Runoff mitigation via micro‐dams and conservation tillage—Numerical modeling of runoff and erosion from maize field trials

INTRODUCTION

In the European risk assessment framework for plant protection products (European Parliament and the Council of the European Union [EU], 2009), surface runoff from sloped agricultural fields is one building block for the estimation of risks for the aquatic environment, that is, surface water bodies adjacent to agricultural fields. The Forum for Co-ordination of pesticide fate models and their Use (FOCUS; FOCUS, 2001, 2015) set the framework to assess predicted environmental concentrations (PECs) in standard water bodies for surface water (PECsw) and sediment (PECsed). For this process, the calculation of runoff water, erosion, and pesticide mass loadings to surface water bodies is conducted using the Pesticide Root Zone Model (PRZM; US Environmental Protection Agency [USEPA], 2006). In PRZM, water runoff and erosion are quantified with the US Department of Agriculture (USDA) Soil Conservation Service curve number (CN) methodology (US Department of Agriculture [USDA], 2004) and a watershed-scale variation of the Universal Soil Loss Equation, that is, the MUSS equation (Williams, 1995), respectively. Based on the PRZM output, PECs are then calculated using the model TOXic substances in Surface Waters (TOXSWA; Beltman et al., 2014). There, additional entries via spray drift are accounted for to calculate maximum concentrations as well as time-weighted averages. An analogous approach exists in US surface water exposure assessments with the PWC suite (US Environmental Protection Agency [USEPA], 2021), where PRZM is used in conjunction with the surface water body scenarios implemented in the variable volume water model (VVWM) to calculate estimated environmental concentrations for ecological risk assessments (EECs).

Mitigation becomes increasingly important to reduce the environmental impact of pesticides (Fox et al., 2021) and to comply with the objectives of the EU Green Deal (European Commission [EC], 2019) and the Farm-to-Fork strategy (European Commission [EC], 2020a). To prevent surface runoff and erosion, it is advised to minimize runoff generation both in-field (via increasing infiltration) and at the edge-of-field (e.g., via vegetated filter strips [VFS]). However, measures that prevent runoff are much more effective than measures to stop already moving runoff (Alix et al., 2017).

One method is to create so-called micro-dams between the ridges of potato fields (Olivier et al., 2014) or in maize cultivation (Sui et al., 2016), also with the purpose of improving irrigation efficiency (Keshavarz et al., 2020). The required equipment is largely available commercially or available as advanced prototypes. An overview of the effects of quantitative reductions of runoff, erosion, and pesticide fluxes along with calculated effects on CN and resulting PECs is given in Sittig et al. (2020). Another method for runoff mitigation is conservation tillage. This term refers to a variety of practices, in contrast to conventional tillage. Gebhardt et al. (1985) group all tillage practices that reduce soil or water loss, when compared with mold-board plowing, as conservation tillage. This results in little or no soil disturbance (Cueff et al., 2021). In their review paper, Alletto et al. (2010) characterize conservation tillage generally by the condition that more than 30% of the soil remains covered by crop residues after planting. The mitigating effect of conservation tillage on runoff and erosion needs further investigation (Du et al., 2021). Nevertheless, this practice not only prevents soil erosion but also compaction (Borrelli & Panagos, 2020; Gaynor et al., 1995) and reduces the emission of carbon dioxide (Reicosky, 1997).

In environmental risk assessment, either fixed default mitigation effectiveness values for specific measures can be used or a higher-tier modeling can be conducted, for example, by deriving CNs for PRZM belonging to specific measures. A lowering of the CN by three points following the application of micro-dams, other in-field bunds, and reduced tillage was proposed by the SETAC MAgPIE workshop (Alix et al., 2017). More recent work generated additional robust data and demonstrated that this proposed reduction is too small (Sittig et al., 2020). The objective of the present work is to further enlarge the underlying database for the effect of micro-dams and to add further data for conservation tillage as well as the combination of both strategies. The aim is to provide a better-founded recommendation for the use in regulatory risk assessment and management.

In this study, in addition to quantitatively reporting the results of maize field trials, CN values and estimates for the C-factor of the MUSS erosion equation (representing crop type and tillage method) have been derived. CN numbers were estimated both event-wise, based on reported rainfall–runoff relationships and, alternatively, via simulating complete seasons with PRZM. By this means, we numerically quantified the effects of the dedicated mitigation measures micro-dams and conservation tillage (with subsoiling, as defined in Alletto et al., 2010) in terms of reduction in runoff, and consequently, erosion and pesticide loadings. The field trial data allow both the abstraction and transfer to other conditions and a conservative and protective risk assessment because they were gained under very vulnerable conditions with large slopes (9%–16%) and large amounts of precipitation (191–337 mm). Example calculations within the European and the US framework for pesticide legislation were conducted to demonstrate the effect in the risk assessment.

MATERIALS AND METHODS

The effects of the mitigation measures micro-dams and conservation tillage in three maize field trials from 2018, 2019, and 2013 were assessed. Furthermore, these results were evaluated numerically to derive runoff CNs and parameter values for the C-factor of the MUSS erosion equation. Example calculations demonstrate the impact of these measures on concentrations in surface water and sediment for the risk assessment with the European runoff scenarios (the PECs in surface water and sediment: PECsw and PECsed) and with the US scenario of Illinois corn (estimated environmental concentration: EECs).

For the trial from 2013, a PRZM simulation for the complete season is presented as an alternative to the event-wise evaluation of micro-dam application already presented for this trial in Sittig et al. (2020). The event-wise evaluation is presented here again, this time with the conservation tillage practice and, for comparison, with the simulations of the complete season.

Trials under consideration in this study

The field trials with maize cultivation (see Table 1 for details) were conducted on the Bayer “ForwardFarm” in Huldenberg, Belgium, in 2018, 2019, and 2013. The aim was to compare runoff and erosion under conventional tillage and conservation tillage with or without the installation of micro-dams. After each runoff event, samples were taken for the quantification of the complete runoff water volume and erosion mass from each subplot.

TABLE 1. Details of the maize field trials under investigation, conducted on the Bayer ForwardFarm in Huldenberg (Belgium) Trial 2018 Trial 2019 Trial 2013 Devices “ERuiStop” drum plow for micro-dams, “Micheltand” for conservation tillage Micro-dams: disc plow, drum plow, “Micheltand” for conservation tillage Soil type Sandy loam (clay 15%, silt 74%) Sandy loam (clay 15%, silt 74%) Irrigation No No Plot length (m) 24 18 24 Plot area (m2) 72 54 72 Slope (%) 9 9, 16

In the 2018 and 2019 trials, micro-dams were applied to both conventionally tilled fields and to conservation tillage using the “ERuiStop” micro-dam device from LSM (Jurbise, Belgium; Figure S1). The resulting pattern on the maize field is shown in Figure 1. The earthen micro-dams had a rectangular shape of 40 times 20 cm, opposite to the slope, and a depth of 12 cm. They were spaced approximately 15 cm apart along the furrow. The “ERuiStop” anti-erosion drum is a further development from the one used in the 2013 trial and is commercially available. In the 2013 trial, both a disc and a drum plow were applied to create micro-dams of similar dimensions on conventionally tilled fields (images of the devices and the resulting patterns on the field are shown in Sittig et al., 2020).

image

Resulting pattern of micro-dams on the field (2018 and 2019 trials)

Figure 2 shows the historical field that had been treated since 1997 with conventional tillage or conservation tillage on different subplots. In all trials reported here, the so-called “Micheltand” from Steeno (Vichte, Belgium; Figure S2) was applied for conservation tillage. This device follows the principle of “subsoiling,” as defined in Alletto et al. (2010) in which the soil surface remains nearly covered with crop residues, and the soil is just lifted and the (sub)soil aerated. Hence, the whole surface is treated.

image

Historical field on the Bayer ForwardFarm in Huldenberg, undergoing conventional tillage (left) and conservation tillage (right) since 1997

Numerical evaluation of runoff and erosion Theoretical background

The implementation of the USDA runoff CN concept (USDA, 2004) was conducted as described in Sittig et al. (2020). In short, the runoff CN relates precipitation events with corresponding runoff under the assumption of a certain fraction of precipitation that infiltrates immediately (Hortonian runoff).

Erosion is often quantified with the universal soil loss equation (USLE; Wischmeier & Smith, 1960). It was designed to calculate annual amounts of soil erosion and corresponding sediment yields from a watershed. The basis is a rainfall distribution. The application of the USLE to small plots and the calculation of event-wise erosion results in large uncertainties. To overcome these, modified versions of the USLE were developed. For example, the MUSS equation (Williams, 1995) uses runoff amounts as direct input and is designed for small watersheds.

The model PRZM uses MUSS to quantify erosion (US Environmental Protection Agency [USEPA], 2014): urn:x-wiley:15513777:media:ieam4546:ieam4546-math-0001(1)where Xe (t/day) is the event soil loss, Vr (mm) the volume of daily runoff event, qp (mm/h) the peak storm runoff rate, A (ha) the field size, K (-) the soil erodibility factor, LS (-) the length-slope factor, C (-) the soil cover factor, and P (-) the conservation practice factor.

Many of the factors in Equation (1) stem from other sources and are not taken from the field plots under consideration. The dimensionless factors K, LS, C, and P are listed in handbooks (Stone & Hilborn, 2015; US Environmental Protection Agency [USEPA], 1978) or defined in the corresponding FOCUS guidance documents (FOCUS, 2001, 2015), respectively.

The C-factor is a combined measure that contains the influence of crop type and tillage method by multiplying these factors (Borrelli & Panagos, 2020). The crop type and tillage method used by default are related to a continuously fallow and tilled land. Hence, the objective is to derive relative numbers to compare different cropping and tillage systems (Stone & Hilborn, 2015). Panagos, Borrelli, Meusburger, Alewell, et al. (2015) present the combined effect that is considered for the estimation of the then called “management” C-factor. They describe it as the product of the influences of the crop type, tillage, residue management, and soil cover. In this framework, as applied in the work herein, the effect of micro-dams and/or conservation tillage were considered in the factor as being summarized in the concept of tillage.

Parameter estimation—Two alternative strategies

First, an event-based calculation of the runoff CN was conducted as described in detail in Sittig et al. (2020). Briefly, the reported collected rainfall and the runoff water amounts were used to calculate the runoff CN. To arrive at a meaningful average value, the events were weighted with the corresponding amount of precipitation.

Second, the complete seasons were simulated with PRZM and values for CN and the C-factor from the MUSS equation were derived. This approach has the advantage that environmental conditions such as weather, soil texture, and daily soil moisture are inherently considered.

Simulations with PRZM SW v4.3 in an automated calibration mode were conducted. The properties of the trial sites were parameterized in the PRZM input files (.inp; see Supporting Information). To this end, these files were set up following the PRZM manual (USEPA, 2006) with the appropriate MUSS parameters (except for C), according to the USLE fact sheet (Stone & Hilborn, 2015) to: K = 0.12 (sandy loam; OM > 2%); LS = 1.05 (trial of 2018), 0.90 (2019), 2.25 (2013: 16% slope), 1.05 (2013, 9%); p = 1. Additionally, the field sizes (e.g., 0.0054 ha), and slopes (9% or 16%, respectively) were required. The precipitation data were taken from stations near the trial sites for the 2018 and 2019 trials. In the study reports, rain amounts were given cumulatively (i.e., as they were gathered in the reservoirs) after an event of runoff and (if applicable) erosion. For the simulations, these reported amounts were adjusted with the information from the weather stations to better represent the actual temporal distribution of rain. The remaining meteorological data were obtained from the Meteorological Archival and Retrieval System (MARS), Grid No. 101094 (European Commission [EC], 2020b).

In a first step, the CN most suitable for the measured total annual runoff amounts were inferred. To this end, simulations with varying values for CN were conducted and the value fitting best was deducted. Subsequently, values for the MUSS C-parameter were derived analogously to best match the measured annually eroded masses. Additionally, plots hitting the annual sums of runoff water and eroded sediment, together with their dynamics, were drawn.

Example calculations of PECsw, PECsed, and EECs

The standard scenarios were adapted in terms of CN and MUSS C-factor following the findings in this evaluation. For the European risk assessment, the simulations were conducted with the FOCUS standard scenarios R1 to R4 (FOCUS, 2001, 2015); the PECs considering mitigation were compared with the PECs from a corresponding regular assessment. For the US assessment, the Illinois corn scenario was taken.

Three substances were chosen for example calculations: FFA, TCM, and IMS, which represent substances of relative immobility (FFA: Kom = 125 L/kg; DT50,water = 49.6 days, DT50,soil = 18.3 days), intermediate mobility (TCM: Kom = 48 L/kg; DT50,water = 26.1 days, DT50,soil = 11.6 days), and high mobility (IMS: Kom = 19 L/kg; DT50,water = 19.8 days, DT50,soil = 2.7 days).

The concentrations of the European PECsw for the stream water body were evaluated as described below. The remaining values (i.e., PECsw for the pond scenario, PECsed, and EECs) were taken unchanged from the model outputs.

Procedure for PECsw stream scenarios

The PECsw values according to FOCUS and calculated with the coupled models PRZM/TOXSWA need to be further processed if in-field mitigation measures such as micro-dams and conservation tillage in combination with a stream scenario are considered (in the pond scenario, no modification is required). In this scenario, a 1-ha field is treated with a pesticide and equipped with a mitigation measure. The exposure endpoint is the concentration in surface water (and sediment) in an edge-of-field stream adjacent to this very field. Further dilution of the surface water is provided by baseflow from an upstream catchment that is only partially treated and equipped with the mitigation measure under consideration. Therefore, a post processing of the TOXSWA results must be conducted. Figure 3 shows the recalculation process, using Equations (2)–(5) below.

image

Procedure to calculate PECsw,maxmitigated for the FOCUS stream scenarios using FOCUS PRZM/TOXSWA after the application of micro-dams and/or conservation tillage

The procedure is based on the concept of the upstream catchment, in which only 20 ha of the 100-ha large area receive mitigation measures (and pesticide application). This leads to a greater dilution of the edge-of-field PECsw as if the unprocessed output were directly considered (which would assume that the entire upstream catchment was mitigated as well). The strategy is in accordance with the handling of VFS as a mitigation measure in the context of the FOCUS landscape and mitigation framework (FOCUS Step 4; FOCUS, 2007), where only the treated field next to the water body and 20% of the upstream catchment have a VFS (ter Horst et al., 2009).

The fractional reduction in runoff (pesticide) mass flux (RFLX; from the PRZM output [g/cm2/day]) by the mitigation is calculated with: urn:x-wiley:15513777:media:ieam4546:ieam4546-math-0002(2)whereas the fractional reduction in runoff volume (RUNF; from the PRZM output [L/m2/day]) is given by: urn:x-wiley:15513777:media:ieam4546:ieam4546-math-0003(3) The relationship between the mitigated and the unmitigated PECsw (mg/L) is defined by: urn:x-wiley:15513777:media:ieam4546:ieam4546-math-0004(4)where fa(-) is the fraction area treated = 0.208 (default FOCUS stream). The calculation of urn:x-wiley:15513777:media:ieam4546:ieam4546-math-0005 in an adaptation of Adriaanse et al. (2017) is conducted with: urn:x-wiley:15513777:media:ieam4546:ieam4546-math-0006(5)where n(-) is the number of hours per day with simulated runoff (for calculation, the total precipitation per day must be divided in n 2-mm steps); BASF (L/m2/day) the baseflow; default = 0.2 (approximate median of FOCUS R stream scenarios); raindr (-) the fraction of INFL going into base flow (TOXSWA input); user-defined: default = 0.1 (FOCUS definition); INFLMON (L/m2/day) the monthly average of INFL for the corresponding month (from PRZM output).

This procedure constitutes a refinement of the one presented in Sittig et al. (2020), in which the fractional reductions of runoff mass flux fr and the runoff volume fv were assumed to be identical. It is demonstrated later that those two variables have similar values.

Example calculation: The global maximum PEC for the R1 scenario (TOXSWA output) is 2.72 µg/L, occurring on 7 May 1984. Reducing the CN in the PRZM input leads to a global maximum on 20 May 1984 of 1.34 µg/L. This PEC was calculated under the intrinsic consideration of reduced runoff via micro-dams in the complete scenario, including the entire upstream catchment. Therefore, the correction is required to assume only mitigation on the proportion of the upstream catchment that receives pesticide application in compliance with the defined scenario conditions.

The runoff fluxes or runoff volumes in the regular or modified scenario are 1.3E–2 and 4.6E–5 mg/m2/day and 1.6 and 0.2 L/m2/day, respectively. This leads to values for the fractions fr and fv of 0.997 and 0.89. Using Equation (5), a PECsw for the standard scenario on 20 May 1984, with the precipitation on that day being 19.9 mm (i.e., n = 10) and the monthly average of INFL in May 1984 of 1.3 L/m2/day, a PECsw of 1.61 is calculated. Inserting this into Equation (4) leads to a PEC with mitigation of 6.82E–3 µg/L.

RESULTS AND DISCUSSION Effects of micro-dam application and conservation tillage on the measured quantities and the calculated runoff CNs in the event-based evaluation

Table 2 lists the effects of micro-dams and conservation tillage. The measures led to a decrease in runoff from the fields, which is reflected in lower runoff CNs (here: calculated for each event). Consequently, following the reduced runoff water amounts, the eroded sediment quantities were reduced. Table S1 (2018 and 2019 trials) and Table S2 and Table S3 (trial of 2013) in the Supporting Information provide the corresponding experimental data, together with the event-wise derived runoff CN. Table S4 and Table S5 in the Supporting Information list the erosion amounts for the 2018, 2019, and 2013 trials, respectively.

TABLE 2. Reductions of runoff volumes, erosion quantities, and derived runoff curve numbers (CN; means, calculated event-wise) after the application of micro-dams (MD) and/or conservation tillage (CsT) compared with conventional tillage (CvT) CvT CsT Red. CsT versus CvT Red. MD + CsT versus CvT Red. MD + CsT versus MD + CvT Red. CsT versus MD + CvT No MD MD Red. (%) No MD MD Red. (%) (%) (%) (%) (%) 2018 Runoff (mm) 7.3 4.2 43 1.2 0.7 46 83 91 83 71 CN 75 72 4.0 67 66 1.9 11 13 8 7 Erosion (kg/ha) 2371 1046 56 62 34 43 97 99 97 94 2019 Runoff (mm) 9.7 3.3 66 3.0 1.5 49 69 84 55 9 CN 73 70 4.4 66 65 1.9 10 12 7 6 Erosion (kg/ha) 6655 1203 82 599 204 66 91 97 83 50 2013 (slope 16%) Runoff (mm) 7.3 5.5/2.3a 24/68a 0.78 NA NA 89 NA NA 86/66a CN 69 68/63a 1.6/8.6 63 NA NA 8.6 NA NA 5/0a Erosion (kg/ha) 1030 480/260a 54/75 17 NA NA 98 NA NA 96/93a 2013 (slope 9%) Runoff (mm) 4.8 2.5/1.4a 47/71a 0.62 NA NA 87 NA NA 75/56a CN 67 66/64a 2.1/3.9a 63 NA NA 5.9 NA NA 3/1a Erosion (kg/ha) 420 170/80a 59/81a 17 NA NA 96 NA NA 90/79a Mean reductions CN points % MD + CvT versus CvT (n = 6) 3 (±1.7) 4.0 (±2.4) MD + CsT versus CsT (n = 2) 1 (±0.0) 1.5 (±0.0) CsT versus CvT (n = 4) 6 (±1.5) 8.7 (±1.7) MD + CsT versus CvT (n = 2) 9 (±0.5) 11.5 (±0.5) MD + CsT versus MD + CvT (n = 2) 6 (±0.5) 7.7 (±0.6) CsT versus MD + CvT (n = 6) 3 (±1.9) 4.4 (±2.7) Notes: Mean reductions are calculated over all corresponding experimental setups where applicable, standard deviations are given in brackets. aFor disc and drum plow micro-dam creating technology, respectively. Abbreviations: NA, not applicable; Red., reduced.

Generally, conservation tillage led to larger decreases in both runoff and erosion than micro-dams: The overall mean reductions (and standard deviations) by micro-dams versus conservation tillage were 53% (±17%) versus 82% (±17%) for runoff and 68% (±12%) versus 96% (±3%) for erosion.

Quantitative runoff

In the 2018 trial, in one of the runoff events (on 24 May), the reservoir for the collection of runoff had overflown; this event was not considered in the further evaluations for both runoff and erosion. On another date, runoff was observed from the control plot only (0.24 mm on 30 August)—both micro-dams and conservation tillage completely suppressed runoff generation. On other occasions, micro-dams or conservation tillage also completely stopped runoff.

On 14 August 2018, a collected rainfall event of 26 mm led to a runoff event of 4.38 mm (in the control plot), whereas on 2 May 2018, 34 mm of rain led to 0.58 mm runoff only. Due to the evaluation methods, these divergences in the observations were not considered here, because both the general condition of the micro-dams and differences in rain intensities could be the reason, and neither aspect was considered in this assessment.

In the 2019 trial, the first five reported rain events (with a collected rain amount up to 20 mm on 6 May) did not initiate runoff generation. This was presumably because of the relatively new condition of the micro-dams because, for example, a rainfall of only 11 mm on 3 September led to a runoff of 0.57 mm on the control plot.

Generally, over all trials, the application of micro-dams reduced the amount of runoff considerably—between 24% (2013 trial, usin

留言 (0)

沒有登入
gif