Combined influence of fire and salvage logging on carbon and nitrogen storage in boreal forest soil profiles

Forest Ecology and Management 326 (2014) 133–141

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Combined influence of fire and salvage logging on carbon and nitrogen storage in boreal forest soil profiles Vincent Poirier a,⇑, David Paré b, Juliette Boiffin a, Alison D. Munson a a Centre d’étude de la forêt, Département des sciences du bois et de la forêt, Faculté de foresterie, géographie et géomatique, Université Laval, 2405 de la Terrasse, Québec, Québec G1V0A6, Canada b Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Center, 1055 du P.E.P.S., P.O. Box 10380 Stn. Sainte-Foy, Quebec, Quebec G1V4C7, Canada

a r t i c l e

i n f o

Article history: Received 21 February 2014 Received in revised form 16 April 2014 Accepted 17 April 2014 Available online 13 May 2014 Keywords: Soil organic carbon Total soil nitrogen Soil horizons Wildfire Salvage logging Black spruce forest

a b s t r a c t Boreal forest soils are a significant component of the global C cycle. Although wildfire and subsequent salvage logging are major disturbances in this ecosystem, their combined influence on soil organic carbon (SOC) and total soil nitrogen (N) storage is poorly understood. Our objective was to investigate the recent influence of fire and post-fire salvage logging on SOC and total soil N stocks and distribution in the profile of boreal forest soils. We measured SOC and total N concentrations, bulk density and pH of organic, surface mineral (0–15 cm depth) and subsurface mineral (15–40 cm depth) soil horizons on 14 different fires (burned 2005–2007) in Quebec, Canada. Each site comprised three treatments: a control stand (CTR), a recently burned (<7 years) stand that was not salvaged logged (B-NL) and a recently burned (<7 years) stand that was salvage logged (B-L) within 2 years after the fire. Our results showed that fire-affected stands had less SOC and total N stored in organic horizons, and that post-fire salvage logging reduced SOC concentration in the organic horizon, but promoted SOC and total N enrichment in the subsurface mineral soil. We conclude that mechanical disturbance of recently burned stands can contribute to the mixing of the forest floor and organic matter with the mineral soil, and influence the depth distribution of SOC and total N in the soil profile. When the entire soil profile was considered, SOC and total N stocks were equivalent in burned versus burned and salvage-logged sites. Further research should focus on how disturbance type and intensity influence the molecular nature of soil organic matter and the mechanisms by which SOC and total soil N are retained in the different soil horizons. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction The boreal forest ecosystem plays a critical role in the global C budget given the large amount of C stored in organic and mineral soil horizons (Apps et al., 1993; Jobbágy and Jackson, 2000; Tarnocai et al., 2009; DeLuca and Boisvenue, 2012). Wildfire is the major natural driver of ecosystem processes controlling C storage in boreal forests (Kasischke et al., 1995) while harvesting is the principal anthropogenic disturbance. However, the combined impact of fire and subsequent harvesting of burned stands (salvage logging) on SOC and total soil N in the boreal forest remains poorly understood (Smith et al., 2000; Seedre et al., 2011). Wildfire and logging as individual disturbances can contribute to decreases in SOC and total N storage through combustion of organic matter, volatilization, deterioration of soil structure and erosion, and alteration of litter inputs and decomposition rate (Smith et al., 2000; Certini, 2005; Bormann et al., 2008). However, ⇑ Corresponding author. Tel.: +1 418 265 2746; fax: +1 418 656 5262. E-mail address: [email protected] (V. Poirier). http://dx.doi.org/10.1016/j.foreco.2014.04.021 0378-1127/Ó 2014 Elsevier B.V. All rights reserved.

fire can potentially increase soil C and N by facilitating greater incorporation and stabilization of charcoal in the mineral soil (Giovannini et al., 1987; Gavin, 1993; DeLuca and Aplet, 2008). In addition, site preparation following logging can increase root penetration, reduce soil bulk density and increase SOC and soil N storage in subsurface soil layers (Nordborg et al., 2006). Logging after fire reduces the vegetation canopy and litter layers, and affects soil microclimate, decomposition, ecosystem carbon uptake capacity, soil microbial communities and nutrient cycling (Brais et al., 2000; Lindenmayer and Noss, 2006; Serrano-Ortiz et al., 2011; Jennings et al., 2012; Marañón-Jiménez and Castro, 2013). Moreover, post-fire salvage logging can reduce organic matter input to the soil by removing the standing dead stems that would eventually fall and contribute to SOC storage (Smith et al., 2000; DeLuca and Aplet, 2008; Moroni et al., 2010; Seedre et al., 2011). In a recent synthesis on the effect of disturbance on Canadian boreal soils, Maynard et al. (2014) found that individually, fire and harvesting were not causing depletion of soil N in coniferous boreal soils. Nevertheless, the cumulative impact of disturbances was identified as a main knowledge gap.

134

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141

Post-fire salvage logging is commonly practiced in the boreal forest within two years after fire, in order to maximize potential fiber harvest before serious deterioration of burnt trees occurs, and to facilitate replanting operations (Lindenmayer and Noss, 2006). However, few studies documented the impact of salvage logging on SOC and soil total N stocks. In the Sierra Nevada mixed conifer forest, contradictory results were obtained, with post-fire logging causing either gains [Johnson et al. (2005) on three sites, due to N-fixing species] or losses [Powers et al. (2013) on one fire] in SOC and total N in surface and subsurface mineral soil horizons. Virtually no data are published in the scientific literature to verify this potential combined impact in managed boreal forests. Fire frequency in this region of North America is expected to increase in response to climate change (Girardin et al., 2013) and therefore, the impact of post-fire salvage logging on SOC and total N storage and distribution in the profile of boreal soils needs further investigation. Major wildfire activity occurred in 2005–2007 in the boreal forest of Quebec (Canada) burning 1.2 M ha of forest, which represents 12% of the total area burned over the past 36 years [Ministère des Ressources Naturelles (MRN), 2010]. About 16% of the area burned in 2005–2007 was salvage logged (MRN, personal communication). This provided a unique occasion to study the impact of post-fire salvage logging on a large number of fires across an east–west transect of over 600 km in the boreal forest; many of the sites were only temporarily accessible. In this context, our objective was to investigate the recent influence (5–7 years) of fire and post-fire salvage logging on SOC and total N in the organic, surface mineral (i.e., 0–15 cm depth) and subsurface mineral (i.e., 15– 40 cm depth) horizons of boreal soils. We hypothesized that fire would reduce SOC and total N storage and concentration and that the combined effect of fire and salvage logging would contribute to enhanced SOC and total N losses. Although we expected a greater response to fire and post-fire salvage logging in organic and surface mineral horizons, we also expected to observe modifications in subsurface mineral horizons.

Fig. 1. Location of study sites. Identification numbers refer to those presented in Table 1.

within each stand before soil sampling and site characterization. Basal area was estimated using tree inventory and diameter at breast height for 10 circular (1.5 m diameter) microplots randomly distributed within the experimental unit. Slope grade and aspect were estimated visually.

2. Methods

2.2. Sampling and analysis of organic and mineral soil horizons

2.1. Description of study sites and experimental design

Soils were sampled during the summer of 2011. We sampled the entire organic horizon [i.e., L, F and H layers according to the Canadian System of Soil Classification (Soil Classification Working Group, 1998) and O layer according to the World Reference Base for Soil Resources (IUSS Working Group WRB, 2006)] and two fixed depth mineral soil horizons [i.e., surface mineral soil (0–15 cm depth) and subsurface mineral soil (15–40 cm depth)]. The surface mineral soil included the Ae horizon according to the Canadian System of Soil Classification (Soil Classification Working Group, 1998), which corresponds with the E horizon of the World Reference Base for Soil Resources (IUSS Working Group WRB, 2006). The subsurface mineral horizon included the Bf and Bhf horizons according to the Canadian System of Soil Classification (Soil Classification Working Group, 1998), corresponding with Bs and Bhs horizons of the World Reference Base for Soil Resources (IUSS Working Group WRB, 2006). Organic horizon sampling for bulk density, organic C, total N and pH analyses was done manually using a 25 cm  25 cm wood frame and a shovel. The bulk density (in g cm3) of the organic horizon was calculated as the oven-dried (60 °C, 72 h) mass of the organic material recovered within the 25 cm  25 cm wood frame divided by its volume. This procedure was done two times on arbitrarily selected samples within the experimental unit, and the average value of the two bulk density measurements was used for calculation and analysis. The thickness (in cm) of the organic horizon was calculated as the mean of 6–10 measurements taken randomly within a grid system of the experimental unit.

Fourteen (14) different fires were chosen along a longitudinal gradient (from 68.98 to 76.56°W) near latitude 50°N in Quebec, Canada (Fig. 1 and Table 1). Black spruce [Picea mariana (Mill.) BSP] dominated all stands, with a variable component of jack pine (Pinus banksiana Lamb.) as a second canopy species. Ericad shrubs and feathermosses dominated the understory, with deciduous shrubs and herbaceous plants being present in less important proportions. Soils of the study sites were classified as Humo-Ferric or Ferro-Humic Podzols according to the Canadian System of Soil Classification (Soil Classification Working Group, 1998), which corresponds with Orthic Podzol according to the World Reference Base for Soil Resources (IUSS Working Group WRB, 2006). All stands were located on mesic sites at an altitude between 268 and 426 m above sea level with mean annual temperature and precipitation of 0.8 °C and 938 mm, respectively. Each site comprised (1) an unburned stand that served as the control (CTR), (2) a burned stand that was not salvage logged (B-NL) and (3) a burned stand that was salvage logged (B-L) within 2 yr after the fire. All fires were of lightning origin (in years 2005– 2007), lasting between 23 and 43 days. Fires occurred in the spring, and burnt areas ranged from 300 to 30,300 ha (Table 1). The 14 burned and salvage logged stands were either harvested (n = 6), scarified following harvesting (n = 3) or scarified and replanted (n = 5) to black spruce or jack pine following harvesting. The experimental unit consisted of a 20 m  20 m plot that was established

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141

135

stored in the entire soil profile was calculated as the sum of the amounts stored in the organic and mineral soil horizons.

Table 1 Description of fire events and study sites. Site ID

Year

Duration (days)

Area (ha)

Longitude (°W)

Latitude (°N)

1 2 3 4 5 6 7 8 9 10 11 12 13 14

2007 2007 2007 2005 2005 2005 2006 2006 2007 2007 2005 2007 2005 2005–2007

36 35 35 25 43 24 22 39 10 32 30 37 27 na

4584 17,940 3406 18,299 30,328 17,288 894 2146 281 8743 13,124 8806 1607 451

71.95 71.94 68.98 74.84 73.93 72.10 73.71 73.90 74.57 74.40 76.27 76.56 76.53 76.54

50.01 50.18 50.87 50.57 49.37 50.31 50.36 50.42 49.98 49.07 50.51 50.44 50.52 50.50

na = data not available.

The concentrations of SOC (in g SOC kg1) and total soil N (in g total N kg1), and soil pH analyses were measured on one composite sample comprised of 6–10 subsamples randomly located within the experimental unit. The composite sample of the organic horizon was manually sieved to 6 mm, air-dried and ground to <1.7 mm using a mini-mill (Thomas WileyÒ, Philadelphia, PA) prior to analyses. Total C and total N concentrations were measured by dry combustion with a CNS-2000 analyzer (LECO Corp., St. Joseph, MI). Total soil C concentration was considered equivalent to SOC concentration. The pH of the organic horizon was measured in a 1:10 soil-to-CaCl2 0.01 M ratio. Surface and subsurface mineral soil horizons were sampled using a manual soil core (5.2 cm diameter). The bulk density of mineral soil horizons was calculated as the oven-dried (105 °C, 72 h) mass of soil (<2 mm) recovered within the core divided by its volume after correcting for the presence of coarse (>2 mm) fragments. This was done by weighing the coarse fragments separated by dry sieving and calculating their volume using a density of 2.65 g cm3. As for the organic horizon, two bulk density samples were taken within the experimental unit, and the mean value of the two density measurements was used for calculation and analysis. Particle-size distribution, the concentrations of SOC (in g SOC kg1) and total soil N (in g total N kg1), and soil pH measurements, were performed on one composite sample comprised of 6–10 subsamples, sampled in a core below the organic horizon samples (see above). The composite sample for surface and subsurface mineral soil was air-dried and sieved to 2 mm. Soil texture was determined after measuring particle-size distribution with the hydrometer method (Kroetsch and Wang, 2007). Air-dried and sieved mineral soil (<2 mm) was ground to < 100 lm in a mortar or with a ball-mill mixer MM 200 (RetschÓ, Haan, Germany) and analyzed for total C and total N concentrations by dry combustion with a CNS-2000 analyzer (LECO Corp., St. Joseph, MI). Since mineral soil samples were free of carbonates, total soil C concentration was considered equivalent to SOC concentration. The pH of mineral soil horizons was measured in 1:2 soil-to-CaCl2 0.01 M ratio. Soil organic C stocks (in Mg SOC ha1) and total N stocks (in Mg total N ha1) in horizons were calculated as follows:

SOCstock ¼ bulk density  SOCconcentration  thickness  0:1 and

Total soil Nstock ¼ bulk density  Total soil Nconcentration  thickness  0:1 where thickness is a measured variable in the organic horizon (see above), but remains constant in surface (i.e., 15 cm) and subsurface (i.e., 25 cm) mineral soil horizons. The amount of SOC and total N

2.3. Statistical analysis Each site (fire) was considered a block containing the three treatments (i.e., CTR, B-NL and B-L) with the exception of site #14 where no CTR stand was available. The statistical design thus comprised 41 experimental units. We performed analysis of variance (ANOVA) with a linear mixed model using the MIXED procedure of the SAS software Version 9.3 (SAS Institue Inc., 2002). We expected basal area, slope aspect and grade, and soil texture to influence the dependent variables analyzed. Thus, we included these four factors in our model as covariables, to take into account possible interactions and reduce variability. However, the exact nature of the relationship between these factors and the dependent variables is not known, hence covariables could be treated as either categorical or continuous. Their categorical or continuous nature was chosen using the lowest Akaike’s Information Criterion corrected for finite sample size (AICc), determined with the maximum likelihood estimation method prior to ANOVA. Basal area was estimated at the site (fire) level. It varied from 4.7 to 50.4 m2 ha1 and was classified in three categories: low (<12.5 m2 ha1), medium (12.5–25 m2 ha1) and high (>25 m2 ha1). Slope grade and aspect were both characterized at the experimental unit level. Slope grade varied from 1% to 30% and was classified in three categories: low (<3%), medium (3–6%), or high (>6%). Slope aspect was N, NE, E, SE, S, W or NW and was transformed to continuous values using sin and cos of corresponding angles, with N set at 0°. The texture of the surface and subsurface mineral soil was measured at the experimental unit level. The concentration of <50 lm particles varied from 199 to 1000 g kg1 and soil texture was classified in three categories: coarse (<333 g kg1), medium (333–500 g kg1) or fine (>500 g kg1). Data for organic, surface and subsurface mineral soil horizons and for the whole soil profile were analyzed individually. The texture of the surface mineral soil was used in conjunction with other co-variables in the analysis of the data for the organic horizon, while the mean texture of both surface and subsurface mineral soil horizons was used in conjunction with other co-variables in the analysis of the data for the whole-soil profile. The analysis was performed on the raw data since they met the ANOVA postulates. Two CONTRAST statements were used; (1) the effect of fire was evaluated by comparing CTR versus burned (B) stands (i.e., averaged across B-NL and B-L) and (2) the effect of post-fire salvage logging was evaluated by comparing B-NL versus B-L stands. Graphical representations were made with the SigmaPlot software Version 10.0 (Systat Software Inc., 2006). Results are presented and discussed considering treatment effects significant at P <0.10. Non-significant differences were explored to see how large the difference would need to be in order to detect a significant treatment effect. This was done using the G Power software Version 3.1.7 (Faul et al., 2009) with an a value of 0.10 and statistical power set at 0.9. 3. Results 3.1. Organic soil horizon The organic horizon of burned stands had a reduced thickness, greater bulk density, lower SOC concentration, lower SOC and total N stocks, and slightly higher pH compared to control stands (Tables 2 and 3, Fig. 2). Salvage logging of burned stands reduced forest floor SOC concentration by 12% when compared to burned stands that were not salvage logged (Table 3). Moreover, forest floor

136

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141

Table 2 Effect of fire and post-fire salvage logging on physical properties and pH of the soil horizons studied. Bulk density (g cm3)

Thickness (cm) a

b

Soil horizon

Mean (SE)

P>F

Organic Control (CTR) Burned (B) Not logged (B-NL) Logged (B-L)

21.8 11.6 12.2 10.9

0.0004

Surface mineral (0–15 cm) Control (CTR) Burned (B) Not logged (B-NL) Logged (B-L)

– – – –

–

Subsurface mineral (15–40 cm) Control (CTR) Burned (B) Not logged (B-NL) Logged (B-L)

– – – –

–

(2.1) (1.1) (1.5) (0.7)

0.43

–

–

pH

Mean (SE)

P>F

Mean (SE)

P>F

0.07 0.10 0.10 0.11

(0.01) (0.01) (0.01) (0.01)

0.005

2.69 2.77 2.75 2.79

(0.06) (0.06) (0.06) (0.06)

0.08

1.03 1.09 1.09 1.09

(0.05) (0.05) (0.05) (0.05)

0.31

3.34 3.36 3.39 3.32

(0.10) (0.08) (0.10) (0.04)

0.85

1.15 1.10 1.09 1.12

(0.06) (0.05) (0.05) (0.05)

0.50

3.75 3.94 4.00 3.88

(0.08) (0.10) (0.14) (0.07)

0.07

0.52

0.97

0.62

0.67

0.50

0.39

a

Least-square means and standard errors (SE) are presented. P value for the two contrasts tested, i.e. between control (CTR) and burned (B) stands, and between burned stands that were not salvage logged (B-NL) and burned stands that were salvage logged within 2 years after the fire (B-L). b

Table 3 Effect of fire and post-fire salvage logging on soil organic carbon (SOC) and total soil N concentrations and C-to-N ratios of the soil horizons studied. SOC (g kg1)

Total soil N (g kg1)

C-to-N ratio

Soil horizons

Mean (SE)a

P > Fb

Mean (SE)

P>F

Mean (SE)

P>F

Organic Control (CTR) Burned (B) Not logged (B-NL) Logged (B-L)

449.9 419.9 444.4 393.5

0.09

9.5 8.9 9.1 8.8

0.40

47.0 46.7 48.6 44.8

(2.1) (2.0) (2.0) (2.0)

0.87

Surface mineral (0–15 cm) Control (CTR) Burned (B) Not logged (B-NL) Logged (B-L)

22.9 18.0 17.6 18.4

(3.6) (3.6) (3.6) (3.6)

0.14

40.2 47.3 48.4 46.2

(9.4) (9.1) (9.4) (8.7)

0.53

Subsurface mineral (15–40 cm) Control (CTR) Burned (B) Not logged (B-NL) Logged (B-L)

29.4 22.5 10.2 34.7

(8.2) (6.2) (4.0) (8.4)

0.44

45.7 38.7 43.0 34.3

(6.2) (6.3) (6.4) (6.2)

0.36

(16.0) (15.5) (15.5) (15.5)

0.02

0.84

0.02

(0.6) (0.6) (0.6) (0.5)

0.71

0.67 0.51 0.50 0.53

(0.15) (0.15) (0.15) (0.15)

0.34

0.85 0.70 0.25 1.15

(0.31) (0.17) (0.13) (0.30)

0.65

0.85

0.02

0.12

0.87

0.36

a

Least-square means and standard errors (SE) are presented. P value for the two contrasts tested, i.e. between control (CTR) and burned (B) stands, and between burned stands that were not salvage logged (B-NL) and burned stands that were salvage logged within 2 years after the fire (B-L). b

Fig. 2. Soil organic carbon (SOC) (a) and total soil N (b) storage in the organic horizon. The left part of the graphs shows comparison between control (CTR) and burned (B) stands [i.e., averaged between burned stands that were not salvage logged (B-NL) and burned stands salvaged logged within 2 years after the fire (B-L)] with the associated P value of the contrast between treatment means. The right part of the graphs shows comparison between B-NL and B-L stands with the associated P value of the contrast between treatment means. Least-square means and standard errors are presented.

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141

SOC and total N stocks were reduced by 22% in burned stands that were salvage logged compared to unlogged ones. This difference was however only significant for SOC (Fig. 2). It would have needed to be slightly greater (i.e., 26%) to be detected for total N. 3.2. Mineral soil horizons Soil organic C and total N concentrations in surface (0–15 cm) mineral soil decreased by 21% and 24% in response to fire, respectively (Table 3). However, these differences were not large enough to be significant when compared to control stands. Differences between control and burned stands would have needed to be 2 times larger to be detected given the high level of variability (CV >60%) of SOC and total N concentrations in the surface mineral soil. Similar observations were made for SOC and total N stocks in control compared to burned stands (Fig. 3). Post-fire salvage logging did not influence the chemical and physical properties of the surface mineral soil. The single effect of fire did not influence soil physical and chemical properties of subsurface (15–40 cm) mineral soil except for soil pH, which was higher in burned than control stands (Tables 2 and 3, Fig. 3). However, post fire-salvage logging significantly increased subsurface mineral soil SOC and total N concentrations and stocks. The SOC concentration and stock of the subsurface mineral soil in burned and salvage logged stands were more than 3 times greater than those in burned stands that were not salvage logged (Table 3, Fig. 4). Subsurface mineral soil total N concentration and storage were respectively 4.6 and 2 times greater in burned and salvage logged stands than in burned stands that were not salvage logged (Table 3, Fig. 4). 3.3. Soil C and N storage in the entire soil profile Neither wildfire nor post-fire salvage logging significantly affected the total amounts of SOC and total N stored in the entire soil profile. Although SOC and total N stocks in burned and salvage logged stands were respectively 20% and 30% greater than in burned and unlogged stands, the differences were not large enough to be significant (Fig. 5a and b). Coefficients of variation were 46% and 53% for SOC and total N, respectively. Only a 30% difference in SOC storage and a 40% difference in total N storage would have resulted in significantly greater storage in burned and salvaged logged stands, compared to burned stands that were not salvage logged. The organic horizon represents a greater contribution to whole-soil SOC and total soil N storage in burned stands that were

137

not salvage logged, compared to those that were logged (Table 4). The contribution of the mineral surface soil remained similar across all treatments and averaged 20% of whole-soil SOC and total N stocks. The subsurface (15–40 cm) mineral soil contributed more importantly than other horizons to whole-soil SOC and total N storage in post-fire salvage logged stands (Table 4). Overall, the contribution of the subsurface (15–40 cm) mineral soil to SOC and total N storage in the entire soil profile was consistently greater than that of the surface (0–15 cm) mineral soil (Table 4). 4. Discussion The boreal soils studied contained 665 Mg C ha1 and 61.4 Mg N ha1 stored in the organic horizon, and 6120 Mg C ha1 and 64.1 Mg N ha1 stored in the mineral soil (down to 40 cm depth). When the entire soil profile was considered, SOC and total N stocks were 6 170 Mg C ha1 and 65.2 Mg N ha1, respectively. These values fall in the range of studies that investigated SOC and total N storage in boreal soils at equivalent soil depth, and under similar vegetation and disturbance types (Smith et al., 2000; Moroni et al., 2010; Nordborg et al., 2006; Paré et al., 2011). 4.1. Effects of fire As expected, our results demonstrate that fire reduced SOC and total N of the organic horizon, which is consistent with previously reported results (Smith et al., 2000; Treseder et al., 2004; Paré et al., 2011). As in the present work, the meta-analysis of Nave et al. (2011) showed a stronger impact of wildfire on SOC and total soil N storage than concentration in the organic horizon. This was likely caused by a decrease of organic horizon mass following fire (Yanai et al., 2003). The reduction of organic horizon SOC and total N stocks might have been caused by losses from combustion and the increased decomposition rate of organic matter in the first years following fire. Because the thickness of the organic horizon has decreased and a portion removed, daytime soil temperature is increased in the first years after the fire (Smith et al., 2000; Treseder et al., 2004), and this could accelerate decomposition and soil respiration. Indeed, Amiro et al. (2010) observed that boreal soils were CO2 sources within 10 years after fire, but became net CO2 sinks thereafter. Our results do not indicate that wildfire significantly reduced SOC and total N in the mineral surface soil, which does not support our hypothesis. We found that SOC and total N concentrations in the surface mineral soil declined by 21% and 24% in response to fire, respectively. Although this effect was not significant, there

Fig. 3. Soil organic carbon (SOC) (a) and total soil N (b) stocks of the surface mineral soil (0–15 cm). The left part of the graphs shows comparison between control (CTR) and burned (B) stands [i.e., averaged between burned stands that were not salvage logged (B-NL) and burned stands salvaged logged within 2 years after the fire (B-L)] with the associated P value of the contrast between treatment means. The right part of the graphs shows comparison between B-NL and B-L stands with the associated P value of the contrast between treatment means. Least-square means and standard errors are presented.

138

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141

Fig. 4. Soil organic carbon (SOC) (a) and total soil N (b) stocks of the subsurface mineral soil (15–40 cm). The left part of the graphs shows comparison between control (CTR) and burned (B) stands [i.e., averaged between burned stands that were not salvage logged (B-NL) and burned stands salvaged logged within 2 years after the fire (B-L)] with the associated P value of the contrast between treatment means. The right part of the graphs shows comparison between B-NL and B-L stands with the associated P value of the contrast between treatment means. Least-square means and standard errors are presented.

Fig. 5. Soil organic carbon (SOC) (a) and total soil N (b) stocks of the whole soil profile. The left part of the graphs shows comparison between control (CTR) and burned (B) stands [i.e., averaged between burned stands that were not salvage logged (B-NL) and burned stands salvaged logged within 2 years after the fire (B-L)] with the associated P value of the contrast between treatment means. The right part of the graphs shows comparison between B-NL and B-L stands with the associated P value of the contrast between treatment means. Least-square means and standard errors are presented. Table 4 Contribution of the soil horizons studied to soil organic C (SOC) and total N storage in the entire soil profile.

1 CTR = control stands; B = Burned stands, i.e. averaged between burned stands that were not salvage logged (B-NL) and burned stands salvage logged within 2 years after the fire (B-L).  Means followed by different capital letter within horizon indicate a significant difference (P < 0.10) between treatments within contrast (i.e., CTR vs B and B-NL vs B-L). Means followed by different lowercase letter within treatment indicate a significant difference (P < 0.10) between soil horizons. Least-square means and standard errors (in parenthesis) are presented.

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141

was an 86% probability that the SOC concentration in the surface mineral soil was truly lower in burned compared to control stands. Consistent effects and direct evidence of fire on SOC and total N in the mineral soil can be difficult to demonstrate (Bormann et al., 2008; Seedre et al., 2011). A recent synthesis showed no significant overall effects of fire on mineral soil N in the Canadian Boreal forest (Maynard et al., 2014). Results from Bormann et al. (2008) indicate that greater SOC and total N losses are expected at higher fire intensity. The fires investigated in the present study were generally characterized by low burn severity (Boiffin and Munson, 2013). This phenomenon, together with the high level of variability, could explain why the effect of fire on mineral soil SOC and total N was not detected. The short-term (<7 years) influence of the fire on SOC and total N largely occurred in the surface organic layer, which was expected. However, the effect of fire is not limited to the forest floor and can translate deeper in the mineral soil. Indeed, the subsurface mineral soil of burned stands had higher pH compared to CTR stands as observed in the organic horizons. This was not expected, but could be explained by the release of exchangeable cations upon biomass combustion (Certini, 2005; Gonzalez-Pérez et al., 2004; Thiffault et al., 2007) that could have been transported deeper in the soil (Johnson et al., 2005; Murphy et al., 2006; Seedre et al., 2011) and stabilized at depth. 4.2. Effects of post-fire salvage logging Our results indicate that post-fire salvage logging reduced SOC concentration and storage in the organic horizon of boreal soils in the 5–7 years following the fire. Accumulation of SOC in the organic horizon occurs rapidly in stands recently (<14 years) affected by the fire (Harden et al., 2012). The organic horizon is a major nutrient pool and the most biologically active portion of boreal soil profile (Weber, 1988; Krause, 1998). Weber (1988) showed that burning and mechanical disturbance of the forest floor lowered its total N content, which concords with the observed (but not significant) lower total N storage in burned and salvage logged stands, compared to burned and unlogged stands. Thus, our results indicate that disturbance of the forest floor by post-fire salvage logging could impact the replenishment of organic horizon SOC and total N in the short-term, which may influence forest regeneration. Indeed, Purdon et al. (2004) showed that post-fire salvage logging had a negative effect on the early stages of understory succession. In addition, harvesting of standing and dead trees removes biomass that would eventually contribute to long-term SOC storage (Smith et al., 2000; Seedre et al., 2011). Brais et al. (2000) estimated a negative balance of forest floor nutrient (Ca, Mg, K and N to a lower extent) storage following post-fire salvage logging over the course of a 110 year rotation in black spruce stands. Thus, further investigations are essential to clarify the longer term impact of post-fire salvage logging on SOC and total N cycling and storage in the organic horizon. Few studies have investigated the combined effect of fire and salvage logging in subsurface mineral soil under similar climatic and edaphic conditions. Still, our results are consistent with those of Johnson et al. (2005) who found, in 2 out of 3 sites in a Sierran forest, that total soil N concentration in subsurface mineral soil increased in post-fire logged sites, compared to an adjacent forest that was underburned and not salvage logged. Johnson et al. (2005) attributed their results to the arrival of an N-fixing shrub species following post-fire salvage logging. However, no N-fixing species was noted on any of our study sites. We thus hypothesize that our observed results could be related to greater incorporation and transport of particulate and soluble compounds and their subsequent stabilization in the subsurface mineral soil of burned and salvage logged stands. Results from Ross and Malcom (1982) and Nordborg et al. (2006) indicate that forest soils subjected to

139

a higher disturbance level had greater water infiltration and depth distribution of organic matter than less disturbed soils. Moreover, Nordborg et al. (2006) observed that the subsurface mineral soil of deeply cultivated forest soils contained greater SOC and total N stocks than the subsurface mineral soil of less disturbed soil. Disturbance of the soil can increase the mobility of dissolved organic C (Schelker et al., 2012) and promote SOC stabilization at depth by association with fine particles (Stemmer et al., 1999). Nitrogenous compounds could be preferentially transported in the soil column (Hilscher and Knicker, 2011) and stabilized at depth given their high affinity for mineral surfaces (Kleber et al., 2007). Nevertheless, the SOC and total N found in the subsurface mineral soil of burned and salvage logged stands could be retained in stable forms due associations with mineral surfaces (Giovannini et al., 1987; Kaiser and Guggenberger, 2003; Rumpel and KögelKnabner, 2011). Indeed, the subsurface mineral horizon of podzolic and acidic soils contain poorly crystalline Al and Fe oxides that can stabilize SOC forming strong organo-mineral complexes (Kleber et al., 2005; Eckmeier et al., 2010). However, this would need to be confirmed using soil fractionation analyses (von Lützow et al., 2007). The combined effect of fire and salvage logging translated more in the subsurface mineral soil than in the surface mineral soil. This result was not expected, and it emphasizes the need to consider subsurface soil horizons when accounting for soil SOC and total N stocks. The dynamics of subsurface soil C is still poorly understood (Rumpel and Kögel-Knabner, 2011; Schmidt et al., 2011) and must be considered when addressing the impact of forest management practices. 5. Conclusions Our results demonstrate that the amounts of SOC and total N stored in the organic horizon and in the surface mineral soil are reduced in the first 5–7 years following the fire. These losses likely occurred from direct combustion or post-fire decomposition of soil organic matter. Our study also shows that the cumulative influence of fire and salvage logging mainly impacted the distribution of SOC and total N in the soil profile. Mechanical disturbance caused by harvesting and site preparation operations in post-fire salvage logged stands promoted SOC and total N enrichment in the subsurface mineral soil. We attribute this observation to greater water infiltration and mixing of soil organic matter in the mineral soil. Mineral soil organic matter is presumably more stable than that contained in organic horizons as showed by studies comparing deciduous and coniferous soils both in North America (Laganière et al., 2013) and Europe (Wiesmeier et al., 2013). Our results suggest that salvage harvesting, while not creating greater soil losses of SOC and N, may confer greater stability to soil organic matter, by promoting storage deeper in the soil profile. Our results could be useful for models predicting temporal SOC and total soil N dynamics at the landscape scale, under different disturbance and management regimes. Efforts should be pursued to model the impact of disturbance type on the vertical distribution of SOC and total soil N in the soil profile and their relative stability in the soil through time. Achieving this goal will require a better understanding of the molecular nature of soil organic matter and the mechanisms by which SOC and total soil N are retained in the different soil horizons. Acknowledgements This project was funded by an NSERC (Natural Sciences and Engineering Research Council of Canada) Strategic Grant and an

140

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141

NSERC Discovery Grant, both to Alison Munson. We are grateful to Alain Brousseau from the Département des sciences du bois et de la forêt at Université Laval and to Serge Rousseau from the Laurentian Forestry Centre at the Canadian Forest Service, Natural Resources Canada, for their technical help with laboratory analysis. We also thank André Beaumont and the graduate and undergraduate students in our research group for their help with field work and thoughtful discussions. Finally, we thank Hélène Crépeau at the Département de mathématiques et de statistiques at Université Laval for her assistance with statistical analysis.

References Amiro, B., Barr, A.G., Barr, J.G., Black, T., Brachos, R., Brown, M., Chen, J., Clark, K., Davis, K., Desai, A., Dore, S., Engel, V., Fuentes, J., Goldstein, A., Goulden, M., Kolb, T., Lavigne, M., Law, B., Margolis, H., Martin, T., McCaughey, J., Misson, L., Montes-Helu, M., Noormets, A., Randerson, J., Starr, G., Xiao, J., 2010. Ecosystem carbon dioxide fluxes after disturbances in forests of North America. J. Geophys. Res. 115, article G00K02 . Apps, M.J., Kurz, W.A., Luxmoore, R.J., Nilsson, L.O., Sedjo, R.A., Schmidt, R., Simpson, L.G., Vincon, T.S., 1993. Boreal forests and tundra. Water Air Soil Pollut. 70, 39– 53. Boiffin, J., Munson, A.D., 2013. Three large fire years threaten resilience of closed crown black spruce forests in Eastern Canada. Ecosphere 4. http://dx.doi.org/ 10.1890/ES13-00038.1, article 56. Bormann, B.T., Homann, P.S., Darbyshire, R.B., Morissette, B.A., 2008. Intense forest wildfire sharply reduces mineral soil C and N: the first direct evidence. Can. J. For. Res. 38, 2771–2783. Brais, S., Paré, D, Ouimet, R., 2000. Impacts of wild fire severity and salvage harvesting on the nutrient balance of jack pine and black spruce boreal stands. Certini, G., 2005. Effects of fire on properties of forest soils: a review. Oecologia 143, 1–10. DeLuca, T.H., Aplet, G.H., 2008. Charcoal and carbon storage in forest soils of the Rocky Moutain West. Front. Ecol. Environ. 6, 18–24. DeLuca, T.H., Boisvenue, C., 2012. Boreal forest soils carbon: distribution, function and modelling. Forestry 85, 161–184. Eckmeier, E., Egli, M., Schmidt, M.W.I., Schlumpf, N., Nötzli, M., Minikus-Stary, N., Hagedorn, F., 2010. Preservation of fire-derived carbon compounds and sorptive stabilisation promote the accumulation of organic matter in black soils of the Southern Alps. Geoderma 159, 147–155. Faul, F., Erdfelder, E., Buchner, A., Lang, A.-G., 2009. Statistical power analyses using G Power 3.1: tests for correlation and regression analyses. Behav. Res. Meth. 41, 1149–1160. Gavin, D.G., 1993. Forest soil disturbance intervals inferred from soil charcoal radiocarbon dates. Can. J. For. Res 33, 2514–2518. Giovannini, G., Lucchesi, S., Giachetti, M., 1987. The natural evolution of a burned soil: a three year investigation. Soil Sci. 143, 220–226. Girardin, M.P., Ali, A.A., Carcaillet, C., Gauthier, S., Hély, C., Le Goff, H., Terrier, A., Bergeron, Y., 2013. Fire in managed forest of eastern Canada: risks and options. For. Ecol. Manage. 294, 238–249. Gonzalez-Pérez, J.A., Gonzalez-Pérez-Vila, F.J., Almendros, G., Knicker, H., 2004. The effect of fire on soil organic matter - a review. Environ. Int. 30, 855–870. Harden, J.W., Manies, K.L., O’Donnell, J., Johnson, K., Frolking, S., Fan, Z., 2012. Spatiotemporal analysis of black spruce forest soils and implications for the fate of C. J Geophys. Res. 117 (G01012). http://dx.doi.org/10.1029/ 2011JG001826. Hilscher, A., Knicker, H., 2011. Degradation of grass-derived pyrogenic organic material, transport of the residues within a soil column and distribution in soil organic matter fractions during a 28 months microcosm experiment. Org. Geochem. 42, 42–54. IUSS Working Group WRB, 2006. World Reference Base for Soil Resources, 2nd edition. World Soil Resources Report No. 103. FAO, Rome, 128 pp. Jennings, T.N., Smith, J.E., Cromack Jr., K., Sulzman, E.W., McKay, D., Cladwell, B.A., Beldin, S.I., 2012. Impact of postfire logging on soil bacterial and fungal communities and soil biogeochemistry in a mixed-conifer forest in central Oregon. Plant Soil 350, 393–411. Jobbágy, E.G., Jackson, R.B., 2000. The vertical distribution of soil organic carbon and its relation to climate and vegetation. Ecol. Appl. 10, 423–436. Johnson, D.W., Murphy, J.F., Susfalk, R.B., Caldwell, T.G., Miller, W.W., Walker, R.F., Powers, R.F., 2005. The effects of wildfire, salvage logging, and post-fire Nfixation on the nutrient budgets of a Sierran forest. For. Ecol. Manage. 220, 155– 165. Kaiser, K., Guggenberger, G., 2003. Mineral surfaces and soil organic matter. Euro. J. Soil Sci. 54, 219–236. Kasischke, E.S., Christensen Jr., N.L., Stocks, B.J., 1995. Fire, global warming, and the carbon balance of boreal forests. Ecol. Appl. 5, 437–451. Kleber, M., Mikutta, R., Torn, M.S., Jahn, R., 2005. Poorly crystalline mineral phases protect organic matter in acid subsoil horizons. Euro. J. Soil Sci. 56, 717–725.

Kleber, M., Sollins, P., Sutton, R., 2007. A conceptual model of organo-mineral interactions in soils: self-assembly of organic molecular fragments into zonal structures on mineral surfaces. Biogeochemistry 85, 9–24. Krause, H., 1998. Forest floor mass and nutrients in two chronosequences of plantations: jack pine vs black spruce. Can. J. Soil Sci. 78, 77–83. Kroetsch, D., Wang, C., 2007. Particle size distribution. In: Carter, M.R., Gregorich, E.G. (Eds.), Soil Sampling and Methods of Analysis, 2nd edn. CRC Press, Boca Raton, pp. 713–725. Laganière, J., Paré, D., Bergeron, Y., Chen, H.Y.H., Brassard, B.W., Cavard, X., 2013. Stability of soil carbon stocks varies with forest composition in the Canadian boreal biome. Ecosystems 16, 852–865. Lindenmayer, D.B., Noss, R.F., 2006. Salvage logging, ecosystem processes, and biodiversity conservation. Conserv. Biol. 20, 949–958. Marañón-Jiménez, S., Castro, J., 2013. Effect of decomposing post-fire coarse woody debris on soil fertility and nutrient availability in a Mediterranean ecosystem. Biogeochemistry 112, 519–535. Maynard, D.G., Paré, D., Thiffault, E., Lafleur, B., Hogg, K.E., Kishchuk, B., 2014. How do natural disturbances and human activities affect soils and tree nutrition and growth in the Canadian boreal forest? Environ. Rev. 22. http://dx.doi.org/ 10.1139/er-2013-0057. Ministère des Ressources Naturelles (MRN), 2010. Direction de la protection des forêts, Québec, 2010. Location, origin and area of fires recorded between 1972 and 2012 in the intensive protection zone of Quebec, Canada. . Moroni, M.T., Shaw, C.H., Otahal, P., 2010. Forest carbon stocks in Newfoundland boreal forests of harvest and natural disturbance origin I: field study. Can. J. For. Res. 40, 2135–2145. Murphy, J.G., Johnson, D.W., Miller, W.W., Walker, R.F., Carroll, E.F., Blank, R.R., 2006. Wildfire effects on soil nutrients and leaching in a Tahoe basin watershed. J. Environ. Qual. 35, 479–489. Nave, L.E., Vance, E.D., Swanston, C.W., Curtis, P.S., 2011. Fire effects on temperate forest soil C and N storage. Ecol. Appl. 21, 1189–1201. Nordborg, F., Nilsson, U., Gemmel, P., Örlander, G., 2006. Carbon and nitrogen in soil, trees and field vegetation in conifer plantations 10 years after deep soil cultivation and patch scarification. Scand. J. For. Res. 21, 356–363. Paré, D., Banville, J.L., Garneau, M., Bergeron, Y., 2011. Soil carbon stocks and soil carbon quality in the upland portion of a boreal landscape, James Bay, Quebec. Ecosystems 14, 533–546. Powers, E.M., Marshall, J.D., Zhang, J., Wei, L., 2013. Post-fire management regimes affect carbon sequestration and storage in a Sierra Nevada mixed conifer forest. For. Ecol. Manage. 291, 268–277. Purdon, M., Brais, S., Bergeron, Y., 2004. Initial response of understorey vegetation to fire severity and salvage-logging in the southern boreal forest of Québec. Appl. Veg. Sci. 7, 49–60. Ross, S.M., Malcom, D.C., 1982. Effects of intensive forestry ploughing practices on an upland healt soil in South East Scotland. Forestry 55, 155–171. Rumpel, C., Kögel-Knabner, I., 2011. Deep soil organic matter – a key but poorly understood component of terrestrial C cycle. Plant Soil 338, 143–158. SAS Institue Inc., 2002. SAS software for Windows v. 9.3, Cary, North Carolina. Schelker, J., Eklöf, K., Bishop, K., Laudon, H., 2012. Effects of forestry operations on dissolved organic carbon concentrations and export in boreal first-order streams. J. Geophys. Res. 117 (G1011). http://dx.doi.org/10.1029/ 2011JG001827. Schmidt, M.W.I., Torn, M., Abiven, S., Dittmar, T., Guggenberger, G., Janssens, I.A., Kleber, M., Kögel-Knabner, I., Lehmann, J., Manning, D.A.C., Nannipieri, P., Rasse, D.P., Weiner, S., Trumbore, S.E., 2011. Persistence of soil organic matter as an ecosystem property. Nature 478, 49–56. Seedre, M., Shrestha, B.M., Chen, H.Y.H., Colombo, S., Jõgiste, K., 2011. Carbon dynamics of North American boreal forest after stand replacing wildfire and clearcut logging. J. For. Res. 16, 168–183. Serrano-Ortiz, P., Marañón-Jimenez, S., Reverter, B.R., Sánchez-Cañete, E.P., Castro, J., Kowalski, A.S., 2011. Post-fire salvage logging reduces carbon sequestration in Mediterranean coniferous forest. For. Ecol. Manage. 262, 2287–2296. Smith, K.C., Coyea, M.R., Munson, A.D., 2000. Soil carbon, nitrogen, and phosphorus stocks and dynamics under disturbed black spruce forests. Ecol. Appl. 10, 775– 788. Soil Classification Working Group, 1998. The Canadian system of soil classification. Third Edition. Agriculture and Agri-Food Canada, Publication No. 1646. NRC research Press, Ottawa, pp. 187. Stemmer, M., von Lützow, M., Kandeler, E., Pichlmayer, F., Gerzabek, M.H., 1999. The effect of maize straw placement on mineralization of C and N in soil particle size fractions. Euro. J. Soil Sci. 50, 73–85. Systat Software Inc., 2006. SigmaPlot software for Windows v. 10.0, Point Richmond, California. Tarnocai, C., Canadell, J.G., Schuur, E.A.G., Kuhry, P., Mazhitova, G., Zimov, S., 2009. Soil organic carbon pools in the northern circumpolar permafrost region. Glob. Biogeochem. Cycles 23, article GB2023 http://doi:10.1029/2008GB003327. Thiffault, E., Bélanger, N., Paré, D., Munson, A.D., 2007. How do forest harvesting methods compare with wildfire? a case study of soil chemistry and tree nutrition in the boreal forest. Can. J. For. Res. 37, 1658–1668. Treseder, K.K., Mack, M.C., Cross, A., 2004. Relationships among fires, fungi, and soil dynamics in Alaskan boreal forest. Ecol. Appl. 14, 1826–1838. von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Flessa, H., Guggenberger, G., Matzner, E., Marschner, B., 2007. SOM fractionation methods: relevance to functional pools and to stabilization mechanisms. Soil Biol. Biochem. 39, 2183– 2207.

V. Poirier et al. / Forest Ecology and Management 326 (2014) 133–141 Weber, M.G., 1988. Fire and ecosystem dynamics in Eastern Canadian Pinus Banksiana forest. In: Verhoeven, J.T.A., Heil, G.W., Werger, M.J.A. (Eds.), Vegetation Structure in Relation to Carbon and Nutrient Economy. SPB Academic Publishing, The Hague, pp. 93–105. Wiesmeier, M., Prietzel, J., Barthold, F., Spörlein, P., Geuß, U., Hangen, E., Reischl, A., Schilling, B., von Lützow, M., Kögel-Knabner, I., 2013. Storage and drivers of

141

organic carbon in forest soils of southeast Germany (Bavaria) – Implications for carbon sequestration. For. Ecol. Manage. 295, 162–172. Yanai, R.D., Stehman, S.V., Arthur, M.A., Prescott, C.E., Friedland, A.J., Siccama, T.G., Binkley, D., 2003. Detecting changes in forest floor carbon. Soil Sci. Soc. Am. J. 67, 1583–1593.