Improved estimation and prediction of the wind-erodible fraction for Aridisols in arid southeast Tunisia

Wind erosion is a serious environmental threat. However, for Northern Africa estimates or predictions of the wind erodible fraction (WEF) in arid soils are rare. This study was conducted to (i) compare the flat sieve (WEFfs) and the Fryrear equation (WEFc) in order to estimate WEF (ii) fuse the directly measured WEFfs values to other measured soil parameters, i.e. soil organic matter (SOM), carbonate (CaCO3), sand, silt and clay content, to obtain a new improved WEF equation prediction. Samples were collected at ten locations near Gabe`s (Southern Tunisia) at three depths (0-2.5, 2.5-5 and 5-15 cm) in cultivated or native soils. The obtained WEFfs were > 80% and therefore the soils would be classified as highly wind erodible (WEF > 50%). The lowest WEFfs values were found in the cultivated soils (82.5%), ranging from 82.5 to 96.9%. Our WEFfs estimates were positively correlated to sand (r = 0.78; p < 0.001), but negatively correlated to CaCO3 content (r =-0.82; p < 0.001). Interestingly, neither SOM nor clay content was strongly related to WEFfs. However, the WEFfs in cultivated soils was significantly linked to Sigma(SOM + clay content). The Fryrear estimation predicted all WEFc values to be less than 60%. This clear mismatch to our data relates to CaCO3 presence, high sand and low SOM content in our Tunisian soils, when compared to US soils used to establish the Fryrear method. Therefore, we established a new equation for the wind erodible fraction, WEFmod (%) = 125.83-0.59*Clay(%) -0.22*Silt(%) -0.25*Sand(%) + 0.29*SOM(%) -0.69*CaCO3(%) (r = 0.85; p < 0.001). In the studied regions of Tunisia, this equation predicts much better soil WEF. The present paper is a first study to improve WEF estimation and prediction in arid North Africa. Better predictions of WEF are crucial in combating wind erosion of soils both in North Africa and in comparable soils globally.