- Email: [email protected]
Biochar made from low density wood has greater plant available water than biochar made from high density wood
Journal Pre-proof Biochar made from low density wood has greater plant available water than biochar made from high density wood
Joerg Werdin, Tim D. Fletcher, John P. Rayner, Nicholas S.G. Williams, Claire Farrell PII:
S0048-9697(19)35851-6
DOI:
https://doi.org/10.1016/j.scitotenv.2019.135856
Reference:
STOTEN 135856
To appear in:
Science of the Total Environment
Received date:
17 September 2019
Revised date:
11 November 2019
Accepted date:
28 November 2019
Please cite this article as: J. Werdin, T.D. Fletcher, J.P. Rayner, et al., Biochar made from low density wood has greater plant available water than biochar made from high density wood, Science of the Total Environment (2018), https://doi.org/10.1016/ j.scitotenv.2019.135856
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2018 Published by Elsevier.
Journal Pre-proof Title: Biochar made from low density wood has greater plant available water than biochar made from high density wood Authors: Joerg Werdin*, Tim D. Fletcher, John P. Rayner, Nicholas S.G. Williams and Claire Farrell Authors affiliation: School of Ecosystem and Forest Sciences, Faculty of Science, The University of Melbourne, 500 Yarra Boulevard, Richmond 3121, Victoria, Australia *
Corresponding author: E-mail address: [email protected]
of
Abstract
ro
Soil water limitations often restrict plant growth in unirrigated agricultural, forestry and urban
-p
systems. Biochar amendment to soils can increase water retention, but not all of this additional
re
water is necessarily available to plants. Differences in the effectiveness of biochar in ameliorating
lP
soil water limitations may be a result of differences in feedstock cell structure. Previous research has shown that feedstock cell structure influences the pore structure of biochar and consequently the
na
volume available for water storage. The availability of this water for plant uptake will be determined by biochar pore diameters, given its role in determining capillary forces which plants must overcome
Jo ur
to access pore water. Therefore, we hypothesised that differences in hardwood feedstock cell structure would result in differences in the plant available water holding capacity of biochar. Before pyrolysis, we measured the wood morphology of 18 Eucalyptus species on three replicates of equal age on a gradient of wood density (572 – 960 kg m-3). Wood samples were then pyrolysed (550 °C) and the resulting biochars were sieved and their particle size distribution was standardised before their physical properties, including water holding capacity, plant available water and bulk density were measured. Our results show that biochar made from lower density eucalypt wood had up to 35% greater water holding capacity and up to 45% greater plant available water than biochar made from higher density eucalypt wood. Further, feedstock wood density related well to fibre cell wall thickness and fibre lumen diameter. Therefore, wood density could be used as a proxy for wood cell
1
Journal Pre-proof structure, which can in turn be used to predict plant available water in biochar. The simple measure of feedstock wood density can inform feedstock choices for producing biochars with greater plant available water, optimal for the use as soil amendment in water limited environments. Keywords biochar pore structure; hardwood; plant available water; water holding capacity; wood cell structure; wood density
of
1. Introduction
ro
Biochars are highly porous solids produced by the thermochemical conversion of biomass under
-p
oxygen-depleted conditions in a process known as pyrolysis; which has been traditionally used to
re
produce charcoal (Lehmann and Joseph, 2015). They are considered a useful soil amendment for improving soil water retention and plant growth in rainfed systems in agricultural (Jeffery et al.,
lP
2011; Obia et al., 2016), forestry (Tryon, 1948) and urban landscapes (Cao et al., 2014). As well as
na
improving water retention in soils (Omondi et al., 2016), biochars have also been shown to increase crop yields by increasing nutrient retention (Major et al., 2010), improving soil structure (Lim et al.,
Jo ur
2016; Obia et al., 2016) and microbial activity (Wang et al., 2016) and reducing the bioavailability of heavy metals (O'Connor et al., 2018) whilst adding a stable source of soil carbon for sequestration (Lehmann, 2007). However, the effects of biochar as a soil amendment can be highly variable due to (i) differences in feedstock materials, (ii) their interaction with different soil types and (iii) differences in pyrolysis conditions (Atkinson, 2018; Jeffery et al., 2011; Li et al., 2019; Rasa et al., 2018; Ronsse et al., 2013; Sohi et al., 2010; Weber and Quicker, 2018).
While biochar can be made from a wide range of organic materials including crop residues, biosolids or food waste, the most commonly used feedstocks for biochar production worldwide are woodbased (Jirka and Tomlinson, 2015). When wood feedstocks are pyrolysed, their chemical and physical make up changes, influencing their water retention properties, water uptake behaviour 2
Journal Pre-proof (degrees of hydrophobicity) and longevity. Chemically, wood is mainly composed of hemicellulose, cellulose and lignin (Walker, 2006). During pyrolysis, these compounds decompose, releasing volatile liquid and gaseous fractions and leaving behind solid aromatic carbon compounds (Zeriouh and Belkbir, 1995). This results in wood biochars generally having a greater than 80% (w/w) carbon content when pyrolysed above 500 °C (Weber and Quicker, 2018). The aromatic carbon structure of biochar contributes to its long-term stability. This makes biochar a better soil ammendment with sustained improvement for water retention than other organic or synthetic water retention
of
amendments. For instance, compost or hydrogels decompose orders of magnitude faster than
ro
biochar (Al-Harbi et al., 1999; Bolan et al., 2012; Singh et al., 2012). Another important biochar
-p
property governing its water uptake behaviour is the degree of hydrophobicity. Generally, low pyrolysis temperatures (< 500 °C) can result in hydrophobic, water repellent biochar (Zornoza et al.,
re
2016). This is caused by incomplete carbonisation of chemical wood components, resulting in
lP
remaining aliphatic functional groups on the biochar surface which repel water (Das and Sarmah, 2015; Kinney et al., 2012). Therefore, greater pyrolysis temperature generally results in reduced
Jo ur
of biochar.
na
hydrophobicity of biochar, enabling greater soil water uptake and increased water holding capacity
The cellular wood structure remains largely intact after pyrolysis and forms the macroporous structure of biochar (Baltrėnas et al., 2015; Ehrburger et al., 1982; Gray et al., 2014; Hyväluoma et al., 2018; Wildman and Derbyshire, 1991). Biochar macropores are very important for water retention, as they make up the main pore volume for water storage (Brewer et al., 2014; Gray et al., 2014; Hyväluoma et al., 2018; Lu and Zong, 2018). This has been demonstrated by Zhang and You (2013) who showed a strong positive correlation between macropore size and water holding capacity (WHC) for biochars produced from poplar (Populus davidiana) and pine (Pinus sylvestris var. mongolica) wood. The biochar made from poplar wood (hardwood) had greater macropore diameters (1 – 40 µm) and greater WHC (69 – 72%) when compared to the biochar made from pine 3
Journal Pre-proof wood (softwood), with smaller macropore diameters (1 – 10 µm) having a WHC of 29%. The authors concluded that biochars with greater macropore diameters and therefore greater pore volumes will lead to increases in water retention and are therefore best suited for improving soils in waterlimited environments. The wood of hardwood species (generally non-coniferous) is made up of fibre, vessel and parenchyma cells (Wilkes, 1988). Fibre cells can be described as hollow tubes with thickened cell walls (Downes et al., 1997) and as they make up a large proportion of the wood structure, fibre properties greatly influence wood density (Abruzzi et al., 2013; Zbonak et al., 2007).
of
In general, higher density hardwoods have greater fibre wall thickness and smaller fibre diameters
ro
than lower density hardwoods (Watson and Dadswell, 1961). Therefore, it is likely that biochar
-p
produced from lower density hardwoods will retain more water.
re
However, water holding capacity (WHC) does not always reflect how much water is accessible for
lP
plant growth (Cao et al., 2014; Masiello et al., 2015). A better measure is plant available water (PAW), which accounts for the amount of water that is held in pores at tensions which can be
na
extracted by plants. PAW, also referred to as available water content (AWC), is mainly determined
Jo ur
by macropore size, with water stored in pores with diameters between 0.2 to 30 µm being plant available, while water stored in smaller pores is held too tightly for plant uptake, mainly due to capillary and adsorption forces (Atkinson, 2018; Cassel and Nielsen, 1986). The total amount of water in pores with diameters smaller than 0.2 µm is held with a negative pressure greater than 1500 KPa which supersedes most plants ability to extract it. It is commonly referred to as permanent wilting point (PWP), whereas water in pores with diameters greater than 30 µm drains under gravity and is only briefly available for plant uptake after a rain event (Atkinson, 2018). The number of pores in biochar in the relevant diameter range for PAW is largely influenced by feedstock type (Lu and Zong, 2018). Therefore, the number of pores in the size range for maximising PAW in biochar should relate to the cell structure of the woody feedstock material, and biochar produced from lower density hardwood should hold more PAW than biochar made from high density hardwood. 4
Journal Pre-proof
While there is an extensive body of research on how biochar influences water retention in soils (Abel et al., 2013; de Melo Carvalho et al., 2014; Hardie et al., 2013; Tryon, 1948), few studies have investigated how feedstock fibre properties and wood density influences water retention and availability of biochar (Hyväluoma et al., 2018; Rasa et al., 2018). Given this gap, in this study we investigated how feedstock wood density was related to biochar WHC and PAW. We hypothesized that lower density hardwood feedstock would result in biochar with greater overall WHC and more
of
of this water would be plant available (PAW) when produced under the same conditions. Our
ro
research objectives were: (i) to investigate if wood density can be used as a low cost and easy to
-p
measure proxy for feedstock cell structure (ii) to determine how feedstock cell structure influences WHC and PAW of the resulting biochar and (iii) to assess which proportion of the overall water
re
retention is plant available. The water retention measurements (WHC and PAW) in our study were
lP
performed on pure biochars without being incorporated into a soil to avoid complications due to biochar-substrate interactions. Our findings need to be validated in future studies with amended
Jo ur
na
soils.
2. Materials and Methods
To determine how hardwood feedstocks with different wood densities, influence biochar water retention and availability, we compared biochars made from 18 Eucalyptus species under the same pyrolysis conditions. The 18 Eucalyptus species represent a gradient of wood density (Table 1) from low density e.g. E. pauciflora (572 kg m-3) to high density e.g. E. polybraktea (960 kg m-3). 2.1. Wood feedstock and sample preparation The wood samples were taken from 18 Eucalyptus species (Table 1), grown under equal environmental conditions in a field experiment at the Burnley Campus of the University of
5
Journal Pre-proof Melbourne in November 2015. All trees were 4.25 years of age (51 months) and a minimum of three replicate samples were taken per species (1 sample per tree).
The samples were taken from approximately 150 mm long cross-sections of each trunk. To ensure comparability among samples in regard to their wood density, all samples were cut 200 mm above the soil surface as wood density tends to decrease with increasing distance from the base of the stem (Downes et al., 1997) (Figure 1). Stem diameters ranged from 52 mm (E. tricarpa) to 229 mm
of
(E. viminalis). The wood samples were air-dried in a polytunnel for five months before being
ro
debarked. From three air-dried wood samples per species, 7 mm long radial strips were cut from
-p
pith to bark with a standard width of 2 mm (Figure 1) and one of their transverse surfaces was
2.2. Wood analysis
lP
method described in Evans et al. (2017).
re
polished with sanding paper (coarse 400 – fine 1500 grit) for analysis on Silviscan3 according to the
na
Prepared wood sections were analysed using Silviscan3, a wood analysing instrument combining optical microscopy, X-ray densitometry and automated image analysis to measure wood cell
Jo ur
properties (Figure 1) (Lawrence and Woo, 2005). The wood cell properties, fibre cell wall thickness and radial and tangential fibre lumen diameter were obtained in 25 μm radial increments for each of the three replicates per species and automatically analysed by integrated image analysis software on Silviscan3 (only one sample could be analysed for E. dives and two samples for E. macrohyncha and E. radiata due to internal ruptures and extensive tension wood which prohibited meaningful analysis). Detailed information on within- and between species variability are provided in the supplementary information (Figure 6). Values for each sample (in 25 μm sections) were then averaged over the radial wood profile from pith to bark before a mean of the three samples of each species was calculated. At each end of the wood samples, 5 mm was excluded from the analysis to avoid errors caused from cracks and ruptures in the wood due to post-harvest drying.
6
Journal Pre-proof
Wood density was determined by water displacement for a minimum of five replicates per species, prior to pyrolysis (Figure 1) and (Figure 6 supplementary information). Debarked and air-dried wood samples were weighed before being briefly submerged in a water-filled PVC cylinder (diameter 150 mm) equipped with a water-level gauge. The change in water level was recorded and the air-dry wood density calculated: 𝑤𝑒𝑖𝑔ℎ𝑡 𝑎𝑖𝑟−𝑑𝑟𝑦 𝜋∗𝑟 2 ∗𝛥ℎ
∗ 1000
(1)
of
𝑤𝑜𝑜𝑑 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 =
ro
where wood density is the air-dry wood density in kg m-3, weight air-dry is the air-dried weight in g, r
-p
is the radius of the PVC cylinder in cm and 𝛥ℎ is the change in water level in cm.
re
2.3. Biochar pyrolysis
lP
Prior to pyrolysis, wood samples (excluding sections taken for wood analysis on Silviscan3 as described in section 2.1) were labelled and individually placed on stainless-steel trays. The wood was
na
then pyrolysed (January 2017) in a mobile charmaker (Charmaker MPP20, Earth Systems Pty., Melbourne, Australia), using slow-batch pyrolysis at 550 °C (± 50 °C) and a 4-hr residence time. Each
Jo ur
of the species-specific biochars was then air-dried for three months in a polytunnel before being crushed manually and passed through a nest of sieves of decreasing aperture (12.5 mm, 10 mm, 6.3 mm, 4 mm, 2 mm, 1 mm). Particles greater then 12.5 mm in diameter were discarded. Each of the particle size fractions for each sample were then weighed and their volume determined using a measuring cylinder at a standardised compaction to determine the bulk density (g cm-3) of each particle size fraction. The biochar samples were then remixed to ensure that the particle size fractions were identical between each of the 18 biochar types before testing the biochar properties (Table 2). 2.4. Biochar properties
7
Journal Pre-proof Biochar bulk density and water holding capacity (WHC) were determined on three samples for each of the 18 eucalypt species. Biochar dry bulk density was determined according to the Australian Standard for Potting Mixes (AS 3743—2003) as the weight of oven dried (105 °C until weight equilibrium) material divided by its volume using standardised measuring cylinders at a standardised compaction (free fall from 5 cm height repeated five times).
WHC was determined by using a tension table (Topp et al., 2008). Known quantities of each biochar
of
type were submerged in deionised water for 24 hr and then placed onto a tension table at -10 KPa
ro
suction (field capacity) until constant weight. The samples were weighed and oven dried at 105 °C
𝑊𝐹𝐶 −𝑊𝑑𝑟𝑦 𝑉
∗ 100
(2)
re
𝑊𝐻𝐶 =
-p
until constant weight. The volumetric water content was calculated:
lP
Where WHC is the water holding capacity in % v/v, WFC is the sample weight at field capacity (-
na
10KPa) in g, Wdry is the oven-dry sample weight in g and V is the sample volume in cm3.
Jo ur
To determine PAW, moisture release curves for each of the 18 eucalypt biochars were obtained using the filter paper method described by Greacen et al. (1989) and modified by Farrell et al. (2013). For each biochar, samples were oven dried at 105 °C to constant weight and divided into 10 equal portions (by weight), before deionised water was added to each sample to obtain samples of 6 – 33% of their WHC, at 3% increments. Samples were packed into glass jars with three filter papers (Whatman No. 42) placed between half the sample volume, so that one filter paper was sandwiched between the other two to reduce weighing errors due to adhering biochar particles. The jars were then sealed and placed in a temperature-controlled environment to equilibrate for seven days. The gravimetric water content of the middle filter paper of each sample was determined, and the corresponding matric suction was calculated according to Greacen et al. (1989). Moisture release
8
Journal Pre-proof curves were then plotted and the volumetric water content of biochar at -1.5 MPa calculated from the intersect of the exponential model. Biochar PAW was calculated:
PAW = WHC – PWP
(3)
Where PAW is the plant available water in % v/v, WHC is the water holding capacity in % v/v and PWP is the water content at permanent wilting point (-1.5 MPa) in % v/v. 2.5. Statistical analyses
of
All statistical analyses were performed in R version 3.4.1 (R Core Team, 2017). Correlations among all
ro
measured variables were determined using Pearson’s Correlations to determine the variables best
-p
explaining differences in WHC and PAW. Linear regressions were performed to relate (i) feedstock
re
fibre wall thickness and radial fibre lumen diameter with wood density; (ii) wood density with biochar WHC and PAW and (iii) biochar WHC with biochar PAW. Variability among species for air-dry
lP
wood density, radial fibre lumen diameter and fibre wall thickness was analysed using one-way
0.05).
Jo ur
3. Results and Discussion
na
ANOVA and significant differences among means were determined using Tukey’s post-hoc test (p <
Overall, our study showed that eucalypt feedstock with lower wood density resulted in biochar with up to 35% greater water holding capacity (WHC) and up to 45% greater plant available water (PAW) than biochar made from higher wood density feedstock. Further, WHC was strongly related to PAW, indicating that the additional water retained in biochar produced from lower density feedstocks is also available for plant growth. As feedstock wood density was also related to feedstock cell structure, we suggest that lower density feedstock wood with greater fibre diameters and thinner fibre walls results in biochar with a greater number of pores which retain water that is subsequently available to plants. These results support our hypothesis that differences in feedstock wood density cause differences in biochar WHC and PAW.
9
Journal Pre-proof 3.1. Relationships between wood fibre properties with wood density Wood density is mainly determined by fibre wall thickness and fibre lumen diameter. Thinner cell walls and greater fibre diameter are found in lower density species, whereas thicker cell walls and smaller fibre diameters are found in higher density species (Figure 3). Variability in wood density across species was best explained by fibre wall thickness (R2 = 0.88) and to a lesser degree by radial fibre lumen diameter (R2 = 0.69). The eucalypt feedstock with the greatest average fibre wall thickness (1.96 µm across the radial diameter from pith to bark, SE± 0.03) was E. polybractea, which
of
was also the heaviest (960 kg m-3, SE± 40.53); whereas E. pauciflora had an average fibre wall
ro
thickness of 1.08 µm (SE± 0.02) and was the species with the lowest wood density (572 kg m-3, SE±
-p
18.82). The range in average radial fibre diameter was small among the investigated feedstock
re
species with E. regnans (11.25 µm, SE± 0.31) wood having the greatest average diameter and E. microcarpa (9.72 µm, SE± 0.06) the smallest. This was well aligned with their wood density, as these
lP
species were among the lightest (599 kg m-3, SE± 24.36) and heaviest (895 kg m-3, SE± 10.13) species
na
investigated in our study. Ziemińska et al. (2013) reported similar relationships between wood density and fibre wall thickness in a study of the branch-wood structure of 24 Australian
Jo ur
angiosperms across different genera. Our study shows that wood density can be used as a proxy for wood cell structure. This has important implications for the rapid assessment of wood properties for biochar feedstock selection, as air-dry wood density can be easily determined, whereas the direct measurement of cell properties is time and cost intensive and requires elaborate research equipment. 3.2. Relationships between feedstock wood density and biochar WHC and PAW Since wood density is mainly determined by fibre cell wall thickness and fibre lumen diameter, it can be regarded as a predictor of porosity. The cell walls form the solid structure, whereas the lumen diameters are open void spaces. In other words, wood density is a measure of the cell wall material per volume of wood (Phillips, 1941). These wood cell properties are largely retained in the pore
10
Journal Pre-proof structure of biochar with the cell walls forming the pore boundaries within biochar and the cell lumen forming the biochar pore space (Gray et al., 2014; Hyväluoma et al., 2018). Therefore, lighter wood with a greater proportion of open void space (cell lumen) and a lower proportion of solid material (cell wall) results in biochar with a greater proportion of open void space (pore volume) and a lower proportion of solid material (pore boundaries). When biochar is exposed to water, the airfilled biochar pores fill with water and the total amount of water retained in a volume of biochar is the WHC. In our study, wood density ranged between 572 – 960 kg m-3. Biochar produced from
of
eucalypts with lower density wood had up to 35% greater WHC than biochar made from higher
ro
density wood (Figure 4). However, greater overall water retention (WHC) due to greater internal
-p
porosity does not necessarily result in greater water availability for plants (PAW), as some of this
lP
the potential benefit for plant growth.
re
water might be too tightly bound for plant uptake. Therefore, PAW is a better measure to evaluate
PAW is stored in biochar pores that range between 0.2 µm – 30 µm and therefore, biochars made
na
from lower density Eucalyptus wood are likely to have a greater proportion of pores in the plant-
Jo ur
available range. Hardie et al. (2013) investigated the pore structure of an Acacia green waste biochar and showed that the median pore diameter ranged between 0.4 – 13 µm, thus most pores residual from the feedstock cell structure were in the diameter range relevant for PAW. Even though the Acacia species from which the biochar was made from was not specified, the general hardwood cell structure of acacias is similar to Eucalyptus species. A Scanning Electron Microscope (SEM) study by Kumar and Gupta (1995) that investigated the morphological changes of Acacia and Eucalyptus wood undergoing carbonisation came to the conclusion that both species underwent similar morphological changes. The biochar with the greatest amount of PAW in our study was made from Eucalyptus radiata (42.2%, SE± 0.64), which is also a relatively light wood species with a density of 689 kg m-3(SE± 25.27) (Table 1). Conversely, biochar made from E. polybractea had the lowest amount of PAW (28.6%, SE± 0.08) and the greatest wood density of all investigated species (960 kg 11
Journal Pre-proof m-3, SE± 40.53). The greater wood density of E. polybractea wood is due to thicker cell walls (1.97 µm, SE± 0.03) and smaller fibre lumen diameters (9.92 µm, SE± 0.07), when compared to E. radiata (1.12 µm; 10.83 µm), resulting in a greater proportion of solid, lignocellulose material and less open void space per volume of wood. 3.3. Relationship between biochar PAW and WHC WHC and PAW were strongly correlated in the 18 biochars made from different eucalypt feedstocks
of
under equal pyrolysis conditions (Figure 5). The difference between WHC and PAW determines the
ro
amount of water which is unavailable to plants due to the high tensions at which water is held in pores smaller than 0.2 µm (Atkinson, 2018; Cassel and Nielsen, 1986). Among the 18 biochars there
-p
was an average 6% difference between WHC and PAW, which is the amount of water unavailable for
re
plant uptake. The biochar with the greatest amount of unavailable water was E. microcarpa (8%, SE±
lP
0.53), whereas biochar made from E. pauciflora had the smallest amount of unavailable water (3%, SE± 0.45). Compared to the differences in PAW across biochar types (28.6 – 42.2%), the differences
Jo ur
na
in unavailable water were marginal.
The small differences in unavailable water among the 18 biochars are likely due to the fact that they were pyrolysed at the same temperature in one batch. Gray et al. (2014) proposed a biochar pore size classification based on the pore origin and diameter size range. Residual macropores in the range between 1-100 µm are inherited from the feedstock cell structure and pyrogenic nanopores in the nanometer range develop with increasing pyrolysis temperature above 500 °C. Since all biochars used in this study were produced at a pyrolysis temperature only marginally greater (550 °C ± 50 °C), extensive pyrogenic nanopore development is unlikely. This is supported by the relatively small amount of unavailable water per volume of biochar since water stored in these nanopores is bound to tightly for plant uptake. Therefore, the equal pyrolysis conditions used to create all the biochars tested likely resulted in similar pyrogenic nanopore development, regardless of feedstock cell 12
Journal Pre-proof structure. This supports our hypothesis that differences in WHC and PAW can be primarily attributed to differences in feedstock cell structure.
Multiple studies have highlighted that the macroporous structure of biochar (i) is inherited from the feedstock cell structure and (ii) determines its water retention (WHC and PAW) (Gray et al., 2014; Hyväluoma et al., 2018; Kumar and Gupta, 1995; Wildman and Derbyshire, 1991; Zhang and You, 2013). However, studies relating WHC and PAW of biochar with feedstock cell structure are rare,
of
which might be due to the fact that wood cell structure is heterogenous and commonly used
ro
analytical methods such as SEM techniques or more recently X-ray tomography, rely on small sample
-p
sizes (< 1 mm2) (Hyvaluoma et al., 2018; Kumar and Gupta, 1995). These small samples are usually not representative of the species-specific cell structure and mechanistic relationships between
re
feedstock cell structure and biochar pore structure cannot easily be made. For example, Hyvaluoma
lP
et al. (2018) compared the cell structure of 1 mm2 samples of dried willow wood with the pore structure of the resulting biochar using x-ray tomography images and concluded that their results
na
were not representative for willow wood in general, due to the small sample size and heterogeneity
Jo ur
of willow wood structure. Generally, wood density tends to increase from the centre of the tree (pith) to bark and from the tree base to the top, due to changing fibre properties (lumen diameters and cell wall thickness) (Downes et al., 1997). In our study, we accounted for the heterogeneity in wood fibre properties by measuring the cell structure across the full diameter from pith to bark, and by taking all wood samples at similar distance from the tree base. Furthermore, all trees were planted at the same time, grown under the same conditions and were of equal age when harvested which controls for differences caused by environmental conditions (i.e. precipitation, temperature and soil conditions). This gives us confidence in the relationships we found between wood properties and biochar properties amongst the 18 Eucalyptus species.
We investigated the water retention behaviour of pure biochars to explore if there is a mechanistic 13
Journal Pre-proof relationship between wood cell structure and resulting biochar water retention. However, it remains unknown how different levels of WHC and PAW in the pure biochar affect the water holding properties of amended soil and substrate. A metanalysis conducted by Omondi et al. (2016) based on 74 datasets reporting changes of PAW in soils after biochar amendment reported an overall increase in PAW after biochar amendment of 15%. However, in contrast other studies have reported no significant changes in PAW after biochar addition (i.e. (Hardie et al., 2013)). Therefore, the reported differences in WHC and PAW of biochar made from feedstocks with different wood
of
densities in our study need to be validated in studies with amended soils.
ro
4. Conclusions
-p
For the 18 eucalypt species evaluated in this experiment, the WHC and PAW of their resulting
re
biochar was strongly related to their feedstock wood density. Biochar produced from eucalypt
lP
species with lower wood density had greater WHC and PAW than biochar produced from higher density feedstocks. Wood density was strongly related to feedstock cell structure. Therefore, we
na
suggest that lower density wood species, with greater fibre diameters and thinner fibre walls, result in biochar with a greater proportion of pores which retain water available to plants. As biochar
Jo ur
water retention could be estimated from the simple measure of wood density, this can be used to select feedstocks to produce biochars with high plant available water, without the need for rigorous testing. However, as our results are for pure biochar, further research is needed to test whether the increased WHC and PAW of biochar produced from lower density feedstocks also translate to increased WHC and PAW of biochar amended soils in agricultural, forestry and urban applications. 5. Acknowledgements This research was funded by an Australian Research Council Linkage Grant (LP30100731) supported by Melbourne Water and the Inner Melbourne Action Plan (IMAP) group of local governments. Joerg Werdin was supported by a Research Training Program Scholarship and a Graduate Research Studentship. We thank Robert Evans for his help with the wood analysis, Adrian Morphett and 14
Journal Pre-proof Moana Quiatol from Earth Systems for their help producing the biochar, Richard Conn and Petra Katona for their help with the experiment. We thank the anonymous reviewers for their valuable comments and constructive criticism that helped to improve the quality of our manuscript.
References
ro
of
Abel S, Peters A, Trinks S, Schonsky H, Facklam M, Wessolek G. 2013. Impact of biochar and hydrochar addition on water retention and water repellency of sandy soil. Geoderma; 202203: 183-191, https://doi.org/10.1016/j.geoderma.2013.03.003.
-p
Abruzzi RC, Dedavid BA, Pires MJR, Ferrarini SF. 2013. Relationship between density and anatomical structure of different species of Eucalyptus and identification of preservatives. Materials Research; 16: 1428-1438, https://doi.org/10.1590/s1516-14392013005000148.
lP
re
Al-Harbi A, Al-Omran A, Shalaby A, Choudhary M. 1999. Efficacy of a hydrophilic polymer declines with time in greenhouse experiments. HortScience; 34: 223-224, https://doi.org/10.21273/HORTSCI.34.2.223.
na
Atkinson CJ. 2018. How good is the evidence that soil‐applied biochar improves water‐holding capacity?. Soil use and management; 34(2): 177-186, https://doi-org.ezp.lib.unimelb.edu.au/10.1111/sum.12413
Jo ur
Baltrėnas P, Baltrėnaitė E, Spudulis E. 2015. Biochar from Pine and Birch Morphology and Pore Structure Change by Treatment in Biofilter. Water, Air, & Soil Pollution; 226, https://doi.org/10.1007/s11270-015-2295-8. Bolan NS, Kunhikrishnan A, Choppala GK, Thangarajan R, Chung JW. 2012. Stabilization of carbon in composts and biochars in relation to carbon sequestration and soil fertility. Science of the Total Environment; 424: 264-70, https://doi.org/10.1016/j.scitotenv.2012.02.061. Brewer CE, Chuang VJ, Masiello CA, Gonnermann H, Gao X, Dugan B, et al. 2014. New approaches to measuring biochar density and porosity. Biomass and Bioenergy; 66: 176-185, https://doi.org/10.1016/j.biombioe.2014.03.059. Cao CTN, Farrell C, Kristiansen PE, Rayner JP. 2014. Biochar makes green roof substrates lighter and improves water supply to plants. Ecological Engineering; 71: 368-374, https://doi.org/10.1016/j.ecoleng.2014.06.017. Cassel D, Nielsen D. 1986. Field capacity and available water capacity. Methods of soil analysis: Part 1—Physical and mineralogical methods: 901-926, https://doi.org/10.2136/sssabookser5.1.2ed.c36. Das O, Sarmah AK. 2015. The love–hate relationship of pyrolysis biochar and water: a perspective. Science of the Total Environment; 512: 682-685, https://doi.org/10.1016/j.scitotenv.2015.01.061 15
Journal Pre-proof de Melo Carvalho MT, de Holanda Nunes Maia A, Madari BE, Bastiaans L, van Oort PAJ, Heinemann AB, et al. 2014. Biochar increases plant-available water in a sandy loam soil under an aerobic rice crop system. Solid Earth; 5: 939-952, https://doi.org/10.5194/se-5-939-2014. Downes GM, Hudson IL, Raymond CA, Dean GH, Michell AJ, Schimleck LR, et al. Sampling plantation eucalypts for wood and fibre properties: CSIRO publishing, 1997. https://doi.org/10.1071/9780643105287. Ehrburger P, Lahaye J, Wozniak E. 1982. Effect of carbonization on the porosity of beechwood. Carbon; 20: 433-439, https://doi.org/10.1016/0008-6223(82)90044-6.
of
Evans R, Downes G, Menz D, Stringer S. 2017. Rapid measurement of variation in tracheid transverse dimensions in a radiata pine tree. Appita Journal: Journal of the Technical Association of the Australian and New Zealand Pulp and Paper Industry; 70: 283,
ro
Farrell C, Ang XQ, Rayner JP. 2013. Water-retention additives increase plant available water in green roof substrates. Ecological Engineering; 52: 112-118, https://doi.org/10.1016/j.ecoleng.2012.12.098.
re
-p
Gray M, Johnson MG, Dragila MI, Kleber M. 2014. Water uptake in biochars: The roles of porosity and hydrophobicity. Biomass and Bioenergy; 61: 196-205, https://doi.org/10.1016/j.biombioe.2013.12.010.
lP
Greacen EL, Walker G, Cook P. Procedure for filter paper method of measuring soil water suction: Adelaide, SA, CSIRO Division of Soils, 1989. https://doi.org/10.4225/08/585ac5eadab2b.
na
Hardie M, Clothier B, Bound S, Oliver G, Close D. 2013. Does biochar influence soil physical properties and soil water availability? Plant and Soil; 376: 347-361, https://doi.org/10.1007/s11104-013-1980-x.
Jo ur
Hyväluoma J, Hannula M, Arstila K, Wang H, Kulju S, Rasa K. 2018. Effects of pyrolysis temperature on the hydrologically relevant porosity of willow biochar. Journal of Analytical and Applied Pyrolysis; 134: 446-453, https://doi.org/10.1016/j.jaap.2018.07.011. Hyvaluoma J, Kulju S, Hannula M, Wikberg H, Kalli A, Rasa K. 2018. Quantitative characterization of pore structure of several biochars with 3D imaging. Environmental Science and Pollution Research; 25: 25648-25658, https://doi.org/10.1007/s11356-017-8823-x. Jeffery S, Verheijen FGA, van der Velde M, Bastos AC. 2011. A quantitative review of the effects of biochar application to soils on crop productivity using meta-analysis. Agriculture, Ecosystems & Environment; 144: 175-187, https://doi.org/10.1016/j.agee.2011.08.015. Jirka S, Tomlinson T. 2015. State of the biochar industry 2014. International Biochar Initiative. Kinney TJ, Masiello CA, Dugan B, Hockaday WC, Dean MR, Zygourakis K, Barnes RT. 2012. Hydrologic properties of biochars produced at different temperatures. Biomass and Bioenergy; 41: 3443, https://doi.org/10.1016/j.biombioe.2012.01.033 Kumar M, Gupta R. 1995. Scanning electron microscopic study of acacia and eucalyptus wood chars. Journal of materials science; 30: 544-551, https://doi.org/10.1007/BF00354423. Lawrence V, Woo K. 2005. SilviScan: an instrument for measuring wood quality. Paprican Special Report, PSR; 550, 16
Journal Pre-proof Lehmann J. 2007. A handful of carbon. Nature; 447: 143, https://doi.org/10.1038/447143a. Lehmann J, Joseph S. Biochar for environmental management: science, technology and implementation: Routledge, 2015. https://doi.org/10.4324/9780203762264. Li S, Harris S, Anandhi A, Chen G. 2019. Predicting biochar properties and functions based on feedstock and pyrolysis temperature: A review and data syntheses. Journal of Cleaner Production; 215: 890-902, https://doi.org/10.1016/j.jclepro.2019.01.106. Lim TJ, Spokas KA, Feyereisen G, Novak JM. 2016. Predicting the impact of biochar additions on soil hydraulic properties. Chemosphere; 142: 136-44, https://doi.org/10.1016/j.chemosphere.2015.06.069.
of
Lu S, Zong Y. 2018. Pore structure and environmental serves of biochars derived from different feedstocks and pyrolysis conditions. Environmental Science and Pollution Research; 25: 30401-30409, https://doi.org/10.1007/s11356-018-3018-7.
-p
ro
Major J, Rondon M, Molina D, Riha SJ, Lehmann J. 2010. Maize yield and nutrition during 4 years after biochar application to a Colombian savanna oxisol. Plant and Soil; 333: 117-128, https://doi.org/10.1007/s11104-010-0327-0.
re
Masiello CA, Dugan B, Brewer CE, Spokas KA, Novak JM, Liu Z, et al. Biochar effects on soil hydrology. Biochar for Environmental Management. Routledge, 2015, pp. 575-594, https://doi.org/10.4324/9780203762264.
lP
Obia A, Mulder J, Martinsen V, Cornelissen G, Børresen T, 2016. In situ effects of biochar on aggregation, water retention and porosity in light-textured tropical soils. Soil and Tillage Research; 155: 35-44, https://doi-org.ezp.lib.unimelb.edu.au/10.1016/j.still.2015.08.002
Jo ur
na
O'Connor D, Peng T, Zhang J, Tsang DC, Alessi DS, Shen Z, Bolan NS, Hou D, 2018. Biochar application for the remediation of heavy metal polluted land: a review of in situ field trials. Science of the total environment; 619: 815-826, https://doi.org/10.1016/j.scitotenv.2017.11.132 Omondi MO, Xia X, Nahayo A, Liu X, Korai PK, Pan G. 2016. Quantification of biochar effects on soil hydrological properties using meta-analysis of literature data. Geoderma; 274: 28-34, https://doi.org/10.1016/j.geoderma.2016.03.029. Phillips E. 1941. The inclination of the fibrils in the cell wall and its relation to the compression strength of timber. Empire Forestry Journal: 74-78, Rasa K, Heikkinen J, Hannula M, Arstila K, Kulju S, Hyväluoma J. 2018. How and why does willow biochar increase a clay soil water retention capacity?. Biomass and Bioenergy; 119: 346-353, https://doi.org/10.1016/j.biombioe.2018.10.004. Ronsse F, van Hecke S, Dickinson D, Prins W. 2013. Production and characterization of slow pyrolysis biochar: influence of feedstock type and pyrolysis conditions. GCB Bioenergy; 5: 104-115, https://doi.org/10.1111/gcbb.12018. Singh BP, Cowie AL, Smernik RJ. 2012. Biochar carbon stability in a clayey soil as a function of feedstock and pyrolysis temperature. Environmental Science & Technology; 46: 11770-8, https://doi.org/10.1021/es302545b.
17
Journal Pre-proof Sohi SP, Krull E, Lopez-Capel E, Bol R. A Review of Biochar and Its Use and Function in Soil, 2010, pp. 47-82, https://doi.org/10.1016/s0065-2113(10)05002-9. Topp GC, Parkin G, Ferré TP, Carter M, Gregorich E. 2008. Soil water content. Soil sampling and methods of analysis’. 2nd edn.(Eds MR Carter, EG Gregorich) pp: 939-962, https://doi.org/10.1201/9781420005271. Tryon EH. 1948. Effect of charcoal on certain physical, chemical, and biological properties of forest soils. Ecological Monographs; 18: 81-115, https://doi.org/10.2307/1948629. Walker JC. Primary wood processing: principles and practice: Springer Science & Business Media, 2006. https://doi.org/10.1007/1-4020-4393-7.
of
Wang Y, Zhang L, Yang H, Yan G, Xu Z, Chen C, et al. 2016. Biochar nutrient availability rather than its water holding capacity governs the growth of both C3 and C4 plants. Journal of Soils and Sediments; 16: 801-810, https://doi.org/10.1007/s11368-016-1357-x.
ro
Watson A, Dadswell H. 1961. Influence on fibre morphology on paper properties. Part 1. Fibre length. Appita J; 14: 168-178,
re
-p
Weber K, Quicker P. 2018. Properties of biochar. Fuel; 217: 240-261, https://doi.org/10.1016/j.fuel.2017.12.054.
lP
Wildman J, Derbyshire F. 1991. Origins and functions of macroporosity in activated carbons from coal and wood precursors. Fuel; 70: 655-661, https://doi.org/10.1016/0016-2361(91)901819.
na
Wilkes J. 1988. Variations Inwoodanatomy Within Species of Eucalyptus. IAWA Journal; 9: 13-23, https://doi.org/10.1163/22941932-90000461.
Jo ur
Zbonak A, Bush T, Grzeskowiak V. Comparison of tree growth, wood density and anatomical properties between coppiced trees and parent crop of six Eucalyptus genotypes. Improvements and culture of eucalypts, IUFRO conference. Citeseer, 2007, pp. 22-26, Zeriouh A, Belkbir L. 1995. Thermal decomposition of a Moroccan wood under a nitrogen atmosphere. Thermochimica acta; 258: 243-248, https://doi.org/10.1016/00406031(94)02246-K. Zhang J, You C. 2013. Water Holding Capacity and Absorption Properties of Wood Chars. Energy & Fuels; 27: 2643-2648, https://doi.org/10.1021/ef4000769. Ziemińska K, Butler DW, Gleason SM, Wright IJ, Westoby M. 2013. Fibre wall and lumen fractions drive wood density variation across 24 Australian angiosperms. AoB PLANTS; 5, https://doi.org/10.1093/aobpla/plt046. Zornoza R, Moreno-Barriga F, Acosta JA, Muñoz MA, Faz A. 2016. Stability, nutrient availability and hydrophobicity of biochars derived from manure, crop residues, and municipal solid waste for their use as soil amendments. Chemosphere; 144: 122-130, https://doi.org/10.1016/j.chemosphere.2015.08.046.
18
Journal Pre-proof Figure captions Figure 1. Top view of a stem sample cut from a eucalypt tree (E. macrohyncha) for pyrolysis and airdry density measurement (left) and for wood cell structure analysis using the Silviscan3 (right, dashed strip) Figure 2. Transmitted-light cell scan micrographs obtained from Silviscan3 cell scanner (100 x) of three species of Eucalyptus with different wood densities. A: low density (653 kg m-3; E. nitens), B: medium density (788 kg m-3; E. polyanthemos), C: high density (960 kg m-3; E. polybraktea). White to light grey circular structures are cell walls, dark grey circular structures are fibre lumen, dark grey horizontal band structures are radial parenchyma. Fibre lumen diameters decreasing and fibre cell wall thickness increasing from low density wood (A) to high density wood (C). Fibre lumen almost invisible as cell walls take up almost all area (C).
ro
of
Figure 3. Relationships between fibre wall thickness and radial fibre lumen diameter with wood density. Values for fibre wall thickness (n=3), radial fibre lumen diameter (n=3) and wood density (n≥5) represent means, bars represent ±mean standard error and R2 and p-values are for regression fit. Shaded areas indicate 95% confidence regions for regression fit.
re
-p
Figure 4. Relationships between feedstock wood density with biochar maximum water holding capacity and plant available water. Values for wood density (n≥5), water holding capacity (n=3) and plant available water (n=3) represent means, bars represent ±mean standard error and R2 and pvalues are for regression fit. Shaded areas indicate 95% confidence regions for regression fit.
Supplementary material
na
lP
Figure 5. Relationship between biochar water holding capacity and biochar plant available water. Values represent means (n=3), bars represent mean standard error and R2 and p-values are for regression fit. Shaded area indicates 95% confidence region for regression fit.
Jo ur
Figure 6. Boxplots of species air-dry wood density (top), radial fibre lumen diameter (middle) and fibre wall thickness. Different letters indicate significant differences among species means as determined from one-way ANOVA. Data points represent individual sample means (all P < 0.001).
19
Journal Pre-proof Table 1. Mean air-dry wood density of the feedstock material and resulting dry bulk density, water holding capacity (WHC) and plant available water (PAW) for biochar produced from 18 Eucalyptus species grown in a field experiment. Species listed in order of increasing wood density. Mean ± SE in parentheses.
Biochar bulk
Biochar WHC
Biochar PAW
Species
(kg m-3)
density dry (kg m-3)
(% v/v)
(% v/v)
E. pauciflora
571.9 (18.82)
140.3 (2.88)
44.5 (0.45)
41.6 (0.45)
E. regnans
598.8 (24.36)
158.7 (2.21)
41.6 (0.10)
36.9 (0.10)
E. delegatensis
628.3 (20.66)
157.7 (2.80)
43.6 (0.22)
38.3 (0.22)
E. nitens
653.2 (21.76)
186.0 (2.82)
E. cypellocarpa
666.9 (24.61)
E. muelleriana
41.3 (0.44)
159.1 (0.84)
44.2 (0.29)
40.0 (0.29)
666.9 (23.89)
179.5 (0.99)
43.5 (0.16)
38.7 (0.16)
E. dives
685.1 (22.70)
189.8 (11.2)
43.2 (0.48)
37.3 (0.48)
E. obliqua
686.5 (22.19)
158.6 (3.63)
40.8 (0.24)
35.8 (0.24)
E. radiata
688.5 (25.27)
195.6 (7.65)
47.8 (0.64)
42.2 (0.64)
E. macrohyncha
698.6 (22.51)
178.2 (3.36)
42.8 (0.60)
37.9 (0.60)
717.3 (17.64)
189.7 (4.14)
44.2 (0.35)
38.1 (0.35)
738.5 (19.66)
280.9 (38.1)
44.0 (0.71)
36.0 (0.71)
755.7 (20.71)
190.7 (4.51)
41.5 (0.61)
35.6 (0.61)
E. leucoxylon
778.3 (15.79)
210.4 (8.04)
40.4 (0.06)
35.1 (0.06)
E. polyanthemos
788.3 (18.65)
249.2 (10.9)
39.0 (0.19)
33.0 (0.19)
E. microcarpa
894.6 (10.13)
322.9 (9.84)
38.0 (0.53)
30.0 (0.53)
E. tricarpa
925.4 (10.91)
304.8 (11.5)
36.9 (0.18)
29.6 (0.18)
E. polybractea
959.9 (40.53)
294.3 (2.45)
35.5 (0.08)
28.6 (0.08)
E. melliodora E. globoidea
lP
na
Jo ur
E. viminalis
-p
47.4 (0.44)
re
ro
of
Wood density air-dry
20
Journal Pre-proof Table 2. Particle size distribution (% w/w) for all biochar types after remixing
12.5-10 mm
10-6.3 mm
6.3-4 mm
4-2 mm
2-1 mm
<1 mm
biochar
16.9
26.4
19.9
15.3
7.4
14
Jo ur
na
lP
re
-p
ro
of
`
21
Journal Pre-proof Graphical abstract
Highlights Biochars differ greatly in their properties due to feedstock differences We studied the effect of feedstock cell structure on biochar water retention Biochar from low density eucalypt wood has greater plant available water retention Greater wood density is due to greater cell wall thickness and smaller cell diameters
Jo ur
na
lP
re
-p
ro
of
Wood density can be used as a proxy for wood cell structure
22
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Joerg Werdin, Tim D. Fletcher, John P. Rayner, Nicholas S.G. Williams, Claire Farrell PII:
S0048-9697(19)35851-6
DOI:
https://doi.org/10.1016/j.scitotenv.2019.135856
Reference:
STOTEN 135856
To appear in:
Science of the Total Environment
Received date:
17 September 2019
Revised date:
11 November 2019
Accepted date:
28 November 2019
Please cite this article as: J. Werdin, T.D. Fletcher, J.P. Rayner, et al., Biochar made from low density wood has greater plant available water than biochar made from high density wood, Science of the Total Environment (2018), https://doi.org/10.1016/ j.scitotenv.2019.135856
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2018 Published by Elsevier.
Journal Pre-proof Title: Biochar made from low density wood has greater plant available water than biochar made from high density wood Authors: Joerg Werdin*, Tim D. Fletcher, John P. Rayner, Nicholas S.G. Williams and Claire Farrell Authors affiliation: School of Ecosystem and Forest Sciences, Faculty of Science, The University of Melbourne, 500 Yarra Boulevard, Richmond 3121, Victoria, Australia *
Corresponding author: E-mail address: [email protected]
of
Abstract
ro
Soil water limitations often restrict plant growth in unirrigated agricultural, forestry and urban
-p
systems. Biochar amendment to soils can increase water retention, but not all of this additional
re
water is necessarily available to plants. Differences in the effectiveness of biochar in ameliorating
lP
soil water limitations may be a result of differences in feedstock cell structure. Previous research has shown that feedstock cell structure influences the pore structure of biochar and consequently the
na
volume available for water storage. The availability of this water for plant uptake will be determined by biochar pore diameters, given its role in determining capillary forces which plants must overcome
Jo ur
to access pore water. Therefore, we hypothesised that differences in hardwood feedstock cell structure would result in differences in the plant available water holding capacity of biochar. Before pyrolysis, we measured the wood morphology of 18 Eucalyptus species on three replicates of equal age on a gradient of wood density (572 – 960 kg m-3). Wood samples were then pyrolysed (550 °C) and the resulting biochars were sieved and their particle size distribution was standardised before their physical properties, including water holding capacity, plant available water and bulk density were measured. Our results show that biochar made from lower density eucalypt wood had up to 35% greater water holding capacity and up to 45% greater plant available water than biochar made from higher density eucalypt wood. Further, feedstock wood density related well to fibre cell wall thickness and fibre lumen diameter. Therefore, wood density could be used as a proxy for wood cell
1
Journal Pre-proof structure, which can in turn be used to predict plant available water in biochar. The simple measure of feedstock wood density can inform feedstock choices for producing biochars with greater plant available water, optimal for the use as soil amendment in water limited environments. Keywords biochar pore structure; hardwood; plant available water; water holding capacity; wood cell structure; wood density
of
1. Introduction
ro
Biochars are highly porous solids produced by the thermochemical conversion of biomass under
-p
oxygen-depleted conditions in a process known as pyrolysis; which has been traditionally used to
re
produce charcoal (Lehmann and Joseph, 2015). They are considered a useful soil amendment for improving soil water retention and plant growth in rainfed systems in agricultural (Jeffery et al.,
lP
2011; Obia et al., 2016), forestry (Tryon, 1948) and urban landscapes (Cao et al., 2014). As well as
na
improving water retention in soils (Omondi et al., 2016), biochars have also been shown to increase crop yields by increasing nutrient retention (Major et al., 2010), improving soil structure (Lim et al.,
Jo ur
2016; Obia et al., 2016) and microbial activity (Wang et al., 2016) and reducing the bioavailability of heavy metals (O'Connor et al., 2018) whilst adding a stable source of soil carbon for sequestration (Lehmann, 2007). However, the effects of biochar as a soil amendment can be highly variable due to (i) differences in feedstock materials, (ii) their interaction with different soil types and (iii) differences in pyrolysis conditions (Atkinson, 2018; Jeffery et al., 2011; Li et al., 2019; Rasa et al., 2018; Ronsse et al., 2013; Sohi et al., 2010; Weber and Quicker, 2018).
While biochar can be made from a wide range of organic materials including crop residues, biosolids or food waste, the most commonly used feedstocks for biochar production worldwide are woodbased (Jirka and Tomlinson, 2015). When wood feedstocks are pyrolysed, their chemical and physical make up changes, influencing their water retention properties, water uptake behaviour 2
Journal Pre-proof (degrees of hydrophobicity) and longevity. Chemically, wood is mainly composed of hemicellulose, cellulose and lignin (Walker, 2006). During pyrolysis, these compounds decompose, releasing volatile liquid and gaseous fractions and leaving behind solid aromatic carbon compounds (Zeriouh and Belkbir, 1995). This results in wood biochars generally having a greater than 80% (w/w) carbon content when pyrolysed above 500 °C (Weber and Quicker, 2018). The aromatic carbon structure of biochar contributes to its long-term stability. This makes biochar a better soil ammendment with sustained improvement for water retention than other organic or synthetic water retention
of
amendments. For instance, compost or hydrogels decompose orders of magnitude faster than
ro
biochar (Al-Harbi et al., 1999; Bolan et al., 2012; Singh et al., 2012). Another important biochar
-p
property governing its water uptake behaviour is the degree of hydrophobicity. Generally, low pyrolysis temperatures (< 500 °C) can result in hydrophobic, water repellent biochar (Zornoza et al.,
re
2016). This is caused by incomplete carbonisation of chemical wood components, resulting in
lP
remaining aliphatic functional groups on the biochar surface which repel water (Das and Sarmah, 2015; Kinney et al., 2012). Therefore, greater pyrolysis temperature generally results in reduced
Jo ur
of biochar.
na
hydrophobicity of biochar, enabling greater soil water uptake and increased water holding capacity
The cellular wood structure remains largely intact after pyrolysis and forms the macroporous structure of biochar (Baltrėnas et al., 2015; Ehrburger et al., 1982; Gray et al., 2014; Hyväluoma et al., 2018; Wildman and Derbyshire, 1991). Biochar macropores are very important for water retention, as they make up the main pore volume for water storage (Brewer et al., 2014; Gray et al., 2014; Hyväluoma et al., 2018; Lu and Zong, 2018). This has been demonstrated by Zhang and You (2013) who showed a strong positive correlation between macropore size and water holding capacity (WHC) for biochars produced from poplar (Populus davidiana) and pine (Pinus sylvestris var. mongolica) wood. The biochar made from poplar wood (hardwood) had greater macropore diameters (1 – 40 µm) and greater WHC (69 – 72%) when compared to the biochar made from pine 3
Journal Pre-proof wood (softwood), with smaller macropore diameters (1 – 10 µm) having a WHC of 29%. The authors concluded that biochars with greater macropore diameters and therefore greater pore volumes will lead to increases in water retention and are therefore best suited for improving soils in waterlimited environments. The wood of hardwood species (generally non-coniferous) is made up of fibre, vessel and parenchyma cells (Wilkes, 1988). Fibre cells can be described as hollow tubes with thickened cell walls (Downes et al., 1997) and as they make up a large proportion of the wood structure, fibre properties greatly influence wood density (Abruzzi et al., 2013; Zbonak et al., 2007).
of
In general, higher density hardwoods have greater fibre wall thickness and smaller fibre diameters
ro
than lower density hardwoods (Watson and Dadswell, 1961). Therefore, it is likely that biochar
-p
produced from lower density hardwoods will retain more water.
re
However, water holding capacity (WHC) does not always reflect how much water is accessible for
lP
plant growth (Cao et al., 2014; Masiello et al., 2015). A better measure is plant available water (PAW), which accounts for the amount of water that is held in pores at tensions which can be
na
extracted by plants. PAW, also referred to as available water content (AWC), is mainly determined
Jo ur
by macropore size, with water stored in pores with diameters between 0.2 to 30 µm being plant available, while water stored in smaller pores is held too tightly for plant uptake, mainly due to capillary and adsorption forces (Atkinson, 2018; Cassel and Nielsen, 1986). The total amount of water in pores with diameters smaller than 0.2 µm is held with a negative pressure greater than 1500 KPa which supersedes most plants ability to extract it. It is commonly referred to as permanent wilting point (PWP), whereas water in pores with diameters greater than 30 µm drains under gravity and is only briefly available for plant uptake after a rain event (Atkinson, 2018). The number of pores in biochar in the relevant diameter range for PAW is largely influenced by feedstock type (Lu and Zong, 2018). Therefore, the number of pores in the size range for maximising PAW in biochar should relate to the cell structure of the woody feedstock material, and biochar produced from lower density hardwood should hold more PAW than biochar made from high density hardwood. 4
Journal Pre-proof
While there is an extensive body of research on how biochar influences water retention in soils (Abel et al., 2013; de Melo Carvalho et al., 2014; Hardie et al., 2013; Tryon, 1948), few studies have investigated how feedstock fibre properties and wood density influences water retention and availability of biochar (Hyväluoma et al., 2018; Rasa et al., 2018). Given this gap, in this study we investigated how feedstock wood density was related to biochar WHC and PAW. We hypothesized that lower density hardwood feedstock would result in biochar with greater overall WHC and more
of
of this water would be plant available (PAW) when produced under the same conditions. Our
ro
research objectives were: (i) to investigate if wood density can be used as a low cost and easy to
-p
measure proxy for feedstock cell structure (ii) to determine how feedstock cell structure influences WHC and PAW of the resulting biochar and (iii) to assess which proportion of the overall water
re
retention is plant available. The water retention measurements (WHC and PAW) in our study were
lP
performed on pure biochars without being incorporated into a soil to avoid complications due to biochar-substrate interactions. Our findings need to be validated in future studies with amended
Jo ur
na
soils.
2. Materials and Methods
To determine how hardwood feedstocks with different wood densities, influence biochar water retention and availability, we compared biochars made from 18 Eucalyptus species under the same pyrolysis conditions. The 18 Eucalyptus species represent a gradient of wood density (Table 1) from low density e.g. E. pauciflora (572 kg m-3) to high density e.g. E. polybraktea (960 kg m-3). 2.1. Wood feedstock and sample preparation The wood samples were taken from 18 Eucalyptus species (Table 1), grown under equal environmental conditions in a field experiment at the Burnley Campus of the University of
5
Journal Pre-proof Melbourne in November 2015. All trees were 4.25 years of age (51 months) and a minimum of three replicate samples were taken per species (1 sample per tree).
The samples were taken from approximately 150 mm long cross-sections of each trunk. To ensure comparability among samples in regard to their wood density, all samples were cut 200 mm above the soil surface as wood density tends to decrease with increasing distance from the base of the stem (Downes et al., 1997) (Figure 1). Stem diameters ranged from 52 mm (E. tricarpa) to 229 mm
of
(E. viminalis). The wood samples were air-dried in a polytunnel for five months before being
ro
debarked. From three air-dried wood samples per species, 7 mm long radial strips were cut from
-p
pith to bark with a standard width of 2 mm (Figure 1) and one of their transverse surfaces was
2.2. Wood analysis
lP
method described in Evans et al. (2017).
re
polished with sanding paper (coarse 400 – fine 1500 grit) for analysis on Silviscan3 according to the
na
Prepared wood sections were analysed using Silviscan3, a wood analysing instrument combining optical microscopy, X-ray densitometry and automated image analysis to measure wood cell
Jo ur
properties (Figure 1) (Lawrence and Woo, 2005). The wood cell properties, fibre cell wall thickness and radial and tangential fibre lumen diameter were obtained in 25 μm radial increments for each of the three replicates per species and automatically analysed by integrated image analysis software on Silviscan3 (only one sample could be analysed for E. dives and two samples for E. macrohyncha and E. radiata due to internal ruptures and extensive tension wood which prohibited meaningful analysis). Detailed information on within- and between species variability are provided in the supplementary information (Figure 6). Values for each sample (in 25 μm sections) were then averaged over the radial wood profile from pith to bark before a mean of the three samples of each species was calculated. At each end of the wood samples, 5 mm was excluded from the analysis to avoid errors caused from cracks and ruptures in the wood due to post-harvest drying.
6
Journal Pre-proof
Wood density was determined by water displacement for a minimum of five replicates per species, prior to pyrolysis (Figure 1) and (Figure 6 supplementary information). Debarked and air-dried wood samples were weighed before being briefly submerged in a water-filled PVC cylinder (diameter 150 mm) equipped with a water-level gauge. The change in water level was recorded and the air-dry wood density calculated: 𝑤𝑒𝑖𝑔ℎ𝑡 𝑎𝑖𝑟−𝑑𝑟𝑦 𝜋∗𝑟 2 ∗𝛥ℎ
∗ 1000
(1)
of
𝑤𝑜𝑜𝑑 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 =
ro
where wood density is the air-dry wood density in kg m-3, weight air-dry is the air-dried weight in g, r
-p
is the radius of the PVC cylinder in cm and 𝛥ℎ is the change in water level in cm.
re
2.3. Biochar pyrolysis
lP
Prior to pyrolysis, wood samples (excluding sections taken for wood analysis on Silviscan3 as described in section 2.1) were labelled and individually placed on stainless-steel trays. The wood was
na
then pyrolysed (January 2017) in a mobile charmaker (Charmaker MPP20, Earth Systems Pty., Melbourne, Australia), using slow-batch pyrolysis at 550 °C (± 50 °C) and a 4-hr residence time. Each
Jo ur
of the species-specific biochars was then air-dried for three months in a polytunnel before being crushed manually and passed through a nest of sieves of decreasing aperture (12.5 mm, 10 mm, 6.3 mm, 4 mm, 2 mm, 1 mm). Particles greater then 12.5 mm in diameter were discarded. Each of the particle size fractions for each sample were then weighed and their volume determined using a measuring cylinder at a standardised compaction to determine the bulk density (g cm-3) of each particle size fraction. The biochar samples were then remixed to ensure that the particle size fractions were identical between each of the 18 biochar types before testing the biochar properties (Table 2). 2.4. Biochar properties
7
Journal Pre-proof Biochar bulk density and water holding capacity (WHC) were determined on three samples for each of the 18 eucalypt species. Biochar dry bulk density was determined according to the Australian Standard for Potting Mixes (AS 3743—2003) as the weight of oven dried (105 °C until weight equilibrium) material divided by its volume using standardised measuring cylinders at a standardised compaction (free fall from 5 cm height repeated five times).
WHC was determined by using a tension table (Topp et al., 2008). Known quantities of each biochar
of
type were submerged in deionised water for 24 hr and then placed onto a tension table at -10 KPa
ro
suction (field capacity) until constant weight. The samples were weighed and oven dried at 105 °C
𝑊𝐹𝐶 −𝑊𝑑𝑟𝑦 𝑉
∗ 100
(2)
re
𝑊𝐻𝐶 =
-p
until constant weight. The volumetric water content was calculated:
lP
Where WHC is the water holding capacity in % v/v, WFC is the sample weight at field capacity (-
na
10KPa) in g, Wdry is the oven-dry sample weight in g and V is the sample volume in cm3.
Jo ur
To determine PAW, moisture release curves for each of the 18 eucalypt biochars were obtained using the filter paper method described by Greacen et al. (1989) and modified by Farrell et al. (2013). For each biochar, samples were oven dried at 105 °C to constant weight and divided into 10 equal portions (by weight), before deionised water was added to each sample to obtain samples of 6 – 33% of their WHC, at 3% increments. Samples were packed into glass jars with three filter papers (Whatman No. 42) placed between half the sample volume, so that one filter paper was sandwiched between the other two to reduce weighing errors due to adhering biochar particles. The jars were then sealed and placed in a temperature-controlled environment to equilibrate for seven days. The gravimetric water content of the middle filter paper of each sample was determined, and the corresponding matric suction was calculated according to Greacen et al. (1989). Moisture release
8
Journal Pre-proof curves were then plotted and the volumetric water content of biochar at -1.5 MPa calculated from the intersect of the exponential model. Biochar PAW was calculated:
PAW = WHC – PWP
(3)
Where PAW is the plant available water in % v/v, WHC is the water holding capacity in % v/v and PWP is the water content at permanent wilting point (-1.5 MPa) in % v/v. 2.5. Statistical analyses
of
All statistical analyses were performed in R version 3.4.1 (R Core Team, 2017). Correlations among all
ro
measured variables were determined using Pearson’s Correlations to determine the variables best
-p
explaining differences in WHC and PAW. Linear regressions were performed to relate (i) feedstock
re
fibre wall thickness and radial fibre lumen diameter with wood density; (ii) wood density with biochar WHC and PAW and (iii) biochar WHC with biochar PAW. Variability among species for air-dry
lP
wood density, radial fibre lumen diameter and fibre wall thickness was analysed using one-way
0.05).
Jo ur
3. Results and Discussion
na
ANOVA and significant differences among means were determined using Tukey’s post-hoc test (p <
Overall, our study showed that eucalypt feedstock with lower wood density resulted in biochar with up to 35% greater water holding capacity (WHC) and up to 45% greater plant available water (PAW) than biochar made from higher wood density feedstock. Further, WHC was strongly related to PAW, indicating that the additional water retained in biochar produced from lower density feedstocks is also available for plant growth. As feedstock wood density was also related to feedstock cell structure, we suggest that lower density feedstock wood with greater fibre diameters and thinner fibre walls results in biochar with a greater number of pores which retain water that is subsequently available to plants. These results support our hypothesis that differences in feedstock wood density cause differences in biochar WHC and PAW.
9
Journal Pre-proof 3.1. Relationships between wood fibre properties with wood density Wood density is mainly determined by fibre wall thickness and fibre lumen diameter. Thinner cell walls and greater fibre diameter are found in lower density species, whereas thicker cell walls and smaller fibre diameters are found in higher density species (Figure 3). Variability in wood density across species was best explained by fibre wall thickness (R2 = 0.88) and to a lesser degree by radial fibre lumen diameter (R2 = 0.69). The eucalypt feedstock with the greatest average fibre wall thickness (1.96 µm across the radial diameter from pith to bark, SE± 0.03) was E. polybractea, which
of
was also the heaviest (960 kg m-3, SE± 40.53); whereas E. pauciflora had an average fibre wall
ro
thickness of 1.08 µm (SE± 0.02) and was the species with the lowest wood density (572 kg m-3, SE±
-p
18.82). The range in average radial fibre diameter was small among the investigated feedstock
re
species with E. regnans (11.25 µm, SE± 0.31) wood having the greatest average diameter and E. microcarpa (9.72 µm, SE± 0.06) the smallest. This was well aligned with their wood density, as these
lP
species were among the lightest (599 kg m-3, SE± 24.36) and heaviest (895 kg m-3, SE± 10.13) species
na
investigated in our study. Ziemińska et al. (2013) reported similar relationships between wood density and fibre wall thickness in a study of the branch-wood structure of 24 Australian
Jo ur
angiosperms across different genera. Our study shows that wood density can be used as a proxy for wood cell structure. This has important implications for the rapid assessment of wood properties for biochar feedstock selection, as air-dry wood density can be easily determined, whereas the direct measurement of cell properties is time and cost intensive and requires elaborate research equipment. 3.2. Relationships between feedstock wood density and biochar WHC and PAW Since wood density is mainly determined by fibre cell wall thickness and fibre lumen diameter, it can be regarded as a predictor of porosity. The cell walls form the solid structure, whereas the lumen diameters are open void spaces. In other words, wood density is a measure of the cell wall material per volume of wood (Phillips, 1941). These wood cell properties are largely retained in the pore
10
Journal Pre-proof structure of biochar with the cell walls forming the pore boundaries within biochar and the cell lumen forming the biochar pore space (Gray et al., 2014; Hyväluoma et al., 2018). Therefore, lighter wood with a greater proportion of open void space (cell lumen) and a lower proportion of solid material (cell wall) results in biochar with a greater proportion of open void space (pore volume) and a lower proportion of solid material (pore boundaries). When biochar is exposed to water, the airfilled biochar pores fill with water and the total amount of water retained in a volume of biochar is the WHC. In our study, wood density ranged between 572 – 960 kg m-3. Biochar produced from
of
eucalypts with lower density wood had up to 35% greater WHC than biochar made from higher
ro
density wood (Figure 4). However, greater overall water retention (WHC) due to greater internal
-p
porosity does not necessarily result in greater water availability for plants (PAW), as some of this
lP
the potential benefit for plant growth.
re
water might be too tightly bound for plant uptake. Therefore, PAW is a better measure to evaluate
PAW is stored in biochar pores that range between 0.2 µm – 30 µm and therefore, biochars made
na
from lower density Eucalyptus wood are likely to have a greater proportion of pores in the plant-
Jo ur
available range. Hardie et al. (2013) investigated the pore structure of an Acacia green waste biochar and showed that the median pore diameter ranged between 0.4 – 13 µm, thus most pores residual from the feedstock cell structure were in the diameter range relevant for PAW. Even though the Acacia species from which the biochar was made from was not specified, the general hardwood cell structure of acacias is similar to Eucalyptus species. A Scanning Electron Microscope (SEM) study by Kumar and Gupta (1995) that investigated the morphological changes of Acacia and Eucalyptus wood undergoing carbonisation came to the conclusion that both species underwent similar morphological changes. The biochar with the greatest amount of PAW in our study was made from Eucalyptus radiata (42.2%, SE± 0.64), which is also a relatively light wood species with a density of 689 kg m-3(SE± 25.27) (Table 1). Conversely, biochar made from E. polybractea had the lowest amount of PAW (28.6%, SE± 0.08) and the greatest wood density of all investigated species (960 kg 11
Journal Pre-proof m-3, SE± 40.53). The greater wood density of E. polybractea wood is due to thicker cell walls (1.97 µm, SE± 0.03) and smaller fibre lumen diameters (9.92 µm, SE± 0.07), when compared to E. radiata (1.12 µm; 10.83 µm), resulting in a greater proportion of solid, lignocellulose material and less open void space per volume of wood. 3.3. Relationship between biochar PAW and WHC WHC and PAW were strongly correlated in the 18 biochars made from different eucalypt feedstocks
of
under equal pyrolysis conditions (Figure 5). The difference between WHC and PAW determines the
ro
amount of water which is unavailable to plants due to the high tensions at which water is held in pores smaller than 0.2 µm (Atkinson, 2018; Cassel and Nielsen, 1986). Among the 18 biochars there
-p
was an average 6% difference between WHC and PAW, which is the amount of water unavailable for
re
plant uptake. The biochar with the greatest amount of unavailable water was E. microcarpa (8%, SE±
lP
0.53), whereas biochar made from E. pauciflora had the smallest amount of unavailable water (3%, SE± 0.45). Compared to the differences in PAW across biochar types (28.6 – 42.2%), the differences
Jo ur
na
in unavailable water were marginal.
The small differences in unavailable water among the 18 biochars are likely due to the fact that they were pyrolysed at the same temperature in one batch. Gray et al. (2014) proposed a biochar pore size classification based on the pore origin and diameter size range. Residual macropores in the range between 1-100 µm are inherited from the feedstock cell structure and pyrogenic nanopores in the nanometer range develop with increasing pyrolysis temperature above 500 °C. Since all biochars used in this study were produced at a pyrolysis temperature only marginally greater (550 °C ± 50 °C), extensive pyrogenic nanopore development is unlikely. This is supported by the relatively small amount of unavailable water per volume of biochar since water stored in these nanopores is bound to tightly for plant uptake. Therefore, the equal pyrolysis conditions used to create all the biochars tested likely resulted in similar pyrogenic nanopore development, regardless of feedstock cell 12
Journal Pre-proof structure. This supports our hypothesis that differences in WHC and PAW can be primarily attributed to differences in feedstock cell structure.
Multiple studies have highlighted that the macroporous structure of biochar (i) is inherited from the feedstock cell structure and (ii) determines its water retention (WHC and PAW) (Gray et al., 2014; Hyväluoma et al., 2018; Kumar and Gupta, 1995; Wildman and Derbyshire, 1991; Zhang and You, 2013). However, studies relating WHC and PAW of biochar with feedstock cell structure are rare,
of
which might be due to the fact that wood cell structure is heterogenous and commonly used
ro
analytical methods such as SEM techniques or more recently X-ray tomography, rely on small sample
-p
sizes (< 1 mm2) (Hyvaluoma et al., 2018; Kumar and Gupta, 1995). These small samples are usually not representative of the species-specific cell structure and mechanistic relationships between
re
feedstock cell structure and biochar pore structure cannot easily be made. For example, Hyvaluoma
lP
et al. (2018) compared the cell structure of 1 mm2 samples of dried willow wood with the pore structure of the resulting biochar using x-ray tomography images and concluded that their results
na
were not representative for willow wood in general, due to the small sample size and heterogeneity
Jo ur
of willow wood structure. Generally, wood density tends to increase from the centre of the tree (pith) to bark and from the tree base to the top, due to changing fibre properties (lumen diameters and cell wall thickness) (Downes et al., 1997). In our study, we accounted for the heterogeneity in wood fibre properties by measuring the cell structure across the full diameter from pith to bark, and by taking all wood samples at similar distance from the tree base. Furthermore, all trees were planted at the same time, grown under the same conditions and were of equal age when harvested which controls for differences caused by environmental conditions (i.e. precipitation, temperature and soil conditions). This gives us confidence in the relationships we found between wood properties and biochar properties amongst the 18 Eucalyptus species.
We investigated the water retention behaviour of pure biochars to explore if there is a mechanistic 13
Journal Pre-proof relationship between wood cell structure and resulting biochar water retention. However, it remains unknown how different levels of WHC and PAW in the pure biochar affect the water holding properties of amended soil and substrate. A metanalysis conducted by Omondi et al. (2016) based on 74 datasets reporting changes of PAW in soils after biochar amendment reported an overall increase in PAW after biochar amendment of 15%. However, in contrast other studies have reported no significant changes in PAW after biochar addition (i.e. (Hardie et al., 2013)). Therefore, the reported differences in WHC and PAW of biochar made from feedstocks with different wood
of
densities in our study need to be validated in studies with amended soils.
ro
4. Conclusions
-p
For the 18 eucalypt species evaluated in this experiment, the WHC and PAW of their resulting
re
biochar was strongly related to their feedstock wood density. Biochar produced from eucalypt
lP
species with lower wood density had greater WHC and PAW than biochar produced from higher density feedstocks. Wood density was strongly related to feedstock cell structure. Therefore, we
na
suggest that lower density wood species, with greater fibre diameters and thinner fibre walls, result in biochar with a greater proportion of pores which retain water available to plants. As biochar
Jo ur
water retention could be estimated from the simple measure of wood density, this can be used to select feedstocks to produce biochars with high plant available water, without the need for rigorous testing. However, as our results are for pure biochar, further research is needed to test whether the increased WHC and PAW of biochar produced from lower density feedstocks also translate to increased WHC and PAW of biochar amended soils in agricultural, forestry and urban applications. 5. Acknowledgements This research was funded by an Australian Research Council Linkage Grant (LP30100731) supported by Melbourne Water and the Inner Melbourne Action Plan (IMAP) group of local governments. Joerg Werdin was supported by a Research Training Program Scholarship and a Graduate Research Studentship. We thank Robert Evans for his help with the wood analysis, Adrian Morphett and 14
Journal Pre-proof Moana Quiatol from Earth Systems for their help producing the biochar, Richard Conn and Petra Katona for their help with the experiment. We thank the anonymous reviewers for their valuable comments and constructive criticism that helped to improve the quality of our manuscript.
References
ro
of
Abel S, Peters A, Trinks S, Schonsky H, Facklam M, Wessolek G. 2013. Impact of biochar and hydrochar addition on water retention and water repellency of sandy soil. Geoderma; 202203: 183-191, https://doi.org/10.1016/j.geoderma.2013.03.003.
-p
Abruzzi RC, Dedavid BA, Pires MJR, Ferrarini SF. 2013. Relationship between density and anatomical structure of different species of Eucalyptus and identification of preservatives. Materials Research; 16: 1428-1438, https://doi.org/10.1590/s1516-14392013005000148.
lP
re
Al-Harbi A, Al-Omran A, Shalaby A, Choudhary M. 1999. Efficacy of a hydrophilic polymer declines with time in greenhouse experiments. HortScience; 34: 223-224, https://doi.org/10.21273/HORTSCI.34.2.223.
na
Atkinson CJ. 2018. How good is the evidence that soil‐applied biochar improves water‐holding capacity?. Soil use and management; 34(2): 177-186, https://doi-org.ezp.lib.unimelb.edu.au/10.1111/sum.12413
Jo ur
Baltrėnas P, Baltrėnaitė E, Spudulis E. 2015. Biochar from Pine and Birch Morphology and Pore Structure Change by Treatment in Biofilter. Water, Air, & Soil Pollution; 226, https://doi.org/10.1007/s11270-015-2295-8. Bolan NS, Kunhikrishnan A, Choppala GK, Thangarajan R, Chung JW. 2012. Stabilization of carbon in composts and biochars in relation to carbon sequestration and soil fertility. Science of the Total Environment; 424: 264-70, https://doi.org/10.1016/j.scitotenv.2012.02.061. Brewer CE, Chuang VJ, Masiello CA, Gonnermann H, Gao X, Dugan B, et al. 2014. New approaches to measuring biochar density and porosity. Biomass and Bioenergy; 66: 176-185, https://doi.org/10.1016/j.biombioe.2014.03.059. Cao CTN, Farrell C, Kristiansen PE, Rayner JP. 2014. Biochar makes green roof substrates lighter and improves water supply to plants. Ecological Engineering; 71: 368-374, https://doi.org/10.1016/j.ecoleng.2014.06.017. Cassel D, Nielsen D. 1986. Field capacity and available water capacity. Methods of soil analysis: Part 1—Physical and mineralogical methods: 901-926, https://doi.org/10.2136/sssabookser5.1.2ed.c36. Das O, Sarmah AK. 2015. The love–hate relationship of pyrolysis biochar and water: a perspective. Science of the Total Environment; 512: 682-685, https://doi.org/10.1016/j.scitotenv.2015.01.061 15
Journal Pre-proof de Melo Carvalho MT, de Holanda Nunes Maia A, Madari BE, Bastiaans L, van Oort PAJ, Heinemann AB, et al. 2014. Biochar increases plant-available water in a sandy loam soil under an aerobic rice crop system. Solid Earth; 5: 939-952, https://doi.org/10.5194/se-5-939-2014. Downes GM, Hudson IL, Raymond CA, Dean GH, Michell AJ, Schimleck LR, et al. Sampling plantation eucalypts for wood and fibre properties: CSIRO publishing, 1997. https://doi.org/10.1071/9780643105287. Ehrburger P, Lahaye J, Wozniak E. 1982. Effect of carbonization on the porosity of beechwood. Carbon; 20: 433-439, https://doi.org/10.1016/0008-6223(82)90044-6.
of
Evans R, Downes G, Menz D, Stringer S. 2017. Rapid measurement of variation in tracheid transverse dimensions in a radiata pine tree. Appita Journal: Journal of the Technical Association of the Australian and New Zealand Pulp and Paper Industry; 70: 283,
ro
Farrell C, Ang XQ, Rayner JP. 2013. Water-retention additives increase plant available water in green roof substrates. Ecological Engineering; 52: 112-118, https://doi.org/10.1016/j.ecoleng.2012.12.098.
re
-p
Gray M, Johnson MG, Dragila MI, Kleber M. 2014. Water uptake in biochars: The roles of porosity and hydrophobicity. Biomass and Bioenergy; 61: 196-205, https://doi.org/10.1016/j.biombioe.2013.12.010.
lP
Greacen EL, Walker G, Cook P. Procedure for filter paper method of measuring soil water suction: Adelaide, SA, CSIRO Division of Soils, 1989. https://doi.org/10.4225/08/585ac5eadab2b.
na
Hardie M, Clothier B, Bound S, Oliver G, Close D. 2013. Does biochar influence soil physical properties and soil water availability? Plant and Soil; 376: 347-361, https://doi.org/10.1007/s11104-013-1980-x.
Jo ur
Hyväluoma J, Hannula M, Arstila K, Wang H, Kulju S, Rasa K. 2018. Effects of pyrolysis temperature on the hydrologically relevant porosity of willow biochar. Journal of Analytical and Applied Pyrolysis; 134: 446-453, https://doi.org/10.1016/j.jaap.2018.07.011. Hyvaluoma J, Kulju S, Hannula M, Wikberg H, Kalli A, Rasa K. 2018. Quantitative characterization of pore structure of several biochars with 3D imaging. Environmental Science and Pollution Research; 25: 25648-25658, https://doi.org/10.1007/s11356-017-8823-x. Jeffery S, Verheijen FGA, van der Velde M, Bastos AC. 2011. A quantitative review of the effects of biochar application to soils on crop productivity using meta-analysis. Agriculture, Ecosystems & Environment; 144: 175-187, https://doi.org/10.1016/j.agee.2011.08.015. Jirka S, Tomlinson T. 2015. State of the biochar industry 2014. International Biochar Initiative. Kinney TJ, Masiello CA, Dugan B, Hockaday WC, Dean MR, Zygourakis K, Barnes RT. 2012. Hydrologic properties of biochars produced at different temperatures. Biomass and Bioenergy; 41: 3443, https://doi.org/10.1016/j.biombioe.2012.01.033 Kumar M, Gupta R. 1995. Scanning electron microscopic study of acacia and eucalyptus wood chars. Journal of materials science; 30: 544-551, https://doi.org/10.1007/BF00354423. Lawrence V, Woo K. 2005. SilviScan: an instrument for measuring wood quality. Paprican Special Report, PSR; 550, 16
Journal Pre-proof Lehmann J. 2007. A handful of carbon. Nature; 447: 143, https://doi.org/10.1038/447143a. Lehmann J, Joseph S. Biochar for environmental management: science, technology and implementation: Routledge, 2015. https://doi.org/10.4324/9780203762264. Li S, Harris S, Anandhi A, Chen G. 2019. Predicting biochar properties and functions based on feedstock and pyrolysis temperature: A review and data syntheses. Journal of Cleaner Production; 215: 890-902, https://doi.org/10.1016/j.jclepro.2019.01.106. Lim TJ, Spokas KA, Feyereisen G, Novak JM. 2016. Predicting the impact of biochar additions on soil hydraulic properties. Chemosphere; 142: 136-44, https://doi.org/10.1016/j.chemosphere.2015.06.069.
of
Lu S, Zong Y. 2018. Pore structure and environmental serves of biochars derived from different feedstocks and pyrolysis conditions. Environmental Science and Pollution Research; 25: 30401-30409, https://doi.org/10.1007/s11356-018-3018-7.
-p
ro
Major J, Rondon M, Molina D, Riha SJ, Lehmann J. 2010. Maize yield and nutrition during 4 years after biochar application to a Colombian savanna oxisol. Plant and Soil; 333: 117-128, https://doi.org/10.1007/s11104-010-0327-0.
re
Masiello CA, Dugan B, Brewer CE, Spokas KA, Novak JM, Liu Z, et al. Biochar effects on soil hydrology. Biochar for Environmental Management. Routledge, 2015, pp. 575-594, https://doi.org/10.4324/9780203762264.
lP
Obia A, Mulder J, Martinsen V, Cornelissen G, Børresen T, 2016. In situ effects of biochar on aggregation, water retention and porosity in light-textured tropical soils. Soil and Tillage Research; 155: 35-44, https://doi-org.ezp.lib.unimelb.edu.au/10.1016/j.still.2015.08.002
Jo ur
na
O'Connor D, Peng T, Zhang J, Tsang DC, Alessi DS, Shen Z, Bolan NS, Hou D, 2018. Biochar application for the remediation of heavy metal polluted land: a review of in situ field trials. Science of the total environment; 619: 815-826, https://doi.org/10.1016/j.scitotenv.2017.11.132 Omondi MO, Xia X, Nahayo A, Liu X, Korai PK, Pan G. 2016. Quantification of biochar effects on soil hydrological properties using meta-analysis of literature data. Geoderma; 274: 28-34, https://doi.org/10.1016/j.geoderma.2016.03.029. Phillips E. 1941. The inclination of the fibrils in the cell wall and its relation to the compression strength of timber. Empire Forestry Journal: 74-78, Rasa K, Heikkinen J, Hannula M, Arstila K, Kulju S, Hyväluoma J. 2018. How and why does willow biochar increase a clay soil water retention capacity?. Biomass and Bioenergy; 119: 346-353, https://doi.org/10.1016/j.biombioe.2018.10.004. Ronsse F, van Hecke S, Dickinson D, Prins W. 2013. Production and characterization of slow pyrolysis biochar: influence of feedstock type and pyrolysis conditions. GCB Bioenergy; 5: 104-115, https://doi.org/10.1111/gcbb.12018. Singh BP, Cowie AL, Smernik RJ. 2012. Biochar carbon stability in a clayey soil as a function of feedstock and pyrolysis temperature. Environmental Science & Technology; 46: 11770-8, https://doi.org/10.1021/es302545b.
17
Journal Pre-proof Sohi SP, Krull E, Lopez-Capel E, Bol R. A Review of Biochar and Its Use and Function in Soil, 2010, pp. 47-82, https://doi.org/10.1016/s0065-2113(10)05002-9. Topp GC, Parkin G, Ferré TP, Carter M, Gregorich E. 2008. Soil water content. Soil sampling and methods of analysis’. 2nd edn.(Eds MR Carter, EG Gregorich) pp: 939-962, https://doi.org/10.1201/9781420005271. Tryon EH. 1948. Effect of charcoal on certain physical, chemical, and biological properties of forest soils. Ecological Monographs; 18: 81-115, https://doi.org/10.2307/1948629. Walker JC. Primary wood processing: principles and practice: Springer Science & Business Media, 2006. https://doi.org/10.1007/1-4020-4393-7.
of
Wang Y, Zhang L, Yang H, Yan G, Xu Z, Chen C, et al. 2016. Biochar nutrient availability rather than its water holding capacity governs the growth of both C3 and C4 plants. Journal of Soils and Sediments; 16: 801-810, https://doi.org/10.1007/s11368-016-1357-x.
ro
Watson A, Dadswell H. 1961. Influence on fibre morphology on paper properties. Part 1. Fibre length. Appita J; 14: 168-178,
re
-p
Weber K, Quicker P. 2018. Properties of biochar. Fuel; 217: 240-261, https://doi.org/10.1016/j.fuel.2017.12.054.
lP
Wildman J, Derbyshire F. 1991. Origins and functions of macroporosity in activated carbons from coal and wood precursors. Fuel; 70: 655-661, https://doi.org/10.1016/0016-2361(91)901819.
na
Wilkes J. 1988. Variations Inwoodanatomy Within Species of Eucalyptus. IAWA Journal; 9: 13-23, https://doi.org/10.1163/22941932-90000461.
Jo ur
Zbonak A, Bush T, Grzeskowiak V. Comparison of tree growth, wood density and anatomical properties between coppiced trees and parent crop of six Eucalyptus genotypes. Improvements and culture of eucalypts, IUFRO conference. Citeseer, 2007, pp. 22-26, Zeriouh A, Belkbir L. 1995. Thermal decomposition of a Moroccan wood under a nitrogen atmosphere. Thermochimica acta; 258: 243-248, https://doi.org/10.1016/00406031(94)02246-K. Zhang J, You C. 2013. Water Holding Capacity and Absorption Properties of Wood Chars. Energy & Fuels; 27: 2643-2648, https://doi.org/10.1021/ef4000769. Ziemińska K, Butler DW, Gleason SM, Wright IJ, Westoby M. 2013. Fibre wall and lumen fractions drive wood density variation across 24 Australian angiosperms. AoB PLANTS; 5, https://doi.org/10.1093/aobpla/plt046. Zornoza R, Moreno-Barriga F, Acosta JA, Muñoz MA, Faz A. 2016. Stability, nutrient availability and hydrophobicity of biochars derived from manure, crop residues, and municipal solid waste for their use as soil amendments. Chemosphere; 144: 122-130, https://doi.org/10.1016/j.chemosphere.2015.08.046.
18
Journal Pre-proof Figure captions Figure 1. Top view of a stem sample cut from a eucalypt tree (E. macrohyncha) for pyrolysis and airdry density measurement (left) and for wood cell structure analysis using the Silviscan3 (right, dashed strip) Figure 2. Transmitted-light cell scan micrographs obtained from Silviscan3 cell scanner (100 x) of three species of Eucalyptus with different wood densities. A: low density (653 kg m-3; E. nitens), B: medium density (788 kg m-3; E. polyanthemos), C: high density (960 kg m-3; E. polybraktea). White to light grey circular structures are cell walls, dark grey circular structures are fibre lumen, dark grey horizontal band structures are radial parenchyma. Fibre lumen diameters decreasing and fibre cell wall thickness increasing from low density wood (A) to high density wood (C). Fibre lumen almost invisible as cell walls take up almost all area (C).
ro
of
Figure 3. Relationships between fibre wall thickness and radial fibre lumen diameter with wood density. Values for fibre wall thickness (n=3), radial fibre lumen diameter (n=3) and wood density (n≥5) represent means, bars represent ±mean standard error and R2 and p-values are for regression fit. Shaded areas indicate 95% confidence regions for regression fit.
re
-p
Figure 4. Relationships between feedstock wood density with biochar maximum water holding capacity and plant available water. Values for wood density (n≥5), water holding capacity (n=3) and plant available water (n=3) represent means, bars represent ±mean standard error and R2 and pvalues are for regression fit. Shaded areas indicate 95% confidence regions for regression fit.
Supplementary material
na
lP
Figure 5. Relationship between biochar water holding capacity and biochar plant available water. Values represent means (n=3), bars represent mean standard error and R2 and p-values are for regression fit. Shaded area indicates 95% confidence region for regression fit.
Jo ur
Figure 6. Boxplots of species air-dry wood density (top), radial fibre lumen diameter (middle) and fibre wall thickness. Different letters indicate significant differences among species means as determined from one-way ANOVA. Data points represent individual sample means (all P < 0.001).
19
Journal Pre-proof Table 1. Mean air-dry wood density of the feedstock material and resulting dry bulk density, water holding capacity (WHC) and plant available water (PAW) for biochar produced from 18 Eucalyptus species grown in a field experiment. Species listed in order of increasing wood density. Mean ± SE in parentheses.
Biochar bulk
Biochar WHC
Biochar PAW
Species
(kg m-3)
density dry (kg m-3)
(% v/v)
(% v/v)
E. pauciflora
571.9 (18.82)
140.3 (2.88)
44.5 (0.45)
41.6 (0.45)
E. regnans
598.8 (24.36)
158.7 (2.21)
41.6 (0.10)
36.9 (0.10)
E. delegatensis
628.3 (20.66)
157.7 (2.80)
43.6 (0.22)
38.3 (0.22)
E. nitens
653.2 (21.76)
186.0 (2.82)
E. cypellocarpa
666.9 (24.61)
E. muelleriana
41.3 (0.44)
159.1 (0.84)
44.2 (0.29)
40.0 (0.29)
666.9 (23.89)
179.5 (0.99)
43.5 (0.16)
38.7 (0.16)
E. dives
685.1 (22.70)
189.8 (11.2)
43.2 (0.48)
37.3 (0.48)
E. obliqua
686.5 (22.19)
158.6 (3.63)
40.8 (0.24)
35.8 (0.24)
E. radiata
688.5 (25.27)
195.6 (7.65)
47.8 (0.64)
42.2 (0.64)
E. macrohyncha
698.6 (22.51)
178.2 (3.36)
42.8 (0.60)
37.9 (0.60)
717.3 (17.64)
189.7 (4.14)
44.2 (0.35)
38.1 (0.35)
738.5 (19.66)
280.9 (38.1)
44.0 (0.71)
36.0 (0.71)
755.7 (20.71)
190.7 (4.51)
41.5 (0.61)
35.6 (0.61)
E. leucoxylon
778.3 (15.79)
210.4 (8.04)
40.4 (0.06)
35.1 (0.06)
E. polyanthemos
788.3 (18.65)
249.2 (10.9)
39.0 (0.19)
33.0 (0.19)
E. microcarpa
894.6 (10.13)
322.9 (9.84)
38.0 (0.53)
30.0 (0.53)
E. tricarpa
925.4 (10.91)
304.8 (11.5)
36.9 (0.18)
29.6 (0.18)
E. polybractea
959.9 (40.53)
294.3 (2.45)
35.5 (0.08)
28.6 (0.08)
E. melliodora E. globoidea
lP
na
Jo ur
E. viminalis
-p
47.4 (0.44)
re
ro
of
Wood density air-dry
20
Journal Pre-proof Table 2. Particle size distribution (% w/w) for all biochar types after remixing
12.5-10 mm
10-6.3 mm
6.3-4 mm
4-2 mm
2-1 mm
<1 mm
biochar
16.9
26.4
19.9
15.3
7.4
14
Jo ur
na
lP
re
-p
ro
of
`
21
Journal Pre-proof Graphical abstract
Highlights Biochars differ greatly in their properties due to feedstock differences We studied the effect of feedstock cell structure on biochar water retention Biochar from low density eucalypt wood has greater plant available water retention Greater wood density is due to greater cell wall thickness and smaller cell diameters
Jo ur
na
lP
re
-p
ro
of
Wood density can be used as a proxy for wood cell structure
22
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6