Plant Nutrient Availability and pH of Biochars and Their Fractions, with the Possible Use as a Component in a Growing Media

: Biochar has the potential to be used as a growing media component, and therefore plays a role in reducing peat usage. It has unique properties apart from the ability to sequester carbon. Here we investigated the nutrient contents of four commercial biochars and their fractions. The biochars’ feedstock was wood waste, except for one with paper ﬁbres and husk. The ﬁne or ﬁner fractions in wood waste biochars contained higher levels of nutrients that were available to plants. The coarse fraction of the biochar derived from husk and paper ﬁbre feedstock had a higher level of total N, P and K in contrast to the other three biochars. The pH of the ﬁner fraction (pH of 9.08) was also higher compared with coarse fraction (pH of 8.71). It is important that when biochar a is used as a component of a peat based growing media, particle size information should be provided, as fractions from the same biochar can have di ﬀ erent levels of total extractable nutrients and pH levels. If biochar is used to replace or reduce lime application rates of a peat-biochar mixtures, one must take into account the levels of total and extractable Ca and Mg levels, as these can vary. The variation of these elements was not only between biochars’ feedstocks, even at similar pH-values, but within di ﬀ erent fractions in the same biochar. We concluded that biochars should be characterized from the feedstock as well as from the particle size aspect, as it could have a profound e ﬀ ect on nutrient availability of Ca and Mg. This could lead to nutrient imbalances in cultivating plants on substrate mixtures. In addition to nutrient ratios, the suitable pH-level for a given grown species should be adjusted.


Introduction
Biochar is an organic carbon-rich solid by-product, which is gaining great interest in research for its utilization under the environmental and agricultural management [1][2][3]. Therefore, in addition to the common use as soil amendment material, biochar is being explored in terms of use for soil remediation [4][5][6][7], water filtration [8] and soilless substrates [9,10]. Recently there has been great interest in the use of biochar as a bioresource and growing media material [11][12][13][14]. Biochar and hydrothermal carbonization might play more important roles as constituents of growing media [15]. find differences in pH and nutrient contents in the different fractions from the same biochar. This is in contrast to peat, as no difference in pH was reported between fine, medium and coarse peat [42]. In the present study, we studied the pH, total and available macronutrients and micronutrient contents of biochar fractions from a number of commercial biochars that have the potential for use as growing media. The nutrient availability test that we used has been found to be very strongly correlated to plant nutrient uptake in a series of publications [43], and on this basis has been accepted as a European test [44]. We particularly looked at both total and extractable (available) Ca and Mg in relation to pH in the biochar fractions.

Materials and Methods
For our investigations we selected four commercial grade biochars: three from Europe and one from China. One of them derived from having cereal husk and paper fibre as a feedstock (Biochar A-DU). Three were derived from wood-based materials; namely, bamboo (Biochar B-No commercial name), wood screenings (Biochar C-Verora) and forest wood (Biochar D-Carbon Terra). Following pyrolysis by different manufactures, the biochars were chunky and had been milled to approximately less than 10 mm. The four biochars were characterized for pH [45] and electrical conductivity [46] in water extract at a 1:5 (v:v) ratio. The materials were characterised by extractable NH 4 -N and NO 3 -N; total N, K and P [44]; bulk density; and specific surface areas of the different biochars, including pore volume and diameter that were measured with a Quantachrome Autosorb-1 surface area Analyser, using N 2 (Brunauer, Emmett and Teller-BET) sorption methods [47].
Each material was sieved to three fractions as < 1 mm, 1-2 mm and 2-4 mm (see Figure A1). We took care to ensure each entire fraction that passed through the screens for the various fractions was obtained. For each particle size fraction, 20 to 30 g of biochar was obtained, dried and mixed well prior to physic-chemical analyses. They were then analysed for total macronutrients and micronutrients, as described by Chrysargyris et al. [48]. Nitrogen was analysed using Kjeldahl method (Buchi Digest Automat K 439 and Distillation Kjelflex K 366, Switzerland). Other macro and micronutrients were ashed and the ash was digested with hydrochloric acid (2N HCl). Ash extracts were analysed using atomic absorption spectroscopy (PG Instruments AA 500 FG, Leicestershire, UK). The determinations of K and Na were made using a flame photometer (Lasany Model 1832, Lasany International, Haryana, India). Results are expressed in g kg −1 and mg kg −1 for macronutrients and micronutrients, respectively. Biochars' fractions were extracted for available nutrients using BaCl 2 /DTPA extraction on volume basis according to EN 13651 [44]. BaCl 2 was used at the same concertation as the standard CaCl 2 /DTPA, EN 13651 [44]. The extracts were analysed for macro and micronutrients. The four biochar materials by three particle sizes were assessed for pH [45] in water extracts, at 1:5 (v:v) ratios, and BaCl 2 /DTPA extractable NH 4 -N and NO 3 -N [44,49]. Determination of BaCl 2 /DTPA extractable nutrients P, K, Ca, Mg, Fe, Na, Zn, Cu, Mn and B were determined in the filtered extract by inductively coupled plasma atomic emission spectrometry (Perkin Elmer ICP-OES, Waltham, MA, USA). Sulphate was measured in the extract by separating the anions by their affinity for an anion exchange resin packed in the anion separator column. The concentration of the sulphate anion was determined by measuring the conductance as it passed through conductivity cell.

Statistical Methods
Data were statistically analysed with the IBM SPSS version 22 (IBM Corp., Armonk, NY, USA) and results are expressed as means (n = 3) ± standard errors (SEs). Differences between treatment means were compared at p = 0.05 with ANOVA, followed by Duncan's multiple range test (DMRT).

Results
The characterization of the examined commercial, and an unscreened biochar, showed high pH, relatively high electrical conductivity, high extractable K and very low extractable N in all four biochars (Table 1). Extractable P was significantly (p < 0.05) higher in Biochar B compared with the other Biochars. The surface area tended to be highest with Biochar D, and lowest with Biochar B. The bulk density was highest for Biochar D. The pore volume and pore size followed the same trend as the surface area (Table 1).
The total macronutrients and micronutrients of different fractions of four biochars are given in Figures 1 and 2, respectively. The pH values and extractable contents of elements of the three fractions are presented in Figure 2. The extractable macronutrients and extractable micronutrients of three fractions of four biochars are given in Figures 3 and 4, respectively, while selected correlation analysis among some elements are presented in Figure 5.

Biochar's Total Elemental Contents and Correlations
3.1.1. Total Macronutrients N, K, Ca, Mg, P and Na The fine fraction had significantly (p < 0.05) higher levels of N (25.9%, 19.0% and 18.9%) for the Biochars B, C and D, respectively, compared with the coarse fraction (Figure 1), while the opposite was found in case of the Biochar type A. The coarse fraction of 2-4 mm of Biochar A showed 26.41 mg kg −1 N, which was 77% higher than the fine fraction, and that high N level was present only in Biochar A, which was the opposite to the other three biochars ( Figure 1). The total K levels followed the same pattern as N, except that Biochar C that showed no significant effects of particle size ( Figure 1). The biochar derived by husks and paper fibres, named Biochar A, revealed 17.50 mg kg −1 of K, which was the highest content at 2-4 mm fraction. Similar to N and K, the higher levels of total P were found in Biochar B, Biochar C and Biochar D in the finer fraction. An exception was Biochar A, which had the opposite trend with higher levels in the coarser fraction ( Figure 1).
The Ca levels were significantly higher in the fine (<1 mm) and medium (1-2 mm) fractions for the examined biochars, except for Biochar B, which had no significant differences between fractions ( Figure 1). Similarly, to Ca, the Mg levels were higher in the fractions of < 1 mm and 1-2 mm for the Biochar A, whereas in Biochar C and Biochar D, greater Mg contents were observed in their fractions < 1 mm. The total contents of Mg in Biochar B were the same among fractions ( Figure 1). No differences were evident between the fractions for the total Na content, as presented in Figure 1. bulk density was highest for Biochar D. The pore volume and pore size followed the same trend as the surface area ( Table 1). The total macronutrients and micronutrients of different fractions of four biochars are given in Figure 1 and Figure 2, respectively. The pH values and extractable contents of elements of the three fractions are presented in Figure 2. The extractable macronutrients and extractable micronutrients of three fractions of four biochars are given in Figure 3 and Figure 4, respectively, while selected correlation analysis among some elements are presented in Figure 5.

Biochar's Total Elemental Contents and Correlations
3.1.1. Total Macronutrients N, K, Ca, Mg, P and Na The fine fraction had significantly (p < 0.05) higher levels of N (25.9%, 19.0% and 18.9%) for the Biochars B, C and D, respectively, compared with the coarse fraction ( Figure 1), while the opposite was found in case of the Biochar type A. The coarse fraction of 2-4 mm of Biochar A showed 26.41 mg kg −1 N, which was 77% higher than the fine fraction, and that high N level was present only in Biochar A, which was the opposite to the other three biochars ( Figure 1). The total K levels followed the same pattern as N, except that Biochar C that showed no significant effects of particle size ( Figure  1). The biochar derived by husks and paper fibres, named Biochar A, revealed 17.50 mg kg −1 of K, which was the highest content at 2-4 mm fraction. Similar to N and K, the higher levels of total P were found in Biochar B, Biochar C and Biochar D in the finer fraction. An exception was Biochar A, which had the opposite trend with higher levels in the coarser fraction ( Figure 1).
The Ca levels were significantly higher in the fine (< 1 mm) and medium (1-2 mm) fractions for the examined biochars, except for Biochar B, which had no significant differences between fractions ( Figure 1). Similarly, to Ca, the Mg levels were higher in the fractions of < 1 mm and 1-2 mm for the Biochar A, whereas in Biochar C and Biochar D, greater Mg contents were observed in their fractions < 1 mm. The total contents of Mg in Biochar B were the same among fractions ( Figure 1). No differences were evident between the fractions for the total Na content, as presented in Figure 1.

Total Micronutrients Cu, Mn and Zn
Higher levels of total Cu were found in the finer fractions of two biochars-A and D. There was a clear-cut trend with the fine fraction with higher levels of Zn for the examined biochars ( Figure 2). This was similar to the results that we got with extractable Zn (Figure 4). Only in Biochar C, the trend of higher levels of Mn in the fine fraction was present (Figure 2), and the results did not reflect in extractable Mn (Figure 4).

Total Micronutrients Cu, Mn and Zn
Higher levels of total Cu were found in the finer fractions of two biochars-A and D. There was a clear-cut trend with the fine fraction with higher levels of Zn for the examined biochars ( Figure 2). This was similar to the results that we got with extractable Zn (Figure 4). Only in Biochar C, the trend of higher levels of Mn in the fine fraction was present ( Figure 2), and the results did not reflect in extractable Mn ( Figure 4).

Total Micronutrients Cu, Mn and Zn
Higher levels of total Cu were found in the finer fractions of two biochars-A and D. There was a clear-cut trend with the fine fraction with higher levels of Zn for the examined biochars ( Figure 2). This was similar to the results that we got with extractable Zn (Figure 4). Only in Biochar C, the trend of higher levels of Mn in the fine fraction was present (Figure 2), and the results did not reflect in extractable Mn ( Figure 4).  The effects of biochar (A, B, C, D) particle size on total micronutrients. Non-significant differences are indicated by ns, while significant differences (p < 0.05) among particle size for each Biochar are indicated by different Latin letters according to Duncan's Multiple Range test. Error bars show standard errors-SE. Table 1. Characterization of commercial biochars as is (pre-screening). Values are means ± standard errors of two measurements made on pH, electrical conductivity (µS cm −1 ) and minerals (mg L −1 ). Means values followed by the same letter do not differ significantly at p < 0.05.

Biochars' Extractable Elemental Contents
3.2.1. The pH and Extractable Macronutrients N, K, P, Ca, Mg, Na and SO 4 The pH was higher in the fine fraction for all four biochars ( Figure 3). The pH levels revealed the same values for the medium (1-2 mm) and coarse (2-4 mm) fractions for Biochar A, Biochar B and Biochar D.

Biochars' Extractable Elemental Contents
3.2.1. The pH and Extractable Macronutrients N, K, P, Ca, Mg, Na and SO4 The pH was higher in the fine fraction for all four biochars ( Figure 3). The pH levels revealed the same values for the medium (1-2 mm) and coarse (2-4 mm) fractions for Biochar A, Biochar B and Biochar D. Levels of NH4-N and NO3-N were very low, with values < 2 mg L −1 (data not presented). There was a clear trend of increasing K in the fine fraction ( Figure 4). The two biochars, C and D, based on wood waste material, had a similar trend. Biochar B from paper fibre and cereal husk had higher (2535 mg L −1 ) levels of K ( Figure 2).
The finer fraction (< 1 mm) had significantly higher levels of extractable P in three (A, B and C) of the four biochars, while no significant differences were found among 1-2 mm and 2-4 mm fractions. Biochar D had very low values of extractable P in all the three fractions and there was no significant effect of biochar fraction ( Figure 4).
Regarding extractable Ca, there was also a clear significant trend of the fine fraction containing higher levels calcium, with the woody materials, Biochars C and D, being very similar. The level of calcium in the Biochar A was high (up to 1094 mg L −1 at < 1 mm), while in the Biochar B, even in the fine fraction it was very low (up to 271 mg L −1 at < 1 mm) ( Figure 4).
The fine fraction had significantly higher levels of magnesium but there was no significant difference between 1-2 mm and 2-4 mm. As with potassium the levels, in the biochar derived by bamboo (B), Mg was high, reaching 145.7 mg L −1 of extractable Mg (Figure 4).
Only in two types of the biochars, namely, A and D, did the fine fraction have significantly higher levels of Na, while biochars B and C did not differ among the fractions examined ( Figure 4). There were no significant differences between the fine and coarser fractions among the biochars in relation to extractable SO4 (Figure 4). Levels of NH 4 -N and NO 3 -N were very low, with values < 2 mg L −1 (data not presented). There was a clear trend of increasing K in the fine fraction ( Figure 4). The two biochars, C and D, based on wood waste material, had a similar trend. Biochar B from paper fibre and cereal husk had higher (2535 mg L −1 ) levels of K ( Figure 2).
The finer fraction (<1 mm) had significantly higher levels of extractable P in three (A, B and C) of the four biochars, while no significant differences were found among 1-2 mm and 2-4 mm fractions. Biochar D had very low values of extractable P in all the three fractions and there was no significant effect of biochar fraction (Figure 4).
Regarding extractable Ca, there was also a clear significant trend of the fine fraction containing higher levels calcium, with the woody materials, Biochars C and D, being very similar. The level of calcium in the Biochar A was high (up to 1094 mg L −1 at < 1 mm), while in the Biochar B, even in the fine fraction it was very low (up to 271 mg L −1 at < 1 mm) (Figure 4).
The fine fraction had significantly higher levels of magnesium but there was no significant difference between 1-2 mm and 2-4 mm. As with potassium the levels, in the biochar derived by bamboo (B), Mg was high, reaching 145.7 mg L −1 of extractable Mg (Figure 4).
Only in two types of the biochars, namely, A and D, did the fine fraction have significantly higher levels of Na, while biochars B and C did not differ among the fractions examined ( Figure 4). There were no significant differences between the fine and coarser fractions among the biochars in relation to extractable SO 4 ( Figure 4).

Extractable Micronutrients (Zn, Mn, Fe, B and Cu)
There was a clear significant trend of higher levels of Zn in the fine fraction for all four biochars, but the differences between 1-2 mm and 2-4 mm were not significant ( Figure 5). Biochars derived by bamboo revealed the higher (4.72 mg L −1 ) Zn levels in the < 1 mm fraction. Manganese levels were particularly low in Biochar A and Biochar C, while both Biochar B and Biochar D revealed high (ranged from 18.82 to 20.85 mg L −1 ) levels of Mn in their fine fractions ( Figure 5).
Higher levels of extractable Cu were found in the finer fractions of two biochars, A and D, those being the same biochars with higher levels of total Cu (Figure 2). There was no clear trend regarding the four biochars in the fractions regarding the Fe and B levels ( Figure 5).

Extractable Micronutrients (Zn, Mn, Fe, B and Cu)
There was a clear significant trend of higher levels of Zn in the fine fraction for all four biochars, but the differences between 1-2 mm and 2-4 mm were not significant ( Figure 5). Biochars derived by bamboo revealed the higher (4.72 mg L −1 ) Zn levels in the < 1 mm fraction. Manganese levels were particularly low in Biochar A and Biochar C, while both Biochar B and Biochar D revealed high (ranged from 18.82 to 20.85 mg L −1 ) levels of Mn in their fine fractions ( Figure 5).
Higher levels of extractable Cu were found in the finer fractions of two biochars, A and D, those being the same biochars with higher levels of total Cu (Figure 2). There was no clear trend regarding the four biochars in the fractions regarding the Fe and B levels ( Figure 5).

Correlations
The pH value was correlated with total Ca (r = 0.518; Figure6), and with total Mg (r = 0.364; Figure 6), respectively. However, the correlation was poor with extractable Ca (r = 0.272; Figure 6) and extractable Mg (r = 0.191; Figure 6). There was a strong correlation between total K and total N (r = 0.870) at level of p ≤ 0.01 (Figure 7). There was a very strong correlation between total Ca and total Mg (r = 0.862) at level of p ≤ 0.001 (Figure 7), indicating when Ca is high or low, the Mg content would follow a similar pattern. There was a very strong correlation (p ≤ 0.001) between extractable Ca and total Ca (r =0.711) and total Mg (r = 0.823, Figure 7). Moreover, there was a very strong correlation (p = 0.001) between extractable K and total Mg (r =0.941). There was fairly a strong correlation (p = 0.001) between extractable Zn and extractable SO4 (r =0.616), extractable Fe (r =0.602), extractable Cu (r = 0.857), respectively. Additionally, there was a strong correlation (p ≤ 0.001) between extractable Mn and extractable Zn (r = 0.795), and a fairly good correlation (r = 0.640) between extractable Mn and extractable Cu.

Correlations
The pH value was correlated with total Ca (r = 0.518; Figure 6), and with total Mg (r = 0.364; Figure 6), respectively. However, the correlation was poor with extractable Ca (r = 0.272; Figure 6) and extractable Mg (r = 0.191; Figure 6). There was a strong correlation between total K and total N (r = 0.870) at level of p ≤ 0.01 (Figure 7). There was a very strong correlation between total Ca and total Mg (r = 0.862) at level of p ≤ 0.001 (Figure 7), indicating when Ca is high or low, the Mg content would follow a similar pattern. There was a very strong correlation (p ≤ 0.001) between extractable Ca and total Ca (r = 0.711) and total Mg (r = 0.823, Figure 7). Moreover, there was a very strong correlation (p = 0.001) between extractable K and total Mg (r = 0.941). There was fairly a strong correlation (p = 0.001) between extractable Zn and extractable SO 4 (r = 0.616), extractable Fe (r = 0.602), extractable Cu (r = 0.857), respectively. Additionally, there was a strong correlation (p ≤ 0.001) between extractable Mn and extractable Zn (r = 0.795), and a fairly good correlation (r = 0.640) between extractable Mn and extractable Cu. Agronomy 2019, 9, x FOR PEER REVIEW 4 of 17 Figure 6. Correlation analysis of pH with total and extractable elements from four biochars over three fractions.

Discussion
It is well known that feedstock, pyrolysis temperature and residence time have a major effect on pH and nutrient content of a biochar [1,14,35]. We investigated four commercial biochars and found a high variation between them with regard to total nutrients. This variation can be partially explained on the basis of different processing conditions and differences in feedstock. However, due to the propriety nature of the materials used in the current study, exact information on processing and feedstocks are lacking. In the present study, the levels of total N, P and K were in the same range as previously reported [20,28,32,49]. Biochar A was an exception, having particularly high nutrient levels. The feedstock of this material was different from the other three, as it was based on paper fibre and cereal husk.
The finer fraction has higher levels of total macronutrients, and to a lesser extent, micronutrients, for the three-wood based biochar. The percentage of finer fraction of a biochar is of importance through the dependence of nutrients on the particle size. Regarding the biochars used in the present study, their particle sizes have been investigated previously; they have fine particles less than 1 mm in the range of 23% to 64% [30]. Biochar B in that study had 49.5% less than 1 mm. Other authors who investigated biochars derived from woody material, reported levels of less than 1 mm of 40%-81% [29], 21%-79% [37], 38%-73% [20] and 2%-98% [38]. It is most likely that the partial removal of the fine fraction (< 1 mm) could make it suitable as a component of a growing media. This could be explained on the basis that different particle size of feedstock may lead to differential heating during pyrolysis and the finer fraction could have been more carbonized than the coarser fraction. Kloss et al. [35] have shown that the levels of cations Ca, Mg and K increase with greater carbonization. They, however, found that total N levels fall as carbonization increases. However, other authors have found that N levels can increase with carbonization [41,[50][51][52].
He et al. [41] studied the chemical properties of biochar fractions of 2-5 mm, 1-2 mm and 0.5-1 mm but also seven fractions below 0.5 mm made from pine wood. In the three fractions, which were somewhat similar to our fractions, they found an increase in total N, Ca and Mg contents in the finer fraction. The authors argued that longitudinal and transverse heterogeneity of biochar and the dominant cleavage during the preparation process may be responsible for the significant differences in properties initiated by particle size. In our investigation, N, P and K of Biochar A derived from cereal husk and paper fibre, were higher in the coarser fraction.
Generally, all the extractable nutrients were higher in the fine fraction. The fact that the availability of nutrients based on extraction with BaCl2/DTPA increases as the fraction size decreases can be explained to some extent, as it is understandable, due to the greater surface area of the fine material that allows the extractant to extract more nutrients. It may also be due to higher bulk density.
In the present study, we showed that particle size of the same biochar can have different levels on pH and total and extractable macro and micro nutrients in the different fractions (i.e., higher in the fine fraction) in the context of its use as a component of a growing media. This would ultimately affect the acidity and macro and some micronutrient content and availability of peat-biochar growing

Discussion
It is well known that feedstock, pyrolysis temperature and residence time have a major effect on pH and nutrient content of a biochar [1,14,35]. We investigated four commercial biochars and found a high variation between them with regard to total nutrients. This variation can be partially explained on the basis of different processing conditions and differences in feedstock. However, due to the propriety nature of the materials used in the current study, exact information on processing and feedstocks are lacking. In the present study, the levels of total N, P and K were in the same range as previously reported [20,28,32,49]. Biochar A was an exception, having particularly high nutrient levels. The feedstock of this material was different from the other three, as it was based on paper fibre and cereal husk.
The finer fraction has higher levels of total macronutrients, and to a lesser extent, micronutrients, for the three-wood based biochar. The percentage of finer fraction of a biochar is of importance through the dependence of nutrients on the particle size. Regarding the biochars used in the present study, their particle sizes have been investigated previously; they have fine particles less than 1 mm in the range of 23% to 64% [30]. Biochar B in that study had 49.5% less than 1 mm. Other authors who investigated biochars derived from woody material, reported levels of less than 1 mm of 40%-81% [29], 21%-79% [37], 38%-73% [20] and 2%-98% [38]. It is most likely that the partial removal of the fine fraction (<1 mm) could make it suitable as a component of a growing media. This could be explained on the basis that different particle size of feedstock may lead to differential heating during pyrolysis and the finer fraction could have been more carbonized than the coarser fraction. Kloss et al. [35] have shown that the levels of cations Ca, Mg and K increase with greater carbonization. They, however, found that total N levels fall as carbonization increases. However, other authors have found that N levels can increase with carbonization [41,[50][51][52].
He et al. [41] studied the chemical properties of biochar fractions of 2-5 mm, 1-2 mm and 0.5-1 mm but also seven fractions below 0.5 mm made from pine wood. In the three fractions, which were somewhat similar to our fractions, they found an increase in total N, Ca and Mg contents in the finer fraction. The authors argued that longitudinal and transverse heterogeneity of biochar and the dominant cleavage during the preparation process may be responsible for the significant differences in properties initiated by particle size. In our investigation, N, P and K of Biochar A derived from cereal husk and paper fibre, were higher in the coarser fraction.
Generally, all the extractable nutrients were higher in the fine fraction. The fact that the availability of nutrients based on extraction with BaCl 2 /DTPA increases as the fraction size decreases can be explained to some extent, as it is understandable, due to the greater surface area of the fine material that allows the extractant to extract more nutrients. It may also be due to higher bulk density.
In the present study, we showed that particle size of the same biochar can have different levels on pH and total and extractable macro and micro nutrients in the different fractions (i.e., higher in the fine fraction) in the context of its use as a component of a growing media. This would ultimately affect the acidity and macro and some micronutrient content and availability of peat-biochar growing media due to fact that each fraction, particularly the fine fraction has significantly different levels of nutrients and acidity.
Numerous studies have been published showing the beneficial effects on crops due to the addition of biochar to peat; however, only in a limited number of studies has the particle size of the biochar been presented. There have been a few publications that showed the particle size of the biochar has a marked effect on physical properties of the growing media [27,32,38]. Our results indicate that in the use of biochars in growing media, particle size should be taken into serious consideration not only for physical properties, but just as importantly, chemical properties. This information should be essential when formulating nutrients/lime addition and subsequent nutrient management during cropping.
The higher nutrient availability from the four biochars in fine fractions as found in the present study, are in agreement with previous reports [39]. However, Angst and Sohi [39] examined the effect of particle size in context of a soil application of biochar, which is different from the growing media conditions. They found that the availability and the release of limited number of nutrients, namely, magnesium, potassium and phosphorus, were affected by particle size with the fine fraction material showing greater water-soluble contents of these nutrients, and better release. The fraction sizes they looked at were smaller than ours, as they studied fractions of 0.15 µm to 0.60 µm, 0.60 µm to 1.8 mm, 1.8 mm to 4 mm and > 4 mm. For K and Mg, the differences between the finer fractions were relatively small but the differences were greater between the fine fractions, that was, < 4 mm, and > 4 mm. For P the differences in the finer fraction were even less and the difference between the finer fraction and > 4 mm was less than for K and Mg [39]. The extraction solution might affect the nutrients extracted derived by different fractions of biochar. Angst and Sohi [39] carried out six extractions and for K there was little change but for Mg and P the amount and the difference between the fractions increased. They used water as an extractant, unlike our extractant which had a cation in it. One is aware that biochar has a high cation exchange capacity and also has the ability to bind nitrate nitrogen, and as such a stronger extraction, which includes a cation and contains DTPA, which may give a bigger difference between the fractions and perhaps better picture on plant availability in growing media. In any case the extractant used here has been strongly correlated the plant uptake of some macro nutrients and micronutrients in a growing media situation [30,43].
Due to the nature of biochar coming from woodchips, thereby the need for crushing and relatively easy break down of particles, variation in particle size is likely to occur between batches, and this could possibly have reflected variation in physical and chemical properties of growing media, when biochar is added to peat. Our findings point to the need to be vigilant about particle size for each batch. However, the use of dolomitic lime was added not only to neutralize the acidity of peat but also to supply essential nutrients such as calcium and magnesium. The pH values in biochar can vary according to the pyrolysis temperature, as biochars produced from sewage sludge at low temperatures (300 and 400 • C) were acidic, whereas at high temperatures (500 and 600 • C), they were alkaline [53].
The levels for the Cu, Zn and Mn were in the same range as found by Altland and Locke [28] and Bedussi et al. [32]. The average extractable K, Ca and Mg as a percentage (%) of the total K, Ca and Mg for Biochar A, Biochar B, Biochar C and Biochar D were as follows: for K-Approximately 18%, 34%, 31% and 48%, respectively; for Ca-24%,16%,16% and 19%, respectively; and for Mg-14%, 9%, 11% and 28%, respectively. Angst and Sohi [39] found greater water extractability of K compared to Mg, probably reflecting the feedstock.
It is well known that most biochars have high pH-values [31], and peats have very low pH-values, ranging from 3 to 4 [54]. Therefore, peat requires additional lime, e.g., in form of a dolomitic lime, to adjust the pH to values to around 5.5, while other substrates, e.g., wood fibres, do not need it [55]. A number of researchers suggested that biochars could be used to replace lime application [30,34]. However, surprisingly, biochar-peat blends can contain up to 80% biochar without raising the pH above 7 [34]. Unfortunately, these studies do not give any information on particle size; one can surmise it was a coarse material. Zaccheo et al. [33] reported that particle size can have an effect in increasing the pH of growing medium. Furthermore, the same authors found that the finer fraction, 0-3.3 mm, was more effective at raising the pH of a peat-biochar mixture [33]. This is in agreement with our findings, as the finer fraction had a higher pH and a higher content on Ca and Mg. However, in the same study of Zaccheo et al. [33], the premise that dolomitic lime or ordinary lime is added not only to increase the pH but also supply Ca and Mg, which are essential plant nutrients with very low levels in peat, was not taken into consideration [32]. In our study, despite the minor differences in pH, the levels of extractable Ca varied enormously between Biochar A with very high extractable Ca and Biochar B with very low extractable Ca. In addition, most biochars contain high levels of K. This could depress and reduce Mg uptake by plants, due to the cation antagonism among them. Increased K-rates, varied from 4.61-5.39 g kg −1 by different biochars, were also reported by Gasco et al. [31], being in agreement with the findings of the present study.
Although generally there is good relationship between pH and total calcium and magnesium content, in our case the relationship would not be good enough to predict Ca and Mg levels. This is related to the poorer relationship of pH and available and extractable Ca and Mg content in biochars. The example of a good relationship or otherwise between Ca and Mg is important, as in peat dolomitic lime is added not only to adjust pH but to supply Ca and Mg. Extractable levels of Ca are generally very low in peat. The good relationship found in our study indicates that biochar is a similar material to dolomitic lime rather than calcitic lime. The good relationship between extractable Ca and total Ca gives confidence on the supply of Ca, whereas a poor relationship would invoke doubt, either from total Ca or extractable Ca: extractable Ca gives an indication of short term availability, while total Ca gives an indication of long term availability. P levels are too low to have any significance as a biochar is added at low rates with peat. However, K would be of importance. For instance, data from Bedussi et al. [32] shows that the finer biochar from poplar using pyrogasification has higher contents of total N, P, Ca, Mg and many micronutrients. The poplar biochar had higher contents of total Ca and Mg, despite having a lower pH. In addition, Bedussi et al. [32] measured water-soluble N, P and K which were higher for these nutrients, but found lower levels of Ca and no difference for Mg. Therefore, a blanket recommendation or a suggestion that biochar can replace lime in peat-based media needs serious rethinking. We feel it is essential that the total Ca and total Mg and available Ca and available Mg, should be analysed in biochar and taken into account when recommending lime application rates if biochar has been added to peat. This is in agreement with our previous works. As the biochar component increases in a peat/biochar-mixtures, this is affecting nutrient levels in plants; for example, the levels of Mg content in a leaf dropped due to the antagonistic effect of K on Mg [30]. This is primarily due to the excessive level of K, which is a feature of most but not all biochars. Additionally, in our resent study, K, P and Cu accumulation and Mg deficiency in cabbage leaves were related to the biochar presence and feedstock [10]. In that study, the biochar's feedstock, rate and the addition of fertilizers could affect the cabbage seedling performance.

Conclusions
The results of the present study have shown that within the same batch of biochar, total macronutrient and micronutrient levels are different in different fraction sizes. One can clearly state from our data that the use of a biochar when added to peat and when it is considered as a substitute for lime, due to its high pH, may not be valid in many cases, as Ca levels in biochars do not always equate to a pH value. Extractable Ca and perhaps total Ca in biochars need to be considered when advising biochar rates and lime rates to peat. Moreover, the particle size of biochar is very important regarding electrical conductivity and nutrient availability, especially K. This has far reaching consequences regarding nutrient imbalances in growing media and in formulating a base dressing and a liquid feeding programme. To our knowledge, this is the first time that it has been reported for biochar in context of growing media. The variation in total nutrients in biochar fractions of the same biochar was an unexpected result for a growing media. Our recommendation is to partially take out the biochar fraction < 1 mm if the EC levels and nutrients are particularly high and use it for other purposes. The reduction of EC and extractable K would, thus, make the biochar suitable as a component of a growing medium.