Organic Capillary Barriers for Soil Water Accumulation in Agriculture: Design, Efficiency and Stability

Abstract

Acute shortage of water resources and high unproductive water losses are the key problems of irrigated agriculture in arid regions. One of the possible solutions is to optimize soil water retention using natural and synthetic polymer water absorbers. Our approach uses the HYDRUS-1D design to optimize the placement of organic water absorbents such as peat and composite hydrogels in the soil profile in the form of water-storing capillary barriers. Field testing of the approach used a water balance greenhouse experiment with the cultivation of butternut squash (butternut squash (Cucurbita moschata (Duchesne, 1786)) under sprinkler irrigation with measurement of the soil moisture profile and unproductive water losses in the form of lysimetric water outflow. In addition, the biodegradation rate of organic water absorbents was studied at the soil surface and at a depth of 20 cm. Organic capillary barriers reduced unproductive water losses by 40–70%, retaining water in the topsoil and increasing evapotranspiration by 70–130% with a corresponding increase in plant biomass and fruit yield. The deepening of organic soil modifiers to the calculated depth not only allowed capillary barriers to form, but also prevented their biodegradation. The best results in soil water retention, plant growth and yield according to the “dose-effect” criterion were obtained for a composite superabsorbent with peat filling of an acrylic polymer matrix. The study showed good compliance between the HYDRUS design and the actual efficiency of capillary barriers as an innovative technology for irrigated agriculture using natural and synthetic water absorbents.

1. Introduction

“Water is life” is an idiom that is especially clear for arid climates with severe water shortages. The threshold for sustainable development is estimated at 500 tons of water per capita annually [1]. This seemingly colossal amount of water for one person is explained primarily by the large water needs in agriculture. Photosynthesis is an extremely water-intensive process where 300–500 tons of water or more is required for every ton of plant production [2]. The water requirements of the main crops (rice, wheat, alfalfa, maize, soybean) of arid irrigated agriculture amount to 8000–10,000 tons of water per hectare annually [3]. However, the annual irrigation rate in arid conditions is usually doubled, reaching 18,000–20,000 tons of water per hectare [3]. The reason is high unproductive losses of irrigation water due to evaporation and, especially, rapid infiltration beyond the topsoil in the arid soils with low water retention typical for deserts. The downside of these losses is the development of secondary salinization of irrigated areas [4,5]. The scale of soil degradation from salinization according to the FAO global mapping is approaching a billion hectares, including 424 million hectares affected by salinization of the 0–30 cm root layer and 833 million hectares where salinization extends to deeper to 30–100 cm deep soil horizons [6]. A close expert assessment of the susceptibility to salinization of about 955 million hectares of the world’s land resources is given in review [7]. Accelerated anthropogenic salinization of the arid environment is caused both by a direct negative impact in the form of imperfect irrigation technologies that contribute to the development of secondary salinization and by indirect causes of land aridization due to global climate change [8]. These negative trends are especially relevant and dangerous in drylands with intensive irrigated agriculture, where up to 38% of the world’s population lives and where from 20 to 80% of irrigated lands are subject to active secondary salinization, while, for a number of Asian and African countries, the scale of these trends may be underestimated due to a lack of quantitative information [6,8,9].

Solving these most significant environmental problems of irrigated agriculture requires optimizing soil water retention to reduce non-productive water losses and enhance the efficiency of water use in crop production. Water retention in topsoils with maximum root concentration optimizes root water absorption and, consequently, biomass growth and plant yield [3,4]. Irrigation rates in southern Chernozems (Russia, Ukraine) with high water retention are 1.3–1.5 times lower than in arid coarse-textured soils with low water retention (Central Asia, Kazakhstan) and high unproductive water losses of up to 10,000–15,000 m³·ha⁻¹, or more than 50% of annual irrigation [3]. Soil water retention and interception of irrigation water by plant roots reduce subsurface runoff, and the risk of irrigation water coming into contact with saline groundwater is the main cause of secondary salinization for irrigated lands [4,7].

Natural and synthetic organic materials are successfully used to increase water retention in soil both directly through their own water absorption and indirectly through optimization of soil structure [5,10,11,12]. Among the most frequently used natural materials and their modifications are organic fertilizers and plant residues [13,14,15], composts [16], and vermicomposts [17,18], biochar, and peat [19,20,21]. These materials are commonly applied in large doses (1–5% of the soil mass or 10–50 t·ha⁻¹ and more) to increase the organic carbon content of the soil and, accordingly, its water retention [12]. The greatest effect of organic treatment is usually observed in coarse-textured soils with initially low water-holding capacity [10,20]. Thermodynamic analysis of water retention in soils of different genesis and texture shows that each percentage of organic carbon adds an average of 19 J·kg⁻¹ of integral water retention energy [10]. The maximum water retention energy (800–1200 J·kg⁻¹) as well as the maximum total water capacity (80–95 vol%) are observed in detritus (forest litter, peat), therefore peat is often used in urban landscaping and greenhouse agriculture to create artificial soils with specified properties (constructozems) [21].

Synthetic and composite superabsorbent polymers (SAPs) for improving soil water retention are attractive due to significantly lower working doses in the range of 0.1–0.8 mass%, i.e., 10–50 or more times lower than the norms of conventional organic fertilizers and ameliorants [22,23,24,25,26,27]. These doses effectively increase the field capacity and active water range by 1.5–2 times, reduce water evaporation by 1.3–3 times and unproductive infiltration losses by 3–10 times, extend the inter-period irrigation period and, in general, significantly optimize the efficiency of water use [23,24,25,26,27,28]. Additionally, along with improving water retention, synthetic, biopolymer and composite gel-forming SAPs enhance plant growth, stabilize fertilizers in the soil and prevent the leaching out of active ingredients (pesticides, agrochemicals) to the groundwater [28,29,30].

Despite the abundance of publications on the effective impact of SAPs on soil water retention, in most cases, they are based on the results of laboratory studies or small-scale vegetation experiments with soil in pots. Under these conditions, it is easy to assess individual elements of water balance and plant productivity under the influence of hydrogels [23,31]. Larger-scale field experiments, as a rule, are limited to evaluating the effectiveness of water use (accounting for irrigation and yield), and, less often, to comparative monitoring of the soil water regime [24,25,32,33]. Field lysimetric studies with a detailed assessment of water balance elements, including unproductive water losses under the influence of organic soil modifiers, are very rare [21]. The same is typical for preliminary soil design based on modern models of energy and mass exchange in the “soil–plant–atmosphere” system of the HYDRUS type, which is based on experimental information about the water retention curves and saturated hydraulic conductivity of the soil under the influence of organic soil modifiers [24,34]. However, it is evident that the water-retaining effect of a soil modifier, as demonstrated in a laboratory setting on a small-scale sample, does not necessarily translate to effective water retention at the soil profile level. To ascertain the efficacy of organic materials in the soil, it is essential to consider a multitude of soil ecological factors. These include the depth of the soil modifier, its hydraulic conductivity, the specific crop requirements and root distribution, the frequency and efficiency of irrigation, the biodegradability of polymers and numerous other variables that can influence the effectiveness of organic materials in the soil. In particular, the layered arrangement of soil modifiers with contrasting hydraulic properties at a strictly defined calculated depth leads to a synergistic effect of a “capillary barrier”, which allows water to be retained and accumulated in topsoil [35,36,37]. Conversely, the incorporation of greater quantities of organic matter facilitates the reduction in its biodegradability, thereby extending the beneficial impact on soil water retention [38]. Prevention of biodegradation also means reducing CO₂ emissions from agricultural soil, which is a priority soil and environmental task in connection with the “4 per 1000” initiative adopted at the UNFCCC in Paris in 2015 [39]. It is therefore evident that preliminary smart soil design is of paramount importance, with the caveat that it must be tailored to the specific requirements of each sustainable agriculture project that utilizes natural or synthetic organic soil modifiers [24].

Taking into account the above problems, we aimed this study at the technological smart design of capillary barriers and water balance verification of their efficiency in a greenhouse lysimetric experiment in order to increase soil water retention using natural (peat) and synthetic (composite gel-forming SAPs) organic materials. The main objectives of our investigation included the following:

(i)

Laboratory studies of the hydraulic properties of natural and synthetic organic soil modifiers for soil water retention;

(ii)

Computer design of organic capillary barriers based on HYDRUS-1D version 4.15 software and the author’s nomograms for assessing the biodegradation of polymers in soil;

(iii)

Field testing of the effectiveness and stability of organic capillary barriers in a closed lysimetric experiment under complete monitoring of the main elements of the water balance (irrigation, soil water distribution and storage, transpiration, groundwater runoff).

The scientific novelty and practical value of the study for agriculture consist of the technological smart design of organic capillary barriers, allowing for efficient saving of irrigation water, reducing its unproductive losses by 40–70%, while stimulating crop yields and reducing the emission of gaseous carbon in treated soils. The smart technological development was successfully tested in a greenhouse irrigation system for growing pumpkins and verified by the results of a lysimetric experiment with a complete water balance. The results of the study are of interest to specialists in the field of soil hydrophysics and agrophysics, developing the theory of bulk soil water retention with the effect of layered capillary barriers, as well as for practitioners of irrigated agriculture.

2. Materials and Methods

2.1. Tested Materials and Their Characteristics

The study tested Russian natural and synthetic organic materials to improve soil water retention and save water in smart agricultural and landscaping technologies. The eutrophic peat, a natural biopolymer material enriched with biophilic elements (produced by Russkie Gazony LLC, Moscow, Russia), contained 50% carbon, 4% total nitrogen, 180 mg·kg⁻¹ mobile nitrogen in nitrate and ammonium forms, 430 mg·L⁻¹ potassium and 150 mg·L⁻¹ phosphorus. Its total water capacity in a water-saturated state reached 200 mass% or 95 vol% at a bulk density of 0.5 g·cm⁻³. Synthetic gel-forming SAPs were presented by experimental products of the LLC “RoLiUz” (Perm, Russian Federation), patented in the Russian Federation (see Section 6). In the first polymer soil ameliorant, PSA 5420, an acrylic polymer matrix with a ratio of acrylamide and sodium acrylate of 30/70 was filled by 23 mass% dispersed peat. Its maximum swelling degree (SD) in pure water reached 600–660 g of H₂O per gram of dry gel (

Table 1. Composition and some properties of natural and synthetic soil ameliorants.

Material	Composition	SD, g·g−1	C, %	N, %	Wh, %	pH	T0.5, yrs
Peat	Detritus, mineral fertilizers	2.0 ± 0.1	49.8 ± 0.7	4.1 ± 0.2	24 ± 2	7.1 ± 0.3	1.3 ± 0.2
PSA 5420	PAA*, AcNa, peat	630 ± 30	45.0 ± 0.6	5.2 ± 0.2	34 ± 3	7.2 ± 0.1	3.2 ± 0.2
PSA 5127	PAA, AcAm, peat, silver	675 ± 25	47.5 ± 0.3	20.1 ± 0.5	49 ± 3	7.4 ± 0.1	6.6 ± 0.3

Denotation: PAA*—polyacrylamide; AcNa, AcAm—sodium and ammonium acrylates; SD—the maximum swelling degree in distilled water; C, %; N, %; W_h, %—percentage of carbon; nitrogen; and hygroscopic water; T_0.5—half-life of organic materials; ± is the confidence interval at p = 0.05.

). The SD indicator in a mineralized (1–3 g·L⁻¹) solution of potassium chloride decreased by up to 100–150 g·g⁻¹. External pressure in the range of 1–4 kPa suppressed the SD value to 100–200 g·g⁻¹. In the second innovative composite, PSA-5407, an acrylic polymer matrix with a ratio of acrylamide and ammonium acrylate of 30/70 was filled by 23 mass% dispersed peat with 1% silver in the form of nitrate as a biocide. The SD parameter reached 650–700 g·g⁻¹ in distilled water and 180–250 g·g⁻¹ in a saline solution with a concentration of 1–3 g·L⁻¹. Suppression of swelling by an external pressure of 1–4 kPa reduced the SD to 120–250 g·g⁻¹.

The gel-forming superabsorbents contained 45–48% organic carbon and 5–20% nitrogen and had a slightly alkaline reaction (pH = 7.2–7.4). Their half-life, determined in laboratory conditions [24,38], ranged from 1.3 (peat) to 6.6 (gel with silver biocide) years.

Before application to the soil, gel structures were obtained by 24 h free swelling of dry hydrogels. Due to the high hygroscopicity, the calculation of working doses of soil modifiers (D, [kg]) was carried out according to the following formula:

$$D=m_s{\displaystyle \frac{100+W_h}{100}},$$

(1)

where m_s [kg] is the required oven-drying weight (dose) of the soil conditioner, and W_h is its hygroscopic water content (Table 1).

2.2. Laboratory Experiments and Equipment; HYDRUS-1D Model Setup

The laboratory experiments with soil samples and organic amendments included determination of their hydraulic properties, required for HYDRUS-1D process modeling, and some basic soil properties. The particle size distribution in the soil samples was determined by laser diffractometry [40] using the Mastersizer 3000 (Malvern, UK) instrument. The organic carbon content (C, [%]) in the soil and polymer materials was determined by coulometric titration using the AN-7529 analyzer (Measurer PO, Gomel, Belarus). Peat samples were pre-ground with a mechanical coffee grinder and sieved through a 2 mm sieve. Before application to the soil, gel structures were obtained by 24 h free swelling of dry hydrogels in pure water. After that, they were mixed with soil to achieve a 0.1% concentration, considering the initial hygroscopy of dry gels, according to Formula (1). For example, for a 0.1% dose of PSA 5127 hydrogel in a 100 g sample of dry sand, it was necessary to take 149 mg of air-dry hydrogel with a hygroscopic water content of 49% (Equation (1), Table 1). Next, 14.9 g of pure water was added to the dry gel to obtain a swollen 1:100 gel structure. The resulting jelly was evenly mixed with 100 g of dry sand. This procedure guaranteed uniform distribution of the gel in the soil sample, whereas it is practically impossible to uniformly mix 149 mg of dry gel with 100 g of soil. Water retention curves (WRCs) were obtained by the method of equilibrium centrifugation using a modification taking into account the effect of gravity [10], implemented on the basis of the thermostatic centrifuges FC5515R (OHAUS Corp., Greifensee, Switzerland) and Sigma 2-KHL (Sigma Laborzentrifugen, Osterode am Harz, Germany) with a speed range from 100 to 12,000 rpm or a corresponding range of soil water potential from 0 to 2800 J·kg⁻¹ (pressure head 0–285.4 m). Before centrifugation, the saturated hydraulic conductivity of the samples was determined in centrifuge tubes using the transient flow method with an alternating positive pressure head [21]. The experimental WRC data were fitted by the van Genuchten (VG) model, standard for the HYDRUS-1D software [34]:

$$\Theta ={\displaystyle \frac{{\Theta }_s-{\Theta }_r}{{\left[1+{(\alpha P)}^n\right]}^m}}+{\Theta }_r,$$

(2)

where Θ [cm³·cm⁻³] is the volumetric water content in the soil, P [cm] is the pressure head and Θ_s, Θ_r, α, n and m = 1−1/n are the VG empirical parameters.

2.3. Preliminary HYDRUS-1D Modeling for Field Experiment

The computer design of capillary soil barriers preceding the field experiment used the HYDRUS-1D software [34] and nomograms [38] to assess the transfer in the “soil–plant–atmosphere” system and the biodegradation of organic materials depending on the depth of their location in the soil. The technological HYDRUS-1D modeling included a simulation of the soil water regime and root water absorption, which was dependent on the dose of soil modifier, the method of its application (mixing with soil or without), the depth of the layer and its thickness. The initial condition simulated complete water saturation of the soil (P → 0) or the so-called water-charging irrigation. Evapotranspiration flux and free drainage were used as upper and lower boundary conditions. Water consumption was described by the Feddes model without salt stress with the parameters for cantaloupe most suitable for the selected test culture [34].

2.4. Location, Design and Equipment of the Field Experiment

A field lysimetric experiment testing organic soil conditioners was conducted in 2022 at the experimental station of the Institute of Forest Science, Russian Academy of Sciences (Moscow Western Administrative District; 55.77090274 N, 37.39450062 E), in a ventilated greenhouse. The study used the following 5-block randomized duplicate experimental design. Each block, in the form of a 300 L plastic barrel with a height of 130 cm and a diameter of 54 cm, was equipped with a vertical carbon fiber tube for the Delta-HH2 PR-2 Profile Probe TDR moisture meter (Delta-T Devices LTD, Cambridge, UK) and sprinkling irrigation timers (Green Helper CA-322N, Green Helper, Nanjing, China), as well as mechanical water meters for irrigation and lysimetric water flow from the bottom of the block (VLFG-15U, Valtec, Moscow, Russia). The first block, No. 1, was an untreated control in the form of the original coarse-textured pine forest Cambisol containing 95–98% sand (mineral particles greater than 0.05 mm) and 0.2–2.3% organic carbon with an acidic reaction (pH = 5.4–6.2) (

Table 2. Some physical and chemical properties of the soil used.

Horizon	C, %	ρb, g·cm−3	Sand, %	Silt, %	Clay, %	pH
A1 (0–10 cm)	2.1 ± 0.2	1.3 ± 0.1	94.9 ± 0.5	2.6 ± 0.2	2.5 ± 0.2	5.4 ± 0.4
B (10–50 cm)	0.3 ± 0.1	1.5 ± 0.1	96.3 ± 0.4	2.0 ± 0.2	1.6 ± 0.3	5.8 ± 0.5
C (70–150 cm)	0.2 ± 0.1	1.6 ± 0.1	98.5 ± 0.4	0.8 ± 0.2	0.7 ± 0.2	6.2 ± 0.5

Denotation: ρ_b—soil bulk density; C, %; Sand, %; Silt, %; Clay, %—percentage of carbon and soil particle fractions.

).

The next two blocks included peat treatment in the form of a single-layer structure (block No. 2, 20 cm peat layer at a depth of 20–40 cm) and a two-layer structure (block No. 3, 10 cm peat layer at a depth of 25–35 cm and 10 cm (reserve) peat layer at a depth of 70–80 cm). The last two blocks had a similar design, where, instead of peat in the sand, swollen hydrogels PSA 5420 (block No. 4) and PSA 5127 (block No. 5) were used in doses of 0.1 mass% of the parent sand. The tested crop was butternut squash (Cucurbita moschata (Duchesne, 1786)), the Guitar variety. Plant monitoring included weekly morphometry of aboveground biomass (growth), as well as weight analysis of plant biomass and fruit yield at the end of the experiment. Schemes and photographs of the experimental blocks with growing vegetation are presented in Appendix A. Soil monitoring in the blocks included weekly TDR measurements of volumetric water content at depths of 10, 20, 30, 40, 60 and 100 cm and an assessment of soil respiration (CO₂ emissions), as well as the recording of water inflow during irrigation and its lysimetric runoff. Soil respiration was measured by a closed-chamber method with NDIR CO₂ detection using a portable gas analyzer AZ 7752 (AZ Instrument Corp, Taiwan, China). Synchronous monitoring of temperature (T, [K]) and atmospheric pressure (P_a, [Pa]) by a portable weather station, the La Crosse WS2812 (LaCrosse Technology, Strasbourg, France), was used to calculate the CO₂ emission flux (E, [mgCO₂·m⁻²·h⁻¹]) according to the following equation [38]:

$$E=A{\displaystyle \frac{3600·P_{a}M·V}{{10}^{6}R·T·S}}·1000,$$

(3)

where A [ppm·s⁻¹] is the slope of the linear trend of the increase in the volumetric content of CO₂ in the chamber, recorded by the gas analyzer; 3600 is the conversion factor from seconds to hours; M = 44 g·mol⁻¹ is the molar mass of CO₂; R = 8.314 J·(mol K)⁻¹ is the universal gas constant; V is the volume of the chamber; S is the cross-sectional area of the chamber; 1000 is the conversion factor from grams to milligrams.

Along with CO₂ emission, we used a direct assessment of carbon losses from biodegradation of organic soil modifiers in a field experiment using the application method [38]. Samples of peat or sand with 0.1% SAP with known dry weight and carbon content were placed in a 200 mL glass filter (crucibles G1: 50–70 µm), installed vertically on the surface or in the soil layer. The material was covered from above with a disk of 1.5 mm mosquito net. After the end of the vegetation experiment, the samples with the disk removed from the surface were dried in the laboratory at 95 °C and analyzed for organic carbon content. The percentage of carbon losses (CL, [%]) characterizing the stability of the organic soil modifiers was calculated using the following formula:

$$CL=100·{\displaystyle \frac{m_1·C_1-m_2·C_2}{m_1·C_1}},$$

(4)

where m₁, C₁ and m₂, C₂ are the dry mass of the material and its carbon content before and after the experiment.

Experimental data from water balance monitoring were used for a comparative assessment of unproductive water losses (lysimetric runoff) and plant productive evapotranspiration water flow according to the following water balance equation [21]:

ΔW_S = (I – ET − L)·Δt,

(5)

where ΔW_S [mm] is the change in the soil water storage, I [mm·day⁻¹] is the irrigation rate, ET [mm·day⁻¹] is the evapotranspiration rate and L [mm·day⁻¹] is the intensity of lysimetric runoff for the accounting period Δt.

Automatic monitoring of microclimatic conditions inside and outside the greenhouse every 2 h used DS 1923 dataloggers (Dallas Semiconductor, Dallas, TX, USA).

2.5. Mathematical and Statistical Data Processing

All experiments were carried out in triplicate with subsequent statistical processing of data. Statistical and mathematical processing of the results, including WRC data approximation by the VG model and numerical integration of soil water content distributions in water storage calculations, was carried out using MS Excel, Microsoft Office 2016 (Microsoft, Redmond, DC, USA), the R (3.5.3) program (USA, RStudio PBC, Boston, MA, USA) and S-Plot 11 (Systat Software GmbH, Erkrath, Germany) computer software.

3. Results

3.1. Laboratory Analysis of Hydraulic Properties in Soil and Organic Soil Conditioners

The first illustration shows the WRCs of coarse-textured initial soil with organic soil modifiers (Figure 1, left part) and their saturated hydraulic conductivity (Figure 1, right part).

The sandy samples had low water retention capacity. Despite the water content of the saturated soil being 40–50%, the relatively small suction of −10 kPa, corresponding to a 100 cm soil column (horizontal dotted line in Figure 1, left part), removed almost all this water from the topsoil. The total water capacity of peat compared to sand was twice as high and reached 90–92%. An important advantage of the peat soil modifier was high water retention in the entire range of pressure head from 0 to 10⁴ cm (Figure 1, left part). In particular, the residual water content at the pressure head of −100 cm exceeded 60%, i.e., it was greater than the total water capacity of the parent sand. Therefore, in the peat layer located in the topsoil, after the gravitational discharge of water from the pressure head of 100 cm, the volumetric water content should be more than in the water-saturated parent sand. The water-retaining effect of gel-forming SAPs was somewhat less. The total water capacity was close to 60%, and the water content after the suction at −100 cm varied from 40 to 55%, which is actually equal to the total water capacity of the parent sandy soil. It should be noted that this is the effect of a small dose of SAPs, 0.1%, i.e., the mass of dry SAPs is 1000 times less than the mass of dry peat soil modifier. This remark is important from a logistical point of view considering the significant reduction in costs for transportation and application of synthetic superabsorbents. Both soil modifiers are highly dispersed polymeric materials (specific surface area of 300–600 m²·g⁻¹ or more [10,24]) with high surface energy, which is the main factor in their powerful water retention. In contrast to coarse-textured, capillary-porous sands, their compositions with 0.1% hydrogels, as well as 100% peat in the swollen state, have a high water-holding capacity and fairly high (pressure head = 30–70 cm) values of the intake point, as a conditional boundary between the saturated and unsaturated state of the soil, according to [40]. This feature indicates the dominance of the non-capillary mechanism of water retention for such a swollen two-phase system by analogy with the contraction of fine-textured clay samples and micro-powders in [40], where the intake point reaches 10–100 kPa (pressure head ~100–1000 cm).

In the pressure head range of 0–10⁴ cm, all WRCs are successfully fitted by the standard VG model (solid lines in Figure 1, left part) with determination coefficients R² = 0.993–0.999 and standard errors of estimate s = 0.2–2.2%. The parameters of the VG model, as well as the values of saturated hydraulic conductivity, used in the HYDRUS-1D process modeling are given in

Table 3. Parameters of the VG model and unsaturated hydraulic conductivity (K) for the parent sandy soil and organic soil modifiers.

Material	Θs, %	Θr, %	α, cm−1	n	R2	s, %	K, m·day−1
Sand, A1	1.9	50.9	1.051	1.304	0.997	0.87	4.1
Sand, B	2.6	47.5	0.716	1.601	0.999	0.16	4.3
Sand, C	1.4	43.4	0.367	1.639	0.999	0.38	1.2
100% Peat	0.2	91.6	0.038	1.217	0.995	1.89	4.8
0.1% PSA 5420	2.7	55.5	0.183	1.501	0.996	1.42	0.6
0.1% PSA 5127	0.2	68.5	0.203	1.290	0.993	2.23	0.5

Denotation: R², s—coefficient of determination and standard error of estimate; K, saturated hydraulic conductivity; other parameters, see Equation (2).

.

The saturated conductivity of different horizons of coarse soil and peat soil modifier was high and usually varied in the range of 5–10 m·day⁻¹ (Figure 1, right part). Therefore, in case of unexpected overwatering, coarse-textured soils with peat should quickly drain excess water. However, the layered structure with alternating sand and peat can greatly (up to 10 times or more) reduce the saturated hydraulic conductivity, probably due to the Jamin effect at the boundary of sand and peat layers [38,41]. In the hydraulic engineering practice of the USSR, the method of introducing peat lenses into sand with subsequent compaction of the soil made it possible to significantly reduce infiltration losses from water channels [42]. A small dose of gel (0.1 mass%) in sand reduces the unsaturated conductivity of the composite to 0.3–0.6 m·day⁻¹, or 20–40 times compared to the parent mineral soil (Figure 1, right part). Higher doses of 0.2–0.3% are sufficient to transform sand into an aquiclude with a saturated hydraulic conductivity of 3–10 cm·day⁻¹ or lower [24].

The high water retention of peat or SAPs obviously does not guarantee the optimization of the water regime of the entire soil column since it depends on the dose of the soil modifier, the thickness of the capillary barriers and the depth of their location. Therefore, the next step of our research was the HYDRUS-1D design of organic capillary barriers aimed at optimizing water retention and root water absorption at the level of the entire soil column planted with an agricultural crop.

3.2. HYDRUS-1D Design of Capillary Organic Barriers for Bulk Soils

Figure 2 shows the selected numerical stimulation of water dynamics in the parent sand and in the layered soil constructions with water-accumulative capillary barriers from natural and synthetic soil modifiers. As expected from the results of the thermodynamic water retention assessment, water in the saturated parent sand leaves the topsoil very rapidly (in less than a day) due to gravity drainage with a suction of −100–130 cm. The water consumption of the agricultural crop at a potential evapotranspiration rate of 5 mm·day⁻¹ with linearly decreasing root distribution in the 10–80 cm zone drops sharply after 11–12 days. On the twentieth day, it decreases to the conditional wilting limit of 1 mm·day⁻¹. The introduction of a layer of peat or 0.1% gel-forming SAP into topsoil at a depth of 20–40 cm radically changes the water regime of the soil column, prolonging the root water uptake. Despite active water consumption and intra-soil water outflow, the water content in the layer with organic soil modifier remains at the level of 20–40% even at the end of the 30-day period, i.e., 10–20 times more than in the parent sandy soil. The deepening of organic soil modifiers by 20 cm from the soil surface, on the one hand, protects the accumulated water from evaporation, and, on the other hand, increases their resistance to biodegradation by up to 8–10 times, according to the nomograms for sustainable soil engineering [38]. In addition, the removal of organic soil modifiers to a depth greater than the usual depth of mechanical cultivation (plowing, loosening) obviously guarantees their inviolability in real agricultural conditions, different from the experiment in the greenhouse. In comparison with the parent sandy soil, the root uptake is prolonged by up to 30 days, remaining at a high level of 3–5 mm·day⁻¹, necessary for plant productivity. The placement of organic soil modifiers on the soil surface can give an even greater effect, as, for example, in the case of forest litter or urban peat-based constructozems [21]. However, in this case, all the above-mentioned advantages of protecting organic capillary barriers from water evaporation and biodegradation are completely lost.

For fast-growing crops with deep root systems, an alternative soil design can be recommended based on the results of HYDRUS-1D process modeling (Figure 2, left part). Here, soil modifiers are divided into two smaller (10 cm) layers. One remains in the topsoil at a depth of 20–25 cm, and the other (reserve) is removed from the surface by up to 70–80 cm. The efficiency of this split structure in specific water retention remains just as high (20–40% water content after 30 days), but the stimulation of root consumption is reduced compared to a powerful single-layer composite since the water storage in the topsoil is halved. The prolongation of root consumption is accompanied by a decrease in its rate of up to 2 mm·day⁻¹ at the end of the 30-day period, which is 2 times higher than in the parent sand and 1.5 times lower than in the single-layer structure. At the same time, this process modeling did not consider the possibility of activating root water absorption during the growth of an agricultural crop with a concentration of roots in the reserve layer and, consequently, with the possibility of a more active water supply. These and some other advantages of double-layer capillary barriers were demonstrated directly in a lysimetric vegetation experiment.

3.3. Field Experimental Results of Water Regime and Water Balance for Soil Constructions with Organic Capillary Barriers

The results of the water content monitoring depending on the soil depth after spline interpolation were presented in the form of 2D diagrams of the soil water regime in the experimental blocks. Their numerical integration over the soil profile gave the water storage depending on time. After that, using the water balance Equation (3), we calculated the productive evapotranspiration water consumption for plants and compared it with unproductive losses in the form of lysimetric runoff for different experimental blocks. Figure 3 presents the results of this assessment for the untreated control (block No. 1). The diagram of the water regime shows that the topsoil in this block to a depth of 30 cm and more was almost constantly characterized by a lack of water (water content less than 10%). At the same time, the irrigation rate was the largest among the blocks and periodically reached 15–20 mm·day⁻¹. Due to the low water retention of the soil, all irrigation water quickly left the topsoil and accumulated in the lower part of the block, where the volumetric water content reached 30–40%, i.e., was close to the total water capacity of the sand. The low water retention of the coarse-textured soil determined high non-productive water losses in the form of cumulative lysimetric runoff, reaching 64.3% of the total water input to the soil (Figure 3, lower part). Even during periods of decreasing irrigation flow, the lysimetric flow remained high, reaching 5 mm·day⁻¹. With high irrigation, it increased to 10–15 mm·day⁻¹ or more. Against this background, only 35.7%, or slightly more than 1/3, of the water entering the soil remained for the plant productive evapotranspiration flow.

The introduction of a 20 cm layer of peat into the topsoil significantly improves its water retention capacity and redistributes the water balance towards the evapotranspiration flow (Figure 4). The water content in the layer with the soil modifier reached 60–70 and more, which also correlated well with the results of preliminary technological modeling in HYDRUS-1D. Clear boundaries of the peat capillary barrier were diagnosed by a sharp decrease in water content in the adjacent sand layers located above and below. In these layers, the soil dried out to a water content of 10% or less, which periodically increased due to irrigation. In the lower soil layers at the base of the block, as in the previous case, the water content was consistently high (30–40% or more) despite free drainage and the frequency of irrigation water supply at different rates.

The water-retaining effect of the capillary peat barrier in the topsoil caused a significant reduction in non-productive water losses in this block. Lysimetric runoff, as a rule, did not exceed 5–10 mm·day⁻¹, and, in certain periods with a low irrigation rate, it dropped to zero. Its cumulative share in the water balance was 37.5%, that is, slightly more than 1/3 of the irrigation water entering the soil. Plant productive evapotranspiration consumption in this block, accordingly, increased to 62.5%. Thus, compared to the previous block, the water balance has changed from unproductive to productive for the agricultural crop. This experimental result fully confirmed the effectiveness of a single-layer peat capillary barrier of great thickness placed in the topsoil, as predicted by the by the HYDRUS-1D simulation.

The current situation with the two-layer peat capillary barrier was somewhat different from the forecast at the stage of technological modeling (Figure 5). Regarding the water regime, everything corresponded to the forecast, and the two-layer composite effectively (water content of 30–50% or more) retained irrigation water both in the topsoil at a depth of 25–35 cm and in the lower part of the block (depth of 60–80 cm). However, the reduction in unproductive water losses and the increase in evapotranspiration (root consumption) here turned out to be not less than expected from the HYDRUS-1D simulation but more compared to the previous block. Leaving the hypothetical reason for such discrepancies with the forecast for Section 4 (Discussion), here, we only state the fact that there was an increase in evapotranspiration consumption to78.9% with a corresponding decrease in unproductive water losses by lysimetric runoff to 21.1% of the total amount of irrigation water received during the vegetation experiment (Figure 5).

The subsequent experimental block, comprising two capillary barriers based on the gel-forming soil conditioner PSA 5420, yielded results that were no less effective (Figure 6). Due to the lower porosity (total water capacity) of the 0.1% gel composite in sand compared to peat, the water content in the water-accumulating layers was slightly lower than in the previous blocks, reaching 30–50%. However, this moisture, as predicted by the HYDRUS-1D simulation, was close to the saturated water content in the parent sandy soil. As a result, the lysimetric runoff in this block with only 0.1% dry hydrogel content in sand was reduced by up to 20.6%, i.e., as effectively as in a two-layer composite based on 100% peat. After a small irrigation rate, the runoff fell to zero. In the case of a large watering, it rose to 2–4 mm·day⁻¹. In general, the dynamics of the lysimetric runoff adequately reflected the flow of irrigation water, unlike in previous experimental blocks, where there was often a delay in the lysimetric flow relative to the water input to the surface. Effective water retention by the gel capillary barrier resulted in an increase in plant productive evapotranspiration to 79.4% of the total amount of irrigation water. Equally successful results were obtained for the second innovative gel-forming soil conditioner, PSA 5127, with a peat filler, acrylic polymer matrix and silver ions as a biocide (Figure 7).

Due to the planned reduction in irrigation in this block, the water content of the soil was lower than in the other experimental blocks. In the layers with soil modifier, it was mainly in the range of 20–40%, and, in the surrounding sandy layers, it was about 10–15% and lower. This circumstance increased the effect of the capillary barrier, so the cumulative lysimetric runoff here was reduced by up to 17.0%, or the lowest value among all experimental blocks.

The lysimetric runoff was observed to drop to zero at regular intervals, albeit with a lag relative to the irrigation pulses. Its surges did not exceed 2–3 mm·day⁻¹, i.e., they were minimal in comparison with the runoff rate in other experimental blocks. Minimizing unproductive water losses led, according to the water balance Equation (3), to the highest results for evapotranspiration flow among all experimental blocks. Their cumulative assessment reached 83%, which was more than two times higher than the total evapotranspiration water consumption for the untreated control. In general, for all blocks with organic capillary barriers, the experimental results showed an effective reduction in unproductive water losses of 40–70% and an increase in plant productive evapotranspiration water consumption of 75–130% relative to the untreated control. Since the productivity of agricultural crops is directly related to their water consumption, organic capillary barriers, shifting the water balance towards increased evapotranspiration, contribute to an increase in plant biomass and yield.

3.4. Comparative Analysis of Plant Productivity

A comparison of plant growth and yield indicators under the action of capillary soil barriers is shown in Figure 8 (upper and middle parts).

The growth of agricultural stems was well fitted (R² = 0.994–0.998; s = 6.6–19.3 cm) by the well-known Verhulst–Pearl logistic model in the form of the following equation from the “Regression Wizard” toolbox of the S-Plot-11 program:

$$y=y_0+{\displaystyle \frac{a}{1+\exp \left(-\frac{t-t_0}{b}\right)}},$$

(6)

where y₀, a [cm] are the initial and maximum plant length; and t₀, b [day] are empirical growth parameters. The half-life growth index (T_0.5) for the model (6) can be estimated using the following formula:

$$T_{0.5}=b·\ln \left({\displaystyle \frac{2}{1-2·y_0/a}}\right)+t_0$$

(7)

The approximation indicators of agricultural crop growth by model (6) and its half-life calculated by Equation (7) are presented in

Table 4. Parameters of the Verhulst–Pearl model (6) for the growth of agricultural crop stems under the influence of organic soil modifiers.

Block Number	y0, cm	a, cm	b, day	t0, day	T0.5, day	R2	s, cm
No. 1	22.5 ± 2.1	246.2 ± 6.0	5.7 ± 0.4	28.3 ± 0.4	30.4	0.997	6.6
No. 2	50.7 ± 20.6	731.4 ± 25.4	8.5 ± 0.7	29.6 ± 0.8	31.9	0.995	19.3
No. 3	20.8 ± 10.7	828.7 ± 16.6	8.1 ± 0.4	39.1 ± 0.5	39.9	0.998	15.4
No. 4	57.1 ± 9.7	617.2 ± 15.8	10.6 ± 0.4	39.0 ± 0.5	42.9	0.998	9.9
No. 5	47.4 ± 14.7	514.8 ± 18.5	7.4 ± 0.7	31.4 ± 0.9	34.2	0.994	17.4

Denotation: R², s—coefficient of determination and standard error of estimate; other parameters, see Equations (6) and (7).

.

The logistic growth curves according to model (6) are shown in Figure 8 (upper part) by solid lines. Analysis of the growth curves and their parameters confirms the significant influence of capillary organic barriers. Plants in the untreated control had the shortest stems (2.2–3.0 m), and, in blocks with peat soil modifier, the longest (7.3–8.6 m). For blocks treated with gel-forming conditioners, this indicator occupied an intermediate position (5.1–6.8 m). The growth rates, characterized by the b and T_0.5 indicators, were at a maximum in the control (b = 5–6 days, T_0.5 = 30 days). Treatment with organic soil modifiers slowed down the growth of the agricultural crops (b = 8–11 days, T_0.5 = 32–43 days). The greatest slowdown, with an increase in the half-life of 10–13 days, was observed in blocks No. 3 and No. 5 with two-layer peat and PSA 5420 hydrogel with a peat filler of an acrylic polymer matrix. In the case of hydrogels, a similar result in terms of slowing down plant growth was obtained earlier for potatoes [24]. Despite the longer growth, treatment with organic soil modifiers significantly (up to 4–6 m) increased the size of the crop stems compared to the untreated control.

The dry biomass of plants, excluding the mass of the vegetable harvest, responded to organic capillary barriers somewhat more weakly, but with the same tendency as in the case of stem length (Figure 8, middle part). The maximum values of aboveground biomass (430–500 g·m⁻²) and dry roots (220–260 g·m⁻²) were observed in experimental blocks with peat. Lower values (200–360 g·m⁻² for tops and 260–280 g·m⁻² for roots) were obtained for plots with hydrogels. The excess over the untreated control in the case of total biomass for blocks with organic soil modifiers was 1.7–2.5 times. ANOVA also confirmed the influence of the treatment factor on the biomass of the agricultural crop at a high level of significance (p-value = 3.1 × 10⁻⁵). The preliminary Shapiro–Wilk, Pearson and Kolmogorov–Smirnov tests were passed successfully (p-value = 0.175–0.532), as was the Bartlett test for homogeneity of variance, so we used the parametric ANOVA. A posteriori post hoc analysis using the Fisher LSD criterion showed the difference between all blocks with soil modifiers and the untreated control. The blocks also differed significantly from each other, except for variants No. 2 compared to No. 3 and No. 3 compared to No. 4, which is clearly visible from the diagrams of biomass (Figure 8, middle part). The more stringent Tukey test revealed no statistically significant differences between any treatments (p-value = 0.150–0.914), while their differences from the untreated control were statistically significant (p-value = 4 × 10⁻⁴ –1.6 × 10⁻²).

The fresh yield of vegetable crops showed a slightly different reaction to organic capillary barriers. The maximum yield (15–20 kg·m⁻²) was in blocks No. 3 and No. 5 with two-layer barriers based on peat and innovative gel-forming composite PSA 5127. In blocks with a single-layer peat barrier and a two-layer barrier made of polymer composite, PSA 5420, the fresh yield was lower (12–14 kg·m⁻²). The parametric ANOVA, selected by the successful (p-value = 0.705–0.881) passing of the corresponding normality tests, also confirmed a significant (p-value = 1.9 × 10⁻⁴) effect of treatment on yield, but the differences between treatment variants here were different compared to for dry biomass. The Fisher LSD criterion did not reveal statistically significant differences between blocks No. 2 and No. 4 and No. 3 and No. 4. The Tukey criterion showed the absence of a statistically significant (p-value = 0.185) excess of yield in block No. 2 relative to the control. According to this criterion, variant Nos. 3, 4 and 5 significantly differed from the untreated control (p-value = 1.1 × 10⁻⁴ –4.2 × 10⁻²). Between blocks No. 2 and No. 3, No. 2 and No. 4, No. 3 and No. 4 and No. 3 and No. 5, the Tukey criterion did not reveal significant differences in fresh yield (p-value = 0.129–0.862).

Note that only crop yield showed the soil dependence of water retention and root water use predicted by the HYDRUS-1D simulation (Figure 2) and not herbaceous biomass and its growth. Overall, the experiment confirmed the increase in root water consumption and, accordingly, crop yields by 1.5–2 times or more under the influence of water-accumulative capillary barriers based on natural and synthetic organic materials, as predicted by technological modeling. The results of long-term (more than 1 year) storage of vegetables under room air conditions showed a stronger desiccation of pumpkins from the untreated control group (Figure 7, lower part). Vegetable mass losses were well fitted by an exponential trend, with the slope for the untreated control being 1.33 higher than for the variants using organic soil modifiers. Probably, the faster weight loss for vegetables from the untreated control was caused not only by their desiccation, but also by biodegradation, since some of the vegetables rotted at 7–8 months of storage. The harvest from the treated blocks was successfully preserved until the end of the year and longer, without signs of rot damage.

3.5. Comparative Analysis of CO₂ Soil Emission and Biodegradability of Organic Soil Modifiers

Monitoring of CO₂ emissions showed the advantage of gel-forming SAPs compared to peat soil modifiers in the potential optimization of the C-balance of agricultural crops (Figure 9). Carbon dioxide emission in blocks No. 2 and No. 3 with peat soil modifier was consistently higher compared to the untreated control and blocks treated with hydrogels, reaching 1–2 g CO_2·m⁻²·h⁻¹ and higher. The average emission values for the entire duration of the experiment in blocks No. 2 and No. 3 were 1267 ± 332 and 1225 ± 81 g CO_2·m⁻²·h⁻¹, respectively, while, in the control, this indicator was 512 ± 71 g CO_2·m⁻²·h⁻¹. Individual emission peaks of about 1 CO_2·m⁻²·h⁻¹ in blocks with peat occurred in September and even in October, when there was no longer any vegetation in the blocks. This indicates a purely soil-based microbiological component of the emission rather than root respiration. The strong (2–3 times or more) excess of CO₂ emission compared to the control is apparently the result of not only the stimulation of soil biological activity under the influence of peat, but also its own biodegradation. The average emission in blocks No. 4 and No. 5 with hydrogels was 355 ± 81 and 456 ± 157 g CO_2·m⁻²·h⁻¹, respectively. These values did not differ statistically significantly (p = 0.01) from each other or from the emission in the untreated control. The increase in emission under the influence of peat, and the absence of this effect and the sometimes weak inhibition in the case of hydrogels, is explained by both the stronger resistance of synthetic composite polymers to biodegradation [24,26,27,28,29] and their small doses, 100 times lower compared to peat.

An approximate characteristic of the carbon balance can be the difference between net productivity and soil respiration (ΔPhE). If it is positive, the site loses carbon to the atmosphere; if it is negative, it sequesters carbon. Obviously, this approximate criterion is more stringent compared to the generally accepted one, in which net productivity is compared with only the heterotrophic component of soil respiration [43]. In the present case, the ΔPhE criterion, estimated by the average values of emissions and total biomass, including the carbon of the crop, indicated a negative balance (carbon source) for the untreated control (ΔPhE = −20 g per block) and block No. 2 with peat (−50 gC). The two-layered soil structure with peat (block No. 3) had a small positive balance, i.e., a carbon sink (ΔPhE = +12 gC). In the case of hydrogels, the positive balance was the maximum (ΔPhE = 170–190 gC per block). The carbon sink in block No. 3 with high soil emissions was obviously the result of maximum biomass and good yield. In blocks with hydrogels, the significantly higher carbon sink was determined primarily by low soil CO₂ emissions. Therefore, the use of gel capillary barriers to optimize soil water retention and crop productivity provides the additional important advantage of improving the carbon balance of the soil–plant system.

Soil CO₂ emissions indirectly reflect the biodegradation of organic materials and hence their stability. The results of the direct assessment of the biodegradability of organic soil modifiers using the application method in glass filters are shown in Figure 10 (lower part) along with data from the automatic monitoring of air temperature and humidity in the greenhouse and outside it during the experiment (Figure 10, upper part).

The greenhouse’s microclimatic conditions were comfortable not only for vegetation but also for soil microorganisms. The average temperature for the entire period of the experiment was 20 °C, or 5 degrees higher than the air temperature outside the greenhouse. The temperature variation range inside and outside the greenhouse was 0–46 °C and −1.4–35 °C, respectively. The average air humidity inside the greenhouse (RH = 75.9%) was also 10% higher than the humidity outside (RH = 75.9%). Optimal hydrothermal conditions in the greenhouse during the summer season (T = 25 °C, RH = 70%) determined the high intensity of biodegradation for organic soil modifiers located on the soil surface (Figure 8, lower part). Losses from biodegradation for peat varied from 26 to 35%, while, for synthetic composite hydrogels, they were consistently lower (15–22%). At the same time, real soil conditions smoothed out the stronger differences in the biodegradability of peat and hydrogels obtained in laboratory experiments with swelling in distilled water (Table 1). A similar result for hydrogels was obtained in experiments [38] with the addition of extracts from soil and rotting plants that sharply stimulate biodegradation. The introduction of silver biocide did not slow down biodegradation as much as the laboratory experiments suggested. The most effective way to increase resistance to biodegradation in real soil conditions is not the application of biocides, but the removal of organic material from the active surface deep into the soil. The biodegradation of peat removed to a depth of 20 cm from the soil surface decreased to 3.5–5% per season, i.e., actually 7 times. For hydrogels, the CL indicator varied from 1.2 to 2.7%, or 8–12 times lower than on the soil surface. Its transformation into the half-life biodegradation of the material according to the standard exponential model [38] gives 1.6–2.3 years for peat on the soil surface, and 2.7–4.4 years at a depth of 20 cm. Similar indicators for composite synthetic SAPs were 2.7–4.4 years (soil surface) and 26–46 years and longer at 20 cm depth. Disassembling the blocks 1.5 years after the end of the experiments and slow drying of the soil in a greenhouse without watering and vegetation confirmed the preservation of layers with organic soil modifiers removed from the soil surface (photo in Appendix B). The samples from the subsoil were especially well preserved. Hydrogels here bound sand grains into strong conglomerates that did not crumble after air drying. Generally, the experimental results are in good agreement with the theoretical nomograms [38], predicting a decrease in the biodegradation rate of up to 8–10 times for organic materials removed 20 cm from the soil surface.

4. Discussion

The high efficiency of polymers and especially synthetic and composite SAPs in increasing soil water retention is well known. The laboratory results (Figure 1) are in complete alignment with previously published data [10,44] on the high total water-holding capacity (porosity) of peat, which is in excess of 200–250 mass.% or 90 vol.%. Additionally, the moisture content at 100 cm pressure head (pF = 2) is also high, reaching 150–180 mass.% or 50–60 vol.%. Also, our data on water retention in coarse soil under the influence of synthetic gel-forming SAPs were completely consistent with the studies [45,46,47,48] reporting an increase in saturated soil water content of up to 60–70 and in water content at pF2 of up to 30–50%. However, in this study (Figure 1) and in our previous publications [10,24,37], this effect was achieved using hydrogel doses of 0.1–0.3 wt%, whereas, in most known sources [45,46,47,48], the hydrogel doses were higher (0.4–1%). The reason is presumably not only methodological differences in the instrumental assessment of soil water retention (centrifugation or the method of pressure and suction plates) or differences in the quality of hydrogels, but, above all, the method of their application.

Field results on soil water retention (Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7) and plant productivity (Figure 8) also agree well with previously published information in this area. Topsoil amendment or mulching with organic and composite materials results in 30–50% more water content than untreated soil and 20–50% greater plant productivity and crop yields (vegetables, grains, grasses), as shown in publications [23,24,33,37,49,50,51,52]. Abd El-Aziz et al. [23] report greenhouse experiments with tomatoes with a similar high efficacy of a composite biopolymer gel, reducing irrigation volume by 25–50% and increasing soil water retention by 2–2.8 times. In a three-year field study [33], soybean and wheat yields increased to 18% with 100% irrigation under the influence of a composite acrylic hydrogel with cellulose and kaolin. In the case of economical irrigation and in rainfed areas, a similar increase in yield under the influence of hydrogels reached 21–53% compared with the untreated control. The same authors showed an increase in the duration of phenological stages of growth under the influence of hydrogels, similar to in our experiments for pumpkin (Figure 8) and potatoes [24,37]. Various sources report yield increases in wheat and oats of 5–18% [31,33,53,54,55], in rice of 12–31% [56], in corn and soybeans of 31–46% [32,57,58] and in oilseed grasses of 10–25% [59] under the influence of the gel-forming SAPs. The data [49,60] for oats and tomatoes confirm our results regarding the positive effect of hydrogels not only on aboveground biomass and fruit yield, but also on root growth (Figure 8).

Unlike the majority of known field experiments with organic soil modifiers, our study used smart soil design or multivariate computer modeling of water retention, root water uptake, subsurface runoff, biodegradation and other processes to determine optimal doses and placement depths for soil modifiers according to [24]. It showed the advantage of not introducing polymeric materials onto the soil surface, but deepening them in the form of separate layers (capillary barriers). Another effect, discussed below, contributes in this case to an increase in soil water retention in addition to the action of organic soil modifiers. In all experiments, regardless of soil modifiers and rate of irrigation, we observed significant accumulation of water in the lower part of the block despite free drainage (Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7). This additional accumulation of water is explained by the effect of a perfect capillary barrier according to [37]. The bottom of the plastic barrel with a siphon drain completely breaks the capillary hydraulic connection of water in the sand and outside the bottom of the block. As a result, water is “suspended” in the sand at the lower boundary of the block since there is no capillary suction from below, and the gravitational pressure of the water is insufficient to overcome the capillary forces of soil water retention. This mechanism is successfully imitated by a long single capillary. If it is completely filled with water and placed in a vertical position, the water will be removed from the upper part, and will be retained at the bottom because the weight (gravitational pressure) of the remaining part of the water does not exceed the Laplace capillary pressure for the curved surface at the upper and lower boundaries of the “suspended” water column. Additional portions of water flowing from above will replace this column, but will not be able to decrease or increase its height, determined by the well-known Jurin formula for the equilibrium of liquid in a vertical cylindrical capillary [61]. Therefore, in fact, regardless of the irrigation rate, the water content in the lower part of the experimental unit should remain consistently high, which was confirmed by both process modeling (Figure 2) and lysimetric experiments (Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7). A similar perfect barrier can be constructed in the field by laying coarse-textured material (crushed stone, pebbles, wooden fascines, etc.) or by hydrophobizing the soil at a certain estimated depth. In this case, not only additional water accumulation is achieved, but also 100% perfect protection of topsoils from secondary salinization, since salt water from subsoils cannot enter into topsoils due to capillary rupture [24,37]. If heavy precipitation occurs suddenly, as, for example, unexpected heavy (over 10 inches) rainfall in Oman and Dubai this spring, such a subsoil layer of coarse-textured material will serve as drainage.

The direction of soil smart design and smart irrigation using process HYDRUS modeling has been actively developing since the millennium [24,36,37,62,63,64]. Almost all of these studies show good agreement between the predicted HYDRUS data and the experimental results of the water regime and infiltration runoff. However, our study revealed some discrepancies. In particular, HYDRUS-1D predicted a relatively small reduction in cumulative lysimetric runoff from the lower soil boundary under the influence of organic capillary barriers made of peat or hydrogel of no more than 10–20% relative to untreated coarse-textured soil. In reality, the lysimetric runoff decreased by 40–70%, i.e., 2–3 times more (Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7). Apparently, the algorithms for calculating the hydraulic conductivity of layered systems in HYDRUS use an analog of the resistance of electrical circuits [65] and do not take into account the possibility of a sharp decrease in the conductivity of contrasting capillary-porous structures due to the Jamin effect [41]. The alternative explanation for the different moisture conditions (water-charging irrigation with full soil saturation in the HYDRUS simulations and periodic irrigation with different intensity in the field experiments) is, in our opinion, less likely due to the close soil water content and low pressure head in the numerical calculations and field experiments (Figure 2, Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7). This problem requires further study in the field of process modeling of water flows in layered soil constructions with contrasting hydraulic properties.

Biodegradation is one of the main soil processes limiting the efficiency and profitability of organic soil conditioners [38]. The service life of peat, often used in urban landscaping and greenhouses, rarely exceeds 1–2 years due to rapid microbial biodegradation, enhanced by the urban “heat island” and greenhouse microclimate [38,66,67]. As a result, areas with topsoils from introduced peat usually have a negative carbon balance with increased CO₂ emission into the atmosphere, which devalues the benefits of this material as an effective soil modifier [66,67]. Information about the stability of synthetic hydrogels is still quite contradictory. The traditional viewpoint classifies synthetic cross-linking superabsorbents as stable, non-biodegradable materials in contrast to biodegradable polysaccharide gels [29,30,68,69]. In particular, Wilske et al. [69], using an incubation experiment with a 13C-label, reported small losses of polyacrylate superabsorbent of 0.12–0.24% over 6 months of incubation that were practically independent of temperature in the range of 20–30 °C. The review [68] estimates the half-life of synthetic superabsorbents in soil as 5–7 years since the high mass due to cross-linking makes them resistant to breakdown by bacteria. However, the authors [68] do not exclude an increase in the biodegradation of synthetic hydrogels due to the action of other soil microorganisms, in particular, fungal rots. Our study [38] confirmed this possibility and showed high biodegradability of acrylic gels (half-life from 0.2 to 3.5 years for the temperature range of 20–37 °C) if they were swollen not in clean water but in aqueous extracts from rotting plants and humus or mixed with fresh soil. The study [70] reports half-life data from 0.13 to 1.31 for a superabsorbent based on co-polymers of acrylamide and potassium acrylate. The authors [70] also mention a rather strong (about 25%) loss of water-holding capacity of this material during the 8-month biodegradation, which is consistent with some other similar data, including those from our studies [38]. It can be confidently stated that, in real biologically active soil environments, in contrast to the almost “sterile” laboratory conditions, biodegradation causes serious damage to cross-linking superabsorbents, impairing their functionality as soil conditioners. Field experiments (Figure 10, lower part) fully confirm this conclusion, indicating high losses of acrylic gels on the soil surface, close to the rates of peat biodegradation. This serious problem poses the challenge of increasing the resistance of organic soil conditioners to biodegradation.

A promising way to reduce the biodegradability of polymeric materials can be the introduction of biocidal components into their composition [38,71,72]. In this study and in previous works [24,38], we, probably for the first time for soil conditioners, proposed the use of silver ions and nanoparticles as biocides. Known similar developments with silver biocides in gels mainly concern medical preparations and antiseptics [71,72]. However, despite a significant (5–20 times or more) increase in the half-life of polymer superabsorbents under the influence of silver ions and nanoparticles in laboratory experiments [38], field tests revealed a relatively small (1.2–2.4 times) increase in half-life for the PSA 5420 composite (4.0 ± 0.5 years on the soil surface and 56.8 ± 12.5 years at 20 cm depth) with a biocide compared with the pure hydrogel PSA 5127 (3.1 ± 0.6 years on the soil surface and 36.1 ± 6.8 years at 20 cm depth). The deepening of organic soil modifiers to a depth of 20 cm was a more effective method, reducing biodegradation (increasing half-life) by 7–15 times or more (Figure 10, lower part). Therefore, the removal of organic materials to the calculated depth, according to the smart design of capillary barriers, guarantees not only their efficiency in water retention, but also resistance to biodegradation, contributing to the minimization of CO₂ emissions and carbon neutrality of the reclaimed area.

5. Conclusions

This study, using smart soil design and water balance lysimetric experiments in a greenhouse with vegetable vegetation, showed the high efficiency of capillary soil barrier technology based on natural and synthetic polymer soil modifiers in increasing soil water retention and crop productivity. Organic capillary barriers in the soil allowed the following goals to be achieved:

• To increase by 2 times or more the total water capacity and field capacity of coarse-textured soils;
• To reduce by 40–70% unproductive water losses in the form of infiltration runoff;
• To increase by 70–130% plant productive evapotranspiration water consumption and achieve a 20–50% reduction in total irrigation;
• To increase by up to 1.7–2.5 times the dry aboveground (photosynthetic) and root biomass and by up to 1.5–2 times the fresh vegetable yield;
• To achieve a neutral or positive carbon balance in the case of two-layer organic barriers, especially effective if synthetic composite water superabsorbents are used.

The computer smart design of capillary barriers based on HYDRUS-1D and the author’s nomograms for process description of the water regime, root water absorption and biodegradation of organic materials in the “soil–plant–atmosphere” system revealed the advantage of a layered arrangement of organic soil modifiers with distance from the soil surface. This soil design ensures high soil water retention and stimulation of root water absorption along with a strong (up to 7–12 times) reduction in the biodegradation of organic soil modifiers, prolonging their service life in the soil. The use of two-layer barriers is more effective in water retention and reduction in unproductive water losses compared to one layer of the same amount of organic material. The HYDRUS-1D process modeling predicts well the qualitative characteristics of organic capillary barriers in soil, but, in quantitative terms, it may underestimate their effectiveness in reducing unproductive water losses in the form of infiltration runoff. The possible causes of this discrepancy, including the Jamin effect, need further quantitative study.

6. Patents

The synthesis of composite gel-forming SAPs was based on the technology of filed hydrogels (superabsorbents) for soil conditioning, patented in the Russian Federation: patent RU no. 2726561 (https://findpatent.ru/patent/272/2726561.html; accessed on 15 June 2024) and patent RU 2639789 (http://www.findpatent.ru/patent/263/2639789.html; accessed on 15 June 2024).