Application of a Soil Quality Index to a Mediterranean Mountain with Post-Fire Treatments

Abstract

In Mediterranean areas, fire has increased soil degradation and erosion. For this reason, the application of soil quality indices can help to determine soil recovery and degradation levels. By using a multiparameter soil quality index fitted to undisturbed forest soils, we can show the right information on soil functionality. In this study, the objectives were to evaluate soil functionality after suffering a fire, to subsequently treat with various soil treatments (wood mulching), and then check a soil quality index (SQI) to assess the soil quality recovery in burned Pinus nigra stands. For this purpose, a burned area was selected in the Cuenca Mountain range (Spain) under a Mediterranean climate. Experimental plots were established in the study area, with three slope gradients and applying three methods of covering soil using: (1) wood chips; (2) piles of branches; and (3) trunks of contour-felled logs. The experiment was conducted for 4 years. Results showed that the properties of soil were enhanced under wood chips and logs as a surface-covering material, and in a short time (less of 3 years). In consequence, the values of the SQI index were higher after applying these two treatments, thus reflecting the effectiveness of the SQI for monitoring post-fire recovery.

1. Introduction

After creating a multiparameter soil quality index, it is necessary to validate its efficacy and sensitivity, both in undisturbed and disturbed soils [1]. A high-standard soil quality index must be able to both measure the imbalance produced among the soil’s edaphic properties [2], while being sensitive to detecting disturbances.

In the validation process of a quality index, it is important to consider its functionality [3]. Andrews et al. [4] focused on productivity, nutrient recycling, or environmental protection. Bastida et al. [5] based their study on the microbial functionality of the soil to carry out basic processes related to its sustainability. Other authors have highlighted other functions such as the transfer and absorption of water, the ability to resist degradation, or the capacity for plant growth [6]. In our case, the index was based on the soil’s biological functionality, and consequently on its capacity to maintain microbiological and physicochemical stability.

Forests in the Mediterranean area have a special relationship with disturbances caused by fire [7], it being also one of the main triggers which allow vegetation and landscape dynamics processes to occur [8]. This phenomenon can modify the soil’s chemical and physical content, the structure of it nutrient cycles, organic matter, and pH or hydrophobicity, in addition to its microbiological and biochemical properties. In this scenario, where erosion is a risk, it is possible to control the effects of forest fires by applying emergency post-fire soil treatments [9,10,11]. Success in using various management techniques to aid natural regeneration after a forest fire depends on the own fire’s nature and the original soil’s characteristics [12]. Addition of organic matter allows an increase in the amount of retained carbon by the soil as well as in its productivity and biodiversity. In general, addition of organic matter to degraded soils can cause an enhancement in physicochemical soil properties (greater infiltration, more water retention capacity, a better structural and chemical stability), which can be also translated as an improvement of its microbiological and biochemical properties and, therefore, as an increase in its productivity [13]. Controversy appears when determining the role of burned wood in this matter. For this reason, post-fire wood management has been a topic of intense and continuous scientific discussion. In Spain, the felling and removal of leftovers after a fire is a common practice [14]. However, several studies show that the extraction of burned wood can produce several negative impacts from an environmental perspective: seed contribution, and changes in microclimatic conditions [15,16].

Furthermore, the period of time between the fire and the post-fire treatment must be considered as an element to properly assess the impact of fires and post-fire treatments on a soil’s functionality, particularly in Mediterranean forest ecosystems [17,18].

Based on the hypothesis that microbiological, biochemical, and physicochemical properties of burned soils—and therefore the quality index that relates them—are improved with the application of post-fire forest remains, the aims of this research were: to evaluate soil properties after having suffered a fire, to subsequent treat with various silvicultural treatments (wood mulch) in order to achieve soil recovery, and to check a soil quality index (SQI) to assess the soil recovery in burned Pinus nigra stands subjected to the soil treatments. This research will allow us to make post-fire treatment recommendations to help improve soil properties in the post-fire stages.

2. Materials and Methods

2.1. Study Area and Experimental Design

The studied area is located within the “Dehesa de Don Juan” mountain (between 40°0′17.9″ N–40°1′23.48″ N and 1°50′55.49″ W–1°50′41.45″ W), south of the Cuenca Mountain range (eastern central Spain) (Figure 1).

The annual average temperature is 12.2 °C while average rainfall is 384.7 mm, hence this type of climate is classified as Mediterranean continentalized. High temperatures and drought (with eventual storm events) characterize the climate in summer. Seasonal data of average temperature and precipitation for the study period (2012–2015) are shown in

Table 1. Seasonal data of values of temperature and precipitation.

Year	Season	Mean Temperature (°C)	Rainfall (mm)
2012	Summer	18.11	37
2013	Spring	14.98	162
2014	Summer	24.81	56
2015	Spring	18.48	98

(SIAR; Advisory Service to the Irrigator of Castilla-La Mancha Region, Albacete, Spain).

According to the soil classification, the dominant substrate is categorized as Lithic Xerorthent [19], formed in the upper Cretaceous of the Turonian–Coniacian by dolomites and white dolomitic marls. Forests are mainly dominated by Spanish black pine (Pinus nigra Arn. ssp. salzmannii) and the dominant shrub species are junipers (Juniperus thurifera L.), Portuguese oak (Quercus faginea) and holly (Ilex aquifolium).

In September 2011, the Dehesa de Don Juan mountains suffered a wildfire, considered of high severity. Three months later, post-fire management trials were initiated. The most frequent treatments were assessed, including the application of various covers or mulch on soil [20]: (i) contour-felled logs or log erosion barriers, where burned trees are cut down and delimbed boles are placed on the slope, (ii) covering soil with wood chips (homogenous distribution of chips, 1 cm deep), and (iii) crown debris in piles of burned branches. These treatments can be built to reduce erosion and runoff in burned pine stands, because they trap runoff and sediment.

To develop the experimental design, three study plots of 1 ha each were established depending on the slope: No Gradient (NG; with a slope < 1% and an altitude of 1035 m.a.s.l.); Medium Gradient (MG; with a 13% gradient and an altitude of 1049 m.a.s.l.); and High Gradient (HG; with a 30% gradient and an altitude of 1089 m.a.s.l.). Within these plots, another 4 subplots of 30 × 30 m were established, according to the 4 types of post-fire treatment (Figure 2 and Figure 3): NT, control zone without treatment with untouched wood from burned trees; TC, zone clearcutting and chippings of posterior leftovers spread on the ground (“wood chips”); TB, zone clearcutting and posterior dispersal of branches on the ground (“piles of branches”); TL, zone clearcutting and posterior dispersal of 2 m logs in different orientations (“contour-felled logs”). Each plot was spaced out enough to avoid pseudoreplication.

2.2. Soil Sampling

Soil sampling was carried out from of 2012 to 2015. Three soil samples in every subplot were taken, each composed of 6 homogeneously mixed subsamples to minimize the spatial variability of the soil. Soil samples were taken from the upper 10–15 cm layer, in summer 2012, spring 2013, summer 2014 and spring 2015. The first sampling was conducted seven months after the post-fire mulching, in order to observe the treatments’ first effects. The following soil samplings were planned annually but taking into account that spring is a very active season from the point of view of soil microbiological and enzymatic activities. For this reason, the soil samplings were carried out in late spring or early summer (to include the effects of rainfall). By sampling in 2 seasons for 4 years, the great variability of the Mediterranean climate was included in the research. Results in this paper reflect the mean values of the sampling sites’ plots with different slopes. Vegetal remains were removed to prevent them from influencing the measurements made, and the soil samples were sieved (2 mm). Soil moisture content was also measured immediately after sieving, followed by determination of its biological and biochemical properties. Lastly, physicochemical, microbiological, and biochemical soil properties were determined (cited in Section 2.3). For the time period and study area, the total number of samples that were taken was 144 (36 for each subplot and year).

2.3. Physicochemical, Microbiological, and Biochemical Characterization of Treated Soils

A gravimetric method was used to determine soil moisture content (M) (115 °C for 24 h) while in an aqueous solution (1:5 soil) a pH-meter (Navi Horiba model, HORIBA Ltd., Kyoto, Japan) was used to obtain electrical conductivity and pH values. Walkley–Black’s and Kjeldhal’s (modified by Bremmer) methods [21,22] were chosen to determine total organic carbon (TOC) and total nitrogen content (N), respectively. The available phosphorus (P) was extracted by using Olsen and Sommers’ method [23]. Determination of carbonates (CARB) was conducted in an acid-base colorimetric titration, using phenolphthalein as an indicator (HCl was added to the sample, reacting with carbonates. Then, excess acid was neutralized with a base, NaOH; when the color of the solution turns to fuchsia, the milliliters of NaOH used in the process are the basis for obtaining the carbonate values contained in the grams of analyzed soil). Recorded data related to soil temperature were obtained with the help of a soil temperature sensor (TMC6-HD, Tempcon Instrumentation Ltd., Ford, UK). To obtain basal soil respiration (BR), a 20-day incubation was carried out at 28 °C, measuring released CO₂ regularly with aid from an infrared gas analyzer (Toray PG-100, Toray Engineering Co. Ltd., Tokyo, Japan) (readings were collected once a day for the first four days, then weekly until the end). A fumigation–extraction method adapted by García et al. [24] was followed to measure microbial biomass carbon (MBC). The dehydrogenase activity (DHA) procurement process entailed reducing 2-p-iodophenyl-3-p-nitro-phenyl-5-phenyltetrazolium chloride (INT) to iodonitrophenyl formazan (INTF) and calculating its value aided by a spectrophotometer at 490 nm (Helios Alpha 100–240, Thermo Electron Corporation, Gloucester, UK) [24]. The buffered hydrolysis reaction method modified by Kandeler et al. [25] was applied to obtain urease activity (UA); NH₄⁺ released from the reaction was analyzed and the absorbance of the supernatant at 690 nm was determined with a spectrometer. Alkaline phosphatase (APA) and β-glucosidase activity (β-GLU) were determined as described by Tabatabai and Bremner [26], but without using p-nitrophenyl phosphate p-nitrophenyl-β-D-glucopyranoside as a substrate. The C/N ratio, the metabolic coefficient [(q (CO₂) = BR/MBC], and the mineralization coefficients [q(mC) = BR/TOC] are obtained from the quotient between the parameters that form these simple indices.

2.4. Soil Quality Index Applied (SQI)

The quality index that was validated in these study areas was the one obtained by Andres-Abellán et al. [27] on undisturbed Mediterranean forest soils through the polynomial function expressed in Equation (1):

$$\begin{array}{c}\mathrm{S}\mathrm{Q}\mathrm{I}=0.576\times \left[0.489\times \left(\frac{1}{1+{\left(\frac{1308-16.31}{\mathrm{M}\mathrm{B}\mathrm{C}-16.31}\right)}^2}\right)+0.459\times \left({\mathrm{e}}^{-\left(\frac{{\left(\mathrm{M}-39\right)}^2}{2\mathrm{}·\mathrm{}{11.2}^2}\right)}\right)+0.445\times \left(\frac{1}{1+{\left(\frac{50-19.61}{\mathrm{T}\mathrm{O}\mathrm{C}-19.61}\right)}^2}\right)\right]+\\ 0.228·\left[-0.602\times \left(\frac{1}{1+{\left(\frac{121.9}{\mathrm{A}\mathrm{P}\mathrm{A}}\right)}^{1.7}}\right)+0.510\times \left(\frac{1}{1+{\left(\frac{197.2}{\mathsf{\beta }-\mathrm{G}\mathrm{L}\mathrm{U}}\right)}^{1.7}}\right)\right]+0.196\times \left[0.831\times \left({\mathrm{e}}^{-\left(\frac{{\left(\mathrm{p}\mathrm{H}-6\right)}^2}{2\mathrm{}·\mathrm{}{0.59}^2}\right)}\right)\right]\end{array}$$

(1)

Andrews et al.’s method [4] was the foundation used to conceive the SQI, replicating a similar procedure employed by other authors over different types of soil analyses [28]. This method consists of: (i) selection of the parameters, (ii) transformation and the weighting of the values, and (iii) the weighted combination in an index.

This soil index (SQI) was obtained from unaltered soil samples in the Cuenca Mountain range, where Monte Dehesa de Don Juan is included. Thus, the experimental area in the present research had similar environmental characteristics and principal species (Pinus nigra) to those used to obtain the SQI. The index incorporates a number of variables which were selected as having both the largest eigenvector values on these axes and weak correlations between them. In order to accomplish this, the following procedure was followed (see Andrés et al. [27] for further details). (I) Recorded data for the set of variables (moisture M, pH, electrical conductivity EC, carbonates CARB, total organic carbon TOC, total nitrogen N, phosphorus P, basal soil respiration BR, microbial biomass carbon MBC, urease activity UA, alkaline phosphatase activity APA, β-glucosidase activity β-GLU, and dehydrogenase activity DHA). (II) A general principal component analysis (PCA) was carried out initially with the entire group of variables (13 of them) selecting just those showing the largest eigenvectors for each of the three principal components (PCs). The followed criteria included the selection of variables that accounted for 90% or more of the maximum eigenvector observed for each PC [29]. (III) Uncorrelation between variables was not assured if this criterion was used. Later, a second PCA was realized with only the group of variables selected previously, by specifying which were strongly correlated in order to eliminate them and retaining the most suitable ones. (IV) Lastly, a third PCA was conducted with the last set of chosen variables remaining. The objective was focused on obtaining the variables that would shape the SQI from the largest eigenvectors of the first three PCs. Previously selected variables were standardized using a transformation functions “the more the better” [30] and “defines a Gaussian-type curve” [31]. (V) The SQI was defined by a linear combination of the PC scores applied to the transformed variables and weighted by the variance explained by each PC.

2.5. Statistical Analysis

An analysis of variance GLM was computed over soil properties and SQI values. The effects of 3 main factors on dependent variables were analyzed: (i) treatment, with 4 levels (NT: Non treatment; TC: Treatment with wood chips; TB: Treatment with piles of branches; TL: Treatment with logs), (ii) time, with 4 levels (RT: Recent treatment, within 7 months; T1: second year, 2013; T2: third year, 2014; T3: fourth year, 2015), and (iii) gradient, with 3 levels (NG: No gradient, 1%; MG: Medium gradient, 13%; HG: High gradient, 30%). Fisher’s least significant difference (LSD procedure) was used to test the differences among the means of levels (95% probability; p < 0.05). To ensure the equality of variances and normal distribution, the square root was applied to the different variables when necessary. Statgraphics Centurion version XVI^® (Stat-Point Technologies Inc., Warrenton, VA, USA) was used as the software for data analysis.

3. Results

3.1. Physicochemical, Microbiological, and Biochemical Characterization of Treated Soils

In general, important physicochemical properties of soils (M: soil moisture content; TOC: total organic C; N: nitrogen content) were improved in the treated areas (especially in the wood chip mulching), with different degrees of significance. On the contrary, pH significantly decreased when covering the soil with branches and logs (

Table 2. Mean ± standard error (ANOVA analysis) values for each variable and experimental condition. For each soil property and factor (treatment, time, and gradient), the means with the same letter were not significantly different (at the 0.05 level, according to the least significant difference, LSD test).

Soil Properties a	Treatment b				Time c				Gradient d
	NT	TC	TB	TL	RT	T1	T2	T3	NG	MG	HG
M (%)	13.17 ± 0.48 a	15.57 ± 0.48 bc	14.59 ± 0.48 bc	16.67 ± 0.48 c	2.57 ± 0.47 a	21.24 ± 0.47 c	20.68 ± 0.47 c	15.4 ± 0.47 b	15.76 ± 0.41 b	14.09 ± 0.41 a	15.13 ± 0.41 b
pH	8.33 ± 0.01 b	8.31 ± 0.01 b	8.28 ± 0.01 a	8.30 ± 0.01 ab	8.11 ± 0.01 b	7.94 ± 0.01 a	8.46 ± 0.01 c	8.68 ± 0.01 d	8.31 ± 0.01 a	8.31 ± 0.01 a	8.31 ± 0.01 a
EC (mS·m−1)	27.98 ± 1.12 a	23.64 ± 1.12 ab	25.42 ± 1.12 ab	25.81 ± 1.12 b	43.30 ± 1.12 c	22.30 ± 1.12 b	20.99 ± 1.12 b	16.26 ± 1.12 a	23.86 ± 0.90 a	26.19 ± 0.90 ab	27.09 ± 0.90 b
TOC (%)	8.77 ± 0.33 a	9.75 ± 0.33 b	8.99 ± 0.33 ab	9.49 ± 0.33 ab	10.92 ± 0.33 c	13.46 ± 0.33 d	7.92 ± 0.33 b	4.76 ± 0.33 a	10.03 ± 0.29 b	9.19 ± 0.28 b	8.52 ± 0.29 a
N (%)	0.26 ± 0.01 a	0.29 ± 0.01 b	0.29 ± 0.01 b	0.30 ± 0.01 b	0.24 ± 0.01 a	0.38 ± 0.01 b	0.26 ± 0.01 a	0.25 ± 0.01 a	0.30 ± 0.01 b	0.30 ± 0.01 b	0.26 ± 0.01 a
P (ppm)	33.64 ± 2.59 a	37.65 ± 2.59 a	39.06 ± 2.59 a	35.82 ± 2.59 a	76.85 ± 2.59 c	19.07 ± 2.59 a	29.62 ± 2.59 b	20.63 ± 2.59 a	38.39 ± 2.24 a	35.79 ± 2.24 a	35.45 ± 2.24 a
CARB (mg CO32−·g−1)	0.18 ± 0.01 a	0.18 ± 0.01 a	0.18 ± 0.01 a	0.19 ± 0.01 a	0.19 ± 0.01 b	0.24 ± 0.01 c	0.15 ± 0.01 a	0.15 ± 0.01 a	0.17 ± 0.01 a	0.19 ± 0.01 a	0.18 ± 0.01 a
BR (µg C-CO2 g−1·day−1)	9.09 ± 0.47 a	11.58 ± 0.47 b	11.67 ± 0.47 b	11.92 ± 0.47 b	15.40 ± 0.48 c	8.80 ± 0.47 a	10.78 ± 0.47 b	9.26 ± 0.47 a	11.38 ± 0.40 b	11.28 ± 0.40 b	10.52 ± 0.40 a
MBC (μg C·g−1)	388.1 ± 45.7 a	524.8 ± 45.7 c	441.1 ± 45.7 b	417.8 ± 45.7 b	114.7 ± 45.7 a	257.1 ± 45.7 b	675.2 ± 45.7 c	924.7 ± 45.7 d	493.5 ± 39.9 c	421.1 ± 39.9 b	414.2 ± 39.9 b
DHA (μmol (INTF)·g−1·h−1)	0.02 ± 0.001 a	0.02 ± 0.001 a	0.02 ± 0.001 a	0.02 ± 0.001 a	0.03 ± 0.001 b	0.02 ± 0.001 a	0.02 ± 0.001 a	0.02 ± 0.001 a	0.02 ± 0.001 a	0.03 ± 0.001 a	0.02 ± 0.001 a
UA (μmol (N-NH4+)·g−1·h−1)	2.11 ± 0.18 ab	2.32 ± 0.18 b	1.74 ± 0.18 a	1.95 ± 0.18 ab	1.77 ± 0.18 ab	1.42 ± 0.18 a	2.13 ± 0.18 b	2.80 ± 0.18 c	1.99 ± 0.16 a	2.12 ± 0.16 a	1.98 ± 0.16 a
APA (μmol (PNP)·g−1·h−1)	10.36 ± 0.80 b	8.43 ± 0.80 a	10.08 ± 0.80 b	8.51 ± 0.80 a	13.34 ± 0.08 d	5.08 ± 0.08 a	8.31 ± 0.09 b	10.64 ± 0.80 c	8.31 ± 0.70 a	8.98 ± 0.70 ab	10.75 ± 0.70 b
β-GLU (μmol (PNP)·g−1·h−1)	39.79 ± 0.95 b	40.40 ± 0.95 b	25.85 ± 0.95 a	38.84 ± 0.95 b	45.98 ± 1.00 b	88.91 ± 1.00 c	5.40 ± 0.71 a	5.40 ± 0.71 a	49.26 ± 0.82 b	30.61 ± 0.82 a	28.79 ± 0.82 a
C/N	33.63 ± 1.21 b	33.51 ± 1.21 b	31.56 ± 1.21 a	31.56 ± 1.21 a	45.43 ± 1.21 d	36.25 ± 1.21 c	29.52 ± 1.21 b	19.05 ± 1.21 a	30.80 ± 1.05 a	32.47 ± 1.05 b	34.43 ± 1.05 b
q (CO2) (µg C-CO2·µg−1 C·day−1)	0.28 ± 0.001 a	0.22 ± 0.001 a	0.23 ± 0.001 a	0.26 ± 0.001 a	0.08 ± 0.001 b	0.06 ± 0.001 b	0.02 ± 0.001 a	0.01 ± 0.001 a	0.02 ± 0.001 a	0.03 ± 0.001 b	0.03 ± 0.001 b
q (mC) (µg C-CO2·mg−1 COT·day−1)	1.15 ± 0.06 a	1.30 ± 0.06 ab	1.49 ± 0.06 c	1.41 ± 0.06 c	0.87 ± 0.06 a	1.14 ± 0.06 b	1.45 ± 0.06 c	1.90 ± 0.06 d	1.39 ± 0.05 b	1.36 ± 0.05 b	1.27 ± 0.05 a

^a Soil properties—M: moisture; pH: soil acidity; EC: electrical conductivity; TOC: total organic carbon; N: total nitrogen; P: phosphorus; CARB: Carbonates; BR: basal soil respiration; MBC: microbial biomass carbon; DHA: dehydrogenase activity; UA: urease activity; APA: phosphatase activity; β-GLU: β-glucosidase activity; C/N: carbon–nitrogen ratio; q (CO₂)): metabolic coefficient; q (mC): coefficient of mineralization. ^b Treatments—NT: Non treatment; TC: Treatment with wood chips; TB: Treatment with branches; TL: Treatment with logs. ^c Time—RT: Recent treatment (7 months); T1: Treatment in the second year (2013); T2: Treatment in the third year (2014); T3: Treatment in the fourth year (2015). ^d Gradient—NG: Not gradient (1%); MG: Medium gradient (13%); HG: High gradient (30%). The slope is in %.

).

Referring to microbiological and biochemical properties, the variables BR (basal respiration), MBC (microbial biomass C), and UA (urease activity) were also significantly higher in plots treated with some type of wood mulch, compared to non-treated ones. The increment in these variables revealed an increase in soil activity in the treated soils with wood mulching (Table 2). In this sense, the most consistent result within treatments showed that wood chips improved all the referred variables, and with the maximums reached for BR (basal respiration; 11.58 ± 0.47 µg C-CO₂ g⁻¹·day⁻¹), MBC (microbial biomass C; 524.8 ± 45.7 μg C·g⁻¹), and UA (urease activity; 2.32 ± 0.18 μmol (N-NH₄⁺)·g⁻¹·h⁻¹). However, these positive results were not significantly distinct for the variables BR and UA when contour-felled logs on soil were applied. These results highlighted that wood mulching methods (especially wood chips as covering material) can help to improve soil activity in burned Pinus nigra stands. In addition, wood acts as a surface-covering material forming a post-fire treatment, which allows soil functions to recover.

Analyzing the sampling time, it was observed that M (soil moisture content), pH, MBC (microbial biomass C), UA (urease activity), and APA (alkaline phosphatase) increased over the sampling time. Thus, a general improvement in the microbiological activity of treated soils was confirmed at the end of the research (year 2015).

Focusing on slope effects, a significant decrease in the mean values of TOC (total organic C), N (nitrogen content), BR (basal respiration), MBC (microbial biomass C), and β-GLU (β-glucosidase) was observed while the gradient increased. For the rest of the physicochemical properties and enzymatic activities analyzed, their mean values maintained constant (pH, DHA, UA) or slightly increased values along with the slope (Table 2).

In addition, treatment and time caused an increase in the mineralized coefficient [q(mC)], therefore decreasing the C/N ratio (indicating a higher relative liberation of nitrogen), which also resulted in a reduction in the metabolic coefficient [q (CO₂)] in function of time. The slope produced opposite effects on the average values of these coefficients; higher slopes led to an increase in the q (CO₂) and C/N ratios, hence causing a reduction in q(mC) (Table 2).

3.2. Application of Soil Quality Index (SQI)

The results showed that the factors “treatment”, “time period” and “gradient” have a significant influence (p < 0.05) on the soil quality index (SQI) values, explaining 72.51% of its total variation (

Table 3. Significance level of three factors (Treatment, Time, Gradient) affecting the SQI variable. Fitness level of the full model (F: F-Snedecor, R²: coefficient of determination, SEE: standard error of estimation). The factors (LSD test) were significant when p < 0.05.

Variables	Treatment		Time		Gradient		Full Model
	F	p	F	p	F	p	F	R2 (%)	SEE
SQI	5.4	<0.005	97.78	<0.001	4.12	<0.05	23.74	72.51	0.04

).

In accordance with the ANOVA and the results included in Table 3, it was observed that in the treated areas with the addition of fire remnants, the SQI values were higher than in the untreated area (NT; Figure 4a). In addition, those treatments using wood chippings and logs on the ground showed higher values of the SQI (SQI = 0.135 ± 0.0067 and 0.133 ± 0.0067, respectively; mean ± standard error). Thus, the soil quality given by cover with wood chips was similar to that given by cover with felled logs, verifying that both soil treatments are adequate for improving soil quality in accordance with the results obtained in the physicochemical, microbiological, and biochemical characterization of treated soils.

Moreover, two years after the treatments, the average value of the quality index increased significantly (Figure 4b), from 0.022 ± 0.0067 (RT; recent treatment, 7 months) to 0.150 ± 0.0071 (T1; treatment in the second year, 2013), showing the highest value three years after treatment (0.169 ± 0.0067; T2, treatment in the third year, 2014). These results add to the growing weight of evidence that mulch applications are effective when reducing previous soil disturbances in the initial post-fire stage.

In the case of the gradient factor (slope), the highest mean value of the SQI was found close to zero for the slope (NG; 0.134 ± 0.006), decreasing its value in areas with medium or high slopes, between which the difference was not significant (0.111 ± 0.0058 in medium gradient, and 0.118 ± 0.0060 in high gradient) (Figure 4c).

In general, all the values obtained for the SQI index were low (<0.20; Figure 4). Based on our previously published work [27] and the results obtained for the SQI in the undisturbed areas, percentiles of 0.25, 0.5, and 0.75 were established for different levels of soil quality. Thus, values for the SQI index of 0.204, 0.274, and 0.345 were set, respectively, as interval limits for soil quality. In the present research, when the SQI was applied in the treated plots, the soil quality values were less than 0.20 in all cases (Figure 4). Due to that, these values indicated low-quality soils because of forest fire impacts. However, it should be highlighted that the SQI was higher in treated plots (wood mulching, and especially in the treatment with a cover of wood chips). This is accordance with the results obtained in the physicochemical, microbiological and biochemical characterization of treated soils. In conclusion, our results demonstrated that post-fire management treatments on soils improved soil quality (although their effects were clearer at areas with lower slope), and the soil recovery in function of soil treatment was also reflected in the value of the soil index.

4. Discussion

4.1. Physicochemical, Microbiological, and Biochemical Characterization of Treated Soils

Those plots treated after the fire with burned wood remains have shown a general improvement in the average values of their main properties, which is directly related to an increase in the amount of plant biomass, thus modifying the amount of organic matter in the plots. A higher content of organic matter in the soil has a positive effect on soil biology, increasing the concentration of nutrients, which triggers enzyme activity, BR, and MBC associated with the time elapsed since treatment [32]. In addition, microclimatic conditions of the soil change with less direct light radiation, along with variations in temperature and humidity, improving climatic conditions for microorganisms [27].

The addition of larger wood remains produces a beneficial effect on the soil properties as they improve soil moisture and reduce erosion and compaction, while maintaining an optimal temperature without fluctuations [33,34]. Comparatively, wood chips are slender, having a lower C/N ratio than branches or trunks, and this causes quicker decomposition [35]. The application of woodchips has an effect similar to mulching, as it lowers soil temperature and evapotranspiration [35,36]. The cover of wood chips also increases soil nitrogen [36,37].

Time also produces an increase in the availability of nutrients after the fire, which is associated with ash depositions when organic matter combustion occurs. Because of that, physicochemical properties and enzymatic activity are raised significantly post-fire, since both properties are very sensitive to changes in the soil environment [32,38]. Organic matter decomposition causes enzyme activity and microorganisms to activate [39].

Khalili-Rad et al. [40] have shown that the biological and biochemical properties of semi-arid soils depend strongly on slope. Furthermore, it was concluded that low-slope areas generally provide greater sources of carbon and energy for soil microorganisms, thus resulting in greater microbial biomass and activity (according to this paper’s results).

The variation taking place in the simple quality indices (q (CO2), q(mC) and C/N) suggests that certain soils had similar degrees of post-fire stress, decreasing over time [41]. Furthermore, the contribution of plant remains in the treated areas favored the mineralization process (release of nutrients). Being in its initial stages, the C/N ratio was still high in relation to the untreated area, which indicates that there was a temporary retention of mineral nutrients in the microbial biomass (immobilization) [34,42].

4.2. Evaluation of Soil Quality Index (SQI)

The SQI has been obtained from undisturbed soils with similar characteristics to the studied area, and it is made up of a set of indicators that reflect the balance between different soil properties. Based on our previously published work [27] values <0.204 indicate low quality soil, values between 0.205 and 0.274 indicate medium-quality soil, while values between 0.275 and 0.345 reflect high quality. Thus, in the present research, the low SQI values obtained after the fire (<0.20 in all the treatments) indicated that forest fire affects soil functioning in Mediterranean forests, in accordance with Pausas et al. [7]. In consequence, the SQI is an adequate index to apply in the evaluation of post-fire recovery in forest soils. The loss of vegetation modifies microclimatic conditions and evapotranspiration and changes the parameters that control runoff and infiltration [43,44]. This fact produces an important increase in soil erosion [45,46].

Soil quality index increased in treated areas (especially in the wood chip treatment) and is due to the contribution of fire remnants coming from burned trees. Those remnants, left in the soil, can mitigate extreme temperatures and erosion, and also help natural regeneration by increasing the germination of seeds [12,16] as well as nutrient availability, modifying microbial communities and their activity [17]. We confirm that woodchips also helped to reduce the crusting effect at the surface of soil, in accordance with Martínez García et al. [18]. As other studies have shown [47], the slight increase in the mean values of the SQI in areas treated with wood chippings and logs in the ground compared to the treatment with branches, may be due to the increase in sun protection and reduction in water stress. These positive effects of covering the soil with wood chips have also been demonstrated to enhance seedling emergence in other Mediterranean pine species such as Pinus halepensis [20]. In this sense, similar studies show that the increase of the SQI in the treatments’ succeeding year indicates that the vegetation remnants help to reduce soil erosion and increase humidity, favoring the mineralization process (release of nutrients) [17,48]. On the other hand, a higher slope enables soil-washing action by meteorological agents, producing a decrease in nutrients, an increase in acidity, and a reduction in salts caused by dragging ashes and organic matter. This fact led the index to decrease [49].

5. Conclusions

Our results demonstrated that post-fire management treatments such as woodchips and logs left on soil significantly improved Pinus nigra stands after a severe wildfire, even in a short period of time (less than 3 years).

In consequence, the soil quality index (SQI) reflected the changes in the functionality of Mediterranean forest soils under the action of external impacts (such as forest fires) as well as the post-fire treatments on soil. Therefore, the SQI is a valuable tool to consider in decision-making processes.

Finally, considering that forest management has the objective of increasing pine resilience after fire, we think that our study contributes to the evidence that soil treatments might be implemented in burnt Mediterranean pine forests, and soil quality indexes should be included in forest planning to assess soil recovery in the post-fire stages.