labile, recalcitrant, and inert organic matter in mediterranean forest soils

ARTICLE IN PRESS

0038-0717/$ - se

doi:10.1016/j.so

�CorrespondE-mail addr

Soil Biology & Biochemistry 39 (2007) 202–215

www.elsevier.com/locate/soilbio

Labile, recalcitrant, and inert organic matter in Mediterraneanforest soils

Pere Roviraa,�, V. Ramon Vallejob

aDepartment of de Biologia Vegetal, Facultat de Biologia, Universitat de Barcelona, Diagonal 645, 08028 Barcelona, SpainbCEAM, Parc Tecnologic, Charles Darwin 14, 46980 Paterna, Valencia, Spain

Received 5 December 2005; received in revised form 12 July 2006; accepted 19 July 2006

Available online 12 September 2006

Abstract

The biochemical quality of soil organic matter (SOM) was studied in various profiles under Quercus rotundifolia Lam. stands on

calcareous parent material. Special attention was paid to the question of how biochemical quality is affected by position within the soil

profile (upper versus lower horizons). The following global SOM characteristics were investigated: (a) overall recalcitrance, using

hydrolysis with either hydrochloric or sulphuric acid; (b) hydrolyzable carbohydrates and polyphenolics; (c) extractability by hot water

and quality of the extract; and (d) abundance of inert forms of SOM: charcoal and soot-graphite. The recalcitrance of soil organic carbon

(OC) decreases with depth, following the order: H horizons4A horizons4B horizons. In contrast, the recalcitrance of nitrogen is

roughly maintained with depth. The ratio carbohydrate C to total OC increases from H to B horizons, due to the increasing importance

of cellulosic polysaccharides in B horizons, whereas other carbohydrates are maintained throughout the soil profile at a relatively

constant level, 12–15% of the total OC in the horizon. Whereas the quality of the hydrolyzable carbon (measured by the carbohydrate to

polyphenolic C ratio) decreases with depth from H to B horizons, the quality of the hot-water extractable organic matter is much higher

in B horizons than in A or H horizons. The relative importance of both charcoal and soot-graphitic C and N tends to increase with depth.

The ratio black/total is usually higher for N than for C, a result that suggests that inert SOM may represent a relevant compartment in

the nitrogen cycle. Overall, our data suggest that in Mediterranean forest soils the organic matter in B horizons could be less stable than

often thought.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Mediterranean forest; Black carbon; Soil organic matter; Recalcitrant organic matter; Acid hydrolysis; Available organic matter;

Carbohydrates; Polyphenolics; Nitrogen

1. Introduction

Soil organic matter (SOM) consists of a variety ofcompounds with different chemical characteristics andphysical availability. These features combine with theprevailing climatic constraints to determine the rate ofSOM decomposition, i.e., on the one hand, the repartitionof SOM between easily biodegradable and refractorycompounds and, on the other hand, the repartition ofSOM between physically available and physically protectedcompounds.

To quantify free and protected SOM pools, the use ofphysical fractionation procedures is widespread (Christen-

e front matter r 2006 Elsevier Ltd. All rights reserved.

ilbio.2006.07.021

ing author. Tel.: +3493 402 14 62; fax: +3493 411 28 42.

ess: [email protected] (P. Rovira).

sen, 1992). In contrast, there is less agreement about howto quantify SOM biochemical quality. Ideally, this ‘quality’must reflect biodegradability in the absence of physicalprotection and hence be based on chemical compositionindependent of physical position within the mineral matrix.Standard humus fractionation methods (Stevenson, 1982)are not completely satisfactory because they are time-expensive and the fractions obtained (fulvic acids, humicacids, humin) do not clearly correspond to functionallydifferent fractions. Advanced methods like pyrolysis (Py-GC, Py-MS) or

13

C-CPMAS-NMR can provide importantinformation on SOM structural features (Schnitzer, 1990),but translating this information to a quantitative measureof SOM quality is complicated.A useful alternative is acid hydrolysis, proposed by Stout

et al. (1981) as a simple method to evaluate SOM quality;

www.elsevier.com/locate/soilbio

dx.doi.org/10.1016/j.soilbio.2006.07.021

mailto:[email protected]

ARTICLE IN PRESSP. Rovira, V. Ramon Vallejo / Soil Biology & Biochemistry 39 (2007) 202–215 203

this procedure is easy to perform and can be applied to thelarge series of samples generally employed in ecologicalresearch. The non-hydrolyzable residue may include youngSOM (e.g., Bottner and Peyronel, 1977), but most radio-carbon studies have shown that the residue from acidhydrolysis is consistently older than the hydrolyzablefraction, whether the hydrolysis is applied to the wholeSOM or to a given chemical fraction such as humic acids orhumin (Goh and Pullar, 1977; Goh and Molloy, 1978;Leavitt et al., 1997; Paul et al., 1997). Refractory plantcompounds include lignin, suberin, and fatty acids (Mind-erman, 1968), all of which are resistant to acid hydrolysis.Pelz et al. (2005) studied the isotopic composition ofseveral SOM fractions, in a context of crop change (fromC3 to C4 plants), and found that the non-hydrolyzableresidue has clearly the lowest penetration of recent C.

The biochemical quality of SOM, as measured by itshydrolyzability, decreases with decomposition (Rovira andVallejo, 2002). Since SOM is older in deep horizons (Guillet,1979), biochemical quality should decrease with depth.Surprisingly, when acid hydrolysis was applied to mineralsoil at several depths the opposite result was obtained, i.e.,biochemical quality increased with depth (Goh et al., 1984;Joergensen and Meyer, 1990; Tan et al., 2004). The scarceavailable data suggest that the relationship between SOMbiochemical quality and position within the soil profile canbe complex and thus deserves further attention.

For a more complete picture of SOM biochemicalquality, other features must be taken into account. Thebiochemical quality of the whole SOM could not translateto the immediately available fractions, and hence the latterhad to be included in the study. Also, in addition torecalcitrant compounds, the presence of an inert pool hasoften been mentioned, either from theoretical approaches(RothC model: Jenkinson, 1990), or from analytical studies(Poirier et al., 2000). Owing to wildfires, black carbon (BC)is expected to be a main component of such an inertfraction in Mediterranean ecosystems (Poirier et al., 2000).

The aim of this paper is to characterize SOM inMediterranean forest soils from the point of view of itsbiochemical quality. We applied various chemical treat-ments (acid hydrolysis, BC analysis, hot water extraction)to a set of soil profiles in the Mediterranean area to obtaininsight on: (i) the abundance of labile, recalcitrant, andinert pools of SOM; (ii) how SOM biochemical qualitydepends on the position in the soil profile; (iii) how theoverall SOM quality affects the quality of the mostavailable SOM pools.

2. Materials and methods

2.1. Soils

The soil profiles studied (Table 1) were all located in thearea of Lleida (Catalonia, NE Spain), under Quercus

rotundifolia stands. All parent materials were calcareous(limestone, calcareous sandstone, and marl). Since position

within the soil profile was one of the main issues in this study,soil horizons were classified in three large groups: (i) organicL and F horizons were not included in the study; (ii) the firsthorizon under the F horizon, usually dark, highly organicand poor in recognizable plant debris, was considered the ‘H’horizon if its organic carbon (OC) content was 420%.Otherwise, it was considered the ‘A’ horizon; (iii) the firsthorizon below the H horizon was considered the ‘A’ horizon;and (iv) any horizon below the A horizon was considered a‘B’ horizon. A soil profile can have several B horizons, butonly a single H and a single A horizon.The profiles were usually poorly differentiated; the A

horizon was always present, whereas the B and/or Hhorizons may be absent. The sampled horizons were air-dried and crushed to pass through a 2mm mesh. Thetexture of the fine earth (o2mm) was obtained by thepipette method; total OC by dichromate oxidation, usingan aluminum digestion block (Nelson and Sommers, 1996);total N by using a CARLO ERBA analyzer; carbonates bythe calcimeter method; pH in water, in a suspension1:2.5 w/v.

2.2. Acid hydrolysis

To obtain the most complete picture possible, we appliedtwo different protocols:

(a)
Single-step hydrolysis with HCl. 100mg (in H horizons)or 400–500mg (in A and B horizons) were hydrolyzedwith 20ml of 6M HCl in sealed Pyrex tubes, at 105 1Cfor 18 h. The hydrolysate was discarded. The unhy-drolyzed residue was washed in deionized water withrepeated centrifugations and decantations, and thentransferred to pre-weighed vials, dried at 60 1C toconstant weight, and analysed for C and N using aCARLO ERBA analyzer.
(b)
Two-step hydrolysis with H2SO
4(Oades et al., 1970;

Rovira and Vallejo, 2000). 100mg (in H horizons) or400–500mg (in A and B horizons) were hydrolyzedwith 20ml of 2.5M H

2SO

4in sealed Pyrex tubes, at

105 1C for 30min. The hydrolysate (labile pool I) wasrecovered by centrifugation. The residue was washedwith water and dried. Then, 2ml of 13M H2SO4 wasadded, and the tubes were placed in an end-over-endshaker overnight. After diluting the acid with water to1M, the residue was hydrolyzed 3 h at 105 1C. Thehydrolysate (labile pool II) was also recovered bycentrifugation. The residue (recalcitrant pool) waswashed again, dried at 60 1C, weighed, and analysedfor C and N with a CARLO ERBA analyzer.

In both methods, we define the ‘recalcitrance index’ (RI)as follows:

RIC ¼ ðunhydrolyzed C=total OCÞ � 100,

RIN ¼ ðunhydrolyzed N=total NÞ � 100.

ARTICLE IN PRESS

Table 1

Main characteristics of the studied profiles

St Profile Parent material Horizon Thick pH H2O Gravel OC N CaCO3 Texture

* CS1 Calcareous sandstone A 2 7.27 32 112.0 7.9 206 Loam

* B 3 7.27 52 18.6 2.5 292 Sandy loam

* CS2 Calcareous sandstone A 2 6.78 32 132.5 10.1 104 Clay loam

* B 5 7.57 71 23.1 2.6 125 Sandy loam

– CS3 Calcareous sandstone H 3 6.9 39 204.4 15.8 157 Clay loam

* A 8 7.48 109 11.2 1.2 302 Sandy loam

* CS4 Calcareous sandstone A 2 7.18 29 98.7 7.3 227 Clay loam

* B 5 7.69 51 27.7 3.1 323 Sandy loam

* CS5 Calcareous sandstone A 2 6.15 13 84.8 5.5 0 Sandy clay loam

* B 23 6.52 24 8.2 1.3 0 Sandy loam

* CS6 Calcareous sandstone A 1.5 7.29 201 122.7 8.9 223 Clay loam

* B 5.5 7.9 81 32.05 3.5 334 Sandy loam

* CS7 Calcareous sandstone A 6 7.48 6 22.5 2 106 Loamy sand

* B 31 8.22 40 6.2 1.1 111 Loamy sand

* LI1 Limestone A 1.5 6.88 268 119.9 8.5 170 Silty loam

– B 13.5 7.20 56 11.0 1.1 9 Silty loam

* LI2 Limestone A 5 7.55 433 160.3 10.7 112 Silty clay loam

* B 27 7.62 558 34.7 3.6 121 Clay loam

* LI3 Limestone H 1 6.3 8 285.7 19.4 59 Clay

* A 8 7.6 386 26.0 3.1 39 Silty clay loam

* B 25 7.93 99 16.1 2.7 113 Clay loam

* LI4 Limestone H 2 6.35 59 238.1 22.3 109 N.D.

* A 2 7.25 397 99.5 9.0 154 Silty clay loam

* B1 6 6.71 444 34.3 3.0 24 Clay loam

* B2 23 7.42 366 24.4 2.4 83 Clay loam

* LI5 Limestone H 3 6.82 145 247.8 17.2 63 Silty loam

* A 12 7.33 535 63.7 4.1 75 Clay loam

* B 13 7.7 389 11.7 1.3 119 Clay loam

– MR1 Marl A 4 7.36 297 150.7 9.8 393 Clay

* B 12 8.05 233 17.4 1.4 698 Silty clay loam

* MR2 Marl A 8 6.62 616 198.9 14.8 258 Silty clay

* B1 21 7.91 354 21.7 2.7 487 Loam

* B2 16 8.14 449 10.9 1.0 646 Silty clay loam

* MR3 Marl A 3 7.7 65 90.5 7.5 328 Clay loam

* B1 7 7.68 74 20.4 2.8 371 Loam

– B2 18 8.40 306 9.9 1.0 400 Loam

– C 15 8.22 552 11.9 1.2 457 Silty loam

* MR4 Marl H 2 6.9 589 258.1 18.1 198 N.D.

* A 8 7.73 530 46.8 4.0 586 Loam

* MR5 Marl A 3 7.53 696 175.8 10.2 170 Silty clay

* B 14 8.02 686 33.2 2.4 275 Silty loam

N.D.: not determined; St: status of the horizon (*: analysed; –: not included in the study). Thickness given in cm. Gravel (fraction42mm) given as g/100 g

of fine earth (fraction o2mm). OC, N and CaCO3 given as g kg�1 in the fine earth (o2mm). Reproduced from Rovira & Vallejo (2003). Soil Biol

Biochem 35, 245–261, with permission of Elsevier.

P. Rovira, V. Ramon Vallejo / Soil Biology & Biochemistry 39 (2007) 202–215204

In the sulphuric acid method, the labile pools wereanalysed: (a) for total sugars by the phenol–sulphuricmethod (Dubois et al., 1956), using glucose as a standard,after elimination of Fe

3+

with Na2CO

3to avoid interfer-

ences (Martens and Frankenberger, 1993), and (b) for totalpolyphenolics by the Folin–Denis method (Ribereau-Gayon, 1968), using tannic acid as a standard. To convertcarbohydrates to carbohydrate C, we assumed a weight/Cratio of 2.5. To obtain the polyphenolic C, a weight/C ratioof 1.86 was assumed: this value was obtained from theformula for tannic acid (C

76H52O46).

The carbohydrate C/polyphenolic C ratio was thencalculated for each labile pool as an estimation of the

quality of the hydrolyzable C. This ratio is meant to reflectthe balance between easily biodegradable compounds (ofwhich carbohydrates are the most typical) and refractorycompounds (of which polyphenolics are the most repre-sentative).For both carbohydrates and polyphenolics, we obtained

the II/total ratio, i.e., [pool II/(pools I+II)]� 100. Asshown by Oades et al. (1970), carbohydrates of labile poolII correspond to cellulose, whereas those of labile pool Iinclude polysaccharides of both plant origin (hemicellu-loses, starch residues) and microbial origin (microbial cellwalls). Hence the II/total ratio for sugars is equivalent tothe cellulose-to-total-carbohydrates ratio.


2.3. Hot-water extraction

Extraction with hot water was applied as an estimationof the SOM available for microflora in the short term(Leinweber et al., 1995; Landgraf et al., 2005). Air-drysamples (500mg for H horizons, 1000mg in A and Bhorizons) were extracted with 20ml of distilled water insealed Pyrex tubes placed in an aluminum tube digestor, at105 1C for 1 h. The extract was recovered by centrifugationand decantation, filtered through Whatman 2 filter paper,and analysed for OC using a Shimadzu TOC analyzer, forcarbohydrates by the phenol–sulphuric method (Dubois etal., 1956), using glucose as a standard, and for poly-phenolics by the Folin–Denis method (Ribereau-Gayon,1968), using tannic acid as a standard. Carbohydrate andpolyphenolic C were obtained as explained above. Toobtain the extracted N, an aliquot of the extract wasoxidized with sulphuric acid and H2O2 (Hossain et al.,1993) and analysed for NH4

+ by the nitroprussiate method(Baethgen and Alley, 1989).

2.4. Inert (black) organic matter

Charcoal was estimated in the samples after oxidation ofthe non-BC organic matter. Briefly, 500mg of the groundsoil sample was placed in a Pyrex reaction tube andrefluxed with 5ml of concentrated HNO3 at 105 1C for 2 h.After cooling, the residue was washed with deionizedwater, transferred to pre-weighed centrifuge tubes andrecovered by centrifugation and decantation. Four addi-tional washings were made to eliminate any remainingHNO3. The residue was dried at 60 1C for 2 d, weighed, andanalysed for C and N with a CARLO ERBA analyzer.Basically, this method follows that of Winkler (1985),except that in the original Winkler method the amount ofSOM in the residue was analysed for loss on ignition.Henceforth, the term charcoal will appear between quota-tion marks ‘‘charcoal’’, because we assume that the methodused provides only an approximation of the true charcoalcontent of the sample (see Section 4).

For soot-graphitic SOM, which is assumed to be themost inert form of SOM, we applied a slightly simplifiedversion of the method of Gelinas et al. (2001). The groundsamples (about 5 g) were placed in 60ml polypropylenebottles, suspended in 20ml of deionized water, anddecarbonated by adding 6M HCl (in 2ml doses) until noeffervescence was observed. The suspensions were thencentrifuged and the liquid discarded: in the mineral samplesthis was done by decantation and in the organic horizonsby siphonation, taking special care to avoid losses offloating organic debris. The solid residue was re-suspendedin 30ml of a mixture 1:1 of 6M HCl+37% HF, and theclosed bottles were submitted to end-over-end shakingovernight; water was then added up to the 50ml mark, andthe bottles were centrifuged and the liquid discarded. Theresidue was treated again with HCl+HF, washed (re-suspended, centrifuged and decanted off) twice with

deionized water, dried at 60 1C for 2 d, and transferred toPyrex reaction tubes. The samples were hydrolyzed twice,first with 20ml of 1M HCl 3 h, then with 30ml of 6M HCl18 h, in an aluminum block digestor at 105 1C and underrefluxing. The hydrolysates were separated by centrifuga-tion and discarded. The residue was finally washed twicewith water (resuspension, centrifugation, decantation),transferred to pre-weighed vials, dried, and weighed. Asubsample of the residue was calcinated at 375 1C for 18 hin a muffle furnace. During the first 4 h of calcination, thedoor of the muffle was periodically opened for a fewseconds to permit the intake of O2, thus reducing the risk ofartifacts (generation of BC-like compounds). Finally, thecalcinated residue was analysed for total C and N in aCARLO ERBA elemental analyzer.

2.5. Size fractionation

For the size fractionation, samples were submitted to atwo-step dispersion: (1) 10 g- samples were put into plasticbottles and shaken with 50ml of water overnight in arotatory agitator at 50 revmin�1 and (2) thereafter, bottleswere put in an ice bath, and ultrasonically dispersed for15min (100W output) using a BRANSON probe-typedisintegrator. The dispersed samples were passed through acolumn of meshes (200, 50, and 20 mm), under magneticstirring and flow of distilled water. Materials retained inthe meshes were collected in crucibles and dried at 60 1C.Subsamples were ground for chemical analyses.Clays (o2 mm) were separated from fine silt (20–2 mm) by

repeated sedimentation in water and siphonation of thesuspended matter (clays), which was flocculated with aminimum amount of potassium aluminum sulphate, andrecovered by centrifugation. Both fine silt and clay weretransferred to crucibles, dried at 60 1C, and ground forchemical analyses.The particle-size fractions were analysed for OC and N

as above mentioned. A subsample was hydrolyzed with6M HCl to obtain the RIC and RIN indices. Only the HClmethod was applied to the size fractions because thismethod gave the best differentiation between horizons withrespect to SOM quality (see below).

2.6. Statistical methods

The relationship of most of the variables with the totalOC or N content of the horizons was analysed by curve-fitting techniques, using the SigmaPlot 2000 Packageroutines (Jandel Scientific). When appropriate, the bio-chemical characteristics of the horizons were compared bystandard one-way ANOVA. If the analysed data werepercentages, they were previously transformed by arcsinesquare root.

ARTICLE IN PRESSP. Rovira, V. Ramon Vallejo / Soil Biology & Biochemistry 39 (2007) 202–215206

3. Results

3.1. Organic matter recalcitrance (H2SO4 and HCl

methods)

C and N recalcitrance behave differently (Fig. 1). RICvalues increase with total OC in the horizon; they are highestin H horizons, and lowest in B horizons. In contrast, RINvalues are roughly maintained as total N decreases; theyshow a strong increase only at very low N values (o0.3%).Indeed, this increase occurs only in a few B horizons; inmost B horizons RIN values are maintained at levels similarto those of H or A horizons. For both C and N, these trendsare more apparent with the HCl method (Fig. 1c and d).Irrespective of the hydrolysis method used, the recalcitranceof N is much lower than that of C.

We did not observe a very clear relationship between sizefraction and OC or N recalcitrance (Table 2). For RICvalues, some consistent patterns appear: (i) RIC values tendto be lowest in the clay fraction (o2 mm), with theexception of the coarse sand fraction; and (ii) in the

Total OC in the horizon (%)

0 5 10 15 20 25 3030

40

50

60

70

0.

Rec

alci

tran

ce in

dex

for

C (

RI C

)

30

40

50

60

70(a)

(c)

2 = 0.217**r

2 = 0.802***r

Fig. 1. Recalcitrance indexes for carbon and nitrogen (RIC, RIN), plotted again

the fitting curves are given in Appendix A: (a) RIC values, obtained by the sulp

(c) RIC values, obtained by the hydrochloric acid method; and (d) RIN values

fractions o50 mm, which can form organomineral com-plexes, the RIC values decrease with depth, i.e., are in theorder H4A4B horizons. The trends observed for the RINvalues of the size fractions are less clear, overall. In general,RIN values are highest in B horizons and lowest in Ahorizons. An exception is the clay fraction of B horizons,with the lowest RIN in the whole data set. In contrast, in Hand A horizons RIN values are highest in the clay fraction.In most cases, the high variability prevents any statisticalsignificance of these trends.

3.2. Quality of the labile organic matter (H2SO4 method)

The amount of carbohydrates in both labile poolsincreases with the total OC in the horizon. In labile poolI this relationship is nearly linear, with little data dispersion(Fig. 2a); as a result, the carbohydrate C in this poolrepresents a quite constant percent of the total OC(Fig. 2b). In contrast, in labile pool II a less regularrelationship appears; this carbohydrate accounts forjust 2–4% of the total OC in H and A horizons, but it

Rec

alci

tran

ce in

dex

for

N (

RI N

)0

10

20

30

40

50

60

70

80

Total N in the horizon (%)

0 0.5 1.0 1.5 2.0 2.50

10

20

30

40

50

60

H horizons

A horizons

B horizons

(b)

(d)

2 = 0.617***r

2 = 0.152 nsr

st total OC or N in the horizon. Points have been fitted to parabolic curves;

huric acid method; (b) RIN values obtained by the sulphuric acid method;

, obtained by the hydrochloric acid method.

ARTICLE IN PRESS

Table 2

Recalcitrance indices for carbon (RIC) and nitrogen (RIN) obtained by the hydrochloric acid method for the several size fractions

Size fraction (mm)

Horizon 2000–200 200–50 50–20 20–2 o2 Whole sample

RIC H 53.3 (6.1) 1 57.7 (27.8) 1 65.5 (14.4) a 1 56.7 (3.4) 1 56.6 (20.9) a 1 60.5 (5.1) a

A 50.9 (18.0) 1,2 60.6 (13.7) 2 57.9 (12.5) ab 1,2 55.8 (11.6) 1,2 46.1 (10.9) ab 1 53.6 (5.6) a

B 65.5 (15.5) 1 46.3 (8.9) 2 49.1 (12.4) b 2 46.7 (5.7) 2 33.4 (4.1) b 3 42.9 (5.5) b

RIN H 16.3 (2.2) a 20.3 (6.7) 21.6 (5.1) a 18.3 (2.5) 27.2 (24.4) 15.2 (1.0)

A 18.6 (5.9) ab 16.4 (6.0) 19.6 (4.4) a 17.9 (3.3) 22.8 (21.0) 15.1 (1.3)

B 30.5 (19.6) b 1 26.5 (17.3) 1,2 37.6 (17.0) b 1 23.1 (9.9) 1,2 14.9 (6.3) 2 18.2 (8.1)

Results for whole (unfractionated) soil samples are also given for comparison. Data are averages; values between parentheses standard deviations.

Statistical comparison was done with data previously transformed per arc sine of square root. Within a column, and for a given parameter (either RIC or

RIN, not both), values followed by the same letter are not statistically different (P ¼ 0.05). Within a row (excluding the ‘whole sample’ column), values

followed by the same number are not statistically different. If no letter or number is given, no significant differences were detected.

P. Rovira, V. Ramon Vallejo / Soil Biology & Biochemistry 39 (2007) 202–215 207

can account for up to 12% of total OC in B horizons(Fig. 2c and d).

In contrast with the behaviour observed for carbohy-drates, that of extracted polyphenolics was quite similar forboth labile pools, i.e.: (i) the amount of extractedpolyphenolics consistently increases with total OC (Fig.3a and c), and (ii) the polyphenol C/total OC ratioincreases with decreasing total OC in the horizon, and thisincrease is very strong in B horizons (Fig. 3b and d).Unrealistic values may appear (e.g., in some B horizonspolyphenolic C in labile pool I is calculated to be morethan 40% of the total OC in the horizon), which suggeststhat tannic acid might not be the appropriate standard forall samples. As observed for carbohydrates, the amount ofextracted polyphenolics was much higher in labile pool Ithan in labile pool II.

A result of this increase is that in both labile pools thecarbohydrate C/polyphenol C ratio consistently increaseswith total OC in the horizon (Fig. 4). While in many Bhorizons this ratio may be o1, in H horizons it can reachvalues 44.

As for the II/total ratio, owing to the above-mentionedtrends, the cellulose/total carbohydrates ratio decreaseswith increasing total OC: it is highest in B horizons, andlowest in H horizons (Fig. 5a). In spite of this increase indeep layers, in most cases cellulose accounts for less than40% of the total carbohydrates; i.e., non-cellulosiccarbohydrates are almost always dominant. In contrast,no trend is detectable for phenolic compounds, whosedistribution between the two labile pools seems quiteconstant and not clearly related to the total OC in thehorizon (Fig. 5b).

3.3. Available organic matter (hot-water extract)

Hot-water-extracted OC accounts for 4–10% of thetotal, the ratio slightly increasing with total OC (Fig. 6a).The carbohydrate C/total OC ratio in the extract wasroughly similar in all horizons (Fig. 6b), whereas thepolyphenolic C/total OC ratio showed a curvilinearrelationship with total OC in the horizon: it was roughly

maintained at about 4 in H and A horizons, but droppeddown to 2 or less in B horizons (Fig. 6c). As a result, thequality of the extract (carbohydrate C/polyphenol C ratio),which stayed at values o10 in most H and A horizons,increased in B horizons, being 480 in one extreme caseoutside the graph (Fig. 6d).N extractability also increased with total N in the

horizons (Fig. 7). This trend is clearer when plotted againsttotal OC, i.e., it seems to depend more on the total OC thanon the total N in the horizon. No relationship was detectedbetween the C/N of the extract and the total OC or N in thehorizon (not shown).

3.4. Inert organic matter

Both inert pools showed a similar relationship with totalOC or N in the horizon; i.e., the relative amount of OC orN in the fraction increases in the order HoAoB‘‘charcoal’’ or H AoB (soot-graphitic) (Fig. 8). In H andmost A horizons ‘‘charcoal’’ accounts for less than 5% ofthe total OC, while in B horizons it can account for morethan 15%. The soot-graphitic fraction usually accounts forless than 2% of total OC, but in some B horizons it canbecome much higher.Nitrogen was detected in both ‘‘charcoal’’ and soot-

graphitic SOM. In agreement with the general term BC, wecall this pool of N black nitrogen (BN). Their amountswere always low; in the soot-graphitic residue, BN was notdetected in some cases (N in the calcinated residue belowthe detection limit) (Fig. 8b and d). Nevertheless, in most‘‘charcoal’’ and in some soot-graphitic SOM, the ratio BN/total N is higher than the ratio BC/total OC.

4. Discussion

4.1. Recalcitrance of C and N

There is little information on how hydrolyzabilitychanges with the position in the soil. The decrease in Crecalcitrance with depth has been observed previously

ARTICLE IN PRESS

Labile Pool II


0 5 10 15 20 25 300

5

10

15

20

25

Labile Pool I

Car

bohy

drat

e C

(%

of t

otal

OC

in th

e ho

rizon

)

0

5

10

15

20

25

30

Labile Pool II

0 5 10 15 20 25 300

3

6

9

12

15

Labile Pool I

Car

bohy

drat

e co

nten

t, as

glu

cose

(m

g g-1

)

0

20

40

60

80

100

120 H horizons

A horizons

B horizons

(a) (b)

(c) (d)

2 = 0.923***r

2 = 0.615***r

2 = 0.009 nsr

2 = 0.979***r

Fig. 2. Soil carbohydrates, obtained by hydrolysis with sulphuric acid (two-step hydrolysis): (a) carbohydrates of labile pool I, as milligrams of glucose

equivalent per gram of soil horizon; (b) carbohydrate C in labile pool I, as % of total OC in the horizon; (c) carbohydrates of labile pool II, as mg of

glucose equivalent per g of soil horizon; and (d) carbohydrate C in labile pool II, as % of total OC in the horizon. Points have been fitted to linear (a) and

(b), third-order polynomial (c), or exponential curves (d). The fitting curves are given in Appendix A.


(Joergensen and Meyer, 1990; Tan et al., 2004) and can beexplained by several simultaneous processes:

(a)
The biodegradation of lignin (the main unhydrolyzableplant polymer) as plant-derived organic debris areintroduced downwards in the soil profile. Ligninoxidation and fragmentation will result in hydrolyzablephenolic compounds, which in our method are releasedto the labile pools (Fig. 3b, d); this in turn will result ina decrease in the quality of these pools (Fig. 4). Theincreased oxidation of lignin as SOM is introduceddownwards in the profile has been well-described intemperate forest soils (Kogel, 1986; Kogel-Knabner etal., 1991), although there is a lack of data forMediterranean forest soils.
(b)
The release of fresh debris of high quality, namely fineroots. Fine roots are usually more abundant in upperlayers, but in A horizons the characteristics of SOM arealso greatly affected by the more humified SOM
coming from H horizons. In contrast, in B horizons fineroots may constitute the main source of SOM,especially in some Mediterranean forest soils in whichfine roots apparently accumulate in the middlehorizons (Lopez et al., 2001). In small debris carefullymixed with soil samples the cellulose/total carbohy-drates ratio decreases with decomposition (Rovira andVallejo, 2002), likely because of: (i) the breakdown ofthe cellulose polymer, which results in the release ofmono- or oligosaccharides to the non-cellulosic pool(labile pool I), and (ii) the increased accumulation ofmicrobial polysaccharides, also included in labile poolI. Hence the increase in the cellulose/total carbohy-drates ratio in B horizons (Fig. 5) suggests a relevantpresence of fresh plant contributions therein.

(c)
The contrasted mobility of hydrolyzable and non-hydrolyzable fractions within the soil profile. Thehydrolysate is expected to move down the soil profilemore easily than the unhydrolyzable fraction (Goh

ARTICLE IN PRESS

Labile Pool II


0 5 10 15 20 25 300

1

2

3

4

5

6

Labile Pool I

Pol

yphe

nol C

(%

of t

otal

OC

)

0

20

40

60

80

Labile Pool II

0 5 10 15 20 25 300

2

4

6

8

10

Labile Pool I

Pol

yphe

nol c

onte

nt, a

s ta

nnic

aci

d (m

g g-1

)

0

5

10

15

20

25

H horizons

A horizons

B horizons

(a) (b)

(c)

2 = 0.633***r

2 = 0.724***r

2 = 0.928***r

2 = 0.833***r

(d)

Fig. 3. Soil polyphenolics, released by acid hydrolysis with sulphuric acid (two-stephydrolysis): (a) polyphenolics of labile pool I, as milligrams of tannic

acid equivalent per gram of soil horizon; (b) polyphenol C in labile pool I, as % of total OC in the horizon; (c) polyphenolics of labile pool II, as milligrams

of tannic acid equivalent per gram of soil horizon; and (d) polyphenol C in labile pool II, as % of total OC in the horizon. Points have been fitted to

hyperbolic curves, given in Appendix A.


et al., 1984). Thus, even with the probable partialbiodegradation of the hydrolyzable fraction during theprocess, this would result in the passage of hydrolyz-able compounds from the surface to the deep horizons,and hence in decreased recalcitrance in the latter.

The low RIC in clay-associated SOM, especially in Bhorizons (Table 2), is an additional explanation for thedecrease in RIC values with depth, because in deephorizons the SOM associated to the finest fractionsbecomes dominant (Rovira and Vallejo, 2003). However,the decrease in RIC values from H to B horizons is alsoobserved in most size fractions, suggesting that the positionin the soil profile itself is the dominant factor.

As for the changes in RIN values with depth, there areconflicting results in the literature. In soils of theArgentinian pampas, N recalcitrance decreases with depthin most cases, in both burnt and unburnt plant commu-nities (Sanchez and Lazzari, 1999). However, from the dataprovided by Tan et al. (2004), we can calculate an overall

maintenance of RIN with depth in agricultural andmeadow soils, whereas in forest soils RIN tends to decreasewith depth: from 17.4 in 0–5 cm to 13.1 in 10–20 cm.Differences with our results may be partly due todifferences in the method of hydrolysis, in particular inthe time of reaction. In our study N recalcitrance is roughlyconstant throughout the soil profile, except for some N-poor B horizons, in which RIN values can be much higher.An obvious reason for such an increase could be the clay-fixed ammonium, which increases in relative importance indeep horizons (Stevenson, 1982) and is only partly releasedby acid hydrolysis (Freney and Miller, 1970; Greenfield,1992). Nevertheless, the very low recalcitrance of clay-associated N in B horizons (Table 2) does not substantiatethis explanation. Rather, black nitrogen seems to beresponsible for this, since the B horizons showing thehighest RIN values also have the highest amounts of N inthe ‘‘charcoal’’ fraction.The inconsistent relationship between particle size and

RIN value (Table 2) disagrees with the results from

ARTICLE IN PRESS

Labile Pool II

0 5 10 15 20 25 30

Labile Pool I


0 5 10 15 20 25 30

Car

bohy

drat

e C

/ P

olyp

heno

l C

0

1

2

3

4

5 H horizons

A horizons

B horizons

(a) (b)

2 = 0.912***r

2 = 0.604***r

Fig. 4. Carbohydrate C to polyphenol C ratio: (a) in labile pool I, and (b) in labile pool II. Points have been fitted to power curves, given in Appendix A.

Hydrolyzed polyphenolics

0 5 10 15 20 25 30P

olyp

heno

lics:

II /

tota

l rat

io (

%)

0

10

20

30

40

50

60Carbohydrates


0 5 10 15 20 25 30

Cel

lulo

se /

Tota

l car

bohy

drat

es (

%)

0

10

20

30

40

50

60

H horizons

A horizons

B horizons

(a) (b)

2 = 0.702***r

2 < 0.001 nsr

Fig. 5. Labile pools I and II: ratio pool I/pools I+II for carbohydrates and polyphenolics: (a) cellulose/total carbohydrates (pool II/pools I+II) and (b)

polyphenolics in labile pool II/total hydrolyzed polyphenolics. Points in Fig. 5(a) have been fitted to an hyperbolic curve (given in Appendix A).


Anderson et al. (1981), who observed a general decrease inN recalcitrance with decreasing particle size, but agreeswith most of the results from Leinweber and Schulten(1998), who observed a lack of relationship, except thatRIN of clays is consistently the lowest one, a fact that inour study occurs only in B horizons. The fact that theresults from Anderson et al. (1981) and Leinweber andSchulten (1998) refer to agricultural soils, and not to forestsoils as ours do, may explain these differences with ourdata.

4.2. Hot-water extract

Water extraction has been used to predict both OC andN pools mineralizable in the short term (Stanford, 1982;Leinweber et al., 1995; Landgraf et al., 2005). Haynes

(2000) takes the sum: light fraction+microbial biomass+-soluble SOM, to be the overall ‘‘labile’’ pool; because of thehigh correlation found between these factors, they areassumed to reflect a single soil property. Ghani et al. (2003)stated that the hot-water extract integrates many of thefeatures of the soil samples, and suggested using it as anindicator of soil quality.According with this view, our results show that B

horizons are highly favourable to microbial activity: OCextractability in them is only slightly lower than in H or Ahorizons, but the quality of the extracted OC is muchhigher (Fig. 6). This result is opposite to that obtained inthe hydrolysates, whose quality decreases with depth (Fig.4); i.e., the qualities of the available and the labile fractionsare not closely related. There are several explanations forthis. The decrease in quality of the B-horizon hydrolysates

ARTICLE IN PRESS

HW

OC

(%

of t

otal

OC

)

0

3

6

9

12

15 H horizonsA horizonsB horizons

Car

bohy

drat

e H

WC

(%

of

HW

C)

0

10

20

30

40

50

60

70


0 5 10 15 20 25 30

Pol

yphe

nolic

HW

C (

% o

f HW

C)

0

1

2

3

4

5

6

7

0 5 10 15 20 25 30

Car

bohy

drat

e /

Pol

yphe

nolic

0

5

10

15

20

25

(a) (b)

(c) (d)

2 = 0.596***r

2 = 0.177*r 2 = 0.07 nsr

Fig. 6. Organic matter extracted with hot water, plotted against total OC in the horizon: (a) extracted OC, as percent of total OC in the horizon; (b) hot-

water extracted carbohydrate C, as percent of total extracted OC; (c) hot-water extracted polyphenolic C, as percent of total extracted OC; and (d)

carbohydrate C/polyphenolic C ratio in the hot-water extract. Points in Fig. 6 (a) and (b) have been fitted to linear graphs; in Fig. 6(c) to a hyperbolic

curve. The fitting curve in Fig. 6(d) is the ratio between the fitting graph in Fig. 6(b) and the fitting curve in Fig. 6(c). Equations of the fitting graphs are

given in Appendix A.


seems to be attributable to the increased amount ofphenolic compounds in them. Phenolic compounds aremore hydrophobic than saccharides, and this characteristicgenerates a higher retention in soil by adsorption to soilactive surfaces (Guo and Chorover, 2003). This should alsoprevent the leaching of phenolic compounds from theupper horizons, in contrast to the much more solublecarbohydrates, which can reach B horizons in this way.

A slight decrease in OC extractability with depth couldbe a widespread pattern, since it was also observed byLandgraf et al. (2005), for forests of Saxony (Germany). Incontrast, these authors did not observe any consistentpattern for hot-water extractable N.

4.3. Inert organic matter

The maintenance of the term inert to refer to ‘‘charcoal’’and soot-graphitic SOM is due to the lack of specificity in

the analyses, which is a problem with most of the methodsfor quantifying BC in soil samples. Recent comparativestudies on BC analysis (Schmidt et al., 2001) clearly showthat this problem has not yet been solved, probablybecause the available methods differ in the quantifiedfraction within the BC continuum.In Winkler’s (1985) original method samples are oxidized

with HNO3 in a hot water bath (temperature not given) for1 h. However, in our own trials (unpublished), even after24 h at 105 1C, concentrated HNO3 did not completelydestroy the ground plant debris while charcoal was partlylost. Oxidation with HNO3 is a non-specific treatment, andthe high correlation obtained by Winkler (1985) in his trialsbetween expected and measured charcoal contents could bemisleading. To destroy more of the non-charcoal SOM, weapplied harsher conditions (2 h, 105 1C) than in the originalWinkler method, but not enough to destroy a substantialamount of charcoal. We assume the method as conven-

ARTICLE IN PRESS


0.0 0.5 1.0 1.5 2.0 2.5

HW

N (

% o

f tot

al N

)

0

2

4

6

8

10

12

14 H horizons

A horizons

B horizons


0 5 10 15 20 25 30

(a) (b)

2 = 0.398***r 2 = 0.477***r

Fig. 7. Hot-water extracted N, as percent of total N: (a) plotted against total N in the horizon, and (b) plotted against total OC in the horizon. Points have

been fitted to hyperbolic curves, given in Appendix A.


tional: it gives a highly resistant organic fraction consistinglargely, but not exclusively, of BC. The use of the term‘‘charcoal’’ in quotation marks must be taken in this sense.As for soot-graphitic SOM, the method by Gelinas et al.(2001) seems to be the most specific, but even in this caseorganic compounds other than soot and graphite canpartially resist the treatment, and thus be included in thesoot-graphitic pool. At any rate, radiocarbon studies haveshown that the residue from the nitric acid treatment isconsistently the oldest SOM fraction (Goh, 1978; Goh andPullar, 1977; Goh and Molloy, 1978; Hammond et al.,1991), and thus it seems logical to take it as the most inertone.

The very high relevance of BC in the overall SOM hasbecome widely accepted in recent years. In chernozemicsoils of Germany, for instance, charred OC can account forup to 45% of total OC (Schmidt et al., 1999), whereas inblack chernozemic soils of Canada it can account for up to80% in an extreme case (Ponomarenko and Anderson,2001), or up to 35% in US agricultural soils (Skjemstad etal., 2002). Our data are very far from all these. In ourprofiles, rarely is more than 5% of the total OC accountedfor in the ‘‘charcoal’’ pool, and that figure is usually lessthan 2% in the soot-graphitic pool. Even taking intoaccount that our data are not directly comparable to thoseobtained by contrasted methods, they are low enough tosuggest that in our profiles BC is poorly relevant for theaccumulation and stabilization of OC in the long term. Thereasons for these low amounts could be two (or acombination of both): either BC is more biodegradablethan often assumed (Bird et al., 1999), or, simply, thecurrent generation of BC in these forests is small. Thealmost irrelevant amount of these inert forms in H or Ahorizons, where SOM is younger, suggests that theirgeneration has been low in recent years because of the

reduced number of wildfires. Our results may not beextrapolative to all Mediterranean ambients, however. OurQuercus rotundifolia stands (the climatic forest in the area)can be considered representative of relatively untouchednative forests since they suffered few wildfires in the pastcentury. In other types of Mediterranean terrestrialecosystems (shrublands, pine forests), often rich inpyrophitic plant species, wildfires are more frequent, andBC could be more abundant than in our profiles.The presence of N in the inert SOM was expected, since

N stabilizes together with C in condensed organicstructures when burning organic debris, trunks orbranches. Nevertheless, the high amount of N in the‘‘charcoal’’ pool, and in some cases also in the soot-graphitic pool, was unexpected. The ratio black/total isusually higher for N than for OC. This would imply that,as a pool in which a chemical element almost goes out ofcycling, the inert SOM could be more relevant for N thanfor C, an implication that deserves further attention.

5. Conclusions

A widely extended concept indicates that SOM stabilizesin deep horizons because: (i) it is physically more protectedand (ii) it is in a more advanced state of biochemicalstabilization relative to the SOM of surface horizons. ForMediterranean forest soils, while our previous workconfirms the first statement (Rovira and Vallejo, 2003),the results of the current study do not confirm the second.In fact, the chemical indicators applied suggest a quitecomplex situation. Overall, recalcitrance clearly decreaseswith depth. The quality of the hydrolysate (measured by itscarbohydrate/polyphenol ratio) also decreases with depth,but this does not translate to the most extractable fraction

ARTICLE IN PRESS

'Cha

rcoa

l' C

(%

of t

otal

OC

)

0

5

10

15

20

25

30

'Cha

rcoa

l' N

(%

of t

otal

N)

0

5

10

15

20

25

30


0 5 10 15 20 25 30

Soo

t-gr

aphi

tic C

(%

of t

otal

OC

)

0

2

4

6

8

10


0.0 0.5 1.0 1.5 2.0 2.5

Soo

t-gr

aphi

tic N

(%

of t

otal

N)

0

2

4

6

8

10

(a)

H horizons

A horizons

B horizons

(b)

(c) (d)

2 = 0.320**r 2 = 0.581***r

2 = 0.900***r 2 = 0.908***r

Fig. 8. Content of inert organic matter, plotted against total OC or total N in the horizon: (a) ‘‘charcoal’’ C, as percent of total OC; (b) ‘‘charcoal’’ N, as

percent of total N in the horizon; (c) soot-graphitic C, as percent of total OC; and (d) soot-graphitic N, as percent of total N: the points for which N was

not detectable have been omitted. In (a) and (b), points have been fitted to hyperbolic curves; in (c) and (d) to exponential curves. The equations of the

fitting curves are given in Appendix A.


(hot-water extract), whose quality is much higher in Bhorizons.

The increased relative importance of inert SOM fractionswith depth is the only feature that agrees with an increasedbiochemical stabilization with depth. However, the amountof either ‘‘charcoal’’ or soot-graphitic OC stays almostalways at a quite low level, not high enough to distort theoverall trends observed.

There are several possible explanations for the abun-dance of labile compounds in B horizons: (i) the input offresh organic matter, e.g., dead fine roots, in deep layers,(ii) an in situ evolution due to differences in thepedoclimate between upper and lower horizons, or (iii)the differential translocation of labile and recalcitrantcompounds, with the former being more easily leacheddown the soil profile. Further research is needed toelucidate the relative importance of each of these explana-tions.

Under Mediterranean climate, deep horizons are morefavourable to microbial activity than shallower ones, owingto the higher drought levels in the latter (Rovira andVallejo, 1997). This, in addition to the lower recalcitranceof SOM in B horizons, suggests that in Mediterraneanforests the SOM of B horizons could be much less stablethan often thought.

Acknowledgement

We acknowledge the collaboration of Pilar Fernandez,of the Scientifical and Technical Services of the Universitatde Barcelona, for N analyses.

Appendix A

Equations for the fitting curves of the figures can be seenin Table A1.

ARTICLE IN PRESS

Table A1

Equations for the fitting curves of the figures

Fig. Function type Equation

1(a) Power y ¼ 49.4481 x0.0461

1(b) Hyperbolic decay (l.s.) y ¼ 18.8809+1.2761/(0.0227+x)

1(c) Power y ¼ 39.1493 x0.1402

1(d) Exponential decay y ¼ 14.4338+290.4377 exp(�25.2757x)

2(a) Linear y ¼ 0.7671+3.0491 x

2(b) Linear (n.s.)

2(c) Polynomial, order 3 y ¼ 0.1196+1.4188 x�0.0945 x2+0.0025 x3

2(d) Inverse polynomial, order 1 y ¼ 2.3819+5.2024 x�1

3(a) Hyperbolic y ¼ 17.6557 x/(2.3901+x)

3(b) Hyperbolic decay y ¼ 3.1843+36.9617/(0.0412+x)

3(c) Hyperbolic y ¼ 3.2801 x/(1.9506+x)

3(d) Hyperbolic decay y ¼ �0.2924+22.0885/(2.4888+x)

4(a) Power y ¼ 0.3480 x0.7159

4(b) Power y ¼ 0.9938+0.1619 x0.8346

5(a) Hyperbolic decay y ¼ 15.8001+24.8056/(0.1277+x)

5(b) Linear (n.s.)

6(a) Linear y ¼ 5.9318+0.0707 x

6(b) Linear (n.s.) y y ¼ 32.8135�0.0953 x

6(c) Hyperbolic y ¼ 4.7484 x/(1.7471+x)

6(d) Ratio 6(b)/6(c) y ¼ (32.8135�0.0953 x)/(4.8133 x/(1.7884+x))

7(a) Hyperbolic y ¼ 9.4964 x/(0.2487+x)

7(b) Hyperbolic y ¼ 8.9241 x/(2.0822+x)

8(a) Hyperbolic decay y ¼ 40.7057/(0.5044+x)

8(b) Hyperbolic decay y ¼ 1320/(0.0028+x)

8(c) Exponential decay y ¼ 0.1258+17.7106 exp(�0.8251 x)

8(d) Exponential decay y ¼ 0.1995+7.8878 exp(�6.7081 x)

The equation is not given when the fit is not significant (n.s.), except in one case (y: equation 6b) in which this equation was used later to calculate the best

fit to another set of data (equation 6d). The equation obtained for Fig. 1(b) is given, even though it was in the limit of significance (l.s.), with P ¼ 0.0606.


References

Anderson, D.W., Saggar, S., Bettany, J.R., Stewart, J.W.B., 1981. Particle

size fractions and their use in studies of soil organic matter. I. The

nature and distribution of forms of carbon, nitrogen, and sulfur. Soil

Science Society of America Journal 45, 767–772.

Baethgen, W.E., Alley, M.M., 1989. A manual colorimetric procedure for

measuring ammonium nitrogen in soil and plant digests. Communica-

tions in. Soil Science and Plant Analysis 20, 961–969.

Bird, M.I., Moyo, C., Veenendaal, E.M., Lloyd, J., Frost, P., 1999.

Stability of elemental carbon in a savanna soil. Global Biogeochemical

Cycles 13, 923–932.

Bottner, P., Peyronel, A., 1977. Dynamique de la matiere organique dans

deux sols mediterraneennes etudiee a partir de techniques de datation

par le radiocarbone. Revue d’ecologie et de Biologie du Sol 14,

385–393.

Christensen, B.T., 1992. Physical fractionation of soil and organic matter

in primary particle size and density separates. In: Stewart, B.A. (Ed.),

Advances in Soil Science, vol. 20. pp. 1–90.

Dubois, M., Giles, K.A., Hamilton, J.K., Rebers, P.A., Smith, F., 1956.

Colorimetric method for the determination of sugars and related

substances. Analytical Chemistry 28, 350–356.

Freney, J.R., Miller, R.J., 1970. Investigation of the clay mineral

protection theory for non-hydrolyzable nitrogen in soil. Journal of

the Science of Food and Agriculture 21, 57–61.

Gelinas, Y., Prentice, K.M., Baldock, J.A., Hedges, J.I., 2001. An

improved thermal oxidation method for the quantification of soot/

graphitic black carbon in sediments and soils. Environmental Science

and Technology 35, 3519–3525.

Ghani, A., Dexter, M., Perrott, K.W., 2003. Hot-water extractable carbon

in soils: a sensitive measurement for determining impacts of fertiliza-

tion, grazing and cultivation. Soil Biology & Biochemistry 35,

1231–1243.

Goh, K.M., 1978. Removal of contaminants to improve the reliability of

radiocarbon dates of peat. Journal of Soil Science 29, 340–349.

Goh, K.M., Molloy, B.P.J., 1978. Radiocarbon dating of paleosols using

soil organic matter components. Journal of Soil Science 29, 567–573.

Goh, K.M., Pullar, W.A., 1977. Radiocarbon dating techniques for

tephras in central North Island, New Zealand. Geoderma 18, 265–278.

Goh, K.M., Stout, J.D., O’Brien, B.J., 1984. The significance of

fractionation in dating the age and turnover of soil organic matter.

New Zealand Journal of Science 27, 69–72.

Greenfield, L.G., 1992. Acid hydrolysis and the release of fixed

ammonium from soils. Soil Biology & Biochemistry 24, 271–273.

Guillet, B., 1979. etude du renouvellement des matieres organiques des

sols par les radioisotopes (14C). In: Bonneau, M., Souchier, B. (Eds.),

Pedologie, vol. 2. Constituants et Proprietes du Sol., Masson, Paris,

pp. 210–226.

Guo, M.X., Chorover, J., 2003. Transport and fractionation of dissolved

organic matter in soil columns. Soil Science 168, 108–118.

Hammond, A.P., Goh, K.M., Tonkin, P.J., Manning, M.R., 1991.

Chemical pretreatments for improving the radiocarbon dates of peats

and organic silts in a gley podzol environment: Grahams terrace,

North Westland. New Zealand Journal of Geology and Geophysics 34,

191–194.

Haynes, R.J., 2000. Labile organic matter as an indicator of organic

matter quality in arable and pastoral soils in New Zealand. Soil

Biology & Biochemistry 32, 211–219.

Hossain, A.K.M.A., Khanna, P.K., Field, J.B., 1993. Acid-peroxide

digestion procedure for determining total nitrogen in chloroform-

fumigated and non-fumigated soil extracts. Soil Biology & Biochem-

istry 25, 967–969.


Jenkinson, D.S., 1990. The turnover of organic carbon and nitrogen in

soil. Philosophical Transactions of the Royal Society of London, Series

B (Biological Sciences) 329, 361–368.

Joergensen, R.G., Meyer, B., 1990. Chemical change in organic matter

decomposing in- and on a forest Rendzina under beech (Fagus

sylvatica L.). Journal of Soil Science 41, 17–27.

Kogel, I., 1986. Estimation and decomposition pattern of the lignin

component in forest humus layers. Soil Biology & Biochemistry 18,

589–594.

Kogel-Knabner, I., Hatcher, P.G., Zech, W., 1991. Chemical structural

studies of forest soil humic acids: aromatic carbon fraction. Soil


Landgraf, D., Wedig, S., Klose, S., 2005. Medium- and short-term

available organic matter, microbial biomass, and enzyme activities in

soils under Pinus sylvestris L. and Robinia pseudoacacia L. in a sandy

soil in NE Saxony, Germany. Journal of Plant Nutrition and Soil

Science 168, 193–201.

Leavitt, S.W., Follett, R.F., Paul, E.A., 1997. Estimation of slow and fast

cycling soil organic carbon pools from 6N HCl hydrolysis. Radio-

carbon 38, 231–239.

Leinweber, P., Schulten, H.R., 1998. Nonhydrolyzable organic nitrogen in

soil size separates from long-term agricultural experiments. Soil


Leinweber, P., Schulten, H.R., Korschens, M., 1995. Hot water extracted

organic matter: chemical composition and temporal variations in a

long-term field experiment. Biology and Fertility of Soils 20, 17–23.

Lopez, B., Sabate, S., Gracia, C.A., 2001. Vertical distribution of fine root

density, length, density, area index and mean diameter in a Quercus

ilex forest. Tree Physiology 21, 555–560.

Martens, D.A., Frankenberger, W.T., 1993. Soil saccharide extraction and

detection. Plant and Soil 149, 145–147.

Minderman, G., 1968. Addition, decomposition and accumulation of

organic matter in forests. Journal of Ecology 56, 355–362.

Nelson, D.W., Sommers, L.E., 1996. Total carbon, organic carbon, and

organic matter. In: Sparks, D.L. (Ed.), Methods of Soil Analysis. Part

3. Chemical Methods. ASA-SSSA, Madison, USA, pp. 961–1010.

Oades, J.M., Kirkman, M.A., Wagner, G.H., 1970. The use of gas–liquid

chromatography for the determination of sugars extracted from soils

by sulfuric acid. Soil Science Society of America Proceedings 34,

230–235.

Paul, E.A., Follett, R.F., Leavitt, S.W., Halvorson, A., Peterson, G.A.,

Lyon, D.J., 1997. Radiocarbon dating for determination of soil

organic matter pool sizes and dynamics. Soil Science Society of

America Journal 61, 1058–1067.

Pelz, O., Abraham, W.R., Saurer, M., Siegwolf, R., Zeyer, J., 2005.

Microbial assimilation of plant-derived carbon in soil traced by isotope

analysis. Biology and Fertility of Soils 41, 153–162.

Poirier, N., Derenne, S., Rouzaud, J.N., Largeau, C., Mariotti, A.,

Balesdent, J., Maquet, J., 2000. Chemical structure and sources of

macromolecular, resistant, organic fraction isolated from a forest soil

(Lacadee, south-west France). Organic Geochemistry 31, 813–827.

Ponomarenko, E.V., Anderson, D.W., 2001. Importance of charred

organic matter in Black Chernozem soils of Saskatchewan. Canadian

Journal of Soil Science 81, 285–297.

Ribereau-Gayon, P., 1968. Les Composes Phenoliques des Vegetaux.

Dunod, Paris.

Rovira, P., Vallejo, V.R., 1997. Organic carbon and nitrogen mineraliza-

tion under Mediterranean climatic conditions: the effects of incubation

depth. Soil Biology & Biochemistry 29, 1509–1520.

Rovira, P., Vallejo, V.R., 2000. Evaluating thermal and acid hydrolysis

methods as indicators of soil organic matter quality. Communications

in Soil Science and Plant Analysis 31, 81–100.

Rovira, P., Vallejo, V.R., 2002. Labile and recalcitrant pools of carbon

and nitrogen in organic matter decomposing at different depths in soil:

an acid hydrolysis approach. Geoderma 107, 109–141.

Rovira, P., Vallejo, V.R., 2003. Physical protection and biochemical

quality of organic matter in Mediterranean calcareous forest soils: a

density fractionation approach. Soil Biology & Biochemistry 35,

245–261.

Sanchez, J.P., Lazzari, M.A., 1999. Impact of fire on soil nitrogen forms in

Central Semiarid Argentina. Arid Soil Research and Rehabilitation 13,

81–90.

Schmidt, M.W.I., Skjemstad, J.O., Gehrt, E., Kogel-Knabner, I., 1999.

Charred organic carbon in German chernozemic soils. European

Journal of Soil Science 50, 351–365.

Schmidt, M.W.I., Skjemstad, J.O., Czimczik, C.I., Glaser, B., Prentice,

K.M., Gelinas, Y., Kuhlbusch, A.J., 2001. Comparative analysis of

black carbon in soils. Global Biogeochemical Cycles 15, 163–167.

Schnitzer, M., 1990. Selected methods for the characterization of soil

humic substances. In: P. MacCarthy, P., Clapp, C.E., Malcolm, R.L.,

Bloom, P.R. (Eds.), Humic Substances in Soil and Crop Sciences:

Selected Readings. ASA-SSSA, Madison, USA, pp. 65–89.

Skjemstad, J.O., Reicosky, D.C., Wilts, A.R., McGowan, J.A., 2002.

Charcoal carbon in US agricultural soils. Soil Science Society of

America Journal 66, 1249–1255.

Stanford, G., 1982. Assessment of soil nitrogen availability. In: Stevenson,

F.J. (Ed.), Nitrogen in Agricultural Soils. ASA-SSSA, Madison, USA,

pp. 651–688.

Stevenson, F.J., 1982. Humus Chemistry. Wiley, New York.

Stout, J.D., Goh, K.M., Rafter, T.A., 1981. Chemistry and turnover of

naturally occurring resistant organic compounds in soil. In: Paul, E.A.,

Ladd, J.N. (Eds.), Soil Biochemistry, vol. 5. Marcel Dekker, New

York, pp. 1–73.

Tan, Z.X., Lal, R., Izaurralde, R.C., Post, W.M., 2004. Biochemically

protected soil organic carbon at the North Appalachian experimental

watershed. Soil Science 169, 423–433.

Winkler, M.G., 1985. Charcoal analysis for paleoenvironmental inter-

pretation: a chemical assay. Quaternary Research 23, 313–326.

labile, recalcitrant, and inert organic matter in mediterranean forest soils

Documents