Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain)

CATENA-02271; No of Pages 10 Catena xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Catena journal homepage: www.elsevier.com/locate/catena

Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain) A. Lombao a,⁎, A. Barreiro a, T. Carballas a, M.T. Fontúrbel b, A. Martín a, J.A. Vega b, C. Fernández b, M. Díaz-Raviña a a b

Departamento de Bioquímica del Suelo, Instituto de Investigaciones Agrobiológicas de Galicia (IIAG-CSIC), P.O. Box 122. Avda. Vigo s/n, 15780 Santiago de Compostela, Spain Centro de Investigación Forestal-Lourizán, Consellería do Medio Rural e do Mar, Xunta de Galicia, P.O. Box 127, 36080 Pontevedra, Spain

a r t i c l e

i n f o

Article history: Received 4 February 2014 Received in revised form 1 August 2014 Accepted 12 August 2014 Available online xxxx Keywords: Wildfire Biochemical properties PLFA pattern Quercus Eucalyptus Soil depth

a b s t r a c t The impact of a wildfire on some selected physicochemical, chemical (water retention, pH, electrical conductivity, free Fe and Al oxides, total C, extractable C), biochemical (microbial C, soil respiration, bacterial activity, βglucosidase, urease and phosphatase activities) and microbiological properties (analysis of phospholipid fatty acids, PLFA pattern) was evaluated in Fragas do Eume Natural Park (NW Spain). Soil samples were collected three months after the wildfire from the A horizon (0–2.5 and 2.55 cm) of the unburnt and burnt soil under climax vegetation (Quercus) and non-autochthonous vegetation (Eucalyptus). The results indicated that, independent of the vegetation considered, the wildfire induced short-term modifications of most soil properties analysed, more accentuated changes being those related to labile fractions of the soil organic matter (extractable C and microbial biomass C, negative effects) as well as those in pH and bacterial growth values (positive effects). The fire effect was often more noticeable in the 0–2.5 cm layer than in the 2.5–5 cm layer. The results of a principal component analysis performed with the matrix of the physicochemical and biochemical data showed that vegetation was the most important factor controlling the overall quality of these soils and that wildfire is also an important source of variation in soil quality. This is in agreement with the PLFA pattern, differentiating clearly the Quercus soil samples from the Eucalyptus ones and, to a lesser extent, the burnt soil samples from the corresponding unburnt ones. Medium- and long-term consequences of these microbial changes in the functioning of the plant–soil system should be investigated in order to preserve the biodiversity of the Natural Park. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The frequency and extent of wildfires increased dramatically in the European Mediterranean region from the 1960s, aided by a general warming and drying trend, but driven primarily by socio-economic changes, including rural depopulation, land abandonment and afforestation with flammable species (Shakesby, 2011). Wildfire effects on the soil environment are highly variable; however, in general, fires cause partial or complete combustion of organic matter, deterioration of soil structure, depletion of nutrients through volatilization and leaching, altered aggregate stability and water repellency, together with marked quantitative and qualitative alterations of soil microbial communities (Almendros and González-Vila, 2012; Carballas et al., 2009; Certini, 2005; Díaz-Raviña et al., 2010; Fernández et al., 2001; Holden and Treseder, 2013; Neary et al., 1999). Furthermore, fires cause vegetation destruction, which favors erosion processes that can produce enormous irreversible losses of soil (Díaz-Fierros et al., 1987; Shakesby, 2011; Vega et al., 2013a). The impact of wildfires on the ⁎ Corresponding author. Tel.: +34 981590958; fax: +34 981592504. E-mail address: [email protected] (A. Lombao).

environment (direct effects on the plant–soil system and indirect effects such as post-fire erosion) is commonly considered to be especially harmful causing very serious ecological, economic and social problems. The extent of fire effects can be highly variable due to the large amount of controlling factors such as fire regime (severity, duration and recurrence) as well as local conditions such as type of soil, vegetation composition, topography or regional climate (Neary et al., 1999). Thus, the impact of fire on soils from a temperate humid forest is very different compared to its impact on soils in an arid or Mediterranean ecosystem (Almendros and González-Vila, 2012; Carballas et al., 2009). Studies on fire impacts focused on Mediterranean areas are fairly abundant; however, information on the effect of wildfires in Atlantic ecosystems is still scarce (Santín et al., 2008). Galicia (NW Spain) and the north of Portugal are the European areas most affected by forest wildfires, and worldwide they are among the areas with the greatest number of fires per hectare and inhabitant (Carballas et al., 2009). Galicia is a very mountainous region with 2,000,000 ha of forestland with scrub and tree stands (69% of the surface), developed on a complex mosaic of soil types and parent materials, most soils being acid and sandy. Quercus robur forest is the climax vegetation, but nowadays the majority of the territory is covered by

http://dx.doi.org/10.1016/j.catena.2014.08.007 0341-8162/© 2014 Elsevier B.V. All rights reserved.

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

2

A. Lombao et al. / Catena xxx (2014) xxx–xxx

Pinus and Eucalyptus stands. Approximately half of the wildfires in Spain occur in this region that represents less than 6% of the national territory; during the last ten years the number of fires was around 9000 per year on average, and the burnt area was about 40,000 ha. Thus, since forest fires are common events in Galicia causing the destruction of vegetation and soil degradation as well as enormous losses of soil and nutrients due to runoff and erosion processes (Carballas et al., 2009; Díaz-Fierros et al., 1987; Vega et al., 2013a), there is a need to determine their effects on terrestrial ecosystems, particularly when fires occur in environmentally sensitive areas, natural areas and areas set aside for wildfire protection. Soil quality has been defined as the capacity of a specific kind of soil to function, within natural or managed ecosystem boundaries, to sustain plant and animal productivity, maintain or enhance water and air quality, and support human health and habitation (Karlen et al., 1997). Several parameters have been proposed for the assessment of soil quality. The indication given by individual properties is often limited and subjected to seasonal and spatial variations: an integration of basic physical, chemical and biological indicators is required for the evaluation of climate and management effects on soil function. Several parameters have been measured for the evaluation of the status of burnt soils (Cerda and Jordán, 2010). Main physical and chemical properties are often included in these investigations but microbiological characterization is one of the aspects that has received less attention (Certini, 2005; Díaz-Raviña et al., 2010; Mataix-Solera et al., 2009; Neary et al., 1999); on top of that, integrative studies combining basic physical, chemical and biological properties by means of multivariate statistical techniques are even more uncommon. In addition, vegetation is not considered in most studies concerning wildfire impacts on soil quality. The aim of the present study was to evaluate the short-term impact of a wildfire that occurred in 2012 in the Fragas do Eume Natural Park (Galicia, NW Spain) on soil quality using several physicochemical, chemical and biological properties (biomass, activity and community structure measurements [PLFA pattern] as well as enzyme activities connected to the C, N and P-cycles) aggregated by means of a principal component analysis.

vegetation) both in the unburnt and burnt areas were selected at different locations throughout the park. Thus, a total of 16 plots (4 unburnt Quercus, 4 burnt Quercus, 4 unburnt Eucalyptus, 4 burnt Eucalyptus), each plot covering a surface area of about 1000 m2, were established for the field experimental design. The soil is developed over granite and the slope of the plots is 30–70%. Soil sampling was performed 3 months after the wildfire and several physicochemical, biochemical and microbiological properties were analyzed. From each plot, after removing the litter in the case of unburnt plots, multiple soil subsamples were taken from the 0 to 5 cm (0–2.5 cm, 2.5–5 cm) of the A horizon top layer; they were mixed to form one representative composite soil sample per depth and per plot and refrigerated (4 °C) until processing in the laboratory. 2.2. Methods The following soil properties were monitored in the b 2 mm fraction: moisture content and water retention capacity, pH (in water and KCl), electrical conductivity, free Fe and Al oxides, total C, extractable C, microbial biomass, soil respiration, bacterial activity and soil enzymes related to the C, N and P cycles, and phospholipid fatty acid analysis (PLFA pattern). Previous studies performed with soils from the same region with similar characteristics (coarse texture, high organic matter content, acid pH) showed that soil physical properties such as texture, aggregate stability and water repellence were unaffected or slightly affected by wildfire (Díaz-Raviña et al., 2012) and therefore these parameters were excluded from this study. The methods described by Guitián-Ojea and Carballas (1976) were utilized to determine the following physical and chemical properties: moisture content by oven-drying soil samples at 105 °C for 6–7 h; water retention capacity using Richard's pressure plate apparatus (pF = 2); pH in H2O and KCl in a soil:solution ratio of 1:2.5 and electrical conductivity in a soil/water extract of 1:5; the organic C content was determined by combustion in a Carmhograph 12 (Wosthoff OHG, Bochum, Germany) and free Fe and Al oxides by extraction with a mixture of Tamm's reagent and sodium dithionite.

2. Material and methods 2.3. Microbial biomass 2.1. Experimental site The study was performed in one of the six natural parks in Galicia, Fragas do Eume Natural Park (A Coruña, NW Spain), which extends along the valley of the Eume river within the Ferrolterra municipalities of Pontedeume, Cabanas, A Capela, Monfero, Pontedeume and As Pontes de García Rodríguez. The area was declared a natural park (a level of protection lower than national park) in 1997 and it is an example of a temperate rainforest recognized by the European Union as a Site of Community Importance in which climax vegetation is dominated by “Fragas”, natural woodland with a mixture of species such as Q. robur, Corylus avellana, Castanea sativa, Betula alba, Laurus nobilis, Ulmus glabra, Salix atrocinerea, Fraxinus excelsior, Fraxinus angustifolia and Alnus glutinosa. However, the protected area also included vegetation dominated by non-autochthonous species such as Eucalyptus globulus and, to a lesser extent, Pinus radiata. On March 2012 a wildfire destroyed for 3 days the heart of the park and approximately 1000 ha of the Natural Park was affected, 750 ha dominated by non-autochthonous vegetation, mainly E. globulus, and 350 ha dominated by climax vegetation, mainly Q. robur. The extent of fire can be partly explained by the different susceptibilities of the two tree species to combustion. The level of combustion of the organic layer and the deposition of ash from the aboveground combustion of the biomass suggested that fire severity had been moderate to high (Vega et al., 2013b). In order to evaluate the impact of this wildfire, plots with different species compositions named according the dominating species and representatives of these two types of vegetation affected by the wildfire (Quercus, climax vegetation; Eucalyptus, non-autochthonous

The microbial biomass C was determined using the fumigation extraction method with some modifications (Díaz-Raviña et al., 1992). After soil fumigation with CHCl3 for 24 h, the organic C was extracted from the unfumigated and fumigated samples with 0.05 M K2SO4 using a 1:4 soil:extract ratio. The microbial biomass C values were calculated from the equation: biomass C = 2.64EC, where EC is the extractable C flush (difference between the extractable organic C from the fumigated and unfumigated samples). Extractable C from the unfumigated samples was used as a measurement of available C (soil solution). 2.4. Microbial activity The soil respiration, an overall index of activity of heterotrophic microorganisms, and the measurement of three specific enzyme activities related to the C (β-glucosidase), N (urease) and P (phosphatase) cycles were used as indicators of soil microbial activity. The soil respiration was determined by the incubation of fresh soil samples (75% of field capacity) at 22 °C during a 10 day period measuring the CO2 trapped in a NaOH solution, which was then titrated with HCl (Díaz-Raviña et al., 1993b). The β-glucosidase activity was measured following the procedure of Eivazi and Tabatabai (1988), which determines the released pnitrophenol after the incubation of the soil with a p-nitrophenyl glucosidase solution for 2 h at 37 °C. The urease activity was estimated by incubating the soil samples with an aqueous urea solution and extracting the NH+ 4 with 1 M KCl and 0.01 M HCl followed by the

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

A. Lombao et al. / Catena xxx (2014) xxx–xxx

colorimetric NH+ 4 determination by a modified indophenol reaction (Kandeler and Gerber, 1988). The phosphatase activity was assayed following the method described by Trasar-Cepeda et al. (1985), which determines the p-nitrophenol released after soil incubation with pnitrophenyl phosphate for 30 min at 37 °C. The bacterial activity was also determined by means of the incorporation of labeled leucine into bacteria extracted after homogenization–centrifugation (Bååth et al., 2001). 2.5. Microbial community structure The microbial biomass and the microbial community structure were estimated by phospholipid fatty acid (PLFA) analysis using the procedure described by Frostegård et al. (1993). Briefly, lipids were extracted from the soil with a chloroform:methanol:citrate buffer mixture (1:2:0.8 v/v/v) and separated into neutral lipids, glycolipids and phospholipids using a pre-packed silica column. The phospholipids were subjected to a mild alkaline methanolysis and the fatty acid methyl esters were identified by gas chromatography (flame ionization detector) by the relative retention times of the fatty acids, using methyl nonadecanoate (19:0) as an internal standard. The total microbial biomass [total PLFAs (TotPLFAs)] was estimated as the sum of all the extracted PLFAs. The quantity of 10Me18:0, 10Me17:0 and 10Me16:0 was used as an indicator of actinomycete biomass (ActPLFAs); the sum of the PLFAs considered to be predominantly of bacterial origin (i15:0, a15:0, 15:0, i16:0, 16:1ω9, 16:1ω7t, i17:1ω8, i17:0, a17.0, 17:0, cy17:0, 18:1ω7 and cy19:0), was used as an index of the bacterial biomass (BactPLFAs), and the quantity of the 18:2ω6, 18:1ω9 and 16:1ω5 PLFAs was used as an indicator of the fungal biomass (FungPLFAs). The i14:0, i15:0, i16:0 and 10Me18:0 PLFAs are predominantly found in Gram-positive (G+) bacteria (G+PLFA), and the cy17:0, cy19:0, 16:1ω7c and 18:1ω7 PLFAs characterize Gram-negative (G−) bacteria (G−PLFA) (Díaz-Raviña et al., 2006). 2.6. Statistical analysis In order to evaluate the effect of the wildfire on the soil in two representative park ecosystems (climax vegetation and non-autochthonous vegetation), the values of four plots with the same vegetation and soil depth were averaged (mean ± SE). The data were analyzed by a three way analysis of variance (ANOVA3) to determine the percentage of variation attributable to the factors vegetation, burning and soil depth. For each depth, the data were also analyzed by a standard analysis of variance (ANOVA1) and, in the cases of significant F statistics; the Tukey's minimum significant difference test was used to separate the means. Data corresponding to the concentrations of all the individual PLFAs, expressed in mole percent and logarithmically transformed, were also subjected to a principal component analysis (PCA) to elucidate the main differences in the PLFA patterns. In addition, another principal component analysis (PCA) was carried out on physicochemical, chemical and biochemical data for the evaluation of the soil status. All statistical analyses were made using SPSS 15.0 statistical package. 3. Results and discussion 3.1. Physicochemical, biochemical and microbiological properties of the study soils The minimum and maximum values of the selected physicochemical and biochemical properties analyzed in the Natural Park are summarized in Table 1 and mean values and standard deviation are shown in Table 2 and Figs. 1–3. The results indicated that although the values of the burnt soil samples differed from those obtained for the unburnt soil samples, in general, the order of magnitude was similar. The studied soils are very acid with high soil organic matter (SOM) content and very high water retention capacity due to the high content of SOM. These

3

Table 1 Minimum and maximum values of selected physicochemical and biochemical properties of the studied soils in the Fragas do Eume Natural Park (n = 16 samples for each vegetation, 8 unburnt samples and 8 burnt samples). Quercus Unburnt pH H2O pH KCl Moisture (%) Water retention at field capacity (g water kg−1) Total C (g kg−1) Electrical conductivity (μS cm−1) Al2O3 (g kg−1) Fe2O3 (g kg−1) Extractable C (μg g−1) Microbial C (μg g−1) Soil Respiration (μg CO2 g−1 day−1) Bacterial activity (mol leu 10−14 g−1 h−1) Glucosidase activity (μg pnitrophenol g−1 h−1) Urease activity −1 −1 (μg NH+ h ) 4 g Phosphatase activity (μg pnitrophenol g−1 h−1) Total PLFA (nmol g−1) Fungal PLFA (nmol g−1) Bacterial PLFA (nmol g−1) Actinomycete PLFA (nmol g−1) Gram-negative PLFA (nmol g−1) Gram-positive PLFA (nmol g−1)

Eucalyptus Burnt

Unburnt

Burnt

3.37–3.91 3.71–4.50 3.53–4.09 3.88–4.81 2.55–3.27 2.93–3.59 2.56–3.35 2.84–3.56 38–58 31–49 33–45 24–35 765–1635 646–1175 674–1026 461–751 111–400 110–180

111–262 74–148

71–240 70–164

82–255 68–130

5.7–14.5 8.7–15.9 8.7–14.3 7.5–18.2 21.4–56.6 22.0–59.6 34.1–71.1 19.4–58.8 23.0–53.4 12.3–23.1 25.0–42.6 9.0–22.9 1542–2239 364–1275 930–2943 401–1014 30–247 36–146 28–145 30–62 6.6–17.6

9.9–43.8

3.7–12.1

10.4–28.3

96–557

104–350

64–446

29–180

134–252

217–251

97–238

45–160

495–1731

428–1503

245–1106

157–740

285–832 50–122 115–402 39–76

306–688 42–108 148–298 31–67

405–673 48–97 175–287 41–84

319–777 39–98 133–390 37–78

60–258

95–200

102–165

78–265

39–76

31–67

36–65

30–69

characteristics are representative of forest soils developed on acid rocks in the Atlantic humid temperate zone of the NW of Spain, the SOM being the main agent responsible for fairly high productivity of these soils supporting the establishment of good forests (Carballas et al., 2009; González-Prieto et al., 1996; Martín et al., 2012; Santín et al., 2008). The microbial biomass C and activity values, expressed as absolute values (per g−1 soil), fell in the reported range for Galician soils (Alvárez et al., 2009; Basanta et al., 2002; Díaz-Raviña et al., 1988, 1993a, 1993b, 1996, 2012; Fontúrbel et al., 2012; Mahía et al., 2006; Trasar-Cepeda et al., 2000) and were much higher than those reported for arid and semiarid zones (Bárcenas-Moreno et al., 2011; Hernández et al., 1997). The different range of SOM values observed in different climatic regions explained this behavior since most ecosystems are resource-limited (metabolisable C, available nutrients) (Wardle, 1992) and therefore the size and microorganisms activity is mainly determined by both quantity and quality of the SOM. This is supported by positive and significant correlations (p b 0.001, n = 32) between SOM and most microbial indices here analyzed (r = 0.57 for microbial C, r = 0.91 for respiration, r = 0.84 for glucosidase, r = 0.49 for urease and r = 0.89 for phosphatase). The microbial C was negatively correlated with bacterial activity (r = −0.46) and positively correlated with the other indices of metabolic activity (r = 0.62 for respiration and r = 0.67 for glucosidase, r = 0.41 for urease and r = 0.52 for phosphatase). A positive and significant relationship (p b 0.001) between glucosidase and urease (r = 0.64) and phosphatase (r = 0.89) and between urease and phosphatase (r = 0.72) activities were also found. These findings indicated that the mass and activity of the microorganisms in these soils are linked, emphasizing the interdependence of the activity of the biogeochemical cycles of C (respiration, glucosidase), N (urease) and P (phosphatase) cycles, as well as the lack of relationships with the bacterial activity determined by the leucine incorporation technique.

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

4

A. Lombao et al. / Catena xxx (2014) xxx–xxx

Table 2 Mean values (±SE; n = 4 field plots) of selected biochemical properties of the studied soils in the unburnt and burnt soil samples studied under different vegetation properties at two depths (1, 0–2.5 cm; 2, 2.5–5 cm). For each depth, different letters show significant differences (ANOVA1, p b 0.05 level). For each parameter ANOVA3 (V, vegetation; B, burning; D, depth) was performed, but only the proportions of a variance explained by the significant factors (p b 0.05 level) are indicated. ANOVA3

Depth

Extractable C/total C (%)

B (26%); V (11%); D (15%)

Microbial C/total C (%)

B (68%); V (6%)

1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2

Quercus

Eucalyptus

Unburnt

Soil respiration (μg CO2 g−1 OM day−1) Bacterial activity (×10−14 mol leu g−1 OM h−1) Glucosidase activity (μg p-nitrophenol g Urease activity (μg

NH+ 4

g

−1

OM h

−1

−1

OM h

B (35%) −1

)

B (34%); V (10%)

)

Phosphatase activity (μg p-nitrophenol g−1 OM h−1)

V (26%)

Total PLFA (nmol g−1 OM)

D (15%)

Bacterial PLFA (nmol g

−1

OM)

D (17%)

Fungal PLFA (nmol g−1 OM) Actinomycete PLFA (nmol g−1 OM) Gram-negative PLFA (nmol g Gram-positive PLFA (nmol g

−1

−1

D (19%)

OM)

D (13%)

OM)

D (19%)

The total microbial biomass values, estimated as the total PLFA, in the soil samples studied ranged from 285 to 832 nmol g−1 soil (Table 1). The amount of PLFAs that was chosen to represent bacteria, fungi and actinomycete PLFAs comprised 40–53, 10–19 and 9–13 mol% of the total amount of PLFAs, respectively. The amount of PLFAs representative of G− bacteria and G+ bacteria comprised 21–36 and 8–12 mol% of the total amount of PLFAs, respectively. The results fell in the reported range given for other forest ecosystems from the same region (Barreiro et al., 2010; Díaz-Raviña et al., 2006, 2012). The unburnt and burnt soil samples showed quite similar values; likewise, no appreciable differences between soil samples under different vegetation were observed. The PLFAs of specific groups (fungi, bacterial, Grampositive bacteria, Gram-negative bacteria, actinomycetes) followed the same trend as that observed for the total amount of PLFAs, which can be explained by the positive and significant relationships observed between the total PLFA and the biomass of the specific microbial groups (r = 0.896 for FungPLFA, r = 0.997 for BactPLFA, r = 0.936 for ActPLFA, r = 0.963 for G−PLFA and r = 0.927 for G+PLFA, p b 0.001, n = 32). 3.2. Fire effects on the different soil properties analyzed In order to facilitate the interpretation of the fire effects data, the values of four plots with the same vegetation and soil depth were averaged and the percentage of attributable variation to the three factors considered (vegetation, burning and soil depth) was analyzed by ANOVA3; the results are indicated in Figs. 1–3. Except soil pH, the rest of the physicochemical and chemical properties exhibited lower values in the burnt soil samples than in the corresponding unburnt ones, indicating that the fire had a negative effect on the soil quality (Fig. 1). The results agree with those obtained by other authors in a wide range of Galician forests affected by wildfires of medium severity and high severity (Carballas et al., 2009; Martín et al., 2009, 2012; Vega et al., 2013b). It should be noted, however, that differences are not significant in many cases due to the variability of data from field plots with the same vegetation and disturbance. The different fire severity and/or the high spatial variability of the different plots with very high SOM content located at

0.17 0.20 0.91 1.1 292 231 27.3 55.0 940 838 494 545 3081 2629 1520 2364 687 1122 230 305 147 241 424 694 138 222

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

Burnt 0.02b 0.02ab 0.06b 0.16ab 35a 29a 6.8c 8.5a 150c 145a 80a 63a 403b 112a 538a 623a 267a 308a 75a 74a 47a 60a 185a 202a 43a 58a

0.09 0.12 0.40 0.56 251 217 77.5 58.9 558 655 507 718 2813 2439 1705 1951 770 938 252 266 164 187 477 595 159 178

Unburnt ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.01a 0.01a 0.10a 0.08a 30a 26a 14.1b 13.9a 81b 97a 58a 56a 155b 227a 196a 171a 96a 72a 27a 41a 20a 15a 65a 51a 20a 15a

0.19 0.35 1.39 1.51 268 229 32.3 46.8 875 596 623 715 2588 1851 1751 2701 778 1259 286 348 180 303 476 761 158 256

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

Burnt 0.04b 0.07b 0.09b 0.14b 23a 14a 5.8a 17.0a 73c 42a 146a 197a 193a 172a 393a 348a 161a 121a 68a 52a 37a 48a 96a 58a 40a 36a

0.11 0.18 0.42 0.54 205 204 92.9 81.0 383 472 375 605 1426 2448 1829 2380 858 1142 245 255 196 274 574 675 168 243

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.02ab 0.07ab 0.09a 0.09a 29a 24a 16.3b 15.6a 23a 106a 108a 147a 279a 247a 552a 386a 293a 204a 63a 39a 52a 42a 205a 145a 48a 35a

different sites throughout the park can explain this behavior. The ANOVA3 showed that the fire had a significant negative effect on total C (29% of variance explained), water retention (15–22% of variance explained), electrical conductivity (17% of variance explained) and extractable C (68% of variance explained) and a positive significant effect on pH values (38–49% of variance explained). Positive effects of the fire on extractable C and electrical conductivity could have been detected immediately after the fire but these effects are transitory and with time the opposite trend was observed (Carballas et al., 2009; Couto-Vázquez and González-Prieto, 2006; Hernández et al., 1997; Martín et al., 2012). Our data also indicated that the labile fraction of the organic matter (extractable C) rather than the total organic C is adequate to evaluate the short-term impact of the wildfire. This is also consistent with studies showing that the labile fractions of the organic matter increased notably following the fire but that with time they tend to decrease to lower levels than those initially present in the corresponding unburnt soil samples whereas the recalcitrant organic matter pool increased (Almendros and González-Vila, 2012; Barreiro et al., 2010; Fernández et al., 2001; Martín et al., 2009; Rovira et al., 2012). Though the unburnt soil samples exhibited slightly higher values of the total biomass and the biomass of specific microbial groups (fungi, bacteria, actinomycetes, Gram-positive bacteria, and Gram-negative bacteria) than the burnt soil samples (Fig. 3), the results of ANOVA3 indicated that the fire had no significant effect on the biomass data. Therefore, biomass estimations by means of phospholipids fatty acids are not adequate to determine the short-term fire impact on these forest ecosystems. Bárcenas-Moreno and Bååth (2009) and Bárcenas-Moreno et al. (2011) have also detected problems in using biomass estimates by means of phospholipid fatty acid analysis to evaluate the effects of forest fires due to their lower sensitivity to temperature with respect to other microbial indices (heating only destroys the PLFAs at very high temperatures), as well as the presence of confounding factors (a decrease in PLFAs due to the death and degradation of organisms and an increase in PLFAs resulting from the emergence of a new community growing under post-fire conditions). Our study also showed the improved

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

A. Lombao et al. / Catena xxx (2014) xxx–xxx

pH H2O

V (10%) c B (49%)

5 bc

a

ab

a

Qb

Eu

Eb

ab ab

6

4

2

2

1

1

0

0-2.5 cm

Qu Qb Eu Eb

V (31%) B (22%) D (9%)

b

b ab

b

40

20

b ab

b ab

800

a

ab a

0

0-2.5 cm

2.5-5 cm

0-2.5 cm 450

Electrical conductivity (µS cm-1

(

200

Qu Qb Eu Eb

V (14%) B (17%)

a 150

a

a

a

a

a

a a

100

Total C (g kg-1

2.5-5 cm

375 300 a

225

Qu Qb Eu Eb

V (11%) B (29%)

a a a

a a

150 50

a

a

75 0-2.5 cm

Al2O3 (g kg-1

( D (20%)

15

Qu

Qb

70

Eu

Eb

60

a

50

a a

a a

a

0

2.5-5 cm

a

0-2.5 cm

Qu Eu

Fe2O3 (g kg-1

a a

a

a

a

Qb Eb

a

a

a

40

a

2.5-5 cm

(

0

10

Qu Qb Eu Eb

400

0

20

a

a

V (31%) B (15%) D (10%)

1600 1200

Qb Eb

2.5-5 cm

ab a

a

Water retention (g water kg-1

ab

a

a

ab

0-2.5 cm

2000

(% ) Moisture

b

a

0

2.5-5 cm

60

250

b

ab 3

Qu Eu

B (38%) D (10%)

b

3

80

pH KCl

5

(

4

Qu

(

6

5

30 20

5

10 0

0-2.5 cm

2.5-5 cm

0

0-2.5 cm

2.5-5 cm

Fig. 1. Physicochemical and chemical soil properties at two depths in the unburnt (u) and burnt (b) soil samples studied under Quercus (Q) and Eucalyptus (E) (mean ± SE; n = 4 field plots). For each depth, different letters show significant differences (ANOVA1, p b 0.05 level). For each parameter ANOVA3 (V, vegetation; B, burning; D, depth) was performed, but only the proportions of the variance explained by significant factors (p b 0.05 level) are indicated.

sensitivity when using standard biomass measurements such as fumigation–extraction methods rather than estimations by means of PLFAs (see Fig. 2). The data support that the microbial C rather than the total organic matter content is adequate for detecting fire effects on soil quality at short- and medium-term time scales (Basanta et al., 2004; Prieto-Fernández et al., 1998; Villar et al., 2004). Except for the bacterial activity values, the microbial C and several indices of the metabolic activity were negatively affected by the fire independent of the vegetation considered (Fig. 2). The direct impact of the fire on the microorganisms (death due to high temperatures) and the indirect effects due to observed variations in the physicochemical

and chemical properties (Fig. 1, for example reduction in quantity and quality of SOM) can explain these effects. The results agree with those obtained by other authors, the extent of the effects depending on the temperature reached during the fire (Barreiro et al., 2010; Díaz-Raviña et al., 2012; Fontúrbel et al., 2012; Vega et al., 2013b). The ANOVA3 data showed that the fire had a negative effect on the microbial C (67% of variance explained), respiration (9% of variance explained) and soil enzyme activities (7–23% of variance explained) and a significant positive effect on the bacterial activity (29% of variance explained). This behavior can be explained on the basis of the different information obtained from the microbial index (overall microbial biomass, overall

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

A. Lombao et al. / Catena xxx (2014) xxx–xxx

Extractable C (μg g-1

(

700

500

B (68%)

b

400

b

b

b

4000

Qu Qb Eu Eb

600

3000

a

a

b

b

a

a

Qu Qb Eu Eb

B (67%) D (5%)

b

b

2000

300 200

Microbial C (μg g-1

(

6

a

1000

a

a

a

100

a

Qu Qb Eu Eb

V (13%) B (9%) D (21%)

150 a

a a

a 50 0

0-2.5 cm

a

a

2.5-5 cm

Glucosidase activity (µg p-nitrophenol g-1 h-1

(

a 100

700

40 35 30 25 20 15 10 5 0

Soil respiration (μg CO 2 g-1 day-1

V (16%) B (23%) D (15%)

b

400

ab

Qu Qb Eu Eb

b

ab

300

ab

200

a

ab

250

Bacterial activity (x10-14 mol Leu g-1 h-1 a Qu a

Qb Eu Eb

B (29%)

a a

a

a

a a

0-2.5 cm

2.5-5 cm

Urease activity (µg NH4+ g-1 h-1

0-2.5 cm

V (21%) B (7%) D (8%)

ab ab

Qu Qb Eu Eb

b ab

150

a

ab

a

100 a

50 0

0

b

200

100 2.5-5 cm

0-2.5 cm

2.5-5 cm

Phosphatase activity (µg p-nitrophenol g-1 h-1

(

2500

2.5-5 cm

300

600 500

350

0-2.5 cm

(

200

0

2.5-5 cm (

250

0-2.5 cm

(

0

2000 1500

V (30%) B (7%) D (21%)

b b

1000

ab

a a

500

Qu Qb Eu Eb

a a

a

0

0-2.5 cm

2.5-5 cm

Fig. 2. Soil biochemical properties at two depths in the unburnt (u) and burnt (b) soil samples studied under Quercus (Q) and Eucalyptus (E) (mean ± SE; n = 4 field plots). For each depth, different letters show significant differences (ANOVA1, p b 0.05 level). For each parameter ANOVA3 (V, vegetation; B, burning; D, depth) was performed, but only the proportions of the variance explained by significant factors (p b 0.05 level) are indicated.

microbial activity, specific enzyme activity of C, N or P cycles, activity of bacteria). The data clearly indicated that the microbial C rather than the microbial activity parameters is a good index for evaluating the shortand medium-term impacts of wildfires, which is in agreement with other studies (Díaz-Raviña et al., 2012; Hernández et al., 1997; Holden and Treseder, 2013; Mataix-Solera et al., 2009; Pourreza et al., 2014). The results also support that bacteria rather than fungi are favored under post-fire conditions (Bárcenas-Moreno et al., 2011; Lie et al., 2014; Mataix-Solera et al., 2009; Pourreza et al., 2014; Vázquez et al., 1993); thus, the burnt soil samples exhibited higher microbial growth values estimated by the leucine incorporation technique than the corresponding unburnt ones because only bacteria are able to incorporate this substrate. In contrast, the microbial parameter values quantifying both bacteria and fungi (microbial biomass, respiration, enzyme activities) showed a negative fire effect due to the fact that fungi, the main group contributing to both biomass and metabolic activity (eukaryote), decreased notably masking the positive bacterial fire effect. As previously noted, biochemical properties are largely dependent on SOM content (see correlation coefficients) and fire decreased organic matter levels; therefore, in order to discard the influence of this factor

on biochemical properties, values of selected properties were also expressed as relative values in relation to SOM content (Table 2). The results showed a similar trend, and hence fire effect, than that observed when values were expressed as absolute values (Figs. 2 and 3); however, it should be noted that while in some cases the importance of burning as a source of variation was maintained (microbial C, 68% variation) or even increased (bacterial growth and glucosidase activity, 33–35% variation) for others it decreased (extractable C, 26% variation) and even was not significant (respiration, urease and phosphatase activity). The results clearly confirmed that induced short-term changes in biochemical properties as a consequence of wildfire were not only attributed to decreases of SOM levels but also related to changes in other soil properties (e.g. SOM quality, pH, and nutrient availability), which is consistent with previous studies (Certini, 2005; Mataix-Solera et al., 2009). 3.3. Soil quality of the studied forest ecosystems Individual soil properties can fail to give an appropriate estimation of the quality of the burnt soils, particularly when analyzed soil properties

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

A. Lombao et al. / Catena xxx (2014) xxx–xxx

(

TotPLFA (nmol g-1 800

a

a

a

600

a

a

a

a

Qu Qb Eu Eb

BactPLFA (nmol g-1

(

500

1000

7

Qu Qb Eu Eb

400 a 300

a a a

a

a

a

a 400

200

200

100

0-2.5 cm 100

FungPLFA (nmol g-1

Qu Qb Eu Eb

120 100

0

2.5-5 cm

(

140

0-2.5 cm

V (12%) D (3%)

a a

a

80

a a

a

ActPLFA (nmol g-1

60 a

Qu Qb

80

Eu

a a 60

a

2.5-5 cm (

0

a

a

a

a

a a

Eb

a

40

40 20 20

300

G+bactPLFA (nmol g-1

a 60

0

2.5-5 cm

(

80

0-2.5 cm

a

a

Qu

Qb

Eu

Eb

0-2.5 cm

2.5-5 cm

G-bactPLFA (nmol g-1

(

0

a a

a

a

200 a

40

Qu Qb Eu Eb

250

150

a

a a

a a

a

a a

100 20 50 0

0-2.5 cm

2.5-5 cm

0

0-2.5 cm

2.5-5 cm

Fig. 3. Total biomass (TotPLFA), bacterial (BactPLFA), fungal (FungPLFA), actinomycete PLFA (ActPLFA), Gram-positive bacteria PLFA (G+bact) and Gram-negative PLFA (G−bact) at two depths in the unburnt (u) and burnt (b) soil samples studied under Quercus (Q) and Eucalyptus (E) (mean ± SE; n = 4 field plots). For each depth, different letters show significant differences (ANOVA1, p b 0.05 level). For each parameter ANOVA3 (V, vegetation; B, burning; D, depth) was performed, but only the proportions of the variance explained by the significant factors (p b 0.05 level) are indicated.

are affected in a different way (positive, negative or no effect) and extent as in the present study. Multivariate statistical techniques such as principal component analysis can by successfully used in the analysis of the soil status in burnt ecosystems (Barreiro et al., 2010; Díaz-Raviña et al., 2006; Martín et al., 2012). Therefore, in order to compare the soil quality in the samples collected in the Fragas do Eume, all the physicochemical and biochemical properties should be used together. Consequently, the principal component analysis was used to analyze the 15 soil variables of the whole data set of the samples analyzed (n = 32 samples). The main three factors identified accounted for 86% of the variance altogether (Table 3). Factor 1, which accounted for 46% of the total variance, is defined at its positive extreme by all the variables related to the SOM (total C, water retention, microbial biomass, phosphatase activity, urease activity, glucosidase activity and soil respiration) and at its negative extreme by pH. Factor 2, which accounted for 24% of the total variance, is defined at its positive arm

by microbial biomass C and extractable C and at its negative arm by pH and bacterial activity. Factor 3, which accounted for 16% of the variance, is defined at its positive extreme by free Al and F oxides and at its negative part by total C. The distribution of the samples on the planes defined by Factors 1 and 2 (Fig. 4A) and by Factors 1 and 3 (Fig. 4B), makes it possible to separate: a) soils under Quercus from the soils under Eucalyptus; b) the unburnt soil samples from the burnt ones; and c) the 0–2.5 samples from the 2.5 to 5 cm samples, suggesting that Factor 1 is related to the vegetation effect; Factor 2 is related to the fire effect; and Factor 3 is related to the soil depth. The results of this study show that tree species is the major factor influencing the soil quality in the temperate humid region, which can be mainly attributed to its influence on the C dynamics in the forest ecosystems (tree biomass, forest floor and mineral soil) (Alvárez et al., 2009; Martín et al., 2011; Pérez-Cruzado et al., 2012). Likewise the data clearly showed the usefulness of the principal component analyses to quantify the importance

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

8

A. Lombao et al. / Catena xxx (2014) xxx–xxx

Table 3 Loadings on the factors identified by principal component analysis (PCA) performed on physicochemical, chemical and biochemical properties of the studied soils in the Fragas do Eume Natural Park (15 variables, n = 32 samples).

Phosphatase activity Water retention at field capacity Moisture Soil respiration Total C Urease activity Electrical conductivity Glucosidase activity Bacterial activity Microbial C pH H2O pH KCl Extractable C Al2O3 Fe2O3

Factor 1

Factor 2

Factor 3

0.93 0.93 0.90 0.83 0.80 0.79 0.78 0.77 0.01 0.43 −0.64 −0.60 0.40 −0.10 −0.06

0.09 0.25 0.33 0.13 0.20 0.17 0.44 0.33 −0.84 0.76 −0.71 −0.70 0.70 −0.14 0.42

−0.23 −0.01 0.09 −0.43 −0.48 0.29 −0.05 −0.39 −0.34 −0.21 −0.08 0.22 −0.06 0.92 0.83

of the vegetation and the fire as sources of variation in the soil environment. To our knowledge, this is the first study evaluating the impact of wildfire on the quality of soil under different vegetation combining such a wide range of soil properties (physico-chemical, chemical, biochemical and microbiological properties). Principal component analysis performed on the PLFA pattern data was also used for soil status evaluation and exploring the factors

influencing the soil quality of the studied forest ecosystems. Factors 1 and 2 accounted for 46% of the variance and allowed us to differentiate among microbial communities according to tree species, soil depth and burning (Fig. 5). The first component, explaining 25% of the variance, was defined on its positive arm by the PLFAs 18:2ω6, 16:1ω7c, 16:1ω7t, 16:0, 17:0, a15:0, 14:0 and x6; and on its negative part by PLFAs 10Me16b:0, i17:0, br18:0, cy19:0, 18:1ω7 and 10Me18:0. The second component, explaining 21% of the variance, was defined by the PLFAs 10Me16a:0, i16:1, br17:0 and i16:0 at its positive part and by the PLFAs 16:1ω5, 16:1ω9, 18:1, 18:0 a17:0, 18:1ω7 and 10Me18:0 at its negative part. Soil samples under the same vegetation are grouped together and separated according to the soil depth. The Quercus soil samples (having positive values along PC1 and negative values in PC2) were mainly characterized by high concentrations of the PLFAs 18:2ω6 and 16:1ω5, indicative of fungi, monosaturated PLFAS 16:1ω7, 16:1ω9, 18:1ω7, indicative of Gram-negative bacteria and saturated PLFAS 16:0, 17:0 and 18:0. The Eucalyptus soil samples (having negative values in PC1 and positive values en PC2) had relatively high concentrations of mid-chain branched saturated PLFA, 10Me16:0, 10Me17:0, and 10Me18:0, characteristic of actinomycetes, the PLFAs br17:0, i16:1, i16:0, and i17:0, indicative of Gram-positive bacteria and the cy19:0,

2 1.5

A

Eu1

Eu2

Qu1

0.5 Factor 2 (24%)

Factor 2 (21%)

Qu2

Eb2

0.5

1.0

0.0 Eb2

0

Eu2 Qu2

-0.5

Qu1 Qb1

Qb2

Eb1

Qb2 -1.5

Qb1

-1.5

-2 -1.5

-2.0 -1.5

Eu1

-1

-0.5 -1.0

Eb1

1

1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

-1

-0.5

0 0.5 Factor 1 (25%)

1

1.5

2

2.0

Factor 1 (46%) 1.5

B

Factor 3 (16%)

1.0

0.5

Qb2 Eu2 Qu2

Eb2 0.0

Qb1 Eu1

-0.5

Qu1 -1.0

Eb1

-1.5 -1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

Factor 1 (46%)

Fig. 4. Distribution of the soil samples (mean ± SE; n = 4 field plots) in the plane defined by Factors 1 and 2 (A) and 1 and 3 (B) from principal component analysis performed on the physicochemical, chemical and biochemical properties from the unburnt (u) and burnt (b) soil samples studied under Quercus (Q) and Eucalyptus (E) at two depths (1, 0–2.5 cm; 2, 2.5–5 cm).

Fig. 5. Score (mean ± SE; n = 4 field plots) and loading plots from principal components analysis performed on the PLFAs of the unburnt (u) and burnt (b) soil samples studied under Quercus (Q) and Eucalyptus (E) at two depths (1, 0–2.5 cm; 2, 2.5–5 cm) (mean ± SE; n = 4 field plots).

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

A. Lombao et al. / Catena xxx (2014) xxx–xxx

indicative of Gram-negative bacteria. For the same vegetation, higher values along Factor 1 were exhibited by the 0–2.5 cm layer samples with respect to the 2.5–5 cm layer; and by the unburnt soil samples compared with the burnt samples. The results showed that the relative importance of the three factors considered on the PLFA pattern was as follows: soil vegetation ≫ soil depth = soil burning. The results of the biochemical properties indicating that the vegetation is one of the most important factors for determining variance of the parameters related with the microbial activity (13–30% of the variance explained, Fig. 2) seem to support these data. This is also consistent with earlier investigations showing that the microbial component can vary greatly with vegetation (Alvárez et al., 2009; Grayston et al., 1998; Mahía et al., 2006; Priha et al., 2001) and soil depth (Fierer et al., 2003; Fritze et al., 2000; González-Prieto et al., 2013; Mahía et al., 2007; Matinizadeh et al., 2008: Yang et al., 2010). Likewise, other studies also showed immediate or short-term fireinduced changes in microbial community structure were attributed mainly to variations in the soil environment, particularly soil pH and C and nutrient availability, following soil heating, prescribed fires or wildfires (Bååth et al., 1995; Bárcenas-Moreno et al., 2011; Barreiro et al., 2010; Díaz-Raviña et al., 2006). Our data also showed that these shifts in the microbial composition after the fire were accompanied by marked changes in the microbial biomass detected by the fumigation– extraction method and less pronounced changes in the microbial activity measured on the basis of different indices (bacterial growth, respiration, enzymes involved in the C, N and P cycles). These microbial changes (mass, activity, structure) may have important implications for the soil functioning of these protected forest ecosystems, particularly if the changes persist over time; therefore, studies concerning the medium-term evolution of the microbial parameters should be performed in order to preserve soil quality of these forest ecosystems and hence to promote protection of park biodiversity. In addition, it should be noted that the negative fire effects can be more accentuated at medium-term time scale if soil suffers post-fire erosion processes (slope 30–70%, abundant precipitation), which can lead to important soil and nutrient losses. The knowledge about soil microorganisms and soil biological processes may improve the scientific basis for forest management decisions, e.g. the types of species for planting in the natural area and the need to adopt post-fire stabilization and rehabilitation strategies. Soil quality evaluation is a complex undertaking and soil microorganisms are widely recognized as integrative components of the soil quality because of their crucial involvement in many ecosystems processes such as breakdown of organic matter and the net fluxes and amounts of soil organic carbon and nutrients via decomposition, mineralization and immobilization (Nannipieri et al., 2003). In Galicia (NW Iberian Peninsula) examples of biological indicators included several parameters based on the size and activity of microorganisms (Díaz-Raviña et al., 1988; Leirós et al., 2000; Trasar-Cepeda et al., 2000), while measurements of the microbial community structure assessed by means of phospholipid fatty acid pattern (PLFA pattern) and DNA/RNA analysis have received less attention (Mahía et al., 2011). In these forest ecosystems representative of the Fragas do Eume Natural Park, the usefulness of the PLFA data treated with multivariate statistical analysis in exploring the importance of different factors controlling soil quality is shown since data are complementary with the results obtained by the independent measurement of 15 different physicochemical, chemical and biochemical properties combined with principal component analysis (Figs. 4 and 5). It should be noted that the PLFA pattern provides information at the community level on both microbial community structure and biomass estimates reflecting “in situ” conditions; in addition, soil environmental conditions responsible for the observed microbial changes can be identified (Frostegård et al., 2011). Although data should be interpreted with caution, our results support that this quite fast, simple and cheap PLFA method can be considered as a promising early indicator for monitoring changes in soil quality due to forest management.

9

4. Conclusions In the present study the overall quality of the soils of the Fragas do Eume Natural Park following a wildfire was evaluated by multivariate statistical techniques in two ways, using either physicochemical, chemical and biochemical properties (a total of 15 variables) or analysis of the phospholipid fatty acids (35 biomarkers, PLFA pattern). Similar results were observed independently of the multivariate soil status assessment indicating that the quality of the soils is mainly determined by vegetation and only to a lesser extent by soil depth and burning. Soils under Quercus exhibited a higher quality and different microbial community composition than those under Eucalyptus. Most selected soil physicochemical, chemical and biochemical properties have experienced short-term modifications following wildfire; however, the magnitude of these changes and/or high spatial variability among the field plot replicates, makes it difficult to detect significant changes. More marked fire effects were detected on pH, extractable C, microbial biomass C, glucosidase activity and bacterial growth values. Shifts in the structure of the microbial communities (as deduced from their PLFAs patterns) were also observed in the burnt soil samples as compared with the unburnt ones, although fire induced changes had only a minor effect compared to vegetation. The importance of a fire as a disturbance agent in these forest ecosystems and the relevance of the vegetation in determining the soil quality are demonstrated. Acknowledgments This study was supported by the Ministerio Español de Economía y Competitividad (AGL2012-39686-C02-01) and by Fundación MAPFRE (MA-12-AYU-374). A. Barreiro and A. Lombao are recipients of FPU grants from Spanish Ministry of Education (AP2010-2284, AP2010-1900). References Almendros, G., González-Vila, F.J., 2012. Wildfires, soil carbon balance and resilient organic matter in Mediterranean ecosystems. A review. Span. J. Soil Sci. 2, 8–33. Alvárez, E., Torrado, V.M., Fernández-Marcos, M.L., Díaz-Raviña, M., 2009. Microbial biomass and activity in a forest soil under different tree species. Electron. J. Environ. Agric. Food Chem. 8, 878–887. Bååth, E., Frostegård, A., Pennanen, T., Fritze, H., 1995. Microbial community and pH response in relation to soil organic matter quality in wood ash fertilized, clear-cut or burned coniferous forest soils. Soil Biol. Biochem. 27, 229–240. Bååth, E., Petterson, M., Söderberg, K.H., 2001. Adaptation of a rapid and economical microcentrifugation method to measure thymidine and leucine incorporation by soil bacteria. Soil Biol. Biochem. 33, 1571–1574. Bárcenas-Moreno, G., Bååth, E., 2009. Bacterial and fungal growth in soil heated at different temperatures to simulate a range of fire intensities. Soil Biol. Biochem. 41, 2517–2526. Bárcenas-Moreno, G., García-Orenes, F., Mataix-Solera, J., Mataix-Beneyto, J., Bååth, E., 2011. Soil microbial recolonisation after a fire in a Mediterranean forest. Biol. Fertil. Soils 47, 261–272. Barreiro, A., Martín, T., Carballas, T., Díaz-Raviña, M., 2010. Response of soil microbial communities to fire and fire-fighting chemicals. Sci. Total Environ. 408, 6172–6176. Basanta, M.R., Díaz-Raviña, M., González-Prieto, S.J., Carballas, T., 2002. Biochemical properties of forest soils as affected by a fire retardant. Biol. Fertil. Soils 36, 377–383. Basanta, M.R., Díaz-Raviña, M., Cuiñas, P., Carballas, T., 2004. Field data of microbial response to a fire retardant. Agrochimica 48, 51–60. Carballas, T., Martín, A., Díaz-Raviña, M., 2009. Efecto de los incendios forestales sobre los suelos de Galicia. In: Cerdà, A., Mataix-Solera, J. (Eds.), Efectos de los incendios forestales sobre los suelos en España. El estado de la cuestión visto por los científicos españoles. Cátedra Divulgación de la Ciencia. Universitat de Valencia, Valencia, Spain, pp. 269–301. Cerda, A., Jordán, A., 2010. Actualización en métodos y técnicas para el estudio de los suelos afectados por incendios forestales. Cátedra de Divulgación de la Cienica. Universidad de Valencia. FUEGORED 2010, Valencia, Spain. Certini, G., 2005. Effects of fire on properties of forest soils: a review. Oecologia 143, 1–10. Couto-Vázquez, A., González-Prieto, S., 2006. Short- and medium-term effects of three fire-fighting chemicals on the properties of a burnt soil. Sci. Total Environ. 371, 353–361. Díaz-Fierros, F., Benito, E., Pérez, R., 1987. Evaluation of the USLE for prediction of erosion in burnt forest areas in Galicia (NW Spain). Catena 14, 189–199. Díaz-Raviña, M., Carballas, T., Acea, M.J., 1988. Microbial biomass and metabolic activity in four acid soils. Soil Biol. Biochem. 20, 817–823. Díaz-Raviña, M., Prieto, A., Acea, M.J., Carballas, T., 1992. Fumigation–extraction method to estimate microbial biomass in heated soils. Soil Biol. Biochem. 24, 259–264.

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007

10

A. Lombao et al. / Catena xxx (2014) xxx–xxx

Díaz-Raviña, M., Acea, M.J., Carballas, T., 1993a. Seasonal fluctuations in microbial populations and available nutrients in forest soils. Biol. Fertil. Soils 16, 205–210. Díaz-Raviña, M., Acea, M.J., Carballas, T., 1993b. Microbial biomass and its contribution to nutrient concentrations in forest soils. Soil Biol. Biochem. 19, 220–226. Díaz-Raviña, M., Prieto, A., Bååth, E., 1996. Bacterial activity in a forest soil after soil heating and organic amendments measured by the thymidine incorporation techniques. Soil Biol. Biochem. 28, 419–426. Díaz-Raviña, M., Bååth, E., Martín, A., Carballas, T., 2006. Microbial community structure in forest soils treated with a fire retardant. Biol. Fertil. Soils 42, 465–471. Díaz-Raviña, M., Fontúrbel, M.T., Guerrero, C., Martín, A., Carballas, T., 2010. Determinación de propiedades bioquímicas y microbiológicas de suelos quemados. In: Cerdá, A., Jordán, A. (Eds.), Actualización en métodos y técnicas para el estudio de los suelos afectados por incendios forestales. Cátedra de Divulgación de la Ciencia. Universidad de Valencia. FUEGORED, Valencia, Spain, pp. 465–497. Díaz-Raviña, M., Martín, A., Barreiro, A., Lombao, A., Iglesias, L., Díaz-Fierros, F., Carballas, T., 2012. Mulching and seeding treatments for post-fire soil stabilisation in NW Spain: short-term effects and effectiveness. Geoderma 191, 31–39. Eivazi, F., Tabatabai, M.A., 1988. Glucosidases and galactosidases in soils. Soil Biol. Biochem. 20, 601–606. Fernández, I., Cabaneiro, A., Carballas, T., 2001. Thermal resistance to high temperatures of different organic fractions from soils under pine forest. Geoderma 104, 281–298. Fierer, N., Schimel, J.P., Holden, P.A., 2003. Variations in microbial community structure through two soil depth profiles. Soil Biol. Biochem. 35, 167–176. Fontúrbel, M.T., Barreiro, A., Vega, J.A., Lombao, A., Martín, A., Jiménez, E., Carballas, T., Fernández, C., 2012. Effects of an experimental fire and post-fire stabilisation treatments on soil microbial communities. Geoderma 191, 51–60. Fritze, H., Pietikånen, J., Pennanen, T., 2000. Distribution of microbial biomass and phospholipid fatty acids in Podzol profiles under coniferous forest. Eur. J. Soil Sci. 51, 561–573. Frostegård, A., Bååth, E., Tunlip, A., 1993. Shifts in the structure of soil microbial communities in limed soils as revealed by phospholipid fatty acid analysis. Soil Biol. Biochem. 25, 723–730. Frostegård, A., Tunlip, A., Bååth, E., 2011. Use and misuse of PLFA measurements in soils. Soil Biol. Biochem. 43, 1621–1625. González-Prieto, S.J., Villar, M.C., Carballas, M., Carballas, T., 1996. Effect of soil characteristics on N mineralization capacity in 112 native and agricultural soils from the northwest of Spain. Biol. Fertil. Soils 22, 252–260. González-Prieto, S.J., Díaz-Raviña, M., Martín, A., López-Fando, C., 2013. Effects of agricultural management on chemical and biochemical properties of a semiarid soil form central Spain. Soil Tillage Res. 134, 49–55. Grayston, S.J., Wang, S., Campbell, C.D., Edwards, A.C., 1998. Selecting influence of plant species on microbial diversity in the rhizosphere. Soil Biol. Biochem. 30, 369–378. Guitián-Ojea, F., Carballas, T., 1976. Técnicas de Análisis de Suelos. Editorial Pico Sacro, Santiago de Compostela, Spain. Hernández, T., García, C., Reinhardt, I., 1997. Short-term effect of wildfire on the chemical, biochemical and microbiological properties of Mediterranean forest soils. Biol. Fertil. Soils 25, 109–116. Holden, S., Treseder, K.K., 2013. A meta-analysis of soil microbial biomass responses to forest disturbances. Front. Microbiol. 4 (163), 1–17. Kandeler, E., Gerber, H., 1988. Short-term assay of soil urease activity using colorimetric determination of ammonium. Biol. Fertil. Soils 6, 68–72. Karlen, D.L., Mausbach, M.J., Doran, J.W., Cline, R.G., Harris, R.F., Schuman, G.E., 1997. Soil quality: a concept, definition and framework for evaluation. Soil Sci. Soc. Am. J. 61, 4–10. Leirós, M.C., Trasar-Cepeda, C., Seoane, S., Gil-Sotres, F., 2000. Biochemical properties of acid soils under climax vegetation (Atlantic oakwood) in an area of European temperate-humid zone (Galicia, NW Spain): general parameters. Soil Biol. Biochem. 32, 733–745. Lie, G., Ye, L., Bao, H., Chen, H., Xue, L., 2014. The effects of a wildfire on soil chemical and biological properties in an Eucalyptus urophylla forest. Adv. Mater. Res. 864–867, 2492–2497. Mahía, J., Pérez-Ventura, L., Cabaneiro, A., Díaz-Raviña, M., 2006. Soil microbial biomass under pine forests in the north-west Spain: influence of stand age, site index and parent material. Investig. Agrar. Sist. Recur. For. 15, 152–159. Mahía, J., Martín, A., Carballas, T., Díaz-Raviña, M., 2007. Atrazine degradation and enzyme activities in an agricultural soil under two tillage systems. Sci. Total Environ. 378, 187–194.

Mahía, J., González-Prieto, S.J., Martín, A., Bååth, E., Díaz-Raviña, M., 2011. Biochemical properties and microbial community structure of five different soils after atrazine addition. Biol. Fertil. Soils 47, 577–589. Martín, A., Díaz-Raviña, M., Carballas, T., 2009. Evolution of composition and content of soil carbohydrates following forest wildfires. Biol. Fertil. Soils 45, 511–520. Martín, A., Díaz-Raviña, M., Carballas, T., 2011. Seasonal changes in the carbohydrates pool of an Atlantic forest under different vegetation types. Span. J. Soil Sci. 1, 38–53. Martín, A., Díaz-Raviña, M., Carballas, T., 2012. Short- and medium-term evolution of soil properties in Atlantic forest ecosystems affected by wildfires. Land Degrad. Dev. 23, 427–439. Mataix-Solera, J., Guerrero, C., García-Orenes, F., Bárcenas, G.M., Torres, M.P., 2009. Forest fire effects on soil microbiology. In: Cerdá, A., Robichaud, P.R. (Eds.), Fire Effects on Soils and Restoration Strategies. USA, Science Publishers, Enfield, pp. 133–175. Matinizadeh, M., Korori, S.A.A., Teimouri, M., Praznik, W., 2008. Enzyme activities in undisturbed and disturbed forest soils under oak (Quercus brantii var. persica) as affected by soil depth and seasonal variation. Asian J. Plant Sci. 7, 368–374. Nannipieri, P., Ascher, J., Ceccherini, M.T., Landi, L., Pietramellara, G., Renella, G., 2003. Microbial diversity and soil functions. Eur. J. Soil Sci. 54, 655–670. Neary, D.G., Klopatek, C.C., DeBano, L.F., Ffolliott, P.F., 1999. Fire effects on belowground sustainability: a review and synthesis. For. Ecol. Manag. 122, 51–71. Pérez-Cruzado, C., Mansilla-Salinero, P., Rodriguez-Soalleiro, R., Merino, M., 2012. Influence of tree species on carbon sequestration in afforested pastures in a humid temperate region. Plant Soil 353, 333–353. Pourreza, M., Hosseini, S.M., Sinegani, A.A.S., Matinizadeh, M., Dick, W.A., 2014. Soil microbial activity in response to fire severity in Zagros oak (Quercus brantii Lindl.) forests, Iran, after one year. Geoderma 213, 95–102. Prieto-Fernández, A., Acea, M.J., Carballas, T., 1998. Soil microbial and extractable C and N after wildfire. Biol. Fertil. Soils 27, 132–142. Priha, O., Grayston, S.J., Hiukka, R., Pennanen, T., Smolander, A., 2001. Microbial community structure and characteristics of the organic matter in soils under Pinus sylvestris, Picea abies and Betula pendula at two forest sites. Biol. Fertil. Soils 33, 17–24. Rovira, P., Romanyà, J., Duguy, B., 2012. Long-term effects of wildfires on the biochemical quality of soil organic matter: a study on Mediterranean shrublands. Geoderma 179–180, 9–19. Santín, C., Knicker, H., Fernández, S., Menéndez-Duarte, R., Álvarez, M.A., 2008. Wildfires influence on soil organic matter in an Atlantic mountainous region (NW of Spain). Catena 74, 286–295. Shakesby, R.A., 2011. Post-wildfire soil erosion in the Mediterranean: review and future research directions. Earth-Sci. Rev. 105, 71–100. Trasar-Cepeda, M.C., Gil Sotres, F., Guitián Ojea, F., 1985. Determinación de la actividad de la fosfatasa en suelos gallegos. Precisiones al método de Saratchandra y Perrott. An. Edafol. Agrobiol. 44, 987–991. Trasar-Cepeda, C., Leirós, M.C., Gil-Sotres, F., 2000. Biochemical properties of acid soils under climax vegetation (Atlantic oakwood) in an area of European temperatehumid zone (Galicia, NW Spain): specific parameters. Soil Biol. Biochem. 32, 747–755. Vázquez, F.J., Acea, M.J., Carballas, T., 1993. Soil microbial population after wildfire. FEMS Microbiol. Ecol. 13, 93–104. Vega, J.A., Fontúrbel, T., Fernández, C., Arellano, A., Díaz-Raviña, M., Carballas, M.T., Martín, A., González-Prieto, S., Merino, A., Benito, E., 2013a. Acciones urgentes contra la erosión en áreas forestales quemadas. Guía para su planificación en Galicia. Tórculo Artes Gráficas, Santiago de Compostela, Spain. Vega, J.A., Fontúrbel, T., Merino, A., Fernández, C., Ferreiro, A., Jiménez, E., 2013b. Testing the ability of visual indicators of soil burn severity to reflect changes in soil chemical and microbial properties in pine forests and shrubland. Plant Soil 369, 73–91. Villar, M.C., Petrikova, V., Díaz-Raviña, M., Carballas, T., 2004. Changes in soil microbial biomass and aggregate stability following burning and soil rehabilitation. Geoderma 122, 73–82. Wardle, D.A., 1992. A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soil. Biol. Rev. 67, 321–358. Yang, K., Zhu, J., Zhang, M., Yan, Q., Sun, O.J., 2010. Soil microbial biomass carbon and nitrogen in forest ecosystems of Northeast China: a comparison between natural secondary forest and larch plantation. J. Plant Ecol. 3, 175–182.

Please cite this article as: Lombao, A., et al., Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain), Catena (2014), http://dx.doi.org/10.1016/j.catena.2014.08.007