Management of oak forests: striking a balance between timber production, biodiversity and cultural services

ABSTRACT

Identification of the ecosystem services provided by oak-dominated forests in southern Sweden is a prerequisite for ensuring their conservation and sustainable management. These forests seem well-suited for multiple-use forestry, but knowledge is limited regarding how to manage them for multiple uses. Management for the production of high-value timber species like oaks and management to conserve biodiversity, or for cultural services can be in conflict with each other. This study evaluates the capacity of three contrasting management regimes to provide societies with economic revenue from timber production, habitats for biodiversity and cultural services, and the study analyses associated trade-offs and synergies. The three regimes were: intensive oak timber production (A), combined management for both timber production and biodiversity (B) and biodiversity conservation without management intervention (C). We synthesized relevant scientific literature, governmental statistics and grey literature. Our assessments identified that Regime A provided the highest levels of economic returns and the lowest level of biodiversity. Regime C provided higher levels of habitat provision but at expense of wood production and cultural services. In contrast, Regime B provided a balanced delivery of timber production, biodiversity conservation and cultural services. We identified several stand-management options which provide comparatively synergistic outcomes in ecosystem services delivery. The use of these management options in combination with more traditional stand-management approaches may be a more effective means of achieving sustainable forest goods and services.

EDITED BY Nicholas Brokaw

KEYWORDS:

• Ecosystem services
• oak forests
• Sweden
• synergies
• trade-offs
• multiple-use forestry
• silviculture

1. Introduction

Forest ecosystems dominated by oak (Quercus spp.) are common throughout Eurasia and the Americas, where they are valued for providing ecosystem services (Johnson et al. 2009). Oak-dominated forests provide high value timber for industry, biomass for bio-energy production, key habitats for biodiversity and valued environments for recreation and other cultural services. The recognized capacity of oak-dominated forests to provide for multiple ecosystem services aligns well with growing societal expectations that production forests are managed for multiple products and services (Gustafsson et al. 2012; Schwenk et al. 2012). Referred to as multiple-use or multi-functional forests, these forest lands are managed in a way that recognizes the importance of balancing non-timber values with sustainable timber production (Thompson et al. 2011). Within their native range, oak-production forests appear well-suited to meet the challenge of multi-use forest management, and in the northern edges of their distribution, can also aid in the adaptation of forest management to warmer and possibly more extreme climates (Bolte et al. 2009; Felton et al. 2010a; Löf et al. 2012). Despite this potential, current knowledge is limited regarding how to manage oak forests for such multiple values.

Figure 1. Distribution of standing volume of oak in southern Sweden south of ‘Limes Norrlandicus’, the northern limit of the temperate forest zone. The data are derived from Swedish University of Agricultural Sciences and the Swedish National Forest Inventory (SLU-SNFI).

In Sweden, oak forests are limited to the south, temperate region of the country, and have been widespread in the region since the Holocene (Figure 1). Currently however, due to a variety of biogeographical changes and anthropogenic impacts, oak comprises only 2% of the standing volume (Swedish Forest Agency 2014), and is much less abundant than in previous centuries (Lindbladh & Foster 2010). Instead, and partially as a result of conifer-dominated production forestry, the landscape of southern Sweden primarily consists of a mixture of conifer plantations and agriculture (Löf et al. 2012). Most of the remaining oak forests are managed or have been managed for timber production by non-industrial private forest owners in a variety of ways. Given appropriate management, these stands can produce highly valuable timber and other wood products providing substantial economic returns (Werner et al. 2000). Given that a substantial percentage of oak wood used in Sweden is imported, there appears to be untapped potential for the Swedish oak resource (Nylinder et al. 2006).

In southern Sweden and elsewhere, managed oak stands also provide a range of benefits in addition to economic returns and wood products. For example, stands of oak are among the most preferred forest habitats for recreation, especially near urbanized areas (Norman et al. 2010). Furthermore, these forests are also considered important environments for biodiversity conservation (Götmark 2013). With the vast majority of Sweden’s forest outside of protected areas (Swedish Forest Agency 2014), production forests are necessarily enlisted as part of biodiversity conservation efforts. Because of the high biodiversity value of the oak forests which remain (Berg et al. 1994), these forest areas often receive disproportionate pressure from the state to be set aside exclusively for conservation, which often leads to conflict between forest owners and the authorities (Götmark 2009). Irrespective of these conflicts, the protected area consisting of oak-dominated forest lands is expected to increase (Götmark 2013). Management for the production of high-value timber species like oaks and management to conserve biodiversity can be in conflict with each other. This paper examines the scientific basis for positive synergy between managing both for high-value oaks and for biodiversity.

There is thus a range of distinctive and potentially competing or synergistic societal goals influencing the management of oak production stands, but there is little information available regarding the relative capacity of different management regimes to simultaneously balance the habitat requirements of forest biodiversity and delivery of associated ecosystem services for societal wellbeing. By reviewing the available literature, we attempt to fill some of the relevant knowledge gaps by contrasting three distinctive management regimes for oak forests in southern Sweden, two of which are currently applied while one may become a valuable alternative in the future. Our objective is to evaluate the capacity of these alternatives to provide societies with timber (including economic revenue), habitat for biodiversity, and cultural services, while analyzing associated trade-offs and synergies. We hope our conclusions stimulate the development of new management guidelines for oak-dominated lands in southern Sweden and elsewhere that are more effective at fulfilling multiple societal goals. We also hope that our study stimulates field research comparing delivery of ecosystem services by alternative forest management plans, including our proposed new management regime. Such comparative field experiments would provide a platform for more formalized approaches to ecological and socio-economic valuation of ecosystem functions and services (De Groot et al. 2002; Fontana et al. 2013).

2. The southern Swedish context

The two oak species found in southern Sweden, Quercus robur and Q. petraea, overlap considerably in range, forest types and ecological characteristics and we consequently do not differentiate between them in this article. The history of oak in this region is defined by a gradual decline over recent millennia in the prevalence of oak-dominated forests (Lindbladh & Foster 2010). Today, oaks are distributed south of Limes Norrlandicus, the northern limit of the temperate forest zone (Figure 1). Oak stands are mainly found on sites with favorable climate and soil conditions, often in the transition between farmland and coniferous forest, but also on nutrient-poor and dry sites in coastal areas (Diekmann 1999). In the region, around 60.000 – 70.000 ha consists of oak forests (>50% oak by basal area), but a considerable and additional proportion of oak occurs as scattered trees in pastures and as mixtures with other tree species in forests (Almgren et al. 1984; Swedish Forest Agency 2014). Common mixtures are oak/Norway spruce (Picea abies) and oak/Scots pine (Pinus sylvestris) with or without other broadleaf species, such as birch (Betula spp.), aspen (Populus tremula), beech (Fagus sylvatica) and lime tree (Tilia cordata), or mixtures of oak and other broadleaf species (Drössler et al. 2012). Many of these pure oak or oak-rich mixed forests derive from historical land uses such as woodland pasture, coppice woodland or, more recently, abandoned fields and pasture (Götmark 2013). The mean annual increment in pure oak stands ranges between 3 and 6 m³ ha⁻¹ year⁻¹ on relatively good sites, with top heights from 20 to 26 m after 100 years (Carbonnier 1975).

Most temperate broad-leaved forests are owned by small private forest owners (Löf et al. 2012), and although these forests may produce highly valuable timber, wood production often only plays a minor role in the economy of each forest owner and in the overall forestry economy due to relatively slow growth and the limited timber resource. The associated industry (saw-mills, furniture manufacturers, etc.) is under-developed, and attempts to improve the timber resource are hampered by the small size of stands, their scattered placement in the landscape and associated harvesting inefficiencies (Werner et al. 2000). In contrast, the forest industry is heavily dependent on the intensive, production-oriented management of a limited number of highly productive tree species, resulting in a forest landscape in southern Sweden dominated by Norway spruce, and to a lesser degree, Scots pine and birch (Swedish Forest Agency 2014). Norway spruce is heavily utilized due to its rapid growth, ease of establishment and management, and its lower degree of palatability to browsing ungulates (Bergquist et al. 2009).

Policies relating to temperate broad-leaved forests have primarily concentrated on protecting remaining stands from conversion to other land uses, primarily Norway spruce-dominated forestry. As a result, forest owners are not allowed to convert oak forest to conifer forest. To compensate forest owners for the associated silvicultural costs of managing such stands, the regeneration and early stand management of these stands are subsidized (Löf et al. 2012). In addition, throughout the production–forest matrix, the conservation of temperate broad-leaved forests is encouraged by the Forestry Act, which requires timber and environmental values to be given equal consideration. These requirements and the widespread adoption of third-party forest certification programs have resulted in the use of green tree retention practices which often prioritize retention of broadleaf tree species (Simonsson et al. 2015).

Additional policies of relevance to oak stand management include the Right of Public Access which allows people to freely visit and experience forest environments regardless of forest ownership (Bergfors 1990). This right is considered essential for the outdoor recreation experiences of Swedes. As a result a typical Swede visits forests for recreation at least once every two weeks, and over 40% of the Swedish population would prefer a shorter distance to the forest from their homes (Lindhagen & Hörnsten 2000). Oak trees and forests are associated with a range of aesthetic, symbolic, religious, recreational and historical values (Garrido 2014) and are considered important in providing stress relief and other health benefits (Annerstedt et al. 2010). Broad-leaved forests are generally favored by the public; however, mixed forests are the most preferred (Nielsen et al. 2007; Norman et al. 2010).

3. The three management regimes

All three management regimes considered are either already applied or may be feasibly applied to the remaining oak-dominated forest lands in southern Sweden that originate from afforestation or reforestation of abandoned agricultural fields or grazing pastures, or from oak-rich woodland pastures. To simplify comparisons, we assume that that the three regimes all have the same starting point. Thus, when the regeneration is established, any living or dead trees are removed but there is a sparse layer of shrubs. This approach is adopted not to strictly mimic realistic establishment circumstances, but to minimize confounding variables when making our comparisons. In all cases, the dense (2000–5000 seedlings per hectare) natural or artificial (direct seeding or planting) regeneration of oaks is initially developed under open conditions. Normally, other species of trees and shrubs also colonize the area through natural regeneration so that an oak-rich mixture is developed with a composition depending on the surrounding seed sources.

Two of the management regimes considered (A and C) are already typical for southern Sweden, whereas the third (B) is hypothetical (see Figure 2). Alternative A is targeted at the production of high-value oak timber and follows contemporary silvicultural practices for oak management, mainly with the objective of maximizing the economic return. Alternative B is also targeted at the production of high-value oak timber, but combined with concerns for biodiversity (Jensen & Skovsgaard 2009; Wilhelm & Rieger 2013). Alternative C is targeted at the conservation of biodiversity, without management interventions (Götmark 2013). From a Swedish policy perspective, Option B could become increasingly attractive to forest owners in the future. In this study, A serves as a reference condition for evaluating the production potential of B and C, and C provides a reference condition for evaluating the biodiversity potential of A and B. The production potential and economic return will be evaluated for the entire rotation period, whereas the biodiversity and cultural potential will be evaluated mainly at the end of the rotation period, that is, at the age of 120 years or more. The fact that C stands will undergo natural processes of growth, disturbances and decay processes after this period is nevertheless acknowledged, something that can have important positive consequences for biodiversity.

Figure 2. Schematic illustration of the three management regimes in oak forests at the end of the rotation for which the crowns of crop trees are indicated with light green circles and areas without management intervention in dark green. The crowns of harvested crop trees occupy the majority of the stand in Regime A, parts of the stand in Regime B and are absent in stand C for which no trees are harvested. Each regime is also illustrated with photos. The A regime alternative that lacks an understory is illustrated. Photos: Magnus Löf, Lars Drössler and Jörg Brunet.

Option A is targeted at the production of high-value oak timber over the whole area of the stand (Figure 2) and corresponds to contemporary silviculture practices for oak in Sweden (Carbonnier 1975). In the young stand, management Regime A involves early pre-commercial thinnings to remove undesired tree species and wolf oak trees and to promote the growth of the remaining oaks. Trees that do not interfere with oaks of good stem quality are usually not removed. With time and growth of the oaks the share of other tree species usually decreases so that, at the end of the rotation, oak trees dominate the overstorey. At the age of 30–40 years, when the stand is around 10–15 m tall and the oak stem density has been reduced to around 500 trees per ha, 60–100 potential future crop trees per hectare will be selected and marked (Almgren et al. 1984). Subsequently, regular thinnings will be conducted to promote the selected potential oak crop trees to ensure their optimal crown development and diameter growth of stems. If other tree species are not interfering with the crowns of the oaks, they may be kept as an understorey. This understorey may be dominated by broad-leaved tree species or Norway spruce and may help prevent the development of epicormic shoots on the oaks. At the age of 30–40 years and onwards, epicormic shoots may be pruned at regular intervals to produce premium quality timber. After a rotation period of 120–150 years, around 50–70 of these oaks dominate the overstorey and remain in the final crop, which may or may not possess a well-developed understorey. By then, the oaks will be around 25 m tall, have a crown diameter of 10–14 m, a target diameter at breast height of 60–70 cm and a clear bole of 6–8 m. In this option, only small amounts of dead wood or dead trees are present during the rotation due to frequent management interventions that remove trees. From the management, some slash and thinning residues will be left on site.

Option B (Figure 2) is targeted at the production of high-value oak timber and biodiversity conservation. To achieve this, the production aspect is limited to a subsection of the stand, with the rest of the stand left for natural development without management interventions (Jensen & Skovsgaard 2009; Wilhelm & Rieger 2013). The production-dedicated areas of the stand are established and maintained essentially in the same way as in Option A, but with fewer crop trees (for example, 15–35 per ha) managed for timber. Regular thinning will occur only to promote the timber crop trees. At the end of the rotation, after 120–150 years, the managed oaks will occupy approximately 20–70% of the overstorey. If other tree species are not interfering with the crowns of the timber oaks, they may be kept as an understorey. The unmanaged mixed parts of the stand will gradually become denser and contain more dead wood due to competition and self-thinning. Here, oaks and other light-demanding tree species may suffer from competition during stand development, and at the end of the rotation more shade-tolerant tree species can be expected to dominate the overstorey in these parts of the stand. Canopy gaps may occur due to disturbance from wind and tree diseases, but will most likely remain small and close rapidly. Overall, the stand will likely be denser than in Option A, but will contain gaps especially where the managed oaks grow.

Option C is targeted at the conservation of biodiversity, and the whole stand is left for natural forest development without management interventions (Götmark 2013). Competition and lack of regular management interventions within this dense mixture will disfavor oaks and other light-demanding species. Therefore, the proportion of crown cover provided by oaks after 120–150 years will most likely be lower than in Option B. Similar to the unmanaged parts of Option B, some small gaps may temporarily be present, and these stands will be denser and more dead wood will be present compared to Options A and B. However, in the absence of major disturbances or targeted intervention, dead wood accumulation is a slow process requiring extended time periods (Vandekerkhove et al. 2009).

4. Assessment of the management regimes

4.1. Timber production

A larger variety of marketable wood (small – and large diameter timber) will be produced and harvested per area in management Regime A compared to Regime B. In Regime C, no marketable wood will be harvested.

We evaluated the economic consequences of the three management regimes (Table 1) in terms of economic indicators such as cash flow (age at first positive cash flow and pay-back period of initial investment), net stumpage value (accumulated undiscounted cash flow) and internal rate of return (the discount rate at which the net present value of the investment equals zero) (Klemperer 1996). The net stumpage value of thinning revenues relative to the total revenue (thinning and final harvest) was used as a supplementary indicator. The scenarios presented here for production of high-quality oak timber all rely on suitable site conditions and a continuously optimal management throughout the rotation. Deviation from these conditions may reduce the economic outcome.

Table 1. Rough estimates of economic indicators for scenarios under Option A: oak managed for high-quality timber with a target diameter (dbh) of 70 cm and an expected rotation length of 120 years, with or without an admixture and with or without pruning. Legend: CF_pos = cash flow positive from the indicated age-class, Pay-back = the number of years before complete pay-back of establishment costs, NSV = net stumpage value (NSV_thin refers to thinning revenues), IRR = internal rate of return (IRR_marg refers to IRR of the marginal investment of a specified management action). Missing values (–) indicate that no specific information was available. All estimates are approximations and based on analyses by Hermansen (1956), Holten (1980, 1986), Ståål (1986), Staun (1989), Jensen (1989, 1993), Madsen (1991), Lindén (2003), Skovsgaard (2004), Lomholt (2006) and Jørgensen (2013).

Admixture	Pruning	Pos. CF (age)	Pay-back (years)	NSV k€/ha	NSVthin/ NSVtotal	IRR %	IRRmarg %
None	No	50	90	50–60	.40–.50	2.5–3.0	–
Conifer	No	25	40	55–65	–	3.5	–
None	Yes	50	100	55–75	.25–.35	–	2.0–3.5

Note: The examples all include selection and marking of 50–100 potential future crop trees per ha. Estimates of NPV are updated to the price level of 2015.

The most relevant alternative silvicultural scenarios for timber production under Option A include the possible admixture of other tree species (hardwoods, conifers or both), the use of permanently marked potential future crop trees and high-pruning. A stand managed without an understorey and without high-pruning is considered the base scenario against which the economic outcome of other alternatives under Regime A will be compared (Table 1). The admixture of one or more hardwood species (instead of a conifer) does not significantly change any of the economic indicators and this option is therefore not considered in the comparison.

The cash-flow profile of the base scenario for Regime A is strongly skewed because of establishment and tending costs that are not compensated for by early thinning revenues. In contrast, the admixture of a fast-growing conifer (for example Norway spruce) provides early revenues from timber harvesting, which provides a potentially positive cash flow after 25 years. After approximately 50 years, net income from the oak takes over as the conifer trees are gradually removed from the stand. The pay-back period with an admixture of Norway spruce may be less than half of the base scenario.

The net stumpage value of the investment in growing oak for high-quality timber in heavily thinned stands under Regime A is approximately 55,000 € ha⁻¹ or more, depending mainly on market conditions and the quality of the timber at final harvest (Table 1). The ratio of revenues from thinnings to that of the total revenue from thinnings and final harvest depends heavily on the quality of timber and on prices for premium timber and lower classes of grading. This is indicated by the variation in NSV_thin/NSV_total (Table 1)

The magnitude of the added revenue from an admixture of conifers obviously depends on the growth and proportion of conifers in the stand and on the final quality of the oak stems. The example in this review has been estimated to provide an additional 10% compared to a pure stand of oak (Madsen 1991). The main economic effect of the conifers, however, is improved (positive) cash-flow earlier in the rotation. Mixed stands (i.e., oak and spruce) are also considered less sensitive to changes in economic conditions than monocultures (Lindén 2003).

Based on sources used in this analysis (see references in text for Table 1), the internal rate of return on the investment in the base scenario has remained relatively constant at 2.5–3.0% since the 1950s, indicating a low investment risk. Considering changes in technologies, work salaries and society at large during this period, this can also be taken as an indication of a stable balance between supply and demand in the market for oak timber.

The main consequence of high-pruning is the increased revenue that results from the improved quality of final crop trees. High-pruning remains profitable at an internal rate of return of 2.0–3.5% when considering only the added value on the oak timber (Jensen 1989, 1993; Skovsgaard 2004). This is based on conservative estimates and corresponds to an added income of 20% or more from pruning.

Finally, the practice of leaving a few retention trees at regeneration as prescribed by Swedish legislation is considered of no or only marginal importance for the economic outcome of management Option A. Retention trees in oak stands will usually be selected among those of inferior stem quality or located at stand margins and are consequently of little monetary value.

There is essentially no information available regarding the actual costs and incomes of management Regime B or the factors upon which they depend. Option B is currently being explored in some commercial forest stands in the states of Rheinland-Pfalz and Saarland in Germany (Wilhelm & Rieger 2013). In addition, in Denmark three thinning experiments in young oak include treatments similar to Option B (Jensen & Skovsgaard 2009). These experiments were installed in 2002–2003 and no results are as yet available. Nevertheless, we can expect that the reduction in income will be proportional to the number of final timber trees negatively affected by Option B plus some additional costs due to affected logistics during forest operations (Wilhelm & Rieger 2013). In the event of a larger impact comprising half or more of the potential final timber trees, the income reduction would probably be marginally larger than the actual share of negatively affected potential final timber trees. In addition, no timber trees will be managed or extracted from management Regime C which is why this regime will produce zero costs and incomes.

4.2. Biodiversity

4.2.1. Vascular plants

By increasing light availability and providing disturbed ground for plant recruitment, canopy thinning of oak forest (Regime A) generally increases understorey plant species richness. This increase may occur at least up to a decade into the rotation (Brunet et al. 1996, 1997). Also, partial cutting, corresponding to Option B, will probably have a clear effect on the herbaceous layer and will likewise increase plant species richness during the first growing season after cutting (Götmark et al. 2005). Most typical forest plants are tolerant of canopy thinning, or may even increase in abundance or frequency with its occurrence (Brunet et al. 1996; Götmark et al. 2005). The abundance of many other plant species (ruderal, grassland and habitat generalists) also increases with thinning intensity, which results in a short-term increase of total species richness with increasing management intensity (Brunet et al. 1996, 1997). The composition of the vegetation emerging after cutting also depends on previous land use history and its effects on the seed bank, and on patches of remnant grassland vegetation in the forest (Milberg 1995; Jonason et al. 2014).

Conversely, the gradual loss of light-demanding herbaceous species under closing canopies in unmanaged forest (C) is usually not compensated by an equivalent gain of shade tolerant plant species, which results in an overall decrease of species richness (Von Oheimb & Brunet 2007). Likewise, when comparing different types of oak-hazel woodland, Hansson (2001) found that areas managed as wooded meadow or wood pastures maintained a significantly higher total number of vascular plant species than unmanaged woodlands. Herb layer plants greatly vary in abundance along the light gradient (Tyler 1989), and we conclude that total vascular plant diversity at the end of the rotation period is maximized in oak forests with both open and closed canopy parts. Therefore, vascular plant diversity is probably greatest under Regime B which provides the highest structural heterogeneity due to partial understorey clearing (Table 2).

Table 2. Relative value of the different management options in relation to their promotion of structures important for biodiversity. Maximum value for positive effects at stand age of ca 120 years is +++.

Stand characteristics	Option A	Option B	Option C	Species groups
Stand structural heterogeneity	+	+++	++	Plants, Epiphytes, Birds
Tree species diversity	+	+++	++	Epiphytes, Birds, Beetles
Live trees with cavities and dead branches	+	++	+++	Birds, Beetles
Dead wood	+	++	+++	Beetles, Saproxylic fungi

Similarly to the herbaceous vegetation, the diversity of regenerating woody plant species in the herbaceous layer generally increases after canopy thinning (Brunet et al. 1996). Concerning oak itself, seedling density was found to be threefold higher in partially cut oak stands (Option B) than in unmanaged control plots (Option C) three years after cutting, and survival and growth were positively related to canopy openness (Götmark 2007). Repeated understorey cutting may lead to changes in the composition of the woody understorey and may favor species with rapid and vigorous sprouting such as hazel and lime tree (Leonardsson & Götmark 2015).

4.2.2. Lichens and bryophytes

Previous studies have found that the thinning of trees and shrubs around large oaks (Options A and B) helped to maintain the species density of epiphytic lichens and bryophytes on oak trunks, including several species of conservation concern, relative to unmanaged forest (Nordén et al. 2012). In a related study by Paltto et al. (2008), the composition of lichens and bryophytes on tree stumps shifted towards species adapted to drier dead wood as a result of thinning, implying an increase of lichen richness and a trend of decreasing bryophyte richness. Similar trends were also recorded on tree logs. However, no increases or decreases in species of conservation concern could be detected (Paltto et al. 2008). Similar to vascular plants, thinning had a positive effect on bryophytes associated with the forest floor (Götmark 2013). We conclude from these studies that overall epiphyte diversity may be highest at the stand level in Option B, by providing the highest diversity of micro-habitats in terms of tree species and micro-climate (Table 2). The proportion of lichens compared to bryophytes will probably decrease with decreasing management intensity from A to C.

4.2.3. Birds

Bird diversity and composition is strongly influenced by the structural heterogeneity of a forest stand, and in particular the understorey density (Hinsley et al. 2009; Hewson et al. 2011). In a study of understorey importance for bird communities in the oak-rich urban woodlands of western Sweden, 90% removal of the understorey had a significant negative affect on total breeding bird species densities, while 50% removal did not affect densities compared to the unmanaged controls (Heyman 2010). This indicates that a moderate disturbance of the understorey consistent with B (and some A alternatives) may not affect the avifauna negatively. Another Swedish study showed that closed oak-hazel woodlands derived from former wooded grassland, corresponding to Option C, supported a lower number of bird species than woodlands of similar origin yet retaining some degree of disturbance and open areas, and thus most closely resembling Option B (Hansson 2001). Even if Option B does not involve a creation of grazed and fully open sections of the forests, repeated disturbances from the thinning and management of the crop trees are probably beneficial to some species as well as for the overall diversity of the stand. To summarize, management Options C and B would produce more heterogeneous structure, with a more developed understorey, compared to A. To distinguish between B and C in this regard is more difficult. Even though C represents a less disturbed habitat, B would include more structural heterogeneity due to the juxtaposition of two relatively distinct habitat types, including unmanaged areas. Furthermore, the crop trees will probably be taller and coarser as compared to most of the trees in the unmanaged sections of the stand, creating an extra structural dimension (Table 2).

Tree species composition is also an important determinant of forest bird diversity (Poulsen 2002). Many bird species are exclusively associated with broad-leaved, coniferous or mixed forests, thereby causing an important divergence in the composition of bird communities associated with either forest type (Bibby et al. 1989; Roberge & Angelstam 2006). Several studies have shown that the bird communities within mixed stands of conifers and broadleaf trees are composed of a mixture of bird species representing both the broadleaf- and coniferous-associated fauna (Donald et al. 1998; Hausner et al. 2002). Hence, the mixed overstories of Options B and C will probably result in higher bird diversity than the base Option A, which is managed towards a pure oak overstorey (Table 2).

Old trees are another important structural feature for birds (Nilsson et al. 2001; Poulsen 2002), in particular for cavity nesters (Carlson et al. 1998). A recent study from southern Sweden comparing oak plantations of different ages with natural oak-rich forests shows that mid-age oak plantations with a woody understorey and managed according to Option A had a comparable bird diversity to that of older natural forests. However, the community composition differed primarily due to a lack of many tree hole-nesting species (Felton et al., unpublished results). Neither of the options assessed here include trees older than the stand itself, and few of the trees will develop large amounts of ‘old-tree structures’ during the 120 years considered. Nevertheless, a few trees in the unmanaged stand (C) and unmanaged parts of the stand in B can be expected to develop cavities during the time frame considered, and increasingly as time progresses, as large branches die from shading and eventually break off.

4.2.4. Saproxylic beetles and fungi

Oak is the most important tree in northern Europe for saproxylic beetles. Over 500 different species, many of them red-listed, use oak during part or all of their life cycle (Jonsell et al. 1998; Dahlberg & Stokland 2004). There is a large body of literature showing the importance of the type and quantity of dead wood for saproxylic species richness (e.g., Jonsson et al. 2005; Franc et al. 2007). Semi-natural oak-rich stands in southern Sweden contained on average 14.3 m³/ha coarse dead wood, which is twice as much as the mean value for all forests in Sweden (Nordén et al. 2004). In the same study, fine dead wood made up another 12 m³/ha. The stands evaluated in that study are similar to our models for Option C (and parts of B), namely abandoned woodland pastures. These stand types can be expected to accumulate more dead wood through self-thinning, and thus have higher beetle diversity compared to the pure production Option A. However, if logging residues are left after thinnings in A and B, this can readily benefit many species (Jonsell et al. 2007).

Stand openness is also important for many oak-associated beetles (Ranius & Jansson 2000; Widerberg et al. 2012; Gough et al. 2014). In a study of retained oaks in nine spruce production forests in southern Sweden, oaks exposed to intermediate insolation harbored more species than shaded oaks (Widerberg 2013). Similar results were obtained in an experiment in 22 oak-rich stands in southern Sweden where about 25% of the tree basal area was cut in 1 ha plots and then compared with uncut plots (Franc & Götmark 2008). Saproxylic beetle species richness increased by about 35% in the harvested plots, relative to the reference plots. The results of both of these studies indicate that Option B could be more beneficial for saproxylic species than the alternatives considered (Table 2), due to the combination of dead wood and favorable light conditions. In addition, the larger tree diversity in Options C and B will most probably be beneficial for saproxylic beetles compared to the potential monocultures of some A alternatives (Jonsell et al. 1998; Felton et al. 2010b).

Species richness of fungal fruiting bodies on fine woody debris (1–10 cm in diameter) was negatively affected by thinning but remained unaffected on coarse dead wood (>10 cm) in a study of experimental partial cutting in southern Swedish oak-rich forests. Overall species composition did not change significantly as a result of partial cutting, but red-listed species tended to decrease in thinned plots (Nordén et al. 2008). The available data are too scarce to allow general conclusions. However, the assessment made here indicates that the diversity of saproxylic fungi may be highest under Option C, where a moister microclimate is maintained and also where the most dead wood will accumulate due to self-thinning. As long as substantial parts of the stand remain unmanaged in Option B, negative effects of tree harvest may remain negligible, while Option A probably will cause a considerable decrease of saproxylic fungal diversity (Table 2).

4.3. Cultural services

The nonmaterial benefits people obtain from ecosystems are often referred to as cultural services (MEA, 2005). These services include spiritual and religious values, inspiration, aesthetic values, recreational values and cultural heritage values, and have implications for social relations and one’s sense of place. Their importance for society and thus need for integration into decision-making has been increasingly acknowledged (Bowler et al. 2010; Daniel et al. 2012). The most studied cultural services in forests are recreational and aesthetic values. Recreation values increases with the share of temperate broad-leaved trees, such as oak found within a stand (Norman et al. 2010; Johansson et al. 2014). Structural composition of the stand is, however, often considered more important than tree species composition (Edwards et al. 2012). Overall, tree size, openness and visual penetration into the stand as well as visual diversity of the stand are the most common predictors of public preferences (Ribe 1989; Gundersen & Frivold 2008).

The development stage of the stand has also been shown to be an important contributor to its recreational value. According to a Danish study (Jensen & Koch 2000), this effect is more prominent in broad-leaved stands than in conifers. Lower recreational and aesthetic values of young stands can be attributed to their relatively high density (Ribe 1989). The same effect is often seen in relation to the presence of a dense understorey and is probably linked to accessibility and safety considerations (Jørgensen et al. 2009). For forest management this suggests the potential positive effects of thinning young and medium-aged stands, especially when leaving the most vigorous and attractive trees on the site. However, high-intensity thinnings in young oak stands could also decrease the impression of accessibility due to the presence of high volumes of thinning slash on the site (Jensen & Skovsgaard 2009).

Management Regime A is likely to be preferred especially in the later stages of the stand development, providing big trees that are well distributed, open environments and visual penetration (Table 3). However, from the point of visual diversity, the structure of monocultures is much simpler than that in natural or mixed forests. Thus, the presence of an understorey may be beneficial for recreational and aesthetic values. In addition, aesthetic and recreational values also increase with presence of retention trees, if they are in good condition (Bostedt & Mattsson 1995; Tönnes et al. 2004). Regime A is the most intensively managed out of the three alternatives. Whereas public preferences regarding forest management can often vary among studies and with the characteristics of the sample population, people tend to favor managed forests containing few visible traces of human activities or large disturbances (Ribe 1989; Mattsson & Li 1994; Gundersen & Frivold 2008). Therefore, in regard to anthropogenic disturbances, clear-cuttings are believed to have the strongest and the longest negative impact on scenic beauty.

Table 3. Relative recreational and aesthetic values in relation to stand characteristics for the three management regimes. Maximum value for positive effects at stand age of ca. 120 years is +++.

Stand characteristics	Option A	Option B	Option C
Number of medium/large trees	+++	++	+
Openness	+++	++	+
Visibility into the stand	+++	++	+
Visual/structural diversity	+	+++	++
Positive anthropogenic disturbance	+	+++	+

Management Regime C may be perceived as ‘messy’ by the general public (Table 3). The presence of large amounts of dead wood (from self-thinning) coupled with a high density of trees and thus low visual penetration and absence of forest management results in negative attitudes and low recreation and aesthetic values of such environments (Hultman 1983). However, variation in preferences exist between people with different levels of knowledge about forests and their functioning (Gundersen & Frivold 2008; Eriksson et al. 2012). Potentially countering these aspects, Regime C also contains a variety of features often considered favorable, including medium to large trees and structural diversity. Fallen twigs and branches associated with older trees may diminish recreational and aesthetic values, but a number of more recent studies suggest that since the 1990s natural forests have been considered more suitable for recreation than they were in previous decades (Lindhagen & Hörnsten 2000; Tahvanainen et al. 2001). Relatedly, there seems to be an increasing acceptance and appreciation of dead wood (Tyrväinen et al. 2003; Nielsen et al. 2007; Heyman 2012). This development has been attributed to increasing levels of knowledge among the general public about biodiversity.

Management Regime B probably provides many desirable features for the general public such as medium and large trees, lower overall tree density of some sections of the stand and structural diversity (Table 3). Semi-open stands increase access to more distant areas and openings and provide a higher sense of security compared to dense stands (Kaplan & Kaplan 1989; Gundersen & Frivold 2008; Heyman et al. 2011). Additionally, management targeting only specific trees and leaving other areas untouched contributes to a patchy and visually diverse environment, thus potentially increasing recreational and aesthetic values of the forest (Ribe 1989; Lindgren 1995; Nielsen et al. 2007; Heyman 2012). Uneven-aged stands with a mixture of trees of different sizes are not only preferred over monocultures, but this effect is maintained throughout the rotation (Hultman 1983; Lindhagen & Hörnsten 2000). Overall, this regime seems to avoid and/or mitigate some of the negatively perceived features of Regime A and Regime C, by excluding large disturbances such as clear-cuttings in the end of the rotation, reducing stand density and providing higher levels of visual diversity. Thus, Regime B is expected to be slightly more favored by the public than Regime A, and much more favored than Regime C.

5. Discussion

Any management decision that alters the tree species composition, structure or disturbance regimes of a forest has the corresponding capacity to alter the range and balance of ecosystem services derived from that forest. Previous studies have highlighted, for example, that an unbalanced prioritization of intensive timber production will reduce forest biodiversity (Lindenmayer & Franklin 2002; Gossner et al. 2014), as well as the provisioning of regulatory and cultural services (Bennett et al. 2009; Raudsepp-Hearne et al. 2010). The potential for trade-offs is, however, tempered by the existence of synergistic relationships among biodiversity and some ecosystem services derived from production forests (Duncker et al. 2012). Our assessment lends support to the existence of both trade-offs and synergies among the ecosystem service deliverables of oak-dominated stands, with some management alternatives providing outcomes far more balanced in terms of the overall delivery of ecosystem services.

With respect to trade-offs, the most intensive oak-production forest management regime (A) provided the highest levels of economic returns, slightly decreased cultural values and the lowest biodiversity. Notably however, the intensive production alternative assessed did not consist of only one silvicultural option, but included a range of management prescriptions with different implications for both biodiversity and recreational values. Oak production alternatives considered in A ranged from monocultures to mixed-species forests, and also varied in terms of the extent of understorey vegetation permitted to develop. Thus the intensive production forest alternative (A) assessed varied in terms of tree species diversity and structural heterogeneity allowed for, two aspects of direct relevance to evaluations of the biodiversity and cultural value of a stand. As our assessments indicate, the biodiversity and cultural value of these stands could therefore be increased with minimal reductions to production and economic outcomes, by adopting silvicultural approaches that increase tree species composition and understorey structural heterogeneity (Bostedt & Mattsson 1995; Lindenmayer & Franklin 2002; Tönnes et al. 2004; Brockerhoff et al. 2008).

As expected, our evaluation of the free development regime (C) indicated substantially higher levels of habitat provision for a range of taxonomic groups, which came at the expense of contributions to wood production and cultural services. The long-term trajectory of oak biodiversity values in these stands is however somewhat uncertain, as it depends on the extent of canopy disturbance and associated light conditions beneficial to the natural regeneration of oak. The availability of light is a primary determinant of oak seedling and sapling growth following canopy disturbance (e.g. Jones 1959; Collet et al. 1997; Lüpke 2008; Parker & Dey 2008; Gardiner et al. 2009). For this reason the occurrence of oak has often been linked to various natural and anthropogenic forest disturbances, (e.g., livestock grazing, harvesting, fire and windthrow) that provide increased insolation for oak regeneration and stump sprouting (Vera 2000; Johnson et al. 2009; Brose et al. 2013).

There is thus a potential synergistic interaction between oak management which provides conditions beneficial for oak regeneration and associated habitats for biodiversity. If sufficient openness is maintained, due to natural or anthropogenic disturbance, to ensure the continued persistence of an oak-dominated canopy, then a range of structural features of relevance to biodiversity in oak reserves can be expected to increase as these stands mature beyond 120 years. This is due to the natural processes of growth; aging and senescence facilitating an increased abundance of structural features (e.g., large diameter dead wood, tree holes) often closely associated with high biodiversity values (Berg et al. 1994; Remm & Lõhmus 2011). The importance of maintaining such features within landscapes like southern Sweden, where land-use practices have largely reduced their occurrence, requires that oak reserves continue to be a key component of protected broadleaf areas in southern Sweden. This may require that in oak reserves, in which the occurrence of natural disturbance regimes is unlikely to maintain an oak-dominated forest system, active intervention takes place to ensure the continuation of oak-favorable successional dynamics (Götmark 2013).

In contrast to the categorically distinct production and biodiversity regimes (A & C), the ‘combined goal’ management regime (B) unites both of these management priorities. Option B appears to provide the same potential value for biodiversity as C, although over a smaller spatial extent, combined with approximately the same cultural services as A. Because of this, Option B provides forest owners with a high degree of flexibility with respect to the relative prioritization of different ecosystem services. Forest owners can for example determine the extent to which different areas of a stand are allocated to management practices more conducive to production, biodiversity and cultural values. Despite these capabilities, this management alternative is so far primarily a hypothetical construct in southern Sweden. Our assessments suggest, however, that this management alternative may be a suitable means of balancing the provision of a more comprehensive suite of the ecosystem services within a stand than the currently adopted management alternatives in this region. This balanced provision appears to be consistent with increasing demands that forests are managed for multiple goals (Gustafsson et al. 2012; Gamfeldt et al. 2013). An alternative approach would require a continuation of more categorically distinctive oak production stands and reserves that provide for the ecosystem requirements of societies at landscape scales. The landscape scale balancing of ecosystem services does, however, raise complications (Duncker et al. 2012), especially in regions such as southern Sweden dominated by a large number of small-scale forest owners (McDermott et al. 2010; Gustafsson et al. 2015).

Despite the apparent capacity of alternative B to provide suitable habitat for a substantial proportion of a region’s flora and fauna, we nevertheless wish to emphasize that protected areas are an essential part of any comprehensive framework for the protection of forest biodiversity. This is because the most demanding species often require relatively contiguous areas of complex and heterogeneous forest lands for which disturbance regimes are allowed to approximate natural patterns and processes (Hunter et al. 1988; Kuuluvainen et al. 2012; Götmark 2013). Caution is thus warranted when evaluating the capacity of production forest to provide a balanced delivery of ecosystem services. In relation to biodiversity, extensive reliance on forest management alternatives that strike an optimal balance may nevertheless compromise the regional viability of populations only sustained in protected areas.

In relation to timber production and associated economic returns, the outcome of Regime B will depend on the number of crop trees produced. Thus, Regime B will often produce less marketable timber per hectare compared to A. Therefore, an extensive reliance on only Regime B would probably reduce the economic returns from oak production in the region. The small size of stands, and their scattered placement in the landscape already increases harvesting inefficiencies and reduces available timber for the industry. Thus, Regime A and Regime B are not interchangeable, but instead are complementary management alternatives. Furthermore, if decisions were taken to manage the majority of forests similarly, the resultant homogeneity would likely decrease their recreational and aesthetic values (Axelsson-Lindgren & Sorte 1987). We would also like to emphasize that little is known about the effects of ‘combined goal management regimes’ on timber production economy.

When evaluating the relative biodiversity and cultural value of any production forest alternative, a strong determinant of outcomes is the reference condition chosen. In this study, we contrasted intensely managed oak production forests (A) with alternatives possessing a substantial (B) or encompassing extent (C) of total area allocated to unmanaged development. Whereas the intensive oak production management alternative fared poorly with respect to biodiversity in this comparison, this outcome would likely be reversed if comparison was made with the most common silvicultural alternative in the region: Norway spruce-dominated production forests. This difference only increases if the management regime allows for retention forestry and a biodiversity-rich understorey. This is because even the most intensive oak production alternative is nevertheless consistent with regional biodiversity goals for the increased use of broadleaf tree species, as advocated by the Swedish Forest Agency and forest certification programs (Gustafsson et al. 2010; Gustafsson & Perhans 2010; Johansson et al. 2013). Likewise, broadleaf forests are consistently preferred over spruce-dominated alternatives in terms of their recreational and aesthetic values (Norman et al. 2010).

Whereas oak timber production values have been relatively stable since at least the 1950s, their recreational and aesthetic values may be less stable. For example, demographic differences between various categories of users exist, even though public preferences exhibit high degree of consensus (Stamps 1999). Children find dense forests more fascinating, and mushroom and berry pickers favor more natural-looking environments (Ribe 1989; Eriksson et al. 2012). Forestry professionals and students tend to prefer medium- to high-density stands, whereas landscape architects favor low-density stands (Petucco et al. 2013). In addition, variation in preferences exist between people with different levels of knowledge about forests and their functioning (Gundersen & Frivold 2008; Eriksson et al. 2012). When respondents in Denmark and southern Sweden were informed about the importance of biodiversity, they provided higher values for its conservation (Bakhtiari 2014). Thus, values can change relatively rapidly through education, with corresponding influences on preferences for forest use.

Management for the production of high-value timber species like oaks and management for cultural services, or to conserve biodiversity can be in conflict with each other. The results from our review have identified a management regime with combined management goals for the oak-dominated forests which appears to provide a balanced delivery of timber production, biodiversity conservation and cultural services. We would, however, like to stress that insufficient studies have been conducted to determine the precise capacity of management Regime B to provide for production, biodiversity and cultural values. Instead, our assessments had to rely on information derived from management regimes which closely approximated this management alternative. To validate and expand upon our findings, targeted field experiments are necessary to determine the precise capacity of this management alternative to provide for ecosystem services of societal interest. These findings would help stakeholders evaluate the feasibility of the three options in terms of implementation capacity and their respective desirability in output of ecosystem services. This may be achieved using a systematic framework such as multi-criteria decision analysis, which quantifies ecosystem services while highlighting the contribution of selected services. (Fontana et al. 2013). Such a process of stakeholder involvement would help test the accuracy and comprehensiveness of ecosystem service projections, and in addition, may identify additional silvicultural pathways for optimizing the delivery of goods and services from oak forests at stand and landscape scales.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was funded by the Broadleaves for the Future program. Anna Filyushkina was funded by European Commission under Erasmus Mundus joint Doctoral Programme ‘Forest and Nature for Society’ (FONASO), and Adam Felton was partially funded by Future Forests, a research program supported by the Foundation for Strategic Environmental Research (MISTRA).