Organic Matter Changes Immediately After a Wildfire Fernandez

7/30/2019 Organic Matter Changes Immediately After a Wildfire Fernandez

1/11

Soil Liiol. Ei0ch.m. Vol. 29, No. I, pp. 1-I 1, 19970 199 7 Elsevier Science Ltd. All rights reservedPII: soo38-0717(!X)oo28!M Printed in &at Britain0038-0717/9717.00 + 0.00

ORGAN IC MATTER CHANG ES IMMEDIATELY AFTER AWILDFIRE IN AN ATLANTIC FOREST SOIL AND

CO:MPARISON WITH LABORATO RY SOIL HEATINGI. FERNANDE Z, A. CABANEIRO and T. CARBALLA S*

Instituto de Investigaciones Agrobiol6gicas de Galicia, C.S.I.C. Apartado 122, 157 80 Santiago deCompostela, Spain(Accepted 6 September 1996)

Summary-The quantity, chemical composition and mineralization kinetics of the organic matter of anacid Humic Cambisol, developed over granite, under Pinus sylvesfris L. were determined in O-5 and 5-10 cm samples collected im mediately after a high-intensity wildfire, and compared with those of anunaffected riite nearby. Organic m atter w as characterized by different chemical fractionation methods,and the C mineralization was determined by aerobic incubation. A similar unburnt soil located in thesame area was heated at the laboratory at 150, 220, 350 and 49 0C to measure the losses of C content;the samples heated at 220 and 350C were selected to determine chemical changes in organic mattercomposition. Surface and subsurface soil layers lost about 50% of their C content during w ildfire. TheC mineralized decreased in the surface layer; however, the percentage of total C mineralized increasedin both layers. The cumulative CO*-C mineralized fitted a double exponential first-order kinetic model,but the fire affected the kinetic parameters, increasing both the labile pool of the potentially mineraliz-able C and the mineralization rate constants of the recalcitrant and labile pools. Cellulose + hemicellu-loses declined significantly after the burn ing, whereas lipids did not vary. The fire decreased the amountof unhumifised organic matter and the alkali-soluble compounds, particularly hmnic acids, but there wasa net increase of humin. The organic matter bound to Fe and especially to Al was much higher afterthe burning. In the soil heated und er laboratory conditions the changes observed at 150 C were verylow, whereas at 490C almost all the organic m atter disappeared. The changes exhibited by the samplesheated at 2:!OCwere the most similar to those observed in the samples from the wildfire. At 220 and350C the humification and metal complexation percentages of the organic matter increased, similar tothe trend observed in the samples from the wildfire. 0 199 7 Elsevier Science Ltd

INTRODUCTIONThe changes induced by fires on vegetation and soilcharacteristic s have been studied for a long time.Vegetation fires are considered to be a naturalphenomenon in many parts of the world, whererepeate d fires have resulted in maintaining particu-lar ecosystems (Ahlgren and Ahlgren, 1960). Insuch areas controlled fires are widely used to reduceaccumu lated fuel and thus limit the risk of high-intensity wildfires. Rep orts are available on con-trolled fires (Chandle r et al., 1983; Khanna andRaison, 1986; Marion et al., 1991) and soil heatingunder laboratory conditions (Kang andSajjapong se, 1980 ), and their influence on soil prop-erties such as pH, loss and availability of nutrients(Feller, 198 8). Studies on high-intensity wildfires arescarce and, moreover, most were undertaken aftersome time had lapi3ed since the fire.

The northwest of Spain, which is located in thetemperate-h umid climate zone, contains soils whic hare rich in organic matter, and are acidic and

*Author for correspondence.

sandy. Th is region had been considered in the pastto be a relatively low-risk wildfire area, because ofits climate and the nature of floristic composition ofits forests. Deciduous species such as oak, chestnut,birch and alder in the natural forests of this areaposs ess an intrinsic high resistance to fires. Duringthe last two centuries the landscape of this regionwas greatly modified by human activities whichincluded growin g evergreen trees, particularly resi-nous, typ es which are more vulnerable to fire.Consequen tly, in recent decades the number ofwildfires has dramatically increased. These firesadversely affect vegetation cover and soils. In som ecases soil erosion p roduced irreversible damage,with consequent economic losses and landscape de-terioration. Diaz-Fierros et a l . (1982, 1987) foundthat the erosion risk increased considdrably duringthe first year after a fire. These autho rs reported anaverage soil erosion of 56 t ha- for this period, therisk declining exponentially to negligible amountsafter 12 months.

In sandy soils the main factor of cohesion andprotection against erosion when the fires destroythe vegetation cover is a high content of organic


2/11

2 1. Fernandez et al .matter whic h, in this area, is maintained even afterhigh-intensity wildfires (Bara, 198 1). The organicmatter also produces th e adequate conditions forthe establishment of new vegetation contributing tosoil regeneration. Studies on the effects of fires onorganic matter content are quite common (e.g.Kang and Sajjapong se, 1980 ; Giovannini et al.,1987; Kutiel and Kutiel, 1989); however, those onchanges in organic matter composition (Shindo andUrabe, 1993), structure (Almendros et al., 1992)and biological characteristics (Vazquez et al., 1993;Fritze et al., 1994; Acea and Carballas, 1996) areless so. There are studies on organic matter compo-sition and mineralization in the soils of the north-west of Spain (Carballas et al ., 1979, 1980, 1983;Beloso et al., 1993) but until now the effect of fireson these characteristics had not been studied.

Organic matter is responsible for forest ecosys-tems productivity because o f its nutrient contentand its influence on physico-ch emical and biologicalcharacteristics. Because of this the aim of the studywas to evaluate the effects of fire on the compo -sition, stability and C mineralization of the soil or-ganic matter immediately after a high-intensitywildfire.

MATERIALS AND METHODS

The soils selected for study were two HumicCambiso ls under a pine forest of Pinus sylvestris,developed over granite and located in Cabeza deManzaneda (Orense, northwest Spain) at 1740 mabove sea level. The undergrowth consisted o fChamaespartium tridentatum (L.) P. Gibbs,Vaccinium myrt i lus L., Erica spp and species ofGramineae. The burnt soil was samp led inSeptem ber, 1 d after a high-intensity wildfire; simul-taneously an unburnt soil from the same edaphicand pine stand located close to the burnt area wassampled and used as a control. Soil samples w erecollected at O-5 and 5-10 cm depth all along a con-tinuous profile 10 m in length. They were sieved(4 mm), and the fraction less than 4 mm was hom-ogenized. Fresh samples were used for the study ofC mineralization. Air-dried sample s were employ edfor the determination of organic ma tter and theother soil characteristics.

For the experime nts on soil heating under labo ra-tory conditions a different unburnt soil from thesame area, but far from the burnt zone, was used.The sample w as collected in October, in the sameway as the others from O-5 cm, using a subsampleas control. The soil was heated in an oven with atimer and a heating rate programme. Temperaturein the soil was controlled by a thermocouple locatedinto the bulk of soil. Conditions of heating were asfollows: an increase of 3C min- to prevent thesudden combustion produced when reaching thesoil ignition temperatu re; 30 min of heating after

reaching the selected heating temperature; and asample size of less than 1 cm depth to avoid a tem-perature gradient in the profile. The soil was treatedat the ignition temperature (220C; Salgado et al.,1995) and at temperatures lower (15OC) and higher(350 and 490C) than the ignition temperature.Losses of organic matter were determined in all thesamples. A ccording to these losses the samplesheated at 220 and 350C were selected for sub-sequent organic matter fractionation. The samefractionation procedure was followed in the controlsample.

The methods described by Guitian Ojea andCarballas (1976) were used to determine the follow-ing soil properties : field capacity (at 1 0 KPa in aRichards membrane-plate extractor); pH in Hz0(1:2.5); pH in 1 N KC1 (1:2.5); extractable Fe andAl oxides (by extraction with a mixture of hydrosul-phite and Tamms reagent). C content was deter-mined by combustion and measurement of the CO2in a Carmhog raph 12 with the primary oven at900C and the secondary oven at 400C. The factorused to convert C into organic m atter was 1.724(Guitiin Ojea and Carballas, 197 6). Total N con-tent was determined by Kjeldahl digestion using themethod developed by Bremner (1965). The resultsare the average of three replicate determinations.

Analysis of C-containing compo unds was per-formed by the method of Kononova (1961) excep tfor water-soluble compounds that were extracted asrecommended by Stevenson (1965). The water-sol-uble fraction was extract ed by refluxing with boilingwater for 2 h. Lipids were extracted from the resi-due with ethanol-benzene for 24 h in a Soxhlet. Theresidue was treated with HzS04 to solubilize cellu-lose and hemicelluloses leaving lignin and mineralmaterial as a residue. After ignition of this residuein an oven at 50 0C for 6 h the lignin content wascalculated by difference from th e final residueweight (Gonzalez-Prieto et al., 1993). The resultswere expressed as percentage of either soil weightor total organic matter.

To study the humified organic matter of the O-5 cm soil layer a modified method of Duchaufourand Jacquin (1966) was used. After a densiometricseparation of the unhumified and the humified or-ganic matter using bromo form-ethano l with a den-sity of 1.8, the humified o rganic com pounds wereextracted successively by 1% Na4 P207 (pH 9.8) and0.1 N NaO H (pH 13). The extraction was performedby agitating soil with the correspond ing reagent fo r1 h, centrifugation and repetition of the cycle an ad-ditional time. The final residue, whic h was com-posed o f the humin fraction (H), was dried at WC,ground and homogenized. In each extract humicacids (HA) were separated from fulvic acids (FA)by precipitation at pH 1 with HzS 04 and centrifu-gation after more than 4 h of standing (Gonzdlez-Prieto et al., 1993). C was determined in each


3/11

Soil organic matter changes after wildfires 3extract by Sauerlandts meth od (Guitiin Ojea andCarballas, 1976) ,which used Cr207K2 + H2S04 fordigestion and further titration with(NH& S04~FeS0 ,16H20. The residue C contentwas assessed using the Carmhog raph method. Feand Al bound to the organic m atter was assessedby determining these elements in the alkali-extractsby atomic spectrophotom etry. The per cent extrac-tion (CFA + HA total C-l), per cent humification(C,, + HA + H total C-l) and indices of polymeriz-ation, fulvic acids to humic acids ratio (FA-to-HA),and metal comp lexation, (Fe + Al)-to-C ratio, werecalculated.

Fresh samples or samples stored at 4C for lessthan 7 d were used for the determination of the Cmineralization kinetics by aerobic incubation of theunburnt and the burnt soil sample s from the wild-fire. The incubation method was that of Guckert etal. (1968) modified by Chant e t a l . (1974), whoused an intermittent air flow for aeration. Fivereplicates of 50 g soil from each sam ple were placedinto 50 0 ml Erlenmeyer flasks and maintained at28C and m oisture at 75% of field capacity for11 wk. The flask atmosph eres were periodicallyrenewed with CO z-free air, and the CO* evolvedwas trapped by bubbling for 2 h in 40 ml of 2 NNaO H. From 5 ml NaO H solution, the Na$03formed was precipitated by 20% BaCl* solutionand then the remaining NaO H was measured by ti-tration against 0.2 N HCl using a Metrohm 682titrator. The CO2 was then quantified by subtrac-tion, using two empty flask s incubated under thesame conditions as the control (Carballas e t a l . ,1979; Gondlez-Prieto et uf., 1991; Beloso e t a l . ,1993). After 77 d of incubation, pH in water andKC1 were determined in the burnt and th e unburntsamples. T he C mineralization, expressed as g CO*-C evolved kg- so:il, and the C mineralization coeffi-cient expressed as a percentage of the total C werecalculated.

The cumulative curves of C02-C release overtime followed the double exponential model pro-posed by And& and Pauskian (1987) andemployed by Updegraff et a l . (1995) to quantify the

kinetics of C release from different soils underdifferent conditions:

C, = Co(l -e-k)+(TC-Co)(l -eeh) (1)where C, is the cumulative C released after time t(days), Co is the potentially mineralizable C in alabile pool with a instantaneous mineralization ratek (d-l), TC is the total amou nt of C present in thesoil sample (g kg-) and (TC - CO) s the amount ofC in a recalcitrant pool with a instantaneous miner-alization rate h (d-l). This model explains the over-all C mineralization proce ss by considering twopools o f C that dec ay independently and in parallel,with different degradation rates. The non-linear par-ameter estimation procedures in SPSS.5.0.1. forWindows were used to fit the experimental data tothe model. To avoid errors in the parameter esti-mation the convergence criteria indicated byUpdegraff et al. (1995) were used.Two-way ANO VA test was applied to weekly Cmineralization data to follow the kinetic variationsas well as to quantify the variation of C mineraliz-ation attributable to the burnt soil and to the lo-cation of the layer in the profile. The lea stsignificant difference (LS D) test (Tukey, 1953) atthe 95 and 99 % probability levels was applied tothe results. The x2 test was used for comparison ofthe results obtained in the different org anic m atterfractionations.

RESULTS AND DISCUSSIONTable 1 show s the main characteristics of the soil

sample s. The unburnt soil varied little in depth , wa shighly acid, had a high content of organic C and N ,high values of cation exchange capacity (CEC),with a low percentage of base saturation and amoderate content of extractable Fe and Al oxides.

Immediately after the fire the increase of pH w asmoderate (Table 1) due to the high value of thebuffer capacity index (Guiti6n Ojea and Carballas,1976) of this soil (5.99) and the short time that hadelapsed after the fire. There fore, it is probably thatpH had not yet reached its maximum value. This

Table 1. Main characteristics of the unburnt (UB), the burnt (B), the unheated (UH) and the heated (H) soil samples (mean values & SD)Sample Depth (cm) pH Hz0 pH KC1 C (g kg-) C-to-N N (g kg-) FazOl AlzOj CEC Base

(g kg- ) (g kg- ) (meq 100 saturationg-9 (%)UB. O-5UB, 5-10B. O-SB, 5-10UH O-5H% O O-5HHZ

O-5O-5

Hi90 O-5

4.19 f 0.024.25 f 0.034.74 f 0.034.58 f 0 .034.35 * 0.034.00 f 0.024.05 f 0.025.70 f 0.056.35 f 0.04

3.29 f 0 .013.45 f 0.014.00 f 0.003.78 f 0.033.30 * 0.013.30 * 0.033.15 f 0.024.75 f 0.035.45 f 0.05

158.2 f 1.8111.0 f 1.8

74.4 * 1.051.1 *2.1

102.9 * 1.1105.1 f 0.1

64.8 f 0.110.7 f 0.00.3 * 0.0

18 8.73 f 0.1018 6.30 f 0.1415 4.87 f 0.1515 3.31 f 0.0719 5.49 f 0.0619 5.50 f 0.0212 5.40 * 0.018 1.40 * 0.01ND 0.10 f 0.00

5.5*0.16.5 kO.14.9 f 0.34.7 * 0.05.5 * 0.15.6 k 0.15.0 * 0.25.4 f 0.15.6 f 0.1

6.3 f 0.4 51.8 f0.1 5.7 f 0.47.8 f 0.5 49.7 f 1.5 3.3 f 0.46.1 k 0.3 24.2 f 1.5 13.2 f 0.36.1 kO.1 33.8 f 1.1 3.6 k 0.26.3 zt 0.1 43.9 * 0.4 3.1 f 0.06.8 k 0.9 ND ND6.2 f 0.0 ND ND7.1 fO.O ND ND8.6rtO.1 ND ND

Heating temperatures (C).ND = not determined.


4/11

4 I. FernPndez et al .Table 2. Carbon mineralization ap acity (mean values f SD), kinetic param eters or C mineralization model based on equation I(estimate values f asymptotic SE) and pH of the surface (s) and subsurface layer (ss) of the unburnt (UB) and the burnt (B) soil samples

after 11 weeks of incubation (mean values f SD)

Sample

Kinetic parameters PHPercentage ofMineralized C C mineralized

(g kg-) (g C 100 g&C,) G (gkg-) k hx lo4 R2 Hz0 KC1UBs 3.41 * 0.09 2.15 k O.OSb 0.67 + 0.02 0.105 f 0.005 2.29 f 0.02 0.9998 4.00 f 0.0 3.15f0.02UBSS 1.97 0.078 1.77 + 0.06 0.34+0.01 0.159 * 0.011 1.94 f 0.02 0.9995 3.95 f 0.0 3.26 f 0.02B, 2.58 + 0.07b 3.47 f o.loc 1.06 k 0.02 0.184 f 0.008 2.80 f 0.06 0.9986 5.12+0.0 4.05 * 0.01BW 1.99 * 0.06 3.89 k 0.12d 0.48 k 0.01 0.173 f 0.008 3.98 It 0.04 0.9996 5.00 f 0.0 3.45 f 0.03-%alues within columns with dissimilar letters are significantly different (P < 0.01).Co, potentially mineralizable C of the labile pool.k, instantaneous mineralization rate of the labile C pool.h, instantaneous mineralization rate of the recalcitrant C pool.

agrees with the higher p H in water exhibited by theburnt soil after incubation which was not found inthe unburnt soil (Table 2). Chand ler et al. (198 3)attributed the increase of pH after burning not onlyto fire intensity, but also to soil characteris tics.Ulery et al. (1993) gave the loss of organic acidsduring the fire and the addition of hydroxides andcarbonates by the ashes as some of the reasons forthe increase in pH. Th e percentage of base satu-ration increased and the CEC d ecreased as a conse-quence of the disappearance of organic matter dueto fire (Table 1).

In the soil heated at 150C there were no signifi-cant changes in most of the soil characteristicsexcept a low decrease of pH which was alsoobserved at 220C (Table 1). At 350C the pH wasmore than one unit higher and at 490C it was twounits high er than that o f the control soil. The lowdecrease o f pH at the lower temperatures as well asthe high increase at the higher temperatures agreedwith the findings of Nishita and H aug (1972),Giovannini et al . (1990) and Diaz-Ravifia et al.(1992). Extractable Al oxides also decreased at22O C, which was similar to that of the wildfire,and sho wed an increase at 350 and 490C . In thecase of Fe, a decrease was only observed in the soilsample heated at 220C.

Organic matter contentThe C content had decreased considerably after

the fire. The decrease was approximately 50% inthe whole depth studied (Table 1); however, it islikely that C losses were even higher in the surfacelayer, which was much more morphologicallyaltered, and that they had been partially compen-sated by deposition of plant material not completelyincinerated. Chandler et al. (1983) reported thegreat variability of the effect of fire on organic mat-ter content ranging from a total destruction of theorganic matter due to wildfires to increases that canreach 30 % in the surface layers as a consequence oforganic material redistribution in the proflle. Theloss of C was accompanied by a lower N decreaseand consequently the C-to-N ratio also decreased

from 18 (characteristic of a moder forest humus) to15 after the fire. The soil heated at 150C did notexhibit a significant loss of C, whereas at 490Calmost all the organic matter had disappeared(Table 1). In the soil heated at 22O C, 37% of Cwas lost and at 350C the loss was 90% of the totalC. The C-to-N ratio dramatically decreased in thesoil heated at 22o C, declining to a value of 8 in thesoil heated at 350C.

C mineralization activityFigure 1, which show s the cumulated percentage

of C mineralized, indicates that the wildfire notonly affected the percentage of C mineralized by theend of incubation, but also altered the C mineraliz-ation kinetics du ring the incubation.The C mineralization coefficient was relativelylow (x2% of total C after 11 wk) in the unburntsoil. At the end of the incubation this coefficient aswell as the total amount of C mineralized were stat-istically different (PI 0.01) in both soil layers(Table 2). Low values for endogenous C mineraliz-ation coefficient are usual in soils over granite ofsimilar characteristics from the same area(Carballas et al., 1979; Beloso et al., 1993).The C02-C evolved from the surface layer of theburnt soil was significantly lower (PI 0.01) thanthat o f the same layer from the unburnt soil(Table 2). This result agrees with that found byAlmendros et al. (1984a) for soils under a mixedforest, collected 2 y after the fire. The C mineraliz-ation coefficient in the surface layer o f the burntsoil was significantly highe r (PI 0.05) than that ofthe correspond ing layer of the unburnt soilthroug hout the incubation period (Fig . 1) as a con-sequence of the high loss of C after the fire. In thesubsurface soil, fire did not significantly affect theamount of organic matte r mineralized (Table 2).Neve rtheless, like the surface soil, the C mineraliz-ation co efficient increased significantly (P 50.01 )and after 11 wk incubation was twice that of theunburnt soil (see Fig. 1 and Table 2).After burning, the surface soil exhibited a highinitial loss of COI-C that declined sharply after


5/11

Soil organic matter changes after wildfires

L.S.D. (PsO.05)

-0 7 14 21 28 35 42 49 58 53 70 77DAYS

Fig. 1. Cumulative C mineralization co efficient during incubation of the O-5 and 5 -10 cm layers of theunburnt (IJB) and the burnt (B) soils just after the w ildfire. Leas t significant difference (PI 0.05)shown as bars at week intervals.14 d of incubation; by this period the slope of thecurve diminished and was practically parallel tothat o f the control soil (Fig. 1). The subsurfac e soilalso exhibited an initial high value of CO z evolved,although lower than that observed in the O-5 cmlayer. After 42 d the CO 2 evolved by the subsurfacesoil was higher than that evolved by the surfacesoil, the difference between both values were signifi-cant from week 8 (PI 0.05) until the end of the in-cubation (PI 0.01). The two-way AN OVA testindicated a significant interaction between theburnt-unburnt factor and the depth of soil (O-5 cm/%10 cm).

The cumulative! curve of COz-C evolved by thedifferent so il sample s significantly fitted the first-order equation proposed by Stanford and Smith(1972), which was successfully used by Zak et al .(199 3) to describe th e mineralization kinetics innative soils and b:y Prieto-Fernandez et al . (1993) innative and b urned soils. Ho wev er, as in the case ofDiaz-Fierros et al . (198 8) for N mineralization andUpdegralI et al . (1995) for C and N mineralizationkinetics, the CO& data evolved by the soils stu-

died show ed a superior fit to a double exponentialmodel; the kinetic param eters obtained using themode1 proposed by Andrtn and Pauskian (1987)are shown in Table 2. The values of these par-ameters showed the differences between the kineticsof the burnt and unburnt soils. The fire increasedthe potentially mineralizable C of the labile pool,which was 1.6- and 1 Cfold higher in the surfaceand the subsurfac e layer, respectively, of the burntsoil than th at in the correspond ing layers of theunburnt soil. In genera1 the mineralization rate con-stants w ere also increased by the burning. For thelabile pool the kinetic con stant (k) was multipliedby 1.8 in the surface layer of the burned soil,whe reas it did not vary in the subsurface layer. Incontrast, for the recalcitrant pool the mineralizationrate (h) duplicated after the fire in the subsurfacelayer and increased only 1.2-fold in the surfacelayer.

C-containi ng compoundrThe organic m atter content in the surface and the

subsurface soil layers from unburnt site was 27.3


6/11

I. FernPndez et al .

- - 25%

0UNBURNT BURNT UNBURNT BURNT

SOIL SOIL SOIL SOIL\ / \ /

0-5cm 5-10cmFig. 2. Content (foreground of the figure) and changes in the percentage distribution (background) ofthe main groups of C-containing comp ounds of the organic matter from the O-5 and 5-10 cm layers ofthe unburnt and the burnt soils.

and 1 9.1%, respectively, and the composition of theorganic matter was similar in both soil layers asshown by the percentage distribution of the differ-ent fractions (Fig. 2). Lignin, whic h representedmore than 50% of the total organic matter, was thepredominant fraction follo wed by cellulose + hemi-celluloses and a much sm all proportion of lipidsand water-soluble compounds that did not reach10% of the total organic m atter. Therefore themost recalcitrant fractions to microbial attack pre-domina ted. This finding a greed with the C mineral-ization activity of this soil.

Burning affected all the organic matter fractionsstudied, althoug h to a different ex tent (Fig. 2). Inthe surface soil, 72% of the cellulose + hemicellu-loses disappeared, this fraction being the mostaffected by the fire followed by the water-solublecompounds (56% of losses), lignin (46%) and lipids(25%). The same trend was observed in the subsur-face soil, but the disapp earance of cellulose + hemi-celluloses w as higher (81 %) and that of water-soluble compounds lower (46%), lignin andlipids presenting similar losses to thos e of the

surface layer (46 and 26% , respectively). Thereforethe fire increased the more labile fractionsand decreased the more resistant fractions(lignin + cellulose + hemicellulose s). The x2 testindicated no significant d ifferences between the sur-face layers from the burnt and unburnt soils, butsignificant differences in subsurface layers(PI 0.05). Almendros et al. (1988) observed thatfire increased the proportion of the shorter chainmolecules of the lipid fraction, whic h may explainthe increase of the C mineralization coefficient inboth surface and subsurface soils immediately afterthe fire.

Although the number of samples is not enoughto show possible correlations among the kineticparam eters (Table 2) and the different fractions ofthe organic matter, there was a high correlation(r = 0.9420) between the content of the water-sol-uble compo unds and the C mineralization rate ofthe labile C pool (k), although this was at a lowlevel of significance (P = 0.0 58).

Heating of the soil in the laboratory had similareffects on the distribution of the C compounds as


7/11

35

30

/

I

,

,

,

i

/

Soil organic matter changes after wildfires I% O.M.

UNtlEAl-ED SOIL SOIL HEATED SOIL HEATEDAT 22ooc AT sow

100%

75%

25%

Fig. 3. Content (foregroun d of the figure) and chang es in the percentage distribution (backgrou nd) ofthe main groups of C-containing compounds of the organic matter from the surface layer of theunheated soil or soils heated at 220 and 350 C in the laboratory.

those of wildfire (Fig . 3). At 220C lignin, wh ich 50% was quite uniformly distributed between fulvicwas the mos t resistant fraction to the wildfire, only acids and humin. The extraction and humificationdecreased by 11%. Lipids declined in the same pro- percentages as well as the FA-to-HA ratio (Table 3)portion, whereas cellulose + hemicelluloses w as the were of the same order of magnitude as thosefraction most affected by the heating-85% of this found for other soils of the same area located atfration disappearing during the heating process. At similar altitude and topo graph y (Carballas et al.,350C when only 10% of the initial organic m atter 1983). The organic matter was quite well humifiedremained, lipids were the least affected fraction and moderately polymerized with a predominancebecause only 33% of them disappeared, whereas of HA according to the FA-to-HA ratio. The90% of water-solu.ble compoun ds, cellulose + hemi- extractable Fe and Al content of the different frac -celluloses and lignin fractions were destro yed. The tions was not very high, which implies a low degreepercentage distribution clearly showed the decrease of complexation of the organic m atter. The extrac-of cellulose + hemicelluloses and the increase oflipids when the temperature increased. The x2 test table Al content in each fraction was always higherthan that of extractable Fe, agreeing w ith the find-showed that the organic matter composition of the ings of Carballas et al. (1980) who studied thesoils heated at :!20 or 350 C were significantly organo-metallic complexes in soils developed overdifferent (P I 0 .05) from that of the unheated soil. granite from the same region.Humus

Figure 4 show s the percentage composition of thehumified organic matter as well as the C contentfor the different fractions expressed as a percentageof dry soil. In the unburnt soil, althoug h the con-tent of unhumified organic m atter w as high, thehumified fraction clearly predominated.Approximately 50% of the humified organic mattercorresponded to humic acids, whereas the other

The decrease of organic matter after fire washigher for the unhumified than for the humifiedfraction, and consequently the percentag e of humifi-cation dramatically increased (see Fig. 4 andTable 3). There was a net increase of the huminfraction, as also observed by Almendros et al.(1990) n an oak forest affected by a fire. Huminwas the predominan t fraction in percenta ge d istri-bution of the humic compounds, whereas HA andFA fractions decrea sed not only in percenta ge dis-


8/11

8 I. FernPndez et al.

UNBURNT SOIL

liiiidh.0.u. 1l.rmL - .__-* BURNT SOIL

Fe

Fig. 5. Fe and Al bound to humic (HA ) and fulvic (FA)acids from the O-5 cm layer of the unburnt and the burntsoils.Fig. 4. Content and distribution of the different humicfractions from the O-5cm layer of the unburnt and the

burnt soils.

tribution, but also in absolute amoun ts-thedecrease being higher for HA. The effect of the fireon the organic matter was similar t o a climaticmaturing (Almendros et al., 1984b) becausealthough the extraction percentage decreasedslightly, the percentag e o f humification increasedconsiderably as well as the degree of polymeriz-ation, all pointing to a more stab ilized organic mat-ter. The fire increased the Fe and Al bounded toorganic matter (from 39 to 91 mg Fe g- C and

from 79 to 223 mg Al g- C considering the FA andHA fractions together) (Fig. 5). The complexationcapacity of FA was higher than that of HA; more-over organo-Al comp lexes predominated overorgano-Fe com plexes. The se findings also contribu-ted to an increase in organic matter stabilization(Jacquin et al., 1978; Cabaneiro et al., 1987).

Heating the soil in the laboratory at 220Cresulted in a decrea se of humified and unhumifiedorganic matter, the losses being 22 and 55% of thetotal organic matte r, respectively (Fig. 6). Althou ghHAS were still the predominant fraction, theyunderwent the highest decrease, in contrast to

Table 3. Characteristics of the humified organic matter after a high intensity wildfire and after heating under laboratory conditions

Sample Depth (cm)Percentage

Percentage extraction humification(C(FA + HA,C ;I) (C,,, + HA + H)C ;I) FA-to-HA (Fe + AI)/C

UB. O-5 51 61 0.58 0.12B. &5 41 81 0.57 0.32UH O-5 44 54 0.46 0.16

:: 48 670 0.52.60 12.89.29

UB, unburnt sample;B, burnt sample; UH, unheated sample; H, heated sample.Heating temperatures (C).FA, fulvic acids; HA, humic acids; H, humin.


9/11

Soil organic matter changes after wildfires 9

n: o((

Minml fmc lbn82.26%

. *: . .Unliuk .11..l_,y . . . . * . - . t l k s mb5.12 gc Img b.CHqm. O.M. 6 .6%

UNHEATED SOIL

was similar to that found for the wildfire, whereasat 350C it was considerably increased. Neverthelessthe high increase of organic m atter bound tometals, particularly to Al (Al-to-C = 12.29 )observed at 35O C, could not be a real effect butrather a conseque nce of the alteration of Al inor-ganic forms that became soluble at high pH (NaOHat pH 13, in which fraction 93% of extractable Alwas obtained). This did not occur for Fe complexa-tion capacity since at 350 C the Fe-to-C ratioaccounted for only 0.60.

l-The x2 test showed that the percentage variation

of the different humus fractions from the wildfiresoil was significantly different (PI 0.05) from thatof the unburnt soil. Comparison of the unheatedsoil with the soil heated at 220C at the laboratoryshow ed that their compo sition were not significantlydifferent, w hereas the composition of the organicmatter of the soil heated at 350C was significantlydifferent (P I 0.01 ) from that of the unheated soil.

_ . - . -A: 47%z.wgC~ODpd*SOIL HEATED AT 220C

Acknowledgements-The authors thank Mr J. Salmonteand MS B. Am& for technical assistance. This researchwas supported by Conselleria de Educaci6n y Ordenaci6nUniversitaria de la Xunta de Galicia and by Comisi6nInterministerial de Ciencia y Tecnologia (CICY T), Spain.

REFERENCES

SOIL HEATED AT 350C

Fig. 6. Content and percentage distribution of the differenthumic fractions from the surface layer of the unheated soilor soils heated at 220 and 350C in the laboratory.

humin w hich increased; h owever, humin was nothigh enough to be the predominant fraction as itwas after the wildfire. At 350C the loss of organicmatter was so high that the humin fraction was theonly component of the organic matter thatremained after the heating. Com plexation of the or-ganic matter with metals followed the same trend asin the case o f the wildfire (Table 3). It increasedafter the heating, the association with Al being alsofavoured. At 220C the increase of complexation

Acea M.J. and Carballas T. (1996 ) Changes in physiologi-cal groups of microorganisms in soil following wildfire.FEMS M icrobiology Ecology 20, 33-39.Ahlgren Y.F. and Ahlgren C.E. (196 0) Ecological effectsof forest fires. The Botani cal Review 26, 483-533.Almendros G., Polo A., Ibaiiez J.J. and LoboMC. (198 4a) Contribuci6n al estudio de la infuencia de10s incendios forestales en las caracteristicas de la mate-ria organica de1 suelo. Revue d kcologi e et Biologic duSol 21, 7-20.Almendros G., Polo A., Lobo M.C. and IbailezJ.J. (1984b) Contribucibn al estudio de la infuencia de10s incendios forestales en las caracteristicas de la mate-ria orgLnica de1 suelo. II. Transformaciones de1 humuspor ignic$m en condiciones controladas de laboratorio.Revue d Ecologi e et Biologic du Sol 21, 145-160.Almendros G., Martin F. and Gonzilez-Vila F.J. (198 8)Effects of fire on humic an d lipid fractions in a dystricxerochrept in Spain. Geoderma 42, 115-127.Almendros G., Gonzllez-Vila F.J. and Martin F. (199 0)Fire-induced transformation of soil organic matter froman oak forest: an experimental approach to the effectsof fire on humic substances. Soil Science 149, 158-168.Almendros G., GonzLlez-Vila F.J., Martin F., Friind R.Liidemann H.-D. ( 1992 ) Solid state NM R studies offire-induced changes in the structure of humic sub-stances. The Science of th e Total Environmenr 117, 63-74 .And& 0. and Pauskian K. (1987) Barley straw decompo-sition in the field: a comparison of models. Ecology 68,1190-1200.Bar6 S. (1981 ) Evolucibn de la composici6n de 10s suelosincendiados. I Jomadas TCnicas sobre prevenci6n deincendios forestales. Lou&&n, Pontevedra.Beloso MC., Villar M.C., Cabaneiro A., Carballas M.,Gonz&lez-Prieto S.J. and Carballas T. (1993) Carbonand nitrogen mineralization in an acid soil fertilized


10/11

10 I. Fernandez et al.with composted urban refuses. Bioresource Technology45, 123-129.Bremner J.M. (19 65) Nitrogen availability indexes. InMethods of Soil Analysis. Part 2. Chemical andMicrobiological Properties (C. A. Black, D. D. Evans, J.L. White, L . E. Ensminger, F. E. Clark, Eds), pp. 132 41345 . American Society of Agronomy, Madison.Cabaneiro A., Chont Th. and Andreux F . (198 7) The in-fluence of seasonal changes and clay mineral on the bio-degradation of 14Ct 5N labelled proteinaceous materialfrom Chlorella pyrenoidosa in an acidic soil. In CurrentPerspectives in Environmental Biogeochemistry (G.Giovannozzi-Sermanni and P. Nannipieri, Eds), pp. 57-74. Consigho Nazionale delle Ricerche, Rom a.Carballas M., Carballas T. and Jacquin F. (1979)Biodegradation and humification of organic matter inhumiferous atlantic soils. Anales de Edafologia yAgrobioiogia 38, 1699-1717.Carballas M ., Cabaneiro A., Guitian-Rivera F. andCarballas T. (1980) Organo-metallic complexes in atlan-tic humiferous soils. Anales de Edafologia y Agrobiologia39, 1033-1043.Carballas M., Carballas T., Cabaneiro A., Villar M.C.,Leiros M.C. and Guitian Ojea F. (1983) Suelos ACsobre granitos de Galicia (NO de Espaiia) con especialreferencia al ranker atlantico. III Fraction organica.Anales de Edafoiogia y Agrobiologia 42, 1781-1814.Chandler C., Cheney Ph., Thom as Ph., Trabaud L. andWilliams D. (1983). Fire in Forestry. Vol. I. Forest FireBehavior and Effects. Wiley, New York.Choni Th., Jacquin F. and Yagui M. (1974) Emploi de14C et 45Ca comme elements traceur de Ihum ification.Bulletin de IEcole Nationale Superieure dAgronomie etdes Industries Alimentaires de Nancy 15, 69-85.Diaz-Fierros F., Gil Sotres F., Cabaneiro A., CarballasT., Leiros de la Peiia M.C. and Villar CelorioM.C. (198 2) Efectos erosivos de 10s incendios forestalessuelos de Galicia. Anales de Edafologia yygrobioiogia 41, 627-639.Dial-Fierros F., Benito Rueda E. and Perez MoreiraR. (1987) Evaluation of the U.S.L.E. for the nredictionof erosion in burnt forest areas in Galicia (N.W. Spain).Catena 14, 189-199.Diaz-Fierros F., Villar M.C., G il F., Carballas M., LeirosM.C., C arballas T. and Cabaneiro A. (1988) Effect ofcattle slurry fractions on nitrogen mineralization in soil.Journal of Agricultural Science, Cambridge 110, 491-491.Diaz Raviiia M., Prieto A., Acea M.J. and CarballasT. (199 2) Fumigation-extraction method to estimatemicrobial biomass in heated soils. Soil Biology andBiochemistry 24, 259-264.

Duchaufo ur P. and Jacquin F. (19 66) Nouvelles recherchessur Iextraction et le fractionnement des composts humi-ques. Bulletin de IEcole Nationale SuperieuredAgronomie et des Industries Alimentaires de Nancy 8,l-24.Feller M.C. (19 88) Relationships between fuel propertiesand slashburning-induced nutrient losses. Forest Science34,998-1015.Fritze H., Smolander A., Levula T., Kitunen V. andMalkonen E. (199 4) Wood-ash fertilization and firetreatmen ts in a Scats pine forest stand: effects on the or-ganic layer, microbial biomass, and microbial activity.Biology and Fertility of Soils 11, 57-63.Giovannini G., Lucchesi S. and Giachetti M. (1987 ) Thenatural evolution of a burned soil: a three-year investi-gation. Soil Science 143, 220-226.Giovannini G., Lucchesi S. and Giachetti M. (1990)Effects of heating on some chemical parameters relatedto soil fertility and plant growth. Soil Science 149, 344-350.

Gonzalez-Prieto S.J., Carballas M . and Carballas T. (199 1)Mineralization of a nitrogen-bearing organic substratemodel 14C 15N -glycine in two acid soils. Soil Biologyand Biochemistry 23, 53-63.Gonzalez-Prieto S.J., Carballas M., Villar M.C., BelosoM.C.. C abaneiro A. and Carballas T. (1993) Carbon-and nitrogen-containing compounds in compostedurban refuses. Bioresource Technology 45, 115-12 1.Guckert A., Roger P. and Jacquin F. (1968) Contributiona Iutilisation de techniques radioisotopiques pourT&de de la matiere organique du sol. Bulletin deIEcoie Nationale Superieure dAgronornie et desIndustries Alimentaires de Nancy 10, 69-100.Guitian Ojea F. and Carballas T. (1976) Tecnicas deAnalisis de Suelos. Pica Sacro, Santiago de Compostela.Jacquin F., Carballas M. and Carballas T. (197 8)Interaction entre les ions aluminium et la mineralisationde la mat&e organique dans les sols humiferes atlan-tiques. Comptes Rendues de IAcademie de Sciences, pp.511-514. Paris 28 6, SCrie D.Kang B.T. and Sajjapongse A. (1980 ) Effects of heatingon properties of some soils from Sou thern Nigeria andgrowth of rice. Plant and Soil 55, 85-95.Khanna P.K. and Raison R.J. (1986 ) Effect of fire inten-sity on solution chemistry of surface soil under aEucalyptus pauciJora forest. Australian Journal of SoilResearch 24, 423-434.Kononova M. M. (1961) Soil Organic Matter. Its Nature,its Role in Soil Formation and in Soil Fertility.Pergamon, Oxford.Kutiel P. and K utiel H . (1989 ) Effects of a wildfire on soilnutrients and vegetation in an Aleppo pine forest onMount Carmel, Israel. Pirineos 134, 59-74.Marion G .M., Moreno J.M. and Gechel WC. (1991) Fireseverity, ash deposition, and clipping effects on soilnutrient in chaparral. Soil Science Society of AmericaJournal 55, 235-240.Nishita H. and Haug R.M. (1972) Some physical andchemic al characteristic cs of heated soils. Soil Science113,430442.Prieto-Fernindez A., Villar MC., Carballas M. andCarballas T. (19 93) Short-term effects of a wildfire onthe nitrogen status and its mineralization kinetics in anatlantic forest soil. Soil Biology and Biochemistry 25,1657-1664.Salgado J., Gonzalez MI., Armad a J., Paz-Andrade MI.,Carballas M. and Carballas T. (1995) Loss of organicmatter in Atlantic forest soils due to wildfires.Calculation of the ignition temperature. ThermochimicaActa 259, 165-175.Shindo H . and Urabe M. (1993) Changes in the humuscomposition of volcanic ash soils by heating at varioustemperatures. Soil Science and Plant Nutrition 39, 189-192.Stanford G. and Smith S.J. (197 2) Nitrogen mineralizationpotentials of soils. Soil Science Society of AmericaProceeding 36,465-472.Stevenson F. J. (1965 ) Gross chemical fractionation of or-ganic matter. In Methods of Soil Analysis. Part 2.Chemical and Microbiological Properties (C. A. Black,D. D. Evans. J. L. White. L. E. Ensminaer and F. E.Clark, Ed& pp. 1409-i414. American Society ofAgronomy, Madison.Tukev S. W. (1953) The Problem of Muitiole Comoarisons.Princeton University Press, Prin&ton. - -Ulery A.L., Graham R.C. and Amrhein C . (1993) Wood-ash composition and soil pH following intense burning.Soil Science 156, 358-364.Updegraff K., Pastor J., Bridgham S.D. and JohnstonA. (199 5) Environmental and substrate controls overcarbon and nitrogen mineralization in northern wet-lands. Ecological Applications 5, 151-163.


11/11

Soil organic m atter changes after wildfires 11VgZquez F.J., A cea M.J. and Car ballas T. (1993 ) Soil Zak D.R., Grigal D.F. and Ohm ann L.F. (199 3) Kineticsmicrobial populations after wildfire. FEMS of microbial respiration and nitrogen mineralization inMicrobiology Ecology 13, 93-104. Great Lakes Forests. Soil Science Society of AmericaJournal 57. 1100-l 106.

Organic Matter Changes Immediately After a Wildfire Fernandez

Documents

Transcript of Organic Matter Changes Immediately After a Wildfire Fernandez