Biochar addition to organo-mineral fertilisers delays nutrient leaching and enhances barley nutrient content

ABSTRACT Biochar, a carbon-rich solid produced from biomass pyrolysis, has attracted growing interest as a fertiliser ingredient due to its ability to non-permanently retain nutrients. A greenhouse pot experiment was set up to compare three commercial organo-mineral fertiliser formulations (NPK, NP and K) with the corresponding formulations containing a slow-pyrolysis wood biochar (NPK+B, NP+B and K+B) (6 replications each). Nutrient leaching as well as crop growth and nutrient uptake was monitored using barley as model species. Nutrient leaching was slowed down in the NPK+B compared to the NPK fertiliser. The most responsive ions were nitrate and potassium, whose leaching during the two first weeks was reduced by 28% and 22%, respectively, while this trend reversed from the third week on. One plausible explanation would be a microbial nutrient immobilisation mediated by the concurrent NPK and biochar habitat provision. NPK+B significantly enhanced barley straw biomass (23.43% increase respect to NPK), whereas all the biochar-based fertilisers showed increases in nutrient content and export (involving potassium, sulphur, calcium and manganese), possibly indicating that biochar acted as a nutrient source. These results provide some evidence of the potential use of the studied biochar in biochar-based fertilisers to meet nutrient availability with plant demands.


Introduction
Nitrogen (N), phosphorus (P) and potassium (K) are the main limiting nutrients for crop growth, but due to the high solubility of traditional mineral fertiliser formulations, there is a poor synchrony between N, P and K release rate and crop uptake needs (Pang et al. 2018). Apart from solubilisation and subsequent leaching losses, other processes such as NH 3 volatilisation and denitrification, in the case of N (Cameron et al. 2013), or P retrogradation (Vanderdeelen 1995), also contribute to low nutrient use efficiency by plants. As a matter of fact, plant nutrient uptake from mineral fertilisers has been reported to be as low as 33-50% for N (Raun and Johnson 1999;Hirel et al. 2011) and 10-25% for P (Syers et al. 2008), which results in economic costs for farmers and has environmental risks associated. Namely, N and P leaching and runoff to water bodies provoke eutrophication (Vitousek et al. 1997) and have been linked to human health risks, such as the carcinogenic effect of N-nitroso compounds formation after nitrate-polluted drinking water exposure (Ward et al. 2018).
all of them contained maize flour in their formulation and differed in their macronutrient content (NPK, NP and K), and in the presence of biochar (+B). The six resulting combinations were designated as NPK, NP, K, NPK+B, NP+B and K+B. The exact formulation is not shown for being subject to business confidentiality, but it should be noted that N was provided as a mixture of ammonium and organic sources while P and K were supplemented in an inorganic form (carbon (C) content, and C/N ratio of fertilisers are shown in Table 1).
The biochar used for NPK+B, NP+B, and K+B formulations was produced by slow pyrolysis (400-550°C, 120 min) from a mixed feedstock of Quercus ilex, Quercus suber and Eucalyptus sp., and obtained by Corchos Oliva S.L. (Oliva de la Frontera, Badajoz, Spain). A detailed characterisation of the biochar is presented in Table 2. For more information about this biochar refer to Martí et al. (2021).
Pots (13 cm height, 16.5 cm diameter) were filled with 1.8 kg of soil sieved to 5 mm. On the onset of the experiment (16 October 2018), the amount of fertiliser tablets to provide 66.67 mg N kg −1 (equivalent to 173.3 kg N ha −1 ) for N-containing fertilisers, and equivalent K contents for K- Table 1. Carbon content (%); C/N ratio; dosing of fertiliser (g pot −1 ) and application rates of N, P, K and biochar (mg kg −1 soil) for each fertiliser. containing fertilisers (Table 1), were added to each pot at a 5 cm depth. Biochar addition rates were not possible to even out (Table 1) since each fertiliser formulation differed in its biochar proportion for stability purposes. Aside from fertilisation, pots were watered to field capacity (around a 20% moisture content) and fifteen barley seeds (Hordeum vulgare L. Graphic variety) were sown (thinned to three plants after a fortnight period). Plants were grown under a photoperiod of 14 h light:10 h dark. Climatic conditions during the experimental period (October 2018-April 2019) ranged from 1 to 30°C with a mean of 12°C.

Leaching assay
Pots contained a 2 mm mesh gauze at the bottom to allow proper drainage while preventing soil losses. Soil water content was determined gravimetrically before the leachate sampling and used for the estimation of the water to be added to achieve a target leaching volume of 100 ml per pot. Distilled water was used for leaching, and, to simulate natural rainfall, water was slowly added to the pots using perforated bottles. The leachate was collected by placing each pot on a glass tray but raised 1.3 cm to ensure drainage. A total of nine leaching events were carried out. Leachates were collected once per week during the first five leaching events; then, the sampling frequency was extended to two weeks and, finally, to three weeks (see sampling schedule in Figure 1). The collected leachate was then filtered (using Whatman no. 42 filters) under laboratory conditions and stored at −20°C until analysis. Water-soluble ions in leachates were measured as proxy for nutrient availability. Although Na + is not generally considered an essential element for plant nutrition, it was ascribed under the 'plant nutrient' category for being a beneficial element (Brown et al. 2021). Liquid chromatography was carried out to determine leachate inorganic ions content on a Dionex ICS-1100 ion chromatograph (Dionex, Sunnyvale, USA) using a AS4A-SC Dionex anion column for the quantification of Cl − , NO 2 − , NO 3 − , HPO 4 2and SO 4 2-, and a CS12A Dionex cation column for Na + , NH 4 + , K + , Mg 2+ and Ca 2+ determination. All the ion concentrations were estimated using linear calibration except for SO 4 2-, NH 4 + , Mg 2+ and Ca 2+ , in which quadratic regression substantially increased the fitting (R 2 ). Detection limit (LOD) estimation was set as three times the standard deviation of five blank values. NO 2 − concentration values were not considered in this study for being almost always below detection limits (i.e. < 0.10 mg NO 2 − L −1 ). Additionally, HPO 4 2was also discarded from analysis since its signalto-noise ratio in the chromatogram did not exceed the set value of 3. Leachate ionic concentration was quantified as both absolute and cumulative leaching and was expressed as mg per dry weight (kg) of soil, according to the following equation (Eq. (1)): Where LIC soil stands for leachate ionic concentration expressed on a soil dry weight basis, and LIC LC for leachate ionic concentration as obtained from liquid chromatography.

Plant measurements
Above-ground barley biomass was harvested after plants were fully grown and senescent (after a 6month growth period) and dried at 60°C for 48 h. The total number of ears and the number of grains per ear were quantified. Then, straw and grain were weighed separately to determine their biomass. After straw and grain were ground in a ball-mill, nutrient content (N, P, K, Ca, Mg, S, Fe, Mn and Zn) was determined through near infrared spectrometry (NIRS) by scanning the grounded samples from 1100 to 2500 nm in a NIRSystems 5000 scanning monochromator (FOSS, Hilleröd, Denmark). The calibrations used were developed in a previous study (see the Supplementary Materials in Martos et al. 2020 for more details). P content in straw presented some negative values that were set to 0. Finally, nutrient export was calculated by multiplying each nutrient concentration measurement by its corresponding biomass.

Statistical analyses
Since the objective of this research is to compare fertilisers with and without biochar inclusion in its formulation, statistical tests involved comparisons between each organo-mineral fertiliser (without biochar) and its counterpart with biochar (i.e. NPK vs NPK+B; NP vs NP+B, and K vs K+B). Longitudinal data, i.e. variables for which exist a between-subjects factor (biochar inclusion in the fertiliser), and a within-subjects factor (sampling dates) were analysed using two-way mixed ANOVAs, which were computed with the rstatix package v0.2.0 (Kassambara 2019b). Before, Shapiro-Wilk and Levene tests were used to ensure that data had a normal distribution and homogeneous variances, respectively. When these assumptions were not met, the test was run on the log 10 -transformed variable. The assumption of sphericity was checked using the Mauchly's test, and when violated, the Greenhouse-Geisser correction was applied. Finally, homogeneity of covariances was tested by Box's M. Statistical results of the mixed ANOVA are shown in Supplementary Table S1 (S1.1.-S1.48.). Pairwise comparisons were tested with t-test with Bonferroni adjustment, and the significance level was set at p < 0.05. These statistical analyses and results visualisation were performed using R software v. 3.6.1 (R Core Team 2021) using the packages ggplot2 (2016) and ggpubr v 0.2.3 (Kassambara 2019a).
For absolute nutrient leaching data the Principal Response Curves (PRC) method 1999), developed for biological community data, was conducted using the CANOCO software version 5.12 (Ter Braak and Šmilauer 2012). The PRC describe the trajectory over time of the community response (nutrients in this study) in each treatment group, expressed as coefficient of community response, i.e. the canonical coefficient C dt , relative to the control (which response is set to 0); whereas the weight (b k ) indicates the affinity of the response of each attribute (here ions in leachates) to the overall community response. Positive b k values indicate attributes whose response pattern follows the PRC, by contrast, negative values indicate attributes whose response pattern is in the opposite fashion to the overall PRC pattern. Near zero b k values represent weak responses or response patterns unrelated to the PRC. Given the disparity of nutrient concentration ranges, data was standardised to zero-mean and unit-variance prior analysis. Significance of the first axis was checked with a Monte Carlo permutation test, while significances at each time point were evaluated by performing redundancy analysis (RDA) on subsets of different sampling dates.
Finally, those parameters analysed at a single sampling date (plant measurements) were assessed by means of Student t-test, Welch's t-test (if homoscedasticity assumption was not met) or the Mann-Whitney-U test (if normality assumption was not met), all of them with Bonferroni correction (rstatix package v 0.2.0 (Kassambara 2019b)). When mixed ANOVA analysis of longitudinal data was not significant, differences at each sampling date were also checked with this approach.

Nutrient leaching
The pair of fertilisers NPK vs NPK+B (Figure 2a) was the only one showing significant results on nutrient leaching patterns as measured by the PRC analysis ( Figure 2), with 80.18% of the total variance being explained by time and 5.78% by biochar treatment. A significant proportion of the variance (52.21%) was captured by the first canonical axis of the PRC (Monte Carlo permutation test, 499 permutations, F = 1.8, p = 0.018). The RDA analyses at each sampling time showed contrasted and significant temporal effects associated to biochar inclusion in the formulation. At the first leaching events (days 8 and 15 since experimental onset) the NPK+B leached less nutrients in contrast to the NPK. However, at the third leaching event (day 22) this trend was inverted, and from that moment onwards, NPK+B leached more nutrients than NPK (significantly at days 22, 36, 85 and 106). As shown by b k values of the PRC, the most responsive nutrients were NO 3 − -N (b k = 1.50) followed by K + (b k = 1.26), whereas NH 4 + was the only ion showing an inverse pattern with respect to the PRC (b k = −0.53).
Regarding the absolute leaching of each ion plotted separately (Supplementary Figs. S2 and S3), the results of mixed ANOVA showed significant interactions between biochar treatment and time for NO 3 − -N, Na + , K + and Ca 2+ (Supplementary Tables S1.1-4 and S6) in the NPK vs NPK+B fertiliser pair. In addition, Mg 2+ , Cl − and SO 4 2also showed significant effects at some sampling dates for the same pair of fertiliser (Supplementary Figs. S2 and S3). These ions mostly followed the same trend explained in the PRC analysis, i.e. more leaching in the NPK fertiliser at the first weeks and the inverse trend in the subsequent weeks. As an example, in the NPK+B treatment, NO 3 − -N and K + decreased their leaching at the two first weeks by 27.99% and a 22.01%, respectively. That was not the case for ammonium, whose leaching was slightly higher (0.5 mg kg −1 ) in the NPK+B compared to NPK at the first sampling time. Concerning the NP and K fertiliser pairs, only sparse significant results were found and thus there was no clear pattern on nutrient leaching as mediated by biochar.
The results of mixed ANOVA evidenced time effects always being significant (Supplementary Tables S1.1-24), i.e. there was a generalised plunge in all ions for every fertiliser pair at about the fifth leaching event. Concerning the nitrogen forms, it should be noted that, NO 3 − -N content was about two orders of magnitude larger than that of NH 4 + -N at the first samplings, while from the fifth leaching event ionic content was similar (and lower) for the two ions (Supplementary Figs. S2 and S3).
Finally, in the matter of cumulative nutrient leaching, whereas sparse significant differences were found at the first sampling dates, there were no significant differences on the final cumulative amount of nutrients leached associated to biochar inclusion in fertiliser formulation (Supplementary Figs. S4 and S5).

Barley growth parameters
Barley straw weight was significantly enhanced in the NPK+B fertiliser (23.43% increase) in contrast with NPK according to Welch's t-test, t (5.68) = −2.61, p = 0.04. However, the higher straw biomass in the NPK+B fertiliser did not concur with significantly higher grain yield on this same treatment, although it was marginally improved (Welch's t-test, t (6.0) = −2.25, p = 0.065) (Figure 3a, b). Ear number and grain number per ear were not significantly different within any pair of fertilisers ( Figure  3c, d).

Barley nutrient content and export
Grain nutrient content was unaffected by any of the treatments within any fertiliser pair (Table 3). However, the NP+B fertiliser did increase Ca (t-test, t (10) = −2.59, p = 0.03) and Mn export in grain (Mann-Whitney test, U = 5, z = −2.04, p = 0.04) compared to the NP fertiliser (Table 4). Concerning straw nutrient content and export, some beneficial effects were found for the biochar-based fertilisers. Precisely, K content (Welch's t-test, t (5.33) = −2.53, p = 0.049) and export (Welch's t-test, t (5.43) = −3.11, p = 0.02), together with S export (Welch's t-test, t (5.69) = −4.39, p = 0.01), were higher at the NPK+B in contrast with the NPK fertiliser (Tables 3 and Table 4). In addition, Mn content . Mean (± SE) (n = 6) straw weight (g pot −1 ) (a); grain weight (g pot −1 ) (b); ear number per pot (c); and grain number per ear averaged per pot. Different letters indicate statistically significant differences within a fertiliser pair (NPK vs NPK+B, NP vs NP +B, and K vs K+B). Table 3. Nutrient content in grain and straw of harvested barley for the six organo-mineral fertilisers. Reported values are mean ± standard errors (n = 6). Different letters indicate statistically significant differences within a fertiliser pair (NPK vs NPK+B, NP vs NP +B, and K vs K+B).

NPK+B formulation delayed most of the nutrients release
The fact that nutrient release was only delayed in the NPK+B fertiliser and not in the other biocharbased fertilisers plausibly discards any biochar direct nutrient retention mechanisms such as chemisorption, physisorption or water-pore retention. Namely, despite the higher rate of biochar applied in the NPK+B formulation in relation to NP+B and K+B, we would have expected some kind of effect (albeit weaker) also in the other two biochar fertilisers. Although biochar could also have altered leaching patterns by shifting pH (Laird and Rogovska 2015), it seems unlikely in the current study, where both the soil (pH 1:2.5 8.2; Marks et al. 2016) and biochar (pH 1:20 8.5) had an alkaline pH, and this same soil has been proved to resist biochar-induced pH shifts (Marks et al. 2016). On the other hand, redox status alterations that could drive nutrient availability are less common at alkaline pH (Chacón et al. 2017). By contrast, as both biochar and macronutrients in NPK+B could influence the microbial potential to grow and store nutrients, microbial nutrient immobilisation could be behind the observed temporary nutrient retention. Concretely, microbial biomass can grow and store nutrients while C and nutrient provision is available, but when microbes are devoid of such resources, nutrients immobilised within them are returned to the soil phase via decomposition of dead cells, which can, in turn, contribute to the pool of available nutrients for plants (Anderson and Domsch 1980). Regarding the causative mechanisms of the potential microbial nutrient immobilisation, different drivers could have played a role. Since microorganisms in agricultural soils are usually C limited (Schimel 1986), biochar labile C provision could be one explanatory factor. In particular, lowtemperature biochars (< 500°C), thus akin to our biochar produced at 400-550°C, are more likely to induce net nutrient immobilisation, because they contain higher concentrations of bioavailable C and residual bio-oils (DeLuca et al. 2015). Nevertheless, all fertilisers contained labile C in the form of maize flour, so this mechanism could be mostly disregarded. In this regard, the effect of biochar providing suitable habitat for microorganisms could have played a role, i.e. the porous structure of Table 4. Nutrient export (nutrient content*biomass) in grain and straw of harvested barley for the six organo-mineral fertilisers. Reported values are mean ± standard errors (n = 6). Different letters indicate statistically significant differences within a fertiliser pair (NPK vs NPK+B, NP vs NP+B, and K vs K+B  2.13 ± 0.0 2.42 ± 0.1 2.14 ± 0.1 2.30 ± 0.1 1.57 ± 0.1 1.67 ± 0.1 Straw N (g) 0.74 ± 0.1 1.55 ± 0.5 1.51 ± 0.3 1.68 ± 0.3 0.46 ± 0.2 0.51 ± 0.2 P (g) 0.00 ± 0.0 0.76 ± 0.3 0.00 ± 0.0 0.69 ± 0.4 0.41 ± 0.2 0.00 ± 0.0 K (g) 2.97 ± 0.3 a 8.00 ± 1.6 b 3.75 ± 0.4 7.15 ± 2.3 5.21 ± 1.0 3.00 ± 0.5 S (g) 1.12 ± 0.0 a 1.61 ± 0. 1.67 ± 0.1 2.14 ± 0.2 2.02 ± 0.1 2.07 ± 0.1 1.10 ± 0.1 1.14 ± 0.1 biochar could have protected microbes from grazers, increased the water retention (and thus meet microbial moisture requirements) or prevented microbes from leaching as they adsorb on biochar surfaces (Lehmann et al. 2011;Ennis et al. 2012). Indeed, slow-pyrolysis biochars (as our biochar) are more prone to boost microbial communities (Gul et al. 2015). Although microbial colonisation is not expected to be a substantial process in the short term, biochar blending in a powdered form could have promoted such an effect (Quilliam et al. 2013). This biochar habitat effect in combination with commensurate nutrient supply (NPK instead of reduced nutrient combinations) could fulfil the microbial stoichiometric requirements (Ashraf et al. 2020) and thus allow for microbial growth and retention of nutrients within their biomass. Interestingly, in the meta-analysis of Melo et al. (2022), significant increases in crop productivity were found for biochar-based fertilisers when combined with NPK but not for those with NP. This could fit with the hypothesis that the concurrent provision of both biochar and the three macronutrients is determinant for biochar to show effects. Furthermore, the timing of nutrient release in the NPK+B fertiliser, i.e. retaining nutrients the first two weeks and releasing them from the third week onwards, may support the microbial immobilisation mechanism. Specifically, and although microbial turnover time can vary greatly (normally in the range of days to months; Schmidt et al. 2007), similar values of microbial turnover have been found in cultivated soils (Cheng 2009).
Additionally, the microbial immobilisation mechanism could also explain the nutrient release delay found for multiple ions (NO 3 − -N, K + , Na + , Ca 2+ , Mg 2+ , Cl − , and SO 4 2-) since although microbial immobilisation mainly provokes N and P retention (Malik et al. 2013), other nutrients such as K, S and Ca are also substantially stored in microbial biomass (Anderson and Domsch 1980;Brookes 2001;Yamashita et al. 2014). In accordance with this, PRC results of the NPK+B vs NPK pair of fertilisers pointed to major effects on NO 3 − -N and K + retention, plausibly as a result of their high relative abundance in microbial biomass or by their direct addition with fertiliser application.
On the other hand, the low recovery in leachates of added P in both NPK+B and NPK is not surprising as phosphate is immobile in most soils because of precipitation reactions and adsorption to mineral surfaces (Haygarth et al. 2013). In our experiment, the negligible P release is mainly attributed to P precipitation with Ca due to the high Ca content and high soil pH (Hopkins and Ellsworth 2005) found in the soil used for this experiment. Regarding the Cl − leaching delay, it is less straightforward to understand because its retention in microbial biomass is not expected (Kanwar et al. 1997). As a possible explanation, nitrification of the ammoniacal N provided in fertilisers is known to cause acidification and therefore to displace basic cations from the cation exchange complex (Poss and Saragoni 1992;Bouman et al. 1995), but this process can also affect anion availability and leaching. In detail, when cations like Ca 2+ or Mg 2+ are displaced by H + resulting from nitrification, anions (as Cl − ) could be concurrently released (i.e. they can be directly bounded to these polyvalent cations by bridging or be present in the diffuse double layer around the cation exchange complex). Then, if microbial immobilisation was promoted in the NPK+B fertiliser, less NH 4 + would have been available for nitrification and, as a result, acidification and leaching of ions (including Cl − ) could have been prevented to some extent at this treatment. It is worth mentioning that if this process occurred, it could imply complex interactions between microbial immobilisation, nitrification and nutrient leaching, making difficult to discern why NH 4 + was the only ion whose release was not delayed by the NPK+B formulation, with its leaching indeed being higher at the NPK+B than in NPK at the first sampling date. However, the low quantities found for soluble NH 4 + even at the first sampling date, especially in comparison with NO 3 − , make us disregard NH 4 + release as an important process for understanding nutrient dynamics.

All biochar-based fertilisers enhanced barley nutrient status and NPK+B increased straw biomass
As expected given the more gradual nutrient release achieved in the NPK+B fertiliser, thus potentially improving plant uptake, this fertiliser was the only one significantly enhancing barley straw biomass, whereas grain biomass was only marginally improved. However, it should be noted that the high number of leaching events performed in this experiment, which led to a soluble nutrient shortage at about the fifth sampling, could have diluted the possible NPK+B beneficial effects on grain biomass.
Regarding barley nutrient content and export, all biochar-based fertilisers exerted some beneficial effects compared to the corresponding non-biochar formulations. The majority of effects became apparent in straw nutrient and export, i.e. the NPK+B fertiliser enhanced K content and export, and S export in contrast with NPK, while K+B improved Ca content and export and Mn content compared to the K fertiliser. On the other hand, the NP+B fertiliser provoked effects on grain export, i.e. enhanced Ca and Mn export in comparison with the NP fertiliser. Nutrient content and export improvement caused by the NPK+B fertiliser might result from its more gradual nutrient release pattern, by directly providing nutrients, i.e. biochar acting as a source of nutrients itself (Chan and Xu 2009), or a combination of these two mechanisms. On the other hand, regarding the NP+B and K+B fertilisers, since nutrient leaching patterns were not found to improve in relation to its conventional counterpart, the beneficial effects on plant nutrient status would mostly fit the direct nutrient provision hypothesis. In support of this, the studied biochar has been proven to release K in water extracts (unpublished results). In addition K, Ca and S are nutrients commonly reported to be released after biochar application (Lehmann et al. 2003;Laird et al. 2010b;Marks et al. 2016). Finally, although less commonly reported, biochar could also have released Mn (Elmer and Pignatello 2011); alternatively, soil Mn solubilisation due to redox properties of biochar has also been reported (Graber et al. 2014), but only at pH levels below 8, and therefore this mechanism is not likely in our alkaline system.
Our results could be in line with those of Güereña et al. (2013), who reported not significant biochar effects on grain yield whereas N leaching was reduced in biochar treatments, this result being potentially attributable to a three-fold increase in microbial biomass. However, if the microbial immobilisation mechanism is confirmed, caution must be laid on fertiliser development and application, since this positive effect might be soil-specific and highly dependent on the native microbial activity, biochar dosing and biochar composition. For instance, no microbial immobilisation might be observed in soils with already high soil organic carbon contents and microbial biomass (Subedi et al. 2016;Yanardağ et al. 2017). On the other hand, excessive microbial immobilisation on larger time frames could cause competition for nutrients between plants and soil microorganisms (Anderson and Domsch 1980). Despite these potential constraints, biochar-based fertilisers are of high interest for being able to provide immediate benefits for soil fertility by supplying and, more importantly, by retaining nutrients. Additionally, biochar-based fertilisers can also provide multiple environmental benefits widely proposed in the literature, such as a higher C sequestration (Shackley et al. 2013), greenhouse gas mitigation (Kammann et al. 2017) or water retention (Omondi et al. 2016). Further research is needed to avoid unintended effects and to maximise the number of services which biochar can deliver as a fertiliser ingredient.

Conclusions
The inclusion of biochar in fertiliser formulation did not alter the final cumulative amount of nutrients leached; nevertheless, there were distinct temporal patterns of nutrient release in the NPK+B fertiliser. While most fertilisers contributed to a large flush of available nutrients upon application, the NPK+B biochar-based fertiliser caused a delayed and more sustained release of nutrients throughout the growing season. Microbial nutrient immobilisation could be behind this response since the effects were only found in the biochar fertiliser which supplied simultaneously the three macronutrients (reduced nutrient combinations could have limited microbial growth). Although there were no significant differences on grain production, straw biomass was indeed increased in the NPK+B treatment, potentially indicating that this fertiliser might aid in the synchronisation of nutrient availability and crop requirements. Furthermore, nutrient content and export were enhanced in all biochar-based fertilisers, likely as a result of biochar acting as a direct nutrient source. To conclude, the slow nutrient release of NPK+B might facilitate the gradual provision of nutrients for plants and holds promise for the development of new generation biochar-based fertilisers.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was funded by the project FERTICHAR (AGL2015-70393-R) of the Spanish Ministry of Economy and Competitiveness and is supported by a Margarita Salas grant under the European Union-NextGenerationEU funds.

Data availability statement
Data will be available upon request.