In this case study of Swedish boreal forestry, we analyze the implementation in practical management of a sustainable forest management that includes the continued capacity of ecosystems to also provide non-timber services. In the Swedish national forest policy, the equal goals of timber production and biodiversity conservation have increased the responsibility of landowners and managers to modify practical management. We compiled written information from three larger FSC-certified forestry companies combined with semi-structured interviews with forest management personnel at regional offices to map what data and information the companies collect and use. We examined to what extent obtainable information from the forest companies captures composition, structure, and function of boreal ecosystems at patch, stand and landscape scale. This was done using 47 indicators compiled from literature and grouped into five categories based on nationally identified deficiencies in the managed boreal forest compared to unmanaged forest. We found that the recording of data describing different aspects of the forest ecosystem were separated in two largely uncoordinated systems, Ecological Landscape Plans and stand registers. While there is a need for conservation-oriented data across scales, collected statistics was largely production-oriented and related to the stand scale. We also identified information gaps regarding different ecosystem structures and their spatial distribution such as dead wood and habitat networks. The knowledge base available to future management decisions also seems to lack information on connectivity in the landscape, habitat at smaller spatial scales and the effectiveness of reserved areas. To reach the ecological goals of sustainable forest management, there is a need to further coordinate existing data and information but also to fill gaps in particular regarding cross-scale information.

We combine historical maps and satellite derived data to reconstruct the development of a Swedish boreal landscape over the past 300 years. The aim is to understand legacies from past use patterns in present-day forest composition and consequences for conservation objectives from a landscape perspective. We analyze landscape development in cross-tabulation matrixes, building change trajectories. These trajectories are tested in linear models to explain the distribution of present-day landscape composition of coniferous, mixed, and deciduous forests >110 years. Of 49 tested change trajectories, 11 showed a significant association. Associations for mixed and coniferous forests were similar and linked to characteristics such as forest continuity, which characterized the studied landscape. Deciduous older forests did not show any association to forest continuity but were more likely to occur on areas that specifically shifted from forests with grazing in the 1720s to open impediment (likely indicating low tree coverage) in the 1850s. There were large shifts and spatial redistribution in ownerships over time. Use patterns and legacies varied between small- and large-scale ownership categories as well as within small-scale categories. The legacies found in the study indicate a complex origin of heterogeneous landscape elements such as older deciduous forests. Additionally, the origin of the legacies indicates a potential need to diversify conservation management based on the influence of past use patterns. Despite large inconsistencies in historical and contemporary data we argue that this type of analysis could be used to further understand the distribution of landscape elements important for conservation objectives.

The process of recreating historical land cover dynamics, needed to nderstand trends and transient states in ecosystems, includes ifficulties such as the sensitivity of applied spatial analyses to eterogeneities in historical material. This paper compares the handling f quantitatively dominating categories in two matrix analyses of land over change within a Swedish boreal landscape (1725-1859). The focus is n how inconsistencies between historical maps can be handled without iolating the inherent semantic potential. The study shows that analyses f land cover support different indications of change depending on the reatment of dominating categories. The type of landscape and research uestions in focus should therefore be part of choosing matrix method nd classification scheme. The observed patterns need to be evaluated gainst drivers of change and semantic plasticity in classification chemes to separate ecological change from semantic confusion. This aper recommends aggregated classification schemes with maintained riginal relationships between categories in comprehensive analyses. owever, no pathway is persistent over time and categories should be llowed to disappear and new to appear. Analysis of historical dynamics ith extended transition matrixes is recommended to account for the ynamics of small categories in relation to dominating categories within landscape.

In this study we reconstruct forest composition during the 1860s for a 71000 ha area in southern boreal Sweden. The aim is to show how historical ownership and associated anthropogenic disturbances act as a source of heterogeneity in the present-day distribution and composition of coniferous and deciduous forest within the commercial production forest. We use older (>110 years) and mature (70-110 years) forest as response variables in generalized linear models with a binominal error distribution. The explanatory variables include size of zone, historical type of ownership zone (village, company, and farm), amount of forest, and forest type. We focus in particular on investigating effects from interacting explanatory variables. The significant statistical associations in the study indicate that patterns of deciduous and coniferous older patches differ, and that deciduous patches differ in relation to age interval. The oldest deciduous patches, for example, are today more likely on areas that had deciduous cover also in the past and stood on forestland managed by farmers, but less likely on the same habitat managed by companies. We show that there are strong effects on present forest composition from historical ownership and forest composition. We argue that by including local data on past ownership combined with knowledge on use patterns management could be better adapted to local landscape dynamics compared to the application of overly generalized patterns or models of boreal dynamics that excludes interactions with management.

Habitats of Swedish conservation interests are in general small and fragmented following the extensive and intensive forest management history. This study covering 71 000 ha of boreal Sweden investigates how history influences present-day distribution and composition of forests identified as high conservation value habitats and how they are protected. We also investigated if the habitat criteria used to describe reservations differed between reservation types and if habitat criteria were associated with the size of Woodland Key Habitats. The results show strong effects from historical ownership and historical forest type on the probability of an area being set aside as formally protected or as voluntary protected Woodland Key Habitats. We also found that both formal reservations and Woodland Key Habitats primarily cover coniferous forest in the age interval 70-110 years but not the presumably most valuable oldest coniferous category >110 or deciduous forests, which are as common in reservations as in other areas. Old deciduous forests (>110 years) are significantly more rare in formal reservations compared to the forest matrix. When viewed in a context of fragmentation and edge effects the results underline the importance of evaluating reserved areas and Woodland Key Habitats in a wider temporal and larger spatial perspective to optimize conservation management efforts. Maximal representation and biodiversity can be better achieved if new reservations are chosen to represent different ownership and forest history, and if they are selected in a landscape context related to present reservations and the present surrounding production forest.