Influence of landscape position and climatic seasonality on soil water and gas conductivity properties in agricultural soils

Abstract Agricultural landscape management and climate seasonality can influence soil structure, hydraulic conductivity, and air permeability within the context of soil water and soil gas mobility. To investigate this, in situ and laboratory-based data were collected from three agricultural landscape positions within a watershed in eastern Ontario, Canada during a growing season. Macropore classification, water infiltration tests, and air permeability measurements were conducted in situ and standard soil characterizations were carried out on soil samples. Hydraulic conductivity of the soil matrix, based on grain size data, indicated that the highest values were consistently measured in the B horizon at each landscape setting. Macropores were found to be more abundant within uncultivated drainage ditch bank soils, compared to the adjacent cropped fields. Macropores in the ditch bank soils were exclusively consisted of circular biopores, while both circular and linear macropores were observed in the cultivated field soils. Air permeability, vertical hydraulic conductivity, and horizontal hydraulic conductivity were also greater in the uncultivated soils, relative to the cultivated soils. Field saturated hydraulic conductivity measurements offered evidence of anisotropy, likely due to the vertical nature of the macropore features. Macropore disposition and extent varied over the growing season, especially in the cultivated field soils where tillage and field trafficking are physically disruptive. Seasonality of macropore development will influence temporal changes in advection-based mass exchange of gas and water in the vadose zone. Modeling of mass exchange in agricultural soils should consider time variability in macroporosity to more realistically characterize infiltration and soil gas emissions.


Introduction
Soil characteristics, soil management, and weather are all factors that influence the mobility of water, contaminants, colloids, and gases in the vadose zone (Mohanty et al. 2016;Simunek and van Genuchten 2016;Pla et al. 2017;Glaesner et al. 2018;Orozco-Lopez et al. 2018).There has been considerable interest in how these factors influence the fate and transport of nutrients, pesticides, and pathogens in the shallow subsurface and associated "downstream" impacts on the environment and human health (Kanwar et al. 1985;Edwards et al. 1992;Stehouwer et al. 1994;Larsson and Jarvis 1999).More recently, as concerns over global climate change continue to grow, land management and soil characteristics have become of increasing interest for their potential role in controlling greenhouse gas (GHG) emissions from agricultural soils (Nangia et al. 2013;Oertel et al. 2016;Shakoor et al. 2021;Bhattacharyya et al. 2022;Virk et al. 2022).GHG exchange at the agricultural soil/atmosphere interface is anticipated to be both temporally and spatially variable depending on the time of year and the landscape position, yet the time-varying soil physical factors (i.e., air permeability, pore connectivity, hydraulic conductivity etc.) that govern effluxes and GHG concentrations in soils, are usually not rigorously characterized (Chang et al. 2018;Piccoli et al. 2019;Blume et al. 2022;Jia et al. 2023).However, more accurate quantification of GHGs from agricultural soils is critical with respect to refining atmospheric N and C balance (Wang et al. 2021).Therefore, an improved understanding and quantification of the factors that control agriculturally engendered GHG emissions, and how, when, and where to mitigate these emissions is required (Gregorich et al. 2005).
Soil structure and textural characteristics influence the transient movement of soil water and soil gas in the vadose zone (Hillel 2013).Grain size, mineralogy, carbon content, and bulk density of the soil matrix control the permeability, reactivity, and capillarity of the soil.In addition, the role of secondary permeability related to macroporosity has been a topic of considerable interest as it represents a pathway of least resistance for water and gas advection through soils (Blackwell et al. 1990;Schjønning et al. 2002;Arthur et al. 2013;Katuwal et al. 2015).
Soil macropores are utilized in many mass exchange modeling activities and the quantification of their physical characteristics is required to inform their incorporation in simulation experiments (Jarvis 2007;Ball 2013;de Vries et al. 2017;Hussain et al. 2019;Jia et al. 2023).By virtue of their size and connectivity, soil macropores can dominate mass exchange and critical reactions within the soil profile (Koestel and Larsbo 2014;Jarvis 2020;Hussain et al. 2021;Jia et al. 2023).Soil pores larger than 0.3-0.5 mm in diameter (assuming cylindrical pore geometry) are conventionally classified as macropores (Jarvis 2007).They can form due to root growth (dead root channels) and burrowing by biota such as worms and detritus feeders, cracking and fissuring due to wettingdrying and freezing-thawing cycles, and preferential subsurface soil erosion (Beven and Germann 1982;Iversen et al. 2012).Macropore characteristics are controlled by, among other time-varying factors, temperature, moisture, surface vegetation, slope changes, soil fauna, and soil management practices.Therefore, the role macropores play in mass exchange varies temporally and spatially in response to these drivers (Turpin et al. 2007;Zhang et al. 2016;Demand et al. 2019;Budhathoki et al. 2022).
Agricultural landscape position, soil management, and degree of soil drainage can influence the nature of soil hydraulic properties.Cultivation results in the cyclical growth and removal of surface vegetation, which influences root formation, soil surface exposure, and surface vehicular activity as compared to uncultivated land.Soil drainage, which can occur during the course of the growing season or be enhanced through tiles and surface water drains, can result in dryer shallow soil conditions, which may enhance formation of macroporosity for example.
Soil physical parameters that control water and gas movement include texture, moisture content, hydraulic conductivity (saturated and unsaturated), matric flux potential, and capillarity.Related process monitoring and modeling studies require estimates of these key parameters; however, the majority of the measurement techniques are designed for soil water applications, and the required parameters controlling soil gas mobility are seldom measured directly.Conventional in situ methods include soil lysimeters, soil water content measurement methods, and infiltration devices such as the Guelph permeameter (GP) and the tension infiltrometer, which are commercially available and widely used.At the laboratory bench scale, various approaches for estimating soil matrix parameters impacting soil water mobility such as grain size distribution, mineralogy, and hydraulic parameters, including porosity and hydraulic conductivity are routinely employed (Carter and Gregorich 2007).For example, the HYPROP2 method (Shokrana and Ghane 2020) provides a convenient and accurate approach for determining the soil water characteristic curves for laboratory samples.The highly advance Rosetta 3 software platform based on pedotransfer functions is also an invaluable tool for determining representative soil properties, including the characteristic curves (Zhang and Schaap 2017).However, few methods are available to measure physical parameters that control soil gas movement directly.One instrument that has been used specifically to measure soil gas permeability is the air permeameter (Iversen et al. 2003).As the mobility of soil gas is controlled by the drained or air-filled pore network, information derived from the approaches used to quantify soil water movement may provide valuable insight into the nature of the soil parameters that control soil gas movement.Essentially, the gas content and available transport pathways will be the inverse of those controlling soil water under a given soil water content.
Although methodologies for measurement of critical soil matrix (or bulk soil) parameters are well established (Carter and Gregorich 2007), characterizing and quantifying macropore networks still remains a significant challenge (Cheik et al. 2019;Hlavacikova et al. 2019).In many investigative applications, the macropore domain is often treated as static or spatially explicit to simplify modeling parameterization and reduce computational complexity (Frey et al. 2016;Weiler 2017).Frey and Rudolph (2011) demonstrated that the occurrence, size distribution, and density of macropores within the shallow subsurface tends to decrease with depth, with the majority of the macropores often concentrated in the rooting zone and slightly below it.However, characterizing the in situ spatio-temporal varying disposition of soil macropores is enormously time consuming and often destructive to the field site (Perret et al. 1999;Frey and Rudolph 2011;Katuwal et al. 2015;Fan et al. 2016;Min et al. 2020).As such, in situ characterization and quantification of the macropore variability and the incorporation of this information in predictive analytics of soil mass exchange is still an evolving area of research (Taft et al. 2019;Cheng et al. 2021;Jia et al. 2023).
The primary objective of this study is to characterize and quantify the spatial and temporal variability of the main physical soil parameters that influence soil water and soil gas mobility within cultivated and uncultivated soils in an agricultural landscape and utilizing parameterization techniques specifically used in investigations of soil water mobility.Field investigations were conducted at a typical agricultural setting located near Ottawa, ON, Canada.In considering the soil gas phase, an initial hypothesis was made that the micro-and macroporosity occupied by soil gas can be estimated from the soil water parameter information and soil water distributions through inversion methods.Utilizing this approach, changes in the physical, hydraulic, and pneumatic properties of cultivated field soils and adjacent uncultivated soils along a drainage ditch bank, were tracked throughout the unfrozen growing season to assess key soil parameter changes over time.It was further hypothesized that the characteristics of the physical parameters controlling soil gas mobility would vary during the course of the unfrozen seasons and also with the degree and nature of field cultivation, or lack of it in the case of the drainage ditch banks.
Field and laboratory measurements were undertaken to (i) characterize the soil grain size, carbon content, bulk density, porosity, and permeability; (ii) quantify the temporal variability of macropore types, depth-density relationships, and connectivity over unfrozen seasons; (iii) quantify and relate soil hydraulic properties to macropore properties; and finally, (iv) quantify air permeability of the soils and Fig. 1.Location of the study site in Ontario, Canada.Information for the map was derived from the following websites: http://www.cec.org/north-american-environmental-atlas/political-boundaries-2021/ and https://www.glc.org/greatlakesgis.
relate them to macropore properties.The results of this study attempt to provide insight into the differences and evolution of soil gas/water flow pathways in cultivated and noncultivated soils with a common parent material during an annual growth cycle.The work also aims to provide unique information and data to inform numerical analysis of dynamic soil water mobility and GHG emission behaviour and specifically help define the time-varying nature of "preferential flow" in soil-based mass exchange modeling activities (i.e, Jia et al. 2023).

Study sites
Field investigations were conducted in eastern Ontario, Canada, ∼45 km southwest of Ottawa, Canada (Fig. 1), on a small experimental watershed (Wilkes et al. 2014;Sunohara et al. 2015).The field measurements took place on cropped fields and on an adjacent uncultivated drainage ditch bank.Within the South Nation River basin where the experimental watersheds are located, it is estimated that over 3000 km of agricultural drainage ditches exist (Rideout et al. 2022).The ditches examined in this study were excavated over 40 years ago and vegetation along banks and channels were not managed prior to the beginning of the study.The drainage ditch bank site was never cultivated and consisted of native vegetation.At the nearby Russell, Ontario meteorological station (approximately 14 km east of the study sites), between 1981 and 2010, mean total annual precipitation was 981 mm, while average daily air temperature was 6.5 • C, with an interannual range from −10.2 • C in January to 20.8 • C in July (https://climate.weather.gc.ca/climate_normals/results_1 981_2010_e.html).Fields at the study site are generally flat and systematically tile drained (20 m separation) to a depth of approximately 1 m at the outlets.The tiles discharge into the adjacent constructed drainage ditches, which are approximately 2-3 m in depth below field level.The field soils are uniformly classified across the study site as Bainsville Series silt loam; fine sandy loam soils with layered silt and fine sand parent materials with soil horizons defined by decreasing organic matter and increasing clay content with depth (Wicklund and Richards 1962).
Soil measurements were conducted on a tilled corn field site, which was moldboard plowed in fall (∼15-20 cm interval depth), cultivated in the spring prior to planting, and used for corn (Zea mays) production for three consecutive years (2016, 2017, and the 2018 study year).Select soil measurements were also carried out on a no-till field under alfalfa production for over 3 years prior to the study (no-till alfalfa field site), located approximately 500 m south of the tilled field.Soil measurements were also conducted on the top (shoulder) of the adjacent agricultural drainage ditch (ditch bank site), approximately 2 m south of the no-till alfalfa field.The relative location of each of the field investigation sites is shown in Fig. 2.
The study was conducted over the course of the 2018 growing season with three data collection campaigns taking place in May/June (following field cultivation and planting), July/August, and October/November, referred to as spring, summer, and fall seasons, respectively.Some initial field data were also collected during the fall of 2017.Additional weather data, including air temperature and precipitation, were collected hourly with an onsite weather station (HOBO Weather Station Data Logger, Onset Computer Corp., MA) located within the no-till field site.

Experimental plot locations and preparation
At both the tilled corn field site and ditch bank site, replicate plot-scale investigations and excavations were conducted during each of the three field monitoring campaigns during the unfrozen seasons in 2018.For logistical reasons, it was not possible to conduct the plot-scale work at the notill alfalfa field site.The field plots measured 80 cm × 80 cm square on both the soil surface and on 15 cm deep excavated soil "benches".During each site visit, measurements were conducted on three ditch bank site plots and on two tilled corn field site plots (Fig. 2).At ground surface, the plots were prepared by trimming vegetation at the stem base, taking care to avoid pulling out roots and disturbing the soil structure.The soil surface was further prepared by removing between 0.5 and 3 cm of soil from the uneven surface by creating vertical incisions with the side of a sharpened trowel and flicking the soil up to expose its structure while minimizing smearing.During the surface preparation stage, a small bubble-level was used to ensure the plot surfaces were level.Finally, the plot surfaces were vacuumed to clear any remaining loose debris.After all soil surface experiments were completed (macropore counting, air permeability, pressure and tension infiltration tests, and coring, as described in the following sections), the surface plots were excavated down to approximately 15 cm depth to expose soil benches undisturbed by previous measurements.At one of the plots at the tilled corn field site, an additional bench was excavated to a depth of 30 cm to examine the continuity of macropore features below the plow depth.The bench surfaces were prepared following the same procedure as that used for the surface experiments.The specific locations of the hydraulic and air permeability measurements (described in detail in the following sections) on the benches were laterally offset from those conducted on the surface to avoid influence from surface experiment-induced soil moisture variations.Plot locations were replicated immediately adjacent to the previous location from spring to summer to fall such that measurements during each season were completed on undisturbed soils.

Soil matrix characterization
To determine soil particle and compositional and hydraulic characteristics of the shallow soil matrix at the study site, soil cores were collected during the 2018 fall field campaign from the surface and 15 cm benches in the A horizon of the both the tilled corn field site and ditch bank site.Additionally, a combination of core and hand-auger samples were collected from the A, B, and C horizons at the no-till alfalfa field site and from B and C horizons underlying the plots at the tilled corn and ditch bank sites to more completely examine the nature of the shallow soil profile.Soil texture was visually inspected during hand augering to confirm the vertical positions of the primary A, B, and C horizons.The undisturbed cores were extracted using a thin-walled steel tube with 4.6 cm inner diameter and 7.5 cm length (volume of 124.6 cm 3 ) that was carefully pressed into the ground and then removed by excavating around the tube and sliding a trowel under the tube base to minimize soil loss.Core samples were extruded from the core tubes and placed in sealable bags and kept frozen prior to submission to the laboratory.
Soil samples were analyzed for grain size distribution, organic matter and carbon content, and dry bulk density.Grain size distribution was analyzed using the sieve-pipette method (Carter and Gregorich 2007).Organic matter content was determined using the Walkley and Black (1934) method.Soil dry bulk density and carbon content were measured using the combustion method where soil samples are heated to temperatures up to 1000 • C, converting all carbon species to carbon dioxide.These analyses were completed by the Agriculture and Food Laboratory at the University of Guelph.The grain size distribution and bulk density data were further used to calculate soil hydraulic conductivity and porosity, respectively utilizing an Excel-based software tool Hydro-geoSieveXL developed by Devlin (2015).

Pressure infiltration measurements
During both the 2017 and 2018 sampling campaigns, in situ infiltration tests were conducted at each of the three landscape positions using a commercially available GP (Soilmoisture Equipment Corp.).The GP was operated in two modes per Carter and Gregorich (2007): (1) with a surface ring to estimate vertical field hydraulic conductivities and (2) with a well outflow tube to estimate, primarily, horizontal field saturated hydraulic conductivities.Hydraulic conductivity estimates using the GP rely on either performing two infiltration tests at each location using two different hydraulic head settings or using only one hydraulic head setting in addition to using information about the soil texture, which informs assignation of an empirically derived sorptive number of the porous medium, α * (Elrick et al. 1989;Reynolds and Elrick 1990).As the single-head GP methodology and equations developed by Reynolds and Elrick (1990) were employed in this study, a fixed α * value of 12 m −1 was used, as it has been found most suitable for many structured and medium textured materials, including structured clayey and loamy soils (Reynolds and Elrick 1990).Previous work has shown that hydraulic conductivities derived using a single hydraulic head setting and estimating α * results in errors within acceptable limits when compared to other potential sources of errors, such as soil heterogeneity (Elrick et al. 1989;Reynolds et al. 1992).Vertical hydraulic conductivities were measured in triplicate at the ground surface and at the 15 cm deep bench within each of the tilled corn field and ditch bank sites.These were conducted during each seasonal site visit.Horizontal hydraulic conductivity values were measured adjacent to each of the plots within the ditch bank site and the tilled corn field site and at three locations within the no-till alfalfa field site.The horizontal hydraulic conductivity measurements were only conducted during the fall season, so no attempt to distinguish a seasonal trend was made for those data.Boreholes were augured between 18 and 80 cm in depth to target the A and B horizons for the well outflow tube tests.The infiltration tests and data analyses were carried out as described in Reynolds et al. (1985) and Reynolds and Elrick (1986).

Air permeability measurements
Direct measurements of air permeability were conducted at ground surface and at the 15 cm benches in each plots within the ditch bank and tilled corn field sites during the summer and fall site visits in 2018 using a commercially available air permeameter device (Umwelt-Geräte-Technik).The air permeameter consisted of a 7.1 cm diameter cylindrical surface chamber and a 7.5 cm long time domain reflectometer attachment for measuring volumetric soil moisture content.In preparation for measurement, the surface chamber was placed over a representative area of soil within the plot and then gently pushed approximately 1 cm into the soil to prevent air short-circuiting around the chamber-soil interface.A circular area, 14.5 cm in diameter, around the surface chamber was sealed with a viscous gel (commercially available wallpaper glue) to further reduce air short-circuiting.The instrument measures the pressure difference between a calibrated pressure sensor and the pressure within the surface chamber as air is drawn through the soil and into the instrument through the surface chamber at a constant rate.The rate of air flow into the instrument is adjusted by the user in response to soil conditions (lower flow rates for low permeability soils and vice versa).Air permeability measurements were completed in five locations on each bench plot at the two locations.

Macropore characterization
As part of the in situ site characterization strategy, direct mapping of visible macropore features was undertaken within the plots at the tilled corn field and ditch bank field sites during the spring, summer, and fall field programs.The macropore counting procedure was adapted from Frey and Rudolph (2011), with macroporosity defined as the sum of biogenic pores, formed by root growth and invertebrate activity, and linear macropores such as desiccation cracks with apertures greater than 0.5 mm.Macropore counting was conducted using a grid scale and a round-machined pin with 5, 8, 10, and 12 mm diameter sections.The grid scale comprised a wooden frame divided into eight 10 cm rows and columns delineated by strings, which was placed over the prepared soil surface to subdivide the areas to be counted (Fig. 3).The number of macropores present in each 10 cm cell was tallied for the following diameter ranges: 0.5-5, 5-8, 8-10, and 1012 mm.Any visible macropores with diameters smaller than 5 mm were counted manually and listed as being between 0.5 and 5 mm.The lengths of cracks with aperture widths between approximately 0.5 and 5 mm were measured with a ruler.Following Frey and Rudolph (2011), macropore counts were used to calculate the macropore area fraction (MAF), which is the ratio of the area of macropores (biopores and cracks) to the area of soil.The specific proportion of the various sizes of the biogenic macropores to the MAF was calculated by multiplying the arithmetic mean value of each circular macropore size interval by the number of macropores in that interval.Similarly, the proportion of the cracks was calculated by multiplying the widths of each observed crack by its length.MAF was calculated for each cell within the grid scale.

Weather and soil moisture conditions
Precipitation and soil moisture content measured within the alfalfa no-till field site between 1 May and 31 October 2018, are presented in Fig. 4.Over the 6-month study interval, 254.4 mm of rainfall was recorded, with the highest intensity observed on 23 July when 6.4 mm fell within a 15 min span.An extended dry period was observed from mid-August through to the end of September.Soil moisture at 0.15 m depth varied from 0.36 m 3 m −3 on 3 May to 0.26 m 3 m −3 on 22 July, while at 0.45 m depth, it varied from 0.37 m 3 m −3 on 3 May to 0.21 m 3 m −3 on 20 September.Air temperature ranged from 34 • C on 4 July to −7 • C on 26 October.

Soil texture, composition, and hydraulic conductivity
Physical and textural characteristics for the soils at all three sites, based on laboratory analysis of the soil core samples are presented in Table 1.The grain size distribution of the A horizon samples indicates that the soils should be characterized as clay loam and silty clay loam.The A horizon of the no-till alfalfa and ditch bank sites had very similar grain size distributions.The tilled corn field soils had a higher silt fraction, smaller sand fraction, higher dry bulk density, and slightly lower porosity than the soils at the other two sites.Organic and inorganic carbon fractions as well as total organic matter content were the lowest in the tilled corn field soil.Based on the grain size distribution data of soil samples taken from  the B horizon, the soils range between very fine sandy loam and clay loam, whereas the C horizon soils are predominantly clay.The B and C horizons also had much lower organic carbon and organic matter contents than the A horizon as would be anticipated.
The calculated porosities for each soil horizon and spatial location are also summarized in Table 1.The A horizon of the ditch bank soils displayed a larger average porosity as well as a greater range of porosity values than the cultivated soils.The B horizon porosity values did not display a discernable pattern and the variation in values in the different fields is likely due to spatial heterogeneity.The calculated porosity values of these soils are in reasonable agreement with representative porosity ranges for various geological materials presented by Davis (1969).
Soil grain size, bulk density, and porosity data obtained from laboratory analysis on soil cores were also used to estimate hydraulic conductivities with HydrogeoSieveXL (Devlin 2015).The software uses data sets noted above to calculate hydraulic conductivities of the soil matrix based on over a dozen published equations while checking that the applicable criteria for each equation have been met.The soils met the conditions required for calculating hydraulic conductivities us-ing equations developed by I.I.Sauerbrei (1932) (presented in Vukovic and Soro (1992)), Shepherd (1989), Alyamani and Sen (1993), and Barr (2001), and the resulting hydraulic conductivities derived from the different equations are presented in Table 2.A wide range in magnitude of hydraulic conductivity values between the different grain size equations for a given soil is relatively common, especially for finer grained sediments (Devlin 2015).The data can be useful for relative comparative analysis when considering spatial variability as explored below.The combined data from each of the grain size analyses can be averaged to provide an estimated representative matrix hydraulic conductivity for the soil.
The Barr (2001) method resulted in the lowest estimates of hydraulic conductivity, followed by Sauerbrei (1932), Alyamani and Sen (1993), and lastly Shepherd (1989) with the highest estimated hydraulic conductivities.In the order listed, each method resulted in one order of magnitude higher hydraulic conductivities than the previous; however, trends with depth in the soil profile are consistent between the different methods.Based on the combined grain size data, the uncultivated ditch bank site soils appear to have the highest hydraulic conductivity in the B horizon with the A and C horizon soils being lower but very similar to each other.In the no-till and tilled field sites, the B horizon again was estimated to have the highest matrix hydraulic conductivity, followed by the A horizon, with the C horizon having the lowest hydraulic conductivity as measured beneath the tilled corn field.The soil textural classifications presented in Table 1 realistically relate to the relative hydraulic conductivity estimates from the grain size distribution equations, and averaging of the data from the different methods can provide useful estimates of hydraulic conductivities in the soil matrix portion of the bulk soils.

In situ hydraulic conductivity
Vertical and horizontal hydraulic conductivity values were measured in situ with the GP by conducting controlled infiltration tests with a surface ring to estimate vertical hydraulic conductivities in the spring, summer, and fall and with a well outflow tube to estimate horizontal hydraulic conductivities during the fall season only.The hydraulic conductivity values were calculated using the infiltration rate from one hydraulic head setting by assuming a sorptive number of the porous medium, α * of 12 m −1 (Elrick et al. 1989;Reynolds and Elrick 1990;Carter and Gregorich 2007).
Table 3 presents the minimum and maximum, and arithmetic mean of vertical hydraulic conductivity and matric flux potential values calculated at the ditch bank and tilled corn field sites.At the ditch bank plots, the lowest average value of vertical hydraulic conductivity was measured in the spring, while the highest value was observed in the summer.Hydraulic conductivity variability at these plots was also highest during the summer.The ditch bank plot surface soil displayed higher average hydraulic conductivity values in the spring and summer and lower values in the fall, relative to the soil at 15 cm depth.
The tilled corn field soil surface displayed the same pattern of hydraulic conductivity changes throughout the different seasons as the ditch bank soils, with the lowest average values measured in the spring and the highest in the summer.The variability in the tilled corn field measurements was lower in the spring and fall than that in the ditch bank measurements, while in the summer, the tilled corn field soil displayed a similar degree of variability to that of the ditch bank soils.In the spring, the tilled corn field soil surface had higher average hydraulic conductivity than at 15 cm depth, while in the fall there was little difference between the two sites.All but one infiltration test using the surface ring GP attachment at the 15 cm deep bench during the summer field campaign resulted in water short-circuiting through soil cracks and hence could not be completed.
In the spring and fall, the average hydraulic conductivity values at the soil surface and 15 cm depth are approximately 90% greater in the ditch bank soil compared to the tilled corn field soil.In the summer, hydraulic conductivity values at the soil surface are approximately 50% greater in the ditch bank soil than those in the tilled corn field soil.A meaningful comparison of summer hydraulic conductivity values in the tilled corn field soil at 15 cm depth is difficult to make due to the limited number of results.
Horizontal hydraulic conductivity values in the A and B horizons at the ditch bank site, the no-till corn field site, and the tilled alfalfa field site during the fall of 2017 and 2018 are presented in Table 4. Horizontal hydraulic conductivity values as well as the variability in measurements were highest in the ditch bank soils followed by the no-till alfalfa field soil, and lastly the tilled field soil.This may be due to compaction caused by vehicle traffic taking place on the tilled field soils during plowing, planting, and harvesting activities.Compared to the average vertical hydraulic conductivity values, the average horizontal hydraulic conductivity values in the A horizon are approximately two times smaller in the ditch bank soil and approximately eight times smaller in the tilled corn field soils.This result counters the frequently made assumption that horizontal hydraulic conductivity is 10 times greater than vertical hydraulic conductivity in natural porous media (Freeze and Cherry 1979) and potentially highlights the role of macropores in controlling anisotropy of bulk hydraulic conductivity in structured soils.

Macropore characteristics
The majority of observed macropores at all sites were circular in form, which points to biogenic origins.Circular macropore classification data are presented in Table 5.Overall, macropores with diameters between 0.5 and 5 mm were the most abundant compared to macropores with larger diameters, and as the macropore diameter increased, the abundance generally decreased.When interpreting the data, it is important to note that the corn field was tilled (including moldboard plowing) and corn was planted approximately 1 week prior to commencing measurements.The tilling would have destroyed much of the near-surface macroporosity.Tillage depth in these soils can be over 20 cm in places.
For the ditch bank plots, surface macropores with diameters between 0.5 and 5 mm steadily increased in abundance throughout the growing season, while those in the tilled corn field plots were least abundant in the summer and most abundant in the fall.Subsurface macropores at 15 cm depth, with diameters between 0.5 and 5 mm were present in similar numbers during all three seasons in two of three ditch bank plots, while in the third plot and in each tilled corn field plot, the number of these macropores remained steady between spring and summer and increased between summer and fall.The total number of macropores 0.5-5 mm in diameter was greater on the ditch bank plots compared to the tilled corn field plots, at surface and at 15 cm depth, during all three measurement campaigns (Table 5).
The abundance of macropores with diameters between 5 and 8 mm at the soil surface and at 15 cm depth in the ditch bank plots varied between the monitoring locations such that seasonal trends were difficult to discern.In general, however, the number of macropores 5-8 mm in diameter were greater in the ditch bank plots than the tilled corn field plots.In the tilled corn field, the abundance of macropores in this diameter range remained relatively constant throughout the three seasons for two surface plots and one subsurface plot.The second tilled corn field subsurface plot displayed a decrease in the abundance of macropores with diameters between 5 and 8 mm between spring and fall.
At the ditch bank plots, the total number of macropores with 8-10 mm diameters counted at the surface and 15 cm Table 6.Linear macropores observed in the tilled corn field soil.depth was about three times less abundant than that in the 5-8 mm diameter category, and similarly, did not display clear seasonal trends.The occurrences of macropores with 10-12 mm diameters were low, making seasonal trends at the soil surface and at 15 cm depth difficult to discern.At the tilled corn field plots, macropores with diameter ranges of 8-10 mm and 10-12 mm were rare during in the spring and fall at both the surface and 15 cm depth.While slightly more abundant in the summer, the occurrences were too low to make meaningful conclusions about seasonal trends for these two larger macropore diameter ranges in the tilled corn field plots.
The linear macropore occurrence, represented as an estimated aperture size and area observed at the tilled corn field plots during the summer and fall seasons, summarized relative to depth and time of the year is presented in Table 6.Linear macropores were not observed within the ditch bank plots during any of the macropore characterization campaigns.Similarly, linear macropores were not observed at the tilled corn field plots during spring.In the summer, linear macropores were observed at tilled corn field plots, with a greater total linear macropore area found at 15 cm depth relative to the soil surface.These linear macropores are thought to be a result of desiccation cracking during warm and dry summer conditions and water uptake by the shallower rooting corn (in relation to ditch vegetation).The reduced frequency of cracking at ground surface may be the result of near-surface soil swelling during summer rain events and greater surface soil friability at this time resulting from spring cultivation.
In the fall, no cracking was observed on the soil surface, again, likely due to swelling associated with increased soil moisture and direct rain impact following harvest.At 15 cm depth, linear macropores were observed but with an overall open area 83% lower than that observed in summer.It is important to note that the changes in the number of macropores and cracks found at each of the plots between seasons, in both the ditch and tilled field soils, are likely affected to some extent by soil heterogeneity in addition to seasonal changes.
Additional insight into the nature and variability of the macroporosity at the tilled field sites can be gained from the calculated MAF data.Results from the ditch bank soil MAF analysis are presented in Fig. 5.In the spring, ditch bank plots 1, 2, and 3 had total MAF of 0.0092, 0.0168, and 0.0084, respectively; while at 15 cm depth, plots 1, 2, and 3 had MAF of 0.0042, 0.0198, and 0.0122, respectively.Plot 1 at 15 cm depth in spring had the lowest ditch bank MAF observed throughout the duration of the study, with a value of 0.0042.
In the fall, MAFs at ditch bank surface plots 1, 2, and 3 were 0.0393, 0.0345, and 0.0322, respectively, which represent changes of 160%, 35%, and 55% in relation to summer.Despite displaying different MAF starting points in the spring and evolving at different rates through the summer, all three surface plots in the ditch soil displayed similar MAF values by the fall.At 15 cm depth in the fall, the total MAFs for plots 1, 2, and 3 were 0.0225, 0.0210, and 0.0150, respectively, reflecting changes of 360%, 75%, and −6%, relative to summer.
As noted earlier, circular macropores comprised the entirety of the macroporosity in the ditch bank soil as linear macropores were not observed.At the soil surface, MAF increased steadily throughout the growing season and interplot variability was greatest in the spring.At 15 cm depth, the MAF was lower than at the soil surface and remained relatively stable throughout the growing period.One plot was an exception, where MAFs in the spring and summer were low, but evolved to match the other plots by the fall season.Overall, interplot variability at 15 cm depth tended to be lower than on the surface.
MAFs in the tilled corn field plots over the growing season are presented in Fig. 6.In the spring, surface plots 1 and 2 had total MAF of 0.0053 and 0.0076, respectively.At 15 cm depth, plots 1 and 2 had MAF of 0.036 and 0.048, respec-Fig. 5. Distribution of macropore area fraction of circular macropores in each 10 × 10 cm grid cell within plots 1, 2, and 3 at the soil surface and 15 cm depth in the ditch bank soil.The × marks the mean of the distribution, while points mark outliers.tively.As the field soil was tilled and planted approximately 1 week prior to macropore characterization, soil structure and macroporosity in approximately the top 20 cm of the soil profile was effectively reset (Ouellet et al. 2008).For consistency, macropores counted in the spring included only those identified as biopores and did not include spaces between soil peds formed during tillage.
In the summer, surface plots 1 and 2 had total MAF of 0.0046 and 0.0078, respectively, reflecting a change of −13% and 3% relative to spring.At 15 cm depth, the summer total MAFs for plots 1 and 2 were 0.0088 and 0.0092, respectively, representing 144% and 92% increase from spring.The MAF evolution in the field soil appears to be predominantly occurring at depth rather than at the soil surface between spring and summer.
In the fall, MAFs at surface plots 1 and 2 were 0.0230 and 0.0211, respectively, representing changes of 400% and 171% relative to summer.The majority of the macropore formation at the surface of the tilled corn field soil occurred between the summer and fall.At 15 cm depth, the total MAFs for plots 1 and 2 were 0.0151 and 0.0146, which are increases of 72% and 59% relative to summer.
As noted earlier, an additional bench was prepared at 30 cm depth at one of the tilled corn field site plots to observe how macroporosity transitions below the plow pan.At about 25 cm below the surface, the soil transitioned from A horizon to B horizon, changing in colour from dark greyish brown to orangey tan and becoming sandier as outlined in Table 1.As seen in Fig. 6, the total MAF at 30 cm depth was 0.0017, which is over four and five times lower than at the surface, and at 15 cm depth at plot 2, respectively.Linear macropores were not observed at this 30 cm deep soil bench.
Overall, at the tilled corn field plots, linear macropores made up 53%-69% of the total MAF in the summer and 8%-15% in the fall (Table 5).At both the soil surface and 15 cm depth, the MAF increased steadily throughout the growing season (Fig. 6).However, the abundance of circular macropores on the soil surface decreased and remained relatively unchanged at 15 cm depth between the spring and summer (Table 6), suggesting that in the summer, linear macropores drive the overall MAF increase in the tilled field soil in response to soil moisture changes.

Air permeability
Air permeability measurement data collected from the ditch bank plots and tilled corn field plots at surface and at 15 cm depth are shown in Fig. 7.It is important to note that the air permeability encountered in these soil settings was often higher than the upper recommended accuracy range of the air permeameter of 30 cm s −1 , and as such, the data pre-Fig.6. Distribution of macropore area fraction of circular and linear macropores in each 10 × 10 cm grid cell within plots 1 and 2 at the soil surface, 15, and 30 cm depth in the tilled corn field soil.The × marks the mean of the distribution, while the points mark outliers.sented should be viewed in terms of the trends and orders of magnitude rather than exact values.
At the ditch bank plots, the magnitude of the air permeability was generally greater at the surface than at 15 cm depth at a given soil moisture content, and the values were generally higher in the fall than in the summer.At the tilled corn field plots, the air permeability trends followed the same patterns as that of the ditch bank plots.Interestingly, air permeability measurements at 15 cm depth at the tilled corn field plots did not follow a clear decreasing trend with increasing soil moisture content, in contrast to the tilled corn field soil surface and all depths at the ditch bank plots.Overall, the ditch bank plots had higher air permeability than the tilled corn field plots at both the soil surface and 15 cm depth.Accord-ing to the air permeameter manual, the air permeability rates measured at tilled field and ditch bank locations, at surface and 15 cm depth, are considered very high.
At the 30 cm deep bench excavated at the tilled corn field site in the summer, air permeability was measured to be 0.03 cm s −1 at a soil moisture content of 29% in an area of the soil with no macropores, and 1.4 cm s −1 at a soil moisture content of 25% in an area with one circular macropore with a diameter in the range of 8-10 mm.While only providing a single set of measurements, the results from the 30 cm bench highlight that macropores can play a significant role in governing air (oxygen) movement/transport at depth and therefore be important in the context of critically regulating GHG production in the deeper soil layers (Jia et al. 2023).

Discussion
The combined data sets resulting from laboratory analysis of soil core samples and the in situ field measurements can be used to inform a wide range and scale of modeling experiments concerned with soil water, solute, and gas fate and transport in the shallow vadose zone in this type of environment.Critical controlling matrix parameters such as grain size distribution, carbon content, porosity, and hydraulic conductivity were observed to vary significantly between the uncultivated and cultivated field settings and also within the A, B, and C horizons.
A finer grain size component was documented to be present within the tilled corn field soils and the overall carbon content was the lowest of the three landscape settings investigated.At each field site (ditch bank, no-till alfalfa, and tilled corn field sites), the carbon content was the highest within the A horizon, decreasing with depth through the B and C horizons.Based on the grain size analysis approach, the soil matrix within the B horizon was the most permeable at each site.This soil variability is likely due to the land management practices and vegetative growth cycles within uncultivated and cultivated soils, something that would need to be considered when investigating both physical flow and reactive transport processes.
Soil macropores, which can act as paths of least resistance for advective flow of water/gases in the soil environment, were found to be more abundant at the uncultivated drainage ditch bank plots as compared to the tilled corn field plots during spring, summer, and fall.Overall, macropore abundance at the ditch bank plots increased steadily between spring and fall at the soil surface, likely in response to seasonal earthworm (endogeic) activity (Perreault and Whalen 2006;Potvin and Lilleskov 2017;Ruiz et al. 2021), given that only biopores were evident in this landscape setting.Deeper in the soil profile, at 15 cm depth in the ditch bank soils, the abundance of macropores remained relatively steady throughout the growing season perhaps due to these biopores being dominated by anecic earthworms with more temporally stable burrows (Potvin and Lilleskov 2017).In contrast, macropores at both the surface and at 15 cm depth in the tilled corn field soils were destroyed by field plowing and planting activities in the spring and appeared to evolve faster at depth than at the soil surface.At 15 cm depth, the macropore abundance appears to increase steadily between spring and fall, while at the soil surface, macropore abundance increases mainly between summer and fall.This may be due to the tilled field soil surface experiencing more direct impact from rainfall while the crop was small between spring and summer.In comparison, along the drainage ditch bank, the soil was shielded by vegetation detritus on the soil surface (surface detritus protection), which can enhance earthworm activity (Ouellet et al. 2008).Warmer and drier conditions could contribute to the formation of desiccation cracks in soils, which were more pronounced in the tilled corn field, especially in the summer.
Circular macropores comprised all of the macropores found in the drainage ditch bank soil for all seasons; no linear macropores were evident.In contrast, both circular and linear macropores were observed in the tilled corn field soil, as mentioned previously.During the summer, linear macropores were observed to comprise over half of the total soil macroporosity at each bench depth in the tilled corn field, some of which reached apertures of up to 5 mm.Linear macropore features did not appear to extend as deeply in the soil profile as circular macropores.At 30 cm depth in the tilled corn field soil, below the plow pan and in the upper portions of the B horizon, only circular macropores were observed.
Like the soil macroporosity, values of air permeability, vertical hydraulic conductivity, and horizontal hydraulic conductivity were greater in the ditch bank soil than those in the tilled corn field soils, likely as a result of biopore connectivity/continuity to undisturbed surface soils (Heard et al. 1988).In situ air permeability measurements show that air permeabilities in the ditch bank soil were up to three times greater than in the tilled corn field soil.Air permeability measurements conducted on the soil surface and at 15 cm depth in the two landscape positions were considered very high.The air permeability remained high even at relatively high soil moisture contents, suggesting that macropores at this site remain effectively "open" and provided more or less temporally consistent pathways for gas advection over the growing seasons.
In general, vertical hydraulic conductivities measured in situ at the soil surface and at 15 cm depth in the ditch bank plots appeared to increase throughout the growing season with increasing total MAF, displaying most variability in the summer and fall.In the tilled corn field plots, vertical hydraulic conductivity was similar in the spring and fall with little variability.During the summer, vertical hydraulic conductivity in the tilled corn field was up to three and two orders of magnitude greater than in the spring and fall, respectively, and displayed a large degree of variation between measurements.At 15 cm depth, all but one of the attempts at vertical hydraulic conductivity measurements failed as a result of excessively rapid infiltration.This is likely caused by the high abundance of macropore features present in the tilled corn field soil in the summer.
Horizontal hydraulic conductivity measurements were approximately two times smaller than vertical hydraulic conductivity in the ditch bank soil and approximately eight times smaller in the tilled corn field soil.Pulido-Moncada et al. ( 2021) also found that saturated hydraulic conductivity was higher in the vertical direction, relative to horizontal direction, irrespective of soil compaction treatment.Yet, Soracco et al. (2010) found horizontal hydraulic conductivity was around five times higher than vertical hydraulic conductivity in no-tillage soils likely due to compaction as a result of field trafficking and lack of annual disturbance, which normally occurs during tillage.Frey and Rudolph (2011), in their study conducted on cultivated loam soil in Ontario, Canada, found that the total MAF and MAF attributable to macropores in the 0.5-5 mm diameter range was greatest in the top 0.2 m of the soil profile, and that the majority of the macroporosity appeared to terminate at a textural boundary between the A and B horizons.These observations are mirrored in this study.It was also noted that few linear macropores were present in their study plot, with the highest contribution of linear macropores to the total MAF being 15% at a depth of 20 cm.In the current study, linear macropores comprised up to 15% of the total MAF in the fall and up to 69% of the total MAF in the summer at a depth of 15 cm.Frey and Rudolph (2011) reported total MAF ranged from 0.0036 to 0.0058 in their surface plots and from 0.0026 to 0.0087 at 20 cm depth.Comparatively, the total MAF found in the tilled field soil at the surface and at 15 cm depth in this study was within a similar range.
MAF was measured to a maximum depth of 15 cm in the ditch bank soil and 30 cm in the tilled corn field soil in this study.Because macropores were relatively abundant at these depths and deeper benches were not prepared, it remains to be confirmed at what depth in the soil profile macropores terminate in these soil settings.However, Frey and Rudolph (2011) found that, at their site, the total MAF began to terminate at depths between 45 and 75 cm in the B horizon.Frey and Rudolph (2011) found that hydraulic conductivity in the A and B horizons of their study plot appeared to increase with increasing total MAF.The results of this study show that hydraulic conductivity in the tilled field soil was perhaps controlled more by the abundance of linear macropores rather than total MAF as it did not appear to follow the same trend.
The combined observations derived from these field investigations provide insight into the transient nature of soil characteristics that influence soil water and soil gas mobility in the vadose zone.Within the agricultural landscape, the dynamic movement of solutes such as nutrients and pesticides is often explored through numerical simulations conducted at scales ranging from field plots to watersheds.Conventionally, soil parameters are considered to be static during long simulation periods.This was demonstrated to not be the case in the current study.The realistic representation of soil characteristics in the shallow vadose zone is critical in establishing representative process-based and predictive models.With considerable attention currently being focused on quantifying GHG sources and global balance relative to agricultural cropping practices, the understanding of soil parameters influencing soil gas mobility is also critical and the current study provides new insight in this regard.

Conclusions
At the study site in eastern Ontario, Canada, soil macropores were found to be more abundant along an uncultivated drainage ditch bank compared to a conventionally tilled corn field.The drainage ditch macropores were exclusively consisted of circular macropores inferred to be a result of biological activity such as root growth and earthworm burrowing.In contrast, both circular and linear macropores were observed in the tilled corn field soil, with linear macropores inferred to be a result of soil desiccation fracturing.During the summer, linear macropores were observed to comprise over half of the total soil MAF in the tilled corn field but did not appear to extend as deeply in the soil profile as circular macropores.Overall, the macropore disposition and extent was seasonally variable, especially in the cultivated soils, which can significantly influence advection-based mass exchange of gas and water in the vadose zone.
Similar to soil macroporosity, air permeability, vertical hydraulic conductivity, and horizontal hydraulic conductivity were greater in ditch bank soils than in the tilled corn field soils.Air permeability remained high even at relatively high soil moisture contents in both soils, suggesting that macropores remain transmissive and provide paths of least resistance for gas advection in these soils for at least 7 months of the growing season; periods of time when fertilizer amendments, which can be sources of GHG emissions, are applied to fields.Vertical and horizontal hydraulic conductivity measurements offered clear evidence of field saturated hydraulic conductivity anisotropy in the soils under investigation.The vertical component of hydraulic conductivity was greater than the horizontal, especially in the tilled corn field soil, which was inferred to be driven by the ubiquitous vertical macropore features.It is concluded that laboratory and in situ hydraulic parameter estimations based on conventional approaches related to soil water variability can provide significant insight into the hydraulic conditions that will control soil gas mobility.These types of measurements can inform studies of GHG mobility in shallow agricultural soils.
It remains challenging to parameterize the spatial and temporal variability of soil properties; and by extension, soil mass exchange models often assume many soil properties are effectively static in space and time.Yet, the appropriate inclusion of parameter variability would make model outputs more bio-physically realistic.The current study provides data that can help in the development of empirical models designed to predict the spatial and temporal dynamics of soil properties, which can subsequently be used to inform a wide range of process-based models used to study the exchange of gas, liquid, and mass in the vadose zone.In addition, the physical and hydraulic characteristics of riparian zone soils are often not distinguished as being different from the adjacent cultivated field soils in agriculture-focused water quantity and quality modelling studies.However, this study clearly demonstrates significant differences between these two landscape settings and if not considered, could impart deleterious behavior into agricultural landscape models designed to simulate fluxes from both settings and across the field/riparian zone interface.

Fig. 3 .
Fig.3.Macroporosity measurements conducted on a prepared field soil surface during the summer.Some macropores and cracks are highlighted in the inset to the right to demonstrate the macropore identification process.

Fig. 4 .
Fig. 4. Measured daily precipitation and soil moisture at 0.15 and 0.45 m depth at the study site during the 2018 field season.

Fig. 7 .
Fig. 7. Air permeability estimates at the soil surface and 15 cm deep benches measured at the ditch bank and tilled corn field soils.

Table 1 .
Physical soil parameters and textural classification of soils at the study site.
* Three soil cores were used for dry bulk density and porosity estimation of the C horizon.

Table 2 .
Hydraulic conductivities of uncultivated drainage ditch bank and field soils calculated based on grain size distributions.

Table 3 .
Hydraulic conductivities of ditch bank and tilled field soils estimated from single ring pressure infiltrometer tests.

Table 4 .
Horizontal hydraulic conductivities estimated using the Guelph permeameter.
* Number of infiltration tests.† Arithmetic mean.

Table 5 .
Total circular macropore occurrences in the ditch bank and tilled corn field settings, along with seasonal trends.