Ash samples from co-combustion of chicken (broiler) manure with wood chips at different proportions were chemically analysed to determine their nutrient concentrations. Ash samples were obtained directly from the combustion system built in a chicken farm to obtain thermal energy. The burner (50 kW) was adapted to the starting material with a fixed grill, but with a mobile cleaning system to avoid the accumulation of the combustion material at the entry area and to facilitate the correct distribution of the aeration throughout the grill. Mixtures with different proportions of chicken manure and wood chips were burnt in the farm system. The proportions were (by weight): manure at 100%, manure:wood at 50:50 and 60:40 and wood at 100%. The resulting bottom ashes were sampled. An initial ash sample of the 50:50 manure:wood mixture was obtained during the optimisation of the burning system, while the rest of the ash samples were obtained after equipment optimisation. The ash samples were fractionated into three particle sizes: > 4 mm, 2–4 mm and < 2 mm. The >4 mm fraction was discarded due to the presence of impurities, mainly from unburned material. The 2–4 mm fraction was ground to <2 mm and incorporated into the smaller fraction, to obtain a homogeneous sample.

The phytotoxicity of the ashes was determined by a germination test with seeds of Lepidium sativum, using 1:10 (w:v) water extracts and dilutions with distilled water to obtain extract concentrations of 12.5, 25, 50 and 75%, in addition to the undiluted extracts (100%). Ten seeds were placed in a Petri dish (11 cm in diameter) containing filter paper with 2 mL of the corresponding extract or distilled water (control); ten Petri dishes were used for each extract concentration and the control. After 48 h of incubation in the dark at 25°C, the germinated seeds were counted and the root length of each germinated seed measured. The germination index was calculated by multiplying the percentages of root length and germination, both with respect to the control (Zucconi et al., 1981).

Ashes from the 50:50 manure/wood mixtures were used in a pot experiment as a P fertiliser. The selection was based on the difficulties found for the combustion of 100% manure associated to the low bulk density, making the mixture with wood more feasible from a practical point of view. Pots (14.5 cm in diameter) were filled with 1.5 kg of an agricultural soil. The soil was a sandy loam of pH 7.5 with 7.8% CaCO₃, electrical conductivity of 0.230 dS m⁻¹, 6.7% organic matter, 3.9 g kg⁻¹ total organic carbon (TOC), 2.03 g kg⁻¹ total-N and 15.8 mg kg⁻¹ available-P. Six treatments were run: non-fertilised control soil, mineral fertiliser and four different application rates of ash, equivalent to 50, 65, 80 and 100% of the phosphorus (P) supplied in the mineral fertiliser treatment; these doses were equivalent to 262, 350, 437 and 525 kg of ash per ha, respectively. The mineral fertiliser treatment comprised a 15:15:15 N:P:K fertiliser and KNO₃, equivalent to 130 kg N/ha, 40 kg P₂O₅/ha and 205 kg K₂O/ha (García-Serrano et al., 2010). The ash treatments were complemented with NH₄NO₃ and KNO₃, to achieve the same N and K fertilisation as the mineral fertiliser treatment. Four pots per treatment were prepared and in each pot three plants of lettuce were grown: plantlets of variety Almadraba baby (1 month-old), acquired commercially, were transplanted from germination trays. The experiment was run for 60 days in a growth chamber under controlled conditions of temperature, humidity and light (12 h photoperiod, 25/18°C day/night and 60/70% humidity day/night). The soils were watered at least twice per week using tap water. The above-ground parts of the plants were harvested, weighed fresh, washed with distilled water and dried at 60°C to determine the dry weight. The plant material was then ground to 0.5 mm for chemical analysis. Soil samples were taken from each pot at harvesting, air-dried and sieved to 2 mm for analysis.

From the results for the nutrient concentrations in the plants and their production (plant biomass as dry weight) in each treatment and in the unfertilised control, the efficiency in the use of the nutrients was determined. The apparent recovery efficiency (RE) of P was calculated as Fixen et al. (2015):

R E = Uptake treatment − Uptake control / P applied

Where Uptake_treatment is the plant uptake of P in the treatment ([P] × dry weight), Uptake_control is the plant uptake of P in the unfertilised control and P_applied is the amount of P applied (all in g per pot).

The equivalent fertiliser efficiency (EFE) of the ash at the different application rates was calculated as the percentage of the apparent recovery efficiency of the mineral fertiliser.

E F E = R E ash / R E fert × 100

Where RE_ash is the apparent recovery efficiency of the ash treatment and RE_fert the apparent recovery efficiency of the mineral fertiliser treatment.

The ash and soil samples were analysed for pH (Crison Basic 20 pH meter) in a water saturated paste for soil and in a 1:5 water extract (w:v) for ash, electrical conductivity (EC) (Crison GLP 31 Conductimeter) in a 1:5 water extract (w:v) and TOC and total nitrogen (also for plants) using an automatic elemental microanalyser (EuroVector elemental analyser, Milan, Italy). The total concentrations of macronutrients, micronutrients and heavy metals were determined by inductively coupled plasma-optical emission spectrometry (ICP-OES; ICAP 6500DUO ONE FAST, Thermo Scientific, Waltham, MA, United States) after microwave (Ethos-1, Milestone, Sorisole, Italy) acid digestion using HNO₃ + HCl for ash samples and H₂O₂ + HNO₃ for plant samples. The soluble cations (K, Na, Ca and Mg) and anions (chloride, nitrite, nitrate, phosphates and sulphates) in ash were determined in water extracts (1:10 w:v), by ICP-OES and ion chromatography (IC), respectively. In soils, nitrate was determined using a selective electrode in 1:5 water extracts (USEPA, 2007), available-P was measured colorimetrically in 0.5 M NaHCO₃ extracts (Watanabe and Olsen, 1965) and available-K was measured by flame photometry after extraction with 1 N ammonium acetate at pH 7 (Kudsen et al., 1985). Available micronutrients and heavy metals in the soil were analysed by ICP-OES after extraction with 0.005 M DTPA + 0.01 M CaCl₂ + 0.1 M triethanolamine (pH 7.3; 1:5 w/v; 1 h) (Lindsay and Norvell, 1978). The availability of P and K in the ash samples was analysed both in water extracts (room temperature and 65°C) and in ammonium citrate extracts (UNE EN 15957). The neutralising capacity of the ashes was determined by titration according to UNE-EN 12945. All results for soil and ash refer to dry matter (105°C, 24 h).

The results were evaluated by one-way ANOVA and the differences between means were determined by the post-hoc Tukey’s test, at a level of significance of p < 0.05. The data were tested for normality using the Shapiro–Wilk test. The statistical software IBM SPSS Statistics version 26 for Windows was used for the statistical analysis.

The ash samples showed high pH and EC values (Table 1), with the initial sample (the one taken during the stabilisation of the combustion system) having the highest EC. The concentrations of TOC and TN were low in all samples, which indicates the loss of organic material during the combustion process; the highest values were again observed in the initial sample, while there were no significant differences among the rest of the samples. Contrastingly, the concentration of nitrate (NO₃-N), which is directly available to plants, was lower in the initial sample than in the rest of the ashes. A different situation occurred with the concentration of total P, which was highest in the initial sample. Despite the ash from 100% manure was the second richest in total-P, it showed the highest concentration of soluble phosphates (> 330 mg kg⁻¹), which was unexpected and possibly a consequence of the peculiar characteristics of that particular manure sample. Generally, P showed rather low solubility in water (below 1.5% of total P), whereas the concentration of available-P (extracted in ammonium citrate) reached 4.15% of total-P in the sample from 100% manure, and less than 1% in the rest of the samples (data not shown).

The main plant nutrients that could be provided by the ashes were P (as indicated) and K (Table 1). In contrast to P, almost 50% of total K was water-soluble (at room temperature) and directly available to plants. Potassium solubility was higher in both hot water and ammonium citrate extracts, with values reaching 80.1 > 73.8 > 64.4 ≥ 63.9% of total-K in hot water extracts for ashes from 100% manure, 100% wood and the manure:wood 60:40 and 50:50 mixtures, respectively, and 94.9 > 87.7 > 84.3 > 74.8% of total-K, respectively, in ammonium citrate extracts. The high soluble K and Na concentrations were mainly responsible for the high EC of the ashes. Sodium can be considered a potentially toxic element when it reaches high levels, causing toxicity in plants. However, the concentration of total-K was 7.4 times higher (on average) than that of Na, which may prevent Na toxicity in plants. Furthermore, there is a predominance of soluble K over soluble Na (Table 1), so the risk of Na toxicity is very low, although the high EC may cause phytotoxicity problems due to osmotic stress in plants.

Other main elements present in the ashes include Ca and Mg, with very similar values for all samples, the ranges being 126–146 g kg⁻¹ for Ca and 36–38 g kg⁻¹ for Mg. These elements can contribute significantly to the high pH values of the ashes, and may also be responsible for their high neutralising capacity, mainly regarding those derived from mixtures of chicken manure and wood chips (Figure 1). These data indicate that combustion ash has a CaO substitution capacity of 34% by weight, which can be very relevant for the use of ash in acidic soils, partially replacing the lime required for pH correction.

The ash materials were rich in essential plant micronutrients (Fe, Cu, Mn and Zn). Small differences in the concentrations of these elements were observed between samples (Table 2); the greatest differences occurred for Zn, with the initial M + W 50:50 sample having the highest value, followed by ash from 100% manure. This may indicate the possible transfer of this element from the metallic part of the burner to ash prepared of the system, but could be associated also with the presence of Zn in the chicken manure. Potentially toxic elements, such as heavy metals and metalloids (As, Cd, Cr, Ni and Pb), were all low, with the only exception of Pb in the initial sample, that reached 133 mg kg⁻¹, again probably due to an early contamination with this element, coming from the burner system during its optimisation period. In agreement with this, the concentration of Pb in the rest of the samples was low. The ash from chicken manure combustion had the lowest concentrations of potentially toxic elements such as Cr, Tl and V (Table 2).

The most relevant issue regarding the agricultural use of the combustion ash samples analysed was their high salt content, as shown by their very high EC values. The potential toxicity associated with the salinity of the ashes was determined in a germination test (Figure 2). Water extracts of the ashes, whether pure (100%) or partially diluted (75%), were highly phytotoxic and completely inhibited seed germination (germination index (GI) value of zero). Greater dilution of the extracts (50%) allowed the germination of some seeds, but the GI was still very low (<4%). Further dilution was necessary to obtain an adequate number of germinated seeds and the GI increased as the concentration of the extract decreased: 81 ± 7.8% for a dilution of the extract to 25% and 102 ± 6.9% for a dilution to 12.5%. Then, it was necessary to dilute the extract to 25% to obtain adequate germination without symptoms of toxicity (> 80%). The 12.5% extract may even have been beneficial for germination, probably due to the supply of nutrients from the ashes.

The plant growth was adequate with all ash treatments; it was lowest in the control without fertilisation and highest with the mineral fertiliser, followed by the ash that supplied 100% P (Figure 3). The treatments with ash at 50 to 80% P produced similar plant yields, greater than the control but lower than for the mineral fertiliser treatment. The concentration of N in the plants was highest in the mineral fertiliser and ash at 100% P treatments, being lowest for the control soil without fertilisation; the rest of the treatments with ash gave a N concentration in the plants that was significantly lower than for the mineral fertiliser (Table 3). However, the P concentration in the plants was not significantly different among the different fertilising treatments (with mineral fertiliser or with ash), but all showed higher values than the unfertilised control. These results indicate that the ash can supply enough P to plants for their growth.

The concentration of K was highest in the plants receiving the mineral fertiliser and lowest in those grown in the control soil. Ash treated plants showed intermediate K concentrations, with no significant differences among them (Table 3). Contrastingly, Na levels in the plants were not affected by the addition of ash, in agreement with the fact that the total and soluble concentrations of this element were lower than those of K in the ashes (Table 1). The concentrations of the different micronutrients in the plants were all very similar for all treatments—Fe and Zn showed no statistically significant differences, while small differences were observed for Cu and Mn between treatments (Table 3) – and all can be considered normal for plant growth. No accumulation of heavy metals was observed in the plants after the use of combustion ash as a fertiliser substitute (data not shown).

The most relevant changes in the characteristics of the soil were found for the pH values at plant harvest, with a concomitant increase as the ash rate increased, from 7.7 in the control to 8.0 in the treatment with the highest ash rate (Table 4). This effect clearly indicates the usefulness of ash to regulate the pH in acidic soils, as it would decrease the amount of lime required for soil pH adjustment.

Despite the high EC of the ash, this parameter did not increase significantly in the soil (Table 4). Parameters such as the concentrations of TOC and total N (and also nitrate) in the soil showed only very small differences among the treatments, which could be associated with their low concentrations in the ash, and also to the fact that N was supplied at the same rate to all treatments (except the control). The concentration of available P in the soil was increased significantly in the treatment with the highest ash rate, followed by the treatment with the lowest rate of ash and the mineral fertiliser. These results might be showing that the P potentially provided by the ash in excess of the plant requirements was fixed in the soil, due to its calcareous character and the low solubility of this element in the ash, as previously discussed. Furthermore, the concentration of available K in the soil was significantly lower in all the ash treatments than for the mineral fertiliser and the unfertilised control (Table 4), which suggests that the K provided by the ashes and not assimilated by the plants (if any) did not remain in the soil in plant available forms.

Despite the presence of certain heavy metals in the ashes, their available concentrations (extractable with DTPA) in the soil were all quite low (the highest being 2 mg kg⁻¹, for Pb), which indicates a scarce risk of metal accumulation in the soil or plant toxicity. In fact, the total amount of heavy metals added to the soil with the ash at the highest application rate was (mg kg⁻¹): 1.12 Zn, 0.28 Cu and < 0.1 for the rest of heavy metals (Cd, Co, Cr, Ni, Pb), which support the assessment of scarce risk of metal accumulation. These values are far below the limit values for amounts of heavy metals which may be added annually to agricultural land, based on a 10-year average (Cd 0.15, Cu 12, Ni 3, Pb 15 and Zn 30 mg kg⁻¹) according to the European legislation (Council Directive, 1986).

Using the values of the concentration of P in the plants, the plant biomass (the dry weight of the aerial part) and the amount of P added in each treatment, the use of the P applied through fertilisation was calculated as its apparent RE. The apparent RE of P from the mineral fertiliser was 72.8% of the applied P, greater than for the ash treatment at 100% P (61%). The values diminished as the application rate of ash increased, following the order 125 > 69 > 64 ≈ 61% in the ash treatments at 50, 65, 80 and 100% P, respectively. The high P efficiency at the lowest ash rate indicates that the amount of P added was not enough for the plant requirements, and the plants had to take up part of the P already available in the soil.

The fertilisation efficiency of the ash treatments in comparison with the mineral fertiliser indicates the equivalent fertiliser efficiency and therefore the fertiliser replacement value (Figure 2). The ash treatments were able to replace 79% of the mineral P fertiliser when the ash was applied at the same P rate as the mineral fertiliser (ash treatment 100% P), with greater values for 80 and 65% P application by ash (82 and 90% of the mineral P fertiliser, respectively). All this indicates that the ash P fertilising capacity is capable of providing 80% of the P supplied by the mineral fertiliser for the P nutrition of the plants.