A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production

Forest Ecology and Management xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

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

A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production Gustaf Egnell Department of Forest Ecology and Management, Swedish University of Agricultural Sciences, SE-90183 Umeå, Sweden

a r t i c l e

i n f o

Article history: Received 7 June 2016 Received in revised form 23 August 2016 Accepted 15 September 2016 Available online xxxx Keywords: Whole-tree harvest Full-tree harvest Sustainable forest yield Stump lifting Bioenergy Biomass Site productivity

a b s t r a c t In the Nordic countries, emerging markets for renewable energy have resulted in increased harvest intensity, i.e., branches, tops, and stumps are now also harvested. This increased harvest intensity changes site conditions in a way that may impact future forest production. In this review published results are compiled from long-term field experiments in the Nordic countries. The objectives are to identify general patterns or inconsistences, to identify possible causes, and to discuss the practical implications of the results. I summarize 16 publications where short to medium-term forest production data were presented from 72 experimental sites. Data on growth of the subsequent stand following slash harvest in final felling indicate a moderate negative growth effect in Norway spruce, whereas growth in Scots pine appears unaffected, as compared to stem-only harvested control plots. Spruce data also showed a trend suggesting that poorer sites are more sensitive. Stump harvest in final felling did not have a negative effect on growth – rather the opposite – particularly on poor Scots pine sites. Trends in the data suggest that the positive growth effect in pine is stronger on poorer sites at higher latitudes. Slash harvest in thinnings resulted in more consistent growth reductions in the residual stand in both pine and spruce. There was a weak trend suggesting that poorer spruce sites are more susceptible. Seedling survival rates or stem numbers following slash and stump harvest were either unaffected or positively affected by the treatments – particularly by stump harvest. Trends in data suggest a stronger positive effect on more fertile sites. High relative survival or stem numbers coincided with high relative growth. Thus, survival rates or stem number may partly explain the lack of consistency in growth responses in field experiments. Management of natural regeneration in the experiments is discussed as potentially critical. Both short- and medium-term growth responses have been reported in individual studies. It is therefore recommended that a final evaluation should be based on longer-term data. The recommended next step is to combine all available data into a formal meta-analysis. Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction The European Union has set high targets for the proportion of renewable energy sources (Renewable Energy Directive, 2009/28/EC), which have created a market for biomass energy in Europe. These targets were driven mostly by climate change concerns. However, the shift towards renewable energy had already started in the Nordic countries after the oil crises in the 1970s, driven at that time by security of supply and employment concerns. In Finland and Sweden, where most of the biomass originates from forests, policies to support a development towards more renewables have been crucial for the development (Ericsson et al., 2004). Initially this resulted in better use of industrial residues, not the least as process energy in the industry itself. Thereafter the growth has E-mail address: [email protected]

been on the power and heat market, where biomass-based combined heat and power plants connected to district heating networks have increased the market for biomass further. As a result of this market growth the demand side became bigger than what could be supplied by industrial residues, therefore primary residues available following forest harvest operations were targeted. This includes logging residues such as slash, small-diameter trees, non-industrial wood due to damages or species, and stumps. The idea of using logging residues raised questions within the scientific community already in the 1970s. The prime questions at that time were potential impact of increased harvest intensity on site and thereby stand productivity, and whether long-term site and stand productivity could be maintained. This was particularly the case for harvest of the nutrient rich slash, where a moderate increase in biomass removal coincides with a substantial increase in nutrient loss from the site (Mälkönen, 1976; Nykvist and

http://dx.doi.org/10.1016/j.foreco.2016.09.019 0378-1127/Ó 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

2

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

Rosén, 1985). Therefore a number of long-term field experiments were established already in the 1970s, and new ones have been added over time. These experiments have now, from a longrotation forestry perspective, generated short- to medium-term data (with long-term here referring to one rotation period or more). Published results from some of these experiments were used in a review by Thiffault et al. (2011), where it was concluded that: ‘‘There are no consistent, unequivocal and universal effects of forest biomass harvesting on soil productivity. However, climate and microclimate, mineral soil texture and organic C content, the capacity of the soil to provide base cations and phosphorus, and tree species autecology appear to be critical determinants of site sensitivity to biomass harvesting”. Published data from some of the experiments were also included as a part of the global data set used in a meta-analysis by Achat et al. (2015), with results suggesting a tree growth reduction by 3–7% in the short- to medium-term when slash was harvested in final felling. Most studies on slash harvest analyse productivity effects by analysing the growth of the subsequent stand or in the case of thinnings, growth of the residual stand. Since stand productivity is a function of site productivity + silvicultural practices + genetics + random events such as frost, browsing, pests, storm and snow events, it is difficult to sort out effects on site productivity based on these kind of data. Furthermore, management of the experiments over time could be important for the outcome. E.g. if an increased harvest intensity stimulates natural regeneration it will be important how that regeneration is managed over time. If it is included in growth estimates they may fall out high, but if only growth of planted seedlings is included in the analyses, management of the natural regeneration becomes critical. If the natural regeneration was removed late it may have hampered establishment and growth of the planted seedlings. Overall there is much more than just the direct impact of additional nutrient withdrawal that may have an impact on growth of the subsequent crop. This adds to possible reasons why Thiffault et al. (2011) failed to find an unequivocal and universal response pattern in their review. Another could be that there is no such unambiguous treatment response, but rather site- or species-specific responses. For slash harvest in thinnings there is less to consider when analysing the data, but important factors are thinnings strength and standing stock after thinning. Both are valid covariates in an analysis of growth effects following slash harvest in thinnings. This could also change over time due to self thinning, e.g. through snow, wind, pest- and disease-caused mortality, that may not be linked directly to the treatments, but will have an impact on stand production and the interpretation of the results. This review is based on stand productivity response data following slash and stump harvest in final felling and slash harvest in thinnings from long-term field experiments in the Nordic countries published in the peer-reviewed literature. In addition, some more information on site characteristics and metadata has been gathered through professional networks. Metadata are used in the discussion on possible explanations for altering treatment effects in the experiments. The aim has been to identify response patterns. In the absence of such patterns possible site and species-specific responses or response differences due to experimental design, management, and random events are discussed. The results are compared and discussed together with published results from similar studies from other countries. Finally, the results are discussed in relation to practical forest operations with suggested practical implications. The review focuses on forest production and does not consider other possible effects of increased harvest intensity on other ecosystem services and biodiversity or possible trade-offs between them.

2. Material and methods The figures presented in this review are based on data from long-term field experiments in the Nordic countries published in the peer-reviewed literature. Data includes short- to mediumterm data (5–34 years) on seedling survival and growth of the subsequent stand following slash and/or stump harvest in final felling and growth of the residual stand following slash harvest in thinnings. In all experiments but one, stem-only harvest is used as a reference. The exception is Karlsson and Tamminen (2013), where stump, slash, and stem harvested plots were compared with slash and stem harvested plots. In total 16 studies were found presenting data from 72 experimental sites. Data from one site includes stand growth in the residual stand following slash harvest in two thinnings (Helmisaari et al., 2011) and for the subsequent stand following slash harvest in final felling (Tamminen and Saarsalmi, 2013). Slash is harvested on all sites, whereas stump harvest is restricted to 13 sites. Presented growth data varied between studies including volume (most studies), basal area (Egnell, 2011; Egnell and Leijon, 1999), and carbon in tree biomass (Jurevics et al. 2016). In general each treatment from an experimental site is only presented once in the figures. In the case where data have been collected multiple times only the most recent data have been used in the figures with some exceptions. Data from the 22 experimental sites with slash harvest in thinnings in Helmisaari et al. (2011) are presented twice in the figures since the publication reports growth data between year 0–10 and between year 11–20. The latter follows a second thinning (10 years after the first thinning) with slash harvest on 11 out of the 22 sites. Studies reporting data for both the planted tree species and the planted tree species together with natural regeneration are also represented twice in the presented figures (Karlsson and Tamminen, 2013; Tamminen and Saarsalmi, 2013; Wall and Hytonen, 2011). In those where data are presented twice in a figure this is indicated with a dashed arrow connecting the two data points. The individual experiments are described briefly in Table 1. To facilitate the graphical presentation, reported responses in the studies were normalized by dividing treatment response values with the corresponding value for stem-only harvested control plots (with the exception of the stump harvest study by Karlsson and Tamminen (2013), where slash was harvested also on control plots). Thus, control plots have a normalized value of 1 in all experiments represented by the dashed horizontal line in the presented figures. These values were then plotted over common site characteristics presented in the studies, e.g., site index and latitude. These plots together with metadata from the experiments were used as a basis for a discussion on consistencies and inconsistencies in the results and possible reasons for this. As an objective support to that discussion Spearmans rank-order correlation coefficient was calculated for the plot variables by means of Minitab 17 (2010). No further statistical analyses of the data set have been performed. 3. Results and discussion 3.1. Slash harvest in final felling Slash harvest effect data in final felling was collected from 10 different publications covering 31 long-term experiments in Finland and Sweden. From some experiments presented data includes data with and without natural regeneration (Tamminen and Saarsalmi, 2013; Wall and Hytonen, 2011), thus, the same treatment on a site is in some cases represented by more than one data point. Furthermore, presented data following slash harvest also includes a treatment where both slash and stumps were harvested (Egnell, 2016; Jurevics et al., 2016).

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

y y

x x x

x

s, m x x x x x x

x x x x x

m.p m.p h m.p p m.p x x x x

x x t t 10 19–25 4 5 1990 1972–1977

2 1 1 4 8 1 2 1 6 10 1 4 4 22

2 4

Egnell and Leijon (1999) Egnell and Valinger (2003) Egnell (2011) Egnell (2016) Jurevics et al. (2016) Karlsson and Tamminen (2013) Saarsalmi et al. (2010) Sikström (2004) Smolander et al (2015) Tamminen and Saarsalmi (2013) Wall and Hytonen (2011) Egnell and Leijon (1997) Egnell and Ulvcrona (2015) Helmisaari et al. (2011)

Nord-Larsen (2002) Tveite and Hanssen (2013)

S S S S S F F S F F F S S F, S, N D N

1974, 1975 1975 1976 1982, 1983 1981–1983 1977 1978, 1979 1995 2002, 2003 1999, 2000 1973 1982–1985 1983–1985 1977–1986

4 4 4 4 1 8 12, 24 2 3 3–9 10 4 3–4 2–4

12, 15 24 21 24–27 31–34 34 22, 24 5, 6 10 10 30 10 18–21 10, 20

f.f. f.f. f.f. f.f. f.f. f.f. f.f. f.f. f.f. f.f. f.f. t t t

x x x x x x x x x x x x x x

x

x x x

x x x x

m.p x x

x

x x x x x x x x x x

y y/n y

y y/n

y y/n y

Pre-commercial thinning Supplementary planting Planting Site preparation Scots pine Stumps

Slash & stumps Slash

Norway spruce

Additional treatments Species Treatments

Harvest form Observation period (years) No of replicates Established year Country No of sites Publication

Table 1 Publications from which forest production data have been extracted together with brief information about the long-term field experiments they are based on. Countries are abbreviated F, S, N, and D, which denotes Finland, Sweden, Norway, and Denmark, respectively. Harvest forms are final felling (f.f.) and thinning (t). Treatments show the additional biomass harvested in the experiments and species shows the tree species under study. Additional treatments give information relevant for the interpretation of the results, where site preparation methods reported are scalping (s), harrowing (h), manual patch (m.p), mounding (m), and ploughing (p). y = yes, n = no, y/n = yes or no depending on site in that particular study. Open cells under additional treatments mean that the information is not provided in the publications.

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

3

3.1.1. Regeneration success following slash harvest in final felling Data on number of stems following slash harvest show a positive trend over site index in Norway spruce (q = 0.508, p = 0.045), with higher stem numbers following slash harvest (ratio > 1) in 11 out of 16 cases (mean ratio = 1,08). No such trend was found Scots pine (q = 0.061, p = 0.887), with higher stem numbers in 5 out of 8 cases (mean ratio = 1.11) (Fig. 1). For the individual spruce studies, statistically significant treatment effects are reported from six experiments, with two showing higher and four lower stem numbers following slash harvest (Smolander et al., 2015; Tamminen and Saarsalmi, 2013). For all four sites with lower stem numbers, only the planted tree species were accounted for in the data, whereas in one of the two sites with higher stem numbers, also naturally regenerated trees of other species were included in the data. It should be noted that in all four experiments where data also included natural regeneration, the stem numbers were higher following slash harvest despite the fact that most of the deciduous natural regeneration was removed 4–6 years after harvest in tree out of the four experiment (Tamminen and Saarsalmi, 2013). The second site with a significantly higher spruce stem number following slash harvest is site 738 in Tamminen and Saarsalmi (2013) with a low, and according to the authors, not even satisfactory survival rate of the planted spruce seedlings in all treatments. This low survival rate was explained by a high suppression of planted seedlings from naturally regenerated birch and other deciduous trees. In another study from Finland, not represented in Fig. 1 (no site index information provided), slash harvest resulted in significantly fewer spruce stems 30 years after harvest (Wall and Hytonen, 2011). In pine, statistically significant treatment effects were reported from two experiments, both with a higher 15-year survival rate for the planted pine seedlings following slash harvest (Egnell and Leijon, 1999). As with the spruce experiments, one of the pine experiments with a significant positive effect of slash harvest was an experiment with overall low and unsatisfactory survival rates suggesting that the high mortality rate was not primarily due to the abundance or absence of slash. Although seedling survival or stem numbers have been affected in individual experiments the total picture is that there is no consistent effect of slash harvest. Considering that natural regeneration has been removed at least once in a majority of the experiments, there may be a trend towards more natural regeneration following slash harvest as a result of more exposed soil. But possibly more important for the input from natural regeneration is the site preparation applied in the experiments and the resulting area of suitable mineral-soil seedbeds that are exposed. Support for this comes from a study showing that mounding had a significant and positive effect on natural regeneration of birch (Betula pubescens Ehrh. and B. pendula Roth), Scots pine, and Norway spruce, with a three times higher stem density as compared to control plots (Karlsson et al., 2002). The study also reported more naturally regenerated birch seedlings following slash harvest. If slash harvest has favoured natural regeneration in the experiments, the timing of any pre-commercial thinning will be critical for survival and performance of the planted seedlings. In addition, the exclusion or inclusion of natural regeneration in the analysed data will be critical for the interpretation of the results. Similar results as reported here have been presented from the North American long-term site productivity study (LTSP, Powers, 2006), with no effects of slash harvest on seedling survival for planted black spruce (Picea mariana (Mill.) B.S.P.) (Morris et al., 2014) and jack pine (Pinus banksiana Lamb.) (Fleming et al., 2014) seedlings over a number of different soil types after 15 years. For the jack pine sites Fleming et al. (2014) reported that slash harvest without subsequent scarification reduced survival of planted pines and input from natural regeneration. Thus, site

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

4

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

Fig. 1. Relative survival rates/stem no for the subsequent tree crop planted following slash harvest (SLH) in relation to stem-only harvest (SOH, dashed horizontal line) plotted over site index. Norway spruce (left, filled) and Scots pine (right, open). Squares indicate data from plots where the stumps were also harvested and triangles data where natural regeneration was also included. Arrows connect data from the same experimental site and treatment with or without natural regeneration included. Data from Egnell (2016), Egnell and Leijon (1999), Smolander et al. (2015), and Tamminen and Saarsalmi (2013).

preparation seems to be more critical than slash removal for seedling survival and establishment of natural regeneration. This is in line with experiences from regeneration research in northern Europe (Nilsson et al., 2010). Large analyses including data from more than 40 LTSP sites including different tree species, soil types, and site types, also report minor and inconsistent effects of slash harvest on seedling survival (Fleming et al., 2006; Ponder et al., 2012). The full LTSP design (i.e., split plots with herbicide and no herbicide control) is ideally suited for evaluating natural regeneration effects although naturals have been removed in some of the experiments to promote growth of the planted seedlings. 3.1.2. Forest production following slash harvest in final felling Data on forest production in the subsequent Norway spruce stand following slash harvest in final felling show a positive trend over site index (q = 0.491, p = 0.045), but no trend over latitude (q = 0.014, p = 0.943). With the stump- and slash-harvested experiments excluded from the data the trend becomes even stronger (q = 0.641, p = 0.003). In Scots pine no trend could be shown over site index (q = -0.281, p = 0.260), but there was a weak positive trend over latitude (q = 0.403, p = 0.098), suggesting that slash harvest could be positive for growth in hash climates (Fig. 2).

However, the latitude trend in Scots pine was strongly supported by the slash- and stump-harvested plots and q was reduced to 0.284 (p = 0.371) with that data excluded. Data from Norway spruce experiments support the suggestion that poor sites could be more vulnerable to more intense harvest practices as suggested by guidelines (Stupak et al., 2007), whereas Scots pine data do not. The ratios show reduced production following slash harvest (ratio < 1) in 18 cases out of 28 in spruce (mean ratio = 0.96) and in 8 cases out of 18 in pine (mean ratio = 1.06). Based on this, the trend suggests that growth is negatively affected by slash harvest in spruce and positively affected in pine. For the individual spruce studies, statistically significant reductions in growth following slash harvest were reported from four experiments (Egnell, 2011; Egnell and Leijon, 1999; Smolander et al., 2015; Wall and Hytonen, 2011) and a significant increase during the first ten years was reported from one experiment (Tamminen and Saarsalmi, 2013). The experiment (737) with a significant increase in growth was based on 10-year data that included growth of both the planted tree species and natural regeneration of other tree species. They also analysed the data with only growth of the planted tree species included, and then failed to show a treatment effect. The authors made a note that planted

Fig. 2. Relative growth (volume, basal area or C-content in biomass) for the subsequent tree crop during 5–34 years following slash harvest (SLH) in relation to stem-only harvest (SOH, dashed horizontal line) plotted over site index and latitude. Left column Norway spruce (filled) and right column Scots pine (open). Small symbols indicates a growth period 610 years and large symbols a growth period >10 years. Squares indicate data from plots where also the stumps were harvested and triangles indicates data where also natural regeneration was included. Arrows connect data from the same experimental site and treatment with or without natural regeneration included. Data from Egnell (2016, 2011), Egnell and Leijon (1999, 2003), Karlsson and Tamminen (2013), Saarsalmi et al. (2010), Smolander et al. (2015), Tamminen and Saarsalmi (2013), and Wall and Hytonen, (2011).

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

spruce seedlings were badly suppressed by birch and other deciduous trees in the experiment. With this remark the trend that slash harvest in final felling reduce stand growth in the subsequent spruce stand remains. Another, possibly important, observation is that a majority of the studies with ratios below one has observation periods longer than 10 years (Fig. 2). This is in line with a study by Egnell (2011), giving some evidence suggesting that effects on site and stand productivity following slash harvest in final felling is delayed with significant effects only after more than 10 years. Longer-term data for final determination of growth effects of slash harvest have been suggested in many publications (e.g. Tamminen and Saarsalmi, 2013). This will allow for residual slash to decompose and thereby release its nutrients and the stands will reach a stage in their growth when the need for nutrients is at its highest. A meta-analysis on slash harvest effects on subsequent growth by Achat et al. (2015) showed a 3–7% growth reduction in the short- to medium-term. However, their supplementary material showed that a majority of the data included were from experiments that were ten years or younger. That may be too short a time to reveal the full effect of the treatment. Furthermore, the meta-analysis included studies where the forest floor was also raked off the site and studies with control plots receiving double amounts of slash. These extreme treatments are not relevant for current practical slash harvest operations and recovery rates of slash in practical forestry are lower than those applied in most experiments (Thiffault et al., 2015). For the individual pine studies statistically significant growth reductions following slash harvest were reported from two sites (Egnell, 2016; Egnell and Valinger, 2003) and increases from two sites (Egnell, 2016; Egnell and Leijon, 1999). Both experiments where growth was increased represent lower site indices (SI = 19 and 20) and sites in northern Sweden with a harsher climate (latitude 64), whereas the experiments with reduced growth represent more fertile sites (SI = 24) in central and southern Sweden (latitude 60 and 57). However, in one of the experiments with increased growth (Lund), seedling survival was low in all treatments, with a survival rate around 50% on slash-harvested plots and 35% on control plots. Whether higher survival rate was related to slash harvest is therefore important information for the interpretation of the results. Egnell and Leijon (1999) mention frost and snow blight (Phacidium infestans Karst.) as major causes of seedling mortality in the experiment. For frost damage slash could give the necessary shade from direct sunlight after a clear night with frost during the vegetation period and thereby reduce mortality (Lundmark and Hällgren, 1987). For snow blight slash retention operates in the opposite direction by enhancing the infection (Hansson, 2006). In the other experiment with increased growth, survival rates were high and similar between treatments, but slash harvest was combined with stump harvest. It is therefore impossible to separate slash-harvest effects from stump-harvest effects. For the two sites with significantly reduced growth, survival rates were higher following slash harvest, statistically significant in one case (Egnell and Leijon, 1999). On the other site slash harvest was combined with stump harvest (Egnell, 2016) complicating the interpretation of slash harvest effects. Altogether pine data from the experiments suggests that slash harvest in final felling does not have a negative impact on growth of the subsequent pine stand. One possible reason for this is that the nutrient content in slash often is lower in pine stands than in spruce stands (Palviainen and Finer 2012). These results suggest that there is a correlation between regeneration success and growth of the subsequent stand. This is visualised in Fig. 3, where relative regeneration success (seedling survival rates or number of stems with or without natural regeneration) is plotted against relative growth. Although not a straight line the correlation is obvious with q = 0.699 (p < 0.001). Unless

5

Fig. 3. Relative survival rate or stem no (SSH/SOH) plotted against relative growth (SSH/SOH) for the subsequent tree crop planted following slash harvest (SSH) in relation to stem-only harvest (SOH). Filled symbols from experiments planted with Norway spruce and open symbols from experiments planted with Scots pine. Data includes plots where also the stumps were harvested and where also the natural regeneration was included. Data from Egnell (2016), Egnell and Leijon (1999), Kaarakka et al. (2014), Karlsson and Tamminen (2013), Smolander et al. (2015), Tamminen and Saarsalmi (2013), and Wall and Hytonen, (2011).

there is a direct link between regeneration success and nutrient withdrawal, this suggests that it is not primarily the withdrawal of nutrients that is critical for growth of the subsequent stand following slash harvest in final felling – but rather the regeneration success. If that is the case the necessary question to ask is if it is the slash harvest per se that determines the regeneration success or if other factors become more important? One such factor has already been mentioned – namely how natural regeneration has been managed in the experiments. If slash harvest stimulates natural regeneration as a result of more exposed seedbeds suitable for seed germination, this may result in a higher stem density following slash harvest. On the other hand, if natural regeneration is removed in a late pre-commercial thinning, planted seedlings may have experienced competition during establishment resulting in increased mortality and reduced growth. Slash could also have an impact on seedling survival and growth through its impact on i.e. microclimate and competing vegetation (Trottier-Picard et al. 2014; Lundmark and Hällgren, 1987), forest pests (Örlander et al., 2001) and diseases (Hansson, 2006). 3.2. Slash harvest in thinnings Evidence of the slash harvest effect in thinning was collected from five different publications covering 41 long-term experiments in Finland, Sweden and Denmark. When slash is harvested in thinnings nutrients are removed from the stand and thereby withheld from the residual stand immediately ready to take advantage of it. Residual slash has the potential to also change nutrient availability for the residual stand by suppressing competing vegetation and by changing the conditions for mineralisation of litter and humus. Thus, compared to slash harvest in final felling it is a more straightforward case, with an established stand ready to respond to any change in nutrient availability. The results from long-term field experiments in the Nordic countries presented here show a common pattern with reduced growth following slash harvest in most experiments and in both spruce and pine (Fig. 4). A growth ratio lower than one is found in 13 out of 16 experiments in spruce (mean ratio = 0.95) and in 19 out of 24 experiments in pine (mean ratio = 0.95). For pine there are no obvious trends in the response pattern when plotted over site index (q = 0.099, p = 0.555), latitude (q = 0.232,

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

6

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

Fig. 4. Relative volume growth for the subsequent stand during 10–25 years following slash harvest (SLH) in thinning in relation to stem-only harvest (SOH, dashed horizontal line) plotted over site index, latitude, and the estimated additional amount of nitrogen (N) harvested with the slash. Left column Norway spruce (filled) and right column Scots pine (open), where the smaller symbols indicate a growth period of 10 years and large symbols a growth period between 18 and 25 years following harvest. Diamonds indicate experiments where a second thinning with slash harvest has been performed. Growth data from Helmisaari et al. (2011) are included twice (connected with a dashed line) in the figure, i.e., growth 0–10 years after first thinning and growth 11–20 years after first thinning, with a second thinning with slash harvest performed in some of the experiments after 10 years. Data also from Egnell and Leijon (1997), Egnell and Ulvcrona (2015), Nord-Larsen (2002), and Tveite and Hanssen (2013).

p = 0.160), or additional N harvested with the slash (q = 0.021, p = 0.898). In spruce there are trends suggesting that the relative growth loss increases with latitudes (q = 0.453, p = 0.026), decreases with increased amount of additional N harvested with the slash (q = 0.402, p = 0.052), whereas no trend over site index could be reviled (q = 0.189, p = 0.377). Although growth rates were lower during ten years after slash harvest in thinning, they were not statistically different in two pine and two spruce experiments (Egnell and Leijon, 1997). But, ten years may be a too short a period to experience the full effect of slash harvest. Helmisaari et al. (2011) reported statistically significant growth reduction in spruce (5, 13%) for the two growth periods analysed (0–10 and 11–20 years after thinning). For the pine experiments a statistically significant growth reduction (8%) was revealed only for the second growth period suggesting a long-term response in pine. Longer-term data (18–21 years after thinning) from four pine experiments were analysed by Egnell and Ulvcrona (2015), with a statistically significant growth reduction (4%) following slash harvest, and by Tveite and Hanssen (2013) from four spruce experiments (25 years after thinning), with volume production reduced by 11%. Both Egnell and Ulvcrona (2015) and Tveite and Hanssen (2013) pointed out statistically significant early growth responses when analysing different

growth periods. Also Nord-Larsen (2002), reported statistically significant early growth reductions during the first four years following slash harvest in thinning in two spruce experiments in Denmark – a reduction that was not detectable after ten years. Considering the nutrient release dynamics of slash (Hyvönen et al. 2000), a delayed growth response should be expected (c.f. Egnell, 2011). Possible explanations for short-term growth responses are suppression of competing vegetation and changes in soil conditions important for mineralisation rates (temperature and moisture) by the slash left on site. In the case of Egnell and Ulvcrona (2015) the whole trees were left on control plots (pre-commercial thinning), resulting in an even stronger potential mulching effect. Also Sterba (1988) reported higher growth rates on control plots during the first three years following precommercial thinning (whole trees left on control plots) of Norway spruce stands in Austria. Soil nutrient data following slash harvest in thinning do not indicate serious problems with nutrient availability with detectable reductions of exchangeable nutrients only for Ca, K, and Mg in the organic soil horizon (Tamminen et al. 2012; Olsson, 1999). Nutrients that normally do not limit stand growth on upland soils in the Nordic countries. A study on needle nutrient data based on 12 long-term experiments gave similar results with minor effects of slash harvest and no general trend

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

although some treatment differences were documented (Luiro et al., 2010). With this uncertainty on effect dynamics it is here suggested that longer-term data, including potential effects of nutrients slowly released from slash, should be at hand before any final conclusions are drawn. 3.3. Stump harvest in final felling The number of published papers based on long-term experiments with stump harvest in the Nordic countries and the world is limited. Presented data are based on three publications and in total 13 experimental sites in Sweden and Finland. More experiments that will deliver data in the future have been established in both countries. 3.3.1. Regeneration success following stump harvest in final felling Following stump harvest relative survival rates/stem number of the planted tree species tends to be unaffected or rather stimulated by stump harvest with a weak positive trend over site index (q = 0.517, p = 0.126) and no trend over latitude (q = 0.025, p = 0.946) (Fig. 5). For the individual studies, significantly higher stem numbers of the planted tree species were reported by Karlsson and Tamminen (2013), with relative stem number ratios of 1.2 and 1.3 for Norway spruce and Scots pine, respectively. In that study no additional site preparation was done on any of the plots, including the control plots, and slash was harvested in addition to stumps on both stump-harvested and control plots. They also reported significantly more natural regeneration, primarily of birch, on stump-harvested plots despite the fact that many of them were removed in a pre-commercial thinning 10 years after planting. For two spruce and two pine experiments a slightly, but statistically significant, higher survival rate following stump harvest was reported for the least productive pine site in northern Sweden (Egnell, 2016). In the same study one of the spruce experiments showed a low, but not statistically different, survival rate following stump and slash harvest (c.f. Fig. 5, survival ratio = 0.7). Metadata from that experiment indicates a high abundance of natural regeneration on stump-harvested plots that may have suppressed the planted seedlings? Survey studies following practical stump harvest operations in Finland support the idea that stump harvest promotes natural regeneration of pioneer tree species, particularly on moist and fine textured soils (Hyvönen et al., 2016; Saksa, 2013). Notable here is that in practical operations also slash is harvested on stump-harvested sites. 3.3.2. Forest production following stump harvest in final felling The trend is that growth is positively affected by stump harvest, at least in experiments with Scots pine, but less so when also slash is harvested (Fig. 6). The correlation analyses suggests that the positive effect is stronger on poorer sites when only the stumps

7

are harvested (q = -0.530, p = 0.076), whereas the correlation with SI is weak when both slash and stumps are harvested (q = 0.103, p = 0.705). For latitude growth ratios tend to increase with increasing latitude following both stump harvest (q = 0.517, p = 0.085) and slash and stump harvest (q = 0.612, p = 0.012). For the individual studies a statistically significant treatment effect was only revealed in experiments with Scots pine, with increased growth on two sites (Egnell, 2016; Karlsson and Tamminen, 2013) and decreased growth on one site (Egnell, 2016). Metadata from the site with decreased forest production give information on severe browsing damages by moose (Alces alces (L.)) on the planted pines. If applied treatments have had an impact on browsing damages as a result of differences in abundance of natural regeneration or palatability effects on the pines (c.f. Edenius 1993), this may have had an impact on the results. Palviainen et al. (2010) showed that the nitrogen content in stumps increased after harvest and was substantially larger as long as 40 years after harvest. Thus, one possible explanation for reduced forest production on control plots could be that decomposers exploiting residual stumps and slash immobilize important nutrients limiting forest production on poor sites. This could explain why the positive effect of stump harvest is more pronounced on less fertile sites or at higher latitudes (Fig. 6). However, this could also be a species effect since the pine experiments primarily are located on poor sites. The low ratio (0.4) in Fig. 6 originates from the spruce site with reduced survival rate (c.f. Fig. 5), hence, there may be a relation between survival rates/stem number and forest production also in stump harvest. This opens the question whether survival rate primarily is affected by the increased harvest intensity or if other causes are more important. With stump harvest stimulating natural regeneration (Hyvönen et al., 2016; Saksa 2013), management of that natural regeneration becomes critical for the development of the planted seedlings and thus forest production. I.e. with a late removal of competing natural regeneration planted seedlings may have suffered from competition and thereby demonstrated reduced forest production over time. This may have counteracted early positive growth responses following stump harvest as shown in a study by Kataja-Aho et al. (2012). In experiments where growth of natural regeneration was included forest production may have increased as a result of stump harvest. One observation in the survey study by Saksa (2013), where 37 stump-harvested sites were compared with 10 control sites with the stumps remaining, was that natural regeneration of pioneer species like pine and birch was higher on stump-harvested sites, whereas the number of naturally regenerated spruce seedlings was lower. This suggests two things. Firstly, that the positive effect on forest production in pine experiments could be due to recruitment of naturally regenerated pine seedlings that were wrongly identified as planted seedlings during attempts to remove natural regeneration. Secondly,

Fig. 5. Relative survival rate/stem number for the subsequent tree crop planted following stem, slash, and stump harvest (SSSH, filled symbols) and stem and stump harvest (SSH, open symbols) in relation to stem-only harvest (Egnell, 2016) or stem and slash harvest (Karlsson and Tamminen, 2013) plotted over site index and latitude. Squares show data from experiments in Scots pine and triangles data from experiments in Norway spruce. The dashed horizontal line shows the level for control plots.

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

8

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

Fig. 6. Relative growth for the subsequent tree crop during 24–34 years following stem and stump harvest (SSH, left column) and stem, stump and slash harvest (SSSH, right column) in relation to stem-only harvest (SOH, Egnell, 2016) or stem and slash harvest (SLH, Karlsson and Tamminen, 2013) plotted over site index and latitude. Open symbols are from experimental sites planted with Scots pine and filled symbols are for sites planted with Norway spruce. The dashed horizontal line shows the level for control plots. Both data for the planted tree species only and data for all tree species, including natural regeneration (triangles), from Karlsson and Tamminen (2013) are included.

that natural regeneration of spruce to some extent may originate from beeted seedlings already in place at the time of harvest, with survival rates related to the degree of soil disturbance as a result of harvest and site preparation. Increased soil disturbance following stump harvest as compared to stem-only harvesting has been reported by Kataja-Aho et al. (2012) and Tarvainen et al. (2015). 4. Conclusions and practical implications Results presented here suggest a moderate negative effect of slash harvest in final felling on forest production of the subsequent stand in Norway spruce, whereas Scots pine is unaffected. This would suggest that pine stands should be targeted rather than spruce stands. However, in practical operations focus is on spruce-dominated sites since they produce substantially more slash than pine stands. This is important for the economy of the operation since biomass for the energy marked is a low priced commodity. This could explain the species difference in response since harvest of substantially more slash means that substantially more nutrients is exported and that slash left on site will have a stronger mulching effect. Data from experiments in final felling show a trend suggesting that growth is more negatively affected by slash harvest on poorer sites in Norway spruce, whereas no such trend could be seen in Scots pine. Thus, guidelines suggesting that poor sites should be avoided are supported for one of the two tree species. But since pines are confined to poorer sites, such guidelines could still be misleading. Furthermore, in practical operations poor sites are usually not targeted since they carry less biomass and are therefore not economically suitable. Results from experiments with slash harvest in thinnings are more consistent, with growth reductions in the subsequent stand in most experiments for both spruce and pine. In spruce there is also a weak trend suggesting that the growth response in relative terms is more pronounced at higher latitudes. The reason for this consistency in thinnings may be that it is a case where production

of the residual stand primarily is nutrient limited, whereas in the case of final felling production of the subsequent stand to a larger extent depends on e.g. regeneration success. From a forest production point of view it seems more critical to harvest slash in thinnings and the recommendation would be to harvest more in final felling. However, results from thinning experiments show that compensation fertilization offset the growth reduction (Helmisaari et al., 2011). Thus, if the revenue from harvested slash could pay for fertilization that could counteract the loss. Furthermore, if markets for pulpwood drops, thinning operations are still needed in order to improve growth of stems of higher quality. During such conditions and a bioenergy market in place, harvesting of the whole tree makes more sense and profit. The limited amount of studies on stump harvest effects suggest that forest production is unaffected or in the case of pine even stimulated. The stimulation effect tends to be higher when only the stumps are harvested, i.e. the when the slash is left on site. Thus, stump harvest may counteract growth losses due to slash harvest. The positive effect also tends to be more pronounced on poorer sites at higher latitudes – possibly as an effect of nutrient immobilisation in logging residues on control plots. From a forest production point of view the recommendations based on these results would be to harvest more stumps than slash and to focus more on pine stumps than on spruce stumps. In practice, with current technology, it is more expensive to procure stumps than slash – therefore slash will be targeted before stumps (Lundmark et al., 2015). Furthermore, spruce stumps often have a more superficial root system making them easier to harvest. Presented data show a correlation between seedling survival (or stem number) and growth in experiments with additional biomass harvest in final felling. This may be one reason why the growth response does not show any consistent results, although there are evidence suggesting that slash harvest has a negative effect on site productivity – at least for a while (Egnell, 2011). Input from natural regeneration complicates this further and there is

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

empirical evidence supporting that natural regeneration is stimulated by slash and stump harvest (Karlsson et al. 2002; Saksa, 2013; Hyvönen et al., 2016). If that additional natural regeneration is included as part of the stand growth, this could result in improved forest growth following slash and stump harvest. On the other hand, if it is not included, management of the natural regeneration in the experiments becomes critical. Natural regeneration could have suppressed the planted seedling resulting in both increased mortality and growth. This could instead have exaggerated treatment effects. From a practical forestry perspective the latter case is more relevant since foresters tend to care more for the planted, genetically improved, seedlings. Furthermore, natural regeneration is often dominated by less productive tree species, such as birch, with limited market opportunities as compared to pine and spruce. This suggests that in a practical case increased natural regeneration in many cases will be removed in precommercial thinnings rather than something that adds to forest production. This behaviour could be changed in a changed market situation where more of the small diameter and low quality trees are marketed on a bioenergy/biomass market rather than on the traditional forest industry market. When interpreting the results it is also important to bear in mind that in many experiments recovery rates of biomass is much higher than in practical operations (Thiffault et al., 2015). Removing slash from the clearcut also has the potential to make site preparation and planting more efficient and with better quality (Saarinen, 2006). With this in mind, together with the results presented here, impacts on future forest production is not a strong argument against using slash and stumps for energy purposes. Only short- to medium-term data are presented here. Although there are some evidence suggesting a transient effect on site productivity (Egnell, 2011), it is recommended to continue to follow the development in long-term experiments and to get more data from experiments where logging residues have been harvested multiple times. To evaluate if there are any hidden consistency in growth responses following increased biomass harvest intensities next step would be a formal meta-analysis including a large data set that takes all possible meta-information into account. Acknowledgement This study was financed by the Swedish Energy Agency. The author directs his gratitude to Anna Saarsalmi and Aino Smolander for sharing data from experiments in Finland and to the field based staff at institutions in the Nordic countries responsible for maintaining long-term field experiments. References Achat, D.L., Deleuze, C., Landmann, G., Pousse, N., Ranger, J., Augusto, L., 2015. Quantifying consequences of removing harvesting residues on forest soils and tree growth – a meta-analysis. For. Ecol. Manage. 348, 124–141. Edenius, L., 1993. Browsing by moose on Scots pine in relation to plant resource availability. Ecology 74, 2261–2269. Egnell, G., 2016. Effects of slash and stump harvesting after final felling on stand and site productivity in Scots pine and Norway spruce. For. Ecol. Manage. 371, 42–49. http://dx.doi.org/10.1016/j.foreco.2016.03.006. Egnell, G., 2011. Is the productivity decline in Norway spruce following whole-tree harvesting in the final felling in boreal Sweden permanent or temporary? For. Ecol. Manage. 261, 148–153. Egnell, G., Leijon, B., 1997. Effects of different levels of biomass removal in thinning on short-term production of Pinus sylvestris and Picea abies. Scand. J. For. Res. 12, 17–26. Egnell, G., Leijon, B., 1999. Survival and growth of planted seedlings of Pinus sylvestris and Picea abies after different levels of biomass removal in clearfelling. Scand. J. For. Res. 14, 303–311. Egnell, G., Ulvcrona, K.A., 2015. Stand productivity following whole-tree harvesting in early thinning of Scots pine stands in Sweden. For. Ecol. Manage. 340, 40–45. Egnell, G., Valinger, E., 2003. Survival, growth, and growth allocation of planted Scots pine trees after different levels of biomass removal in clear-felling. For. Ecol. Manage. 177, 65–74.

9

Ericsson, K., Huttunen, S., Nilsson, L.J., Svenningsson, P., 2004. Bioenergy policy and market development in Finland and Sweden. Energy Policy 32, 1707–1721. Fleming, R.L., Leblanc, J.D., Hazlett, P.W., Weldon, T., Irwin, R., 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. Fleming, R.L., Powers, R.F., Foster, N.W., Kranabetter, J.M., Scott, D.A., Ponder, F., Berch, S., Chapman, W.K., Kabzems, R.D., Ludovici, K.H., Morris, D.M., PageDumroese, D.S., Sanborn, P.T., Sanchez, F.G., Stone, D.M., Tiarks, A.E., 2006. Effects of organic matter removal, soil compaction, and vegetation control on 5-year seedling performance: a regional comparison of long-term soil productivity sites. Can. J. For. Res. 36, 529–550. Hansson, P., 2006. Effects of small tree retention and logging slash on snow blight growth on Scots pine regeneration. For. Ecol. Manage. 236, 368–374. Helmisaari, H.-S., Hanssen, K. Holt, Jacobson, S., Kukkola, M., Luiro, J., Saarsalmi, A., Tamminen, P., Tveite, B., 2011. Logging residue removal after thinning in Nordic boreal forests: long-term impact on tree growth. For. Ecol. Manage. 261, 1919– 1927. Hyvönen, R., Olsson, B.A., Lundkvist, H., Staaf, H., 2000. Decomposition and nutrient release from Picea abies (L.) Karst. and Pinus sylvestris L. logging residues. For. Ecol. Manage. 126, 97–112. Hyvönen, R., Kaarakka, L., Leppälammi-Kujansuu, J., Olsson, B.A., Palviainen, M., Vegerfors-Persson, B., Helmisaari, H.-S., 2016. Effects of stump harvesting on soil C and N stocks and vegetation 8–13 years after clear-cutting. For. Ecol. Manage. 371, 23–32. Jurevics, A., Peichl, M., Olsson, B.A., Strömgren, M., Egnell, G., 2016. Slash and stump harvest have no general impact on soil and tree biomass C pools after 32–39 years. For. Ecol. Manage. 371, 33–41. http://dx.doi.org/10.1016/j.foreco.2016.01.008. Kaarakka, L., Tamminen, P., Saarsalmi, A., Kukkola, M., Helmisaari, H.-S., Burton, A.J., 2014. Effects of repeated whole-tree harvesting on soil properties and tree growth in a Norway spruce (Picea abies (L.) Karst.) stand. For. Ecol. Manage. 313, 180–187. Karlsson, K., Tamminen, P., 2013. Long-term effects of stump harvesting on soil properties and tree growth in Scots pine and Norway spruce stands. Scand. J. For. Res. 28, 550–558. Karlsson, M., Nilsson, U., Örlander, G., 2002. Natural regeneration in clear-cuts: effects of scarification, slash removal and clear-cut age. Scand. J. For. Res. 17, 131–138. Kataja-Aho, S., Pennanen, T., Lensu, A., Haimi, J., 2012. Does stump removal affect early growth and mycorrhizal infection of spruce (Picea abies) seedlings in clear-cuts? Scand. J. For. Res. 27, 746–753. Luiro, J., Kukkola, M., Saarsalmi, A., Tamminen, P., Helmisaari, H.S., 2010. Logging residue removal after thinning in boreal forests: long-term impact on the nutrient status of Norway spruce and Scots pine needles. Tree Physiol. 30, 78–88. Lundmark, R., Athanassiadis, D., Wetterlund, E., 2015. Supply assessment of forest biomass – a bottom-up approach for Sweden. Biomass Bioenergy 75, 213–226. Lundmark, T., Hällgren, J.E., 1987. Effects of frost on shaded and exposed spruce and pine seedlings planted in the field. Can. J. For. Res. 17, 1197–1201. Mälkönen, E., 1976. Effect of whole-tree harvesting on soil fertility. Silva Fennica 10, 157–164. Minitab 17 Statistical Software, 2010. [Computer software]. Minitab, Inc., State College, PA. . Morris, D.M., Kwiaton, M.M., Duckert, D.R., 2014. Black spruce growth response to varying levels of biomass harvest intensity across a range of soil types: 15-year results. Can. J. For. Res. 44, 313–325. Nilsson, U., Luoranen, J., Kolstrom, T., Orlander, G., Puttonen, P., 2010. Reforestation with planting in northern Europe. Scand. J. For. Res. 25, 283–294. Nord-Larsen, T., 2002. Stand and site productivity response following whole-tree harvesting in early thinnings of Norway spruce (Picea abies (L.) Karst.). Biomass Bioenergy 23, 1–12. Nykvist, N., Rosén, K., 1985. Effect of clear-felling and slash removal on the acidity of northern coniferous soils. For. Ecol. Manage. 11, 157–169. Olsson, B.A., 1999. Effects of biomass removal in thinnings and compensatory fertilization on exchangeable base cation pools in acid forest soils. For. Ecol. Manage. 122, 29–39. Örlander, G., Nordlander, G., Wallertz, K., 2001. Extra food supply decreases damage by the pine weevil Hylobius abietis. Scand. J. For. Res. 16, 450–454. Palviainen, M., Finer, L., 2012. Estimation of nutrient removals in stem-only and whole-tree harvesting of Scots pine, Norway spruce, and birch stands with generalized nutrient equations. Eur. J. For. Res. 131, 945–964. Palviainen, M., Finér, L., Laiho, R., Shorohova, E., Kapitsa, E., Vanha-Majamaa, I., 2010. Carbon and nitrogen release from decomposing Scots pine, Norway spruce and silver birch stumps. For. Ecol. Manage. 259, 390–398. Ponder Jr., F., Fleming, R.L., Berch, S., Busse, M.D., Elioff, J.D., Hazlett, P.W., Kabzems, R.D., Kranabetter, J.M., Morris, D.M., Page-Dumroese, D., Palik, B.J., Powers, R.F., Sanchez, F.G., Scott, D.A., Stagg, R.H., Stone, D.M., Young, D.H., Zhang, J., Ludovici, K.H., McKenney, D.W., Mossa, D.S., Sanborn, P.T., Voldseth, R.A., 2012. Effects of organic matter removal, soil compaction and vegetation control on 10th year biomass and foliar nutrition: LTSP continent-wide comparisons. For. Ecol. Manage. 278, 35–54. Powers, R.F., 2006. Long-term soil productivity: genesis of the concept and principles behind the program. Can. J. For. Res. 36, 519–528. Saarinen, V.M., 2006. The effects of slash and stump removal on productivity and quality of forest regeneration operations-preliminary results. In: Proceedings of the Third Annual Workshop of Task 31 ’Sustainable Production Systems for Bioenergy: Impacts on Forest Resources and Utilization of Wood for Energy’, October 2003, Flagstaff, Arizona, USA. Biomass Bioenergy, 30, pp. 349–356.

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019

10

G. Egnell / Forest Ecology and Management xxx (2016) xxx–xxx

Saarsalmi, A., Tamminen, P., Kukkola, M., Hautajarvi, R., 2010. Whole-tree harvesting at clear-felling: impact on soil chemistry, needle nutrient concentrations and growth of Scots pine. Scand. J. For. Res. 25, 148–156. Saksa, T., 2013. Regeneration after stump harvesting in southern Finland. For. Ecol. Manage. 290, 79–82. Sikström, U., 2004. Survival, growth and needle element concentrations of Picea abies (L.) Karst. seedlings after brash removal in a previously N fertilized stand. For. Ecol. Manage. 203, 123–134. Smolander, A., Saarsalmi, A., Tamminen, P., 2015. Response of soil nutrient content, organic matter characteristics and growth of pine and spruce seedlings to logging residues. For. Ecol. Manage. 357, 117–125. Sterba, H., 1988. Increment losses by full-tree harvesting in Norway spruce (Picea abies). For. Ecol. Manage. 24, 283–292. Stupak, I., Asikainen, A., Jonsell, M., Karltun, E., Lunnan, A., Mizaraite, D., Pasanen, K., Parn, H., Raulund-Rasmussen, K., Roser, D., Schroeder, M., Varnagiryte, I., Vilkriste, L., Callesen, I., Clarke, N., Gaitnieks, T., Ingerslev, M., Mandre, M., Ozolincius, R., Saarsalmi, A., Armolaitis, K., Helmisaari, H.S., Indriksons, A., Kairiukstis, L., Katzensteiner, K., Kukkola, A., Ots, K., Ravn, H.P., Tamminen, P., 2007. Sustainable utilisation of forest biomass for energy-possibilities and problems: policy, legislation, certification, and recommendations and guidelines in the Nordic, Baltic, and other European countries. Biomass Bioenergy 31, 666–684. Tamminen, P., Saarsalmi, A., 2013. Effects of whole-tree harvesting on growth of pine and spruce seedlings in southern Finland. Scand. J. For. Res. 28, 559–565.

Tamminen, P., Saarsalmi, A., Smolander, A., Kukkola, M., Helmisaari, H.-S., 2012. Effects of logging residue harvest in thinnings on amounts of soil carbon and nutrients in Scots pine and Norway spruce stands. For. Ecol. Manage. 263, 31–38. Tarvainen, O., Hekkala, A.-M., Kubin, E., Tamminen, P., Murto, T., Tolvanen, A., 2015. Soil disturbance and early vegetation response to varying intensity of energy wood harvest. For. Ecol. Manage. 348, 153–163. Thiffault, E., Béchard, A., Paré, D., Allen, D., 2015. Recovery rate of harvest residues for bioenergy in boreal and temperate forests: a review. Wiley Interdisc. Rev. – Energy Environ. 4, 429–451. Thiffault, E., Hannam, K.D., Pare, 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. Trottier-Picard, A., Thiffault, E., DesRochers, A., Pare, D., Thiffault, N., Messier, C., 2014. Amounts of logging residues affect planting microsites: a manipulative study across northern forest ecosystems. For. Ecol. Manage. 312, 203–215. Tveite, B., Hanssen, K.H., 2013. Whole-tree thinnings in stands of Scots pine (Pinus sylvestris) and Norway spruce (Picea abies): short- and long-term growth results. For. Ecol. Manage. 298, 52–61. Wall, A., Hytonen, J., 2011. The long-term effects of logging residue removal on forest floor nutrient capital, foliar chemistry and growth of a Norway spruce stand. Biomass Bioenergy 35, 3328–3334.

Please cite this article in press as: Egnell, G. A review of Nordic trials studying effects of biomass harvest intensity on subsequent forest production. Forest Ecol. Manage. (2016), http://dx.doi.org/10.1016/j.foreco.2016.09.019