrence probability levels of 25, 50, 75, 85, and 95 percent. Next, the average air temperature, the possibility of occurrence of adverse temperatures, precipitation and ratio of effective precipitation to potential evapotranspiration of wheat for each growth stage and total growth period were calculated for whole weather stations at different occurrence probability levels. Finally, by providing zoning maps of these variables and other features including elevation, land slope, land type, soil salinity and overlaying 22 layers for each occurrence probability level, agro-ecological zoning of agricultural lands of Golestan province for rainfed wheat cultivation at different occurrence probability levels was performed.
 
Results and Discussion
 Results showed that increasing occurrence probability level from 25 to 95 percent due to delaying suitable planting date from late autumn to early winter and its effect on the climatic-environmental conditions and the length of the plant growth stages improved the degree of land suitability grade for rainfed wheat cultivation in terms of mean air temperature in stages of flowering, maturity and total growth season, the occurrence possibility of temperatures higher than 14 and less than 9 °C in germination and flowering and maturity stages, respectively, and the amount of precipitation in the germination stage. Adversely, it resulted in degradation in the degree of land suitability in term of mean air temperature in the germination stage, the occurrence possibility of temperatures higher than 25 and 30 °C in flowering and maturity stages, respectively, and the amount of precipitation and ratio of effective precipitation to potential evapotranspiration in stages of flowering and maturity and the entire season. The results of agro-ecological zoning showed that the extent of suitable and moderate zones was between 50-56 and 44-50 percent of the agricultural land, respectively, for the occurrence probability levels of 50, 75 and 85 percent and the occurrence probability did not have a significant effect on zoning. In comparison, at the occurrence probability levels of 25 and 95 percent (equal to the early and late planting dates, respectively), about 91 and 78 percent of the province's agricultural land was located in a suitable and moderate zone, respectively. These results are accordant with the findings of Hoseini et al. (2018), who showed that the maximum extent of suitable land for cotton cultivation in Southern Khorasan province occurred in the early planting date. Considering the majority of agricultural land of Golestan province is located in the middle strip of the province, most changes of zoning maps by occurrence probability level have also occurred in this region.
 
Conclusion
A low reduction in cultivation risk (increasing the occurrence probability from 25 to 50 percent) significantly degraded the land suitability for rainfed wheat in Golestan province while more reduction in the farming risk (increasing the occurrence probability up to 85 percent) led to no significant effect on land suitability. However, if the minimum risk (95% probability) be selected as a management decision for rainfed wheat farming, the lowest land suitability for rainfed wheat cultivation will occur.