1 Introduction

The papermill industry in the US and perhaps in other parts of North America uses biomass to generate energy and, in the process, generates large quantities of biochar, often incorrectly referred to as “boiler ash”. In the past, the biomass was fully combusted and true ash was generated which is fine in texture consisting mainly of salts, oxides and hydroxides of Ca, Mg, Na, and K and trace amounts of metals (Pugliese et al. 2014). In recent years, however, in order to reduce CO2 emissions, many plants have changed their practices to combusting the biomass under limited O2 which leads to the generation of biochar by pyrolysis, which contains very high carbon instead of the typical boiler ash (https://www.babcock.com/home/about/resources/success-stories/converting-a-recovery-boiler-to-a-bubbling-fluid-bed). Much of this biochar is disposed in landfills (Pitman 2006; Staples and Van Rees 2001; Vance 1996). Sometimes, the material is spread over farmlands not necessarily as a valuable soil amendment but as a way of disposing it considering the increasing costs and the difficulties in acquiring new landfill sites (Arshad et al. 2012; Miner and Unwin 1992; Ochecova et al. 2014).

Biochar, however, is considered a valuable soil amendment that has gained substantial research attention in recent years. Much of the research to date has shown biochar as a potential technology to increase soil organic carbon sequestration and to concurrently improve soil physical and hydraulic parameters, such as improved soil structure, water retention and elevated soil physical quality (Antwi et al. 2020; He et al. 2020; Knoblauch et al. 2021). Many studies have shown that biochar amendments improve soil structure and aeration as well as increase total porosity (Islam et al. 2021; Liang et al. 2021). Recent incubation studies in the laboratory reported increased aggregate stability following biochar addition (Zheng et al. 2021; Sun et al. 2021). Also, biochar has high porosity and specific surface area, and thus its addition to soil can improve the total pore structure and gas transport at the soil-atmosphere interface (Yu and Lu 2020). The improvement in soil porosity and aggregation upon biochar addition leads to increases in soil water retention and plant available water (Huang et al. 2021; Liang et al. 2021; Yang and Lu 2021). Thus, biochar amendment is deemed as an effective strategy to improve the structure of degraded soils and increase rainfall and irrigation water use efficiency in the field.

But the effect of biochar has not always been positive as no effects or even negative effects have also been reported in some cases (Aller et al. 2021; Hardie et al. 2014; Mukherjee and Lal 2013). Biochar for the most part has high pH, with the potential to increase soil pH (Pandian 2021), and this can potentially increase clay dispersibility due to a dominance of repulsive forces between clay minerals (Roth and Pavan 1991; Saffari et al. 2020). This effect implies biochar can decrease aggregation and disrupt soil structure. Busscher et al. (2010), for example, reported a significant decrease in soil aggregation when biochar produced from pecan shells was applied to a loamy sand soil.

Other studies show that biochar does not affect soil physical or hydraulic properties. Fungo et al. (2017), for example, reported that the application of 2.5 t ha−1 wood biochar pyrolyzed at 550 °C to a Typic Kandiudult failed to affect aggregate stability of the soil. Similarly, rice-straw biochar, when applied to an Ultisol, had no effect on soil structural stability (Peng et al. 2011). In addition, some studies found biochar addition had no effect on water holding capacity (Jeffery et al. 2015) and even could decrease the saturated hydraulic conductivity of the soil. Hardie et al. (2014) applied 47 Mg·ha−2 gum Arabic greenwaste biochar to soil and found no effect on the soil moisture content after 30 months because of the high hydrophobicity of biochar that prevented its effect on soil water retention (Jeffery et al. 2015). Indeed, an increase in the permanent wilting point after biochar addition (Herath et al. 2013) may lead to an overall reduction of plant available water despite increases in field capacity (Bayabil et al. 2015; Liu 2020).

Many of the results thus far reported on biochar effect on soil physical and hydraulic properties are derived from pot experiments with sieved and repacked soil under climate-controlled incubation conditions rather than under in situ conditions. Thus, the applicability of existing results to real-world soil conditions is questionable (Alghamdi et al. 2020; Chen et al. 2020; Głąb et al. 2016). Furthermore, whether this knowledge that applies to biochar derived from papermill boilers applies to real-world farms on a long-term basis is not known. The objective of this study was to determine whether biochar derived from a papermill boiler applied to cotton fields yearly for as long as 10 years could improve soil physical and soil water retention properties.

2 Materials and methods

2.1 Field sites and biochar characteristics

The study was conducted on a commercial farm that grows cotton on a continuous basis in northern Mississippi, USA (Fig. 1). The study area is characterized by humid and wet winter months and hot summers. The region receives an average annual precipitation of 1385 mm with annual mean air temperature of 16.8 °C based on NOAA meteorological records from 1938 to 2020 (https://www.ncdc.noaa.gov/cdoweb/datasets/GHCND/stations/GHCND:USC00227111/detail). Much of the rainfall in the region is concentrated in the fallow season during October through April (Feng et al. 2016).

Fig. 1
figure 1

Map showing the study location and soil sampling points. Field 0, Field 2, Field 3, Field 5 and Field 10 received 6.7 Mg ha−1 year−1 biochar for 0, 2, 3, 5, and 10 years, respectively

Full size image

The study consisted of five fields ranging in size from 3.5 to 15 ha located within a 5-km radius except for one field that was 15 km away (Fig. 1). The soil type in all five fields was silt loam. The fields received 6.7 Mg ha−1 year−1 biochar for 0, 2, 3, 5, and 10 years as a surface broadcast application every spring before planting cotton. The last application was made in spring 2018. The 10-year field started receiving biochar application in 2009; the 5-year field started receiving biochar 5 years later. The last field (2 year) received biochar in 2017 and 2018. The field that received biochar for 0 years served as the control.

Once broadcast-applied, the biochar was left on the soil surface and not incorporated to the soil since the farm practices a no-till production system. Cotton was the only commercial crop grown in all the fields every year throughout the study period and in the previous decade. The cotton was managed following the regional commercial cotton production practices each year including fertilizer applications, pest control, and soil conservation practices.

The biochar applied in all years was obtained from the same papermill plant located within 5 km from the majority of the fields. The biochar was produced by the slow pyrolysis method (≈ 500 °C) from mixed pine biomass as a byproduct of energy generation for the plant. A sample taken in 2019 from the papermill plant had the following properties: 79.3% total C and 1.3% total N on a dry weight basis, a pH of 12.2, and 47.8% solids content.

2.2 Soil sampling

Undisturbed and disturbed soil core samples were collected from all five fields in March 2019. The samples were taken from 10 to 16 random sampling stations throughout each field that represented the bulk of the field. At each sampling station, undisturbed core samples were taken from 0–5 and 5–10 cm layers using 10 cm long PVC rings with 7.7 cm inner diameter. One end of each tube was sharpened to help drive the tube into the soil. Once extracted, the ends of the 10-cm core were wrapped with a plastic film, transported to the lab, and stored in the lab until analyzed for soil hydraulic properties and bulk density. The 10 cm PVC rings filled with soil were cut into 0–5 and 5–10 cm cores using a sharp electric cutting saw without removing the soil core from the PVC tubes to make sure the soil structure didn’t get disturbed when cutting the soil samples.

The disturbed soil core samples were taken from each sampling station adjacent to the undisturbed sampling points using a standard soil probe. Three to four core samples were taken from 0–5 and 5–10 cm depths and composited by the depths within a 0.5 m radius of each sampling location. These samples were air-dried on a greenhouse bench then stored in sealed plastic bags until they were used to measure soil physical properties and total carbon (TC).

2.3 Measurement of soil properties

The disturbed soil samples were crushed and passed through 2-mm sieve, then were used to measure TC content by an automated dry combustion method using an Elementar Vario MAX CN analyzer (Elementar Instrument, Mt. Laurel, NJ). Soil particle size distribution was also determined on these samples using Mastersizer 3000 (Malvern Instruments Ltd, UK). To determine particle size distribution, carbonates and organic matter were removed from the soil samples using a 3% (w/w) hydrogen peroxide solution. The ultrasonic wave feature of the Mastersizer 3000 analyzer was then used to further disperse soil granular structure to single soil particles.

The results of TC and particle size distribution were employed to calculate structural stability index (SI) (Reynolds et al. 2009):

$${\text{SI}} = \frac{{1.724{\text{TOC}}}}{{\left( {{\text{Silt}} + {\text{Clay}}} \right)}} \times 100; 0 \le {\text{SI}} \le \infty$$
(1)

where TOC (wt. %) is the total organic carbon content, Silt and Clay (wt. %) is the silt and clay content of the sample. Measured TC was used in place of TOC because the soil in these fields contains very low or no inorganic C (https://www.nrcs.usda.gov/resources/data-and-reports/web-soil-survey).

Soil water-stable aggregates of the disturbed samples with > 0.25 mm diameter (SAG) were measured using a wet sieving apparatus (Eijkelkamp Agrisearch Equipment, Giebeek, the Netherlands) following the methods described by Kemper (1986) and Li et al. (2021).

The undisturbed soil core samples were used for determining soil water retention curves (SWRCs), which describe the relationship between pressure head (water potential) and water content, using a high-speed refrigerated centrifuge (Model CR21G, Hitachi Co., Ltd., Japan) (Reatto et al. 2008; Fu et al. 2011; Gao and Shao 2015). The centrifuge method has a few advantages compared to the traditional method which uses chember cells on pressure plates. First, it is fast and saves much time taking only a few days to achieve the same results that would take 6–8 weeks by the traditional pressure plate method (Rahardjo et al. 2019; Zheng et al. 2022). Second, the range of soil water suction that can be applied on a sample is greater with a high-speed centrifuge than with the conventional method (Lozano et al. 2020; Reatto et al. 2008). Third, the accuracy of high-speed centrifuge method has been validated against results obtained with the pressure plate method (Reatto et al. 2008). Most of previous studies showed that soil water content obtained with the centrifuge method is highly linearly correlated with that of the pressure plate method and their results from both methods are similar (Reatto et al. 2008; Centurion et al. 1997). Therefore, establishing SWRCs using the centrifuge method could be considered as a fast, reliable, and accurate method of testing soil hydraulic properties. In fact, it has been widely adopted in many studies (Khanzode et al. 2002; Gao and Shao 2015; Jotisankasa and Sirirattanachat 2017; Khaksar et al. 2022).

In our study, each undisturbed soil core sample was transferred from the PVC tubes to a stainless steel holder with an open end (2.5 cm inner diameter and 5 cm long) and saturated with deionized water. Then the saturated samples were loaded onto the centrifuge with a 205-µm filter paper at the bottom of the sample holder that facilitated drainage but retained soil particles during saturation and centrifugation. The dehydration process started at a given rotation speed corresponding to a particular pressure head. After 1 h of centrifugation when soil samples reached equilibrium at each rotation speed, the samples were taken out from the centrifuge and weighed using an electronic balance (precision of 0.01 g) to determine their soil gravimetric water content. The compressed height of soil (H, mm) was measured using a caliper (0.02 mm precision) to calculate the pressure head (h, cm). Then, the soil sample cylinder was placed into the centrifuge and centrifuged at the next higher rotation speed corresponding to its pressure head. This procedure was sequentially repeated up to the highest rotation speed of 10,000 rad min−1. Each core sample was subjected to a total of 10 rotation speeds (300, 350, 400, 560, 730, 1220, 2050, 4590, 7250, 8400 and 10,000 rad min−1) which corresponded to the pressure head of 0, 2, 3, 4, 8, 13, 33, 104, 509, 1250, 1500, and 2692 kPa, respectively. After the highest speed was run, the soil samples were oven-dried at 105 °C for two days and the dry mass of the soil was measured. The bulk density (BD) of the samples was calculated from the oven-dried (105℃) mass and volume of the undisturbed soil core. The volumetric water content at each matric potential was calculated by the respective difference in weight of the oven-dried samples multiplied by the BD.

The degree of soil compactness (SDC, %) was also calculated as [SDC = (BD/BDmax) × 100]. Maximum soil bulk density (BDmax, g cm−3) was estimated using a pedotransfer function described by Marcolin and Klein (2011), in which SOM and clay content are the input parameters.

Water holding capacity at field capacity (FC, cm3 cm−3) and permanent wilting point (PWP, cm3 cm−3) was calculated as the volumetric water content in the soil at the pressure head of 33 kPa and 1500 kPa, respectively (Aller et al. 2017; Lu et al. 2014). Plant available water (PAW, cm3 cm−3) was computed as the difference in soil volumetric water content at FC and PWP (White 2013).

Air capacity (AC, cm3 cm−3), which affects soil aeration, was calculated as the difference between saturated water content (θs) and FC (White 2013). Water-filled pore space (WFPS) was computed by dividing volumetric moisture content at 6 kPa by total porosity (TP) as described by Wienhold et al. (2009).

Soil water storage capacity (SWSC) defined as the ratio of water content at field capacity (FC) to TP (SWSC = FC/TP) was also calculated. Soil aeration capacity (SAC) was calculated as the ratio between AC and TP (SAC = AC/TP) (Reynolds et al. 2002).

The soil water release curves (SWRCs), estimated pore size distribution characteristics, and Ksat of the undisturbed samples were created with the USDA-ARS software RETC for analyzing hydraulic properties of unsaturated soils (Van Genuchten et al. 1992). The RETC tool allows to select several soil water release curve fitting functions, among which the following van Genuchten equation (VG) (Van Genuchten 1980) is the most widely used function.

$$\theta \left( {\text{h}} \right) = \theta r + \frac{\theta s - \theta r}{{\left[ {1 + \left( {\alpha {\text{h}}} \right)^{{\text{n}}} } \right]^{{1 - {\text{m}}}} }}$$
(2)

where α, n and m are parameters that define the curve shape. θs and θr are saturated water content and residual water content, respectively; θ(h) is the soil water content at a given pressure head. The reciprocal of α accounts for the air entry pressure. Higher α−1 value indicates that the soil consists of a wider capillary saturation zone; whereas n represents the slope of the curve and it increases for coarser soil textures. According to Mualem (1976), to reduce the number of parameters being estimated and simplify the integration of Eq. (2), m is set to 1–1/n. Following a fitting process, the RETC tool finds the set of parameters that minimizes the sum of squared residuals (RMSE) between model-predicted and observed water retention data (Van Genuchten et al. 1991). The RETC tool minimizes the RMSE iteratively by means of a weighted least-squares approach based on Marquardt’s maximum likelihood method (Marquardt 1963).

The SWRCs function was also employed to estimate the pore size distribution based on the capillary rise equation by approximating the relationship between h (cm) and the equivalent pore diameter (d, μm). The d parameter was estimated as follows (Warrick 2002):

$$\text{d} = \frac{2980}{{\text{h}}}\text{; h>}{0}\text{ (cm)}$$
(3)

A “normalized” pore volume distribution function, S*(h), was calculated as follows (Reynolds et al. 2009):

$${\text{S}}*\left( {\text{h}} \right) = \frac{{{\text{m}}\left( {\alpha {\text{h}}} \right)^{n} \left[ {1 + {\text{m}}^{ - 1} } \right]^{{\left( {{\text{m}} + 1} \right)}} }}{{\left[ {1 + \alpha {\text{h}}^{{\text{n}}} } \right]^{{{\text{m}} + 1}} }}$$
(4)

The volume of different pore categories was determined according to the pore classification developed by (Greenland 1981), which characterizes pores as a bonding space (< 0.005 μm), residual pores (0.005–0.5 μm), storage pores (0.5–50 μm), transmission pores (50–500 μm), and fissures (> 500 μm). Volumetric water content within different pore size classes was determined by interpolation between SWRCs. Then the total porosity (TP) was calculated as the sum of < 0.5, 0.5–50 and > 50 μm pores using the value of volumetric water content when pressure head was equal to 0 kPa (Igaz et al. 2018).

2.4 Calculation of soil physical quality index

An overall soil physical quality index (SPQI) was calculated to quantify biochar effects on the soil physical health in the study. The first step was to select appropriate SPQI indicators. Based on published literature and the authors’ experience, fourteen indicators including BD, SDC, SAG, SI, TP, WFPS, SAC, FC, PWP, PAW, SWSC, Ksat, θs and θr were selected and used as a minimum dataset to determine SPQI and assess how much biochar affected soil physical properties.

The second step involved transforming each indicator into a unitless value ranging from 0 to 1 for inclusion in the SPQI. The transformation was performed using a linear technique as described by Andrews et al. (2002). Indicators were ranked in ascending or descending order depending on whether a higher value was considered “good” or “bad” in terms of soil function. The indicators SAG, SI, TP, FC, PAW, Ksat and θs were assigned ascending values (‘more is better’) and the indicators BD, SDC, PWP and θr were assigned descending values (‘less is better’) as follows (Wang et al. 1948):

$$\mathrm{Positive\, indicators}:{x}_{ij}^{^{\prime}}=\frac{xij-\mathit{min}\left\{{x}_{1j},\hspace{0.33em}\cdots \hspace{0.33em},\hspace{0.33em}{x}_{nj}\right\}}{\mathit{max}\left\{{x}_{1j},\hspace{0.33em}\cdots \hspace{0.33em},\hspace{0.33em}{x}_{nj}\right\}-\mathit{min}\left\{{x}_{1j},\hspace{0.33em}\cdots \hspace{0.33em},\hspace{0.33em}{x}_{nj}\right\}}$$
(5)
$$\mathrm{Negative\, indicators}:{x}_{ij}^{^{\prime}}=\frac{\mathit{max}\left\{{x}_{1j},\hspace{0.33em}\cdots \hspace{0.33em},\hspace{0.33em}{x}_{nj}\right\}-xij}{\mathit{max}\left\{{x}_{1j},\hspace{0.33em}\cdots \hspace{0.33em},\hspace{0.33em}{x}_{nj}\right\}-\mathit{min}\left\{{x}_{1j},\hspace{0.33em}\cdots \hspace{0.33em},\hspace{0.33em}{x}_{nj}\right\}}$$
(6)

where, \({{x}_{i}^{^{\prime}}}_{j}\) is the value of the jth indicator of the ith treatment (i = 1, 2 …, n; j = 1, 2 …, m).

For indicators such as WFPS, SAC and SWSC that do not fall into one of these two categories and have an optimum ideal value instead, the following algorithm was used to standardize and then rank the data (Wang et al. 1948):

$$\mathrm{Optimum \,indicators}:{{x}_{i}^{^{\prime}}}_{j}=M/\left[M+\left|{{x}_{i}}_{j}\right|\right]$$
(7)

where M is the optimal value, the range of standardized data was changed from 0 to 1.

In the third step, the weight of each soil indicator was calculated using principal component analysis (PCA), which is the variance of the eigenvectors derived from the first four principal components based on the inflection point from Scree plot and Kaiser’s cut-off (eigenvalues > 1). Each transformed indicator value was multiplied by its weight to get the indicator score of each treatment. These indicator scores for each sample were summed to obtain the soil physical health score. The overall soil physical quality index (SPQI) was calculated as follows:

$${\text{SPQI}} = \mathop \sum \limits_{i = 1}^{n} {\text{S}}_{i} \times {\text{W}}_{i}$$
(8)

where, Si is the ith soil indicator, Wi is the weight of each soil indicator, and n is the number of soil indicators.

2.5 Statistical analysis

All data were subjected to analysis of variance (ANOVA) to test the effects of the number of biochar applications on the measured parameters in the same soil layer. Differences in mean values among treatments were compared using least significant difference test (LSD) at p ≤ 0.05. The Pearson correlation coefficient was calculated to estimate the linear relationships among the soil properties and the correlation coefficient matrix was drawn using R software package (v.4.0.2) (R Development Core Team, 2014). Other figures were created using SigmaPlot 13.0 (Systat Software Inc., California, USA).

3 Results

3.1 Soil texture, structure, and total organic carbon

Biochar addition applied yearly for 2–10 years had no effect on soil texture at 0–5 and 5–10 cm depths relative to the control. The soil texture of all fields was classified as silt loam based on our measurements (Table 1) which is consistent with the classification in Fig. 1 as published on Web Soil Survey (https://websoilsurvey.sc.egov.usda.gov/). Our measurement showed that amending the soil with biochar for as long as 10 years did not affect the texture which is expected.

Table 1 Effect of biochar on soil texture, total carbon, and physical properties at the 0–5 and 5–10 cm depths
Full size table

Applying biochar, regardless of how many years, resulted in the accumulation of total organic carbon (TOC) at both depths (Table 1). At the 0–5 cm layer, the TOC averaged across the biochar treatments was 1.36%, an increase of 45% relative to the control. The increase in TOC was proportional to the number of applications. Total organic C increased with increasing number of biochar applications in the 5–10 cm depth also, although the values were generally lower than the 0–5 cm layer (1.08%). The soil in the 0–5 cm layer had 26% greater TOC than the soil in the 5–10 cm layer, suggesting that the biochar accumulated on the soil surface.

Biochar addition reduced bulk density (BD) and the degree of soil compactness (SDC) at both soil layers. Bulk density in the 0–5 cm layer decreased from 1.37 for the control to 1.26 g cm−3 for the treatment that received yearly biochar for 10 consecutive years (Table 1). The corresponding decrease of SDC was from 67.7% to 62.8%. Both of these parameters were linearly and negatively correlated with the number of years when biochar was applied (x) as follows: BD = –0.038x + 1.422 (R2 = 0.94, p < 0.05) and SDC =  −0.519x + 67.816 (R2 = 0.92, p < 0.05). That biochar did not significantly affect BD or SDC in the deeper soil layer (5–10 cm) probably is an indication that its mass movement below the 0–5 cm layer was restricted in this no-till crop** system.

Biochar also enhanced > 0.25 mm soil aggregate stability (SAG) and structural stable index (SI). Relative to the control, biochar application across all treatments increased SAG and SI by an average of 21% and 67% in the 0–5 cm depth and by 8% and 52% in the 5–10 cm depth, respectively (Table 1). Similar to BD, the least increase of SAG and SI occurred for the fewest applications and the greatest increase occurred with the most applications. The results indicate that, after biochar application, soil particles aggregated and were more difficult to be disrupted during the wetting process due to the improvement of soil structure from biochar applications. However, yearly application for at least 5 years may be necessary to secure substantial improvements.

3.2 Soil pore characteristics

Biochar addition altered soil pores in the 0–5 cm soil depth. Relative to the control, biochar application significantly increased the amounts of pores with < 10 μm diameter at the depth of 0–5 cm (Fig. 2). The percentage of residual pores (< 0.5 μm) increased from 15.6% for the control to 24.1% for the treatment that received biochar for 5-years. The corresponding increase for storage pores (0.5–50 μm) was 7% relative to the control. This shows that biochar increased smaller pores (< 0.5 μm) more than it did larger pores (0.5–50 μm). Biochar addition had no significant effect on transmission pores (> 50 μm).

Fig. 2
figure 2

Normalized pore volume distributions in the soil that received yearly biochar application for 0 to 10 years at the depths of 0–5 and 5–10 cm

Full size image

At the 5–10 cm depth, biochar amendment increased soil pores of < 0.5, 0.5–50 and > 50 μm in diameter. The increase averaged across all treatments relative to the control was 24%, 10% and 9% for pore sizes of < 0.5, 0.5–50 and > 50 μm, respectively (Table 2). The results overall show that biochar application increases soil porosity, but this effect mostly occurs for smaller pores (< 0.5 μm) at both 0–5 and 5–10 cm depths.

Table 2 Functional pore classes and total porosity (TP) in soils that received yearly biochar application for 0 to 10 years at the depth of 0–5 and 5–10 cm
Full size table

Biochar increased soil total porosity (TP) at both depths as a result of its effect on the different pore size classes. Total porosity linearly increased with the number of years of biochar application (Table 2). The increase averaged across all treatments of biochar application relative to the control was 11% at the 0–5 cm depth and 12% at the depth of 5–10 cm. The 5-year application at the 0–5 cm and the 10-year application at the 5–10 cm depth resulted in the greatest TP increase .

3.3 Soil water retention

Biochar application notably affected soil water retention curves (SWRCs) (Fig. 3). The differences among the treatments in SWRCs occurred across the full range of soil matric potential. Biochar-amended soil across the full range of pressure heads had greater soil water content than the control at both 0–5 and 5–10 cm soil layers. The differences were greater and more consistent at the 0–5 cm than at the 5–10 cm depth.

Fig. 3
figure 3

Soil water retention curves (SWRCs) at 0–5 and 5–10 cm depths of soil from fields that received yearly biochar application for 0 to 10 years

Full size image

Table 3 shows saturated (θs) and residual (θr) water contents, and shape parameters (α and n) of the Van Genuchten model (Van Genuchten 1980). Relative to the control, the values of θs and θr averaged across all treatments that received biochar increased by 12% and 9% in the 0–5 cm layer and by 5% and 1% in the 5–10 cm layer, respectively. In the top layer (0–5 cm), biochar increased θs and θr as the number of applications increased, with the greatest increase occurring with the 5-year application (up to 24% and 12%, respectively).

Table 3 Statistics for the estimated values of residual water (θr) and van Genuchten shape parameters (α−1 and n) for the five treatments
Full size table

The effect of biochar on the soil water holding properties can be seen on the changes in shape parameters of the SWRCs. On average, fitted α values (inverse of air entry value) ranged from 0.0071 to 0.0155 cm−1. Biochar addition slightly reduced α values, although the effect wasn’t significant compared to the control at both soil depths. The low α values caused by biochar addition means high air entry value and strong adsorption of soil water. The n values (related to the pore size distribution) ranged from 1.22 to 1.37. The n values of all biochar amendments were less than those of the control treatment at the 0–5 cm depth. At the 5–10 cm depth, only 10 years of biochar addition increased n; none of the other treatments significantly affected n and α at the 5–10 cm depth. The above results confirm biochar addition increased total porosity and changed soil pore size distribution.

Other soil hydraulic properties including Ksat, FC, PWP, and PAW were also affected by biochar application depending on the number of years of application (Fig. 4). Ksat at the 0–5 cm depth ranged from 42.2 to 99.1 cm min−1, FC ranged from 0.28 to 0.39 cm3 cm−3, PWP ranged from 0.10 to 0.16 cm3 cm−3, and PAW ranged from 0.18 to 0.23 cm3 cm−3. At the 5–10 cm depth, Ksat, FC, PWP and PAW were 40.40 to 44.29 cm min−1, 0.26 to 0.33 cm3 cm−3, 0.10 to 0.14 cm3 cm−3 and 0.16 to 0.19 cm3 cm−3. Relative to the control, biochar application averaged across all years increased Ksat, FC, PWP and PAW by 94%, 26%, 43% and 17% at the 0–5 depth and by 4%, 13%, 18% and 16% at the 5–10 cm depth, respectively. The effect was greater at the 0–5 cm than at the 5–10 cm depth. Biochar addition resulted in a smaller PAW increase relative to the increase in FC because biochar also significantly increased soil water content at PWP.

Fig. 4
figure 4

Saturated hydraulic conductivity (Ksat), volumetric water content at field capacity (FC) and at permanent wilting point (PWP), and plant available water (PAW) in soil that received yearly biochar for 0 to 10 years at the depths of 0–5 and 5–10 cm. Error bars show standard deviation. Data points are significantly different (p < 0.05) from each other if denoted by different small (0–5 cm depth) or capital (5–10 cm depth) letters by Fischer's LSD test

Full size image

3.4 Soil aeration

Despite notable variability, biochar affected both air capacity (AC) and water-filled pore space (WFPS) at both 0–5 and 5–10 cm depths (Fig. 5). Compared to the control, biochar averaged across all years of application reduced AC by 25% at the 0–5 cm depth and by 17% at the 5–10 cm depth. It also increased WFPS by 94% at the 0–5 cm depth and by 20% at the 5–10 cm depth. The greatest WFPS increase occurred with the most applications at both depths.

Fig. 5
figure 5

Air capacity (AC) and water-filled pore space (WFPS) in soil that received yearly biochar for 0 to 10 years at the depths of 0–5 and 5–10 cm. Error bars show standard deviation. Data points denoted by different letters at the same depth are significantly (p < 0.05) different from each other by Fischer’s LSD test

Full size image

Overall, biochar addition increased SWSC and decreased SAC (Fig. 6). The increase of SWSC was up to 118% at the 0–5 cm depth but only 9% at the 5–10 cm depth. Biochar addition, relative to the control, reduced SAC by 53% for the 0–5 cm depth and by 12% for the 5–10 cm depth when averaged across all treatments.

Fig. 6
figure 6

Soil aeration capacity (SAC) and soil water storage capacity (SWSC) of soil that received yearly biochar for 0 to 10 years at 0–5 and 5–10 cm depths. Mean values of SAC or SWSC within the same depth followed by the same letter are not significantly different according to Fischer’s LSD test (p < 0.05)

Full size image

3.5 Soil physical quality assessment

Biochar addition increased the scores of 12 soil indicators and the overall soil physical quality index (SPQI). Biochar increased individual soil indicator scores of BD, SDC, SAG, SI, TP, WFPS, SAC, SWSC, θr, θs, Ksat, FC, PWP and PAW above the control, which also led to the increase of overall score (Fig. 7). This increase was greater as the number of applications increased, with the greatest scores being obtained from the 10-year application at both 0–5 and 5–10 cm depths. The scores of all physical indicators increased from 0.17 to 0.23 at the 0–5 cm depth and from 0.11 to 0.17 at the 5–10 cm depth. Biochar also increased the SPQI from 0.16 for the control to 0.26 for the 10-year biochar application at the 0–5 cm depth, with a corresponding increase at the 5–10 cm depth from 0.18 to 0.22.

Fig. 7
figure 7

Individual soil physical quality indicator scores and an overall soil physical quality index (SPQI) scores at 0–5 and 5–10 cm depths of soil that received yearly biochar application for 0 to 10 years. BD, bulk density; TP, total porosity; SDC, soil degree of compactness; SAG: > 0.25 mm water stable aggregate; SI, structural stability index; WFPS, water-filled pore space; SAC, soil aeration capacity; SWSC, soil water storage capacity; θr, residual water content; θs, saturated water content; Ksat, saturated hydraulic conductivity; FC, field capacity; PWP, permanent wilting point; PAW, plant available water. Mean values followed by the same letter are not significantly different according to Fischer’s LSD test (p < 0.05)

Full size image

4 Discussion

This study investigated whether biochar generated from papermill plants as a waste byproduct can gainfully be applied to farmlands. This research showed that applying such biochar to no-till agricultural soils positively affects the soil and likely benefits crop production in addition to the “disposal” benefits for the papermill plants. Yearly application of about 6.7 Mg ha−1 year−1 by spreading on the soil surface of a no-till land for as long as 10 years increased values of all soil physical and hydraulic indices to a depth of 10 cm.

4.1 Soil structure and porosity

Soil texture, defined by the composition of different soil particles, namely silt, clay and sand, has close relationship with soil physical properties. It is a static property and thus will not be changed by the addition of organic amendments. Our results confirmed this property and showed biochar addition had no effect on soil particle size distribution.

Soil bulk density (BD) which is considered to be the main driving force of soil physical properties depicts the potential function of the soil with regards to soil aeration, water and gaseous movement, water infiltration and structural support (Huan et al. 2021). Our results showed that yearly application of 6.7 Mg ha−1 biochar reduced the bulk density (BD) and degree of compactness (SDC) of this soil, although it was applied on the soil surface and was never incorporated with the soil due to the no-till practice. Ten consecutive yearly applications of biochar reduced the BD from 1.37 g cm−3 to 1.26 g cm−3 and SDC from 67.7% to 62.8% at the 0–5 depth and from 1.40 to 1.37 g cm−3 and SDC from 73.2% to 67.7% at depth of 5–10 cm. The reduction in BD and SDC at both depths shows that biochar particles somehow moved from the surface into the soil profile and were effective in loosening, an effect that may be attributed to two mechanisms. The first mechanism is a simple dilution effect where the biochar with much lower bulk density than the soil mixes with the soil leading to a lower BD of the mixture. The second and more likely mechanism for the reduction of BD and SDC may be derived from the effect of the biochar on soil aggregation which enhanced the formation of slaking-resistant macroaggregates and increased the resistance of soil aggregates against fast and slow wetting (Table 1). The enhanced soil aggregation and stability due to biochar amendment could be partly related to the increased TOC in the soils as shown by the significantly positive relationships between TOC and AG (Fig. 8). The increased TOC in biochar amended soil could promote soil aggregation and stability via cementing soil particles together. Higher TOC content in the soil is known to alter soil physical conditions, creating a conducive environment for microbial activity, thus enhancing soil aggregation, stability (Palansooriya et al. 2019), and SI (Table 1). Unlike our results, published research shows that the effect of biochar on soil aggregate stability is inconsistent likely because most of the experiments were performed utilizing different methodologies, soil types, and biochar feedstock materials (Herath et al. 2013). Aggregate stability in soils is critical for maintaining the soil’s resistance to mechanical stresses such as the impacts of rainfall, puddling, and compaction.

Fig. 8
figure 8

Correlation coefficient matrix of soil total carbon, structure, porosity, and water retention. Asterisks denote a significant correlation between two parameters at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***). Absence of an asterisk denotes lack of significant correlation between two parameters at p < 0.05. TOC total organic carbon, BD bulk density, TP total porosity, SDC soil degree of compactness, SAG > 0.25 mm water stable aggregate, SI structural stability index, WFPS water-filled pore space, SAC soil aeration capacity, SWSC soil water storage capacity, θr residual water content, θs saturated water content, Ksat saturated hydraulic conductivity, FC field capacity, PWP permanent wilting point, PAW plant available water

Full size image

The enhancement of soil aggregation (AG) and SI as well as the decrease of soil compaction (lower BD and SDC) after biochar application in our study may further lead to the increase in pore volume and total porosity (TP) (Table 2). The results are in good agreement with Baiamonte et al. (2019), who reported an increase in the percentage of porosity in a desert sandy soil with biochar addition, as compared to no biochar addition. Biochar has a highly porous structure (Qian et al. 2020), which may directly increase soil porosity. The positive correlation between AG and TP in our study (Fig. 8) also showed that biochar, which has high adsorption and cohesion (Qiang et al. 2019) because of its large surface area, combines with fine soil particles to form a larger aggregate structure and increases the soil porosity upon coming into contact with the soil.

However, much of the increase in soil TP belonged to small pores in this study. Our study showed that small pores account for 55% of total porosity increase due to biochar addition. This may be attributed to two mechanisms. The first mechanism is that biochar itself is highly porous although 95% of its pores have less than 0.002 μm diameter (Major et al. 2009). So, when added to the soil it may contribute to the porosity of the soil. The second mechanism is that the surface-applied small biochar particles may have moved into the soil profile probably by the action of rainfall and filled pore spaces larger than itself and seal or reduce the size of the existing larger soil pores (Blanco-Canqui 2017; Zhang et al. 2016).

Consistent with our findings, others also reported that much of the increase in soil total porosity by biochar is due to the creation of small pores. Liu et al. (2016) observed an increase in smaller pores (0.1–10 μm) relative to larger pores (10–1000 μm) when they applied 16 t ha−1 of commercial straw biochar to a loamy soil in experimental plots. Abel et al. (2013) also reported an increase in smaller pore size fractions and a decrease in larger fractions as they applied maize biochar at rates of 1%, 2.5% and 5% by weight to a loamy sand soil in a field study. Głąb et al. (2016) found an increased volume of small pores (< 50 μm in diameter) and a decreased volume of larger pores (50–500 μm) as they applied biochar at rates of 0.5%, 1%, 2% and 4% by weight to a loamy sand soil.

Others, however, have reported results not consistent with ours. For example, Obia et al. (2016) found the addition of crushed maize cob biochar at the rate of 0.8 and 2.5% (w/w) to sandy loams caused an increase in soil porosity in four months, particularly macropores with > 1.5 mm radius. Hseu et al. (2014) also reported that 2.5%, 5%, and 10% biochar application induced the formation of macropores in mudstone slope lands in an incubation experiment. Amoakwah et al. (2017) found that 10 and 20 t ha−1 biochar with or without phosphate fertilizers numerically increased the number of large pores in addition to increasing the fraction of small pores in a sandy loam soil. The inconsistency of results regarding the increase in pore sizes by biochar amendment may be due to soil types, the method of application and whether the biochar is incorporated with the soil, and other unidentified soil factors. The properties of the biochar such as particle size and the kind of biomass feedstock may also affect the pore size created in the soil. But one fact common to all reported results including ours is that biochar increases soil porosity.

The effect of biochar on soil structure depends on how long the biochar has been applied. Less than three years of 6.7 Mg ha−1 year−1 biochar application likely was not long enough to change soil BD, SDC, AG and SI to a depth of 10 cm in a no-till farming system where no mechanical mixing of the biochar took place (Table 1). Soil structure in this soil was slightly affected by the yearly biochar application of 6.7 Mg ha−1 year−1, an amount likely too small to significantly change soil structure with applications fewer than 3 years. We observed distinct effects on soil structure with yearly applications of 5 years or longer.

Results similar to ours on soil physical properties have been reported by Fu et al. (2019) and Moragues-Saitua et al. (2017). They applied 0–6 kg m−2 biochar to a soil in field trials and did not observe significant differences in soil macrostructure and water-stable aggregates at the depth of 0–30 cm. But they reported significant increase in water-stable aggregates and total porosity after 9–12 kg m−2 biochar was applied to soil once.

In our study, the greatest effect of biochar on the soil structure occurred on the top layer of 0–5 cm. But it also affected the next 5–10 cm layer although to a lesser degree than the 0–5 cm depth. The biochar each year was applied on the soil surface using a mechanical spreader but was not incorporated into the soil since all fields were managed as no-till cotton production systems. The improvement in soil structure at the 5–10 cm depth could be attributed to the movement of smaller biochar particles with rain water to lower soil profiles. The effect of the biochar on the 0–5 cm depth may be important but the changes in the 5–10 cm depth may even be more important because substantial root activity takes place at this depth.

4.2 Soil water retention and aeration

Crop growth, microbial activities, and gas exchange dynamics are processes that are significantly influenced by the ability of the soil to retain water. In our study, application of biochar derived from a papermill plant improved the water retention properties of the soil that was under continuous cotton production for many years. The increase in soil water retention after biochar application led to the increase of FC, PWP and PAW, which are consistent with previous studies (Liang et al. 2021; Ling et al. 2021) suggesting that the application of biochar can increase PAW. These results indicate that the effect of biochar on increasing the PAW under field conditions could lead to reduced irrigation frequency and cost during dry spells and allow plants to survive intermittent drought due to uneven rainfall distribution (Feng et al. 2016) and produce more under rainfed production situations.

However, the increase of PAW was small compared to FC and PWP. This lack of PAW increase proportional to the FC is attributable to the dynamics in total soil porosity associated with biochar application. Biochar application increased soil pores in all three size classes [transmission (> 50 μm), storage (0.5–50 μm), and residual (< 0.50 μm)], but the effect was greatest for residual pores. According to Greenland (1981), residual pores are responsible for the retention and diffusion of ions in soil solutions whereas storage pores and transmission pores are responsible for water stored for movement and plant uptake. Thus, although biochar application greatly increased FC, the increase of PAW was relatively small because much of the water was held at PWP. These dynamics imply that increasing soil water content by applying biochar may improve plant water use but not in proportion to the increase in water held at FC.

Our results overall show that applying biochar improved hydraulic properties of this soil at both 0–5 and 5–10 cm layers. The biochar was applied to the soil surface with no mechanical incorporation as the fields were all no-till systems, so the improvement in the 0–5 cm depth is expected. The improvement in the 5–10 cm depth likely occurred because smaller biochar particles moved down into this layer by natural process such as rainfall.

Soil aeration has been shown to be effective in promoting soil permeability, water use efficiency and crop yields. In general, the soil AC depends directly on the soil porosity and soil water release characteristics (Dexter 2004). Recent studies have shown that a soil AC > 0.14 cm3·cm−3 indicates good soil aeration and minimal damage to crop roots (Reynolds et al. 2009). As shown in Fig. 6, biochar addition reduced AC and made the soil aeration lower than 0.14 cm3·cm−3 due to a high increase in SWSC. It indicated although biochar addition increased the TP, more space was available for water instead of air.

4.3 Soil physical health and recommendation

The effect of biochar on soil physical health in this study was assessed using SPQI calculated based on fourteen soil physical and hydraulic datasets. Overall, biochar application improved soil structure, porosity, and hydraulic scores, thus leading to beneficial changes in soil physical quality. The results show that yearly application of biochar to this no-till soil improved the physical and hydraulic properties of the soil although it was applied to the surface and left without incorporation. Its greatest impact was on the top 5 cm soil layer, but it also improved the next 5 cm layer although to a lesser degree than the upper 5 cm layer. This suggests biochar particles moved down in the soil profile likely with rainwater.

 Continuous cultivation on the same land for decades leads to a serious soil quality decline such as high bulk density, low organic matter content, less aggregate stability, and low soil water retention (Soil Survey Staff, 1999). Average annual precipitation in this region amounts to 1, 307 mm, of which approximately 60% is received in the fallow season during October through April (Feng et al. 2016). Only 40% of the annual total precipitation falls during summer crop growing period from May to September. Further, substantial amount of the annual rainwater is either lost by surface runoff or deep percolation due to slow infiltration and poor water holding capacity of the soil. Our results show that biochar can improve water infiltration and storage capacity of these soils, implying biochar from papermill plants has the potential to mitigate intermittent drought and increase rainfall and irrigation water use efficiency in the field.

Biochar is generated in many localities that house paper mills throughout North America. The disposal of this biochar can be a burden for the local industry. Considering the increasing costs and the difficulties in acquiring new landfill sites (Arshad et al. 2012; Miner and Unwin 1992; Ochecova et al. 2014) as well as the benefits of the biochar for improving the soil structure and water retention (Islam et al. 2021), applying it on cropland near such plants may be an option that benefits the papermill plants as well as the owners of the farmland. An additional benefit that was not the focus of this study is C sequestration. The benefit to the papermill plants of applying biochar on farmlands is reduced cost. The benefit to farmland owners and farmers is improved soil physical and hydraulic properties with potential for improved productivity.

5 Conclusions

Our results showed that the long-term application of biochar derived from papermill plants to the field surface had beneficial effects on soil physical and hydraulic properties. Biochar application reduced soil bulk density and degree of soil compactness depending on how many times it was applied. It promoted the formation of > 0.25 mm aggregates and enhanced aggregate stability in soil. It modified soil pore size distribution and increased total porosity, greatly increased water content held at field capacity and permanent wilting point, and led to a smaller increase of plant available water. Overall, biochar improved soil health scores and SPQI, showing that applying biochar generated as a waste byproduct of papermill plants on commercial farms has considerable potential as a useful soil amendment.