Effects of tillage measures on runoff
We verified that crops could act as a type of vegetation cover (Table 1 and 2) and play an important role to mitigate runoff and soil loss on sloping farmlands in agreement with previous studies (Cerdà et al., 2017; Keesstra et al., 2016; Prosdocimi et al., 2016a, b; Wang et al., 2018). Different tillage systems had different impacts on soil erosion associated with processes occurring in slope farmlands (Liu et al., 2011; Xu et al., 2018). The Vr treatment has already been verified to increase soil erosion because of microtopography changes (Liu et al., 2011; Zhang et al., 2009). Conservation tillage could significantly postpone runoff yield and slow runoff velocity compared to conventional tillage (Table 3). Our results indicated that horizontal ridges, mulching, or seedling corn canopy were effective in controlling runoff generation (Table 1, 2 and 3), especially at 50 mm/h, at the maize seedling stage. This might be because the studied conservation measures could enhance the infiltration capacity of water (Table 5) or increase soil surface roughness (Rodríguez-Caballero et al., 2012; Vermang et al., 2015; Wang et al., 2018), and crop leaves could intercept rainfall and change raindrop diameter and energy (Table 1 and 2) (Ma et al., 2013; Zhang et al., 2015). As there are only limited chances for extreme precipitation in the region (Zhang et al., 2010), adopting Hr and Cm would be able to restrict runoff generation. In addition, these two tillage measures were also efficacious in reducing the runoff-flow velocity, which is a key factor affecting runoff energy and erosiveness (Vermang et al., 2015); both Hr and Cm performed better at 50 mm h-1 than at 100 mm h-1. Our results were in accordance with previous studies on other soil types (Prosdocimi et al., 2016a, b; Xu et al., 2017). The runoff generation was postponed and the surface-flow velocity decreased mainly because both Hr and Cm treatments changed the microtopography of the soil with increasing surface roughness (Vermang et al., 2015; Wang et al., 2018) and the infiltration of conservation tillage was higher than that of conventional measures (Table 3). The outcome provided more water storage microstructure for the surficial soil, causing the rainwater to infiltrate rather than flowing downhill (Liu et al., 2015; USDA-ARS, 2008, 2013). This outcome also increased the friction between rainwater and land, thereby reducing runoff velocity. Comparing the effects of Hr and Cm, Hr set a higher threshold for runoff yield, as it could lead to more water storage between ridges. However, once the runoff had occurred, Cm performed better, since the presence of cornstalk could reduce the flow velocity to a very low level (Table 1). Thus, Hr+Cm is the optimal treatment from the perspective of postponing runoff-yield and restricting the destruction of runoff, once generated.
The runoff loss rate significantly increased following a low start during the runoff generation period and then remained stable at a certain level, based on the rainfall intensity. The results correspond to a study on purple soil (Xu et al., 2008). Hr and Cm could effectively constrain the runoff loss rates and decrease the runoff amount, especially at 50 mm h-1. The Hr+Cm treatment, which combined horizontal ridging and mulching, always had an effective influence on runoff under all rainfall types, especially for preventing runoff throughout the experiment under the rainfall intensity of 50 mm h-1. As runoff is the main vector affecting both soil loss and agricultural nonpoint source pollution (Hudson, 2015; Zhang et al., 2007), Hr+Cm should be recommended as an effective tillage practice in the region.
However, this recommendation would engender extremely higher outliers for runoff rate as a real-time response to ridge rupture when the plots were treated with Hr, especially under heavy rainfall conditions (Li et al., 2016; Lu et al., 2016). In this case, the water held by the two adjacent ridges drained immediately after ridge rupture and outrushed into the next inter-ridge area, causing either successive ridge ruptures or runoff overflow, both of which could prompt a sudden upsurge in runoff rate, in accordance with the results of Xu et al. (2018). Consequently, the total runoff loss amount also ascended. The rising magnitude caused by ridge rupture depended on the rupture time and location of the initially ruptured ridge. For instance, in the Hr-treated plot, ridge rupture occurred relatively earlier and closer to the top of the plot under the rainfall intensity of 100 mm h-1 than under 50 mm h-1 (Figure 6 and 7), thereby resulting in greater runoff loss. Thus, enhancing the quality of ridges to improve their water pressure tolerance capability is vital when applying horizontal ridges (Liu et al., 2014a). It was found that mulching could directly lead to water absorption and protection of a ridge from saturation and erosion by raindrop and runoff (Cerdà et al., 2016; Jordán et al., 2010), thereby reducing the risk of ridge rupture. The results confirmed that the Hr-treated plots suffered three times as many ridge ruptures, while the Hr+Cm plots suffered only one ridge rupture. Moreover, no successive ridge ruptures were observed in the Hr+Cm plots, because mulching and soil blocks would likely be obstructed by the next ridge with the presence of cornstalk, rather than triggering successive ridge ruptures, even if one of the ridges happened to rupture. Moreover, ridge-furrow planting under mulching conditions played an effective role in reducing surface runoff with an increase in soil-water infiltration (Gholami et al., 2013; Kader et al., 2017).
Vr could increase the runoff loss rate and amount under light rainfall conditions, as shown by Shen et al. (2005) and Zhang et al. (2009) on black soil, and by Xu et al. (2008) on purple farmlands compared to the runoff between contours and downslope ridges. Therefore, vertical ridges should be avoided on slope croplands in the region.