1 Introduction

The global society generates a vast amount of organic wastes, which should be managed adequately to avoid their accumulation and the corresponding negative effects on human health and the environment (Tyagi et al. 2018). Fruit and vegetable wastes are produced in large quantities worldwide, in markets, supermarkets, restaurants and homes, usually collected together with other food wastes for recycling. It is estimated that 95–115 kg year−1 of food waste is generated per inhabitant in Europe (Bulmer et al. 2021), Spain being the European country with the fifth-greatest production of this type of waste (Eurostat Database 2020). Food wastes are very heterogeneous and their characteristics and composition vary greatly, depending on the place where they are generated (Phun Chien Bong et al. 2018) . Due to their rapid biodegradation and the complicated logistics of their collection, the treatment and recycling of food wastes have become a challenge for society (Scano et al. 2014; Esparza et al. 2020). Recently, the implementation of separate bio-waste collection systems has increased the proportion of the organic fraction in bio-wastes, improving the efficiency of waste management systems.

Among the existing technologies for organic waste management, the biological treatments (composting and anaerobic digestion (AD)) are considered the best options in terms of economic viability and reduction of environmental impacts (Mozhiarasi et al. 2019). In addition, AD produces biogas, a source of renewable energy from biomass, which minimises the environmental problems arising from the use of fossil fuels (Grippi et al. 2020), while providing a solution to the growing global problem of waste management (Srivastava et al. 2014). In AD, the complex organic matter (OM) present in the residues is degraded in the absence of oxygen by the action of a set of microorganisms (Deublein and Steinhauser 2008). The process reduces the volume of the starting material, ensures the elimination of pathogens and the reduction of odours and gaseous emissions and yields two products: biogas and digestate (Deublein and Steinhauser 2008). Biogas consists of methane (50–75%), CO2 (25–50%) and other volatile organic substances in smaller quantities (Deublein and Steinhauser 2008). Digestate is the semi-liquid effluent from AD that, due to its concentrations of plant nutrients, can be used in agriculture, reducing the use of mineral fertilisers (Möller 2015; Nkoa 2014). The agricultural use of digestate can improve soil conditions and thus generate both economic savings and a decrease in the environmental impact provoked by the production and use of chemical fertilisers (Tyagi et al. 2018; Alburquerque et al. 2012a).

The fertiliser potential of digestates is related mainly to their richness in NH4+-N, which is immediately available to plants, and their stable OM content (Makádi et al. 2012), although they can also provide significant amounts of plant-available P and K, which help to satisfy the fertiliser needs of crops (Nkoa 2014). In fact, digestate is included as a component material category (CMC4 and 5) for EU fertilising products in the European legislation (Regulation (EU) (2019)) . The final composition of the digestate, including the nutrient and stable OM concentrations, depends on the characteristics of the feedstock transformed through AD (Risberg et al. 2017; Möller and Müller 2012; Zirkler et al. 2014).

The bio-waste collected separately is largely (more than 60%) composed of fruit and vegetable waste (Liu and Liao 2019). Also, fruit and vegetable waste can be easily collected from larger producers, such as markets. This fraction can be treated by AD due to its high biodegradability and moisture content (75–90%), although its rapid degradation could lead to acidification with high production of volatile fatty acids, limiting the anaerobic process (Ganesh et al. 2022). Most of the information available for digestate composition and agricultural use refers to anaerobically digested animal manures, corn silage, sewage sludge or biowaste, all with high NH4+-N concentrations (Alburquerque et al. 2012a; 2012b; de la Fuente et al. 2013; Häfner, et al. 2022; Reuland et al. 2022). But, little is known about the digestates produced from the AD of fruit and vegetable wastes. Therefore, the study of the characteristics of digestate from fruit and vegetable residues is of particular importance to identify its agronomic value for the partial substitution of chemical fertilisers (Du et al. 2018).

We hypothesised that the digestate from AD of fruit and vegetable wastes can provide essential nutrients and stable OM for the soil–plant system and that the recycling of these wastes in agriculture can promote the contribution of the biodigestion process to the circular economy. Therefore, the objective of this study was to define the agronomic value of digestate from fruit and vegetable wastes, in terms of the dynamics of C and N mineralisation processes in soil and the nutrient supply to crops, to establish its capacity to substitute mineral fertilisers.

2 Materials and Methods

2.1 Origin and Characteristics of the Digestate

For AD, fruit (apple, grapefruit, lemon, mandarin, orange, pear and pomegranate; 35.5%) and vegetable (green beans, broccoli, green cabbage leaves, carrot leaves, cauliflower leaves, celery, coriander, courgette, cucumber, eggplant, fennel, onion leaves, parsley, pepper, potato and tomato; 64.5%) wastes were mixed and crushed to two particle sizes (4 and 10 mm). The AD was performed in continuous stirred-tank type reactors (CSTR), operating in semi-continuous mode under mesophilic conditions (35–37 °C), with a hydraulic retention time (HRT) of 21 days and an organic load rate (OLR) of 1 g VS L−1 day−1 (VS: volatile solids), with a total running time of 84 days (equivalent to 4 retention cycles). Digestates prepared from wastes of 4 (D1) and 10 (D2) mm particle size were sampled after AD (without any post-treatment). They were divided into two fractions: one was used immediately for biological stability and pathogenic microorganisms determination and the other was stored at 4 °C for chemical analysis and for the different studies.

The characteristics of the digestates obtained from both particle sizes were very similar, with low concentrations of heavy metals (Table 1) and organic pollutants, the PAH16 being < 0.16 mg kg−1 (PAH16, sum of the 16 polycyclic aromatic hydrocarbons in the priority list: naphthalene, acenaphtylene, acenaphtene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, indeno[1,2,3-cd]pyrene, dibenzo[a,h]anthracene and benzo[ghi]perylene). Also, they had high microbial stability (average residual biogas potential of 80 mL g−1 VS), in all cases below the limits fixed for CMC5 in the European legislation for EU fertilising products (Regulation (EU) (2019)). However, the high EC and total K and Na concentrations in the digestates may limit their agricultural use. Salmonella was absent from both digestates (in 25 g fresh matter (fm)), but E. coli (3.6 × 104 and 7.8 × 103 CFU g−1 fm for D1 and D2, respectively) was above the limit of 1000 CFU g−1 fm established (Regulation (EU) (2019)) , indicating insufficient sanitisation.

Table 1 Chemical characteristics of the digestates D1 and D2, obtained from anaerobic digestion of fruit and vegetable residues at particle sizes 4 and 10 mm, respectively (mean ± sd). Significant differences between digestates obtained from residues at different particle size are shown by one-way ANOVA. The total concentration of the different elements is shown
Full size table

2.2 Dynamics of C and N in Soil

The mineralisation of the organic C from the digestate in the soil was determined in a 56-day laboratory incubation experiment under controlled temperature and moisture conditions (25 °C and 70% water holding capacity (WHC)). For this, 40 g of previously characterised agricultural soil (S soil, Table 2) was weighed and placed in hermetically closed glass containers of 500 mL capacity, and 1.6 mL of each digestate was added separately, which resulted in the following treatments (in triplicate): D1, D2 and Control (without any amendment); a blank without soil was also tested. Distilled water was added to reach 70% of the WHC of the soil and the control samples received only distilled water. A small vial with 10 mL of 0.1 M NaOH was placed inside each vessel to trap the CO2 evolved during the incubation. The NaOH vials were replaced, to maintain adequate aerobic conditions, after 2, 4, 7, 14, 28, 42 and 56 days, and the CO2 released was determined by titration of the excess NaOH with 0.1 M HCl in the presence of an excess of BaCl2 (Alburquerque et al. 2012b; de la Fuente et al. 2013). The results from the treatments were compared with the data obtained from the empty vessels used as blanks. The daily CO2-C emission rate was calculated by dividing the result at each sampling time by the number of days between samplings.

Table 2 Characteristics of the soils used in the different experiments
Full size table

The cumulative CO2-C evolved was calculated as the sum of the measurements at consecutive sampling times. The results for the mineralisation of C (cumulative CO2-C emission) in the soils were fitted to a first-order kinetic function (de la Fuente et al. 2013; Häfner et al. 2022) (Eq. 1) by the non-linear least-square technique (Marquardt–Levenberg algorithm), using the Sigma-Plot computer program (SPSS Inc.):

$${\mathrm{C}}_{\mathrm{m}} = {\mathrm{C}}_{0} (1-{\mathrm{e}}^{-k\mathrm{t}})$$
(1)

where Cm is the production of CO2-C (mg kg−1) at time t (d), C0 indicates the potentially mineralisable C (mg kg−1 ) and k (day−1) is the rate constant. The statistical significance of the curve-fitting (residual mean square (RMS) and F-values) was also calculated. The mineralisation of the organic-C from the digestates (Cdig) was calculated as the difference between the CO2-C evolved from the amended soils and that produced in the control (unamended) soil and was expressed as a percentage of the C added with the digestate, calculated as (Eq. 2):

$${\mathrm C}_{\mathrm{dig}}=100\times\lbrack(({\mathrm{CO}}_2{-}{\mathrm C}_{\mathrm{treatment}})-({\mathrm{CO}}_2{-}{\mathrm C}_{\mathrm{control}}))/({\mathrm{TOC}}_{\mathrm{added}})\rbrack$$
(2)

The mineralisation of N from the digestate was studied in a 56-day incubation experiment under the same controlled conditions of temperature (25 ± 1 °C) and soil moisture (70% WHC), but using two digestate application rates. The same agricultural soil (S) was used (Table 2). Aliquots (5 g) of the soil were placed in 50-mL plastic centrifuge tubes and 3 treatments were applied: D1, D2 and Control (unamended). In the low-rate incubation, the amount of digestate added to each soil sample was 0.2 mL, equivalent to the amount applied in the previous C mineralisation experiment. In the high-rate incubation, 0.8 mL of each digestate was added to each 5-g soil sample. Distilled water was added to reach 70% of the WHC of the soil; only distilled water was added in the control treatment. Each treatment was run in triplicate. The tubes were covered with Parafilm (which allows gas exchange) to retain soil moisture and avoid anaerobic conditions inside the tubes (de la Fuente et al. 2010). The evolution with time of the inorganic-N (IN) in the soil (NH4+-N and NO3−-N) was determined by destructive sampling on days 0, 2, 7, 14, 28, 42 and 56. The concentration of NH4+-N was determined spectrophotometrically after a sequential extraction using first distilled water and then 2 M KCl. The concentration of NO3−-N was determined by potentiometric determination in an aqueous extract using a selective electrode (USEPA 2007). The digestate samples were mixed homogeneously with the soil at the time of application to avoid any NH3 losses, and there was no airflow at the soil surface during incubation (de la Fuente et al. 2010).

The IN formed in the soil from the total nitrogen (TN) added in the digestate was calculated as a percentage (INd), using the difference between the concentration of IN (NO3−-N + NH4+-N) in the treated soil and that in the unamended control soil (Reuland et al. 2022) (Eq. 3):

$${\mathrm{IN}}_{\mathrm{d}} (\mathrm{\%}) = 100 \times [({\mathrm{IN}}_{\mathrm{treatment}} - {\mathrm{IN}}_{\mathrm{Control}})/{\mathrm{TN}}_{\mathrm{added}}]$$
(3)

Furthermore, the net N mineralisation was calculated as a percentage of the organic-N added to the soil in the amendment (Reuland et al. 2022) (Eq. 4):

$${\mathrm{N}}_{\mathrm{min}} (\mathrm{\%}) = 100 \times [({({\mathrm{IN}}_{56\mathrm{d}}- {\mathrm{IN}}_{0\mathrm{d}})}_{\mathrm{treatment}} - {({\mathrm{IN}}_{56\mathrm{d}}- {\mathrm{IN}}_{0\mathrm{d}})}_{\mathrm{Control}})/{\mathrm{organic}}{\text-}{\mathrm{N}}_{\mathrm{added}}]$$
(4)

2.3 Pot Experiment

A pot experiment was performed using two different agricultural soils, La Pinilla (S1) and La Alberca (S2) (Table 2), and two crops: maize (Zea mays L.) variety Lleno de España and cardoon (Cynara carduculus L.) variety Rustler. Pots (12 cm in diameter and 11 cm high) were filled with 950 g of dry soil. Four seeds were sown per pot.

The treatments applied to each soil were as follows: control without fertilisation (CT), inorganic fertiliser (F), digestate D1 and digestate D2. The digestates were applied on the soil surface and homogenised with the soil in the pots immediately before seeding. The digestate application rate was calculated based on the TN concentration of the digestate (1.2 and 1.3 g N L− 1 for D1 and D2, respectively; Table 1) and according to the requirements of the crops, which were 152.3 and 120.5 mL of D1 and D2, respectively, for cardoon and 183.1 mL of D1 and 144.8 mL of D2 for maize. Fertilisation in treatments F, D1 and D2 was, in all cases, equivalent to 270, 75 and 340 kg ha−1 of N, P (P2O5) and K (K2O), respectively, for cardoon and 324, 120 and 240 kg ha−1 of N, P (P2O5) and K (K2O), respectively, for maize (García-Serrano et al. 2010). In the digestate treatments, the P and K requirements that were not satisfied by the digestate were provided with inorganic fertilisers (CaHPO4 and K2SO4). The inorganic fertilisation consisted of 434 and 693 mg pot−1 of NPK fertiliser (15:15:15) for cardoon and maize, respectively, complemented with NH4NO3 and K2SO4 (965 and 423, and 1012 and 193 mg pot−1 for cardoon and maize, respectively).

The pots were watered with tap water to reach 70% of the WHC of each soil, considering the initial moisture provided by the digestates. The pots were distributed in a growth chamber under controlled conditions of temperature, humidity and light (12 h of light, 25/18 °C day/night and 60/70% humidity day/night). The soils were watered four times per week to maintain the soil moisture conditions. The duration of the experiment was 58 days, at which point the aerial parts of the plants were collected to determine their fresh and dry (60 °C for 48 h) weights as yield parameters. The dried plant samples were ground for analysis, and the soil of each pot was left to air dry and sieved to 2 mm for further analysis.

2.4 Analytical Methods

The digestate samples were analysed for electrical conductivity (EC) and pH directly after sample homogenisation, while for the soil samples, the pH was evaluated by the saturated paste method (Crison Basic 20 pH meter) and EC in a 1:5 (w/v) soil:water extract (Crison GLP 31 Conductimeter). The moisture content of both soils and digestate was determined after drying to constant weight at 105 °C. The volatile solids (VS) in the digestate were determined by loss on ignition at 550 °C for 24 h. The total organic C (TOC) and total N (TN) concentrations in the different samples were measured using an automatic elemental microanalyser (EuroVector elemental analyser, Milan, Italy). The concentration of NO3−-N in the soil was measured in a water extract (1:5 w/v) using a nitrate-ion selective electrode (USEPA 2007), while the NH4+-N was extracted in 2 M KCl (1:5 w/v) and determined by a colorimetric method based on Berthelot’s reaction (Sommer et al. 1992). The concentrations of P, K, Ca, Mg, Na, micronutrients and other trace elements were analysed by inductively coupled plasma-optical emission spectrometry (ICP-OES; ICAP 6500DUO ONE FAST, Thermo Scientific, Waltham, MA USA) after microwave (Ethos‐1, Milestone, Sorisole, Italy) acid digestion using HNO3 + HCl for soils and H2O2 + HNO3 for plants and digestates. In soil, available-K was determined by flame photometry after extraction with 1 N ammonium acetate at pH 7, and available-P was extracted with 0.5 M NaHCO3 (1:10, w/v) for 30 min and measured colorimetrically (Watanabe and Olsen 1965). All values refer to soil dried at 105 °C for 24 h.

The stability of the digestate was evaluated by testing the residual biogas potential using the ANKOM Gas Production System with 310-mL-capacity bottles (ANKOMRF, ANKOM Technology, Macedonia, NY, USA), which automatically records the increase in pressure caused by the generation of biogas under anaerobic incubation conditions at 37 °C. Anaerobic conditions were obtained by purging the headspace of each bottle with N2 and the experiment ended when the biogas production reached stability (12 days, approximately). The biogas produced was expressed as volume of biogas (mL) per unit of VS (g). The biogas accumulated in the bottles at the end of the incubation period was analysed using a gas chromatography system (Agilent 8860 GC System) to determine the methane percentage. To determine the hygienic conditions of the digestate, the presence of Salmonella sp. and E. coli was evaluated, according to official methods ISO 6579–1:2017 and ISO-7251, respectively. The concentration of PHA16 in the digestate samples was determined by gas chromatography–mass spectrometry (GC–MS). All the analyses were performed at least in duplicate.

2.5 Statistical Analysis

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

3 Results

3.1 Carbon Mineralisation in Soil

The addition of digestate to the soil caused a rapid increase in CO2-C emission during the first days of the incubation (Fig. 1a), the maximum emission rate being reached on the second day of incubation (47 ± 0.1, 49 ± 1.2 and 24 ± 1.3 mg CO2-C kg − 1 soil day−1 for D1, D2 and the control, respectively). The CO2–C production rate gradually decreased in all treatments (Fig. 1a) to reach stable and similar values after 42 days of incubation for both digestate treatments, these values being close to those recorded in the unamended control soil.

Fig. 1
figure 1

a CO2-C emission rate (mg C g−1 day−1) in soils with digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively), and unamended control soil (mean ± sd, n = 3); b cumulative CO2-C release in soil during the incubation of soil amended with digestates (D1 and D2) and unamended control soil. The symbols represent experimental values (mean ± se, n = 3) and the lines (in b) represent the C-mineralisation predicted by the first-order kinetic model (see Material and Methods for details)

Full size image

The cumulative CO2-C release showed significant differences between the digestate treatments and the control soil (Fig. 1b). In both treatments with digestate, CO2-C production increased quickly at the beginning of the incubation, and the differences between the digestate-amended and control soils remained at the end of the incubation (56 days). The cumulative CO2-C release reached 256 mg CO2-C kg−1 soil in the control without amendment and 413 and 459 mg CO2-C kg−1 soil in the D1 and D2 treatments, respectively, without significant differences between them. The dynamics of C mineralisation followed a first-order kinetic model (Table 3). The values of potentially mineralisable-C (C0) were very close in the two digestate treatments and higher than in control soil, while similar values for the rate constant (k) were found in all treatments.

Table 3 Carbon mineralisation in the soil treated with the digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively), parameters of the first-order kinetic model used to describe C-mineralisation in the soil with digestate treatments (mean ± sd; n = 3) and the statistical significance of the non-linear curve fitting (RMS, residual mean square, F-value of the ANOVA)
Full size table

3.2 Nitrogen Dynamics in Soil

The digested materials studied showed good levels of the nutrients required for plant growth (Table 1), with, on average, 15.9% of the TN being in the form of NH4+-N. When expressed in dry matter, the concentration of TN of the digestate was 68.6 and 76.9 g kg−1 for D1 and D2, respectively.

The mineralisation of the organic-N fraction was revealed by the evolution of the inorganic forms in the soil. In the low-rate incubation (Fig. 2), a homogeneous dynamic was observed. The concentration of NH4+-N slightly decreased during the first 2 weeks, reaching low values in both digestate treatments and in the control, which remained virtually stable during the rest of the incubation time. This decrease in the NH4+-N concentration was concomitant with increases in the NO3−-N concentration, which, in the case of soils treated with digestate, increased from the second day of incubation to day 28, when the concentration reached its highest value (66.3 ± 1.6 and 65.9 ± 1.9 mg NO3−-N kg−1 soil for D1 and D2, respectively). The values slightly decreased at the end of the incubation, to 54.5 ± 2.1 and 55.9 ± 2.8 mg NO3−-N kg−1 soil for D1 and D2, respectively, values quantitatively higher than those in the control soil (40.0 ± 1.9 mg NO3−-N kg−1 soil).

Fig. 2
figure 2

Evolution of NH4+-N, NO3−-N and inorganic-N (NH4+-N + NO3−-N) in the low-rate soil incubation with digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively) (mean ± se, n = 3)

Full size image

Initially, the IN concentration slightly decreased in soil treated with digestates, but from day 2 to day 28, N mineralisation (an increase in the IN concentration) occurred in the soil with both digestates, with slight decreases afterwards. The concentration of IN at the end of the experiment was lower in control soil (43.8 ± 2.0 mg inorganic-N kg−1 soil) than in soil with D1 (60.0 ± 1.3 mg N kg−1 soil) or D2 (60.9 ± 3.2 mg N kg−1 soil). The proportion of the TN from the digestate present in inorganic forms (INd) that can be available to the crops was (Eq. 3): in D1 38.9 and 32.9% for days 28 and 56, respectively, and in D2 35.5 and 32.5% for the same days, without statistical differences between the digestates. The net N-mineralisation (Nmin) of the digestate (Eq. 4) was very similar for the two materials (33.2 ± 2.5% for D1 and 31.3 ± 3.7% for D2).

In the high-rate incubation (Fig. 3), the dynamic was more erratic than in the low-rate incubation, with an initial predominance of NH4+-N over NO3−-N in the soil, mainly in the treatment with D2. The concentration of NH4+-N increased during the first 7 days in the digestate treatments. After that, the concentration of NH4+-N started to decrease, but NO3−-N concentration only increased after 14 days of incubation. The delay in NO3−-N formation provoked decreases IN concentration until day 14th, followed by increases until the end of the incubation, which could reflect periods of immobilisation-mineralisation of N. At the end of the experiment, the INd as a percentage of TN was statistically higher in D1 (31.7%) than in D2 (16.7%), indicating the higher fertilising value of digestate D1 in terms of plant-available N. The net N-mineralisation was 22.4 ± 4.4% for D1 and 16.3 ± 1.0% for D2, values significantly lower than those obtained in the low-rate experiment.

Fig. 3
figure 3

Evolution of NH4+-N, NO3−-N and inorganic-N (NH4+-N + NO3−-N) in the high-rate soil incubation with digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively) (mean ± se, n = 3)

Full size image

3.3 Plant Production and Composition

Plant production (fresh weight and dry weight in g pot−1) only showed statistically significant differences between the treatments applied for cardoon plants in soil S1 (the values being higher in both digestate treatments), while for maize plants, no significant differences were found between any of the treatments applied to the soils (Fig. 4).

Fig. 4
figure 4

Biomass (fresh weight) of cardoon (a) and maize (b) plants in soils (S1 and S2), with treatments with digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively), mineral fertiliser (F) and control (mean ± se, n = 3). Bars with different letters indicate significant differences between treatments according to the Tukey test at p < 0.05 for each soil

Full size image

In the present pot experiment, the treatment with digestate resulted in the lowest concentration of N in cardoon in soil S1 and in maize in soil S2 (Table 4), and the highest values for cardoon in S1 with mineral fertiliser (without statistical differences from control). Similarly, the concentrations of P and K were highest in the fertiliser treatment for cardoon plants in soil S1, in which the plants grew less than in the other treatments. Contrastingly, the addition of digestate resulted in the highest concentrations of P and K in maize plants in soil S1. The concentrations of Ca and Mg in both plants, and for both soils, were generally lower in the treatments with digestates than with inorganic fertiliser or in the control (Table 4). In contrast, the concentration of Na in both plants was highest in both digestate treatments. The concentrations of micronutrients and potentially toxic elements in both plants were within normal ranges (data not shown), indicating that the digestates did not provoke any negative effects on the plants in this regard.

Table 4 Composition of cardoon and maize plants (g kg−1) grown in different soils and treatments with digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively), mineral fertiliser and control soil without fertilisation (mean ± sd; n = 3). The results of each element for a soil and crop were evaluated by one-way ANOVA and two-way ANOVA for each crop, using treatment and soil as factors
Full size table

3.4 Effects of the Digestate Application in Soils

The digestate can affect the soil chemical properties and nutrient availability. The pH of the soils (S1 and S2) was significantly higher in both digestate treatments (Tables 5 and 6), while the EC showed significant differences among the treatments only in the cardoon crop for S1, but in both crops for S2 (Tables 5 and 6). In both soils, the fertiliser treatment caused the greatest increase in EC, followed by the two digestate treatments. Although digestate can be considered an organic soil amendment, the TOC concentration in the soils was not affected by the application of the digestate, and only the S1 soils from the fertiliser treatment showed significantly higher concentrations of TN than the controls (Tables 4 and 5). The available P and K in the soil were significantly higher in the digestate treatments than in the control soils (Tables 5 and 6).

Table 5 Characteristics of soil S1 after plant harvesting with the fertiliser treatments using digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively), mineral fertiliser and control soil without fertilisation (mean ± sd; n = 3). The results of each element for a crop were evaluated by one-way ANOVA (treatment factor) and two-way ANOVA with treatment and plant as factors
Full size table
Table 6 Characteristics of soil S2 after plant harvesting with the fertiliser treatments using digestates obtained after anaerobic digestion of fruit and vegetable residues at two particles sizes 4 and 10 mm (D1 and D2, respectively), mineral fertiliser and control soil without fertilisation (mean ± sd; n = 3). The results of each element for a crop were evaluated by one-way ANOVA (treatment factor) and two-way ANOVA with treatment and plant as factors
Full size table

4 Discussion

The CO2-C emission from amended soil indicates the C-mineralisation process and the stability of the amendment. During the first days of the incubation of soil amended with digestate from fruit and vegetable wastes, the CO2-C emission was enhanced due to the presence in the digestate of OM easily available to soil microorganisms under aerobic conditions, which activates the soil microbial population (Alburquerque et al. 2012b; Barduca et al. 2020; Häfner et al. 2022); subsequently, the microbial respiration rate declined along the incubation, following a pattern similar to that found for other digestates (de la Fuente et al. 2013; Reuland et al. 2022). Such behaviour is characteristic of a first-order kinetic process. Hence, the values of the kinetic parameters for the soil amended with the digestates studied here show certain similarities to those for digestates of different origin. For example, Alburquerque et al. (2012b) found slightly higher C0 concentrations (ca. 620 mg CO2-C kg−1 soil) and higher k values (from 0.27 ± 0.02 to 0.28 ± 0.04 day−1) in soil amended with digestates from cattle or pig slurry mixed with agro-industrial wastes. Values of k similar to those of the present work were found by de la Fuente et al. (2013) (from 0.04 ± 0.01 to 0.10 ± 0.01 day−1) and Häfner et al. (2022) (0.087–0.098 day−1) in soils amended with cattle slurry digestate and with a range of feedstocks, respectively.

The cumulative CO2-C release values found here were quantitatively lower than those obtained for digestates of pig manures with agro-industrial residues (639–1679 mg CO2-C kg−1 soil; Alburquerque et al. 2012b; de la Fuente et al. 2013) and for food waste digestate (1900 mg CO2-C kg−1 soil; Häfner et al. 2022). However, those digestates had higher concentrations of TOC (5.9–42.8 g L−1 for the pig manures and agro-industrial residues digestates; 36.2% of the dry matter, equivalent to 13.87 g kg1 fresh matter, for the food waste digestate) than the materials studied here, so that the TOC application rate to the soil was lower in the present experiment. Then, expressed as a percentage of digestate TOC (Eq. 2), the mineralisation of the TOC from the digestates exceeded 100% at the end of the incubation. This indicates that the TOC of the digestates is easily degradable in soils and cannot be considered microbiologically stable under aerobic soil conditions (Alburquerque et al. 2012b; Huygens et al. 2020). The low recalcitrance of the OM from food waste can be responsible for the high C-mineralisation of digestate from food waste found by Häfner et al. (2022), which was linked to the low C/N ratio in the digestate (2.18), intermediate between those of the present digestates from fruit and vegetable wastes. Although the residual biogas potential indicates high microbial stability of the digestates under anaerobic conditions, the aerobic conditions prevailing in the soil enhanced the degradation processes. During AD, if the hydrolysis and methanogenesis are not completely balanced, simple organic compounds and intermediate products, which are easily degradable by microorganisms under aerobic conditions, can accumulate in the digestate, leading to the production of unstable materials (Alburquerque et al. 2012a; 2012b). The presence of such intermediate products could enhance soil microbial activity and the oxygen demand after soil application of the digestate, which can result in oxygen depletion and the consumption of the C provided by the digestate (Huygens et al. 2020); even part of the soil OM can be degraded, provoking the so-called “priming effect.” This must be taken into account in soil conservation strategies, since soil TOC (or OM) conservation is crucial to ensure good soil fertility and to promote soil conservation (Martín-Lammerding et al. 2021; Prescott et al. 2021). Thus, digestate composting is recommended over direct soil application for soil C conservation strategies (de la Fuente et al. 2013). In line with the above-discussed factors, in the pot experiment, the addition of digestate did not have any significant effect on soil TOC (Alburquerque et al. 2012c; Huygens et al. 2020; Möller 2015; Riva et al. 2016). An increase in soil TOC should be associated with the vegetation (root exudates or plant material debris) (Mendoza et al. 2022).

The TN concentrations in the digestates (dry matter basis) are within the range found by Reuland et al. (2022) for digestates of different origins. However, the proportion of TN present as NH4+-N was greater in those digestates (on average, 61% of TN) than in the present samples (12.5 and 19.2% in D1 and D2, respectively), making the organic-N the dominant N-fraction in the digestates studied here. This may be related to the low protein concentration of the fruit and vegetable wastes used as a feedstock for AD. With regard to digestates from animal manures, the presence of ≥ 90% of total N as inorganic forms, or TOC/TN ≤ 3, was a requirement for being considered as a RENURE material (Huygens et al. 2020; Reuland et al. 2021); the latter criterion was fulfilled by the digestate studied here. Therefore, the mineralisable-N will constitute a key fraction of the plant-available N of the digestate (inorganic-N initially present and that produced by organic-N mineralisation) and will be controlled by the mineralisation-immobilisation processes in the soil.

Initial microbial N-immobilisation occurred in both experiments (shown by the early decrease in the IN concentration) due to the initial boost of the microbial population (Manu et al. 2021), as confirmed by the quick release of CO2 in the C-mineralisation experiment. Afterwards, a net increase in the IN concentration (NO3−-N increase concomitant with NH4+-N decrease) occurred from day 2 to day 28 in the low-rate experiment, resulting from microbial N-mineralisation in the soil. In the high-rate incubation, ammonification occurred during the first 7 days and nitrification was delayed until the 14th day. Partial nitrification also occurred in a field experiment with digestate and cattle manure (Alburquerque et al. 2012c). The initial decrease in IN observed may indicate microbial N-immobilisation (Reuland et al. 2022), which can delay the development of nitrifying populations (Kuzyakov et al. 2000; Bertrand et al. 2007; Alburquerque et al. 2012b; Waggoner et al. 2021). Re-mineralisation of the immobilised N, with the formation of NO3−-N, was later found. At high digestate application rates, N-losses by NH3-N volatilisation can be relevant in soils with alkaline pH values. However, the pH of the soil used in the incubation experiments was below 7, limiting NH3-N volatilisation as the NH4+ –NH3 equilibrium tends towards the protonated form, and the sandy texture of the soil makes the NH4+-N fixation negligible. Also, aerobic incubation conditions may not favour N-losses by denitrification.

In both the low-rate and high-rate experiments, the proportion of the TN from the digestate present in inorganic forms (INd) that can be available to the crops (35.5–38.9) was lower than those found in NH4+-N-rich digestates (63–81%), which also had high proportions of TN present as IN (53.3–70.3% of TN; Reuland et al. 2022). Therefore, the net N-mineralisation (Nmin) of the digestate was of the same magnitude as the values reported by Reuland et al. (2022) for digestates of different origins (20–39%), in spite of their lower TOC/TN ratio, which may have led to greater net N-mineralisation.

At high application rates, the abundance of labile C available as a microbial energy source may induce strong N-acquiring microbial growth with intense respiration (Alburquerque et al., 2012b), giving a reduction in the net N-mineralisation in comparison with low application rates. Therefore, immobilisation and a delay in nitrification should be taken into account when defining a fertilisation strategy involving digestates, in order to avoid plant nutrient deficiency and negative effects on the environment, and the application should be adjusted, in terms of the timing and amount, to the crop’s N requirements (Makádi et al. 2012; Reuland et al. 2022). Regarding the fertiliser capacity, the organic-N pool from the digestate remaining in the soil at the end of the incubation (66.8–83.7%) will be subject to further mineralisation, releasing IN for the plants later, during the crop cycle. These results give an idea of the short-term and long-term availability of N from digestates and therefore the synchrony or asynchrony with plant N requirements.

In the pot experiment, the rate of application of the digestate to the soils was based on the digestate TN and the crop requirements. The lower N concentrations in plants of cardoon (soil S1) and maize (soil S2) may be associated with a dilution effect caused by greater plant growth (Riedell 2010). Also, they may be a consequence of the partial mineralisation of the organic-N fraction from the digestate, so that the IN supply could have been insufficient for optimal plant nutrition. In addition, the TN concentrations in the soils did not increase with the digestate treatments, but increases in NO3−-N occurred in soils receiving digestate and used to grow cardoon, due to the fast nitrification in the soil, as found in the incubation experiment. Häfner et al. (2022) considered the concentration of NH4+-N in the digestate as the compositional factor most affecting yield, and the C/N ratio the most important one for plant N uptake.

Contrastingly, the highest concentrations of P and K in maize plants (soil S1) were found in the digestate treatments, as reported also by Barbosa et al. (2014). In general, the concentrations of N, P and K in maize were higher than those found by Pampuro et al. (2017) in maize plants fertilised with pellets from pig slurry compost (mean values in the control/fertiliser treatment of: N 1.28/1.91%, P 1.60/1.98 g kg−1 and K 27.92/34.74 g kg−1). In our work, the application rate of both the digested materials and the mineral fertiliser was calculated according to the crop nutrient requirements, but the crop growth period in the pot experiment was shorter than is usual for agricultural production. Thus, the nutrients supplied by the digestate and fertiliser probably exceeded the plant nutrient uptake capacity, resulting in the accumulation of available P and K (and NO3−-N) in the soil. The relatively high K concentration in the digestates from fruit and vegetable wastes, with respect to Ca and particularly Mg, could be responsible for the lower concentrations of Ca and Mg in both plant species (for both soils), in comparison with the inorganic fertiliser or the control. The presence of high K+ concentrations in the soil can depress Ca2+ and Mg2+ uptake (Mengel and Kirkby 2001). Regardless of the possible nutrient imbalances, the concentrations of nutrients in maize found here are considered adequate to high for N, low to adequate for P, high for K and adequate for Ca and Mg (Mengel and Kirkby 2001). Since digestates cannot be considered balanced fertilisers, their application often increases the concentrations of available nutrients in soils (Makádi et al. 2012; Alburquerque et al. 2012c), but at a different rate for each element (Makádi et al. 2012; Riva et al. 2016). The main factor of concern is the salt content of the digestates, which may cause a high accumulation of Na in the vegetative tissues of the plants. This is particularly relevant for cardoon due to its capacity to accumulate Na+ in response to salinity, as this species can be considered a facultative halophyte (Benlloch-González et al. 2005). In cardoon, Na+ and K+ show distinct effects: high concentrations of the latter produce stronger adverse effects than the same concentrations of Na+, so that high accumulation of K+ in the plant is more toxic (Benlloch-González et al. 2005). However, maize is considered a salt-sensitive crop plant, and its yield is affected even by low concentrations of salt (Chellamma and Pillai 2013). Salinity causes osmotic and ionic stresses in maize plants, leading to reduced shoot extension and limited plant growth as well as nutrient imbalances. Although the EC threshold for maize yield decrease (1.7 dS m−1; Chellamma and Pillai 2013) was not reached in the soils, the contribution of K+ and Na+ from the digestates could have limited plant growth and provoked nutrient imbalance (high K+:Ca2+ and K+:Mg2+ ratios).

The optimisation of the fertiliser application rate is needed to make agriculture profitable and to improve the competitiveness and sustainability of the sector (Caamal-Pat et al. 2014) by reducing fertilisation costs. The potential substitution of inorganic fertilisers by digestates has been evaluated using the “Recovery Efficiency” index (RE). This index shows the apparent recovery efficiency for the N applied in comparison with the mineral fertiliser. Because the same amount of TN was applied in the digestates and mineral fertiliser treatments, the following equation was used:

$$\mathrm{RE }= ({\mathrm{N}}_{\mathrm{D}} \times 100) / {\mathrm{N}}_{\mathrm{F}}$$

where ND is the plant N uptake (concentration × dry weight) from the digestate (D1 and D2) and NF is the plant N uptake from the fertiliser (F), both in mg N pot−1.

The RE of the digestates applied to soil S1 exceeded 100%, indicating that all the N provided by the fertiliser to both crops can be substituted by the digestate N, increasing or maintaining plant production. This may well be associated with the immediate availability of the nutrients provided by the digestates in this soil of low fertility. In S2, the values almost reached 100% for cardoon (94.8 and 99.5% in D1 and D2, respectively). These values indicate that digestate from fruit and vegetable residues can be considered a RENURE material, with a N efficiency similar to that of chemical fertilisers (Huygens et al. 2020; Reuland et al. 2021).

For maize, the values obtained were 82.6 and 44.5% for D1 and D2, respectively, probably due to the salinity effect of the digestate on this salt-sensitive crop (Chellamma and Pillai 2013) and the greater fertility of S2 with respect to S1. High N replacement values for digestate applied to a maize crop were reported by Grillo et al. (2021), but Häfner et al. (2022) reported mineral fertiliser equivalents for digestates of between 27.6 and 49.0% in the first year for a maize crop in a field experiment (the greatest value was for food waste digestate, 69.9% in the second harvest). In the case of P and K, in the present work, the values obtained generally exceeded 100% (except for P in maize grown in S2), indicating a positive effect of the digestate treatment on plant growth.

The economic savings per hectare that would be achieved by the application of the digestates studied (and supplementary inorganic P and K) as substitutes for the mineral fertilisation of cardoon and maize crops were also calculated. The official fertiliser prices paid by the farmers in Spain in 2019 were used for this (0.36 € kg−1 for fertiliser 15:15:15, 0.34 € kg−1 for NH4NO3, 0.20 € kg−1 for CaHPO4 and 0.68 € kg−1 for K2SO4; MAPA 2021). The economic savings were estimated to be 549 € ha−1 and 652 € ha−1 for the cultivation of cardoon and maize, respectively. However, other economic considerations must be taken into account, such as the cost of digestate transport and application, the savings in the application of fertiliser and other labour and machinery costs.

5 Conclusions

The quality of the digestates produced from fruit and vegetable residues was adequate for agricultural soil application, according to their concentrations of nutrients and their nutrient recovery efficiencies for the potential substitution of mineral fertilisers. However, the Na+ concentrations of the digestates could limit their use in soil, since prolonged digestate application may cause salt accumulation in soil, and so their use in salt-sensitive crops should be restricted.

The dynamics of C and N mineralisation of the digestates from fruit and vegetable wastes in the soil mirrored the behaviour of digestates from other waste materials. The digestate TOC is easily degradable in the soil and could even promote the partial mineralisation of the soil OM. Therefore, while the provision of plant nutrients is the main benefit deriving from the soil application of these digestates, the decision to use them as soil organic amendments should be taken with care. These digestates can be considered RENURE materials. However, high application rates must be avoided as nitrification can be delayed and organic-N mineralisation reduced.

Digestates from fruit and vegetable wastes can be an adequate source of plant nutrients for salt-tolerant species such as cardoon, since they can substitute 100% of mineral fertiliser. This implies a clear economic advantage, in addition to the environmental implications of the use of residual materials in agriculture, which promotes the circularity of their nutrients and OM in the soil–plant system.