Distribution of charred plant fragments in particle size fractions of Japanese volcanic ash soils

Abstract

To gain a better understanding of the distribution of charred plant fragment C (CPFC) and its contribution to organic C (OC) in the particle size fractions of Japanese volcanic ash soils, each of four soil samples was divided into six particle size fractions, namely three sand-sized aggregate (20–53, 53–212 and 212–2,000 µm) fractions, one silt-sized aggregate (2–20 µm) fraction, and two clay-sized aggregate (< 0.2 and 0.2–2 µm) fractions. Furthermore, after HCl–HF treatment of these aggregate fractions, sub-fractions of less than specific gravity (s.g.) 1.6 g cm⁻³ (< 1.6 fraction) were isolated using s.g. 1.6 g cm⁻³ sodium polytungstate solution. Microscopic observation indicated that the charred plant fragments, which are black or blackish brown, were the main components in the < 1.6 fractions. Therefore, the OC in this fraction was designated as CPFC. In all the soils studied, the quantitative distribution of the CPFC of the silt-sized aggregate fractions to total CPFC of whole soils, ranging from 59 to 84%, was greatest among the aggregate fractions. The sum of the distribution (%) values of the CPFC in the three sand-sized aggregate fractions varied from 6.9 to 33%, while that in the two clay-sized aggregate fractions ranged from 1.1 to 9.4%. Similar to the CPFC, in all soils, the quantitative distribution of the OC in the aggregate fractions was greater in the silt-sized aggregate fractions (52–76%) than in the other aggregate fractions (0.1–20%). In all soils, the quantitative contribution of total CPFC to total OC of whole soils ranged from 10 to 28%. The CPFC/OC values in the aggregate fractions were 21% or more in 10 samples from a total of 24 fractions, with a maximum value of 34%. On the basis of the findings obtained in the present study, it is assumed that in Japanese volcanic ash soils the silt-sized fraction is an important reservoir of CPFC and OC, and CPFC merits attention as one of the constituents of OC in particle size fractions.

Key words:

• charred plant fragment
• Japanese volcanic ash soil
• particle size fraction
• silt
• specific gravity method

INTRODUCTION

In terrestrial ecosystems, charred plant materials are produced during the burning of vegetation by human activity and wildfires. Charred and buried plant fragments have been widely detected in various soils. In Japanese volcanic ash soils (Shindo et al. 2004a), European Chernozemic soils (Schmidt et al. 1999) and US agriculture soils (Skjemstad et al. 2002), it has been suggested that charred plant fragments should not be overlooked as one of the constituents of soil organic matter, which affects the physical, chemical and biological properties and fertility of soils as well as the global carbon cycle. Furthermore, in Japanese volcanic ash soils, which contain predominately black (Type A) humic acids with a high degree of darkening and a graphite-like structure (Kumada 1987), it has been assumed that charred plants could be an important source of Type A humic acids (e.g. Nishimura et al. 2006; Shindo et al. 2004a,b). After burning, charred plant materials that remain on the soil surface may be distributed into soil fractions with different particle sizes under the soil conditions, as is the case with other organic constituents. However, this behavior and the contribution of charred plants to organic matter in soil fractions are poorly documented. The objective of this study was, therefore, to gain a better understanding of the distribution of charred plant fragment C (CPFC) and its contribution to organic C (OC) in the particle size fractions of Japanese volcanic ash soils using four soil samples with different OC contents.

Table 1 Brief description of the soils used

Soil sample^†	Sampling site^†	Classification (US Soil Taxonomy)	Depth (cm)^†	Horizon^†	Particle size fraction^†			Total organic C content (g kg^−1 whole soil)^‡	Note^†
					Sand	Silt	Clay
					(g kg^−1 whole soil)
S-3-2	Fujinomiya, Shizuoka	Typic Melanudand	15–27	A12	N.D.	N.D.	N.D.	181	Forest
M-1-2	Ohta, Shimane	Pachic Melanudand	22–47	A12	N.D.	N.D.	N.D.	146	Grassland
I-3-4	Kanagasaki, Iwate	Pachic Melanudand	35–63	3A11	133	300	567	115	Forest
H-1-3	Hiroshima, Hokkaido	Typic Melanudand	24–35	3A3	340	310	350	70.0	Forest
^†From Wada (1986) except for M-1-2.
^‡From Shindo et al. (2004a). N.D., not determined.

MATERIALS AND METHODS

Soil samples

In this study, to reduce the influence of fresh plant remains or living plants on the particle size fractionation and subsequent determination of charred plant fragments, four volcanic ash soil samples that were collected at depths of 15–63 cm were selected (Table 1). These soil samples, which differed in OC content, horizon and location, were used previously by Shindo et al. (2004a). Each sample was air-dried and passed through a 2-mm mesh sieve, and then the plants were visually and carefully removed.

Preparation of the particle size fractions

Physical fractionation of the soils was carried out using a modified procedure of the fractionation method described by Skjemstad et al. (1996), who investigated the fractional distribution of soil organic matter. Soil samples (10.00 g) were suspended in 40 mL of 2 mol L⁻¹ Na₂SO₄ in centrifuge tubes, shaken for 1 h and then centrifuged at 1,700 g for 20 min. The supernatant obtained was filtered through a glass microfiber filter (GF/A; Whatman, Kent, UK). The filtration removed plant remains that might otherwise have appeared in the subsequent sediment fractions, particularly in the silt-sized and clay-sized fractions. In contrast, soil residues in the centrifuge tubes were suspended in 40 mL of 2 mol L⁻¹ Na₂SO₄ solution, shaken, centrifuged and then filtered as described above. A small amount of plant remains was detected on the glass filters, but no charred plant fragments were detected. The soil residues in the centrifuge tubes were further mixed with 40 mL of 2 mol L⁻¹ Na₂SO₄ solution before being transferred to cellulose dialysis tubing (Sanko Pure Chemical, Tokyo, Japan) and dialyzed against deionized water (water) until no SO₄ ^2– was detected by the precipitation reaction with BaCl₂ solution.

After dialysis, the soil samples were transferred to 300 mL beakers with 100 mL of water. The suspensions were subjected to ultrasonic treatment for 1 min at 20 kHz and 200 W using an ultrasonic disrupter (UD-201; Tomy, Tokyo, Japan). After allowing the samples to stand for 5 min, the supernatant was poured into 1-L measuring cylinders through a 212-µm and a 53-µm mesh sieve in series. The ultrasonic treatment was repeated five times until clay was virtually absent from the supernatant. The remaining materials in the beakers were then transferred to the 212-µm and 53-µm mesh sieves in this order and washed with water. Washings were combined in the 1-L measuring cylinder. The retained materials (212–2,000 µm and 53–212 µm) on the sieves were freeze-dried and weighed.

The fraction of < 2 µm (specific gravity [s.g.] 2.65) was collected by sedimentation, and then the 2–20 µm (s.g. 2.65) fraction was collected. For the collection of the former fraction, the dispersion and sedimentation were repeated 15 times, and for the latter fraction, the procedure was repeated 10 times. Subsequently, the < 2 µm fraction collected was subdivided into < 0.2 and 0.2–2 µm fractions by centrifugation at 1,700 g for 22 min (calculated value) using a swing rotor. The < 0.2 and 2–20 µm fractions obtained were concentrated below 40°C using a rotary evaporator, freeze-dried and then weighed. In contrast, the 0.2–2 and 20–53 µm fractions were freeze-dried and weighed.

In this study, each whole soil was divided into six particle size fractions. As described later, because highly recalcitrant complexes, which are resistant to an ultrasonic disruption, exist in the soils studied, the fractions obtained were designated as sand-sized aggregate I (212–2,000 µm), sand-sized aggregate II (53–212 µm), sand-sized aggregate III (20–53 µm), silt-sized aggregate (2–20 µm), clay-sized aggregate I (0.2–2 µm), and clay-sized aggregate II (< 0.2 µm) fractions.

Isolation of charred plant fragments from the particle size fractions

Charred plant fragments in the particle size fractions were isolated using the method described in Nishimura et al. (2006). Particle size fractions in the Teflon beakers (100 mL) were treated twice with a mixed solution of 12 mol L⁻¹ HCl and 46–48% (approximately 16.75 mol L⁻¹) HF on a hotplate at 200°C, and then the residues were washed with water and were subjected to the specific gravity method using s.g. 1.6 g cm⁻³ sodium polytungstate solution (pH 4.7, prepared at a ratio of sodium polytungstate : water of 1 : 1.25 [w/v]). The isolated sub-fraction of less than s.g. 1.6 g cm⁻³ (< 1.6 fraction) was used for microscopic observation and for the determination of the OC content.

Microscopic observation of the < 1.6 fraction

The < 1.6 fractions were observed under a stereomicroscope and a scanning electron microscope as reported previously (Nishimura et al. 2006; Shindo et al. 2004a).

Determination of organic C

The OC contents of the particle size fractions and < 1.6 fractions were determined using the dichromate–sulfuric acid oxidation method described in Nishimura et al. (2006). The digestion method used was valid for OC determination in this study because the digestion caused the black color of OC in the charred plant fragments and particle size fractions to disappear, and this method has often been used to determine OC in Japanese volcanic ash soils (e.g. Wada and Higashi 1976).

RESULTS AND DISCUSSION

In a previous study, Nishimura et al. (2006) found that the < 1.6 fractions, which were obtained after HCl–HF treatment of soils S-3-2, M-1-2 and I-3-4 (Table 1), contained humic acids belonging to Type A and fulvic acids, although the optical properties of the humic acids extracted from the < 1.6 fractions were different from those extracted from the whole soils. Furthermore, according to Nishimura et al. (2006), the degree of the quantitative contribution of NaOH-extractable humic acids and fulvic acids in the < 1.6 fractions, which were isolated after HCl–HF treatment, to those in the whole soils ranged from 12 to 44% and from 3.8 to 9.6%, respectively. In this study, using soils S-3-2, M-1-2, I-3-4 and H-1-3 (Table 1), the distribution of CPFC and its contribution to OC in the soil fractions were investigated.

The average recoveries of mass weight of six particle size fractions (sand-sized aggregate I, sand-sized aggregate II, sand-sized aggregate III, silt-sized aggregate, clay-sized aggregate I, and clay-sized aggregate II fractions) by duplicate physical fractionation were 98.9% for soil H-1-3, 102% for soil M-1-2, 106% for soil I-3-4 and 107% for soil S-3-2. Thus, the percentage distribution of the mass weight in the particle size fractions (W) was corrected to a total of 100%, as shown in Table 2. In all soils used, the percentage distribution of the W was much greater in the silt-sized aggregate fraction, ranging from 45 to 58%, than in the other aggregate fractions, which ranged from 0.8 to 23%. The sum of the distribution (%) values of the W in the three sand-sized aggregate fractions ranged from 26 to 46%, while that in the two clay-sized aggregate fractions ranged from 8.7 to 16%. These results suggest that highly recalcitrant organic–inorganic complexes, which are resistant to ultrasonic disruption, exist in the soils studied because larger amounts of clay-sized fractions (< 2 µm) than silt-sized fractions (2–20 µm) were recovered (Table 1) when the organic matter in soils I-3-4 and H-1-3 was digested with hot H₂O₂ (Wada 1986).

Table 2 Percentage distribution of mass weight (W), organic C (OC) and charred plant fragment C (CPFC) in six particle size fractions of four soils

Particle size fraction (µm)	S-3-2			M-1-2			I-3-4			H-1-3
	W (%)	OC (%)	CPFC (%)	W (%)	OC (%)	CPFC (%)	W (%)	OC (%)	CPFC (%)	W (%)	OC (%)	CPFC (%)
Sand-sized aggregate I (212–2,000)	7.79	6.63	3.48	22.8	0.29	–	0.77	0.14	–	8.32	0.86	–
Sand-sized aggregate II (53–212)	23.1	20.3	20.4	13.0	1.83	1.09	5.26	0.90	1.81	12.1	5.85	19.4
Sand-sized aggregate III (20–53)	12.0	6.94	7.16	9.89	6.18	5.79	19.9	17.2	31.1	22.6	20.1	13.4
Silt-sized aggregate (2–20)	44.7	51.7	59.8	45.7	75.6	83.7	58.0	64.9	66.1	44.8	55.5	58.7
Clay-sized aggregate I (0.2–2)	9.97	11.7	6.52	7.69	14.4	7.82	13.9	14.8	1.05	10.3	12.9	8.54
Clay-sized aggregate II (< 0.2)	2.51	2.76	2.67	0.98	1.62	1.56	2.18	2.11	–	1.99	4.71	–
Total	100	100	100	100	100	100	100	100	100	100	100	100
–, not determined because the amount of CPFC was very small.

The s.g. values of charcoals (charred woods) and soil mineral particles were approximately 1.7 g cm⁻³ and 2.6 g cm⁻³, respectively (Nishida et al. 1929). Shindo et al. (2004a) isolated the < 1.6 fractions before and after HCl–HF treatment in 24 volcanic ash soil samples involving the four soil samples used in this study and observed the fractions under a stereomicroscope and a scanning electron microscope. The observation results indicated that the charred plant fragments, which are black or blackish brown, are the main components in the < 1.6 fractions before and after treatment. Furthermore, it was suggested that most of the fragments in the < 1.6 fractions isolated before treatment originated from the xylem of woody plants, while the morphology of the fragments after the treatment was mainly amorphous. As expected, the results of the microscopic observation (not presented as photographs) of the < 1.6 fractions isolated from the particle size fractions in the present study were very similar to those isolated from the whole soils described above and reported previously by Nishimura et al. (2006). Therefore, in this study, the OC in the < 1.6 fractions was designated as CPFC.

Similar to the W, in all soils, the quantitative distribution of the CPFC of particle size fractions to total CPFC of whole soils was greatest in the silt-sized aggregate fractions, ranging from 59 to 84% (Table 2). In the present study, after the clay-sized aggregate fraction was collected by sedimentation, the silt-sized aggregate fraction was obtained. This means that free (not connected with minerals) and fine charred plant fragments may be recovered in the clay-sized aggregate fraction. Accordingly, higher distribution values of the CPFC in the silt-sized aggregate fraction compared to the other aggregate fractions may be because many charred fragments were connected to the minerals (including aggregated clay-sized minerals) in the silt-sized aggregate fraction and preserved there. The connected fragments can be recovered only after HCl–HF treatment of the fraction. The sum of the distribution (%) values of the CPFC in the three sand-sized aggregate fractions varied from 6.9 to 33%, while that in the two clay-sized aggregate fractions ranged from 1.1 to 9.4%. The sum of the distribution values of the CPFC in the silt-sized aggregate fraction and the two clay-sized aggregate fractions was greater in soil M-1-2 than in soils S-3-2, I-3-4 or H-1-3. This finding suggests that the charred plant fragments could be more weathered and degraded in soil M-1-2 than in soils S-3-2, I-3-4 and H-1-3, because it is considered that the size of the charred plants became smaller and that the amount of mineral particles adhering and/or being associated with them increased with the progression of weathering and degradation after burning. According to Skjemstad et al. (1999, 2002), most charcoal (char) from Australian and US soils is detected in particle size fractions of less than 53 µm. Ponomarenko and Anderson (2001) reported that significant amounts of char particles were present in silt-sized fractions of Black Chernozem soils in Western Canada.

As in the cases of the W and CPFC, in all soils, the quantitative distribution of OC in the particle size fractions to total OC of whole soils was greater in the silt-sized aggregate fractions (52–76%) than in the other aggregate fractions (0.1–20%) (Table 2). The sum of the distribution (%) values of the OC in the three sand-sized aggregate fractions varied from 8.3 to 34%, while that in the two clay-sized aggregate fractions ranged from 15 to 18%.

The total CPFC contents of whole soils (g kg⁻¹) were 42.8 for soil S-3-2, 34.0 for soil M-1-2, 17.1 for soil I-3-4 and 7.49 for soil H-1-3. In the particle size fractions of soils S-3-2 and M-1-2, the CPFC contents were highest in the silt-sized aggregate fraction, while in soils I-3-4 and H-1-3 the contents were highest in the sand-sized aggregate III and sand-sized aggregate II fractions, respectively (Table 3). In contrast, the total contents of OC recovered through the fractionation (g kg⁻¹ whole soil), which were calculated fromTables 2 and 3, were 190 for soil S-3-2, 123 for soil M-1-2, 118 for soil I-3-4 and 73.5 for soil H-1-3. In the particle size fractions of all soils, the OC contents of the silt-sized aggregate and clay-sized aggregate fractions were greater than those of the sand-sized aggregate fractions (Table 3).

The proportion (%) of the quantitative contribution of total CPFC to total OC of whole soils was 22.5 for soil S-3-2, 27.6 for soil M-1-2, 14.5 for soil I-3-4 and 10.2 for soil H-1-3. The proportion in soils S-3-2 and M-1-2 was higher than that in soils I-3-4 and H-1-3, as in the case of total CPFC and total OC contents. The higher proportion in soils S-3-2 and M-1-2 compared to soils I-3-4 and H-1-3 is ascribed to the fact that the total CPFC contents were much greater in soils S-3-2 and M-1-2 than in soils I-3-4 and H-1-3. In the particle size fractions of soils S-3-2 and M-1-2, the quantitative contribution of the CPFC to the OC were highest in the silt-sized aggregate fraction, while in soils I-3-4 and H-1-3, the contribution was highest in the sand-sized aggregate II fraction (Table 3). This result suggests that behavior such as weathering and degradation of the CPFC and the OC in soils S-3-2 and M-1-2 was different from that in soils I-3-4 and H-1-3. The CPFC/OC values of the particle size fractions were 21% or more in 10 samples from a total of 24 fractions, with a maximum value of 34%.

Table 3 Charred plant fragment C (CPFC) contents, organic C (OC) contents and CPFC/OC values of six particle size fractions in four soils

Particle size fraction (µm)	S-3-2			M-1-2			I-3-4			H-1-3
	OC^†	CPFC^†	CPFC/OC (%)^‡	OC	CPFC	CPFC/OC (%)	OC	CPFC	CPFC/OC (%)	OC	CPFC	CPFC/OC (%)
Sand-sized aggregate I (212–2,000)	162	19.1	11.8	1.58	–	–	20.7	–	–	7.63	–	–
Sand-sized aggregate II (53–212)	167	37.7	22.6	17.3	2.84	16.4	20.2	5.86	29.0	35.5	12.0	33.8
Sand-sized aggregate III (20–53)	110	25.5	23.2	77.0	19.9	25.8	102	26.7	26.2	65.4	4.41	6.74
Silt-sized aggregate (2–20)	220	57.2	26.0	204	62.4	30.6	132	19.5	14.8	91.0	9.83	10.8
Clay-sized aggregate I (0.2–2)	224	28.0	12.5	232	34.6	14.9	126	1.30	1.03	91.8	6.17	6.72
Clay-sized aggregate II (< 0.2)	209	45.4	21.7	204	54.1	26.5	114	–	–	174	–	–
^†Carbon content is expressed as g kg^−1 particle size fraction.
^‡CPFC/OC value is for the weight percentage of CPFC to the weight of OC in each particle size fraction. –, not determined because the amount of CPFC was very small.

On the basis of the findings obtained in the present study it is assumed that in Japanese volcanic ash soils the silt-sized fraction is an important reservoir of CPFC and OC, and CPFC merits attention as one of the constituents of OC in particle size fractions.

ACKNOWLEDGMENTS

We thank the members of the Cooperative Research Projects on Ando soils (volcanic ash soils), especially Dr T. Honna and Dr S. Yamamoto, Faculty of Agriculture, Tottori University, for supplying the soil samples, and we also thank Dr H. Honma, Forensic Science Laboratory, Yamaguchi Prefectural Police Headquarters, for his assistance.