Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Next Article in Journal
Dams as Hydrosocial Infrastructure: Attributes and Drawbacks from a Structural and Relational Perspective
Previous Article in Journal
Rapid Prediction and Inversion of Pond Aquaculture Water Quality Based on Hyperspectral Imaging by Unmanned Aerial Vehicles
Previous Article in Special Issue
Research on Flood Storage and Disaster Mitigation Countermeasures for Floods in China’s Dongting Lake Area Based on Hydrological Model of Jingjiang–Dongting Lake
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 17
Issue 4
10.3390/w17040518
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Comparison of Three Methodologies for Determining Soil Infiltration Capacity in Thicketized Oak Woodlands and Adjacent Grasslands

by
Furkan Atalar
1,2,†,
Pedro A. M. Leite
2,*,† and
Bradford P. Wilcox
2
1
Institute of Graduate Studies, Istanbul University-Cerrahpasa, 34320 Istanbul, Turkey
2
Department of Ecology and Conservation Biology, Texas A&M University, College Station, TX 77843, USA
*
Author to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Water 2025, 17(4), 518; https://doi.org/10.3390/w17040518
Submission received: 17 December 2024 / Revised: 4 February 2025 / Accepted: 10 February 2025 / Published: 12 February 2025
(This article belongs to the Special Issue Advances in Ecohydrology in Arid Inland River Basins)
Browse Figures
Review Reports Versions Notes

Abstract

:
This study had two primary objectives: (1) to determine relative differences in soil infiltration capacity between native grasslands and thicketized oak woodlands and (2) to compare the effectiveness of three infiltration measurement techniques—rainfall simulation, an automated Simplified Steady Beerkan Infiltration (SSBI) method, and the Saturo dual-head infiltrometer. The study was conducted at three sites with clay, loamy sand, and sandy soils. Rainfall simulation captured significant infiltration differences between vegetation covers at all three sites, while SSBI did so at two sites, and Saturo failed to detect significant differences. Consistent with past studies, rainfall simulation results showed significantly higher infiltration capacity in thicketized woodlands compared to adjacent grasslands, with mean infiltration capacity an order of magnitude greater in clay soils (67 mm h−1 vs. 7.5 mm h−1) and more than twice as high in sandy (144.5 mm h−1 vs. 69 mm h−1) and loamy sand (106 mm h−1 vs. 49 mm h−1) soils. Across sites, rainfall simulation and SSBI showed strong positive correlations between infiltration capacity and dead biomass (R2 = 0.74 and 0.46, respectively; p < 0.001 for both), as well as significant negative correlations with live biomass and bulk density. In contrast, the Saturo method exhibited higher variability, overestimating infiltration capacity by an average of 34.3 mm h−1 compared to rainfall simulation, and did not capture significant relationships with biomass or bulk density. Our findings have twofold importance: first, they demonstrate that thicketization of oak savannahs results in higher soil infiltration capacity; and second, they show that for determining soil infiltration capacity, the SSBI methodology is an accurate and practical alternative to the labor-intensive rainfall simulation.

1. Introduction

Thicketization is a form of woody plant encroachment (WPE) that, over time, leads to the closing of woody plant canopies. It is a common process in oak woodlands and savannahs in the central and eastern United States [1] that has been facilitated and accelerated by a combination of fire suppression and agricultural land abandonment [2,3]. It is especially prevalent in the Post Oak Savannah ecoregion of Texas, with the expansion of understory shrubs, such as Yaupon (Ilex decidua, Ilex vomitoria), as well as of Eastern red cedar trees (Juniperus virginiana) [4].
The impact of thicketization on the hydrological aspects of oak woodlands has received relatively little attention. The work that has been carried out suggests that the impact is significant. For example, in the Cross Timbers region of Oklahoma, field and modeling work has demonstrated that thicketization will lead to significant decreases in both groundwater recharge [5] and streamflows [6,7]. Similarly, Basant et al. [3] found that in sites overlying the regionally important Carrizo–Wilcox Aquifer, groundwater recharge was effectively eliminated in thicketized woodlands, whereas in open areas annual recharge rates ranging from 3 to 18 cm were recorded.
One of the most important factors that regulate the terrestrial water cycle is the infiltration capacity of soils. Commonly termed soil infiltrability, it is directly related to the soil’s hydraulic conductivity and is influenced by soil properties, such as texture, bulk density, pore structure, aggregate stability, and organic matter content [8]. Changes in vegetation cover can alter these properties, consequently altering infiltration rates and important hydrological processes, such as soil water storage, groundwater recharge, and streamflow [9]. For this reason, soil infiltration capacity is a key parameter in many hydrological and Earth system models [7].
With respect to how WPE may influence water infiltration into the soil, it has been broadly reported that, in general, infiltration capacity and soil porosity (especially macroporosity) are higher under trees or shrubs than in adjacent open areas [10,11,12,13]—most likely owing to the formation of channels by decayed roots and the incorporation of leaf litter and organic matter [13,14]. To date, little, if any, work has evaluated the influence of thicketization on soil infiltration capacity in the Post Oak Savannah. However, in the Cross Timbers region of Oklahoma, which is ecologically similar, Zou et al. [6] found that soil infiltration capacity was three times higher under Eastern red cedar canopies than in open grasslands.
Various instruments have been developed to determine soil infiltration capacity in a variety of settings—the most commonly employed being rainfall simulators and ring infiltrometers [15]. Both have advantages and limitations in measuring infiltration rates and providing data useful for understanding the underlying processes. Rainfall simulators, for example, can provide information not only on soil infiltration capacity at the plot scale [16,17,18], but also on runoff and soil erosion rates [19,20,21]. Because rainfall simulation is generally applied to areas larger than those for which ponding methods are used, it is better able to capture the spatial variability of surface and subsurface conditions [22,23]. At the same time, rainfall simulation often requires heavy equipment, large amounts of water, and more than one operator—factors that can make multiple measurements difficult.
A more practical alternative for estimating infiltration capacity is the ring infiltrometer. Double-ring infiltrometers are well established and widely used for this purpose; however, they can be labor-intensive and consume more water than single-ring devices, without necessarily providing greater accuracy [24]. A relatively simple and inexpensive single-ring methodology is the Beerkan Estimation of Soil Transfer parameters (BEST) [25,26]. This technique involves inserting a ring into the soil at a shallow depth and sequentially applying small water volumes (ensuring negligible head within the ring) until infiltration rates stabilize, indicating the establishment of a steady-state condition. To reduce manual effort and human error, constant-head infiltrometers can be employed to automate the water application and data collection processes [27,28]. Various algorithms can then be applied to the cumulative infiltration data to estimate field saturated hydraulic conductivity (Kfs, mm h−1), a widely used measure for assessing soil infiltration capacity for different vegetation covers [13,29,30]. Most of these algorithms also require soil texture, moisture, and bulk density data [26]. However, the Simplified Steady Beerkan Infiltration (SSBI) method proposed by Bagarello et al. [31] requires only an estimation of the α* value, which represents capillary length and takes into account the three-dimensional aspects of infiltration. The α* value of 0.012 mm was considered a good approximation for most field soils, but uncertainties in α* estimates can propagate into the resulting Kfs value. Another potential limitation of this method is that an incorrect assumption of steady-state conditions could result in an overestimation of the soil’s infiltration capacity [32].
Only a few commercially available infiltrometers do not require constant monitoring or post-processing of data. One notable example is the Saturo device, produced by Meter Group, which employs the two-ponding head technique developed by Reynolds and Elrick [33] to calculate Kfs. While this instrument does not require the estimation of other parameters, such as bulk density or α* value, the user must set the configurations—including pressure heads, soak and hold times, and number of pressure cycles—which can also introduce bias. Additionally, the two-ponding head method often generates invalid (negative) Kfs results, especially in more heterogeneous soils with high microporosity [34]. Although fully automated and user-friendly, the Saturo device can be costly and might not be the most feasible option for many users.
Different techniques for measuring soil infiltration capacity can yield varying results. Many studies have compared different methods, but the conclusions drawn are often contrasting and highly site-specific [15]. Rainfall simulators, for instance, have shown results both similar to [22,35,36] and differing from [23,37] those of ring infiltrometers. While such studies provide important insights into the differences between methods, comprehensive comparisons across diverse soil textures and plant communities remain limited. For example, to our knowledge, no studies have compared multiple measurement techniques in woodland and grassland areas with different soil textures. This leaves a substantial gap in understanding how contrasting natural or unmanaged vegetation covers influence the performance, accuracy, and reliability of these methods. Comparative studies addressing this knowledge gap are essential for improving our ability to select the most appropriate techniques for specific environmental conditions.
Our study had two primary objectives: (1) to test the hypothesis that thicketization of oak savannahs increases soil infiltration capacity, driven by changes in soil porosity and addition of organic matter; and (2) to compare three different methodologies for estimating soil infiltration capacity, each of which has distinct advantages and disadvantages. The three methodologies we compared are rainfall simulation with a drip rainfall simulator [38], the SSBI method employing a constant-head infiltrometer [28,39], and measurement via a Saturo dual-head infiltrometer (Meter Group Inc., Pullman, WA, USA). Rainfall simulation was selected because it is a well-established and widely accepted method for accurately measuring soil infiltration capacity. The SSBI method was included for its simplicity, affordability, and positive comparisons with more complex single-ring methods. Finally, the Saturo infiltrometer was chosen as it is likely the only fully automated commercially available infiltrometer, though its performance in field soils has not been extensively evaluated. We hypothesize that these methods will yield significantly different results due to differences in their measurement scale, underlying assumptions, and calculation approaches.

2. Study Sites

This research was conducted at three sites, each representing a different soil textural class (clay, loamy sand, and sandy) and each consisting of paired grassland- and woodland-cover zones. The clay site was located at the Texas A&M University Beef Cattle Center in College Station, Texas (30°32′49″ N, 96° 25′03″ W, 68 m above sea level); the loamy sand site was located at the Texas A&M University Ecology and Natural Resources Teaching Area in College Station, Texas (30°34′39″ N, 96°21′04″ W, 88 m above sea level); and the sandy site was located at the Gus Engeling Wildlife Management Area, 32 km northwest of Palestine, Texas (31°56′21″ N, 95°53′45″ W, 104 m above sea level). The average annual temperature and precipitation are 21 °C and 1011 mm, respectively, at the clay and loamy sand sites, and 19.6 °C and 1083 mm, respectively, at the sandy site. All three sites lie within the Post Oak Savannah ecoregion of Texas and have a humid, subtropical climate.
According to the USDA/NRCS Web Soil Survey, the clay site is a Chromic Hapluderts (Ships series), the loamy sand site is a Chromic Vertic Albaqualfs (Boonville series), and the sandy site is an Arenic Plinthic Paleudults (Lilbert series). These soil types are characteristic of the Post Oak Savanna region of Texas, and our site selection aimed to capture some of this natural soil variability. While not exhaustive, selecting three distinct soil types and contrasting vegetation covers improves the study representativeness within this landscape.
The overstory component of the woodland zones at all three sites is dominated by post oak (Quercus stellata), blackjack oak (Quercus marilandica), winged elm (Ulmus alata), and a few large individuals of Eastern red cedar (Juniperus virginiana). The dense shrub understory consists mainly of encroaching species—Yaupon holly (Ilex vomitoria), the exotic Chinese privet (Ligustrum sinense), and young individuals of Eastern red cedar. The little herbaceous cover in the woodland zones is mostly C3 grasses such as Inland sea oats (Chasmanthium latifolium) and a few forb species.
The grassland zones at the loamy sand and sandy sites are dominated by native C4 grasses, including little bluestem (Schizachyrium scoparium) and switchgrass (Panicum virgatum), and many forb species. Neither of these sites has any recent history of grazing. At the clay site, the grassland zones are composed mainly of the introduced bermudagrass (Cynodon dactylon). This site experiences rotational grazing annually at relatively low stocking densities (approximately 1 animal unit/ha).

3. Materials and Methods

3.1. Experimental Design

At each site and within each cover type, we selected five measurement locations, spaced three meters apart along a linear transect. The transects for grassland and woodland zones were drawn parallel to each other and approximately 10 m from the boundary separating the two cover types. All measurements were conducted between April and May during the early growing season to minimize potential effects of seasonal variability. At each measurement location, a 65 cm × 65 cm runoff plot was established, enclosed by galvanized steel sheets inserted to a shallow depth (approximately 5 cm), for the rainfall simulation experiments. After the simulations had been completed, two corners of the plot were randomly selected for infiltration capacity measurements with the SSBI method, and a third corner was chosen for measurements with the Saturo infiltrometer (Figure 1). Figure 2 shows photographs of the three methodologies deployed in two different vegetation cover settings.

3.2. Infiltration Capacity Measurement Methods and Instruments

3.2.1. Rainfall Simulation Method

We utilized a drip-type rainfall simulator that has been widely employed by previous researchers [12,40,41,42] and is described in detail by Blackburn et al. [38]. The simulator (see Figure 2a) is placed approximately 40 cm above the soil surface and delivers droplets of water to a 90 cm × 90 cm surface area at a maximum rate of 154 mm h−1. At each plot, we conducted a rainfall simulation over the 65 cm × 65 cm surface area. Because we expected high infiltration capacity, especially in the woodland zones, we applied the maximum rainfall intensity of 154 mm h−1 in order to maximize the chances of obtaining runoff. Runoff was collected for 30 s immediately after it commenced, followed by 30 s collections every 5 min until four consecutive runoff volumes were approximately the same. These four volumes were averaged to calculate a terminal runoff value, which was then subtracted from the rainfall intensity to obtain the steady-state infiltration rate [43,44]. The steady-state infiltration rate can be considered a good approximation of the soil’s infiltration capacity, or Kfs [43], particularly under the assumption that lateral matrix flow is negligible owing to the buffering effect created by water falling outside the plot [45].

3.2.2. Automated Simplified Steady Beerkan Infiltration (SSBI) Method

The automated infiltrometers described in Leite et al. [28] were used to determine the Kfs values, which were calculated from steady-state infiltration rates obtained via the SSBI method [31]. A 10 cm diameter ring was inserted to a depth of 1 cm and the infiltrometers (see Figure 2b), filled with 2.2 L of water, were used to sustain a hydraulic head of about 1 cm within the ring. The test was conducted for a period of 60 min or until all water drained from the infiltrometer. Pressure readings were recorded every five seconds by a HOBO U20 datalogger (Onset, Bourne, MA, USA) and were subsequently used to generate cumulative infiltration curves via the calibration function provided in Leite et al. [28]. Steady-state infiltration rates (is; mm h−1) were then derived from the linear (stable) segments of the infiltration curves and used to calculate Kfs (mm h−1) according to Bagarello et al. [31], as follows
K f s = i s y y w r α ∗ + 1 ,
where r denotes the radius of the ring in millimeters, y and y w are dimensionless constants, and α* is an empirical parameter that captures the effects of gravitational and capillary forces. We adopted an α* value of 0.012 mm, which is considered a good approximation for most field soils [31]. We performed two tests per plot one week after the rainfall simulations. The Kfs values for each plot represent the average of these two measurements.

3.2.3. Saturo Method

A Saturo dual-head infiltrometer (Meter Group Inc., Pullman, WA, USA) with a 14.4 cm (inner diameter) ring and a 5 cm insertion depth was employed to measure Kfs at one point per plot (see Figure 2c). An increasing number of studies have been adopting this instrument for evaluating soil infiltration capacity in engineered soils [46,47] and agricultural settings [48,49]. Because it is a relatively new instrument, there are few studies comparing the Saturo with other methods; however, one recent laboratory study on homogenous soil beds concluded that it yields results comparable with those obtained from rainfall simulators [36].
The instrument is fully automated and allows for setting different configurations—such as pressure heads, soak and hold times, and number of pressure cycles. We used pressure heads of 10 and 100 mm H2O, a soak time and hold time of 10 min, and three pressure cycles, for a total test duration of 70 min. These configurations were selected to strike a balance between water consumption and test duration (higher pressure heads and longer hold times would have increased both water usage and the total duration of each test). Additionally, we sought to align the test duration with those of the other two methods to improve comparability across methodologies.
The final Kfs value is automatically calculated by the instrument by means of the simplified version of the two-ponding head technique developed by Reynolds and Elrick [33], as follows:
K f s = ∆ ( i 1 − i 2 ) D 1 − D 2 ,
where D1 represents the higher-pressure head (mm H2O), D2 denotes the lower-pressure head (mm H2O), and Δ is calculated as 0.993 times the ring insertion depth (mm) plus 0.578 times the ring radius (mm). The variables i1 and i2 are the steady-state infiltration rates (mm h−1) at D1 and D2, respectively.

3.3. Additional Data

For each test, we also collected 5 cm diameter soil cores at the 0 to 5 cm and 5 to 10 cm depth intervals for measuring bulk density via a standard oven-drying technique (105 °C for 48 h). After the rainfall simulations and prior to the other infiltration tests, we collected all the live (clipped at ground level) and dead plant tissues from the surface of each plot to determine live and dead plant biomass. The samples were taken to the laboratory and oven-dried at 65 °C for 48 h.

3.4. Data Analysis

Infiltration capacity differences between vegetation cover types (woodland vs. grassland) for the three sites (clay soil, loamy sand soil, and sandy soil) were tested by means of t-tests, and comparisons of the three measurement methodologies were analyzed with one-way ANOVA and Tukey’s HSD post-hoc test. Prior to these tests, skewness and heteroscedasticity were taken into account by the application of log transformations that met parametric assumptions. We then conducted simple linear regressions to evaluate how vegetation biomass (live and dead) and bulk density correlated with infiltration capacity values obtained with each method. Finally, we conducted paired t-tests to assess whether, across all sites and treatments, the infiltration capacity values obtained from the automated SSBI and Saturo methods differed significantly from those obtained via rainfall simulation. All analyses were performed using R version 4.4.1.

4. Results

4.1. Rainfall Simulation Method

The results obtained with the rainfall simulator showed significantly higher infiltration rates for the thicketized woodland zones than for the grassland zones at all three sites (clay, loamy sand, and sandy) (Figure 3, Table 1). Specifically, at the clay soil sites, mean infiltration capacity values in the WPE-affected zones were an order of magnitude higher than those in the grassland zones; and at the loamy sand and sandy soil sites they were more than twice as high as those in grassland zones. At the sandy site, infiltration capacity for the thicketized woodland exceeded rainfall intensity (154 mm h−1) in four of the five plots. The mean from the rainfall simulation is likely underestimated, as there was no runoff from those four plots, even at the maximum simulated rainfall intensity of 154 mm h−1. For this reason, statistical comparisons between methodologies were not performed for the woodland tests at this site.

4.2. Automated SSBI Method

With the automated SSBI method, we found significantly higher infiltration capacity for the woodland-cover zones than for the grassland-cover zones at the clay and sandy soil sites, but no significant difference at the loamy sand site (Figure 4, Table 1). For the clay and sandy soils, mean infiltration capacity in woodland zones was an order of magnitude higher than in grassland zones.

4.3. Saturo Method

Measurements with the Saturo infiltrometer indicated higher median infiltration capacity in thicketized woodlands for the clay and sandy soil sites (Figure 4), but in both cases the difference was not significant (Table 1) because of the higher variability of this instrument. Similarly, for the grassland zone at the clay soil site, the Saturo-obtained mean infiltration capacity was an order of magnitude higher than those obtained with the other two methods; and for the grassland zone at the sandy soil site, the Saturo-obtained mean value was 1.6 times higher than that obtained via rainfall simulation and 8 times higher than that obtained with the automated SSBI method (Table 1).

4.4. Influence of Biomass and Bulk Density on Infiltration Capacity

With respect to the influence of biomass on infiltration capacity, we found that the rainfall simulation and automated SSBI methods both indicated a significant negative correlation (adj R2 = 0.32, p < 0.001 and adj R2 = 0.28, p = 0.002, respectively) between live biomass and infiltration capacity (Figure 5)—which suggests that higher live biomass was associated with reduced infiltration rates. The Saturo method, however, showed an insignificant correlation (adj R2 = -0.03, p = 0.874), indicating that this method was not sensitive to differences in live biomass. Additionally, both the rainfall simulation (adj. R2 = 0.74, p = 0.001) and the automated SSBI (adj. R2 = 0.46, p < 0.001) methods showed strong positive correlations between dead biomass and infiltration capacity, suggesting that an increase in dead biomass significantly enhances soil infiltration. This effect is likely due to decomposed organic matter and aggregate breakdown, which slow runoff and facilitate infiltration. The Saturo method, on the other hand, showed a result similar to that for live biomass, i.e., no significant relationship between dead biomass and infiltration capacity (adj. R2 = 0.02, p = 0.213).
Finally, with respect to the effects of bulk density on infiltration, both rainfall simulation and the automated SSBI method yielded significant negative correlations, on the basis of measurements at two depth intervals (0–5 cm and 5–10 cm). For the 0 to 5 cm depth, the adjusted R2 values for rainfall simulation and for automated SSBI were, respectively, 0.25 (p = 0.003) and 0.65 (p < 0.001); and in the 5 to 10 cm depth, they were 0.37 (p < 0.001) and 0.65 (p < 0.001)—implying that higher bulk density reduces infiltration capacity. In contrast, the Saturo method showed non-significant relationships at both the 0 to 5 cm and 5 to 10 cm depths, with adjusted R2 values of 0.08 (p = 0.067) and 0.03 (p = 0.166), respectively.

4.5. Differences Among Methodologies

With respect to variability, rainfall simulation generally showed the least of the three methods. Coefficients of variability ranged from 15.5% to 54.3% for the rainfall simulator, from 15.6% to 128.9% for the automated SSBI, and from 39.3% to 152.5% for the Saturo. When we compared the rainfall simulation and automated SSBI infiltration capacity values for all sites and both cover types, the paired t-test yielded a t-value of −0.49, a p-value of 0.627, and a mean difference of −5.42 mm h−1—that is, there was no significant difference between the two methods. Conversely, when we compared the rainfall simulation and the Saturo methods in the same way, the paired t-test revealed a significant difference (t = 2.15, p = 0.041), with the Saturo method producing higher infiltration capacity values on average (mean difference = 34.28 mm h−1). In other words, the Saturo method may overestimate infiltration capacity relative to rainfall simulation.

5. Discussion

We designed this study with two objectives in mind. The first was to evaluate differences in soil infiltration capacity in thicketized woodlands vs. more open grasslands in the Post Oak Savannah, and the second was to compare three field-based methods for determining soil infiltration capacity.

5.1. Effects of Thicketization on Soil Infiltration Capacity

We found that, regardless of soil texture, soil infiltration capacity was higher in thicketized woodlands than in open grasslands. This finding is based primarily on the results obtained via rainfall simulation, which (as we explain below) we consider the gold standard for determining soil infiltration capacity. In addition, this finding is generally supported by the other methodologies we tested.
Higher infiltration capacity in the thicketized woodland zones is broadly consistent with what has been demonstrated elsewhere [6,12,13,20,30]. The closest comparison to our site would be the Cross Timbers region, where juniper is encroaching upon grasslands and open woodlands. As noted previously, Zou et al. [6] found that infiltration capacity was three times higher in juniper-encroached areas than in open grasslands—most likely as a result of the addition of organic matter from leaf litter.
Notably, we found that the highest contrast in infiltration capacity between encroached woodlands and grasslands occurred at the clay site, which was also the only site subject to grazing. This aligns with previous studies showing that grazing reduces infiltration capacity [12,40] and that the contrast between woody and grass cover is most pronounced in the presence of grazing [50].
While increased infiltration capacity may benefit soil water retention and aboveground biomass productivity, it does not necessarily translate into higher streamflow or groundwater recharge, particularly in thicketized landscapes. For example, in regions like the Cross Timbers, thicketization has been linked to lower streamflows due to enhanced evapotranspiration and reduced surface runoff, which is the primary contributor to streamflow in the region [5,6]. Similarly, Basant et al. [3] found that in a Post Oak Savannah site overlying deep sands, thicketization was associated with greater rooting depths and effectively eliminated groundwater recharge. Nonetheless, higher infiltration under woody plants can have positive aspects. At our clay site, for instance, the extremely low infiltration capacity of the grassland (<10 mm/h) can lead to substantial runoff generation during storms, contributing to high erosion rates, nutrient and contaminant transport, and flash floods. Therefore, maintaining areas of thicketized woodlands within grazed grasslands and pastures could be beneficial for capturing runoff, particularly near riparian corridors, swales, and drainage depressions prone to gully erosion.
As shown in Figure 5, soil infiltration capacity at our sites was highly correlated with dead biomass or leaf litter, a finding that also aligns with those of numerous other studies reporting that litter enhances soil infiltration capacity [51,52,53]. The lower bulk densities for soils in thicketized areas (Figure 5) are also likely a reflection of additions of organic matter. The negative relationship between bulk density, which is an expression of soil compaction, and soil infiltration capacity has been shown repeatedly [13,50,54,55].
One counter-intuitive result that we found was a negative correlation between infiltration capacity and live biomass (Figure 5). For the thicketized zones, the fact that there was relatively little live biomass on the ground surface compared with the grassland zones explains the negative correlation. Wilcox et al. [16] found grass biomass to be positively correlated with soil infiltration capacity, whereas a meta-analysis by Thompson et al. [15] reported both positive and non-significant relationships, depending on climate. In contrast, findings from other studies have suggested that the dense root systems associated with high levels of herbaceous cover may contribute to reduced free pore spaces, which impedes the movement of water into the soil [56,57].

5.2. Comparison of Methodologies

Rainfall simulators have been widely used to measure surface runoff and infiltration rates across areas that differ with respect to land use and vegetation cover. As noted above, rainfall simulation is often considered the gold standard for determining soil infiltration capacity [22,58,59,60] because it is more comparable to natural rainfall than ponding methods and its gradual wetting of the soil avoids problems such as air entrapment and rapid soil slaking [58]. Another advantage over ring infiltration methods is that rainfall simulations are normally performed over a larger soil surface, enabling them to capture greater spatial variability of factors that influence infiltration capacity (e.g., microtopography and infiltration hotspots such as soil macropores). The net result is less variability in infiltration capacity values [22,23,61]. In the case of our study, the larger measurement area of the rainfall simulations offers a plausible explanation for the generally higher values than those obtained with the SSBI method, as well as the lower variability in values than observed with both the SSBI and Saturo methods.
Some types of rainfall simulators attempt to mimic natural rainfall events to take into account the effect of raindrops’ kinetic energy on aggregate breakdown and soil sealing. In our study, we opted to perform simulations from a short distance above the ground surface to make results more comparable with those obtained by the two ring infiltrometers. Nonetheless, one potential reason for differences between the rainfall simulation results and those of the other methods could be the protective effect of leaf litter and herbaceous cover. Because leaf litter and herbaceous cover—which are known to protect the soil against surface sealing by soil aggregate breakdown [52,53,62]—were removed for the infiltration tests with the automated SSBI and Saturo instruments, only the rainfall simulations would have benefited from this protective effect. This could partially explain why the automated SSBI infiltrometers yielded lower infiltration capacity values for the sandy soils. Previous studies have demonstrated that surface sealing can lead to major reductions in infiltration capacity [44], and that the effect can be accentuated for coarser-textured soils (which tend to have lower aggregate stability than finer-textured soils) [63].
One drawback of rainfall simulators is that their maximum rainfall intensity imposes an upper limit on measurable infiltration capacity values—in our case, 154 mm h−1. As a result, soils with exceptionally high infiltration capacity, such as those observed in the woodland zone at our sandy site, may exceed this limit, leading to imprecise measurements and potentially undermining meaningful statistical analysis. Additionally, they are labor-intensive and often require large volumes of water, making them less suitable for use in areas with difficult access and limiting the number of replicates.
The Saturo infiltrometer, while fully automated and capable of providing Kfs values without the need for additional data processing, shares some of these drawbacks. Like a rainfall simulator, it is relatively bulky and water-demanding, limiting its applicability in certain field situations. Further, with the Saturo device it is crucial to monitor the equipment carefully; if the water runs out, the test can be interrupted and the results lost. Additionally, the Saturo’s complexity and reliance on interconnected components make it prone to malfunctions requiring specialized servicing, further increasing costs and downtime.
Consistent with what has been found elsewhere, the Saturo infiltrometer generally yielded higher infiltration capacity values than the other methods [64]. This difference could be explained by the methodological approach of the Saturo. In contrast to rainfall simulation and the automated SSBI method, with which ponding and water-pressure heads are mostly negligible, the Saturo requires at least one relatively high-pressure head (10 cm in our study)—which could exaggerate the role of preferential flow. Higher-pressure heads could cause water to flow to a disproportionately greater extent through preferential pathways, such as macropores or cracks in the soil, than through the soil matrix [65]. And the fact that these pathways are more common in finer-textured soils [66] would explain why the difference in values was particularly striking at the clay soil site.
Compared with the SSBI method, the Saturo method involves the insertion of heavier-gauge rings deeper into the soil, which causes greater disturbance to the soil structure, potentially masking effects of vegetation cover on surface hydraulic properties. This may also explain the lack of correlation between Saturo data and variables such as vegetation biomass or bulk density. In a study conducted across 124 sites in North America, [49] used the Saturo to measure Kfs and found that it was not significantly affected by land management or soil properties—a finding that contradicts a substantial body of research [12,39,67,68,69,70]. While the authors acknowledged that the limited sampling size (one point per experimental unit) and the inherently high variability of Kfs likely influenced their results, we further propose that their use of the Saturo may also have contributed to these unexpected findings.
For the automated SSBI method, a potential source of uncertainty in the Kfs values obtained is the choice of the α* value. However, for most applications, large variations in the α* value would have relatively small impacts on Kfs estimates [71]. For example, even a 10-fold overestimation of the α* value would result in just a 50% error in Kfs estimates [72]. Within each site, the relative differences in Kfs between vegetation covers should be mostly unaffected by the choice of α* value, as the soil class remains the same. For our clay and loamy sand sites, the Kfs values obtained with the SSBI method were within the same range as those obtained by other methods, suggesting that the recommended first-approximation α* value of 0.012 mm was appropriate. However, for the sandy site, the SSBI method yielded Kfs values lower than those of the other methods, and this may have been caused in part by underestimation of the α* value. Thus, with the SSBI method, caution is warranted in using a fixed α* value to compare Kfs across sites having different soil types.
Despite potential uncertainties, the automated SSBI method’s ability to collect more replicates in a shorter time offers a clear advantage over both rainfall simulation and the Saturo method. Being lightweight, requiring less water, and producing results within a range comparable to those of the other methods—while generally aligning more closely with rainfall simulation than with Saturo—it offers a practical compromise between the two.
As expected from a single-ring method, the infiltration capacity data obtained with the SSBI method exhibit greater variability than those from rainfall simulation. However, the automated SSBI remains more affordable and significantly lighter than the other two instruments. During our study, one operator using 10 automated SSBI infiltrometers (at the cost of a single Saturo unit) was able to collect all 20 data points per site in less than four hours. In contrast, collection of 10 data points with either a Saturo or a rainfall simulator would take two days because of the time required for setup and testing. Additionally, the automated SSBI required significantly less water per test (2.2 L or less) than either the Saturo (5–30 L) or the rainfall simulator (60–80 L). This reduced water demand is especially advantageous in remote or water-scarce environments, where transporting large volumes of water may be impractical.

6. Summary and Conclusions

This study compared infiltration capacity determined through the use of three different measurement methodologies: rainfall simulation, automated SSBI, and the Saturo dual-head infiltrometer method. The tests were carried out at three experimental sites having soils of different textures (clay, loamy sand, and sandy) and each having both a grassland and a woodland component. Measurements done with the rainfall simulator revealed significantly higher infiltration capacity for the woodland-cover zones than for the grassland zones across all three soil textures. For the woodland zones of the sites with loamy sand and sandy soils, mean infiltration capacity was more than double that found in the grassland zones, and for the woodland zone of the clay soil site it was nearly ten times higher than for the grassland zone.
Paired t-tests using data from all three sites showed no significant difference in infiltration capacity between the rainfall simulation and automated SSBI methods (t = −0.49, p = 0.627). In contrast, the test results showed a significant difference for the Saturo method (t = 2.15, p = 0.041), indicating a slight overestimation compared with rainfall simulation. For both the rainfall simulation and automated SSBI methods, regressions analyses revealed significant correlations between infiltration capacity and influencing factors (e.g., live and dead biomass, bulk density), whereas no such correlations were observed for the Saturo.
Each of the three methods tested in this study has its advantages and disadvantages and was able to generate infiltration capacity values within a comparable range. However, with respect to ability to capture differences between woodland and grassland covers, the three differed substantially. The Saturo infiltrometer, while fully automated, showed high variability and failed to capture significant differences between covers. Rainfall simulation yielded measurements with lower variability and showing significant differences between grassland and woodland cover on all three sites—but its labor-intensive and water-demanding nature limits its practicality. The automated SSBI method, despite also showing high variability, captured differences in infiltration capacity between cover types at two of the three sites. This method also offers several advantages over the other two, including the ability to collect a larger number of replicates in less time, portability, and reduced water requirements. For soils with natural vegetation cover and high spatial heterogeneity, the SSBI method may be a practical alternative to rainfall simulation.

Author Contributions

Conceptualization, P.A.M.L. and B.P.W.; Methodology, P.A.M.L.; Formal analysis, F.A. and P.A.M.L.; Investigation, F.A. and P.A.M.L.; Writing—original draft, F.A. and P.A.M.L.; Writing—review & editing, P.A.M.L. and B.P.W.; Supervision, B.P.W.; Project administration, B.P.W.; Funding acquisition, B.P.W. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Institute of Food and Agriculture, grant number 13386861.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Scifres, C.J.; Kelly, D. Range Vegetation Response to Burning Thicketized Live Oak Savannah. In Texas FARMER Collection; Texas Agricultural Experiment Station: Weslaco, TX, USA, 1979. [Google Scholar]
  2. Dyksterhuis, E. The Savannah Concept and Its Use. Ecology 1957, 38, 435–442. [Google Scholar] [CrossRef]
  3. Basant, S.; Wilcox, B.P.; Parada, C.; Wyatt, B.M.; Newman, B.D. Thicketized Oak Woodlands Reduce Groundwater Recharge. Sci. Total Environ. 2023, 862, 160811. [Google Scholar] [CrossRef] [PubMed]
  4. Olariu, H.G.; Wilcox, B.P.; Popescu, S.C. Examining Changes in Woody Vegetation Cover in a Human-Modified Temperate Savanna in Central Texas between 1996 and 2022 Using Remote Sensing. Front. For. Glob. Change 2024, 7, 1396999. [Google Scholar] [CrossRef]
  5. Acharya, B.S.; Hao, Y.; Ochsner, T.E.; Zou, C.B. Woody Plant Encroachment Alters Soil Hydrological Properties and Reduces Downward Flux of Water in Tallgrass Prairie. Plant Soil 2017, 414, 379–391. [Google Scholar] [CrossRef]
  6. Zou, C.B.; Turton, D.J.; Will, R.E.; Engle, D.M.; Fuhlendorf, S.D. Alteration of Hydrological Processes and Streamflow with Juniper (Juniperus virginiana) Encroachment in a Mesic Grassland Catchment. Hydrol. Process. 2014, 28, 6173–6182. [Google Scholar] [CrossRef]
  7. Qiao, L.; Zou, C.B.; Will, R.E.; Stebler, E. Calibration of SWAT Model for Woody Plant Encroachment Using Paired Experimental Watershed Data. J. Hydrol. 2015, 523, 231–239. [Google Scholar] [CrossRef]
  8. Hillel, D. Soil and Water: Physical Principles and Processes; Elsevier: Amsterdam, The Netherlands, 2012; ISBN 0-323-15670-3. [Google Scholar]
  9. Ding, J.; Eldridge, D.J. Contrasting Global Effects of Woody Plant Removal on Ecosystem Structure, Function and Composition. Perspect. Plant Ecol. Evol. Syst. 2019, 39, 125460. [Google Scholar] [CrossRef]
  10. Eldridge, D.J.; Maestre, F.T.; Maltez-Mouro, S.; Bowker, M.A. A Global Database of Shrub Encroachment Effects on Ecosystem Structure and Functioning: Ecological Archives E093-234. Ecology 2012, 93, 2499. [Google Scholar] [CrossRef]
  11. Bargués Tobella, A.; Reese, H.; Almaw, A.; Bayala, J.; Malmer, A.; Laudon, H.; Ilstedt, U. The Effect of Trees on Preferential Flow and Soil Infiltrability in an Agroforestry Parkland in Semiarid Burkina Faso. Water Resour. Res. 2014, 50, 3342–3354. [Google Scholar] [CrossRef]
  12. Basant, S.; Wilcox, B.P.; Leite, P.M.; Morgan, C.L. When Savannas Recover from Overgrazing, Ecohydrological Connectivity Collapses. Environ. Res. Lett. 2020, 15, 054001. [Google Scholar] [CrossRef]
  13. Leite, P.A.M.; Wilcox, B.P.; McInnes, K.J. Woody Plant Encroachment Enhances Soil Infiltrability of a Semiarid Karst Savanna. Environ. Res. Commun. 2020, 2, 115005. [Google Scholar] [CrossRef]
  14. Wu, G.-L.; Cui, Z.; Huang, Z. Contribution of Root Decay Process on Soil Infiltration Capacity and Soil Water Replenishment of Planted Forestland in Semi-Arid Regions. Geoderma 2021, 404, 115289. [Google Scholar] [CrossRef]
  15. Thompson, S.E.; Harman, C.J.; Heine, P.; Katul, G.G. Vegetation-infiltration Relationships across Climatic and Soil Type Gradients. J. Geophys. Res. 2010, 115, 2009JG001134. [Google Scholar] [CrossRef]
  16. Wilcox, B.P.; Wood, M.K.; Tromble, J.M. Factors Influencing Infiltrability of Semiarid Mountain Slopes. J. Range Manag. 1988, 41, 197. [Google Scholar] [CrossRef]
  17. Cerda, A.; Schnabel, S.; Ceballos, A.; Gomez-Amelia, D. Soil Hydrological Response under Simulated Rainfall in the Dehesa Land System (Extremadura, SW Spain) under Drought Conditions. Earth Surf. Process. Landf. 1998, 23, 195–209. [Google Scholar] [CrossRef]
  18. Leite, P.A.M.; de Souza, E.S.; dos Santos, E.S.; Gomes, R.J.; Cantalice, J.R.; Wilcox, B.P. The Influence of Forest Regrowth on Soil Hydraulic Properties and Erosion in a Semiarid Region of Brazil. Ecohydrology 2018, 11, e1910. [Google Scholar] [CrossRef]
  19. Seyfried, M.S. Infiltration Patterns from Simulated Rainfall on a Semiarid Rangeland Soil. Soil Sci. Soc. Am. J. 1991, 55, 1726–1734. [Google Scholar] [CrossRef]
  20. Taucer, P.I.; Munster, C.L.; Wilcox, B.P.; Owens, M.K.; Mohanty, B.P. Large-Scale Rainfall Simulation Experiments on Juniper Rangelands. Trans. ASABE 2008, 51, 1951–1961. [Google Scholar] [CrossRef]
  21. Williams, C.J.; Pierson, F.B.; Kormos, P.R.; Al-Hamdan, O.Z.; Nouwakpo, S.K.; Weltz, M.A. Vegetation, Hydrologic, and Erosion Responses of Sagebrush Steppe 9 Yr Following Mechanical Tree Removal. Rangel. Ecol. Manag. 2019, 72, 47–68. [Google Scholar] [CrossRef]
  22. Gupta, R.K.; Rudra, R.P.; Dickinson, W.T.; Patni, N.K.; Wall, G.J. Comparison of Saturated Hydraulic Conductivity Measured by Various Field Methods. Trans. ASAE 1993, 36, 51–55. [Google Scholar] [CrossRef]
  23. Vlček, L.; Šípek, V.; Zelíková, N.; Čáp, P.; Kincl, D.; Vopravil, J. Water Retention and Infiltration Affected by Conventional and Conservational Tillage on a Maize Plot; Rainfall Simulator and Infiltrometer Comparison Study. Agric. Water Manag. 2022, 271, 107800. [Google Scholar] [CrossRef]
  24. Verbist, K.; Torfs, S.; Cornelis, W.M.; Oyarzún, R.; Soto, G.; Gabriels, D. Comparison of Single- and Double-Ring Infiltrometer Methods on Stony Soils. Vadose Zone J. 2010, 9, 462–475. [Google Scholar] [CrossRef]
  25. Lassabatère, L.; Angulo-Jaramillo, R.; Soria Ugalde, J.M.; Cuenca, R.; Braud, I.; Haverkamp, R. Beerkan Estimation of Soil Transfer Parameters through Infiltration Experiments—BEST. Soil Sci. Soc. Am. J. 2006, 70, 521. [Google Scholar] [CrossRef]
  26. Angulo-Jaramillo, R.; Bagarello, V.; Di Prima, S.; Gosset, A.; Iovino, M.; Lassabatere, L. Beerkan Estimation of Soil Transfer Parameters (BEST) across Soils and Scales. J. Hydrol. 2019, 576, 239–261. [Google Scholar] [CrossRef]
  27. Di Prima, S. Automated Single Ring Infiltrometer with a Low-Cost Microcontroller Circuit. Comput. Electron. Agric. 2015, 118, 390–395. [Google Scholar] [CrossRef]
  28. Leite, P.A.M.; Di Prima, S.; Schmidt, L.M.; Wilcox, B.P. A Simple Infiltrometer Automated with a User-friendly Pressure Datalogger. Vadose Zone J. 2024, 23, e20366. [Google Scholar] [CrossRef]
  29. Wilcox, B.P.; Breshears, D.D.; Turin, H.J. Hydraulic Conductivity in a Piñon-Juniper Woodland: Influence of Vegetation. Soil Sci. Soc. Am. J. 2003, 67, 1243–1249. [Google Scholar] [CrossRef]
  30. Eldridge, D.J.; Wang, L.; Ruiz-Colmenero, M. Shrub Encroachment Alters the Spatial Patterns of Infiltration. Ecohydrology 2015, 8, 83–93. [Google Scholar] [CrossRef]
  31. Bagarello, V.; Di Prima, S.; Iovino, M. Estimating Saturated Soil Hydraulic Conductivity by the near Steady-State Phase of a Beerkan Infiltration Test. Geoderma 2017, 303, 70–77. [Google Scholar] [CrossRef]
  32. Bagarello, V.; David, S.M. Run Duration Effects on the Hydrodynamic Properties of a Loam Soil Estimated by Steady-State Infiltration Methods. J Agric. Eng. 2020, 51, 229–238. [Google Scholar] [CrossRef]
  33. Reynolds, W.D.; Elrick, D.E. Ponded Infiltration From a Single Ring: I. Analysis of Steady Flow. Soil Sci. Soc. Am. J. 1990, 54, 1233–1241. [Google Scholar] [CrossRef]
  34. Bagarello, V.; Iovino, M.; Lai, J.-B. Field and Numerical Tests of the Two-Ponding Depth Procedure for Analysis of Single-Ring Pressure Infiltrometer Data. Pedosphere 2013, 23, 779–789. [Google Scholar] [CrossRef]
  35. Lewis, J.; Amoozegar, A.; McLaughlin, R.A.; Heitman, J.L. Comparison of Cornell Sprinkle Infiltrometer and Double-ring Infiltrometer Methods for Measuring Steady Infiltration Rate. Soil Sci. Soc. Am. J 2021, 85, 1977–1984. [Google Scholar] [CrossRef]
  36. Naik, A.P.; Norbu, T.; Pekkat, S. Comparison of Flux-Based and Head-Based Methods for Determination of near-Surface Saturated Hydraulic Conductivity. Hydrol. Sci. J. 2024, 69, 275–293. [Google Scholar] [CrossRef]
  37. Di Prima, S.; Bagarello, V.; Lassabatere, L.; Angulo-Jaramillo, R.; Bautista, I.; Burguet, M.; Cerdà, A.; Iovino, M.; Prosdocimi, M. Comparing Beerkan Infiltration Tests with Rainfall Simulation Experiments for Hydraulic Characterization of a Sandy-Loam Soil. Hydrol. Process. 2017, 31, 3520–3532. [Google Scholar] [CrossRef]
  38. Blackburn, W.H.; Meeuwig, R.O.; Skau, C.M. A Mobile Infiltrometer for Use on Rangeland. Rangel. Ecol. Manag./J. Range Manag. Arch. 1974, 27, 322–323. [Google Scholar] [CrossRef]
  39. Leite, P.A.M.; Castellanos, A.E.; Wilcox, B.P.; Vega-Puga, M.; Martínez, E.; Dennis, S.; Choza, S.; Acuña-Acosta, D.M. Contrasting Effects of Native and Exotic Vegetation on Soil Infiltrability in the Sonoran Desert. Sci. Total Environ. 2022, 852, 158544. [Google Scholar] [CrossRef] [PubMed]
  40. McCalla, G.R.; Blackburn, W.H.; Merrill, L.B. Effects of Livestock Grazing on Infiltration Rates, Edwards Plateau of Texas. J. Range Manag. 1984, 37, 265. [Google Scholar] [CrossRef]
  41. Knight, R.W.; Blackburn, W.H.; Merrill, L.B. Characteristics of Oak Mottes, Edwards Plateau, Texas. J. Range Manag. 1984, 37, 534. [Google Scholar] [CrossRef]
  42. Wilcox, B.P.; Wilding, L.P.; Woodruff, C.M. Soil and Topographic Controls on Runoff Generation from Stepped Landforms in the Edwards Plateau of Central Texas. Geophys. Res. Lett. 2007, 34, L24S24. [Google Scholar] [CrossRef]
  43. White, I.; Sully, M.J.; Melville, M.D. Use and Hydrological Robustness of Time-to-Incipient-Ponding. Soil Sci. Soc. Am. J. 1989, 53, 1343–1346. [Google Scholar] [CrossRef]
  44. Di Prima, S.; Concialdi, P.; Lassabatere, L.; Angulo-Jaramillo, R.; Pirastru, M.; Cerdà, A.; Keesstra, S. Laboratory Testing of Beerkan Infiltration Experiments for Assessing the Role of Soil Sealing on Water Infiltration. CATENA 2018, 167, 373–384. [Google Scholar] [CrossRef]
  45. Bowyer-Bower, T.A.S.; Burt, T.P. Rainfall Simulators for Investigating Soil Response to Rainfall. Soil Technol. 1989, 2, 1–16. [Google Scholar] [CrossRef]
  46. Ebrahimian, A.; Sample-Lord, K.; Wadzuk, B.; Traver, R. Temporal and Spatial Variation of Infiltration in Urban Green Infrastructure. Hydrol. Process. 2020, 34, 1016–1034. [Google Scholar] [CrossRef]
  47. Tecca, N.P.; Nieber, J.; Gulliver, J. Bias of Stormwater Infiltration Measurement Methods Evaluated Using Numerical Experiments. Vadose Zone J. 2022, 21, e20210. [Google Scholar] [CrossRef]
  48. Norris, C.E.; Bean, G.M.; Cappellazzi, S.B.; Cope, M.; Greub, K.L.H.; Liptzin, D.; Rieke, E.L.; Tracy, P.W.; Morgan, C.L.S.; Honeycutt, C.W. Introducing the North American Project to Evaluate Soil Health Measurements. Agron. J. 2020, 112, 3195–3215. [Google Scholar] [CrossRef]
  49. Bagnall, D.K.; Morgan, C.L.S.; Bean, G.M.; Liptzin, D.; Cappellazzi, S.B.; Cope, M.; Greub, K.L.H.; Rieke, E.L.; Norris, C.E.; Tracy, P.W.; et al. Selecting Soil Hydraulic Properties as Indicators of Soil Health: Measurement Response to Management and Site Characteristics. Soil Sci. Soc. Am. J. 2022, 86, 1206–1226. [Google Scholar] [CrossRef]
  50. Wilcox, B.P.; Caldeira, M.C.; Leite, P.A.M.; Lobo-do-Vale, R.; Bugalho, M.N. Understory Shrubs Improve Soil Infiltrability in Overgrazed Mediterranean Oak Woodlands, but Have Little Impact on Ungrazed Woodlands. For. Ecol. Manag. 2024, 569, 122186. [Google Scholar] [CrossRef]
  51. Marín-Castro, B.E.; Negrete-Yankelevich, S.; Geissert, D. Litter Thickness, but Not Root Biomass, Explains the Average and Spatial Structure of Soil Hydraulic Conductivity in Secondary Forests and Coffee Agroecosystems in Veracruz, Mexico. Sci. Total Environ. 2017, 607–608, 1357–1366. [Google Scholar] [CrossRef]
  52. Xia, L.; Song, X.; Fu, N.; Cui, S.; Li, L.; Li, H.; Li, Y. Effects of Forest Litter Cover on Hydrological Response of Hillslopes in the Loess Plateau of China. CATENA 2019, 181, 104076. [Google Scholar] [CrossRef]
  53. Wang, L.; Zhang, G.; Zhu, P.; Wang, X. Comparison of the Effects of Litter Covering and Incorporation on Infiltration and Soil Erosion under Simulated Rainfall. Hydrol. Process. 2020, 34, 2911–2922. [Google Scholar] [CrossRef]
  54. Jabro, J.D. Estimation of Saturated Hydraulic Conductivity of Soils From Particle Size Distribution and Bulk Density Data. Trans. ASAE 1992, 35, 557–560. [Google Scholar] [CrossRef]
  55. Hu, W.; Shao, M.A.; Si, B.C. Seasonal Changes in Surface Bulk Density and Saturated Hydraulic Conductivity of Natural Landscapes. Eur. J. Soil Sci. 2012, 63, 820–830. [Google Scholar] [CrossRef]
  56. Liu, Y.; Guo, L.; Huang, Z.; López-Vicente, M.; Wu, G.-L. Root Morphological Characteristics and Soil Water Infiltration Capacity in Semi-Arid Artificial Grassland Soils. Agric. Water Manag. 2020, 235, 106153. [Google Scholar] [CrossRef]
  57. Liu, Y.-F.; Zhang, Z.; Liu, Y.; Cui, Z.; Leite, P.A.M.; Shi, J.; Wang, Y.; Wu, G.-L. Shrub Encroachment Enhances the Infiltration Capacity of Alpine Meadows by Changing the Community Composition and Soil Conditions. CATENA 2022, 213, 106222. [Google Scholar] [CrossRef]
  58. Ogden, C.B.; Van Es, H.M.; Schindelbeck, R.R. Miniature Rain Simulator for Field Measurement of Soil Infiltration. Soil Sci. Soc. Am. J. 1997, 61, 1041–1043. [Google Scholar] [CrossRef]
  59. Szabó, J.A.; Centeri, C.; Keller, B.; Hatvani, I.G.; Szalai, Z.; Dobos, E.; Jakab, G. The Use of Various Rainfall Simulators in the Determination of the Driving Forces of Changes in Sediment Concentration and Clay Enrichment. Water 2020, 12, 2856. [Google Scholar] [CrossRef]
  60. Simelane, M.P.Z.; Soundy, P.; Maboko, M.M. Effects of Rainfall Intensity and Slope on Infiltration Rate, Soil Losses, Runoff and Nitrogen Leaching from Different Nitrogen Sources with a Rainfall Simulator. Sustainability 2024, 16, 4477. [Google Scholar] [CrossRef]
  61. Lai, J.; Ren, L. Assessing the Size Dependency of Measured Hydraulic Conductivity Using Double-Ring Infiltrometers and Numerical Simulation. Soil Sci. Soc. Am. J. 2007, 71, 1667–1675. [Google Scholar] [CrossRef]
  62. Lado, M.; Paz, A.; Ben-Hur, M. Organic Matter and Aggregate Size Interactions in Infiltration, Seal Formation, and Soil Loss. Soil Sci. Soc. Am. J. 2004, 68, 935–942. [Google Scholar] [CrossRef]
  63. Lado, M.; Ben-Hur, M.; Shainberg, I. Soil Wetting and Texture Effects on Aggregate Stability, Seal Formation, and Erosion. Soil Sci. Soc. Am. J. 2004, 68, 1992–1999. [Google Scholar] [CrossRef]
  64. Radinja, M.; Vidmar, I.; Atanasova, N.; Mikoš, M.; Šraj, M. Determination of Spatial and Temporal Variability of Soil Hydraulic Conductivity for Urban Runoff Modelling. Water 2019, 11, 941. [Google Scholar] [CrossRef]
  65. Zhang, J.; Lei, T.; Chen, T. Impact of Preferential and Lateral Flows of Water on Single-Ring Measured Infiltration Process and Its Analysis. Soil Sci. Soc. Am. J. 2016, 80, 859–869. [Google Scholar] [CrossRef]
  66. Grant, K.N.; Macrae, M.L.; Ali, G.A. Differences in Preferential Flow with Antecedent Moisture Conditions and Soil Texture: Implications for Subsurface P Transport. Hydrol. Process. 2019, 33, 2068–2079. [Google Scholar] [CrossRef]
  67. Zimmermann, B.; Elsenbeer, H.; De Moraes, J.M. The Influence of Land-Use Changes on Soil Hydraulic Properties: Implications for Runoff Generation. For. Ecol. Manag. 2006, 222, 29–38. [Google Scholar] [CrossRef]
  68. Singh, B.P.; Cowie, A.L.; Chan, K.Y. (Eds.) Soil Health and Climate Change; Soil Biology; Springer: Berlin/Heidelberg, Germany, 2011; Volume 29, ISBN 978-3-642-20255-1. [Google Scholar]
  69. Jarvis, N.; Koestel, J.; Messing, I.; Moeys, J.; Lindahl, A. Influence of Soil, Land Use and Climatic Factors on the Hydraulic Conductivity of Soil. Hydrol. Earth Syst. Sci. 2013, 17, 5185–5195. [Google Scholar] [CrossRef]
  70. Sun, D.; Yang, H.; Guan, D.; Yang, M.; Wu, J.; Yuan, F.; Jin, C.; Wang, A.; Zhang, Y. The Effects of Land Use Change on Soil Infiltration Capacity in China: A Meta-Analysis. Sci. Total Environ. 2018, 626, 1394–1401. [Google Scholar] [CrossRef] [PubMed]
  71. Bagarello, V.; Di Prima, S.; Iovino, M.; Provenzano, G. Estimating Field-Saturated Soil Hydraulic Conductivity by a Simplified Beerkan Infiltration Experiment. Hydrol. Process. 2014, 28, 1095–1103. [Google Scholar] [CrossRef]
  72. Stewart, R.D.; Abou Najm, M.R. A Comprehensive Model for Single Ring Infiltration II: Estimating Field-Saturated Hydraulic Conductivity. Soil Sci. Soc. Am. J. 2018, 82, 558–567. [Google Scholar] [CrossRef]
Water 17 00518 g001
Figure 1. A schematic diagram showing the layout of the five measurement locations along each transect and the infiltration tests performed with the three methods. Green-outlined squares = rainfall simulation plots; black X = measurement via automated SSBI; blue circle = measurement via Saturo infiltrometer.
Figure 1. A schematic diagram showing the layout of the five measurement locations along each transect and the infiltration tests performed with the three methods. Green-outlined squares = rainfall simulation plots; black X = measurement via automated SSBI; blue circle = measurement via Saturo infiltrometer.
Water 17 00518 g001
Water 17 00518 g002
Figure 2. Soil infiltration capacity was measured via three different types of instruments under two vegetation cover conditions: (a) rainfall simulator in a thicketized woodland; (b) automated SSBI system in a grassland; and (c) Saturo infiltrometer in a grassland.
Figure 2. Soil infiltration capacity was measured via three different types of instruments under two vegetation cover conditions: (a) rainfall simulator in a thicketized woodland; (b) automated SSBI system in a grassland; and (c) Saturo infiltrometer in a grassland.
Water 17 00518 g002
Water 17 00518 g003
Figure 3. Infiltration rate curves from tests conducted with the rainfall simulator. The red dashed line represents the average steady-state infiltration rate (infiltration capacity) for each site. For the woodland zone in the sandy soil site, from which there was no runoff in four of the five plots, the infiltration rates shown (measured at the maximum simulated rainfall intensity of 154 mm h−1) are likely underestimations.
Figure 3. Infiltration rate curves from tests conducted with the rainfall simulator. The red dashed line represents the average steady-state infiltration rate (infiltration capacity) for each site. For the woodland zone in the sandy soil site, from which there was no runoff in four of the five plots, the infiltration rates shown (measured at the maximum simulated rainfall intensity of 154 mm h−1) are likely underestimations.
Water 17 00518 g003
Water 17 00518 g004
Figure 4. Box plots of infiltration capacity values for grassland and woodland zones obtained at the three sites via the three measurement methodologies.
Figure 4. Box plots of infiltration capacity values for grassland and woodland zones obtained at the three sites via the three measurement methodologies.
Water 17 00518 g004
Water 17 00518 g005
Figure 5. Correlations between infiltration capacity values (mm h−1) obtained with each method (rainfall simulation, automated SSBI, and Saturo) and influencing variables: live biomass and dead biomass (g m−2), and soil bulk density (g cm−3) in the 0 to 5 cm and 5 to 10 cm depth intervals. Note that the regression line, adjusted R2, and p-values are based on a log–log relationship, but the scales on both axes are presented in their original form for better visualization.
Figure 5. Correlations between infiltration capacity values (mm h−1) obtained with each method (rainfall simulation, automated SSBI, and Saturo) and influencing variables: live biomass and dead biomass (g m−2), and soil bulk density (g cm−3) in the 0 to 5 cm and 5 to 10 cm depth intervals. Note that the regression line, adjusted R2, and p-values are based on a log–log relationship, but the scales on both axes are presented in their original form for better visualization.
Water 17 00518 g005
Table 1. Infiltration capacity values (mean ± standard deviation, mm h−1) obtained via the three measurement methodologies for each soil textural class and cover type. Capital A or B represents a significant difference (revealed by t-tests) in the mean values obtained by each methodology for the two cover types (grassland vs. woodland) within each site, while lowercase a or b represents a significant difference (revealed by Tukey’s HSD test) between the mean values obtained by different methodologies within the same site and cover type.
Table 1. Infiltration capacity values (mean ± standard deviation, mm h−1) obtained via the three measurement methodologies for each soil textural class and cover type. Capital A or B represents a significant difference (revealed by t-tests) in the mean values obtained by each methodology for the two cover types (grassland vs. woodland) within each site, while lowercase a or b represents a significant difference (revealed by Tukey’s HSD test) between the mean values obtained by different methodologies within the same site and cover type.
Site/
Soil Texture		Cover
Type		Methodology
Rainfall SimulationAutomated SSBISaturo
ClayGrassland7.5 ± 3.4 A b 4.3 ± 3 A b 47.4 ± 18.6 A a
Woodland66.8 ± 18.9 B a 71.7 ± 92.4 B a 153.3 ± 138.2 A a
Loamy sandGrassland48.9 ± 26.6 A a 101.6 ± 94.3 A a 113 ± 63.6 A a
Woodland106.3 ± 27.1 B a 79.2 ± 57.4 A a 56.3 ± 85.8 A a
SandyGrassland68.8 ± 14.1 A a 13.4 ± 2.1 A b108.4 ± 81.6 A a
Woodland144.5 ± 22.4 B120.9 ± 28.6 B183.4 ± 84.3 A
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Atalar, F.; Leite, P.A.M.; Wilcox, B.P. A Comparison of Three Methodologies for Determining Soil Infiltration Capacity in Thicketized Oak Woodlands and Adjacent Grasslands. Water 2025, 17, 518. https://doi.org/10.3390/w17040518

AMA Style

Atalar F, Leite PAM, Wilcox BP. A Comparison of Three Methodologies for Determining Soil Infiltration Capacity in Thicketized Oak Woodlands and Adjacent Grasslands. Water. 2025; 17(4):518. https://doi.org/10.3390/w17040518

Chicago/Turabian Style

Atalar, Furkan, Pedro A. M. Leite, and Bradford P. Wilcox. 2025. "A Comparison of Three Methodologies for Determining Soil Infiltration Capacity in Thicketized Oak Woodlands and Adjacent Grasslands" Water 17, no. 4: 518. https://doi.org/10.3390/w17040518

APA Style

Atalar, F., Leite, P. A. M., & Wilcox, B. P. (2025). A Comparison of Three Methodologies for Determining Soil Infiltration Capacity in Thicketized Oak Woodlands and Adjacent Grasslands. Water, 17(4), 518. https://doi.org/10.3390/w17040518

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop