Microbial biomass and nitrogen transformations in surface soils strongly acidified by volcanic hydrogen sulfide deposition in Osorezan, Japan

Authors


M. WATANABE, Water and Soil Environment Division, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan. Email: watanabe.mirai@nies.go.jp

Abstract

Volcanic acidification has created unique ecosystems that have had to adapt to the acidic environments in volcanic regions. To characterize the primary microbial properties of strongly acidified soils in such environments, we investigated microbial biomass, nitrogen transformations and other relevant chemical properties in the surface soils of solfatara and forests from Osorezan, a typical volcanic region in Japan, and compared the results to common Japanese forest soils. Soil microbial biomass C (MBC) and N (MBN) were determined using the chloroform fumigation–extraction method. Potential net N mineralization and net nitrification were measured in aerobic laboratory incubations. Long-term acidification in the Osorezan soils by volcanic hydrogen sulfide deposition caused low soil pH (3.0–3.8), base cation deficiency and increased concentrations of toxic ions such as Al3+. The proportions of MBC to total carbon (MBC/TC ratio) and MBN to total nitrogen (MBN/TN ratio) were lower than those in common Japanese forest soils. The extreme acidic conditions may have inhibited microbial survival in the Osorezan acid soils. Net N mineralization occurred at rates comparable to those in common Cryptomeria japonica forest soils, probably because of the presence of acid-tolerant soil microorganisms. Net nitrification was completely inhibited and autotrophic ammonia oxidizers were not detected by the MPN method. The inhibition of nitrification prevents nitrogen leaching from the soils, thus maintaining a nitrogen cycle in the volcanic acid region in which inline image (and NH3) is recycled among microorganisms and plants.

Introduction

Japan, located in the Pacific Ring of Fire, has over 100 active volcanoes (Japan Meteorological Agency 2005). Some volcanoes, such as Sakurajima and Miyakejima, erupt frequently and there are many regions where acidic gases, including hydrogen sulfide (H2S), sulfur dioxide and hydrogen chloride, are emitted from fumaroles. Long-term exposure to these gases has acidified the soils and freshwaters in volcanic regions. This acidification increases the concentrations of Al3+ and heavy metals in the soil solution. These increased concentrations may be toxic to plants and soil microorganisms (Takamatsu et al. 1992; Yoshitake et al. 2007). Because only a few organisms can survive in these extreme volcanic environments, special ecosystems exist that have adapted to acidic conditions (Satake et al. 1995; Yoshitake et al. 2007). For example, solfatara plants, such as Carex angustisquama and Deschampsia flexuosa, are distributed around fumaroles (Tsujimura 1977; Yoshida and Wakamatsu 1991). Yoshitake et al. (2007) recorded reduced soil microbial biomass and different microbial communities in a solfatara soil compared with an adjacent forest soil.

Chemical and microbiological investigations of strongly acidified soils in volcanic regions are important for understanding soil elemental cycles in soil ecosystems adapted to acidic environments. Osorezan (in Aomori, Japan) is a typical volcanic region acidified by H2S gas emitted from fumaroles (Satake et al. 1995; Takamatsu et al. 1992). Many researchers have studied the plant and fish ecology of acidic ecosystems in Osorezan (Hirata et al. 2003; Satake et al. 1995; Tsujimura 1977). Takamatsu et al. (1992) studied exchangeable cations and sulfur in Osorezan acid soils with and without vegetation. However, soil microbiology in this area remains unstudied.

The aim of the present study was to characterize the primary microbial properties of acid soils from Osorezan. For this purpose, we analyzed microbial biomass and the rate of nitrogen transformations (mineralization and nitrification), together with relevant chemical properties. We compared the microbial biomass and N mineralization rate of Osorezan acid soils to those in common Japanese forest soils to characterize the fertility status and nitrogen cycle in these acid soils.

Materials and methods

Study area and soil sampling

The study area is the Osorezan volcanic region (41°19′N, 141°6′E) located in the city of Mutsu, Aomori, Japan, 200–300 m a.s.l. (Fig. 1). The mean annual temperature and precipitation in this area are approximately 10°C and 1,400 mm, respectively (Japan Meteorological Agency 2009). The area has a central crater lake (Lake Usorisan) and is dotted with many fumaroles that emit high H2S concentrations and several hot springs (Takamatsu et al. 1992). The Osorezan volcano erupted intermittently between 240,000 and 300,000 years ago, with a few explosive eruptions after that, but there have been no eruptions in the last 10,000 years (Kuwabara and Yamazaki 2001).

Figure 1.

 Map showing the sampling sites. Sites 1–5 are acidified sites located within the crater of Osorezan and Site 6 is a reference site. Asterisks (*) indicate the fumaroles.

Soil samples were collected from six sites at different distances from the fumaroles. The soil type and vegetation at each site are summarized in Table 1. Three sites in Oniishi (Sites 1–3) are located in the solfatara field near the shore of the crater lake, near which are distributed many fumaroles (Fig. 1). The soil surface at Site 1, which is closest to the fumaroles, is exposed to H2S at a level of 2.3 mg S cm−2 month−1 (Takamatsu et al. 1992). This site is almost bare land, although some solfatara plants (e.g. Carex angustisquama and Deschampsia flexuosa) are present in the vicinity (Tsujimura 1977). Although Sites 2 and 3 are adjacent to Site 1, a dense vegetation cover of dwarf bamboo (Sasa kurilensis) is present at these sites. Sites 4 and 5 are located in artificial coniferous forests of Cryptomeria japonica and Thujopsis dolabrata, respectively. The stands in these sites appeared healthy, although their location is relatively near the fumaroles. Site 6, located outside the Osorezan crater, was selected as a reference site. This site is the least affected by volcanic gases because of interception by the somma and has the same vegetation (Cryptomeria japonica) as Site 4 and the same soil type (Dystric Cambisol) as Sites 4 and 5.

Table 1.   Site characteristics in the present study
SiteSoil class (FAO/UNESCO taxonomy)VegetationDistance from the nearest fumaroles (m)Soil depth collected (cm)†
  1. Soil samples were collected after removing fresh litter and rhizomes.

Site 1RegosolBare land (solfatara plants are sparsely distributed around the site)80–7 
Site 2Ochric AndosolsSasa communities (Sasa kurilensis)150–15
Site 3Ochric AndosolsSasa communities (Sasa kurilensis) with sparsely distributed deciduous broad-leaved trees (Quercus crispula)300–15
Site 4Dystric CambisolConiferous forest (Cryptomeria japonica)1000–10
Site 5Dystric CambisolConiferous forest (Thujopsis dolabrata)6000–12
Site 6Dystric CambisolConiferous forest (Cryptomeria japonica)29000–8 

After removing the fresh litter and the rhizomes of dwarf bamboo at Sites 2 and 3, surface soil samples (primarily F/H/A horizons) were collected because these were expected to be most affected by volcanic acid deposition. The depth of the soil collected at each site is shown in Table 1. Because the soil had no well-differentiated horizons, a yellowish-brown color was considered to be the B-horizon and the upper part of the soil above the B-horizon was collected. Soil samples at Sites 2–5 were collected predominantly from the F/H horizons because of the thickness of these horizons. Sampling was conducted in summer from 1998 to 2000 and the collected samples were transported to the laboratory in a cooling box.

Soil chemical analysis

Aliquots of the soil samples were used to measure the pH and the concentrations of water-soluble elements. Soil pH was measured in both water and 1 mol L–1 KCl suspensions (1:2.5 w/v) with a glass electrode. Soils (10 g) were extracted with 50 mL deionized water by shaking for 30 min on a reciprocal shaker. The resulting extracts were filtered (pore size: 0.45-μm, 13AI; GL Sciences, Tokyo, Japan). The concentrations of inline image, Cl, inline image, inline image and inline image in the extracts were measured by ion chromatography (DX-100; Dionex, Osaka, Japan), and the concentrations of Al, K, Na, Ca and Mg were measured by inductively coupled plasma-atomic emission spectrometry (ICP-AES) (ICAP-750; Nippon Jarrell-Ash, Tokyo, Japan) after adjusting the pH of extracts to <2 with HNO3. The remaining soil samples were sieved (<5 mm), mixed and then stored at 4°C until use (generally within 1 month).

Aliquots of the stored soil samples were air-dried, ground and homogenized for analysis of total carbon (TC), nitrogen (TN) and sulfur (TS) contents. Total C and TN were measured after combustion with a CN corder (MT-500, Yanagimoto, Kyoto, Japan). For TS analysis, the soil samples (10 mg) were digested with mixtures of 60% HNO3 (2 mL) and 60% HClO4 (1 mL) in stainless-steel high-pressure digestion bombs (140°C, 6 h) and subsequently with 50% HF (1 mL) on a hot plate (at ∼190°C) (Hou et al. 2006; Takamatsu et al. 1992). The digests obtained were diluted with deionized water, filtered (pore size: 0.45-μm) and analyzed for sulfur by ICP-AES (61E-Trace; Thermo Jarrell-Ash, MA).

Soil microbial analysis

Determination of soil microbial biomass C, N and S

The microbial biomass carbon (MBC), nitrogen (MBN) and sulfur (MBS) in the soils were determined using the chloroform fumigation–extraction method (Brookes et al. 1985;Saggar et al. 1981; Vance et al. 1987). Before analysis, the stored soil samples were adjusted to 60% of their respective water-holding capacities and pre-incubated at 25°C for 10 days in the dark to stabilize microbial activities (Vance et al. 1987; Wu et al. 1994). Pre-incubated soil samples (10 g fresh weight) were fumigated with alcohol-free chloroform for 24 h at 25°C in a sealed desiccator. After removal of the chloroform, the fumigated soil samples were extracted with 50 mL of 0.5 mol L–1 K2SO4 by shaking for 30 min and the extract was filtered. For the analysis of MBS, 0.1 mol L–1 CaCl2 was used as the extractant. Unfumigated soil samples were also subjected to extraction using the same method. Organic carbon in the K2SO4 extracts was measured by an automatic combustion TOC analyzer (TOC-5000; Shimadzu, Kyoto, Japan). The MBC was calculated as EC/kEC, where EC = (organic C extracted from fumigated soils) − (organic C extracted from unfumigated soils) and kEC (conversion factor) = 0.45 (Joergensen 1996; Wu et al. 1990). Total nitrogen in the K2SO4 extracts was determined using the hydrazine reduction method (Hayashi et al. 1997; Kamphake et al. 1967) following alkaline persulfate oxidation (Cabrera and Beare 1993). The MBN was calculated as EN/kEN, where EN = (total N extracted from fumigated soils) − (total N extracted from unfumigated soils) and kEN = 0.54 (Brookes et al. 1985; Joergensen and Mueller 1996). Total S in the CaCl2 extracts was measured by ICP-AES following HNO3 digestion, which was conducted to convert organic S to inline image and/or to prevent the formation of precipitates. The MBS was calculated as ES/kES, where ES = (total S extracted from fumigated soils) − (total S extracted from unfumigated soils) and kES = 0.35 (Banerjee and Chapman 1996; Wu et al. 1994).

Compilation of data on MBC, MBN and MBS in common Japanese forest soils

To compare the soil parameters obtained in the present study with those from common Japanese forest soils, previously published data on TC, TN, TS, MBC, MBN and MBS were compiled. All compiled data for microbial biomass were determined using the chloroform fumigation–extraction method. However, the conversion factors for MBC and MBN (kEC and kEN) in the literature differed occasionally from those used in the present study. In such cases, the data were corrected using our values (i.e. kEC = 0.45 and kEN = 0.54). When compiling the data, values influenced by fertilization or weeding were excluded. In addition, the data used were taken only from texts and tables and not from figures.

Data on MBC and MBN were abundant for mineral soil horizons (e.g. Ohse et al. 2003b; Tokuchi et al. 2000), but limited for organic soil horizons (Chihara et al. 2000; Oyanagi et al. 2002a). Therefore, data for both MBC and MBN were used without distinction in our discussion, although some Osorezan acid soils had thick organic soil horizons. Soil microbial biomass varies significantly with several soil factors, such as vegetation, soil type, temperature and water conditions (Bauhus et al. 1998; Devi and Yadava 2006; Oyanagi et al. 2002a; Sato et al. 2000); thus, only tentative comparisons of soil microbial biomass and related parameters may be possible between Osorezan acid soils and common Japanese forest soils.

Determination of N mineralization and nitrification rates

Potential net N mineralization and net nitrification rates were determined by aerobic incubation of the soils. Because the soil samples had been stored at 4°C for 1–6 months, the soils were pre-incubated as previously described to allow the microbial activity to restore and stabilize. Pre-incubated soil samples (5 g dry weight) were then placed into 100 mL bottles and sealed loosely. The bottles were incubated at 30°C for 4 weeks in the dark, with the moisture maintained at 60% of water-holding capacity by periodic addition of deionized water (control plots). Soil samples spiked with ammonium (200 mg N kg−1 dry soil) as (NH4)2SO4 solution prior to incubation were also incubated in a similar manner (+inline image plots). The +inline image plots were used to study the effect of inline image accumulation on rates of N mineralization and nitrification. Fifteen bottles were prepared for both the control plots and the +inline image plots. Three replicate bottles were retrieved from each plot weekly and inline image and inline image were extracted with 50 mL of 1 mol L–1 KCl by shaking for 30 min. The extracts were filtered and analyzed colorimetrically for inline image-N and inline image-N using the Berthelot reaction (salicylate–nitroprusside–hypochlorite method) (Anderson and Ingram 1993; Searle 1984) and the hydrazine reduction method (Hayashi et al. 1997; Kamphake et al. 1967), respectively. Rates of net N mineralization and net nitrification during the 4-week incubation were calculated from the differences in total inorganic N (sum of inline image-N and inline image-N) and inline image-N concentrations, respectively, before and after incubation in the control plots. inline image-N was not detected in the extracts throughout the incubation.

Enumeration of autotrophic ammonia oxidizers

Autotrophic ammonia oxidizers in pre-incubated soil were enumerated using a microtechnique for the most-probable-number (MPN) method (Rowe et al. 1977). Until recently, ammonia oxidation, the first step in nitrification, was considered to be carried out almost exclusively by chemolithoautotrophic ammonia-oxidizing bacteria in soils. However, recent molecular ecological studies have suggested that ammonia-oxidizing archaea may also play an important role in nitrification (Hayatsu et al. 2008; Leininger et al. 2006; Nakaya et al. 2009; Prosser and Nicol 2008). The MPN method is expected to detect both groups of microorganisms as ammonia oxidizers because it is based on the chemical detection of potential ammonia oxidation. Five grams of pre-incubated soil samples and 45 mL of sterile water were placed into sterile bottles and shaken vigorously for 10 min on a reciprocal shaker. Serial 10-fold dilutions were made by adding 1 mL of the suspension to 9 mL of sterile water. The neutral inorganic medium for ammonia oxidizers contained (in g L–1): (NH4)2SO4, 0.5; K2HPO4, 1.0; MgSO4·7H2O, 0.3; NaCl, 0.3; FeSO4·7H2O, 0.03; CaCO3, 7.5. The medium (0.1 mL) was dispensed into each well of a 96-well microtiter plate and the serial dilution (0.1 mL) was inoculated into eight replicate wells for each dilution. The pH values in the medium inoculated with 10-fold dilutions ranged from 6.3 to 6.9. The inoculated media were incubated at 30°C for 4 weeks in the dark. Growth of ammonia oxidizers was determined by the production of nitrite and/or nitrate in the medium. Nitrite was detected using the Griess–Ilosvay reaction following the reduction of nitrate to nitrite by the addition of zinc powder. The numbers of wells that were positive or negative to the growth test were noted, and the MPN values were calculated from a MPN table (Rowe et al. 1977).

Statistical analysis

All statistical calculations were carried out using StatView 4.5j (Abacus Concepts, CA).

Results and Discussion

Soil chemical properties

The chemical properties of the Osorezan acid soils are shown in Table 2. The soil pH (H2O) at Sites 1–5 within the crater ranged from 3.0 to 3.8, and the soil pH at Site 6 was 5.4 (Table 2). The soil pH at Sites 1–5 was much lower than that recorded in Japanese forest soils, where the soil pH (H2O) of mineral soils ranged from 3.5 to 8.1, with a median of 5.1 (Takahashi et al. 2001). The pH at Site 6 (the reference site) located outside the crater was close to the median of the Japanese forest soils. The surface soil at Site 1 adjacent to the fumaroles was the most strongly acidified. Direct exposure of soil and vegetation surfaces to volcanically generated H2S and its subsequent oxidation to H2SO4 has been shown to cause soil acidification at Osorezan (Takamatsu et al. 1992). Low soil pH around fumaroles has also been reported in other Japanese volcanic regions, for example, pH 2.4–4.2 at Mt Garandake in Oita prefecture (Yoshitake et al. 2007) and pH 2.5–3.9 at Hachimantai in Iwate prefecture (Yoshida and Wakamatsu 1991).

Table 2.   Chemical and microbial properties of the Osorezan acid soils
 pHTotal contentsMicrobial biomassAAO‡
(H20)(KC1)CN (g kg−1 DS)SC/NCN (mg kg−1 DS)SC/N(log10MPN g−1  DS)
  1. Values were expressed on an oven-dry soil basis (DS). AAO indicates autotrophic ammonia oxidizers. §Microbial biomass values at Site 1 are semi-quantitative because these values are close to the determination limits. Values are the means of duplicate or triplicate measurements. N.D., not detected using the most-probable-number method.

Site 13.02.5313.41.09.17.1§12§3.8§0.6§N.D.¶
Site 23.32.635217.82.220838206304.1N.D.
Site 33.72.737718.72.0201011322623.1N.D.
Site 43.62.633815.11.42251593235.5N.D.
Site 53.82.737314.01.7271318230735.7N.D.
Site 65.43.8957.70.712264168241.60.9

Total C and TN were lowest at Site 1, which has little or no vegetation. At the other sites with vegetation, Sites 2–5 had higher TC and TN than Site 6 (the reference site) (Table 2). The thick F/H horizons at Sites 2–5 result from the accumulation of organic matter on the soil surface, probably resulting from the suppressed decomposition of litter and plant residues under strongly acidic conditions (Anderson and Joergensen 1997). The TS at Sites 1–5 (1.0–2.2 g kg−1) was comparable to that in organic and mineral soils in Japanese forests (∼0.1–2.2 g kg−1) (Sakuma et al. 1994; Takamatsu et al. 1992; Tanikawa et al. 2003), despite the continuous exposure of the soils to H2S.

Takamatsu et al. (1992) showed that over 70% of the sulfur is present as organic S and that inorganic S is present exclusively as inline image in acid soils from Oniishi (corresponding to Sites 1–3). In fact, inline image was the major anion in the soil extracts from Sites 1–5, and the proportion of inline image to total anions increased from Site 5 to Site 1, with increasing closeness to fumaroles (Fig. 2). In contrast, at Site 6 (the reference site), inline image was the major anion and was also the dominant form of inorganic N in the soil extracts (Fig. 2). However, at Sites 2–5 inline image was not detected except at Site 4, and inline image was the dominant form of inorganic N. Base cations (Ca2+, Mg2+, K+ and Na+) were dominant at Site 6, whereas acidic cations (H+ and Al3+) increased at Sites 1–5 with decreasing soil pH (Fig. 2). The high concentrations of K+ at Sites 2 and 3 and Ca2+ at Sites 4 and 6 might have been influenced by the vegetation because Sasa kurilensis and Cryptomeria japonica accumulate K (Takamatsu et al. 1997) and Ca (Baba et al. 2004), respectively.

Figure 2.

 Concentrations of anions and cations in water extracts from Osorezan acid soils. The H+ concentrations were calculated from pH values of the extracts.

Cation exchange and aluminum hydrolysis are the main pH-buffering systems in acid soils (Ulrich 1991). However, the soils from Oniishi (corresponding to Sites 1–3) were depleted not only in exchangeable base cations, but also in aluminum (exchangeable Al3+ and Al-oxides) as a result of long-term exposure to acidic volcanic gases (Takamatsu et al. 1992). Thus, the pH-buffering system has progressed to the next stage, in which iron hydrolysis and H+-exchange at strongly acidic sites of soluble and insoluble organic matter and clay minerals are initiated (Takamatsu et al. 1992). Similarly, the acid soils from Osorezan demonstrate a unique inorganic environment.

Soil microbial properties

Microbial biomass C, N and S

The MBC, MBN and MBS are shown in Table 2. These values tended to be higher at Sites 2–5 where the soils were rich in organic matter, although a significant correlation was found only between TC and MBC (r = 0.884, < 0.01, = 6). Soil pH showed no significant correlation with MBC, MBN or MBS (r = −0.129, 0.208 and 0.065, respectively, > 0.1, = 6). Soil organic matter contents and vegetation types are significant parameters controlling soil microbial biomass (Banerjee and Chapman 1996; Bauhus et al. 1998; Chowdhury et al. 1999; Ohya et al. 2000; Oyanagi et al. 2002a). The ratios of MBC to MBN (biomass C/N) were correlated with those of TC to TN (soil C/N) (r = 0.962, < 0.01, = 6). Thus, biomass C/N ratios in the Osorezan acid soils might be affected by soil C/N ratios. The MBC, MBN, and MBS were well correlated with each other (r = 0.810–0.940,  0.05, = 6).

To characterize the soil microbial biomass at Osorezan, the relationships between MBC and TC and between MBN and TN were compared with literature data from common Japanese forest soils (Fig. 3a,b). In the common forest soils, MBC was positively correlated with TC (r = 0.746, < 0.01, = 49) (Fig. 3a). The MBN was also correlated with TN (r = 0.488, < 0.01, = 41) (Fig. 3b). The proportion of MBC to TC (the MBC/TC ratio) ranged from 0.5 to 4.2%, with an arithmetic mean and standard deviation of 1.7 ± 0.9%, and the MBN/TN ratio was 3.9 ± 2.8% (range: 0.7–16.4%). In contrast, the MBC/TC ratio in the Osorezan acid soils was low (<0.5%) even at Sites 2–5 where the TC was high (Fig. 3a). The MBN/TN ratio in the Osorezan acid soils was relatively low even at Sites 2–5 with high TN, and ranged from 0.3 to 1.7% at Sites 1–5 (Fig. 3b). These results indicate that the microbial biomass is less in the Osorezan acid soils than in common Japanese forest soils.

Figure 3.

 Relationships between (a) total C (TC) and microbial biomass C (MBC), (b) total N (TN) and microbial biomass N (MBN), (c) MBN and MBC, and (d) TN and TC in Osorezan acid soils (bsl00001) and common Japanese forest soils (○). Values are expressed on an oven-dry soil basis (DS). The site number is labeled near each plot. Values in common Japanese forest soils were cited from: Billore et al. 1995; Chihara et al. 2000; Chowdhury et al. 1999; Guan et al. 1997; Furusawa et al. 2005; Mabuhay et al. 2006; Ohse et al. 2003a,b; Ohya et al. 2000; Oyanagi et al. 2002a,b; Sato et al. 2000; Sato and Seto 1995; Tamai et al. 2003; Tokuchi et al. 2000; Tripathi et al. 2005, 2006.

Soil microbial biomass generally decreases with acidification. Anderson and Joergensen (1997) showed a decreasing MBC/TC ratio with decreasing soil pH in a German beech forest. Thirukkumaran and Morrison (1996) and Pennanen et al. (1998) found that the ratio of MBC/TC in forest soils decreased with the application of simulated acid rain. Only a few microbial communities can adapt to extremely acidic soil environments that are deficient in nutrients such as base cations and rich in potentially toxic ions such as Al3+ (Pennanen et al. 1998; Yoshitake et al. 2007). Thus, microbial biomass and survival may have been restricted in the Osorezan acid soils, resulting in low MBC/TC and MBN/TN ratios.

Relationships between MBC and MBN and between TC and TN were also compared between the Osorezan acid soils and the common forest soils. In the latter, MBC was highly correlated with MBN (r = 0.934, < 0.01, = 27) as shown in Fig. 3c and the biomass C/N ratio was 6.6 ± 1.8 (range: 4.8–14.0). These values were similar to those (7.0 ± 0.3) found in forest soils around the globe (Cleveland and Liptzin 2007). Relative to these values, biomass C/N ratios in the Osorezan acid soils were slightly low, although this difference was not significant (Fig. 3c). Total C was correlated with TN (r = 0.865, < 0.01, = 63) in common forest soils (Fig. 3d), and the soil C/N ratio was 17.2 ± 5.2 (range: 8.2–33.4). Soil C/N ratios in the Osorezan acid soils were almost equivalent to this (Fig. 3d). Thus, biomass C/N and soil C/N ratios were similar in the Osorezan acid soils to those in common Japanese forest soils.

Although soil C/N ratios in the Osorezan acid soils were equivalent to those in common Japanese forest soils, the concentrations of elements other than C and N in dwarf bamboo and litter have been found to be significantly different in the Osorezan area from those in other non-acidic areas (Takamatsu et al. 1997). Therefore, the change in quality (availability and decomposability) of soil organic matter may have influenced the microbial biomass (Bauhus et al. 1998; Ohse et al. 2003b; Raubuch and Beese 2005). At Sites 2 and 3, it is also possible that competitive nutrient uptake by the dense dwarf bamboo inhibited any increase in microbial biomass (Takamatsu et al. 1997; Tripathi et al. 2006).

Although information on MBS is limited, Chowdhury et al. (1999) showed that MBS and the MBS/TS ratio in two different Japanese forest soils were 6.4 and 8.0 mg kg−1 and 4.0 and 3.4%, respectively. When including grassland and arable land, the values were 0.8–13.4 mg kg−1 and 1.1–4.0%, respectively. Banerjee and Chapman (1996) found that MBS in global forest soils ranged from 2.9 to 140 mg kg−1. The MBS values in the Osorezan acid soils (3.8–73.2 mg kg−1) were higher than those found by Chowdhury et al. (1999), but were within the range found in global forest soils. However, the MBS/TS ratios (0.4–4.3%) were similar to those found by Chowdhury et al. (1999).

Nitrogen mineralization and nitrification

To clarify the nitrogen cycle in the Osorezan acid soils, the potential rates of N mineralization and nitrification were measured in aerobic laboratory incubations. Figure 4 shows the changes in inline image-N and inline image-N concentrations during the course of the incubation. At Sites 1–5, inline image-N increased almost linearly throughout the incubation period, whereas inline image-N did not increase. A similar increase in inline image-N was observed in the +inline image plots. In contrast, at Site 6 (the reference site), although inline image-N was always present in trace amounts, inline image-N increased during the incubation and at day 28 the proportion of inline image-N to total inorganic nitrogen was 97%. In the +inline image plots, inline image-N decreased and inline image-N increased concomitantly. These results indicated that net N mineralization occurred at approximately constant rates at all sites, whereas net nitrification occurred only at Site 6 and was inhibited completely at Sites 1–5.

Figure 4.

 Changes in inline image-N and inline image-N concentrations in Osorezan acid soils during aerobic incubation. Values are expressed on an oven-dry soil basis (DS). In the +inline image plots, the soil was spiked with ammonium (200 mg N kg−1) before the incubation. Concentrations of inline image-N and inline image-N in the control plots at the start of the incubation were set to zero.

At Sites 1–5, autotrophic ammonia oxidizers were not detected in the soil by the MPN method (Table 2). In addition, inline image-N was not detected in the soil extracts except at Site 4 (Fig. 2). These results suggest that ammonia oxidizers are absent or that activity is seriously inhibited in the Osorezan acid soils. Ammonia oxidation is conducted mainly by a relatively narrow range of bacteria and maybe by archaea, although it remains unclear as to what extent archaea actually contribute to nitrification activity in soils (Hayatsu et al. 2008; Leininger et al. 2006; Nakaya et al. 2009; Prosser and Nicol 2008). Ammonia-oxidizing bacteria (e.g. Nitrosospira and Nitrosomonas) prefer NH3 to inline image as a substrate, but the availability of NH3 decreases exponentially with decreasing pH (NH3+ H+inline image; pKa = 9.25), and there are few acid-tolerant strains that can use inline image at low pH (De Boer and Kowalchuk 2001; Frijlink et al. 1992). Although pH-neutral medium were used for MPN enumeration, ammonia oxidizers were not detected in the Osorezan acid soils. In addition to nitrogen availability, other factors such as the presence of natural inhibitors and water stress suppress the activity of autotrophic nitrifiers (De Boer and Kowalchuk 2001; Hirobe et al. 1998; Paavolainen et al. 1998; Tokuchi et al. 2000). In contrast, heterotrophic or fungal nitrification also contributes to nitrification in some acid forest soils (Brierley et al. 2001; De Boer and Kowalchuk 2001). At Site 4, the concentration of inline image-N did not increase, but certain amounts (∼38 mg-N kg−1) of inline image were detected throughout the incubation period, although inline image was not detected in the other acid soils. This inline image might have been produced by heterotrophic or fungal nitrification in view of the absence of autotrophic ammonia oxidizers (Table 2).

The rates of net N mineralization in the control plots were in the order: Site 3 (139 mg N kg−1 4 weeks−1) ≥Site 2 (132) ≥ Site 5 (128) ≥ Site 4 (122) > Site 6 (62) > Site 1 (25). The corresponding values in the +inline image plots ranged from 32 to 203 mg N kg−1 4 weeks−1, and were 1.2–1.6-fold higher than those in the control plots. This result suggests that inline image accumulation probably accelerates N mineralization in the soils. Because inline image was accumulated during the storage and pre-incubation periods in the control plots (21–273 mg N kg−1), there is a possibility that the N mineralization rates were overestimated. The N mineralization rate was significantly correlated with both TC (r = 0.989, < 0.001, = 6) and TN (r = 0.979, < 0.001, = 6), without the influence of soil pH. A similar dependence of N mineralization rate on the amount of soil organic matter has been reported by Hirobe et al. (1998) and Hirai et al. (2006). The N mineralization rates in the Osorezan acid soils were compared with the rate in common Cryptomeria japonica forest soils. The latter ranged from 12.5 to 245.3 mg N kg−1 4 weeks−1 and the arithmetic mean ± standard deviation were 93.7 ± 58.0 mg N kg−1 4 weeks−1 (= 85) (Hirai et al. 2006). The N mineralization rates observed in the Osorezan acid soils (25–139 mg N kg−1 4 weeks−1) were within the range of those found in common forest soils. In contrast to nitrification, N mineralization is conducted by various microorganisms, some of which may have been able to adapt to the acidic conditions in Osorezan and maintain mineralization rates.

Inorganic N is essential for plants and soil microorganisms, but inline image-N is readily leached from soils. A lack of nitrification prevents nitrogen leaching from the acid soils, and thus an optimum nitrogen cycle, in which inline image (and NH3) is recycled among soil microorganisms and plants, may have been established in Osorezan.

Acknowledgments

The authors thank Dr M. Nishikawa, Mrs R. Kumata and Mrs M. Okawa of the National Institute for Environmental Studies for assistance with the ICP-AES analysis.

Ancillary