Reduced global warming potential after wood ash application in drained Northern peatland forests

Forest Ecology and Management 328 (2014) 159–166

Contents lists available at ScienceDirect

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

Reduced global warming potential after wood ash application in drained Northern peatland forests Tobias Rütting a,⇑, Robert G. Björk a, Astrid Meyer a, Leif Klemedtsson a, Ulf Sikström b a b

Department of Earth Sciences, University of Gothenburg, Box 460, 405 30 Gothenburg, Sweden The Forestry Research Institute of Sweden, Uppsala Science Park, 751 83 Uppsala, Sweden

a r t i c l e

i n f o

Article history: Received 12 March 2014 Received in revised form 5 May 2014 Accepted 20 May 2014

Keywords: Forestry Greenhouse gas Mitigation option Land use change Tree growth

a b s t r a c t Past land use change has converted vast areas of Northern peatland by drainage to agricultural or forested land. This change often reduces the greenhouse gas (GHG) sink strength of peatlands or turns them even from sinks to sources, which affects the global climate. Therefore, there is a need for suitable mitigation options for GHG emissions from drained peatlands. Addition of wood ash to peatland forests has been suggested as such a measure, but the overall effect on the global warming potential (GWP) of these ecosystems is still unclear. In order to fill this knowledge gap, we investigated three drained peatland forests in Sweden that had been fertilized with wood ash and monitored stand growth as well as the GHG emissions from soil, i.e. net effluxes of carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O). Our results show that over the first five to eight years after wood ash application, tree growth was enhanced at all sites. This was accompanied by generally little changes in the GHG emissions. Overall, we found that wood ash application reduced the GWP of drained peatland forests. Even though that our study was limited to eight years after wood ash application, we can conclude that in the short term wood ash application may be a suitable mitigation option for GHG emissions from Northern drained peatland forests. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction Northern peatlands have accumulated large amounts of carbon (C) as peat from the atmosphere since the last glaciation (Yu et al., 2010). However, past and present land use have altered their greenhouse gas (GHG) balance, resulting in high emissions of carbon dioxide (CO2) and nitrous oxide (N2O) after drainage (Maljanen et al., 2010). Forestation of drained peatlands may be a mitigation option for increased GHG emissions due to accumulation of C in the aboveground and belowground tree biomass. However, reviewing existing studies, Maljanen et al. (2010) concluded that there is a lack of data and that, hence, large uncertainties exist regarding the potential of different land-use options to mitigate GHG emissions from drained peatlands. On drained peatlands with sufficient aeration in the (hemi-)boreal zone, tree growth is commonly limited by nutrient availability, especially by phosphorus (P) or potassium (K), but sometimes also by nitrogen (N) or boron (B) (Ferm et al., 1992; Moilanen et al., 2010). Wood ash contains all essential elements for plant growth, although only trace amounts of N (Demeyer et al., 2001). Provided that the drainage and supply of N from mineralization is sufficient, ⇑ Corresponding author. Tel.: +46 31 786 1874. E-mail address: [email protected] (T. Rütting). http://dx.doi.org/10.1016/j.foreco.2014.05.033 0378-1127/Ó 2014 Elsevier B.V. All rights reserved.

wood ash application has proved beneficial for tree growth on drained peatlands (Hökkä et al., 2012; Moilanen et al., 2005; Sikström et al., 2010). However, previous studies have mostly been conducted in low- to medium-productive peatland sites forested with Scots pine (Pinus sylvestris L.). Hence, there is a lack of studies on high-productive drained peatlands and of peatlands forested with other tree species. As fertile afforested peatlands have been identified as hot spots for GHG emissions (Alm et al., 2007; Ernfors et al., 2007; Klemedtsson et al., 2005), there is a particular need to evaluate the potential of wood ash application to mitigate GHG emissions from these forests. Apart from increasing tree growth, wood ash application may also affect the exchange of GHGs between the soil and atmosphere. An increased tree growth can lower the ground water table (Hökkä et al., 2008) and thereby affect the decomposition of organic matter, leading to increased CO2 emissions. Wood ash usually increases the soil pH, which can stimulate both soil microbial activity (Fritze et al., 1994; Zimmermann and Frey, 2002) and decomposition rates (Moilanen et al., 2002, 2012), but see Björk et al. (2010). However, increased pH may also affect the product ratio of denitrification, hence reducing N2O emission (Šimek and Cooper, 2002). The few studies that investigated the effect of wood ash application on soil GHG emissions are inconsistent, suggesting either unaffected (Ernfors et al., 2010; Maljanen et al., 2006),

160

T. Rütting et al. / Forest Ecology and Management 328 (2014) 159–166

increased (Moilanen et al., 2012) or reduced (Klemedtsson et al., 2010) CO2 and N2O emissions for up to five years after wood ash application. Methane (CH4) fluxes have been found to be unaffected in these studies. However, in the long-term the effects of wood ash on GHG emissions may differ from the short-term responses. Indeed, 14–50 years after wood ash application to three Finnish pine forests, the soil CO2 emissions increased, while decreased CH4 and no consistent effect on N2O emission was observed (Maljanen et al., 2006). Application of wood ash on drained peat soil usually contributes to increased C sequestration in terms of increased tree growth (as referred to above) and can additionally be a profitable silvicultural investment (Väätäinen et al., 2011). However, a pertinent question is the overall effect on the global warming potential (GWP), i.e. if the increased forest growth can offset the increased GHG emissions. Therefore, in the present study we evaluated the effects of wood ash application on tree growth and soilatmosphere GHG exchange (CO2, CH4 and N2O) in drained peatland forests, by combining new and earlier published data (Ernfors et al., 2010; Klemedtsson et al., 2010). The overall aim was to quantify how wood ash application alters the total GWP of drained peatland forests. 2. Materials and methods The present study was conducted at two experimental sites, an oligotrophic (Anderstorp) and a mesotrophic drained peatland (Skogaryd), and at one catchment study with oligotrophic peatlands (Bredaryd) (Table 1). Tree growth and GHG emissions following wood ash application for the first five years at Anderstorp and for the first two years at Skogaryd have been reported previously (Ernfors et al., 2010; Klemedtsson et al., 2010). Here, we monitored tree growth in Anderstorp, and both tree growth and GHG emissions in Skogaryd over three additional years and compiled a total GWP for these sites. In addition, data on tree growth over seven years following wood ash application are reported for Bredaryd (Table 1). 2.1. Sites, experimental designs and treatments The experiments at Anderstorp (Ernfors et al., 2010) and Skogaryd (Klemedtsson et al., 2010) were designed as randomized blocks,

with four respectively three replicates (blocks). At Bredaryd two small catchments were investigated without replication on wood ash application (Ring et al., 2011). The Anderstorp site is a former bog, which was drained in the late 1980s and has a naturally generated tree stand dominated by Scots pine. The Skogaryd site is a former fen, which was drained in the 1870s and used for agriculture until 1951, when it was planted with Norway spruce (Picea abies L. Karst.). Both experiments included three treatments; a control and application of crushed wood ash at a rate of 3.3 and 6.6 t d.w. ha1 (hereafter referred to as low and high wood ash treatment). The wood ash applications were conducted in September 2003 (Anderstorp) and in August 2006 (Skogaryd). Although the doses were the same in both experiments, the amounts of added nutrients varied due to differences in element concentrations in the wood ashes (Ernfors et al., 2010; Klemedtsson et al., 2010). Chemical and physical soil characteristics, microbial community structure and biomass in the peat, as well as some microbial processes at the two sites are reported by Björk et al. (2010). The Bredaryd site consists of two small catchments, both containing drained peatland, referred to as ‘‘Bredaryd North’’ and ‘‘Bredaryd South’’. Both peatlands were originally bogs that were drained in the late 1980s. The peatlands were about 400 m apart and forested by naturally generated tree stands dominated by Scots pine. In October 2004, Bredaryd North was treated with 3.1 t d.w. ha1 of crushed wood ash, while the southern peatland was left as control (Ernfors et al., 2010). On each of the two peatlands, nine permanent measurement plots were arranged in a grid-based pattern within 3.5 ha (Bredaryd North) and 6 ha (Bredaryd South), respectively. The plots were located halfway between two ditches. 2.2. Tree growth In Anderstorp and Skogaryd, biomass data of all trees in each plot were collected before wood ash application (September 2003 and June 2006, respectively) and at the time of revision (September 2011). Stem diameters at breast height (1.3 m above the ground) were measured with callipers in two perpendicular directions. Tree heights were measured using a hypsometer (Vertex III, Haglöf Scandinavia AB, Avesta, Sweden). In addition, increment cores were sampled in September 2011 from each tree at all three

Table 1 Basic site and soil properties for three Swedish peatland forests (control plots). Data are presented as means ± standard error. Anderstorp

Skogaryd

Bredaryd

57°150 N, 13°350 E

58°230 N, 12°090 E

57°110 N, 13°440 E

98% P. sylvestris 2% P. abies

99% P. sylvestris 1% P. abies

90% P. sylvestris 9% P. abies 1% B. pubescens 722 ± 38 757 ± 76 14.8 ± 0.2 15.1 ± 0.2 131 ± 6 132 ± 13 October 2004–September 2011

4.4 ± 0.0 n.d. 33 ± 3c

North Coordinates Stand properties Tree speciesa

a

No. stems (ha1) Tree height (m) Tree stem volume (m3 ha1) Monitoring period

742 ± 32 13.3 ± 0.3 109 ± 5 September 2003–September 2011

3% P. sylvestris 95% P. abies 3% B. pubescens 797 ± 55 22.3 ± 0.8 404 ± 20 June 2006–September 2011

Soil properties (0–0.05 m)b pH SOM (%) C/N

4.9 ± 0.3 96.2 ± 0.2 23.4 ± 1.0

4.5 ± 0.1 79.4 ± 9.6 23.1 ± 1.0

South

By stem volume. Data from Björk et al. (2010) for Anderstorp and Skogaryd and from Ernfors et al. (2010) for Bredaryd. pH measured in 1:10 water extract; SOM is soil organic matter content measured by loss on ignition; C/N is carbon to nitrogen ratio of bulk soil. c 0.05–0.15 m soil depth. b

T. Rütting et al. / Forest Ecology and Management 328 (2014) 159–166

sites. The tree ring widths of the cores were measured under a microscope (resolution 0.01 mm), and tree diameters and the annual basal area increment was reconstructed until five years before treatment. All trees with a breast height diameter >50 mm at the start of the experiment were included in the study. For each plot, mean tree height (basal area weighted), basal area, stem volume and total tree biomass (aboveground + belowground) were calculated based on the tree diameter measurements. The stem volume was calculated using empirical functions (Näslund, 1947). Aboveground tree stand biomass, excluding stumps, was calculated according to Marklund (1988), while belowground biomass, including stumps and roots P2 mm in diameter, was calculated according to Petersson and Ståhl (2006). Mean annual increments in the tree stand variables over the observed period (5–8 years) were calculated from the differences between the two assessment dates, divided by the number of years between the assessments. In the winter 2009/2010 in Skogaryd, a substantial part of the trees stems were broken due to frozen heavy snow in the tree crowns in combination with strong wind. Therefore, the height (h) of the trees was estimated by height curves derived based on undamaged trees only:

161

affected by wood ash application. For that, average GHG emissions were calculated based on the annual fluxes of the individual years after wood ash fertilization. For this N2O and CH4 fluxes were transformed to CO2-equivalents based on a GWP of 298 and 25, respectively (IPCC, 2007). The GWP (as CO2-equivalents) was then calculated according to:

GWP ¼ F CO2  fRh þ F CH4 þ F N2O  BI  LP;

ð3Þ

where a, b and c are empirical constants and d is diameter at breast height. In Bredaryd North, height curves were derived for Scots pine, and in Bredaryd South for Scots pine, Norway spruce and Downey birch (Betula pubescens Ehrh.), respectively.

where FCO2, FCH4, FN2O are mean annual net fluxes of the respective GHG, fRh is the fraction of heterotrophic to total soil respiration and BI is mean annual biomass C increment over five/eight years and LP is annual litter production. Thereby, a positive flux means emission from (source) and a negative flux uptake of (sink) the soil. For the GWP budget, only the heterotrophic respiration from decomposition of peat and soil organic matter has to be considered. For six peatland forests in Finland, Mäkiranta et al. (2008) found that heterotrophic respiration contributed on average by 42% to total respiration and Ojanen et al. (2010) found for 68 peatland forest a value of 46%. We therefore assume here that 44% of soil respiration is heterotrophic (fRh) and, moreover, that this percentage is not affected by wood ash application. Aboveground litter production (LP) was not measured in the current study. At Skogaryd, Meyer et al. (2013) reported for an adjacent spruce stand on the same former fen a LP of 2.9 t C ha1 yr1 and a BI of 8.3 t C ha1 yr1. We used this LP/BI ratio to calculate the aboveground litter production from the ash experiment plots. Estimation of litter production for the pine stand at Anderstorp is based on the investigation by Albrektson (1988), who measured needle litter fall and tree growth in 16 pine stands in Sweden. Based on the average litter production and average biomass growth, Albrektson (1988) reported that 203 kg needle litters are produced per m3 stem volume increment. We, however, recalculated this ratio for each individual stand investigated by Albrektson (1988) and found an average litter production of 247 kg m3, which was used to estimate litter production at Anderstorp. We, moreover, assume that needle litter contributed by 75% to the total litter (Albrektson, 1988).

2.3. Greenhouse gas flux measurements

2.5. Statistical analyses

In Skogaryd, net soil fluxes of CO2, N2O and CH4 were measured every second week between September 2009 and September 2011, i.e. in the fourth and fifth year after wood ash application, with manual chambers (Weslien et al., 1998). The measurements were conducted at the same locations as in Klemedtsson et al. (2010), with four collars permanently installed along a 20 m transect in the center of each plot (total of 12 chambers per treatment). Air samples from the chambers were taken 5, 15, 25 and 35 min after closure and the samples were brought to the laboratory for analysis on a Varian 3800 gas chromatograph (Klemedtsson et al., 1997). Due to snow and ice cover of the soil, flux measurements could not be conducted for a 3–4 month period in January–April in both years. For these periods, fluxes were interpolated between the last early winter and first spring measurements. Soil temperature and groundwater level were measured at the time of gas measurements. Soil temperature was measured at 0.1 and 0.2 m depth, approximately 0.3 m from each collar. The depth to groundwater was determined at wells inserted to a depth of 1.5 m at the center of each plot (Klemedtsson et al., 2010).

A statistical significance level (a) of 0.05 was used for all tests. The tree growth data in Anderstorp and Skogaryd were analyzed using a two-way analysis of variance (ANOVA) with block and treatment as factors, followed by Tukey’s HSD post hoc test for pairwise comparisons between treatments. As covariates, the initial values at the outset for number of trees per hectare, mean height of the trees (weighted by basal area), basal area, stem volume, the proportion of Scots pine and basal area increment over the five years before treatment were tested individually and together in the model. The GLM procedure of the SASÒ program version 9.2 (http://support.sas.com) was used for the calculations. For Bredaryd, the tree growth response to the wood ash application was calculated as the difference between the mean values for the nine plots in Bredaryd North and South, respectively. No statistical test was carried out since the experimental design did not include any true replicates of the ash treatment. Data on GHG fluxes in Skogaryd were analyzed using a nested three-way model. The factors included in the model were block (random), treatment (fixed) and chamber (random; nested within treatment) and the ground water level was included as a covariate (PASW Statistics 18). Tukey’s HSD post hoc test was used when the effect of treatment was significant. The gas flux data was analyzed separately for each year. The flux data were, after addition of a constant, log-transformed and concomitantly scaled to unit variance to achieve a normal distribution and to eliminate skewness and ensure homogeneity of variances according to Økland et al. (2001).

H ¼ a þ b  d;

ð1Þ

where a and b are empirical constants and d is diameter at breast height (procedure REG in SASÒ software, version 9.2). In Bredaryd, the tree growth was measured in a similar way, with the exception that tree height was measured only for a randomly selected subsample of the trees. Height curves were derived according to: 2

h¼aþbdþcd ;

ð2Þ

2.4. Compilation of total global warming potential Combining the measurements of the present study with previously published data on GHG emissions from Skogaryd and Anderstorp (Ernfors et al., 2010; Klemedtsson et al., 2010) enables us to provide a total GWP for the two sites and to reveal how this is

162

T. Rütting et al. / Forest Ecology and Management 328 (2014) 159–166

3. Results and discussion 3.1. Tree growth In the oligotrophic Scots pine stand at Anderstorp, tree growth variables were generally not affected during the first five years after wood ash application (Ernfors et al., 2010). In contrast, considering the period of eight years after application all growth variables were significantly higher in the ash plots compared to the control plots (Table 2), with the exception of height increment in the low ash dose (p = 0.08). This finding agrees well with the Scots pine stand at Bredaryd where, over the seven-year period after wood ash application, a higher height increment, basal area increment and stem volume increment was observed for the ash treatment (Table 2), which could, though, not be statistically analyzed. In Anderstorp, the mean stem volume increment was 1.4 (+38%) and 1.7 (+46%) m3 ha1 yr1 higher in the low and high ash dose, respectively, compared to the control, and in Bredaryd 2.0 (+61%) m3 ha1 yr1 higher. The similar responses between the high dose at Anderstop and the treatment at Bredaryd was likely due to the fact that comparable amounts of P and K were added (not shown), even when ash amounts differed. Moreover, the magnitudes of responses are in agreement with earlier findings on young Scots pine stands after wood ash application and PKfertilization in drained N-rich peatland forests (Hökkä et al., 2012). In addition, the biomass increment in Anderstorp was 700 (+26%) and 1160 (+43%) kg d.w. ha1 yr1 higher in low and high ash dose treatments, respectively, compared to control. The relative increase in growth variables is lower than what has been found for another oligotrophic Scots pine peatland in southern Sweden (Perstorp), in which 26 years after wood ash application (2.5 t ash ha1) basal area increment was 28-times and volume increment 38-times higher than in the control (Sikström et al., 2010). However, at Perstorp the trees on control plots showed very severe P-deficiency and tree growth was negligible during the observation period, whereas the ash treatment promoted both the number of trees established and the growth of individual tree Table 2 Annual height increment (HI), basal area increment (BAI), volume increment (VI) and biomass increment (BI) for three Swedish peatland forests with and without application of wood ash. Data at Anderstorp and Skogaryd for eight and five years, respectively, following wood ash application at two doses (3.3 and 6.6 tonnes d.w. ash ha1) and for Bredaryd seven years following wood ash application of 3.1 tonnes d.w. ash ha1. Different letters by rows indicate significant differences (p < 0.05) between the means (±standard error) of treatments. Treatment Control

Low dose

High dose

Anderstorpa HI (m yr1) BAI (m2 ha1 yr1) VI (m3 ha1 yr1) BI (kg d.w. ha1 yr1)

0.14 ± 0.01 a 0.39 ± 0.02 a 3.7 ± 0.15 a 2680 ± 90 a

0.18 ± 0.01 a 0.53 ± 0.02 b 5.1 ± 0.15 b 3380 ± 90 b

0.23 ± 0.01 b 0.60 ± 0.02 b 5.4 ± 0.15 b 3840 ± 90 b

Skogaryda HI (m yr1) BAI (m2 ha1 yr1) VI (m3 ha1 yr1) BI (kg d.w. ha1 yr1)

0.15 ± 0.01 a 0.99 ± 0.08 a 12.4 ± 0.92 a 7620 ± 510 a

0.28 ± 0.01 b 1.19 ± 0.08 a 17.2 ± 0.95 b 9060 ± 520 a

0.25 ± 0.01 b 1.28 ± 0.09 a 18.2 ± 1.07 b 9410 ± 580 a

0.08 ± 0.01 0.37 ± 0.03 3.3 ± 0.33

0.19 ± 0.01 0.53 ± 0.04 5.3 ± 0.32

Bredarydb HI (m yr1) BAI (m2 ha1 yr1) VI (m3 ha1 yr1)

a Least-square means. BAI over five years before treatment was included as a covariate in the statistical model (p < 0.002–0.14) in all cases. At Anderstorp, the proportion of Scots pine was included as a covariate for HI (p = 0.062) and the number of trees ha1 for BAI (p = 0.063) and for BI (p = 0.054). b No statistical comparison was conducted, as designed as catchment study without replication (see text for further explanation).

Fig. 1. Relative basal area increment (BAI) at the Anderstorp, Skogaryd and Bredaryd experimental sites, five years before and five–eight years after wood ash application (year 0); means (±standard error) for four (Anderstorp) and three (Skogaryd) replicates, and no replicate (Bredaryd). The post-treatment values (years 1–8; a) were adjusted for pre-treatment growth rates, i.e. by the ratio between the growth rates on the treated plots (t) and the control plots (c) during the five-year period prior to treatment (b) [(annual BAIt,a/annual BAIc,a)/(5-yr BAIt,b/5-yr BAIc,b)]. Dashed vertical line represents year of wood ash application.

rendering in the large relative differences. Also, the difference in time after fertilization may influence the magnitude of the growth response. In Anderstorp we observed a tendency (p = 0.053) to a higher biomass increment in the high compared to low ash dose (Table 1). This finding is in agreement with the study by Moilanen et al. (2005) on seven pine peatland forests in Finland, receiving varying doses of wood ash. Stand growth was improved 15 years after fertilization for all ash doses, but volume increment was positively correlated with the added nutrient contents given in the ash. In the mesotrophic Norway spruce stand at Skogaryd, calculated over the five-year period, the height and stem volume increment was significantly higher in both ash doses than in the control, but there was no difference between the two ash doses (Table 2). The biomass increment was 19% and 23% higher in the low and high ash dose, respectively, compared to the control, but this effect

163

T. Rütting et al. / Forest Ecology and Management 328 (2014) 159–166

Fig. 2. Groundwater levels (above) and soil temperature (below) at 10 cm and 20 cm depth, respectively, at the Skogaryd experimental site (means ± standard error). Measurements were conducted manually at the dates of soil greenhouse gas measurements. During winter month no measurements were conducted due to snow/ice cover of the ground.

was not statistically significant. The increased growth is lower than what has been found for mesotrophic pine peatlands in Finland, where the addition of 5 t wood ash ha1 led to a three-fold volume increment after 12 years (Moilanen et al., 2012). This was likely due to severe P and K deficiency at that site, which does not seem to be the case for Skogaryd. In addition, another peatland pine forest showed 48 years after wood ash fertilization at a rate of 8 t ha1 a 13-times higher total wood production compared to controls (Moilanen et al., 2002), which can, though, be explained by the virtually treeless control plots, as also found by Sikström et al. (2010). Again, such differences may be due to substantial differences in stand development, different times after fertilization, but also due to the amount of nutrients added with wood ash. Based on the tree ring analysis, we observed an increased basal area increment (BAI) in all three sites at both wood ash doses (Fig. 1). After a time lag of 2–3 years the growth increase reached a constant level, which was sustained over the investigated period of 5–8 years after wood ash application. Hökkä et al. (2012) reported a temporal growth response to PK-fertilization (ash or mineral fertilizers) for at least 35 years, suggesting a continuous growth response at our experimental sites, at least valid for the Scots pine stands. 3.2. Greenhouse gas exchange (Skogaryd) The ground water level did not significantly influence the fluxes of CO2, CH4 and N2O. Indeed, the water table showed only moderate fluctuation (Fig. 2), but was significantly deeper in the high ash compared to the other two treatments. This was, however, not an effect of wood ash addition, but persisted already prior to wood ash application, likely due to spatial heterogeneity in topography (Klemedtsson et al., 2010). 3.2.1. Carbon dioxide During the fourth and fifth year after wood ash application, the annual CO2 emission was not significantly different between

Table 3 Annual net soil exchange of carbon dioxide (dark respiration), methane and nitrous oxide from a drained peatland forested with Norway spruce at the Skogaryd site with and without application of wood ash at two dozes (3.3 and 6.6 tonnes d.w. ha1, respectively). The wood ash was applied on 7–8 August 2006. Different letters by rows indicate significant differences (p < 0.05) between the means (±standard error) of treatments. Perioda

Treatment Control

CO2 (kg m2 yr1) 2006–2007b 2007–2008b 2009–2010c 2010–2011c

1.70 ± 0.02 1.76 ± 0.10 1.37 ± 0.06 1.53 ± 0.10

CH4 (g m2 yr1) 2006–2007b 2007–2008b 2009–2010c 2010–2011c

0.43 ± 0.06 0.46 ± 0.06 0.27 ± 0.03 0.26 ± 0.03

N2O (g m2 yr1) 2006–2007b 2007–2008b 2009–2010c 2010–2011c

0.44 ± 0.08 0.33 ± 0.06 0.26 ± 0.05 0.21 ± 0.04

Low dose b b a a b b ab ab b b a b

High dose

1.41 ± 0.06 1.47 ± 0.10 1.20 ± 0.05 1.56 ± 0.04

a a a a

1.38 ± 0.19 1.35 ± 0.19 1.18 ± 0.19 1.51 ± 0.17

a a a a

0.45 ± 0.03 0.50 ± 0.03 0.32 ± 0.03 0.32 ± 0.03

b c b b

0.23 ± 0.07 0.23 ± 0.07 0.14 ± 0.05 0.15 ± 0.03

a a a a

0.29 ± 0.10 0.21 ± 0.06 0.21 ± 0.06 0.20 ± 0.05

a a a b

0.31 ± 0.15 0.21 ± 0.08 0.12 ± 0.05 0.09 ± 0.03

a a a a

a The respective period were: July 2006–July 2007; July 2007–June 2008; September 2009–September 2010; and September 2010–September 2011. b Data presented by Klemedtsson et al. (2010). Values for N2O fluxes are corrected. c Based on measurements conducted during snow-free period only.

treatments (Table 3), which contrasts the first two years, in which both doses significantly decreased CO2 emissions (Klemedtsson et al., 2010). This indicates that the ash-induced reduction in soil respiration may only be a short term effect. Indeed, Moilanen et al. (2012) reported for a mesotrophic pine peatland 13 years after application of 5 or 15 t ash ha1 that the soil CO2 efflux from peat decomposition increased by 77–100% compared to control. This indicates that the ash application in Skogaryd may in the

164

T. Rütting et al. / Forest Ecology and Management 328 (2014) 159–166

Fig. 3. Net soil exchange of carbon dioxide (dark respiration; top), methane (middle) and nitrous oxide (bottom) from a drained peatland forested with Norway spruce at the Skogaryd experimental site (means ± standard error). Treatments are control and wood ash application at a rate of 3.3 or 6.6 tonnes ha1 (applied on 7–8 August 2006). During winter months no measurements were conducted due to snow/ice cover of the ground.

future also lead to a stimulation of soil respiration rather than a decrease, which in turn would increase the soil GHG source strength. The soil CO2 efflux from the control plots was somewhat lower during the studied two year period (2009–2011) compared to the years 2006–2008 measured by Klemedtsson et al. (2010) (Table 2),

which was also true for CH4 uptake and N2O emission. This could be due to the fact that in both investigated winters we could not conduct flux measurements for a 3–4 month period, due to snow and ice cover of the soil (Fig. 3), but fluxes were instead interpolated. However, as winter fluxes at the Skogaryd site in earlier years were either constant or decreased linearly (Ernfors et al.,

T. Rütting et al. / Forest Ecology and Management 328 (2014) 159–166

165

Fig. 4. Overall global warming potential (symbols; means ± standard error) of the Anderstorp and Skogaryd experimental sites as affected by wood ash application at two doses (3.3 and 6.6 tonnes d.w. ash ha1, respectively). Bars present the contribution of the various ecosystem fluxes to the overall GWP.

2011; Klemedtsson et al., 2010), this is not likely an important bias and cannot explain the differences between years. More likely, the lower fluxes during 2009–2011 are a result of lower annual temperature. At the meteorological station Vänersborg (www.smhi.se), about 20 km away from Skogaryd, the year 2010 was particularly cold with an annual temperature of 4.8 °C, while the period 2006–2008 had an annual temperature of 7.7–7.9 °C. Also the years 2009–2011 had 11% less precipitation than the years 2006–2008. We, hence, conclude that the lower CO2 emissions during 2009– 2011 are due to different meteorological conditions. This also agrees with the fact that the soil CO2 efflux during the non-frozen periods (Fig. 3) followed the observed pattern in soil temperatures at 0.1 and 0.2 m depth (Fig. 2). 3.2.2. Methane The investigated peatland forest at Skogaryd was in general a sink for CH4, with only occasional emissions (Fig. 3). For the high ash dose, a lower CH4 uptake was observed, which in the latter years was only significant in comparison with the low ash treatment (Table 3). The reason for that is not fully clear, as the lower water table depth in that treatment would suggest higher CH4 oxidation and, hence, uptake. As discussed in Klemedtsson et al. (2010) a likely explanation is water stress of the microbial community. The CH4 fluxes did generally not significantly differ between the low ash dose and control (Table 3), which is in agreement with other, short-term studies (Ernfors et al., 2010; Maljanen et al., 2006). However, for three peatland forests 14–50 years after wood ash application (5–8 t ha1) it was found that, compared to control, the CH4 emissions decreased or turned into an uptake (Maljanen et al., 2006), which agrees with the tendency for higher CH4 uptake in the low wood ash treatment (Table 2). 3.2.3. Nitrous oxide The N2O emission showed no consistent temporal (annual) pattern, but two main peaks were observed in early winter 2009 and autumn 2011 (Fig. 3). Notably, these high peaks were not observed in the high ash dose treatment. There were no significant differences between the N2O emissions from the treatments during the fourth year after wood ash application (September 2009– September 2010), but the high ash dose plots emitted less than half

the amount of N2O during the fifth year compared to both, low ash dose and control (Table 3). Our results, hence, partly contrast the findings of the first two years after wood ash application of lower N2O emission in both ash treatments (Table 3), but agree with five Finnish peatland forests (Maljanen et al., 2006). The lower N2O emission in the first two years after wood ash application was explained by an increase in soil pH (Klemedtsson et al., 2010), a soil property known to strongly affect N2O emission from organic forest soils (Weslien et al., 2009). During the latter years of the experiment, we did not observe differences in soil pH between treatments (data not shown), which could explain why wood ash application did not affect N2O emissions. However, as in the fifth year after wood ash application the N2O emission from the high ash dose was again significantly lower compared to the other two treatments, more long-term investigations are needed to follow the effects of wood ash. In order to assess the environmental impact of land use and produced goods, it is not only of interest to evaluate GHG emissions on areal basis, but in relation to the amount of produced goods. In agriculture, specific N2O emissions have been used (Ma et al., 2010), that is N2O emissions per tonne produced grain, but such measures have so far not been used for forest products. In the present study, the specific N2O emission over the five year period from the control treatments was 0.41 (±0.10) g N2O kg1 d.w. biomass increment. The addition of wood ash decreased the specific N2O emission by 40–60%, with fluxes of 0.25 (±0.09) and 0.19 (±0.09) g N2O kg1 for the low and high ash dose, respectively. This means that the addition of wood ash leads to the production of wood with a lower N2O footprint.

3.3. Total GWP sink–source strength Averaged over all investigated years, the controls at both sites acted as an overall sink for GHG, with a GWP of 0.27 (±0.09) kg CO2_eq m2 yr1 at Anderstorp and 1.07 (±0.17) kg CO2_eq m2 yr1 at Skogaryd. This finding agrees with 68 peatland forests in Finland, which all showed a negative GWP (Ojanen et al., 2013). Irrespective of dose and site, wood ash applications decreased the GWP and, hence, increased the GHG sink strength (Fig. 4). For Anderstorp, the pine forest had a GWP of 0.48 (±0.08) and

166

T. Rütting et al. / Forest Ecology and Management 328 (2014) 159–166

0.59 (±0.09) kg CO2_eq m2 yr1 for low and high ash dose, respectively. In Skogaryd the wood ash addition also increased the overall GHG sink by 42% and 54% for low and high dose, respectively (Fig. 4). The reduced GWP at both sites was almost entirely an effect of the enhanced biomass increment and litter production, as both non-CO2 GHGs made only small contributions to the overall GWP. Our finding of a reduced GWP by wood ash applications agrees with a study on a Finnish peatland forest, in which 13 years after wood ash addition (5–15 t ha1) a reduced CO2 source strength by 72–79% was found (Moilanen et al., 2012). This was due to a higher stimulation of biomass increment than soil respiration, which was also found in our study. Overall we conclude that the addition of wood ash to peatland forests, not growth limited by plant-available N, leads to an increase in tree growth and, hence, C sequestration. This has the potential to decrease the total GHG source strength of these ecosystems or turn them even into GHG sinks. However, as studies were limited to <10 years after wood ash application, more longterm, decadal data are needed to investigate the persistence of the reduced GWP by wood ash application. Acknowledgements We sincerely thank Mikael Westerlund (Skogforsk, The Forestry Research Institute of Sweden), David Allbrand (University of Gothenburg) and Jonas Persson for their assistance in the field. This work was funded by the Thermal Engineering Research Institute (Värmeforsk) as part of the programme ‘‘Environmentally correct utilisation of ashes’’ and by the Swedish Energy Agency. We also acknowledge the strategic research area Biodiversity and Ecosystem services in a Changing Climate (BECC, http://www.cec.lu.se/ research/becc). References Albrektson, A., 1988. Needle litterfall in stands of Pinus sylvestris L. in Sweden, in relation to site quality, stand age and latitude. Scand. J. Forest Res. 3, 333–342. Alm, J., Shurpali, N.J., Minkkinen, K., Aro, L., Hytönen, J., Laurila, T., Lohila, A., Maljanen, M., Martikainen, P.J., Mäkiranta, P., Penttilä, T., Saarnio, S., Silvan, N., Tuittila, E.-S., Laine, J., 2007. Emission factors and their uncertainty for the exchange of CO2, CH4 and N2O in Finnish managed peatlands. Boreal Environ. Res. 12, 191–209. Björk, R.G., Ernfors, M., Sikstrom, U., Nilsson, M.B., Andersson, M.X., Rütting, T., Klemedtsson, L., 2010. Contrasting effects of wood ash application on microbial community structure, biomass and processes in drained forested peatlands. FEMS Microbiol. Ecol. 73, 550–562. Demeyer, A., Voundi Nkana, J.C., Verloo, M.G., 2001. Characteristics of wood ash and influence on soil properties and nutrient uptake: an overview. Bioresour. Technol. 77, 287–295. Ernfors, M., Rütting, T., Klemedtsson, L., 2011. Increased nitrous oxide emissions from a drained organic forest soil after exclusion of ectomycorrhizal mycelium. Plant Soil 343, 161–170. Ernfors, M., Sikström, U., Nilsson, M., Klemedtsson, L., 2010. Effects of wood ash fertilization on forest floor greenhouse gas emissions and tree growth in nutrient poor drained peatland forests. Sci. Total Environ. 408, 4580–4590. Ernfors, M., Von Arnold, K., Stendahl, J., Olsson, M., Klemedtsson, L., 2007. Nitrous oxide emissions from drained organic forest soils – an up-scaling based on C: N ratios. Biogeochemistry 84, 219–231. Ferm, A., Hokkanen, T., Moilanen, M., Issakainen, J., 1992. Effects of wood bark ash on the growth and nutrition of a Scots Pinus sylvestris L. afforestation in central Finland. Plant Soil 147, 305–316. Fritze, H., Smolander, A., Levula, T., Kitunen, V., Mälkönen, E., 1994. Wood-ash fertilization and fire treatments in a Scots Pinus sylvestris L. forest stand: effects on the organic layer, microbial biomass, and microbial activity. Biol. Fertil. Soils 17, 57–63. Hökkä, H., Repola, J., Moilanen, M., 2012. Modelling volume growth response of young Scots pine (Pinus sylvetris) stands to N, P, and K fertilization in drained Peatland sites in Finland. Can. J. Forest Res. 42, 1359–1370. Hökkä, H., Repola, J., Laine, J., 2008. Quantifying the interrelationship between tree stand growth rate and water table level in drained peatland sites within Central Finland. Can. J. Forest Res. 38, 1775–1783. IPCC (Ed.), 2007. Climate Change 2007: The Physical Science Basis. Cambridge University Press, Cambridge.

Klemedtsson, L., Ernfors, M., Björk, R.G., Weslien, P., Rütting, T., Crill, P., Sikström, U., 2010. Reduction of greenhouse gas emissions by wood ash application to a Picea abies forest on a drained organic soil. Eur. J. Soil Sci. 61, 734–744. Klemedtsson, L., Kasimir Klemedtsson, Å., Moldan, F., Weslien, P., 1997. Nitrous oxide emission from Swedish forest soils in relation to liming and simulated increased N-deposition. Biol. Fertil. Soils 25, 290–295. Klemedtsson, L., von Arnold, K., Weslien, P., Gundersen, P., 2005. Soil CN ratio as a scalar parameter to predict nitrous oxide emissions. Global Change Biol. 11, 1142–1147. Ma, B.L., Wu, T.Y., Tremblay, N., Deen, W., Morrison, M.J., McLaughlin, N.B., Gregorich, E.G., Stewart, G., 2010. Nitrous oxide fluxes from corn fields: on-farm assessment of the amount and timing of nitrogen fertilizer. Global Change Biol. 16, 156–170. Mäkiranta, P., Minkkinen, K., Hytönen, J., Laine, J., 2008. Factors causing temporal and spatial variation in heterotrophic and rhizospheric components of soil respiration in afforested organic soil croplands in Finland. Soil Biol. Biochem. 40, 1592–1600. Maljanen, M., Nykänen, H., Moilanen, M., Martikainen, P.J., 2006. Greenhouse gas fluxes of coniferous forest floors as affected by wood ash addition. For. Ecol. Manage. 237, 143–149. Maljanen, M., Sigurdsson, B.D., Guðmundsson, J., Óskarsson, H., Huttunen, J.T., Martikainen, P.J., 2010. Greenhouse gas balances of managed peatlands in the Nordic countries – present knowledge and gaps. Biogeoscience 7, 2711–2738. Marklund, L.-G., 1988. Biomassafunktioner för gran, tall och björk i Sverige. Report 45. Department of Biometrics, Swedish University of Agricultural Sciences, Umeå (In Swedish). Meyer, A., Tarvainen, L., Nousratpour, A., Björk, R.G., Ernfors, M., Grelle, A., Kasimir Klemedtsson, Å., Lindroth, A., Räntfors, M., Rütting, T., Wallin, G., Weslien, P., Klemedtsson, L., 2013. A fertile peatland forest does not constitute a major greenhouse gas sink. Biogeoscience 10, 7739–7758. Moilanen, M., Hytönen, J., Leppälä, M., 2012. Application of wood ash accelerates soil respiration and tree growth on drained peatland. Eur. J. Soil Sci. 63, 467– 475. Moilanen, M., Saarinen, M., Silfverberg, K., 2010. Foliar nitrogen, phosphorus and potassium concentrations of Scots pine in drained mires in Finland. Silva Fenn. 44, 583–601. Moilanen, M., Silfverberg, K., Hökkä, H., Issakainen, J., 2005. Wood ash as a fertilizer on drained mires – growth and foliar nutrients of Scots Pinus sylvestris L. Can. J. Forest Res. 35, 2734–2742. Moilanen, M., Silfverberg, K., Hokkanen, T.J., 2002. Effects of wood-ash on the tree growth, vegetation and substrate quality of a drained mire: a case study. Forest Ecol. Manage. 171, 321–338. Näslund, M., 1947. Funktioner och tabeller för kubering av stående träd - tall, gran och björk i södra Sverige samt i hela landet (functions and tables for computing the cubic volume of standing trees – pine, spruce and birch in southern Sweden, and in the whole of Sweden). Meddelanden från Statens Skogsforskningsinstitut, Stockholm, vol. 36, pp. 1–81. (In Swedish, with English summary). Ojanen, P., Minkkinen, K., Alm, J., Penttilä, T., 2010. Soil-atmosphere CO2, CH4 and N2O fluxes in boreal forestry-drained peatlands. Forest Ecol. Manage. 260, 411– 421. Ojanen, P., Minkkinen, K., Penttilä, T., 2013. The current greenhouse gas impact of forestry-drained boreal peatlands. Forest Ecol. Manage. 289, 201–208. Økland, R.H., Økland, T., Rydgren, K., 2001. Vegetation-environment relationships of boreal spruce swamp forest in Østmarka Nature Reserve, SE Norway. Sommerfeltia 29, 1–190. Petersson, H., Ståhl, G., 2006. Functions for below ground biomass of Pinus sylvestris, Picea abies, Betula pendula and B. pubescens in Sweden. Scand. J. Forest Res. 21, 84–93. Ring, E., von Brömssen, C., Losjö, K., Sikström, U., 2011. Water chemistry following wood-ash application to a Scots pine stand on a drained Peatland in Sweden. Forestry Stud. 54, 54–70. Sikström, U., Almqvist, C., Jansson, G., 2010. Growth of Pinus sylvestris after the application of wood ash or P and K fertilizer to a peatland in Southern Sweden. Silva Fenn. 44, 411–425. Šimek, M., Cooper, J.E., 2002. The influence of soil pH on denitrification: progress towards the understanding of this interaction over the last 50 years. Eur. J. Soil Sci. 53, 345–354. Väätäinen, K., Sirparanta, E., Räisänen, M., Tahvanainen, T., 2011. The costs and profitability of using granulated wood ash as a forest fertilizer in drained peatland forests. Biomass Bioenergy 35, 3335–3341. Weslien, P., Kasimir Klemedtsson, Å., Börjesson, G., Klemedtsson, L., 2009. Strong pH influence on N2O and CH4 fluxes from forested organic soils. Eur. J. Soil Sci. 60, 311–320. Weslien, P., Klemedtsson, L., Svensson, L., Kasimir Klemedtsson, Å., Gustafsson, A., 1998. Nitrogen losses following application of pig slurry to arable land. Soil Use Manage. 14, 200–208. Yu, Z., Loisel, J., Brosseau, D.P., Beilman, D.W., Hunt, S.J., 2010. Global peatland dynamics since the last glacial maximum. Geophys. Res. Lett. 37, L13402, doi:13410.11029/12010GL043584. Zimmermann, S., Frey, B., 2002. Soil respiration and microbial properties in an acid forest soil: effects of wood ash. Soil Biol. Biochem. 34, 1727–1737.