Dead wood is an important microsite for seedling regeneration in forest ecosystems. Several abiotic and biotic environmental conditions found in dead wood benefit seedlings of especially small-seeded tree species in boreal, subalpine, temperate, and tropical ecosystems.
For example, in boreal to temperate forests, the seedlings of the genera Picea, Abies, Tsuga, Cryptomeria, and Chamaecyparis among gymnosperms, and Betula, Clethra, and Sorbus among angiosperms, are often reported to regenerate on dead wood.
The growth is stimulated by several factors, including more sunlight, stable moisture levels, fewer pathogens, less litter cover, and less root competition, which allow the seedlings to grow and survive better on dead wood than on the ground soil.
Increasing the understanding of the factors that drive seedling establishment on dead wood and their mechanisms is essential for enabling the improved prediction of forest dynamics.
Recently, Fukasawa reported that properties of dead wood associated with fungal species-specific wood decay activities, (i.e., decay type) were critical to determining seedling density and community composition on dead wood.
Fungi, particularly in the class basidiomycetes, are strong decomposers of wood structural components, and their decay type is traditionally categorised into brown rot and white rot. This reflects different fungal species’ preferences for wood lignin and carbohydrates.
In brown rot, fungi decay only carbohydrates with little modification of lignin. In white rot, fungi decay both lignin and carbohydrates. It is quite typical that brown rot wood becomes blocky in texture, brown in color due to the accumulation of lignin, and acidic because of organic acid production by brown rot fungi.
In contrast, white rot wood becomes fiber-like in texture, and pale in color due to delignification. In addition to texture, pH, and nutrient content, fungal and bacterial communities differ between brown and white rot wood.
However, it is unclear how and which of these properties affect seedling regeneration on dead wood.
A seedling’s mycorrhizal type is a critical factor in determining its preference and performance in regeneration sites, and affects their population dynamics.
In forest trees, the two important mycorrhizal types, arbuscular mycorrhiza and ectomycorrhiza, exhibit regeneration niche preferences among their seedlings in relation to mycorrhizal capacity for providing nutrients and protection from soil-borne diseases.
Fukasawa reviewed the results of previous field studies on seedling regeneration on dead wood and found a pattern that seedling densities of the species associated with arbuscular mycorrhizal fungi were greater on brown rot logs than on white rot logs, whereas the densities of seedlings associated with ectomycorrhizal fungi were lower on brown rot logs than on white rot logs.
Since ectomycorrhizal fungi have the ability to decompose organic matter, it is reasonable to consider that they colonise and utilise the white rot logs which have abundant delignified carbohydrates.
In contrast, arbuscular mycorrhizal fungi are poor decomposers of organic matter and therefore cannot be dominant in environments with abundant available carbohydrates where strong decomposers dominate, i.e., white rot wood in this study.
Thus, they may prefer brown rot wood for colonisation. Since communities of mycorrhizal fungi develop well in logs in late stages of decay, we predict that these mycorrhizal communities affect seedling establishment and are associated with wood decay type. However, this idea has not yet been tested in controlled experiments.
In the present study, pot experiments were conducted for the seedlings of two arbuscular mycorrhizal tree species, Cryptomeria japonica and Chamaecyparis obtusa, and two ectomycorrhizal tree species, Betula ermanii and Abies veitchii, to evaluate their growth on three substrates (brown rot wood and white rot wood of Pinus densiflora, and soil from a pine-dominated forest) in laboratory conditions.
Seedlings of all these four trees are known to regenerate on dead wood. Nutrient ion concentrations and fungal communities were also analyzed in the three substrates and their potential effects on seedling growth were discussed.
We hypothesised that growth of arbuscular mycorrhizal trees (Cr. japonica and Ch. obtusa) would be better on brown rot wood than on white rot wood, whereas the growth of ectomycorrhizal trees (B. ermanii and A. veitchii) would be better on white rot wood than on brown rot wood. Forest soil was prepared as a standard substrate for purpose of comparison.
Tree Species Used in Pot Experiment
Seeds of two ectomycorrhizal tree species (A. veitchii, B. ermanii), and seeds of two arbuscular mycorrhizal tree species (Ch. obtusa, Cr. japonica) were used for pot experiments. All the seeds were obtained from the Forest Tree Breeding Center of FFPRI (Hitachi, Ibaraki, Japan).
The seeds were surface sterilised with hydrogen peroxide: submerged in 70 percent ethanol (v/v) for 1 minutes to wet the surface, sterilised for 50 seconds in a solution of 30 percent hydrogen peroxide (v/v), and then rinsed with sterile deionised water.
The surface sterilised seeds were then kept in darkness at two deg C in moistened autoclaved cotton for 1 month to break dormancy. At the point seeds were surface sterilised again as described above, and placed onto moistened autoclaved cotton to induce germination (20 °C, 13 h light, 20 percent luminescence) in a Biotron system (NK system, Osaka, Japan). Aseptically germinated seedlings were then individually transported to pots prepared as described below.
In July 2019, well-decayed (decay class IV in the five-decay class system), pine (P. densiflora) dead wood was collected from two sites of mixed secondary forest dominated by P. densiflora and Quercus serrata in Miyagi, Japan (38°37.2 N, 140°48.6 E).
The dead wood was then categorised into white rot and brown rot by visual criteria after Araya. These dead wood samples were then sieved through six mm of mesh twice.
Samples from three dead woods were combined and mixed well to make one sample for each of white rot wood and brown rot wood.
Samples were also collected from the top 10 cm of soil at three locations in the same collection sites of as the dead wood. These were sieved through five mm mesh, and then mixed to make one soil sample. Invertebrates in the sample were carefully hand-sorted and removed.
Half volumes of each sample of white rot wood, brown rot wood, and soil were then autoclaved for 20 minutes at 121 deg C. Autoclaving was repeated three times with one day intervals. These six types of substrates (sterilised and non-sterilised soil, white rot wood, and brown rot wood) were then mixed with equal volumes of vermiculite and used as substrates for the pot experiment.
Individual germinants were planted into pots (4 cm diameter, 4.2 cm height; polypot, Tokorozawa Ueki Bachi Center, Saitama, Japan) filled with substrates prepared as described above.
The pots were incubated at 20 deg C, 13 h light, 100 percent luminescence (average photon density = 345 µmol m−2•s−1) in a Biotron for 193 days, except for A. veitchii which was incubated for 188 days.
For B. ermanii, Cr. japonica, and Ch. obtusa, 15 replicates were prepared, although only eight replicates were prepared for A. veitchii due to a low germination rate. Pots were watered every four days during the incubation period.
Seedling shoot length (length from the substrate surface to the top growth point) was measured at three time periods during incubation: day 20, day 141, and day 193 (end of incubation) for Cr. japonica, Ch. obtusa, and B. ermanii, and day 15, day 136, and day 188 for A. veitchii.
Seedling shoot growth was measured as the difference from the first measurement. After the incubation period, seedlings were harvested, weighed in fresh state, and dried at 70 deg C to measure dry weight. Before drying, the mycorrhizal colonisation rate was measured as follows.
For arbuscular mycorrhizal trees (Cr. japonica and Ch. obtusa), root samples were separated into two subsamples: one for measuring mycorrhizal colonisation rate, and the other for measuring water content of the roots to calculate dry weight of the whole root system using their fresh weight.
Mycorrhizal Colonisation Rate of Root Systems
The mycorrhizal fungi colonisation rate (%) of root systems of individual germinants was measured after incubation.
For arbuscular mycorrhizal trees (Cr. japonica and Ch. obtusa), roots (>10 cm length) were washed with 0.005 percent aerosol OT (Wako, Osaka, Japan) in vortex for 1 min, then washed with hypersonic for 10 min, rinsed with deionised water twice, and cleared by heating in 10 percent KOH at 100 deg C for more than one hour.
Cleared roots were rinsed with deionised water, bleached in 0.5 percent H2O2 for 20 minutes, rinsed with deionised water and fixed in two percent HCl for more than 10 minutes. The fixed roots were stained with trypan blue and stored in lactoglycerol (lactic acid 525 mL, glycerin 37.8 mL, deionised water 37.2 mL). Colonisation was assessed under 200× magnification to obtain percentages of root length colonised by arbuscular mycorrhizal fungal structures, specifically arbuscules, coils, and vesicles.
Among the ectomycorrhizal trees (A. veitchii and B. ermanii), the colonisation rate (%) of ectomycorrhizal fungi was measured as the percentage of ectomycorrhizal root tips relative to the total number of root tips (120 on average) by direct observation of root systems under binocular of <45× magnification (SZ2–ILST, Olympus, Tokyo, Japan).
Chemical Properties and Fungal Metabarcoding of Substrates
For each of the six substrates, pH and concentrations of cations (Na+, NH4+, Ca+, K+, Mg+) and anions (NO3−, SO42−, Cl−) were measured. Fresh subsamples (ca. 80 mL, five replicates for each substrate) were extracted with 200 mL deionised water in 250 mL polyethylene bottles and shaken for one hour (100 rpm; Shaker MK201, Yamato Scientific, Tokyo, Japan).
The pH of the extract was measured using a potable pH meter (LAQUAtwin-pH-11B, HORIBA, Kyoto, Japan), and was filtered using filter paper (5C; ADVANTEC, Tokyo, Japan) and a syringe filter (DISMIC25CS; ADVANTEC, Tokyo, Japan). The filtrate was then analyzed using an Ion Chromatography system (Shim-pack IC, Shimadzu, Kyoto, Japan) with 0.6 mM Na2CO3/12 mM NaHCO3 as the anion eluent and 2.5 mM oxalic acid as the cation eluent, 40 deg C at the separation column. Concentrations of ions were calculated as 100 g-substrate bases.
Whole DNA was extracted from 0.2 g white rot wood and brown rot wood subsamples, and 0.3 g soil subsamples (five replicates for each substrate, not including sterilised ones) using ISOIL for Beads Beating (Nippon Gene, Tokyo, Japan) following the manufacturer’s protocol.
Prior to extraction, freeze-dried wood samples were milled using Multi-beads Shocker (Yasui Kikai, Osaka, Japan) for 30 seconds three times each. For sequencing of the fungal internal transcribed spacer 1 (ITS1) region using the MiSeq sequencing platform with 250 × 2 paired-end reads (Illumina, San Diego, CA, USA), a two-step PCR protocol with ITS1F_KYO1/ITS2_KYO2 primers in the primary amplification containing tails for adding indices and Illumina flow cell adapters in the secondary amplification was conducted.
Both positive and negative controls were used for PCR, and positive controls were used for MiSeq sequencing. Note that the ITS region has been proposed as the formal fungal barcode.
A total of 675,854 reads were obtained after MiSeq sequencing and chimera check (deposited in the Sequence Read Archive of the DNA Data Bank of Japan, accession number DRA014309).
The sequence reads were trimmed with a minimum quality value of 30, and the 5'- and 3'-primer sequences were then removed from the trimmed reads.
Next, the trimmed reads were denoised with Claident using Assams-assembler 0.2.2015.05.10. A chimera check was conducted with Claident software using the UNITE database.
Then, the quality-filtered sequences were classified into molecular operational taxonomic units (OTU), and taxonomically identified using the Claident software original fungal ITS database (fungi_its_genus) which is structured after the International Nucleotide Sequence Database, at a threshold similarity of 97 percent, which is widely used for the fungal ITS region.
One soil sample with less than 1000 reads was removed. For each sample, OTUs with less than 0.1 percent of the total number of reads per sample were removed.
After the filtering process, a total of 567,516 reads were obtained. The OTU numbers were saturated in all samples after the filtering process.
Singleton OTUs were removed from the downstream analyses. Consequently, each of the 133 filtered OTUs was searched using the FUNGuild database and assigned to one of eight functional groups: arbuscular mycorrhizal, brown-rot, ectomycorrhizal, plant pathogen, soft-rot, undefined saprotroph, white-rot, and wood decay with unknown decay type.
After the analysis of statistics, nutrient ion concentrations were found to be significantly different among the six substrates, and sterilisation was seen to increase the nutrient ion concentrations in all except for NO3− in soil. Cl− concentration was significantly higher in white rot compared to soil, regardless of sterilisation.
The NO3− concentration was significantly higher in soil than in brown rot in unsterilised substrates, but this difference disappeared after sterilisation.
In contrast, the SO42− concentration was significantly increased by sterilisation in brown rot, but was not in white rot or soil. Na+ concentrations in wood samples (brown rot and white rot) were higher than seen in soil, and the difference was significant after the samples were sterilised.
NH4+ concentration was higher in white rot compared to soil, but this difference disappeared after sterilisation. Similar to Na+, concentrations of K+, Mg2+, and Ca2+ in wood samples were not significantly different from those in soil, but were significantly higher than in soil after sterilisation.
Concentrations of NO2−, Br−, and PO43− were too low to be detected as meaningful data. The five replicates showed almost equal pH values within substrates: 3.6–3.7, 4.2, and 5.0 for unsterilised brown rot, white rot, and soil, respectively, and 3.4–3.5, 4.0–4.1, and 5.5–5.6 for sterilised brown rot, white rot, and soil, respectively.
While we did not perform statistical analysis, the pH increased in this order, and was altered by sterilisation (slightly decreased in wood samples and increased in soil).
Fungal Community in Substrates
In total, 132 fungal OTUs were detected. The mean OTU richness per log ranged from 29.2 in brown rot logs to 54.3 in soil, and was larger in the soil than in the logs (BM test, brown rot vs. soil p = 0.048; white rot vs. soil p = 0.048).
No significant difference was detected in total OTU richness between brown rot and white rot logs. Among the 132 OTUs, 52 OTUs were assigned to one of the eight functional groups.
Ectomycorrhizal fungi included the largest 19 OTUs, mainly detected among soil samples, followed by undefined saprotrophs (15 OTUs), and soft rot fungi (seven OTUs).
White rot fungi and wood decay fungi with an unknown decay type include four OTUs for each, and arbuscular mycorrhizal, brown rot fungi, and plant pathogens include one OTU for each. Eighty OTUs were functionally unknown. Results of PERMANOVA analysis suggested that fungal communities were significantly different among the three substrates, while dispersions were not significantly different among the substrates.
Indicator species analyses suggested that seven OTUs were indicative of brown rot wood, 11 OTUs were indicative of white rot wood, and 39 OTUs were indicative of soil. Seven OTUs indicative of brown rot wood consisted of one brown rot fungi (Pseudomerulius curtisii), two soft rot fungi in the genus Scytalidium, and four OTUs functionally unidentified.
Eleven OTUs indicative of white rot wood consisted of three white rot fungi (Gymnopilus picreus, Xeromphalina sp., and Sistotremastrum sp.), a wood decay fungi with an unknown decay type (Botryobasidium sp.), a soft rot fungi (Scytalidium album), and six OTUs which were functionally unidentified.
Among the 39 OTUs associated with soil, 16 OTUs were ectomycorrhizal fungi, six OTUs were undefined saprotrophs, one of each were arbuscular mycorrhizal (Archaeospora sp.) and soft rot fungi (Trichoderma sp.), and 15 OTUs were functionally unidentified. Ectomycorrhizal OTUs associated with soil were dominated by the genera Russula (three OTUs), Inocybe (two OTUs), Sebacina (two OTUs), and Tomentella (two OTUs). Cenococcum geophilum, Clavulina sp., Cortinarius sp., Delastria sp., Elaphomyces sp., Laccaria bicolor, and Thelephora sp. were also detected.
Seedling Growth from the Study Conclusion
The growth of the seedling shoots during 121 days from day 20 to day 141 (day 15 to day 136 for A. veitchii) is shown from the survey.
In Cr. japonica, the greatest growth rate was observed in unsterilised brown rot and in sterilised soil (1.4 cm). However, the shoot growth of Cr. japonica was not significantly different among unsterilised substrates, and growth in soil was significantly greater than that in white rot only when the substrates were sterilised.
Ch. obtusa growth was not significantly different among the substrates regardless of sterilisation, although the growth was stimulated by sterilisation in soil, and the greatest growth (2.1 cm) was recorded in sterilised soil.
B. ermanii growth was significantly greater in white rot (max. 0.8 cm) than in brown rot and soil in unsterilised substrates. However, growth in white rot was significantly reduced when the substrates were sterilised, and growth in soil became significantly greater than that in wood (brown rot and white rot).
A. veitchii growth (max. 1.1 cm) was not different among the substrates. Seedling shoot growth during 173 days from day 20 to day 193 (day 15 to day 188 for A. veitchii), covering the entire period.
The significant differences in shoot growth described above were disappeared except for the negative effect of sterilisation on B. ermanii seedlings in white rot wood.
Dried masses of the seedlings are also shown. In Cr. japonica, seedling mass was significantly greater in soil than in wood regardless of the sterilisation.
Similarly, seedling mass of Ch. obtusa was greater in soil than in wood, particularly in sterilised substrates. Seedling mass of A. veitchii did not differ among substrates. No effect of sterilisation was observed in all substrates for all species. The seedling mass of B. ermanii was too small to measure.
Colonisation of arbuscular mycorrhizal fungi was almost entirely restricted to the seedlings in unsterilised soil, although there was some colonisation of arbuscule and coil in Cr. japonica roots in brown rot, and of coil in Ch. obtusa roots in sterilised soil.
Colonisation of ectomycorrhizal fungi was also entirely restricted to the seedlings in unsterilised soil. All observed ectomycorrhizal root tips were Cenococcum-type.
There was no significant relationship observed between seedling growth (shoot growth and dry mass) and mycorrhizal colonisation rate (data not shown).
The present study showed significant effects of substrate and sterilisation on seedling growth. Because the nutrient concentrations differ greatly between sterilised and unsterilised substrates, it would be difficult to discuss the biotic factors without these large differences. Rather, it would be better to sort out the differences in growth responses within the sterilised and unsterilised conditions, respectively. The greater shoot growth seen in B. ermanii in unsterilised white rot wood, compared with brown rot wood and soil, might be attributable to some positive effects of microorganisms in white rot wood.
The effects of substrates on the growth of arbuscular mycorrhizal trees Ch. obtusa and Cr. japonica was clear in seedling weight, with both being greater in soil than in wood. The higher colonisation rate of arbuscular mycorrhizal fungi in soil might be a cause of this effect.
Brown rot wood substrates used in the present study appear to be decayed primarily by a single brown rot fungus, Pseudomerulius curtisii, because this fungus was detected in 100 percent of the brown rot samples and there were no other brown rot fungi detected. Whereas in white rot wood, decay appears to be the result of a combination of several white rot fungi.
Additionally, the physicochemical properties and microbial communities might differ among the wood decayed by the different fungi, even within the same decay type.
Our previous study using substrates from another study site, where different species of brown rot fungi Neolentinus lepideus dominated on P. densiflora logs, showed that white rot wood contains higher concentrations of potassium, magnesium, and calcium than brown rot wood, and soil contained higher concentrations of SO42− than wood. These findings are inconsistent with the findings of the present study. These results suggest that studies using substrates decayed by different fungal species may have different effects on the seedlings, and thus wood decayed by known fungal species should be used as the substrates in future research.
To summarise, seedling growth of the two arbuscular mycorrhizal trees Cr. japonica and Ch. obtusa, and an ectomycorrhizal tree B. ermanii, were significantly different among the three substrates: white rot wood, brown rot wood, and soil. Particularly, the shoot growth of B. ermanii was significantly greater in white rot wood than in the other two substrates, supporting our hypothesis.
However, the growth of Cr. japonica and Ch. obtusa was not significantly different between brown rot and white rot wood. These results suggest that good growth of B. ermanii on white rot wood could be a reason for their regeneration success on white rot logs observed in the field, although its mechanism is still unknown.
In contrast, the dominance of Cr. japonica and Ch. obtusa seedlings on brown rot logs might be attributable to mechanisms in other stages of seedling development, such as germination, or survival, and should be tested in future field experiments.
Furthermore, we did not detect any effects associated with the substrates and sterilisation in A. veitchii seedlings, suggesting that responses to the substrates and sterilisation differ among seedling species within the same mycorrhizal type. Therefore, further research is needed involving many tree species, including ectomycorrhizal and arbuscular mycorrhizal trees, to test their performance on different substrates.