Soil CO2 efflux and net ecosystem exchange following biomass harvesting: Impacts of harvest intensity, residue retention and vegetation control

Forest Ecology and Management 360 (2016) 181–194

Contents lists available at ScienceDirect

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

Soil CO2 efflux and net ecosystem exchange following biomass harvesting: Impacts of harvest intensity, residue retention and vegetation control K.L. Webster a,⇑, S.A. Wilson a, P.W. Hazlett a, R.L. Fleming a, D.M. Morris b a b

Canadian Forest Service, Great Lakes Forestry Centre, 1219 Queen St. East, Sault Ste. Marie, ON P6A 2E5, Canada Ontario Ministry of Natural Resources and Forestry, Centre for Northern Forest Ecosystem Research, Thunder Bay, ON P7B 5E1, Canada

a r t i c l e

i n f o

Article history: Received 4 August 2015 Received in revised form 13 October 2015 Accepted 19 October 2015

Keywords: Bioenergy Biomass harvesting Soil respiration Net ecosystem exchange Forest management Carbon

a b s t r a c t Biomass harvesting removes more woody material than would be taken with conventional forest harvesting. Harvesting residues, left on site are an important substrate for micro-organisms that maintain nutrient cycles essential for future forest productivity by mineralizing organic matter, and releasing carbon dioxide (CO2) as a respiratory bi-product. We assessed the impact of biomass removal intensity (stem-only [SO], full-tree biomass [FT], full-tree biomass plus stumping [FT + S], full-tree biomass plus stumps and forest floor removed [FT + B]), and herbicide application on soil respiration and net ecosystem exchange of carbon (C) in a harvested 40-yr-old jack pine stand. Soil respiration (surface CO2 efflux) normalized to 15 °C (R15) was lower in biomass harvest treatments than in the uncut stand and a mature 80-yr-old fire-origin natural stand. Among harvest treatments, R15 was positively related to the amount of C retained, with the general pattern of FT + B < FT + S < FT
SO. Differences in R15 among treatments were primarily related to residue and soil organic matter quantity and quality (i.e., presence of mineral soil and forest floor polysaccharide). Herbicide application further reduced R15 by diminishing root respiration, although herbicide treatments in the SO, FT and FT + S resulted in greater net CO2 fluxes to the atmosphere in August because herbaceous photosynthesis was greatly reduced. We suggest that criteria for determining site-specific biomass retention should take into account the amount and type of residue required to maintain microbial soil respiration driving nutrient cycling. Crown Copyright Ó 2015 Published by Elsevier B.V. All rights reserved.

1. Introduction There is growing interest in the use of forest harvest residues and non-merchantable biomass (i.e., coarse and fine woody debris, non-target and undersized trees) for bioenergy production. Greater utilization of residues can partially replace the use of fossil fuels, reducing longer-term greenhouse gas emissions (i.e. carbon [C] offsets) and diversify a country’s energy portfolio (Roach and Berch, 2014; Ter-Mikaelian et al., 2015). However, energy diversification and economic development should not compromise ecological sustainability (Lattimore et al., 2009). Downed woody debris, including harvest residue, have important on-site roles in sustaining soil nutrient cycles (Nambiar, 1996; Powers et al., 2005; Wall, 2012). Biomass harvesting directly reduces the amount of labile residue, such as leaves, needles and ⇑ Corresponding author. E-mail address: [email protected] (K.L. Webster). http://dx.doi.org/10.1016/j.foreco.2015.10.032 0378-1127/Crown Copyright Ó 2015 Published by Elsevier B.V. All rights reserved.

fine woody debris (Ewel et al., 1987; Marshall, 2000; Wall, 2012), by removing smaller diameter trees than conventional harvesting for traditional wood products (i.e., saw logs and pulpwood). Residues provide a substrate for soil microbes that mineralize organic matter and release nutrients (e.g., nitrogen [N]) with carbon dioxide (CO2) released as a respiratory by-product (Raich and Tufekcioglu, 2000). Harvesting and silvicultural prescriptions affect the near surface micro-climate (e.g., solar radiation, temperature, moisture and wind [Fleming et al., 1998; Proe et al., 2001]) and soil environment (e.g., compaction, profile turnover and mixing [McNabb et al., 2001; Marshall, 2000; Williamson and Neilsen, 2000]). If substrate or environmental conditions are limiting, decomposition slows and heterotrophic soil respiration declines, resulting in less mineralization of nutrients for plant uptake (Fox, 2000; Grigal, 2000; Marshall, 2000; Thiffault et al., 2011). Harvesting intensity and reforestation practices affect net ecosystem exchange (NEE) of CO2 with the atmosphere. NEE is a function of CO2 released through soil respiration (Rs), which

182

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

includes both heterotrophic (from microbes) and autotrophic (from plants) respiration, minus CO2 sequestered in regenerating vegetation. Factors affecting the re-establishment of ground vegetation and seedlings such as mineral soil exposure, changes to the soil environment (e.g., temperature, moisture) and application of herbicide (Roberts, 2007) also affect NEE following harvest. Most studies evaluating the impacts of harvesting on soil respiration and NEE have not examined a broad gradient of biomass removal intensities and post-harvest reforestation practices (e.g., site preparation and herbicide application). Studies examining soil respiration following forest harvesting (typically conventional harvesting, occasionally whole-tree harvesting) have been few and have produced inconsistent findings (Peng et al., 2008). Studies report Rs in the first few years following clearcut harvesting in the boreal to increase (Gordon et al., 1987; Mallik and Hu, 1997), decrease (Striegl and Wickland, 1998; Pumpanen et al., 2004; Moroni et al., 2009) or have little change (Fleming et al., 2006). Clear patterns have been elusive due to the time between harvest and Rs measurements, variation in tree species, understory composition, stand age, site fertility, and amount of residues (Peng et al., 2008). However, harvesting generally results in sites being a net CO2 source to the atmosphere after harvest (Liski et al., 1998; Pypker and Fredeen, 2002a; Zha et al., 2009) though some report a small sink (Pypker and Fredeen, 2002b). This study examines the impacts of a range of biomass harvest intensities, including intensive bioenergy harvesting, of a 40-yearold second growth, irregularly spaced jack pine stand (Pinus banksiana Lamb.), in the boreal forest of northeastern Ontario. Jack pine woodlands are an important component of the boreal forest, covering over 2 million km2 of predominantly well-drained uplands in northern North America (Law and Valade, 1994; Lowe et al., 1994). These deep, well-drained coarse-textured soils support productive forests but have limited water-holding capacity and nutrient reserves, which raises long-term sustainability concerns (Foster, 1996). These low quality mid-rotation stands are potential candidates for biomass harvest to allow for stand rehabilitation, yet there is little data on impacts on these sites and results from mature stands may not be applicable given the historical legacy of build-up of coarse woody debris and forest floor over time. Rs and NEE are expected to change with increasing biomass removal due to changes in the quantity and quality of residues and in environmental conditions. The key question is how does intensification of biomass removal and regeneration practices affect the (1) magnitude, (2) source, and; (3) physical vs chemical controls on Rs and NEE. This information is essential to aid in determining the optimal level of biomass removal that still ensures sufficient decomposition of organic matter to mineralize nutrients that ensures future tree productivity while minimizing the site’s C source to the atmosphere. 2. Material and methods

mature natural stand at Nimitz (MN). The soils are Dystric Brunisols (Soil Classification Working Group, 1998), formed over rapidly-drained, coarse textured, glacial–fluvial deposits. Surface organic horizons (forest floor) are classified as HumiFibrimors having an average depth of 10 cm – and support a continuous carpet of feathermoss and understory herbs and shrubs (Kwiaton et al., 2014). For the Chapleau region the mean annual temperature (MAT) is 2.0 °C, with 1444 growing degree days (>5 °C) and 92 frost free days, typically from early June to early September. Daily maximum temperatures are highest in July and coldest in January. Mean annual precipitation is 827 mm (545 mm in rainfall, 282 cm in snowfall) with September being the wettest month and February being the driest (Environment Canada, 2014). During 2012 the months of May to October were warmer (by 1.3 °C) and slightly drier (by 17 mm) than the climate normal. 2.2. Island Lake biomass harvest experiment At the Island Lake site a 50 ha area was set aside for harvesting (Fig. 1). Harvesting occurred during December 2010 and January 2011, with 70  70 m plot treatments laid out during July to September 2011. Plots were positioned at least 50 m from uncut forest to minimize forest edge effects, with individual plots separated by at least 20 m. There are five blocks (randomized complete block design) containing each of the four harvest treatments of stem-only harvest [SO], where only the bole of each tree was removed leaving stumps and upper branches on site after harvest; full tree biomass harvest [FT], where the entirety of each tree (including traditionally non-merchantable trees) upwards from the stump was removed; stumped [FT + S], where the full tree biomass harvest was followed by stump removal; bladed [FT + B], which consisted of a full-tree biomass harvest, stumping and removal of the forest floor by blading (Fig. 1). The applied treatments resulted in a broad gradient of removals, with the amount of C removed increasing with biomass removal intensity from SO (30.5 Mg ha1) to FT (54.6 Mg ha1) to FT + S (74.6 Mg ha1) and finally FT + B (108.5 Mg ha1) (Fig. 2; Kwiaton et al., 2014). Each of the harvested plots was subdivided into four 35 m by 35 m sub-plots, using a split plot design. Two sub-plots were sprayed with a glyphosate herbicide (VisionÒ, at 4 L of product ha1) to control vegetation, while the other two sub-plots had no vegetation control. Disc trenching within SO, FT, and FT + S created repetitive rows of flat areas, trenches and debris from trenches. At UC and MN there were 5 replicate 70  70 m plots within each site. In May 2012, the buffer area was planted to jack pine, while sub-plots in each of the treatment plots were split between jack pine and black spruce (Picea mariana [Mill.] Britton), with each species planted in one herbicide-treated and one non-treated quadrant at approximately 1.8 by 2 m spacing. Seedlings were overwintered planting stock grown in jiffy pots with improved seed from Ontario seed zone 24.

2.1. Study sites 2.3. Soil respiration and NEE sampling The study was carried out at two boreal jack pine-dominated boreal sites near Chapleau, Ontario (Fig. 1). The first site (Island Lake N 47.7° W 83.6°) was a 40-year old second growth jack pine stand established following clearcut harvesting during the 1960s of a mature fire-origin stand. The site was mechanically site prepared with Young’s teeth, hand seeded and subsequently fill planted. The second site currently supports an 84-year old fire origin jack pine stand (Nimitz N 47.6° W 83.3°). Details related to the study sites are given in Kwiaton et al. (2014) and Fleming et al. (2014) and summarized briefly here. The height at age 50 (site index) is 19.3 m for the uncut control adjacent to the harvested area at Island Lake (UC) and 18.8 m for

Rs was measured monthly during the first growing season (May to October 2012) after harvest in 3 of the 5 treatment blocks, and at 3 of the 5 plots within UC and MN. In each plot, soil respiration was monitored in each of the 4 sub-plots, and for disc-trenched treatments, within the area of undisturbed forest floor between the trench and trench debris rows. All measurements were made between 1000 h and 1400 h using the static chamber method (Livingston and Hutchinson, 1995). This involved placing a portable acrylic flux chamber (49.5  49.5  40 cm = 90.2 L volume) over permanently installed square aluminum collars (0.21 m2 measurement area; installed in fall 2011 to allow equilibration

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

183

Fig. 1. Map of the Island Lake biomass harvest research and demonstration area. Within each plot (square) the number indicates the block and the letter indicates the harvest treatment: forest floor removal (B), stumped (S), full-tree (F) and tree-length (T) harvest treatments and uncut control (C). The inset map shows the location of the Island Lake study site and the mature natural (MN) stand (‘‘Nimitz”). The plot enlargement shows the split-plot locations of the herbicide (h) and no herbicide (nh) treatments, and the planting of jack pine (Pj) and black spruce (Sb) seedlings.

Fig. 2. Components of carbon retained (Mg ha1; stacked bar) and removed (Mg ha1; line) from the biomass removal treatments. Bars indicate standard deviation of total carbon retained and removed.

and prevent flushing from root or soil disturbance) using a waterfilled channel to create an air-tight seal. Chambers were equipped with a mixing fan placed vertically at the top of the chamber to facilitate mixing without disturbing the air–soil boundary layer. CO2 concentrations within the chambers were measured with a

Vaisala CARBOCAPÒ CO2 Probe GMP343 (Vaisala, Helsinki, Finland) attached to an infrared gas analyzer (IRGA) and logged with a Vaisala MI70 controller. The controller had an internal compensation for chamber humidity (50%), oxygen concentration (20.95%), pressure (101.3 kPa) and temperature (built-in temperature sensor,

184

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Vaisala HM70). Measurements of Rs that include heterotrophic respiration and respiration of root and aboveground biomass (where it occurred) were taken with the chamber opaque. NEE measurements were taken once in August (August 8 and 9) during peak vegetation growth for the season using a clear chamber (photosynthesis minus respiration). CO2 fluxes for Rs and NEE were calculated by linear regression of the slope of increasing CO2 concentration in the chambers with time. CO2 fluxes were scaled according to the total volume determined by summing the volume of the chamber with the volume of each collar, adjusting for the topography of the surface within the collar, and corrected for chamber temperature and ambient pressure. 2.4. Partitioning soil respiration sources and R15 Soil respiration gas fluxes were partitioned into root, forest floor and mineral soil sources. Mineral Rs, derived from the mineral soil surface, was assumed equal to the respiration from the herbicide FT + B treatment, where the forest floor was removed, and thus assumed constant among treatments. Forest floor respiration was the difference between herbicide respiration for any treatment and that of the FT + B. Autotrophic respiration (root plus aboveground respiration of herbs and small shrubs) was the difference between non-herbicide and herbicide plot for each treatment. Normalized soil respiration at 15 °C (R15) was estimated by fitting an exponential curve based on temperature dependent enzyme kinetics of soil temperature at 5 cm depth to observed soil respiration measurements from May to October at each collar, and then calculating the soil respiration at 15 °C from the regression equation. 2.5. Soil environment Soil (which included forest floor and mineral layer) environmental variables were measured synoptically at the same time as gas measurements. Soil temperature was measured at a depth of 5 cm into the mineral soil using a hand-held portable electronic temperature probe. Soil moisture was measured using Delta T theta probe, which integrates soil moisture over the top 6 cm from the soil surface. Soil moisture, measured as a voltage, was converted to volumetric moisture (%) (Delta T Devices, Ltd., 1999).

2.6. Substrate quality Substrate quality was based on C, N, P and K concentrations, C:N, and organic functional groups. The forest floor and top 10 cm of mineral soil were sampled in late June by compositing 6 forest floor grab samples (including LFH horizons and fine logging residues) and 6 mineral soil core samples from each plot. These samples were analyzed for C and N by high temperature combustion with a Vario EL III Elementar analyzer. Carbon functional groups (Fig. 3) were determined using the Diffuse Reflectance Infrared Fourier Transformation (FTIR) method on a Varian Scimitar FTS 800 midIR FTIR spectrometer with a 60° PermaLine interferometer and ambient DTGS detector. Samples were mixed with IR grade KBr (1:6 ratio – S:KBr) and hand-ground to a fine homogenous mixture, and analyzed on the Diffuse Reflectance Accessory used for direct analysis of solid samples. For fine logging residues total C was determined using a LECO C/N/S analyser, total N was determined using the semi-micro Kjeldhal procedure, and phosphorus (P), potassium (K), calcium (Ca) and magnesium (Mg) were determined by inductively coupled argon plasma emission spectrometry after acid digestion.

2.7. Statistical analyses Differences in plot averaged (average of 4 sub-plots) R15, respiration components, soil temperature, soil moisture and soil chemistry for harvest and unharvested treatments were examined using a linear, mixed effects model ANOVA with customized orthogonal contrasts. The effect of herbicide application in the harvested treatment plots (average of 2 sub-plots) was examined using a splitplot linear mixed effects model ANOVA with customized orthogonal contrasts. The lme function was used for the linear mixed effects model using the R package nlme (Pinheiro et al., 2014). Differences in chemistry among residue types were analyzed using one-way ANOVA on ranks using Dunn’s method for multiple comparisons in SigmaPlot12 (Systat Software, Inc., 2011). Plot average non-herbicided R15 was related to C retention using a sigmoidal function and to substrate quality parameters using multiple linear regression with forward step-wise variable selection within SigmaPlot12 (Systat Software, Inc., 2011).

0.14

0.10 0.08 0.06

fats and lipids

0.12

polysaccharide

Forest floor Mineral soil

lignin, aromatics or carboxylates

0.16

cellulose

Relative abundance (normalize absorbance)

0.18

0.04 0.02 0.00 4000

3000

2000

1000 -1

Wave number (cm ) Fig. 3. Relative intensities from FTIR scans of forest floor samples (solid line) and mineral (dashed line) samples with different functional group peaks identified.

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Fig. 4. Mass (Mg ha1 dry weight) and composition of post-harvest fine residue on FT, FT + S and SO treatments.

3. Results 3.1. Residue retention, composition and chemistry The biomass harvest treatments resulted in a gradient of C retention, ranging from 133.0 Mg ha1 (SO) to 46.0 Mg ha1 (FT + B) (p < 0.001; Fig. 2). Harvesting resulted in a large increase in coarse (>5 cm in diameter) and fine residues (<5 cm in diameter) in SO, and increases primarily in fine residue in FT and FT + S. Twigs (0–2 cm diameter) represented the dominant component of the fine residue with branches, foliage and cones contributing less (Fig. 4). The chemistry of the residue components varied with needles having significantly (p < 0.001) higher K, P, K, Ca and Mg and lower C:N than large twigs and small twigs having intermediate values (Fig. 5). 3.2. Carbon exchange 3.2.1. Soil respiration Respiration was highest during the summer months of June, July and August in the SO, FT, and FT + S treatments and continued to be sustained into September in UC and MN plots (Fig. 6). Respiration remained low (<1 lmol m2 s1) and consistent throughout

185

the growing season (May to October) in the FT + B treatment (Fig. 6). Calculated R15 values for the different treatments had coefficients of determination (r2) ranging from 0.36 to 0.99 with a median value of 0.81 (data not shown). R15 was highest in the MN and the unharvested plots (MN and UC) had higher R15 than in the biomass harvest plots (Tables 1 and 2). Among the biomass treatment plots R15 decreased in the general trend of SO
FT > FT + S > FT + B (Tables 1 and 2). Herbicide application in SO, FT, and FT + S treatments resulted in significant declines in R15 of 0.23, 0.28, and 0.38 lmol m2 s1 respectively (Table 2). Respiration of CO2 from the soil surface was dominated by respiration from within the forest floor (accounting for 46–74% across months and treatments), with the highest forest floor respiration rates during June, July and August (Fig. 7). During these summer months forest floor respiration was significantly higher in FT and SO than in FT + S (p = 0.002). Root respiration was equal to or greater than mineral soil respiration in SO, FT and FT + S for the months of July, August and September but lower in May, June and October (Fig. 7). 3.2.2. August net ecosystem exchange The SO, FT and FT + S treatments with herbicide application, along with FT + B, UC understorey and MN understorey, were net sources of CO2 to the atmosphere on August 8 and 9 due to low rates of photosynthesis compared to respiration (Fig. 8). In contrast, the non-herbicided plots of FT + S, FT, and SO were net CO2 sinks due to high rates of photosynthesis compared to respiration (Fig. 8). 3.3. Soil environment Average summer mid-day soil temperatures were cooler in MN and UC plots than the harvested treatments (Tables 1 and 2), with FT + B having the warmest average temperature. FT + B also had the highest average summer soil moisture, with the other harvest treatment plots and MN and UC having lower soil moisture (Tables 1 and 2). Soil moisture was also affected by vegetation control and was significantly higher in the herbicide treatment (Tables 1 and 2). 3.4. Substrate quality A series of orthogonal contrasts (Table 3) showed that there were differences between the unharvested plots in soil chemistry.

Fig. 5. Chemical composition of fine residues post-harvest. Bars denote standard deviation.

186

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Fig. 6. Mean (±standard deviation) of soil surface respiration (lmol CO2 m2 s1) for each of the harvest treatments for each of the sampling dates from May to October 2012. In harvested treatments, non-herbicided plots are indicated by solid lines and herbicided plots with dashed lines.

In the forest floor, MN had higher lignin, aromatics or carboxylates, C, N, P, and lower polysaccharide levels than UC. In the mineral layer MN had higher cellulose and K and lower polysaccharide and C:N levels than UC (Table 1). Differences were also observed between the unharvested plots (MN and UC) and the harvested plots (SO, FT, FT + S, FT + B) soil chemistry (Table 3). In the forest floor, the harvested treatments had higher polysaccharide and lower fats and lipids (symmetric and antisymmetric), C and N. In the mineral layer the harvested treatments had higher polysaccharide and K and lower fats and lipids (symmetric and antisymmetric) and cellulose than unharvested (Table 1). Among the harvested treatments, FT + B plots had had lower polysaccharide and K and higher lignin, aromatics

or carboxylates in the mineral soil compared to the other harvested treatments (SO, FT, FT + S) (Table 1). The stump removal treatment increased C:N in forest floor and polysaccharide and lignin, aromatics or carboxylates in the mineral layer and decreased the K in the mineral layer compared to SO and FT treatments (Table 1). There was no change in forest floor or mineral soil chemistry between SO and FT treatments, with the exception of higher P in the mineral layer of SO. Herbicide effects were not very evident in the soil chemistry, with the exception of the C:N of the forest floor being lower in the herbicide plots (Table 3). There was also an interaction effect of herbicide with treatment. In FT + S and FT + B treatments, herbicide plots had lower fats and lipids (symmetric and antisymmetric)

187

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Table 1 Average ± standard deviation of R15 (lmol m2 s1 at 15 °C) soil temperature at 5 cm (°C), soil moisture integrated over the top 6 cm in harvest treatments (%), and forest floor and mineral soil properties (% or relative abundance). Average includes herbicided and non-herbicided plots of all three blocks for SO, FT, FT + S, FT + B (n = 6 per treatment) or uncut plots for UC and MN (n = 3 per site). Treatment

Respiration and environment R15 (lmol m2 s1 at 15 °C)

Average Sd Average Sd Average Sd

Temperature (°C) Moisture (%) Forest floor N (%) C (%) C:N K (%) P (%) Polysaccharide (relative abundance @ 1030 cm1) Lignins, aromatics or carboxylates (relative abundance @ 1622 cm1) Fats and lipids – symmetric (relative abundance @ 2851 cm1) Fats and lipids – antisymmetric (relative abundance @ 2920 cm1) Cellulose (relative abundance @ 3310 cm1) Mineral N (%) C (%) C:N K (%) P (%) Polysaccharide (relative abundance @ 1030 cm1) Lignins, aromatics or carboxylates (relative abundance @ 1622 cm1) Fats and lipids – symmetric (relative abundance @ 2851 cm1) Fats and lipids – antisymmetric (relative abundance @ 2920 cm1) Cellulose (relative abundance @ 3310 cm1)

FT + B

FT + S

FT

SO

UC

MN

0.4 0.03 19.2 1.97 14.1 0.99

1.67 0.49 16.7 1.31 8.5 0.17

2.16 0.49 16.6 4.01 7.1 0.26

2.25 0.61 16.2 2.91 7.7 0.35

3.29 0.63 11.9 1.34 6.7 2.78

4.29 0.37 13.6 3.17 8.6 0.48

0.70 0.32 20.9 9.4 29.8 4.3 231 94 42.4 0.0917 0.0121 0.0344 0.0079 0.0428 0.0096 0.0509 0.0113 0.0610 0.0123

0.68 0.25 18.1 8.1 26.1 2.0 254 73 51.6 0.1029 0.0135 0.0358 0.0071 0.0372 0.0093 0.0446 0.0113 0.0517 0.0126

0.54 0.20 13.9 5.1 26.0 1.2 215 57 35.0 0.1024 0.0046 0.0364 0.0063 0.0388 0.0048 0.0469 0.0060 0.0543 0.0068

0.78 0.23 21.9 6.0 28.3 1.4 262 112 37.9 0.0944 0.0075 0.0348 0.0024 0.0448 0.0055 0.0541 0.0069 0.0599 0.0063

1.01 0.10 27.9 3.9 27.6 2.1 319 29 58.2 0.0757 0.0129 0.0418 0.0032 0.0456 0.0013 0.0535 0.0014 0.0638 0.0034

0.10 0.02 2.0 0.5 19.6 0.7 26.7 5.1 6.5 1.8 0.1150 0.0018 0.0189 0.0016 0.0149 0.0014 0.0189 0.0017 0.0358 0.0026

0.11 0.03 2.1 0.5 19.5 1.4 31.5 5.3 6.1 1.9 0.1086 0.0058 0.0170 0.0033 0.0140 0.0034 0.0177 0.0042 0.0354 0.0072

0.11 0.01 2.2 0.2 19.7 0.9 36.3 3.6 9.3 3.4 0.1044 0.0017 0.0149 0.0012 0.0165 0.0012 0.0200 0.0014 0.0329 0.0021

0.12 0.04 2.6 1.5 20.9 4.7 19.4 1.0 7.5 0.4 0.1056 0.0052 0.0152 0.0025 0.0185 0.0045 0.0222 0.0059 0.0347 0.0066

0.10 0.01 1.7 0.1 17.5 0.4 26.8 5.4 7.0 2.2 0.0980 0.0050 0.0171 0.0012 0.0197 0.0019 0.0235 0.0023 0.0422 0.0043

Average Sd Average Sd Average Sd Average Sd Average Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd Average Sd

and cellulose in mineral layer, but in SO and FT treatments herbicide plots had higher values for those constituents (Table 3). FT also had a larger increase in those constituents with herbicide, than SO (Table 3). 3.5. Potential drivers of soil respiration

0.10 0.01 1.9 0.3 18.1 1.7 23.7 5.1 5.6 1.5 0.1018 0.0015 0.0117 0.0025 0.0141 0.0033 0.0163 0.0042 0.0309 0.0057

Forest floor C:N had a positive effect on R15, and its inclusion increased the explained variation to 74%. Adding average plot summer soil moisture and temperature to the regression did not significantly explain any additional variation in R15. 4. Discussion

2

R15 increased with the amount of C retained (r = 0.71; Fig. 9). Stump plus coarse root plus forest floor organic matter explained most of the variation (64%) of the variation in C retained. R15 was also related to the quality of soil C. R15 had a negative correlation to the abundance of polysaccharide in both the forest floor and mineral soil, which in combination could explain 60% of the variation in R15 (p < 0.001; Eq. (1)).

R15 ¼ 15:1  30:5  Organic polysaccharide  90:9  Mineral polysaccharide

ð1Þ

4.1. Changes in dominant drivers of soil respiration with different levels of biomass removal Soil respiration in the UC and MN sites had peak rates of summer soil respiration of 3–4 lmol m2 s1 comparable to other boreal forest stands (Weber, 1985; Striegl and Wickland, 1998; Pumpanen et al., 2004). Soil respiration decreased in the biomass harvested plots compared to the UC and MN, indicating that CO2 from the decomposition of residue and forest floor was insufficient to compensate for the loss of root and rhizosphere respiration

188

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Table 2 P values for a series of orthogonal contrasts for R15 (lmol m2 s1 at 15 °C) soil temperature at 5 cm (°C), soil moisture integrated over the top 6 cm in harvest treatments (%) for all treatments (including uncut MN and UC) and for only harvest treatments including herbicide as a factor. Bolded values indicate p 6 0.10. R15 All treatments (P value) MN vs UC MN and UC vs SO, FT, FT + S and FT + B FT + B vs SO, FT and FT + S FT + S vs SO and FT FT vs SO Treatment

Temperature

Moisture

0.03 <0.001 0.004 0.14 0.83 <0.001

0.23 0.001 0.33 0.45 0.65 0.01

0.34 0.16 0.002 0.55 0.77 0.03

Harvest treatments and herbicide (P value) FT + B vs SO, FT and FT + S <0.001 FT + S vs SO and FT 0.05 FT vs SO 0.74 Treatment 0.001 Herb 0.03 Treatment:Herbicide 0.82

<0.001 0.45 0.45 0.003 0.62 0.67

0.004 0.51 0.75 0.02 0.01 0.46

(Striegl and Wickland, 1998; Pumpanen et al., 2004; Moroni et al., 2009). Among the biomass treatments, R15 was related primarily to substrate quantity (amount retained) and quality as opposed to physical environment conditions, thus the presence of a forest floor and an abundance of labile residues were keys to high rates of soil respiration. Within the FT + B treatment, CO2 production was limited because of the absence of forest floor and harvesting residue, consistent with low CO2 efflux observed from screefed plots (Mallik and Hu, 1997) and in studies where forest floor was removed (Pumpanen et al., 2004; Fleming et al., 2006). The lower polysaccharide abundance and higher lignin, aromatics or carboxylates in the FT + B mineral soil may reflect the removal of input of substrates from the forest floor. The FT + B treatment also experienced greater diurnal fluctuations in temperature due to the absence of an overlying insulating forest floor (Webster; unpublished data), which has been observed in other studies (Huhta et al., 1967; Keenan and Kimmins, 1993; Marra and Edmonds, 1996) and may decrease metabolic efficiency of microbial communities in the mineral soil (Schimel et al., 2007). The forest floor also reduces evaporation and thermal admittance, creating a mesic moisture regime and temperaturemoderated soil environment favourable for decomposition (Hendrickson et al., 1989; De Luca and Boisvenue, 2012) and creates habitats for arthropods and other decomposer organisms (Peng et al., 2008). Our estimate of 46–76% of soil respiration originating from the forest floor is in line with estimates of Pumpanen et al. (2004) who reported 69% of respiration was from the humus (forest floor) layer and Striegl and Wickland (1998) who reported 40% of CO2 was produced from near-surface sources. Similarly, Pietikäinen et al. (1999) found the highest respiration rates in the uppermost 10 cm of soil on unharvested sites. In the FT + S treatment, although the forest floor was present, soil respiration was constrained compared to FT and SO. This may have been due to large amounts of stump and coarse root biomass removed (46.3 Mg ha1; Kwiaton et al., 2014) or from other inadvertent impacts of the treatment. Stump removal has an impact on the forest floor and mineral soil, through mineralexposed divots left by stump and root removal, but also due to crushing of forest floor and fine residues from frequent passes of heavy machinery during stump removal. Although this impact was not quantified there were increases in the C:N in the forest floor and increases of polysaccharides and lignin, aromatics or carboxylates in the mineral soil which may reflect physical crushing and mixing of material in the near surface layers of soil. Soil compaction has been previously shown to affect site organic matter

content and porosity (Powers et al., 2005) and increase variability of pool size estimates (Huntington and Ryan, 1990). Furthermore, if care is not taken to shake stumps upon removal, forest floor and mineral soil may be inadvertently removed (Walmsley and Godbold, 2010). However this latter explanation does not appear to be an important factor at this site since on average stumping removed 1.2% and 0.8% of the forest floor and upper mineral soil C, respectively (Kwiaton et al., 2014). The impacts of stump and coarse root removal on soil respiration are likely to have impacts into the future. Walmsley and Godbold (2010) showed that stump removal and scarification resulted in reduced soil C stores 10 years after harvest and Strömgren et al. (2013) found that lowered soil C stocks following stump removal could persist 25 years after harvest. While greater CO2 effluxes were observed in the SO and FT than the FT + S, there were no clear differences between SO and FT treatments. This is in contrast to a previous study in which retention of residues, particularly fine residues expected in SO was shown to affect soil respiration (Moroni et al., 2009). Thus it is surprising that more dramatic differences between SO and FT were not observed. However, there may have been a trade-off in the quality of residues and environmental conditions with the SO having more optimal substrate chemistry for respiration, but FT having optimal environmental conditions for respiration. However, higher P was observed in the mineral soil of SO which may, in part, reflect the decomposition of P-rich needles and small twigs abundant on those plots, but that pre-dated or was too transient an effect for our sampling to capture. For example, Pumpanen et al. (2004) estimated that 23% of the total C pool in the aboveground logging residue was released during the first year after clear-cutting. The UC and MN plots had the highest soil respiration rates. The undisturbed forest floor in the UC and NM had a different substrate quality profile compared to each other and compared to the harvest treatments. The MN soil reflected a more nutrient rich and complex soil (higher C, N, P and lignin, aromatics or carboxylates) compared to the younger UC. The unharvested plots had more C, N in forest floor, fats and lipids in mineral soil and lower polysaccharides (both forest floor and mineral soil) contributing to higher heterotrophic Rs than the treated plots. Furthermore, an intact understorey of moss, herbs and low shrubs contribute to higher rates of understorey autotrophic Rs. Root respiration is an important component of soil respiration in uncut forests, contributing 35–47% of total soil respiration in boreal systems (Striegl and Wickland, 1998 [65–90 year old jack pine]; Pumpanen et al., 2004 [130 year old Scots pine – Norway spruce]; Gaumont-Guay et al., 2008 [125 year old black spruce]). In our harvested treatments, root respiration accounted for 5–28% of total respiration, comparable to the 5% value observed by Bond-Lamberty et al. (2004) in recently burned stands. Across all harvest and unharvested plots, R15 could be explained by forest floor and mineral soil polysaccharide abundance and forest floor C:N, indicators of the ease in which C can be degraded. R15 increased with decreases in recalcitrant complex polysaccharides, also observed in forested swamp peat profiles (Webster et al., 2014), but also increased with higher C:N ratios (more recalcitrant), suggesting there is enough N to allow decomposition of high C:N materials. Pietikäinen and Fritze (1995) also detected differences in humus structure among uncut, clearcut and burned coniferous forests. Thus the herbaceous litter, along with labile needle and bark residues serve as an important short-term substrate for microbial respiration, while coarser residues (larger branches and stumps) will more gradually decompose over the medium-term (5–15 + years) bridging a gap in substrate availability in time until the regenerating forest is old enough to produce substantial quantities of litter (Liski et al., 1998; Jandl et al., 2007).

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

189

4.2. Changes in net ecosystem exchange In terms of net exchange of C with the atmosphere, herbicide application reduced the amount of soil respiration by eliminating or diminishing root respiration (Raich and Tufekcioglu, 2000), but resulted in a net CO2 source due to absence of photosynthesis. The FT + B are also small sources of CO2, with low respiration rates offset by even lower rates of photosynthesis. This is consistent with other studies showing clearcuts to be CO2 sources and not changing to sinks until 20–30 years later (Striegl and Wickland, 1998; Liski et al., 1998; Zha et al., 2009; Pypker and Fredeen, 2002a,b). The understories of the UC and MN were large sources of CO2 because of large rates of respiration that were not countered by understorey photosynthesis. At the stand scale, however, the photosynthesis by trees (not quantified) would undoubtedly result in these stands becoming sinks of CO2 (Zha et al., 2009). Nonherbicided plots of FT + S, FT and SO were sinks of CO2 during August because moderate rates of soil respiration were offset by high rates of understory photosynthesis, although these sites would likely be CO2 sources during different times of the year. Understorey vegetation development increases soil respiration, as well as reducing the release of CO2 to the atmosphere through photosynthesis. The deciduous herbaceous litter also contributes a new source of labile soil OM which improves soil quality (Lister et al., 2004) and primes the decomposition process (Kuzyakov, 2010). This is reflected in the higher C:N in the non-herbicided forest floor indicating greater litter inputs. However, over the longer term, competition from the understorey may compromise the growth of trees and near-term benefits of not applying herbicide might be offset (Wagner et al., 1999).

4.3. Trade-offs between off-site C offsets and on-site ecosystem services Decomposition can be considered both good and bad. High rates of decomposition translate into higher mineralization rates of organic bound nutrients, such as N, and this nutrient recycling is essential in sustaining forest productivity. However, high rates of decomposition also result in greater emission of CO2, a greenhouse gas strongly implicated in climate change (De Luca and Boisvenue, 2012). Residues produced from forest harvesting that are left to decompose on site could be used as bioenergy sources to offset fossil fuel use (NRCan, 2010). Theoretically, there should be an optimum level of C retained to allow for sufficient decomposition to maintain nutrient cycling and sustain productivity while diverting ‘‘excess” residue into the bioenergy stream to offset fossil fuel consumption. A sigmoidal function represents the decrease in R15 with decreased on-site C retention (Fig. 9). A mirror image of this curve would indicate the CO2 that could be generated from the biomass when used for off-site for bioenergy production (Fig. 10). A greater amount of C removed from the site could result in more CO2 produced from bioenergy, reducing fossil fuel use. A key consideration in determining the amount of residue to be removed vs retained would be to assess the minimum level of soil respiration (i.e., CO2 production) required to sustain nutrient recycling. Using a seasonal integrated CO2 value and C:N of residues, we predicted the amount of mineralized N (Fig. 11), showing that the harvested treatments have a gross mineralization of only 50% of the UC (taking into account a baseline of mineralization from mineral soil, e.g., FT + B). The lower range of gross N mineralization of 50 mg N m2 d1 as observed within boreal black spruce forests of northwestern Ontario (Stottlemyer and Toczydlowski, 1999) converts to 142 kg N ha1 (over 283 days), close to that observed in the UC.

Fig. 7. Components of monthly average soil respiration (mineral soil, forest floor and root) (lmol CO2 m2 s1) for each of the harvest treatments.

Sites that are less fertile (as may occur on coarse-textured soils such as the jack pine woodlands), or sustain greater nutrient losses through intensive short rotation forestry may require a higher

190

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Fig. 8. Mean (±standard deviation) photosynthesis, respiration and net ecosystem exchange (by difference) for August 2012 (lmol CO2 m2 s1) for each of the harvest and herbicide treatments [forest floor removal (FT + B), stumped (FT + S), full-tree (FT) and tree-length (SO) harvest treatments for herbicided (h; grey bars) and non-herbicided (nh; white bars) plots and uncut control (UC) and mature natural (MN) stand]. Negative values indicate a net sink and positive values a net source of CO2.

minimum level of decomposition (i.e., closer to the maximum onsite respired CO2) to maintain future productivity, thus increasing the recommended amount of residue retained on site. In contrast, for sites that are more fertile (fine-textured soils) or managed on

longer rotations the minimum CO2 to ensure mineralization requirements may be much lower (i.e., at the point of inflection on the curve), suggesting that more residue could be removed from the site. A diagram, such as that presented in Fig. 10, provides a

191

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Table 3 P values for a series of orthogonal contrasts for forest floor and mineral soil chemistry (functional groups reflect relative abundance, refer to Table 1 and elements as%, and C:N unitless) for all treatments (including uncut MN and UC) and for only harvest treatments including herbicide as a factor. Bolded values indicate p 6 0.10. Polysaccharide

Lignins, aromatics or carboxylates

Fats and lipids (symmetric)

Fats and lipids (antisymmetric)

Cellulose

C

N

C:N

K

P

0.02 0.02

0.04 0.22

0.51 0.08

0.72 0.10

0.30 0.13

0.05 0.03

0.06 0.04

0.81 0.56

0.20 0.15

0.07 0.46

0.14 0.95 0.04

0.56 0.87 0.24

0.23 0.71 0.29

0.28 0.68 0.36

0.14 0.66 0.27

0.18 0.31 0.08

0.50 0.35 0.11

0.04 0.96 0.26

0.95 0.50 0.44

0.92 0.14 0.22

Harvest treatments and herbicide (P value) FT + B vs SO, FT and FT + S FT + S vs SO and FT 0.11 FT vs SO 0.92 Treatment 0.23 Herb 0.65 Treatment:Herbicide 0.63

0.67 0.91 0.89 0.78 0.40

0.31 0.75 0.54 0.69 0.44

0.37 0.72 0.60 0.76 0.46

0.23 0.72 0.42 0.72 0.48

0.27 0.40 0.37 0.79 0.85

0.52 0.38 0.53 0.82 0.91

0.09 0.96 0.20 0.02 0.24

0.93 0.38 0.65 0.99 0.53

0.84 0.03 0.07 0.80 0.89

Mineral soil All treatments (P value) MN vs UC FT + B vs SO, FT and FT + S FT + S vs SO and FT FT vs SO Treatment

0.26 0.003 0.06 0.21 0.02

0.59 0.58 0.87 0.29 0.13

0.66 0.30 1.00 0.45 0.21

0.09 0.26 0.64 0.55 0.19

0.13 0.60 0.82 0.94 0.67

0.26 0.73 0.69 0.95 0.87

0.05 0.29 0.98 0.91 0.37

0.04 0.01 0.02 0.14 0.01

0.74 0.18 0.33 0.05 0.23

0.003 0.04 0.15 0.01 0.22 0.20

0.14 0.15 0.71 0.37 0.31 0.01 0.08

0.05 0.24 0.97 0.30 0.35 0.01 0.06

0.06 0.60 0.11 0.35 0.15 0.03 0.09

0.34 0.68 0.90 0.74 0.42 0.45

0.66 0.62 0.94 0.92 0.70 0.48

0.05 0.96 0.80 0.20 0.25 0.83

0.03 0.05 0.19 0.04 0.41 0.64

0.20 0.34 0.07 0.14 0.63 0.61

0.02 0.04

0.02 0.04

0.04 0.05

Forest floor All treatments (P value) MN vs UC MN and UC vs SO, FT, FT + S and FT + B FT + B vs SO, FT and FT + S FT + S vs SO and FT FT vs SO Treatment

0.03 0.01 0.01 0.18 0.004

Harvest treatments and herbicide (P value) FT + B vs SO, FT and FT + S 0.002 FT + S vs SO and FT 0.002 FT vs SO 0.06 Treatment 0.002 Herb 0.52 Treatment:Herbicide 0.34 FT + B vs SO, FT and FT + S: Herb FT + S vs SO and FT:Herb FT vs SO:Herb

Fig. 9. Relationship between carbon retained on plot (Mg C ha1) and average non-herbicided R15 (lmol CO2 m2 s1), including fit regression model and equation. Treatments names are truncated to F, S, B, T, and C for FT, FT + S, FT + B, SO, and UC respectively.

conceptual model for comparing trade-offs in wood removal for bioenergy to offset fossil fuel emissions vs wood retention to promote nutrient cycling. An approach to estimating the economic

feasibility of harvesting a site is to determine how much residue should be retained, estimate the total stock and then calculate the difference to determine the amount that can be removed.

192

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Fig. 10. Conceptual graph of the trade-off of CO2 for nutrient mineralization on-site (solid line) and CO2 from residue taken off-site (dotted line). Shaded area indicates the optimal range of C retained on site depending on site fertility.

4.4. Best management practices and future work A precautionary approach would err on retaining residues at levels higher than the theoretical optimum in Fig. 10. However, completely excluding infertile sites, 2nd rotation and under stocked sites from bioenergy production is not always necessary since such practices may not be detrimental to soil productivity as long as sufficient residue remains and specific sites may benefit from stand improvement through harvest and silviculture. However our results and those of others (e.g., Lamers et al., 2013) suggest that the type of material taken vs retained (including C quality and nutrient content) is an important factor that is not taken into account in the conceptual model presented in Fig. 10. Since biomass harvesting impacts are highly variable across different forest and soil types, there is a need for harvesting guidelines aimed at encouraging retention of ecologically important

structural attributes (Thiffault et al., 2011; Littlefield and Keeton, 2012). Fine residues have low energy content for bioenergy, but are high in nutrients that could improve soil productivity. Best management practices could focus on leaving fine residues on site, as well as residues having mixed size classes of stems. A mixture of materials would enhance soil physical properties, maintain biological refugia (e.g., for saproxylic fungi, insects and higher trophic levels) (Bengtsson et al., 1998) and ensure sources of labile materials and slower degrading C residues that would result in continuous OM inputs and nutrient cycling over the short- to mediumterm (Marshall, 2000). Fertilization is often proposed as a solution to short-term reductions in soil fertility that may occur with greater residue removal (Eisenbies et al., 2009). However, although fertilizers and other amendments such as ash would add nutrients, their positive benefits are likely to be short-term and not maintain longer-term nutrient cycling or produce the physical structure necessary to accommodate biodiversity. We established that the quantity and quality of forest floor and residue are important to Rs, NEE and, by inference, to nutrient cycling (Fig. 11). Further research is needed, however, to better understand how different combinations of residue types, their quality, and size classes affect Rs, NEE and soil nutrition through stand development. Key to this understanding will be to understand connections between the decomposition of substrates and diversity and function of soil micro and macro fauna.

5. Conclusion A holistic view of the ecological consequences of energy forestry recognizes that C offsets off-site are achieved at the potential expense of future forest sustainability. We showed that intensification of biomass removal and regeneration practices in a low quality, mid-rotation stand of jack pine affected magnitude, source and physical vs chemical drivers of Rs and NEE. The presence of a forest floor and absence of complex polysaccharides were essential to ensuring high levels of Rs that drive nutrient mineralization. Therefore, Rs may be a useful indicator to determine the optimal level of biomass removal and a metric for comparing trade-offs between nutrient cycling on-site and C offsets to fossil fuel off-site.

Fig. 11. Relationship between summer integrated Rs (lmol m2 from May 1 to October 31, 2012) and N mineralization (kg ha1) calculated from C:N of residues. New residues include needles, twigs and branches left from harvesting and CWD is coarse woody debris.

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Acknowledgments This research was funded by Canadian Forest Service A base funds and EcoEii funding to KLW, PWH, and RLF. We acknowledge Johanna Curry (Canadian Forest Service) and Ravi Kanipayor (Ontario Ministry of Natural Resources) for soil analyses, and Rob Irwin, Mike Adams, Mark Primavera and Gord Brand (Canadian Forest Service) for assistance in the field.

References Bengtsson, J., Lundkvist, H., Saetre, P., Sohlenius, B., Solbreck, B., 1998. Effects of organic matter removal on the soil food web: forestry practices meet ecological theory. Appl. Soil Ecol. 9, 137–143. Bond-Lamberty, B., Wang, C., Gower, S.T., 2004. Contribution of root respiration to soil surface CO2 flux in a boreal black spruce chronosequence. Tree Physio. 24, 1387–1395. De Luca, T.H., Boisvenue, C., 2012. Boreal forest soil carbon: distribution, function and modelling. Forestry 85, 161–184. Delta T Devices, Ltd., 1999. Theta Probe Soil Moisture Sensor Type ML2x User Manual ML2x-UM-1.21. Eisenbies, M.H., Vance, E.D., Aust, W.M., Seiler, J.R., 2009. Intensive utilization of harvest residues in southern pine plantations: quantities available and implications for nutrient budgets and sustainable site productivity. BioEnergy Res. 2, 90–98. Environment Canada, 2014. National Climate Data and Information Archive. (accessed 26 June 2013). Ewel, K.C., Cropper Jr., W.P., Gholz, H.L., 1987. Soil CO2 evolution in Florida slash pine plantations. I. Changes through time. Can. J. For. Res. 17, 325–329. Fleming, R.L., Black, T.A., Adams, R.S., Stathers, R.J., 1998. Silvicultural treatments, microclimatic conditions and seedling response in Southern interior clearcuts. Can. J. Soil Sci. 78, 115–126. Fleming, R.L., Laporte, M.F., Hogan, G.D., Hazlett, P.W., 2006. Effects of harvesting and soil disturbance on soil CO2 efflux from a jack pine forest. Can. J. For. Res. 36, 589–600. Fleming, R.L., Leblanc, J.D., Hazlett, P.W., Weldon, T., Mossa, D.S., 2014. Effects of biomass harvest intensity and soil disturbance on jack pine stand productivity 15 year results. Can. J. For. Res. 44, 1566–1574. Foster, N.W., 1996. Sustainability of nutrient supply to short-rotation jack pine forests for energy. In: Proceedings of the Canadian Energy Plantation Workshop, May 2–4, 1995, Gananoque, Ontario. Compiled by J. Karau. Natural Resources Canada, Canadian Forest Service, Headquarters, Ottawa, pp. 99–106. Fox, T.R., 2000. Sustained productivity in intensively managed forest plantations. For. Ecol. Manage. 138, 187–202. Gaumont-Guay, D., Black, T.A., Barr, A.G., Jassal, R.S., Nesic, Z., 2008. Biophysical controls on rhizospheric and heterotrophic components of soil respiration in a boreal black spruce stand. Tree Physiol. 28, 161–171. Gordon, A.M., Schlentner, R.E., Cleve, K.V., 1987. Seasonal patterns of soil respiration and CO2 evolution following harvesting in the white spruce forests of interior Alaska. Can. J. For. Res. 17, 304–310. Grigal, D.F., 2000. Effects of extensive forest management on soil productivity. For. Ecol. Manage. 138, 167–185. Hendrickson, O.Q., Chatarpaul, L., Burgess, D., 1989. Nutrient cycling following whole-tree and conventional harvest in northern mixed forest. Can. J. For. Res. 19, 725–735. Huhta, V., Karppinen, E., Nurminen, M., Valpas, A., 1967. Effects of silvicultural practices upon arthropod, annelid and nematode populations in coniferous forest soil. Ann. Zool. Fennici 4, 87–143. Huntington, T.G., Ryan, D.F., 1990. Whole-tree-harvesting effects on soil nitrogen and carbon. For. Ecol. Manage. 31, 193–204. Jandl, R., Lindner, M., Vesterdal, L., Bauwens, B., Baritz, R., Hagedorn, F., Johnson, D. W., Minkkinen, K., Byme, K.A., 2007. How strongly can forest management influence soil carbon sequestration? Geoderma 137, 253–268. Keenan, R.J., Kimmins, J.P., 1993. The ecological effects of clear-cutting. Environ. Rev. 1, 121–144. Kuzyakov, Y., 2010. Priming effects: interactions between living and dead organic matter. Soil Biol. Biochem. 42, 1363–1371. Kwiaton, M., Hazlett, P., Morris, D., Webster, K., Venier, L., Aubin, I., 2014. Island Lake Biomass Harvest Research and Demonstration Area: Establishment Report. NRCan-CFS Info. Rep. No. GLC-X-11. Lamers, P., Thiffault, E., Paré, D., Junginger, M., 2013. Feedstock specific environmental risk levels related to biomass extraction for energy from boreal and temperate forests. Biomass Bioenergy 55, 212–226. Lattimore, B., Smith, C.T., Titus, B.D., Stupak, I., Egnell, G., 2009. Environmental factors in wood fuel production: opportunities, risks, and criteria and indicators for sustainable practices. Biomass Bioenergy 33, 1321–1342. Law, K.N., Valade, J.L., 1994. Status of the utilization of jack pine (Pinus banksiana) in the pulp and paper industry. Can. J. For. Res. 24, 2078–2084. Liski, J., Ilvesniemi, H., Mäkelä, A., Starr, M., 1998. Model analysis of the effects of soil age, fires and harvesting on the carbon storage of boreal forest soils. Eur. J. Soil Sci. 49, 407–416.

193

Lister, T.W., Burger, J.A., Patterson, S.C., 2004. Role of vegetation in mitigating soil quality impacted by forest harvesting. Soil Sci. Soc. Am. J. 68, 263–271. Littlefield, C.E., Keeton, W.S., 2012. Bioenergy harvesting impacts on ecologically important stand structure and habitat characteristics. Ecol. Appl. 22, 1892– 1909. Livingston, G.P., Hutchinson, G.L., 1995. Enclosure-based measurement of trace gas exchange: applications and sources of error. In: Matson, P.A., Harriss, R.C. (Eds.), Biogenic Trace Gases: Measuring Emissions from Soil and Water. Blackwell Scientific, Oxford, UK, pp. 14–51. Lowe, J.J., Power, K., Gray, S.L., 1994. Canada’s forest inventory 1991. Can For. Serv. Petawawa Nat. For. Res. Inst. Inf. Rep. No. PI-X-115. Mallik, A.U., Hu, D., 1997. Soil respiration following site preparation treatments in boreal mixedwood forest. For. Ecol. Manage. 97, 265–275. Marra, J.L., Edmonds, R.L., 1996. Coarse woody debris and soil respiration in a clearcut on the Olympic Peninsula, Washington, USA. Can. J. For. Res. 26, 1337– 1345. Marshall, V.G., 2000. Impacts of forest harvesting on biological processes in northern forest soils. For. Ecol. Manage. 133, 43–60. McNabb, D.H., Startsev, A.D., Nguyen, H., 2001. Soil wetness and traffic level effects on bulk density and air-filled porosity of compacted boreal forest soils. Soil Sci. Am. J. 65, 1238–1247. Moroni, M.T., Carter, P.Q., Ryan, D.A., 2009. Harvesting and slash piling affects soil respiration, soil temperature, and soil moisture regimes in Newfoundland boreal forests. Can. J. Soil Sci. 89, 343–355. Nambiar, E.K., 1996. Sustained productivity of forests is a continuing challenge to soil science. Soil Sci. Soc. Am. J. 60, 1629–1642. NRCan, 2010. Is Forest Bioenergy Good for the Environment. Natural Resources Canada Canadian Forest Service Science-Policy Notes. December 2010. ISBN 978-1-100-17372-6. Cat. No. Fo93-1/6-2010E-PDF. Peng, Y., Thomas, S.C., Tian, D., 2008. Forest management and soil respiration: implications for carbon sequestration. Environ. Rev. 16, 93–111. Pietikäinen, J., Fritze, H., 1995. Clear-cutting and prescribed burning in coniferous forest: comparison of effects on soil fungal and total microbial biomass, respiration activity and nitrification. Soil Biol. Biochem. 27, 101–109. Pietikäinen, J., Vaijärvi, E., Ilvesniemi, H., Fritze, H., Westman, C.J., 1999. Carbon storage of microbes and roots and the flux of CO2 across a moisture gradient. Can. J. For. Res. 29, 1197–1203. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., R Core Team, 2014. nlme: Linear and Nonlinear Mixed Effects Models. R Package Version 3.1-117. . Powers, R.F., Scott, D.A., Sanchez, F.G., Voldseth, R.A., Page-Dumroese, D., Elioff, J.D., Stone, D.M., 2005. The North American long-term soil productivity experiment: findings from the first decade of research. For. Ecol. Manage. 220, 31–50. Proe, M.F., Griffiths, J.H., McKay, H.M., 2001. Effect of whole-tree harvesting on microclimate during establishment of second rotation forestry. Agr. For. Met. 110, 141–154. Pumpanen, J., Westman, C.J., Ilvesniemi, H., 2004. Soil CO2 efflux from a podzolic forest soil before and after forest clear-cutting and site preparation. Boreal Environ. Res. 9, 199–212. Pypker, T.G., Fredeen, A.L., 2002a. The growing season carbon balance of a subboreal clearcut 5 years after harvesting using two independent approaches to measure ecosystem CO2 flux. Can. J. For. Res. 32, 852–862. Pypker, T.G., Fredeen, A.L., 2002b. Ecosystem CO2 flux over two growing seasons for a sub-Boreal clearcut 5 and 6 years after harvest. Agr. For. Meteorol. 114, 15–30. Raich, J.W., Tufekcioglu, A., 2000. Vegetation and soil respiration: correlations and controls. Biogeochemistry 48, 71–90. Roach, J., Berch, S.M., 2014. A Compilation of Forest Biomass Harvesting and Related Policy in Canada. Prov. B.C., Victoria, B.C. Tech. Rep. 081. (accessed 27 June 2014). Roberts, M.R., 2007. A conceptual model to characterize disturbance severity in forest harvests. For. Ecol. Manage. 242, 58–64. Schimel, J., Balser, T.C., Wallenstein, M., 2007. Microbial stress-response physiology and its implications for ecosystem function. Ecology 88, 1386–1394. Soil Classification Working Group, 1998. The Canadian System of Soil Classification, third ed. Agriculture Canada. (accessed July 18, 2014). Stottlemyer, R., Toczydlowski, D., 1999. Nitrogen mineralization in a mature boreal forest, Isle Royal, Michigan. J. Environ. Qual. 28, 709–720. Striegl, R.G., Wickland, K.P., 1998. Effects of a clear-cut harvest on soil respiration in a jack pine – lichen woodland. Can. J. For. Res. 28, 534–539. Strömgren, M., Egnell, G., Olsson, B.A., 2013. Carbon stocks in four forest stands in Sweden 25 years after harvesting of slash and stumps. For. Ecol. Manage. 290, 59–66. Systat Software, Inc., 2011. SigmaPlot, Version 12 for Windows. Chicago, IL. Ter-Mikaelian, M.T., Colombo, S.J., Chen, J., 2015. The burning question: does forest bioenergy reduce carbon emissions? A review of common misconceptions about forest carbon accounting. J. For. 113, 57–68. Thiffault, E., Hannam, K.D., Paré, D., Titus, B.D., Hazlett, P.W., Maynard, D.G., Brais, S., 2011. Effects of forest biomass harvesting on soil productivity in boreal and temperate forests – a review. Environ. Rev. 19, 278–309. Wagner, R.G., Mohammed, G.H., Noland, T.L., 1999. Critical period of interspecific competition for northern conifers associated with herbaceous vegetation. Can. J. For. Res. 29, 890–897. Wall, A., 2012. Risk analysis of effects of whole-tree harvesting on site productivity. For. Ecol. Manage. 282, 175–184.

194

K.L. Webster et al. / Forest Ecology and Management 360 (2016) 181–194

Walmsley, J.D., Godbold, D.L., 2010. Stump harvesting for bioenergy – a review of the environmental impacts. Forestry 83, 17–38. Weber, M.G., 1985. Forest soil respiration in eastern Ontario jack pine ecosystem. Can. J. For. Res. 15, 1069–1075. Webster, K.L., Creed, I.F., Malakoff, T., Delaney, K., 2014. Potential vulnerability of deep carbon deposits of forested swamps to drought. Soil Soc. Am. J. 78, 1097– 1107.

Williamson, J.R., Neilsen, W.A., 2000. The influence of forest site on rate and extent of soil compaction and profile disturbance of skid trails during ground-based harvesting. Can. J. For. Res. 30, 1196–1205. Zha, T., Barr, A.G., Black, T., McCaughey, J.H., Bhatti, J., Hawthorne, I., Krishnan, P., Kidston, J., Saigusa, N., Shashkov, A., Nesic, Z., 2009. Carbon sequestration in boreal jack pine stands following harvesting. Global Change Biol. 15, 1475– 1487.