This field study was conducted at two Arkansas Discovery Farms (ADF) [20] research sites located near the cities of Dumas (33.8188, −91.3386214) in 2024 and 2025 and Stuttgart (34.3620372, −91.5213238) from 2023-2025, along with the UADA Jackson County Extension Center in Newport, Arkansas (35.5712, −91.2601) in 2023 and 2024. Soil taxonomic classifications for the study areas at each location are reported on Table 1.

Table 1. Summary of soil series, slope class, and taxonomic description of the soils mapped at the three field sites in eastern Arkansas.

At each location, three biochar rate treatments, 0 (0B), 2000 (2B), and 4000 (4B) kg·ha⁻¹, were established in plots with dimensions of four rows that were 10-m long, with a row spacing of 96 cm at Dumas and 76 cm at Stuttgart and Newport. Plots were arranged in a randomized complete block design (RCBD), where treatments were replicated with four blocks for a total of 12 plots at each location.

Soil samples were collected from the 0 to 10-cm depth interval in each unamended plot at Newport on May 3 and Stuttgart on May 17 and 0 - 15 cm at Dumas on March 22 prior to the first biochar applications in 2023, sent to the Arkansas Soil Testing Laboratory in Marianna, Arkansas, and analyzed for pH and estimated cation exchange capacity (CEC) using a 1:2 soil-water volume mixture ratio. The soil samples were then transported to the Altheimer Soil and Plant Diagnostic Laboratory in Fayetteville, Arkansas for determination of organic matter (OM) concentration by the weight-loss-on-ignition method [21] in a muffle furnace for two hours at a starting temperature of 105˚C and an ending time of 360˚C.

A separate set of soil samples were collected from the unamended control plots in 2024 at all three locations using a 4.8-cm-diameter soil core. Samples were dried for approximately 48 hours (h) at 70˚C [22] and weighed for bulk density determinations. Table 2 summarizes initial soil properties from the unamended control plots at the three research sites.

Table 2. Summary of initial 2023 unamended plot soil textures and mean (± standard error) bulk density (BD) for the unamended control plots in 2024, soil pH, Mehlich-3 phosphorus (M3-P), Mehlich-3 potassium (M3-K), total nitrogen (N), estimated cation exchange capacity (CEC) and organic matter (OM) prior to the first biochar applications in 2023 for the three sites in eastern Arkansas.

Each ADF site focused on row-crop agriculture with emphasis on cotton (Gossypium hirsutum L.) and corn (Zea mays L.) production in Dumas, soybean (Glycine max L.) and corn production in Stuttgart, and soybean production in Newport. Yearly crop production information for each research location is reported in Table 3.

Table 3. Summary of annual crop production information for the three sites in eastern Arkansas.

^aNA, not available.

Due to the relatively small plot dimensions, field applications of biochar were broadcast by hand at a walking speed of approximately 4.8 km·hr⁻¹. Biochar was applied each project year, shortly before each location’s planting. The mechanical soil manipulations from planting lightly incorporated the biochar to a depth of ~ 5 to 10 cm. Pre-plant and in-season fertilizer applications were made for each location according to soil-test results and UADA fertilizer recommendations for cotton [23], corn [24], and soybean [25]. Estimated crop yields were determined from hand-harvested samples collected at Dumas on September 4, 2024, and August 12, 2025; at Newport on October 3, 2023, and September 30, 2024; and at Stuttgart on October 12, 2023, August 12, 2024, and October 14, 2025. Cumulative biochar applications for each site and project year were as follows: biochar treatments 2B and 4B at Dumas in 2024 received 7.76 and 15.52 kg·plot⁻¹ respectively and in 2025, the cumulative applications were 15.52 and 31.04 kg·plot⁻¹ respectively; treatments 2B and 4B at Newport in 2023 received 6.08 and 12.16 kg·plot⁻¹ respectively and in 2024, cumulative rates were 12.16 and 24.32 kg·plot⁻¹ respectively; treatments 2B and 4B for Stuttgart in 2023 received 6.08 and 12.16 kg·plot⁻¹ respectively and cumulative rates rose to 12.16 and 24.32 kg·plot⁻¹ in 2024 and 18.24 and 36.48 kg·plot⁻¹ for treatments 2B and 4B respectively.

The Dumas and Stuttgart sites were managed using furrow irrigation according to producer-standard practices. Irrigation water was supplied using two irrigation sets with an approximate capacity of 4921 liters per minute (LPM). Irrigation was typically applied at 7- to 8-day intervals depending on rainfall and crop water demand during the various stages of crop development throughout a growing season. Cumulative rainfall totals at each site during each project year’s data range were as follows: Dumas in 2024 and 2025 had cumulative rainfall totals of 39.9 cm and 39.2 cm respectively (University of Arkansas at Monticello, AR US USC00037315), Newport in 2023 and 2024 had cumulative rainfall amounts of 25.3 cm and 41.5 cm respectively (Newport Army Air Field, AR US USC00035189), and Stuttgart in 2023, 2024, and 2025 had cumulative rainfall totals of 18.8 cm, 36.5 cm, and 22.8 cm respectively (Stuttgart Municipal Airport, AR US USW00013925).

The biochar used in this study was produced by enviraPAC (Monticello, Arkansas) from short-leaf pine (Pinus echinata), a species the company recommended for agricultural usage, and was produced at temperatures > 600˚C. The resulting particle-size distribution of the biochar was mixed, ranging from fine to 1.6 cm in characteristic dimension. The biochar was not inoculated prior to field application. Biochar sub-samples were sent to the Altheimer Soil and Plant Diagnostic Laboratory in Fayetteville, Arkansas and evaluated for various initial chemical characteristics (Table 4).

Evaluation of biochar rates on soil water content was made by installing three soil water pressure sensors (200SS-15, IRROMETER Company Inc., Riverside, CA) in each plot by excavating holes with a slide hammer and soil probe to depths of 15, 30, and 45 cm to continuously monitor soil water pressure throughout the calendar year. Soil water pressure sensors were connected to data loggers (900 M, IRROMETER Company Inc.) placed within the 3-m wide alleys between treatment blocks for continuous data recording and exporting. Soil water pressure sensors recorded data every 12 hours (hr), but averaged at 12-hr intervals for statistical analyses.

Table 4. Chemical analyses of enviraPAC (Monticello, AR) biochar for total nitrogen (TN), total phosphorus (TP), total potassium (TK), total calcium (TCa), pH, and electrical conductivity (EC) along with N, P₂O₅, K₂O, and Ca.

Measured soil water pressures were converted to soil water contents using the Soil-Plant-Air-Water (SPAW) computer tool [26] using initial sand and clay percentages and average OM concentration data from the unamended control plots at each location from the 0 - 10 cm at Newport and Stuttgart and 0 - 15 cm at Dumas to reflect the appropriate soil sample depth according to UADA recommendations for soybeans [25] and cotton [23], respectively. Soil texture data used were as follows: 15% sand and 6% clay at Dumas, 46% sand and 12% clay at Newport, and 17% sand and 15% clay at Stuttgart based on on-house measurements using the 12-hr hydrometer method [27]. Soil OM concentrations were averaged from soil samples collected from unamended control plots in 2023 (i.e., 2.3% for Dumas, 1.4% for Newport, and 2.2% for Stuttgart). Conversion of measured soil water pressures to soil water contents allowed for a clearer inference in determining differences between the three biochar treatments and soil depths by project year along with any observed statistical interactions.

Dataloggers and sensors were removed from each field plot prior to each location’s primary harvesting event and reinstalled shortly thereafter. The dataloggers and sensors were then removed prior to the planting of next year’s crop at each location and subsequently reinstalled. Data was not collected at Dumas in 2023 and at Newport in 2025 due to repeated equipment malfunctions that impeded effective data collection.

The effects of biochar rate, soil depth, and their interaction on soil water contents were evaluated by a two-factor ANOVA, separately by location and separately by year, using a generalized linear mixed model with a lognormal distribution in JMP Student Edition 18 and SAS Version 9.4 (SAS Institute, Inc.). Biochar rate and soil depth were fixed effects, while datalogger measurement date in each project year was a random effect, used as a term of repeated measures and not officially tested as a fixed effect in all analyses. Soil water content values for each biochar rate and soil depth treatment combination were calculated by averaging soil water contents from individual biochar treatment plots across all four blocks at each location. The 12-hour sampling intervals for each site were included in each sampling date random effect and subsequently incorporated into the covariance structure. Means for all analyses were determined by a protected least significant difference (LSD) procedure at P < 0.05.

Soil water contents over time for Dumas in 2024 were generally consistent among the three biochar treatments at the 15- and 30-cm depths for each instance of low and large soil water contents throughout that soybean growing season (Figure 1). The only inconsistency between biochar rates at 45 cm occurred between days of year 160 - 180 when soil water contents in the 2B treatment were visually greater when compared to the other two treatments (Figure 1). In contrast to 2024, soil water contents over time for 2025 showed inconsistencies between biochar treatments throughout the corn growing season. Soil water contents for all biochar rate/soil depth combinations in 2025 varied for most of the measurement period, with days of year 100 - 120 providing the largest variation in soil water contents among treatment combinations (Figure 1).

Figure 1.Soil water contents over time throughout the 2024 (top) and 2025 (bottom) growing seasons at the Dumas, AR site by biochar (B) treatment (0, 2, and 4 Mg·ha⁻¹) and soil sensor depth combination. Cotton was grown in 2024 and corn in 2025.

At Dumas, in southeast Arkansas, soil water contents differed among biochar treatments by soil depth (P < 0.001) in 2024 and 2025 (Table 6). In 2024, soil water content in the 2B/45- was greater than all other biochar-depth treatment combinations. Soil water contents in the 2B/30-, 4B/45-, and 0B/45- were greater than in the 0B/30- and 4B/30-cm biochar-depth treatment combinations, which, in turn, had greater soil water contents than in the 4B/15-, 2B/15-, and 0B/15-cm biochar-depth treatment combinations.

In 2025, in contrast to 2024, soil water contents in the 0B/30- were greater than in all other biochar-depth treatment combinations. Slightly similar to 2024, soil water contents were also greater in the 0B/15-, 0B/45-, and 2B/45- than in the 4B/30- and 4B/45-, which were, in turn, greater than in the 2B/15-, 4B/15-, and 2B/30-cm biochar-depth treatment combinations. Data for Dumas in 2023 was not reported due to repeated equipment malfunctions which deterred data collection.

Table 6. Soil water contents at Dumas, Newport, and Stuttgart in percent by volume (%Vol.) for biochar treatments equivalent to 0 (0B), 2000 (2B), and 4000 (4B) kg·ha⁻¹ in 2023, 2024, and 2025 at the 15-, 30-, and 45-cm soil depths.

^aColumns not sharing the same letter are significantly different (P < 0.05).

Figure 2.Soil water contents over time throughout the 2023 (top) and 2024 (bottom) growing seasons at the Newport, AR site by biochar (B) treatment (0, 2, and 4 Mg·ha⁻¹) and soil sensor depth combination. Soybeans were grown in both years.

Soil water contents at Newport in 2023 were generally similar between biochar treatments at all depths, except for two periods: between days of year 170 - 180 and 185 - 200 (Figure 2). Soil water contents in the 2B treatment were consistently larger at all depths during these intervals. In contrast to 2023, soil water contents in 2024 showed inconsistencies between biochar treatments at all depths (Figure 2). The inconsistencies were more pronounced in the intervals between peaks representing periods of large soil water contents, with soil water contents in the 2B and 4B treatments greater than in the 0B treatment at the 15- and 30-cm depths throughout the 2024 soybean growing season (Figure 2). At the 45-cm soil depth, soil water content inconsistencies among biochar treatments occurred again during peak intervals, however the inconsistencies fluctuated between treatments (Figure 2).

At Newport, in northeast Arkansas, soil water contents differed among biochar treatments (P < 0.001) and differed among soil depths (P < 0.001) in 2023 (Table 7). Soil water contents, averaged across soil depths, was greater in the 2B than in the 4B treatment and soil water contents in both biochar treatments were greater than in the 0B treatment (Table 7). Averaged across biochar treatments, soil water contents at the 45-cm depth were greater than at the 30-cm depth, which was, in turn, greater than at the 15-cm depth (Table 7).

Table 7.Soil water contents at Newport in percent by volume (%Vol.) for biochar treatments equivalent to 0 (0B), 2000 (2B), and 4000 (4B) kg·ha⁻¹ and for the 15-, 30-, and 45 -cm soil depths in 2023.

^aColumns not sharing the same letter are significantly different (P < 0.05).

In contrast to 2023, in 2024, soil water contents differed among biochar treatments by soil depth (P < 0.001) (Table 5). Soil water contents in the 2B/45- was greater than in all other biochar-depth treatment combinations (Table 6). Soil water contents in the 0B/45- was greater than in the 2B/30- and 4B/45-cm combinations, with the 4B/30-cm combination intermediate (Table 6). Soil water contents decreased in the following order 0B/30- = 4B/15- > 0B/15- > 2B/15-cm for the remaining biochar depth combinations (Table 6). Data for Newport in 2025 was not reported due to consistent equipment breakdowns which impeded data collection.

Soil water contents for Stuttgart in 2023 were generally consistent among biochar treatments at all soil depths throughout the soybean growing season (Figure 3). However, in contrast to 2023, there were visible differences in soil water contents in the intervals between peak water contents in 2024, with soil water contents in the 0B treatment consistently greater at 45 cm than for all other treatment combinations from days of year 130 - 160 (Figure 3). From day of year 160 to 220, little soil water content consistency occurred, where soil water contents in the 0B and 4B treatments were consistently greater than the 2B treatment at all soil depths (Figure 3).

Figure 3. Soil water contents over time throughout the 2023 (top) and 2024 (bottom) growing seasons at the Stuttgart, AR site by biochar (B) treatment (0, 2, and 4 Mg·ha⁻¹) and soil sensor depth combination. Differences in x-axes values on both graphs are the result of different planting/harvesting dates and growing season durations for soybean (2023) and corn (2024).

Similar to Dumas, but in contrast to Newport, soil water contents at Stuttgart, in east-central Arkansas, differed among biochar treatments by soil depth (P < 0.001) in 2023, 2024, and 2025 (Table 6). In 2023, soil water contents in the 4B/30- and 0B/30- were greater than in all other biochar-depth treatment combinations (Table 6). Soil water contents in the 4B/45- were also greater than in the 2B/30-, with soil water contents in the 4B/15- and 0B/45-cm combinations intermediate (Table 6). Soil water contents in the 2B/45- and 2B/15- were greater than in the 0B/45-cm combination (Table 6).

In contrast to 2023, soil water contents in the 0B/45- were greater than in all other biochar-depth combinations (Table 6) in 2024. Soil water contents in the 0B/30- were also greater than in the 4B/30- and 4B/45-, which, in turn, had greater soil water contents than in the 2B/15- and 0B/15-cm combinations (Table 6). Soil water contents in the 2B/45- and 4B/15- were both greater than in the 2B/30-cm combination (Table 6).

In 2025, similar to 2024, soil water contents in the 0B/45- were again greater than in all other biochar-depth treatment combinations (Table 6). In contrast to 2023/2024, soil water contents in the 0B/30-, 4B/30-, and 2B/15- were greater than in the 4B/45-cm combination, with soil water contents in the 2B/45-cm combination intermediate (Table 6). Soil water contents in the 4B/45- were greater than in the 0B/15-cm combination, with soil water contents in the 4B/15- intermediate followed by the 2B/30-cm combination (Table 6).