Modeling Tillage Effects on Plant-Available Water by Considering Changes in Soil Structure

Sayantan Samanta^1,2, Dianna Bagnall³, Srinivasulu Ale^2,*, Cristine L.S. Morgan³, Christine C. Molling³

Published in Journal of the ASABE 67(3): 589-599 (doi: 10.13031/ja.15695). Copyright 2024 American Society of Agricultural and Biological Engineers.

¹    Texas A&M University, College Station, Texas, USA.

²    Texas A&M AgriLife Research and Extension Center, Vernon, Texas, USA.

³    Soil Health Institute, Morrisville, North Carolina, USA.

^*    Correspondence: sriniale@ag.tamu.edu

Submitted for review on 4 June 2023 as manuscript number NRES 15695; approved for publication as a Research Article and as part of the Regenerative Agriculture Collection by Community Editor Dr. Meetpal Kukal of the Natural Resources & Environmental Systems Community of ASABE on 12 January 2024.

Highlights

• No-tillage improved plant-available water (PAW) as compared to conventional tillage, especially in dry years.
• Soil structure showed 4.5 times higher impact on PAW than surface disturbance due to tillage.
• Consideration of changes in soil structure is important in modeling the effects of soil management.
• PALMS was found to be an effective model to assess the effects of changes in soil management on soil water.

Abstract. Management practices such as no-tillage (NT) have the potential to provide many benefits, such as reduced runoff and soil erosion and increased infiltration and soil water holding capacity. Most hydrological models that are used to simulate the effects of soil management are built based on empirical relationships between management and hydrology outcomes, and they tend to ignore or oversimplify the effects of soil structure. However, soil structure is management dependent and is a driver of water movement and storage in soil. The goal of this study was to better understand the effects of differences in soil structure between NT and conventional tillage (CT) on field-scale hydrology and plant available water (PAW). This study employed in-field measurements of soil structure in NT and CT fields in the Texas Blackland Prairies and used the Precision Agricultural-Landscape Modeling System (PALMS), which can simulate the effects of differences in soil structure. Regression analysis was performed on simulated soil water to understand the relative contributions of variations in surface roughness and macropore properties due to tillage on PAW. Results from this study showed that NT accumulated 44.8, 20.4, and 5.7 cm more PAW than CT in the top 150 cm of the soil profile during the summer growing season in the years 2006, 2008, and 2011, respectively, all of which encountered considerable dry spells. It was also found that the changes of soil structure due to tillage had about 4.5 times more impact on PAW than surface roughness. This study highlights the benefits of adopting NT over CT and showcases the importance of considering soil structure in modeling the effects of soil management on PAW.

Keywords.Macropores, No-till, Preferential flow, Surface disturbance, Surface roughness, Texas Blackland Prairies.

Regenerative agricultural practices that promote soil health are receiving wider attention because of their potential to mitigate climate change through soil carbon sequestration (O’Donoghue et al., 2022; Newton et al., 2020). Soil health-promoting practices such as no-tillage (NT) can help crops build climate resilience through improved soil water balance, specifically by increasing water storage and infiltration (DeLaune et al., 2016; Hansen et al., 2012) and reducing runoff (DeLaune and Sij, 2012) and evaporation (Baumhardt and Lascano, 1999; Mellouli et al., 2000; Baumhardt et al., 2013). Schwartz et al. (2010) reported that tillage increases evapotranspiration by reducing soil albedo and increasing the absorption of radiant energy. In contrast, NT increases soil albedo due to the presence of surface residue, thereby reducing evapotranspiration and increasing soil water content (De Vita et al., 2007). It has also been found that the bulk density of soils can be lower and macropores can be larger under NT systems when compared to conventional tillage (CT) systems (Alhameid et al., 2019; Bordovsky et al., 1999). This improvement in soil physical characteristics with the change in management from CT to NT is more prominent in soils susceptible to surface sealing (Ramos et al., 2019). Blevins et al. (1971) reported that crops can sustain periods of short-term droughts under an NT system because of greater plant-available soil water storage resulting from reduced evapotranspiration and higher infiltration.

Hydrologic models are useful to simulate the long-term effects of soil management on water balances rapidly and inexpensively in comparison to field experiments. After a thorough evaluation based on measured data from field experiments, they provide insights into the physical processes associated with the system being modeled (Epstein, 2008) and enable assessment of the effects of changes in soil management. However, modeling efforts to quantify the effects of soil management can fall short when the models do not represent how soil health-promoting practices affect soil functioning. Some common physical indicators of soil health related to soil water benefits are soil water holding capacity, plant-available water (PAW, which is the difference between daily soil water content and the permanent wilting point), infiltration, bulk density, porosity, and saturated hydraulic conductivity (Basche and DeLonge, 2019; Stewart et al., 2018). A better quantification of the effects of soil health promoting practices on biophysical processes of carbon, nutrients, and on-farm and off-farm water cycling (Schulte et al., 2019; Staes et al., 2018) is essential for understanding the potential benefits of NT over CT.

Conversion from CT to NT is known to improve soil structure (Arai et al., 2014; Bagnall and Morgan, 2021), but not many hydrologic models explicitly address changes in soil structure due to changes in soil management (Fatichi et al., 2020; Lepore et al., 2009) and associated variation in preferential flows. Some hydrologic models, such as HYDRUS (Simunek et al., 1998), the Soil Water Atmosphere Plant (SWAP; Kroes et al., 2017), the Root Zone Water Quality Model (RZWQM; Hanson et al., 1998), and the Precision Agricultural-Landscape Modeling System (PALMS; Molling et al., 2005), have the ability to simulate preferential flow. However, HYDRUS, SWAP, and RZWQM models are often difficult to parameterize (Larsbo and Jarvis, 2005), and some macropore related parameters in these models lack field-observable physical meaning. As such, inverse modeling is the only way to simulate preferential flow accurately. In contrast, PALMS simulates macropore/preferential flow by representing soil structure with cubic peds. PALMS treats the intersection of the cubic peds as macropores (Beven and Germann, 2013), assuming that the soil structural cleavage planes act as macropores (Lepore et al., 2009) to facilitate the preferential flow of water through the geometry of the planes. PALMS is also capable of modeling agricultural processes of interest such as tillage, planting direction, and fertilizer and manure applications. PALMS does not directly simulate the changes in soil structure due to changes in soil management, but it uses robust process-based components for surface disturbances caused by tillage and allows the implementation of differences in soil water redistribution within the profile when the model is parameterized with the structural differences associated with the management practice.

The primary focus of this study was to assess the effects of tillage practices on PAW using PALMS by considering changes in soil surface roughness and soil structure due to changes in tillage. The first objective was to evaluate the performance of PALMS in simulating volumetric water content (VWC) while considering the differences in soil structure to answer the research question, “Can PALMS accurately simulate VWC based on field measurements?” Measured soil structural properties from a past study (Bagnall and Morgan, 2021) were used to parameterize a perennial grassland plot and to assess PALMS’ simulation efficiency. The second objective was to quantify the on-farm variability of PAW due to tillage management. An important research question that this objective addresses is, “How much additional water can a soil hold if a farmer switches from CT to NT practice?” The CT and NT models were developed to assess the variability in PAW in dry, normal, and wet years. The third objective of this study was to quantify the relative importance of soil structure and surface roughness in modeling the effects of soil management practices. This objective addresses the research question, “What is the relative importance of soil structure over surface roughness in simulating PAW?” The individual contributions of measured soil structure and simulated surface disturbances were assessed utilizing PALMS’ ability to simulate variation in PAW due to differences in surface roughness and soil structure due to tillage.

Materials And Methods

Model Description

PALMS is a field-scale agricultural model that is intended for use in situations where the robustness of physical processes is of utmost importance. This rainfall-runoff biophysical model is a combination of a 2D diffusive wave model and a 1D point-column model, the Integrated Biosphere Simulator (IBIS) (Kucharik et al., 2000). The diffusive wave model in PALMS is activated when precipitation exceeds the infiltration rate and water is routed as runoff or runon (Molling et al., 2005; Nelson et al., 2013). PALMS operates on a three-dimensional grid-based approach and simulates pedon-scale water transport processes such as surface detention, infiltration, runoff, runon, plant water uptake, and subsurface drainage, using both topography and soil horizon information. Currently, PALMS can simulate maize (Zea mays L.), soybean (Glycine max Merr.), and cotton (Gossypiumhirsutum L.) crops (Molling et al., 2005; Booker et al., 2015). It employs the Mesopore and Matrix (M&M) module, which uses a combination of Hagen-Poiseuille law for water flow through macropores and Darcy’s law for moving the water from macropores to the soil matrix. Some important aspects of PALMS are its intricate adoption of process-based components, physically meaningful parameters, and ease of parameterization, which in general eliminates the need to calibrate the model.

Representation of Changes in  Soil Management in PALMS

We used two methods to represent management changes in PALMS. The first method involved selecting management practices within the PALMS settings file in accordance with those reported at the site. For example, tillage events which consequently altered surface roughness and Manning’s coefficients were included following the type of tillage implement used. Surface roughness was altered in PALMS based on the date of tillage, the type of tillage equipment, and the angle of tillage. The second method of representing soil management in PALMS involved the use of the M&M subroutine. The M&M subroutine was parameterized using field-measured soil structure sizes to a depth of 30 cm for each management (CT and NT).

The diffusive wave model in PALMS uses Manning’s equation to relate the depth of surface water flow to the discharge rate. There are two types of roughness affected by tillage events: random roughness, or non-directional roughness, and oriented roughness. Both random and oriented roughness affect depression storage, which is a critical aspect of runoff-infiltration partitioning. Runoff in PALMS begins only when the depression storage is full. PALMS uses equation 1 reported in Onstad (1984) to define depression storage as follows:

        (1)

where S_D is the depression storage (cm), R_R is random roughness (cm), and S_L is the slope (percent). This is valid for slopes up to 12%, and for tillage parallel to the direction of the slope. However, this relationship was modified (eq. 2) in PALMS to account for tillage.

        (2)

where

S_DA = modified depression storage affected by tillage angle (cm)

C_AI = anisotropic roughness coefficient (dimensionless; Molling et al., 2005)

A_C = angle between the axis of tillage direction and a line perpendicular to the slope.

The coefficients for directional and oriented roughness as well as Manning’s coefficient for different tillage implement types are obtained from published values (Engman, 1986; Molling et al., 2005; Zobeck and Onstad, 1987).

PALMS also has a component for soil surface reconsolidation, which is a function of time and rainfall (eqs. 3 and 4). The random roughness and Manning’s coefficients are smoothed by precipitation as follows:

        (3)

        (4)

where

R_RP = modified random roughness (cm)

R_R = original random roughness for a freshly tilled soil (cm)

P_tot = accumulated precipitation (mm) since the last tillage event

n_p = modified Manning’s roughness coefficient

n = original Manning’s roughness coefficient (Molling et al., 2005).

The M&M infiltration approach (Lepore et al., 2009) uses a preferential flow routine alongside matrix flow or two-domain flow. Infiltration takes place through the slits between imaginary cubic peds and through the six faces of the cubic peds. The modified Hagen-Poiseullie law (Bird et al., 2006; eq. 5) for flow through a planar slit is used to describe this unsaturated laminar flow in the M&M model.

        (5)

where

Q_mp = flow rate through the mesopores (m s^-1)

B_ped = half of the mesopore slit width

w_ped = width of the ped

?_w = density of water (kg m^-3)

µ = viscosity of water (kg m^-1 s^-1)

g = acceleration due to gravity (9.81 m s^-2).

The w_ped is a function of soil water content (?) and depth below the ground (z) and is defined as (eq. 6).

        (6)

where

w_ped,0 = ped size at the surface (m)

w_ped,max = maximum ped size in the profile (m)

h = slope of curve (-)

c = depth at which the ped size is exactly halfway between minimum and maximum (m).

When standing water is not present on the soil surface, Richards’ equation runs on quarter-hourly timesteps to redistribute water in the soil and accounts for matrix storage and root water uptake. PALMS uses a root water uptake model that draws water from layers where moisture is available and is limited by average plant stress (Molling et al., 2005). When water stands on the soil surface, the slits between the cubic peds allow movement of water through them and wet the surfaces of the ped to allow matrix flow. When the water application rate on the surface is faster than the flux through the mesopores, the excess water is available for runoff, which is then available to infiltrate or contribute to depression storage in neighboring modeled units. PALMS can also be operated without the macropore/preferential flow component by using the Green & Ampt method (1911) for infiltration. Both cases use Richard’s equation for water movement, but when the M&M model is on, water moving in the macropores becomes a source for all layers of the soil profile, whereas the Green & Ampt method (1911) only supplies soil water to the surface layer. As a result, when the M&M model is used in PALMS, water is distributed quickly throughout the profile, and soil wetting behavior is less Darcian (Lepore et al., 2009; Bagnall et al., 2019).

Study Area and Parameterization of PALMS

Two experimental fields located at the Grassland Soil and Water Research Laboratory (GSWRL) at Riesel, TX, in the Blackland Prairies Major Land Resource Area were selected for this study (fig. 1). The SW17 field was a 1.2-ha perennial grassland that was managed for improved grasses (e.g., coastal Bermuda grass) and was rotationally grazed by cattle. We selected this field due to the availability of measured VWC data (Bagnall et al., 2019) for validation of PALMS’s ability to simulate VWC. The second field, W13, was a 4.7-ha conventionally tilled field where corn (Zea mays L.) was planted during the 2008-2009 period, and this field was selected to model the effects of tillage practices. The PALMS model requires topography, vegetation type, and soil profile information consisting of the horizon depths and soil texture in each of the horizons. Topography information was taken from a 1-m gridded digital elevation model that was developed as a part of the United States Geological Survey (USGS) lidar data collection program in 2017. We used this data because of the unavailability of a high-resolution digital elevation model for the simulation period. The lidar data was collected between January and March, when there was no crop on the ground, and hence surface topography might have been captured very well. The downloaded elevation model was resampled to 10 m resolution and smoothed with a lowpass filter to remove any noise.

Figure 1. Study Area: USDA ARS Watershed in Riesel, TX. Field W13 is a conventionally tilled cornfield, and Field SW17 is a perennial grassland.

To parameterize soil structure and soil properties, a total of 18 observations from production agriculture fields in three counties across the central Blackland Prairie Major Land Resource Area in Texas (Bagnall and Morgan, 2021) were used. These fields represented six CT, six NT, and six perennial grasslands. For each field observation, three rectangular pits were dug to expose the soil profile to a depth of 40 cm. The pits were dug with a pickaxe and shovel to expose the A horizon and any horizon directly below it. Care was taken not to walk on the pit side that would be described, to avoid compacting the soil. After each pit was dug, it was left to dry for at least 24 h and then picked. Picking removed shovel marks to expose structural units. The picking method used is outlined in Haddad et al. (2009), though the pit faces were not vacuumed. While picking out the structure, care was taken not to leave any knife or other artificial marks on the face. Larger structural units were exposed with a knife, and smaller ones with a dental pick. If an extensive root system was present, roots were burned with a propane self-igniting torch. Further details about the structure measurements can be found in Bagnall et al. (2020). One composite soil sample per field representing 0 to 10 cm depth and one composite soil sample representing 10 to 30 cm depth were analyzed for particle-size distribution (Gee and Bauder, 1986). For horizons deeper than 30 cm, the official soil characterization data of Houston Black clay (Fine, smectitic, and thermic Udic Haplustert) was used. Sub-daily weather data (hourly air temperature, relative humidity, wind velocity, and 15-minute interval rainfall amount) and land management data were obtained from the United States Department of Agriculture – Agricultural Research Service (USDA-ARS) hydrologic database.

The M&M parameters used in this study were estimated using a combination of measured and soil survey manual-reported (National Cooperative Soil Survey, 2018) ped sizes of the soil profile. The soil structure was measured in three soil pits in each field to a depth of 30 cm. Soil structure descriptions beyond 30 cm depth were based on the soil survey manual and are management independent. This combination approach was used in this study because PALMS requires ped size information for the entire soil profile. The use of this approach is justified because the effect of tillage management is most prominent only in the plough layer (approximately 20-30 cm depth). For each management practice, the 75^th percentile of the measured ped sizes was used (table 1). The values used for parameters w_ped,0, w_ped,max, c, and h were obtained by fitting the measured ped sizes along the soil profile to equation 6 (table 2). Figure 2 compares the vertical distribution of ped sizes calculated by fitting the measured ped sizes across the soil profile to equation 6. It is to be noted that the changes in soil structural properties are a result of the long-term adoption of the respective management practices. The simulated scenarios in this study used representative soil structural properties from the Blackland Prairies.

Model Evaluation

Bagnall et al. (2019) reported an acceptable simulation of VWC and infiltration by the model in the study region without any calibration. To gain further confidence in the model, we evaluated the performance of the model in simulating VWC at 20, 40, 60, and 80 cm depths in the SW17 field at the GSWRL by comparing the simulated values with measured values (Harmel et al., 2014) during the years 2008 and 2009. Volumetric water content was measured using a neutron moisture meter (TH2O portable soil moisture probe; Delta-T Devices) at five different locations in the field at roughly two-week intervals between July 2008 and December 2009. Details about the calibration of the moisture meter can be found in Bagnall et al. (2019). The coefficient of determination (r²) and the root mean squared error (RMSE) were used to evaluate the model's performance in predicting VWC. The performance of the model could not be assessed on CT or NT plots due to the lack of measured VWC data from these plots. Since the primary goal of this study was to quantify the difference in PAW due to a change in tillage management, the change in PAW is more important than the absolute values of PAW under each management.

Figure 2. Comparison of PALMS (Precision Agricultural-Landscape Modeling System) generated and measured (OBS) soil ped size distributions along the soil profile for perennial grassland (P), conventional tillage (CT), and no tillage (NT) managements.

Simulation of CT and NT Scenarios

The evaluated PALMS model was used to simulate PAW and other water balance components for the CT and NT scenarios over six years, from 2006 to 2011, for the W13 field at GSWRL. The field operations performed in W13 in 2008 (table 3) were applied to all years for the CT model to quantify the effects of tillage management on PAW (i.e., by eliminating the effects of changes in other management practices/field operations on PAW). In the case of NT, planting and harvest are the only field operations, and they were implemented on the same dates as CT. The soil texture was kept the same for CT and NT models based on the underlying assumption that soil texture remains unaffected by management, but the structural information was changed based on soil structure measurements from CT and NT fields in the Blackland Prairies (table 2). Soil peds were larger at the surface under CT, whereas the ped sizes under NT were relatively smaller at the surface and reached the maximum size at about a depth of 0.8 m (fig. 2). This is because of the presence of large clods at the surface that are formed by mechanical disturbance under CT management.

Simulated PAW was summed up to a depth of 150 cm and was used to assess the effects of switching management from CT to NT. The effect of tillage on PAW was analyzed for different weather conditions by classifying the simulation period of 2006 to 2011 into dry, normal, and wet years based on the Standardized Precipitation Index (SPI), which is a meteorological drought index estimated from daily long-term precipitation data (McKee, 1995; McKee et al., 1993). The SPI uses daily precipitation data for long periods (30 to 50 years), which are transformed into a normal distribution to calculate the number of standard deviations by which the precipitation deviates from the long-term mean. The value of this index ranges between -2 (extremely dry weather) and +2 (extremely wet weather) (Amrit et al., 2019). In this study, a 6-month SPI was calculated using 30 years of daily precipitation records (between 1960 and 1990) to classify the years 2006 and 2011. The selected years covered a range of wet and dry conditions and were used to understand the effects of the adoption of NT in these conditions on PAW. The years classified as extreme wet and severe wet years based on SPI were grouped as wet years; extreme dry and severe dry years were grouped as dry years; and moderately wet, near normal, and moderately dry years were grouped as normal categories to reduce the number of classes.

Evaluating the Impacts of Interpedal  Macropores (Soil Structure) on  Modeling Outcomes

The relative contributions of (1) surface roughness caused by tillage events and (2) macropore properties (driven by measured ped size differences) to simulated PAW were also quantified using the evaluated model. PALMS uses Manning’s coefficient approach to alter the soil surface (roughness) upon tillage, which affects the runoff-infiltration partitioning directly. On the other hand, tillage also destroys the soil structure in the upper horizons of the soil profile, which is reflected by the changes in ped size and the resulting macropore parameters in PALMS (table 2, fig. 2). To quantify the individual contributions of surface roughness and macropores on differences in PAW due to differences in soil management, four scenarios were developed and run in the PALMS model (table 4). Each scenario includes combinations of surface disturbance (represented by surface roughness parameters) and macropore parameters for both CT and NT. Simulations S-1 and S-4 reflect realistic CT and NT scenarios, respectively, whereas S-2 and S-3 are hypothetical scenarios run with interchanged surface and macropore properties to isolate the effects of surface roughness and macropore size on PAW. These hypothetical scenarios enabled us to estimate the difference in PAW within the extremities of the parameter space, specifically among the conditions when the variations between the parameter values were the highest.

Simulation groupings (table 5, fig. 3) were used to demonstrate how surface roughness and macroporosity contribute to water capture and storage, both individually and in combination. The first two simulation groupings, 1 and 2, illustrate the contribution of macropore parameters to changes in PAW. Groups 1 and 2 consist of scenarios that have constant surface roughness parameters, but macropore parameters vary between CT and NT (table 2). For example, group 1 has CT surface roughness parameters and macropore parameters of CT and NT; group 2 holds surface roughness consistent for NT. Groupings 3 and 4 each hold consistent macropore parameters, while surface roughness varies between CT and NT. Groups 3 and 4 show the contribution of roughness parameters from surface disturbances. Simulation group 5 alters both roughness and macropore parameters due to changes in tillage management. The metric used for comparison between groups is PAW, calculated for 0 to 150 cm soil profile.

Figure 3. Schematic representation of the five scenario groups from table 5. The notation of either CT (light color) or NT (dark color) at the top of the profile represents the surface roughness properties. The notation of either CT or NT at the bottom of the profile represents the soil structural properties.

A multiple linear regression was used to determine whether surface roughness or macropore parameters had a relatively greater impact on PAW. The dependent variable in the regression was the change in PAW due to the combined effect of macropores and surface disturbance. We used two independent variables as predictors in the regression: (1) the difference in PAW due to differences in surface disturbance, and (2) the difference in PAW due to differences in macropore properties. Both coefficients for the independent variables were statistically significant (p-values <0.0001), and so their magnitude indicated the relative contribution of surface disturbance and macropore properties on predicted PAW.

Results And Discussion

PALMS Model Performance Evaluation

PALMS performed well in predicting VWC at depths of 20, 40, 60, and 80 cm during the years 2008 and 2009 (fig. 4). An overall r² of 0.71 and an RMSE of 0.05 m³ m^-3 (table 6) indicate a good fit of the simulated data to the measured data. The shallow layers showed greater r² values and smaller RMSE, indicating better performance of the model for shallow layers than the deeper layers. This trend is consistent with that of Bagnall et al. (2019), as the model drains water out of the soil profile too quickly for it to move from the macropores to the soil matrix. The PALMS model was previously evaluated for the Blackland Prairies of Texas by Bagnall et al. (2019) and used to assess the effect of soil structure parameters on VWC and water redistribution in the soil profile. They found that PALMS adequately predicted soil water content at multiple depths, and predictions improved when the M&M component (soil structure representation) was used.

The Effect of Tillage Practices  on Plant-Available Water

A considerable gain in PAW occurred in normal and dry years under NT compared to CT management (fig. 5). For example, simulated average PAW during the growing season in the top 150 cm of the soil profile was higher by 22.6 and 9.2 cm under NT compared to CT in dry (2006) and normal (2008, 2010, 2011) years, respectively. Averaged over all the years, NT stored 11.7 cm more PAW as compared to CT during the corn growing season. Similar PAW improvement was noted by Govaerts et al. (2007), who reported that a more stable soil structure under NT allowed greater and rapid infiltration. The most prominent improvement in PAW was seen in the first half of 2006, which was classified as a severe dry year (table 7), when NT captured 45 cm more PAW as compared to CT during that growing season. Similar outcomes were found in the years 2008 and 2011, with an improvement in PAW of 20.4 and 5.7 cm, respectively, during the growing season. For the remaining years, 2007, 2009, and 2010, no noticeable improvement in PAW was simulated. In general, during dry periods, NT held more PAW compared to CT. However, during wet conditions (2007), the effects were occasionally reversed, and NT held less PAW than CT.

Figure 4. Measured versus simulated volumetric water content (VWC) at 20, 40, 60, and 80 cm depths at the perennial grassland (SW17) during 2008 (left) and 2009 (right). The black dots represent the mean measured VWC, and the black error bars represent the standard deviation. The blue solid line represents the daily simulated VWC. The vertical gradient bars represent the magnitude of daily rainfall.

Figure 5. The difference in simulated plant-available water (0 to 150 cm) between conventional tillage and no-tillage managements under dry, normal, and wet years from 2006 to 2011 averaged under each year class. The area shaded green represents the time of year when corn is in the field, and the black radial lines indicate tillage events. The concentric circles represent the difference in PAW between CT and NT management. A positive difference in PAW indicates that no-tillage holds more water in the soil than conventional tillage.

Plant-available water is largely controlled by infiltration/runoff partitioning. In PALMS, runoff occurs when the rainfall rate exceeds the M&M infiltration rate, leading to puddling and ponding in the field. Once the puddles fill up, runoff begins, and all the water that accumulates in the puddle is available for macropore infiltration, and the excess water is subjected to runoff. Figure 6 shows the effects of changing management from CT to NT on water balance components, with a positive difference indicating that the component is greater under NT. Simulated infiltration rates increased considerably when a precipitation event occurred after tillage, thereby reducing the runoff. For example, a tillage event occurred on Mar 17^thof 2007, and it was followed by a 56-mm rainfall event on 26 March which caused 0.7 mm higher infiltration, leading to a 0.2 mm higher PAW under CT. On the other hand, when a 54-mm rainfall event occurred on June 29^th of 2009, 106 days after the last tillage event, 3 mm lower infiltration occurred under CT as compared to NT, and consequently, the PAW was higher under NT by 0.2 mm. With greater simulated infiltration after tillage under CT, more water entered the matrix, thereby increasing the soil water content. As a result, whenever there was a precipitation event immediately after tillage, PAW was higher under CT as compared to NT. From field experiments, Myers et al. (1995) and Zeimen et al. (2006) also reported that NT experienced a higher volume of runoff water as compared to CT in the early stages of crop growth due to lower infiltration. This condition is especially true in years that receive rainfall early in the growing season. This effect became less prominent when the soil surface disturbance was reconsolidated due to several rainfall events following equations 3 and 4 within PALMS.

Figure 6. Change in monthly water balance components compared to the Standardized Precipitation Index (SPI). ? PAW is difference in plant-available water between CT and NT; CT is conventional tillage; NT is no-tillage. A positive change indicates that the component has a higher magnitude under no-tillage than conventional tillage.

The years 2008 and 2011 marked some of the worst drought years in Texas. In 2008, cotton and sorghum acreages were abandoned in more than one-third of Texas, mostly in the south-central part of Texas, due to dry moisture conditions, leading to a total loss exceeding $3.6 billion (Nielsen-Gammon and McRoberts, 2009). In 2011, the loss was even more with estimates as high as $7.62 billion. In comparison to CT, NT provided an additional 20.4 and 5.72 cm of water as PAW during the growing seasons of 2008 and 2011, respectively. The extra water held by the NT system in 2008 and 2011 was equivalent to 50% and 14%, respectively, of the average (1941-2010) rainfall received during the growing season (41 cm). The average water requirement for corn is 65 cm (McDaniels, 1962), and the additional available water under NT would be beneficial for the crop to endure drought conditions and reduce irrigation needs. These results from this study indicate that NT can serve as a valuable climate change adaptive practice to sustain dryland crops under extreme weather conditions and the increasing number of droughts that are expected in the future.

Distinguishing Between the Contributions  of Macropore Size and Surface Disturbance  to Changes in PAW

The differences in PAW under CT and NT are driven by: (1) the differences in surface roughness caused by tillage events (or their absence), and (2) the differences in infiltration resulting from the differences in ped sizes under CT and NT management, which influence macropore parameters. The changes in PAW due to the changes in surface roughness and macropore parameters are reported in table 8. A positive difference in PAW indicates that NT held more water than CT, either due to surface roughness or macropore size.

Simulations indicated that macropore parameters had a greater relative effect on PAW than surface roughness. For example, when using parameters for the CT surface roughness and CT macropore sizes (Condition 1 of simulation group 3), the macropores drained out substantial water. Even when the surface disturbance was removed (e.g., surface roughness parameters changed from CT to NT, Condition 1), the difference in PAW was still negative (since CT infiltered more water into the soil after tillage). This indicated that increasing the surface roughness alone was not enough to create a positive difference in PAW. However, changing the macropore properties had a net positive effect on PAW, regardless of surface disturbance. This emphasizes an important aspect of hydrologic modeling to assess the effect of changes in soil management on soil water dynamics. These results imply that changes in both soil structure and macropores should be considered while simulating the effects of soil/tillage management on soil water dynamics. Neglecting soil structure and the macropores they create means that the effects of management practices are not adequately modeled.

A multiple linear regression was used to predict the combined effects of macropores and surface disturbance (y) based on the isolated effects of changes in surface disturbance on soil with CT macropores (x₁) and the effect of macropore changes on a CT surface (x₂). The y term was estimated from x₁and x₂ with an r² of 0.95. The following equation (eq. 7) was obtained from the regression:

        (7)

Both predictors were statistically significant, with p-values <0.0001. The coefficient of x₂ is 4.5 times greater than that of x₁, suggesting a greater impact of macropores over surface disturbances. This implies that changing only tillage settings in PALMS and neglecting the associated changes in soil structure, such as those estimated in our study, could underestimate the ability of crops to overcome water stress in dry years under NT management.

Several modern technological improvements have been made recently to quantify the changes in soil structure. Bagnall et al. (2020) used a multistripe laser triangulation scanning technique and reported a significantly higher Dirichlet tessellation feature area in CT than in NT. X-ray tomography (Jarvis et al., 2017; Luo et al., 2008; Sandin et al., 2017; Singh et al., 2020) is another commonly used technique to quantify the changes in soil structure and preferential flow. These advancements, if put together with hydrological models, can provide better quantification of the benefits of soil management.

Conclusions

The effects of management practices are often misrepresented in the models due to the omission of soil structure related processes. In this study, the effects of CT and NT management on PAW were simulated using the PALMS model based on measured soil texture and structure information from 18 farms in the Texas Blackland Prairies. The NT management was found to improve PAW over CT during the corn growing season, especially during dry and normal years, due to greater infiltration. No-tillage showed the most prominent improvement in terms of higher PAW as compared to CT during dry years 2006 and 2008, when NT held 44.8 and 20.4 cm of more PAW. With a higher drought risk under a changing climate, this additional PAW under NT can be extremely valuable in sustaining a crop. However, during wet years, CT occasionally held higher PAW, especially when rainfall occurred after a tillage event, as compared to NT. Overall, PALMS was found to be an effective tool for simulating the effects of tillage management on PAW, although it tended to slightly underpredict soil water in deeper layers owing to a higher drainage rate causing inadequate macropore to matrix flow. An analysis of the individual contribution of surface roughness and macropore parameters on PAW indicated that soil structure had about 4.5 times higher impact than surface disturbance due to tillage on PAW in the soil profile for the study area. This highlights the importance of considering changes in soil structure and its effects on macropore flow in modeling the effects of changes in soil management and shows that when these parameters are neglected, inaccurate conclusions can be drawn on the effect of NT on PAW.

Acknowledgments

This material is based upon work that is supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, under award numbers NIFA-2018-67019-27975 and 2021-68012-35897. The Water Management and Hydrologic Sciences (WMHS) program at Texas A&M University has also provided partial support towards graduate research assistantship to the first author. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture.

References

Alhameid, A., Singh, J., Sekaran, U., Ozlu, E., Kumar, S., & Singh, S. (2020). Crop rotational diversity impacts soil physical and hydrological properties under long-term no- and conventional-till soils. Soil Res., 58(1), 84-94. https://doi.org/10.1071/SR18192

Amrit, K., Mishra, S. K., Pandey, R. P., Himanshu, S. K., & Singh, S. (2019). Standardized precipitation index-based approach to predict environmental flow condition. Ecohydrology, 12(7), e2127. https://doi.org/10.1002/eco.2127

Arai, M., Minamiya, Y., Tsuzura, H., Watanabe, Y., Yagioka, A., & Kaneko, N. (2014). Changes in water stable aggregate and soil carbon accumulation in a no-tillage with weed mulch management site after conversion from conventional management practices. Geoderma, 221-222, 50-60. https://doi.org/10.1016/j.geoderma.2014.01.022

Bagnall, D. K., & Morgan, C. L. (2021). SLAKES and 3D Scans characterize management effects on soil structure in farm fields. Soil Tillage Res., 208, 104893. https://doi.org/10.1016/j.still.2020.104893

Bagnall, D. K., Jones, E. J., Balke, S., Morgan, C. L., & McBratney, A. B. (2020). An in situ method for quantifying tillage effects on soil structure using multistripe laser triangulation. Geoderma, 380, 114642. https://doi.org/10.1016/j.geoderma.2020.114642

Bagnall, D., Morgan, C. L., Molling, C. C., Heilman, J. L., & Moore, G. W. (2019). Testing a water redistribution model in a cracked vertisol at two scales. Vadose Zone J., 18(1), 180173. https://doi.org/10.2136/vzj2018.09.0173

Basche, A. D., & DeLonge, M. S. (2019). Comparing infiltration rates in soils managed with conventional and alternative farming methods: A meta-analysis. PLoS One, 14(9), e0215702. https://doi.org/10.1371/journal.pone.0215702

Baumhardt, R. L., & Lascano, R. J. (1999). Water budget and yield of dryland cotton intercropped with terminated winter wheat. Agron. J., 91(6), 922-927. https://doi.org/10.2134/agronj1999.916922x

Baumhardt, R.L., Schwartz, R., Howell, T., Evett, S.R., & Colaizzi, P. (2013). Residue management effects on water use and yield of deficit irrigated corn. Agron J., 105, 1035-1044. https://doi.org/10.2134/agronj2012.0361

Beven, K., & Germann, P. (2013). Macropores and water flow in soils revisited. Water Resour. Res., 49(6), 3071-3092. https://doi.org/10.1002/wrcr.20156

Bird, B. R., Stewart, W. E., & Lightfoot, E. N. (2006). Transport phenomena (2nd ed.). John Wiley & Sons.

Blevins, R. L., Cook, D., Phillips, S. H., & Phillips, R. E. (1971). Influence of no-tillage on soil moisture. Agron. J., 63(4), 593-596. https://doi.org/10.2134/agronj1971.00021962006300040024x

Booker, J. D., Lascano, R. J., Molling, C. C., Zartman, R. E., & Acosta-Martínez, V. (2015). Temporal and spatial simulation of production-scale irrigated cotton systems. Precis. Agric., 16(6), 630-653. https://doi.org/10.1007/s11119-015-9397-6

Bordovsky, D. G., Choudhary, M., & Gerard, C. J. (1999). Effect of tillage, cropping, and residue management on soil properties in the Texas Rolling Plains. Soil Sci., 164(5), 331-340. https://doi.org/10.1097/00010694-199905000-00005

De Vita, P., Di Paolo, E., Fecondo, G., Di Fonzo, N., & Pisante, M. (2007). No-tillage and conventional tillage effects on durum wheat yield, grain quality and soil moisture content in southern Italy. Soil Tillage Res., 92(1), 69-78. https://doi.org/10.1016/j.still.2006.01.012

DeLaune, P. B., & Sij, J. W. (2012). Impact of tillage on runoff in long term no-till wheat systems. Soil Tillage Res., 124, 32-35. https://doi.org/10.1016/j.still.2012.04.009

DeLaune, P., Mubvumba, P., & Lewis, K. (2016). Influence of management practices on soil health in cotton cropping systems. Proc. 2016 Beltwide Cotton Conf. National Cotton Council.

Engman, E. T. (1986). Roughness coefficients for routing surface runoff. J. Irrig. Drain. Eng., 112(1), 39-53. https://doi.org/10.1061/(ASCE)0733-9437(1986)112:1(39)

Epstein, J. M. (2008). Why model? J. Artif. Soc. Soc. Simul., 11(4). Retrieved from https://www.jasss.org/11/4/12.html

Fatichi, S., Or, D., Walko, R., Vereecken, H., Young, M. H., Ghezzehei, T. A.,... Avissar, R. (2020). Soil structure is an important omission in Earth System Models. Nat. Commun., 11(1), 522. https://doi.org/10.1038/s41467-020-14411-z

Gee, G. W., & Bauder, J. W. (1986). Particle-size analysis. In A. Klute (Ed.), Methods of soil analysis, Part 1: Physical and mineralogical methods (2nd ed., pp. 383-411). https://doi.org/10.2136/sssabookser5.1.2ed.c15

Govaerts, B., Fuentes, M., Mezzalama, M., Nicol, J. M., Deckers, J., Etchevers, J. D.,... Sayre, K. D. (2007). Infiltration, soil moisture, root rot and nematode populations after 12 years of different tillage, residue and crop rotation managements. Soil Tillage Res., 94(1), 209-219. https://doi.org/10.1016/j.still.2006.07.013

Green, W. H., & Ampt, G. A. (1911). Studies on soil phyics. J. Agric. Sci., 4(1), 1-24. https://doi.org/10.1017/S0021859600001441

Haddad, N. I., Lawrie, R. A., & Eldridge, S. M. (2009). Improved method of making soil monoliths using an acrylic bonding agent and proline auger. Geoderma, 151(3), 395-400. https://doi.org/10.1016/j.geoderma.2009.05.012

Hansen, N. C., Allen, B. L., Baumhardt, R. L., & Lyon, D. J. (2012). Research achievements and adoption of no-till, dryland cropping in the semi-arid U.S. Great Plains. Field Crops Res., 132, 196-203. https://doi.org/10.1016/j.fcr.2012.02.021

Hanson, J. D., Ahuja, L. R., Shaffer, M. D., Rojas, K. W., DeCoursey, D. G., Farahani, H., & Johnson, K. (1998). RZWQM: Simulating the effects of management on water quality and crop production. Agric. Syst., 57(2), 161-195. https://doi.org/10.1016/S0308-521X(98)00002-X

Harmel, R. D., Haney, R. L., Smith, D. R., White, M., & King, K. W. (2014). USDA-ARS Riesel Watersheds, Riesel, Texas, USA: Water quality research database. Water Resour. Res., 50(10), 8374-8382. https://doi.org/10.1002/2013WR015191

Jarvis, N., Larsbo, M., & Koestel, J. (2017). Connectivity and percolation of structural pore networks in a cultivated silt loam soil quantified by X-ray tomography. Geoderma, 287, 71-79. https://doi.org/10.1016/j.geoderma.2016.06.026

Kroes, J. G., Van Dam, J. C., Bartholomeus, R. P., Groenendijk, P., Heinen, M., Hendriks, R. F.,... van Walsum, P. E. (2017). SWAP version 4, Theory description and user manual. ESG Report 2780. Wageningen Environmental Research. https://doi.org/10.18174/416321

Kucharik, C. J., Foley, J. A., Delire, C., Fisher, V. A., Coe, M. T., Lenters, J. D.,... Gower, S. T. (2000). Testing the performance of a dynamic global ecosystem model: Water balance, carbon balance, and vegetation structure. Global Biogeochem. Cycles, 14(3), 795-825. https://doi.org/10.1029/1999GB001138

Larsbo, M., & Jarvis, N. (2005). Simulating solute transport in a structured field soil: Uncertainty in parameter identification and predictions. J. Environ. Qual., 34(2), 621-634. https://doi.org/10.2134/jeq2005.0621

Lepore, B. J., Morgan, C. L., Norman, J. M., & Molling, C. C. (2009). A Mesopore and Matrix infiltration model based on soil structure. Geoderma, 152(3), 301-313. https://doi.org/10.1016/j.geoderma.2009.06.016

Luo, L., Lin, H., & Halleck, P. (2008). Quantifying soil structure and preferential flow in intact soil using X-ray computed tomography. Soil Sci. Soc. Am. J., 72(4), 1058-1069. https://doi.org/10.2136/sssaj2007.0179

McDaniels, L. L. (1962). Consumptive use of water by major crops in Texas. Texas Board of Water Engineers Bulletin 6019. Texas Board of Water Engineers.

McKee, T. B. (1995). Drought monitoring with multiple time scales. Proc. 9th Conf. on Applied Climatology.

McKee, T. B., Doesken, N. J., & Kleist, J. (1993). The relationship of drought frequency and duration to time scales. Proc. 8th Conf. on Applied Climatology.

Mellouli, H. J., van Wesemael, B., Poesen, J., & Hartmann, R. (2000). Evaporation losses from bare soils as influenced by cultivation techniques in semi-arid regions. Agric. Water Manag., 42(3), 355-369. https://doi.org/10.1016/S0378-3774(99)00040-2

Molling, C. C., Strikwerda, J. C., Norman, J. M., Rodgers, C. A., Wayne, R., Morgan, C. L.,... Mecikalski, J. R. (2005). Distributed runoff formulation designed for a precision agricultural landscape modeling system. JAWRA, 41(6), 1289-1313. https://doi.org/10.1111/j.1752-1688.2005.tb03801.x

Myers, J. L., Wagger, M. G., & Leidy, R. B. (1995). Chemical movement in relation to tillage system and simulated rainfall intensity. J. Environ. Qual., 24(6), 1183-1192. https://doi.org/10.2134/jeq1995.00472425002400060019x

National Cooperative Soil Survey. (2018). National Cooperative Soil Characterization Database. Lincoln, NE: National Cooperative Soil Survey. Retrieved from http://ncsslabdatamart.sc.egov.usda.gov

Nelson, J. R., Lascano, R. J., Booker, J. D., Zartman, R. E., & Goebel, T. S. (2013). Evaluation of the precision agricultural landscape modeling system (PALMS) in the semiarid Texas Southern High Plains. Open J. Soil Sci., 3(4), 169-181. https://doi.org/10.4236/ojss.2013.34020

Newton, P., Civita, N., Frankel-Goldwater, L., Bartel, K., & Johns, C. (2020). What is regenerative agriculture? A review of scholar and practitioner definitions based on processes and outcomes. Front. Sustain. Food Syst., 4. https://doi.org/10.3389/fsufs.2020.577723

Nielsen-Gammon, J., & McRoberts, B. (2009). An assessment of the meteorological severity of the 2008-09 Texas drought through July 2009. Office of the State Climatologist.

O’Donoghue, T., Minasny, B., & McBratney, A. (2022). Regenerative agriculture and its potential to improve farmscape function. Sustainability, 14(10), 5815. https://doi.org/10.3390/su14105815

Onstad, C. A. (1984). Depressional storage on tilled soil surfaces. Trans. ASAE, 27(3), 729-732. https://doi.org/10.13031/2013.32861

Ramos, M. C., Pareja-Sánchez, E., Plaza-Bonilla, D., Cantero-Martínez, C., & Lampurlanés, J. (2019). Soil sealing and soil water content under no-tillage and conventional tillage in irrigated corn: Effects on grain yield. Hydrol. Process., 33(15), 2095-2109. https://doi.org/10.1002/hyp.13457

Sandin, M., Koestel, J., Jarvis, N., & Larsbo, M. (2017). Post-tillage evolution of structural pore space and saturated and near-saturated hydraulic conductivity in a clay loam soil. Soil Tillage Res., 165, 161-168. https://doi.org/10.1016/j.still.2016.08.004

Schulte, R. P., O’Sullivan, L., Vrebos, D., Bampa, F., Jones, A., & Staes, J. (2019). Demands on land: Mapping competing societal expectations for the functionality of agricultural soils in Europe. Environ. Sci. Policy, 100, 113-125. https://doi.org/10.1016/j.envsci.2019.06.011

Schwartz, R. C., Baumhardt, R. L., & Evett, S. R. (2010). Tillage effects on soil water redistribution and bare soil evaporation throughout a season. Soil Tillage Res., 110(2), 221-229. https://doi.org/10.1016/j.still.2010.07.015

Simunek, J., Sejna, M., Van Genuchten, M. T., Šimunek, J., Šejna, M., Jacques, D.,... Sakai, M. (1998). HYDRUS-1D. Simulating the one-dimensional movement of water, heat, and multiple solutes in variably-saturated media, ver. 2.

Singh, J., Singh, N., & Kumar, S. (2020). X-ray computed tomography-measured soil pore parameters as influenced by crop rotations and cover crops. Soil Sci. Soc. Am. J., 84(4), 1267-1279. https://doi.org/10.1002/saj2.20105

Staes, J., Vrebos, D., Georgoulas, A., Meire, P., Jones, A., Creamer, R. E.,... O Sullivan, L. (2018). Demand scenarios: Quantification of the demand for soil functions. LANDMARK Report 4.1. Retrieved from www.landmark2020.eu

Stewart, R. D., Jian, J., Gyawali, A. J., Thomason, W. E., Badgley, B. D., Reiter, M. S., & Strickland, M. S. (2018). What we talk about when we talk about soil health. Agric. Environ. Lett., 3(1), 180033. https://doi.org/10.2134/ael2018.06.0033

Zeimen, M. B., Janssen, K. A., Sweeney, D. W., Pierzynski, G. M., Mankin, K. R., Devlin, D. L.,... McVay, K. A. (2006). Combining management practices to reduce sediment, nutrients, and herbicides in runoff. J. Soil Water Conserv., 61(5), 258-267. Retrieved from https://www.jswconline.org/content/jswc/61/5/258.full.pdf

Zobeck, T. M., & Onstad, C. A. (1987). Tillage and rainfall effects on random roughness: A review. Soil Tillage Res., 9(1), 1-20. https://doi.org/10.1016/0167-1987(87)90047-X