1
MSc THESIS
1Roots foraging phosphorus preference in Scots pine (Pinus sylvestris)
23
Details of Student
4Name: Kaiyu Lei
5
Student ID: 12739774
6
Major: Earth Science - Environmental Management track
7
Institute: Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam
8 9
Details of Thesis
10Number of credits: 30EC
11
Period: 01/11/2020 – 01/05/2021
12
Research institute: Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam;
13
School of Geoscience, The University of Edinburgh 14
Examiner: Prof. Dr. Bol Roland
15
Assessor: Prof. Dr. Albert Tietema
16
Daily supervisor: Prof. Dr. Sohi Saran, Hamish Creber
17
Date: 05/2021
CONTENTS
19CONTENTS ... 2
20Abstract ... 4
21Introduction ... 5
22Soil phosphorus in forest systems ... 5
23
ECM fungi in forest systems ... 5
24
Biochar as a latent phosphorus fertilizer ... 6
25
Roots response to phosphorus patches ... 8
26
Roots response to biochar derived phosphorus with ECM fungi ... 8
27
Materials and Methods ... 11
28Conifer tree species ... 11
29
Rhizobox experiment design ... 11
30
Sampling areas and methods ... 13
31
Growing media properties ... 14
32
Biochar production and properties ... 14
33
Root morphological tracking ... 15
34
Optical microscopy ... 16
35
Scanning electron microscopy ... 16
36 Fluorescent microscopy ... 17 37 Chemical analysis ... 17 38 Data analysis ... 17 39
Results ... 18
40Root growth responses ... 18
41
Chemical property alterations ... 20
42
Fungi occupation ... 25
43
Root colonization of biochar particles ... 27
44
Discussion ... 34
45The priority effect and avoidance effect ... 34
46
Relationship among root system, ECM fungi, biochar and soils ... 34
47
Multiple strategies for ECM roots to colonize wood-derived biochar ... 38
3
Biochar on the management of coniferous forest ... 39
49
Conclusion ... 41
50References ... 42
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90Abstract
91Phosphorus (P) is a limiting nutrient in many forest systems. As the depletion of mineral P fertilizer 92
worldwide and erosion of forest soils, biochar has been considered as a sustainable P fertilizer that 93
benefits those P-deficient forest systems. To further prove biochar’s advantages when plants are 94
faced with P-deficient situations, an understanding on how biochar impacts root systems and the 95
mechanisms for plants to forage P from biochar are required. This study involves a rhizobox 96
experiment to observe the development of root systems in Scots pine (P. sylvestris) and how 97
seedlings respond to biochar derived P and mineral P fertilizer. The colonization strategies can 98
provide a clue in understanding biochar and plant-soil interactions. The vascular cambium zone 99
derived biochar (VCZ), soft wood pellet biochar (SWP), soft wood pellet woody-100
biochar infused with triple superphosphate fertilizer (SWP+P), triple superphosphate (TSP) and no 101
application (Blank) were used. 102
By observing new root growth responses to different types of P sources, there’s a strong priority 103
effect and a strong avoidance effect of roots to VCZ biochar and TSP fertilizer treated areas, 104
respectively. The priority effect to SWP biochar and the avoidance effect to SWP+P biochar is 105
weaker. The dominant factor that causes these two different responses is likely available P. Other 106
factors including pH, micronutrients and physical structure of different P sources may also have 107
some influences. Scots pine is a typical mycotrophic tree species with ectomycorrhizal (ECM) fungi 108
associations. Biochar not only stimulates root generation and occupation by providing available P, 109
but also provides an ideal proliferation bed with an accessible carbon source for ECM fungi. 110
Microscopy images on samples selected from three-year pot experiments show potential strategies 111
for Scots pine roots to colonize biochar may be (a) through surface interaction between biochar and 112
roots which stick to or penetrate inside biochar by following the cracks; (b) roots interacting with 113
biochar with ECM fungi mantle layer which sticks to the surface of biochar with emanating 114
extraradical mycelium strands providing the transition interface; (c) hyphae contact and penetrate 115
biochar directly through cracks and pores; (d) hyphae attract invertebrate grazers to stimulate the 116
fungi-dominated nutrient cycling and obtain additional nutrients by proliferating inside dead grazers. 117
The understanding of the potential strategies for Scots pine roots to colonize biochar particles 118
directly supports biochar’s positive application on the management of coniferous forests in Scotland 119
and other temperate zones. With the improvement to both the root generation and P supply, woody-120
residue-derived biochar can be considered as an effective soil improvement technology in 121
commercial coniferous forests, with P-deficient soils and in the regeneration of coniferous forests. 122
In turn, this can help to close the P cycle in vulnerable forest systems, prevent soil P loss and 123
maximize the potential biomass production. 124 125 126 127 128 129 130 131 132
5
Introduction
133
Soil phosphorus in forest systems
134
Phosphorus (P) is one of the most important nutrients in a forest ecosystem. The P in the soil 135
generally comprises inorganic P and organic P. Trees can generally uptake the dissolved phosphate 136
within inorganic P, so-called available P or soluble P. The increase of atmospheric nitrogen (N) 137
deposition results in N accumulation and saturation in a forest system (Aber et al., 1989), which can 138
cause P to gradually becomes the barrier that prevent the forest from flourishing in some forest 139
systems nowadays (Attiwill & Adams, 1993; Crowley et al., 2012), especially in an intermediated 140
and old aged forest (Davidson et al., 2004) where the P released by weathering from primary 141
minerals occulated with secondary minerals and fixed with soil organic matter particles. In addition, 142
the constant leaching and erosion during soil development at time scales of millennia can increase 143
the P limitation in a forest system despite the P input from precipitation and atmospheric particles 144
(Lambers et al., 2008). Jonard et al. (2015) observed a decreasing trend of P concentration in 145
temperate coniferous forest soils which had an original P deficiency. This enhanced reduction of P 146
may be due to higher N deposition combined with local climatic stress that stimulate the growth of 147
coniferous trees, and therefore increase the demand to support their growth. 148
Available P is labile in the soil. It can decline and be transferred into sorbed phosphate within a short 149
time. With soil acidification in a forest system through time, the pH alteration triggers the 150
transformation from Ca-phosphate to Fe, Al and Mn-phosphate which are significantly insoluble 151
(Schlesinger, 2005). Subsequently, less P is available for trees which can further reduce the trees’ 152
nutrition state and limit growth. Nevertheless, the soil P transformation could be reversed with high 153
concentration of Ca and Mg micronutrients, which could competitively bond the free phosphate and 154
alter the soil pH to a less acidic environment (Moran et al., 2000). In addition to the micronutrients 155
that can inhibit or strength the availability of P in forest soils, other factors including soil moist 156
content (Zhang et al., 2020a) and soil texture (Lambers et al., 2008) are essential to be considered. 157
These factors impact the types and quantity of free radical cations that exist in the soil which have 158
different priority and intensity to bond with free phosphate. 159
Faced with a different availability of P in the soil, forest systems have generated two different P use 160
strategies. In a P-rich environment, forest systems adopt a ‘acquiring system’ in which trees and 161
microorganisms tend to uptake P directly from primary-mineral-released P pools, while in a P-162
deficient situation, a tight P ‘recycling system’ emerges to close the P cycling loop, maximize P pool 163
use efficiency and sustaining the P supply for generations (Lang et al., 2016). Within a mycotrophic 164
forest, a P ‘recycling system’ always highly replies on mycorrhizal fungi despite of the root 165
morphological alterations (Attiwill & Adams, 1993; Lang et al., 2016). 166
ECM fungi in forest systems
167
Ectomycorrhizal (ECM) fungi is a common form of fungi that is symbiotic with trees and shrubs in 168
temperate and boreal zones, and is essential in maintaining the health of the world’s forest 169
ecosystems (Courty et al., 2010; Smith & Read, 2010) despite only being a symbiont of less than 170
2% of plant species (Brundrett & Tedersoo, 2018). ECM fungi comprise of a Hartig net, fungal 171
mantle and extraradical mycelium strands (Smith & Read, 2010). In contrast to arbuscular 172
mycorrhizal fungi which penetrates the cortex cells, ECM fungi roots onto the host plant by 173
penetrating between epidermal and cortical cells to form dense sheath around short lateral roots. 174
The majority of studies prove that the Hartig net and mantle structure cannot directly uptake 175
nutrients from the outside solution (Parladé et al., 2014; Taylor & Peterson, 2005), but perform as 176
the storage and transportation of nutrients between extraradical mycelium strands and the host plants. 177
The function of extraradical mycelia strands is mainly nutrient uptake, but they also have been found 178
to be beneficial for the spread and proliferation of ECM fungi. With a dense mycelium strands below 179
ground, a network may form to facilitate nutrient movement between individual hosts (Cairney, 180
2011). 181
The emergence of ECM fungi stimulates fine root growth in the host plant which extends the P 182
depletion zones by being exposed in a larger proportion of soils. But the distance and the volume of 183
occupation vary across different ECM fungi species, along with the P uptake efficiency (Cairney, 184
2011). Chemically, ECM fungi excrete organic acids including oxalic, citric and malonic acids to 185
accelerate the availability P transformation from both primary minerals and secondary minerals 186
(Cairney, 2011), and then perform as a P transport interface between the soil and the plant, where 187
the soluble P in the soil is taken up and transferred to intraradical hyphae within the root system 188
(Smith & Read, 2010). In addition, ECM fungi can participate in the decomposition of complex soil 189
organic matter along with saprotrophic fungi through releasing organic acids. By releasing 190
phosphatases, different forms of organic P contained in organic matter like phosphomonoesters and 191
phosphodiesters are transferred to inorganic P and become available for plants to uptake (Cánovas, 192
2019). Nevertheless, the degree and acceleration of ECM fungi on releasing organic acids and 193
subsequent degradation vary from different fungal genotypes, soil nutrient properties and bacteria 194
populations (Cairney, 2011). 195
Studies (Cánovas, 2019; Plassard et al., 2011) have proved that more soluble P can be obtained by 196
plants with ECM fungi associations than those without. The degree of improvement can up to 80% 197
but varies from different ECM fungi species according to different P mobilization processes 198
(Cánovas, 2019). In turn, the diversity and abundance of ECM fungi is significantly impacted by 199
the P formations that exist in the soil or are provided by extra input (Twieg et al., 2009; Zavišić et 200
al., 2016). The alteration of P formations in the soil from abiotic and biotic factors, can cause soluble 201
P to decline sharply through time, which leads to a reduction in belowground biomass and 202
scavenging area of the ECM extraradical hyphae (Teste et al., 2016). High available P content in 203
crop-derived biochar results in lower mycorrhizal fungi colonization in the root system or even fungi 204
mortality (Solaiman et al., 2019) which indicates that the emergence of ECM fungi is significantly 205
impacted by the availability of P in the soil and in additional nutrient applications. 206
Biochar as a latent phosphorus fertilizer
207
Biochar is the product of biomass thermochemical processes and consists of different forms of 208
carbon and nutrients (Sohi et al., 2009). The physical and chemical properties of biochar are highly 209
dependent on the nature of feedstock (Lehmann et al., 2011), thermochemical process methods 210
(Bruun et al., 2012) and the degree of oxidation (Cheng et al., 2006). Most types of biochar have 211
been found as an ideal soil amendment both in agriculture (Haider et al., 2017; Saha et al., 2019; 212
Sohi, 2012) and forest systems (Bruun et al., 2013; Thomas & Gale, 2015). Biochar enhances the 213
physical structure of the soil (Haider et al., 2017; Rogovska et al., 2014), increases soil nutrient 214
content and retention ability (Johannes et al., 2003; Lehmann et al., 2011; Zhang et al., 2012; Zhang 215
et al., 2020b), and alleviates soil pollutants (Li et al., 2016; Park et al., 2011). 216
7
With the continuous growing demand for P fertilizer worldwide (FAO, 2019), biochar derived P 217
fertilizer is an attractive option with the promotion for a circular economy to moderate a P crisis to 218
a certain range. Compared with mineral P fertilizer with extremely high available P proportion and 219
concentration, biochar generally has a low to medium available P. But it is still effective for plants 220
improvement on P-deficient soils (Prendergast-Miller et al., 2014) as the amount of dissolvable 221
phosphates and water-soluble P is sufficient within biochar (Zhang et al., 2016). Total P content in 222
biochar can be up to 13.5% (Li et al., 2020), but varies from different types of feedstock, pyrolysis 223
processes and oxidation processes (Kim et al., 2018). 224
Since most types of the biochar, including biochar derived from woody materials, are of 225
comparatively low or medium available P to mineral P fertilizers (Sohi et al., 2009; Warnock et al., 226
2007). Two strategies have been applied. The first method is to alter the pH of biochar (Tumbure et 227
al., 2020). A change in the pH of biochar can trigger the transformation from sorbed P to soluble P. 228
As P is toughly bound to positive cations in a highly acidic or calcareous environment, it is hard for 229
plants to uptake. But it becomes soluble in a slightly acidic condition due to the presence of H+
230
which make Fe-oxides protoned and weakening the P sorption effect. Based on the soil pH, a specific 231
type of biochar can be designed to alter the target soil pH to an appropriate range which will then 232
trigger the transformation from sorbed P to soluble P. 233
The second method is to increase the available P of biochar by adding additional P to the engineered 234
biochar through infusing, soaking in a P-rich media (Pandey et al., 2020; Zheng et al., 2019) or 235
immobilizing a P-rich bacteria (Arun et al., 2020; Moreno-Bayona et al., 2019; Pandey et al., 2020). 236
Engineered biochar can have better adsorption capacity and CEC ability through physical, chemical 237
or other combined methods (Pandey et al., 2020). High adsorption capacity guarantees that the unit 238
of biochar can load more P and P related bacteria while high CEC of biochar accelerates the available 239
P transformation. In P-fixed-biochar, there is initially high available P (Yao et al., 2013), therefore, 240
biochar application can work as an effective fast-efficient nutrient source to stimulate the 241
germination and growth of plants. The large amounts of liming agents, such as Ca, Mg, which are 242
provided by the biochar (Cui et al., 2011) reduce the concentration of Fe and Al in the soil. This 243
reduction avoids latent Al toxicity and results in less sparingly soluble Fe-P and Al-P bonds. The 244
Ca-P and Mg-P bonds are weaker when compared with Fe-P and Al-P bonds and are comparatively 245
readily dissolved into soluble P in a slightly acidic environment. Over time, this may, to some extent, 246
increase the soluble P in the soil and work as a slow-released P fertilizer. 247
Some types of biochar referred to as a potential P fertilizer have also been found to improve the soil 248
environment by altering soil pH, water holding capacity and aeration. These improvements provided 249
by biochar are likely to stimulate both the above-ground plant growth (Jones et al., 2012) and roots 250
development (Prendergast-Miller et al., 2014; Xiang et al., 2017). Biochar-altered soil has 251
advantages in soil tensile strength and bulk density, compare with untreated soils, that enhances root 252
patterns and generation (Lehmann et al., 2011). These root improvements are shown on better 253
primarily root germination and further nutrient mining activity by plant roots. Abiven et al. (2015) 254
have found a significantly increase in the development of primary and lateral root systems. Root 255
biomass, length and density have been found to increase with biochar application by a majority of 256
studies (Xiang et al., 2017). Makoto et al. (2009) found that the presence of biochar not only increase 257
roots biomass and surface area, but also the root tip numbers increased. Through morphological 258
development, a well-developed root system intensifies the nutrients uptake ability and alleviates soil 259
water restriction (Xiang et al., 2017). Combined with the self-promotion effect of biochar-carbon 260
(Lehmann et al., 2011), biochar prompts the plant germination ratio and growth, including pine 261
(Pinus spp.) roots in a forest system (Wardle et al., 1998). In some cases, the combination of 262
mycorrhizal fungi and biochar addition increases the number of nodules and promotes doubled roots 263
colonization compared with untreated samples (Solaiman et al., 2010). 264
Studies have confirmed that biochar can increase the colonization and abundance of ECM fungi, 265
but little difference has been found in the diversity of the ECM fungi community (Robertson et al., 266
2012). The impact of biochar on ECM communities is highly variable and depends on the dosage, 267
biochar type and soil properties (Ogawa & Okimori, 2018). A high dosage of biochar (>75% by 268
volume), which increasing soil P and N concentration, may cause mortality and toxification of ECM 269
fungi community and otherwise negatively affect the growth of the aboveground structures (Choi et 270
al., 2009; Solaiman et al., 2019). However, the influence of P-loaded biochar on the ECM fungi has 271
not been fully studied. 272
Roots response to phosphorus patches
273
Despite the strategies of roots to explore and uptake P in the soil highly depending on plant 274
ontogenetic regulator and inherent genetic factors (Niu et al., 2013), roots of different plants adopt 275
some strategies to explore P-sufficient patches. 276
There are mainly five strategies for plants to mitigate P-deficient stresses (Lambers et al., 2006). 277
First, plants may increase the root-hair production. As the root hair is the initial part for P acquisition, 278
it has been considered the most competitive advantage for roots to uptake P in the soil. Generally, 279
the root hair significantly increases at the beginning of the lifespan under low P availability and 280
decreases in response to higher P availability (Ma et al., 2001). Second, in some cases, root clusters 281
increase with respect to available P reduction in the soil. In addition, the endogenous P status of 282
plants will also perform as a regulator in the formation of cluster roots (Neumann et al., 2000). Root 283
clusters catch available P in the soil by exudation of carboxylates through an anion channel when 284
facing P-deficiency. The mobilization effect of carboxylates on both the inorganic and organic P 285
makes more soluble P available (Lambers et al., 2013). Faced with P-deficient stresses, some plants 286
have developed a remarkable efficiency to resorb P in the soil with the cost of losing P-uptake 287
regulation (de Campos et al., 2013). However, for those plants that have a symbiotic relationship 288
with mycorrhizal fungi, sacrifices are less. Mycorrhizal fungi interact with roots to accelerate 289
available P transformation and increase P depletion zones which significantly mitigate P-deficient 290
stresses. 291
Compared with mineral P fertilizer that can deconstruct and solubilize in the soil during a short time, 292
biochar derived P fertilizer takes years or decades to be fully deconstructed and degraded. In 293
addition, the complex structure of biochar should also be considered as it maybe easier for roots to 294
colonize. Prendergast-Miller et al. (2014) have observed roots contact biochar particles and some 295
of them penetrate inside while a number of studies (Ascough et al., 2010; Hammer et al., 2014; 296
Jaafar et al., 2014) have observed that the hyphae colonize the surface following cracks and 297
penetrate inside to proliferate in those pores. Nevertheless, mechanisms for ECM roots to colonize 298
and uptake P from different types of biochar are limited. 299
Roots response to biochar derived phosphorus with ECM fungi
300
Mycotrophic trees have evolved a nutrient uptake strategy that highly dependent on the development 301
9
of ECM fungi within thier root systems (Verma & Reddy, 2020), especially in their early 302
establishment stages (Nara, 2006). Experiments (Robertson et al., 2012; Verma & Reddy, 2020) 303
have shown that biochar can simulate the early growth of conifer seedlings by providing better soil 304
physicochemical properties (Kluber et al., 2010), soil enzymes activities and ECM fungi 305
colonization. As the benefits of ECM fungi on the health and development of the natural forest 306
system have been illustrated and summarized by Courty et al. (2010), there are still limited studies 307
on its co-benefits when applied with biochar to the forest trees. 308
Biochar-altered soil enhances root patterns and generation (Lehmann et al., 2011), including higher 309
root biomass, longer root length, denser root density (Abiven et al., 2015) and more root tips 310
(Makoto et al., 2009). Through morphological development, the well-developed root system 311
intensifies the nutrients uptake ability and alleviates the water restriction (Xiang et al., 2017). 312
Combined with the self-promotion effect of biochar-carbon (Lehmann et al., 2011), biochar prompts 313
the germination ratio and growth, including pine roots in a forest system (Wardle et al., 1998). In 314
addition, the root morphological alterations and proliferation trigger the reorganization of the plant 315
cytoskeleton which leads to the changes of ECM fungi within root systems (Cánovas, 2019). 316
Symbioses with ECM fungi (Fig. 1), the acquisition efficiency of P for conifer seedling roots is 317
enhanced (Verma & Reddy, 2020), resulting in higher P concentration in aboveground needles (Choi 318
et al., 2009). The possible reason is that with extra ECM root accompanied with extraradical 319
mycelium strands, a large volume of soil is accessible for ECM fungi. The farther distance ECM 320
fungi explores, the more P sources become accessible beyond original depletion zones (Cánovas, 321
2019). What’s more, in P-deficient soil, the signal released by the plant stimulates the excretion of 322
the P by ECM fungi which accelerates the inorganic P transformation from organic P formation 323
(Cánovas, 2019). In turn, the P abundance stimulates the development of ECM fungi and its external 324
hyphae. 325
326
Fig. 1 Schematic diagram of the relationship among root system, ECM fungi, biochar and soil, highly
327
instinct processes in P. sylvestris tree roots in this study are emphasized in thick red. ECM fungi and 328
roots system has a symbiotic relationship. Biochar applied provide fungi with extra oxidable carbon and 329
provide root with extra P and other nutrients. The porosity in biochar is beneficial for them to proliferate. 330
In turn, fungi and roots can contact and penetrate biochar to accelerate its deconstruction and 331
mineralization. The soil is the foundation of root systems while root systems reversely continuous 332
altering the soil. Meanwhile, the soil, to some extent, provide C and N for ECM fungi. ECM fungi 333
accelerate the degradation of organic matter in the soil. 334
Consequently, applying biochar to mycotrophic trees not only directly increases the soil 335
physicochemical properties and nutrient uptake efficiency, but also provides nutrients and shelter 336
which stimulate the abundance of the ECM fungi (Fig. 1). In turn, the total biomass of the root 337
system as well as the aboveground part is strengthened (Choi et al., 2009). 338
In this study, the response of Scots pine (P. sylvestris) roots to biochar derived P sources and mineral 339
P fertilizer is investigated. Based on those advantages of ECM fungi symbiosed trees to uptake P in 340
a P-deficient growing media, this study makes the hypotheses that (a) Scots pine roots have a priority 341
to explore and uptake P from different types of biochar, accompanied with dense extraradical 342
mycelium network, while the areas treated with mineral P fertilizer will have less ECM fungi and 343
roots; In addition, due to the stronger exploration ability of those extraradical mycelium strands, (b) 344
the occupation priority of mycelia will be better than roots. (c) And to maximize the P use efficiency, 345
multiple strategies will be adopted by root systems to fully use those input P sources. 346
11
Materials and Methods
348
Conifer tree species
349
The Scots pine (P. sylvestris) was chosen (Fig. 2) since it is one of typical mycotrophic tree species 350
with ECM fungi and of vital importance for coniferous forest protection and management in 351
Scotland (Leake et al., 2001). To exclude the negative influence of other variables and observe the 352
roots responses to greatest extent within restricted time, one-year-old Scots pine seedlings growing 353
in the same growing media were taken from the same nursery whose roots length was from 20cm 354
to 30cm. Their roots were rinsed carefully to flush the original growing media away, avoid nutrient 355
residue left on them and separate the entangled root system before being transferred into rhizoboxes. 356
However, roots were not sterilized or trimmed to sustain original fungi population as much as 357
possible and dimmish the ‘transplant shock’. 358
359
Fig. 2 The Scots pine (P. sylvestris) seedling used in the rhizobox experiment.
360
Rhizobox experiment design
361
A rhizobox experiment (Fig. 3) was set up to observe roots response to different P sources. 362
Rhizoboxes were made of PVC material with the size of 40cm height x 30cm width x 0.6cm depth 363
(Leake et al., 2001). The size was customized for one-year-old Scots pine seedlings considering the 364
diameter of the stem and the length and density of the root system. The whole rhizobox was 365
comprised of five sheets which were back, sides, bottom and front sheets. With blank and opaque 366
sheets on the other sides, the lights in the glasshouse could not shine in and the internal environment 367
was as the same as in the soil. A transparent front sheet made it easy and convenient for regular 368
observation and photographing. 369
As all sheets were delivered separately, using the glue to stick the sides and bottom sheets to the 370
back sheet and waiting for 24 hours until they were completely fixed. Hereafter, clipping the front 371
sheet to the other parts when the growing media was filled into. All the PVC sheets were rinsed, 372
alcohol sterilized and dried before assembling. 373
The growing media was freshly made and sterilized 2 weeks prior to experimental setup. It was 374
composed of 50% peat and 50% sand to ensure the key nutrients contained within were at a deficient 375
level. The growing media had no added nutrients and had the equivalent soil nutrient regime of a 376
very poor forest soil (Pyatt et al., 2001). After the rhizoboxes were assembled, filtering the growing 377
media through 2mm sieve, filling in rhizoboxes with approximately 720 kg for each and keeping 378
the bulk density, pH and moisture content for each rhizobox within 10% variation range to ensure 379
the same growing environment provided at the beginning. If the pH were not in the certain range 380
(4.5 ± 0.5), using 0.01 mol/L NaOH solution and 0.01 mol/L H2SO4 to regulate. The growing media
381
in the rhizoboxes was fully watered and stabilized for 48 hours before planting. 382
Rizoboxes were averagely divided into twelve unit-blocks with a size of 10cm width x 10cm height 383
(Fig. 3). Four P sources were applied which are soft wood pellet (SWP) biochar with low available 384
P content, vascular cambium zone (VCZ) derived biochar with medium available P content, triple 385
superphosphate infused soft wood pellet (SWP+P) biochar with high available P content and triple 386
superphosphate (TSP) fertilizer with extreme high available P content. The dosage of them are 3.60 387
g, 3.60 g. 4.56 g and 0.96 g, respectively. The treatment materials (biochar & fertilizer) were directly 388
applied in the low middle unit-block for each rhizobox without any plastic meshes (Leake et al., 389
2001). Filling the porosity between treatment materials with growing media to ensure the ratio of 390
treatment materials to growing media as 1:1. Each treatment was arranged with five replicates. With 391
one blank control group, end up with 25 rhizoboxes in total. 392
Subsequently, Scots pine seedlings were planted at a fixed location in the rhizobox approximately 393
in the central unit-block. To ensure the root system could consistently detect the P fragment, one 394
root was placed adjacent to the treatment area and the initial distance between that root and treatment 395
area was 2cm, with orientation pointing to the center of the treatment area. Each rhizobox was 396
watered weekly by removing the front sheet and spraying the growing media evenly with distilled 397
water in a mist sprayer. Growing media was kept moist ensuring that there was no excess water or 398
vertical water movement or pooling. 399
400
Fig.3 Schematic diagram of rhizobox with treated area.
13
Rhizoboxes were covered with the front transparent sheets, fixing four corners with clips. Hereafter, 402
the rhizoboxs were then fully covered with 6cm thick polystyrene boards which were the same size 403
as the front and back sheets, to make sure there was no extra light irradiated inside, with internal 404
conditions having consistent low light penetration. Polystyrene was also used to loosely cover the 405
tops of the rhizoboxes, ensuring soil respiration was not impeded, to reduce soil surface light and 406
temperature gradients. The covered rhizoboxes were stacked at an angle of 60°, grouping them 407
according to different treatment materials, dividing them into five blocks, and placing them 408
separately to ensure that the branches of above-ground parts do not touch with each other. The 409
experiment used a randomized block design, with one replicate of each treatment randomly placed 410
in each block. Setup made sure that all the pine seedlings were horizontal and vertical staggered 411
(Fig. 4). All the rhizoboxes were placed in the glasshouse for 10 weeks. 412
413
Fig. 4 Placement schematic diagram of a group of rhizoboxes in the greenhouse.
414
Sampling areas and methods
415
During experimental setup, all the materials, including growing media, different treated materials 416
and different treated areas were sampled for analysis. After 10 weeks cultivation, all the treatment 417
and sample areas were separately using prepared moldboards to ensure accurate collection (Fig. 5). 418
Then, fully mixing the materials within each sample area to reduce intra-sample area variation and 419
translocating them into the refrigerator under 4°C. Samples were oven dried them at 30°C to 90°C 420
and grind in a ball grinder before further analysis. The temperature of drying various depending on 421
specific testing items. 422
423
Fig. 5 Sampling areas, the sector diameter of Part 1, 2 and 3 are 3cm while the Part 5, 6, 7 and 8 are
1.5cm. 425
Growing media properties
426
The growing media was comprised of 50% peat and 50% sand. The properties of the growing media 427
were listed as below (Table 1): 428
Table 1 Chemical properties of growing media (mean ± SD, n = 3)
429
pH NH4+-N (mg/g) NO3--N (mg/g) Available P (mg/g) Total P 4.41 ± 0.02 1.41 ± 0.06 1.04 ± 0.02 0.046 ± 0.020 -
Fe (mg/kg) Al (mg/kg) Ca (mg/kg) Mg (mg/kg) Cd (mg/kg)
64.93 ± 3.01 9.52 ± 0.54 135 ± 7 468 ± 24 0.040 ± 0.010 Basically, the growing media was slightly acidic and of low total key nutrients content. All the metal 430
contents were in a low level compared with original forest soil. 431
Biochar production and properties
432
All types of biochar used in the rhizobox experiment were derived from woody materials and 433
supplied by the UK Biochar Research Centre, Edinburgh, UK. Production of the biochar was carried 434
out using the stage III continuous feed pyrolysis rotary kiln manufactured by Ansac Pvt Ltd. 435
Feedstock was sourced from the Petersmuir sawmill, BSW Ltd, Scotland. 436
The vascular cambium zone (VCZ) derived material was found to comprise of 40% wood and 60% 437
bark by volume, incorporating the vascular cambium. All material was derived from the sawmill 438
residue produced from Sitka spruce (Picea sitchensis) logs during the process of ring debarking 439
(Rathnayake et al., 2021). 440
The VCZ material was pyrolyzed at nominal HTT of 550°C temperature with a mean residence time 441
of 12.0 minutes. Material was heated at a rate of 78°C min-1 and the residency time at HTT was 3.9
442
minutes. The soft wood pellet (SWP) biochar was produced using Puffin Pellets derived from mixed 443
softwood sources to the same pyrolysis conditions. SWP+P was infused with TSP at a rate of 230 g 444
P (in solution) per kilogram biochar for hours. 445
The property of each biochar type was shown as below (Table 2). In summary, VCZ biochar was of 446
high pH and had comparatively high Ca and Mg content, SWP biochar was slightly alkaline and 447
had higher Fe and Al content but lower P and Ca content, which was similar to the biochar produced 448
by Tammeorg (2014). The Ca, Al and P were very high in SWP+P biochar while the Fe was quite 449
the opposite. 450
Table 2 Chemical properties of different types of biochar (mean ± SD, n = 3)
451 Variables pH NH4 +-N (mg/g) NO3--N (mg/g) Available P (mg/g) Total P (mg/g) VCZ Biochar 10.00 ± 0.01 0 0 22.24 ± 0.84 -
15 SWP Biochar 7.42 ± 0.02 0 0 0.542 ± 0.011 - SWP+P Biochar 3.60 ± 0.02 0.401 ± 0.090 0.080 ± 0.010 258 ± 6 - Variables Fe (mg/kg) Al (mg/kg) Ca (mg/kg) Mg (mg/kg) Cd (mg/kg) VCZ Biochar 22.92 ± 11.25 21.41 ± 10.40 12433 ± 4984 579 ± 231 0.060 ± 0.010 SWP Biochar 66.36 ± 5.66 42.39 ± 3.56 2708 ± 47 231 ± 6 0.050 ± 0.010 SWP+P Biochar 29.11 ± 0.51 148 ± 8 19674 ± 467 836 ± 23 1.49 ± 0.04
Root morphological tracking
452
To determine the new root growth in different treatments and the root preference to different 453
treatments, the root system of each sample was photographed once a week. A darkbox made by PVC 454
sheets covered with thick photographic curtain. A shooting pot was reserved above the box to allow 455
the camera lens in. LED lights were equipped above the inside while a thick black felt was placed 456
bottom the inside. Each rhizobox was placed in a fixed location inside the darkbox. 457
As the growing media was mixed with 50% sand which was at the same grey value as most of roots, 458
a segmentation algorithm based root analysis system including SmartRoot and RootPainter (A. G. 459
Smith et al., 2020) could not be perfectly detected where and how long a single root was in 2D 460
images that were taken weekly. 461
Instead, since the color of new growth roots of Scots pine was slightly lighter than old roots and all 462
the images were taken in the same position, a manually tracking method was applied based on 463
ArcGIS system. 464
The first step was to georeference the TIF images by locating those images into a 30cm x 40cm 465
rectangle, which mirrored the dimensions of the rhizobox. This created a constant geographic 466
coordination of all images to make them detectable by the ArcGIS system. The size of the rectangle 467
was referenced to the internal dimensions of the rhizobox. Hereafter, a feature layer was created to 468
manually tracking the new growth roots following the color differences and former growth routine 469
one week ago (Fig. 6). Only new roots growing on the surface of the media were taken into account 470
as 3D images were lacked and the majority of the new roots had the priority to grow on the surface 471
where the water supply was sufficient. 472
473
Fig. 6 Example of root tracking by ArcGIS system. The left image is the tracking result of new roots in
474
the first week in Blank control No. 4 sample; the right image is the combined tracking result of new roots 475
in the first four weeks in VCZ No.1 sample. 476
The image of each week was recognized as a separate segment. By overlaying the tracking routine 477
of each week, an integrated root growing feature layer could be merged and summarized. Based on 478
which a density analysis, hotspot analysis and other analysis methods could be completed, providing 479
a sequenced series of images showing root growth and directional preferences. 480
Optical microscopy
481
Optical microscopy was used to visually see the strategies for conifer roots to colonize biochar and 482
fertilizer particles. A stereo dissection microscope was used. 483
After the treatment material particles sampled from pots and rhizoboxes, they were stored in samples 484
bags under 4°C until for optical microscopy analysis. Those particles were seen from 1.5x to 4.5x. 485
Internal root and fungi structures were observed by using a scalpel to make a smooth section of the 486
particles. 487
Scanning electron microscopy
488
To determine the pore size of different biochar and the interaction among the root, fungi and biochar 489
particles, a scanning electron microscopy (SEM) was used. 490
Before using SEM, samples were firstly chosen based on the optical microscope results, to select 491
typical particles for further analysis. Chosen samples were fully dried in the oven under 30°C for 3 492
days to completely avoid the interference by free water (Dohnalkova et al., 2011). Thereafter, fitting 493
those particles to the size of the carbon tabs by a scalpel and sticking the particles to the adhesive 494
carbon sheets. Fully fixing them into the sheets and tabs with extra carbon plasticine. Lastly, putting 495
all the prepared samples into the SEM specimen stub storage boxes to transport to SEM lab for 496
further analysis. Samples were outgassed and a sputter coating of gold was applied. 497
The SEM facility (Carl Zeiss SIGMA HD VP Field Emission SEM) was provided by the SEM lab, 498
17
School of Geosciences, University of Edinburgh. All samples were imaged at 15 keV accelerating 499
voltage using the SE2 detector. 500
Fluorescent microscopy
501
To understand the relationship between roots proliferation and fungi population and support the 502
potential strategies for roots to forage P from biochar particles, the abundance of fungi was analyzed 503
by fluorescent microscope. 504
Based on the existed methods (Monheit et al., 1984; Rasconi et al., 2009), all the samples were dried 505
in the oven for 48 hours, grinded and sieved away the sandy particles by 65μm mesh, which may 506
otherwise interfere with the microscope equipment. Prepared samples were accurately weighing 507
0.005 g specimen for each sample and prepared on a clean glass microscope slide. One drop of 508
Calcofluor White Stain (0.025 mL) and one drop of 10% Potassium Hydroxide (0.025 mL) were 509
applied by Stepper Pipette before covering with the cover slide. Samples were let stayed for 5 510
minutes before images were taken microscope in order to provide adequate time for staining of fungi 511
structures to occur. 512
The fluorescent microscope was provided by the Centre Optical Instrumentation Laboratory (COIL), 513
University of Edinburgh. All the samples were observed under 100x lens with immersion oil with 514
GFP transmitted light. For each specimen, four fixed observation spots were chosen. 515
Chemical analysis
516
At the beginning of the rhizobox experiment, the chemical properties of growing media, different 517
biochar and mixed materials in the treatment area were tested. Ten weeks later, except for the mixed 518
materials in the treatment area, the pH of the growing media 3cm beyond the treatment area was 519
additionally tested, along with the same chemical tests completed at the start of the experiment, in 520
order to investigate the differences of P and metals from the treatment materials. Meanwhile, it 521
provided the information on whether the preferences for conifer trees was dominated by pH. 522
The chemical properties tested were pH, active carbon, available P, total P, Fe, Al, Mn, Ca, Mg and 523
other heavy metals. The pH, available P, total P results helped to understand the potential effects 524
provided by biochar as majority of the studies have proved its positive influences. What’s more, 525
since available P is closely related to the pH, Fe, Al, Mn, Ca and Mg, those items were tested as 526
well to understand the P transformation and potential relation among available P, root morphological 527
alteration and fungi colonization. Active carbon was a carbon formation that could be oxidized by 528
potassium permanganate which can be comparatively easier for microorganisms to use (Blair et al., 529
1995). It might affect the colonization rate of fungi and may, to some extent, positively determine 530
the strategies for roots to colonize different treatments. 531
Data analysis
532
The data analysis was mainly focus on qualitative analysis, supplemented by quantitative analysis 533
in this research. 534
For the chemical data processing, the OriginPro 2018 was firstly used to perform simple processing 535
on the original data, mainly including mean and standard deviation. Some linear regression model 536
was applied in OriginPro 2018. ANOVA and T-Test were processed by Python. Based on the changes 537
before and after the experiment, a bar graph could be drawn to compare the differences between 538
treatments. 539
Results
540Root growth responses
541
Generally, there is no significant difference in the total length of new roots (Fig. 7a) and biomass of 542
roots (Fig. 7b) between different treatments after 10 weeks growth. There was one seedling mortality 543
in blank control and the roots deliberately left beside the TSP fertilizer treated area died during the 544
experiment. While in the three-years long-term pot experiment, there is a significant difference 545
between different treatments under the same dosage of P sources applied. The plant biomass 546
accumulation is significantly higher in VCZ biochar treated pots than other treatments (Creber, 547
2021). VCZ biochar applied to Scots pine was observed to have both stronger and healthier roots 548
and aboveground needle biomass. The primary roots had twice the diameter as that in TSP fertilizer 549
treated pots (Creber, 2021). Not only did the TSP fertilizer limit the primary root development, but 550
also caused high root mortality in direct treatment areas. The lagging response for the root biomass 551
to different treatments may owe to the highly resistant for Scots pine to an unhealthy environment. 552
553
Fig. 7 (a) Total length of new roots after 10 weeks application, p = 0.3714 > 0.05; (b) total biomass in
554
dry weight after 10 weeks application, p = 0.1263 > 0.05. Data are mean ± SD, n = 5. 555
According to Kernal Density analysis results (Fig. 8), it’s obvious that VCZ biochar treated areas 556
have significantly higher new roots compared with other treatments while TSP fertilizer treated 557
areas have no new roots growing across. The density difference in treated area plus Part1, 2, 3 areas 558
is significant (Fig. 9a). The mean difference between VCZ biochar and SWP+P biochar treated areas 559
(p = 0.0056), VCZ biochar and TSP fertilizer treated areas (p = 0.0024), SWP biochar and SWP+P 560
biochar treated areas (p = 0.0046), SWP biochar and TSP fertilizer treated areas (p = 0.0020) are 561
significant at the 0.05 level. Combined with the density difference to the opposite side of the treated 562
areas, a quantitively comparison (Fig. 9b) also shows there is a significant difference in the growth 563
of new roots on treated areas and surrounding areas (Blank and TSP fertilizer, p = 0.0474; SWP and 564
SWP+P biochar, p = 0.0345; SWP biochar and TSP fertilizer, p = 0.0016; VCZ and SWP+P biochar, 565
p = 0.0022; VCZ biochar and TSP fertilizer, p = 9.65E-05). The high-density root zones of them 566
locate in the right side and left side of the rhiozobox, respectively. Similar phenomenon is also 567
emerged in SWP and SWP+P biochar treated rhizoboxes, but the degree of difference is slighter. It 568
indicates that there is a strong priority for Scots pine roots to VCZ biochar treated areas and a mild 569
priority to SWP biochar treated areas, while a strong avoidance effect emerges to TSP fertilizer 570
treated areas. The avoidance of roots to SWP+P biochar treated areas was milder than to TSP 571
fertilizer treated areas and the high-density zone of new roots concentrated on the outer edge of 572
treated areas. In untreated rhizoboxes, there is no significant priority or avoidance phenomenon 573
19 being observed.
574
575
Fig. 8 Kernel Density of summarized roots from 5 parallel samples in different treatments, it calculates
576
the density of line features in the neighborhood of each output raster cell. The more reddish, the denser 577
the root is. (a) blank control; (b) TSP fertilizer; (c) SWP+P biochar; (d) SWP biochar; (e) VCZ biochar. 578
579
Fig. 9 (a) The density of new roots in treated area plus Part 1, 2 and 3 areas in different treatment, p =
580
2.51E-04 < 0.05; (b) density difference compared with the same area on the right side of the rhizobox, p 581
= 7.12E-05 < 0.05. Data are mean ± SD, n = 5. 582
Except for one mortality in blank control, all samples were growing well. The TSP fertilizer and 583
SWP+P biochar treated samples have a steady increase throughout all growing weeks (Fig. 10) and 584
there is no significant difference between treatments in each week. But a distinct jump of new 585
growth in blank control and a downward trend in VCZ and SWP biochar treated samples appears 586
starting on the 7th week (Fig. 10) and a significant difference appears between VCZ biochar treated
areas and untreated (Week 7, p = 0.00812; Week 8, p = 0.00231; Week 9, p = 2.13E-04; Week 10, p 588
= 3.31E-04), SWP+P biochar (Week 7, p = 0.00813; Week 8, p = 4.56E-04; Week 9, p = 5.80E-05; 589
Week 10, p = 1.56E-05), TSP fertilizer (Week 7, p = 0.00301; Week 8, p = 9.61E-04; Week 9, p = 590
2.54E-05; Week 10, p = 7.27E-06) treated areas. Meanwhile, the new growth in SWP and VCZ 591
biochar treated areas was seen a weekly increase, especially in VCZ biochar treated areas. It 592
indicates the priority for roots to VCZ biochar treated areas in the late weeks. The same trend can 593
be seen in SWP biochar treated areas, but the degree was weaker. From week 9, the new growth 594
slightly declined a bit in treated areas. It may due to the growth of some roots translocated to the 595
subsurface. The growth trend in TSP and SWP+P treated samples were quite the opposite. As the 596
total length kept increasing, the growth in treated and its surrounding areas were static, which further 597
indicated that roots try to avoid TSP fertilizer treated patches. The growth trend of blank control in 598
the whole rhizobox was consistent with the trend on the left side of the rhizobox which means there 599
was no difference in growth between each side (the same size of areas as SWP+P treated areas). 600
Besides, since all growth of new roots was concentrated in SWP and VCZ biochar treated areas, not 601
the priority effect can be proved for Scots pine roots to those two biochar types, the significantly 602
increase of root length also indicates that SWP and VCZ biochar do improve the root development 603
in the initial establishment state of Scots pine. Creber’s (2021) long-term pot experiment further 604
corroborates this positive influence of biochar on root development. A well-developed root system 605
is more essential for natural trees in a forest system compared with crops in agricultural fields. 606
607
Fig. 10 Weekly total length of new roots from week 3 to week 10 (a) in the whole rhizobox treated area
608
and (b) Part 1, 2 and 3. Data are mean ± SD, n = 5. 609
Chemical property alterations
610
Different P sources led to significantly different chemical properties in the soil. The original soil has 611
seen some alteration in each testing items (Table 2) 612
Table 2 Original property of different treated areas (mean ± SD, n = 3).
613 Variables pH NH4 +-N (mg/g) NO3--N (mg/g) Available P (mg/g) Total P (mg/g) VCZ 8.12 ± 0.01 0.798 ± 0.047 0.654 ± 0.029 1.24 ± 0.86 - SWP 5.58 ± 0.01 0.742 ± 0.031 0.423 ± 0.044 0.314 ± 0.048 - SWP+P 3.65 ± 0.02 0.754 ± 0.087 0.201 ± 0.009 156 ± 24 -
21 TSP 2.90 ± 0.02 1.12 ± 0.11 0.722 ± 0.053 590 ± 37 - Variables Fe (mg/kg) Al (mg/kg) Ca (mg/kg) Mg (mg/kg) Cd (mg/kg) VCZ 20.81 ± 1.24 15.59 ± 0.54 4721 ± 189 569 ± 36 0.060 ± 0.010 SWP 8.91 ± 0.23 11.20 ± 0.37 1206 ± 34 361 ± 15 0.050 ± 0.010 SWP+P 23.45 ± 1.18 107 ± 8 14176 ± 698 782 ± 43 1.11 ± 0.04 TSP 173 ± 4 365 ± 2 50166 ± 1036 1730 ± 24 7.12 ± 0.33
The growing media was alkalized by applying VCZ and SWP biochar and acidified with SWP+P 614
biochar and TSP fertilizer. Though the ending specimen was not enough for the pH analysis, based 615
on the pH results surrounding the treated areas, pH = 5.7 was the critical line (Fig. 11). The areas 616
with a pH lower than 5.7 were around SWP+P biochar and TSP fertilizer treated areas. With a lower 617
pH, the length of new roots was reduced to a comparatively low level. To the opposite, untreated 618
areas and areas treated with VCZ and SWP biochar, with a pH higher than 5.7, saw a quite healthy 619
root growth. On the other hand, with 10 weeks cultivation, the growing media itself tended to alter 620
its pH into a higher available P range (Parladé et al., 2014). The acidic input materials negatively 621
impact this alteration, even beside the treated areas. While the alkaline materials positively enhanced 622
the pH alteration which may, decrease the Fe, Al and Mn phosphate and increase available P content. 623
624
Fig. 11 The relationship between pH and the length of new roots in Part 1,2 and 3. All the samples were
625
presented. 626
In areas farther away from treated areas (Part 5, 6, 7 and 8), to some extent, the growing media was 627
still impacted by the SWP+P biochar and TSP fertilizer. The pH difference of these areas is 628
significant (Table 3). In contrast to the limited new roots grown in those areas accompanied to the 629
treated areas, more roots began to reach these areas. 630
Table 3 pH in different sample parts (Part 1, 2, 3, 5, 6, 7 and 8) after 10 weeks cultivation.
631
Part Treat Mean ± SD (n=5) F ANOVA
1 Blank 5.93 ± 0.11 53.7 p = 2.04E-10 < 0.05 TSP 5.41 ± 0.08 SWP+P 5.56 ± 0.10 SWP 5.98 ± 0.10 VCZ 6.10 ± 0.06 2 Blank 5.96 ± 0.09 39.3 p = 3.28E-09 < 0.05 TSP 5.37 ± 0.08 SWP+P 5.45 ± 0.12 SWP 5.85 ± 0.13 VCZ 5.90 ± 0.03 3 Blank 5.89 ± 0.06 98.3 p = 7.42E-13 < 0.05 TSP 5.35 ± 0.08 SWP+P 5.40 ± 0.06 SWP 5.88 ± 0.07 VCZ 5.95 ± 0.05 5 Blank 5.95 ± 0.09 22.3 p = 8.99E-05 < 0.05 TSP 5.44 ± 0.16 SWP+P 5.63 ± 0.12 6 Blank 5.98 ± 0.05 12.0 p = 1.35E-03 < 0.05 TSP 5.49± 0.15 SWP+P 5.53 ± 0.25 7 Blank 5.92 ± 0.15 24.0 p = 6.37E-05 < 0.05 TSP 5.40 ± 0.13 SWP+P 5.44 ± 0.12 8 Blank 5.92 ± 0.11 22.9 p = 7.97E-05 < 0.05 TSP 5.46 ± 0.14 SWP+P 5.48 ± 0.11
Considering their similar pH range, the pH itself may not the dominant factor that directly influences 632
the growth of new roots. Instead, it may need to count on the available P alteration by pH. The 633
available P in all samples, whether with original low available P or high available P, was declined 634
drastically after 10 weeks cultivation (Fig. 12), which was the same phenomenon as other studies 635
(Steiner et al., 2009). But the degree was various. There’s no significant difference between TSP 636
fertilizer and SWP+P biochar (p = 0.276), SWP biochar (p = 0.102), while the difference between 637
VCZ biochar and other treatments (Blank, p = 3.29E-6; TSP fertilizer, p = 2.94E-04; SWP+P biochar, 638
p = 0.0127; SWP biochar, p = 0.042) are significant. The high difference in variation of blank control 639
is due to one seedling mortality and one extremely slow root development. Both didn’t generate any 640
23
roots to sampled areas and causes any effects on soil alteration. With no extra P source input, the 641
blank control could represent a naturally available P reduction. Therefore, all the extra P sources 642
negatively influenced the maintenance of available P in the early stage. 643
The possible reduction of available P may due to the metals within those input P sources. On the 644
one hand, they provided with extra P to the P pool of the growing media. On the other hand, the 645
metal elements including Al, Fe, Mn, Ca and Mg competitively bonded to the free phosphate 646
compound. Through Ca and Mg may compete with Al, Fe and Mn to form looser Ca-P bound. At 647
the same time, the Al, Fe and Mn phosphate dominated the P pool. The high concentration of Al, Fe 648
and Mn in TSP treated areas may lead rapidly reduction of available P. The same as in SWP+P 649
biochar applied areas. But as the Al, Fe and Mn were comparatively lower, resulting higher ratio of 650
available P retained. In other treatments, the Al, Fe and Mn were low. As VCZ biochar treated areas 651
had a relatively high Ca content and created a slightly acidic environment, a higher ratio of available 652
P was retained. But whether the Fe, Al, Mn or the Ca, Mg were barriers to directly available P uptake 653
for plants cannot be determined based on current study. 654
All extra P sources had a negative influence on the available P alteration. Comparatively, the TSP 655
fertilizer treated areas had the greatest reduction and significantly different from VCZ biochar 656
treated areas which has a slighter reduction (p = 2.94E-04). However, with extremely higher 657
available P within its original state, a 2% left content was still tremendous for the root. 658
659
660
661
Fig. 12 The alteration of available P in treated area. (a) Blank control; (b) TSP treated area; (c) SWP+P
662
biochar treated area; (d) SWP biochar treated area; (e) VCZ biochar treated area; (f) percentage of 663
reduction for each treatment, p = 2.85E-14 < 0.05. Data are mean ± SD, n = 5. 664
The P content in new top needles could reflect the P uptake in the past few months (Fig. 13). At the 665
same starting level, the significant difference emerges a few months later. With the highest available 666
P in TSP fertilizer and SWP+P biochar, the total P content in new top needles was significantly 667
higher than that in blank samples. While in VCZ and SWP biochar with comparatively lower 668
available P, there was no significant difference to the blank samples. Since a P limiting in needle 669
biomass won’t be observed in the second year, long-term experiment (Creber, 2021) shows that 670
enough P can be supplied to Scots pine seedlings by VCZ and SWP biochar. 671
672
Fig. 13 Total phosphorus content in top new needles, p = 0.0039 < 0.05. Data are mean ± SD, n=5.
673
High mount of P uptaken by the root guaranteed the P supply of the plant, but in turn, it inhibited 674
25
the further exploration of roots. In addition, the potential high heavy metal content within TSP 675
fertilizer and SWP+P biochar may restrain the spread of roots beside the treated areas. 676
Kahle (1993) has demonstrated the inhibited effects of Cd on conifer tree root length and total root 677
yield in acidic soil with the increase of concentration. At the same time, the increasing Cd inhibited 678
the ability for roots to uptake micronutrients including K, Mg, Ca. However, as Kahle (1993)’s 679
experiment was based on solution culture, so the exact concentration in unit soil was unknown. Soil 680
experiments (Ismael et al., 2019) have shown that the low Cd concentration generally had no effect 681
on plant growth but had a significantly inhibited effect when the concentration is higher than 5 682
mg/kg. But the direct threshold for the root system was missing. 683
In this case, the concentration of total Cd in TSP was around 7 mg/kg, which was much higher than 684
the standard threshold in most of the countries and higher than the 5 mg/kg threshold (Fig. 14). As 685
the free Cd2+ in the treatment area was not analyzed, the direct toxic effect on root systems as well
686
as the Scots pine aboveground part could not be determined. But as the total Cd in TSP treated areas 687
is almost ten times higher than the value in Dutch standard, there is a potential threat to roots and 688
plants as the proportion of immobilized Cd may transfer into free Cd2+ with the alteration of the soil
689
environment such as water regime, pH, nutrients content and other metals. 690
691
Fig. 14 Cd concentration in treatment areas, Chinese standard (Ministry of Ecology and Environment,
692
2018) value was the target value which indicated desirable maximum levels of total Cd in unpolluted soil 693
while the Dutch value was the maximum permissible value (Crommentuijn et al., 2000). Data are mean 694
± SD, n=3. 695
Fungi occupation
696
Since all the samples were dried and prepared under the same procedure, fungi residues and 697
separated structure could be observed (Fig. 15) and compared (Table 4). 698
699
Fig. 15 Fluorescent microscopy images with distinctive stained fungi on the left and without distinctive
700
stained fungi on the right. 100x lens with immersion oil with GFP transmitted light. 701
Except for TSP fertilizer treated samples which have both few numbers and low occupied areas of 702
distinct fungi structure, there is no significant difference between treatments in part 2 and 3 areas. 703
However, the difference is significant in treated areas (p = 3.96E-05 < 0.05). Namely, different 704
treatments only have a directly impact on its treated area, but little on its surrounding area in the 705
short term. 706
VCZ biochar treated areas have both larger numbers and occupied areas than untreated blank control 707
and those treated with other P sources. When compared TSP fertilizer and SWP+P biochar treated 708
areas, both had an extremely high available P content, larger numbers of fungi structure were found 709
within SWP+P treated areas then in TSP fertilizer treated areas which indicates that biochar porous 710
structure may be positive for fungi proliferation. Other factors including available P content, pH, 711
nutrient retention may also have some influences. 712
Table 4 Distinctive fungi structure numbers and their occupied areas in each treatment (/ = No data; NT
713 = No tested result) 714 Treat Treat NO. Areas/% Numbers (P2+P3+Treated Area) P2 P3 Treated Area Blank 1 1.27 0.944 \ 4 + 4 + \ 2 3.76 0.120 NT 2 +3 + 0 3 2.84 0.776 0.417 4 +2 + 2 4 2.25 2.88 NT 3 +3 + 0 5 0.917 10.97 0.482 4 +3 + 3 TSP 1 NT NT \ 0 +0 + \ 2 0.020 0.641 NT 1 +4 + 0 3 NT 0.315 NT 0 +3 + 0 4 0.175 0.533 \ 1 + 1 + \ 5 NT 0.185 \ 0 + 1 + \ SWP+P 1 1.41 0.551 NT 3 + 3 + 0 2 2.17 2.13 0.549 4 + 4 + 3
27 3 0.920 0.055 0.471 3 + 1 + 2 4 2.04 NT 0.101 2 + 0 + 4 5 4.75 3.02 0.04 4 + 4 + 1 SWP 1 0.830 0.065 0.207 3 + 2 + 4 2 1.29 0.315 0.473 4 + 4 + 4 3 0.031 0.52 3.52 3 + 4 + 3 4 2.67 2.49 0.244 3 + 4 + 4 5 1.11 0.009 0.018 4 + 2 + 4 VCZ 1 2.57 1.48 3.12 3 + 4 + 4 2 NT 7.29 1.28 0 + 3 + 3 3 9.86 0.299 3.52 4 + 4 + 4 4 4.53 2.32 3.95 4 + 4 + 4 5 0.018 1.28 3.59 3 + 4 + 3
Root colonization of biochar particles
715
There’re mainly four strategies observed for ECM root systems to interact with biochar particles, 716
which are: (a) surface interaction between biochar and roots which stick to or penetrate inside 717
following the cracks; (b) roots interacting with biochar through the ECM mantle layer which stick 718
to the surface of biochar and emanate extraradical mycelium strands; (c) hyphae contact and 719
penetrate biochar through cracks and pores; (d) hyphae attract invertebrate grazers to stimulate the 720
fungi-dominated nutrient cycling and obtain additional nutrient by proliferating inside their dead 721
body. 722
Strategy 1 Root contact and penetrate the biochar 723
One of the strategies for root systems to search for P patches is morphological alteration and 724
proliferation (Lynch & Brown, 2008). With a porous structure inside, a coarse and cracked surface, 725
biochar is an ideal media for roots to contact and stick to, let say some types of biochar are infused 726
with sufficient P content. In the early stage, depending on surface structure, roots contact and stick 727
to biochar by generating root hair to penetrate outer layers (Fig.16b) or by penetrating pores 728
following the texture through big cracks (Fig. 16a). However, the successful colonization by roots 729
in the early stage quite depends on the physical characteristic of the surface. In this case, VCZ 730
biochar has a single forward texture and many cracks following the texture. With a larger pore 731
structure within (Rathnayake et al., 2021), it is brittle and fragile which is easier for initial roots to 732
catch and penetrate. Different from VCZ biochar, biochar derived from soft woody materials has a 733
more complex and messy texture and smaller porous structure inside (Fig. 19d). A dense and 734
complex structure may prevent roots from firmly contacting the surface. The same as for the TSP 735
fertilizer infused SWP+P biochar. But the high available P of SWP+P biochar lead to an avoidance 736
effect emerges that may deteriorate roots colonization. 737
In results from samples collected from long-term study by Creber (2021), roots can contact and 738
penetrate all types of biochar, whatever their available P is. The high possibility for the long-term 739
colonization may owe to the deconstruction, mineralization and degradation by the soil and 740
microorganisms, which weaken the structure bond and enlarge the cracks both outside and inside. 741
Hereafter, a breaking point comes into exist for roots to penetrate through. What’s more, as shown 742
in Figure 19f, with extra P nutrients loaded, some pores outside and inside of the biochar are covered 743
and filled by P crystallization. The P crystallization, to some extent, blocks roots from initial 744
catching. Lastly, more available P may transfer into unavailable P in the long term. Then the 745
potential avoidance effect led by availability of P may mitigated. 746
747
748
Fig. 16 Strategy for the root to contact and penetrate different types of woody biochar. (a) root hair stick
749
to the surface of VCZ biochar after three months biochar application; (b) new root penetrates the cracks 750
of VCZ biochar after three months biochar application; (c) root penetrate SWP biochar after three years 751
application; (d) root penetrate SWP+P biochar after three years biochar application. 752
Strategy 2 Root with fungal mantle structure stick to the biochar 753
Not all root tips in contact with biochar have mantle structure, so it is important to separate simple 754
roots from roots with mantle structure. Most studies (Parladé et al., 2014) agree on the fact that the 755
mantle structure is impermeable to common ions (Taylor & Peterson, 2005) and only performs as a 756
nutrient storage and transportation. While a few studies (Vesk et al., 2000) show the permeability of 757
the mantle to some specific ions in soil solution. The common census is that roots with mantle 758
structure seriously blocks their original pathway of nutrient uptake but transfer to highly relying on 759
ECM fungi. When it comes to biochar, there’s no evidence that the mantle layer can uptake nutrients 760
directly from the biochar. However, it is obvious that the mantle structure performs like a drilling 761
platform that thoroughly catches the biochar and gives the extraradical mycelium strand chance to 762
contact and penetrate biochar particles with the least energy consumption (Fig. 17). 763
