Changes in soil nitrogen dynamics caused by ...

Science of the Total Environment 639 (2018) 175–185

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Changes in soil nitrogen dynamics caused by prescribed fires in dense gorse lands in SW Pyrenees Leire Múgica ⁎, Rosa M. Canals, Leticia San Emeterio Dpto. Producción Agraria, Universidad Pública de Navarra, Campus de Arrosadia s/n, 31006, Pamplona, Spain Research Institute on Innovation & Sustainable Development in Food Chain (ISFood), Universidad Pública de Navarra, Campus de Arrosadia s/n, 31006, Pamplona, Spain

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Early spring prescribed fires in abandoned, highly-encroached gorse lands. • Immediate dissolved organic nitrogen and delayed nitrate pulses in burned soils. • Immediate decline of microbial biomass and a drop of urease activity after 7 months. • Even the low fire intensity, effects on N cycle were still detected after 18 months. • Rapid revegetation reduces N loss but post-fire gorse growth questions fire efficacy.

a r t i c l e

i n f o

Article history: Received 10 October 2017 Received in revised form 8 May 2018 Accepted 11 May 2018 Available online xxxx Keywords: Shrub encroachment Ulex gallii Controlled burnings Soil N-cycle Dissolved organic N Nitrate

a b s t r a c t Rural depopulation, abandonment of traditional land uses and decrease of extensive stockfarming is accelerating shrub encroachment in mountain areas. In NW Spain, gorse (Ulex gallii Planch.) is expanding, developing dense shrublands that accumulate high fuel-loads, ignite easily and persist during long periods as alternate stable states. Under this scenario, traditional bush-to-bush farming fires are being replaced by high fuel-load burnings performed by specialised teams to reduce fuels and promote mosaic landscapes. This research analyses the effects on soil function and nitrogen (N)-cycling of these new generation of prescribed fires practiced under similar conditions to traditional fires (winter time, moist soils), but differing in the biomass and the continuity of the surface burnt. The results showed significant changes in N-cycle parameters, such as increases in inorganic N and dissolved organic nitrogen (DON), but declines in N microbial biomass and urease activity. At the ecosystem level, potential N losses were high because the pulse of water-soluble forms, DON and nitrate, following fire overlaps with periods of low biological N retention by microorganisms and plants. Although most effects were similar to those observed in traditional burnings done in the same region, the primary concern is the high potential for DON losses following prescribed burning in highly gorse-encroached areas. In N-limited ecosystems, a crucial issue is to attain an equilibrium between frequent burnings, which may prevent an optimal recovery of the soil function, and uneven burnings, which burn high amounts of accumulated fuel and increase the risk of removing large quantities of dissolved N from the ecosystem in a unique fire event. Overall, the use of different techniques combined with fire are needed to promote and consolidate desired changes in dense gorse lands. © 2018 Elsevier B.V. All rights reserved.

⁎ Corresponding author at: Research Institute on Innovation & Sustainable Development in Food Chain (ISFood), Universidad Pública de Navarra, Campus de Arrosadia s/n, 31006, Pamplona, Spain. E-mail address: [email protected] (L. Múgica).

https://doi.org/10.1016/j.scitotenv.2018.05.139 0048-9697/© 2018 Elsevier B.V. All rights reserved.

176

L. Múgica et al. / Science of the Total Environment 639 (2018) 175–185

1. Introduction Historical fire and grazing regimes are responsible for the mosaic of open landscapes and highly valuable grasslands developed across many natural areas (Vera, 2000; Pausas and Keeley, 2009). The interaction between these two disturbances originated complex landscapes, which ancient humans understood and promoted for their own benefit (Fuhlendorf and Engle, 2004). In European mountain areas, the evolutionary history of fire, herbivory and humanization has been tracked by means of the analysis of coal and pollen fossil records in lacustrine peat sediments (Galop et al., 2013). In the Pyrenees, there is scientific evidence of the use of fire associated to hunting in the Paleolithic (Rius et al., 2012), and to agro-pastoral activities until recent days (Métailié, 2006; Faerber, 2009; Rodríguez et al., 2016), which has caused an impact in the carbon (C) and nitrogen ecosystem stocks (Lo et al., 2015). The socio-economic changes occurring on the half of the 20th century triggered a massive rural exodus, a land-use abandonment and a reduction of extensive stock farming, resulting in an acceleration of shrub encroachment and forest expansion. As a consequence, as in many other European regions, the mosaic of vegetation is being replaced by an homogeneous and more lignified landscape, which affects the diversity and functioning of open ecosystems (Poschlod and WallisDeVries, 2002; Eldridge et al., 2011) and increases the risk of natural fires (Galiana and Lázaro, 2010). In the western Pyrenees and the Cantabrian coast, gorse shrublands (Ulex europaeus L. and U. gallii Planch.) have expanded extensively paralleling land-use abandonment. Both species constitute dense covers in nutrient-poor and acid soils, which remain as a stable vegetation community for long periods and holding a scarcely diverse plant community. In former centuries, farmers harvested gorse and use it for cattle bed combined with cereal straw. Afterwards, cattle dung mixed with gorse, straw and urine was applied to soil as manure for crops (Sineiro García, 1982; Pardo de Santayana et al., 2014). However, the abandonment of herbivory and traditional farming in recent decades has favoured the expansion of both species, allowing the build-up of dense shrublands trapped in an alternate stable state with a low potential of successional dynamics (as “landscape traps” concept by Lindenmayer et al., 2011). Nowadays, prescribed fires (planned use of fire with a preservation purpose), are applied in dense shrublands for pasture reclamation (Armas-Herrera et al., 2016; Faerber, 2009; Fontúrbel et al., 2016). As in the case of traditional burnings, prescribed fires are scheduled in winter and early spring, which ensure dry vegetation, moist soils and low air temperatures. However, the continuity and quantity of fuel load and the large areas affected increase the severity of these fires compared to traditional bush-to-bush burnings (which apply fire to shrubs while maintaining herbaceous vegetation intact; San Emeterio et al., 2016) (Fig. 1). The high fire severity of prescribed burnings in gorsedominated covers may affect soil function differently from traditional fires, alter C and nutrient cycles and influence the whole ecosystem recovery. Consequently, the assessment of the environmental effects of prescribed fires in comparison to traditional burnings is a matter of interest. Even low severity fires have effects on terrestrial C and N cycles. The impact of fires of different intensity on soil organic matter stocks has been studied in depth in the last decades in different ecosystems (González-Pérez et al., 2004; Knicker, 2007; Badía et al., 2014). As a result, detailed balances estimating C stocks and losses depending on temperatures, oxygen and fuel loads have been implemented (Meigs et al., 2009; Hurteau and Brooks, 2011; Keith et al., 2014). Regarding N, the consequences of fire on the terrestrial N cycle have been less studied, even though changes in the soil biological function and in N stocks may entail critical consequences for ecosystem preservation (Covington and Sackett, 1992; Romanyà et al., 2001). To improve the knowledge on the effects of fire on soil N dynamics is crucial to understand the environmental benefits and shortcomings of burning and its

influence on the course of plant succession (Wan et al., 2001; PrietoFernández et al., 2004). Our previous research on traditional bush-tobush burnings showed a transient pulse in N availability, a moderate impact in soil microbial biomass N and a slow-down of N-related enzyme activities (San Emeterio et al., 2016). The matter of this research is to determine whether and how more intense prescribed burnings in abandoned, dense gorselands affect soil function and N cycling differently from traditional burnings. We put in question the hypothesis that current prescribed burning practices, developed as a tool for ecosystem restoration, are environmentally comparable to traditional pastoral burnings. We focus on the short- and mid-term effects of prescribed burnings on soil N dynamics and compare them to the effects reported in traditional fires still practiced in some areas of the same region. 2. Materials and methods 2.1. Study area and prescribed fires The study site was located in Roncesvalles, at the western side of the SCI Roncesvalles-Selva de Irati (protected Natura 2000 site code ES0000126), in the SW Pyrenees (43°1′N 1°19′W). The climate, cold and oceanic, is characterized by snowy winters and cool and misty summers. Mean annual temperature and precipitation are 9.2 °C and 1601 mm, respectively (Espinal climatic station, http://meteo.navarra. es) (Fig. 2). Soils, dominated by clay-loamy textures, are organic, acidic and with high cation exchange capacity (CEC) (Table 2), and are classified as Humic Dystrudept and Typic Udorthent (USDA, 2014). Landscape comprises a mosaic of beech forests, shrublands and grasslands. Gorse (Ulex gallii Planch.) is the dominant species in shrublands, and constitutes dense covers which are accompanied by heath species (such as Calluna vulgaris (L.) Hull., Erica vagans L., Erica tetralix L. and Daboecia cantabrica (Hudson) C. Koch) at best. Grasslands are very diverse and include perennial grasses (such as Festuca rubra gr., Agrostis capillaris L., Agrostis curtisii Kerguélen), perennial forbs (such as Galium saxatile L. and Potentilla erecta (L.) Raeusch.), and a small proportion of legumes (Trifolium repens L.). A sharp reduction of livestock grazing has occurred in the area during the last decades, which has favoured shrub encroachment and gorse expansion in particular. Gorse is a low palatable N2-fixing shrub that develops tall and dense covers with high fuel loads, which are very flammable and have a high calorific power (Elvira and Hernando, 1989; Marino et al., 2011). Therefore, prescribed, winter burnings are promoted and financially supported by the local government with amelioration purposes, in order to promote diversity by the entry of new species and to decrease the risk of uncontrolled fires that may affect valuable nearby forests. At the eastern side of the same SCI, in Aezkoa valley, winter burnings are also frequent in highlands. Since census of extensive livestock are still relevant in that area, gorse encroachment is less intense and a traditional use of the fire prevails, which is based on the bush-to-bush burning technique (Fig. 1; A, B and C). A recent research was done in the area to determine the effects of the bush-to-bush technique in soil N dynamics and function. This research is detailed in San Emeterio et al. (2016) and will be used in the discussion of this manuscript. In early spring 2014, three heavily gorse-dominated areas (U. gallii cover N90%) were selected along a hill in Roncesvalles. They were located at altitudes ranging from 1059 to 1125 m a.s.l. on different slopes (N35%) (Table 1). Selected areas have not been burnt for at least 15–20 years. Prescribed fires were planned and classified depending on the level of difficulty from low difficulty (Level 1) to high difficulty (Level 3, which entailed a previous field mechanical clearing of fragile areas). Three prescribed burnings were carried out on March 21st, April 10th and April 14th 2014, by specialised burning teams. Soil temperatures were recorded during the first fire using three thermistors (TidbiT© UTBI-001 data-loggers, HOBO) located in the same profile at different


177

Fig. 1. Traditional bush-to-bush burning to reduce gorse cover in grasslands (A, B and C; Aezkoa valley) and prescribed burnings in dense gorse lands for ecosystem restoration (D, E and F; Roncesvalles). Both areas belong to the SCI (Site of Community Importance) Roncesvalles-Selva de Irati (SW Pyrenees).

Prec

Temp. Max

Temperature (ºC)

20 15

S1

S3

S2

500 400

F

300

10

200

5

100

0 -5

Precipitation (mm)

600

30 25

depths (5, 7 and 10 cm) and two more thermistors in two different locations at 5 cm depth.

Temp. Min

0

2014

2015

Fig. 2. Air temperature and cumulative monthly precipitation during the study period (March 2014 – October 2015). F, prescribed fire; S1, S2 and S3, first (10 days), second (7 months) and third (18 months) soil sampling after prescribed fires, respectively (Data from Espinal climatic station, 42°58′N 1°11′W, 871 m a.s.l., 6 km from the study area; http://meteo.navarra.es/).

2.2. Soil samplings and laboratory measurement methods Three soil surveys were done 10 days, 7 months and 18 months after the fires. Each prescribed fire determined a survey area, which was constituted by a burned plot and an adjacent unburned control plot. At each date, the three burned plots and the three control plots were surveyed. Control plots in the first two sampling dates were dense gorsedominated shrublands (original community) and in the second year (18 months) were grasslands (target community). In each plot, five sampling sites were randomly selected and four soil samples per site (cores of 9 cm diameter) were collected from the topsoil at two depths (0–5 cm and 5–10 cm). As a whole, in the first two surveys, twenty samples per plot and per depth were collected, making a total of 240 soil samples per date. In the third survey burned soils were sampled at two depths (0–5 cm and 5–10 cm) and control grasslands soils at

178


Table 1 Burned plots characteristics and shrubland description before prescribed burnings. Shrubland description Area

Altitude (m a.s.l.)

Orientation

Burned area (ha)

Total cover (%)

Avg. height (cm)

Shrub species Cover (%)

A

Bizkar

1125

S

2.57

99

70

B

Muruzabal

1113

S-SE

0.52

99

70

C

Lapurtzilo

1059

S–SW

1.80

95–100

25

U. gallii D. cantabrica Pt. aquilinum E. vagans U. gallii D. cantabrica Pt. aquilinum E. vagans U. gallii D. cantabrica E. vagans C. vulgaris

94 5 1 b1 94 5 1 b1 94–100 5 1 b1

0–10 cm depth, making a total of 150 samples. Soils were then stored in polyethylene bags and kept refrigerated at +4 °C. In the laboratory, soil samples were mixed and homogenised to obtain a unique composite sample per site and per depth. Each soil composite sample was divided in two portions by the coning and quartering method, and one of the two portions was sieved at room temperature using a 2 mm mesh. Sieved and non-sieved soils were stored in polyethylene bags and kept refrigerated at +4 °C until analysed. The non-sieved portion was analysed for gravimetric soil water content (SWC grav.), pH (1:2.5 soil-to-water volumetric ratio), + nitrate (NO− 3 ) and ammonium (NH4 ) determined in 2 M KCl extracts by segmented flow colorimetry (analyser AA3, Braun + Luebbe, Norderstedt, Germany), microbial biomass C and N (fumigation-direct extraction method, Brookes et al., 1985) assuming a fumigation efficiency of 0.45 (Joergensen et al., 2011), dissolved organic nitrogen (DON, subtracting the mineral-N pool from N contents in nonfumigated extracts) and dissolved organic carbon (DOC, from nonfumigated extracts). The sieved soils were analysed for total nitrogen by the Kjeldahl method and enzymatic activities of urease, β-glucosidase and phosphatase alkaline (related to N, C and P cycles, respectively). Urease activity was determined following Kandeler and Gerber (1988) method, based on the amount of ammonia released from 1 g soil after incubation at 37 °C for 60 min with urea (820 mM) in borate buffer (100 mM, pH 10). β-glucosidase and alkaline phosphatase enzyme activities were determined using a microplate method (German et al., 2011), based on the release of ρ-nitrophenol from an extraction of soil in water (1:3 soil-to-H2O, 1 h) after incubation at 37 °C for 60 min with a specific substrate (p-nitrophenyl-b-d-glucopyranoside for βglucosidase, and ρ-nitrophenyl phosphate hexahydrate for phosphatase) in modified universal buffer (60 mM, pH 6.0). Absorbances were measured using a spectrophotometer (Multiskan Go, Thermo Scientific). In addition, a fraction of the non-sieved soils in the first survey (10 days after the prescribed fires) was sent to an official laboratory (Nasertic, Pamplona, Spain) and analysed, following standardized methods, for the main physico-chemical parameters: texture, pH, organic matter, total C and N, available phosphorus (P) and potassium (K), total cation exchangeable capacity (CEC), and exchangeable basic potassium (K), calcium (Ca), magnesium (Mg) and sodium (Na)- and acidic -aluminium (Al) and hydrogen (H)- cations.

as a random factor. For burned and control plots comparison in each sampling date, the model included fire and depth as fixed factors and sample nested within plot as a random factor. We chose the best variance structure for the residuals using the likelihood ratio test (restricted maximum likelihood estimation procedures). The significance of the fixed effects were analysed in a similar way by using maximum likelihood estimation procedures (Zuur et al., 2009). Differences between the means of treatments levels were assessed by Tukey's HSD tests with a significance level of p b 0.05 (lsmeans package; Lenth and Hervão, 2015). We applied multivariate approaches to study the relationship between treatments, depth, date and soil properties. We performed a redundancy analysis (RDA) since the gradient length of the detrended correspondence analysis (DCA) was less than two, which indicated a linear gradient (Lepš and Šmilauer, 2003). Since the soil variables were measured at different scales, we standardized the soil data with Hellinger's standardization. We used a forward selection procedure (Ordistep function, Vegan package; Oksanen et al., 2015) to determine the subset of explanatory variables explaining most variation in soil measured properties. The statistical power of the analysis was assessed by Monte Carlo permutation tests (n = 999). Further, Pearson's correlation coefficients were determined among soil variables in burned and unburned soils with a significance level of p b 0.05 (rcorr function, Hmisc package; Harrell, 2017).

2.3. Statistical analyses

We detected slight changes in post-fire soil physico-chemical properties, which were not statistically significant in most cases (Table 2). Soil properties in control and burned plots experienced a strong depth effect, and significant interactions of depth with fire event were reported for soil water content (SWC) and pH. Soil water content decreased immediately after fires, more intensely in surface horizons (0–5 cm). Regarding acidity, soils become more acidic in depth and

Statistical analyses were performed using the R software (R core Team, 2016). Linear mixed models (nlme package; Pinheiro et al., 2015) were used to assess the effects of treatments, depth and date on each soil variable. For the temporal evolution of burned plots the model included depth and sampling date as fixed factors and sample

3. Results 3.1. Soil temperature during prescribed burnings Since soils were cold and moist, temperatures reached during the burnings experienced low increases (see supplementary materials, Fig. 1). At 5 cm deep, a thermic mean rise of 34 °C was recorded (with respect to pre-fire soil temperature, 10 °C), although a high variability among sensors was detected (during the burning temperatures attained 36 °C, 65 °C and 30 °C respectively). Maximum temperatures reached in the soil during the first fire and the short duration of the pulses indicated a low to moderate fire severity. In the case of the thermistor attaining the highest temperature (65 °C), temperatures above 40 °C persisted for 39 min. 3.2. Immediate effects of fire on physical and chemical soil properties


179

Table 2 Soil physical and chemical parameters at the first survey (10 days after the prescribed fires) in both burned and control plots at two depths (0–5 and 5–10 cm). Values are the mean ± stanndard error for 15 soil samples. Likelihood ratio and p-values (tendency: +, p b 0.1; significance: *, p b 0.05; **, p b 0.01; ***, p b 0.001) are shown for comparison between treatments (fire, control), depths (0–5 and 5–10 cm) and their interaction. Different letters indicate significant differences between depths for each treatment (low-case letters) and between treatments for each depth (capital letters). Control

Burned

L. ratio

0-5 cm

5-10 cm

0-5 cm

5-10 cm

Treat.

Depth

T*D

Physical parameters Sand (0.05-2 mm) (%) Silt (0.002-0.05 mm) (%) Clay (b0.002 mm) (%) SWC (%)

29.28 (2.13) 48.43 (1.13) 22.29 (1.52) 49.47 (0.01) Aa

29.59 (2.07) 49.51 (1.27) 20.90 (1.74) 42.67 (0.01) b

26.81 (2.08) 47.13 (1.36) b 26.05 (2.50) a 43.93 (0.01) Ba

24.90 (2.33) 51.54 (2.13) a 23.55 (1.80) b 40.60 (0.01) b

0.59 0.89 0.61 3.95*

0.72 5.43* 4.88* 33.77***

1.46 2.28 0.42 10.71**

Chemical parameters pH in H2O (1:2.5) Total CEC (cmol(+)·Kg-1) Organic Matter (%) C/N ratio Total N (%) Available P (P2O5) (mg·kg-1) Available K (K2O) (mg·kg-1) Exchangeable K (cmol(+)·Kg-1) Exchangeable Ca (cmol(+)·Kg-1) Exchangeable Mg (cmol(+)·Kg-1) Exchangeable Na (cmol(+)·Kg-1) Exchangeable Al (cmol(+)·Kg-1) Exchangeable H (cmol(+)·Kg-1)

4.75 (0.04) 34.18 (2.87) 20.07 (1.92) 11.22 (0.32) b 1.04 (0.08) a 51.84 (5.00) 277.15 (34.67) a 0.59 (0.07) a 5.08 (0.52) a 2.06 (0.19) a 0.85 (0.05) a 4.14 (0.26) b 0.58 (0.11)

4.67 (0.05) 30.21 (2.19) 16.82 (1.66) 12.47 (0.31) a 0.78 (0.04) b 43.84 (4.22) 173.14 (23.45) b 0.37 (0.05) b 2.88 (0.26) b 1.22 (0.09) b 0.66 (0.02) b 5.30 (0.22) a 0.56 (0.09)

4.85 (0.06) a 32.46 (3.74) 17.54 (1.36) a 11.17 (0.55) 0.82 (0.06) a 47.90 (3.89) a 245.58 (22.19) a 0.52 (0.05) a 4.81 (0.59) a 2.05 (0.20) a 0.94 (0.12) a 4.59 (0.33) b 0.43 (0.09)

4.58 (0.06) b 28.10 (2.62) 13.67 (0.73) b 11.94 (0.30) 0.68 (0.03) b 38.96 (4.68) b 151.40 (16.39) b 0.32 (0.03) b 2.36 (0.12) b 1.08 (0.06) b 0.68 (0.02) b 6.02 (0.47) a 0.52 (0.10)

0.01 0.28 1.24 0.58 2.37 0.52 0.29 0.59 0.51 0.14 0.53 1.49 0.37

21.57*** 1.47 9.15** 8.72** 26.15*** 4.69* 38.94*** 38.91*** 25.03*** 28.16*** 13.71*** 16.11*** 0.24

13.87*** b0.01 0.08 0.20 3.50 + 0.02 0.14 0.14 0.11 0.28 0.25 0.24 0.51

tended to respond differently to the burning, increasing pH at 0–5 cm but decreasing at 5–10 cm. In these soils, calcium and aluminium were the most abundant exchangeable cations and, unlike the rest of them, aluminium concentrations increased in depth. Fire did not affect significantly exchangeable cations and P and K concentrations, neither textural parameters nor cation exchange capacity (CEC). Total N showed a tendency (non-significant) to decrease after fires in the surface horizon (0–5 cm).

3.3. Fire effects on soil nitrogen availability After fires, an immediate pulse of DON and a delayed increase of inorganic-N forms were observed (Fig. 3). DON values increased 2.5fold and 1.5-fold at 0–5 and 5–10 cm depth respectively, compared with unburned shrublands on the first date (Fig. 3). Seven months after the fire, NH+ 4 tended to rise in burned plots compared to unburned shrublands, and a high pulse of NO− 3 that comprise an 8-fold and 5.5fold increase of NO− 3 pools at 0–5 and 5–10 cm depth occurred. Although NO− 3 concentrations decreased in burned soils after one year, higher values compared to grasslands soils were still detected in the second autumn (3.4-fold increase at 0–10 cm depth). The statistical selection procedure of explanatory variables (treatment, date and depth) concluded that treatment, date and treatment: date explained most of the variation in soil measured properties, while depth was not a significant explanatory variable. Redundancy analysis (RDA) for this subset of explanatory variables was done for the measured properties at two depths. The RDA analysis supported the patterns detected by the MLM analysis, and stressed the relevance of changes in N availability associated to fire, discriminating soil samples according to the treatment and the soil sampling period (Fig. 4). The first axis of the RDA explained 33.70% of the variance, and mostly separated spring and autumn samplings according to the functional traits of soils, which some traits associated to the burning event (immediate spring pulse of DON, mid-term autumn pulse of NO− 3 ), but not all (phosphatase activity higher in spring than in autumn in all types of soils). The second axis explained 14.79% of the variance and mainly separated recently burned areas (first and second − dates), with higher contents of DON, NH+ 4 and NO 3 , from controls and 18 months-burned areas with higher values of microbial biomass N (MBN) and urease activity.

3.4. Fire effects on soil microbial biomass and enzymes activity Despite the moderate temperatures reached during the burning at 5–10 cm depth (see supplementary materials, Fig. 1), soil microbial biomasses at the first 10 cm were significantly affected (Table 3). Post-fire C in the microbial biomass (MBC) was reduced in the upper 5 cm of the soil, and the effects were still detectable 7 months later. Nitrogen in the microbial biomass (MBN) displayed a similar pattern and the effects were significant at the two sampling depths. However, after 18 months, MBN and MBC contents improved and attained values similar to those in grassland soils (Fig. 3 and Table 3). Regarding enzymes, their activities varied significantly with time and depth (Table 4), and burning influenced them differently. Fire did not affect phosphatase activity but urease activity was significantly reduced in the mid-term compared to control soils (Table 3). Such decline was still detectable 18 months later (Fig. 3). β-glucosidase activity tended to decrease in burned soils 7 months after the fire but differences with unburned soils were not significant (Table 3). The pattern of correlations differed for many key variables between burned and non-burned soils (Fig. 5). In burned soils, MBC and enzymatic activities correlated positively with pH (r ranged from 0.30 to 0.47; p b 0.05), while in control soils MBC, MBN, phosphatase and βglucosidase activities had a stronger positive correlation with SWC (r ranged from 0.37 to 0.75; p b 0.05). Regarding nitrogen, tight interactions occurring between total N and DON, MBN, MBC and NH+ 4 in control soils (r ranged from 0.54 to 0.64; p b 0.05), lessened in burned soils. Besides, other remarkable results were the negative correlations of NO− 3 with microbial biomasses and enzymatic activities (r ranged from −0.44 to −0.52; p b 0.05) and the absence of significant interactions between NH+ 4 and these variables in burned soils, opposite to the reported in the controls (r ranged from 0.37 to 0.47; p b 0.05). In addition, DON in burned soils correlated negatively with NO− 3 contents (r = −0.40; p b 0.05), contrary to control soils where DON correlated positively with inorganic-N (r ranged from 0.36 to 0.57; p b 0.05). 4. Discussion 4.1. Effects of prescribed fires on general soil properties Fire did not have significant effects on most of the physical and chemical soil properties measured ten days after the burnings, which

180


− Fig. 3. Soil mineral N contents (NH+ 4 , NO3 ), dissolved organic N (DON), urease activity (Ur), and N and C in microbial biomass (MBN and MBC) at two depths during the study period in burned (lines) and control (points) plots. Depths differentiated by colours: red (0–5 cm) and green (5–10 cm) in burned and unburned shrubland; yellow, 0–10 cm in grassland soils (18 months). Tendency (+, p b 0.10) and significance (*, p b 0.05; **, p b 0.01; ***, p b 0.001) levels of fire effects for each date. Different letters indicate significant differences between treatments for each sampling date (Spring and Fall 2014) and depth, 0–5 cm (capital letters) and 5–10 cm (low-case letters). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

is consistent with the low fire intensity and the high soil moisture (and consequently the high heat capacity of these soils, as reported by Badía et al., 2017). Fire produced an immediate decrease of SWC at 0–5 cm depth which recovered in the next fall, contrary to the results in other

temperate shrublands which effects lasted for longer (Fontúrbel et al., 2016). Regarding pH, it tended to increase in the top soil, as it has been reported in other studies in temperate shrubland and forest soils (Fontúrbel et al., 2012; Alcañiz et al., 2016). The enhanced pH has

Fig. 4. Redundancy analysis (RDA) of the soil measured variables and the subset of explanatory variables with significant effect on explaining the variation of the soil measured variables (Fire, Date, Fire:Date). A, RDA biplot showing the correlation among soil variables (Nit, Nitrate; Am, Ammonium; DON, Dissolved organic nitrogen; MBN, Microbial biomass nitrogen; Ph, Phosphatase; Gl, Glucosidase; Ur, Urease) indicated by arrows and the centroids of explanatory variables (F, burned; C, control; Sp.14, spring 2014; Fall.14, fall 2014; Fall.15, fall 2015). B, ordination diagram showing burned plots (triangles) and control plots (circles). Sampling date is indicated by colours (blue, spring 2014; orange, fall 2014; green, fall 2015). Confidence ellipses (95%) for burned and control plots in each sampling date are drawn. Solid lines, burned plots; dashed lines, control plots. F Sp.14, burned plots spring 2014; F Fall.14, burned plots fall 2014. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)


181

Table 3 Likelihood ratio tests of fixed effects and their interactions from linear mixed models fitted to evaluate the effects of fire (F) and depth (D) on soil characteristics. SWC, Soil water content; DON, dissolved organic N; MBC, microbial biomass C; MBN, microbial biomass N. Tendency (+, p b 0.10) and significance (*, p b 0.05; **, p b 0.01; ***, p b 0.001) levels. L. ratio and significance level for treatments comparison Spring 2014

Fall 2014

Fall 2015

Soil parameters

Fire

Depth

F*D

Fire

Depth

F*D

Fire

SWC pH Ammonium Nitrate Total N DON MBC MBN Phosphatase Β-Glucosidase Urease

2.95 * 0.01 0.23 0.16 2.37 13.34 *** 6.17 * 6.97 ** 0.31 0.51 0.08

33.77 *** 21.57 *** 30.93 *** 8.60 ** 26.15 *** 14.06 *** 19.79 *** 4.83 * 22.70 *** 8.01 ** 15.09 ***

10.71 ** 13.87 *** 0.87 0.91 3.50 + 5.91 * 3.92 * 16.4 *** 1.90 1.14 0.07

0.01 1.87 3.38 + 7.20 ** 2.77 + 1.62 4.58 * 10.90 ** 2.33 3.10 + 8.01 **

15.07 *** 13.79 *** 2.07 1.64 22.90 *** 28.29 *** 11.53 ** 16.38 *** 7.16 ** 9.96 ** 2.38

0.02 2.10 0.39 5.79 * 3.28 + 4.44 * 0.63 1.63 0.13 0.13 1.00

1.83 1.37 1.87 5.31 * 0.002 0.01 0.58 1.30 0.11 2.62 4.13 *

been related to ash deposition and to organic acid denaturation (Marcos et al., 2009; Gil et al., 2010; Pereira et al., 2014). The absence of noticeable changes in soil OM, C/N ratios, total CEC or exchangeable cations after the burning was likely reflecting the low fire severity and the buffering effect of the SWC, as observed in similar situations for different soil types (Certini, 2005; Vega et al., 2013). 4.2. Effects of prescribed fires on the soil N cycle and the biological function Several authors have reported changes in soil nutrient availability after fire events, particularly N (Certini, 2005; García-Marco and González-Prieto, 2008; Neary et al., 1999; Pereira et al., 2011; PrietoFernández et al., 2004; Rodríguez et al., 2009), and the data presented here also confirm these findings. Our previous research in lightly gorse-encroached areas (at the nearby Aezkoa valley) detected an im− mediate increase of NH+ 4 following fire and a delayed pulse of NO3 and DON (San Emeterio et al., 2016). In this current research in densely gorse-encroached surfaces (Roncesvalles), DON reached high values immediately after fires (2.5-fold at 0–5 cm depth) and pulses of Ninorganic forms were detected seven months later, during next autumn + (8-fold for NO− 3 and circa 2-fold for NH4 at 0–5 cm depth). Comparing both studies in gorse communities, although similar effects were observed (i.e. significant increase of the most available N forms in the soil, high mid-term pulse of autumnal NO− 3 , one-year lasting effects), some distinct patterns emerged between light- and high-encroached areas. For instance, the total pool of DON in burned soils was much higher in high- than in light-encroached areas (60 vs. 20 mg N Kg−1 soil in the top 10 cm) and occurred immediately after burning (Figs. 6 and 7). Compared with unburned soils, DON contents in soils increased Table 4 Likelihood ratio tests of fixed effects and their interactions from linear mixed models fitted to evaluate the temporal evolution of depth (0–5 and 5–10 cm) and sampling date (10 days, 7 months and 18 months) and their interaction (D*D) on soil characteristics in burned plots. SWC, Soil water content; MBC, microbial biomass C; MBN, microbial biomass N. Tendency (+, p b 0.10) and significance (*, p b 0.05; **, p b 0.01; ***, p b 0.001) levels. L. ratio and significance level for burned soils Soil parameters

Depth

Date

D*D

SWC pH Ammonium Nitrate Total N DON MBC MBN Phosphatase Glucosidase Urease

44.88 *** 48.10 *** 1702.50 *** 7.38 ** 73.29 *** 19.68 *** 46.76 *** 31.26 *** 21.63 *** 15.90 *** 14.20 ***

1.99 4.90 + 7.84 * 77.47 *** 46.77 *** 52.65 *** 35.77 *** 43.21 *** 54.01 *** 47.80 *** 13.47 **

5.71 + 3.84 7.85 * 1.04 8.30 * 30.08 *** 4.04 12.53 ** 2.31 3.35 2.88

2.15-fold in high-encroached burned areas and 1.97-fold in lightencroached burned sites. In our two experiences, current study and San Emeterio et al. (2016), the pulse of DON paralleled the reduction of MBN in the soil, suggesting a relationship between both parameters in burned soils. Soil microbiota death by thermal shock is expected, which may cause a liberation of organic compounds and a release of nitrogen. However, the high biomass of gorse accumulated in Roncesvalles (a highly inflammable N2-fixing shrub; Reyes and Casal, 2008; Merino et al., 2015), may also be responsible for the magnitude of the DON pulse after the fire. Unfortunately, we are not able to discern from these studies the relevance of each potential source of soil DON (microbial vs vegetation supply). According to Wan et al. (2001), vegetation and fuel type account for most of the terrestrial N losses produced by fires, while direct fire affections on the total soil N pools are relatively small. Contrary to DON, pools of inorganic-N forms, NO− 3 in particular, remained similar between light- (Aezkoa) and high- (Roncesvalles) encroached areas during the same periods of sampling (spring and fall, Figs. 1 and 6). After a fire, soil N-inorganic pools may increase indirectly, as the left-over of a lower microbial immobilization and a lower plant uptake (Dannenmann et al., 2011; Hanan et al., 2016), and directly, by biological and non-biological processes that transform organic N forms into inorganic N (Prieto-Fernández et al., 2004). While the pulse of NH+ 4 is a direct outcome of biomass burn and enzyme and microbe N-mineralization on simple organic materials, the NO− 3 pulse is mostly due to the biological nitrification induced by the pulse of NH+ 4 (Andersson et al., 2004; Covington and Sackett, 1992; Knicker, 2007). The results presented here suggest the occurrence of biological constraints that regulate the rate of N-inorganic transformations irrespective of the level of the DON pulse. If such hypothesis is certain, the amount and nature of the soil DON pool after the fire would emerge as a key variable to survey due to its high potential of loss by lixiviation. According to Leimer et al. (2016), DON leaching is presumably the most important form of N lixiviated. Jones et al. (2004) identified two distinct DON components in soils, low molecular weight compounds (LMW), such as free amino acids and proteins, which are easily usable by plants and microorganisms to satisfy their N and C demands (Farrell et al., 2014), and high molecular weight (HMW) recalcitrant substances. Because LMW are removed fast, HMW are likely the main component of the DON and may account for the major ecosystem DON losses. Consequently, the post-fire evolution of HMW is crucial, as it is the rate of recovery of the biological retainers, plants and microorganisms, which can take profit of the pulse of more available N forms (LMW and inorganic N). Both factors will affect the amount of ecosystem N losses in the short and the mid-term. From a practical point of view, this is particularly important if fire regimes intensify, i.e., if managers implement controlled burnings more frequently as the only tool to control shrub encroachment. If so, fire-induced N losses from the ecosystem may be

182


Fig. 5. Pearson correlation coefficients between measured variables in burned (left graph) and control (right graph) soils at 0–10 cm depth including the three sample dates (10 days, 7 months and 18 months after the prescribed fires). SWC, soil water content; Nitr, nitrate; Am, ammonium; Ur, urease; Ph, phosphatase; Gl, glucosidase, MBC, microbial biomass carbon; MBN, microbial biomass nitrogen; DON, dissolved organic nitrogen; Ntot; total N. Tendency (+, p b 0.10) and significance (*, p b 0.05; **, p b 0.01; ***, p b 0.001) levels of the correlation between each pair of measured variables.

− Fig. 6. Soil mineral N contents (NH+ 4 , NO3 ), total N, urease activity (Ur), dissolved organic N (DON), dissolved organic C (DOC), and N and C in microbial biomass (MBN and MBC) in densely gorse-encroached areas (High.Encr, Roncesvalles) and lightly gorse-encroached areas (Light.Encr, Aezkoa) in burned (F) and control (C) plots. Lines correspond to lightly gorse-encroached areas and points to densely gorse-encroached areas; red colour, F; green colour, C. Sp-1, Su-1 and Fall-1 (spring, summer and fall in the burning year); Sp-2, Su-2 and Fall-2 (spring, summer and fall in the next year after the burning). Lightly encroached areas' data from San Emeterio et al. (2016). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)


183

high enough to start biological tissue disintegration and induce mortality in the microbial community inhabiting the first centimetres of topsoil (Hernández et al., 1997; Neary et al., 1999). Besides, biome type is important as well, as soil microbiota living in montane cold environments may exhibit a low adaptation to fire (Bárcenas-Moreno and Bååth, 2009; Dooley and Treseder, 2012). MB depression may be temporary and recover in one year (Fontúrbel et al., 2016) or may persist for longer periods (up to two years, San Emeterio et al., 2016). In this research, we reported a temporal MB depression. In low-intensity burnings, the rate of MB recovery after the fire is likely driven by environmental post-fire conditions (fluctuations in soil moisture and temperature, ash deposition on soil) and by the dynamics of the revegetation process, which induce variations in soil labile C necessary for MB recovery (Hart et al., 2005; Capogna et al., 2009; Fontúrbel et al., 2012). With regard to soil enzymatic activities, urease was the enzyme most sensitive to fire. Despite the lack of immediate effects, its activity was reduced 51.2% and 66.1%, seven and 18 months after fire (compared to shrubland and grassland controls respectively), and was positively correlated to MB in burned soils (r = 0.66 to MBC and r = 0.71 to MBN; p b 0.05). Despite the results may suggest a dependence of the urease enzyme activity from the soil microbial community (Lombao et al., 2015), the lack of correlation in control soils may also indicate a similar, parallel response to the same indirect biotic and abiotic parameters affecting soil MB in burned soils (Eivazi and Bayan, 1996; Knicker, 2007). Concerning phosphatase and glucosidase activities, the absence of changes in burned soils may be explained by the low fire intensity, as it has been suggested for other comparable fires (Fontúrbel et al., 2016; Saa et al., 1993). Fine-scale research on enzymatic activities in the first soil centimetres has yield contrasting results in literature (Armas-Herrera et al., 2016; Saa et al., 1993; Goberna et al., 2012). 4.3. Implications of prescribed fire use in highly gorse-encroached areas Fig. 7. Mean extractable N pools (dissolved organic N (DON), ammonium (N-NH+ 4 ) and nitrate (N-NO3−)) in densely gorse-encroached areas (Roncesvalles) and lightly gorseencroached areas (Aezkoa). Sp1, Su1 and Fall1 (spring, summer and fall in the burning year); Sp2, Su2 and Fall2 (spring, summer and fall in the next year after the burning). F +, burned soils; C, control soils. Lightly encroached areas' data from San Emeterio et al. (2016).

higher than the potential for N replenishment (Wan et al., 2001), which will lead to negative consequences for the long-term balance of the ecosystem. Our results also show a negative, significant correlation between NO− 3 content and MB, DON and enzymatic activities in burned soils, coinciding the pulse of NO− 3 with a low MB, enzymatic activity and DON (Fig. 5). Fire caused an immediate, negative affection of the soil MB, which was reduced 30 and 49% (in MBC and MBN) compared to the unburned topsoil, and affected MBN until 10 cm depth. Different from our results, an extensive meta-analysis on the effects of fire on MB did not report significant reductions after prescribed fires, opposite to wildfires (Dooley and Treseder, 2012). However, the authors stressed the variability of the data and the difficulty to draw general conclusions. Depression of MB after prescribed fires in similar temperate shrublands has also been documented by Armas-Herrera et al. (2016) and Fontúrbel et al. (2012), which attribute it to direct and indirect effects. In the particular case of tradicional fires in gorse-shrublands, San Emeterio et al. (2016) reported a time-elapsed MB depression, which they explained by low soil moisture and changing environmental conditions occurring in burned soils. The immediate response of MB reported in this research suggests that a direct depression by thermal shock may occur as well. Fire temperatures may lead to a heat-induced microbiota mortality or to a reproductive capability alteration (Cairney and Bastias, 2007; Hart et al., 2005). Despite the low-intensity burning (65 °C at 5 cm depth in the highest record), temperatures between 40°–70 °C are

Periodicity and timing are key parameters to take into account when scheduling a sustainable plan of prescribed fires. Burning plans must consider the period needed for soil function recovery. The improvement of soil parameters such as MBs and sensitive enzyme activities, such as urease, may help when making management decisions on fire recurrence (Valkó et al., 2014). Regarding timing, winter and early-spring seasons in mountain areas ensure cold temperatures and moist soils, which reduce the intensity and potential fire damage to the soil (Badía et al., 2017). However, the risk of post-fire ecosystem N-losses are high due to the pulses of DON and mobile N-inorganic forms (NO− 3 ). Our results suggest that a rapid development of biological retainers (vegetation and microbiota) that lessen the potential of ecosystem N-losses must be pursued every time after prescribed burning is applied. A rapid resprout or germination by native vegetation ensures an effective N use and a fast improvement of the microbial functions, which is positive for the ecosystem (recover of nutrient cycles and soil functionality). However, prescribed burnings in highly shrub-encroached areas intend to originate a change in vegetation composition, i.e., from dense gorse lands to grasslands. Gorse is a fire-adapted species, which germination and vegetative sprout is stimulated by flames (Reyes and Casal, 2008). Because of the positive feedback between fire and gorse, the chance for a plant community shift using fire as the sole management tool is very low. When fire is not efficient, the risk to intensify the burning regime to control shrub encroachment increases, which may entail negative consequences for the ecosystem in the long-term (Canals et al., 2014). In order to consolidate diverse open communities following prescribed fires and attain a rapid revegetation by native species of grasses and forbs, the combination of fire with other sustainable management tools, such as guided herbivorism, is necessary (Fuhlendorf et al., 2009). A major experimental effort needs to be made in this way in the following years to support the development of integrative measures for the restauration of abandoned grasslands in southern Europe.

184


5. Conclusion Early spring prescribed fires in highly gorse-encroached grasslands, performed with the aim of restoring open landscapes and reducing the risk of wildfires, alter ecosystem N dynamics causing an immediate and high pulse of DON, a delayed increase of NO− 3 , a mid-term reduction of MBC and MBN, and a long-term depression of urease activity in soils. As a consequence, there is a probability of N loss from the ecosystem when soluble N pulses (DON and N-inorganic) parallel periods of low biological N retention by plants and soil microorganisms. The rapid gorse resprout and the consequences of burnings on the N-cycle indicates that more work is required to make prescribed fires an effective tool to consolidate open landscapes, with the aim of applying the technique at low recurrence. Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2018.05.139.

Acknowledgments This experience was partially supported by the Rural Development Programme 2007-2013. L. Múgica was funded through a UPNA's Research Staff Training (1903-2014) Grant. The authors are very grateful to Real Colegiata de Santa María de Roncesvalles, EPRIF (Tragsa), Navarra's Fire Brigade in Orreaga-Roncesvalles, INTIA staff (J. L. Saez, L. Echeverría, P. J. Karrika, I. Vergara and N. Murugarren), I. Iriarte (FANE), UPNA researchers (J. A. Blanco, Y. S. Lo and I. Maquirriain) and the Navarra's Government, for their collaboration in different parts of the project and the preparation of the manuscript. We are also grateful to the anonymous reviewers and the editor for their constructive comments on the manuscript. References Alcañiz, M., Outeiro, L., Francos, M., Farguell, J., Úbeda, X., 2016. Long-term dynamics of soil chemical properties after a prescribed fire in a Mediterranean forest (Montgrí Massif, Catalonia, Spain). Sci. Total Environ. 572:1329–1335. https://doi.org/ 10.1016/j.scitotenv.2016.01.115. Andersson, M., Michelsen, A., Jensen, M., Kjøller, A., 2004. Tropical savannah woodland: effects of experimental fire on soil microorganisms and soil emissions of carbon dioxide. Soil Biol. Biochem. 36:849–858. https://doi.org/10.1016/j.soilbio.2004.01.015. Armas-Herrera, C.M., Martí, C., Badía, D., Ortiz-Perpiñá, O., Girona-García, A., Porta, J., 2016. Immediate effects of prescribed burning in the central Pyrenees on the amount and stability of topsoil organic matter. Catena 147:238–244. https://doi.org/10.1016/ j.catena.2016.07.016. Badía, D., González-Pérez, J.A., Aznar, J.M., Arjona-Gracia, B., Martí-Dalmau, C., 2014. Changes in water repellency, aggregation and organic matter of a mollic horizon burned in laboratory: soil depth affected by fire. Geoderma 213:400–407. https:// doi.org/10.1016/j.geoderma.2013.08.038. Badía, D., López-García, S., Martí, C., Ortíz-Perpiñá, O., Girona-García, A., Casanova-Gascón, J., 2017. Burn effects on soil properties associated to heat transfer under contrasting moisture content. Sci. Total Environ. 601–602, 1119–1128 this issue. 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. https://doi.org/10.1016/j.soilbio.2009.09.010. Brookes, P.C., Kragt, J.F., Powlson, D.S., Jenkinson, D.S., 1985. Chloroform fumigation and the release of soil nitrogen: the effects of fumigation time and temperature. Soil Biol. Biochem. 17:831–835. https://doi.org/10.1016/0038-0717(85)90143-9. Cairney, J.W.G., Bastias, B.A., 2007. Influences of fire on forest soil fungas communities. Can. J. For. Res. 37:207–215. https://doi.org/10.1139/X06-287. Canals, R.M., Pedro, J., Rupérez, E., San Emeterio, L., 2014. Nutrient pulses after prescribed winter fires and preferential patterns of N uptake may contribute to the expansion of Brachypodium pinnatum (L.) P. Beauv. in highland grasslands. Appl. Veg. Sci. 17: 419–428. https://doi.org/10.1111/avsc.12088. Capogna, F., Persiani, A.M., Maggi, O., Dowgiallo, G., Puppi, G., Manes, F., 2009. Effects of different fire intensities on chemical and biological soil components and related feedbacks on a Mediterranean shrub (Philllyrea angustifolia L.). Plant Ecol. 204, 155–171. Certini, G., 2005. Effects of fire on properties of forest soils: a review. Oecologia 143:1–10. https://doi.org/10.1007/s00442-004-1788-8. Covington, W.W., Sackett, S.S., 1992. Soil mineral nitrogen changes following prescribed burning in ponderosa pine. For. Ecol. Manag. 54:175–191. https://doi.org/10.1016/ 0378-1127(92)90011-W. Dannenmann, M., Willibald, G., Sippel, S., Butterbach-Bahl, K., 2011. Nitrogen dynamics at undisturbed and burned Mediterranean shrublands of Salento Peninsula, Southern Italy. Plant Soil 343:5–15. https://doi.org/10.1007/s11104-010-0541-9.

Dooley, S.R., Treseder, K.K., 2012. The effect of fire on microbial biomass: a meta-analysis of field studies. Biogeochemistry 109:49–61. https://doi.org/10.1007/s10533-0119633-8. Eivazi, F., Bayan, M.R., 1996. Effects of long-term prescribed burning on the activity of select soil enzymes in an oak-hickory forest. Can. J. For. Res. 26, 1799–1804. Eldridge, D.J., Bowker, M.A., Maestre, F.T., Roger, E., Reynolds, J.F., Whitford, W.G., 2011. Impacts of shrub encroachment on ecosystem structure and functioning: towards a global synthesis. Ecol. Lett. 14:709–722. https://doi.org/10.1111/j.14610248.2011.01630.x. Elvira, L.M., Hernando, C., 1989. Inflamabilidad y energía de las especies de sotobosque. Colección Monogr. INIA. Faerber, J., 2009. Prescribed range burning in the Pyrenees: from a traditional practice to a modern management tool. Int. For. Fire News 38, 12–22. Farrell, M., Prendergast-Miller, M., Jones, D.L., Hill, P.W., Condron, L.M., 2014. Soil microbial organic nitrogen uptake is regulated by carbon availability. Soil Biol. Biochem. 77, 261–267. Fontúrbel, M.T., Barreiro, A., Vega, J.A., Martín, A., Jiménez, E., Carballas, T., Fernández, C., Díaz-Raviña, M., 2012. Effects of an experimental fire and post-fire stabilization treatments on soil microbial communities. Geoderma 191:51–60. https://doi.org/10.1016/ j.geoderma.2012.01.037. Fontúrbel, M.T., Fernández, C., Vega, J.A., 2016. Prescribed burning versus mechanical treatments as shrubland management options in NW Spain: mid-term soil microbial response. Appl. Soil Ecol. 107:334–346. https://doi.org/10.1016/j.apsoil.2016.07.008. Fuhlendorf, S.D., Engle, D.M., 2004. Application of the fire-grazing interaction to restore a shifting mosaic on tallgrass prairie. J. Appl. Ecol. 41:604–614. https://doi.org/10.1111/ j.0021-8901.2004.00937.x. Fuhlendorf, S.D., Engle, D.M., Kerby, J., Hamilton, R., 2009. Pyric herbivory: rewilding landscapes through the recoupling of fire and grazing. Conserv. Biol. 23:588–598. https:// doi.org/10.1111/j.1523-1739.2008.01139.x. Galiana, L., Lázaro, A., 2010. Potential barriers and factors for success. In: Montiel, C., Kraus, D. (Eds.), Best Practices of Fire Use - Prescribed Burning and Suppression Fire Programmes in Selected Case-Study Regions in Europe. European Forest Institute, Porvoo, Finland, pp. 153–164. Galop, D., Rius, D., Cugny, C., Mazier, F., 2013. A history of long-term human—environment interactions in the French Pyrenees inferred from the pollen data. In: Lozny, L.R. (Ed.), Continuity and Change in Cultural Adaptation to Mountain Environments: From Prehistory to Contemporary Threats. Springer New York, New York, NY: pp. 19–30 https://doi.org/10.1007/978-1-4614-5702-2_3. García-Marco, S., González-Prieto, S., 2008. Short- and medium-term effects of fire and fire-fighting chemicals on soil micronutrient availability. Sci. Total Environ. 407: 297–303. https://doi.org/10.1016/j.scitotenv.2008.08.021. German, D.P., Weintraub, M.N., Grandy, A.S., Lauber, C.L., Rinkes, Z.L., Allison, S.D., 2011. Optimization of hydrolytic and oxidative enzyme methods for ecosystem studies. Soil Biol. Biochem. 43:1387–1397. https://doi.org/10.1016/j.soilbio.2011.03.017. Gil, J., Zavala, L.M., Bellinfante, N., Jordán, A., 2010. Acidez y capacidad de intercambio catiónico en los suelos afectados por incendios. Métodos de determinación e interpretación de resultados. In: Cerdà, A., Jordán, A. (Eds.), Actualización En Métodos Y Técnicas Para El Estudio de Los Suelos Afectados Por Incendios Forestales. Càtedra de Divulgació de la Ciència, Universitat de València, FUEGORED 2010, Valencia, España, pp. 327–348. Goberna, M., García, C., Insam, H., Hernández, M.T., Verdú, M., 2012. Burning fire-prone Mediterranean Shrublands: immediate changes in soil microbial community structure and ecosystem functions. Microb. Ecol. 64:242–255. https://doi.org/10.1007/ s00248-011-9995-4. González-Pérez, J.A., González-Vila, F.J., Almendros, G., Knicker, H., 2004. The effect of fire on soil organic matter-a review. Environ. Int. 30, 855–870. Hanan, E.J., D'Antonio, C.M., Roberts, D.A., Schimel, J.P., 2016. Factors regulating nitrogen retention during the early stages of recovery from fire in coastal chaparral ecosystems. Ecosystems 19 (5):910–926. https://doi.org/10.1007/s10021-016-9975-0. Harrell, F.E., 2017. The R Core Team, R Core Team. Hmisc: Harrell Miscellaneous. Hart, S.C., DeLuca, T.H., Newman, G.S., MacKenzie, M.D., Boyle, S.I., 2005. Post-fire vegetative dynamics as drivers of microbial community structure and function in forest soils. For. Ecol. Manag. 220:166–184. https://doi.org/10.1016/j.foreco.2005.08.012. Hernández, T., García, C., Reinhardt, I., 1997. Short-term effect of wildfire on the chemical, biochemical and microbiological properties of Mediterranean pine forest soils. Biol. Fertil. Soils 25, 109–116. Hurteau, M.D., Brooks, M.L., 2011. Short- and long-term effects of fire on carbon in US dry temperate forest systems. Bioscience 61:139–146. https://doi.org/10.1525/ bio.2011.61.2.9. Joergensen, R.G., Wu, J., Brookes, P.C., 2011. Measuring soil microbial biomass using an automated procedure. Soil Biol. Biochem. 43:873–876. https://doi.org/10.1016/j. soilbio.2010.09.024. Jones, D.L., Shannon, D., Murphy, D.V., Farrar, J., 2004. Role of dissolved organic nitrogen (DON) in soil N cycling in grassland soils. Soil Biol. Biochem. 36, 749–756. Kandeler, E., Gerber, H., 1988. Short-term assay of soil urease activity using colorimetric determination of ammonium. Biol. Fertil. Soils 6:68–72. https://doi.org/10.1007/ BF00257924. Keith, H., Lindenmayer, D.B., Mackey, B.G., Blair, D., Carter, L., McBurney, L., Okada, S., Konishi-Nagano, T., 2014. Accounting for biomass carbon stock change due to wildfire in temperate forest landscapes in Australia. PLoS One 9 (9). Knicker, H., 2007. How does fire affect the nature and stability of soil organic nitrogen and carbon? A review. Biogeochemistry 85:91–118. https://doi.org/10.1007/s10533-0079104-4. Leimer, S., Oelmann, Y., Eisenhauer, N., Milcu, A., Roscher, C., Sheu, S., Weigelt, A., Wirth, C., Wilcke, W., 2016. Mechanisms behind plant diversity effects on inorganic and organic N leaching from temperate grassland. Biogeochemistry 131, 330–353.

L. Múgica et al. / Science of the Total Environment 639 (2018) 175–185 Lenth, R.V., Hervão, M., 2015. lsmeans: Least-squares Means. R package Version 2.19. Lepš, J., Šmilauer, P., 2003. Multivariate Analysis of Ecological Data Using CANOCO. Cambridge Press. Lindenmayer, D.B., Hobbs, R.J., Likens, G.E., Krebs, C.J., Banks, S.C., 2011. Newly discovered landscape traps produce regime shifts in wet forests. Proc. Natl. Acad. Sci. U. S. A. 108: 15887–15891. https://doi.org/10.1073/pnas.1110245108. Lo, Y.H., Blanco, J.A., Canals, R.M., González de Andrés, E., San Emeterio, L., Imbert, J.B., Castillo, F.J., 2015. Land use change effects on carbon and nitrogen stocks in the Pyrenees during the last 150 years: a modelling approach. Ecol. Model. 312, 322–334. Lombao, A., Barreiro, A., Carballas, T., Fontúrbel, M.T., Martín, A., Vega, J.A., Fernández, C., Díaz-Raviña, M., 2015. Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain). Catena 135:409–418. https://doi.org/10.1016/j. catena.2014.08.007. Marcos, E., Villalón, C., Calvo, L., Luis-Calabuig, E., 2009. Short-term effects of experimental burning on soil nutrients in the Cantabrian heathlands. Ecol. Eng. 35:820–828. https://doi.org/10.1016/j.ecoleng.2008.12.011. Marino, E., Guijarro, M., Hernando, C., Madrigal, J., Díez, C., 2011. Fire hazard after prescribed burning in a gorse shrubland: implications for fuel management. J. Environ. Manag. 92:1003–1011. https://doi.org/10.1016/j.jenvman.2010.11.006. Meigs, G.W., Donato, D.C., Campbell, J.L., Martin, J.G., Law, B.E., 2009. Forest fire impacts on carbon uptake, storage, and emission: the role of burn severity in the eastern cascades, Oregon. Ecosystems 12 (8), 1246–1267. Merino, A., Chávez-Vergara, B., Salgado, J., Fonturbel, M.T., García-Oliva, F., Vega, J.A., 2015. Variability in the composition of charred litter generated by wildfire in different ecosystems. Catena 133, 52–63. Métailié, J.P., 2006. Mountain landscape, pastoral management and traditional practices in the northern Pyrenees (France). In: Agnoletti, M. (Ed.), The Conservation of Cultural Landscapes. CAB International, pp. 108–124. Neary, D.G., Klopatek, C.C., DeBano, L.F., Folliott, P.F., 1999. Fire effects on belowground sustainability: a review and synthesis. For. Ecol. Manag. 122:51–71. https://doi.org/ 10.1016/S0378-1127(99)00032-8. Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O'Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H., 2015. vegan: Community Ecology Package. Pardo de Santayana, P., Morales, R., Aceituno-Mata, L., Molina, M. (Eds.), 2014. Inventario español de los conocimientos tradicionales relativos a la biodiversidad. Ministerio de Agricultura, Alimentación y Medio Ambiente, Madrid. Pausas, J.G., Keeley, J.E., 2009. A burning story: the role of fire in the history of life. Bioscience 59, 593–601. Pereira, P., Úbeda, X., Martin, D., Mataix-Solera, J., Guerrero, C., 2011. Effects of a low severity prescribed fire on water-soluble elements in ash from a cork oak (Quercus suber) forest located in the northeast of the Iberian peninsula. Environ. Res. 111: 237–247. https://doi.org/10.1016/j.envres.2010.09.002. Pereira, P., Úbeda, X., Martin, D., Mataix-Solera, J., Cerdà, A., 2014. Wildfire effects on extractable elements in ash from a Pinus pinaster forest in Portugal. Hydrol. Process. 28, 3681–3690. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., 2015. The R Core Team, R Core Team, 2015. nlme: Linear and Nonlinear Mixed Effects Models.

185

Poschlod, P., WallisDeVries, M.F., 2002. The historical and socioeconomic perspective of calcareous grasslands - lessons from the distant and recent past. Biol. Conserv. 104: 361–376. https://doi.org/10.1016/S0006-3207(01)00201-4. Prieto-Fernández, Á., Carballas, M., Carballas, T., 2004. Inorganic and organic N pools in soils burned or heated: immediate alterations and evolution after forest wildfires. Geoderma 121:291–306. https://doi.org/10.1016/j.geoderma.2003.11.016. Reyes, O., Casal, M., 2008. Regeneration models and plant regenerative types related to the intensity of fire in Atlantic shrubland and woodland species. J. Veg. Sci. 19: 575–583. https://doi.org/10.3170/2008-8-18412. Rius, D., Vannière, B., Galop, D., 2012. Holocene history of fire, vegetation and land use from the central Pyrenees (France). Quat. Res. 77:54–64. https://doi.org/10.1016/j. yqres.2011.09.009. Rodríguez, A., Durán, J., Fernández-Palacios, J.M., Gallardo, A., 2009. Short-term wildfire effects on the spatial pattern and scale of labile organic-N and inorganic-N and P pools. For. Ecol. Manag. 257:739–746. https://doi.org/10.1016/j.foreco.2008.10.006. Rodríguez, D., Gassiot, E., Mazzucco, N., Clemente, I., Obea, L., García, D., 2016. Fron natural environments to pastoral landscapes. Human occupations in the high mountain areas of central Pyrenees of Catalonia, since the Mesolithic to bronze age (ca. 9000–1000 cal BC). Muniber Antropología-Arkeologia 67, 325–337. Romanyà, J., Casals, P., Vallejo, V., 2001. Short-term effects of fire on soil nitrogen availability in Mediterranean grasslands and shrublands growing in old fields. For. Ecol. Manag. 147:39–53. https://doi.org/10.1016/S0378-1127(00)00433-3. Saa, A., Trasar-Cepeda, M.C., Gil-Sotres, F., Carballas, T., 1993. Changes in soil phosphorus and acid phosphatase activity immediately following forest fires. Soil Biol. Biochem. 25:1223–1230. https://doi.org/10.1016/0038-0717(93)90218-Z. San Emeterio, L., Múgica, L., Ugarte, M.D., Goicoa, T., Canals, R.M., 2016. Sustainability of traditional pastoral fires in highlands under global change: effects on soil function and nutrient cycling. Agric. Ecosyst. Environ. 235:155–163. https://doi.org/10.1016/ j.agee.2016.10.009. Sineiro García, F., 1982. Aspectos del uso ganadero del monte en Galicia para la producción de carne. Pastos 12, 1–39. Team, R. core, 2016. R package version. A Language and Environment for Statistical Computing. USDA, 2014. Keys to Soil Taxonomy. Twelfth Edition. . Valkó, O., Török, P., Deák, B., Tóthmérész, B., 2014. Review: prospects and limitations of prescribed burning as a management tool in European grasslands. Basic Appl. Ecol. 15, 26–33. Vega, J.A., Fontúrbel, T., Merino, A., Fernández, C., Ferreiro, A., Jiménez, E., 2013. 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. https:// doi.org/10.1007/s11104-012-1532-9. Vera, F.W.M., 2000. Grazing Ecology and Forest History. CABI. Wan, S., Hui, D., Luo, Y., 2001. Fire effects on nitrogen pools and dynamics in terrestrial ecosystems: a meta-analysis. Ecol. Appl. 11, 1349–1365. Zuur, A., Leno, E.N., Walker, N., Saveliev, A.A., Smith, G.M., 2009. Mixed Effects Models and Extensions in Ecology with R. Springer, New York.