1. Introduction
Soil detachment is defined as the separation and dislodging of soil particles from the soil mass surface by the force of raindrops and overland flow (Lautridou, 1990; Wang et al., 2014; Zhang et al., 2002). Scouring by overland flow is the dominant process resulting in detachment and transport of soil particles. The mechanism of soil detachment by open channel flow has become a research focus (Cao et al., 2009; Wang et al., 2016; Xiao et al., 2017). In the case of clear water, maximum soil detachment rate referred to as soil detachment capacity (SDC) is a key parameter used to describe the soil erosion process (Nearing et al., 1991; Sun et al., 2018). SDC is often used to estimate the rate of soil erosion resulting from overland flow and has been widely used in a range of representative erosion models, such as the Water Erosion Prediction Project (WEPP) and other process-based models (Misra and Rose, 1996; Morgan et al., 1998; Nearing et al., 1989). Meanwhile, SDC is to a certain extent affected by the hydraulic parameters of overland flow and the relationship between SDC and hydraulic parameters can be established to quantitatively simulate or predict soil detachment (Nearing et al., 1991; Zhang et al., 2002, 2003). For this reason, past studies have conducted laboratory control experiments to reveal the internal relationship between SDC and hydraulic parameters (Zhang et al., 2002, 2015). The results of these past studies have indicated that soil detachment is strongly influenced, and in some cases controlled, by hydraulic parameters, such as shear stress (Nearing et al., 1991), unit energy of the water carrying section (Hairsine and Rose, 1992a, 1992b), stream power (Wang et al., 2018; Xiao et al., 2017) and unit stream power (Morgan et al., 1998). However, since these studies did not employ standardized experimental conditions, there remains no consensus on the optimal hydraulic parameters to predict SDC and the dynamic mechanism of the soil detachment process remains unclear. For example, Morgan et al. (1998) found that the unit stream power is the best parameter to predict SDC, whereas Nearing et al. (1999) demonstrated the strong relationships of shear stress and stream power with SDC. Therefore, obtaining consensus requires further numerous experiments under different dynamic conditions to reveal the dynamic mechanism of soil detachment driven by overland flow.
Past research has demonstrated that soil type, bulk density, soil moisture, freeze-thaw and the root system all have strong relationships with SDC (Gyssels et al., 2006; Van Klaveren and McCool, 2010; Ye et al., 2017; Zheng et al., 2000). The seasonal soil freeze-thaw process is a common phenomenon in mid-latitude regions, such as the northern part of the Loess Plateau in China. The freeze-thaw process changes soil physical and mechanical properties and reduces the stability of soil aggregates, thus affecting soil SDC (Kværnø and Øygarden, 2006; Sun et al., 2018). And the SDC would be first decreasing and then increasing with an increasing number of freeze-thaw cycles (Sun et al., 2018). The soil freeze-thaw process can result in increased soil and water loss during the spring thawing period of temperate regions to > 50% of annual loss (Chow et al., 2000; Ellison and D., 1945; Froese et al., 2001). An analysis of soil detachment under the freeze-thaw process can provide an important theoretical basis for developing soil erosion control and prevention mechanisms for the seasonal freeze-thaw region. However, most previous studies on the effect of freeze-thaw on SDC were conducted on bare soil, whereas many affected areas have vegetation, and the extent of area with vegetation is continuous increasing with the implementation of a series of ecological restoration projects (Wang et al., 2014). Past studies have gradually highlighted the importance of the effect of the plant root system on soil properties and soil erosion in this region. The root system plays a key role in increasing soil stability, thereby reducing soil erosion (Baets et al., 2007; Jiao et al., 2012). The root system not only promotes the formation of large soil aggregates by consolidating fine soil particles, but also increases soil organic matter and improve soil structure through root exudates and plant residues (Amezketa, 1999; Burylo et al., 2012; Whalley et al., 2005; Ye et al., 2017). Soil with vegetation can have an SDC that is 23.2–55.3 lower than that of bare soil due to the effect of the root system (Wang et al., 2014). Gyssels et al. (2006) found that the SDC can be reduced by > 50% through the presence of herbaceous plant root systems compared with that of soil without roots. However, current research on the root system effect on the soil detachment process is mainly focused on non-freeze-thaw conditions, and therefore the relationship between the root system and SDC under freeze-thaw condition remains unclear. Quantitatively analyzing the effect of the root system and freeze-thaw on the soil detachment process and the development of a high-precision SDC prediction model is of great significance for the improvement of soil erosion prevention and control strategies in the seasonal freeze-thaw region and for the accurate assessment of local soil erosion (Laflen et al., 1991; Lal, 1989).
Therefore, the northern part of the Loess Plateau, China, as a typical seasonal freeze-thaw region, was selected by the current study as the study area. The current study investigated the sand and loessal soils that are widely distributed in this area. A flume experiment was conducted to explore the effect of the root system and the freeze-thaw cycle on SDC. In addition, an SDC prediction model was developed based on hydraulic parameters. The objectives of this study were to: (1) investigate and compare the effects of the root system, freeze-thaw and freeze-thaw combined with the root system on the soil detachment process, (2) develop a model to simulate SDC under the effects of the root system, freeze-thaw and freeze-thaw combined with the root system.