Variable corridor thinning – a cost-effective key to provision of multiple ecosystem services from young boreal conifer forests?

ABSTRACT

The review discusses the potential of mechanized thinning operations with variable corridor patterns as a method to secure multiple ecosystem services. The focus is on young and dense forests, which are increasingly abundant in Northern Europe and a potential source of renewable biomass for the needs of future bioeconomy. Conventional selective (motor-manual) pre-commercial thinning (PCT) without outtake of cut biomass is used as a benchmark to evaluate a new mechanized thinning method: boom corridor thinning (BCT). The paucity of specific studies on the environmental effects of BCT limits systematic and quantitative comparisons. However, information extrapolated from studies on selective or other corridor thinnings suggests that BCT potentiates early outtake of forest biomass for energy or biorefineries while simultaneously maintaining the stand structure's vertical heterogeneity and thereby supporting biodiversity. More experimental evidence is urgently needed to elucidate the detailed environmental consequences of BCT, and especially its effects on biodiversity and carbon balance. The increasing need to evaluate the pros and cons of silvicultural operations against a broad range of ecosystem benefits necessitates a holistic approach and the development of new typologies and indices that better reflect the structural properties of forest stands.

KEYWORDS:

• Young forest management
• pre-commercial thinning
• forest biomass
• boreal forests
• ecosystem services

Introduction

Early successional young forests arise after disturbances caused by wildfires, storms, insect damage, diseases, or logging operations. The effective suppression of disturbances, in particular wildfires, has created deficits of young forests in some areas, notably in regions of the USA (Gallant et al. 2003; Keeling et al. 2006; Pan et al. 2011). Conversely, the proportion of young forests has increased in Nordic countries where forests have been actively managed for several decades, with even-aged stand management regimes predominating (Kuuluvainen et al. 2012; Lundmark et al. 2017). Since even-aged management regimes in boreal regions typically conclude with a clear cut after approximately a century, the proportion of trees of a given age is usually quoted to the nearest percentage point. The total proportion of young forests (i.e. forests up to 20 years old) in Nordic and Baltic countries (Denmark, Norway, Sweden, Finland, Estonia and Latvia) where forests are predominantly subject to even-aged management is reported to be 18% (Rytter et al. 2014, 2015, 2016; Fernandez-Lacruz et al. 2015). Around 9% of the forests in Russia were younger than 30 years of age at the end of 2012 (Loboda and Chen 2017), indicating that Russian boreal forests are subject to different management strategies than those in other parts of Europe. Overall, young forests account for a substantial proportion of the boreal forest as a whole, and today's management decisions regarding these forests will lay the foundation for future forestry.

Traditionally, management regimes for young stands have aimed to secure future revenues, usually through outtake of biomass for timber and fuel. However, the goals of forestry have diversified strongly in recent decades, emphasizing the whole spectrum of ecosystem services provided by forests (e.g. Hörnfeldt and Ingemarson 2006). These services include, among others, carbon sequestration for climate change mitigation and maintenance of habitats for various species (for more detailed information about ecosystem services we refer to, e.g. Bennett et al. 2015; La Notte et al. 2017 and Mori et al. 2017). The provision of many ecosystem services is strongly determined by stand structure (Pretzsch 1997; Pommerening 2002; Bohn and Huth 2017), with vertical complexity being one of the most influential factors regulating the availability of habitats supporting biological diversity (Puettmann et al. 2009, 2014). The vertical diversity of the mature forest is highly sensitive to the management of young stands, and the stand structure-dependence of the relationship between early management and the future provision of multiple ecosystem services means that current-day decisions about young stand management will have important effects for a long time to come (Ahnlund Ulvcrona et al. 2017). The complexity of making these decisions is exacerbated by the need to consider multiple future goals. In recent years, growing interest in the immediate utilization of young, dense stands as sources of biomass (Bergström et al. 2010a, 2010b; Fernandez-Lacruz et al. 2015) has enhanced this complexity by introducing a new short-term objective. This interest is fuelled by the ongoing transition from fossil-based to bio-based economies: the associated demand for biofuels and biomaterials necessitates intensified utilization of forest biomass (Pülzl et al. 2014; De Besi and McCormick 2015). One way to meet this need is to increase the outtake of biomass from young, dense forests by performing early thinnings.

The main silvicultural operations influencing the structural aspects of future stands are commercial or pre-commercial thinnings that aim to improve the stands’ value-growth, with commercial thinning also generating immediate income (Huuskonen and Hynynen 2006). Conventional pre-commercial thinning (PCT) is a form of selective thinning that is usually performed motor-manually at a low stand height (2–5 m) and leaves a stand density of a few thousand stems per ha in a single height layer. It is intended to reduce the competition experienced by final crop trees and increase log dimensions during subsequent thinnings, and it promotes a more uniform distribution of value-trees across the managed forest area (Fahlvik et al. 2015). In PCT, the cut biomass is not utilized. The growth increasing effect of selective thinnings on stem volume has, however, been found to depend on tree species, with Norway spruce responding more readily than Scots pine (Nilsson et al. 2010), and on growth conditions (Skovsgaard 2009).

Mechanical corridor thinning methods have been developed as cost-effective alternatives to selective thinnings in future, as the proportion of manual work may be expected to decrease. One new early thinning method for young dense stands is boom-corridor thinning (BCT) (Bergström 2009; Sängstuvall et al. 2012; Jundén et al. 2013), in which the trees are harvested in narrow (∼1 m wide) corridors, aligned to the strip-road and with a length corresponding to the crane's reach (∼10 m). BCT is a cost-efficient harvesting operation method that allows flexible use of different thinning patterns. For example, using a fan-shaped pattern, “laid out” by the decision of the operator, will give higher degree of tree selection than of using a perpendicular pattern laid out strictly systematically (Bergström 2009). Selective and corridor thinning methods will inevitably have very different effects on vertical stand structure, habitat diversity and light conditions. It is important to note that the current forest management praxis, in which the future stand structure is more permanently adjusted through the first commercial thinning, may annul the impacts of different early thinnings. In future, however, the current silvicultural praxis may be challenged and changed due to the diversified goals for forest management. In that situation, the stands subjected to BCT are likely to allow a wider spectrum of alternative silvicultural trajectories than what is possible in the homogenous stand established through conventional selective PCT. Ahnlund Ulvcrona et al. (2017) recently suggested that BCT should be a cost-effective option for thinning that could also effectively support biodiversity and provisioning of multiple ecosystem services. However, the environmental effects of BCT have not yet been systematically evaluated, so its sustainability benefits are unclear.

The overall goal of this review is to discuss the potential of early thinnings to support the future provision of diverse ecosystem services by the thinned forests. Using conventional motor-manual PCT which aims at homogenous stand structure (Ahnlund Ulvcrona et al. 2014; Fahlvik et al. 2015) as a benchmark, we examine the potential effects of thinnings in flexibly chosen corridors on different ecosystem services. Particular emphasis is placed on gauging the potential consequences of adopting BCT as a new, economically attractive thinning method (with a removal of biomass) capable of meeting the needs of bioeconomy. We consider that a fundamental difference between PCT and BCT is that the latter more cost-effiently maintains the vertical heterogeneity of forests. We acknowledge that because BCT is a new method, there is little experimental evidence about its environmental consequences, which limits the scope for systematic and quantitative comparisons. Predictions regarding benefits to ecosystem services must therefore be made on the basis of knowledge about more conventional thinning practices. Despite the paucity of specific evidence, we argue that extrapolations of the sort presented here are meaningful because ultimately, the consequences of any thinning operation can be linked to changes in the following factors: (1) the density of the tree species; (2) competition for light, nutrients and water by the vegetation; (3) the removal of biomass from the stand; and (4) the architecture of the canopy layer. Despite the paucity of specific literature, we consider a discussion of the environmental impacts of BCT to be very timely for two reasons. First, the topic is relevant for planning of cost-effective forest-based bioeconomy in northern Europe, and is therefore of great interest to decision makers at both operational and policy levels. Second, information from earlier thinning studies can help guide future studies on BCT to ensure that the most important knowledge gaps are addressed.

Approach and literature survey

Although detailed and quantitative analyses of the consequences of BCT are hampered by the lack of empirical studies on this thinning strategy, we outlined plausible ecological, economic, and social impacts of BCT based on related literature, focusing mainly on stand-level processes and phenomena. We used PCT as a benchmark in this analysis because it is the selective thinning regime most commonly applied in Swedish coniferous forests. It is typically performed once the stand height has reached 2–3 m, and the recommended post-PCT stand density is 1500–3500 trees ha–1. Broad-leaved trees are usually only left in gaps or as replacements for severely damaged conifers (Ahnlund Ulvcrona et al. 2014; Fahlvik et al. 2015).

To structure our analysis and discussion, we approached the topic through a modified Leopold matrix – an analytical tool developed for qualitative environmental impact assessments (Leopold et al. 1971). We assumed that the central attributes and traits affecting ecosystem services can be divided into six main groups: 1. Landscape-level and stand structural traits, 2. Biological attributes, 3. Geophysical attributes, 4. Economic attributes, 5. Human environment attributes and 6. Ecological resilience traits (Figure 1). Most of these groups (1, 2, 3, and 6) are strongly linked to supporting ecosystem services (MA 2005). The attributes in groups 3 and 6 are also linked to regulating services. Attributes in group 4 are mainly linked to provisioning services, while those in group 5 are linked to cultural services. When interpreting our results, it is important to recall that while these groups are treated separately in this work, they are strongly interconnected in reality.

Figure 1. A modified Leopold matrix of the different aspects affected by young stand thinning operations, divided into six partly overlapping domains: landscape-level and stand structural traits, biological attributes, geophysical attributes, economic attributes, human environment attributes and ecological resilience traits.

We then examined the peer-reviewed literature published between 1997 and 2018 on thinning effects related to these attribute groups by performing targeted literature searches. The time period was chose to capture most of the studies that acknowledge the importance of multiple goals in forest management and also utilize modern technology in thinning operations. Searches were performed against the Web of Science (WoS), Scopus, Google Scholar, and PubMed databases. Additional references were extracted from the reference lists of the articles captured in the searches. The search strings were constructed by combining the strings (forest* AND “thinning*”) AND (boreal AND conifer*) with a specific word associated with a given attribute group (e.g. “biodiversity”, “carbon”, or “economy*”).

Potential impacts of corridor thinning on forest traits and attributes

Landscape- and stand-level structural traits

In the long term, the capacity of forests to provide a broad range of ecosystem services depends on their biological diversity, i.e. the variability among the living organisms that they support (Brockerhoff et al. 2017 and references therein). To survive, living organisms require suitable habitats that offer nutrition and shelter; habitat fragmentation is a common cause of species becoming threatened or endangered (Fahrig 2003). Habitat fragmentation can occur slowly as a result of geological processes or more quickly as a result of human activities such as forestry practices. Forestry operations, including thinnings, alter the dynamics of landscapes, causing spatial and temporal changes in composition and age structure that may increase the risk of habitat loss and fragmentation (Schmiegelow and Mönkkönen 2002; Brockerhoff et al. 2017). This has been demonstrated in several studies, including that by Thompson et al. (2003), who emphasized that systems with low connectivity may develop simplified stand structures and species mixtures, increasing their vulnerability to insect and mammalian herbivory. Maintaining an adequate balance between conservation and gain of biomass is therefore a crucial goal for twenty-first century forest management (Mori et al. 2017). However, the spatial and temporal aspects of the larger scale phenomena and processes in forests are complex, and authors such as Schmiegelow and Mönkkönen (2002) have highlighted the importance of system-specific and species-specific considerations when assessing the potential effects of habitat loss and fragmentation on regional biota. They also emphasize that unselective application of conservation paradigms may lead to misguided research efforts and poor management guidelines (Schmiegelow and Mönkkönen 2002).

Thinning operations strongly modify the stand structure and habitats. Unlike selective thinnings that aim to homogenize stand structure, density and composition (Puettmann et al. 2009), BCT allows development of vertical heterogeneity and possible different age-classes (as also possible ingrowth trees are kept) and species mixtures at the stand level (Ahnlund Ulvcrona et al. 2017). It can therefore be used as a tool to shape stand structure within the frames established by the planting. Moreover, BCT has the potential to maintain habitats for several canopy-associated organisms (see, e.g. Röder et al. 2010) more effectively than PCT. Furthermore, corridor thinning patterns create gaps that mimic windthrow effects to a certain degree (Christian et al. 1996). These gaps can act as structures that promote biodiversity (Muscolo et al. 2014; Lu et al. 2018) by allowing light-dependent ground-vegetation species to thrive between the strips of dense tree growth.

Another crucial habitat for species conservation is deadwood. There is a risk that the recovery of biomass in BCT treatments could reduce the abundance of smaller diameter deadwood and woody debris at the stand level. This could have negative consequences for nutrient cycling, carbon storage, and the availability of habitats or substrates for certain birds, saproxylic insects, and wildlife (Omari and MacLean 2015). According to Omari and MacLean (2015), the best strategy for providing high quality deadwood and generating deadwood dynamics is to intentionally leave patches of trees at the time of harvesting. Ferguson and Archibald (2002) found that the percentage of dead stems was highest in young forests (18% of stems in 0–60-year-old forests), and concluded that increased live tree density (e.g. due to intensive forest management practices) may secondarily result in greater snag production and structural complexity that favors wildlife biodiversity. Self-thinning (Bravo-Oviedo et al. 2018) can be enhanced in the zones between the corridors, leading to deadwood production, but there is a need to validate this expectation by studying the effects of operational BCT on deadwood accumulation and quality.

Compared to PCT, BCT has a high potential to create stratified forest structures that provide habitats capable of supporting the rich biodiversity associated with young forests. Trajectories for BCT effects on stand-level structures can be estimated with relatively high certainty because these aspects have studied quite extensively in the context of selective thinnings, and BCT-specific analyses have been reported (Ahnlund Ulvcrona et al. 2017). Nevertheless, more empirical evidence is needed; for example, there is a need to elucidate the long-term ecological cascades initiated by corridor thinnings and their effects on landscape-level processes (e.g. fragmentation) and specific habitats such as deadwood.

Biological attributes

The ongoing 6th extinction wave (Ceballos et al. 2017) has prompted considerable concern about the effects of intensified forestry on biological diversity, i.e. the variability among living organisms associated with forest ecosystems (Estavillo et al. 2013). Several cases have been documented in which habitat loss or fragmentation caused by intensive forest management operations negatively influenced species or populations; in some cases, the ecological mechanisms responsible for these changes have been elucidated (Brockerhoff et al. 2017). For instance, Eggers and Low (2014) found that willow tits (Poecile montana) suffered from higher mortality rates than crested tits (Lophophanes cristatus) when exposed to understory removal by thinning. They therefore suggested that the long-term population decline of willow tits in boreal forests could be linked to large scale harvesting of small-diameter spruce trees (Eggers and Low 2014). It has also been suggested that bird populations’ responses to human-induced structural changes in forest structure may be due to habitat loss rather than true fragmentation effects (Schmiegelow and Mönkkönen 2002; Eggers and Low 2014).

Although specific studies on the impact of BCT on individual species are still lacking, several studies have examined the impacts of conventional thinning operations on species and organism guilds in boreal forests. In a meta-analysis of 33 north American studies (Verschuyl et al. 2011) examined the effects of thinning treatments on forest wildlife biodiversity measured in terms of richness, diversity, or the abundance of individual species, taxa or guilds for birds, mammals, reptiles, amphibians, and invertebrates. They found that “fuels treatment thinning”, i.e. manipulation or removal of biomass to reduce near-term fire risk and to had the most favorable effect on bird and mammalian species abundance and diversity, when compared to commercial and pre-commercial thinnings (Verschuyl et al. 2011). Moreover, they concluded that the reported results regarding the responses of reptiles to thinnings have been variable (Renken et al. 2004) and need to be studied more (Verschuyl et al. 2011). While the clearcuttings seem to have a negative effect on salamander populations, thinnings in general retain better the conditions for the moisture sensitive amphibians (Verschuyl et al. 2011). The effect of thinnings on arthropods (herbivores, predators, detritovores) was found to be positive but the mechanisms behind the responses were deemed to be depend on the life history characters of the species (Verschuyl et al. 2011). In general, the authors identified mainly positive or neutral effects on diversity and abundance across all taxa, but stressed that thinning intensity and the type of thinning conducted may at least partially determine the magnitude of the response.

Thinning effect on the diversity of understorey plant species, which stand for the majority of plant diversity in boreal forests, has often been found to be neutral or positive, although it too seems to depend on the treatment intensity and forest type (Verschuyl et al. 2011). The consequences of thinning regimes on understorey vegetation may be largely due to variations in light availability below the canopy, which has critical effects on successional dynamics, influencing seed germination, plant recruitment, and the establishment, early development, and survival of seedlings and young trees (Messier et al. 1998; Angelini et al. 2015). Variable light conditions can support the development of diverse understory vegetation that in turn provides substrates and habitats for other organisms. However, a canopy cover reduction to as much as 30–50% has been suggested to be necessary before noticeable changes in forest understorey occur (Abella and Springer 2015). In young forests, thinnings may also shape the competition for nutrients and water between the plants. All the species richness that is promoted by thinnings is not necessarily positive: Willms et al. (2017) pointed out that thinnings also create conditions for non-native plants that may exhibit invasive character. Given the lack of clear effects of selective thinnings on plant biodiversity, it is difficult to project the possible effects of corridor thinnings on plant species and communities. More research on this topic is thus needed, and the spatial and temporal scales over which the responses are measured (Rossman et al. 2018) should be carefully considered.

In future, technical advances will allow us to look more closely at the total biodiversity of young boreal forests and discover new horizons at both macro- and microscales. Remote sensing using techniques such as airborne LiDAR provides novel possibilities for mapping vegetation structure and obtaining predictive data on wildlife habitats (Vierling et al. 2013). We will also be able to look inside trees with greater accuracy. For example, advances in next generation sequencing and metagenomics have revealed that forest trees harbor a rich diversity of biologically active fungi and bacteria (Newcombe 2011; Witzell et al. 2014). Trees receive these microbes from the environment, air, soil, and other plants, so manipulation of forest structures could greatly influence the assemblages that develop in trees at specific sites. As potent pioneer decomposers in senescing tissues (Müller et al. 2001) and as protective agents (Miller et al. 2008; Martín et al. 2013), these organisms may be functionally important not only for tree growth and health, but also in central processes such as carbon and nutrient cycling. These microbes may thus provide opportunities for ecological engineering: manipulation of the microbiome could be used to adjust forest stand traits in desirable ways, for instance to improve resistance to diseases or pests (Miller et al. 2008; Busby et al. 2017; Witzell and Martín 2018).

Expanded studies that examine both macro- and micro-scale effects will provide important insights into the impacts and sustainability of BCT and other thinning systems. This will enable holistic syntheses and analyses to evaluate the sustainability of different thinning methods, addressing both the mechanistic and functional levels. For instance, we expect that BCT will produce less openness and light exposure than PCT. Consequently, the two thinning methods will have different effects on the anatomical, morphological and physiological traits of needles (e.g. their proportion of mesophyll; Gebauer et al. 2011), which will in turn influence the chemical quality of the green biomass (e.g. its content of lignin and other phenolic substances), potentially influencing the heterotrophic organisms that utilize the biomass and its biorefinery value (Attard et al. 2018). New insights into the effects of thinning methods on forest biodiversity could also be obtained by better integrating silvicultural, physiological, and molecular knowledge and methodologies. The incorporation of ecological concepts such as niche and disturbance theories (Dornelas 2010; Godsoe et al. 2017) will continue to be important in the creation of theoretical frameworks for hypothesis testing in studies on thinnings’ effects on biodiversity.

The impacts of thinning operations on forest-associated organisms are likely to be highly specific and to exhibit spatiotemporal variation. It should also be noted that the reported effects of thinning operations on diversity tend to depend on the time since the operations (Verschuyl et al. 2011). It has been stated that harvests may have negative short-term effects on both species abundance and diversity (Wilson and Puettmann 2007), but that these effects may be neutralized or reversed in the long term due to factors such as ecological feedbacks.

Geophysical attributes

The physical properties of soil can be profoundly affected by mechanical thinning operations, depending on the type and load of the machinery and the number of passes that are made (Spinelli et al. 2010; Picchio et al. 2012). Thinning operations were found to significantly influence the soil infiltration rate and water storage capacity in a mountainous pine-oak mixed forest in China: thinning intensity had a non-linear effect on soil properties, with an intermediate intensity treatment producing the greatest improvements in the soil infiltration rate and water storage capacity (Chen et al. 2014). Spinelli et al. (2014) compared motor-manual and mechanized cut-to-length and whole-tree harvesting methods. They found that mechanized thinning may produce larger increases in soil bulk density than motor-manual thinning, but the difference is small. This is because no matter how the trees are harvested, terrain transport is normally performed using forwarders, which are the machines that apply the greatest pressures to the soil. According to Magagnotti et al. (2012) properly conducted mechanized harvesting does not affect soils more strongly than motor-manual harvesting, and has the benefit of causing less stand damage because mechanized systems are more capable of handling cut trees. In support of this view, Zhang et al. (2016) found that mechanized thinning of a Californian Pinus ponderosa plantation did not compact its soil. Similarly, York et al. (2015) observed no soil compaction due to thinnings in a mixed conifer plantation. Picchio et al. (2012) suggest that skid rails can confine soil compaction, and that careful planning is therefore necessary to minimize negative impacts, especially when working on soils or areas intrinsically prone to erosion. If repeated entries are necessary, it is better to restrict the traffic to permanent tracks instead of creating new ones (Moghaddas and Stephens 2008; Magagnotti et al. 2012). In general, soil compaction due to thinnings can be expected to be rather low. Conventional motor-manual PCT does not include collection of material and thus it does not directly influence the soil, whereas BTC in which harvesters and forwarders are used will have an impact on soil in the strip-roads. However, the severity of soil compaction due to machine entries is site-specific and depends on the slope, season of entry, soil type, and moisture (Tarpey et al. 2008). It may therefore be important to consider factors such as the design of contact organs if multiple entries are needed. Another option would be to use smaller machines, which have recently received a lot of interest due to their potential to give lower impact on soil. However, effective BCT requires high stability of machines at long crane reach (Bergström 2009). As smaller forwarders carry less loads several more turns are required to extract same volumes, which could mean that even though smaller machines are used the soil impact could be greater than of larger machines (Edlund 2012). Further research on the specific effects of machine driving when implementing BCT would be valuable to reach higher accuracy in our analysis.

In northern climates, soil moisture can be expected to increase in gaps that are exposed to rain (Muscolo et al. 2014). However, there is little evidence that thinnings cause major disturbances in soil water balance or erosion in boreal coniferous forests under current precipitation regimes. Following the soil status from May to August after a Norway spruce thinning experiment, Gebauer et al. (2011) found no apparent differences in soil water potential or soil humidity between the thinned plot (5000 trees ha−1) and the control (10,000 trees ha−1). Similarly, Goudiaby et al. (2011) found that soil water content was not affected by thinning of jack pine (Pinus banksiana Lamb.) and black spruce (Picea mariana (Mill.) BSP). Climate may, however, constrain the stability of soil water conditions in boreal forests, e.g. by changing precipitation patterns and increasing the frequency of dry periods.

Thinning operations allow solar radiation to reach the soil surface, which can influence the temperature in the soil and affect mineralization, at least in the short term. Nitrogen (N) is the most important mineral nutrient for tree growth in boreal forests, where its availability is generally considered a limiting factor. Coulombe et al. (2017) studied short-term N dynamics in response to partial harvesting treatments including commercial thinning and small gaps. They observed increases in N mineralization rates and mineral N concentrations and proportions (NO3−-N and NH4+-N) relative to dissolved organic material in gaps but not in thinnings, and linked these changes to increases in soil temperature and water content in gap treatments. Other studies reported that thinning in a coniferous stand can induce a sudden post-logging increase in NO3−-N, NH4+-N, and microbial N that quickly subsides (Thibodeau et al. 2000). Recovery of soil processes is likely to be faster the less intense the disturbance (cf. Kishchuk et al. 2016). The direct exposure of the soil surface is likely to be smaller in BCT than in PCT, implying that the effects of BCT on soil temperature (and thus mineralization) could be less pronounced at the stand level. Foliage left on the thinned areas could also reduce the soil's exposure to solar radiation.

Intensified recovery of biomass through BCT may result in losses of nutrients from forests (Bergström 2009) and influence the enzymatic machinery involved in N, C, and P metabolism. Adamczyk et al. (2015) observed that the activity of proteases, β-glucosaminidases, β-glucosidases, and acid phosphatases in the soil all increased in direct proportion to the quantity of logging residues left at a site. Bergström (2009) therefore emphasized the importance of carefully selecting potential harvesting sites and developing adequate measures to mitigate the risk of nutrient losses. For example, before transporting the harvested material to the roadside, foliage could be left and distributed within the stand to reduce nutrient losses while also increasing the efficiency of off-road transport and enhancing the harvested materials’ fuel properties (Bergström 2009; Mäkipää et al. 2015). Simulations by Mäkipää et al. (2015) showed that the negative effect of whole tree harvesting was most pronounced during the phase when the unperturbed growth rate should have increased rapidly, but the amount of available nitrogen was reduced due to removal of the harvest residues. An additional benefit of leaving foliage on-site is that its high ash content reduces the quality of biomass as fuel in combustion (Demirbas 2005; Bergström 2009). In some cases, fertilization could be used to compensate for nutrient losses due to biomass removal (Jacobson et al. 2000; Eriksson 2006; Bergström 2009).

Boreal forests contain significant stocks of carbon (C) and are regarded as a sink for atmospheric carbon dioxide (Kishchuk et al. 2016). In addition to causing immediate impacts on forest carbon budgets, thinning operations that affect the canopy and species composition of future stands are likely to have major effects on boreal forests’ long-term carbon sequestration capacity. In particular, the post-thinning tree species composition will significantly affect the carbon sequestration capacity of future stands. Conifers (Picea spp. and Pinus spp.) have been found to have higher forest floor C stocks than broadleaves (Populus and Betula spp.) (Vesterdal et al. 2013). In a study focusing on temperate forests, Nave et al. (2010) emphasized that species composition (hardwood vs. coniferous/mixed) significantly influenced forest floor C storage responses to harvesting. Specifically, they found that coniferous/mixed stands lost less forest floor C than hardwoods during harvests (Nave et al. 2010). It thus seems that no particular hazards are associated with either BCT or PCT in terms of the species composition of future stands and their carbon sequestration capacity. However, the timing of thinnings in the rotation cycle may significantly affect the carbon stock: according to Fenton et al. (2013), partial cutting is most likely to maintain a carbon stock similar to that of an unmanaged forest if it is done during the maximal biomass phase before the senescence of the post-fire cohort.

In addition to tree species and timing, an important factor to consider when evaluating a study is the forest layer that it examines. Jurgensen et al. (2012) reported that in red pine or hardwood stands, multiple thinnings did not affect the sizes of the C and N pools in the forest floor or the surface mineral soil layer (i.e. the topmost 30 cm), and concluded that stem-only removal for wildfire risk reduction and bioenergy production would have little impact on the total soil pools of C and N. However, a meta-analysis by Nave et al. (2010) showed that C stored in forest floors is more vulnerable to harvest-induced loss (with the average loss being 30%) than mineral soil C (which exhibited no significant loss). Given the high societal relevance of C sequestration, it would be desirable to create a detailed map of the short-term impacts of BCT on forest soil C stocks and to model the possible long term stand- and landscape-level impacts.

Economic attributes

The biomass production of trees in a stand depends on their absorption of light and efficiency of converting light into biomass. Larger trees in a stand normally absorb more light and use it more efficiently than smaller trees (Binkley et al. 2013). This is often cited as the reason for low thinning procedures, i.e. thinning of smaller trees with the intention of leaving bigger ones to become final crop trees. It is also a reason for performing PCT to make young stands homogenous in spacing and tree size. However, biomass production in conifers also correlates positively with leaf area (e.g. Gspaltl et al. 2013), so even small trees can contribute significantly to the total leaf area if they are numerous enough. Because BCT provides suitable conditions for larger trees until later cuttings while also allowing smaller trees to continue growing, it can support high stand-level biomass production. As a silvicultural technique, BCT allows for flexibility in determining vertical and horizontal stand structure. However, effective utilization of BCT requires further analyses and new thinking about how to construct thinning guidelines. Vertical stand structure and leaf area could be useful parameters in such guidelines, complementing the conventionally used parameters of basal area and tree height. They could also be useful for characterizing stands’ biodiversity and ecosystem services potential.

The possible biomass revenues after applying BCT in young dense stands are highly dependent on species and merchantable tree sizes. For example, in Northern European countries at present, pulpwood logs must have a top diameter under bark of ca. 5 cm and a length of ca 3–5.5 m, and different species may give different prices. If the biomass is instead utilized as biofuel, there are no species or size restrictions. For example, selective thinning of 30% intensity in 30–35-year-old birch-dominated stands yielded 1.6–2.4-fold higher biomass removal and 14% to 75% higher revenues when whole trees were harvested for biofuel compared to harvesting for pulpwood (Di Fulvio et al. 2011). Additionally, in a 24-year old stand, harvesting for biofuel increased the biomass removal five-fold and the revenue almost three-fold compared to pulpwood harvesting. Only trees of >5 cm DBH (diameter at breast height) were considered in this analysis; the reported effects would have been even greater if smaller trees had been included. BCT with suitable harvesting techniques in dense conifer stands permits the harvesting of very small trees, which may increase the biomass extracted per ha. (Bergström et al. 2010a, 2010b). In addition, the yield of extractives such as fatty and resin acids is significantly higher in non-pulpwood fractions such as bark, fine branches and foliage (Backlund 2013).

According to Bergström and Di Fulvio (2014), BCT with new felling and handling technologies would reduce supply costs by up to 12–15% and reduce energy requirements by 22–32% compared to conventional supply systems. These effects would increase with decreasing tree sizes (denser stands). The same authors also show that the whole tree revenues at road-side landing achieved by using BCT systems are comparable to the cost of performing a motor-manual PCT in stands with a cut tree stem volume of ca. 10 dm3 or less. Thus, changing the thinning strategy converted a cost into a source of revenue, and enabled the extraction of biomass instead of thinning with no outtake. Therefore, in economic terms, BCT allows cost-effective young stand management and extraction of valuable biomass (i.e. revenue) in early stages of rotations without jeopardizing major ecosystem services. These benefits of operational BCT could be increased further if new measures of stand value were available to support operational management. Karlsson et al. (2015) showed that management regimes that included a BCT instead of a PCT generally yielded higher land expectation values, and overall harvest yields of pulpwood and timber were not greatly affected by the silvicultural regime. More studies are, however, needed to better understand the post-BCT wood production potential during the whole rotation period.

Human environment attributes

Gundersen et al. (2016) presented a thorough survey-based analysis on the effects of bioenergy extraction on visual preferences in boreal forests. One of their central findings was that the quantity of logging residues has important effects on people's perception of managed forests: logging residues are seen as a negative element that is associated with poor stewardship and disturbances and contributes to unattractive messiness (Gundersen et al. 2016 and refs. therein). Open stands such as those formed by cuttings that preserve shelter trees are generally considered to be a positive trait because they permit high visibility and clear sightlines, enhancing feelings of safety (Holgén et al. 2000). While some studies (Rydberg and Falck 1998) have reported a preference for dense stands, dense young stands do not usually score highly in preference tests (Gundersen et al. 2016 and refs. therein). Stratification has been identified as a positive attribute in many studies but not in all (Tyrväinen et al. 2003; Gundersen et al. 2016). BCT treatment may thus provide both positive and negative visual cues. The corridors it creates may be appreciated because they increase accessibility and visibility in certain directions, but the dense stand structure between corridors could be experienced as an unsafe environment that is uninviting for recreational purposes. Since BCT involves removing cut trees, it should generally improve walking conditions when compared to PCT. In northern areas, snowmobile traffic related to reindeer herding could benefit from openings in young stands, but visibility features are highly dependent on corridor width, and the ca. 1 m wide openings could be harder to locate among the trees than broader machine-width strips. However, as BCT is based on mechanized systems, the narrow ca. 1 m-wide boom-corridors will always will be aligned in a spatial pattern system of 3–4 m-wide strip-road openings.

Ecological resilience traits

Abiotic disturbances in boreal coniferous forests are usually due to high winds, frost and snow, or wildfires (Moore and Allard 2011). Whether flexible BCT regimes increase or reduce the vulnerability of coniferous forest is likely to vary considerably depending on the general environmental conditions, the type and intensity of disturbances, and the time that has passed since the disturbance episode. For instance, the higher structural heterogeneity in stands subjected to BCT (Ahnlund Ulvcrona et al. 2017) could buffer the negative effects of abiotic stressors such as frosts or moisture imbalances. However, under certain conditions (e.g. during winters with heavy snow cover), the dense stripes left by BCT could efficiently capture snow, increasing the risk of snow damage (Valinger et al. 1993). Trees in young stands with a DBH ≤ 80 mm have been suggested to have an increased risk of damage (e.g. broken stems or windthrow), especially after the thinning of dense stands (Abetz and Klädtke 2002). Similarly, during severe drought, dense patches could promote and intensify the spread of wildfires (Song and Lee 2017). The reduced stem density created by conventional PCT can act as a structure that suppresses the spread of fires, but logging residues left in the forest could provide fuel for wildfires (Granström and Niklasson 2008). It should be emphasized that predictions regarding the effects of thinning regimes on the resilience of future forests are highly uncertain because climate change is altering natural disturbance regimes (O’Hara and Ramage 2013) in a highly unpredictable manner. In northern forests, for instance, the probability of heavy rain and drought episodes is expected to increase in future (EEA 2017).

It is difficult to generalize the expected influence of BCT on the risk of biotic damage in young and future forests: prognoses are complicated because different pathogen and pest species have very different spatiotemporal habitat and substrate requirements. Some trends can, however, be assumed based on case studies. For instance, Hood et al. (2016) found that thinning increased the resistance of ponderosa pine to both wildfire risks and the Mountain Pine Beetle (MPB). They also reported a strong density effect at the individual level: trees in thinned treatments exhibited a greater basal area increment and larger resin ducts than those in the control and burn-only treatments (Hood et al. 2016). These changes in tree character may explain the increased resistance to beetles (Hood et al. 2016; see also Kolb et al. 2007, and Fettig et al. 2014). If fertilization is used to compensate for removal of biomass through BCT (Bergström 2009), this may have consequences for the defensive capacity of the trees, often reducing allocation of resources to defences at the cost of vegetative growth (Edenius et al. 2012). The low vitality trees that are left standing in BCT may be susceptible to pests and pathogens and support the spore load or pest population at stand level. On the other hand, the stratification and vertical complexity of the stand structures created by BCT might better protect ecological networks that could strengthen the trees’ resistance to biotic damage.

Because it supports connectivity and reduces fragmentation, BCT would not be expected to have pronounced negative effects or to impose any great bias on genetic resources in tree populations (cf. Bacles and Jump 2011). It can also be expected to effectively maintain the ecological networks connecting organisms. However, to our knowledge there have been no empirical studies addressing this issue. In general, a thorough understanding of species’ ecology and evolution at the genetic level is needed to predict environmentally driven trends in the gene pool (Bacles and Jump 2011), especially in complex and dynamic forest ecosystems.

Conclusions

Climate instability and the increasing diversification of targets for forest management necessitate the development of flexible silvicultural strategies that allow adaptive management and support the provision of diverse ecosystem services by forests (Gunn and Buchholz 2018; Sing et al. 2018). Although this review could only cover a fraction of the relevant literature, it illustrates the complex effects of structural heterogeneity in young forests on the functions of forest ecosystems and hence the provisioning of ecosystem services from future forests. There are clearly large gaps in our current understanding of the specific effects of variable boom-corridor thinnings on the different dimensions of sustainability, especially during the whole forest management circle. Nevertheless, BCT may allow us to in a cost-effective and flexible (Ahnlund Ulvcrona et al. 2017) way shape stand structure, and the heterogeneity in forest structures is likely to support a broad range of habitats for different organisms. Because biodiversity, human well-being, and the success of the bioeconomy are interlinked (Braat and de Groot 2012; Bennett et al. 2015), further studies in this area are needed. An ecosystem services approach is needed to identify synergies and potential conflicts between different objectives and to design silvicultural measures to enable sustainable utilization of forest resources. A major challenge in evaluating the sustainability of new silvicultural methods is that traditional sustainability criteria cannot meaningfully describe and address spatial and temporal complexity (Wolfslehner et al. 2016; La Notte et al. 2017). Future studies in this area should therefore adopt cross-disciplinary approaches and concurrent theoretical frameworks such as nexus thinking (Fürst et al. 2017), i.e. recognition of the “big picture” and the interconnectedness between human actions (e.g. spatial planning), environmental (climatic) variation, and the biological responses at different functional levels (including consideration of ecological networks) that operate in young, dense forests now and in the future.

Acknowledgements

We are grateful to Dr. Kristina Ahnlund Ulvcrona for her valuable input in the early stage of the work. We also thank SEES-editing Ltd., North Somerset, UK, for revision of the English language.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Johanna Witzell http://orcid.org/0000-0003-1741-443X

Additional information

Funding

The research was supported by the Swedish Research Council Formas [Svenska Forskningsrådet Formas], grant number 2015-1790, and the EFFORTE project, which has received funding from the Bio Based Industries Joint Undertaking under the European Union’s Horizon 2020 research and innovation program under grant agreement No. 720712.