Phytoavailability and Leachability of Heavy Metals and Metalloids in Agricultural Soils Ameliorated with Coal Fly Ash (CFA) and CFA-Treated Biosolids

Abstract

Application of CFA-treated biosolids (NVS) offers multiple benefits to agricultural soils, including fertilizer replacement, soil rehabilitation, and disinfection. It also poses a heavy metal(loid)s threat to the agro-environment. NVS (and CFA to some extent) was tested in lysimeter and field trials, using soils differing in physicochemical properties and a large selection of crops. Consistently, As, Pb, and Cd concentrations in leachate were at or below detection limit, and these and other heavy metal(loid)s (and P) were within the permitted range in plant tissue. Foliage Mo (occasionally also Se, P) concentrations often increased significantly, especially in crops (legumes, potatoes) grown on marginal soils, which also displayed significantly higher yields. CFA and NVS reduced lettuce and legumes foliage Mn (and occasionally Zn) concentrations, which remained, however, adequate. NVS (214 and 642 mT ha⁻¹), digested sewage sludge (ADS) and its compost (24 and 72 mT ha⁻¹), temporarily increased the DTPA-extractability of some elements (NVS: B, Cr; ADS: Cu, Ni, Zn; Compost: Zn) 10–30-fold. The extractabilities of Fe and P increased by up to six times. These increases vanished soon after additive application, supporting the hypothesis of ‘self-attenuation’ by applied biosolids. Our data indicate that long-term application of NVS (and CFA) to calcareous soils poses no heavy metal(loid)s-related threat to the agro-environment.

1. Introduction

Coal-fired energy production increases constantly, contributing to approximately one-third of the world’s electricity generation at ca. 17.5 Giga Joules in 2022 [1,2]. CFA comprises up to ca. 40% of the combusted coal, and CFA containing <7% and >20% CaO is classified as Class F and Class C, respectively. The world’s annual production of CFA in 2021 was ca. 109 mT, most of which was used beneficially [3,4]. Of the ~28 million mT CFA produced in the USA in 2021, 67% was utilized, primarily as a pozzolan substitute for Portland cement or as clinker, and another 530,000 mT were applied to soils, either as is or mixed with organic wastes [4,5]. Stabilization of raw sewage sludge in Israel in 2017/18 consumed a similar amount of CFA, and the alkaline stabilized biosolids (patented as N-Viro Soil^© or NVS) product was applied to agricultural soils for the benefit of the urban, agricultural, and environmental sectors [6,7,8]. Unutilized CFA is stored in landfills and ponds, spread on soils, dumped at sea, etc., potentially causing environmental damage.

CFA may contain a wide range of heavy metal(loid)s of environmental concern. The more common are As, B, Cd, Co, Cu, Cr, Hg, Ni, Pb, V, Se, and Zn. It may also contain modest amounts of radioactive elements, including Uranium and Thorium [9,10,11]. In a long-term leaching study of 32 Class F fly ash (the kind of CFA used herein) samples, [12] showed that the solubility in naturally occurring fluids of most all cations (including Cd, Fe, Pb, and Se) was very limited, that of Cr, Cu, Mn, Ni, and Zn were slightly soluble in acid (<20% of total metal content), and that As was soluble in a basic solution.

In nutrient-deficient soils and under inadequate fertilization, CFA application constitutes a source of essential micro- and macro-elements such as Ca, Mg, Mo, Se, B, Zn, K, and S, but also of soluble salts and of nonessential elements (e.g., Al, Cs, Rb, As, Hg, and Sr) (Ibid.). Yet, the predominantly mineral CFA can potentially sequester heavy metals and metalloid species. Thus, Kumpiene et al. [13] demonstrated a drastic decrease in Cu and Pb solubility in mine-spoil soil after amending it with an alkaline CFA and peat (each at 5% w/w load) in a lysimeter study. This trend persisted even 10 years after the mixing was performed [14].

One of the uses of alkaline CFAs is for the stabilization and disinfection of sewage sludge to ‘Class A’ [15]. The resulting NVS product can be applied to edible crops that are eaten fresh. Furthermore, it is considered an “exceptional quality biosolid”, which warrants unrestricted application to soil, in the event that heavy metals of concern are below specified concentration thresholds (which is practically always the case). Mixing CFA with sewage sludge might enhance heavy metals and metalloids solubility, transport, and phytoavailability due to binding to sludge-derived DOC [16,17]. Dissolution of sludge OC is likely to occur at the high pH (≥12) and temperature (≥52 °C) required for the sludge disinfection and stabilization to occur [15], thus facilitating the leaching of DOC-complexed elements [18,19].

Still, Logan et al. [20] tested the effect of a single application of a 500 mT NVS ha⁻¹ dose in mini-plots (a Typic Hapludalf) on the uptake of Cd, Cu, Ni, Pb, and Zn by six vegetable crops for three years. The uptake by the crops reflected the elements’ total content in the amended soil and their relative solubility, and it declined in successive years. Lettuce contained the highest metal concentrations, yet all were within the permissible range. Logan et al. [20] concluded that NVS “poses no threat to the food chain from trace element uptake by vegetable crops”. It is noteworthy, however, that the phytoavailability of oxyanions (especially of B and Mo) may increase significantly following CFA or NVS application due to both their elevated content and their high solubility at high pH [19,21,22].

The primary agronomic use of CFA and NVS is as liming agents [23]. Thus, the application to calcareous soils is rather counterintuitive, mostly due to the diminishing solubilities of P and Fe [24]. The pH of calcareous soils rises immediately after NVS application and wetting, and the extent and duration of this rise in pH depend on the NVS load and the buffer capacity of the soil [25,26]. Still, NVS (and CFA) is widely applied for its fertilizer replacement and crop quality improvement values, as well as for the resultant increase in water content at the field capacity of sandy soils [27], and the recovery of the structure of crusting sodic soils which improves inter alia and seed emergence and establishment [6,28,29]. Note that upon the shift to Irrigation with WWE originating from domestic use of desalinized water that is common in Israel (or direct irrigation with such water) stunted growth of vegetable crops was encountered even when grown on calcareous soils. This is because solubilization of soil-inherent carbonates might prove too slow to compensate for leaching and plant uptake [30,31]. In such an event, NVS constituted a readily available calcium source in potato nutrition (Fine and Zig, unpublished).

NVS is also an alternative ‘green’ soil disinfectant, shown to be effective against soil-borne pathogens and weeds [6,25,26], and plant-parasitic nematodes [32,33,34,35,36,37]. For this purpose, the NVS load (and composition) was adjusted to raise the soil solution pH by at least 1 pH unit above the pKa of the Henderson–Hasselbalch NH₃/NH₄ relationship equation. This ensured the dominance in the soil of the biocidal NH₃ species. Hadas et al. [7] estimated the annual weighted potential benefits and costs incurred from NVS application to be 324 US $ ha⁻¹ and 131 US $ ha⁻¹, respectively. The benefits are highest when applying NVS to suppress soil borne diseases in cash crops.

Vast practical experience was gained regarding the use of CFA and NVS in agriculture, especially with respect to plant uptake of heavy metals in confined systems [11,14,20,38]. Yet, data on the fate of heavy metals and metalloids in CFA- and NVS-amended soils from field studies is scarce [27,39]. Herein, we present data collected from lysimeter and field studies on heavy metals and metalloids leaching, fate in the soils, and crop uptake and yield, encompassing a wide range of soil characteristics. The ultimate goal of our study was to help assess the risk of CFA and NVS application to agricultural products and the environment.

2. Materials and Methods

2.1. Experimentation

In our agronomic trials we tested NVS and CFA potential to supply macro- and micro-nutrients, their effect on crop quality and on the physicochemical properties of sodic and crust-forming soils. We were also interested in testing the possible reduction in soil-borne diseases, pests, and weeds using the combination of NVS and NH₃ [6]. The test crops in these plant protection studies were vegetables including potatoes, carrots, peanuts, lettuce, and bell peppers. Full commercial fertilizer suite was provided, regardless of the fertilizer value of the NVS. In fertilizer replacement and soil structure improvements studies, feed and fodder crops (corn, wheat, legumes, and chickpeas) were used as test crops. ADS and ADSC were sometimes included for comparison. The biosolids yearly loading rate was determined on the basis of their total N content, aiming to provide 500 kg total-N ha⁻¹, the maximum manure-derived N load permitted in Israel. Fertilized controls, not-amended by biosolids or CFA, were included in all the experiments.

2.2. Amendments, Soils, Irrigation Water, Analyses

The elemental composition of the soil amendments used in the current study is presented in

Table 1. Components of the amendments used in the lysimeter and field studies.

Component	Units (Dry Weight)	CFA (Fresh)	NVS	ADS	ADSC	WWE (1) (mg L−1 and kg ha−1y−1)
Specific density	kg L−1		0.8	0.85	0.57
Dry weight	kg kg−1	1.0	0.71	0.20	0.59
Specific density dry			0.57	0.17	0.34
Ash	g kg−1	~950	880	240	530
TOC	-”-	2.31	91.8	387	222	14 (56)
IC	-”-	0.29	15.5	0.67	2.8
DOC	mg kg−1	1420	23,150	70,560	15,400
DOC/TOC	Ratio	0.61	0.25	0.18	0.07
Total N	g kg−1	0.064	7.0	63	21	6.8 (27)
Total P	-”-	0.128	3.6	23	13	2.4 (9.6)
K	-”-	0.380	1.8	2.2	5.9	32 (120)
pH(1:5, solid–water)	−log[H+]	12.0	11.5	6.5	6.6	8.4
EC(1:5, solid–water)	dS m−1	4.4	2.6	7.8	6.6	1.80
Cl	mg kg−1	217	40	360	650	340 (1360)
As	-”-	32	13	1.9	2.1
B	-”-	270	330	30	50	0.35 (1.4)
Ca	-”-	31,235	56,000	39,400	70,500	70 (28)
Cd	-”-	1.5	0.5	5.0	1.1	bdl
Co	-”-	21	0.5	76	6.4	bdl
Cr	-”-	134	77	153	111	bdl
Cu	-”-	44	40	540	230	0.01 (0.04)
Fe	-”-	26,600	15,250	6150	5500	0.20 (0.8)
Hg	-”-	<0.5	<0.5	<0.5	<0.5	bdl
Mg	-”-	6280	6400	7100	10,600	30 (120)
Mn	-”-	225	200	240	200	0.04 (0.16)
Mo	-”-	12	4.2	9.6	3.1	0.03 (0.13)
Na	-”-	5550	825	1700	1900	280 (1120)
Ni	-”-	49	42	108	63	bdl
Pb	-”-	44	30	32	39	bdl
S	-”-	2915	2650	11,700	7700	35 (140)
V	-”-	135	65	21	23	bdl
Zn	-”-	72	150	3400	1200	0.06 (0.24)

⁽¹⁾ The first value is the concentration and the value in the parentheses is the amount in kg applied with the irrigation water (taken as 4000 m³ ha⁻¹y⁻¹).

. The additives that were applied in the current study are NVS, CFA, ADS, and ADSC. The Israeli NVS comprises a mixture of raw sewage sludge (80% water content), CFA (Class F), and burnt lime at 60:35:5 w/w ratio on moist basis (23:67:10 dry basis). Approximately 1 million mT of such NVS were produced in Israel during the 2010–2018 period. The two biosolids, ADS and the ADSC, are commonly used as amendments to agricultural soils, and were used here for comparison. They are richer in N, P, and K as compared to the NVS (and CFA), and poorer in As, B, Cr, Mo, and V (and have a much lower pH, Table 1). The ADS had higher contents of Cd, Co, Cr, Cu, Mo, and Zn. The DOC/TOC ratio of the CFA (0.61) and NVS (0.25) is considerably higher, which probably derives from their high pH.

The soils and the locations of the field experiments represent a climatic gradient along the Israeli Coastal Plaines (ca. 200–550 mm average annual rainfall) and range physicochemical properties with plow layer textures varying from sand to fine clayey (

Table 2. Soils ⁽¹⁾ and locations of the experiments.

Soil Classification (USDA-ARS)	Location Name	Location (Coordinates)	Texture
Quartz sand	Sand dunes at Yavne and Bnei-Darom	31°53′08.66″ N 34°43′19.26″ E and 31°49′38.30″ N 34°41′50.12″ E	Sand
Xeric Torripsamments	Bsor R&D Center, Sde-Nitsan, Nir Oz, Nirim, Ein HaBsor, Gevulot, Yevul (locations in the NW Negev)	31°15′41.51″ N 34°23′29.01″ E	Sand and sandy loams
Calcic Haploxeralf	Nahal Oz	31°28′30.82″ N 34°29′32.87″ E	Sandy clay loam
Chromic Haploxerert and Sodic Haploxerert	Revadim	31°47′01.99″ N 34°49′34.40″ E	Fine clay
Lithic Xerorthent	Mishmar David (two locations)	32°01′07.44″ N 35°27′09.60″ E	Loamy pale Rendzina with ca. 70% w/w CaCO3
Typic Haplargid	Tomer (the lower Jordan valley)	32°01′28.80″ N 35°27′41.65″ E	Calcareous clay (35% clay)
Typic Haploxerepts and Typic Haploxeralfs	Beit Ezra, Nir Elyyahu; Kfar Haim (locations in the Coastal Plains)	Between 31°44′02.59″ N 34°39′24.51″ E and 32°21′09.19″ N 34°53′49.96″ E	Sand to sandy clay loam

⁽¹⁾ The sand and the soils are calcareous, at least to some extent. The Haploxerepts and Haploxeralfs have traces of carbonates. The pH value ranges from 7.4 to 8.5 (1:5 water extract). The clay fraction mineralogy in most soils is dominated by randomly interlayered Illite-smectite (I/Sr) and illite. The I/Sr in the Haploxererts (≥90%) is highly crystallized montmorillonite. The Xeralfs are dominated by kaolinite (~65%) and illite (25%). The temperature regime of all the soils is thermic and their moisture regime is xeric (which is imbedded in the classification). The nomenclature is according to Soil Survey Staff (2022).

). The soils belong to the orders Aridisols, Inceptisols, Alfisols, and Vertisols, mainly from the suborder Xeralf [40,41,42]. Edible plants were drip-irrigated with local tap water and feed and fodder crops were irrigated with high quality WWE (Table 1). Fertilization was in accordance with commercial practices while considering the WWE chemical composition [43].

Harvested plant material was separated into canopy (usually leaves) and marketable organs (fruit, seeds, bulbs, taproots). Plant material was thoroughly washed (tap water, dilute HCl, deionized water), dried, and milled. Seeds and grains were not washed. Bulbs and taproots were peeled and rinsed. Outer leaves of lettuce were trimmed off. Plant material and soils were digested in concentrated nitric acid [44,45]. The methods of collection, storage, and analysis of lysimeter leachates are detailed elsewhere [18,45,46]. Evaluation of “potentially phytoavailable” heavy metal(loid)s in soils was performed using the DTPA-TEA extraction [47]. The mode of determination of heavy metal(loid)s and other elements is delineated below.

Fe and P were also measured because of their environmental and plant nutritional significance and because liming (by CFA or NVS) is expected to diminish their solubility and phytoavailability [24]. Calcareous soils are especially susceptible in this regard. On the other hand, Hg concentration in the CFA and NVS was below detection (Table 1) and was not measured herein.

2.3. Mini-Lysimeter Study

Heavy metal(loid)s uptake by Romain lettuce and their leachability were studied in a glasshouse in a mini-lysimeters (2 L; 3.3 kg sand/lysimeter) experiment. Quarzitic sand was mixed with fresh CFA at 3.5% (115 g), 11% (360 g), and 20% (660 g) weight ratios and packed into the lysimeters. These mixing ratios were equivalent to approximately 90, 270, and 450 mT ha⁻¹. Each treatment consisted of fifteen planted and five not-planted mini-lysimeters, one plant per mini-lysimeter. Computer-controlled fertigation was daily, using tap water (pH 7.2) augmented with N (80–120 mg L⁻¹; 75% as nitrate) and P (15–20 mg L⁻¹). Leachate was sampled by attaching a 10 L urine collection plastic bag to the lysimeter drainage outlet. The confined atmosphere prevented degasification and pH changes and possible co-precipitation of carbonates and heavy metals [43]. The leachate was samples and bags replaced every week (7 times altogether).

2.4. The 220 L Lysimeter Study

The leachability of heavy metal(loid)s and their uptake by Romain lettuce were tested in 200 L lysimeters over a 3 y period. Of the fifty-two lysimeters used, twenty-four were packed with quarzitic sand, twenty with the non-sodic Revadim fine-clay, and eight with the Nahal Oz sandy clay loam (Table 2). NVS, ADS, and ADSC were mixed with the 0–15 cm topsoil layer, leaving the underlying 65 cm-thick layer untreated. Four crops of Romain lettuce were grown over the three years of the study (2011–2013). The biosolids were applied in May of each year, each at a 500 kg N ha⁻¹ equivalent (the maximal yearly agronomic rate permitted in Israel), and the NVS was also applied at triple this load. The yearly NVS application was equivalent to 71 and 214 mT ha⁻¹, and those of the ADS and ADSC were 8 and 24 mT ha⁻¹, respectively, accumulating to 214, 642, 24, and 72 mT ha⁻¹ in the three years. Treatments were in four replicates. Forty lysimeters were irrigated with WWE and an additional twelve (all packed with the quarzitic sand) were irrigated with tap water. WWE is widely used in Israel for irrigation of field crops and orchards (but not for vegetables), often in fields applied with biosolids (and other manures). Due to the load of elements and DOC in the WWE, enhanced solubility and mobility of heavy metal(loid)s might occur [17]. Hence, the two water types were applied to NVS-treated quarzitic sand as a worse-case scenario with respect to leaching. The design of the 220 L lysimeter study was fully factorial with respect to ADSC, to the lower (agronomic) NVS dose, and to the non-amended control. Finally, the Revadim soil was treated with all three biosolids (and with NVS also at a triple dose). This was performed to complement the Revadim field study (below). Additional experimental details are provided elsewhere [46].

2.5. Field Studies

More than 60 full-scale experiments and observations were performed in commercial fields under common cultivation practices. The edible crops tested were potatoes, carrots, peanuts, chickpeas, lettuce, and bell peppers. The NVS was applied for soil disinfection at 50–120 mT ha⁻¹, often in combination with ammonium sulfate. The main experiment on feed and fodder crops was a 5-year study at Revadim (Table 2) that started in 2011, which was run to reveal the NVS fertilizer replacement efficacy for a summer corn–winter wheat rotation. Application rate was based on the total N loads of 500 kg ha⁻¹ applied each year, and in addition 1500 kg N ha⁻¹ was applied on the first and fourth years of the study. ADS and ADSC were applied on the same TN basis for comparison. The cumulative NVS, ADS, and ADSC loads in the fourth year were equivalent to 285, 32, and 96 mT ha⁻¹ at the lower application rate, and 428, 48, and 144 mT ha⁻¹ at the triple application rate, respectively. Control treatments received zero and commercial fertilization (250 kg N ha⁻¹ y⁻¹). Based on soil tests, K and P were not applied. This experiment consisted of 96 plots of 72 m² each (6 m × 12 m). Each additive was tested in eighteen replicates, and the control in twenty-four. Note that the 220 L lysimeters and the biosolid plots in the field study received the same loads during the first three years. The exception was the high NVS application rate. While in the field it was applied only in the first year, in the lysimeters it was applied in every year of the three.

NVS impact on legumes (a 1:1 mixture of common vetch and clover) was tested in comparison to a commercially fertilized control. Seeding was in November 2014 on a hilly (2–5% inclination) shallow calcareous soil at Mishmar David (Table 2). NVS was applied at 150 mT ha⁻¹. NVS-free, control plots received basal superphosphate and potash dressing at rates determined by soil tests. Whole plant sampling was performed before the flowering stage in the following April.

2.6. Analyses and Quality Control

Microwave-assisted method was used for soil and plant digestion. Triplicate samples were digested in six ml nitric acid in Teflon tubes at 750 watts microwave radiation for a 20 min period. The internal temperature was maintained at 190 °C. Dilution to volume use RO water. Elemental determination used by End-On-Plasma ICP-AES model ‘ARCOS’ from Spectro Gmbh, Kleve, Germany. The measurement was performed according to EPA standard method 6010c. Calibrations used Merck standards. Certified reference materials for soil analysis (IRMM, Geel, Belgium) was analyzed for validation. Quality control in plant material determinations used standard additions of certain analytes (mainly Pb and As) to the digested samples. Assessment of detection and quantitation levels used 40 CFR Appendix B to Part 136 methodology. Briefly, twenty samples of background solutions were analyzed, and the detection and quantitation levels were set at 3 and 10 times the standard deviations.

2.7. Statistical Analysis

Statistical analyses were performed using JMP 16 [48]. One-way ANOVA tests by pairs were performed to identify significant treatment differences (at min. Prob > F < 0.05). When significant differences were detected, comparisons for all pairs were performed using the Tukey–Kramer HSD procedure.

3. Results and Discussion

3.1. Heavy Metal(loid)s in Lettuce Grown in CFA-Containing Mini-Lysimeters

3.1.1. Biomass

The lettuce plants were grown for 73 days after planting and were cut at flowering. The average dry weights were ca. 45 g/plant, rather similar in all four treatments. The cumulative irrigation volume was 21 L/mini-lysimeter, and the average overall leaching fraction (LF) of the not-planted and planted lysimeters was 77% ± 5 and 17% ± 6, respectively. This was maintained by a near-daily adjustment of the irrigation volume to plant consumption. Although the uniform yield and crop water uptake among all treatments may indicate that the CFA did not adversely affect the plants, the root system at the highest CFA load (450 mT ha⁻¹) seemed retarded. The root/canopy dry weight ratio at this load was 0.6 while in the control and the two lower loads (90 and 270 mT CFA ha⁻¹) it was 1.3.

3.1.2. Leachate pH

Mixing CFA with the sand raised the pH of the leachate from ~7.8 in the not-amended sand to ca. 11–12 (at CaO equilibrium) (Figure 1). Irrigation of the planted lysimeters ultimately reduced the pH to that of the non-amended sand. Since most of the change in pH in the mini-lysimeters was relatively quick and started soon after the seedlings’ establishment (Figure 1), we assume that the main pH-neutralizing factor was CO₂ emitted from root and rhizosphere respiration and the equimolar release of H⁺ by CO₂ dissolution [49] and the resulting CaO carbonation. Organic acids that exude from roots [50] must have also contributed directly to the drop in pH. Hence, the pH of the leachate from the planted lysimeters decreased much faster than that of the leachate from the non-planted counterparts. Obviously, the excess CaO at the higher CFA loads required more acidity and thus a longer period of time to complete the pH neutralization. Also, as mentioned above, the roots at the highest CFA load were considerable smaller.

The investigated CFA consumed 1.2 mmole H⁺/g upon titration from pH 12 to 7. Complete pH neutralization occurred at all three planted CFA loads and hence the amounts of CO₂ consumed could be calculated and were found to be ca. 1656, 5184, and 9504 mg C/mini-lysimeter (i.e., per plant), respectively, at the 90, 270, and 450 mT CFA ha⁻¹ equivalent loads (or 115, 360 and 660 g CFA/lysimeter, respectively). The respective durations for complete neutralization were 28, 46, and 58 days (Figure 1).

The pH declined with time also in the non-planted, irrigated mini-lysimeters, albeit considerably slower, being completed only at the 90 mT CFA ha⁻¹ equivalent load after 73 days (Figure 1). The mean daily CO₂ release, that caused this pH decline due to the irrigation, was thus estimated to be 23 mg C/day/lysimeter. This suggests that the net average daily amounts of CO₂ released by the roots and their surrounding rhizosphere (assuming that organic root exudates were mineralized), in the planted lysimeters were ~36 mg C/plant/day in the first 28 days after planting, ~173 mg C/plant/day in the 28–46 days period, and ~337 mg C/plant/day in the 46–58 days period after planting. Similarly, the cumulative CO₂ emission from the lettuce field’s soil over 73 days of growth, as reported by Wang et al. [51], ranged between 33 and 45 g C/plant, 20–50% of which was attributed to the roots and rhizosphere (ibid). Hence, the mean daily emission ranged between 90 and 315 mg C/plant/day, in good agreement with the range of daily CO₂ emissions reported in the present study.

3.1.3. Heavy Metals and Metalloids in the Lettuce Foliage and Leachate

The concentrations of Cd and Pb in the leachate were below quantitation (

Table 3. Mini-lysimeter study: content of elements in the CFA, lettuce foliage, and leachates.

Treatment	As	Cd	Pb	Cr	Cu	Ni	Zn	B	Mn	Mo	Se	V
(a)Concentration in the CFA (mg kg−1)
CFA	10	0.7	70	165	75	95	85	200	400	8.5	3	160
(b)Concentration in lettuce leaves at harvest (mg kg−1) (1)
Sand	bdl b	0.17	0.96	6.6	2.8	2.9	91	25 c	75 a	0.16 d	0.12	1.1 b
90 T ha−1	0.34 a	0.21	1.05	18.2	3.2	5.4	64	53 bc	89 a	0.73 c	0.1	1.8 ab
270 T ha−1	0.52 a	0.31	1.18	11.7	3.4	4.1	69	80 b	60 b	1.23 b	0.21	1.9 ab
450 T ha−1	0.64 a	0.22	1.11	13.2	3.1	4.5	55	143 a	48 b	2.16 a	0.14	2.1 a
Prob > F (1)	<0.001	ns	ns	ns	ns	ns	Ns	<0.001	<0.0001	<0.0001	ns	<0.05
Market plants	0.2	0.19	0.13	4.2	9.6	0.71	48	45	68	0.37	bdl	0.19
Max allowed (2)	5	1	1.5
(c)Average (±std) concentration of all 10 periodical leachates-planted and not-planted mini-lysimeters (3)
	µg L−1							mg L−1	µg L−1
Sand–Planted	14 ± 2	bdl	bdl	bql	7 ± 3	9 ± 8	79 ± 25	0.4 ± 0.1	3 ± 3	3 ± 1	15 ± 6	4 ± 1
Sand–Not Pl.	10 ± 0.4	bdl	bdl	bql	2 ± 0.2	4 ± 0.0	30 ± 5	0.15 ± 0.01	1 ± 0.4	1 ± 0.3	10 ± 2.4	6 ± 0.2
90 T ha−1–P	29 ± 4	bdl	bdl	410 ± 90	28 ± 6	19 ± 4	65 ± 25	8.0 ± 1.9	2 ± 2	193 ± 40	16 ± 3	196 ± 17
90 T ha−1–NP	14 ± 2	bdl	bdl	45 ± 14	3 ± 1	4 ± 0	8 ± 1	1.26 ± 0.14	2 ± 1	19 ± 3	8 ± 3	129 ± 21
270 T ha−1–P	19 ± 4	bdl	bdl	470 ± 240	12 ± 3	13 ± 6	60 ± 23	7.3 ± 2.0	14 ± 10	351 ± 109	16 ± 5	62 ± 5
270 T ha−1–NP	10 ± 2	bdl	bdl	370 ± 60	7.4 ± 1.3	4.4 ± 0.8	30 ± 14	0.75 ± 0.09	0.3 ± 0.0	158 ± 36	8 ± 3	44 ± 2
450 T ha−1–P	28 ± 4	bdl	bdl	4000 ± 540	9 ± 1	115 ± 13	51 ± 21	10.7 ± 1.6	15 ± 19	1065 ± 34	26 ± 5	45 ± 12
450 T ha−1–NP	9 ± 1	bdl	bdl	86 ± 7	3.6 ± 0.3	3.5 ± 0.0	21 ± 8	2 ± 0.1	0.3 ± 0.1	99 ± 22	9 ± 2	23 ± 1
DL/QL
WHO [52]	10	3	10	50	2000	70	-	2.4	-	-	40	-

⁽¹⁾ Comparisons for all pairs used Tukey–Kramer HSD. Levels not connected by the same letter are significantly different. ⁽²⁾ https://www.gov.il/BlobFolder/policy/fcs-reg-01022007/he/files_databases_fcs_regulations_Reg_01022007.pdf (in Hebrew; last accessed 1 December 2025) ⁽³⁾ DL/QL—detection and quantitation levels in solutions and QL in plant material (µg L⁻¹): As 7.5/75/2100; Cd 0.3/3.0/80; Cr: 0.6/6/170; Pb 2.8/28/780; Se: 11/111/3110; V: 1.1/11/310.

). The concentrations in the leachate of most all the other elements positively correlated with the CFA load and was higher in the planted than in the not-planted mini-lysimeters. The leachate salinity has increased in the planted lysimeters because of the ~4.5 times lower average LF. Arsenic concentrations in the leachate from all the treatments, with/without CFA, w/wo plant, namely even from the not-amended quarzitic dune sand, was at or above the WHO [52] drinking water limit (10 µg L⁻¹). However, it was below the max concentration permitted for long-term irrigation [53]. Cr concentrations in the leachate were high in all the CFA treatments, and this was aggravated in the presence of a plant. This was probably due to the presence of some Cr as Cr(VI) [54]. Soil oxides and organic matter transform it to the benign Cr(III) form, a process hastened in of the OM-rich NVS. B concentrations in the leachates from the planted mini-lysimeters also exceeded the WHO limit, while the concentrations of B in the not-planted lysimeters were below this limit. At any rate, the concentrations that did occur in the leachates are rather common in the soil solution of irrigated soils of the arid and semi-arid regions of Israel [55].

The concentrations of Cd and Pb in the lettuce leaves were not affected by CFA addition at all the application rates (Table 3). Arsenic concentrations did significantly increase upon CFA application compared with the non-amended control, yet they were not significantly different in all the loads, being ca. 10 times lower than the maximum value allowed for leafy vegetables. Despite the high Cr concentration in the leachates, its concentrations in the lettuce foliage did not significantly differ among all the treatments, including the not amended control (Table 3). The leachate and foliage concentrations of the other oxyanions, B, Mo, and V (but not Se), did significantly increase in the foliage, positively responding to an increase in CFA load.

The concentrations in the leachate of the metallic elements, Cu, Ni, Mn, and Zn, were all below the WHO thresholds for drinking water, and with the exception of Mn, their foliage concentrations were not affected by the presence of CFA. Mn foliage concentration was significantly lower at the two higher amendment rates than in the control plants, perhaps due to its lower solubility at a pH above 8 [56,57].

3.2. Heavy Metal(loid)s in Lettuce Grown in 220-L Lysimeters

3.2.1. Effect of the Amendments on the Content of Elements in Soil, Plants, and Leachate

The calculated amounts of the thirteen elements that were added to the 15-cm soil layer during the three years of the study are shown in

Table 4. The 220 L lysimeter study: the overall elemental loading (a) and extractability (b), and the concentrations in the lettuce foliage (c) and in the leachate (d).

Treatment (1)	As	Cd	Pb	Cr	Cu	Ni	Zn	B	Mn	Mo	V	P	Fe
(a) Calculated amount of elements loaded at the 15 cm lysimeters’ soil layer over the 3 y period (mg kg−1) (2)
NVS-214 mT ha−1	1.5	0.06	3.4	9	4	5	16	37	23	0.5	7	406	1719
NVS-642 mT ha−1	4.5	0.17	10	26	12	14	49	112	68	1.4	22	1217	5157
ADS-24 mT ha−1	0.02	0.06	3.6	17	61	12	379	0	27	1.1	2.4	2480	692
ADSC-72 mT ha−1	0.08	0.04	4.4	13	26	7	132	2	22	0.3	2.6	1465	665
(b) Concentration of elements (mg kg−1) in the DTPA-TEA extract of the 0–15 cm soil layer: a 3 y average
Control	0.03	0.02	0.3	0.01	1.1	0.25	1.7	0.7	11	bdl	0.6	9	9
NVS-214 mT ha−1	0.03	0.02	0.3	0.04	1.6	0.31	3.9	3.0	10	bdl	0.9	23	18
NVS-642 mT ha−1	0.35	0.03	0.6	0.10	3.5	0.87	7.2	10.3	22	bdl	2.0	22	37
ADS-24 mT ha−1	0.02	0.03	0.1	0.02	3.3	0.32	13.7	0.2	11	bdl	0.1	17	10
ADSC-72 mT ha−1	0.02	0.04	0.4	0.01	2.9	0.40	15.2	0.8	15	bdl	0.5	21	16
QL (3)	0.02	0.006	0.06	0.012	0.04	0.022	0.004	0.02	0.002	0.024	0.02	0.08	0.01
(c) Concentrations in the lettuce foliage (mg kg−1)-average of the plants from all the four seasons (5)
Commercial (4)	0.20	0.19 b	0.14	0.40	9.6 ab	0.71 b	48 bc	45 b	68 c	0.37 bc	0.19	7200 a	136
Control	0.16	0.43 ab	0.17	1.02	5.7 c	0.97 ab	32 c	48 b	149 b	0.47 c	0.43	4485 b	174
NVS-214 mT ha−1	0.32	0.27 b	0.19	0.86	6.6 bc	0.76 b	37 c	59 a	79 c	0.78 ab	0.39	4400 b	158
NVS-642 mT ha−1	0.28	0.41 ab	0.20	1.15	9.0 a	0.90 ab	44 bc	51 ab	49 c	0.67 abc	0.40	5100 ab	168
ADS-24 mT ha−1	0.13	0.77 a	0.16	0.62	5.6 bc	1.39 a	84 a	49 ab	354 a	0.89 a	0.48	6201 a	176
ADSC-72 mT ha−1	0.16	0.36 b	0.13	0.91	5.8 bc	0.77 b	50 b	47 b	127 bc	0.53 bc	0.36	5500 a	151
Prob > F	bql	<0.0001	bql	ns	<0.0001	0.01	<0.0001	<0.0001	<0.0001	0.0002	ns	<0.0001	ns
(d) Highest discrete concentration measured in the leachates from the lysimeters throughout the three years
	µg L−1							mg L−1	µg L−1
Control	bdl	1	bdl	2	11	70	37	2.1	3	5	17	0.13	bdl
NVS-214 mT ha−1	bdl	1	bdl	bdl	18	71	58	4.6	4	8	8	0.08	bdl
NVS-642 mT ha−1	bdl	2	bdl	bdl	18	77	44	2.0	6	4	11	0.09	bdl
ADS-24 mT ha−1	bdl	1	bdl	6	15	58	28	1.8	6	12	16	0.96	bdl
ADSC-72 mT ha−1	bdl	1	bdl	bdl	10	53	33	2.0	2	7	8	0.13	bdl
WHO [51]	10	3	10	50	2000	70	-	2.4	80	-	-	-	-

⁽¹⁾ Data in each treatment represents all the receiving lysimeters, which consisted of up to 3 soils and 2 water types (4 to 16 lysimeters per treatment—see Table 5); ⁽²⁾ Based on data in Table 1, multiplication by 50 returns mg/lysimeter; ⁽³⁾ QL—quantitation level, bdl/bql—below detection/quantitation level; ⁽⁴⁾ Lettuce plants purchased at the market; ⁽⁵⁾ For explanation of statistical analysis see Table 3 comment #1.

a. The NVS enriched the soils with practically all elements tested by far more than did the ADS or ADSC. This was because the application was made on the basis of adding a quantity equal to 500 kg of total N ha⁻¹ addition. Hence, the amounts of As and B added with the NVS were ca. 100 folds larger than those with the other biosolids (Table 4a), and the difference in the added amount of the other elements was also rather substantial. The only exceptions were Zn and Cd which was a result of their higher concentrations in the other additives (Table 1). Note that the amounts of metals added are given in Table 4a in mg per lysimeter. Division by 50 (the weight of the 15-cm soil layer) is required to obtain the mg kg⁻¹ value.

The amounts of the elements extracted by the DTPA-TEA reagent were averaged for each of the additives over the 3-year period and all the participating soils (see

Table 5. The average concentrations of As, Cd, and Pb (µg kg⁻¹ dry matter) ⁽¹⁾ in the foliage of the four lettuce crops from the 3 y, 220 L lysimeter study: as affected by types of soil, amendment, and irrigation water.

Treatment and Load (mT h−1)	As (QL (2) = 700 µg kg−1)				Cd (QL = 40 µg kg−1)				Pb (QL = 300 µg kg−1)
	1st	2nd	3rd	4th	1st	2nd	3rd	4th	1st	2nd	3rd	4th
Quartz sand (tap water)
Control	80	90	10	10	240 b	350 bcd	110 c	170 bcd	230	150	170	8
NVS-214	90	190	70	40	170 b	200 cd	130 c	170 bcd	220	130	300	bdl
ADSC-72	70	140	30	90	120 b	210 cd	120 c	210 bcd	170	bdl	200	77
Quartz sand (WWE)
Control	150	150	160	40	600 ab	640 abc	1430 a	440 b	170	30	350	bdl
NVS-214	150	150	200	60	370 ab	370 bcd	410 bc	190 bcd	130	90	360	31
ADSC-72	80	130	230	80	310 ab	780 ab	580 bc	420 bcd	bdl	bdl	230	46
Calcic Haploxeralf (WWE)
Control	40	20	140	20	100 b	70 d	60 d	100 cd	30	80	370	15
NVS-214	70	150	140	70	100 b	110 d	70 d	60 d	160	130	360	9
Non-sodic Chromic Haploxerert (WWE)
Control	130	130	200	50	320 ab	200 cd	190 c	190 bcd	410	290	480	13
NVS-214	100	220	170	90	260 b	270 cd	200 c	180 bcd	340	230	450	33
NVS-642	100	130	190	60	160 b	270 cd	170 c	180 bcd	90	320	370	14
ADS-24	180	130	90	40	810 a	980 a	930 ab	940 a	350	80	210	15
ADSC-72	50	90	70	70	220 b	240 cd	160 c	360 bcd	190	310	230	75
Prob > F (3)	-	-	-	-	<0.001	<0.0001	<0.0001	<0.0001	ns	ns	ns	ns

⁽¹⁾ Upper limits for leafy vegetables of As, Cd, and Pb are 5, 1, and 1.5 mg kg⁻¹ (Table 3); ⁽²⁾ QL, the quantitation level. Detection level equals to QL/30; ⁽³⁾ Explanation of the statistical analysis see Table 3 comment #1. ns—not significantly different.

the treatments). Notably, concentrations in the NVS treatment were not significantly different from those in the other treatments, including the control (Table 4b). This includes the elements (Fe, Mn, Zn, and Cu) whose “potential phytoavailability” the DTPA extraction was developed to assess [47]. In fact, the procedure is widely used for risk assessment of other elements too (e.g., [58]). Albeit the DTPA-TEA extraction was not intended to determine the solubility of oxyanions (as they do not complex with DTPA), they do solubilize in the TEA-buffered solution of pH 7.3. We included P, Mn, and Fe in this assessment because of their role in plant nutrition and their known problematic phytoavailability at higher soil pH values, such as are under CFA and NVS application.

The concentration of the elements in the lettuce foliage is presented in Table 4c. As above (Table 4b), the data were averaged for each additive over the four growing seasons and participating soils. In addition, lettuce purchased at four different markets was also analyzed for an independent comparison. The concentrations of As and Pb were below the quantitation levels, and values given are the estimated highest concentrations measured and these were ca. 6% and 20% of the maximum permissible values. Cadmium concentration was significantly higher in the ADS treatment and without significant difference between its concentration in the other treatments, including the non-amended control and the commercial plants. Likewise, the foliage concentrations of Cd, Ni, Zn, Mn, P, and Mo were highest in the ADS-treated plants, and by and large the same in the other treatments. B was the only element that displayed a significantly higher concentration in an NVS treatment (the lower dose) as compared with the other treatments. Despite the differences in the amount of the elements added (70 and 200 times more) with the ADS and ADSC. Yet, the difference between the control and the NVS treatment was small (47 and 59 mg kg⁻¹, respectively). The foliage concentrations of Cr, V, and Fe were not significantly different between all the treatments, despite the significant differences in the amounts applied.

It is important to note that a very low correlation was observed between the “potential phytoavailability” of the elements (DTPA-TEA extractability) and the elemental content in the lettuce foliage. In fact, the same was true also with regard to the water extraction of these soils. Furthermore, the correlation between the amounts extracted by the two extractants, DTPA, and water, was also very low.

The highest single concentration measured in the leachates from the lysimeters at each of the treatments throughout the three years of the experiment is presented in Table 4d. The concentrations of As, Pb, Fe, and usually also of Cr were below the detection limits. The concentrations of the other elements were usually very low, although, the concentrations of Cu, Zn, and B were relatively high. Cu concentrations were well below the WHO limit for drinking water, B bordered or exceeded its upper limit (while Zn lacks a WHO limit). Ni peak concentration bordered or slightly exceeded its maximum allowed value, yet this concentration was the same in the NVS treatments and the not amended control. Ni, Zn, and Cu tend to become mobilized by DOC [17,59,60,61]. The concentration of all three metals was below the levels acceptable for long-term unrestricted irrigation [53]. The maximum B concentrations measured in the leachate were rather high in all the treatments, bordering on the WHO drinking water limit. Note that the high NVS load was tested only in the clayey soil while the low load was tested in all the four soils. This can explain the lower B concentration in the leachate from the higher load. Note also that the WWE used for irrigation contained B (Table 1).

An experimental setup very similar to ours was studied by Richards et al. [19] and Kim et al. [22]. This long-term column study examined the fate of biosolids-derived Cu, Zn, and Mo. Their ~17 L columns were packed with intact cores of two loamy soils (of pH 5 and 7) and loaded, at one time, with NVS, ADS, and ADSC at 663, 215, and 249 mT ha⁻¹, respectively, based on an equal organic matter load. The amounts of Cu, Zn, and Mo applied were, respectively, 7.7, 6.7, and 1.5 times larger than those applied to our 220 L lysimeters. The biosolids in these 17 L columns were mixed with the upper 10 cm soil layer, and the 20 cm soil underneath was left intact. Although the water-extractable fractions of Cu, Zn, and Mo in the NVS were rather high (45%, 20%, and 12% of the total metal content, respectively [18]), their content in the leachate was within acceptable levels, with Mo concentrations peaking at 0.03 mg L⁻¹ in the third (of the successive eight) cropping cycle. The authors suggested that the intact bottom soil layer intercepted these and other heavy metal(loid)s that leached from the applied upper soil layer. To conclude, despite the differences between our 220 L lysimeter study and that of Richards et al. [19] and Kim et al. [22], the similarity in the results is remarkable.

3.2.2. As, Cd, and Pb Contents in the Foliage of the Four Lettuce Crops

A detailed presentation, by soil and treatment, of the concentrations of the elements of concern, As, Cd, and Pb, in the lettuce foliage throughout the three years and four growing cycles is given in Table 5. The yearly and cumulative heavy loads notwithstanding, the content of As and Pb in the foliage were either below or on the threshold of quantitation, and well below the upper levels permissible for human consumption of leafy vegetables (Table 3, comment #2). However, Cd concentrations in the lettuce grown in the Sand-ADS-WWE combination bordered on that upper level (1 mg kg⁻¹) in each of the four seasons (Table 5). The concentration of Cd was also relatively high in the foliage from the first two seasons of the Sand–Control–WWE treatment, and in the third season, it was 1.4 mg kg⁻¹, exceeding the upper limit (Table 5). The Cd source was not the WWE where it was below detection (Table 1). Hence, it seems that DOC contained in the WWE complexed Cd that was already present in the growing medium and facilitated its phytoavailability [17,43].

The concentration of Cd in the lettuce foliage at the fourth season (which was a summer) was similar or lower than in the former three seasons. The phenomenon of low foliage concentration throughout four growing cycles and the further reduction in the fourth, was common to all the three soils used, and it occurred despite the three repeated applications of the additive. Thus, the cumulative NVS load in the fourth season was 170 and 510 mT ha⁻¹ at the lower and higher loads, respectively. This observation suggests that the risk of enhanced phytoavailability and leaching of heavy metal(loid)s following repeated NVS application to soils is negligible.

3.3. Heavy Metal(loid)s in Field Crops

A large amount of data on heavy metal(loid) concentrations in a wide selection of crops was compiled from field experiments (Table 6). The crops, site locations, soils, loads of NVS (or CFA), and yields are presented. Wherever a range of yields is provided, these are the average yields of the treatments with the lowest and highest yields. They refer either to a single experiment or to all the experiments in which a specific crop was used. The concentrations of the elements are displayed in Table 6 as follows: the average concentrations in the NVS treatment(s) is placed in front of parentheses, which hold the range of values (minimum and maximum) that were measured in both the amended and not-amended treatments. In one case (Table 6a; potatoes at the Bsor R&D Center), a single average or two averages are given for each element. In the first case, the treatments did not statistically differ, and in the other case, one average is of the non-amended control and the other is of the NVS treatment.

The concentrations of As, Cd, and Pb in all the crops tested, edible and non-edible alike (Table 6a,b), were well below the upper threshold allowed for human consumption. Often, they were below their quantitation levels. This is due to both their low contents in the NVS and CFA and due to their limited uptake and transport within the plant from root to shoot and to their redistribution via the phloem [62,63,64]. The only exception was the relatively high Pb content in silage wheat (average 1.4 mg kg⁻¹, peaking at 2.8 mg kg⁻¹), following the application of 300 mT CFA ha⁻¹ to the Bnei-Darom dune sand (Table 6b).

Mo uptake was the most affected of all other elements examined. Its concentration in almost all the crops tested increased both under CFA and NVS application, regardless of the soils concerned (Table 6). The crops most affected were the legumes, fodder, and edibles. Enhanced Mo uptake was probably the reason for the more than 3-fold increase in the yield of the legumes upon NVS application (Table 6b). The yield of the fodder legumes, vetch, and clover mixture, responded strongly and positively to the application of NVS (Table 6b). The corn crop that was sawn on a Sodic Chromoxerert soon after the application of 800 mT CFA ha⁻¹, displayed a significant increase vs. the control of Mo (1.9 vs. 0.9 mg kg⁻¹) in the whole plant (canopy + ears) and of Cr in the seeds (4.1 vs. 2.1 mg kg⁻¹) (Table 6b). The whole plant average Cr concentration was ~11 mg kg⁻¹, both with CFA and without.

Legumes, in particular, strongly respond to Mo application [65,66,67], Mo being an essential element with which NVS-treatment significantly enriched the soil (e.g., Table 4). The yield of another mixed-legumes crop (data not presented) on the same soil (Lithic Xerorthent), in a nearby field, still strongly significantly responded to a 50 mT ha⁻¹ NVS load even six years after application. This shallow, calcareous Lithic Xerorthent must have lost much of its essential macro- and micro-nutrients by leaching, runoff, and erosion. Manure (NVS or other) applications could effectively replenish these nutrients and contribute significantly to the legumes’ growth and yield. Note that the Mo/Cu ratio of these legumes was too low to induce hypocuprosis in livestock [65,67].

Untypically, one potato crop (at the Bsor R&D Center) out of the many tested, showed a remarkable statistically significant enhancing effect of NVS application (45 mT ha⁻¹) compared to the control (Table 6a). Affected were the weight per tuber (272 vs. 212 g/tuber), the number of tubers (77 vs. 68/m²⁻), and the overall yield (65 vs. 55 mT ha⁻¹). The concentration of several elements in the tubers (Mo, Se, Cr, Cu, and even P) were significantly higher in the NVS treatments than in the control plots (Table 6a). This was despite of the fact that the field was profusely fertilized with N, P, and K. This desert sandy soil (a Xeric Torripsamment; Table 2) was not cultivated for several years before and probably became depleted of essential trace elements. These elements are not regularly provided to soils unless manure is applied. Interestingly, the heavy metal(loid)s contents of the potato crops presented in Table 6 fits to a large extent with the data provided by Dramićanin et al. [68] gathered from a wide assemblage of cultivation systems, varieties, genotypes, harvesting time and environmental conditions. Note that Mo, Cu and Se content significantly increased also in the CFA-treated chickpeas (Table 6a). However, their Zn content significantly decreased as compared to the control, and it also decreased in the feed legumes under NVS application as compared to the control (Table 6b), which thus seems to be a phenomenon related to legumes in general. Zn content decreased significantly in the NVS-treated lettuce. Such decreases did not occur in crops other than lettuce (Mn) and legumes (Zn). In fact, Mn content significantly increased in the corn canopy from the Revadim long-term study under 426 mT NVS ha⁻¹ four-year cumulative load (Table 6b). The Zn and Mn values are within the levels of micro-nutrients required in dairy cattle feed [69]. In a four-year field study, Sajwan et al. [27] applied CFA at 280–1120 mT ha⁻¹ to a sandy soil, all in one application. Interestingly, the contents of Mn and Zn have significantly decreased, and Mn content inversely related to the CFA load.

Table 6. (a): Average concentrations of heavy metal(loid)s in field trial with edible crops. (b): Average concentrations of heavy metal(loid)s in field trial with nonedible crops. In parentheses are the minimum and maximum values pertinent to a crop, and the value in front is the average of all the relevant NVS treatments.

(a)
Crop	Potatoes (var. Vivaldi, Winston)	Potatoes (1) (var. Winston)	Carrots (var. Nairobi)	Lettuce (var. Iceberg)	Bell pepper (var. Gilad)	Peanuts (var. Hanoch; Seeds)	Chickpeas (var. Kabuli; Seeds)
Location and soil classification	Bsor, Nir-Oz, Yevul (Torripsamments) Nir-Eliyahu (Haploxeralfs)	Bsor R&D Center (Torripsa-mments)	Nir-Eliyahu (Haploxeralfs)	Ein-Habsor, Sde- Nitsan, Bet-Ezra, Kfer-Hayim (Torripsamments, Haploxerepts and Haploxeralfs)	Tomer (Haplargids)	Nir-Oz, Nirim (Torripsamments)	Revadim (a Sodic Haploxerert)
NVS load (dry mT ha−1)	50–120	45	120	50–120	100	50	CFA 800
Yield (fresh mT ha−1)	37–58	(2) 55 vs. 65	9.1	30–50	92	6.5–8.1 (pods)	4.2 (pods)
As (µg kg−1)	9 (bdl-45)	Bdl	40 (20–60)	10 (bdl-80)	bdl	210 (bdl-780)	70 (bdl-0.16)
Cd (µg kg−1)	35 (25–60)	20	50 (30–130)	160 (60–560)	60 (bdl-100)	40 (bdl-100)	2 (bdl-20)
Pb (µg kg−1)	130 (80–370)	140	Bdl	bdl-1.3	bdl	60 (bdl-700)	bdl
B (mg kg−1)	27 (7–57)	36	70 (45–100)	60 (37–102)	40 (10–90)	16 (12–21)	14 (11–17)
Cr (mg kg−1)	0.17 (0.06–0.37)	0.12	1.7 (0.7–4.5)	0.44 (0.19–1.0)	0.5 (0.2–1.8)	0.08 (bdl-0.31)	0.56 (0.25–0.27)
Cu (mg kg−1)	5 (4.3–6.0)	(2) 3.8 vs 4.5	5.4 (3–10)	14 (4–104)	10 (7–14)	8 (5–10)	(2) 6 (5–8)
Fe (mg kg−1)	16 (7–25)	25	160 (70–330)	97 (55–140)	45 (26–134)	17 (13–41)	48 (42–58)
Mn (mg kg−1)	6.5 (5.3–7.0)	7.4	9 (6–20)	(3) 21 (14–63)	13 (11–14)	15 (11–20)	26 (22–30)
Mo (mg kg−1)	0.3 (0.16–0.50)	(2) 0.11 vs. 0.90	(2) 0.7 (0.15–1.7)	(2) 0.3 (0.12–0.42)	0.7 (0.4–1.5)	5.6 (2.2–10.1)	(2) 9 (4.4–14)
Ni (mg kg−1)	0.29 (0.17–1.0)	0.31	1.2 (0.49–2.8)	1.0 (0.32–5.2)	1.1 (0.7–1.6)	0.52 (0.25–1.12)	1.0 (0.8–1.5)
P (g kg−1)	0.238 (0.15–0.4)	(2) 0.14 vs 0.18	0.44 (0.31–0.64)	0.83 (0.60–1.14)	0.29 (0.21–0.40)	0.42 (0.30–0.48)	0.41 (0.39–0.43)
Se (mg kg−1)	-	(2) 0.009 vs 0.56	-	-	0.74 (bdl-1.4)	-	(2) 0.04 (bdl-1.4)
V (mg kg−1)	0.01 (bdl-0.03)	0.10	0.53 (0.2–1.1)	0.14 (0.8–0.28)	0.08 (bdl-0.3)	0.03 (bdl-0.13)	0.009 (bdl-0.03)
Zn (mg kg−1)	25 (13–52)	19	19 (12–30)	54 (35–80)	20 (13–24)	43 (27–50)	(3) 19 (16–23)
(b)
Crop	Corn Canopy (+Cobs and Husks)	Corn Kernels	Wheat Canopy	Wheat Grains	Silage Wheat (at Wax Ripening)	Vetch and Clover	Corn (4)
Location and soil classification	Revadim (a Chromic Haploxerert)				Bnei-Darom (dune sand)	Mishmar David (a Lithic Xerorthent)	Revadim (a Sodic Haploxerert)
Load (dry mT ha−1)	284 and 426 mT NVS ha−1, the 4-y cumulative loads at the low and high rates; double cropping: corn in summers, wheat in winters.				CFA: 300 and NVS: 100	NVS: 95	CFA: 800
Yield (mT ha−1)	34.1 (moist)	6.8 (dry)	29.3 mT (moist)	3.1 (dry)	10.6 (moist)	(2) 5.0 vs. 14 (dry)	44 (total, moist)
As (µg kg−1)	970 (130–7000)	bdl	43 (bdl-160)	bdl	60 (bdl-170)	90 (10–250)	bdl
Cd (µg kg−1)	22 (bdl-240)	30 (bdl-620)	30 (bdl-70)	7 (0–130)	32 (15–43)	50 (30–100)	8; 2
Pb (µg kg−1)	160 (bdl-430)	bdl	160 (bdl-700)	60 (0–750)	(2) 1400 (700–2800)	200 (30–360)	120; 40
B (mg kg−1)	(2) 55 (27–127)	23 (5–54)	17 (3–37)	13 (3–27)	39 (17–63)	46 (10–82)	24; 7
Cr (mg kg−1)	1.5 (bdl-6.0)	0.17 (0.06–0.40)	0.5 (0.3–1.3)	0.41 (0.11–9.0)	0.60 (0.33–1.2)	0.9 (0.5–2.7)	6; (2) 3
Cu (mg kg−1)	(2) 5.5 (2.4–9.6)	1.4 (0.7–2.3)	2.7 (1.4–4.0)	5 (4–11)	3.4 (2.6–4.7)	8 (6–10)	7.3; 2.6
Fe (mg kg−1)	135 (56–415)	12 (7–22)	134 (76–270)	31 (19–63)	188 (72–485)	220 (117–440)	150; 40
Mn (mg kg−1)	(2) 51 (36–82)	4 (3–5)	68 (48–103)	69 (58–82)	29 (17–62)	35 (16–54)	29; 8
Mo (mg kg−1)	(2) 0.07 (bdl-0.50)	0.15 (0.05–0.35)	(2) 2.7 (1.5–4.0)	(2) 0.62 (0.3–2.8)	(2) 1.6 (0.4–2.3)	(2) 2.2 (0.4–5)	(2) 0.9; 0.6
Ni (mg kg−1)	0.20 (bdl-2.3)	0.28 (0.13–0.64)	0.21 (0.12–0.47)	0.52 (0.3–2.5)	0.34 (0.15–0.68)	1.1 (0.7–1.9)	2.4; 1.4
P (g kg−1)	0.14 (0.04–0.35)	0.28 (0.23–0.32)	0.15 (0.08–0.21)	0.38 (0.34–0.43)	0.24 (0.21–0.27)	0.29 (0.23–0.39)	0.26; 0.40
Se (mg kg−1)	-	-	-	-	(2) 0.53 (bdl-0.72)	-	-
V (mg kg−1)	0.39 (0.19–0.76)	0.01 (bdl-0.03)	0.35 (0.19–0.72)	0.06 (0.01–0.21)	0.49 (0.18–1.3)	0.56 (0.25–1.3)	0.22; 0.012
Zn (mg kg−1)	28 (13–61)	21 (17–27)	11 (4.5–22)	33 (24–57)	23 (14–29)	(3) 41 (39–56)	11; 19

⁽¹⁾ One average represents the control and NVS treatments if not significantly different. If they are different, a separate average is provided for each; ⁽²⁾ Elements whose average concentration in the NVS (or CFA) treatment was statistically significantly larger than in the not-amended control; ⁽³⁾ Elements whose average concentration in the not-amended control was statistically significantly larger in than in the NVS treatment. ⁽⁴⁾ The control and CFA treatments were averaged together. The first datum refers to the corn cobs and husks, and the second to the corn kernels; With the exceptions delineated above, it is evident that the uptake of the cationic trace elements (Mn, Zn, Cu, Fe, and Ni) was not significantly affected by the application of CFA and NVS (Table 6). This could be due to the quick return of the soil solution pH of NVS-treated soils to the value of the untreated soil or even to a lower value (Figure 1 and [25]). However, the abundant DOC in the NVS (and the further OC solubilization at the NVS pH ≥ 12) and in and the other two biosolid types, could effectively complex trace elements, rendering them soluble at the pH of carbonate equilibrium [70]. This was perhaps the reason for Kirchmann et al. [71]’s observation that significantly more Cu and Fe accumulated in wheat grains under manure application than under NPK fertilization. Despite of that, the phytoavailability of the above-mentioned elements in the current study was not affected.

Table 6. (a): Average concentrations of heavy metal(loid)s in field trial with edible crops. (b): Average concentrations of heavy metal(loid)s in field trial with nonedible crops. In parentheses are the minimum and maximum values pertinent to a crop, and the value in front is the average of all the relevant NVS treatments.

(a)
Crop	Potatoes (var. Vivaldi, Winston)	Potatoes (1) (var. Winston)	Carrots (var. Nairobi)	Lettuce (var. Iceberg)	Bell pepper (var. Gilad)	Peanuts (var. Hanoch; Seeds)	Chickpeas (var. Kabuli; Seeds)
Location and soil classification	Bsor, Nir-Oz, Yevul (Torripsamments) Nir-Eliyahu (Haploxeralfs)	Bsor R&D Center (Torripsa-mments)	Nir-Eliyahu (Haploxeralfs)	Ein-Habsor, Sde- Nitsan, Bet-Ezra, Kfer-Hayim (Torripsamments, Haploxerepts and Haploxeralfs)	Tomer (Haplargids)	Nir-Oz, Nirim (Torripsamments)	Revadim (a Sodic Haploxerert)
NVS load (dry mT ha−1)	50–120	45	120	50–120	100	50	CFA 800
Yield (fresh mT ha−1)	37–58	(2) 55 vs. 65	9.1	30–50	92	6.5–8.1 (pods)	4.2 (pods)
As (µg kg−1)	9 (bdl-45)	Bdl	40 (20–60)	10 (bdl-80)	bdl	210 (bdl-780)	70 (bdl-0.16)
Cd (µg kg−1)	35 (25–60)	20	50 (30–130)	160 (60–560)	60 (bdl-100)	40 (bdl-100)	2 (bdl-20)
Pb (µg kg−1)	130 (80–370)	140	Bdl	bdl-1.3	bdl	60 (bdl-700)	bdl
B (mg kg−1)	27 (7–57)	36	70 (45–100)	60 (37–102)	40 (10–90)	16 (12–21)	14 (11–17)
Cr (mg kg−1)	0.17 (0.06–0.37)	0.12	1.7 (0.7–4.5)	0.44 (0.19–1.0)	0.5 (0.2–1.8)	0.08 (bdl-0.31)	0.56 (0.25–0.27)
Cu (mg kg−1)	5 (4.3–6.0)	(2) 3.8 vs 4.5	5.4 (3–10)	14 (4–104)	10 (7–14)	8 (5–10)	(2) 6 (5–8)
Fe (mg kg−1)	16 (7–25)	25	160 (70–330)	97 (55–140)	45 (26–134)	17 (13–41)	48 (42–58)
Mn (mg kg−1)	6.5 (5.3–7.0)	7.4	9 (6–20)	(3) 21 (14–63)	13 (11–14)	15 (11–20)	26 (22–30)
Mo (mg kg−1)	0.3 (0.16–0.50)	(2) 0.11 vs. 0.90	(2) 0.7 (0.15–1.7)	(2) 0.3 (0.12–0.42)	0.7 (0.4–1.5)	5.6 (2.2–10.1)	(2) 9 (4.4–14)
Ni (mg kg−1)	0.29 (0.17–1.0)	0.31	1.2 (0.49–2.8)	1.0 (0.32–5.2)	1.1 (0.7–1.6)	0.52 (0.25–1.12)	1.0 (0.8–1.5)
P (g kg−1)	0.238 (0.15–0.4)	(2) 0.14 vs 0.18	0.44 (0.31–0.64)	0.83 (0.60–1.14)	0.29 (0.21–0.40)	0.42 (0.30–0.48)	0.41 (0.39–0.43)
Se (mg kg−1)	-	(2) 0.009 vs 0.56	-	-	0.74 (bdl-1.4)	-	(2) 0.04 (bdl-1.4)
V (mg kg−1)	0.01 (bdl-0.03)	0.10	0.53 (0.2–1.1)	0.14 (0.8–0.28)	0.08 (bdl-0.3)	0.03 (bdl-0.13)	0.009 (bdl-0.03)
Zn (mg kg−1)	25 (13–52)	19	19 (12–30)	54 (35–80)	20 (13–24)	43 (27–50)	(3) 19 (16–23)
(b)
Crop	Corn Canopy (+Cobs and Husks)	Corn Kernels	Wheat Canopy	Wheat Grains	Silage Wheat (at Wax Ripening)	Vetch and Clover	Corn (4)
Location and soil classification	Revadim (a Chromic Haploxerert)				Bnei-Darom (dune sand)	Mishmar David (a Lithic Xerorthent)	Revadim (a Sodic Haploxerert)
Load (dry mT ha−1)	284 and 426 mT NVS ha−1, the 4-y cumulative loads at the low and high rates; double cropping: corn in summers, wheat in winters.				CFA: 300 and NVS: 100	NVS: 95	CFA: 800
Yield (mT ha−1)	34.1 (moist)	6.8 (dry)	29.3 mT (moist)	3.1 (dry)	10.6 (moist)	(2) 5.0 vs. 14 (dry)	44 (total, moist)
As (µg kg−1)	970 (130–7000)	bdl	43 (bdl-160)	bdl	60 (bdl-170)	90 (10–250)	bdl
Cd (µg kg−1)	22 (bdl-240)	30 (bdl-620)	30 (bdl-70)	7 (0–130)	32 (15–43)	50 (30–100)	8; 2
Pb (µg kg−1)	160 (bdl-430)	bdl	160 (bdl-700)	60 (0–750)	(2) 1400 (700–2800)	200 (30–360)	120; 40
B (mg kg−1)	(2) 55 (27–127)	23 (5–54)	17 (3–37)	13 (3–27)	39 (17–63)	46 (10–82)	24; 7
Cr (mg kg−1)	1.5 (bdl-6.0)	0.17 (0.06–0.40)	0.5 (0.3–1.3)	0.41 (0.11–9.0)	0.60 (0.33–1.2)	0.9 (0.5–2.7)	6; (2) 3
Cu (mg kg−1)	(2) 5.5 (2.4–9.6)	1.4 (0.7–2.3)	2.7 (1.4–4.0)	5 (4–11)	3.4 (2.6–4.7)	8 (6–10)	7.3; 2.6
Fe (mg kg−1)	135 (56–415)	12 (7–22)	134 (76–270)	31 (19–63)	188 (72–485)	220 (117–440)	150; 40
Mn (mg kg−1)	(2) 51 (36–82)	4 (3–5)	68 (48–103)	69 (58–82)	29 (17–62)	35 (16–54)	29; 8
Mo (mg kg−1)	(2) 0.07 (bdl-0.50)	0.15 (0.05–0.35)	(2) 2.7 (1.5–4.0)	(2) 0.62 (0.3–2.8)	(2) 1.6 (0.4–2.3)	(2) 2.2 (0.4–5)	(2) 0.9; 0.6
Ni (mg kg−1)	0.20 (bdl-2.3)	0.28 (0.13–0.64)	0.21 (0.12–0.47)	0.52 (0.3–2.5)	0.34 (0.15–0.68)	1.1 (0.7–1.9)	2.4; 1.4
P (g kg−1)	0.14 (0.04–0.35)	0.28 (0.23–0.32)	0.15 (0.08–0.21)	0.38 (0.34–0.43)	0.24 (0.21–0.27)	0.29 (0.23–0.39)	0.26; 0.40
Se (mg kg−1)	-	-	-	-	(2) 0.53 (bdl-0.72)	-	-
V (mg kg−1)	0.39 (0.19–0.76)	0.01 (bdl-0.03)	0.35 (0.19–0.72)	0.06 (0.01–0.21)	0.49 (0.18–1.3)	0.56 (0.25–1.3)	0.22; 0.012
Zn (mg kg−1)	28 (13–61)	21 (17–27)	11 (4.5–22)	33 (24–57)	23 (14–29)	(3) 41 (39–56)	11; 19

⁽¹⁾ One average represents the control and NVS treatments if not significantly different. If they are different, a separate average is provided for each; ⁽²⁾ Elements whose average concentration in the NVS (or CFA) treatment was statistically significantly larger than in the not-amended control; ⁽³⁾ Elements whose average concentration in the not-amended control was statistically significantly larger in than in the NVS treatment. ⁽⁴⁾ The control and CFA treatments were averaged together. The first datum refers to the corn cobs and husks, and the second to the corn kernels; With the exceptions delineated above, it is evident that the uptake of the cationic trace elements (Mn, Zn, Cu, Fe, and Ni) was not significantly affected by the application of CFA and NVS (Table 6). This could be due to the quick return of the soil solution pH of NVS-treated soils to the value of the untreated soil or even to a lower value (Figure 1 and [25]). However, the abundant DOC in the NVS (and the further OC solubilization at the NVS pH ≥ 12) and in and the other two biosolid types, could effectively complex trace elements, rendering them soluble at the pH of carbonate equilibrium [70]. This was perhaps the reason for Kirchmann et al. [71]’s observation that significantly more Cu and Fe accumulated in wheat grains under manure application than under NPK fertilization. Despite of that, the phytoavailability of the above-mentioned elements in the current study was not affected.

Likewise, Fe content, as well as that of P and B in all the field crops treated with CFA and NVS, was not affected compared with their respective controls (Table 6). The single exception was a higher P content (on top of a profuse P application) in the one potato crop at the Besor R&D farm (Table 6a). B content did increase in the lettuce foliage grown in sand-CFA mixtures in the mini-lysimeters (Table 3) but it was not affected in the 220-lysimeters loaded with NVS (Table 4), or in the field studies in either CFA- or NVS-treated soils. Seemingly, the quick decline in the pH after the NVS application rendered P more available while enhancing B adsorption. Furthermore, unlike CFA, NVS contains a large biosolids component (23% w/w dry) that may bind B, on the one hand [55], and release plant essential micro-elements and P, on the other hand [57].

3.4. Heavy Metal(loid)s Content and Potential Phytoavailability in the Treated Soils

3.4.1. Heavy Metal(loid)s Content

The application of biosolids is expected to affect the content of the investigated elements in the applied soils depending on their concentration in the biosolids relative to that in the receiving soils, and on the loading rates. The overall elemental content in the 0–15 cm layer of the Revadim fine-clayey Chromic Haploxerert was determined at the end of the third year of this field study. The cumulative amounts of NVS, ADS, and ADSC applied by the third year, were 214, 24, and 72 mT ha⁻¹.

The NVS application significantly increased the contents of As, B, and V compared to the other treatments (

Table 7. Concentration of elements (mg kg⁻¹) in the upper 20 cm layer of the fine-clayey soil at the end of the third year of the corn–wheat rotation field experiment at Revadim ⁽¹⁾.

Element	Control	NVS	ADS	ADSC	Prob > F	Max (2)
As	1.0 b	1.6 a	1.1 b	1.1 b	<0.0001	20
Cd	0.20 b	0.22 ab	0.21 ab	0.22 a	<0.05	2
Pb	15.5 b	16.4 a	15.9 ab	16.2 a	<0.05	100
B	35 b	41 a	35 b	35 b	>0.001	20
Cr	34	35	35	35	ns	100
Cu	13.9 c	15.0 bc	16.0 ab	17.6 a	<0.0001	100
Fe	18,152	18,156	18,284	18,208	ns	-
Mn	472 a	459 b	471 a	471 a	<0.001	2000
Ni	24	24	25	25	ns	100
P	555 c	749 ab	638 bc	819 a	<0.0001	-
V	40 b	43 a	40 b	41 b	<0.0001	-
Zn	42 c	46 bc	55 ab	59 a	<0.0001	250

⁽¹⁾ NVS, ADS, and ADSC were applied at 214, 24, and 72 mT ha⁻¹, respectively; ⁽²⁾ The upper concentration limit in agricultural soils (Israeli Ministry of Environmental Protection). See Table 3 for statistical analysis.

). This is in accord with these elements’ higher concentrations and loads in the NVS. Notably, although NVS added merely 1.5 mg As kg⁻¹ soil, this more than doubled its content in the soil. Similarly, the minute amounts of Cd (0.04–0.06 mg kg⁻¹ soil) and Pb (3.4–4.4 mg kg⁻¹ soil), which were added with the ADSC and NVS, significantly increased their content in the soil (Table 7). The above resolution of elemental concentration indicates that the analysis of the soil samples was very precise. In addition, the large number of replicates (18–24 plots per treatment) enhanced the precision of the statistical assessment of the data. Yet, the changes in the measured content of some of the other elements tested were inconsistent with their additions to the soil. Most conspicuously were Cr, Ni, and Fe whose concentrations in the soil remained constant (Table 7) while their addition was up to 50% (Cr and Ni) and 10% (Fe) of their initial content in the soil (Table 4a).

The ADSC- and ADS-treated soils had significantly higher Cu and Zn concentrations than the control soil, and Cu content in the ADSC-treated soils was significantly higher than in the NVS-amended soils (Table 7). This was consistent with their larger input, although the ADS’s input was twice as large as that of the ADSC’s (Table 4a). Finally, the content of P was highest in the ADSC-treated soils, significantly higher than in the control and the ADS-treated soils but not compared with the NVS, and the NVS soils had significantly more P than the control soils. The 70% more P that was added to the soil with the ADS compared with the ADSC was indiscernible (Table 7). Strangely, the NVS seemed to have significantly reduced the Mn content in the treated soil compared with the other treatments, including the control (Table 7). This was despite the fact that the three biosolids added the same amounts of Mn (Table 4a).

As can be expected, the concentrations in the soil of added elements increased following the heavy application of some or all the additives (e.g., As, Cd, and Pb, B, V, Zn). However, the concentration of some elements remained constant (Cr, Ni, and Fe) or even decreased (Mn in NVS applied soils). Similarly, Sajwan et al. [27] reported that Mehlich-1 extracted less Mn from an NVS (but not from a CFA-treated counterpart).

Note that even at the excessive application rates of the three biosolids tested, the content of the elements of concern was well below the upper levels permitted in agricultural soils (Table 7). Similar results have already been reported [19,20,22]. Strangely, had the Israeli threshold for B been applied, it would have excluded these tested representative soils from agricultural use.

3.4.2. Potential Phytoavailability’ of Heavy Metal(loid)s

The DTPA-TEA extraction [47] was used to assess the ‘potential phytoavailability’ of the heavy metal(loid)s and other elements in samples from the top 15 cm soil layer of the 200 L lysimeters. All the fifty-two lysimeters were sampled three times during the span of the experiment: two weeks after the first (June 2011) and third (May 2013) applications, and seven months after the latter (November 2013), which was three months after the cutting of the last lettuce crop. Five sub-plots in Figure 2a–e present the extractability data related to one additive and the non-amended control. Each data point is the average change (in percentage) in the extractability of an element across all pertinent soils tested with that additive. The changes are with respect to the initial concentration in the not-amended control soil counterpart. The number of replicates per element were between four and sixteen, depending on the number of soils and water types examined per additive (see Table 5 for details). The sixth sub-plot (Figure 2f) presents the relative extractabilities of an element per each sampling date, averaged over all the additives, soils, and water type combinations tested (assimilating data from all the fifty-two lysimeters).

The relative DTPA-TEA extractability of the 11 elements in the not-amended control soils (spanning sand to fine clay) changed little during the 2.5-year period (Figure 2a). Yet, Zn extractability approximately tripled (from 0.9 to 2.5 mg kg⁻¹) and those of B, Fe, and Pb nearly doubled. Cu and Cr extractability decreased by ~50%, while P, Cd, and Ni barely changed. All these changes were statistically insignificant. NVS was tested at two annual application rates that accumulated to 214 and 642 mT ha⁻¹ by the third application. The low NVS dose was tested at all four soil–irrigation water combinations while the high dose was tested only with the clayey soil (Table 5). Still, the elemental extractabilities were by and large similar between the two treatments, both in pattern and magnitude (Figure 2b,c). Following the first application the extractability of most of the elements tested had increased in both treatments. The most affected elements were Cr and B, whose extractabilities increased more than 10-fold. The extractability of Fe, P, and Zn also increased by 3–7 times, more at the lower load (probably because 15 of the replicates were in the lighter textures soils). The other elements (Mn, Cu, Cd, Pb) were less affected or not at all. Seven months after the third and last application, the extractabilities of almost all elements steeply decreased, more prominently at those elements that were initially strongly enhanced (Figure 2b,c). At the higher NVS dose, the extractability of Cr and B decreased to nearly the value of the not-amended soil, or even below it (Figure 2c). The extractability of most other elements also decreases to nearly the initial value even below it. Noteworthy, are the substantial extractabilities of Fe and P, elements whose phytoavailabilities are very susceptible to high pH. The extractability of practically all the elements, whether enhanced or not enhanced, decreased during the seven months past the third addition. This steep and thorough reduction in elemental extractabilities in the NVS-treated soils could be due to surface reactions and to superfluous carbonation and coprecipitation.

The ADS was applied only to the clayey soil. Significant enhancement of the extractabilities of Cu and Zn (25–30-fold), Ni (15-fold), and P (~5-fold) occurred, which was in accord with their high content in the ADS (Table 4a). The extractability of most other elements has also increased to some extent, including Cd (whose content in the ADS was 5 mg kg⁻¹), but not that of B, Fe, and Pb. The extractability of all the elements did not change much seven months after the last application, and that of the four elements (Cu, Zn. Ni, P) continued to remain high (Figure 2, ADS). Zn (and Cd) tends to be relatively soluble, and a significant proportion of it resides in the exchangeable soil complex, while Cu and Ni tend to bind to soil OM and become mobilized by the DOC [17,18,60,72].

The ADSC had a similar effect on metal extractabilities as the ADS, except that the increases following the first application were small, and that only the increase in Zn extractability was just as large as with the ADS. Still, the changes in the extractability of most elements following the third application were similar in the ADS and its compost. However, seven months after the cessation of application, the extractability of all the elements in the ADSC-treated soils steeply decreased. Similarly, Haas [73] applied ADS to seven Israeli soils at a load equivalent to 100 mT ha⁻¹ (dry). He showed that the DTPA-TEA extractabilities of Cd, Pb, Cu, Zn, and Ni following 0, 12, 24, and 36 months of incubation of these soil–ADS mixtures gradually and significantly decreased, especially in those soils that were relatively rich in Fe and Mn oxyhydroxides. Using these soils and the soil–ADS mixtures, Oliver et al. [74] demonstrated that Cu activity in the soil solution extracts and Cu isotopic exchange properties were similarly low and constant even after seven years of incubation. This was despite an up to 50% OC loss during the 7 y incubation period.

To summarize, the effect of the additives on the elemental extractability depended on the additives biochemical activity, the chemical speciation of the elements, and on their loads, as well as on the properties of the receiving soils [14,17]. Overall, the “potential phytoavailability” of Zn and Cu were enhanced relatively more than the other elements and, to a lesser extent, also Ni, Fe, and P (Figure 2f). The NVS strongly enhanced the extractability of B and Cr. Although NVS loaded the soils with large amounts of B, this was not the case regarding Cr (Table 4a). These elements quickly lost their high extractability, probably due to adsorption (B) and oxidation (Cr). Fe loads were similar between ADS and ADSC application; however, while Fe extractability remained steady in the ADS-treated soil, it strongly decreased in the ADSC-treated soils after the cessation of application. Possibly, FeS precipitates that might have dominated Fe solubility in the ADS, oxidized in the amended soil thus releasing extractable ferric and ferrous Fe. Interestingly, P extractability was strongly enhanced with all three additives. Regarding NVS application, this seemingly counterintuitive effect probably resulted from its substantial non-stabilized sludge moiety [75].

A prominent feature was the steep reduction in elemental extractabilities of most elements that occurred seven months after the cessation of application of NVS and ADSC. This phenomenon was less prominent in the ADS-treated soil, perhaps due to being still biochemically active. The fate of heavy metal(loid)s in biosolids-amended soils is a matter of ongoing dispute. While some advocate a ‘time-bomb’ hypothesis, based on the release of heavy metal(loid)s bound to mineralizing organic matter [76]. Others endorse the ‘self-attenuation’ hypothesis and show evidence that heavy metals become sequestered in amended soils and thus the risk they present decreases over time (e.g., [57]). Our results on heavy metal(loid)s’ potential phytoavailability, which are a product of a study that covered three amendments applied to three widely different soil types and lasted for 3.5 years, seem to support the self-attenuation hypothesis.

4. Summary and Conclusions

NVS is a mixture of CFA and dewatered sewage sludge. Sludge disinfection is achieved by the heat of dissolution of the CaO, the high pH (>12), and lethal NH₃ concentrations. The fertilizer replacing value of NVS depends mainly on the quality and degree of anaerobic digestion of its sludge component. The least treated sludge provides the highest value, especially with respect to phytoavailability of N, P, and micronutrients. Another innovative application of NVS is the disinfection of fungi and nematode-infested light-textured soils [6,33,36]. The proper use of NVS improves both the profitability of farming and the quality of agro-products, while simultaneously reducing costs to society at large and the exploitation of diminishing resources [7].

An assessment of the levels of As, Cd, and Pb, as well as other elements of concern, accompanied the application of NVS to Israeli agriculture. Mercury was exempt because of its minute concentrations in CFA and NVS. The experiments presented herein focus on the environmental aspects of CFA and NVS applications to agricultural soils, encompassing leaching, plant uptake, and the behavior of heavy metal(loid)s in amended soils. The CFA and NVS loading rates tested were up to 800 and 642 mT ha⁻¹, which were applied to soils of a wide range of physicochemical properties. The experimental systems encompassed 2 L mini-lysimeters packed with CFA–sand mixtures (up to 450 mT ha⁻¹), a three-year 220 L lysimeter study where NVS was loaded to three soils at up to 642 mT ha⁻¹, and a large number of full-scale field experiments performed on a wide selection of soils which extended for up to four years. These tested CFA and NVS loads of up to 800 and 426 mT ha⁻¹. ADS and ADSC were tested too, for comparison, in the 220 L lysimeters and in the 4-year field experiment.

In all these experiments, the extent of leaching and plant uptake were acceptable, and the As, Cd, and Pb contents in the plants were well below the allowed upper limit. Moreover, the foliage concentrations of Cd and Pb were not significantly different from those in plants grown in the non-amended control sand.

In the 220 L lysimeter study, the amended upper 15 cm soil layer laid on top of a four-times thicker buffer zone, and tap water and secondary WWE, which is the primary irrigation water type used for field crops in Israel (although never for edible crops), were used for irrigation. Four cycles of lettuce were grown as a test crop, and the NVS, ADS, and ADSC were applied three times, once every spring. The cumulative loads were 614, 24, and 72 mT ha⁻¹, respectively. The concentrations of As and Pb in the foliage were all below the quantitation limits, and those of Cd were statistically the same in all the treatments, including the control. The content of Mn in the lettuce was lower in the NVS treatments as compared to the control, while those of Mo and B increased. Despite the heavy biosolid loads, the quantitative changes in the elemental content of the 15 cm-amended layer were usually not measurable. Moreover, the potential phytoavailability of almost all the heavy metal(loid)s in this layer decreased over time. This observation supports the assertion that the application of biosolids has an attenuation effect on metals already present in the soil or recently added (the ‘self-attenuation’ or ‘plateau effect’ [57]).

We conducted a large number of field experiments on NVS application, of which only a select sample was presented. The data presented on the concentrations of heavy metal(loid)s in edible, feed, and fodder crops demonstrates that in all cases, the concentrations of As and Cd were well below the allowed upper limit for human consumption. This is also true for Pb, except for a single case of silage wheat that was grown on a quarzitic sand plot amended with CFA at 300 mT ha⁻¹. The Pb concentration (2.5 mg kg⁻¹) exceeded the value for grains for human consumption. Legumes (and potatoes, in one study) were boosted in NVS amended soils (up to ~3-fold yield increase), especially on soils deficient in micronutrients. Increased availability of Mo was noted in most crops, and it was likely the primary reason for the enhancement in the yield of legumes.

To conclude, we demonstrated that the application of CFA and NVS at agronomic rates (as well as in a considerable excess) does not cause any undesirable changes in the chemical properties of the soils. On the contrary, the extractability of heavy metal(loid)s of concern in the amended soils was shown to diminish following repeated application of NVS (as well as other biosolids). This was demonstrated in a wide range of Mediterranean soils that encompass a broad spectrum of physical and chemical properties.

Still, NVS should be applied with proper caution. This is not just another biosolids product; it is unique in its high lime content, increasing (even if temporarily) the soil pH. Hence, the application of NVS (and CFA) to light-textured calcareous soils warrants special attention, particularly when performed before sowing or planting. Although the likelihood of long-term adverse effects is low, even with repeated application of NVS (or CFA), it is advisable to continue monitoring the impact of application of such additives on the properties of agricultural soils. The increased availability of some oxyanions, especially those of Mo, B, and sometimes also Se and Cr, poses not only a potential risk but also makes NVS and CFA suitable tools for biofortification [77].