Spatial analysis of biological soil crust based on Biological Soil Crust (BSCI) index

Background and Objective:

Biological soil crusts are a collection of lichens, mosses, fungi, cyanobacteria, etc. that are part of the soil ecosystem. Estimation of density and distribution of biological soil crusts in arid and semi-arid regions of Iran, which is the subject of soil erosion and wastage is very important. Methods based on remote sensing techniques are important in terms of cost and time less efficient methods to achieve this goal. Segzi plain is one of the critical points of wind erosion in Iran and identifying and determining the distribution of biological soil crusts as a soil modifier is an effective step in reducing wind erosion in the region. In this research, BSCI (Biological Soil Crust) index has been used to prepare the distribution map of lichen-dominated biological soil crusts.

Materials and Methods :

The study area is part of the Sajzi Desert (Central Deserts of Iran) which is located in Isfahan province of Iran. The study area with an area of 199.5 hectares is spread between the eastern lengths of 51o52'32" to 52o27'41" and the northern widths of 32o33'31" to 32o55'01". The average slope of Segzi plain is 1.08 percent and its average height is 1680 meters. According to the statistics of the East Isfahan Meteorological Station (Shahid Beheshti Station), the average annual rainfall in the region is 106 mm. According to the Dumarten climatic classification, the climate of the region is dry and according to the Amberge classification it is cold. The BSCI index is a combination of the relationships used to estimate vegetation and bare soil surface, and its mathematical relationship is the slope of the soil line. To calculate the soil line in an area, one must first separate the pixels that have bare soil and no vegetation. In order to calculate the soil line equation, in four seasons of a year, images of Landsat OLI 8 satellite related to 2018 were downloaded from the site of the US Geological Survey and 20 to 30 pixels of pure bare soil were extracted by drawing the reflection values ​​of these pixels in the red and infrared band. Red near soil line coefficients was calculated for each season in the Segzi Plain. Based on BSCI index, lichen-dominated biological soil crust are identified using at least VIS-NIR spectral reflection and the slope between the red and green bands compared to bare soil and dry vegetation. Using ENVI software, the distribution shells of biological shells with lichen dominance were prepared in four seasons since 2018 in Segzi plain. Then, the prepared maps were validated based on land points and the total accuracy and kappa index were calculated in all four seasons. The collected lichen samples were identified based on their morphological characteristics and using a stereomicroscope, conventional microscope and common color reagents such as potassium hydroxide (KOH). After applying the BSCI index on the Landsat  OLI 8 satellite image, using ENVI software, spectral profiles related to 4 points of Segzi plain in four seasons of the year were prepared and the spectral reflection in four seasons of the year in different points were examined.

Results and Discussion:

 The slope of the soil line is lower in the rainy season, which coincides with the growth of herbaceous and annual plants, compared to the summer season, which has the least amount of rainfall, and the annual plants have dried up and become extinct. In May, the slope of the soil line was minimal (0.39) and in late summer it has its maximum value (0.78). In fact, the slope of the soil line has decreased from mid-August to May, and then has increased with the loss of annual vegetation and the increase of bare soil surface. The distribution maps of bio-shells in all four seasons of the year were validated during field visits and the year it was found that the highest accuracy of the map related to the map produced from Landsat 8 image is related to summer with 94% total accuracy and Kappa index equal to 0.7412. Interpretation of the spectral profiles of the BSCI index shows that the reflections of the spectra related to the zephyr and strain prepared on the lichen dispersion points are very close to each other and also the spectral profiles of the mid-autumn and early spring are quite consistent. Whereas in the faults, which did not cover the biological crust, the amount of reflection was higher and there was a slight difference between the reflection diagrams of autumn and spring. Although the reflectance values of a range of agricultural lands and the distribution points of biological crusts are very close to each other, the spectral diagrams of all four seasons are very different from each other. But in all seasons of the year and in all places, the least reflection has occurred at the beginning of winter and the most reflection has occurred in summer. The climate of Segzi plain is Mediterranean and precipitation occurs in the cold season of the year. Simultaneously with the increase of precipitation from the middle of autumn, annual plants and mosses at the base of shrubs begin to grow and reach their peak in early winter and again at the beginning of spring. Decreases in rainfall have reduced their density. If the winter spectrum has the least reflection in all places. While in late summer, when the annuals and mosses have dried up, it has had the greatest spectral reflection. Fasaran, which is a barren area and a landfill, it has shown its maximum reflection. Therefore, the BSCI index relative to the percentage of organic matter has a significant error in the detection of biological soil crust, and where the organic matter is high may not provide an accurate diagnosis of soil bioshells. Of course, since the BSCI index is defined for the detection of throat compounds in lichen tissues. The error rate for organic matter is reduced to a minimum. As it has been observed in the final map, there is no cover of biological soil crusts in Fasaran and only soil biological crusts are observed in the areas around Fasaran in the agricultural areas. In agricultural areas, due to human intervention and cultivation, the amount of annual plants is different from the field of natural resources in different seasons of a year have become.

Conclusion:

 Spectral similarity of the most important soil surface, including vegetation, the involvement of human factors in increasing or decreasing soil organic matter, bare soil, etc. limits the efficiency of the BSCI index and therefore in the time period of satellite images and regional conditions have a great impact on It has the accuracy of BSCI index.