Representative boreal forest habitats in northern Europe, and a revised model for ecosystem management and biodiversity conservation

The natural range of variation of ecosystems provides reference conditions for sustainable management and biodiversity conservation. We review how the understanding of natural reference conditions of boreal forests in northern Europe has changed from earlier perceptions of even-aged dynamics driven by stand-replacing disturbances towards current understanding highlighting the role of non-stand-replacing disturbances and the resultant complex forest dynamics and structures. We show how earlier views and conceptual models of forest disturbance dynamics, including the influential ASIO model, provide estimates of reference conditions that are outside the natural range of variation. Based on a research synthesis, we present a revised forest reference model incorporating the observed complexity of ecosystem dynamics and the prevalence of old forests. Finally, we outline a management model and demonstrate its use in forest ecosystem management and show how regional conservation area needs can be estimated. We conclude that attaining favourable conservation status in northern Europe’s boreal forests requires increasing emphasis on ecosystem management and conservation for old forest characteristics. Electronic supplementary material The online version of this article (10.1007/s13280-020-01444-3) contains supplementary material, which is available to authorized users.

Addressing the distribution of disturbance dynamics types across age classes is based on the principle that even-aged dynamics and forest structures are considered most common in post-stand-replacing disturbance sites, but diminish toward older forests due to competition, small-scale and partial disturbances and continuous seedling recruitment (Lilja et al. 2006;Aakala et al. 2009;Ylisirniö et al. 2012). Gap and cohort dynamics as well as uneven-aged structures will for the same reasons increase with time since last major disturbance. We therefore assign the main part (2/3 of 1/3 or 22.2%) of gap dynamics as well as cohort dynamics to old forests (Fig. S1). The remaining part (1/3 of 1/3 or 11.1%) of each type is distributed so that it increases linearly with increasing age from young (1/36 or 2.8% of each type) to mid-aged forests (3/36 or 8.3%). While the remaining part is 11.1%, the proportion must increase from 0% to 0.15%, i.e. by nearly 0.001% per year, across the 150 year time span covered by these two age classes (0.15% over 150 years). The proportion at 75 years, i.e. the age limit between young and mid-aged forests, is therefore ca. 0.07% (0.001% per year×75 years). Hence, the cumulative proportion must be 2.8% (0.07%×75 years/2; corresponding to the area of a right triangle) in young forests and 8.3% in mid-aged forests (11.1%-2.8%). To get the proportions of evenaged dynamics, we subtract the proportions of gap dynamics and cohort dynamics from the total proportion assigned for each age class concerned (e.g. 25%-2.8%-2.8% = 19.4% in young forests). The proportions of even-aged dynamics are 19.4%, 8.3% and 5.6% in young, mid-aged and old forests, respectively.
The exact proportions calculated are finally rounded to even percentages. This is done so that summed proportions across dynamic types and age classes equal the overall targets, but also with regards to the principle that non-stand-replacing disturbance dynamics prevail in old forests while even-aged dynamics dominate in young forests (Fig. S1).
Here, young and mid-aged forests are separated at 75 years, but any other age limit may be used while the share of disturbance dynamic types change linearly until forests become old. For instance, if we change the age limit to 110 years, the proportion of gap dynamics as well as cohort dynamics at that age will be ca. 0.11% (0.001% per year×110 years; see above). Hence, the cumulative proportion of each dynamics type will be ca. 6% (0.11%×110 years/2) in young forests and in 5% in mid-aged forests (11.1%-6%). Finally, while young forests (0-109 years) will cover 36.7% (110/150 or 73% of 50%) and mid-aged forests (110-149 years) will cover 13.3% (40/150 or 26.7% of 50%) of the land area, the proportions of even-aged dynamics will become 24.7%, 3.1% and 5.6% in young, mid-aged and old forests, respectively.

3(6)
Fig. S1. A schematic description of the revised reference model. The dominance (at least 50%) of old forests (≥150 years), but also the equal shares (1/3 or 33.3%) of the three disturbance dynamics types comprise the basic model settings and overall targets (figures in bold) derived from a review of current understanding of reference conditions. The three disturbance dynamic types are then distributed across age classes based on the ecological principle that gap dynamics and cohort dynamics prevail in old forests while even-aged dynamics dominate in young forests. Their respective proportions change linearly with increasing age across young and mid-aged forests. Here, young and mid-aged forests are separated at 75 years, but any other age limit may be used while the share of disturbance dynamic types change linearly until forests become old. The calculated, exact proportions (figures without parenthesis) are finally rounded to even percentages used in the revised reference model presented in the article (figures in parenthesis). This is done so that summed proportions across dynamic types and age classes equal the overall targets, but also with regards to the principle that gap dynamics and cohort dynamics prevail in old forests while evenaged dynamics dominate in young forests.

in the article
The left diagram of Fig. 2 in the article shows the actual age-class distribution across three major site types of mineral soil forests in the northern part of the boreal region in Sweden (the grey area of the map). The underlying statistics are given in Table S1. Table S1. Shares (%) of three major site types on forest land on mineral soils within the boreal region studied (figures used for the left diagram in Fig. 2 in the article; data from the Swedish National Forest Inventory 2020). In this example, wet to moist site types include also flooded forests. The middle diagram of Fig. 2 in the article shows the age-class distribution estimated using the standlevel and bottom-up logic underlying the ASIO model, i.e. assuming that a single specific type of disturbance dynamics; gap dynamics, even-aged dynamics, or cohort dynamics, prevails on each major site type. The expected age-class distribution under natural conditions in the region studied is computed by multiplying the estimated proportions of each major site type (18%, 9% and 73%; Table  S1) with three site-type specific age-class distributions ( Table S2). The results are shown in Table  S3. Table S2. The stand-level and bottom-up ASIO logic underlying previous analyses of reference conditions of naturally dynamic boreal landscapes (Angelstam and Kuuluvainen 2004) and long-term forest reserve needs (Angelstam and Andersson 2001, with details explained in SOU 1997:97 and 1997 in Sweden. The natural occurrence of forest dynamics types are first linked to specific site types. The expected distribution of forest developmental stages after disturbance in terms of age classes is then estimated by using models of equilibrium dynamics. The age-class distributions of gap and cohort dynamics on wet to moist and dry, poor site types are modelled based on expert judgement (cf. Angelstam and Kuuluvainen 2004). The age-class distribution of even-aged dynamics on mesic site types is modelled as an average between a negative exponential and a Weibull distribution (cf. Johnson 1992) resulting from stand-replacing disturbances with a return interval of 100 years.   Fig. 2 in the article), i.e. by multiplying the estimated proportions of each major site type (18%, 9% and 73%; Table S1) with three site-type specific ageclass distributions (Table S2) The right diagram of Fig. 2 in the article shows the results when using the revised reference model ( Table 1 and Fig. 3 in the article and Appendix S1). It emphasizes a prevalence (at least 50%) of old forests (≥150 years) due to a greater importance of non-stand-replacing disturbances and gap and cohort dynamics. Further, the three dynamics types occur in equal proportions (1/3 or 33.3% of each type) and are less strictly related to site type. Here, we adapt the revised reference model to the classification of young (0-109 years) and mid-aged (110-149 years) used in the analysis (see Appendix S1). The expected age-class distribution under natural conditions in the region studied is given by Table S4. Table S4. The revised reference model with an age limit of 110 years between young and mid-age forests (see Appendix S1 for explanation of model set-up and proportions of age classes). The dominance (at least 50%) of old forests (≥150 years), but also the equal shares (1/3 or 33.3%) of the three disturbance dynamics types comprise basic model settings and overall targets (highlighted figures in bold) derived from a review of current understanding of reference conditions. The targeted distribution of dynamics types across age classes (Table S4) can be attained by taking the available site-type distribution (Table S1) and their probable natural dynamics into account. In this example, gap dynamics are assumed to prevail on moist to wet site types (totally ca. 18%) and cohort dynamics on dry, poor site types (totally ca. 9%). Still, gap and cohort dynamics need to occur not only on these two major site types, but also on mesic, intermediate to rich site types to achieve the targeted shares of gap and cohort dynamics across age classes. The resulting distribution of dynamics types across age classes are shown in Table S5. Table S5. The age-class distribution estimated using the revised reference model (used for the right diagram in Fig. 2 in the article). The three disturbance dynamics types are distributed within each age class by taking the available site-type distribution (Table S1) and their probable natural dynamics into account; gap dynamics (GD) prevail on moist to wet site types and cohort dynamics (CD) on dry, poor site types. Still, gap and cohort dynamics need to occur not only on these two major site types, but also on mesic, intermediate to rich site types to attain the targeted shares across age classes (figures underlined). The targeted shares of even-aged dynamics (