Directory UMM :Data Elmu:jurnal:S:Soil Biology And Chemistry:Vol32.Issue10.Sep2000:

(1)

Organic geochemical studies of soils from the Rothamsted

classical experiments Ð VI. The occurrence and source of

organic acids in an experimental grassland soil

Ian D. Bull

a

, Chris J. Nott

a

, Pim F. van Bergen

a, 1

, Paul R. Poulton

b

,

Richard P. Evershed

a,

*

a_{School of Chemistry, University of Bristol, Cantock's Close, Bristol BS8 1TS, UK} b

Soil Science Department, IACR-Rothamsted, Harpenden, Herts AL5 2JQ, UK

Received 23 July 1999; received in revised form 3 February 2000; accepted 23 February 2000

Abstract

Total lipid extracts (TLEs) of grass (aerial and sub-aerial,Holcus lanatus) from a plot on a long-term grassland experiment, and associated soil, along with the organic fraction of the TLE hydrolysates and the hydrolysates of the solvent extracted vegetation have been separated into fractions containing speci®c compound classes and analysed using gas chromatography (GC) and gas chromatography/mass spectrometry (GC/MS). The distributions of n-alkylcarboxylic acids, o-hydroxycarboxylic acids and dicarboxylic acids in the grass and the underlying soil have been determined. Short-chain (<C20) n-alkylcarboxylic acids were designated as having derived from both aerial and sub-aerial vegetation. However, longer-chain n-alkylcarboxylic acids were ascribed to suberin as a predominant source. Moreover,o-hydroxycarboxylic acids and dicarboxylic acids observed in the soil were designated as having predominantly derived from inputs of free, extractable polyesters and suberin intimately associated with plant roots. This study indicates the importance of root material as a predominant source of aliphatic, organic acids in the soil of temperate grassland biomes.72000 Elsevier Science Ltd. All rights reserved.

Keywords:Acid; Gas chromatography; Mass spectrometry; Grassland; Hydroxyl; Root; Rothamsted; Suberin

1. Introduction

In previous studies of the Rothamsted Classical Ex-periments, we have become increasingly aware of sig-ni®cant quantities of refractory, difunctionalised carboxylic acid components present in soil (van Bergen et al., 1998; Bull et al., in press). For example, in a preliminary study of the eect of soil pH on organic matter decay, the occurrence of C22 and C24 o

-hydro-xycarboxylic acids was tentatively ascribed to an input of suberin from overlying plant roots. However, the available data did not preclude inputs from other sources (e.g. cutin) of o-hydroxycarboxylic acids (van Bergen et al., 1998). In light of this, we have made a more detailed study of the organic acid components found in the vegetation and soil constituting a plot of the grass Holcus lanatuson the Park Grass experiment at Rothamsted Experimental Station.

Cutin is a macromolecule composed predominantly of polyesteri®ed hydroxy fatty acids and is intimately associated with the majority of aerial plant tissue. It is found in both the interior and exterior layers of plant cuticles and, along with waxes, as part of an overlying lamellae, which supports a protective epicuticular wax layer (Tulloch, 1976; Walton, 1990 and references

www.elsevier.com/locate/soilbio

1

Present address: Organic Geochemistry Group, Geochemistry, Faculty of Earth Sciences, Utrecht University, P.O. Box 80021, 3508 TA Utrecht, The Netherlands.

* Corresponding author. Tel.: 9287671; fax: +44-0-117-9291295.

(2)

therein). In cases where secondary growth of plant tis-sue occurs, e.g. bark, woody stems and root material, external protection is provided by a complex of an analogous biopolyester, suberin, laired between the plasmalemma and cell wall in the cork cells (phellum) of the periderm (Walton, 1990). Cutin is structurally similar to suberin but there are still a number of dier-ences in monomeric and tertiary structure (Tegelaar et al., 1995). Early phytochemical studies of cutin and suberin have yielded much information regarding the monomeric composition of these two polyesters (e.g. Eglinton and Hunneman, 1968; Holloway and Deas, 1971; Hunneman and Eglinton, 1972; Holloway, 1982, 1984). The predominant monomer constituents in cutin are normal n-alkylcarboxylic acids and C16/C18 hydroxycarboxylic acids. Suberin, whilst exhibiting a somewhat similar content of n-alkyl and mid-chain hydroxycarboxylic acid components, contains a much higher abundance of long-chain o-hydroxycarboxylic acids and dicarboxylic acids (Kollattukudy et al., 1976).

Using a suite of analytical techniques, KoÈgel-Knab-ner et al. (1992a) found the contribution of cutin to forest soil organic matter to be `relatively important when compared with the possible contributions of non-saponi®able aliphatic biomacromolecules such as cutan and suberan'. Humi®cation was concluded to proceed with an increase in cross-linking of cutin and/ or suberin polyesters. Furthermore, in a study of humic acid and humin fractions of an acid soil, (Grasset and AmbleÁs, 1998) were almost able to com-pletely solubilise (87%) initially insoluble humin using a stepwise alkaline hydrolysis under phase transfer conditions. The subsequent production of short and long-chain dicarboxylic acids, monocarboxylic acids and alkanols was interpreted to indicate a macromol-ecular matrix with esteri®ed dicarboxylic acid cross-links. Alkyl carboxylic acids, alcohols and aromatic acids were designated as monosubstituents esteri®ed to the matrix. This emphasises the importance of esteri®-cation as a mode of bonding for organic biopolymers during early diagenesis in terrestrial soils. Since trans-esteri®cation processes have been shown to be active in soil, it follows that such processes must also have a signi®cant eect on the composition of free lipids in the soil environment (AmbleÁs et al., 1994).

Majority of studies examining cutin and/or suberin incorporation into soil organic matter have been con-cerned with forest soil systems (e.g. KoÈgel-Knabner et al., 1992a, 1992b; Riederer et al., 1993; Nierop, 1998). Examples of similar studies in alternative biomes are scarce, although (Schulten et al., 1992 and Saiz Jime-nez et al. (1996)) are two prominent exceptions; both of which make use of pyrolytic techniques of analysis.

This study tests the hypothesis that in a temperate grassland biome whilst both cutin and suberin

mono-mers are sources of organic acids in soil, the latter polyester is by far the major contributor. The site cho-sen to carry out this investigation was a plot on the Park Grass experiment, a long-term grassland exper-iment, located at Rothamsted Experimental Station, Harpenden, Hertfordshire, UK. Being a long-term (>143 years), controlled experiment, this site has a recorded history of inputs (both organic and inor-ganic) and environmental conditions, which enables conclusions to be drawn as to the pedogenic fate of or-ganic matter with some degree of certainty.

2. Sample description

2.1. Park Grass Experiment

The Park Grass Experiment is located on an area of grassland at Rothamsted Experimental Station, Har-penden, Herts, UK. The experiment was laid down in 1856, the ®eld being under pasture for at least a cen-tury before this (Tilman et al., 1994). Control plots show the characteristic mixed plant population of old grassland. The boundaries of the plots are clearly de®ned with the transition between adjacent treatments occupying 30 cm or less. This indicates little sideways movement of nutrients in a ¯at undisturbed soil. At the start of the experiment, the site had received no regular dressings of calcium carbonate and the pH was in the region of 5.7. A test of liming began in 1903, was modi®ed in 1965, and the plots now range in pH from 7.3 to 3.5 depending on the treatment (Johnston et al., 1986).

The soil is a Stagnogleyic paleo-argillic brown earth, classi®ed as Chromic Luvisol (FAO, 1990) or Aquic Paleudalf (USDA, 1992). The mechanical composition of the soil is approximately 19% sand (60 mm±2 mm), 58% silt (2±60 mm) and 23% clay (<2 mm) (Avery and Catt, 1995). The clay fraction is mainly composed of interstrati®ed expanding-layer silicates containing smectite and other layers, with subsidiary mica and kaolinite and small amounts of feldspar, chlorite and crystalline (goethite) or amorphous ferric oxides (Avery and Catt, 1995).

Soils were sampled in May 1995 and 1996 using a 2

Table 1

Characteristics of the samples analysed in this study

Sample 10ÿ3 _Sub-sample _Depth/cm _{% C} _{% N} _C:N

Holcus lanatus Aerial tissue n/a 39.5 4.2 9.4

Root tissue 0±5 43.0 2.2 19.5

Humic horizon Rhizosphere soil 0±5 17.1 1.2 14.3

Humic soil 0±5 23.6 1.6 14.8

(3)

cm diameter auger to a depth of 23 cm and are sum-marised, along with elemental data in Table 1. Samples were collected from a plot (designated 11/1 d) domi-nated by the grass Holcus lanatus. Ecological studies have noted a drastic decrease in the diversity of over-lying plant species for this plot, as the acidity has increased with time. Representative samples taken as far back as 1862 included Agrostis capillaries, Alope-curus pratensis,Arrhenatherum elatius,Dactylis glomer-ata, Festuca rubra, Holcus lanatus and Rumex acetosa

as major species, however, by 1950, the plot supported a virtual monoculture ofHolcus lanatus(Tilman et al., 1994). The soil pH for this plot is low and has remained relatively constant over the past 76 years [1923, (3.8); 1959, (3.7); 1976, (3.7); 1984, (3.5)]. In-homogeneity of the soil and any associated roots was compensated for by taking a number of cores from random positions on the plot and combining the ma-terial thus sampled. Soils were subdivided into a dark brown top layer, (ca. 5 cm) and a bottom layer, desig-nated mineral horizon, (ca. 18 cm). The dark brown top-layer was sub-divided into loose lower soil, desig-nated humic horizonand that soil intimately associated with root matter, designated rhizosphere soil. Holcus

lanatus was sampled from the plot and divided into

aerial vegetation and roots.

3. Experimental

3.1. Sample preparation and solvent extraction

Fresh soil and vegetation samples were initially oven dried at 608C. All the soil samples were crushed with a pestle and mortar and subsequently sieved over 2 mm and 75mm sieves, pH was measured in H2O (soil:H2O 1:2.5 w/v). Dried vegetation was crushed using the same method but with the addition of liquid nitrogen to facilitate the process. This was then sieved over 5 and 2 mm sieves.

All the samples (ca. 0.2 g vegetation, 3 g humic soil, 10 g mineral soil) were Soxhlet extracted for 24 h using 200 ml dichloromethane (DCM)/acetone (9:1 v/ v) to obtain a total lipid extract (TLE). Heptadecanoic acid (180±250 mg gÿ1for vegetation samples and 1±10 mg gÿ1for soil samples) was added as an internal stan-dard. Solvent was removed under reduced pressure. Solvent extracted vegetation residues (ca. 0.015 g) were saponi®ed for 2 h using 3 ml of a 0.5 M methanolic NaOH solution. After cooling and acidi®cation (pH 1), 1 ml of double distilled water was added and free lipids extracted into 3 2 ml DCM. Extracts were combined and passed through an anhydrous MgSO4 column to remove any residual water; solvent was removed under a gentle stream of nitrogen. The repro-ducibility of this extraction method has been tested

previously and found to provide extracts of consistent composition for the same samples (Bull, 1997).

3.2. Fractionation

TLEs were separated into two fractions, `acid' and `neutral', using an extraction cartridge with a bonded aminopropyl solid-phase (500 mg sorbent, 2.8 ml elu-ent capacity; Varian). Extracts dissolved in DCM/iso-propanol (2:1 v/v) were slowly ¯ushed through a cartridge pre-eluted with hexane. After further elution with DCM/isopropanol (2:1 v/v, 8 ml), the collected `neutral' fraction was removed and the cartridge slowly ¯ushed with 2% v/v acetic acid in diethylether (8 ml) thereby eluting the `acid' fraction. Solvent was removed from both the fractions under a gentle stream of nitrogen. Acid fractions were derivatised and ana-lysed using GC and GC/MS techniques.

3.3. Derivatisation

Fractions were methylated by heating with 200 ml of a 14% w/v borontri¯uoride-methanol solution (Sigma) for 10 min at 708C. Residual complex was destroyed by the addition of 1 ml double distilled water and methyl esters were extracted in 3 1 ml hexane. Extracts were combined and passed through an anhy-drous MgSO4 column to remove any residual water. Solvent was removed under a gentle stream of nitro-gen. Free hydroxyl functional groups were derivatised to their respective trimethylsilyl (TMS) ethers by add-ing 30 ml of N,O-bis(trimethylsilyl)tri¯uoroacetamide (BSTFA; Sigma), containing 1% trimethylchlorosilane (TMCS), to sample aliquots and heating for 30 min at 708C. Excess derivatising agent was removed under a gentle stream of nitrogen and samples redissolved in 50ml hexane.

3.4. Gas chromatography (GC)

Organic acids were analysed using a Hewlett-Pack-ard 5890 series II gas chromatograph equipped with a fused-silica capillary column (Chrompack CPSil-5CB, 50 m length0.32 mm i.d.0.12 mm ®lm thickness). Derivatised fractions in hexane were injected (1.0 ml) on-column. The temperature was programmed from 408C (1 min isotherm) to 2008C at a rate of 108C minÿ1 and ®nally to 3008C (20 min isotherm) at 38C minÿ1. The detector temperature was kept at 3208C. Hydrogen was used as carrier gas (10 psi head press-ure).

3.5. Gas chromatography-mass spectrometry (GC/MS)

(4)

capillary column (Chrompack CPSil-5CB, 50 m length

0.32 mm i.d.0.12 mm ®lm thickness) and the tem-perature was programmed from 408C (1 min isotherm) to 2008C at a rate of 108C minÿ1 and ®nally to 3008C (20 min isotherm) at 38C minÿ1; helium was used as carrier gas (10 psi head pressure)] equipped with on-column injection coupled, via a heated transfer line (3208C), to a Finnigan MAT 4500 quadrupole mass spectrometer scanning in the range ofm/z50±650 with a cycle time of 1.0 s. The current was maintained at

300 mA with an ion source temperature of 1908C and an electron voltage of 70 eV.

4. Results

The compounds observed in the `acid' fraction will be assessed initially as three functional classes, namely:

n-alkyl carboxylic acids, o-hydroxycarboxylic acids and dicarboxylic acids.

(5)

4.1. n-alkylcarboxylic acids

Fig. 1 summarises the distributions of n -alkylcar-boxylic acids derived from each of the samples studied. The TLE from the aerial section of Holcus lanatus

exhibits a homologous series ofn-alkylcarboxylic acids ranging from C13 to C30 maximising at C16, C22 and C28 (Fig. 1a). Base hydrolysis of the TLE results in a slightly wider distribution (C10±C30) and a marked increase in the abundance of C16, C18, C22 and C24 n

-alkylcarboxylic acids (Fig. 1b). n-alkylcarboxylic acids obtained by saponi®cation of the solvent extracted aerial vegetation range from C12 to C28 (Fig. 1c). Whilst the C16 and C22 homologues remain abundant components, the distribution is dominated by linoleic acid (C18:2D9,12).

Compared with the aerial TLE that of the root exhi-bits an almost identical distribution of n -alkylcar-boxylic acids albeit at about a tenth of the abundance and over a slightly wider range (C12±C32; Fig. 1d).

(6)

Whilst the hydrolysate of the root TLE yields a higher abundance ofn-alkylcarboxylic acid components (C12± C28), maximising at C18 and C22, it represents only about half the quantity present in the corresponding aerial sample (Fig. 1e). The highest abundance of root derived n-alkylcarboxylic acid is observed to occur in the hydrolysate of the solvent extracted root residue (Fig. 1f). Homologues range from C12 to C30and exhi-bit a similar abundance to the corresponding aerial sample, although the predominant linoleic acid max-ima does not occur. Overall, the total quantity of n -alkylcarboxylic acids derived from the residues, both aerial and sub-aerial, is greater than that obtained from the corresponding TLEs of the identical samples.

The greatest observable dierence between the n -alkylcarboxylic acid components ofHolcus lanatusand those derived from soil TLEs is the higher abundance of the longer homologues (rC22), which dominate each of the three soil distributions (Figs. 1g, h & i). Both the rhizosphere and humic horizon TLEs yield trimodal distributions maximising at C16, C22 and C28. Whilst that of the mineral horizon is bimodal, maxi-mising at C16 and C26. Overall, the rhizosphere and mineral horizon contain a comparable quantity of free, extractable n-alkanoic acids, which is higher than the quantity observed to occur in the humic soil horizon.

4.2.o-hydroxycarboxylic acids

Fig. 2 summarises the distributions ofo -hydroxycar-boxylic acids obtained from each of the samples stu-died. Homologues in the C16±C24 range are observed in the free extractable TLE of aerialHolcus lanatus tis-sue and are centred about a dominant C22 component (Fig. 2a). Saponi®cation of the TLE produces no real change in the distribution of o-hydroxycarboxylic acids but absolute amounts are essentially doubled (Fig. 2b). The hydrolysate of the solvent extracted, aerial Holcus lanatus tissue contained the same range (C16±C24) of o-hydroxycarboxylic acid components as the TLE but at a much lower abundance with a signi®-cantly less dominant C22 component (Fig. 2c); the most abundant hydroxylated acid was 18-hydroxyocta-dec-9-enoic acid (not shown).

Unbound o-hydroxycarboxylic acids present in the TLE of the Holcus lanatus roots are only detectable over the range of C20±C24 with an unpronounced maximum at C22 (Fig. 2d). A drastic change in abun-dance may be observed in homologues obtained by the saponi®ed TLE producing a wider range of detectable compounds (C16±C24) and a more pronounced maxi-mum at C22(Fig. 2e). The largest quantity ofo -hydro-xycarboxylic acids are obtained from the hydrolysate of the solvent extracted root material ofHolcus lanatus

(Fig. 2f). The C22 component can be observed to pre-dominate with peripheral homologues ranging from

C16 to C24 albeit at a higher relative abundance than peripheral components in the other samples. As observed for the hydrolysate of the aerial residue, 18-hydroxyoctadec-9-enoic acid is also present as a major component (not shown).

The TLE extracted from the rhizosphere soil con-tains o-hydroxycarboxylic acids in a higher, narrower range (C20±C26), than what is observed for the ma-jority of vegetation samples (Fig. 2g). The distribution is dominated by the C22 and C24 components, of which, the C22 homologue is the maximum. The same distribution can be observed for the humic horizon TLE, although at about half of the abundance and a small quantity of the C16 homologue is present (Fig. 2h). The o-hydroxycarboxylic acids derived from the mineral horizon TLE range C16±C26 maximising about the C24 homologue. Interestingly, 18-hydroxyoc-tadec-9-enoic acid was not detected in any of the soil samples (not shown).

4.3. Dicarboxylic acids

Fig. 3 depicts the distributions of dicarboxylic acids detected in each of the sample extracts analysed. Of the above-ground vegetation, only the hydrolysed TLE yields any dicarboxylic acid components; a single C22 homologue (0150 mg gTOCÿ1 ; Fig. 3b). The TLE of the root also contains no free dicarboxylic acid com-ponents although there is a small quantity (025 mg gTOCÿ1 ) of the C22 homologue present in the hydrolysed TLE (Fig. 3e). Saponi®cation of the solvent extracted root residue yields a much higher abundance of the C22homologue (0420 mg gTOCÿ1 ) along with a less abun-dant C24component (Fig. 3f).

In each of the three soil TLEs both C22 and C24 dicarboxylic acids are observed to occur albeit at abundances less than 150 mg g_TOCÿ1 (Fig. 3g, h and i). The rhizosphere and humic horizon exhibit an unpro-nounced C22 maximum, whilst the mineral soil has a C24maximum.

5. Discussion

Constituent components observed in the aerial and sub-aerial vegetation shall be discussed and then their subsequent contribution to the lipid content in the underlying soil considered.

5.1. n-alkylcarboxylic acids

(7)

residue, it being an internal component commonly found in higher plant tissue (Fig. 1c; Gunstone et al., 1986). The signi®cant observation that a far greater proportion of extractable components are found in the aerial tissue (Fig. 1a and b) compared with the roots (Fig. 1d and e) is consistent with epicuticular waxes being associated with those parts of the plant exposed to the atmosphere (Eglinton et al., 1967).

The distributions derived from the soil result from an admixture of multisourced components (Fig. 1g, h and i). In the rhizosphere and humic horizon direct

in-corporation of free and hydrolysed n-alkylcarboxylic acids is most likely an important source indicated by the relatively high abundance of C22 and C24 homol-ogues. Another possible source is the oxidation of pri-mary alcohols (AmbleÁs et al., 1994), which constitute the major lipid components in the TLE of the aerial vegetation of Holcus lanatus (van Bergen et al., 1998). Odd n-alkylcarboxylic acid components, are also likely to comprise a component derived from the oxidative degradation of n-alkanes, via methyl ketones, in the soil pro®le (Ambles et al., 1993). Due to the

(8)

sourced nature of these components, their use in the correlation of organic matter inputs is best avoided.

5.2.o-hydroxycarboxylic acids

The two-fold increase in abundance of o -hydroxy-carboxylic acids between the aerial vegetation TLE (Fig. 2a) and its hydrolysate (Fig. 2b) indicates that whilst a signi®cant proportion ofo-hydroxycarboxylic acids exist as freely extractable components nearly the same quantity exists in the form of extractable, neutral complexes. A likely source of these saponi®able o -hydroxycarboxylic acids complexes is neutral, cyclic polyesters. Such polyesters, or estolides, of 14-hydro-xytetradecanoic acid have previously been reported to occur in the wax of leaves of Picea pungens(Colorado spruce; von Rudlo, 1959). Interestingly, the even greater disparity between the abundance of homol-ogues in the root TLE (Fig. 1d) and its hydrolysate (Fig. 1e) reveals that in the sub-aerial Holcus lanatus

tissue the majority of solvent extractable o -hydroxy-carboxylic acids exist as neutral, esteri®ed oligomers/ polymers paralleling, but to a greater extent, the com-partmentalisation of these components in the aerial vegetation. The dominant occurrence of 18-hydroxyoc-tadec-9-enoic acid (2544mg g_TOCÿ1 in the hydrolysate of the aerial vegetation residue agrees with its origin as a common constituent of cutin (Fig. 1c; Kollattukudy et al., 1976). However, the appreciable quantity of this component (1943 mg g_TOCÿ1 observed in the hydrolysate of the root residue indicates an analogous origin in other polyesters, i.e. suberin (Fig. 1f). Overall, the high quantity of o-hydroxycarboxylic acids, which occur in this sample, relative to others, is in concordance with that being previously reported as major components of suberin associated with, amongst other things, plant roots (Kollattukudy et al., 1976; Matzke and Riederer, 1991).

The most prominant dierence between the distri-butions of o-hydroxycarboxylic acids in the soil TLEs and the various samples from Holcus lanatus is the almost total loss of homologuesRC20 and a preferen-tial loss of the C22 component relative to its peripheral homologues (Fig. 2i). Previous work has concluded that organic matter in the humic horizon (and there-fore also the rhizosphere) is dominated by an input from the current vegetation, i.e. Holcus lanatus, hence these dierences are unlikely to derive from another primary input (van Bergen et al., 1998; Bull et al., in press). The retention of a C22 o-hydroxycarboxylic acid maximum in the humic horizon lends further weight to this conclusion. Whilst abiotic processes, such as mineral induced esteri®cation, have been shown to operate in soils, e.g. Jambu et al. (1995), losses are either independent or linearly related to chain length. The preferential loss of the C22

com-ponent is therefore most likely to be a consequence of biotic degradation, which in an oxidative environment will be the major route of loss of lipid components (Heal et al., 1997). Oxidation is accepted as proceeding principally via a b-oxidative pathway which, if adher-ing to ®rst order kinetics, would retain the relative dis-tribution of longer homologues (Minderman, 1968; Voet and Voet, 1995). Physical eects such as water solubility may well increase the rate of loss of more soluble short chain components. One tentative expla-nation for the enhanced degradation of the C22 com-ponent could be an attenuation of the biological soil system to micro-organisms, which preferentially degrade the C22 component, although this has yet to be fully investigated. The non-occurrence of 18-hydro-xyoctadec-9-enoic acid in any of the soil samples is most likely the result of rapid microbial b-oxidation of this shorter component coupled with microbial oxi-dation and/or autoxioxi-dation of the labile double bond (Mlaker and Spiteller, 1996; Frankel, 1998).

5.3. Dicarboxylic acids

Whilst a C22 dicarboxylic acid component does occur in the hydrolysate of the aerial vegetation TLE (Fig. 3b), presumably occurring naturally as a neutral polyester, it is the hydrolysate of the root residue, which contains the greatest abundance of dicarboxylic acids (Fig. 3f). Clearly, the aliphatic acids in Holcus

lanatus are used predominantly as components of

polyesters associated with root material, i.e. suberin; this agrees with the previous work (Kolattukudy et al., 1976; Tulloch, 1976).

In parallel with the observations made about the o -hydroxycarboxylic acids the most signi®cant dierence between soil and vegetation dicarboxylic acids is the loss of the dominant component relative to peripheral homologues. Whilst dicarboxylic acids may well be formed from further oxidation of o-hydroxycarboxylic acids (cf. AmbleÁs et al., 1994), the similarity in chain length of the dominant components will cause little distributional alteration of dicarboxylic acid com-ponents. There is currently no de®nitive explanation of this phenomenon. The occurrence of organic matter from previous vegetation cover in the mineral soil is further con®rmed by a shift in distribution observed to occur between it and the overlying humic horizon and rhizosphere.

5.4. Aerial versus root inputs of organic acids

Since all of the organic acid components reported above occur in both the aerial and root tissue of

Holcus lanatus, the relative contribution of these

(9)

that aerial vegetation contains a greater total con-centration of n-alkyl carboxylic acids than root ma-terial although higher homologues (rC20) are present at about the same concentration. Concentrations ofo -hydroxycarboxylic acids in above- and below-ground tissue are about equal, whilst a greater proportion of dicarboxylic acids may be found in the root tissue. Hence, data concerning the relative input of aerial and subaerial bulk organic matter derived from Holcus

lanatus is required in order to ascertain the primary

source of each organic acid type. Whilst there exist many early studies detailing biomass and net primary production (NPP) estimated for vegetation in dierent biomes, similar studies investigating the aerial:sub-aerial spilt of these variables are scarce. In a study of root turnover and productivity in a coniferous forest (Fogel, 1983) stated that of the organic plant derived residues which enter the soil, between 60 and 70% come from the root system. Furthermore, such detritus or `rhizo-deposits' will vary between dierent plants and dierent biomes depending upon general dier-ences in the structure of the plant sub-system and en-vironmental constraints, respectively. For example, whilst the mean annual NPP of a boreal forest and a temperate grassland are about the same (07.5 t haÿ1 yearÿ1), the predominant amount of NPP in the forest (71%) is diverted to perennial storage in woody tissue, whilst in temperate grasslands a greater percentage (83%) exists as root matter (Swift et al., 1979). Hence, we can say, with some certainty, that in temperate grasslands, whilst inputs of short-chain (<C20) n-alkyl carboxylic acids are largely derived from aerial and sub-aerial tissue, inputs of long-chain n-akylcarboxylic acids, o-hydroxycarboxylic acids and dicarboxylic acids are derived primarily from the root tissue. For the Park Grass experiment, this is certainly the case since the entire experiment is cut from mid-June and made into hay thereby drastically reducing the level of aerial NPP, which is returned to the soil.

In conclusion, this study was instigated to test the hypothesis that in a temperate grassland biome whilst organic acids from both cutin and suberin are inputs to soil organic matter, the latter polyester is by far the major contributor. The most signi®cant ®ndings, which arose from the results of this study were:

1. Short chain n-alkylcarboxylic acids (<C20) in the soils are derived from aerial and sub-aerial veg-etation as both free lipids and polyesteri®ed com-ponents. Higher homologues (rC20) are increasingly derived from free and bound components in sub-erin.

2.o-hydroxycarboxylic acids and dicarboxylic acids in the soils are predominantly derived from the free extractable polyesters and suberin intimately associ-ated with overlying plant roots.

3. The mineral horizon TLE still retains organic mat-ter derived from inputs of previous vegetation cover (i.e. not Holcus lanatus) or other soil dwelling organisms (e.g. micro-organisms).

4. The primary acid components of each vegetation extract lose their dominance relative to peripheral homologues by an as yet unexplained mechanism.

Whilst points 1 and 2 are valid for a temperate grassland, they cannot necessarily be applied directly to other settings. Further study of these components in other biomes is still necessary to fully assess the origin and fate of aliphatic, organic acids in terrestrial soils.

Acknowledgements

This project was undertaken whilst the authors were in receipt of a NERC Grant GR3/9578 to RPE. Jim Carter and Andy Gledhill are thanked for their help with the GC/MS analyses. The use of the NERC Mass Spectrometry Facilities (Grants GR3/2951, GR3/3758, FG6/36/01) is gratefully acknowledged. IACR receives grant-aided support from the BBSRC.

References

AmbleÁs, A., Jambu, P., Jacquesy, J-C., Parlanti, E., Secouet, B., 1993. Changes in the ketone portion of lipid components during the decomposition of plant debris in a hydromorphic forest-pod-zol. Soil Science 156, 49±56.

AmbleÁs, A., Jambu, P., Parlanti, E., Jore, J., Rie, C., 1994. Incorporation of natural monoacids from plant residues into an hydromorphic forest podzol. European Journal of Soil Science 45, 175±182.

Avery, B.W., Catt, J.A., 1995. The soil at Rothamsted. Lawes Agricultural Trust, IACR-Rothamsted, Harpenden.

van Bergen, P.F., Nott, C.J., Bull, I.D., Poulton, P.R., Evershed, R.P., 1998. Organic geochemical studies of soils from the Rothamsted Classical Experiments. IV Ð Preliminary results from a study of the eect of soil pH on organic matter decay. Organic Geochemistry 29, 1779±1795.

Bull, I.D., 1997. New molecular methods for tracing natural and anthropogenic inputs to soils and sediments. Ph.D Dissertation, University of Bristol, UK, 222 p.

Bull, I.D., van Bergen, P.F., Nott, C.J., Poulton, P.R., and Evershed, R.P., in press. Organic geochemical studies of soils from the Rothamsted Classical Experiments. V Ð The fate of lipids in dierent long-term soil experiments. Organic Geochemistry.

Eglinton, G., Hamilton, R.J., 1967. Leaf epicuticular waxes. Science 156, 1322±1335.

Eglinton, G., Hunneman, D.H., 1968. Gas chromatographic-mass spectrometric studies of long-chain hydroxy acids. Part I: The constituent acids of apple cuticle. Phytochemistry 7, 313±322. F.A.O., 1990. F.A.O.-UNESCO soil map of the world: revised

legend. World soil resources report 60, Rome.

Fogel, R., 1983. Root turnover and productivity of coniferous for-ests. Plant and Soil 71, 75±85.

(10)

Grasset, L., AmbleÁs, A., 1998. Structure of humin and humic acid from an acid soil as revealed by phase transfer catalyzed hydroly-sis. Organic Geochemistry 29, 881±892.

Gunstone, F.D., Harwood, J.L., Padley, F.B., 1986. The Lipid Handbook. Chapman and Hall, London, p. 896.

Heal, O.W., Anderson, J.M., Swift, M.J., 1997. Plant litter quality and decomposition: an historical overview. In: Cadisch, G., Giller, K.E. (Eds.), Driven by Nature. CAB International, Wallingford, UK, pp. 3±30.

Holloway, P.J., Deas, A.H.B., 1971. The occurrence of positional isomers of dihydroxyhexadecanoic acid in plant cutins and suber-ins. Phytochemistry 10, 2781±2785.

Holloway, P.J., 1982. The chemical constitution of plant cutins. In: Cutler, D.F., Alvin, K.L., Price, C.E. (Eds.), The Plant Cuticle, Linnean Society Symposium Series, vol. 10. Academic Press, London, pp. 45±85.

Holloway, P.J., 1984. Cutins and suberins, the polymeric plant lipids. In: Mangold, H.K. (Ed.), CRC Handbook of Chromatography, Lipids, vol. 1. CRC Press, Boca Ranton, pp. 321±346.

Hunneman, D.H., Eglinton, G., 1972. The constituent acids of Gymnosperm cutins. Phytochemistry 11, 1989±2001.

Jambu, P., AmbleÁs, A., Magnoux, P., Parlanti, E., 1995. Eects of addition of clay minerals on the fatty acid fraction of a podzol soil. European Journal of Soil Science 46, 187±192.

Johnston, A.E., Goulding, K.W.T., Poulton, P.R., 1986. Soil acidi®-cation during more than 100 years under permanent grassland and woodland at Rothamsted. Soil Use and Management 2, 3± 10.

KoÈgel-Knabner, I., Hatcher, P.G., Tegelaar, E.W., de Leeuw, J.W., 1992a. Aliphatic components of forest soil organic matter as determined by solid-state13_{C NMR and analytical pyrolysis. The} Science of the Total Environment 113, 89±106.

KoÈgel-Knabner, I., de, Leeuw, J.W., Hatcher, P.G., 1992b. Nature and distribution of alkyl carbon in forest soil pro®les: impli-cations for the origin and humi®cation of aliphatic biomacromo-lecules. The Science of the Total Environment 117/118, 175±185. Kolattukudy, P.E., Croteau, R., Buckner, J.S., 1976. Biochemistry of

plant waxes. In: Kolattukudy, P.E. (Ed.), Chemistry and Biochemistry of Natural Waxes. Elsevier Science Publisher, Amsterdam, pp. 289±347.

Matzke, K., Riederer, M., 1991. A comparative study into the chemical constitution of cutins and suberins fromPicea abies(L.) Karst.,Quercus roburL., andFagus sylvaticaL. Planta 185, 233± 245.

Minderman, G., 1968. Addition, decomposition and accumulation of organic matter in forests. Journal of Ecology 56, 355±362. Mlakar, A., Spiteller, G., 1996. Distinction between enzymatic and

nonenzymatic lipid peroxidation. Journal of Chromatography 743, 293±300.

Nierop, K.G.J., 1998. Origin of aliphatic compounds in a forest soil. Organic Geochemistry 29, 1009±1016.

Riederer, M., Matzke, K., Ziegler, F., KoÈgel-Knabner, I., 1993. Occurrence, distribution and fate of the lipid plant biopolymers cutin and suberin in temperate forest soils. Organic Geochemistry 20, 1063±1076.

von Rudlo, E., 1959. The wax of the leaves of Picea pungens

(Colorado spruce). Canadian Journal of Chemistry 37, 1038± 1042.

Saiz Jimenez, C., Hermosin, B., Guggenberger, G., Zech, W., 1996. Land use eects on the composition of organic matter in soil par-ticle size separates, Part III: Analytical pyrolysis. European Journal of Soil Science 47, 61±69.

Schulten, H.R., Leinweber, P., Reuter, G., 1992. Initial formation of soil organic-matter from grass residues in a long-term experiment. Biology and Fertility of Soils 14, 237±245.

Swift, M.J., Heal, O.W., Anderson, J.M., 1979. Decomposition in Terrestrial Ecosystems. Studies in Ecology, vol. 5. Blackwell Scienti®c Publications, Oxford, p. 372.

Tegelaar, E.W., Hollman, G., van der Vegt, P., de Leeuw, J.W., Holloway, P.J., 1995. Chemical characterisation of the permiderm tissue of some angiosperm species: recognition of an insoluble, non-hydrolyzable, aliphatic biomacromolecule (Suberan). Organic Geochemistry 23, 239±250.

Tilman, D., Dodd, M.E., Silvertown, J., Poulton, P.R., Johnston, A.E., Crawley, M.J., 1994. The park grass experiment: insights from the most long-term ecological study. Leigh, R.A., Johnston, A.E. (Eds.) Long-term Experiments in Agricultural and Ecological Sciences, CAB International, pp 287±303.

Tulloch, A.P., 1976. Chemistry of waxes of higher plants. In: Kolattukudy, P.E. (Ed.), Chemistry and Biochemistry of Natural Waxes. Elsevier Science Publisher, Amsterdam, pp. 235±287. U.S.D.A., 1992. Soil Survey Sta. Key to soil taxonomy, SMSS

Technical monograph No. 19; 5th Edition, Pocahontas Press Inc., Blacksburg, Virginia.

Voet, D., Voet, J.G., 1995. Biochemistry, 2nd ed. Wiley, New York, p. 1361.