In order to investigate the possibility of quinoa producing in Garmsar, Iran, a factorial experiment conducted in randomized complete block design with three replications in 2018-2019 growing season at Garmsar Agricultural Research Station. The factors were planting date at three levels (March 6th, April 1st and April 6th) and the three genotypes of quinoa (Q26, Q29 and Titicaca). Results showed that the effect of planting date was significant for all studied traits except the harvest index. Also, all studied traits were significantly different in all genotypes. Titicaca planted on March 6th had the highest yield (2276 kg.ha-1).The grain yield and yield components decreased with the delaying the planting date. Compared to early plantings, Latest date, April 6th, led to reduction of all traits, especially grain yield (about 50%). The results of simple phenotypic correlation between the studied traits showed that grain yield per hectare had the highest correlation with plant yield (0.877) and then with leaf area index (0.832), panicle weight (0.815) and number of branches per plant (0.745) that was significant at the 1% probability level.

Although direct measurement of Silage maize yield is accurate but is difficult, time-consuming, costly despite, and not applicable in large scale. Due to these limitations, the tendency to use indirect methods has increased such as remote sensing in estimating crop yield. However, accurate estimation of crop yield through remote sensing always faces challenges such as the use of images and image parameters suitable for each region. The aim of this study was to increase the accuracy of estimating the fresh weight of Silage maize before the harvest by using Monteith Model and remote sensing. The parameters of the Monteith model were optimized such as the energy efficiency coefficient absorbed in the area, and the biomass of Silage maize was estimated using Landsat 8 satellite data and compared with the biomass measured in the field. The estimated yield with the optimized model showed a significant correlation (R2 = 0. 87) with the actual yield of the fields. The results showed that the estimation error was reduced to an acceptable level (±,10%) by optimizing the parameters with the Monteith model compared with the none-optimization of the parameters. The corn dry weight estimated by the optimized model was converted to fresh weight yield and then by using the NDVI values of satellite images, after the highest NDVI value, a linear relationship with an acceptable correlation coefficient (R2= 0. 81) for fresh weight estimation of fodder corn was presented in during the growing period in the study area. Overall, the yield of Silage maize can be predicted with acceptable accuracy by using NDVI results with one or two pre-harvest images (1-3 weeks before harvest) and using the obtained relationship.

Identifying the constraining factors of production and yield gap is very important. Therefore; this research was performed to identify the production constraining factors of local rice cultivars. All management practices from nursery preparation to harvesting stages for 100 paddy fields of local rice cultivars were recorded through field studies, in Sari, from 2015-2016. In the CPA, the actual and calculated potential yield were 4495 and 5703 kg/ha, respectively and the gap was 1221 kg/ha. The yield gap caused by number of top-dressing variables was 324 kg/ha, equal to 27% of the total yield gap. The yield gap related to previous year of legumes cultivation was 218 kg ha-1, equal to 18% of the total yield variation. Among the 10 variables entered in the CPA model, the effects of top-dress fertilizer application and its application frequency and foliar application were remarkable, which could compensate a significant part of the yield gap (444 kg/ha, 37% of total) in the farmers’ fields by managing these variables. According to boundary line analysis (BLA) finding, actual yield mean on the basis of optimal limit related to 12 variables under study was 5369 kg/ha, with 881 kg/ha yield gap . Mean relative yield and relative yield gap for 12 variables (transplanting date, seedling age, number of seedlings per hill, planting density, nitrogen and phosphorous per hectare, nitrogen before transplanting, harvesting date, lodging problem, pest problem, diseases problem and weeds problem) were 83.64 and 16.35 kg/ha, respectively. Based on the finding, it can be stated that the model precision is appropriate and can be applied for both estimation of the quantity of yield gap and determining the portion of each restricting yield variables.

The increasing demands for agricultural products and pressure on the water and land resources also the problems to generate new data specify the necessity of using suitable models to predict the performance of agricultural products. In this situation, computer models provide the possibility to investigate different management strategies. The objectives of this study were to determine the least important computer input parameters which affecting the silage maize yield using artificial neural networks in different levels of water and nitrogen applications. The experiments included four irrigation levels (0.7, 0.85, 1.0, and 1.13 of crop evapotranspiration, ETc) and three nitrogen fertilization levels (0, 150, and 200 kg N ha-1). The results of artificial neural network analysis showed that when at least three parameters of irrigation, fertilizer and growing degree days (GDD) were introduced as the input of ANN, the model could predict the performance of silage maize with high accuracy. The best validation performance of the model was at step 10 with mean square error of 0.0032. Also the results of sensitivity analysis indicate that the growing degree days with the coefficient of sensitivity of 9.96 is the most important parameter for predicting of silage maize performance and after that is the amount of irrigation with the sensitivity coefficient of 2.07. The results showed that adding the solar radiation and average relative humidity to the input parameter cause reduction in MSE and increasing the accuracy of the model in the process network training.

Increasing demand for agricultural products and lack of appropriate soil and water resources with problems of field research reveals the application of efficient models to predict crop yield. This research aimed to examines the efficiency of artificial neural networks, comparative fuzzy neural network, adaptive nerofuzzy inference system and particle swarm optimization algorithm models for estimating the wheat yield through soil and land properties. For this purpose, 80 soil profiles were drilled in wheat fields’ area in East Azarbaijan province with temperature and moisture regimes of mesic and aridic border to xeric, respectively. Soil samples were collected from each genetic horizon. The results of sensitivity analysis showed that total nitrogen, absorbable phosphorus, slope, gravel, soil reaction and organic matter are effective soil properties in wheat yields. The hybrid model of ANFIS-PSO was the best model from the viewpoint of statistical indices including R2 (0. 89) and RMSE (213. 5). Also, neuro-fuzzy method has a R2 (0. 84) and RMSE (243. 2) and artificial neural networks have a R2 (0. 81) and RMSD (274. 5), respectively. The GMER index also indicated overestimation of artificial neural network (0. 24) and nerofuzzy (0. 53) and underestimation of ANFIS-PSO model (1. 13). The results indicated that the hybrid neuro-fuzzy-swarm particles model performed better than other models that can be used a powerful tool for estimating wheat yield.

Estimates of soil erosion and sediment yield from watersheds are needed to select the best management practices for sediment yield abatement and protection of water quality. The ‘ANSWERS’ model predicts sediment yield from agricultural watersheds for individual rainfall events. Conventional values of the soil erodibility factor, K, for soils containing rock fragments may result in an overestimation of sediment concentration present in the runoff. In this study, the effect of K factor adjustment to predict a more accurate sediment yield by the ANSWERS model was investigated. The value of the K factor was adjusted for the volumetric fraction of rock fragment. This resulted in a higher level of agreement between the predicted and observed values of sediment concentration in the watershed runoff. Therefore, it is concluded that the volumetric fraction of rock fragment should be determined for watershed soils containing rock fragments and be applied for modification of published K values.

Yield prediction before harvesting is one of the tools in order to planning food production supply in future. Yield prediction was carried out for Wheat (Triticum aestivum) using different meteorological variables with agrometeorological indices in Hamedan district during 2003-04 and 2004-05. According to correlation coefficients, standard error of estimate as well as relative deviation of predicted yield from actual yield using different statistical models, the best subset of agrometeorological indices were selected including daily minimum temperature (Tmin), accumulated difference of maximum & minimum temperatures (TD), growing degree days (GDD), accumulated water vapour pressure deficit (VPD), sunshine hours (SH) & potential evapotranspiration (PET). Yield prediction was done two months in advance before harvesting time which was coincide with commencement of reproductive stage of wheat (27th of May). It revealed that in the final statistical models, 83% of wheat yield variability was accounted for variation in above agrometeorological indices.

The aim of the study is determination of the correlation between factors affecting in the wheat yield and preparing of yield mapping of wheat in north of Darab city. In order to determine the relationship between biological and grain yield with some of the important agronomic traits 60 samples in the north of Darab city was investigated. Parameters such as plant height, seed weight, harvest index, tiller number, latitude and longitude for each of the samples was measured. The results show that grain weight has highest correlation with the biological yield (0. 97**). In this study, also using the Kriging (Gaussian models, spherical, circular and exponential models) and average inverse distance (IDW) maps of the biological yield and grain weight was determined. The results of the interpolation showed that kriging method (Gaussian model) with a minimum error (RMSE=0. 98 for biologic yield and RMSE=0. 97 for grain weight) was the best model for preparation of these parameters in the study area. Also the results of biologic yield map showed that areas locating in the North West of the study area had the highest yield.

The main purpose of this study is to predict the future yield of saffron contracts using a modified quadratic model, which is a library documentary research from the aspect of data collection, and from the aspect of results, it is applied and quantitative research. The time period of the quantitative part is a 5-year period from 2019/03/20 to 2023/03/20 in the form of daily frequency of the Ministry of Jihad, Agriculture and Customs of Iran from the website of the Iran Commodity Exchange, which was collected and the modified second-order model in terms of complexity, from The type of nonlinear polynomial problems that the proposed methods are modelled by coding in Matlab software environment with normal data. Overall, the results indicate that the neural network model has a higher reliance on power compared to the adjusted quadratic model in predicting the saffron contract yield, and the calculation results show that price fluctuations, cash price, transaction volume, and liquidity are the most important in order They have the contractual yield of saffron.