Drivers of forest harvesting intensity patterns in Europe
Introduction
Land use provides humanity with essential food, fibre, and bioenergy, but is also a major force of global environmental change (MA, 2005, Haberl et al., 2007, Pereira et al., 2010). As fertile land is getting scarce (Lambin and Meyfroidt, 2011) and further expansion of land use into remaining wildlands incurs high environmental costs, future production increases will, to a large extent, have to rely on sustainably intensifying land already in use (Foley et al., 2011, Tilman et al., 2011). Yet, assessing where future production can be increased and understanding the trade-offs of intensification is currently limited by incomplete knowledge about the spatial patterns and drivers of intensification pathways, especially at broad geographic scales (Verburg et al., 2009, Erb, 2012, Lambin et al., 2001).
This is particularly the case in forestry, where the spatial patterns of forest management intensity and the drivers that produce these patterns remain highly unclear. This is unfortunate, because forest management effects on forest ecosystem functioning vary substantially depending on management intensity. For example, the intensity by which forests are managed affects forest structure (Vilén et al., 2012), soils (Jandl et al., 2007), biogeochemical cycles (Nabuurs et al., 2013, Luyssaert et al., 2012), biodiversity (Paillet et al., 2010), and ecosystem service provisioning (Gamfeldt et al., 2013). Understanding the spatial patterns of forest management intensity and its drivers is therefore important for assessing the environmental trade-offs of forestry and for identifying opportunities for sustainable intensification.
Assessing forest management intensity is challenging because intensity itself is a complex term, encompassing multiple dimensions (Schall and Ammer, 2013). Consequently, forest management intensity has been examined using a wide range of indicators, including harvested timber volumes, forest structural parameters (e.g., the difference between potential and actual biomass storage), stand establishment practices, tree species composition, length of rotation periods, human appropriated net primary production, or the amount of fertiliser, herbicides, and machinery used (Luyssaert et al., 2011, Forest Europe, 2011, Duncker et al., 2012). Intensity metrics, which relate inputs (e.g., capital), outputs (e.g., harvested timber volumes), or system properties (e.g., ecosystem productivity) to each other, can provide insights into land use intensity patterns and drivers (Erb et al., 2013, Kuemmerle et al., 2013). For example, interpreting harvested timber volumes without considering ecosystem productivity could be misleading as the same volume of timber extracted from forest systems with high or low productivity may indicate very different levels of forest harvesting intensity. By expressing harvested timber volumes in relation to the net annual increment, forest harvesting intensity can be assessed across large regions.
Unfortunately, studies assessing forest harvesting intensity have either focussed on the national scale (e.g., relying on national forest resource assessments, (Kuusela, 1994, Forest Europe, 2011), or on small study regions (see Schall and Ammer, 2013 for an overview), both of which precludes understanding spatial patterns of management intensity. Only two studies addressed drivers of forest harvesting patterns at broad spatial scales. Analysing timber harvesting patterns in European Russia showed that road density, forest composition, and total forest area were important determinants of harvesting patterns (Wendland et al., 2011). A range of spatial variables including tree species composition, slope, forest coverage, proximity to cities, and conservation areas allowed mapping different forest management systems in Europe using an expert-based approach (Hengeveld et al., 2012). We know of no study explicitly addressing broad-scale patterns of forest harvesting intensity.
Evidence on the drivers of forest owner’s decisions to manage their forest intensely or not was only derived from local-scale case studies. These studies, mainly focussing on non-industrial, private forest owners, show that a range of policy, forest resource, and market factors are potentially important in determining timber volumes extracted (Beach et al., 2005, Amacher et al., 2003). For example, forest management plans, property size, and income from agriculture determined harvesting decisions in Norway (Størdal et al., 2008), ownership size and type shaped harvesting decisions in the southern US (Arano and Munn, 2006), or the demand for wood products and associated price changes were important drivers of harvesting decisions in the US and Australia (Adams et al., 1991, van Putten and Jennings, 2010). Furthermore, population density, forest size, and distance to urban areas influenced harvesting in the US (Wear et al., 1999, Munn et al., 2002). Yet, none of these studies addressed patterns and drivers of forest harvesting intensity for larger regions. Clearly, there is a research gap at the regional scale, which is unfortunate because of its major importance for policy making and for mitigating the impacts of global environmental change (Wu, 2013).
Regression models are powerful tools to assess drivers and determinants of land use patterns (Müller et al., 2011, Baumann et al., 2011, Wendland et al., 2011). Algorithmic models are particularly promising because they do not impose any a priori relationship between target and predictor variables. Fewer requirements on the data structure make them well-suited to investigate the complex and often non-linear interactions between predictors and response in land systems. Algorithmic models, such as boosted regression trees (BRT), generally attain a higher model fit and predictive accuracy than traditional statistical approaches (Elith et al., 2006, Lakes et al., 2009, Lin et al., 2011). Because of their higher predictive accuracy, better ability to generalise from data, and possibility to handle large heterogeneous data sets, algorithmic models are gaining growing attention in ecology (Leathwick et al., 2006, De’ath and Fabricius, 2000) and land change science (Müller et al., 2013, Gellrich et al., 2008), but no study has so far used BRTs to assess spatial determinants of forest harvesting intensity.
In this study, we sought to quantify and understand broad-scale spatial determinants of forest harvesting intensity patterns across the European Union (EU-27) plus Norway and Switzerland. As intensity metric, we used the ratio of harvested timber volume (fellings and harvest losses) and net annual increment volume (hereafter referred to as “forest harvesting intensity”) because this ratio is an important criterion to assess the sustainability of forest resource use. As explanatory variables, we focused on selected factors that are indirect proxies of the underlying drivers of forest harvesting intensity (hereafter referred to as “spatial determinants”).
Europe is an interesting case for assessing forestry intensity since forest use in Europe has a long history. After centuries of extensive deforestation, Europe’s forests increased in the 19th and 20th century as a result of farmland abandonment, afforestation, and nature protection (Kaplan et al., 2012, Rudel et al., 2005), and forests now cover 37% of Europe’s terrestrial surface. Though forest cover has increased steadily during the last decades (0.37% per year, Forest Europe and UNECE FAO, 2011), forest harvesting intensity also remarkably increased from 58% (1990) to 62.4% (2010) and is expected to increase further (UNECE FAO, 2011, Böttcher et al., 2012). Forest cover is distributed very unevenly across Europe and the region is furthermore characterised by large environmental (e.g., boreal to Mediterranean), historical (e.g., capitalism vs. socialism), ethnic, and economic (highly industrialised vs. less industrialised economies) heterogeneity. How this heterogeneity relates to spatial patterns in forest harvesting intensity remains largely unclear. Understanding forest harvesting intensity is one key aspect for assessing forest management intensity. To ensure the sustainable intensification of forest management in light of growing demands for timber products would, however, require a range of indicators addressing the multidimensionality of forest management intensity.
We compiled time series of sub-national forest harvesting intensity patterns in Europe between 2000 and 2010 and used boosted regression trees to quantify the influence of a set of biophysical, infrastructure, and socioeconomic variables in shaping these patterns. Specifically, we ask the following research questions:
- 1.
What are the spatial patterns of forest harvesting intensity in Europe?
- 2.
What are the most influential spatial determinants of these patterns and what is their relative importance?
- 3.
What is the nature of the relationships between forest harvesting intensity and its spatial determinants?
Section snippets
Forest harvesting intensity
To estimate forest harvesting intensity, we collected sub-national forest harvesting statistics (m3/ha), net annual increment (m3/ha), and forest area (ha) from national forestry reports, statistical yearbooks and databases, and by contacting national experts. Statistics were collected for the entire EU-27 plus Norway and Switzerland and – when possible – for each year between 2000 and 2010 (see Tables A.1–A.3 for a full list of references). Data were collected for administrative units ranging
Patterns of forest harvesting intensity
The spatial patterns of average harvested timber volumes on the one hand, and our forest harvesting intensity index on the other hand differed substantially (Fig. 1). For example, southern Germany had generally high harvested volume levels (i.e., harvested timber volume per hectare forest), but relatively low forest harvesting intensity due to high forest productivity, whereas in southern Finland high forest harvesting intensity occurred despite lower harvest levels. Generally, harvested timber
Discussion and conclusion
Although the importance of forest management intensity to address sustained yield has been recognised long ago (von Carlowitz, 1713), quantitative, broad-scale assessments of the drivers and spatial patterns of forest management intensity have been missing. Here, we derived forest harvesting intensity patterns as one indicator for forest management intensity for all of Europe using, a system metrics relating the outputs from forestry (i.e., harvests) to ecosystem productivity (i.e., net
Acknowledgements
We gratefully acknowledge support by the European Commission (Integrated Project VOLANTE FP7-ENV-2010-265104) and the Einstein Foundation Berlin. We would like to thank all national experts who provided data on regional harvested timber volumes, increment volumes, and forest area, Christoph Plutzar for providing data on terrain ruggedness, and Milenka Hampel and Christoph Israel for their help with data pre-processing. We thank three anonymous reviewers for their comments and suggestions that
References (100)
- et al.
Evaluating forest management intensity: a comparison among major forest landowner types
For. Pol. Econ.
(2006) - et al.
Patterns and drivers of post-socialist farmland abandonment in Western Ukraine
Land Use Pol.
(2011) - et al.
Econometric studies of non-industrial private forest management: a review and synthesis
For. Pol. Econ.
(2005) - et al.
Biodiversity, disturbances, ecosystem function and management of European forests
For. Ecol. Manage.
(2000) How a socio-ecological metabolism approach can help to advance our understanding of changes in land-use intensity
Ecol. Econ.
(2012)Resilience: the emergence of a perspective for social-ecological systems analyses
Glob. Environ. Change
(2006)- et al.
EU-wide maps of growing stock and above-ground biomass in forests based on remote sensing and field measurements
For. Ecol. Manage.
(2010) - et al.
Combining classification tree analyses with interviews to study why sub-alpine grasslands sometimes revert to forest: a case study from the Swiss Alps
Agric. Syst.
(2008) Rationale and methods for conserving biodiversity in plantation forests
For. Ecol. Manage.
(2002)- et al.
How strongly can forest management influence soil carbon sequestration?
Geoderma
(2007)
Forest restitution and protected area effectiveness in post-socialist Romania
Biol. Conserv.
The causes of land-use and land-cover change: moving beyond the myths
Glob. Environ. Change
A multidisciplinary multi-scale framework for assessing vulnerabilities to global change
Int. J. Appl. Earth Obs. Geoinf.
Spatiotemporal modeling of the expansion of mechanized agriculture in the Bolivian lowland forests
Appl. Geogr.
Comparing the determinants of cropland abandonment in Albania and Romania using boosted regression trees
Agric. Syst.
Urbanization’s impact on timber harvesting in the south central United States
J. Environ. Manage.
Pan-European forest/non-forest mapping with Landsat ETM+ and CORINE Land Cover 2000 data
Isprs J. Photogram. Rem. Sens.
Forest transitions: towards a global understanding of land use change
Glob. Environ. Change
Analyzing determinants of forest owners’ decision-making using a sample selection framework
J. For. Econ.
Global food security, biodiversity conservation and the future of agricultural intensification
Biol. Conserv.
Land system architecture: using land systems to adapt and mitigate global environmental change
Glob. Environ. Change
Sensitising rural policy: assessing spatial variation in rural development options for Europe
Land Use Pol.
From land cover change to land function dynamics: a major challenge to improve land characterization
J. Environ. Manage.
The realisable potential supply of woody biomass from forests in the European Union
For. Ecol. Manage.
Reconstructed forest age structure in Europe 1950–2010
For. Ecol. Manage.
The effects of population growth on timber management and inventories in Virginia
For. Ecol. Manage.
Regional- and district-level drivers of timber harvesting in European Russia after the collapse of the Soviet Union
Glob. Environ. Change
Is the level of national forest timber harvest sensitive to price?
Land Econ.
Econometric analyses of nonindustrial forest landowners: is there anything left to study?
J. For. Econ.
Projection of the future EU forest CO2 sink as affected by recent bioenergy policies using two advanced forest management models
GCB Bioenergy
Random forests
Mach. Learn.
Statistical modeling: the two cultures
Stat. Sci.
Plantation forests and biodiversity: oxymoron or opportunity?
Biodivers. Conserv.
Statistical mapping of tree species over Europe
Eur. J. For. Res.
Carbon accumulation in European forests
Nat. Geosci.
A new method for dealing with residual spatial autocorrelation in species distribution models
Ecography
Complex land systems: the need for long time perspectives to assess their future
Ecol. Soc.
Classification and regression trees: a powerful yet simple technique for ecological data analysis
Ecology
Collinearity: a review of methods to deal with it and a simulation study evaluating their performance
Ecography
Classification of forest management approaches: a new conceptual framework and its applicability to European forestry
Ecol. Soc.
Land-sharing versus land-sparing logging: Reconciling timber extraction with biodiversity conservation
Glob. Change Biol.
Novel methods improve prediction of species’ distributions from occurrence data
Ecography
Cited by (140)
Does China's natural forest logging ban affect the stability of the timber import trade network?
2023, Forest Policy and EconomicsMachine Learning Forest Simulator (MLFS): R package for data-driven assessment of the future state of forests
2023, Ecological InformaticsAutomatic forest change detection through a bi-annual time series of satellite imagery: Toward production of an integrated land cover map
2023, International Journal of Applied Earth Observation and GeoinformationGoverning Europe's forests for multiple ecosystem services: Opportunities, challenges, and policy options
2022, Forest Policy and EconomicsCitation Excerpt :They provide European society with multiple forest ecosystem services (FES), which include provisioning services (e.g., wood for construction and energy, non-wood forest products), regulating services (e.g., local and global climate mitigation, hydrological regulation and soil protection) and cultural services (e.g., recreational and health benefits). Most forests in the EU are privately owned (58%), and a large share of the EU's forests is actively managed, in many cases primarily for wood production, but with substantial regional variation (Levers et al., 2014). Forests offer employment and income along various value chains relating to wood and non-wood forest products, and multiple other FES (Winkel, 2017), including recreation and nature-based tourism (Tyrväinen et al., 2017a).
Education and sustainable forest management in the mid-hills of Nepal
2022, Journal of Environmental Management