Geoderma 147 (2008) 178–184

Contents lists available at ScienceDirect

Geoderma

j ourna l homepage: www.e lsev ie r.com/ locate /geoderma

Can terra rossa become water repellent by burning? A laboratory approach

J. Mataix-Solera a,⁎, V. Arcenegui a, C. Guerrero a, M.M. Jordán a, P. Dlapa b, N. Tessler c, L. Wittenberg c

a GEA (Grupo de Edafología Ambiental). Departamento de Agroquímica y Medio Ambiente, Universidad Miguel Hernández, Avenida de la Universidad s/n, 03202-Elche, Alicante, Spainb Department of Soil Science, Faculty of Natural Sciences, Comenius University, Mlynská dolina B-2, 842 15 Bratislava, Slovak Republicc Department of Geography and Environmental Studies, University of Haifa, Haifa 31905 Israel

⁎ Corresponding author. Tel.: +34 966658334; fax: +3E-mail address: jorge.mataix@umh.es (J. Mataix-Sole

0016-7061/$ – see front matter © 2008 Elsevier B.V. Aldoi:10.1016/j.geoderma.2008.08.013

a b s t r a c t

a r t i c l e i n f o

Article history:

Fire usually induces water Received 1 April 2008Received in revised form 4 July 2008Accepted 20 August 2008Available online 27 September 2008

Keywords:Soil hydrophobicityWater repellencyFireRhodoxeralfChromic luvisolKaolinite

repellency (WR) in soils. Reduction in infiltration rates, increase of runoff anderosion are some of the consequences of WR in fire-affected soils. Most forest soils can develop WR byburning; however some previous observations in burned terra rossa soils have shown little changes in WR.Laboratory controlled experiments have been done with samples of terra rossa from 14 different sites. Theobjectives are to confirm whether the observed is a common behaviour of terra rossa and to explore thefactors controlling the wettability of this soil type after burning. Samples from the upper 2.5 cm of terra rossawere collected from 12 forest sites of the Alicante province (Spain), and from 2 sites in the “Mt. Carmel”, Haifa(Israel) with similar environmental conditions. Laboratory burning of samples at 250 °C, 300 °C and 350 °Cwas performed with and without the addition of litter of Pinus halepensis. The results confirm that all soilshave a very low susceptibility to become water repellent by burning. Without the addition of litter, WR wasnot detected in any soil sample at any temperature of burning. With the addition of litter, WR was presentonly in six of the soils after some of the heating treatments. Although all soils had enough soil organic matter(SOM) to develop WR by heating, the ratio between SOM and clay content was considerably lower comparedto other types of forest soils of the region in which WR has been found after forest fires. This could explain inpart the lower susceptibility of terra rossa to become water repellent by burning since, as some authors haveindicated, fine-textured soils are less prone to develop soil WR due to their high specific surface area. Frommineralogical analysis of the clay fraction we found that the dominant clay types in the studied terra rossawere kaolinite and illite, with the exception of one soil where Ca–montmorillonite content is higher thankaolinite and illite. Ca–montmorillonite was present in only three of the soils. Comparing the soil propertiesbetween the group of terra rossa that in no case become water repellent (wettables) with the group that insome cases developed WR (potentially water repellents), some differences were found: the kaolinite contentis higher in the wettables group (Pb0.05), and the soils containing Ca–montmorillonite are in the group ofpotentially water repellents. A clear separation between the 2 groups was found when we compared SOM vskaolinite contents, the kaolinite content being the main factor contributing to this separation. These resultsare in agreement with those obtained in experiments with clay additions to water repellent soils in order toreduce the WR, and also with some studies which found that kaolinite is one of the most effective clayminerals for this purpose.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

Soil water repellency (WR) is a common property in fire-affectedsoils (Doerr et al., 2000). Burning of litter and soil organic matter in arange of temperatures produces hydrophobic substances (Savage,1974; DeBano et al., 1976). In the soil, WR can produce a reduction ininfiltration rates, enhance runoff and erosion, and can also affect there-establishment of vegetation because of the reduction of wateravailability (see review of Doerr et al., 2000).

Studies show that when a forest soil is burned at certain ranges oftemperatures most develop WR (e.g.: DeBano and Krammes, 1966;

4 966658340.ra).

l rights reserved.

Doerr et al., 2004; García-Corona et al., 2004). However there are a fewcases, where no changes in WR after burning have been observed(Giovannini and Lucchesi, 1983; Busse et al., 2005). In these last casesit was argued that this could be due to the relatively high clay contentof the studied soils.

A background of some previous observations under laboratoryconditions (Arcenegui et al., 2007), and in one of the study areas afterwildfires (Arcenegui et al., 2008) in terra rossa soils led us to suspectthat the intrinsic characteristics of this type of soils provide arelatively low susceptibility to develop WR by burning comparedwith other types of forest soils of the region.

Factors such as organic matter and clay content have beendescribed as factors controlling the WR developed by burning(DeBano, 1981, 1991). The mineralogy of clay could also be playing

Table 1Main characteristics of the sampling sites

Soil reference Site Location (UTM) zone/X/Y Vegetation (main species) Geological substratum Soil classification (Soil Survey Staff, 2006)

North Alicante — South-east Spain1 Montgó 1 31/254213/4299993 Pinus halepensis Cretaceous limestone Lithic Rhodoxeralf2 Montgó 2 31/248272/4300592 Pinus halepensis, Quercus coccifera Cretaceous limestone Lithic Rhodoxeralf3 Montgó 3 31/248478/4300471 Pinus halepensis, Quercus coccifera Cretaceous limestone Lithic Rhodoxeralf4 Montgó 4 31/249360/4299160 Pinus halepensis Cretaceous limestone Lithic Rhodoxeralf5 Jalón 31/240800/4291100 Pinus halepensis, Quercus coccifera Cretaceous limestone Lithic Rhodoxeralf6 Lliber 31/241560/4292965 Pinus halepensis Cretaceous limestone Lithic Rhodoxeralf7 Gata 31/242275/4293400 Pinus halepensis Cretaceous limestone Lithic Rhodoxeralf8 Moraig 31/253865/4288813 Pinus halepensis Cretaceous limestone Lithic Rhodoxeralf9 Benifallim 30/726292/4280035 Pinus halepensis, Quercus ilex Cretaceous limestone Lithic Rhodoxeralf10 Mariola 30/707897/4290641 Pinus halepensis, Quercus ilex Cretaceous marl- limestone Lithic Haploxeralf11 Biar 30/700146/4282283 Pinus pinea Cretaceous sandy limestone Lithic Haploxeralf12 Banyeres 30/705827/4284743 Pinus halepensis, Quercus ilex Cretaceous dolostone Lithic Haploxeralf

Mt Carmel — North-west Israel13 Mt. Carmel 1 36/689666/3625017 Pinus halepensis, Quercus calliprinos Cretaceous limestone Lithic Haploxeralf14 Mt. Carmel 2 36/688914/3624165 Pinus halepensis Cretaceous limestone Lithic Rhodoxeralf

Fig. 1. A representative example of terra rossa soil profile of this study. The typicaldiscontinuous layer ranging in thickness can be observed. Soil profile from site 4,Montgó 4 (Lithic Rhodoxeralf). Photography by J. Mataix-Solera.

179J. Mataix-Solera et al. / Geoderma 147 (2008) 178–184

an important role since this has been demonstrated to be a key factorin the alleviation of soil water repellency (McKissock et al., 2000;Dlapa et al., 2004).

Terra rossa is a reddish clayey to silty-clay material, which coverslimestones or dolomites in the form of a discontinuous layer rangingin thickness from a few centimetres to several metres. It is also foundalong cracks and between bedding surfaces of limestones anddolomites. The term “terra rossa” has been commonly used for thered Mediterranean soils derived on hard limestones, and severalnational soil classifications (e.g. Croatian, Italian, Israeli) have retainedthe term. The name currently continues to be used by soil scientists,although the term as a type of soils is no longer used as a separateclassification group in modern soil taxonomies (e.g.: Soil Taxonomy;WRB classification). In Soil Taxonomy (Soil Survey Staff, 2006) terrarossa is classified as Alfisols (Haploxeralfs or Rhodoxeralfs), Ultisols,Inceptisols (Xerochrepts) and Mollisols (Argixerolls or Haploxerolls).According to the WRB classification (FAO, 2006) terra rossa isrecognised as Luvisols (Chromic Luvisols), Phaeozems (Haplic Phaeo-zems or Luvic Phaeozems) and Cambisols. Thick accumulations ofterra rossa likematerial are situated in karst depressions in the form ofpedo-sedimentary complexes (Durn, 2003). Boero and Schwertmann(1989) suggested that the pedoenvironment of terra rossa ischaracterised by an association with Mediterranean climate, highinternal drainage due to the karstic nature of a hard limestone andneutral pH conditions. Moresi and Mongelli (1988) suggested that thenature of the soil formed is a consequence of the dissolution ofcarbonate and accumulation of insoluble residue in situ throughkarstic weathering processes in Mediterranean climatic conditions.However some investigations point out that in some places theformation of terra rossa has an aeolian origin (Yaalon, 1997).

The particle size distribution of the insoluble residue of terra rossais dominated by clay (Durn, 2003). Generally the clay contentincreases with depth in soil profiles. In some cases a relatively highcontent of sand sized particles has been observed, and this has beenattributed to rhizoconcretions which formed in terra rossa as theresult of palaeopedological processes which post-date terra rossaformation (recalcification of terra rossa following its burial) or torecent colluvial additions (Durn et al., 1999).

Scarce data are available concerning the water repellencybehaviour of terra rossa to burning (Arcenegui et al., 2007). Withthe aim of confirming whether the observed response of terra rossa toburning is a common behaviour and to explore the factors controllingthe wettability of this type of soils after burning, laboratory controlledexperiments have been done with samples of terra rossa from 14different sites. Additionally some soil physical, chemical and miner-alogical analyses have been done in samples.

2. Materials and methods

2.1. Study sites, soils used and sampling

Soil samples were collected from 12 forest sites in the north of theAlicante province (southeast of Spain), and from 2 sites in “Mt.Carmel”, in Haifa (northwest of Israel) with similar geological, edaphicand climatic conditions to the north of Alicante province. The maincharacteristics of sampling sites are described in Table 1. The North ofthe province of Alicante (Spain) has a mean annual precipitation thatvaries from 500 mm to 850 mm. Similarly the mean annualprecipitation in Mt. Carmel ranges from 500 mm at the coastal plainto 750 mm at the upper elevations (Wittenberg et al., 2007). Terrarossa soils, as in the entire Mediterranean region are also represented

Table 2Main characteristics of the soil samples used (0–2.5 cm depth from themineral horizon)

Soilreference

Site Sand, silt,clay (%)a,b

SOM(%)a

CaCO3

(%)apHa Munsell

colour (dry)aMunsellcolour(moist)a

1 Montgó 1 23, 62, 15 5.4 9.5 7.9 5YR4/6 5YR3/42 Montgó 2 6, 58, 36 4.8 0.2 7.9 2.5YR4/6 2.5YR2.5/43 Montgó 3 14, 56, 30 8.7 0.3 7.9 2.5YR3/6 2.5YR2.5/34 Montgó 4 45, 26, 29 5.3 21.9 7.9 5YR4/4 5YR3/45 Jalón 29, 40, 31 2.3 1.6 8.2 5YR4/6 5YR4/46 Lliber 23, 38, 39 5.4 1.1 7.8 2.5YR4/6 2.5YR3/67 Gata 33, 40, 27 5.4 5.7 7.8 2.5YR3/6 2.5YR2.5/48 Moraig 27, 44, 29 3.8 4.1 8.1 5YR4/4 5YR3/49 Benifallim 20, 49, 31 4.3 3.3 8.2 2.5YR3/3 2.5YR3/410 Mariola 27, 58, 15 2.9 2.9 8.1 7.5YR4/4 7.5YR3/411 Biar 60, 24, 16 4.5 21.9 8.2 7.5YR4/4 7.5YR3/412 Banyeres 48, 32, 20 3.3 0.8 8.3 7.5YR4/4 7.5YR3/313 Mt. Carmel 1 10, 74, 16 7.1 0.5 7.9 5YR3/3 5YR3/214 Mt. Carmel 2 13, 56, 31 6.3 8.5 7.9 2.5YR3/3 2.5YR2.5/4

SOM: soil organic matter; 7.5YR4/6: strong brown; 7.5YR3/3 and 7.5YR3/4: dark brown;7.5YR4/4: brown; 5YR4/6: yelowish red; 5YR3/2, 5YR3/3, 5YR3/4, 2.5YR2.5/3; 2.5YR2.5/4,2.5YR3/3 and 2.5YR3/4: dark reddish brown; 2.5YR4/6: red; 2.5YR3/6: dark red; 5YR4/4:reddish brown.

a Values based on pooled sample material from 15 sampling places for each site.b Sand: 2–0.05 mm; silt: 0.05–0.002; clay b0.002 mm.

Table 3WDPT values (in seconds) before and after heating at different temperatures withoutand with fuel load (litter of P. halepensis)

Soilreference

Site Unburned Burned without fuelload

Burned with fuel load

250 °C 300 °C 350 °C 250 °C 300 °C 350 °C

1 Montgó 1 b1 b1 b1 b1 b1 b1 b12 Montgó 2 b1 b1 b1 b1 b1 b1 b13 Montgó 3 b1 b1 b1 b1 2±1 688±

77640±64

4 Montgó 4 b1 b1 b1 b1 b1 2±1 b15 Jalón b1 b1 b1 b1 b1 b1 2±16 Lliber b1 b1 b1 b1 b1 1±1 1±17 Gata b1 b1 b1 b1 b1 1±1 b18 Moraig b1 b1 b1 b1 b1 b1 b19 Benifallim b1 b1 b1 b1 b1 2±1 b110 Mariola b1 b1 b1 b1 b1 60±43 2155±

105711 Biar b1 b1 b1 b1 283±

927280±745

3256±3644

12 Banyeres b1 b1 b1 b1 b1 3±3 248±475

13 Mt.Carmel 1

b1 b1 b1 b1 3±2 210±265

254±139

14 Mt.Carmel 2

b1 b1 b1 b1 b1 311±275

b1

Values are mean of 12 measurements (three drops per replicate). Total n=1176. Valuesin bold emphasis indicate water repellent conditions.

180 J. Mataix-Solera et al. / Geoderma 147 (2008) 178–184

in both the North of Alicante Province, and in Mt. Carmel (Israel)(Lavee et al., 1995); most of them classified as Rhodoxeralfs andHaploxeralfs (Soil Survey Staff, 2006). Other dominant types of forestsoils of both regions are Xerorthents, mainly on south slopes, andHaploxerolls, Rendolls, and Calcixerolls on some of the north slopes(Lavee et al., 1995; Mataix-Solera et al., 2002; Zornoza et al., 2007). Allterra rossa soils sampled for this study were classified as Rhodoxeralfsand Haploxeralfs in Soil Taxonomy (Soil Survey Staff, 2006), orChromic Luvisols according to the WRB classification (FAO, 2006) (seeTable 1). All soils are characterized by a discontinuous layer ranging inthickness, with a sequence of horizons A, Bt, R. In most cases theoriginal A horizon has been partially eliminated by erosion processes.An example of a representative soil profile of the study sites is shownin Fig. 1. Main soil characteristics are shown in Table 2.

At each site, soil samples were taken from the first 2.5 cm ofmineral topsoil horizon collected by pooling samples taken fromdifferent points (n=15) and then mixing by hand. The samples wereair-dried in the laboratory at room temperature (~25 °C) to constantweight and passed through a 2 mm aperture sieve to remove stonesand large plant residues. In this way the soil can be used as a factor tostudy in the experiments, increasing replicability and comparability.Litter debris was collected from beneath Pinus halepensis Miller (P),dried at room temperature and hand-mixed.

To compare the organic matter and clay content of the terra rossasoils of this study with other types of forest soils of the region whereWR has developed after forest fires, we have used available data from14 other forest soils of the region that are not terra rossa (Mataix-Solera and Doerr, 2004; Arcenegui et al., 2007, 2008). This group ofsoils, mainly Xerorthents (Soil Survey Staff, 2006) will be called “nonterra rossa” from now onwards, they being comparable as sampleswere taken in a similar way and at a similar depth.

2.2. Analytical methods

Characterization of soils was carried out on air-dried samples(Table 2), which included pH (1:2.5 w/v, distilled water), texturedetermined by Bouyoucos method (Gee and Bauder, 1986), soilorganic matter (SOM) determined by potassium dichromate oxidation(Nelson and Sommers, 1982), and carbonates, determined by themethod of the Bernard calcimeter.

Mineralogical characterization of soil samples was carried out byX-ray diffraction (XRD) usingwhole soil random powder, oriented clayon ceramic plates, and clay random powder. The clay fraction was

separated by sedimentation after treatment with HOAc-NaOAc buffer(pH 5.0) to remove carbonates (Gee and Bauder, 1986) and withhydrogen peroxide to remove organic matter. For the mineralogicalanalysis of soil samples the fractionating method, as described byJackson et al. (1949) and Hathaway (1956) and modified by Jordánet al. (1999), was used. Orientated clay aggregates (normal, heated to550 °C for 2 h and treated with ethylene glycol for 2 h) were alsoprepared. These tests were conducted in a powder diffractometer withBragg–Brentano θ:2θ configuration, using CuKα radiation (40 kV,30 mA), aperture slit =2 mm, divergence slit =2 mm, detectorslit =0.2 mm, antiscatter slit =0.6 mm, and graphite secondarymonochromator. Samples were analysed within the range 3–65°,data taken every 0.04° and step-time of 3.2 s. X-ray diffraction plotswere analysed using the Siemens automatic software for peakrecognition, mineral identification and peak intensity calculations.Chung method (1974a,b) was used for quantitative interpretation ofX-ray diffraction patterns of soils.

The water drop penetration time (WDPT) test (Wessel, 1988) wasused to measure the persistence of WR in soil samples. This involvedplacing three drops of distilled water (~0.05 ml) on the sample surfaceand recording the times required for their complete penetration. Theaverage time for triplicate drops is reported as the WDPT value of asample. Penetration times were classified in intervals according toBisdom et al. (1993), with WDPT≤5 s representing wettable andWDPTN5 s water repellent conditions.

2.3. Heat treatments

Laboratory heating treatments were done for each one of the soilsamples without andwith the addition of a high dose of litter debris ofPinus halepensis.

Sixty grams of soil were placed in a porcelain container (8.5 cmdiameter, 3.5 cm depth). The amount of vegetation used was 5 g. Weselected this quantity based on estimates of the amount of vegetationdebris under field conditions and the results of a previous work(Arcenegui et al., 2007), where the authors found that both the type ofvegetation and the quantity of the litter debris are also factorscontrolling the persistence of WR developed by burning.

The reason for using P. halepensis in a high dose was to createoptimum conditions for the soil to become water repellent by heating,

Table 4Comparison of the mean values of SOM, Clay and the ratio SOM/Clay, between the terrarossa soils of this study and other type of forest soils of the Alicante province (non terrarossa) where WR has been found after burning

Soil parameter Terra rossa Non terra rossaa t-test

SOM (%) 5.0±1.7 8.8±3.7 ⁎⁎

Clay (%) 26±8 14±11 ⁎⁎

SOM/Clay 0.2±0.1 0.8±0.6 ⁎⁎

n 14 14

*, ** and *** indicate significance at Pb0.05, Pb0.01 and Pb0.001, respectively. ns: nonsignificant.

a Other type of forest soils of the region (mainly Xerorthents) where WR has beenfound after burning (data obtained from Mataix-Solera and Doerr, 2004; Arceneguiet al., 2007, 2008). Samples of soils are from 0–2.5 cm of mineral horizon.

Table 6Comparison between the two groups of terra rossa with respect to the wettabilitybehavior after burning with fuel load

Soil parameter Wettables Potentially water repellents t-test

SOM (%) 4.6±1.1 5.5±2.3 nsClay (%) 30±7 21±7 nsSOM/Clay 0.2±0.1 0.3±0.1 nsIllite (% in clay fraction) 50±9 39±7 nsIllite (% in soil) 15±5 8±4 nsKaolinite (% in clay fraction) 42±11 28±3 ⁎

Kaolinite (% in soil) 12±4 6±1 ⁎⁎

SOM/Illite 0.4±0.3 0.7±0.3 nsSOM/Kaolinite 0.4±0.2 1.0±0.3 ⁎⁎⁎

n 8 6

*, ** and *** indicate significance at Pb0.05, Pb0.01 and Pb0.001, respectively. ns: nonsignificant.

181J. Mataix-Solera et al. / Geoderma 147 (2008) 178–184

additionally, this specie is one of the most abundant in the Mediterra-nean forest.

Heat treatments were carried out in a pre-heated muffle furnace(Nabertherm, P320, Bremen, Germany) at 250 °C, 300 °C, and 350 °Cfor 20 min. At each of the specified temperatures, individual aliquotswere exposed to a single temperature. Four replicates were made foreach of the soils and heating treatments (n=336).

2.4. Post-heating measurements of water repellency

After heating, approximately 15 g of soil was placed on separate50-mm diameter plastic dishes and exposed to a controlled laboratoryatmosphere (20 °C, ~50% relative humidity) for one week to eliminatepotential effects of any variations in preceding atmospheric humidityon soil WR and in accordance with the findings of Doerr et al. (2005).The persistence of WR was measured in all the soil samples asdescribed above in Section 2.2.

2.5. Statistical procedures

The t-test for independent samples was used to compare themeans of parameters between the terra rossa and non terra rossa soilgroups, and between wettable and potentially water repellent terrarossa soils. Statistical analyses were performed using the SPSS 14package (© SPSS Inc, 1989).

3. Results

3.1. Heating effects on soil water repellency

All terra rossa soils were initially wettables (see Table 3). Theresults of heating show that all these soils have a very low

Table 5Mineralogical composition (%) of the clay fraction of soils (XRD semi-quantitative analysis)

Soilreference

Site Kaolinitea Illitea Ca–montmorillonitea

Quartza Chloritea

1 Montgó 1 53 33 − 14 −2 Montgó 2 23 46 − 21 103 Montgó 3 25 43 − 22 104 Montgó 4 43 57 − + −5 Jalón 54 46 − + −6 Lliber 47 53 − + −7 Gata 45 55 − + −8 Moraig 39 61 − + −9 Benifallim 27 45 − 28 −10 Mariola 29 26 45 + −11 Biar 29 42 − 29 −12 Banyeres 25 45 − 30 −13 Mt. Carmel 1 31 36 17 16 −14 Mt. Carmel 2 28 43 5 24 −

+Values below 5%.−Not identified.

a Values based on pooled sample material from 15 sampling places for each site.

susceptibility to become water repellent by burning. After heatingthe soils without litter addition, WR was not detected in any of thefourteen soils studied at any temperature of burning. With theaddition of litter, in six of the soils WR was detected after some of theheating treatments (Table 3). Based on these results we distinguishedtwo groups of terra rossa soils from this study: 1 — wettables (thosethat in no case developed WR), and 2 — potentially water repellents(those that in some cases with the litter addition developed WR).

3.2. Key soil properties related to WR: soil organic matter, clay contentand clay mineralogy

Soil organic matter (SOM) content is quite variable in the topsoil ofterra rossa studied ranging from 2.9 to 8.7%. Clay content ranges from15 to 39% (see Table 2). Comparing the data of these terra rossa withother types of forest soils of the region (mainly Xerorthents (SoilSurvey Staff, 2006)) where WR has developed after forest fires(Mataix-Solera and Doerr, 2004; Arcenegui et al., 2007, 2008), wefound some differences (Table 4). Terra rossa show lower SOM andhigher clay contents, resulting in a ratio SOM/clay considerably lowerin the terra rossa soils compared with the other forest soils(statistically different at Pb0.01; Table 4).

From mineralogical analyses of the clay fraction we found that thedominant type of clay in the terra rossa studied soils are kaolinite andillite, with the exception of one of the soils where Ca–montmorillonitecontent is higher than kaolinite and illite. Ca–montmorillonite waspresent only in three of the soils (Table 5).

Some differences were found comparing the soil propertiesbetween the two groups of terra rossa (the wettables versus thepotentially water repellents; Table 6): 1 — the ratio SOM/clay is lowerin the wettables, although the difference is not statistically significant,

Fig. 2. Soil organic matter content versus kaolinite content in terra rossa soilsdistinguishing between wettables (white triangles) and potentially water repellents(black squares). Data represented are the original SOM contents before any heatingtreatment. Numbers indicate the soil reference number (see Table 2).

182 J. Mataix-Solera et al. / Geoderma 147 (2008) 178–184

2 — the kaolinite content in the group of wettables is higher (Pb0.05)than in the potentially water repellents, and 3— the three soils with apresence of Ca–montmorillonite are in the group of potentially waterrepellents (Table 6). In Fig. 2 we show SOM vs kaolinite content foreach terra rossa soil, and a clear separation between the 2 groups canbe observed, the kaolinite content being the main factor contributingto this separation.

4. Discussion

Most studies report that heating of forest soils modifies theirwettability, and the changes in WR are related to the temperaturesreached (e.g.: DeBano and Krammes, 1966; Doerr et al., 2004; García-Corona et al., 2004; Dlapa et al., 2008). In many studies it has beenobserved that fire induced WR in hydrophilic soil, and eitherenhanced or reduced the surface WR in an already water repellentsoil (e.g.: Savage 1974; DeBano et al., 1979; Mataix-Solera and Doerr,2004; Hubbert et al., 2006; Arcenegui et al., 2008). In naturally highlywater-repellent soils, fire may have very little effect on WR if soiltemperatures remain below the threshold of repellency destruction(Doerr et al., 1996). Only in some studies, authors have reported thatno development of WR has been detected after burning of hydrophilicsoils. In this sense Busse et al. (2005) did not find any effects ofburning on WR in a hydrophilic soil with a clay loam texture. Nochanges for soil WR was detected by Giovannini and Lucchesi (1983)in one of their studies of fire effects on soils, and in this case theyargued that the discrepancy found in comparison with most previousstudies was because of the fine texture of the soil tested.

Our results contrast with the previous studies (DeBano et al., 1970;Savage et al., 1972; Savage,1974; Arcenegui et al., 2007) on hydrophilicsoils whereWR can be developed as consequence of burning. On otherhand, they confirm some previous observations of Arcenegui et al.(2007), where two different soils showed very different responses toburning onWR under laboratory conditions, one of them being a terrarossa. The same authors studied the immediate effects of wildfires onsoil WR, finding as a general pattern an increase in WR. However, nosignificant changes on WR were found in one of their study sites witha Rhodoxeralf (Arcenegui et al., 2008).

All terra rossa soils included in this study have enough SOMcontent to develop WR by burning according to DeBano (1991) whosuggested that any soil containing more than 2% of organic matter issusceptible to becoming water repellent when heated. However inour case no soil became water repellent when burned without litterdebris addition. DeBano (1981) also suggested that soil texture is akey factor, the coarser being the most susceptible to develop WR. InMediterranean areas, terra rossa are very old soils rich in claycontent, and although the texture of the topsoils of this study wasquite variable, in general terms the clay content was relatively highcompared with other soil types of the region (Mataix-Solera andDoerr, 2004; Arcenegui et al., 2007, 2008). Although all soils hadenough SOM to develop WR by heating, the relatively low SOM/clayratio found in the terra rossa could explain in part the lowersusceptibility to become water repellent by burning since fine-textured soils have been found to be less prone to develop soil WR.The explanation given in previous studies has been the higherspecific surface area of clayey soils (DeBano, 1981; Giovannini andLucchesi, 1983). In fact, clay minerals have been successfully addedto amend water repellent sandy soils (Ma'shum et al., 1989; Harperand Gilkes, 1994; McKissock et al., 2000; Dlapa et al., 2004).However it is presumable that not only clay content and SOM areresponsible for the low susceptibility of terra rossa to become WRsince some studies reported that soils with 25% to more than 40% ofclay have been found to exhibit extreme WR (Crockford et al., 1991;Chan, 1992; Dekker and Ritsema, 1996a,b). Also some of the terrarossa soils in this study have developed WR by burning with litteraddition, and this group of soils was not statistically different in clay

and SOM content from the wettables group. It has been suggestedthat in clayey soil, WR development may occur as long as the clayforms aggregates, thus causing the surface area to be covered with ahydrophobic film (Wallis et al., 1991; Bisdom et al., 1993). Anadditional factor is the mineralogy of the clay fraction that couldalso be responsible for the wettability behaviour of these soils. Somestudies have found that further than clay content, the mineralogy ofthis fraction is a key factor controlling the WR (Lichner et al., 2006).

The mineralogical analyses of the clay fraction of the terra rossa ofthis study showed that kaolinite is one of the dominant clayminerals instudied terra rossa soils, with the exception of one soil where Ca–montmorillonite content is higher than kaolinite and illite. Ca–montmorillonite was present only in three soils. Similar mineralogicalcomposition has been found in other terra rossa of Mediterraneanregion where kaolinite is one of the dominant clay minerals, e.g.: inwestern Sicily (Bellanca et al., 1996), in “Sierra Gádor” South of Spain(Delgado et al., 2003), in Epirus, Greece (Macleod, 1980), in Apulia,Italy (Moresi andMongelli,1988), in NWMorocco (Bronger and Bruhn-Lobin, 1997), in Istria, Croatia (Durn et al., 1999). The high presence ofkaolinite in terra rossa is justified by the high stability of thismineral inMediterranean soils (Torrent, 1995). These very old soils of the regionhave accumulated the kaolinite by inheritance through the dissolutionof the rock with time, whereas other less stable minerals have beenweathered. Neoformation of kaolinite also can contribute to itspresence in the soil as was demonstrated by Delgado et al. (2003) inthe terra rossa of Sierra of Gádor (south of Spain).

The main difference between the two groups of terra rossa soils(wettables vs potentially water repellents) of this study was that thekaolinite content was higher in the group of wettables than in thepotentially water repellents. The three soils with presence of Ca–montmorillonite are in the group of potentially water repellents. Ourresults are in agreement with some studies that found that kaolinite isone of the most effective types of clay to reduce the WR in waterrepellent soils (Harper and Gilkes, 1994; McKissock et al., 2000;McKissock et al., 2002; Dlapa et al., 2004; Lichner et al., 2006).Ma'shum et al. (1989) using water repellent sands found that clayscontaining illite and kaolinite were more effective in reducing WRthan clays containing montmorillonite. Ward and Oades (1993)examining the effect of monomineralic clays on natural water-repellent sands indicated that kaolinite wasmore effective in reducingthe WR than montmorillonite, despite having larger crystals and alower surface area which are properties that might be consideredconducive to reducing the capacity to adsorb hydrophobic substancesor coat sand grains. For clays containing mixtures of clay minerals,clays containing kaolinite were more effective than clays containingillite or smectite (Ma'shum et al., 1989; Mutter, 1995).

Existing differences in the ability of clay minerals to alleviate soilwater repellency are commonly explained as a result of the differentabilities of clayminerals tomask hydrophobic surfaces and to facilitatewater penetration, the dispersibility of clay being an important factorin defining clay effectiveness. Clay dispersibility is determined bymineralogy and cations such as Na or Ca present on the exchange sites.Na-clays are more effective than Ca-clays for WR reduction (Ma'shumet al., 1989; Lichner et al., 2006). In this sense, some authors haveindicated that kaolinite is more effective than Ca–montmorillonitebecause the latter tends to aggregate, whereas kaolinite remainsdispersed, allowing the retention of an effective covering of hydro-phobic surfaces and facilitating the absorption of water (McKissocket al., 2000; Eynard et al., 2006). Nevertheless, such an explanation isnot in agreement with the conclusions of Singer (1994) who reviewedthe effect of clay mineralogy on soil dispersivity and concluded thatsmectitc soils were the most and kaolinitic soils the least, dispersivewith the dispersivity of illitic soils being intermediate but sometimeexceeding that of smectitic soils.

Due to the above mentioned inconsistency in the existing knowl-edgewe propose an alternative explanation for claymineral effects on

183J. Mataix-Solera et al. / Geoderma 147 (2008) 178–184

soil wettability, which can be derived from the properties of the claysurfaces. It is believed that in clay minerals the oxygen atoms at thebasal surface of the siloxane sheets are hydrophobic whereas thehydroxyls at the basal surface of the octahedral sheets are hydrophilic(Zbik and Smart, 2002). Thus, the hydrophilicity of Ca–montmorillo-nite originates from the hydrated exchangeable cations because itsbasal surface contains only hydrophobic siloxane (Si–O–Si) groups.Replacement of the inorganic cations with hydrophobic organiccations should greatly diminish the hydrophilic nature of the clay.This might facilitate the direct adsorption of hydrophobic organiccompounds to the clay surface (Jaynes and Boyd, 1991). Jouany (1991)observed that the very small amounts of synthetic humic acidadsorbed on Ca–montmorillonite transformed the hydrophilicmineral into a hydrophobic mineral.

On the other hand, kaolinite has two basal surfaces. Thehydrophobic surface with siloxane groups on one basal plane andthe hydrophilic surface with hydroxyl groups of the octahedral sheeton the other side. The significance of polar hydroxyl groups for thehydrophilicity of minerals was emphasized by Bachmann and van derPloeg (2002). Dlapa et al. (2004) noted that higher density of hydroxylgroups on the kaolinite surface may explain the difference in thekaolinite and Ca–montmorillonite effects on the persistence of soilwater repellency. If the surface of octahedral sheet is inherentlyhydrophilic, significant adsorption of hydrophobic organic moleculesshould be restricted due to preferential adsorption of water. In such away kaolinite should be more important for preservation of soilhydrophilicity and more effective in the alleviation of soil waterrepellency compared to Ca–montmorillonite.

5. Conclusions

From our results we can conclude that terra rossa has a very lowsusceptibility to become water repellent by burning. The factors thatseem to control this behaviour are: 1— the relative low ratio SOM/claycontent comparedwith other type of forest soils of the region, and 2—

the high presence of kaolinite.Our results suggest that the probability of finding water

repellency in forest areas affected by fire with this type of soils isrelatively lower than those with other type of soils. This fact reflectsthe advantage of the forest with terra rossa in respect to others tobetter regulate the water into the soil avoiding the usually observedincrease in runoff rate as a consequence of water repellency andthus promoting the restoration of vegetation and preservation of thesoil.

These results stem from controlled heating experiments on terrarossa soils from 14 sites under laboratory condition. The developmentof water repellency has been sought by providing optimum conditionsfor this (Arcenegui et al., 2007) and by adding a high dose of litter fromone of the most abundant species under Mediterranean environment(P. halepensis). This led us to think that the above mentionedconclusions from these laboratory experiments could reflect realityand thus could be transferable to field conditions. However, wesuggest that further research under field conditions in fire-affectedareas with terra rossa soils is needed to verify these findings, also withthe presence of various plant species.

Acknowledgements

This research was supported by the CICYT co-financed FEDERproject CGL2006-011107-C02-01/BOS, the Spanish GovernmentProject for International Cooperation PCI2006-A7-0576 and thecomplementary action CGL2007-28764-E, and the SOLIPHA projectVVCE-0033-07, the VEGA project 1/3058/06. V. Arcenegui acknowl-edges the grant received from the “Caja de Ahorros del Mediterrá-neo”. Authors also thank to the staff of Montgó Natural Park, M.L.Calero for laboratory support, Frances Young for improving the

English, and to the editor and two anonymous reviewers forvaluable comments to improve the manuscript.

References

Arcenegui, A., Mataix-Solera, J., Guerrero, C., Zornoza, R., Mayoral, A.M., Morales, J.,2007. Factors controlling the water repellency induced by fire in calcareousMediterranean forest soils. Eur. J. Soil Sci. 58, 1254–1259.

Arcenegui, V., Mataix-Solera, J., Guerrero, C., Zornoza, R., Mataix-Beneyto, J., García-Orenes, F., 2008. Immediate effects of wildfires on water repellency and aggregatestability in Mediterranean calcareous soils. Catena 44, 219–226.

Bachmann, J., van der Ploeg, R.R., 2002. A review on recent developments in soil waterretention theory: interfacial tension and temperature effects. J. Plant Nutr. Soil Sci.165, 468–478.

Bellanca, A., Hauser, S., Neri, R., Palumbo, B., 1996. Mineralogy and geochemistry of TerraRossa soils, western Sicily: insights into heavymetal fractionating and mobility. Sci.Total Environ. 193, 57–67.

Bisdom, E.B.A., Dekker, L.W., Schoute, J.F.T., 1993. Water repellency of sieve fractionsfrom sandy soils and relationships with organic material and soil structure.Geoderma, 56, 105–118.

Boero, V., Schwertmann, U., 1989. Iron oxide mineralogy of terra rossa and its geneticimplications. Geoderma 44, 319–327.

Bronger, A., Bruhn-Lobin, N., 1997. Paleopedo-logy of Terrae rossae— Rhodoxeralfs fromQuaternary calcarenites in NW Morocco. Catena 28, 279–295.

Busse, M.D., Hubbert, K.R., Fiddler, G.O., Shestak, C.J., Powers, R.F., 2005. Lethal soiltemperatures during burning of masticated forest residues. Int. J. Wildland Fire 14,267–276.

Chan, K.Y., 1992. Development of seasonal water-repellence under direct drilling. SoilSci. Soc. Am. J. 56, 326–329.

Chung, F.H., 1974a. Quantitative interpretation of X-ray diffraction patterns of mixturesI. Matrix-flushing method for quantitative multi component analysis. J. Appl. Cryst.7, 519–525.

Chung, F.H., 1974b. Quantitative interpretation of X-ray diffraction patterns of mixtures II.Adiabatic principle of X-ray diffraction analysis of mixtures. J. Appl. Cryst. 7, 526–531.

Crockford, S., Topalidis, S., Richardson, D.P., 1991. Water repellency in a dry sclerophyllforest — measurements and processes. Hydrol. Proc. 5, 405–420.

DeBano, L.F., 1981. Water repellent soils: a state-of-the-art. US Department ofAgriculture Forest Service General Technical Report, PSW-46, 21. Berkeley, CA.

DeBano, L.F., 1991. The effects of fire on soil properties. US Department of AgricultureForest Service General Technical Report, pp. 151–156. INT-280.

DeBano, L.F., Krammes, J.S., 1966. Water repellent soils and their relation to wildfiretemperatures. Int. Bull. Assoc. Hydrol. Sci. 2, 14–19.

DeBano, L.F., Mann, L.D., Hamilton, D.A., 1970. Translocation of hydrophobic substancesinto soil by burning organic litter. Soil Sci. Soc. Am. J. 34, 130–133.

DeBano, L.F., Rice, R.M., Conrad, C.E., 1979. Soil heating in chaparral fires: effects on soilproperties, plant nutrients, erosion and runoff. USDA Forest Service, Research PaperPSW-145. Pacific Southwest Forest and Range Experimental Station, Berkeley,California. 21 pp.

DeBano, L.F., Savage, S.M., Hamilton, D.A., 1976. The transfer of heat and hydrophobicsubstances during burning. Soil Sci. Soc. Am. Proc. 40, 779–782.

Dekker, L.W., Ritsema, C.J., 1996a. Variation in water content and wetting patterns inDutch water repellent peaty clay and clayey peat soils. Catena 28, 89–105.

Dekker, L.W., Ritsema, C.J., 1996b. Preferential flow paths in a water repellent clay soilwith grass cover. Water Resour. Res. 32, 1239–1249.

Delgado, R., Martín-García, J.M., Oyonarte, C., Delgado, G., 2003. Genesis of the terraerossae of the Sierra Gádor (Andalusia, Spain). Eur. J. Soil Sci. 54, 1–16.

Dlapa, P., Doerr, S.H., Lichner, L., Sir, M., Tesar, M., 2004. Effect of kaolinite and Ca–montmorillonite on the alleviation of soil water repellency. Plant, Soil Environ. 50,358–363.

Dlapa, P., Šimkovic Jr, I., Doerr, S.H., Šimkovic, I., Kanka, R., Mataix-Solera, J., 2008.Application of thermal analyses to elucidate water repellency changes in heatedsoils. Soil Sci. Soc. Am. J. 72, 1–10.

Doerr, S.H., Shakesby, R.A., Walsh, R.P.D., 1996. Soil hydrophobicity variations withdepth and particle size fraction in burned and unburned Eucalyptus globulus andPinus pinaster forest terrain in the Águeda basin, Portugal. Catena 27, 25–47.

Doerr, S.H., Shakesby, R.A., Walsh, R.P.D., 2000. Soil water repellency: its causes,characteristics and hydro-geomorphological significance. Earth Sci. Rev. 51, 33–65.

Doerr, S.H., Blake, W.H., Shakesby, R.A., Stagnitti, F., Vuurens, S.H., Humphreys, G.S.,Wallbrink, P., 2004. Heating effects on water repellency in Australian eucalyptforest soils and their value in estimating wildfire soil temperatures. Int. J. WildlandFire 13, 157–163.

Doerr, S.H., Douglas, P., Evans, R.C., Morley, C.P., Mullinger, N.J., Bryant, R., Shakesby, R.A.,2005. Effects of heating and post-heating equilibration times on soil waterrepellency. Aust. J. Soil Res. 43, 261–267.

Durn, G., 2003. Terra Rossa in the Mediterranean region: parent materials, compositionand origin. Geol. Croat. 56, 83–100.

Durn, G., Ottner, F., Slovenec, D., 1999. Mineralogical and geochemical indicators of thepolygenetic nature of terra rossa in Istria, Croatia. Geoderma 91, 125–150.

Eynard, A., Lal, R., Wiebe, K.D., 2006. Water repellent soils, In: Lal, R. (Ed.), Encyclopediaof Soil Science, Second edition. Taylor and Francis, New York, pp. 1860–1863.

FAO, 2006.World reference base for soil resources.World Soil Resources Report, vol.103.FAO, Rome.

García-Corona, R., Benito, E., de Blas, E., Varela, M.E., 2004. Effects of heating on somesoil physical properties related to its hydrological behaviour in two north-westernSpanish soils. Int. J. Wildland Fire 13, 195–199.

184 J. Mataix-Solera et al. / Geoderma 147 (2008) 178–184

Gee, G.W., Bauder, J.W., 1986. Particle-size analysis, In: Klute, A. (Ed.), Methods of SoilAnalysis 1: Physical and Mineralogical Methods, 2nd Ed. American Society ofAgronomy, Madison, pp. 383–411.

Giovannini, G., Lucchesi, S., 1983. Effect of fire on hydrophobic and cementingsubstances of soil aggregates. Soil Sci. 136, 231–236.

Harper, R.J., Gilkes, R.J., 1994. Soil attributes related towater repellency and the utility ofsoil survey for predicting its occurrence. Aust. J. Soil Res. 32, 1109–1124.

Hathaway, J.C., 1956. Procedure for clay mineral analysis used in the SedimentaryPetrology laboratory of the US Geological Survey. Clay Min. Bull. 3, 8–13.

Hubbert, K.R., Preisler, H.K., Wohlgemuth, P.M., Graham, R.C., Narog, M.G., 2006.Prescribed burning effects on soil physical properties and soil water repellency in asteep chaparral watershed, southern California, USA. Geoderma 130, 284–298.

Jackson, W.W., Whitting, L.D., Pennington, R.P., 1949. Segregation procedure formineralogical analysis of soils. Proc. Soil. Sci. Soc. Am. 14, 77–81.

Jaynes, W.F., Boyd, S.A., 1991. Hydrophobicity of siloxane surfaces in smectites as revealedby aromatic hydrocarbon adsorption fromwater. Clay Clay Miner. 39, 428–436.

Jordán, M.M., Boix, A., Sanfeliu, T., De la Fuente, C., 1999. Firing transformations ofcretaceous claysused in themanufacturingof ceramic tiles. Appl. Clay Sci.14, 225–234.

Jouany, C., 1991. Surface free energy components of clay-synthetic humic acidcomplexes from contact-angle measurements. Clay Clay Miner. 39, 43–49.

Lavee, H., Kutiel, P., Segev, M., Benyamini, Y., 1995. Effect of surface roughness on runoffand erosion in a Mediterranean ecosystem: the role of fire. Geomorphology 11,227–234.

Lichner, L., Dlapa, P., Doerr, S.H., Mataix-Solera, J., 2006. Evaluation of different clayminerals as additives for soil water repellency alleviation. Appl. Clay Sci. 31, 238–248.

Ma'shum, M., Oades, J.M., Tate, M.E., 1989. The use of dispersable clays to reduce water-repellency of sandy soils. Aust. J. Soil Res. 27, 797–806.

Macleod, D.A., 1980. The origin of the red Mediterranean soils in Epirus, Greece. J. SoilSci. 31, 125–136.

Mataix-Solera, J., Doerr, S.H., 2004. Hydrophobicity and aggregate stability in calcareoustopsoil from fire affected pine forests in southeastern Spain. Geoderma 118, 77–88.

Mataix-Solera, J., Gómez, I., Navarro-Pedreño, J., Guerrero, C., Moral, R., 2002. Soilorganic matter and aggregates affected by wildfire in a Pinus halepensis forest in aMediterranean environment. Int. J. Wildland Fire 11, 107–114.

McKissock, I., Walker, E.L., Gilkes, R.J., Carter, D.J., 2000. The influence of clay type onreduction of water repellency by applied clays: a review of some West Australianwork. J. Hydrol. 231–232, 323–332.

McKissock, I., Gilkes, R.J., Walker, E.L., 2002. The reduction of water repellency by addedclay is influenced by clay and soil properties. Appl. Clay Sci. 20, 225–241.

Moresi, M., Mongelli, G., 1988. The relation between the terra rossa and the carbonate-free residue of the underlying limestones and dolostones in Apulia, Italy. ClayMiner. 23, 439–446.

Mutter, A., 1995. The use of clay to ameliorate water repellence. Hons. Thesis, Universityof Western Australia, Perth, Wester Australia, pp. 1–76.

Nelson, D.V., Sommers, L.E., 1982. Total carbon, organic carbon, and organic matter.In: Page, A.L. (Ed.), Methods of soil analysis, Part 2. Chemical and biologicalmethods.

Savage, S.M., 1974. Mechanism of fire-induced water repellency in soils. Soil Sci. Soc.Am. Proc. 38, 652–657.

Savage, S.M., Osborn, J., Letey, J., Heaton, C., 1972. Substances contributing to fire-induced water repellency in soils. Soil Sci. Soc. Am. Proc. 36, 674–678.

Singer, A. 1994. Clay mineralogy as affecting dispersivity and crust formation inAridisols. In Etchevers, J.D. (Ed.), Transactions of the 15th World Congress of SoilScience. Acapulco, Mexico. Vol 8a. Int. Soc. Soil Sci. and Mexican Soc. Soil Sci.,Acapulco, Mexico. pp. 37-46.

Soil Survey Staff, 2006. Keys to Soil Taxonomy, 10th ed. NRCS, Washington, DC.Torrent, J., 1995. Genesis and Properties of the Soils of the Mediterranean Region.

Pubblicazione Università degli studi di Napoli, Federico II, Napoli.Wallis, M.G., Scotter, D.R., Horne, D.J.,1991. An evaluation of the intrinsic sorptivitywater

repellency index of a range of New Zealand soils. Aust. J. Soil Res. 29, 353–362.Ward, P.R., Oades, J.M., 1993. Effect of clay mineralogy and exchangeable cations on

water-repellency in clay amended sandy soils. Aust. J. Soil Res. 31, 351–364.Wessel, A.T., 1988. On using the effective contact angle and the water drop penetration

time for classification of water repellency in dune soils. Earth Surf. Proc. Landforms13, 555–562.

Wittenberg, L., Kutiel, H., Greenbaum, N., Inbar, M., 2007. Short-term changes in themagnitude, frequencyand temporal distribution offlood in the easternMediterraneanregion during the last 45 years— Nahal Oren, Mt. Carmel, Israel. Geomorphology 84,181–191.

Yaalon, D.H., 1997. Soils in the Mediterranean region: what makes them different?Catena, 28, 157–169.

Zbik, M., Smart, R.St.C., 2002. Dispersion of kaolinite and talc in aqueous solution: nano-morphology and nano-bubble entrapment. Miner. Eng. 15, 277–286.

Zornoza, R., Mataix-Solera, J., Guerrero, C., Arcenegui, V., Mayoral, A.M., Morales, J.,Mataix-Beneyto, J., 2007. Soil properties under natural forest in the AlicanteProvince of Spain. Geoderma 142, 334–341.