Modeling of land use effect on soil crust strength in southeastern Ahvaz

Introduction
Crusts are a hard layer on surface soil, that is formed by disaggregation– aggregation process in which particles of soil, air, water, and organic matter are connected to each other and classified into different types including physical, chemical, and biological. Chemical crusts like salt crusts are formed due to intense evaporation on the surface of extremely salty soils. Physical crusts are formed raining or through irrigation of agricultural lands and are divided into three categories including structural, erosional, and sedimentary, depending on the process of their formation. The biological crust which is formed by the function of algae, cyanobacteria, mosses, and lichens, and due to their positive protective roles and restoration ability received much attention so far. However, studies on the effects of non-biological crust and their protective role have been considered less. Various effective factors on crust strength have been investigated but land use has been left out. This research has focused on modeling land use effection crust strength, in dust emission sources in the southeast of Ahvaz.
 
Materials and Methods
In the south-eastern of Ahvaz in Khouzestan province, three land uses including agriculture, agroforestry, and barren land were selected. In order to measure the strength of the surface crust in selected land uses, a handheld penetrometer was used and crust strength was measured in random points in 30 points in each land use. To reduce the influence of other environmental conditions, measurements were done scattered on each land use. Then, to obtain the strength in one point, three measured points were averaged, and finally, 90 measured points were obtained for each. The surface soil moisture of land uses was done by taking soil samples and measuring in the laboratory, and then significant differences between land use groups were tested by analysis of variance. Normality and homogeneity of variances were tested by using the Kolmogrove-Sminov and Levene's tests on soil strength data set. Due to the fact that the soil texture is different in studied land uses and also the soil texture is one of the most important factors affecting the strength of crust in the measured points, the soil texture was extracted from the existing maps. In order to investigate the effect of these two independent factors on the crust strength as well as their interaction, General Linear Modeling (GLM) was chosen to exclude the soil texture effect on crust strength variation and model the land use effects.
 
Results and Discussion
Results showed that soil surface moisture does not have a significant difference in land use groups. By using the General linear model, crust strength was modeled. In the first stage, the effect of land use and soil texture were investigated as the most important factors affecting the hardness of the crust and the results showed that land use and soil texture as well as their interactions are effective in changing the hardness of the crust at the level of 95 % and 99 %, respectively. These factors have an effect on the variance of crust hardness, but the main source of variance is land use, and this factor alone explains about 78 % of the crust strength variance, and the model explains 96 % of the variance of the dependent variable and the presented model is significant at the level of 99 %. In order to check the existence of a significant difference in crust strength in studied land uses, Helmert's Contrast and Bonferroni tests were used. The result showed that there is a significant difference in the average crust strength of barren lands with agriculture and agroforestry at the 99 % level, and no significant difference is observed between agriculture and agroforestry. Then, in order to investigate the single factor of land use, the soil texture was considered as covariance, and its effect on the hardness of the crust was removed. The results showed that there is a significant difference in the average hardness of barren land use with agriculture and agroforestry at the level of 99 %. The presented model explains 86 % of the variance of the hardness of the ridge, and among the factors with a significant level, 99 % of the hardness of the crust in a barren land with 70 % partial effect has the largest role in explaining the variance. With the change of land use from agroforestry to barren land, the hardness of the soil surface increases by 50 %, and with changing to agricultural land, it decreases by 14 %.
 
Conclusion
In agricultural and forestry land uses, with the increase in the traffic of people and heavy machinery, the crust is broken and does not return to its original strength. Based on these results, it can be said that in desert areas, vegetation conservation is not the only way to protect soil from wind erosion, but protecting the crust against traffic and breakage can be an efficient solution that has received less attention. Legal confrontation with land use change and land plowing can be a sustainable solution for these areas. It is suggested that with a general assessment of the surface strength of the crust on bare land, easily can be protected against the wind only with management practices.