Forest vulnerability to extreme climatic events in Romanian Scots [629561]

Forest vulnerability to extreme climatic events in Romanian Scots
pine forests
Cristian Gheorghe Sidora,⁎, J. Julio Camarerob, Ionel Popaa, Ovidiu Badeaa,
Ecaterina Nicoleta Apostola,R a d uV l a da
aNational Institute for Research and Development in Forestry ‘Marin Dr ăcea ’, Calea Bucovinei 73 bis, Câmpulung Moldovenesc, Romania
bInstituto Pirenaico de Ecología (IPE-CSIC), Avda. Montañana 1005, 50192 Zaragoza, Spain
HIGHLIGHTS
•The identi fication of the droughts in flu-
ences on the radial growth and vitality
loss of Romanian Scots pine is quanti-fied.
•A country-wide BAI Scots pine dendro-
chronological network was developed
and used.
•Romanian Scots pine forests were se-
verely impacted by the 2000 and 2012
droughts.
•The high temperatures and low precipi-
tation from April to August were the
main climatic drivers of growth reduc-tion.
•The projected appearance of similar
prolonged and severe droughts in the
future will lead to the damage in someScots pine forests in Romania.GRAPHICAL ABSTRACT
abstract article info
Article history:
Received 7 February 2019Received in revised form 2 May 2019Accepted 3 May 2019Available online 4 May 2019
Editor: Alessandra De MarcoIn the last years, large-scale mass forest withering and dieback have been reported for Scots pine ( Pinus sylvestris )
across eastern Europe, particularly in Romania. In these regions, the climate models forecast an increase in inten-sity and frequency of extreme climate events such as drought. Taking into account these aspects, the exact iden-
tification of the in fluences of drought on the loss of radial growth and vitality in Scots pine stands becomes
mandatory. To achieve this aim, we developed the first country-wide Scots pine dendrochronological network
in Romania consisting of 34 chronologies of basal area increment (BAI), and including 1401 individual tree-
ring width series. Romanian Scots pine forests were severely impacted by the 2000 and 2012 droughts. The
high temperatures and low precipitation from April to August were the main climatic causes of radial-growth re-duction and large-scale withering in some areas. By mapping post-drought growth resilience, we identi fied loca-
tions where resilience was low and could identify foci of future forest dieback and high tree mortality. The
projected appearance of similar prolonged and severe droughts in the future will lead to the damage or local ex-
tinction of some Scots pine forests in Romania, regardless of their age, composition or spatial location. The elab-oration of adaptive forest management strategies to the impact of climate changes, speci fically designed for the
Scots pine stands, is not possible without knowing and understanding these aspects.
© 2019 Elsevier B.V. All rights reserved.Keywords:
Scots pine forestsDroughtResilienceTree ringsSpatial variability
Forest management strategiesScience of the Total Environment 678 (2019) 721 –727
⁎Corresponding author.
E-mail address: [anonimizat] (C.G. Sidor).
https://doi.org/10.1016/j.scitotenv.2019.05.021
0048-9697/© 2019 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv

1. Introduction
Forest dynamics depend on climatic factors and biotic interactions
which change as a function of spatio-temporal gradients ( McKenney
et al., 2007 ;González-de-Andrés et al., 2017 ). Understanding the impact
of climate change on forests is a relevant subject that concerns a large
number of the international scienti fic community ( Andreassen et al.,
2006 ;Bonan, 2008 ;Alkama and Cescatti, 2016 ). For instance, in
Europe recent severe drought episodes have triggered dieback events
and determined a growth decline of some forests located in dry sites,
usually near the southernmost limit of the species distribution area
(Carnicer et al., 2011 ,Sánchez-Salguero et al., 2012 ;Camarero et al.,
2015 ). These climate-driven disturbances have led to productivity de-
clines and shifts in forest composition among other responses
(Galiano et al., 2010 ;Rigling et al., 2013 ;Gazol et al., 2018 ). All climate
scenarios (RCPs) forecast a rapid warming and an increase of droughts'
frequency, duration and severity across Europe ( Rajczak et al., 2013 ).
Thus, it has become imperative to gain an in-depth understanding of
the forest responses to the decrease of water availability and, particu-
larly, to the impact of extreme drought events to elaborate adaptative
forest management strategies ( Temperli et al., 2012 ).
The importance of extreme climatic events has become an essential
ecological research issue in forest ecology, being a key topic on the
global change agenda ( Smith, 2011 ;IPCC, 2013 ). Forest management
depends on the uncertainties due to climatic risks ( Zang et al., 2014 ).
To reduce this uncertainty, the potential of a species can be assessed
based mainly on retrospective studies on its radial growth (tree-ring
width data), mortality and reproduction performance ( Bolte et al.,
2009 ). Therefore, a reconstruction of radial-growth responses to climate
extremes as drought across ample ecological gradients will reduce the
uncertainty associated with future growth projections in response to
projected drought. Objective informations on the way adult trees react
to climate variation, at multiple spatial and temporal levels, can be ob-
tained using tree-rings analyses (dendrochronology), and such long-
term information is not accessible using ecophysiological approaches
often focused on seedlings or saplings ( Carrer and Urbinati, 2006 ). The
annual tree-ring width varies from year to year in a more or less regular
manner, and a large part of this variability depends on changing climatic
conditions ( Fritts, 1976 ;Speer, 2010 ). A pronounced decline in radial
growth in response to a severe drought is considered a proxy of the re-versible loss in tree vitality ( Dobbertin et al., 2007 ).
At European level, one of the most widely distributed and vulnerable
species to the in fluence of climate changes and drought is Scots pine
(Pinus sylvestris L.), and extreme climate events such as dry spells and
heat waves will have a signi ficant negative impact on this species' pop-
ulations, leading to important economic losses ( Hanewinkel et al.,
2013 ). Scots pine is sensitive to drought during the growing season, it
closes stomata to avoid xylem cavitation ( Irvine et al., 1998 ) thus reduc-
ing carbon uptake and growth (
Herrero et al., 2013 ;Liang et al., 2013 ;
Panayotov et al., 2013 ;Fernández-de-Uña et al., 2013 ;Bigler et al.,
2006 ;Rigling et al., 2013 ;Cedro, 2006 ;Pichler and Oberhuber, 2007 ).
In extreme cases, severe drought causes irreversible growth reductions,
reduces the adaptive capacity of trees leading to secondary effects (e.g.
biotic stress), triggers needle shedding and leads to canopy dieback and
tree death ( Sánchez-Salguero et al., 2012 ;Camarero et al., 2015 ). For in-
stance, in Romania severe defoliation and dieback affected on average
1529 ha yr−1of Scots pine forests (ca. 2% from total Romanian Scots
pine forests) during the 2000s; and in 2012 –2013 the withering of the
Scots pine recorded the highest extent in the history of forest records,
affecting all regions in the country, with a total affected area of
8086 ha (ca. 10% of total Romanian Scots pine forests area), far larger
than the average of the preceding decades ( Chira and D ănescu, 2013 ).
The Scots pine is the most widespread conifer in the world, covering
approximately 143 million hectares (3.7% of the total world forest area)
from Siberia to Iberia ( Stănescu, 1979 ). At the European level, the Scots
pine forests cover approximately 28 million ha, which represents 20% ofthe total forest area, and reach their southeastern distribution limit in
the Romanian Carpathians and the Balkan Peninsula ( Houston Durrant
et al., 2016 ). In Romania, the Scots pine forests encompass a wide altitu-
dinal gradient (310 –1600 m) and they are found on diverse geological
substrates ( Alexe, 1964 ). According to the Chiri ță(1981) the total area
of natural and arti ficial Scots pine stands is approximately 89,000 ha
(Ciobanu, 2003 ). Overall, Romanian Scots pine forests show a
fragmented but relatively stretched area and encompass ample ecolog-
ical gradients ( Popescu, 1984 ). These two features and the geographi-
cally marginal situation of Romanian stands make them particularly
interesting as monitors of the effects of droughts on forests.
To improve the available knowledge regarding the vulnerability of
Scots pine to the impact of droughts, here we address the following ques-
tions: (i) Which are the years in which extreme climatic events as
droughts have determined signi ficantly low radial-growth rates in
Romanian Scots pine forests?, (ii) Which are the main climatic variables
responsible for these growth changes?, and (iii) What is the spatial pat-
tern of the Scots pine resilience in Romania after extreme climatic events?
2. Materials and methods2.1. Study sites and tree species
We developed and used the first nation-wide Scots pine dendrochro-
nological network in Romania. This network consists of 34 chronologies
of basal area increment (BAI), corresponding to 1401 individual tree-
ring width series. The sampling sites are located in mixed and pure, natu-
ral and arti ficial, declining and healthy Scots pine forest stands, evenly dis-
tributed across the Romanian forested territory ( Fig. 1 ).
In Romania, climate conditions range from temperate to continental.
The transitional climatic character corresponds to a gradient from the
temperate-oceanic climate, prevailing in Western Europe, to the tem-
perate climate typical of Eastern Europe. The presence in the centre of
the country of the Carpathian Mountains and the orientation of the
mountain peaks and valleys greatly in fluences climate conditions
(Romanian National Meteorological Administration, 2008 ).
To use a homogeneous set of climatic data for the entire study area,
we used the E-OBS ver. 14 climatic database with a resolution of 0.25°
(Haylock et al., 2008 ). For each site, the monthly mean temperature
and total precipitation data were obtained for the period 1985 –2015.
2.2. Field sampling and dendrochronological methods
From each sampling site at least 40 dominant trees were sampled in
the autumn 2015 and during the vegetation season in 2016 ( Table 1 ).
From each tree one core was extracted at 1.3 m using increment borers
(Bosela et al., 2014 ). All cores were prepared according to standard den-
drochronological protocols, including air drying, mounting and sanding
(Cook and Kairiukstis, 1990 ). Tree-ring widths were measured with a
0.01 mm resolution based on the scanned images at 1200 dpi using
the Coorecorder software ( Cybis Elektronik, 2016 ). All individual tree-
ring width series were checked for measurements error and misdating
both by graphical comparison with average chronology, using the
TSAP-Win software (RINNTECH, Inc.), and by correlation statistical anal-
ysis using the COFECHA software ( Holmes, 1983 ). Tree-ring width data
were converted into BAI series because the latter variable are a more
representative characterization of changes in tree biomass than the for-
mer ( Motta and Nola, 2001 ). We computed a BAI chronology for each
sampling site by assuming a circular shape of tree stem at the sampling
height ( Monserud and Sterba, 1996 ).
A smoothing spline with a 50% frequency response cut-off at 67% of
series length was applied to each individual BAI series to eliminate any
size-related trends ( Biondi and Qeadan, 2008 ). Site-speci fic BAI chro-
nologies were calculated by averaging BAI indices, calculated as ratios
between observed and fitted BAI values and removing the first-order
autocorrelation using autoregressive models ( Cook, 1987 ), using722 C.G. Sidor et al. / Science of the Total Environment 678 (2019) 721 –727

biweight robust means. Data processing was done using the dplR ( Bunn
et al., 2017 ) package in R (R Core Team, 2017). For each site we calcu-
lated several descriptive dendrochronology statistics, including the
mean BAI for the common 1985 –2015 period, the correlation coef fi-
cients among BAI series within each site (Rbar) and the mean sensitivity
or relative difference between consecutive BAI indices ( Table 1 ).
2.3. Statistical analyses
The spatial variability and the similarity between BAI series were
analysed using Principal Component Analysis (PCA) calculated on the
covariance matrix resulting of comparing site BAI chronologies. The
BAI chronologies were classi fied using a hierarchical clustering analysis
based on the resulting loadings of the first (PCA1) and second (PCA2)
principal components. These analyses were performed using the
XLSTAT software ( XLSTAT, 2017 ).
Negative (years of extremely low growth) and positive (years of ex-
tremely high growth) pointer years were calculated based on normal-
ized BAI values and considering a five year-long moving window
(Cropper, 1979 ;van der Maaten-Theunissen et al., 2015 ). We consid-
ered only those pointer years with a representation of more than 60%
of the total number of individual series from the chronologies groups,
with a growth reduction. The Weiser software was used to calculate
pointer years ( Gonzalez, 2001 ). For each pointer year, we presented
the monthly average temperatures and total precipitations, which had
an influence on tree growth ( Sidor et al., 2018 ). The climatic variables
were expressed in normalized values with reference to the period
1985 –2015 and considering a temporal window going from the previ-
ous September to the current August, which encompasses months cli-
matically relevant to Scots pine growth.
For each BAI chronology, the degree of decrease and recovery of radial
growth after drought events was determined by calculating the resilience
for each negative pointer year. The resilience index, with values from 0 to
4i no u rc a s e ,r e flects the capacity of trees to recover the growth level after
droughts, and it is quanti fied as the ratio between the BAI value after the
year of the drought and the previous-year BAI value ( Lloret et al., 2011 ).
The impacts of climate on each negat ive pointer year were highlighted
by mapping the spatial interpolation of the resilience indices of the 34 BAI
Scots pine chronologies using the inverse distance weighting interpola-
tion in QGIS software ( QGIS Development Team, 2009 ).
3. Results3.1. Growth data and BAI patterns
The mean BAI had values between 313 and 1725.8 mm
2yr.−1
(Table 1 ). The BAI chronologies showed a wide range of mean sensitivityvalues, from 0.19 to 0.33. The mean correlation coef ficients among the BAI
series (Rbar) within each site vari ed between 0.24 and 0.55. Chronology
length varied from 32 to 194 years with a mean chronology length of
71 years.
The spatial variability and the similarity between the BAI series were
summarized through a PCA ( Fig. 2 A), which showed that all series
showed positive PCA1 loadings and the PCA1 explained 39.67% of the
BAI variability. PCA1 represents the common climatic signal in the ex-
treme negative years. PCA2 accounted for 11.20% of the BAI variability
and expressed the response of the trees to the local climate conditions.
The hierarchical cluster analysis of the BAI chronologies highlighted
three groups of sites ( Fig. 2 B).
Regarding the segregation of the years, the years showing the most
negative loadings along the PCA1 and PCA2 axes were 2003, 2012 and
2013, and 1994, respectively. The highest PCA1 and PCA2 loadings
corresponded to the years 1987 and 2008, respectively.
3.2. Pointer years and their climatic determinants
In the first group of BAI series, four negative pointer years (1992,
2000, 2003 and 2012) and two positive (1999, 2014) pointer years
were identi fied (Fig. 3 A-first group). The low radial growth in 1992
was mainly determined by the very high temperatures in August,
which had values that exceeded two times the standard deviation
(SD) of the 1985 –2015 mean ( Fig. 3 B-first group).
Negative pointer years of the first BAI group are associated with low
rainfall from January to February, and in May and August of the growth
year with potential negative effect on Scots pine growth ( Fig. 3 C-first
group). The pointer year 2000 was characterized by high temperatures
at the beginning of the vegetation season (April) and a low precipitation
regime throughout the growing season. The pointer year 2003 was
mainly driven by high temperatures in May –June (in May the average
temperature exceeded 2.5 times the SD) and low rainfall in the same pe-
riod and in August. The low radial growth in 2012 was determined by
the very high temperatures in the period from April to August, in July
the average temperatures being more than 2.6 times the SD. The high
radial growth in 1999 was determined mainly by a previous autumn
colder than usual, a warmer prior winter and a relatively high precipita-
tion regime at the beginning of the growing season. The positive pointer
year detected in 2014 was associated with warm conditions from Janu-
ary to April and by relatively high rainfall in the early growing season
and in July.
The negative and positive pointer years identi fied in the second
group of BAI were 1987, 1996, 2000, 2003, 2012, 2013, and 2008,
2014, respectively ( Fig. 3 A-second group). The low radial growth in
1987 was related to cold conditions in March –May, and low precipita-
tions in the autumn preceding the formation of the annual ring, and
Fig. 1. A—Locations of the 34 Scots pine study sites (yellow circles); B —The natural spatial distribuion of Scots pine in Europe according to EUFORGEN (green area) and situation of
Romania in Eastern Europe (red outline) ( http://www.euforgen.org/ ).723 C.G. Sidor et al. / Science of the Total Environment 678 (2019) 721 –727

also in current June –July ( Fig. 3 B and C- second group). In 1996, the re-
duced radial growth was associated with cold spring conditions
followed by very high temperatures in May –June and dry June and
July conditions. As in the case of the first group, the 2000, 2003 and
2012 negative pointer years were associated with the same climatic pat-terns. The negative pointer year 2013 was characterized by high tem-
peratures in the prior autumn and in the early growing season
(March –April). The high radial growth observed in 2008 was due to
the high rainfall regime during the prior autumn and in current July.
As in the case of the first group, the positive pointer year in 2014 was as-
sociated with warm January –April conditions and high rainfall in the
early growing season (March –April) and July.
In the case of the third group of BAI series, the negative and positive
pointer years identi fied were 1987, 1996, 2000, 2003, 2012, and 2014,
respectively ( Fig. 3 A- third group). The climatic conditions associated
with these extreme growth values were the same as those aforemen-
tioned for the first and second BAI groups ( Fig. 3 B and C- third group).
3.3. Geographical patterns of resilience in negative pointer years of
Romanian Scots pine forests
The analysis of the Scots pine resilience spatial variability after the
negative pointer year 1987 highlights the fact that, at the level of the
whole country, not all the stands reduced their growth rate after the
drought ( Fig. 4 ). Most of them returned to the previous growth rate in
the next year after the drought, or even reached higher growth levels.
The most affected stands, with low resilience, were located in the cen-
tral area of the country, which needed a longer period to return to the
previous radial growth rate.
The negative impact of the 1992 drought on Scots pine growth led to
a low resilience in the north-west stands. Even if the negative impact ofthe 1996 drought was associated with strong decreases of radial growth
in the case of 74% of the individual BAI series studied, their resilience
was the highest compared with the other negative pointer years. Re-
duced resilience in 1996 was restricted to a few stands, located mainly
in eastern Romania. During the years 2000, 2003 and 2012, the impactof the extreme climate on Scots pine growth showed a very large spatial
extent. In 2000, the growth resilience was low most of the country,
being particularly evident in the western half. The unfavourable climate
conditions in 2000 had a very strong negative impact on growth. Re-
garding the impact of the climate in 2003, the resilience of the stands
was higher compared with the year 2000. As a result, the Scots pine for-
ests located in the eastern half part of the country experienced the
highest loss in BAI resilience. The negative impact of 2012 climate con-
ditions on growth was very strong and widespread, and lead to a reduc-
tion in resilience and a low capacity to recovery the pre-event level of
growth, across Scots pine forests over all the country.
4. Discussion
The necessity and the opportunity of the present study started from
a very practical forest question related to the resilience and post-
drought management of Romanian Scots pine forests showing large-
scale mass decline and withering in the last years in the context of cur-
rent climate change. This is a very relevant question given the silvicul-
tural and ecological importance of this conifer species across Europe,
and particularly near its southeastern marginal distribution. Scots pine
is a species with a high ecological plasticity, facing adverse climate
and soil conditions, reaching at the same time a signi ficant productivity
in some sites ( Costandache et al., 2006 ).
According to the IPCC (2013) , in the next decades, in Romania, the
alternation between very dry years and very rainy years is highly likely.Table 1
Site description and chronology parameters of the study sites. In the ‘composition ’column, P. s. stands for Pinus sylvestris .
Site
codeSite name Latitude
(N)Longitude
(E)Elevation
(m a.s.l.)Chronology
periodChronology
length (yrs)Mean BAI
1985 –2015
(mm2/year)Mean
RbarMean
sensitivityComposition
ALEA Ale șd – Bihor 47.0011 22.3978 500 1975 –2016 40 1057.2 0.48 0.30 Over 70% P. s.
BACA Bac ău – Bac ău 46.4225 27.02 350 1979 –2016 37 1563.7 0.40 0.25 Over 70% P. s.
BARA Bârlad – Vaslui 46.2561 27.5727 100 1979 –2016 37 1543.1 0.41 0.25 Over 70% P. s.
BELA Beli ș- Cluj 46.6561 22.9731 1100 1961 –2016 55 1083.0 0.34 0.24 Mixture stands
BICA Bicaz chei – Neam ț 46.8061 25.8227 1000 1939 –2016 83 453.6 0.28 0.28 Mixture stands
BISA Bicaz sat – Neam ț 46.9015 26.0246 600 1958 –2016 58 632.2 0.51 0.29 Over 70% P. s.
BOTA Boto șani – Boto șani 47.5832 26.8696 200 1978 –2016 38 1670.5 0.49 0.29 Mixture stands
BREA Breaza – Suceava 47.68 25.3178 1300 1820 –2016 194 650.8 0.28 0.23 Mixture stands
BRGA Mure șenii Bârg ăului – Bistri ța 47.2101 24.8733 700 1920 –2016 96 1022.9 0.35 0.27 Over 70% P. s.
BUZA Buz ău – Buz ău 45.1213 26.5482 650 1968 –2016 48 992.3 0.35 0.25 Mixture stands
CARA Caransebe ș–Cara șSeverin 45.3647 22.3696 650 1984 –2015 31 1754.1 0.35 0.24 Over 70% P. s.
CLUA Cluj – Cluj 46.7238 23.6197 600 1982 –2016 34 1053.3 0.42 0.23 Over 70% P. s.
CODA Codlea – Bra șov 45.8198 25.4462 600 1946 –2016 70 1436.7 0.35 0.28 Over 70% P. s.
DUMA Dumitre ști – Sibiu 46.2057 24.6829 420 1906 –2016 110 1056.9 0.33 0.29 Over 70% P. s.
FAGA F ăget – Timi șoara 45.8069 22.2241 200 1975 –2015 40 1432.7 0.40 0.25 Over 70% P. s.
HATA Ha țeg – Hunedoara 45.5428 22.7797 550 1984 –2016 32 876.4 0.44 0.31 Over 70% P. s.
MARA Maramure ș- Maramure ș 47.6669 24.1522 850 1905 –2015 41 932.7 0.39 0.27 Over 70% P. s.
MEHA Topolni ța – Mehedin ți 44.7449 22.7091 500 1969 –2015 46 1082.5 0.32 0.32 Over 70% P. s.
MIEA Miercurea Ciuc – Harghita 46.3312 25.8161 500 1898 –2016 118 473.4 0.40 0.33 Mixture stands
PATA P ătrăuți – Suceava 47.7583 26.1931 400 1890 –2016 126 1583.7 0.30 0.23 Mixture stands
POIA Poiana Cerbului – Arge ș 45.3305 24.6338 900 1968 –2015 47 1094.3 0.40 0.25 Over 70% P. s.
RETA Retezat – Hunedoara 45.3767 22.8122 1600 1875 –2016 141 752.8 0.24 0.19 Mixture stands
RONA Rona – Maramure ș 47.8846 24.0323 450 1979 –2015 36 1042.4 0.36 0.24 Mixture stands
SALA S ăliște – Sibiu 45.8055 23.8258 650 1902 –2016 114 835.1 0.47 0.30 Over 70% P. s.
SILA Silva șu – Hunedoara 45.6972 22.9008 350 1978 –2016 38 1359.8 0.48 0.32 Over 70% P. s.
TGJA Târgu Jiu – Gorj 45.1479 23.4291 380 1972 –2015 43 977.5 0.48 0.33 Mixture stands
TGMA Târgu Mure ș- Mure ș 46.6253 24.4494 500 1975 –2016 41 876.4 0.45 0.32 Mixture stands
TGNA Târgu Neam ț- Neam ț 47.2094 26.3624 500 1922 –2016 94 358.8 0.30 0.35 Mixture stands
TGOA Târgu Ocna – Bac ău 46.2063 26.4248 600 1870 –2015 145 754.6 0.46 0.27 Over 70% P. s.
TOPA Topoloveni – Arge ș 44.8787 25.0498 350 1972 –2015 43 1059.3 0.55 0.32 Mixture stands
VIDA Vidra – Vrancea 45.8924 26.9178 350 1966 –2016 50 558.8 0.30 0.28 Mixture stands
VINA Vintileasca – Vrancea 45.6032 26.6928 800 1890 –2016 126 1453.2 0.38 0.27 Over 70% P. s.
ZALA Zal ău-S ălaj 47.1081 23.1939 300 1904 –2016 112 674.3 0.41 0.31 Mixture stands
ZETA Zetea – Harghita 46.3342 25.3153 700 1936 –2016 80 1119.5 0.44 0.27 Over 70% P. s.724 C.G. Sidor et al. / Science of the Total Environment 678 (2019) 721 –727

Thus, an exact understanding of the impact of climate changes on the
growth and vitality of Scots pine near its southeastern distribution
limit and a knowledge of the most vulnerable areas at the country
level currently represent essential requirements. For that reason, we de-
veloped the first Scots pine dendrochronological network from
Romania and we applied dendroecological methods to gain a better un-
derstanding of forest resilience and vulnerability. This case study:
(i) represents an excellent example on how quantifying and mapping
drought impacts on growth and resiliencie, and (ii) illustrates how
using those measured to forecast future foci of high vulnerability to
those climate extremes.
Our investigations highlight that during the last two decades the
Romanian Scots pine forests were strongly affected by extreme climatic
events like the 2000 and 2012 droughts. High temperature is a common
factor in massive dieback and tree-mortality events, in part through
temperature-driven increase in water de ficit due to increased evapo-
transpiration ( Allen et al., 2010 ;Williams et al., 2013 ). Summer drought
stress hinders radial growth in Scots pine and thus may have profound
effects on the final tree-ring width ( Thabeet et al., 2009 ;Eilmann et al.,
2011 ). Scots pine responds to drought by a fast reduction of transpira-
tion through a rapid stomatal closure ( Irvine et al., 1998 ), but this re-
sponse may vary between coexisting trees as a function of their stress
level ( Hölttä et al., 2012 ). In central Europe, climate data indicate a re-
cent shift in the seasonality of water availability from summer to winter(Begert et al., 2005 ) and this will cause additional periodic drought
stress for those tree species that are not adapted to summer drought
as in the case of Scots pine. These adverse climate conditions may pre-
dispose trees to drought-induced dieback and death once needle shed-
ding, growth decline and vigour loss become irreversible ( Camarero
et al., 2015 ). Recurrent and extended droughts lead to accumulated car-
bon de ficits, which could result in a long-term limitation on hydraulic
conductivity and carbon uptake causing tree death ( McDowell et al.,
2008 ).
Our results highlighted that signi ficant reduction of Scots pine
growth is associated with high spring and summer temperature and
low precipitation during the growing season. Positive pointer years at
the level of Scots pine tree ring network in Romania are assoiated
with high precipitation during the summer and warm late winter and
early spring.
According to our results, the frequency of the extreme years has in-
creased over the past two decades, being a possible indicator of climatechange. If during the period 1985 –1994 just two pointer years were
found, in the period 1995 –2015 the frequency of extreme positive or
negative growth events increased.
The spatial variability of resilience indices of Scots pine growth
varies from year to year according to the prevailing dominant oceanic
or continental summer air circulation. In general, a lower resilience
can be observed in the wet temperate-oceanic western and central
Fig. 2. A—Principal Component Analysis (PCA) biplot of basal area increment indices; B —Hierarchical cluster analysis of BAI indices based on the first (PCA1) and second (PCA2) loadings
of the PCA.
Fig. 3. A-Pointer years of the Scots pine basal area increment (BAI) chronologies (grey bars – percent of individual series having the pointer year, red bars- negative pointer years, orange
bars- positive pointer years, black line- series of BAI indices); B- Average monthly temperatures expressed as standard deviations from the average of the 1985 –2015 period; C- Total
monthly precipitations expressed as standard deviations from the average of the 1985 –2015 period. In B and C the temporal window goes from the previous September up to the
current August (months of the prior and current years are abbreviated by lowercase and uppercase letters, respectively).725 C.G. Sidor et al. / Science of the Total Environment 678 (2019) 721 –727

parts of Romania compared with the temperate eastern regions. Such
reduction in resilience has already been observed in other wet regions
as compared with dry areas ( Gazol et al., 2018 ).
Similar results to those here presented were found in other regions
from east Europe. In Bulgaria, Scots pine stands located at low elevations
showed extremely negative pointer years in 1987, 1996, 2000 and 2003,
the main climatic factor for such growth reduction being droughts dur-
ing the growing season ( Panayotov et al., 2013 ). In Hungary, summer
droughts were the main driver of low radial growth in Scots pine
(Misi and Nafradi, 2016 ). In central European forests, drought con-
strains the physiological activity of Scots pine during summer
(Lebourgeois, 2000 ;Gruber et al., 2010 ;Oberhuber et al., 2011 ;Cuny
et al., 2014 ;Swidrak et al., 2014 ). In low-elevation and dry sites from
the Alps the summer moisture dependency of Scots pine growth was
also underlined ( Bigler et al., 2006 ). In the Mediterranean Basin previ-
ous studies highlighted many dieback events associated with spring
and summer droughts ( Martinez-Vilalta and Pinol, 2002 ;Sarris et al.,
2011 ;Sánchez-Salguero et al., 2012 ,Sánchez-Salguero et al., 2015 ;
Camarero et al., 2015 ). The impact of the 2003 extreme climatic year,
characterized by high summer temperatures, correlated with droughts
pronounced in the summer of 2004 and 2005, leading to a mass drying
of the Scots pine stands in south-east France ( Thabeet et al., 2009 ). In
the Iberian Peninsula the decrease in rainfall and the increase in mean
temperature were the most important factors causing a reduction in
the growth of Scots pine ( Bogino et al., 2009 ). There, low-elevation
stands were those more sensitive to warming and drought stress in
that region ( Candel-Pérez et al., 2012 ). The Polish lowlands Scots pine
populations were also sensitive to water shortage during the growing
season ( Wilczy ński, 1999 ). In Estonia, the most limiting factors for
Scots pine growth were cold winter and spring conditions followed by
dry summer conditions ( Hordo et al., 2009 ).
Future forestry policy strategies should reconsider the site classi fica-
tion of Romanian and possibly Scots pine forests, with the emphasis on
risk factors associated with more frequent extreme climate events such
as dry spells and heat waves. Mapping and quantifying growth re-
sponses to drought are valuable tools to detect and characterized vul-
nerable areas. Given that pure and young Scots pine stands have been
shown to be more vulnerable to climate extremes than mixed and
older stands ( Sidor et al., 2018 ), it is very important that future manage-
ment plans avoid establishing Scots pine plantations in the most vulner-
able areas, mainly because of the negative impact of summer
temperatures on Scots pine growth. In young monocultures, a low den-
sity should be preferred and declining individuals should be replaced
through natural regeneration of resistant or resilient individuals and
tree species ( Costandache et al., 2006 ).5. Conclusions
We developed the first country-level dendrochrnological network of
Scots pine in Romania, near the south-eastern distribution limit of the
species. The Romanian Scots pine stands, regardless of the composition
and spatial location were severely affected by the 2000 and 2012
droughts. In these two years, warm conditions during the growing sea-
son (from April to August) and low precipitation greatly reduced Scots
pine growth and productivity, triggering dieback and mortality events
in some areas. In the context of the current climate changes, successive
and unfavourable extreme climate events could become a common pat-tern in the next decades. The impacts of extreme climate events on for-
est growth and productivity and vitality should be improved by
mapping and quantifying post-event resilience. This would strengthen
forecasts on forest vulnerability including projections of future dieback
or mortality foci. Extreme climate events should be considered to de-
velop adaptive forest management and mitigation strategies across
drought-prone European forested regions.
Acknowledgments
The study was funded by the Romanian National Authority for Scien-
tific Research and Innovation (CNCS-UEFSCDI, project number PN-II-
RU-TE-2014-4-1229), within the project entitled “Scots pine forest de-
cline in Romania under the climate change impact ”. JJ Camarero thanks
the support of the CGL2015-69186-C2-1-R project (Spanish Ministry of
Economy), Spain.
References
Alexe, A., 1964. Pinul silvestru. Editura Agro-Silvic ă,B u c u r e ști (326 p).
Alkama, R., Cescatti, A., 2016. Biophysical climate impacts of recent changes in global for-
est cover. Science 351, 600 –604.
Allen, C.D., Macalady, A.K., Chencouni, H., Bachelet, D., McDowell, N., Vennetier, M.,
Kitzberger, T., Rigling, A., Breshears, D.D., Hogg, E.H., Gonzalez, R.F., Zhang, Z.,
Castro, J., Demidova, N., Lim, J.H., Allard, G., Running, S.W., Semerci, A., Cobb, N.,
2010. A global overview of drought and heat-induced tree mortality reveals emerging
climate change risks for forests. For. Ecol. Manag. 2594, 660 –684.
Andreassen, K., Solberg, S., Tveito, O.E., Lystad, S.L., 2006. Regional differences in climatic
responses of Norway spruce ( Picea abies L. Karst) growth in Norway. For. Ecol. Manag.
222, 211 –221.
Begert, M., Schlegel, T., Kirchhofer, W., 2005. Homogeneous temperature and precipita-
tion series of Switzerland from 1864 to 2000. Int. J. Climatol. 25, 65 –80.
Bigler, C., Bräker, O.U., Bugmann, H., Dobbertin, M., Rigling, A., 2006. Drought as an incit-
ing mortality factor in Scots pine stands of the Valais, Switzerland. Ecosystems 9,330 –343.
Biondi, F., Qeadan, F., 2008. A theory-driven approach to tree-ring standardization: de fin-
ing the biological trend from expected basal area increment. Tree-Ring Res 64, 81 –96.
Bogino, S., José, M., Nieto, F., Bravo, F., 2009. Climate effect on radial growth of Pinus
sylvestris at its southern and western distribution limits. Silva Fennica 43, 609 –623.
Fig. 4. The spatial variability of the Romanian Scots pine growth resilience (colour scale) due to the impact of droughts in years with negative pointer years (1987, 1992, 1996, 2000, 2003
and 2012).726 C.G. Sidor et al. / Science of the Total Environment 678 (2019) 721 –727

Bolte, A., Ammer, C., Leof, M., et al., 2009. Adaptive forest management in central Europe:
climate change impacts, strategies and integrative concept. Scand. J. For. Res. 24,
473 –482.
Bonan, G.B., 2008. Forests and climate change: forcings, feedbacks, and the climate bene-
fits of forests. Science 320, 1444 –1449.
Bosela, M., Sedmák, R., Maru šák, R., Sedmáková, D., Petrá š, R., Barna, M., 2014. Evaluating
similarity of radial increments around tree stem circumference of European beech
and Norway spruce from Central Europe. Geochronometria 41, 136 –146.
Bunn, A., Korpela, M., Biondi, F., Campelo, F., M'erian, P., Qeadan, F., Zang, C., 2017. dplR:
Dendrochronology Program Library in R. R Package Version 1.6.6. https://r-forge.r-
project.org/projects/dplr/ .
Camarero, J.J., Gazol, A., Sanguesa-Barreda, G., Oliva, J., Vicente-Serrano, S.M., 2015. To die
or not to die: early-warning signals of dieback in response to a severe drought. J. Ecol.103, 44 –57.
Candel-Pérez, D., Linares, J.C., Viñegla, B., Lucas-Borja, M.E., 2012. Assessing climate –
growth relationships under contrasting stands of co-occurring Iberian pines along
an altitudinal gradient. For. Ecol. Manag. 274, 48 –57.
Carnicer, J., Coll, M., Ninyerola, M., Pons, X., Sánchez, G., Peñuelas, J., 2011. Widespread
crown condition decline, food web disruption, and ampli fied tree mortality with in-
creased climate change-type drought. Proceedings of the National Academy of Sci-
ences of the USA 108, 1474 –1478.
Carrer, M., Urbinati, C., 2006. Long-term change in the sensitivity of tree-ring growth to
climate forcing of Larix decidua. New Phytol. 170, 861 –872.
Cedro, A., 2006. Comparative dendroclimatological studies of the impact of temperature
and rainfall on Pinus nigra Arnold and Pinus sylvestris in Northwestern Poland. Baltic
For 12, 110 –116.
Chira, D., D ănescu, F., 2013. Uscarea anormal ăa arboretelor de pin, fag, stejar, frasin ș.a –
factori cauzali și metode de gestionare. Scienti fic Report. ICAS, Suceava, Romania
manuscript.
Chiri ță, C., 1981. Pădurile Romaniei 573p.
Ciobanu, D., 2003. Selec ția unor popula ții naturale valoroase de pin silvestru din Carpa ții
Orientali șiC a r p a ții de Curbur ă, apte pentru cultur ăîn sta
țiuni corespunz ătoare din
zonă. Editura Infomarket, Bra șov (133 p).
Cook, E.R., 1987. On the decomposition of tree-ring series for environmental studies. Tree-
Ring Bull. 47, 37 –59.
Cook, E.R., Kairiukstis, L.A. (Eds.), 1990. Methods of dendrochronology. Applications in the
environmental sciences. Kluwer Academic Publishers. Dordrecht. 394 p. Cuny, H. E.,Rathgeber, C. B. K., Frank, D. C., Fonti, P., Fournier, M., 2014. Kinetics of tracheid devel-
opment explain conifer tree-ring structure. New Phytol. 203, 1231 –1241.
Costandache, C., Nistor, S., Ivan, V., 2006. Împădurirea terenurilor degradate ine ficiente
pentru agricultura din sud-estul țării. Analele ICAS 49, 187 –204.
Cropper, J.P., 1979. Tree-ring skeleton plotting by computer. Tree Ring Bulletin, nr. 39, pp.
47–54.
Cuny, H.E., Rathgeber, C.B.K., Frank, D.C., Fonti, P., Fournier, M., 2014. Kinetics of tracheid
development explain conifer tree-ring structure. New Phytol. 203, 1231 –1241.
Cybis Elektronik, 2016. CDendro and CooRecorder. https://www.cybis.se/forfun/dendro/
index.htm .
Dobbertin, M., Wermelinger, B., Bigler, C., Bürgi, M., Carron, M., Forster, B., Gimmi, U.,
Rigling, A., 2007. Linking increasing drought stress to Scots pine mortality and bark
beetle infestations. Sci. World J. 7, 231 –239.
Eilmann, B., Zweifel, R., Buchmann, N., Graf Pannatier, E., Rigling, A., 2011. Drought alters
timing, quantity, and quality of wood formation in Scots pine. J. Exp. Bot. 62 (8),
2763 –2771.
Fernández-de-Uña, L., Fonti, P., Aranda, I., Cañellas, I., Gea-Izquierdo, G., 2013. Impact of
drought on the intra-annual growth of Pinus sylvestris L. The Open Forest Science
Journal 6, 43 –45.
Fritts, H.C., 1976. Tree Rings and Climate. Academic Press, London.
Galiano, L., Martínez-Vilalta, J., Lloret, F., 2010. Drought-induced multifactor decline of
Scots pine in the Pyrenees and potential vegetation change by the expansion of co-occurring oak species. Ecosystems 13, 978 –991.
Gazol, A., Camarero, J.J., Sangüesa-Barreda, G., Vicente-Serrano, S.M., 2018. Post-drought
resilience after forest die-off: shifts in regeneration, composition, growth and produc-
tivity. Front. Plant Sci. 9, 1546.
Gonzalez, I., G., 2001. Weiser: a computer program to identify event pointer years in den-
drochronological series. Dendrochronologia 19: 239 –244.
González-de-Andrés, E., Camarero, J.J., Blanco, J.A., Imbert, J.B., Lo, Y.-H., Sangüesa-
Barreda, G., Castillo, F.J., 2017. Tree-to-tree competition in mixed European beech –
Scots pine forests has different impacts on growth and water-use ef ficiency depend-
ing on site conditions. J. Ecol. 106, 59 –75.
Gruber, A., Strobl, S., Veit, B., Oberhuber, W., 2010. Impact of drought on the temporal dy-
namics of wood formation in Pinus sylvestris. Tree Physiol. 30, 490 –501.
Hanewinkel, M., Cullmann, D.A., Schelhaas, M.J., Nabuurs, G.J., Zimmemann, N.E., 2013.
Climate change may cause severe loss in the economic value of European forest
land. Nat. Clim. Ch. 3, 203 –207.
Haylock, M.R., Hofstra, N., Klein Tank, A.M.G., K lok, E.J., Jones, P.D., New, M., 2008. A European
daily high-resolution gridded dataset of surface temperature and precipitation.
J. Geophys. Res (Atmospheres) 113, D20119. https://doi.org/10.1029/2008JD10201 .
Herrero, A., Rigling, A., Zamora, R., 2013. Varying climate sensitivity at the dry distribution
edge of Pinus sylvestris and P. nigra. For. Ecol. Manag. 308, 50 –61.
Holmes, R.L., 1983. Computer-assisted quality control in tree-ring dating and measure-
ment. Tree Ring Bull 43, 69 –75.
Hölttä, T., Juurola, E., Lindfors, L., Porcar-Castell, A., 2012. Cavitation induced by a surfac-
tant leads to a transient release of water stress and subsequent ‘run away ’embolism
in Scots pine ( Pinus sylvestris ) seedlings. J. Exp. Bot. 63, 1057 –1067.
Hordo, M., Metsland, S., Kiviste, A., 2009. Response of Scots pine ( Pinus sylvestris L.) radial
growth to climate factors in Estonia. Baltic For 15, 195 –205.Houston Durrant, T., de Rigo, D., Caudullo, G., 2016. Pinus sylvestris in Europe: Distribution,
habitat, usage and threats. In: San-Miguel-Ayanz, J., de Rigo, D., Caudullo, G., Houston
Durrant, T., Mauri, A. (Eds.), European Atlas of Forest Tree Species. Publ. Off. EU,
Luxembourg, pp. 132 –133.
IPCC, 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working
Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge, United Kingdom and New York,
NY, USA (1535 p).
Irvine, J., Perks, M.P., Magnani, F., Grace, J., 1998. T h er e s p o n s eo f Pinus sylvestris to
drought: stomatal control of transpiration and hydraulic conductance. Tree Physiol.
18, 393 –402.
Lebourgeois, F., 2000. Climatic signals in earlywood, latewood and total ring width of Cor-
sican pine from western France. Ann. For. Sci. 57, 155 –164.
Liang, W., Heinrich, I., Simard, S., Helle, G., Liñán, I.D., Heinken, T., 2013. Climate signals
derived from cell anatomy of Scots pine in NE Germany. Tree Physiol. 33, 833 –844.
Lloret, F., Keeling, E.G., Sala, A., 2011. Components of tree resilience: effects of successive
low-growth episodes in old ponderosa pine forests. Oikos 120, 1909 –1920.
Martinez-Vilalta, J., Pinol, J., 2002. Drought-induced mortality and hydraulic architecture
in pine populations of the NE Iberian Peninsula. For. Ecol. Manag. 161, 247 –256.
McDowell, N.G., Pockman, W.T., Allen, C.D., Breshears, D.D., Cobb, N., Kolb, T., Plaut, J.,
Sperry, J., West, A., Williams, D.G., Yepez, E.A., 2008. Mechanisms of plant survival
and mortality during drought: why do some plants survive while others succumb
to drought? New Phytol. 178, 719 –739.
McKenney, D.W., Pedlar, J.H., Lawrence, K., Campbell, K., Hutchinson, M.F., 2007. Potential
impacts of climate change on the distribution of North American trees. Bioscience 57,939 –948.
Misi, D., Nafradi, K., 2016. Possibility of identi fication of negative extreme climatic events
using Pinus sylvestris tree-rings in Transdanubia, Hungary. Dendrobiology 75, 45 –54.
Monserud, R.A., Sterba, H., 1996. A basal area increment model for individual trees grow-
ing in even- and unevenaged forest stands in Austria. For. Ecol. Manag. 80, 57 –80.
Motta, R., Nola, P., 2001.
Growth trends and dynamics in sub-alpine forest stands in the
Varaita Valley (Piedmont, Italy) and their relationships with human activities and
global change. J. Veg. Sci. 12, 219 –230.
Oberhuber, W., Swidrak, I., Pirkebner, D., Gruber, A., 2011. Temporal dynamics of non-
structural carbohydrates and xylem growth in Pinus sylvestris exposed to drought.
Can. J. For. Res. 41, 1590 –1597.
Panayotov, M., Za firov, N., Cherubini, P., 2013. Fingerprints of extreme climate events in
Pinus sylvestris tree rings from Bulgaria. Trees Struct. Funct. 27, 211 –227.
Pichler, P., Oberhuber, W., 2007. Radial growth response of coniferous forest trees in an
inner Alpine environment to heat-wave in 2003. For. Ecol. Manag. 242, 688 –699.
Popescu, A.G., 1984. Cercet ări de provenien țe la pin silvestru. Institutul de Cercet ăriși
Amenaj ări Silvice, Bucure ști, Romania (67 p).
QGIS Development Team, 2009. QGIS Geographic Information System. Open Source
Geospatial Foundation URL. http://qgis.osgeo.org .
Rajczak, J., Pall, P., Schär, C., 2013. Projections of extreme precipitation events in regional cli-
mate simulations for Europe and the Alpi ne Region. J. Geophys. Res. – Atmos. 118, 1 –17.
Rigling, A., Bigler, C., Eilmann, B., et al., 2013. Driving factors of a vegetation shift from
Scots pine to pubescent oak in dry Alpine forests. Glob. Chang. Biol. 19, 229 –240.
Romanian National Meteorological Administration, 2008. Clima României. Editura
Academiei Române (365 p).
Sánchez-Salguero, R., Navarro-Cerrillo, R.M. , Camarero, J.J., FernandezCancio, A., 2012. Selective
drought-induced decline of pine species i n southeastern Spain. Clim. Ch. 113, 767 –785.
Sánchez-Salguero, R., Camarero, J.J., Hevia, A., Madrigal-González, J., Linares, J.C.,
Ballesteros-Canovas, J.A., et al., 2015. What drives growth of scots pine in continental
mediterranean climates: drought, low temperatures or both? Agric. For. Meteorol.
206, 151 –162.
Sarris, D., Christodoulakis, D., Körner, C., 2011. Impact of recent climatic change on growth
of low elevation eastern Mediterranean forest trees. Clim. Ch. 106, 203 –223.
Sidor, C.G., Bosela, M., Büntgen, U., Vlad, R., 2018. Mixed effects of climate variation on the
Scots pine forests: age and species mixture matter. Dendrochronologia 52, 48 –56.
Smith, M.D., 2011. An ecological perspective on extreme climatic events: a synthetic def-
inition and framework to guide future research. J. Ecol. 99, 656 –663.
Speer, J., 2010. The Fundamentals of Tree-Ring Research. University of Arizona Press (333 p).
Stănescu, V., 1979. Dendrologie. Editura Didactic ăși Pedagogic ă,B u c u r e ști.
Swidrak, I., Gruber, A., Oberhuber, W., 2014. Xylem and phloem phenology in co-
occurring conifers exposed to drought. Trees Struct. Funct. 28, 1161 –1171.
Temperli, C., Bugmann, H., Elkin, C., 2012. Adaptive management for competing forest
goods and services under climate change. Ecol. Appl. 22, 2065 –2077.
Thabeet, A., Vennetier, M., Gadbin-Henry, C., Denelle, N., Roux, M., Caraglio, Y., Vila, B.,
2009. Response of Pinus sylvestris L. to recent climatic events in the French Mediter-
ranean region. Trees Struct. Funct. 23, 843 –853.
van der Maaten-Theunissen, M., van der Maaten, E., Bouriaud, O., 2015. PointRes: an R
package to analyze pointer years and components of resilience. Dendrochronologia
35, 34 –38.
Wilczy ński, S., 1999. Dendroklimatologia Sosny Zwyczajnej ( Pinus sylvestris L.) z
Wybranych Stanowisk W Polsce (Dendroclimatology of Scots Pine ( Pinus sylvestris
L.) from Range Sites in Poland). Dissertation. Katedra Klimatologii Le śnej, Akademia
Rolnicza w Krakowie, Poland (84 p. (in Polish).
Williams, A.P., Allen, C.D., Macalady, A.K., Grif fin, D., Woodhouse, C.A., Meko, D.M., et al.,
2013. Temperature as a potent driver of regional forest drought stress and tree mor-
t a l i t y .N a t .C l i m .C h .3 ,2 9 2 –297.
XLSTAT, 2017. Data Analysis and Statistical Solution for Microsoft Excel. Addinsoft, Paris,
France (2017).
Zang, C., Hartl-Meier, C., Dittmar, C., Rothe, A., Menzel, A., 2014. Patterns of drought toler-
ance in major European temperate forest trees: climatic drivers and levels of variabil-
ity. Glob. Chang. Biol. 20, 3767 –3779.727 C.G. Sidor et al. / Science of the Total Environment 678 (2019) 721 –727