Sulfur forms in bulk soils and alkaline soil extracts of tropical mountain ecosystems in northern Thailand.
Maintenance of sufficient levels of sulfur (S) for plants and animals involves a complex biogeochemical cycle in which S is converted among various organic and inorganic forms. Nearly 95% of total S in most non-calcareous surface soils exists in organic forms (Biederbeck 1978). Mineralisation of soil organic matter S depends on factors influencing the growth of microorganisms, such as temperature, soil moisture, and pH (Bettany et al. 1973; Stevenson and Cole 1999), and only some of the S is released as sulfate (S[O.sup.2-.sub.4]), which is the main source of S for plant uptake. Most of the S remains in the organic form. Microbially mediated release of S is not uniform but seems to take place at different time scales from various organic matter pools and soil compartments (McLaren and Swift 1977). Despite the close relationship between soil organic matter and soil S, we know little about S pools and how they are connected to the transformation of soil organic matter.
Additionally, most of the investigations on S pools and their transformations are from soils from temperate regions (Tabatabai and Bremner 1972; Bettany et al. 1979; Johnson and Henderson 1979; Zucker and Zech 1985; Autry et al. 1990; Olson and Lowe 1990). Little research has been done in the tropics (McLaren and Swift 1977; Acquaye and Kang 1987; Motavalli et al. 1993), and almost no information is available on the dynamics of soil S pools in tropical mountain ecosystems (Solomon et al. 2001). Our future ability to manage these soils sustainably is limited by this lack of understanding.
The hill site area of northern Thailand has been rapidly deforested and intensively used for agriculture for the past few decades. During the last 20 years, parts of that region were reforested and also secondary forest has developed. The replacement of forests by annual crops results in enhanced breakdown of soil organic matter, whereas re-establishment of forests induces a build-up of soil organic matter in the long term. Consequently, arable and forest management of the sites have opposing effects on soil organic matter and on the transformation and distribution of soil S. Thus, these hill sites offer a range of soils with soil organic matter in various states of transformation, which allows the study of how the transformation of different forms of nutrient elements relates to that of C binding forms under tropical conditions (Moller et al. 2000).
A number of methods exist for the characterisation of S forms in soil (Lowe and DeLong 1963; Kowalenko 1978; Tabatabai 1996). Recently, Kowalenko (1993a, 1993b) introduced a fast and versatile method for fractionation of soil S into organic (C-bonded S and ester S[O.sub.4]-S) and inorganic forms using hydroiodic acid reduction of S[O.sup.2-.sub.4]. The C-bonded S pool in tropical soils represents mainly S forms derived from leaves and other photosynthetic tissues, while amino acid S represents S that has been re-metabolised by soil microorganisms (Stanko-Golden and Fitzgerald 1991). The second major component of organic S, ester S[O.sub.4]-S, is predominantly generated by soil microflora that metabolises organic residues in the presence of adequate S sources (Saggar et al. 1998). Sulfonate-S falls into the C-bonded S pool and may be a good source for the release of S[O.sup.2-.sub.4], and due to its abundance, it possibly serves also as a storage form of S (Stanko-Golden and Fitzgerald 1991). The results of Stanko-Golden and Fitzgerald (1991) agree well with those of Ghani et al. (1991), who showed in their study of New Zealand soils that C-bonded S is the best predictor for mineralisable S and that the majority of mineralised S is derived from the C-bonded S pool.
In our study, we applied this fractionation to bulk samples and alkaline extracts of soils from different land-use systems. The results were related to previous results on the distribution of C binding forms as revealed by [sup.13]C-NMR spectroscopy on the alkaline extracts (Moller et al. 2000). By correlating S and C forms we sought new insights into the coupling of S and C transformations in soil organic matter.
Materials and methods
Experimental sites were on the Khun Sathan Research Station, which is located in the Na Noi district of northern Thailand at an elevation of 1350-1500 m. The soils at the sites were Dystric and Humic Cambisols on lower slopes and Skeletic Umbrisols and Lithic Leptosols on steeper sloping areas. They derived from shale and sand- and silt-stone. Four representative ecosystems in the region were investigated: Pinus reforestation, 20-year-old Pinus kesiya (Royle ex Gordon); secondary forest, natural succession trees with 15-year-old Pinus kesiya (Royle ex Gordon); primary forest; and cabbage cultivation. The reforestation, secondary forest, and cabbage cultivation sites were cleared about 30 years ago. Under the forest vegetation, organic layers covered the mineral soil surface. A more detailed description of the sites and the soils is given in Moller et al. (2000).
Sampling and basic analyses
Soils were sampled in 1997 and 1998. At each site, 2 representative soil profiles were selected and sampled by horizons. Additionally, from at minimum 6 subplots all across the 4 sites, composite samples of O, A, and B horizons were collected from small soil pits in a radial sampling scheme. All samples were air-dried and the mineral samples were sieved to <2 mm.
Soil pH was determined by a glass electrode with 0.01 M Ca[Cl.sub.2] at a soil-to-solution ratio of 1 : 2.5 (w/v). Total S, C, and N were analysed with a CHNS-analyser (Vario EL, Elementar Analysensysteme GmbH, Hanau, Germany). Cation exchange capacity was determined with 1 M N[H.sub.4]OAc (pH = 7.0) according to Avery and Bascomb (1974). The soils were free of carbonates. Texture analysis was conducted after HCl, hot [H.sub.2][O.sub.2], and ultrasonic treatment followed by sieving of sand and determination of silt and clay by the pipette method (Avery and Bascomb 1974). Extraction of oxalate-soluble Al and Fe was carried out with acid 0.2 m ammonium oxalate at pH 3 (Schwertmann 1964). Dithionite-soluble Fe was determined by the sodium dithionite-citrate-bicarbonate method (Mehra and Jackson 1960). The results of the basic analyses are summarised in Table 1.
The inorganic S[O.sub.4]-S fraction was analysed by extracting soluble and adsorbed S[O.sup.2-.sub.4] with K[H.sub.2]P[O.sub.4] (500 mg P/L) followed by reduction of S[O.sup.2-.sub.4] to sulfide using hydroiodic acid (HI) and determination of S as bismuth sulfide colouring complex by spectroscopy at 400 nm (Kowalenko 1993a). Although some of the subsoil horizons from the primary forest site exhibited gleyic properties, analyses for reduced inorganic S using the zinc-hydrochloric acid distillation (Aspiras et al. 1972) followed by determination of S as bismuth sulfide were all negative. Thus, total organic S was estimated as the difference between the total S and inorganic S[O.sup.2-.sub.4] extracted by K[H.sub.2]P[O.sub.4]. HI-reducible S, consisting primarily of ester S[O.sub.4]-S and inorganic S[O.sub.4]-S, was analysed by direct reduction with HI reducing mixture and measured on a spectrophotometer according to Kowalenko (1993b). The difference between HI-reducible S and inorganic S[O.sub.4]-S (K[H.sub.2]P[O.sub.4] extractable) was considered as ester S[O.sub.4]-S. Based on the finding of Strickland et al. (1987) that C-S (amino acid) or C-S[O.sub.3] (sulfonate) linkages will not be cleaved by HI, C-bonded S was considered to be total organic S minus HI-reducible S.
Extraction of soil organic matter
To identify a possible relationship between S fractions and organic C-binding forms, we used an alkaline extractant. The soil samples were extracted 3 times with a solution containing 0.1 M NaOH and 0.4 M NaF at a ratio of soil to extraction solution of 1 : 5 (w/v). The sample size was adjusted to ensure a minimum yield of 200 mg of OC in the extracts. The extraction procedure followed the outline of Schnitzer (1982), modified by Sumann et al. (1998). The extracts were dialysed (molecular weight cutoff, MWCO, 12 000-14 000 daltons) in order to remove salts then freeze-dried. Losses of inorganic and organic S within low-molecular-weight compounds during the extraction procedure were accepted to retain the comparability of the sample structure. The identification of organic C forms was done by liquid-state [sup.13]-CNMR spectroscopy and is described in detail in Moller et al. (2000). The freeze-dried extracts were analysed for inorganic and organic S forms as described above.
Means of concentrations of S constituents (total S, organic S, C-bonded S, ester S[O.sub.4]-S, inorganic S[O.sub.4]-S), OC, and N of the whole data set were compared by horizons among the 4 experimental sites using 1-way analyses of variance (ANOVA), followed by a post-hoc separation (significant at P < 0.05 and in some cases P < 0.1) of means done by a univariate l.s.d, test. The Pearson product moment correlation with a 2-tailed test of significance was used to show the relationship between S fractions, OC, and N, and [sup.13]C-NMR total spectral areas. All statistical analyses were conducted using the software package STATISTICA 5.0 for Windows (StatSoft, Inc., Tulsa, OK).
Results and discussion
Sulfur fractions and pool sizes
The highest concentration of total S was found in the organic layers of the primary forest. It ranged from 1794 to 2705 mg/kg. At the Pinus reforestation, the concentrations were smaller, ranging between 1283 and 1732 mg/kg (Figs 1 and 2). Organic layers of the secondary forest showed the most variable S content, ranging between 1447 and 2410 mg/kg. The large variation in S concentrations can be attributed to the input of litter of differing quality, caused by the alternation of vegetation at the secondary forest site comprising planted Pinus and natural succession trees.
[FIGURE 1-2 OMITTED]
Total S content decreased with soil depth following the same pattern as organic C. It ranged in the mineral soil between 149 and 802 mg/kg (Figs 1 and 2). The only exception was profile A at the secondary forest site where the C horizon contained 2140 mg S/kg. The total S concentrations we found were higher than those reported by Acquaye and Kang (1987) for Ghanaian topsoils (44-281 mg/kg) and Neptune et al. (1975) for Brazilian topsoils (59-398 mg/kg), which also had much lower organic matter levels, but were within the range of values found for topsoils from Puerto Rico (353-1231 mg/kg; Stanko-Golden and Fitzgerald 1991) and Ethiopia (520-1041 mg/kg; Solomon et al. 2001), which had similar organic matter levels compared with the soils in our study. This may indicate that the S status in soil is strongly dependent on the content and turnover rate of organic matter in soil. The ratios C : N, N : S, and C : S of all analysed samples averaged 34, 7.8, and 252 in the organic horizons, 12, 7.7, and 93 in the surface horizons, and 8, 4.7, and 38 in the subsoil horizons, respectively (Figs 1 and 2), demonstrating the advanced organic matter transformation and the preservation of S compared with N and C with soil depth.
Organic S was the predominant S form in the investigated soils, comprising between 75 and 99% of total S. Exceptions were the B1 and B2 horizons of the profiles at the cabbage cultivation site, where organic S accounted for <75% of the total S. Organic S was higher than reported in most studies from tropical lowland regions, but within the range of the results from the temperate zone and tropical highlands (Tabatabai and Bremner 1972; Ghani et al. 1991; Stevenson and Cole 1999; Solomon et al. 2001). This indicated a higher S storage and possibly availability within the hill site systems compared with the tropical lowlands.
Organic S was significantly correlated with total S, C, and N in soil. The correlations (P < 0.001) found in this study between organic S and total S (R = 0.998), C (R = 0.957), and N (R = 0.973) were similar to those reported by other authors (Acquaye and Kang 1987; Ghani et al. 1992; Solomon et al. 2001). This points again to a strong relationship between organic matter decomposition and N and S transformation.
The major organic S form was C-bonded S comprising 81-99% of total organic S in the organic layers and 35-83% in the mineral soil. Its contribution decreased with soil depth, demonstrating the progressive decomposition of plant-derived organic matter and the increased proportion of re-metabolised organic substances with soil depth. This result is similar to observations for many other soils (Ghani et al. 1991; Haynes and Williams 1992; Houle and Carignan 1992; Solomon et al. 2001).
Ester S[O.sub.4]-S, the second major organic S component, played only a minor role compared with C-bonded S in the studied soils. We observed a maximum of ester S[O.sub.4]-S in the A horizons (128 [+ or -] 49 mg/kg). Less ester S[O.sub.4]-S was found in the organic layers (92 [+ or -] 82 mg/kg or 1-12.2% of organic S). In the mineral soil, its concentration decreased in the upper B horizons, and remained almost constant throughout the lower B horizons and the C horizons (64 [+ or -] 34 mg/kg). Keer et al. (1990) showed an increase in percentage of the HI-reducible S forms (which is almost equivalent to ester S[O.sub.4]-S) with increasing molecular weight of the organic fractions. Therefore the maximum of ester S[O.sub.4]-S at the A horizons can be explained by a selective decomposition of organic matter with a higher proportion of materials of low molecular weight or it may originate from microbial re-synthesis of high-molecular-weight compounds.
Tropical systems frequently show a high rate of organic matter decomposition and recycling of nutrients, therefore only low S[O.sup.2-.sub.4] concentrations are found in soil (Saggar et al. 1998). Most of the released S is directly re-metabolised by the soil microflora. Furthermore, the released S[O.sup.2-.sub.4] may be quickly taken up by plants or leached to greater soil depth (Saggar et al. 1998). In this study inorganic S ranged from 3.5 mg/kg in the lower B and the C horizons of the Pinus reforestation up to 192 mg/kg soil in the organic layer of the secondary forest and 252 mg/kg soil in the subsoil of the cabbage cultivation. The latter can be attributed to the addition of ammonium sulfate fertiliser.
Concentrations of S[O.sup.2-.sub.4] were high in the organic horizons (71 [+ or -] 42 mg/kg) then decreased to a minimum in the A horizons (31 [+ or -] 17 mg/kg). The maximum concentrations of S[O.sup.2-.sub.4] occurred in the upper B horizons (81 [+ or -] 53 mg/kg), while the minimum concentrations were found in the lower B horizons and the C horizons (19 [+ or -] 12 mg/kg). The maximum concentrations of extractable S[O.sup.2-.sub.4] in the upper B horizons may be attributed to the strong sorption of S[O.sup.2-.sub.4] to A1 and Fe oxides-hydroxides and kaolinite (e.g. Couto et al. 1979). The only exception was the soil under the Pinus reforestation where there were low S concentrations throughout the whole soil profile. In general, the results are in line with results reported for temperate and tropical regions (David et al. 1982; Stanko-Golden and Fitzgerald 1991; Zhang et al. 1999; Solomon et al. 2001). Also the S[O.sup.2-.sub.4] concentrations in the soil solution at the plots (data not shown) reflected the depth distribution of extractable S[O.sup.2-.sub.4]. We found low S[O.sup.2-.sub.4] concentrations at 15 cm depth and increased concentrations at 30 and 80 cm depth, respectively.
The minimum concentrations of S[O.sup.2-.sub.4] occurred in the mineral A horizons where the concentrations of ester S[O.sub.4]-S were maximal. These results are not in conflict. They rather refer to the rapid transformation of organic matter and nutrient cycling in these tropical surface soils and the losses through leaching into the subsoils.
Influence of vegetation and S availability on the status and forms of S
Changes in a natural tropical ecosystem are usually reflected by an accelerated transformation of organic matter and nutrient cycling in soil. For the region we compared the natural forest as a control with 3 man-made ecosystems (secondary forest, Pinus reforestation, and cabbage cultivation), using comparable soil compartments (organic layer, A horizons, and upper B horizons).
Comparing the S concentrations in the O horizons of the 3 forest plots, there were distinct differences between the primary forest and the Pinus reforestation site (Figs 1 and 2). The secondary forest site showed S concentrations ranging between those under the primary forest and the Pinus reforestation, possibly due to the inhomogeneous vegetation with a mix of Pinus and broad-leaved species. Most other investigations on forest soils have found little or contrary differences in the O horizons of coniferous and deciduous woods. For example, David et al. (1982) found higher S values in the O horizons of a coniferous site compared with a hardwood site in the Adirondack Mountains of New York. This designates that next to the type of vegetation, other factors are controlling the S and N concentrations in the foliage and thus also in the organic layers. Lambert and Turner (1998) reported that in the foliage and other parts of various tree species over a wide geographic area of south-eastern Australia, S and N in organic forms occurred at a constant ratio of 0.03 on a gram atom basis. Foliar total N of coniferous and broad-leafed forest trees has been found to be equivalent to foliar organic N and there is no evidence of nitrate-N accumulation even when S is limiting. Nitrogen appears to be taken up by most forest tree species only at the rate at which S is available (Lambert and Turner 1998). This result may show that the input of S and N from the vegetation cover results in a constant ratio between S and N. In our study, the mean ratio N: S in the O horizons of the 3 forest sites was within the same range (7.4-8.3), but the concentrations of S and N between the sites were significantly different. The availability of S at the reforestation site was more limited since the concentrations of S, and especially of the K[H.sub.2]P[O.sub.4]-extractable S[O.sub.4]-S, were clearly lower in the A horizons and significantly lower in the B horizons than at the other forest sites. Therefore, the limitation of S at the Pinus reforestation site, next to the variable vegetation, seems to be the most suitable reason for the differences among S and N concentrations in the organic horizons of the 3 forest plots.
C-bonded S showed a distribution in the organic layers similar to total S, whereas ester S[O.sub.4]-S was strongly enriched at the Pinus reforestation site (Figs 1 and 2). The differing concentrations at the tree forest sites may result in a different input of S forms into the mineral soil of these plots.
More than 30 years of cultivation at the cabbage cultivation site resulted in topsoils depleted in S by 19% compared with the natural forest (P < 0.05) (Fig. 3). Even after 15 years of reforestation, S was depleted by 11% indicating a slow rebuild of the S stocks at the Pinus reforestation site. Higher losses were calculated for C and N (38-40% and 26-29%, respectively).
[FIGURE 3 OMITTED]
The total S concentrations at the Pinus reforestation site were similar to those under cabbage cultivation (Figs 1 and 2). In contrast to the primary forest, the litter at the Pinus reforestation site accumulated as organic layer and was not incorporated into the mineral soil. This may be the reason for the slow build-up of organic matter in the mineral soil of the reforestation. Consequently, the quality of the Pinus litter might prevent a fast regeneration of organic matter in the topsoil of this site.
The smaller depletion of S than of C and N may be attributed to the higher resistance of S-containing compounds against mineralisation (Solomon et al. 2001) and to the high input of ester S[O.sub.4]-S at the reforestation site, which is more stable against decomposition than C-bonded S (Saggar et al. 1998). At the cabbage cultivation site the input of fertiliser may also play an important role. A frequently used fertiliser is ammonium sulfate. This influence is more visible in the subsoil. Here we found an accumulation of S[O.sup.2-.sub.4] due to leaching from the topsoils and subsequent sorption on Fe and A1 hydrous oxides and edges of clay particles at the upper B horizons (Eriksen et al. 1998). Another possible explanation is the steep slopes and high rainfall, which induce lateral movement of water causing irregular depletion or accumulation of nutrients (Eriksen et al. 1998). As a consequence, the S and especially the inorganic S[O.sub.4]-S concentrations in the subsoil of the cabbage cultivation were higher than in the primary forest. A tendency to accumulate S[O.sup.2-.sub.4] in the upper B horizons could also be observed at the primary forest and the secondary forest site, but it did not appear at the Pinus reforestation site. There we found in the B horizons again significantly lower S and especially inorganic S[O.sub.4]-S concentrations than at any other site. The lower S concentrations of the Pinus reforestation site may be explained by an inhibition of organic matter decomposition and the fast uptake of available S[O.sub.4]-S by plants and soil microorganisms. Therefore, only little S can be leached or accumulated in the mineral soil under the Pinus reforestation.
C-bonded S accounted for <60% of the organic S at the reforestation site and for >75% at all other sites. The high proportion of ester S[O.sub.4]-S in the mineral soil under the Pinus reforestation corresponded with the large proportion of ester S[O.sub.4]-S in the organic layer (see above). David et al. (1982) observed a decrease in C-bonded S and an increase in ester S[O.sub.4]-S during incubation of soils from a coniferous and a hardwood forest site. Visible fungal biomass in the organic horizons of the reforestation as well as the large contribution of fungal-derived amino sugars (data not shown) may suggest that fungal transformation of soil organic S was accompanied by an increase in ester S[O.sub.4]-S.
The depletion in C-bonded S in the topsoil of the different land-use systems relative to the primary forest was generally higher than the losses in ester S[O.sub.4]-S. These results were in agreement with most literature data (McLachlan and De Marco 1975; Solomon et al. 2001). McLaren and Swift (1977) found that a high proportion (75%) of S lost during long-term cultivation consisted of C-bonded S. Ghani et al. (1991) stated that C-bonded forms of S represented the major source of mineralisable S in soils. Inorganic S[O.sub.4]-S showed a strong depletion at the secondary forest and the Pinus reforestation. However, the concentrations are generally low and the internal S transformations and fluxes may also be influenced by seasonal and annual variations due to climatic and edaphic changes (David et al. 1982).
In most cases, the chemical composition of OC and P in NaOH extracts is representative for total soil OC and P (Frund and Ludemann 1989; Gressel et al. 1996). This may be also true for soil organic S and would allow a comparison of C and S forms, which seems to be a promising way to get a deeper insight into the structural composition and turnover of organic S in these soils.
The C to S and N to S ratios of the NaOH extracts of the organic horizons and topsoil horizons were lower than in the respective bulk soil samples (Fig. 4). This indicated that S-enriched organic matter was extracted from the organic layers and topsoil horizons. In the subsoil, no general trend for increased solubility of S-containing compounds could be observed. In the natural forest, the concentrations of organic S were much lower in the extracts than in the bulk soil samples, while at the other sites similar or slightly higher organic S concentrations were found.
[FIGURE 4 OMITTED]
The percentage of extractable C-bonded S in most cases was larger than that of ester S[O.sub.4]-S (Fig. 4). Bettany et al. (1979 and 1980) found that fulvic acids contain high amounts of HI-reducible S (up to 84% of total S). According to the extraction method used in this study, the losses of fulvic acids during the dialyses could be a possible explanation for the reduced ester S[O.sub.4]-S concentrations in the extracts. On the other hand, Keer et al. (1990), using gel filtration, noted that high-molecular-weight organic matter is enriched in ester S[O.sub.4]-S. In general, relative to the C content in the extracts, the concentrations of the S forms correlated well with those in soil (Table 2). So despite certain differences, in most cases the S forms in the NaOH extracts seemed to be representative for the S forms in the bulk soils.
Note that in the C horizon under the secondary forest the concentration of S and of ester S[O.sub.4]-S normalised to organic C was very high. In contrast, the alkaline-extractable organic matter of this horizon showed no unusually high enrichment in S (Fig. 4). One possible explanation for the concentration of S and ester S[O.sub.4]-S might be the inference of some inorganic sulfide. However, no sulfide could be detected in any of the soil samples studied.
C-bonded S was not correlated to C forms as revealed by [sup.13]C NMR. This suggested that C-bonded S forms were not connected to specific C forms of extractable organic substances.
Bettany et al. (1979 and 1980) supposed that ester S[O.sub.4]-S is probably not connected to aromatic units, rather it is largely associated with aliphatic components. This suggestion was not entirely confirmed by our study comparing the [sup.13]C-NMR spectroscopy results with the S fractions in the NaOH extracts. We found for ester S[O.sub.4]-S significant correlations to C in the O-alkyl C region of the [sup.13]C-NMR spectra and negative correlations to aromatic and phenolic C (Table 3). Signals of aliphatic structures occur in the [sup.13]C-NMR spectra in the alkyl region (Kogel-Knabner 1997; Zech et al. 1996) and in the O-alkyl region, a region of easily biodegradable structures (Kogel-Knabner 1997). The O-alkyl region is dominated by OH, O (and N) substituted aliphatic carbons typical of polysaccharides and amino acids (Catroux and Schnitzer 1987), which are possible partners for the formation of ester bondings. For example, structural studies on sulfated polysaccharides and galactans of algae using [sup.13]C-NMR techniques showed that diverse -S[O.sub.4] bondings contribute to the signals in the O-alkyl region (Kjellberg et al. 1995; Liao et al. 1996; Miller and Blunt 2000). Considering the suggestion of McLaren and Swift (1977) and Saggar et al. (1998) that ester S[O.sub.4]-S is more of a transitory nature and is predominantly generated by microorganisms, the significant correlation between the O-alkyl region of the [sup.13]C-NMR spectra and ester S[O.sub.4]-S form may not be indicative of a direct chemical bonding, but it may show similar dynamics of these structures in soil.
The following conclusions could be drawn:
(1) The S status in the soils of the natural mountain forest ecosystem in this region of northern Thailand is higher than those of many tropical lowland areas.
(2) The dominant organic S fraction in these tropical highland soils is C-bonded S. In addition, most of organic S depletion from the bulk soils as a result of land-use changes occurred from C-bonded S.
(3) The composition of S forms in soil is variable due to small-spatial differences within the soils and the sites, indicating a varying biogeochemical transformation of organic matter in these soils.
(4) The NaOH extracts and bulk soil samples showed similar S concentrations. Yet, ester S[O.sub.4]-S seems to be underrepresented in the NaOH extracts, especially in the organic and topsoil horizons. But in general, correlations between S fractions in bulk soil samples and extracts indicate that alkaline-extractable organic S is more or less representative for the composition of total organic S in soil.
(5) Correlations between the S fractions in alkaline extracts and the results of [sup.13]C-NMR spectroscopy indicate that ester S[O.sub.4]-S is mainly bound to O-substituted aliphatic C, whereas C-bonded S did not show any correlations with specific C forms.
Table 1. Range of chemical properties in the investigated soils Samples were by horizons from 2 soil profiles per site and from several soil pits (composite O, A, and B horizons) over the respective sites. N, number of samples; OC, organic carbon; TN, total N; CEC, cation exchange capacity; [Fe.sub.O], acid oxalate-extractable Fe; [Fe.sub.d], dithionite-extractable Fe; [Al.sub.O], acid oxalate-extractable Al N OC TN (%) Pinus reforestation (Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 7 42.59 [+ or -] 3.90 1.10 [+ or -] 0.23 A horizons 5 3.95 [+ or -] 0.36 0.36 [+ or -] 0.03 B and C horizons 9 0.86 [+ or -] 0.29 0.13 [+ or -] 0.03 Secondary forest (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 6 42.81 [+ or -] 4.19 1.51 [+ or -] 0.46 A horizons 5 5.27 [+ or -] 0.86 0.43 [+ or -] 0.08 B and C horizons 7 2.05 [+ or -] 1.05 0.20 [+ or -] 0.09 Primary forest (Humic Cambisols) Organic horizons 6 46.65 [+ or -] 1.77 1.68 [+ or -] 0.21 A horizons 5 6.52 [+ or -] 0.88 0.48 [+ or -] 0.06 B and C horizons 15 1.02 [+ or -] 0.65 0.12 [+ or -] 0.05 Cabbage cultivation (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) A horizons 6 4.08 [+ or -] 0.70 0.34 [+ or -] 0.04 B horizons 10 1.51 [+ or -] 0.99 0.16 [+ or -] 0.06 All samples Total 80 12.37 [+ or -] 17.91 0.51 [+ or -] 0.55 N C : N CEC ([cmol.sub.c]/kg) Pinus reforestation (Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 7 41.9 [+ or -] 17.0 9.6 [+ or -] 1.5 A horizons 5 11.0 [+ or -] 0.4 15.4 [+ or -] 6.0 B and C horizons 9 6.6 [+ or -] 0.8 10.3 [+ or -] 3.4 Secondary forest (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 6 30.0 [+ or -] 7.0 9.9 [+ or -] 1.8 A horizons 5 12.4 [+ or -] 0.5 20.9 [+ or -] 8.6 B and C horizons 7 10.4 [+ or -] 1.4 13.7 [+ or -] 7.8 Primary forest (Humic Cambisols) Organic horizons 6 28.2 [+ or -] 4.4 10.5 [+ or -] 3.3 A horizons 5 13.5 [+ or -] 0.5 25.3 [+ or -] 11.4 B and C horizons 15 7.8 [+ or -] 2.4 11.0 [+ or -] 6.0 Cabbage cultivation (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) A horizons 6 11.8 [+ or -] 0.8 18.1 [+ or -] 3.3 B horizons 10 8.4 [+ or -] 2.9 12.6 [+ or -] 4.4 All samples Total 80 15.3 [+ or -] 12.3 13.4 [+ or -] 6.9 N pH [Fe.sub.O] (Ca[Cl.sub.2]) Pinus reforestation (Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 7 4.8 [+ or -] 0.3 A horizons 5 4.8 [+ or -] 0.4 6.2 [+ or -] 0.3 B and C horizons 9 4.6 [+ or -] 0.6 4.2 [+ or -] 1.3 Secondary forest (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 6 5.6 [+ or -] 0.5 A horizons 5 4.6 [+ or -] 0.5 7.8 [+ or -] 1.8 B and C horizons 7 4.3 [+ or -] 0.3 5.8 [+ or -] 1.9 Primary forest (Humic Cambisols) Organic horizons 6 5.1 [+ or -] 0.4 A horizons 5 3.8 [+ or -] 0.1 12.8 [+ or -] 1.4 B and C horizons 15 4.2 [+ or -] 0.2 5.1 [+ or -] 3.7 Cabbage cultivation (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) A horizons 6 4.7 [+ or -] 0.4 7.2 [+ or -] 0.7 B horizons 10 4.4 [+ or -] 0.2 6.5 [+ or -] 1.1 All samples Total 80 4.6 [+ or -] 0.6 6.4 [+ or -] 3.0 N [Fe.sub.d] [Al.sub.O] (mg/kg) Pinus reforestation (Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 7 A horizons 5 24.6 [+ or -] 6.0 3.8 [+ or -] 0.5 B and C horizons 9 29.2 [+ or -] 8.5 2.2 [+ or -] 0.6 Secondary forest (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) Organic horizons 6 A horizons 5 28.7 [+ or -] 0.4 4.5 [+ or -] 0.9 B and C horizons 7 31.4 [+ or -] 2.4 3.1 [+ or -] 1.3 Primary forest (Humic Cambisols) Organic horizons 6 A horizons 5 37.7 [+ or -] 6.7 6.3 [+ or -] 0.7 B and C horizons 15 35.0 [+ or -] 10.9 2.5 [+ or -] 1.5 Cabbage cultivation (Humic Cambisols-Dystric Cambisols-Skeletic Umbrisols-Lithic Leptosols) A horizons 6 22.1 [+ or -] 6.8 4.4 [+ or -] 1.2 B horizons 10 25.0 [+ or -] 9.0 4.4 [+ or -] 1.7 All samples Total 80 29.7 [+ or -] 10.2 3.6 [+ or -] 1.7 Table 2. Correlations between S fractions in the bulk soil samples and the NaOH extracts of representative soil profiles of the four sites (N = 23) NaOH extracts Bulk samples Ester C-bonded S S[O.sub.4]-S Ester S[O.sub.4]-S Pearson correlation 0.771 0.703 P (2-tailed) 0.000 0.000 C-bonded S Pearson correlation 0.490 0.662 P (2-tailed) 0.018 0.001 Organic S Pearson correlation 0.713 0.785 P (2-tailed) 0.000 0.000 NaOH extracts Bulk samples Organic S Total S Ester S[O.sub.4]-S Pearson correlation 0.746 0.850 P (2-tailed) 0.000 0.000 C-bonded S Pearson correlation 0.619 0.592 P (2-tailed) 0.002 0.003 Organic S Pearson correlation 0.781 0.819 P (2-tailed) 0.000 0.000 Table 3. Correlations between the S fractions and solution [sup.13]C-NMR spectral regions of alkaline-extractable organic matter Results of [sup.13]C-NMR spectroscopy on NaOH extracts are taken from Moller et al. (2000), N = 23 [13.sup.C]-NMR S fractions in the spectral regions NaOH extracts Alkyl C Methoxyl C O-alkyl C Ester S[O.sub.4]-S Pearson correlation -0.362 -0.386 0.909 P (2-tailed) 0.098 0.076 0.000 C-bonded S Pearson correlation -0.076 0.189 0.425 P (2-tailed) 0.737 0.4 0.048 Total S Pearson correlation -0.236 -0.082 0.740 P (2-tailed) 0.291 0.718 0.000 [13.sup.C]-NMR S fractions in the spectral regions NaOH extracts Aromatic Phenolic Carbonyl C C C Ester S[O.sub.4]-S Pearson correlation -0.795 -0.788 -0.306 P (2-tailed) 0.000 0.000 0.166 C-bonded S Pearson correlation -0.518 -0.607 0.059 P (2-tailed) 0.013 0.003 0.793 Total S Pearson correlation -0.739 -0.792 -0.121 P (2-tailed) 0.000 0.000 0.590
We thank the Royal Forest Department (RFD) of Thailand for allowing us to carry out the research at the Khun Sathan RFD Station and for the help of their staff in Bangkok and Nan, especially Montree Puthawong. We also would like to thank the Soil Analyses Division at the Land Development Department (LDD) in Bangkok for access to its laboratory facilities and for the help in analysing the samples. The German Research Foundation (DFG) and the German Academic Exchange Service (DAAD) funded this study. We also gratefully acknowledge the important personal and financial support of the International Board of Soil Research and Management (IBSRAM).
Acquaye DF, Kang BT (1987) Sulfur status and forms in some surface soils of Ghana. Soil Science 144, 43-52.
Aspiras RB, Keeney DR, Chesters G (1972) Determination of reduced inorganic sulfur forms as sulfide by zinc-hydrochloric acid distillation. Analytical Letters 5, 425-432.
Autry AR, Fitzgerald JW, Caldwell PR (1990) Sulfur fractions and retention mechanisms in forest soils. Canadian Journal of Forest Research 20, 337-342.
Avery BW, Bascomb CL (1974) `Soil survey laboratory methods.' Soil Survey Technical Monograph No 6. (Rothamsted Experimental Station, Lawes Agricultural Trust: Harpenden)
Bettany JR, Saggar S, Stewart JWB (1980) Comparison of the amounts and forms of sulfur in soil organic matter fractions after 65 years of cultivation. Soil Science Society of America Journal 44, 70-75.
Bettany JR, Stewart JWB, Halstead EH (1973) Sulfur fractions and carbon, nitrogen and sulfur relationships in grassland, forest and associated transitional soils. Soil Science Society of America Journal 37, 915-918.
Bettany JR, Stewart JWB, Saggar S (1979) The nature and forms of sulfur in organic matter fractions of soils selected along an environmental gradient. Soil Science Society of America Journal 43, 981-985.
Biederbeck VO (1978) Soil organic sulfur and fertility. In `Developments in soil science, Vol. 8: Soil organic matter'. (Eds M Schnitzer, SU Khan) pp. 273-310. (Elsevier Science: Amsterdam)
Catroux G, Schnitzer M (1987) Chemical, spectroscopic, and biological characteristics of the organic matter in particle size fractions separated from an Aquoll. Soil Science Society of America Journal 51, 1200-1207.
Couto W, Lathwell DJ, Bouldin DR (1979) Sulfate sorption by two oxisols and an alfisol of the tropics. Soil Science 127, 108-116.
David MB, Mitchell MJ, Nakas JP (1982) Organic and inorganic sulfur constituents of a forest soil and their relationship to microbial activity. Soil Science Society of America Journal 46, 847-852.
Eriksen J, Murphy MD, Schnug E (1998) The soil sulfur cycle. In `Sulfur in agroecosystems'. (Ed. E Schnug) pp. 39-74. (Kluwer Academic Publishers: Dordrecht)
Frund R, Ludemann HD (1989) The quantitative analysis of solution- and CPMAS-C-13 NMR spectra of humic material. The Science of the Total Environment 81/82, 157-168.
Ghani A, McLaren RG, Swift RS (1991) Sulfur mineralization in some New Zealand soils. Biology and Fertility of Soils 11, 68-74.
Ghani A, McLaren RG, Swift RS (1992) Sulfur mineralization and transformations in soils as influenced by addition of carbon, nitrogen and sulfur. Soil Biology and Biochemistry 24, 331-341.
Gressel N, McColl JG, Preston CM, Newman RH, Powers RF (1996) Linkages between phosphorus transformations and carbon decomposition in a forest soil. Biogeochemistry 33, 97-123.
Haynes RJ, Williams PH (1992) Accumulation of soil organic matter and the forms, mineralization potential and plant-availability of accumulated organic sulfur: Effect of pasture improvement and intensive cultivation. Soil Biology and Biochemistry 24, 209-217.
Houle D, Carignan R (1992) Sulfur specification and distribution in soils and aboveground biomass of a boreal coniferous forest. Biogeochemistry 16, 63-82.
Johnson DW, Henderson GS (1979) Sulfate adsorption and sulfur fractions in a highly weathered soil under a mixed deciduous forest. Soil Science 128, 34-40.
Keer JI, McLaren RG, Swift RS (1990) Acetylacetone extraction of soil organic sulfur and fractionation using gel chromatography. Soil Biology and Biochemistry 22, 97-104.
Kjellberg A, Widmalm G, Jansson PE, Nimmich W (1995) Structural studies of the capsular polysaccharide from Klebsiella Type-7. Carbohydrate Research 273, 53-62.
Kogel-Knabner I (1997) [sup.13]C and [sup.15N]-NMR spectroscopy as a tool in soil organic matter studies. Geoderma 80, 243-270.
Kowalenko CG (1978) Organic nitrogen, phosphorus and sulfur in soils. In `Developments in soil science, Vol. 8: Soil organic matter'. (Eds M Schnitzer, SU Khan) pp. 95-126. (Elsevier Science: Amsterdam)
Kowalenko CG (1993a) Extraction of available sulfur. In `Soil sampling and methods of analysis'. (Ed. MR Carter) pp. 65-74. (Lewis Publishers: Boca Raton, FL)
Kowalenko CG (1993b) Total and fractions of sulfur. In `Soil sampling and methods of analysis'. (Ed. MR Carter) pp. 231-246. (Lewis Publishers: Boca Raton, FL)
Lambert M J, Turner J (1998) Sulfur nutrition and cycling in southern hemisphere temperate and subtropical forest ecosystems. In `Sulfur in the environment'. (Ed. DG Maynard) pp. 263-294. (Marcel Dekker, Inc.: New York)
Liao ML, Chiovitti A, Muro SLA, Craik DJ, Kraft GT, Basic A (1996) Sulfated galactans from Australian specimens of the red alga Phacelocarpus peperocarpos (Gigartinales, Rhodophyta). Carbohydrate Research 296, 237-247.
Lowe LE, DeLong WA (1963) Carbon-bonded sulfur in selected Quebec soils. Canadian Journal of Soil Science 43, 151-155.
McLachlan KD, De Marco DG (1975) Changes in soil organic sulfur fractions with fertiliser additions and cropping treatments. Australian Journal of Soil Research 13, 169-176.
McLaren RG, Swift RS (1977) Changes in soil organic sulfur fractions due to the long term cultivation of soils. Journal of Soil Science 28, 445-453.
Mehra OP, Jackson ML (1960) Iron oxide removal from soils and clays by a dithionite-citrate system buffered with sodium bicarbonate. Clays and Clay Minerals 7, 317-327.
Miller IJ, Blunt JW (2000) New C-13 NMR methods for determining the structure of algal polysaccharides. Part 1-3. Botanica Marina 43,239-271.
Moller A, Kaiser K, Amelung W, Niamskul C, Udomsri S, Puthawong M, Haumaier L, Zech W (2000) Forms of organic C and P extracted from tropical soils as assessed by liquid-state [sup.13]C- and [sup.31]p-NMR spectroscopy. Australian Journal of Soil Research 38, 1017-1035.
Motavalli PP, Duxbury JM, de Souza DMG (1993) The influence of organic soil amendments on sulfate adsorption and sulfur availability in a Brazilian Oxisol. Plant and Soil 154, 301-308.
Neptune AML, Tabatabai MA, Hanway JJ (1975) Sulfur fractions and carbon-nitrogen-phosphorus-sulfur relationship in some Brazilian and Iowa soils. Soil Science Society of America Proceedings 39, 51-55.
Olson BM, Lowe LE (1990) Effects of intensive vegetable production on chemical properties and nutrient mineralization of a British Columbia Humisol. Canadian Journal of Soil Science 70, 445-460.
Saggar S, Hedley M J, Phimsarn S (1998) Dynamics of sulfur transformations in grazed pastures. In `Sulfur in the environment'. (Ed. DG Maynard) pp. 45-94. (Marcel Dekker, Inc.: New York)
Schnitzer M (1982) Organic matter characterization. In `Methods of soil analysis, Part 2--Microbiological and biochemical properties'. (Eds AL Page et al.) pp. 581-594. Agronomy Monograph 9 (American Society of Agronomy, Inc.: Madison, WI)
Schwertmann U (1964) Differenzierung der Eisenoxide des Bodens durch Extraktion mit saurer Ammoniumoxalat-Losung. ZeitschrifT fur Pflanzenernahrung, Dungung und Bodenkunde 105, 194-202.
Solomon D, Lehmann J, Mamo T, Fritzsche F, Zech W (2001) Sulfur fractions in particle-size separates of the sub-humid Ethiopian highlands as influenced by land use changes. Geoderma 102, 41-59.
Stanko-Golden KM, Fitzgerald JW (1991) Sulfur transformation and pool size in tropical forest soils. Soil Biology and Biochemistry 23, 1053-1058.
Stevenson FJ, Cole MA (1999) `Cycles of soil. Carbon, nitrogen, phosphorus, sulfur, micronutrients.' (John Wiley and Sons, Inc.: New York)
Strickland TC, Fitzgerald WJ, Ash JT, Swank WT (1987) Organic sulfur transformations and sulfur pool sizes in soil and litter from a southern Appalachian hardwood forest. Soil Science 143, 453-458.
Sumann M, Amelung W, Haumaier L, Zech W (1998) Climatic effects on soil organic phosphorus in the North American Great Plains identified by phosphorus-31 nuclear magnetic resonance. Soil Science Society of America Journal 62, 1580-1586.
Tabatabai MA (1996) Sulfur. In `Methods of soil analysis, Part 3-Chemical methods'. (Ed. DL Sparks) pp. 921-960. (Soil Science Society of America, Inc./American Society of Agronomy, Inc.: Madison, WI)
Tabatabai MA, Bremner JM (1972) Forms of sulfur, and carbon, nitrogen and sulfur relationships in Iowa soils. Soil Science 114, 380-386.
Zech W, Guggenberger G, Haumaier L, Pohhacker R, Schafer D, Amelung W, Miltner A, Kaiser K, Ziegler F (1996) Organic matter dynamics in forest soils of temperate and tropical ecosystems. In `Humic substances in terrestrial ecosystems'. (Ed. A Piccolo) pp. 101-170. (Elsevier Science: Amsterdam)
Zhang Y, Mitchell MJ, Driscoll CT, Likens GE (1999) Changes in soil sulfur constituents in a forested watershed 8 years after whole-tree harvesting. Canadian Journal of Forest Research 29, 356-364.
Zucker A, Zech W (1985) Sulfur status of four uncultivated soil profiles in northern Bavaria. Geoderma 36, 229-240.
Manuscript received 2 January 2001, accepted 15 June 2001
A. Moller (AF), K. Kaiser (A), N. Kanchanakool (B), C. Anecksamphant (B), W. Jirasuktaveekul (C), A. Maglinao (D), C. Niamskul (E), and W. Zech (A)
(A) Institute of Soil Science and Soil Geography, University of Bayreuth, 95440 Bayreuth, Germany.
(B) Land Development Department, Bangkok 10900, Thailand.
(C) Royal Forest Department, Bangkok 10900, Thailand.
(D) International Board for Soil Research and Management, Bangkok 10900, Thailand.
(E) UI-Consulting, Bangkok 10120, Thailand.
(F) Corresponding author; email: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Author:||Moller, A.; Kaiser, K.; Kanchanakool, N.; Anecksamphant, C.; Jirasuktaveekul, W.; Maglinao, A.; Niam|
|Publication:||Australian Journal of Soil Research|
|Article Type:||Statistical Data Included|
|Date:||Jan 1, 2002|
|Previous Article:||Mineralisation of organic matter in intact versus sieved/refilled soil cores.|
|Next Article:||Measuring peat moisture content using the dual-probe heat pulse technique.|