Increased occurrence of overland flow, erosion and gully transformation
Soil compaction with increased bulk density, increased silt content and decreased infiltration rates due to the combined effects of machine traffic, cattle trampling and liming were identified for croplands and pastures. Substantial increases in pH values were recorded in sampled soils and surface water within croplands and pastures due to extensive liming that is meant to increase biomass yields (Fageria and Barbosa Filho 2008). For Brazilian Ferralsols, Haynes and Naidu (1998) showed that increased soil pH values were related to clay dispersion and decreased the infiltration rates in accordance with our findings. The loss of aggregates with land use change is documented by Resck et al. (1999), who found that after 9 years of soybean growth, the percentage of topsoil aggregates > 2 mm had decreased by 30%. The highest aggregate stability was related to Cerrado, pastures and eucalyptus plantations; interestingly, the stable aggregates were not related to no-till soybean systems (although no-till accumulated the most organic carbon).
Location of the Tenente Amaral catchment where the field campaigns were conducted
Figure. Decreased aggregate stability under croplands (SOY-Soybean, SC-Sugarcane) compared to pastures (PA) and natural Cerrado
(CE). Aggregate stability was determined after the method of Herrick et al. (2001) with n=45 samples per soil depth.
The combined effects of changes to the soil structure and the very high rainfall erosivity factors (Oliveira et al., 2013) commonly cause more frequent overland flow due to high-intensity storms in the rainy season; as a result, extensive soil erosion may occur and gullies may form (voçoroca) (Wantzen et al., 2006; Vrieling and Rodrigues 2004; Nascimento et al., 2012). According to our findings, accelerated erosion is likely to occur in the future. Barreto et al. (2013) used soil erosion modelling to show that high erosion and sedimentation are directly linked to land use changes. Favis-Mortlock and Guerra (1999) observed that merely ten years of conventional soybean cultivation resulted in the entire loss of the A-horizon (up to 20 cm) through soil erosion at a study site in Mato Grosso. No-tillage systems were largely implemented in the Cerrado in the early 1990s (FAO 2001; reviewed by Bernoux et al., 2006); these systems are well known for their positive effects on C-storage (e.g. Carvalho et al., 2009; Neto et al., 2010; Pinheiro et al., 2014) and for stabilising aggregates (Green et al., 2007).
However, even with common no-till practices, high soil erosion and gully formation were directly linked to agriculture (Didoné et al., 2014). Likewise, Wantzen and Mol (2013) pointed out that despite of direct planting and other measures (e.g. construction of contour lines) erosion is still a problem. The study of Jesuz et al. (2014) associated the occurrence of 39 active gullies in a mesoscale catchment to no-tillage soybean-maize/cotton areas, rather than to pasture areas. Their studies investigated active gully systems under agricultural practice but there were no information available if the gullies were initially developed under the current land use or existed before the implemenation of no-tillage. The application of geo-radar (GPR) data established a relationship between piping and gullies (Augustin and Aranha 2006); this link was also found in field studies for pastures (Castro 2005). Based on SAR (Synthetic Aperture Radar) imagery, Vrieling and Rodrigues (2004) assessed similar effects for degraded pastures, which experienced the highest soil losses in the rainy season. Vrieling et al., (2008) found that pasture areas have the highest erosion risk in November and December, when the vegetation cover is reduced, and directly after the application of agrochemicals. Plantations are considered to have a smaller erosion risk than annual crops (Smith et al., 1999). In field experiments, Wantzen (2006) showed that the increased sediment loads from gully erosion significantly reduced benthic invertebrate populations. Wantzen (2003) measured fluxes of up to 60 tons of sediments per day at the outlet of a gully after heavy rainfall.
Our own observations in Mato Grosso confirm that gullies more than 10 metres deep and several hundred metres long are common at pasture and crop sites in the Cerrado; these gullies form rapidly,
even in fairly flat landscapes.
the references of this website can be found in the two publications in Ecohydrology and Geoderma Regional!