Creation of multi-layered canopy structures in young oak-dominated urban woodlands – The ‘ecological approach’ revisited

Urban Forestry & Urban Greening 11 (2012) 147–158

Contents lists available at SciVerse ScienceDirect

Urban Forestry & Urban Greening journal homepage: www.elsevier.de/ufug

Creation of multi-layered canopy structures in young oak-dominated urban woodlands – The ‘ecological approach’ revisited Gustav Richnau a , Björn Wiström a , Anders Busse Nielsen a,∗ , Magnus Löf b a Swedish University of Agricultural Sciences, Faculty of Landscape Planning, Horticulture and Agricultural Science, Department of Landscape Management, Design and Construction, Sweden b Swedish University of Agricultural Sciences, Faculty of Forest Sciences, Southern Swedish Forest Research Centre, Sweden

a r t i c l e Keywords: Afforestation Canopy stratification Forest structure Thinning Understory Urban forestry

i n f o

a b s t r a c t We investigated the stand structure of ten young urban woodlands established in Southern Scandinavia during the 1970s and 1980s according to the ecological approach, which advocated the use of many different species of trees and shrubs to create complex canopy structures as soon as possible after establishment to promote recreation and biodiversity. Tree height and live crown depths were measured and analysed using a combination of quantitative and qualitative approaches to assess the forest structure in terms of canopy stratification. The results show that the current canopy structures could be classified into seven different two- and three-layered structural types which had evolved as a combination of differences in management frequency and the initial species composition. Two layered stands were characterized by lower management frequency compared to three layered stands and stands in transition to three layers. They were also established with a lower proportion of understory species and a higher proportion of shade tree species. The total number of species at the establishment did not influence how stands were categorized. The two main conclusions are that recurrent thinnings is a key factor for successful management of young, species rich forest plantations, and that species composition can increase the resilience towards management neglect. Instead of aiming at maximising total species number it is more reasonable to focus on a few key species in each layer. We conclude that three-layered canopy structures can be created already after twenty five years, which should encourage planners and practitioners to incorporate multilayered stands in future urban woodland creation. © 2012 Elsevier GmbH. All rights reserved.

Introduction During the last decades, increasing the forest cover near cities has become a political priority in many European countries (Konijnendijk, 2000; Weber, 2000). Consequently the area of young woodlands, particularly oak-dominated, has increased significantly in and around towns and cities particularly in the forest poor North-Western parts of Europe (Gundersen et al., 2005; Nielsen and Jensen, 2007; Jensen and Skovsgaard, 2009). Due to the high pressure on forests in and near towns and cities, it is beneficial to take advantage of new forest plantations already during their young stages of development, both for attaining recreationally attractive environments and creating rich biodiversity habitats at an early age. Young forests are however often perceived as unattractive and unsuitable for recreation (Gundersen and Frivold, 2008; Jensen and Skovsgaard, 2009) as well as poor in biodiversity (Christensen and Emborg, 1996; Nilsson et al., 2001). This can be attributed to the lack

∗ Corresponding author. Tel.: +46 40415212. E-mail address: [email protected] (A.B. Nielsen). 1618-8667/$ – see front matter © 2012 Elsevier GmbH. All rights reserved. doi:10.1016/j.ufug.2011.12.005

of specific forest structural attributes, including horizontal and vertical canopy complexity (Niklasson and Nilsson, 2005; Nielsen and Jensen, 2007). The formation of stratified canopy structures is generally associated with the later phases of stand development while young forests have ‘simple, top-loaded, single-layered canopies’ (Franklin and Van Pelt, 2004). This progression is due to several co-interactive factors such as changes in light conditions due to small-scale gap formation and subsequent colonization of shade tolerant understory tree and shrub species. Canopy stratification has enabled forest structure to be classified and conceptualised in different ways (Parker and Brown, 2000). It has for long been an interest to many disciplines in the context of management for timber production, wildlife as well as recreation. Stratification of canopies into overstory and understory layers are of ecological importance for flora and fauna, including various insects and birds (e.g. Ishii et al., 2004; Vance et al., 2007; Gunnarsson et al., 2009). Furthermore, forest preference studies have shown that stratified canopies can be a highly appreciated feature for forest recreation (Silvennoinen et al., 2001; Gundersen and Frivold, 2008). However, the number of strata per se should not be considered as a direct measure of how appealing a certain forest stand is

148

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

(Ribe, 1989; Heyman, 2011). Multi-layered forest stand types have been considered as among the most valuable types for integration of biological and recreational qualities, and thus a keystone stand type to be included in the urban woodland context (Gustavsson, 2004). Departing from a landscape architecture tradition focusing on spatial architecture of the forest, Gustavsson (1986) and Gustavsson and Fransson (1991) developed a conceptual approach, which classified forest stands into different structural types based on the vertical and horizontal distribution of tree crowns and species. This framework allowed for a discussion of different strategies for woodland creation as an alternative to the prevailing use of even-aged monoculture plantations that was considered inappropriate for urban forestry objectives. The structural approach by Gustavsson was part of what has become known as the ‘ecological approach to landscape design’ (Thompson, 1998), which emerged primarily during the 1970s and found favour with practitioners and academics in many North European countries like Denmark, Germany, Holland, Sweden and the UK (Tregay and Gustavsson, 1983; Tregay, 1986; Forbes et al., 1997; Ruff, 2002). By drawing inspiration from natural and semi-natural temperate forest ecosystems, different tree and shrub species with varying life forms, shade tolerance, and successional strategies were mixed in the initial planting. It was argued that with proactive design and management, stratification of tree and shrub crowns into different canopy layers should be achievable already shortly after establishment (Tregay and Gustavsson, 1983; Tregay, 1986). Many of the conceptual ideas of the ecological approach became integrated in the general landscape practice in Europe during the 1990s (Thompson, 1998). Current recommendations for woodland creation include the use of multiple species including both canopy trees and understory shrub species (e.g. Blakesley and Buckley, 2010). However, there is a lack of empirical studies on how forests established according to the ecological approach have developed and how management frequency and species composition influence canopy stratification. In this study, we reconstruct the management history and species mixtures at planting in ten forest stands established according to the ecological approach during the 1970s and 1980s as part of the rapid urban developments in southern Scandinavia. The objectives were to assess the different stand structure and canopy stratification that had developed as a result of (1) differences in management frequency, and (2) differences in species mixtures at planting.

Materials and methods Study sites The study was carried out in the Öresund region of Sweden and Denmark during 2008 and 2009 (Fig. 1). Four major forest establishment projects initiated during late 1970s and early 1980s that had clear references to the ecological approach were identified as case study areas. These were (1) the Alnarp Landscape Laboratory (A) a demonstration forest located adjacent to the campus of the Swedish University of Agricultural Sciences, (2) the Bulltofta Park (B) in Malmö established on a former airfield and today one of the largest and most visited recreational areas of the city, (3) the Filborna forest (F) in Helsingborg established to serve as a green corridor connecting the inner city parks with the recreation landscape on the urban fringe, and (4) the Ishøj Nature park (I), a young frequently visited recreational landscape south of Copenhagen. All four study sites included a variety of forest stand types as well as semi-open and open areas. At each location, forest stands

established with the aim of developing multi-layered canopies were identified. Among those, all stands where the species composition included Quercus robur L. in combination with one or several understory tree or shrub species were selected for inventory. The focus on Quercus was because this species has been favoured for afforestation, and has also increasingly been replacing conifers when long-established forests have been regenerated in Denmark and Southern Sweden (Jensen and Skovsgaard, 2009). The latter is especially the case in publicly owned forests close to cities (Gundersen et al., 2005). Reconstruction of species mixture and management history Information about year of establishment and the species mixtures at planting were acquired from consultation of reports and archives in the authorities owning the individual woodlands. In total ten mixed stands with Quercus robur were identified, having varying stand sizes and original species compositions (Table 1). Henceforth the stands are referred to as A1, A2, B1, B2, F1, F2, F3, I1, I2 and I3. Validity of identified species mixture was controlled (and in two cases corrected; B2, I1) through consultation of present and former managers and field visits. For all stands descriptions of management strategy and longterm aim for stand development had been established at the time of afforestation. For the stands in Alnarp Landscape Laboratory and Bulltofta these have also been published (Qvarnström and Rosenqvist, 1980; Nielsen et al., 2005). However, information about conducted operational management since the stands were established had not been added to the management plans. Thus for all stands the timing and intensity of thinning and other operational management actions had to be reconstructed through interviews with present and former managers, together with whom we also searched in archives. The results of this reconstruction is summarised in Appendix A, and revealed that some of the stands had been thinned 1–2 times (B1, B2, F1, F2, F3, I2) indicating a neglected management, while other stands had been thinned 3–5 times (Table 1). Mixed-method design The oldest and most widely used method to study canopy stratification is the use of profile diagrams (Baker and Wilson, 2000). Profile diagrams, often combined with crown projection diagrams, have gained interest during the last century, especially where mixed-forest management has been practiced, indicating that the more irregular the stand structures are the greater need for integrative and visual tools (e.g. Nielsen and Nielsen, 2005; Nielsen, 2006). Identification of canopy layers from profile diagrams is however visually and qualitatively assessed, and has been criticised as a subjective and non-reproducible method (Parker and Brown, 2000). This has resulted in the development of various mathematical canopy stratification models that allow for a more objective comparison (Latham et al., 1998; Baker and Wilson, 2000; Everett et al., 2008). In this study we combine the two methods in order to develop a more complete and nuanced portrait of the studied stands. Data collection Data collection was carried out in study plots in all stands. The number of plots per stand was determined according to the size of the stand where stand area <0.75 ha = 2 plots, 0.75–1.5 ha = 3 plots and 1.5–3 ha = 4 plots. We chose to locate the plots subjectively to ensure that structural variations within the stands were represented. Plot size was set to 15 m × 15 m except for one of the A2 plots where a plot size of 10 m × 10 m was used to avoid

Table 1 Stand area, management history and species composition at the time of establishment. Species composition is reported as percentages of all individuals and has been grouped according to the layer in which each species was intended to be positioned. Middle and shrub layer species compositions have been combined, as it was not possible to determine the intended position of several of these species according to the documentation of original species composition. Alnarp 1 (A1)

Alnarp 2 (A2)

Bulltofta 1 (B1)

Bulltofta 2 (B2)

Filborna 1 (F1)

Filborna 2 (F2)

Filborna 3 (F3)

Ishøj 1 (I1)

Ishøj 2 (I2)

Ishøj 3 (I3)

Area (ha) Year of establishment Number of thinnings Year of thinnings Species comp.a Nurse trees

0.1 1985

0.2 1984

2.3 1983

0.5 1984

0.5 1983

0.9 1983

1.6 1987

0.9 1977

0.4 1977

0.6 1977

5

5

2

1

1

1

1

3

2

3

−90, −95, −96, −02, −03

−90, −94, −95, −97, −03

−90, −96

−98

−98

−98

−02

−91, −96, −03

−89, −01

−89, −01, −04

A. glutinosa 50%

A. glutinosa 15%

A. glutinosa 20%

A. glutinosa 10%

A. glutinosa 25%

A. platanoides 5% C. betulus 5% F. sylvatica 5% F. excelsior 5% Q robur 10% T. cordata 5% U. glabra 5%

F. sylvatica 5% P. avium 5% Q. robur 30% T. cordata 10%

Q. robur 50%

C. sanguinea 5%

A. campestre 5%

A. campestre 5%

R. canina 5%

A. platanoides 5% B. pendula 10% C. betulus 5% F. sylvatica 5% F. excelsior 5% P. padus 5% Q. robur 30% S. caprea 5% T. cordata 10% C. avellana 5%

B. pendula 10% F. excelsior 10% Q. robur 20%

Middle/shrub layer

A. platanoides 5% B. pendula 5% C. betulus 5% F. sylvatica 5% F. excelsior 10% P. padus 10% Q. robur 15% S. caprea 5% T. cordata 5% C. avellana 15%

A. glutinosa 30% S. alba ‘Liempde’ 14% Q. robur 12%

A. glutinosa 20%

A. platanoides 5% F. excelsior 2% P. communis 3% Q robur 3%

A. glutinosa 5% L. × eurolepis 10% A. platanoides 5% B. pendula 5% C. betulus 5% F. sylvatica 5% F. excelsior 5% P. padus 10% Q. robur 15% S. caprea 5% T. cordata 10% C. avellana 10%

A. glutinosa 15%

Crown layer

A. glutinosa 20% P. simonii 10% A. pseudoplatanus 10% B. pendula 5% C. betulus 10% P. avium 5% P. padus 5% Q. robur 5%

C. avellana 12%

A. campestre 10%

A. campestre 12.5%

C. avellana 20% R. alpinum 5% S. intermedia 2% V. opulus 5%

C. sanguinea 5% C. avellana 10% C. monogyna 5% L. xylosteum 5% R. alpinum 5% S. intermedia 5% V. opulus 5% 16

C. sanguinea 5% C. avellana 10% R. alpinum 5% V. opulus 5%

S. purpurea 5% S. × smithiana 5% S. nigra 5% S. albus 5% V. opulus 5%

R. alpinum 5% S. aucuparia 5%

R. alpinum 5%

S. aucuparia 5%

C. monogyna 14% P. cerasifera 2% R. multiflora 16%

C. avellana 6% C. monogyna 10% F. alnus 4% P. cerasifera 6% R. multiflora 4%

C. avellana 12.5%

10

14

14

12

12

7

10

4

Number of species

10

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

Stand

a Scientific species name with authority: Acer campestre L.; Acer platanoides L.; Acer pseudoplatanus L.; Alnus glutinosa (L.) Gaertn.; Betula pendula Roth.; Carpinus betulus L.; Cornus sanguinea L.; Corylus avellana L.; Crataegus monogyna Jacq.; Fagus sylvatica L.; Frangula alnus Mill.; Fraxinus excelsior L.; Larix × eurolepis Henry; Lonicera xylosteum L.; Popolus simonii Carriére fa. Fastigiata C.K.Schneid.; Prunus avium L.; Prunus cerasifera Ehrh.; Prunus padus L.; Pyrus communis L.; Ribes alpinum L.; Rosa canina L.; Rosa multiflora Thunb.; Salix alba L. ‘Liempde’; Salix caprea L.; Salix purpurea L.; Salix × smithiana Willd.; Sambucus nigra L.; Sorbus aucuparia L.; Sorbus intermedia Ehrh.; Symphoricarpus albus (L.) S.F. Blake; Tilia cordata Mill.; Ulmus glabra Huds.; Viburnum opulus L.

149

150

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

Fig. 1. Location of the four study areas Alnarp, Bulltofta, Filborna and Ishøj in southern Sweden and Denmark.

disturbance from forest edge effects. In each plot, tree height and height of the base of the live crown of all individuals taller than 1 m were measured with a digital clinometer (Haglöf Vertex IV) or a measuring stick. The horizontal distribution of the tree crowns was assessed with crown projections that were made by drawing the vertical projection of the outermost perimeter of all tree crowns on a graph paper in a scale of 1:100 (cf. Koop, 1989). Individuals with stems positioned outside the plot but with crowns partially inside were also included. One profile diagram on a scale of 1:100 was made in each stand on a line transect measuring 25–30 m following the methodology of Koop (1989). A tape measure was laid out through the plot and all plant individuals with tree crowns covering the transect were depicted (Nielsen and Nielsen, 2005). A 10 m high graded measuring stick was used to position major branches, while the position of higher branches was measured with a digital clinometer (Häglöf Vertex IV).

Canopy stratification analysis To identify the number of canopy layers in each stand, we used the Landscape Management System (LMS) stratification algorithm developed by Baker and Wilson (2000; available at http://lms.cfr.washington.edu/). The algorithm identifies canopy strata by comparing tree height and mean height of the base of the live crown of all taller trees. The algorithm operates by first sorting all individuals in descending order of height. Starting with the tallest tree, tree height of all individuals are compared to the mean height of the base of the live crown of all taller trees plus a constant of overlap (k0 ), to define the limit different canopy layers (see Baker and Wilson, 2000 for full description). In accordance with Everett et al. (2008) who concluded that the use of k0 caused the LMS algorithm to miss upper and lower stratum, we set k0 = 0. Furthermore, since the algorithm was not developed for young stands, single individuals were occasionally identified as an individual stratum. Therefore, to avoid identification of a misguidedly exaggerated number of strata in the lower understory, we set a lower limit of 10% canopy cover as a restriction parameter for a LMS-identified stratum to be referred to as a layer. Strata covering less than 10% of the study plot were incorporated in the next superior stratum until this requirement was met beginning with the lowest positioned stratum.

Classification of stand structure The stands were classified according to the Woodland Structural Stand Type framework by Gustavsson and Fransson (1991) (Fig. 2), where vertical stratification of individual crowns is distinguished to overstory trees (crown layer), subordinate trees and large shrubs (middle layer), and saplings and low shrubs (shrub layer). In this study, we refer to canopy as ‘all the foliage from the ground to the top of thee crowns’ in line with Franklin and Van Pelt (2004). Furthermore, in accordance with Gustavsson (1986), we define shrub layer as all foliage below 2 m, middle layer as all foliage positioned between 2 m and <50% of maximum tree height, and crown layer as the foliage of the upper part of the canopy (> 50% of maximum tree height). Classification of stands in structural types was performed using (1) results from the LMS-algorithm, (2) interpretation of crown projections and profile diagrams, and (3) analysing species composition and species distribution in the vertical profile. The results of the LMS-algorithm were used as a first step to distinguish threelayered from two-layered stands and also provided information at what height different layers separated. To further separate the stands in structural types, profile diagrams and crown projections were interpreted by visually comparing the vertical and horizontal profile to the conceptual framework by Gustavsson and Fransson (1991). Stands identified as two-layered by the LMS algorithm but visually considered in a transition stage towards three layers were classified as full-storied or low stands with a shrub layer and emergent standards. Finally, to separate the stands in different subtypes, the species compositions in each LMS-identified layer were evaluated to determine dominating species groups.

Statistical analysis In order to allow for statistical analyses, we grouped the stands into two main categories. The first group consisted of stands classified as two-layered. The second group encompassed stands classified as three-layered, full-storied and low stands with emerging standards. We justify this distinction by arguing that full-storied stands and low stands with emerging standards display a higher level of canopy complexity and can be considered as being in a transition stage to develop three canopy strata. Henceforth, the

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

151

Fig. 2. Conceptual approach of idealised woodland structural types developed from a landscape architectural perspective (modified from Gustavsson and Fransson, 1991). The framework included 8 main categories with a total of 18 different subcategories. The division was based on differences in vertical and horizontal canopy stratification as well as dominant species types. The stratification model used was based on three canopy strata defined at fixed heights: crown layer (>10 m), middle layer (2–10 m), and shrub layer (0–2 m). The stands referred to as full storied in this paper correspond to those referred to as multi-layered in the figure.

152

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

group termed three-layered stands also encompasses the stands in transition to three layers. Differences in management frequency and in species composition at the time of establishment were then compared between the two groups using the statistical analysis software IBM SPSS Statistics 19.0.0. Comparison of differences in management frequency was performed using non parametric Mann–Whitney U test. Comparisons of differences in species composition were performed with independent samples T test, or Mann–Whitney U test in case the conditions for T tests were not fulfilled. Acceptable significance level was set at p < 0.05. Results Structural type classification Based on the classification of stand structures, the stands were categorized in seven structural types (Fig. 3a and b): The stands belonging to the two-layered group (Fig. 3a) included three different structural types categorized as: - Two-layered with a well-developed middle layer dominated by tree species (n = 3), - Two-layered with a well-developed middle layer dominated by shrub species (n = 1), - Two-layered with an understory of tree species (saplings) (n = 1). The three-layered group (Fig. 3b) included four structural stand types categorized as: - Three-layered with a well-developed shrub layer (n = 1), - Three-layered with an indistinct shrub layer (n = 1), - Full-storied stand with scattered or irregularly grouped individuals (n = 1), - Low stands with shrub layer and emerging standards (n = 2). Group 1: two-layered stands F1, F2, F3 and B2 were classed as two-layered stands with a well developed middle layer dominated by either tree species (F1, F2, F3) or shrub species (B2) (Fig. 3a, Table 2). The stand structure of I2 was strongly influenced by an abundance of F. excelsior saplings in the shrub layer, and was therefore classed as a two-layered stands with understory of tree species (saplings). The crown layers of F1, F2 and F3 were closed, while B2 and I2 showed a more semi-open character (Fig. 3a). Species compositions (Table 2) of the crown layers of F1, F2 and F3 were characterized by a high relative proportion of shade tolerant tree species (i.e. A. platanoides, C. betulus, F. sylvatica, P. padus, T. cordata), while B2 and I2 had more light shade casting tree species such as B. pendula, F. excelsior and Q. robur. Except for B2, the understory was relatively dense in all stands, often with an abundance of root suckers of P. Padus or R. multiflora (Table 2). Group 2: three-layered stands and stands in transition to three layers B1 and I3 were classed as three-layered stands (Fig. 3b), the former with a well-developed shrub layer formed by R. alpinum, and the latter with an indistinct shrub layer with mainly saplings of various tree species. A2 was classed as a full-storied stand type as a result of its semi-open crown layer, and the relatively even distribution of trees and shrubs across the vertical profile. Because of their dominant middle layers combined with scattered crown layer trees and shrub layer species A1 and I1 were classed as low stands with shrub layer and emerging standards.

Canopy stratification as affected by management and species composition Differences in both management history and initial species composition between the two main categories of stands were observed (Table 3). The three-layered group had been managed significantly more frequently (p = 0.009) than the two-layered group with a number of thinnings varying between two and five thinnings per stand during the study period, compared to one or two thinnings in the two-layered group (Tables 1 and 3). Initial species composition varied both in terms of number and proportion of different species. The number of crown layer tree species (including nurse species) was significantly lower (p = 0.039) (Table 3) in the three-layered group with between two and eight different species being used, while the stands grouped as twolayered had been established using between five to eleven crown layer species (Table 1). In comparison, no significant differences could be observed in terms of number of understory species or total number of species (Table 3). The relative proportions between crown layer and middle + shrub layer species indicated a tendency for lower understory/crown layer ratio in the two-layered group (p = 0.094). There was also a tendency (p = 0.091) for lower Shannon–Wiener diversity index for the three-layered group, as well as a lower ratio of shade/light tree species (p = 0.093) which to some extent could be attributed to a larger proportions of Q. robur in the crown layer (p = 0.057). Discussion In line with the ambitions set out at the time of planting, all the investigated stands in this study had developed a stratified canopy structure to some extent, even though the ambitions for the future stand development at the time of establishment was a three-layered canopy structure for all stands. The ten stands were distributed among seven different structural types. The distinctions between different structural types described by Gustavsson and Fransson (1991) are fluid, and stands can display characteristics inherent to several structural types during a transition stage from one type to another. Nevertheless, the stands showed considerable variation in canopy stratification and structural character despite their relatively young age. The reasons for this variation were complex and a result of effects from interaction between both species composition and management history. Management Management frequency during the study period clearly affected the canopy structure. The stands that had suffered from management neglect were significantly more often classified as two-layered stand types than those stands that had been thinned more frequently. This is not surprising, as management is well known to reduce canopy density, increase understory light availability and promote understory vegetation development (Crow et al., 2002). Disturbance at recurrent intervals has been shown to be important for the coexistence of tree species (Gravel et al., 2010) and our results indicate that this is applicable for young species rich forest plantations. Trees exposed to strong pressure for light competition are likely to respond by decreasing canopy growth and concentrating only on survival, while if competition is weak active foraging for light by increased canopy growth and canopy displacement has been shown to be a more successful strategy (Muth and Bazzaz, 2003). In line with our results, Crow et al. (2002) found that unmanaged naturally regenerated stands by second-growth hardwood forests developed simple vertical structures with uniform

Table 2 Vertical extension of LMS-identified layers of all stands in 2008 with percentage of individuals per layer displayed in brackets. Species composition is expressed as percentage of individuals of different species per layer. Alnarp 1 (A1)

Alnarp 2 (A2)

Crown layer 3.7–12.8 m [32.9%] 4.4–15.1 m [52.6%] A. platanoides 3.8% A. campestre 9.8% A. platanoides 3.9% C. betulus 5.9% C. sanguinea 9.8% C. avellana 19.6% C. monogyna 3.9% F. sylvatica 2.0% F. excelsior 2.0%

Q. robur 19.6% S. intermedia 9.8% T. cordata 11.8% U. glabra 2.0% 0–4.4 m [47.4%] Middle layer 0–2.4 m [67.1%] A. plananoides 0.9% A. pseudoplatanus 2.2% C. sanguinea 86.1% C. sanguinea 56.5% C. avellana 6.5% C. avellana 0.9% P. communis 2.8% C. monogyna 10.9% F. sylvatica 2.2% R. alpinum 5.6% R. uva-crispa 0.9% L. xylosteum 8.7% R. alpinum 10.9% S. nigra 1.9% V. opulus 2.2% U. glabra 0.9%

Shrub layer

a

Bulltofta 2 (B2)

9.0–18.1 m [50.2%] 9.0–18.1 m [49.6%] A. campestre 1.0% A. pseudoplatanus 18.0% B. pendula 1.0% A. glutinosa 23.0% F. sylvatica 2.0% B. pendula 44.3% P. avium 8.0% P. avium 1.6% Q. robur 72.0% S. × smithiana 13.1% S. caprea 1.0% T. cordata 15.0%

2.1–9.0 m [32.7%] A. campestre 1.5% C. sanguinea 23.1% C. avellana 38.5% F. sylvatica 4.6% Q. robur 7.7% S. nigra 3.1% T. cordata 21.5%

0–9 m [50.4%] A. glutinosa 1.6% C. betulus 12.9% L. xylosteum 3.2% P. padus 16.1% Q. robur 1.6% S. nigra 53.2% S. albus 11.3%

Filborna 1 (F1)

Filborna 2 (F2)

Filborna 3 (F3)

Ishøj 1 (I1)

Ishøj 2 (I2)

Ishøj 3 (I3)

7.3–17.9 m [54.9%] 5.3–16.3 m [34.8%] 3.6–17.3 m [62.9%] 4.2–14.6 m [18.8%] A. platanoides 4.8% A. glutinosa 2.0% A. platanoides 3.8% A. glutinosa 10.6%

8.1–20.6 m [7.2%] 8.3–17.3 m [17.5%] B. pendula 18.5% A. campestre 7.1%

A. glutinosa 8.1% B. pendula 4.9% C. betulus 8.1% C. avellana 8.1% F. sylvatica 3.2% F. excelsior 3.2% P. padus 24.2%

C. avellana 34.0% C. monogyna 4.3 P. avium 4.3% P. cerasifera 4.3% P. padus 6.4% Q. robur 31.9% S. alba ‘Liempde’ 2.1% U. glabra 2.1%

F. excelsior 40.7% Q. robur 40.7%

A. glutinosa 3.6% C. avellana 7.1% Q. robur 82.1%

0–4.2 m [81.2%] A. glutinosa 0.5% C. sanguinea 2.0% C. avellana 5.9% C. monogyna 14.8% P. avium 4.9% P. cerasifera 3.0% P. padus 10.8% R. spicatuma 47.3% R. uva-crispaa 5.9% R. multiflora 3.4% S. nigra 0.5% S. aucuparia 0.5% V. opulus 0.5%

0–8.1 m [92.8%] A. campestre 12.1% A. platanoides 0.3% A. glutinosa 0.6% C. avellana 1.7% C. monogyna 7.2% F. excelsior 7.8% P. avium 1.1% P. cerasifera 10.9% P. padus 23.6% Q. robur 1.7% R. multiflora 27.0% S. nigra 3.4% S. aucuparia 2.0% U. glabra 0.3%

2.0–8.3 m [56.9%] A. campestre 24.2% A. pseudoplatanus 1.1% A. glutinosa 1.1% C. avellana 26.4% C. monogyna 5.5% P. avium 5.5% P. padus 27.5% R. multiflora 2.2% S. nigra 5.5% U. glabra 1.1%

Q. robur 17.7% S. caprea 3.2% S. aucuparia 1.6% T. cordata 12.9% 0–7.3 m [45.1%] A. platanoides 2.0% C. avellana 7.9% F. sylvatica 2.0% P. padus 64.7% R. alpinum 5.9% S. aucuparia 2.0% T. cordata 15.7%

C. betulus 6.1% C. avellana 18.3% F. excelsior 10.2% P. padus 16.3% Q. robur 32.7% S. caprea 8.2% T. cordata 6.1%

A. glutinosa 1.3% B. pendula 3.8% C. avellana 1.3% F. sylvatica 0.6% F. excelsior 1.6% P. padus 69.0% Q. robur 7.7%

0–5.3 m [65.2%] C. betulus 1.1% C. avellana 7.6% F. sylvatica 1.1% P. padus 70.7% P. spinosaa 1.1% R. alpinum 9.8% R. nigrum 2.2% S. aucuparia 3.3% T. cordata 2.2% V. opulus 1.1%

S. caprea 0.3% S. aucuparia 1.6% T. cordata 8.6% U. glabra 0.3% 0–3.6 m [37.1%] A. platanoides 3.2% C. avellana 1.1% Crataegus sp. 1.1% P. avium 0.5% P. padus 80.0% Q. robur 1.6% Rosa sp. 1.6% S. aucuparia 8.1% T. cordata 1.6% V. opulus 1.1%

0–2.1 m [17.1%] C. sanguinea 2.9% C. avellana 2.9% F. sylvatica 2.9% Q. robur 2.9% R alpinum 76.5% S. nigra 8.8% T. cordata 21.5%

0–2.0 m [25.6%] A. campestre 9.8% A. pseudoplatanus 2.4% C. avellana 19.5% C. monogyna 12.2% P. avium 4.9% P. padus 34.1% R. alpinum 2.4% R. spicatuma 7.3% R. multiflora 7.3%

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

A. glutinosa 1.9% C. sanguinea 24.5% C. avellana 49.1% F. excelsior 1.9% P. communis 1.9% Q. robur 5.7% V. opulus 11.3%

Bulltofta 1 (B1)

Scientific species name with authority: Prunus spinosa L.; Ribes spicatum E. Robson; Ribes uva-crispa L. (cf. Table 1 for full names of all other species).

153

154

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

crown layer, few canopy gaps and little understory development. Also in stands established by active planting, like the once studied here, some of the observed differences can thus be explained by an increased competitive pressure from neighbouring trees in the unmanaged stands. The results illustrate the many possibilities to guide young woodland stands in different directions by means of management. The possibility to create stands with highly variable canopy structures offer great possibilities to create forests with high flexibility

and variation, which should be emphasized especially for the urban context where managing for multiple goals and in particular recreational qualities are an important aspect. A similar conclusion was reached by Rydberg and Falck (1998) who demonstrated how different young forest types could be created by using different precommercial thinning strategies to correspond to different needs and preferences of the public. Management frequency and intensity must be adjusted to local climate and soil conditions but based on this study we would

Fig. 3. (a) Profile diagrams of forest stands categorized as ‘Two-layered’. Solid lines indicate the limits of the layers identified by the LMS-algorithm. Length of the transect and tree height are displayed on the X and Y axes. (b) Profile diagrams of stands categorized as ‘Three-layered’. Solid lines indicate the limits of the layers identified by the LMS-algorithm. Length of the transect and tree height are displayed on the X and Y axes.

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

155

Fig. 3. (Continued).

recommend a minimum of three thinnings within the first twenty years of stand development if the ambition is to create multilayered canopy structures. Due to the limited number of stands and their distribution to different structural types and sub-types, the scope for analysing other aspects of management such as timing or thinning intensity at each intervention is limited within the design of this study. These aspects should however also be important and has been highlighted by e.g. Sprugel et al. (2009) or Hanson and Lorimer (2007) who suggest that management systems intended to support development of multi-layered structures would gain from increasing contrasts of harvesting intensity within different areas of the stands to promote small-scale variability.

Species composition In addition to management frequency canopy stratification was also influenced by differences in initial species composition. The decisive factors for canopy stratification were the proportion of understory species, the proportion of shade tree species and the number of crown layer tree species (Table 3). These factors should be regarded as partly intercorrelated. For example, the lower proportion of understory species in the two-layered stands was also reflected in the higher number of crown layer species as well as the higher proportion of shade tree species. Altogether, these three aspects suggest that a better balance concerning the proportion of

156

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

Table 3 Comparison of management frequency during the study period and species composition at the time of establishment between stands categorized as three- or two-layered in 2008. Difference in management frequency was tested with non-parametric Mann–Whitney U test. Differences in initial species compositions were tested with independent samples t-test (df = 8), except for percent Q. robur in the crown layer, for which Mann–Whitney U test was performed. Acceptable significance level was set at p < 0.05.

Total number of species Crown layer species number Understory species number (middle + shrub layer) % Understory (middle + shrub layer species) % Q. robur of crown layer (nurse species excluded) % Shade tree speciesa % Low shade shrubsb Shannon Wiener index Number of thinnings a b

3-Layered group mean values

2-Layered group mean values

p-Values

9.4 4.6 4.8 36.2 64.7 9.0 4.0 1.87 3.6

12.4 8.6 3.8 24.0 27.7 24.0 3.0 2.37 1.2

0.196 0.039 0.475 0.094 0.057 0.093 0.667 0.091 0.009

Includes A. platanoides, A. pesudoplatanus, C. betulus, F. sylvatica, T. cordata, U. glabra, P. padus. Includes R. alpinum, L. xylosteum, S. albus.

each species type at establishment can increase the potential to successfully develop multi-layered canopy structures. The dynamics of coexistence between different forest tree species is complex where light availability is a major determinant operating on different temporal and spatial scales (Gravel et al., 2010). Dekker et al. (2008) have shown that stratification of the canopy is initiated already during the sapling stage as fast growing species start to overtop slower species. A high diversity of species at establishment should therefore indeed be positive for the establishment of three-layered canopy structures. This is however only true if species with different life forms, shade tolerance, and successional strategies are mixed. In the stands grouped as two-layered, the high diversity and proportion of crown layer species is one factor for the negative effect on canopy stratification. This can however not be separated from the higher proportion of shade tree species in this group. Thus Dekker et al. (2008) concluded that in young forest stands, the inherent species morphologies (particularly the shade tolerance of a species), was more important than the competitive pressure from neighbouring tree species for explaining effects of competition and ability to overtop other saplings. In our study, an increased proportion of middle layer and shrub layer species favoured the development of three-layered canopy structures. As mentioned, this can be explained by the reduced competitive pressure from the reduced density of the crown layer, both in terms of number of individuals but also the lesser proportion of shade tolerance crown layer trees. The optimal proportions between crown layer and understory species will vary depending on species composition. But based on the results of this study, an initial species composition with understory species representing 35% (of which the majority are intended for the middle layer) rather than 25% of the total number of individuals at the time of establishment seems to be a reasonable guideline (Table 3). This should however be modified according to for example the shade tolerance or risk of post-establishment mortality of different species. We argue that instead of aiming at maximising total species number it appears to be more reasonable to focus on a few key species in each layer. Many different tree species in the crown layer lead to stronger competition for light and a denser overstory, which could prevent middle and understory plants from developing into vital individuals due to reduced light. This effect will be even more accentuated with the inclusion of a high proportion of shade tree species. In tropical and temperate forests with varying species compositions and stand structures, the relative proportion of shade-tolerant species in the crown layer influence the number of strata (Baker and Wilson, 2000). Shade trees can however contribute to an increased diversity both regarding species richness but also to an increased canopy complexity with deep crowns and the possibility for shade tolerant tree species to become integrated as part of the middle layer. The proportion of shade tree species should however be limited and based on this study we recommend that

shade trees constitute between around 10% of the total population at establishment. Relation between management and species composition Some of the stands clearly illustrate that a balanced or unbalanced species mixture can increase or decrease the resilience towards deficient management. In I2 the overabundance of shrub species that constituted as much as 40% of the population at establishment, resulted in an extremely dense understory soon after plantation. This led to a total removal of all understory species during the first thinning, which resulted in an even denser shrub layer originating from root and stump suckers in combination with natural regeneration of saplings. The B1 stand on the contrary had developed a distinctive three-layered stand structure despite it had been thinned only twice. However, thanks to a high percentage of the main tree species Q. robur, limited share of shade tree species (T. cordata, F. sylvatica), and inclusion of R. alpinum, a three-layered structure had developed. The shrub layer below 2.0 m, where light stress is strongest, was the most difficult layer for species to occupy. The light attenuation effect of the middle layer understory community can be significant and act as a filter and reduce light availability at the forest floor to the same levels irrespective of crown layer light transmission (Bartemucci et al., 2006). A well-developed low shrub layer was however found in some of the stands, for example B1 and A1 as well as A2 to some extent. A commonality for these stands was the occurrence of species belonging to the Ribes genus. These species were sufficiently shade tolerant to survive and also did not react to the low light conditions by developing tall stems with small crowns as was the case for several other shrub species, e.g. V. opulus or C. sanguinea. This was clearly illustrated in F1 and F2 where R. alpinum was still part of the species mixture in 2008 despite the dense canopy in these stands. In three of the stands, A1, I1 and I3, R. spicatum and R. uva-crispa had been established by natural dispersion during the study period. Conclusive discussion Our revisiting to the four urban forests in the Öresund region of Denmark and Sweden established along the ecological principles show that it is possible to create multi-layered canopy structures already within such a relatively short time frame as twenty-five years or less in southern Scandinavia. This said, the results emphasis that the importance of management cannot be underestimated if multi-layered canopy structures are to be developed during the young phase of forest development. In the long term, management neglect would lead to fewer possibilities for managers to guide the woodland towards a multi-layered canopy structure, and likely

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

the necessity to reformulate the overall objective of future stand development and management strategies. Already at the time of the establishment of these type of forest plantations there was a discussion about an assumed higher sensitivity and dependence on both more skilled and well-timed management decisions (Scott et al., 1986; Gustavsson, 2004). This need of adequate thinning operations seems to be true regardless of the initial planting design, but the results also indicate that a balanced or unbalance species mixture very likely may increase or decrease the resilience towards deficient management. With higher management frequency, it is possible that the stands classed as two-layered would have developed a more complex stand structure despite their sometimes less favourable species composition. Management neglect is however not an unusual situation for many young forest plantations where timber production is secondary (Nielsen and Jensen, 2007), and in a wider perspective, the management neglect of young urban woodlands in general is symptomatic for several Scandinavian countries (Gundersen and Mäkinen, 2009). The future development of recreational values in these woodlands has been argued to be an open question (Gundersen et al., 2006). Faced with the varied possibilities for woodland creation, the lack of commitment during the early stages of woodland development should seem surprising. Poor financial resources at the municipal administration level is often mentioned as a major reason for management negligence (Nielsen and Jensen, 2007), but the problem is likely also of attitudinal character. Attitudes among experts in ecology, landscape architecture and forestry has revealed a poor interest in the young forest

157

development phase, which can be traced back to their perceived uninteresting biological processes, uninviting and non-impressive recreational values, and low economical profits respectively (Jönsson and Gustavsson, 2002). Compared with historical times, today’s managers have deeper academic knowledge, better tools and have access to the most elaborated planning systems, but the main question of what kind of landscape to be created is today still perhaps more relevant than ever (Gundersen and Mäkinen, 2009). And as this study show, this question needs to be addressed proactively already in the young stages of forest development. In memoriam Gustav Richnau, the article’s lead author, left us suddenly and tragically while this paper was under review. Gustav had a bright professional and personal future ahead of him when he passed away at only 31 years of age. In his absence we remember Gustav as a thoughtful and intelligent person, with a genuine interest in people, and a well-developed sense of humour. But most of all, we remember Gustav as a dear friend. Acknowledgments The study was financed by the Swedish Research Council FORMAS. We are grateful to thank Jan-Eric Englund for statistical support and to Roland Gustavsson for fruitful discussions. Appendix A.

Stand

Management history

A1

1985: Establishment

1990: Thinning of 50% of nurse trees

1995: Thinning of 50% of remaining nurse trees

A2

1984: Establishment

1990: Thinning of 50% of nurse trees

1994: Thinning of nurse trees, leaving single individuals

B1

1983: Establishment

1990: Thinning of 25–50% of nurse trees

2000: Pruning and thinning of edges along pathways

B2

1984: Establishment

F1

1983: Establishment

F2

1983: Establishment

F3

1987: Establishment

I1

1977: Establishment

1990: Thinning of 25–50% of nurse trees 1998: Thinning reducing tree stock 20% 2000: Thinning, reducing tree stock with 20% 2002: Thinning, reducing tree stock with 15–20% 1991: Thinning of nurse trees and S. caprea, R. multiflora and few C. avellana

1996: Thinning primarily focusing on nurse trees Possibly a second thinning

2003: Thinning primarily of the understory with focus on S. caprea, R. multiflora and C. monogyna

I2

1977: Establishment

1989: coppicing of all understory vegetation to access the stand for thinning of canopy layer

I3

1977: Establishment

1989: Thinning of nurse trees competing with Q. robur. Coppice of 50% of C. avellana

1996: Thinning of nurse trees competing with Q. robur. Thinning of S. caprea, R. multiflora and few C. avellana and C. monogyna 2001: Rigorous thinning of canopy. Coppicing of understory vegetation in case it inhibited work conditions 2001: Thinning of nurse trees leaving single individuals. Coppice of majority of remaining C. avellana

1996: Thinning of 75% F. excelsior, 50% A. platanoides and C. sanguinea. Light thinning of other species 1995: Coppice of majority of shade tolerant species to reinitiate understory formation

2004: Light thinning of the understory focusing on A. glutinosa sprouts

2002: Light thinning of the understory and pruning of the low branches on canopy trees 1997: Rigorous thinning removing 50–75% of the main canopy and understory species 2004: removal of dead wood

2003: Thinning of overstory trees. And coppicing of individual C. avellana. Establishment of field layer 2003: Light thinning of mostly T. cordata, A. platanoides, and A. glutinosa. Establishment of field layer

158

G. Richnau et al. / Urban Forestry & Urban Greening 11 (2012) 147–158

References Baker, P.J., Wilson, J.S., 2000. A quantitative technique for the identification of canopy stratification in tropical and temperate forests. Forest Ecology and Management 127, 77–86. Bartemucci, P., Messier, C., Canham, C.D., 2006. Overstory influences on light attenuation patterns and understory plant community diversity and composition in southern boreal forests of Quebec. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 36, 2065–2079. Blakesley, D., Buckley, G.P., 2010. Woodland Creation for Wildlife and People in a Changing Climate: Principles and Practice. Pisces Publications, Newbury. Christensen, M., Emborg, J., 1996. Biodiversity in natural versus managed forest in Denmark. Forest Ecology and Management 85, 47–51. Crow, T.R., Buckley, D.S., Nauertz, E.A., Zasada, J.C., 2002. Effects of management on the composition and structure of northern hardwood forests in Upper Michigan. Forest Science 48, 129–145. Dekker, M., Verkerk, P.J., den Ouden, J., 2008. Target species identity is more important than neighbor species identity. Forest Ecology and Management 255, 203–213. Everett, R., Baumgartner, D., Ohlson, P., Schellhaas, R., 2008. Defining and quantifying canopy strata. Northwest Science 82, 48–64. Forbes, S., Cooper, D., Kendle, A.D., 1997. The history and development of ecological landscape styles. In: Kendle, T., Forbes, S. (Eds.), Urban Nature Conservation. E & FN Spon, London, pp. 69–113. Franklin, J.F., Van Pelt, R., 2004. Spatial aspects of structural complexity in oldgrowth forests. Journal of Forestry 102, 22–28. Gravel, D., Canham, C.D., Beaudet, M., Messier, C., 2010. Shade tolerance, canopy gaps and mechanisms of coexistence of forest trees. Oikos 119, 475–484. Gundersen, V.S., Frivold, L.H., 2008. Public preferences for forest structures: a review of quantitative surveys from Finland, Norway and Sweden. Urban Forestry and Urban Greening 7, 241–258. Gundersen, V., Frivold, L.H., Löfström, I., Jørgensen, B.B., Falck, J., Øyen, B.-H., 2005. Urban woodland management – the case of 13 major Nordic cities. Urban Forestry and Urban Greening 3, 189–202. Gundersen, V., Frivold, L.H., Myking, T., Øyen, B.-H., 2006. Management of urban recreational woodlands: the case of Norway. Urban Forestry and Urban Greening 5, 73–82. Gundersen, V., Mäkinen, K., 2009. Aldo Leopold and stewardship: lessons for forest planning and management in the Nordic countries? Norwegian Journal of Geography 63, 225–232. Gunnarsson, B., Heyman, E., Vowles, T., 2009. Bird predation effects on bush canopy arthropods in suburban forests. Forest Ecology and Management 257, 619–627. Gustavsson, R., 1986. Struktur i lövskogslandskapet (Structure in the broadleaved landscape). Stad och Land 48, Sveriges Lantbruksuniversitet, Alnarp (in Swedish, with English summary). Gustavsson, R., 2004. Exploring woodland design: designing with complexity and dynamics—woodland types, their dynamic architecture and establishment. In: Dunnet, N., Hitchmough, J. (Eds.), The Dynamic Landscape. Spon Press, London, pp. 184–214. Gustavsson, R., Fransson, L., 1991. Furulunds Fure – en skog i samhällets centrum (Furulunds Fure – a forest in the centre of the town). Stad och Land 96, Sveriges Lantbruksuniversitet, Alnarp (in Swedish, with English summary). Hanson, J.J., Lorimer, C.G., 2007. Forest structure and light regimes following moderate wind storms: implications for multi-cohort management. Ecological Applications 17, 1325–1340. Heyman, E., 2011. Management of urban woodlands: effects on birds communities and recreation values. Doctoral Thesis, University of Gothenburg, Gotherburg. Ishii, H.T., Tanabe, S., Hiura, T., 2004. Exploring the relationships among canopy structure, stand productivity, and biodiversity of temperature forest ecosystems. Forest Science 50, 342–355. Jensen, F.S., Skovsgaard, J.P., 2009. Precommercial thinning of pedunculate oak: recreational preferences of the population of Denmark for different thinning practices in young stands. Scandinavian Journal of Forest Research 24, 28–36.

Jönsson, A., Gustavsson, R., 2002. Management styles and knowledge cultures, past, present and future, related to multiple-use and urban woodlands. Urban Forestry and Urban Greening 1, 39–47. Konijnendijk, C., 2000. Adapting forestry to urban demands role of communication in urban forestry in Europe. Landscape and Urban Planning 52, 89–100. Koop, H., 1989. Forest Dynamics: SILVI-STAR: A Comprehensive Monitoring System. Springer, Berlin. Latham, P.A., Zuuring, H.R., Coble, D.W., 1998. A method for quantifying vertical forest structure. Forest Ecology and Management 104, 157–170. Muth, C.C., Bazzaz, F.A., 2003. Tree canopy displacement and neighborhood interactions. Canadian Journal of Forest Research 33, 1323–1330. Nielsen, A.B., 2006. Understanding and communicating forest stand structures: lifting barriers for nature-based forest management. Ph.D. Thesis, Royal Veterinary and Agricultural University, Frederiksberg. Nielsen, A.B., Jensen, R.B., 2007. Some visual aspects of planting design and silviculture across contemporary forest management paradigms – perspectives for urban afforestation. Urban Forestry and Urban Greening 6, 143–158. Nielsen, A.B., Nielsen, J.B., 2005. The use of profile diagrams for mixed stands in urban woodlands – the management perspective. Urban Forestry & Urban Greening 3, 163–175. Nielsen, A.B., Nielsen, J.B., Gustavsson, R., 2005. Nye artsblandinger til rekreationsskove og parkplantninger – Tor Nitzelius park (New species mixes for recreational forests and park plantations – Tor Nitzelius Park). Videnblad 3.1-50. Skov og landskab, Frederiksberg (in Danish). Niklasson, M., Nilsson, S.G., 2005. Skogsdynamik och arters bevarande (Forest dynamics and species conservation). Studentlitteratur, Lund (in Swedish). Nilsson, S.G., Hedin, J., Niklasson, M., 2001. Biodiversity and its assessment in boreal and nemoral forests. Scandinavian Journal of Forest Research Supplement 3, 10–26. Parker, G.G., Brown, M.J., 2000. Forest canopy stratification – is it useful? American Naturalist 155, 473–484. Qvarnström, H., Rosenqvist, T., 1980. Bulltofta rekreationsområde – den fremtida parken. Förslag till utformning (Bulltofta recreation area – the future park. Proposal for development.). Landskap 57. Swedish University of Agricultural Sciences, Alnarp (in Swedish). Ribe, R.G., 1989. The aesthetics of forestry: what has empirical preference research taught us? Environmental Management 13, 55–74. Ruff, A., 2002. Holland and the ecological landscape. Garden History 30, 239–251. Rydberg, D., Falck, J., 1998. Designing the urban forest of tomorrow: pre-commercial thinning adapted for use in urban areas in Sweden. Arboricultural Journal 22, 147–171. Scott, D., Greenwood, R.D., Moffatt, J.D., Tregay, R.J., 1986. Warrington New Town: an ecological approach to landscape design and management. In: Bradshaw, A.D., Goode, D.A., Thorp, E.H.P. (Eds.), Ecology and Design in Landscape: The 24th Symposium of the British Ecological Society Manchester 1983. Blackwell Scientific Publications, Oxford, pp. 143–160. Silvennoinen, H., Alho, J., Kolehmainen, O., Pukkala, T., 2001. Prediction models of landscape preferences at the forest stand level. Landscape and Urban Planning 56, 11–20. Sprugel, D.G., Rascher, K.G., Gersonde, R., Dovciak, M., Lutz, J.A., Halpern, C.B., 2009. Spatially explicit modeling of overstory manipulations in young forests: effects on stand structure and light. Ecological Modelling 220, 3565–3575. Thompson, I.H., 1998. Environmental ethics and the development of landscape architectural theory. Landscape Research 23, 175–194. Tregay, R., 1986. Design and ecology in the management of nature-like plantations. In: Bradshaw, A.D., Goode, D.A., Thorp, E.H.P. (Eds.), Ecology and Design in Landscape: The 24th Symposium of the British Ecological Society Manchester 1983. Blackwell Scientific Publications, Oxford, pp. 275–284. Tregay, R., Gustavsson, R., 1983. Oakwood’s New Landscape: designing for nature in the residential environment. Stad och Land 15. Sveriges Lantbruksuniversitet, Alnarp. Vance, C.C., Smith, S.M., Malcolm, J.R., Huber, J., Bellocq, M.I., 2007. Differences between forest type and vertical strata in the diversity and composition of hymenpteran families and mymarid genera in northeastern temperate forests. Environmental Entomology 36, 1073–1083. Weber, N. (Ed.), 2000. Proceedings of the Scientific Symposium. 16–17 February, Freiburg, Germany. EFI Proceeding no. 35. EFI, Joensuu.