Nitrogen deposition is the most important environmental driver of growth of pure, even-aged and managed European forests
Introduction
European temperate and boreal forests represent a globally important component of the terrestrial carbon sink and pool (Luyssaert et al., 2010). Their total annual forest volume increment (ΔVol) is approx. 840 × 106 m3 and accounts for an annual increase in biomass carbon stock of 167 Mt between 2005 and 2015 (Forest Europe, 2015). Understanding which factors drive and modulate forest growth is therefore key in ecological and economic perspectives, in Europe and beyond. Uncertainty exists, however, on the role of altered environmental drivers, such as climate and air pollution, in affecting forest production at the continental scale (Zak et al., 2011, Lindner et al., 2014, De Vries et al., 2017a, De Vries et al., 2017), especially in combination (interaction) with site- and stand characteristics.
Increasing average surface temperature in Europe, as well as the increasing frequency and intensity of drought and heat events may lead to highly uncertain and diverse effects on forest productivity (Fischer and Schär, 2010, IPCC, 2014). While tree growth in temperature-limited ecosystems might (Jolly et al., 2005) or might not (Wilmking et al., 2004) benefit from increasing temperatures, water-limited ecosystems, such as the Mediterranean forests, will most likely suffer from the increased drought and heat as indicated by recent studies (Sarris et al., 2011, Fyllas et al., 2017). At the same time, changes in nitrogen deposition (Ndep) may have divergent effects on forest productivity (Waldner et al., 2014, Engardt et al., 2017, Vuorenmaa et al., 2018), although the real size of its impact is still being debated (Sutton et al., 2008). On N-limited sites additional Ndep may be beneficial to growth. On non N-limited sites, additional Ndep may lead to nutrient imbalances, including phosphorus (P) deficiency, nitrate (NO3–) leaching, acidification and loss of base cations, and increased sensitivity to pests or pathogens (De Vries et al., 2014a, Gessler et al., 2017). High Ndep can therefore potentially also reduce forest growth (De Vries et al., 2014a). In this context, not only the actual N deposition is relevant, but also the cumulated Ndep during past decades (Waldner et al., 2014, Engardt et al., 2017, Vuorenmaa et al., 2018), which might have altered the soil and nutrient conditions of the forest site (Aber et al., 1998). Alongside, the impact of ground-level ozone (O3) on forest growth is of concern (e.g. Ainsworth et al., 2012). Ozone is the greenhouse gas with the third strongest radiative force (Forster et al., 2007) and its ground-level concentration doubled globally during the last century. Although more recently ozone trends levelled off in some regions (Cooper et al., 2014), concentration across Europe is high enough to pose a potential risk for forests (Ainsworth et al., 2012). The actual impact of O3 on mature forests growth in Europe, however, remains uncertain (Cailleret et al., 2018) with observational studies providing contradicting results (see Karlsson et al., 2006, Braun et al., 2014, Braun et al., 2017, but also Kint et al., 2012, Verryckt et al., 2017, Ferretti et al., 2018).
Beyond environmental drivers, stand characteristics such as species composition, stand density and stand age, play a crucial role for forest growth. For example, forest growth usually levels off beyond a site- and species-specific threshold of site density due to competition (Long et al., 2004), and is non-linearly (quadratic) related to forest age (Gower et al., 1996). By modulating stand density and species composition, and regulating rotation length, forest management can influence resource availability, competition, stand age, and the response of forests to environmental drivers, such as climate conditions or air pollution, and therefore forest growth and productivity (Bettinger et al., 2017). Not surprisingly, management resulted equal or even more important for recent forest growth than environmental drivers (Altman et al., 2013, Foster et al., 2015, Henttonen et al., 2017, Maes et al., 2019).
Since the majority of European forest is managed (McGrath et al., 2015), it is therefore important to evaluate the impact on forest growth resulting from complex interactions among environmental drivers (e.g. Simpson et al., 2014, Gessler et al., 2017), and between such drivers and management. There are several approaches to evaluate the role of environmental drivers on forest ecosystems, such as controlled experiments.(e.g. Liu and Greaver, 2009, Wittig et al., 2009), field-based observational studies across gradients (e.g. Solberg et al., 2009, Braun et al., 2017), or model simulations predicting responses to environmental change (e.g. Churkina et al., 2009, De Vries et al., 2017). However, many studies are restricted on a smaller regional (sub-national) extent, and multi-factorial studies including interactions of drivers and sufficient replication are scarce. This makes field observational studies essential to detect actual large-scale, long-term patterns. The network of monitoring plots installed under the auspices of the International Co-operative Programme on Monitoring and Assessment of Air Pollution Effects on Forests (ICP Forests) provides a unique opportunity to study the combined and concurrent impacts of multiple factors, including climate, air pollution, site- and stand-properties on growth of managed forest systems (Ferretti et al., 2010).
We conducted a continental-scale analysis of ICP Forests growth data obtained over the period 1995–2010 from nearly 100,000 trees (n = 99,432) distributed in 442 even-aged, almost pure beech-, oak-, spruce- and pine-dominated permanent observation plots distributed in managed forests across 23 countries in Europe (Fig. 1). Even-aged forests represent approx. 70% of all European forest area, with single-tree species forests accounting for approx. 30% of the European forest area (Forest Europe, 2015). We hypothesize that: (i) stand characteristics have a generalized effect on forest growth, and that climate influence is species-specific; (ii) the effect of N deposition on growth is non-linear, turning from a positive relationship with growth at low deposition to a negative one at high deposition by altering site quality (foliar nutrient concentrations, soil pH), especially on N saturated sites (De Vries et al., 2014a); and that (iii) ozone has a generalized negative relationship with forest growth, more pronounced on deciduous broadleaves (hereafter termed in the text as “broadleaves”) (Büker et al., 2015). Its effect may be offset by Ndep or temperature because of multi co-linearity or counteracting effects (See Fig. 2).
We tested these hypotheses by considering 11 different predictors including environmental drivers (climate and air quality), site and stand characteristics (stand density, stand age, altitude) and site quality (foliar nutrients, soil pH) and using independent statistical approaches (non-linear modelling, linear mixed-effects models (LME), and structural equation modelling (SEM) to account for multiple and counteracting pathways, interactions and non-linearity. We formulated a theoretical SEM model (see below) based on existing knowledge, and we tested such a model against observations. Further, in order to see whether results depend on the data source, we considered two data sets, with Ndep data derived from modelled data and from bulk deposition measurements. We expect to detect whether, and to what extent, environmental drivers, such as climate (precipitation, temperature, drought events) and air quality (Ndep, ozone) have direct and/or indirect (e.g. by altering nutrient availability and soil pH) impacts on growth once stand characteristics (stand density, age, altitude) are accounted for. To our knowledge, when considering its spatial, temporal and ecological domains, this is the largest study of this type carried out in Europe.
Section snippets
Study sites
The plots selected for the study were part of the ICP Forests Level II network established under the UNECE Convention on Long-Range Transboundary Air Pollution (Air Convention). These plots have been selected upon various criteria (e.g. coverage of forest types, geographic extent), in order to investigate relationships between environmental, site and stand predictors and response variables linked to forest health, growth, and diversity (Ferretti, 2013). From approximately 760 available plots
Annual volume increment and correlations with predictors
Mean annual forest increment ΔVol from 1995 to 2000 across the examined plots was 7.21 ± 0.22 m3 ha−1 yr−1 (ranged from 1.4 up to 19 m3 ha−1 yr−1), with the highest values for spruce forests (p < 0.05, Table 2) and for central Europe (Fig. S1). No consistent pattern of ΔVol over time (1995–2010) appeared among the four species (Fig. 3). While ΔVol of oak forests continuously increased by almost 3% per year (p < 0.01), there has been a significant decrease for spruce forests since 2005 (–1.2%
Discussion
The set of forest monitoring data collected on a harmonized basis including 23 European countries and over a 15-year period provides a unique opportunity to study patterns of forest growth at continental scale, complementing findings derived from individual case studies and local assessments. Even though observational studies like the present one brings challenges for interpretation such as non-linearity, collinearity and heterogeneity, some clear patterns on the continental-scale emerged that
Conclusions
Our analysis of recent continental-scale forest growth data provided further evidence that management-related stand characteristics like stand density index and stand age are key to explain annual stand volume increment over the period 1995–2010. Besides, we detected for some species a clear relationship of growth with atmospheric N deposition, with an overall positive, but in some cases non-linear (quadratic) response with a tipping point at 24–34 kg N ha−1 yr−1. The most consistent climate
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This study was part of the ÉCLAIRE project (Effects of Climate Change on Air Pollution and Response Strategies for European Ecosystems) funded by the EU's Seventh Framework Programme for Research and Technological Development (FP7, Grant Agreement Number 282910). The evaluation was based on data that was collected by partners of the official UNECE ICP Forests Network (http://icp-forests.net/contributors). Part of the data was co-financed by the European Commission (Data achieved at 20141212).
References (85)
- et al.
Regional differences in climatic responses of Norway soruce (Picea abies L.Karst) growth in Norway
For. Ecol. Manag.
(2006) - et al.
Growth losses in Swiss forests caused by ozone. Epidemiological data analysis of stem increment of Fagus sylvatica L. and Picea abies Karst
Environ. Pollut.
(2014) - et al.
Growth trends of beech and Norway spruce in Switzerland: the role of nitrogen deposition, ozone, mineral nutrition and climate
Sci. Total Environ.
(2017) - et al.
New flux based dose-response relationships for ozone for European forest tree species
Environ. Pollut.
(2015) - et al.
Short and long-term impacts of nitrogen deposition on carbon sequestration by forest ecosystems
Curr. Opin. Environ. Sustain.
(2014) - et al.
Modelling long-term impacts of changes in climate, nitrogen deposition and ozone exposure on carbon sequestration of European forest ecosystems
Sci. Total Environ.
(2017) - et al.
Tree growth in Swiss forests between 1995 and 2010 in relation to climate and stand conditions: recent disturbances matter
For. Ecol. Manag.
(2014) Concepts and Design Principles Adopted in the International Cooperative Program on the Assessment and Monitoring of Air Pollution Effects on Forests (ICP Forests)
- et al.
Tree growth-climate relationships in a forest-plot network on Mediterranean mountains
Sci. Total Environ.
(2017) - et al.
Aboveground net primary production decline with stand age: potential causes
Trends Ecol. Evol.
(1996)