Runoff and soil loss in biocrusts and physical crusts from the Tabernas Desert (southeast Spain) according to rainfall intensity

Abstract

Biological soil crusts (biocrusts) influence hydrological and erosive processes in drylands, and their effects increase with hypothetic successional development. Runoff and raindrops, both dependent on rain intensity, are among the main causes of erosion in these areas. However, little is known about the existence of soil loss nonlinearity in relation to rain intensity and crust types; this nonlinearity could control biocrust succession and dynamics. The assumption of biocrust types as successional stages, which allow space-for-time sampling, makes it advisable to include all the successional stages when exploring possible nonlinearity. We considered seven types of crusts, three physical and four biological. We created four rainfall intensity levels in controlled laboratory conditions: 18, 60, 120, and 240 mm/h. In all but the last, we conducted the experiments at two levels of antecedent soil moisture. Generalized Lineal Models enabled us to test for differences. These analyses confirmed previous knowledge regarding the significant effect of rainfall intensity, crust type and antecedent soil moisture and their interactions on runoff and soil loss, despite the small sample size of the sample units. For example, runoff, and particularly soil loss, decreased along succession. Moreover, some results were novel: the runoff coefficient increased only up to 120 mm/h of rain intensity. A decoupling between runoff and soil loss occurred at high intensities. Soil loss increased as rainfall intensity increased only up to 60 mm/h, and then it decreased, mainly due to physical crusts, because of the formation of a water sheet on the surface due to the incoming rainwater exceeding the drainage capacity. Although soil loss was greater in the incipient cyanobacteria than in the most developed lichen biocrust (Lepraria community), the protection provided by any biocrust against soil loss was great compared to the physical crust, and almost as strong at all rain intensities. Soil loss increased with antecedent soil moisture only in physical crusts. Biocrusts resisted the rain splash even at a rainfall intensity of 240 mm/h.

Keywords: biocrusts; raindrop impact; rainfall simulation; semi-arid; soil crust dynamics; splash erosion.

Figure 1

Characteristic landscape of the Tabernas Desert, Almería, Spain. The high mountains in the right of the background are part of the Sierra Nevada. Lower, in the foreground, it is possible to distinguish a lichenic biocrust.

Figure 2

Undisturbed samples of crusted soil. (A) Sample immediately after being collected, conditioned for transport. Panels (B–H) correspond to the crust types: physical structural, physical depositional, physical island, incipient cyanobacterial, mature cyanobacterial, lichenic dominated by Squamarina and Diploschistes, and lichenic characterized by Lepraria isidiata, respectively. (I) Set of samples on the simulator table after an experiment, showing the collectors and pipes for runoff and sediments.

Figure 3

Average values of runoff coefficient (RC) as related to rain intensity (distinguishing three intensities: 18, 60 and 120 mm/h; graph A), with antecedent soil moisture (dry or wet; graph B), and with crust type, with two (graph C) and seven types (graph D). Graph (E) shows the water content after experiments in relation to the seven crust types. Graphs (F–H) show the interaction of the effects of intensity and crust type on RC.

Figure 4

Average of the times to runoff (T₀, seconds) in relation to the seven crust types (graph A), the three rain intensities (graph B) and the two antecedent soil moistures (graph C).

Figure 5

Average values for runoff coefficient (graph A) and time to runoff (graph B) as related to rain intensity when we plotted four rain intensities (18, 60, 120, and 240 mm/h; only experiments starting on wet soil).

Figure 6

Average of soil loss values (g/L) in relation to the seven crust types (graph A) and the two general crust types (graph B); as well as to the three rain intensities, with all of the data together (graph C) and separating the physical from the biological crusts (D); and to the four rain intensities (graph E).