classification

Predictive models of response of plant species to environmental factorsare important tools for obtaining information about the causes of species establishment. In this research, using multivariate analysis, while determining the ecological factors affecting the changes in plant composition, response of (Vicia persica) to changes in environmental factors was investigated. Sampling of vegetationwas done by systematic-random method during 2016-2018. Vegetation studies, including canopy percentage and species density, were done inside the plots. For this purpose, five transects with a length of 1000 meters were used with the same distance to each other. Then, on each of them, six plots (2 x 2 meters) were established with the same distances. In each of the plots, a soil sample was collected with three repetitions from the depth of 0 to 30 cm of the soil surface. in order to investigate the relationship between effective and significant environmental variables with vegetationand choosingchoose the appropriate linear and non-linear method, DCA was performed on the vegetation data and gradient length was determined. A generalized additive model was used to predict the response of plant species to changes in environmental factors. Canoco software version 4.5 was used for data analysisin this section. The results showed thatthe height range of this plant habitat is 2200-3400 above sea level. Investigating the effect of a set of environmental factors on vegetation changes in the habitat, led to the selection of 12 variables out of 19 primary variables. The response of the species to the variables of sand percentage, electrical conductivity, potassium and soil lime was monoexponential, the response of the species to the variables of soil acidity and stone percentage and gravel was decreasing, And the response of the species to the variables of percentage of organic matter, carbon, soil nitrogen and phosphorus, follows the incremental model. Species response to saturated humidity of soil and soil apparent specific gravity followed the bi-exponential model. In total, the results of this research showed that the generalized additive model provides valuable information for determining the ecological needs of the species, which can be considered invegetation management and rangeland improvement in similar areas.

Knowing the ecological characteristics of plant species and their response to environmental factors provides the necessary information for managing vegetation and improving  of rangelands. Due to the importance of the  Chaerophyllum macropodum and its role in soil protection and forage production, we conducted a study of the ecological needs and determining the ecological factors affecting in plant cover and the response of the species to changes in ecological factors, using the Canonical Correspondence Analysis method ( CCA) and Generalized Additive Models (GAM) in Mazandaran province. The results of conventional comparative analysis showed that environmental factors such as altitude, aspect, phosphorous, organic carbon, organic material, pH, Ec, %sand, saturated moisture percentage and lime  of soil, respectively with the expression of 5.9, 5.6, 2, 2, 1.8, 1.7, 1.7, 1.6 and 1.3 percent of the variance in plant composition had an important role in the changes of plant cover in the studied habitats. The response of the species to the amount of phosphorus and potassium follows the monotonic decrease model. The response of species to the percentage of silt, sand, organic material, organic carbon,nitrogen,line and aspect  follows the bell model (Unimodal) and the optimal growth limit For each of these factors, it is 20%, 55%, 5%, 35,0.45 and north and westnorth slopes. The study of the response of the species  to environmental and soil factors provided valuable information to determine the ecological needs of this species, which can be taken into consideration in the operation of rangelands improvement in similar areas.

One of the fields in vegetation ecology is understanding the relationships between plant species and environmental factors and analyzing species reactions along the slope of environmental factors. Considering the importance of Salsola arbusculiformis Drob in soil conservation and forage production, the reaction of this species to environmental factors was investigated using the Generalized Additive Model in Shoghan region of North Khorasan province. For this purpose, a systematic random-method was used to measure plant parameters and environmental factors. Canonical Correspondence Analysis was used to examine the relationship between vegetation and environmental factors. The results showed that environmental factors such as percentage of sand, sulfate ions and soil bicarbonate, annual rainfall and altitude, respectively, expressed 5.8, 11.4, 5.7, 9.1 and 7% of the existing variance and have an important role in vegetation changes in the studied habitat. The study of the reaction of this species to environmental factors showed that out of 25 environmental variables studied, 14 variables had a significant effect on plant yield (p <0.01). The response pattern of the studied species along the slope of factors such as acidity, percentage of sand and clay, altitude and finally the percentage of slope was unimodal. The optimal growth limit for each of the above factors was 8.4, 66%, 28%, 1452 meters and 5%, respectively. The response of this species to the amount of sulfate and percentage of soil lime followed a decreasing model, so that with increasing the values of these factors, its yield decreased. The response pattern of the desired species was in two modes, along the slope of factors such as chlorine, sodium, organic carbon, nitrogen, soil salinity and average annual temperature. This model indicates a competitive constraint startsrecalong the peripheral slope. The results also showed that this species has the highest yield in the northern slopes. Studying the reaction of the desired species along the slope of topographic and soil factors provided valuable information for determining the ecological needs of this species, which can be considered in rangeland improvement operations in similar areas.

The present study aimed to model potential habitat distribution of A. sieberi, and its ecological requirements using generalized additive model (GAM) and classification and regression tree (CART) in in the Poshtkouh rangelands of Yazd province. For this purpose, pure habitats of the species was delineated and the species presence data was recorded by the systematic-randomize sampling method. Using DEM and geostatistical method, digital layers of environmental variables (soil and physiographic variables) were prepared with the same spatial resolution (pixel size 30×30 meter). Plant distribution modeling was conducted using CART and GAM models in the R.3.3.1 software environment. The prediction performance of the models was evaluated by the AUC (Area Under the Curve) In addition, The TSS (True Skill Statistic) was used to determine the optimal threshold limitof species presence. The classification accuracy of the presence/ absence map was investigated using the Kappa index. Based on the results, The CART model had a better predictive performance than the GAM models (AUC=0.97 and 0.89, respectively). Furthermore, the Kappa coefficient of the CART model was higher than GAM, based on the obtained Kappa coefficient values (0.97 and 0.89, respectively). This study concludes that the CART model were more accurate in estimating the distribution range of A. sieberi in comparison with the GAM model. The analysis of the importance of variables showed that the electrical conductivity (EC) and acidity (pH) of the first soil depth had the highest effect on the distribution of A. sieberi. In general, it can be concluded that application of data-driven methods, such as CART model, can be useful for accurate estimation of the potential habitat distribution of plant species on a local scale. Therefore, the application of these models to introduce the suitable species in vegetation reclamation plans of Iran's desert rangelands is recommended.

Anzali Wetland in Iran as one of the most valuable wetlands registered in the Ramsar Convention is being destroyed by environmental factors and human activities. In the last two decades, among various satellite images, radar images have played a special role in wetland monitoring. Radar is an all-weather sensor and it is sensitive to surface roughness and moisture, they serve as a valuable source for quick and accurate monitoring of wetlands. However, similarities in backscattering coefficients of different wetland classes and relatively difficult processing – in comparison to optical images- are the most important factors that limit their application. In this study, the capabilities of SAR images in the classification of Anzali wetland and the three main land use classes around the wetland (i.e. agricultural lands, reeds, and built-up areas) were evaluated. Two radar images; Advanced Land Observing Satellite/Phased Array L-band Synthetic Aperture Radar (ALOS/PALSAR) and Sentinel 1 captured in 2018 were used. The texture parameters of the two images have been extracted. The images and their extracted texture layers have been fused by the feature-level method and further classified by the random forest method. The overall accuracy of feature-level fusion is equal to 75% and the kappa coefficient is equal to 0.62. The evaluation results related to producer and user accuracy are 100% and 83.33%, respectively, show the high capability of radar images in the classification and detection of wetlands. However, some errors have been observed in the separation of agricultural lands, reeds, and built-up areas.

Estimating the extent of changes in forest and rangelands land cover leads to a clear understanding of the growth or decline of these natural areas and planning for effective protection of these national assets. The aim of current study was to reveal the trend of land-use changes in the Dizmar protected area of Arasbaran vegetative area, using MSS sensor of Landsat-5 for 1984, ETM+ sensor of Landsat-7 for 2000 and OLI sensors of Landsat-7 for 2015 and 2019. For this purpose, the images were classified and supervised with artificial neural network algorithms, support vector machine and maximum probability algorithm for three classes of forest, rangeland and other uses (any other land use except forest and rangeland). The results showed that the support vector machine method with higher overall accuracy than the maximum probability methods and artificial neural network has a higher efficiency in classifying forest and rangeland cover classes in the study area. Estimates showed that during 34 years, the forest cover has decreased by 135.53 square kilometers and the rangeland and other land use cover have increased by 103.19 and 32.34 square kilometers. Also, the most type of land use change during the years 1364 to 2019 with 64.15 square kilometers was related to the conversion of forest cover. The results of the present study clearly indicate the encroachment on the forest lands of the region. This in turn highlights the need for technical rehabilitation operations in this vegetation area.

Water is one of most important factors in the growth and development of human societies, where water resource limitations have always been one of the main barriers to agricultural development as a major basis for achieving food self-sufficiency. One of the main applications of satellite imagery is its utilization in the field of water resources management and agricultural activities, in which managers can benefit from it for studying cultivation levels, crop classification, crop estimation, and agricultural crisis forecasting. Generally, overall consumption estimation, water/irrigation management, and utilization of dams’ storage capacity are among the most important research topics. This study benefits from the Sentinel-2 satellite for classifying the agricultural crops based on the multi-temporal methods. Besides, four classification methods are adopted for classifying, namely, minimum distance, maximum likelihood, fuzzy, and neural network. Due to the spectral changes of goods during the growing period, using the multi-temporal methods based on the crop calendar can play a decisive role in classifying process, such that the classification accuracy increases to 86 percent via the maximum likelihood and neural network methods. Moreover, the normalized Kappa increased to 90.5 percent, when the neural network method parameters are optimized. The results obtained from the simulation indicate that genetic algorithm is the best method for obtaining the optimal results. After selecting the optimized neural network parameters, the classification has been taken into account and observed that Alfalfa has the largest crop surface, while it requires a considerable amount of water and its demand is in a lower value.

The Paliurus spina- christi species grow naturally in three different vegetation regions of Iran (Northern forests, Arsbaran and Zagros forests). But so far no research has been carried out on the sites of this species. Therefore, present study was conducted to investigate the effect of environmental factors on distribution of ecological groups in Paliurus spina- christi site in the Northern forests of Iran. To consider different condition of the study area, sampling was done based on land units and size of square-shaped plots was 400 m2 that it was determined by minimum area method. After recording existing data in each plot, soil samples were collected from the sampling plots. Then, vegetation data were classified by Two-Way Indicator Species Analysis (TWINPAN) using PCORD software and finally, Canonical Correspondence Analysis (CCA) were used to investigate the relation of each plant species and environmental factors, by CANOCO software. According to the results of TWINSPAN, there were four vegetation types in the study area. The results showed that different species, such as Pyrus cerasifera, Berberis vulgaris, Punica granatum, Cupressus arizonica and Carpinus betulus are indicators species of the study area, replace of Paliurus spina- christi. Also, the results of CCA showed that amount of C/N, N, saturated water content, bulk density, porosity, clay, phosphorus and soil potassium are effective factors on distribution of vegetation in Paliurus spina-christi site. On the other hand, C, organic matter, sand, silt, altitude, slope and aspect have not significant effect on plant ecological groups. According to results, the presence of Paliurus spina-christi is directly related to amount of N and bulk density. In this study, it was clearly shown that using CCA, it is possible to determine the relationship between each plant species and the most important environmental factors.

Multilispectral picture classification is one of the most important techniques for separating earth units.The phenomenon of global warming,expansion damming,water storage behind dams and excessive utilization of existing water for human uses has caused the drying of lakes, including Lake Bakhtegan. For this purpose, Landsat images of 1991, 2000, 2010, and 2017 were collected in Bakhtegan Lake and surrounding areas. These images were categorized based on educational samples in four classes of water, septicity, mountain and urban areas after pre-processing and corrections required by the supervised maximum likeness.The same image was then sorted by multi-layer perceptron neural network method in the above classes. Finally, for both methods, the error matrix was extracted and the overall accuracy and kappa coefficient were calculated.For the year 1991, the maximum probability and neural network method was 87% and 93%, and the kappa coefficient was calculated to be 0.86 and 0.90, respectively. . Therefore, due to the higher accuracy of Negative Network, images of the years 2000, 2010 and 2017 were categorized by this method.After classification, in order to evaluate it, Google Earth was considered as the test sample for each information class and the overall accuracy and kappa coefficient were 89% and 0.85, respectively.

Recent advances in precision farming technologies have triggered the need for highly flexible modelling methods to estimate, classificate and map weed population patterns for using in site-specific weed management. In this research, a learning vector quantization neural network (LVQNN) model was used to predict and classify the spatial distribution of Acroptilon repens L. density. This method was evaluated on data of A. repens L. density in a fallow field in Shahrood, Semnan province in 2010. Weed density assessments were performed following a 2 m × 2 m grid pattern on the field and a total of 550 sampling units on field. At each node of grid pattern, the numbers of A. repens L. seedlings were counted in the field within a permanent 50 cm by 50 cm quadrat. Some statistical tests, such as comparisions of the means, variance, statistical distribution as well as coefficient of determination in linear regression were used between the observed point sample data and the estimated weed seedling density surfaces to evaluate the performance of the pattern recognition method. Results showed that in training LVQNN, test and total phase P-value was greater than 0.7, 0.8 and 1 percent respectively, indicating that there was no significant (p<0.05) difference between statsitcal parameters such as average, variance, statistical distribution and also coefficient of determination in the observed and the estimated weed seedling density. This results suggest that LVQ neural network can learn weed density model very well. In addition, results indicated that trained LVQ neural network has a high capability in predicting weed density with recognition accuracy of 2.7 percent at unsampled points. The technique showed that the LVQNN could classify and map A. repens L. spatial variability on the field. Our map showed that patchy weed distribution offers large potential for using site-specific weed control on this field.

Population growth and development of farm lands through forest cutting make it necessary to use the most of land and pay more attention to sustainable agriculture. As a multidisciplinary management technique in lands around forests as Zagros woodlands, agroforestry reduces the human pressure on them and increases land utilization in sustainable development. So this study was performed to recognize available agroforestry systems in Fars province. First the data including the species and other elements of agroforestry systems were gathered and they were geographically located by GPS. Then they were compared with predefined systems and classified finally. The problems were studied and some suggestions were proposed to improve them. Nair classification method (1978) was applied in this research. The results showed that the function of bases, species and management are changed by climate variation. Seven systems was identified in Kazeroun. Generaly they were Agrisilvicultural systems and Agrosilvopastoral systems.