Soil erodibility data, calculated using measured soil loss from standard runoff plots, collected over at least one year and applying the standard requirements for calculating the soil erodibility factor (K) of the Universal Soil Loss Equation (USLE), have been analysed to investigate whether climate affects the susceptibility of soils to water erosion. In total, more than 300 K-values extracted from the literature have been analysed. Due to the limited availability of data related to the characteristics of the soil and the location of the measuring sites, all the analysis has been carried out using only soil textural characteristics, organic matter content, rock fragment content and the some general characteristics of the climatic zone where the plots were located.The first evidence of a strong climate effect on soil erodibility is shown by the seasonal variation of mean monthly soil erodibility (K m ). Using data collected in the USA and Italy an effect of mean monthly air temperature on K m could be identified. Data collected in Indonesia (where mean monthly air temperature remains fairly constant throughout the year) showed comparable variations of monthly soil erodibility. However, it was impossible to explain these variations in K m as no other data than mean monthly air temperature were available.Mean annual soil erodibility shows a clear climate effect. Soil erodibilities can be subdivided into two large groups, one corresponding to soils in cool climates (Df and Cf climate according to the Köppen-Geiger climate classification) and another to soils located in warm climates (tropical Af and Aw climates). Erodibilities of Mediterranean soils (found under Cs climate) plot among the soils found in Af and Aw climates. These subdivisions can be made for both stony and non-stony soils. Limited data suggest that soil aggregate stability is a good predictor for explaining soil erodibility variations between different climate zones.
Aggregate stability testsSoil aggregate stability for four soils was measured accordingly to three different procedures and four different indices were calculated. The procedure described by Imeson and Vis (1984) consists of counting the number of drops (4 mm in diameter, falling from a height of 1 m) needed to break down a soil aggregate with an initial diameter of 4·8 mm into fragments smaller than 2 mm. Tests are repeated for 20 aggregates saturated by capillary rise, and median drop number (i.e. the number of drops at which 50% of the aggregates are broken into fragments smaller than 2 mm) is taken as a soil aggregate stability index.Another procedure is based on immersing 30 g of aggregates (in this case saturated by capillary rise) in 200 cm 3 of distilled water inside a cylinder 10 cm in height and 10 cm in base diameter. The cylinder is than rotated around its height at 33 rotations per minute for a given time duration (with rotation axis kept horizontal). Usually four or more different time durations are used for establishing the whole curve of aggregate destruction until no fur...