Empirical models for predicting the erosion-reducing effects of plant roots during concentrated flow erosion

Recent research indicates that plant roots can reduce soil erosion rates during concentrated flow significantly. Earlier studies revealed the erosion-reducing effects of plant roots under controlled laboratory conditions and presented some equations applicable to the tested soil and flow conditions only. Although an attempt was made to unravel the impact of different environmental, plant and flow properties on the erosion-reducing potential of plant roots during concentrated runoff, significant effects were hard to demonstrate because of the small number of data. Therefore, the objective of this study is to pool different datasets (384 data collected under standardized experimental conditions in total) together for constructing empirically-based models predicting (1) soil detachment rates for both bare and root-permeated topsoils and (2) the erosion-reducing potential of plant roots during concentrated flow erosion. The model that best predicts absolute soil detachment rates from bare and root-permeated silt or sandy loam topsoils (ASD) requires information on root density, topsoil moisture content prior to concentrated flow erosion, bulk density and mean bottom flow shear stress. Although all these variables contribute significantly to the prediction of ASD, the validation results indicate that there is still a large scatter on the data. Especially for small detachment rates, the model tends to overestimate the observed values. Another model predicting the reduction in detachment rates of root-permeated topsoils compared to bare ones (RSD) does not show good validation results either. However, the model for fibrous root systems was found to be very promising. This model explains 79% of the variation using only root density and soil texture information and the validation results show that 69% of the variation in the validation dataset is accounted for by the model. For topsoils permeated with tap root systems, however, the model results were not satisfactory. Only 10% of the variation in the validation dataset could be explained by a model using root density and mean root diameter as input variables. We conclude that the erosion-reducing effect of topsoils permeated with fibrous roots can be predicted very well, whereas relative erosion rates for tap root-permeated topsoils still remain difficult to predict with the studied variables only. More process-based knowledge is needed to distinguish erosion-reducing effects from erosion-accelerating effects for tap root-permeated topsoils.