ArticlePDF AvailableLiterature Review

Strategic roadmap to assess forest vulnerability under air pollution and climate change

August 2022
Global Change Biology 28:5062-5085

DOI:10.1111/gcb.16278

License
CC BY 4.0

Authors:

Pierre Sicard

Argans

Evgenios Agathokleous

Nanjing University of Information Science & Technology

Show all 31 authorsHide

Although it is an integral part of global change, most of the research addressing the effects of climate change on forests have overlooked the role of environmental pollution. Similarly , most studies investigating effects of air pollutants on forests have generally neglected impacts of climate change. We review the current knowledge on combined air pollution and climate change effects on global forest ecosystems and identify several key research priorities as a roadmap for the future. Specifically, we recommend 1) establishment of much denser array of monitoring sites, particularly in the South Hemisphere; 2) further integration of ground and satellite monitoring; 3) generation of flux-based standards and critical levels taking into account the sensitivity of dominant forest tree species; 4) long-term monitoring of N, S, P cycles and base cations deposition together at global scale; 5) intensification of experimental studies, addressing combined effects of different abiotic factors on forests by assuring a better representation of taxonomic and functional diversity across the ~ 73,000 tree species on Earth; 6) more experimental focus on phenomics and genomics; 7) improved knowledge on key processes regulating the dynamics of radionuclides in forest systems; and 8) development of models integrating air pollution and climate change data from long-term monitoring programs.

Content uploaded by Alessio Collalti

Content may be subject to copyright.

Glob Change Biol. 2022;00:1–24.

1wileyonlinelibrary.com/journal/gcb

Received: 6 October 2021

Revised: 2 March 2022

Accepted: 18 May 2022

DOI: 10.1111/gcb.16278

REVIEW

Strategic roadmap to assess forest vulnerability under air

pollution and climate change

Alessandra De Marco1 | Pierre Sicard2 | Zhaozhong Feng3 | Evgenios Agathokleous3 |

Rocio Alonso4 | Valda Araminiene5 | Algirdas Augustatis6 | Ovidiu Badea7,8 |

James C. Beasley9 | Cristina Branquinho10 | Viktor J. Bruckman11 | Alessio Collalti12 |

Rakefet David- Schwartz13 | Marisa Domingos14 | Enzai Du15 | Hector Garcia Gomez4 |

Shoji Hashimoto16 | Yasutomo Hoshika17 | Tamara Jakovljevic18 | Steven McNulty19 |

Elina Oksanen20 | Yusef Omidi Khaniabadi21 | Anne- Katrin Prescher22 |

Costas J. Saitanis23 | Hiroyuki Sase24 | Andreas Schmitz25 | Gabriele Voigt26 |

Makoto Watanabe27 | Michael D. Wood28 | Mikhail V. Kozlov29 | Elena Paoletti16

1ENEA, CR Casaccia, SSPT- PVS, Rome, Italy

2ARGANS, Biot, France

3Key Laboratory of Agro- Meteorology of Jiangsu Province, School of Applied Meteorology, Nanjing University of Information Science & Technology, Nanjing, China

4Ecotoxicology of Air Pollution, CIEMAT, Madrid, Spain

5Lithuanian Research Centre for A griculture and Forestry, Kaunas, Lithuania

6Faculty of Forest Sciences and Ecology, Vytautas Magnus University, Kaunas, Lithuania

7“Marin Drăcea” National Institute for Research and Development in Forestry, Voluntari, Romania

8Faculty of Silviculture and Forest Engineering, “Transilvania” Universit y, Braşov, Romania

9Savannah River Ecology Laboratory and Warnell School of Forestry and Natural Resources, University of Georgia, Aiken, South Carolina, USA

10Centre for Ecology, Evolution and Environmental Changes, Faculdade de Ciências, Universidade de Lisboa, Lisbon, Portugal

11Commission for Interdisciplinary Ecological Studies, Austrian Academy of Sciences, Vienna, Austria

12Forest Modeling Lab., ISAFOM- CNR, Perugia, Italy

13Institute of Plant Sciences, ARO— Volcani Center, Rishon LeTsiyon, Israel

14Instituto de Botanica, Nucleo de Pesquisa em Ecologia, Sao Paulo, Brazil

15Faculty of Geographical Science, Beijing Normal University, Beijing, China

16Depar tment of Forest Soils, Forestry and Forest Products Research Institute, Tsukuba, Japan

17IRET- CNR, Sesto Fiorentino, Italy

18Croatian Forest Research Institute, Jastrebarsko, Croatia

19USDA Forest Service, Research Triangle Park, USA

20Depar tment of Environmental and Biological Sciences, University of Eastern Finland, Joensuu, Finland

21Department of Environmental Health Engineering, Industrial Medial and Health, Petroleum Industry Health Organization (PIHO), Ahvaz, Iran

22Thuenen Institute of Forest Ecosystems, Eberswalde, Germany

23Lab of Ecology and Environmental Science, Agricultural University of Athens, Athens, Greece

24Ecological Impact Research Department, Asia Center for Air Pollution Research (ACAP), Niigata, Japan

25State Agency for Nature, Environment and Consumer Protection of North Rhine- Westphalia, Recklinghausen, Germany

26r.e.m. Consulting, Perchtoldsdorf, Austria

27Institute of Agriculture, Tokyo University of Agriculture and Technology (TUAT), Fuchu, Japan

28School of Science, Engineering and Environment, University of Salford, Salford, UK

29Department of Biology, University of Turku, Turku, Finland

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,

provided the original work is properly cited.

DE MARCO et al.

1 | INTRODUC TION

Forests cover ~30% of the world's land surface, store 45% of terres-

trial carbon (Bonan, 2008), and are home to 80% of global terrestrial

biodiversity (IUCN, 2021). Sustainable socioeconomic development

depends on the proper management of natural resources, includ-

ing forest ecosystems (Badea et al., 2013). Air pollution and climate

change have major impacts on and complex interactions with forest

health and productivity (Augustaitis & Bytnerowicz, 2008; Kozlov

et al., 2009). For example, tropospheric ozone (O3), which is both

a phytotoxic gas and a radiative forcer (Myhre et al., 2013), and ni-

trogen deposition (Du & de Vries, 2018), which causes forest de-

cline due to acidification (Augustaitis et al., 2010) and changes in

the frequency and intensity of climatic extremes (e.g., heat waves,

rainfall, wind storms), may impact the structure, composition, and

functioning of terrestrial ecosystems. These impacts can directly in-

fluence carbon cycling and its feedback to the climate system (Frank

et al., 2015; Matyssek et al., 2012; Paoletti et al., 2007; Serengil

et al., 2011; Sicard et al., 2020).

The future of global forests is a subject of public and politi-

cal concern due to extensive forest degradation worldwide (Hao

et al., 2018; Liu et al., 2018). Recently, environmental pollution

was identified as one of the five main drivers of biodiversity loss

(European Commission, 2020). Although environmental pollution

is an integral part of global change (Dale et al., 2000), most of the

research addressing the biotic effects of climate change do not con-

sider this issue. Furthermore, most studies on both the distribution

of pollutants and the biotic effects of pollution have neglected the

issue of climate change (Sicard, Augustaitis, et al., 2016). As a result,

studies exploring the combined effects of air pollution and climate

change remain uncommon.

A Web of Science search conducted in June 2021 identified only

74 peer- reviewed articles containing the keywords “climat* and pol-

lut*” and “tree* or forest*” in the title, 59 of which were relevant re-

search papers (Figure S1): In all, 11 studies used modeling to explore

the combined effects of air pollution and climate, 27 studies were

based on observations of forest health in either spatial or tempo-

ral gradients of air pollution and climate, and only one reported the

outcomes of a field experiment. The low number of experimental

studies with factorial design involving both airborne pollutants and

climate is alarming because it hampers our ability to identify cause-

and- effect relationships as well as to decipher the mechanisms un-

derlying the combined or interactive effects of pollution and climate

on the health of individual trees and forest ecosystems. As a result,

the quality of our predictions of the combined effects of climate

change and air pollution on future forest health is uncertain. To re-

spond to this global challenge, here we critically review the current

knowledge (and gaps) on air pollution and climate interactions in for-

ests, identify key research priorities, and suggest a strategic road-

map for future studies.

Correspondence

Alessandra De Marco, ENEA Casaccia, Via

Anguillarese 301, 00123 Rome, Italy.

Email: alessandra.demarco@enea.it

Funding information

Academy of Finland, Grant/Award

Number: 276671, 311929 and 316182;

European Commission, Grant/Award

Number: LIFE15 ENV/IT/000183, LIFE19

ENV/FR/000086 and LIFE20 GIE/

IT/000091; JST SICORP, Grant/Award

Number: JPMJSC16HB; US Department of

Energy, Grant/Award Number:

DE- EM0005228

Abstract

Although it is an integral part of global change, most of the research addressing the

effects of climate change on forests have overlooked the role of environmental pollu-

tion. Similarly, most studies investigating the effects of air pollutants on forests have

generally neglected the impacts of climate change. We review the current knowledge

on combined air pollution and climate change effects on global forest ecosystems

and identify several key research priorities as a roadmap for the future. Specifically,

we recommend (1) the establishment of much denser array of monitoring sites, par-

ticularly in the South Hemisphere; (2) further integration of ground and satellite

monitoring; (3) generation of flux- based standards and critical levels taking into ac-

count the sensitivity of dominant forest tree species; (4) long- term monitoring of

N, S, P cycles and base cations deposition together at global scale; (5) intensification

of experimental studies, addressing the combined effects of different abiotic factors

on forests by assuring a better representation of taxonomic and functional diversity

across the ~73,000 tree species on Earth; (6) more experimental focus on phenomics

and genomics; (7) improved knowledge on key processes regulating the dynamics of

radionuclides in forest systems; and (8) development of models integrating air pollu-

tion and climate change data from long- term monitoring programs.

KEYWORDS

air pollution, climate change, forest ecosystem, forest nutrients, forest research roadmap,

forest vulnerability, radioactivity

DE MARCO et al.

2 | ASSESSING AIR POLLUTION:

RESEARCH INFRASTRUCTURES AND

METHODOLOGIES FOR FOREST

MONITORING

Re giona l and nati ona l air qua lit y dir ec tive s and em issi ons contr ol po l-

icies (e.g., Japanese Air Pollution Control Act 1968/1970; European

Council Directive 2008/50/EC; United States Federal Register,

2015) led to the development of air quality monitoring stations.

Monitoring data are collated within national or regional databases,

such as the Acid Deposition Monitoring Network in East Asia, the

European Environment Agency Airbase system, and the Australia Air

Quality Network (AUSAQN; Schultz et al., 2017 ). Despite efforts to

moni tor air quality in Sout h America , the spat ial dist rib uti on of moni-

toring stations is still heterogeneous and insufficient to represent

the pollutant levels (Peláez et al., 2020).

Coordinated research networks of long- term experimental for-

est sites integrating monitoring and state- of- the- art methodological

and conceptual research to assess air pollution and global change

effects are not distributed in a way that represents all forest eco-

system types over the globe. Long- term forest monitoring and in-

frastructure networks are running regionally and worldwide, even

overlapping each other in their geographic expansions, and are likely

to further expand in the future. Here, we introduce some of the larg-

est networks of experimental forest sites, their research aims and

methodologies, and explore their capacities in view of the Supersite

definition (Mikkelsen et al., 2013).

International Long- Term Ecological Research (ILTER) is a “net-

work of networks” with research sites located in a wide array of

ecosystems aimed at developing a global understanding of envi-

ronmental change while also covering socioeconomical aspects

(known as LTSER). Expertise warrants the collection, management,

and analysis of spatiotemporally diverse datasets, such as DEIMS

(Drupal Ecological Information Management System), a central plat-

form providing information on sites and networks with ecological

long- term monitoring and experimentation at European and global

scales. Currently, ILTER encompasses 39 countries which together

operate more than 600 sites (Maass et al., 2016). Some sites main-

tain advanced continuous measurements, such as tower- based eddy

covariance assessments of CO2 and H2O fluxes. The ILTER network

includes the Ter res trial Eco sys tem Re search Netwo rk (TE RN), est ab-

lished in Australia, which provides a comprehensive metadata portal

containing information on continental scale datasets of measure-

ments describing fauna, flora, terrestrial ecosystems, ecological dy-

namics, land surfaces, soils, agricultural ecosystems, coasts, climate

observations and fluxes (Karan et al., 2016). Similarly, the Chinese

National Ecosystem Research Network (CNERN) is an integrated

platform of field stations supervised by various Chinese ministries.

CNERN represents a science and technology system that conducts

network observation and experimentation across China's ecosys-

tems, cutting across governmental departments, industrial sectors,

regions, and jurisdictions, and seeks to integrate observation equip-

ment and data resources and standardize research methods, tools,

and protocols. As a result, CNERN serves as a nexus for national

ecological research, promotes data sharing, and creates an educa-

tional center and collaborative base for ecological research. ILTER

networks are also present in Korea and Taiwan.

Another “network of regional networks” is represented by

FLUXNET, which is coordinating regional and global analyses con-

ducted at micrometeorological tower sites (eddy covariance) to

investigate the exchanges of carbon dioxide (CO2), water vapor,

and energy between terrestrial ecosystems and the atmosphere

(Pastorello et al., 2020). FLUXNET is divided into regional networks,

for example, the European Integrated Carbon Observation System

Research Infrastructure (ICOS RI) with more than 100 measuring

stations including 32 forest stations. In 2021, more than 800 sites

worldwide were operated on a long- term and continuous basis

within this network. Habitats included in this monitoring framework

include temperate conifer and broadleaf (deciduous and evergreen)

forests, tropical and boreal forests, crops, grasslands, chaparral,

wetlands, and tundra.

In Europe, the International Co- operative Programme on

Assessment and Monitoring of Air Pollution Effects was launched

in 1985 under the United Nations Convention on Long- Range

Transboundary Air Pollution (CLRTAP), with several units including

ICP Forests (Michel et al., 2018), ICP Vegetation, ICP Modelling and

Mapping, and ICP Integrated Monitoring (Lundin & Forsius, 2004).

Networks of monitoring stations are established within this frame-

work that continuously assess ecosystem responses to air pollution

and develop the associated modeling and assessment methods

(Forsius et al., 2021). ICP Forests currently monitor forest conditions

in Europe at two intensities: Level I is based on around 6000 ob-

servation plots within a systematic transnational grid of 16 × 16 km2.

Level II comprises around 800 plots in selected forest ecosystems

for clarifying cause– effect relationships, and also assesses foliar and

soil chemistry, tree growth, and conditions of ground vegetation.

Approximately 41 sites, depending on the parameters, also monitor

ambient air quality and meteorology.

ForestGEO is a global network of scientists and forest research

sites dedicated to advancing long- term study of the world's forests,

dedicated to the study of tropical and temperate forest function and

diversity. The multi- institutional network comprises 73 forest re-

search sites across the Americas, Africa, Asia, Europe, and Oceania.

ForestGEO monitors the growth and survival of approximately 6 mil-

lion trees and nearly 13,000 species that occur in the forest research

sites. This network also supports initiatives to monitor attributes

such as climate, carbon flux, vertebrates, insects, and soil microor-

ganisms. ForestGEO increases scientific understanding about the

potential effects of climate change on ecosystems, which is a priority

of the US Climate Change Science Program and highlighted by the

Intergovernmental Panel on Climate Change (IPCC) Working Group

II. Because of ForestGEO's extensive biological monitoring, unique

databases, and the partners' expertise, it promises to enhance soci-

ety's ability to evaluate and respond to the impacts of global climate

change. To date, unfortunately, the distribution of forest monitor-

ing sites within ForestGEO appears non- homogeneous (Figure 1).

DE MARCO et al.

Indeed, boreal and tropical forest s are less represented among mon-

itoring sites and there is a disproportionate number of monitoring

sites in the Northern Hemisphere (NH), particularly in Europe, and

fewer sites in the Southern Hemisphere (SH).

The aforementioned monitoring networks may benefit from

data derived through remote sensing measurements (Lechner

et al., 2020). Remotely sensed imagery provides a synoptic view, and

is potentially available everywhere at a large range of spatial and

temporal scales with a high degree of homogeneity. Furthermore,

remote imagery provides digital images that can easily be integrated

with other spatial datasets in a geographic information system, and

per unit area remote sensing is an inexpensive way to acquire data.

The most used remote sensing sensors for assessing and monitoring

forest conditions are those on- board satellites, followed by airborne

(including Unmanned Aerial Vehicles) and terrestrial systems, or a

combination of these platforms (Torres et al., 2021). Previous studies

have demonstrated the utility of optical remote sensing for assessing

a variety of forest health indices, and are commonly used in forest

monitoring activities (Curran et al., 1992; Huang et al., 2019 ; Parent

& Verbyla, 2010 ). Landsat satellite images are still the most widely

used Earth Observation (EO) data in forest health studies (Torres

et al., 2021), which provide continuous time series data from the

1970s (i.e., Landsat 1 mission) until today (i.e., Landsat 8). Access to

Landsat images has been free since 2008, and the recently launched

Landsat 9 (September 2021) will be publicly available in early 2022.

In addition to Landsat imagery, imagery from sentinel missions

from the European Space Agency is also particularly important for

forest monitoring because of their high spatial and temporal reso-

lution. Furthermore, the availability of both active (Sentinel- 1) and

passive (Sentinel- 2) sensors might increase the precision of previ-

ous analytical methods that rely primarily on optical reflectance

indices. Similarly, forest health monitoring studies are increasingly

using Synthetic Aperture Radar (SAR) sensors. For example, C- band

data are sensitive to variations of Leaf Area Index, which are con-

nected to defoliation and hence forest status (Manninen et al., 2003;

Stankevich et al., 2017). SAR sensors are advantageous not just

because of their sensitivity to forest structural changes (Dobson

et al., 1992; Harrell et al., 1995; Le Toan et al., 1992), but also be-

cause of their ability to monitor the water content of the tree canopy

(Dobson et al., 199 2; Harrell et al., 1995; Le Toan et al., 199 2).

Specific remote sensing techniques that merge different spatial,

spectral, radiometric, and temporal resolutions are being increas-

ingly used to reduce data gaps and to characterize forest ecosystems

(Lausch et al., 2018). For example, Rogers et al. (2018) demonstrated

the potential of derived products based on Landsat, Advanced

Very High- Resolution Radiometer (AVHRR), and MODIS (Moderate

Resolution Imaging Spectroradiometer) data to detect early signals

of tree mortality. Modeling various biophysical indicators based on

aerial or ground- based LiDAR data can further expand the portfo-

lio of remote sensing- derived data, or at the very least allow their

validation in a more efficient manner than by means of traditional

monitoring and inventory. In this regard, a fusion of satellite spectral

data (e.g., Sentinel- 2) and LiDAR data (Global Ecosystem Dynamics

Investigations) could be the next step for global drought- induced

tree mortality assessment (Huang et al., 2019). More recently, the

Copernicus air- pollution monitoring satellite dedicated to trace gas-

ses assessment, such as O3, NO2, SO2, formaldehyde (HCHO), CO,

and CH4 (Sentinel- 5— Precursor/TROPOMI; Inness et al., 2019),

has been used for tracking pollution events and pollution sources

(Mesas- Carrascosa et al., 2020). By merging classical monitoring

techniques and state- of- the- art remote sensing, long- term studies

are facilitated (Tănase et al., 2019). Remote sensing use should be

FIGURE 1 Distribution of the most relevant monitoring network over the forested areas of the globe.

DE MARCO et al.

expand ed to vulnerable regio ns or ecosystem ty pes whi ch need spe-

cial protection from climate change and air pollution.

Highly instrumented forest research infrastructures (supersites)

provide long- term data series and promote integration of research

communities in a transcontinental collaboration network (Fischer

et al., 2011). For these supersites, the use of forest inventory data

together with remote sensing and EO data can provide valuable in-

formation on forest condition (Hartmann et al., 2018). As new forest

change detection algorithms based on EO sensors are developed,

they can be validated using data from long- term monitoring net-

works (Rodman et al., 2021).

To understand climate change and weather extremes, it is im-

portant to have observations of the Earth system going back as far

as possible in time. Reanalysis combines past short- range weather

forecasts with observations through data assimilation (Uppala

et al., 2005). The process mimics the production of day- to- day

weather forecasts. Reanalyses are usually produced at lower resolu-

tion than current weather forecasts, and they use the same modern-

data assimilation system and forecasting model throughout the

reanalysis period. The latest European Centre for Medium- Range

Weather Forecasts (ECMWF) reanalyses are produced through the

EU- funded Copernicus Climate Change Service (C3S). Forecasts are

freely available through the C3S Climate Data Store. The most recent

ECMWF reanalysis dataset is the ER A5 Back Extension, providing

data from 1950 to 1978. The Copernicus Atmosphere Monitoring

Service (CAMS) provides continuous data and information on atmo-

spheric composition. The service describes the current situation,

forecasts the situation a few days ahead, and analyses consistently

retrospective data records for recent years. CAMS supports many

applications in a variety of domains including health, environmen-

tal monitoring, renewable energies, meteorology, and climatology.

CAMS monitors and forecasts European air quality and worldwide

long- range transport of pollutants.

3 | ELEMENT DEPOSITION IN GLOBAL

FORESTS

Various substances emitted from natural or anthropogenic sources

flow from the atmosphere into forest ecosystems by either wet or

dry deposition (Tørseth et al., 2012). Atmospheric deposition may

be harmful or beneficial for trees and other plants (Figure 2). Sulfur

(S) and nitrogen (N) compounds may function as either nutrients or

stressors for forests, even though they are derived from anthro-

pogenic air pollutants, such as sulfur oxides (SOx), nitrogen oxides

(NOx), and ammonia (NH3; Duan et al., 2016; Oksanen & Kontunen-

Soppela, 2021). When traveling through the canopy, acid deposi-

tion can cause direct damage to plant leaves (Du et al., 2017 ). When

deposited to the forest floor, N and S compounds are identified

as a cause of acidification and eutrophication (or N saturation) of

forest ecosystems (de Vries, 2021). Moreover, certain amounts of

phosphorus (P) and basic cations, such as calcium (Ca2+) and mag-

nesium (Mg2+), acting in forests as nutrients, are also derived from

anthropogenic emissions (Du et al., 2016, 2018). Climate change may

FIGURE 2 Main interactions of forest ecosystems with sulfur (S) and nitrogen (N) compounds. They may function as either stressors

(S) or nutrients (N), even when they are derived from anthropogenic air pollutants, such as sulfur oxides (SOx), nitrogen oxides (NOx), and

ammonia (NH3), with direct effects on forest canopy (Du et al., 2017 ) and indirect effects on acidification (Augustaitis et al., 2010) and

eutrophication (de Vries, 2021) including impacts on biodiversity (Clark et al., 2013), growth (Du et al., 2018), volatile emissions (Hansen

et al., 20 17; Liu & Greaver, 2009; Mushinski et al., 2 019; Schindler et al., 2020; Xie et al., 2018), and biogeochemistry (Gaudio et al., 2015;

Nakahara et al., 2010).

DE MARCO et al.

directly or indirectly affect the roles of these substances in forest

ecosystems (e.g., Mitchell & Likens, 2011; Nakahara et al., 2010).

Atmospheric deposition, especially of S and N compounds, has

declined over the last three decades (Sicard, De Marco, et al., 2016;

Tørseth et al., 2012; Zhong et al., 2020), despite many developing

nations still lacking effective SO2 emiss ion cont rol s. In Euro pe, depo-

sition of S and N peaked in the late 1970s and in the 1980s, respec-

tively (Engardt et al., 2017). In North America, deposition of S and

N peaked in the early 1970s and mid- 1990s, respectively (Mitchell

& Likens, 2011), when NH3 emission became more important

(Du, 2016). In South America, most average daily concentrations of

SO2 are below the World Health Organization air quality guidelines

(Peláez et al., 2020), and global atmospheric S deposition is lower

than in Europe, Asia, the United States, and Africa (Gao et al., 2018),

ranging around 4.96 ± 3.45 kgS ha−1 a−1 . In Asia, emissions of SO2

and NOx significantly increased from the early 1980s to the early

2000s (Ohara et al., 2007), 20 or 30 years later than in Europe and

the United States. The emissions of SO2 and NOx in China peaked in

2006 (Lu et al., 2011) and 2011– 2012 (Zheng et al., 2018), respec-

tively, and thereafter started decreasing. In China, emissions of NH3

reached a plateau in 1996 (Kang et al., 2016), although a gradual in-

crease in NH3 emissions in Asia (including China) was observed as of

2015 (Kurokawa & Ohara, 2020). A recent global analysis combined

inventory and modeling data to confirm that the total annual NOx

emissions finally stopped increasing in 2013, largely due to strict

control measures taken in China in recent years (Huang et al., 2017).

However, SO2 emissions in India overtook those in China in 2016 (Li

et al., 2017), and thus a focus should be placed on monitoring atmo-

spheric deposition in India and other developing countries. Major

air pollutants have been changing with industrialization in each re-

gion, from SO2 to NOX and NH3. With temporal changes of major

pollutants relative to industrialization, acidification, photochemical

formation of ozone, and excess N deposition appeared in sequence

as problems for forest ecosystems, as seen in the changes of main

causes of tree and forest decline in Northeast Asia (Takahashi

et al., 2020). Thus, emission reduction of S and/or N has been re-

flected gradually by the conditions of forest ecosystems.

In Europe and the United States, reduced S deposition resulted

in long- term declines in

SO2−

concentrations in soil solutions (Berger

et al., 2016; Johnson et al., 2018) and freshwater (Garmo et al., 2014;

Vuorenmaa et al., 2017 ). However, since S compounds are retained

in forest ecosystems and released with changing environmental con-

ditions, changes in S leaching do not necessarily occur at the same

time as S deposition changes. Therefore, S output in forest catch-

ments often exceeds the atmospheric input due to legacy S pools

derived from past deposition (Vuorenmaa et al., 2017) or due to

changing climate (Mitchell & Likens, 2011), which might delay recov-

ery from acidification. In Asia, much of the deposited atmospheric

SO2−

seems to be retained in forest soils (Duan et al., 2016; Sase

et al., 2019), which may imply a future risk of soil acidification under

changing climate. In fact,

SO2−

concentrations and pH of river wa-

ters are related to the S emission/deposition rate (Duan et al., 2011;

Qiao et al., 2016; Sase et al., 2017, 2021). To understand the S cycle

in forest ecosystems, targeted studies on deposition trend and

changing climate are required (e.g., Mitchell & Likens, 2011; Sase

et al., 2019, 2021; Vuorenmaa et al., 2017 ).

Air pollution abatement may also reduce atmospheric inputs of

base cations (Tørseth et al., 2012), as reported for forest soil solu-

tions (Johnson et al., 2018) and freshwaters (Garmo et al., 2014;

Stoddard et al., 1999). Base cation nutrient s in China forest s neutral-

ized on average 76% of the potential acid load due to acid deposition

during 2001– 2015 (Du et al., 2018). Thus, base cation deposition

should be monitored simultaneously along with S and N deposition

as already done by several networks globally to assess nutrient sta-

tus and recovery from acidification in forest ecosystems.

Excess N inputs from the atmosphere have been disturbing

biogeochemical cycles in forest ecosystems (e.g., Aber et al., 1989;

Nakahara et al., 2010). With reduction in total N deposition mainly

due to NOX emissions, an improvement is expected in the NH.

However, high levels of NH3 deposition are still concerning because

NH3 emissions have not clearly reduced in many of the regions as

described above. Moreover, since emissions of SO2 and NOx have

been reduced resulting in significant decline of particulate forma-

tion (such as (NH4)2SO4 and NH4NO3), air concentrations of NH3

have been increasing and accordingly more localized NH3 deposi-

tion was identified in the United States (Butler et al., 2016). Even

though regional N deposition has gradually decreased, ecosystem

responses to N deposition appeared to show some degree of hyster-

esis (Gilliam et al., 2 019). In fact, there was no large- scale response

in understory vegetation, tree growth, or vitality to reduction of

N deposition in Europe, while a decline in

NO−

concentrations in

soil solutions and foliar N concentrations were partly observed

(Schmitz et al., 2019). In Asia, three decades of increase in N depo-

sition in China have exerted significant impacts on soil and water

acidification, understory biodiversity, forest growth, and carbon se-

questration (Qiao et al., 2016; Tian et al., 2018). However, recovery

from acidification and N saturation has already started following a

reduction in N deposition in Japan (Sase et al., 2019), where high

S and N deposition and climatic anomalies caused acidification and

N saturation in the 1990s (Nakahara et al., 2 010). Nitrogen leach-

ing from forest ecosystems is controlled not only by N deposition,

but also by various factors, including tree age, forest management,

climate, and other limiting nutrients such as phosphorus. Moreover,

emissions of NH3 (e.g., Hansen et al., 2017), N2O (e.g., Schindler

et al., 2020; Xie et al., 2018), and NOy (as NO + NO2 + HONO; e.g.,

Mushinski et al., 2 019) as well as microbial nitrification rate (e.g.,

Fang et al., 2015) in forest areas should be taken into consideration

for actual N fluxes. Since N deposition may increase gas N emissions

from ecosystems (e.g., Xie et al., 2018), a comprehensive study con-

sidering bilateral N fluxes (both deposition and emission) should be

promoted to evaluate whether a forest ecosystem is a sink or source

of reactive N species.

The analysis of N dynamics in Latin America is complex, due to

the enormous diversity of unmanaged and managed ecosystems,

including arid deserts as well as temperate and tropical forests.

Cunha- Zeri and Ometto (2021) stated the major input of N in Latin

DE MARCO et al.

American countries over the past decades occurred via natural bio-

logical fixation, compared to anthropic sources (fertilizers and fos-

sil fuel combustion). Nevertheless, human activities have currently

changed the N cycle of natural ecosystems in Latin America. For in-

stance, the conversion of unmanaged land to agriculture increased

biological N fixation up to twofold (Reis et al., 2020). Although the

highest total N deposition occurs in eastern and southern China,

Japan, Eastern US, and European forests, the highest dry deposi-

tion occurs in tropical forests (Schwede et al., 2018). For instance,

dry N deposition into the Atlantic Forest in the city of São Paulo

(Brazil) can exceed the critical N load found for most forests (Souza

et al., 2020).

Because of the continued increase in NH3 emission in some re-

gions (e.g., Kurokawa & Ohara, 2020) and stagnating values in oth-

ers (Maas & Grennfelt, 2016), N deposition is a per vasive issue that

impacts forest ecosystems. In addition, even relatively low levels of

N deposition affect the mycorrhizal association of trees (Lilleskov

et al., 2019; van der Linde et al., 2018) and may affect biodiversity of

sensitive species, such as lichens (Giordani et al., 2014). The magni-

tude and consequences of these human- induced changes in plant–

soil– microbe interactions as well as potential pathways for recovery

are currently open questions.

Moreover, excess N deposition may induce an imbalance of nutri-

ent ratios, such as N:P ratio (Krüger et al., 2020; Sardans et al., 2016).

However, the observational data on atmospheric P deposition are

still limited for forest areas (e.g., Chiwa, 2020; Du et al., 2016) and

N- P imbalances have been reported from various regions (Boccuzzi

et al., 2021; Krüger et al., 2020; Peñuelas et al., 2013). Taking into

account the global pattern of N and P limitation in forest areas (Du

et al., 2020), N and P deposition should be monitored together. Both

N and P cycles are listed as important Earth- system processes in the

concept of “Planetary boundaries” with N cycle already transgress-

ing its boundary (Rockström et al., 2009; Steffen et al., 2015).

Climate has an important role in regulating the global patterns of

terrestrial N and P limitation (Du et al., 2020). Specifically, there is a

shift from relative P to N limitation at lower mean annual tempera-

ture, temperature seasonality, mean annual precipitation, and higher

precipitation. Future climate change will likely reshape the spatial

pattern of nutrient limitation. For instance, climate warming will im-

prove N availability at mid- to- high latitudes via increasing biologi-

cal N fixation and N mineralization (Zaehle et al., 2010). Moreover,

growth stimulation by rising atmospheric CO2 concentration ([CO2])

will increase nutrient demand and, in turn, result in greater nutrient

limitation (Collalti et al., 2018; Wieder et al., 2015). The changing nu-

trient status under climate change will likely interact with the effects

of S and N deposit ion and thu s they should be considere d simulta ne-

ously when projecting future forest dynamics.

4 | GROUND- LEVEL OZONE IMPACTS

Background O3 concentrations have increased throughout the last

century due to the rising anthropogenic emissions of O3 precursors

from fossil fuel and biomass burning (Cooper et al., 2014; Monks

et al., 2015), although volatile organic compounds (VOCs) also are

major precursors (Wei et al., 20 14). Despite the decreasing trend of

other air pollutants in the last decades (e.g., S and N compounds,

heavy metals), global- scale background O3 concentrations increased

(Jakovljević et al., 2021; Sicard, 2021), but slight regional- scale de-

creases in peak concentrations were observed (Schaub et al., 2018).

Thus, O3 is nowadays one of the main phytotoxic air pollutants with

the potential to affect forest ecosystems worldwide (Agathokleous

et al., 2020; Bytnerowicz et al., 2016; De Marco et al., 2020; Feng,

Shang, Gao, et al., 2 019; Sicard, Augustaitis, et al., 2016).

Ozone burdens are higher in the Northern (O3 mean concentra-

tion 35– 50 ppb) than in the SH (O3 mean concentration <20 ppb;

Sicard et al., 2 017). For example, widespread O3- induced visible in-

jury, a specif ic damage ass ociated wit h O3 exposure, was foun d at 17

forest plots in Europe (Paoletti et al., 2019; Sicard et al., 2020). The

NH is more covered by land and terrestrial ecosystems, and more

inhabited by humans than the SH, and thus is more affected by an-

thropogenic activities. However, the SH is less monitored and thus

O3 burdens and effects may be underestimated. While there are

hundreds of papers on O3 effects on forest plants and forests in the

NH (i.e., Agathokleous et al., 2015; Feng, Shang, Gao, et al., 2019;

Izuta, 2017; Sicard et al., 2020) indicating various effects of O3 in

interaction with climate change (Figure 3), relevant research in the

SH remains scarce.

The analysis of O3 effects in Latin America is complex due to

the enormous diversity of natural and agricultural ecosystems.

Most monitoring studies on O3 effects on forest plants conducted

in the SH come from Brazil. Urban and industrial development has

been more intense along the Atlantic Brazilian coast, especially in

Southeastern region. Consequently, more severe O3 effects on the

Atlantic forest located in this subtropical region (mainly São Paulo

and Rio de Janeiro States) are expected (Domingos et al., 2003;

Moura et al., 2014, 2018). Ozone effects on native tree species from

the Atlantic Forest have recently been determined in the field or

experimentally, pointing to distinct tolerance levels and highlighting

the need to expand knowledge on this topic (Cassimiro et al., 2016;

Engela et al., 2021; Fernandes et al., 2019; Moura et al., 2018). In

the SH, the Amazon spans over 629 million hectares of rainfor-

est, accounting for 54% of the total rainforests left on Earth (Peng

et al., 2020). Recent modeling approaches have shown O3 concentra-

tions have increased above the Amazon and Cerrado biomes in Brazil

as a response to biomass burning and regional air pollution (Gerken

et al., 2016; Pope et al., 2 019). The lowest O3 exposures reported

are in Australia, New Zealand, southern parts of South America, and

some northern parts of Europe, Canada, and the United States (Mills

et al., 2018; Sicard et al., 2017). However, unfortunately, a proper O3

monitoring network does not currently exist. Despite the presence

of ground- level O3 monitoring networks in all the developed coun-

tries (Lefohn et al., 2018), there is still a lack of an integral network

of ground- level O3 monitoring across Asia, although 1500 monitor-

ing stations have recently been installed in China (Feng, Shang, Gao,

et al., 2019).

DE MARCO et al.

Another challenge in monitoring O3 impacts on forests is the

choice of metrics. The AOT40 index (Accumulated Ozone over

Threshold of 40 ppb ozone), describing the exposure of plants to high

O3 concentrations, is the default measure for policy direc tives of the

European Union (Directive 2008/50/EC). However, AOT40 has been

criticized because it is not a proxy of gas uptake through leaf sto-

mata (stomatal flux), and flux- based indices have been applied (Anav

et al., 2022; De Marco & Sicard, 2 019; Paoletti et al., 2019; Sicard

et al., 2020) and showed O3 risks to vegetation would be different

from AOT40 (Anav et al., 2016; De Marco et al., 2015). The new

standard developed in Europe (Emberson et al., 2000) is the stoma-

tal O3 flux, defined as POD (Phytotoxic Ozone Dose). This standard

depends not only on O3 concentration, but also environmental (e.g.,

light intensity, air temperature, relative humidity, soil moisture) and

plant conditions (phenology, leaf morphological, and physiological

traits). A major impact of O3 is reduced aboveground and below-

ground carbon sequestration of forests (Agathokleous et al., 2016;

Gao et al., 2017; Figure 2). Ozone effects on biogenic volatile organic

compounds (BVOCs) are complex, as some compounds may decrease

(e.g., isoprene) while other compounds increase (e.g., monoter-

penes; Feng, Yuan, et al., 2019). Different BVOC compounds have

different capacity to generate O3, with isoprene having higher O3-

forming potential than monoterpenes (9.1 g 03 (g VOC)−1 and 3.8 g 03

(g VOC)−1, respectively; Benjamin & Winer, 1998). However, sesqui-

terpenes and some monoterpenes also contribute to the removal of

O3 at the canopy level and play an important role in the feedback

between stress- induced VOC emissions and O3 or aerosol forma-

tion (Calfapietra et al., 2013). The emission of isoprene, the most

abundant BVOC, can also be decreased by drought and CO2 and in-

creased by warming (Feng, Shang, Li, et al., 2019), indicating complex

O3- climate interactions that remain elusive in real- world forests. Soil

microbial processes contribute to emission of BVOCs and NOx that

act as O3 precursors (Gray et al., 2010). Overall, soils play an im-

portant role in forest VOC exchange, defining also carbon storage

by forest ecosystems, and fluxes depend upon BVOC compounds

and vegetation types (Mäki et al., 2019; Rinnan & Albers, 2020; for

details and values of fluxes in different vegetation types and envi-

ronmental media, see also Tani and Mochizuki (2021)). However, the

FIGURE 3 The Gordian Knot of the Forest– Ozone– Carbon interactions. In the pre- industrial epoch, carbon is stored via photosynthesis

(1) and leads to long- term carbon sequestration into aboveground and belowground (roots and soil) wood biomass (2) (Agathokleous

et al., 2016; Grantz et al., 2006). The higher CO2 levels, alone, in the atmosphere are expected to “feed” forest growth (Koike et al., 2018)

and have beneficial effects. The increased O3 levels, alone, depress forest trees, contributing to “forest decline syndrome,” that is, visible

injury, photosynthesis, carbon sequestration, carbon storage changes (7), and biomass decay, which also releases CO2 in the atmosphere (8)

(Agathokleous et al., 2016; Sandermann et al., 19 97; Sicard et al., 2021; Takahashi et al., 2020). In a positive feedback, the depressed forest

vegetation emits more BVOCs (4), further increasing O3 levels (Peñuelas & Staudt, 2010). Concurrent elevated concentrations of CO2 end

O3 may outcome to a sustained increase in Net Primary Productivity (NPP), while the adverse long- term effect of increased O3 on NPP may

be lesser than projected (Talhelm et al., 2 014). Elevated CO2 levels negate or even overcompensate the negative O3 effect on ecosystem

functions and the cycles of carbon and nitrogen. Anthropogenic emissions of CO2, NOx, and volatile organic compounds (VOCs) (3) as well

as biogenic VOCs (BVOCs) emitted by forests (4) contribute to increased O3 levels in the atmosphere (Yu & Blande, 2021). Soil microbial

processes contribute to soil- emitted BVOCs and NOx (O3 precursors; Gray et al., 2010 ) as well as CO2, N2O and CH4 (Yao et al., 2009; Zhang

et al., 2021) (5). Under advanced climate change, forest fires are expected to be more frequent and larger than in the pre- industrial epoch

(Zhang et al., 2021). These fires release carbon monoxide (CO), organic carbon (OC), NOx (all of which contribute to O3 formation), and

black carbon (BC; which influences photosynthesis by increasing diffuse radiation) as well as CO2 (which further intensifies global warming;

Flannigan et al., 2009; Kumar et al., 2019 ; Pellegrini et al., 2021; Yue & Unger, 2018) (9).

DE MARCO et al.

specific contribution of soil in VOC exchanges and O3 formation re-

mains poorly understood.

5 | TRACE ELEMENTS AND R ADIOACTIVE

CONTAMINATION OF FOREST ECOSYSTEMS

Heavy metal pollution was an important subject in widespread forest

decline in the 1980s– 1990s (Gawel et al., 1996), but more recently has

become a major item in phytoremediation (Pulford & Watson, 2003)

and environmental monitoring (Godzik, 2020). The term “heavy met-

als” is now discouraged, and these elements are now included more

broadly as “trace elements” (Pourret & Bollinger, 2018). Trace ele-

ments are a major component of particulate pollution (Antoniadis

et al., 2017; Grantz et al., 2003; Li et al., 2015; Schlutow et al., 2021;

Tóth et al., 2016). At the global scale, trees are important for their

role in retaining particulates (Yue et al., 2021). Nevertheless, in

some regions, soil contamination by trace elements remains so high

that it continues to kill trees and prevents natural recovery (Kozlov

et al., 2009). Among trace elements, radionuclides display the most

phytotoxic potential.

The use of nuclear energy or nuclear applications in health, ag-

riculture, environmental management, or industry/military resulted

in releases of radionuclides into the environment (Hong et al., 2012).

The first large- scale radioactive contamination from anthropogenic

sources occurred through global radioactive fallout from nuclear

weapons' tests conducted in the atmosphere during 1945– 1980

(Aoyama et al., 2006; United Nations, 2000). A variety of long- and

short- lived radionuclides were released during nuclear incidents; in

particular 137Cs with a relatively long half- life (~30 years) compared

to other radionuclides, such as 134Cs and 131Cs. Other major releases

of radionuclides occurred from the Chernobyl nuclear power plant

accident in 1986 (International Atomic Energy Agency, 2006) and

from the Fukushima Daiichi Nuclear Power plant accident in 2011

(Chino et al., 2011; Terada et al., 2020; Yoshida & Takahashi, 2012).

Radioactive contamination of forests has different types of

impacts (Figure 4). First, direct radiation can affect trees and an-

imals and occur at the level of DNA, cells, individuals, population

to whole ecosystems, and ranges from reparable DNA damage to

death of organisms (Committee on the Biological Effects of Ionizing

Radiation, 1990 ). An example of direct impacts of high radiation

doses to trees is the “Red forest” in the Chernobyl exclusion zone,

where pine trees became reddish brown and died following the ac-

cident (Beresford et al., 2016). Another visible impact of radiation

exposure in trees is the occurrence of morphological abnormalities

(Watanabe et al., 2015; Yoschenko et al., 2011, 2016). Compared

to the effects caused by high doses of radiation, those poten-

tially caused by relatively lower radiation dose are confounded by

many other factors and are still not clearly understood (Beresford

et al., 2020; Ji et al., 2019; Strand et al., 2017). In exposed areas, for-

est ecosystems are released from pressure by human existence, re-

sulting in creation of ecological niches and expansion of populations

of some species (Deryabina et al., 2015; Lyons et al., 2020; Perino

et al., 2019). Through intensive monitoring, it was confirmed that the

overall dynamics of 137Cs within forest ecosystems were similar be-

tween Chernobyl and Fukushima: tree canopies captured the depo-

sition of 137Cs and 137Cs migrated from the canopy to the soil surface

via water and litter fall, and most of it stays in the top layers of soil

(Itoh et al., 2015; Kato et al., 2019; Suchara et al., 2 016). However,

the migration velocity and distribution patterns of 137Cs within for-

ests and tree bodies differ substantially among forests and trees

(Imamura et al., 2017; Ohashi et al., 2017). It is essential to continue

experimental studies to identify the key processes influencing 137Cs

dynamics in forest systems, such as soil potassium concentrations

and fixation processes within soils (Kobayashi et al., 2019 ; Manaka

et al., 2019). Various models have been developed to characterize

137Cs dynamics in forests; however, improvements are necessary

to reproduce variations between forest types and species compo-

sitions (Hashimoto et al., 2020). Another aspect of radionuclide pol-

lution is that deposited radionuclides, which are easy to detect and

measure, provide an unintentional but useful opportunity to track

biogeochemical cycles in forest ecosystems (Fukuyama et al., 2008).

6 | COMBINED EFFECTS OF MULTIPLE

FACTORS ON FOREST ECOSYSTEMS

Our knowledge on combined effects of multiple factors on ecosys-

tem health originated primarily from temperate and boreal forests

of North America and Europe and is limited for tropical forests,

FIGURE 4 Diagram of direct and indirect effects of forest

radioactive contamination. The deposited radionuclides remain in

the forest and continue to circulate in the forest ecosystem, and

radiation can have adverse effects on forest biota (direct effects).

Restrictions on forest use and land abandonment to avoid exposure

can also affect forest ecosystems, including changes in vegetation

and wildlife populations. It has direct and indirect impacts on

ecosystems and local residents.

DE MARCO et al.

especially of those in Africa (Matyssek et al., 2 017). In other words,

areas that have recently experienced the highest risk of forest deg-

radation are studied to a lesser extent than the areas where risk is

low. In addition, many communities whose food security and wealth

generation critically depend on forests are located in geographic re-

gions where our understanding of factors affecting forest ecosys-

tem health is poor. This geographic bias is typical for ecological and

environmental research (Archer et al., 2014), and its consequences

are generally seen as severe, because results obtained with one

study system may appear of little use in predicting the responses of

another, geographically distinct, study system (Haukioja et al., 1994).

Air pollution levels may become more harmful for plants as the

climate warms (Zvereva et al., 2008, 2010). More multi- factorial

manipulative studies are needed because effects of two or more

co- occurring factors on tree growth and forest productivity can-

not be adequately predicted from single- factor experiments

(Niinemets, 2010). The combined effects of two major abiotic aspects

of global change, mostly changes in CO2 and warming, on growth of

forests are studied in detail (Baig et al., 2015; Curtis & Wang, 1998;

Zvereva & Kozlov, 2006), and suggest air temperature may modify

plan t responses to elevate d CO2. Across 42 experiments with woody

plants, aboveground biomass increased significantly with both CO2

(the so called “fertilization effect”) and air temperature (by 21.4%

and 18.1%, respectively), whereas these two factors acting simulta-

neously showed a much smaller effect (8.2%) because of compen-

sating effects (Baig et al., 2015). Nitrogen fertilization enhances the

biomass response to elevated CO2 (Parrent & Vilgalys, 2007) despite

not universally (Terrer et al., 2019). The type of mycorrhiza was also

an important factor related to the effects of soil nutrient availability

on elevated CO2- induced growth enhancement (Baig et al., 2015).

However, two- factorial experiments involving both O3 exposure

and elevated CO2 are limited. Several studies under elevated CO2

showed a reduction in the negative effects of O3 because elevated

CO2 induced stomatal closure leading to lower O

3 uptake (Grams

et al., 1999; Watanabe et al., 2017 ). In contrast, the addition of N

alone exacerbated negative effec ts of O3 on photosynthesis of trees

(Feng, Shang, Li, et al., 20 19), while exposure to drought stress did

(Gao et al., 2 017) or did not protect plants from O3- induced effects

(Alonso et al., 2003, 2014).

Forest health also can be compromised by insect herbivory, in-

cluding both devastating outbreaks of forest pests and changes in

background herbivory. Despite relatively low levels of plant damage

(5%– 7% of leaf biomass annually: Kozlov et al., 2015), background

herbivory greatly reduces growth of woody plants (Shestakov

et al., 2020; Zvereva et al., 2012). Although warming, drought, CO2

increases, N deposition, and air pollution were repeatedly found to

increase herbivory (Lincoln et al., 199 3; Logan et al., 20 03), these

conclusions were likely affected by research and publication biases

(Zvereva & Kozlov, 2010) and/or were derived from results of short-

term laboratory experiments, which tend to overestimate the ef-

fects relative to natural ecosystems (Bebber, 2021). Within forest

ecosystems across the globe, no increase in insect herbivory was

observed from 1952 to 2013 (Kozlov & Zvereva, 2015). Similarly,

long- term monitoring did not reveal the effects of either pollution-

induced disturbance or 2.5°C climate warming on insect herbivory

in subarctic birch forests (Kozlov et al., 2017). Thus, the evidence

regarding combined effects of climate warming and air pollution on

insect herbivory remains somehow contradictory.

Other factors whose effects on forest trees have been stud-

ied in multi- factorial studies include (but are not limited to) cattle/

deer grazing, harvest of non- timber forest products, drought, flood-

ing, soil salinization, spring frost, heat waves, and increased ultra-

violet radiation (e.g., Mac Nally et al., 2011; Pliūra et al., 2019; Sugai

et al., 2019; Varghese et al., 2015). However, a low number of such

studies precludes any generalization regarding effects of these fac-

tors, combined with CO2 and air temperature increases or O3 and

insect herbivory on health of forest ecosystems. Modeling studies

jointly assessing the effects of climate change and air pollution can

greatly help for understanding and predicting future developments

of forests (Akselsson et al., 2016; Dirnböck et al., 2017; Etzold

et al., 2020; Fleck et al., 2 017; Rizzetto et al., 2016).

7 | GENETIC INFORMATION RELATED

TO PHENOTYPES AND PHYSIOLOGY OF

FOREST TREES

Air pollution, climate change, increased pests and pathogens, land-

use changes, and forest fragmentation can all reduce genetic diver-

sity and make forests more fragile and sensitive to other threats

(Gauthier et al., 2015). Current vegetation and forest growth models

are largely parameterized on direct growth and gas exchange meas-

urements or remote sensing, while information from biological and

genetic regulation mechanisms are still scarce. For example, part of

the carbon fixation products (i.e., photosynthates) that is not used

for biomass production is released in soil as root exudates, some is

stored, and some organic carbon is emitted as BVOCs affecting plant

and community ecology and atmospheric chemistry (Blande, 2021;

Collalti et al., 2020; Maja et al., 2015; Naidoo et al., 2019; Šimpraga

et al., 2019). Carbon sink strength of trees is known to be impaired

by limitations in water and nutrient availability, heath spells, air

pollutants, and increased herbivory. However, plant defense pro-

cesses against different abiotic and biotic factors are complex and

involve multiple signaling pathways (He et al., 2018), potentially af-

fecting how carbon is allocated to different organs (Merganičová

et al., 2019). Most of the underlying resistance mechanisms are de-

scribed or predicted from short- living herbaceous model systems,

whereas investigations on mechanisms of defense and adaptation

of forest trees are much more challenging due to long lifetime, high

genetic diversity, and variation of growth environments and climates

(Naidoo et al., 2019). There is an urgent need to intensify studies on

the mechanisms underlying the resilience of forest ecosystems to

current and long- term effects of air pollution and climate change,

utilizing genetic, species, and ecosystem- level functional diversity

as well as adaptive management, resistance breeding, and genetic

engineering (Naidoo et al., 2 019). Mechanistic understanding is

DE MARCO et al.

increasingly important also for efforts in afforestation and protec-

tion of primar y forests. In principle, there are two main approaches

for achieving resistance in forest trees: (i) selection of resistant

phenotypes identified in field experiments (Sniezko & Koch, 2017)

or polluted sites (Eränen et al., 2009; Kozlov, 2005); and (ii) struc-

tured breeding programs relying on multitude of omic techniques

(Naidoo et al., 2019). The databases for genetic information of tree

species have been rapidly increasing, and the most important model

systems for forest trees are Populus, Eucalyptus, Quercus, Castanea,

Pseudotsuga, Pinus, Picea, and Betula genuses (Falk et al., 2018;

Salojärvi et al., 2017). Genetic engineering efforts by forest bio-

technology companies have produced transgenic Eucalyptus and

Populus trees with enhanced growth and disease- resistant proper-

ties (Naidoo et al., 2019). Silver birch (Betula pendula Roth) is an ex-

cellent model system for elucidating the adaptation and acclimation

capacity of forest trees to rapidly changing climate due to its (i) wide

latitudinal and longitudinal distribution; (ii) recent advances in popu-

lation genomics and evolutionary history of birch species (Salojärvi

et al., 2017); and (iii) existence of well- characterized birch genotypes

that have been intensively studied for C and N economy, photosyn-

thetic efficiency, metabolism, chemistry, and phenology (Deepak

et al., 2018; Tenkanen et al., 2020). The population genomic analyses

of silver birch provide insights on natural selection mechanisms, with

candidate genes relevant for adaptation of trees to changing envi-

ronment, biotic stress, and growth regulation (Salojärvi et al., 2017 ).

Studies with birch have also shown the C- sink strength of trees can-

not be explained by physiological or genetic approaches alone, but

there are many negative and positive interactions with pollutants,

climate, pests, pathogens, microbiomes, and between plants that

should be understood in more detail (Naidoo et al., 2019; Silfver

et al., 2020; Wenig et al., 20 19).

Plant phenotypes are strongly affected by the environment, and

often genotype per environment interaction is the factor of greatest

interest. Methodologies have been developed for non- destructive

forest- level and individual tree- level phenotyping with remote sens-

ing techniques, which are particularly useful for identifying superior

genotypes under different stress conditions (Dungey et al., 2018;

Kefauver et al., 2012; Ludovisi et al., 2017 ). Recent advances in

metagenomics and the increasing knowledge of the importance

of microbiomes in plant health offer new opportunities for forest

health management (Imperato et al., 2019; Naidoo et al., 20 19;

Wenig et al., 2019). The regulatory networks of forest trees and the

beneficial non- pathogenic microbes living around and on the sur-

faces of plant roots (rhizosphere), leaves (phyllosphere), or in the

internal plant tissues (endosphere) can be particularly important for

carbon and nutrient dynamics of trees and the development of tree

immunity (Naidoo et al., 2019). Microbes are known to help plants

in water and nutrition acquisition, defense against pathogenic mi-

crobes, tolerance to abiotic stress, adaptation, promotion of the es-

tablishment of mycorrhizal association, and plant growth regulation,

forming a holobiont system with host trees (Imperato et al., 2019;

Naidoo et al., 2019; Wenig et al., 2019). Fungal and bacterial com-

munities in forest soils have been shown to respond to changes in

climate with a shift in their community composition as well as in their

diversity (Dubey et al., 2019; Jansson & Hofmockel, 2020; Milović

et al., 2021; Simard, 2010). For example, under elevated CO2, we

can observe alteration in relative abundances of bacteria and in-

creased bacterial to fungal ratio (Dubey et al., 2019), as well as an

increase in ectomycorrhizal colonization rate but a decrease in ecto-

mycorrhizal diversity (Wang et al., 2015). Warming and elevated O3

reduced ecto- and arbuscular mycorrhizal colonization and shifted

arbuscular mycorrhizal community composition in favor of the genus

Paraglomus, which has high nutrient- absorbing hyphal surface (Qiu

et al., 2021; Wang et al., 2015). At the same time, exposure to higher

levels of O3 is associated with lower soil microbial biomass and with

changes in the overall structure and composition of poplar rhizo-

sphere soil microbial communities (Li et al., 2021). The decreased

growth of roots and decrease in ectomycorrhizal colonization rate

and a shift in species abundance might be an early indicator of the

damaging impacts of O3 in some tree species, occurring prior to visi-

ble responses of aboveground tree parts (Katanić et al., 2014).

8 | MODELING FOREST ECOSYSTEMS FOR

RISK ASSESSMENT

Scientific methods in forestry, including empirical models of tree

growth, were primarily used for optimization of timber harvest

throughout the 20th century (Porté & Bartelink, 2002). A more in-

tegrative modeling approach, acknowledging natural disturbances

(e.g., wind, fires, pests, diseases) as inherent elements of forest

ecosystem dynamics, was developed when computational advance-

ments allowed for the integration of greater complexity (Blanco

et al., 2020; Perera et al., 2015), although some biotic factors of for-

est disturbance such as herbivory are still rarely modeled (De Jager

et al., 2017). Since the forest dieback and acidification debate in

Europe in the 1980s, large efforts were put into improving under-

standing and prediction of anthropogenic disturbances on biogeo-

chemical dynamics of forests. Starting from models mainly targeting

the fate and effects of acid rain in forest ecosystems (Nilsson, 1988;

Sverdrup & De Vries, 1994), simulation tools have broadened to in-

clude other pressures, such as N deposition (De Vries et al., 2010),

O3 (Hoshika et al., 2015), and climate change and forest management

(Collalti et al., 2018). Modeling has been demonstrated to be a valu-

able tool for studying forest responses to present and future distur-

bances, allowing ecologists and foresters to deal with the study of

complex interactions and to evaluate future management strategies

(e.g., Collalti et al., 2018; Fleck et al., 2017) or policy options (e.g.,

Belyazid et al., 2010; Dirnböck et al., 2018).

Existing relationships between forest structure and compo-

sition and environmental variables were initially used to build

empirical models that describe past ecosystem behavior and ex-

trapolate to future conditions (Gustafson, 2013). Subsequent

modeling efforts simulated the causal biogeochemical mecha-

nisms that underlie the responses of ecosystems to these envi-

ronments (Kimmins et al., 2008). These so- called process- based

DE MARCO et al.

models (PBMs) study the ecological processes and are considered

one of the most reliable approaches for modeling forest eco-

system dynamics under global change (Evans, 2012; Maréchaux

et al., 2020). However, forecasting forest growth is still a priority

in many studies, either for planning forestry activities under air

quality and climate change scenarios or as part of carbon storage

calculations (Blanco et al., 2020).

The general trend toward biodiversity conservation in interna-

tional policies (e.g., EU Biodiversity Strategy for 2030; European

Commission, 2020), its importance in preserving ecosystem ser-

vices, and the use of biodiversity metrics as indicators in risk

assessment (Coordination Centre for Effects, 2017) and pol-

icy evaluation (Hein et al., 2018) make the simulation of species

composition changes a decisive function for any model. When

dynamic PBMs are used for forecasting biodiversity shifts, they

are usually combined with vegetation response models based on

species niche suitability and competition (Belyazid et al., 2019;

Dirnböck et al., 2018). PBMs have been used at stand (e.g., Collalti

et al., 2016), landscape (Shifley et al., 2017), regional (Belyazid

et al., 2 019; De Marco et al., 2020; Santini et al., 2014), and global

scales (e.g., Krause et al., 2017 ). However, their implementation is

restricted at larger scales since PBMs need large, detailed input

datasets, which are often not available at national or continen-

tal scales. At these scales, a currently suitable approach is using

new models, mostly empirical, based on currently available large

datasets, such as species distribution models (SDMs; Maréchaux

et al., 2020; Noce et al., 2017). In the same way that PBMs rose

with the increasing computational power during the last decades

of the 20th century, SDMs have improved during the present de-

cade, in parallel to the increase in web available, reliable spatial-

referenced data, including environmental and meteorological

data, forest inventories, habitat distribution, aerial images, and

remote sensing (Pecchi et al., 2019; Urban, 2015). SDMs are usu-

ally statistical models that are currently used to support sustain-

able planning of forests at national and international scales (Zang

et al., 2012), with correlative SDMs using maximum entropy al-

gorithms being most frequently used (Noce et al., 2017; Pecchi

et al., 2019). Using forest decision support systems, climate

change scenarios and the balance of delivered ecosystem services

can be suggested as a methodological framework for validating

forest management alternatives aiming for more adaptiveness in

sustainable forestry (Marano et al., 2019; Mozgeris et al., 2019).

Mo reov er, some of the ve get ati on model s assoc iated wit h PBMs to

assess or forecast biodiversity are SDMs that may be applied from

site to regional scales (e.g., Wamelink et al., 2020). There are some

recent examples of SDMs implemented to assess forest biodiver-

sity respons e to at mospheri c pollution and climate change, such as

Hellegers et al. (2020) and Wamelink et al. (2020). However, these

models still lack essential information to feed their predictions,

since new field observations and experiments with novel set- ups

(e.g., Hansen & Turner, 2019) are needed to address the potential

successional and disturbance dynamics under the forthcoming cli-

mate conditions (McDermott, 2020). Therefore, there are several

possible approaches for different problems that scientists and

managers must deal with (Blanco et al., 2020; Fabrika et al., 2 019;

Maréchaux et al., 2020). The modeling process might be as com-

plex as needed by risk assessment objectives (Figure 5), providing

models and data are available and suitable. In general terms, empir-

ical models are good at predicting biomass and forest structure in

the shorter term, and consequently producing good management

recommendations for the present conditions, but are not reliable

in novel situations (i.e., future air pollution and climate change).

PBMs are good at studying effects and underlying processes of

change, particularly in the context of global change. However,

they still have low feasibility at broad scales, and the calibration

and validation processes are highly time- consuming (particularly

for the less- modeled species or regions). SDMs are appropriate for

early risk assessment on biodiversity conservation at the broad-

est scales, but still too empirical, which diminishes their reliability

in the long term (Urban et al., 2016). Mixing process- based with

empirical approaches (hybrid models), integrating, and connecting

different models (meta- and mega- models; Blanco, 2013) are ex-

cellent strategies to answer specific questions.

9 | NEW DIRECTIONS: INTEGRATING

EXPERTS' OPINIONS

9.1 | Air pollution monitoring network

While conventional field- based monitoring plots will continue

to dominate the mainstay of air pollution and climate change re-

search in forests, they are costly and often logistically difficult to

conduct over large areas. Therefore, remote sensing techniques

will be more and more appropriate for large- scale monitoring pro-

grams, even though a more in- depth approach still needs to be

developed. Finer temporal intervals are required for in- depth un-

derstanding of some responses (e.g., stomatal O3 fluxes require

a continuous- monitoring approach). Highly instrumented field

sites are now cheaper and technically affordable. Integrating data

from transcontinental long- term ecological research infrastruc-

tures in tree- based models would lead to a better understanding

of how ecosystems work (Fischer et al., 2011). Long- term data

series can be integrated in existing big databases such as the

Global Atmosphere Watch (GAW) Program and the international

Tropospheric O3 Assessment Report (TOAR; Schultz et al., 2017;

WMO, GAW, 2003). These raw databases can lead to the devel-

opment of new products for temporal and spatial analysis (data

analysis, maps of data distributions, and data summaries) that are

freely accessible to the scientific community and other stakehold-

ers. Such databases can be used as tools for mechanistic and diag-

nostic understanding and upscaling.

The need for a global forest monitoring is irrefutable, and “su-

persites” promote the integration of research communities in a

transcontinental collaboration network by upgrading existing

ground- based observation networks (e.g., FLUXNET, ICP, NEON)

DE MARCO et al.

covering all biogeographic areas (e.g., tropics, subtropics) and eco-

system types (e.g., woody savannas).

9.2 | Elements deposition in forests

The effects of S and N deposition on forest health have been

reducing gradually in many regions but problems have not been

solved. Legacy S pools remain, which could be affected by chang-

ing climate. Reduction of S deposition is associated with reduc-

tion of base cation deposition, which may alter nutrient status

and increase the risk of further soil acidification. The total inor-

ganic N deposition has been declining due to the implementation

of air pollution control policies, but the relative importance of

NH3 emissions and deposition is now higher (Butler et al., 2016;

Du, 2016), showing a relative increase of 0.38% per year over the

period 1985– 1999 (Du, 2016). Since ecosystem responses to de-

clining N deposition may show hysteresis (Gilliam et al., 20 19) and

key mechanisms of the N- induced changes in forest ecosystems

are not fully understood (Lilleskov et al., 20 19), long- term monitor-

ing of N- , S- , and P- cycles and base cations deposition should be

studied together to better understand biogeochemical processes

and plant biodiversity under climate change. Moreover, interac-

tions between nutrient deposition and rising O3 concentrations

should be considered in future studies (Shi et al., 2017). Long- term

monitoring should be continued even after significant air pollution

reductions to capture and understand the potential long- term ef-

fects of pollution and ecosystem recovery.

9.3 | Ground- level ozone

Surface O3 concentrations are generally higher in rural areas than in

urban areas (Sicard, 2021). However, as O3 levels are rising in cities

(Sicard, 2021), special attention should be paid to urban and peri-

urban forests, which offer services to local communities (Bruckman

et al., 2016) and can help meet air quality standards in cities (Sicard

et al., 2018). Because forest tree species play important (species-

dependent) dual roles as sinks and sources of O3 precursors (Geng

et al., 20 11; Saitanis, Agathokleous, et al., 2020), the O3 forming

potential (OFP) of the best regionally adapted forest tree species

should be investigated and taken into account by decision- makers to

select species with lower OFP for urban planning (Sicard et al., 2018).

The observed high O3 burdens, their high spatial heteroge-

neity, and the differential susceptibility of forest tree species to

O3, as well as their dual role as O3 sinks and precursor sources

(Agathokleous et al., 2020; Li et al., 2018), suggest an urgent

need for the establishment of a globally denser O3 monitoring

network in natural forest ecosystems in particular in the SH.

A new approach to the global O3 monitoring network and alter-

native methods for monitoring O3 are feasible thanks to innova-

tive technologies (Saitanis, Sicard, et al., 2020), which will help

FIGURE 5 Simplified flux of information diagram in modeling approach for risk assessment of air pollution and climate change. The

modeling and risk assessment process might be as complex as the modelers need and the availability and suitability of models and data allow.

There are three main blocks (grey bands): atmospheric and ecological models (top two) are the tools to reach the objective of risk assessment

(bottom). Models are used both in the internal processing and description of the information contained in the boxes, and in the transmission

of information between them (model inputs and outputs). For example, habitat suitability can be modeled using vegetation response models

(such as VEG; Belyazid et al., 2019 or PROPS; Dirnböck et al., 2018), that are particularly designed to process output from process- based

models as input information, but it can be also modelled by species distribution models that are particularly designed for large- scaled input

datasets (e.g., Noce et al., 2017). Information generally needed in any environmental study (such as soil and terrain variables) has been

obviated in this diagram.

DE MARCO et al.

to understand combined effects of O3 with other emerging envi-

ronmental factors. There is also an urgent need to generate flux-

base d standards and criti cal levels for fo res t prote ction taking into

account the sensitivity of dominant forest tree species. Because of

its limitations, the AOT40 index should not be adopted as default

for risk assessment (Agathokleous et al., 2019; Anav et al., 2022;

Sicard, Augustaitis, et al., 2016). Finally, the development of

countermeasures for controlling anthropogenic O3 precursor

emissions is also urgently needed.

Further research is still needed to develop O3- effect indicators

related to other ecosystem ser vices provided by forests such as bio-

diversity, soil protection, and water conservation. Nonlinear models

should be used for establishing cause– effect relationships under

experimental conditions (e.g., Agathokleous et al., 2019 ; De Marco

et al., 2013).

9.4 | Multiple stressors on forest ecosystems

For a better knowledge on combined effects of multiple factors on

ecosystem health, the selection of tree species for future studies

should account for their phylogenetic relatedness with already stud-

ied species. Ecological and environmental studies addressing the re-

sponses of tropical forests to combined effects of climate change and

air pollution should be intensified, in particular in areas at higher risk

of deforestation in the SH. This research domain is strongly biased

toward temperate and boreal forests of the NH. The evolutionary

changes in response to rising global CO2 levels and air temperature

elevation are known to occur in some plants, but the contribution of

evolutionary processes to the forest responses to steady CO2 and

air temperature rises remains unexplored. Experimental studies, ad-

dressing combined effects of different abiotic factors on forests,

should be intensified in the SH and should carefully select tree spe-

cies to assure a better representation of taxonomic and functional

diversity of the approximate 73,000 tree species now found on the

Earth (Cazzolla Gatti et al., 2022).

9.5 | Radioactive contamination of

forest ecosystems

Despite many papers reporting radioactivity effects on forest eco-

systems (Strand et al., 20 17; Tamaoki, 2016), there is still no con-

sensus on the mechanism through which radiation impacts forest

ecosystems or the dose rates at which impacts begin to occur

(Beresford et al., 2020; Strand et al., 2017). More robust and syn-

thesis studies are essential to inform (i) key processes regulating

the dynamics of radionuclides within forests; (ii) models for track-

ing radionuclides and prediction; (iii) holistic assessment of impacts

caused by radioactive contamination and its countermeasure devel-

opment; and (iv) use of 137Cs as a tracer. Furthermore, cost efficient

forest countermeasures must be developed and decisions must in-

clude locals, scientists, stakeholders, and governments.

9.6 | Genetic information of forest trees

More effort should focus on phenomics, combining high- throughput

capture of tree phenotypes, genotype information, data science,

and engineering (Falk et al., 2018; Naidoo et al., 2019). Future work

should include metadata integration and improved visualization for

comparative genomics. Characterizing the root traits and pheno-

types with association to genomics and shoot phenotyping is nec-

essary for whole- plant resistance breeding (Chuberre et al., 2018;

Tracy et al., 2020; Wiley et al., 2020). Rhizosphere phenotyping

opens new opportunities for experimental approaches, includ-

ing stress treatments, repeatability and combined use of imaging

techniques and machine learning to extract new traits from images,

within a systems approach (Tracy et al., 2020). The belowground net

primary production accounts for 40%– 70% of total terrestrial pro-

ductivity (Gherardi et al., 2020); therefore, more studies are needed

to explore response s of tree roots to climate and pol lution and quan-

tify root losses to belowground herbivores.

9.7 | Modeling and risk assessment

Model diversity constitutes a multi- purpose toolkit that can help

society to face the future challenges. Improving and enhancing

scientific communication in forest modeling is required as part of

this enterprise. The development of models integrating air pollution

and climate change data from long- term monitoring programs are

needed to improve forest research assessing interactions between

air pollution and climate change from the individual level to the stand

level. Future challenges include understanding of (i) the impacts of

air pollution on soil chemistry, (ii) the effects of climate change and

air pollution on plant phenol og y and reproductive fitness, (iii) the ca-

pacity of forests to sequester carbon under changing, and extremes,

climatic conditions and co- exposure to elevated levels of pollution,

and (iv) the effects of plant competitiveness (monocultures vs. mixed

cultures, single trees vs. community responses) on plant responses

to stressors.

ACKNOWLEDGMENTS

This work was outlined in the framework of the Research Group

8.04.00 “Air Pollution and Climate Change” under the International

Union of Forest Research Organizations (IUFRO). IUFRO is the

largest international network of forest scientists, promoting global

cooperation in forest- related research and enhancing the under-

standing of the ecological, economic, and social aspects of forests

and trees. M.V.K. was supported by the Academy of Finland (pro-

jects 276671, 311929, and 316182). M.W. was supported by JST

SICORP (JPMJSC16HB). A.D.M., P.S., Y.H., and E.P. were supported

by the LIFE projects MODERN (LIFE20 GIE/IT/000091), MOTTLES

(LIFE15 ENV/IT/000183), and AIRFRESH (LIFE19 ENV/FR/000086).

Contributions of JCB were partially supported through funding from

the US Department of Energy under award number DE- EM0005228

to the University of Georgia Research Foundation. Open Access

DE MARCO et al.

Funding provided by ENEA Agenzia Nazionale per Le Nuove

Tecnologie l'Energia e lo Sviluppo Economico Sostenibile within the

CRUI- CARE Agreement.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVA ILAB ILITY STATE MEN T

Dat a sharing is not applicable to this article as no new dat a were cre-

ated or analyzed in this study.

ORCID

Alessandra De Marco https://orcid.org/0000-0001-7200-2257

Alessio Collalti https://orcid.org/0000-0002-4980-8487

Enzai Du https://orcid.org/0000-0002-5519-0150

Mikhail V. Kozlov https://orcid.org/0000-0002-9500-4244

Elena Paoletti https://orcid.org/0000-0001-5324-7769

REFERENCES

Aber, J. D., Nadelhoffer, K. J., Steudler, P., & Melillo, J. M. (1989). Nitrogen

saturation in Northern forest ecosystems. BioScience, 39(6),

378– 386.

Agathokleous, E., Belz, R. G., Calatayud, V., De Marco, A., Hoshika, Y.,

Kitao, M., Saitanis, C. J., Sicard, P., Paoletti, E., & Calabrese, E. J.

(2019). Predicting the effect of ozone on vegetation via the linear

non- threshold (LNT), threshold and hormetic dose- response mod-

els. Science of the Total Environment, 6 49, 61– 74.

Agathokleous, E., Feng, Z., Oksanen, E., Sicard, P., Wang, Q., Saitanis,

C. J., Araminiene, V., Blande, J. D., Hayes, F., Calatayud, V.,

Domingos, M., Veresoglou, S. D., Peñuelas, J., Wardle, D. A ., De

Marco, A., Li, Z., Harmens, H., Yuan, X., Vitale, M., & Paoletti, E.

(2020). Ozone affects plant, insect, and soil microbial communi-

ties: A threat to terrestrial ecosystems and biodiversity. Scie nce

Advances, 6, eabc1176.

Agathokleous, E., Saitanis, C. J., Satoh, F., & Koike, T. (2015). Wild

plant species as subjects in O3 research. Eurasian Journal of Fores t

Research, 18, 1– 36.

Agathokleous, E., Saitanis, C. J., Wang, X., Watanabe, M., & Koike, T.

(2016). A review study on past 40 years of research on effects of

tropospheric O3 on belowground structure, functioning, and pro-

cesses of trees: A linkage with potential ecological implications.

Water, Air, & Soil Pollution, 227, 33.

Akselsson, C., Olsson, J., Belyazid, S., & Capell, R. (2016). Can increased

weathering rates due to future warming compensate for base

cation losses following whole- tree harvesting in spruce forests?

Biogeochemistry, 128, 89– 105.

Alonso, R., Elvira, S., González- Fernández, I., Calvete, H., García- Gómez,

H., & Bermejo, V. (2014). Drought stress does not protect Quercus

ilex L. from ozone effects: Results from a comparative study of two

subspecies differing in ozone sensitivit y. Plant Biology, 16, 375– 384.

Alonso, R., Elvira, S., Inclán, R., Bermejo, V., Castillo, F. J., & Gimeno, B.

S. (2003). Responses of Aleppo pine to ozone. In D. F. Karnosky,

K. E. Percy, A . H. Chappelka, & C. J. Simpson (Eds.), Air pollution,

global change and forests in the new millenium (pp. 211– 230). Els evier

Science Ltd.

Anav, A., De Marco, A., Collalti, A ., Emberson, L., Feng, Z., Lombardozzi,

D., Sicard, P., Verbeke, T., Viovy, N., Vitale, M., & Paoletti, E. (2022).

Legislative and functional aspects of different metrics used for

ozone risk assessment to forests. Environmental Pollution, 295,

118690.

Anav, A., De Marco, A., Proietti, C., Alessandri, A., Dell'Aquila, A., Cionni,

I., Friedlingstein, P., Khvorostyanov, D., Menut, L., Paoletti, E.,

Sicard, P., Sitch, S., & Vitale, M. (2016). Comparing concentration-

based (AOT40) and stomatal uptake (PODY) metrics for ozone

risk assessment to European forests. Global Change Biology, 22,

1608– 1627.

Antoniadis, V., Levizou, E., Shaheen, S. M., Ok, Y. S., Sebastian, A., Baum,

C., Prasad, M. N. V., Wenzel, W. W., & Rinklebe, J. (2017). Trace

elements in the soil- plant interface: Phytoavailability, transloca-

tion, and phytoremediation– A review. Earth- Science Reviews, 171,

621– 645.

Aoyama, M., Hirose, K., & Igarashi, Y. (2006). Re- construction and up-

dating our understanding on the global weapons tests 137Cs fallout.

Journal of Environmental Monitoring, 8, 431.

Archer, C. R., Pirk, C. W. W., Carvalheiro, L. G., & Nicolson, S. W.

(2014). Economic and ecological implications of geographic bias

in pollinator ecology in the light of pollinator declines. Oikos, 123,

401– 407.

Augustaitis, A., Augustaitienė, I., Kliučius, A., Pivoras, G., Šopauskienė,

D., & Girgždienė, R. (2010). The seasonal variability of air pollu-

tion effects on pine conditions under changing climates. European

Journal of Forest Researc, 129, 431– 441.

Augustaitis, A., & Bytnerowicz, A. (2008). Contribution of ambient

ozone to Scots pine defoliation and reduced growth in the Central

European forests: A Lithuanian case stud. Environmental Pollution,

155, 436– 445.

Badea, O., Silaghi, D., Taut, I., Neagu, S., & Leca, S. (2013). Forest

monitoring- assessment, analysis and warning system for forest

ecosystem status. Notulae Botanicae Horti Agrobotanici Cluj- Napoca,

41, 613– 625.

Baig, S., Medlyn, B. E., Mercado, L., & Zaehle, S. (2015). Does the growth

response of woody plants to elevated CO2 increase with tempera-

ture? A model- oriented meta- analysis. Global Change Biology, 21,

4303– 4319.

Bebber, D. P. (2021). The gap between atmospheric nitrogen deposi-

tion experiments and reality. Science of the Total Environment, 801,

149774.

Belyazid, S., Phelan, J., Nihlgård, B., Sverdrup, H., Driscoll, C., Fernandez,

I., Aherne, J., Teeling- Adams, L. M., Bailey, S., Arsenault, M.,

Cleavitt, N., Engstrom, B., Dennis, R., Sperduto, D., Werier, D., &

Clark, C. (2019). Assessing the effects of climate change and air

pollution on soil properties and plant diversity in northeastern U.S.

hardwood forests: Model setup and evaluation. Water, Air, & Soil

Pollution, 230, 1– 33.

Belyazid, S., Sverdrup, H., Kurz, D., & Braun, S. (2010). Exploring ground

vegetation change for different deposition scenarios and methods

for estimating critical loads for biodiversity using the for SAFE- VEG

model in Switzerland and Sweden. Water, Air, & Soil Pollution, 216 ,

2 8 9 – 3 1 7 .

Benjamin, M. T., & Winer, A. M. (1998). Estimating the ozone e forming

potential of urban trees and shrubs. Atmospheric Environment, 32,

5 3 – 6 8 .

Beresford, N. A., Fesenko, S., Konoplev, A., Skuterud, L., Smith, J. T., &

Voigt, G. (2016). Thirty years after the Chernobyl accident: What

lessons have we learnt? Journal of Environmental Radioactivity, 157,

7 7 – 8 9 .

Beresford, N. A ., Scott, E. M., & Copplestone, D. (2020). Field effects

studies in the Chernobyl Exclusion Zone: Lessons to be learnt.

Journal of Environmental Radioactivity, 211, 105893.

Berger, T. W., Türtscher, S., Berger, P., & Lindebner, L. (2016). A slight

recovery of soils from acid rain over th e last thre e dec ade s is not re-

flected in the macro nutrition of beech (Fagus sylvatica) at 97 forest

stands of the Vienna woods. Environmental Pollution, 216, 6 2 4 – 6 3 5 .

Blanco, J. A. (2013). Modelos ecológicos: Descripción, explicación y pre-

dicción. Ecosistemas, 22(3), 1– 5.

DE MARCO et al.

Blan co, J. A., Ameztegui, A., & Rodríg uez, F. (2020) . Mod ellin g forest eco -

systems: A crossroad between scales, techniques and applications.

Ecological Modelling, 425, 109030.

Blande, J. D. (2021). Effects of air pollution on plant- insect interac-

tions mediated by olfactory and visual cues. Current Opinion in

Environmental Science & Health, 19, 100228.

Boccuzzi, G., Nakazato, R. K., Pereira, M. A. G., Rinaldi, M. C. S., Lopes,

M. I. M. S., & Domingos, M. (2021). Anthropogenic deposition in-

creases nitrogen- phosphorus imbalances in tree vegetation, litter

and soil of Atlantic Forest remnants. Plant and Soil, 461, 341– 354.

Bonan, G. B. (2008). Forests and climate change: Forcings, feedbacks,

and the climate benefits of forests. Science, 320, 1444– 1449.

Bruckman, V. J., Terada, T., Fukuda, K., Yamamoto, H., & Hochbichler, E.

(2016). Overmature periurban Quercus– Carpinus coppice forests in

Austria and Japan: A comparison of carbon stocks, stand charac-

teristics and conversion to high forest. European Journal of Forest

Research, 135, 857– 869.

Butler, T., Vermeylen, F., Lehmann, C. M., Likens, G. E., & Puchalski, M.

(2016). Increasing ammonia concentration trends in large regions

of the USA derived from the NADP/AMoN network. Atmospheric

Environment, 146, 132– 140.

Bytnerowicz, A., Hsu, Y. M., Percy, K., Legge, A., Fenn, M. E., Schilling,

S., Frączek, W., & Alexander, D. (2016). Ground- level air pollution

changes during a boreal wildland mega- fire. Science of the Total

Environment, 572, 755– 769.

Calfapietra, C., Fares, S., Manes, F., Morani, A., Sgrigna, G., & Loreto,

F. (2013). Role of biogenic volatile organic compounds (BVOC)

emitted by urban trees on ozone concentration in cities: A review.

Environmental Pollution, 183, 71– 80.

Cassimiro, J. C., Moura, B. B., Alonso, R., Meirelles, S. T., & Moraes, R. M.

(2016). Ozone stomatal flux and O3 concentration- based metrics

for Astronium graveolens Jacq., a Bra zi li an native forest tr ee species.

Environmental Pollution, 213, 1007– 1015.

Cazzolla Gatti, R., Reich, P. B., Gamarra, J. G. P., Crowther, T., Hui, C.,

Morera, A., Bastin, F., de Miguel, S., Nabuurs, G.- J., Svenning, J.-

C., Serra- Diaz, J. M., Merow, C., Enquist, B., Kamenetsky, M., Lee,

J., Zhu, J., Fang, J., Jacobs, D. F., Pijanowski, B., … Liang, J. (2022).

The number of tree species on Earth. Proceedings of the National

Academy of Sciences of the United States of America, e2115329119.

https://doi.or g/10.1073/pnas.21153 29119

Chino, M., Nakayama, H., Nagai, H., Terada, H., Katata, G., & Yamazawa,

H. (2011). Preliminary estimation of release amounts of 131I and

137Cs accidentally discharged from the Fukushima Daiichi Nuclear

Power Plant into the atmosphere. Journal of Nuclear Science and

Technology, 48(7), 1129– 1134.

Chiwa, M. (2020). Ten- year determination of atmospheric phospho-

rus deposition at three forested sites in Japan. Atmospheric

Environment, 223, 117247.

Chuberre, C., Plancot, B., Driouich, A., Moore, J. P., Bardor, M., Gügi, B.,

& Vicré, M. (2018). Plant immunity is compartmentalized and spe-

cialized in roots. Frontiers in Plant Science, 9, 1692.

Clark, C. M., Bai, Y., Bowman, W. D., Cowles, J. M., Fenn, M. E., Gilliam,

F. S., Phoenix, G. K., Siddique, I., Stevens, C. J., Sverdrup, H.

U., & Throop, H. L. (2013). Nitrogen Deposition and Terrestrial

Biodiversity. In S. A. Levin (Ed.), Encyclopedia of biodiversity (2nd ed.,

pp. 519– 536). Elsevier Inc.

Collalti, A., Ibrom, A., Stockmarr, A., Cescatti, A., Alkama, R., Fernández-

Martínez, M., Ciais, P., Sitch, S., Friedlingstein, P., Goll, D. S., Nabel,

J. E. M. S., Pongratz, J., Arneth, A., Haverd, V., & Prentice, I. C.

(2020). Forest production efficiency increases with growth tem-

perature. Nature Communications, 11, 5322.

Collalti, A ., Marconi, S., Ibrom, A., Trotta, C., Anav, A., D'Andrea,

E., Matteucci, G., Montagnani, L., Gielen, B., Mammarella, I.,

Grünwald, T., Knohl, A., Berninger, F., Zhao, Y., Valentini, R., &

Santini, M. (2016). Validation of 3D- CMCC Forest Ecosystem

Model (v.5.1) against eddy covariance data for 10 European forest

sites. Geoscientific Model Development, 9, 479– 504.

Collalti, A., Trotta, C., Keenan, T., Ibrom, A., Bond- Lamberty, B., Grote, R.,

Vicca, S., Reyer, C. P. O., Migliavacca, M., Veroustraete, F., Anav, A.,

Campioli, M., Scoccimarro, E., Grieco, E., Cescatti, A., & Matteucci,

G. (2018). Thinning can reduce losses in carbon use efficiency and

carbon stocks in managed forests under warmer climate. Journal of

Advances in Modelling Earth Systems, 10(10), 2427– 2452.

Committee on the Biological Effects of Ionizing Radiation. (1990). Health

effects of exposure to low levels of ionizing radiation: BEIR V. National

Research Council, National Academy Press.

Cooper, O. R., Parrish, D. D., Ziemke, J., Balashov, N. V., Cupeiro, M.,

Galbally, I. E., Gilge, S., Horowitz, L., Jensen, N. R., Lamarque, J.- F.,

Naik, V., Oltmans, S. J., Schwab, J., Shindell, D. T., Thompson, A. M.,

Thouret, V., Wang, Y., & Zbinden, R. M. (2014). Global distribution

and trends of tropospheric ozone: An observation- based review.

Elementa: Science of the Anthropocene, 2, 000029.

Coordination Centre for Effects. (2017). European critical loads:

Database , biodiversity and ecosystems at risk. CCE final report 2017

(J.- P. Hettelingh, M. Posch, & J. Slootweg, Eds.). Rijksinstituut voor

Volksgezondheid en Milieu.

Cunha- Zeri, G., & Ometto, J. (2021). Nitrogen emissions in Latin America:

A conceptual framework of drivers, impacts, and policy responses.

Environmental Development, 38, 100605.

Curran, P. J., Dungan, J. L., & Gholz, H. L. (1992). Seasonal LAI in slash

pine estimated with Landsat TM. Remote Sensing of Environment,

39(1), 3– 13.

Curtis, P. S., & Wang, X. Z. (1998). A meta- analysis of elevated CO2 ef-

fects on woody plant mass, form, and physiology. Oecologia, 113,

299– 313.

Dale, V. H., Joyce, L. A., McNulty, S., & Neilson, R. P. (2000). The interplay

between climate change, forests, and disturbances. Science of the

Total Environment, 262, 201– 204.

De Jager, N. R., Drohan, P. J., Miranda, B. M., Sturtevant, B. R., Stout, S. L.,

Royo, A. A., Gustafson, E. J., & Romanski, M. C. (2017). Simulating

ungulate herbivory across forest landscapes: A browsing extension

for LANDIS- II. Ecological Modelling, 350, 11– 2 9.

De Marco, A., Anav, A., Sicard, P., Feng, Z., & Paoletti, E. (2020). High

spati