ArticlePDF Available

Disturbance history is a key driver of tree lifespan in temperate primary forests


Abstract and Figures

Aims We examined differences in lifespan among the dominant tree species (spruce (Picea abies (L.) H. Karst.), fir (Abies alba Mill.), beech (Fagus sylvatica L.), and maple (Acer pseudoplatanus L.)) across primary mountain forests of Europe. We ask how disturbance history, lifetime growth patterns, and environmental factors influence lifespan. Locations Balkan mountains, Carpathian mountains, Dinaric mountains. Methods Annual ring widths from 20,600 cores from primary forests were used to estimate tree life spans, growth trends, and disturbance history metrics. Mixed models were used to examine species-specific differences in lifespan (i.e. defined as species-specific 90th percentiles of age distributions), and how metrics of radial growth, disturbance parameters, and selected environmental factors influence lifespan. Results While only a few beech trees surpassed 500 years, individuals of all four species were older than 400 years. There were significant differences in lifespan among the four species (beech > fir > spruce > maple), indicating life history differentiation in lifespan. Trees were less likely to reach old age in areas affected by more severe disturbance events, whereas individuals that experienced periods of slow growth and multiple episodes of suppression and release were more likely to reach old age. Aside from a weak but significant negative effect of vegetation season temperature on fir and maple lifespan, no other environmental factors included in the analysis influenced lifespan. Conclusions Our results indicate species-specific biological differences in lifespan, which may play a role in facilitating tree species coexistence in mixed temperate forests. Finally, natural disturbances regimes were a key driver of lifespan, which could have implications for forest dynamics if regimes shift under global change.
Content may be subject to copyright.
J Veg Sci. 2021;32:e13069.    
1 of 12
Journal of Vegetation Science
Received:5Januar y2021 
DOI : 10.1111/j vs.130 69
Disturbance history is a key driver of tree life span in
temperate primary forests
Jakob Pavlin1| Thomas A. Nagel1,2 | Marek Svitok3,4| Joseph L. Pettit1|
KrešimirBegović1| Stjepan Mikac5| Abdulla Dikku6| Elvin Toromani7|
Momchil Panayotov8| Tzvetan Zlatanov9| Ovidiu Haruta10| Sorin Dorog10|
Oleh Chaskovskyy11| Martin Mikoláš1| Pavel Janda1|MichalFrankovič1|
Ruffy Rodrigo1|OndřejVostarek1| Michal Synek1| Martin Dušátko1|
TomášKníř1| Daniel Kozák1| Ondrej Kameniar1|RadekBače1|VojtěchČada1|
VolodymyrTrotsiuk1,12,13| Jonathan S. Schurman1| Mélanie Saulnier1,14|
Arne Buechling1| Miroslav Svoboda1
1Depar tment of Forest Ecology, Faculty of Forestr y and Wood S cience s, Czech University of Life Sciences Pra gue, Pr ague, C zech Republic
2Depar tment of Forest ry and Renewab le Fores t Resources, Biotechni cal Faculty, University of Ljubljana, Ljubljana , Slovenia
3Depar tment of Biolog y and Ge neral Ecology, Faculty of Ecology an d Environmental Sciences, Technical Universit y in Zvolen , Zvolen, Slovakia
4Depar tment of Ecosystem Biolo gy, Facult y of Scien ce, Universit y of South Bohemia, Ceske B udejovice, Czech Republic
5Depar tment of Forest Ecology a nd Silviculture, Facult y of Forestry, Universit y of Zagreb, Zagre b, Croatia
8Depar tment of Dendro logy, Universit y of Fores try S ofia, Sofia, Bu lgaria
10Fores try and Fores t Engine ering D epar tment , Univer sity of O radea, Oradea, Romania
11Institute of Forest Management, Ukrainian National Forestry University, Lviv, Ukraine
12Swiss Federal Ins titute for Forest, Snow and Landscape Resea rch WSL , Birme nsdor f, Switze rland
13Depar tmentofEnvironmentalSystemsScience,InstituteofAgriculturalSciences,ETHZurich,Zurich,Switzerland
14CentreNationaldelaRechercheScientifique/FrenchNationalCentreforScientificResearch,UMR5602LaboratoireGéode,Aix-MarseilleUniversit y,Aix-en-
Provence, France
Jakob Pavlin, Departm ent of Forest
Ecology, Faculty of Forestr y and Wood
Science s, Czec h Univer sity of L ife
Prague, Czech Republic.
Email: pavlinj@fld.c
Funding information
This proj ect was suppor ted by th e
institutionalprojec t“EVA4.0”,No.
CZ.02.1.01/0.0/0.0/16 _019/0000803,
MSMT project LTT20016, Internal
Programme Integrated Infrastructure
Aims: Weexamineddifferencesinlifespanamongthedominanttreespecies(spruce,
Picea abies; fir, Abies alba; beech, Fagus sylvatica; and maple, Acer pseudoplatanus)
across primary mountain forests of Europe. We asked how disturbance history, life-
time growth patterns, and environmental factors influence life span.
Locations: Balkan Mountains, Carpathian Mountains, Dinaric Mountains.
Methods: Annual ring widths from 20,600 cores from primary forests were used
to estimate t ree life spans, growt h trends, and dist urbance histor y metrics. Mixed
models wereused to examine species-specific differences inlife span (i.e.,defined
asspecies-specific 90th percentiles ofagedistributions),and how metrics ofradial
2 of 12 
Journal of Vegetation Science
Tree life span plays an important role in the structure and function
spanofcompetingspeciesmaycontributetotree coexistence in a
varietyof forestcommunities(Veblen,1986;Lertzman,1995;Lusk
&Smith,1998;Loehle,2000).Treelifespanalsohas importantim-
plications for carbon sequestration and residence time, particularly
within th e context of glob al change, whe reby faster grow th rates
term car bon storage in fore sts (Bugmann & B igler, 2011; Körner,
provide im portant hab itat for a variet y of forest-dwelling spe cies
Despite the fundamental importance of tree life span, determining
whether there are biological differences in the life span of competing
treespeciesin manyforestcommunitiesremainsachallenge.Forexam-
ple, among the common tree flora of the European temperate region, the
widelyacceptedliteratureindicatesthatbeech(Fagus sylvaticaL.)andfir
(Abies alba Mill.) are among the most long-livedspecies (e.g., 450 year
maximum age), while maple (Acer pseudoplatanus L.) and spruce (Picea
(Leuschner & Ellenberg, 2017; Leuschner & Meier, 2018). However,
a number of site-specific studies carried out in old-growth remnants
throughout temperate mountain forests in Europe often find individual
Motta et al., 2011; Nagel et al., 2014; Di Filippo et al., 2015; Di Filippo
Inaddition to species-specificlife history,elucidating the driv-
ers of life span within and among tree species has also proved chal-
lenging.A relativelylargebody of treering research has examined
the longstanding hypothesized tradeoff between tree growth rates
andlife span(Schulman,1954).Across multiple treetaxaspanning
a range of shade tolerances, past work has generally documented
a consistent negative relationship between radial growth rates and
lifespanwithin speciesforbothlive(Blacket al.,2008;Johnson&
Abrams , 2009; Di Filip po et al., 2012; Cas tagneri et al ., 2013; Di
Filippo et al.,2015; Brienen et al.,2020) and dead trees (Bigler &
In particular, many studies highlight that m aximum tree ages are
during early life stages. Other studies, however, have not found sup-
Aspreviousauthorshave pointedout, it is not entirelyclearto
whatextent the relationshipbetween growthrates and life spanis
above show substantial variation in growth rates and life span
among individual trees within a species at the same site, which likely
reflects the unique grow th history and resource availability of trees
ina given neighborhood ofcompetitors (Black etal., 2008; Bigler,
turbance histor y, which controls the trajectory of canopy accession.
Co-ordinating Editor:KerryWoods
growth, disturbance parameters, and selected environmental factors influence life
Results: While only a few beech trees surpassed 50 0 years, individuals of all four spe-
cies were older than 400 years. There were significant differences in life span among
thefourspecies(beech> fir > spruce >maple),indicatinglifehistorydifferentiation
in life span. Trees were less likely to reach old age in areas affected by more severe
from a weak but significant negative effect of vegetation season temperature on fir
and maple life span, no other environmental factors included in the analysis influ-
enced life span.
Conclusions: Ourresultsindicatespecies-specificbiologicaldifferencesinlifespan,
ests. Finally, natural disturbance regimes were a key driver of life span, which could
have implications for forest dynamics if regimes shift under global change.
Disturbance, European beech, grow th patterns, life span, longevity, Norway spruce, silver fir,
site conditions, sycamore maple
PAVLIN et AL.    
 3 of 12
Journal of Vegetation Science
periodic intermediate severity disturbances drive stand dynamics
(Frelich & Lorimer, 1991;Nagel etal., 2014). In such forests, trees
that avoid suppression, grow fast as juveniles and reach large canopy
stature quickly, may be more likely to die early because residing in
the canopy increases exposureto common disturbance agents (Di
trees will reach old ages on less productive sites sheltered from dis-
turbance(Larson,2001;Lanner,2002;DiFilippoetal., 2015), few
studies have quantified the relationship between local disturbance
In addition to local constraints on growth regulated by stand dy-
fluencing life span within a species, such as climate and topography.
Life span h as been positivel y associated with n orth-facing slo pes
(Bigler &Veblen, 2009; Bigler,2016),south-facing slopes on drier
sites (T herrell & Stahl e, 1998)s lope steepne ss (Bigler, 2016), and
elevation(Splechtnaetal., 2000;DiFilippoetal.,2007;Di Filippo
et al., 2012; Röt heli et al., 2012; D i Filippo et al. , 2015), whereby
harsher site conditions are thought to increase longevity via reduced
growth rates. The link between temperature and life span is less
clear; forbroad-leavedspeciesin the NorthernHemisphere, some
found to reach older ages in colder part s of their range, while this
Previous work focusing on tree life span has of ten relied on
Johnson & Abrams, 2009; Di Filippo et al., 2015; Brienen et al.,
2020), inwhich the samplingobjectives and strategy are not pre-
served in the metadata; in many cases, tree cores may have been
ditions, or tree growth may have been influenced by past land use
histor y. Other st udies that have explicitly sampled live and dead
trees to ex amine life span w ithin specie s have often bee n limited
& Frelich, 1989), and do not permit assessment of environmen-
tal drivers at large scales or variation in life span across a species’
sisting ofplot-levelforest structuraldataincludingrandomlycored
livetrees(N =20,600),sampledwithinextantprimary-forestland-
scapes across the Carpathian Mountains and Balkan peninsula. The
plot network covers the dominant mountain forest communities in
forest t ypes. The data set allows a unique assessment of life span
within and among tree species, across gradients of stand structure
and disturbance history, and from local to subcontinental scales.
We ask if there are differences in life span among dominant
tree spe cies (i.e., spr uce, fir, beech, a nd maple). We first f ind evi-
dence that species-specific differencesinadult lifespan are partly
controlled by life history. We then ask how variation in life span
within species is influenced by a variety of different drivers, includ-
ing local disturbance history, lifetime radial growth patterns, and
environm ental fac tors availabl e in our data set ( i.e., slope, as pect,
predict that older trees will be found on less productive sites, char-
acterizedby higherelevation,lowertemperature, and steepnorth-
span, we predict that old trees will be found in areas with a history
oflow-intensitydisturbance (i.e.,gap dynamics)thatlack evidence
of more severe disturbance over the past few centuries. Individual
tree growth histories in such stands typically show long periods of
slow juvenile growth under the shaded understorey, followed by one
or several growth releases during canopy accession caused by the
weexamine howcommon exceptionallyold treesare withinthese
old-growthforestsites, whichhasimplications forconservation of
forest biodiversity.
2.1  |  Studyareaandsiteselection
This study was conducted in primar y temperate mount ain forests
of the Carpathian Mountains and the Balkan peninsula, spanning
frombeech-dominatedand mixedforests(hereafter referred to as
at higher elevations. These t wo regions contain the largest remnants
of primar y forests i n the temperate zon e of Europe (Janda e t al.,
2019;Mikolášetal., 2019;Nageletal.,2014;Sabatiniet al.,2018).
Primar y forests were characterized as unmanaged forests with natu-
ral stand composition, diverse horizontal, vertical, and age structure,
and a significant amount and diversity of downed and standing dead
trees in different stages of decomposition; most stands were typi-
cally inanold-growthstage of development,butearlyseralstages
developing after more severe natural disturbances were also present
tractsofprimary-forestlandscapes inEurope andlong-term study
of their dynamics. The plot network has a hierarchical sampling
scheme, with plots located within stands, and multiple stands organ-
ized in larger landscapes. For the purposes of this study, we split
the data set into 11 landscapes, based on geographic location and
foresttype. Theyincludesevenbeech-dominatedlandscapes (i.e.,
Albania,Bulgaria,Croatia, CentralSlovakiabeech, EasternSlovakia
four spruce-dominated landscapes (i.e., Central Slovakia spruce,
Ukraine spruce, Northern Romania spruce, and Southern Romania
spruce). Th e landscape s comprise a tota l of 35 spruce-domin ated
stand s, and was occasion ally mixed with fir, Pinus cembra L., and
Sorbus aucupariaL.Inbeech-dominate dstands,be echaccounte dfor
75%ofalltrees onaverage,followedbyfir(14%),spruce(7%),and
maple (2%);other lesscommonspecieswere sporadically present,
4 of 12 
Journal of Vegetation Science
such as Acer platanoides L., Acer obtusatumWaldst.&Kit.exWilld.,
Ulmus glabraHuds.,Fraxinus excelsior L., and Fraxinus ornus L.
2.2  |  Datacollection
Between 2010 and 2018, a net work of circular permanent sampling
plots wa s establis hed within ea ch primar y-fores t stand (A ppendix
S2). A strati fied rando m sampling de sign was used for p lot place-
ment, whereby plot s were randomly located within a regular grid
laid over each stand, with grid size depending on the size of stands.
In spruce l andscape s, a grid with eit her a 1- or 2-ha cell size wa s
used, and a circular plot of 1,000 m2(oroccasionally500m
2 in
recently disturbed areas where tree densities were high) was ran-
1,500 or 2,000 m2.Arandompointwasgeneratedinthe0.5–3.4ha
interior of each grid cell, and a pair of plots was established 4 0 m
from either side of this point along the slope contour, such that plot
centers were separated by 80 m.
jectionarea (neededfor the disturbance reconstruction described
spans, lifetime growth patterns, and disturbance history. We took an
incrementcorefrom 15 (whensmaller plots were used)to30ran-
cored all theliving trees≥20cmDBH and25%randomlyselected
33.9 cores/plot). Allcores were taken 1 m above the ground, and
perpendicular to the terrain slope direction to avoid reaction wood.
2.3  |  Dataanalysis
In total, 20,60 0 increment cores were processed using standard den-
drochronological procedures. Each tree was represented by a single
core. Ring-width series were measured with a LintabTM sliding-
stage measuring system (Rinntech, Heidelberg, Germany; http://
www.rinntech.ds), cross-dated using marker years (Yamaguchi,
1991), and verifie d with COFECHA (Holmes , 1983) and CDe ndro
berof missingringswasextrapolated from the curvature andaver-
agegrowthrates ofinnermost rings (Duncan, 1989). We excluded
all cores that had more than 20 estimated rings missing, or were of
poor quality and therefore did not allow tree rings to be measured
andcross-datedreliably.Intot al,3 ,50 0coreswe reomit tedfromfur-
FIGURE1 Hierarchicaldesignand
spatial distribution of study plots. The
map shows the distribution of spruce
numbers mark the forest landscapes: 1,
Romania beech; 5, North Romania beech;
6, East Slovakia; 7, Central Slovakia beech;
8, Central Slovakia spruce; 9, North
Romania spruce; 10, South Romania
spruce; and 11, Ukraine
PAVLIN et AL.    
 5 of 12
Journal of Vegetation Science
which resulted in a decrease in the overall mean diameter of cored
trees for eachspecies(from 31.4 cm to 28.0cmforfir,37.1cmto
34.4 cm for maple, 36.4 mm to 31.6 cm for beech, and 36.6 cm to
Radial growth pat terns of increment cores were analyzed for
evidence of past disturbance events within each plot. Quantitative
reconstructions of disturbance histories for different regions of the
et al., 2017; Meigs et al., 2017; Nagel et al., 2017; Schurman et al.,
2018;Janda etal., 2019; Schurman, Babst,Björklund, etal., 2019;
Čada et al.,2020; Frankovičetal., 2020),andprovide detailedde-
scriptions of dendroecological methods. We therefore only briefly
summarize the methods used to reconstruct disturbance below. We
used theoriginalapproach of Lorimer and Frelich(1989),in which
each core is screened for (a)abrupt , sustained increases in radial
growth (i.e., releases) and (b) rapid early growth rates (i.e., gap-
ity of a former canopy tree. Following Lorimer and Frelich (1989),
we only included release events before trees reached a diameter
threshold, such that only mortality events that provided access to
the canopy were counted. The diameter threshold was based on
comparisons of diameters of currently suppressed vs released trees
in the plot data. Multiple releases were allowed as long as they oc-
curred before the diameter threshold. The severity of disturbance
was based on the relative canopy area removed on each plot cal-
culated from the current crown area of trees containing evidence
of past disturbance; this approach makes the assumption that the
sum of the current crown areas of such trees is representative of the
prop ortionof thepl otdisturbedinth epast(Lori mer&Frelich,198 9).
Disturbance history reconstruction and lifetime growth patterns
of trees were used to derive several variables that may influence lon-
opyareakilled) ofthereconstructedmaximum severitydisturbance
eve ntonea chplot(Meigsetal.,2017 )( hereafterreferredtoasdistu r-
events per core. To assess the influence of growth rates on longevity,
we used the average growth rate of the first 50 years following Bigler
(2016)(hereafter referredtoas earlygrowth).Assuch, weexcluded
all cores that had less than 50 rings measured or estimated. We also
included metrics of the minimum and maximum 10-year average
(1994)(hereafter referred to as minimumand maximum growth),as
well as the number of releases detected throughout the series.
Several different environmental variables were also compiled to
test their influence of longevity. We used raw values of slope steep-
ness for each plot, while values of slope aspect were transformed
into northness following the formula: northness =cosine[(as-
pect in degrees * π)/180] (J anda et al., 2019). To avoid problems
correlatedwithtemperature(r =−0.752).Meantemperatureofthe
vegetation season for each plot was calculated for the period from
the 1 May until 31 October, by downscaling the Worldclim gridded
by building a linear model of temperature vs the product of altitude,
longitude, andlatitude. Given that we use the 1970–2000 period,
during early life stages of old trees, our temperature variable is more
thereare severalbroad-scaledifferences between the Balkanand
Carpathian study sites that may influence tree growth and longevity,
including higher annual precipitation and temperature in the Balkan
region, as w ell as differenc es in bedrock (Kozák et a l., 2018); we
plore if there are dif ferences in life span across the region.
To estimate life span within a species, we simply use the 90th
percentile of age distributions for each species from pooled data
across th e entire study (N agel et al., 2014). To compare life span
model(GLMM) with the agesoftrees≥thespecies-specific90th
levelrandom-effects structuretoaccount for the potential effects
of geograp hical variab ility: a plot n ested within a p air of plots (in
a stand nested within a landscape. Wald tests were per formed to
assess statistical significance. We used Tukey pairwise comparisons
to test for differences among individual species.
To identify the most influential drivers of life span, a binomial
GLMM was fit for each species. To facilitate interpretation, we con-
verted the 90th percentile ages to a binary variable, such that the
agestatus of a given treeage waseither≥ species-specific 90th
percentile or <species-specific 90th percentile. Age status was
used as a response variable and disturbance history, growth rate,
asanadditionalfixed factor thatmaycause potential differences
in longevit y of spruce growing in these two forest types. The mean
values and ranges of all the variables included in these models are
listedper speciesinAppendixS3.Inthemaple model,some vari-
ables wer e strongly corre lated (i.e., earl y growth with ma ximum
and minimum growth, latitude with the average temperature of the
cluding disturbance year, early growth, number of releases, and lat-
itude from the model. None of the final models showed problems
<3.00)( Zuuretal.,20 09).Weusedthesa merandom-ef fec tsstr uc-
ture as described above. We calculated marginal determination co-
All the analyses were performed inR language version 3.5.1(R
Core Team, 2019) using the following libraries: glmmTMB (Brooks
of-variance calculation, emmeans (R Core Team, R Foundation for
6 of 12 
Journal of Vegetation Science
Statistical Computing , Vienna, Austria) to perform Tukey pairwise
comparisons, DHARMa(R Core Team, R Foundation for Statistical
Computing,Vienna,Austria)toperformresidual diagnostics, perfor-
Austria)to calculate VIFvalues,MuMIn(RCore Team,RFoundation
tion coefficients, and ggplot2(Wickham,2016)forplotting.
3.1  |  Interspecificdifferencesinlifespan
silver fir (218–456 years), spruce (218–449years), to maple (192–
412years)(Figure 2; AppendixS4).The GLMMshowed significant
differences in mean age of the oldest individuals among the four tree
species(χ2 = 280.41, p <0.001).Theoldestbeechtreesweresig-
nificantly older than all the other three species. The oldest fir trees
were significantly older than spruce trees, while the differences be-
tween the oldest maple and spruce trees, and maple and fir trees,
3.2  |  Driversoftreelifespan
The influence of disturbance-related variables on reaching the
90th percentile ages varied among the four tree species (Table 1;
oftrees above the 90th percentile age thresholds) for spruce and
maple,while thecalendaryearofthismaximumseverityeventhad
more recent maximum severityevents had less trees reaching old
age(AppendicesS6, S7). Neither of the two disturbance variables
were significant in the fir model. The number of releases had a sig-
nificant positive effect on life span in all models.
Growth rate variables had strong and consistent effects
on life span a mong the specie s (Table 1; Appendixe s S8–S10).
Minimum ten-year average growth showed a significant neg-
ative influence for each species, while early growth rate had a
significant negative influence on life span for beech and fir, but
not spruce. Maximum growthwas not significantin anyof the
The models did not show strong evidence of environmental
controlontreelifespans. Boththeplotlevel(northnessandslope)
FIGURE2 Agedistributionofthe
represent the interquartile range, with
the median age of the oldest trees given
notches showing 95% confidence inter vals
of the median. The lower whisker of each
percentile of age, while upper whiskers
and points show outliers. Numbers in the
upper right corners indicate the number
percentile, including the tot al numbers
of all trees per species in parentheses.
The letters in the upper left position of
each plot indicate significant differences
amongspecies(p <0.05)basedonTukey's
pairwise comparisons
PAVLIN et AL.    
 7 of 12
Journal of Vegetation Science
TAB LE 1  Summaryofgeneralizedlinearmixedmodels(GLMMs)examiningtheeffectofenvironmental,growth,anddisturbancefactorsontheageoftheoldestindividualsoffourtree
Fagus sylvatica Picea abies Abies alba Acer pseudoplatanus
Est. SE zpEst. SE zpEst. SE zpEst. SE z p
(Intercept) −3.74 0.20 −19.14 <0,001 4.12 0.36 −11. 4 0 <0,0 01 −3.70 0.56 −6.66 <0,0 01 −22 . 36 6 .98 −3.20 0.001
Disturbance severity −0.11 0.09 −1. 22 0 .2 24 −0.70 0.09 −7. 6 0 <0,001 −0.06 0 .17 −0.33 0.738 −9.19 3.81 −2 .41 0.016
Disturbance year −0.39 0.08 −4.72 <0,001 0.29 0.06 −4.73 <0,001 0.25 0.16 1 .61 0 .108
Early growth −0.63 0.11 −5.73 <0,001 0.07 0.08 0.86 0.389 0.98 0.25 −3.90 <0,001
Maximumgrowth −0.05 0.07 −0.70 0.482 0.0 0 0.08 −0.04 0.967 0.06 0.17 0.36 0 .719 2.87 1.90 1. 51 0.131
Minimum growth −1.13 0 .14 −7.99 <0,001 −1 .9 9 0.10 −20.60 <0,001 −1. 11 0.32 −3.48 0.001 −13 .10 5.15 −2 .55 0.011
Number of release s 1.11 0.06 19.70 <0,001 0.58 0.04 14. 27 <0.001 1.08 0.15 7. 42 <0,001
Latitude −0.34 0.20 −1 .72 0.086 0.06 0.13 −0.45 0.650 0.01 0.45 −0.01 0.990
Northness −0.09 0.10 0.85 0.393 0.00 0.07 0.05 0.961 −0.28 0 .17 −1.6 2 0.106 −2.54 2 .47 −1 . 03 0.304
Slope 0.04 0.11 0.39 0.694 −0.03 0.08 −0.35 0.726 0.18 0.20 0.90 0.370 1.49 1.53 0.97 0.331
Mean T of veget ation
−0.20 0.16 −1 . 2 5 0.211 0 .13 0.10 1.27 0.204 −0.50 0.25 −2 .00 0.045 −6.45 3.17 −2.0 4 0.042
Forest t ype 0.29 0.37 0.79 0.432
τ00plot:(pairplot:(stand:landscape)) 0.57 0.47 0.00 464.87
τ00pairplot:(stand:landscape) 0.37 0. 57 0. 41 0.00
τ00 stand:landscape 0.69 0.34 1. 85 0.00
τ00 landscape 0.00 0.00 0.84 0.01
2[%] 48.03 54.56 42.65 40.93
Rc 2[%] 65.25 68.07 70.49 9 9. 59
Note:Resultsshowexplanatoryvariablesusedinthemodels,estimatesoftheregressioncoef ficients(Est.),standarderrors(SE),z-values(z),probabilities(p),variancesofallfourlevelsofrandomeffects
, and conditional determination coefficients
. The significant model parameters are displayed in b old.
8 of 12 
Journal of Vegetation Science
season) wereinsignificantfor mostspecies-level models. The only
of the vegetation season had a weak but significant negative effect
3.3  |  Commonnessofoldtrees
10–20 old tr ees/ha, and 11 stan ds had densities gr eater than 20
old trees/ha. We identified 893 trees that were at least 300 years
old, and 113 trees older than 400 years, of which two were maple
ofallbeech trees).When calculated as a proportionofthecanopy
layer trees, the oldest trees made up from 0% to 23.9% of the can-
We found compelling evidence of biological differences in life span
amongthe four coexistingtreesspeciesacrossmountainregionsof
Europe, yet these dif ferences are not entirely consistent with pub-
lished literature on the ecology of European tree species. We also
foun dth atloc ald ist urb anc ehistory,r ath erthanpl ot-scal ean db roa d-
scale environmental fac tors, was a key driver of life span. The particu-
lar histor y of local disturbance likely contributes to the highly variable
density of the oldest trees that was documented across the study
region. Before we elaborate on these main findings, we highlight sev-
eral caveats that are impor tant for the discussion that follows.
The estimates of tree life span used in our study, in particular the
centiles, provide an indication of the longevity of a given individual,
but not actual longevity, or the number of years from seed germina-
tion to death of an adult tree. Because we cored trees at 1 m in height ,
we underestimate the number of years required to reach 1 m, which
could range from years to decades depending on light conditions,
particularly for shade-tolerant species (Wong & Lertzman, 2001;
Nageletal.,2006).Wealsoonlywor kedwithlivetrees,mainlydueto
the challenges of obtaining a large sample of recently dead trees with
intact wood for tree coring. The species sampled in our study, espe-
difficult(DiFilippoetal.,2012). Finally,giventhatlargertreesoften
had boledecay, morelargerstems were excluded from the sample
compared to smaller stems, which is an additional reason why our
data set may underestimate life span. Despite these limitations, the
largepopulatio n-baseddatasetofr andomlysa mpledtree si nrem ain-
ing primary forests across a subcontinental scale should provide a
robust estimate of interspecific dif ferences in tree life span.
Beech, t he dominant broa d-leaved s pecies, was mar kedly older
than the other species, and fir was older than spruce, but there were
no significant dif ferences in life span between maple and the two
FIGURE3 Densitiesofoldesttrees(≥species-specific90thpercentileofage)byspecieswithineachstandacrossthestudyregion.The
Slovakia; 7, Central Slovakia beech; 8, Central Slovakia spruce; 9, North Romania spruce; 10, South Romania spruce; 11, Ukraine
PAVLIN et AL.    
 9 of 12
Journal of Vegetation Science
ferences in life history traits at both juvenile and adult life stages may
tainforestsofEurope, beechis bothoneof the mostshade-tolerant
species as a juvenileandlong-livedasanadult, likelycontributingto
its dominance across the region. Compared to juveniles of beech and
fir, maple and spruce are less tolerant of shade, but grow more rap-
mountainforest intheDinaricregion, (Nageletal.,2010)foundthat
maplerequiredrelativelylargegaps(>40 0 m2)toaccessthecanopy,
while smaller gaps were captured by beech and fir, which were often
maple may beadisadvantage inforests that lackmoderate-severity
disturbance for long time periods, its relatively high longevit y likely
contributes to its persistence in the landscape. Indeed, there are likely
other interspecific trait differences associated with various life stages
better understand the life history traits and life stages that contribute
It is impor tant to point out that our findings are not consistent
with the classic literature on life span of European tree species. For
example , Korpel’ (1995) report ed life spans of 230 a nd 350 years
for beech and fir, respectively, which is in complete contrast to our
findings. Leuschner and Ellenberg(2017) report maximumages of
450 years for b oth beech and fir, and 300 year s for spruce and mapl e.
older than the median ages of trees above the 90th percentile ages;
in forest s regulated by natural disturbance processes. We believe
that the range of life span estimates provided here may serve as an
important reference for future studies that must prescribe values for
them, such as simulation models of forest dynamics. Relying on esti-
mates from earlier work may lead to spurious conclusions regarding
long-termforestdynamics.Insupport ofourfindings,otherrecent
dendroecological studies that focused on local sites or other regions
Nagel et al., 2014; Di Filippo et al., 2015; Piovesan, Biondi, Baliva,
In addition to life history differences, we also sought to iden-
tify drivers of longevity across the large gradient of environment al
conditions and disturbance histories present in the data set. The
analyses indicated significant links between disturbance metrics
andlongevity in threeofthe four species-levelmodels.There was
verity event occurred further back in time. These findings are not
atall surprising given that both the beech-and spruce-dominated
acterized byrelatively frequent moderate-severity, partial canopy
disturbances that are likelytoremove susceptible individuals(i.e.,
large, old trees)inthecanopy layer (Nageletal.,2014;Čadaetal.,
Disturbance history is also intrinsically linked to lifetime growth
of trees in that it regulates canopy structure, and thereby the grow th
of understorey trees via changes in light. Depending on the size,
location, and timing of disturbance relative to a given understorey
tree, some individuals will gain access to the forest canopy quickly,
others will reach the canopy after multiple periods of suppression
and release, and yet others will die due to prolonged periods of sup-
pression.Acrossallfourofthespeciesstudiedhere,trees thatsur-
vived periods of very suppressed grow th were more likely to reach
oldage.Forexample,averageannual growthratesduringthemini-
mum10-yeargrowthperiodsforallthetreesolderthan 300years
was <0.62 mm/year, <0.36 mm/year for all the trees older than
400 years, and <0.17 mm/year for all the trees older than 500 years
rates were a lso more likely to r each old age (A ppendix S1: Figur e
(Piovesan etal.,2005;DiFilippoetal.,2012; Piovesanetal.,2019)
andspruce (Bigler&Veblen, 2009;Rötheli et al.,2012;Castagneri
et al., 2013). Finally, the results show a positive relationship be-
tween t he number of sup pression-rel ease period s experien ced by
atree and longevity (Appendix S1: Figure S7).Treeswith multiple
suppres sion-r elease perio ds persist in t he understo rey for a large
propor tion of their life span, while trees that access the canopy
quickly, such as those that establish in large gaps, spend a longer
propor tion of their life in the canopy. These results lend support to
the idea that time spent in the canopy, where there is higher risk of
disturbance, may have a stronger influence on longevity than other
factors, such as environmental constraints or genetics.
Indeed, one of the more surprising findings was that we did not
identif y a consistent relationship bet ween life span and environmen-
tal factors, such as temperature and elevation, which has been docu-
DiFilippoetal.,2015;Bigler,2016).A sidefromt henegativecorrela-
tion between the mean temperature of the vegetation season and
longevit y for fir and maple, none of the other environmental factors
four species. The strong influence of disturbance and slow growth
rates on longevity documented here may simply override the influ-
ence of other environmental factors. In a study on mountain pine in
temperature effect on the early growth of mountain pines was likely
due to superimposing effec ts of st and structure. Furthermore, past
studiesthathavedocumentedextremelyoldtrees for agivenspe-
2000),whereas our data set islikely to bemore representative of
site conditions found across mountain forests in the region.
10 of 12 
Journal of Vegetation Science
Anotabledrawbackofour studyis that we cannot teaseapart
how various drivers influence longevit y. Previous work has often
hypothesizedthat slow-growing treesinvest moreindefense, such
as chemicals or other wood properties that promote resistance to
decay (Loehle, 1987),or thatslow radial growth mayincur greater
species studied here, the literature does not provide clear evidence
on how slow grow th or other wood properties may increase longev-
ity. Previous research does not indicate a clear or consistent relation-
shipbetween radial growthand wooddensityfor beech(Bouriaud
found to decrease wooddensity in spruce (Piispanen et al., 2014).
Moreover, in a study of deadwood decay across temperate tree spe-
cies in Europe, wood density was positively correlated with decay
rates for beech, maple, and spruce, whereas chemicals such as phe-
due to light limitation, then presumably they have limited resources
important to note that we cannot rule out genetic control on intra-
specif ic variation in l ife span, or oth er trade-of fs betwe en growth
of old tree s across the primar y-fores t landscapes s ampled in the
than400years.Althoughtheseexceptionally oldtreesarerare,26
out of the 68 stands had more than 10 trees ha−1 that reached the
ation in the density of these trees across the stands likely highlights
how tree longevity is strongly influenced by local disturbance his-
tories, which cover a gradient fromlow-intensitygap dynamics,to
partial canopy disturbance, to severe stand replacement in the study
Schurm an et al., 2018; Čada et a l., 2020; Frankovič et a l., 2020).
Our results imply that disturbance and phenotypic plasticity play a
bance regimes shift toward larger and more intense events under
We thank all who helped collect data in the field and who assisted in
the dendrochronology laboratory.
JP,TAN, andMSvo conceived theidea forthis study. JP,TAN,MSvi,
MSvo located the study sites and dealt with acquiring permissions for
data collection.KB, MM,PJ,MF,RR, MSy,MD, TK,DK, OK,RB, VČ,
analyzedthetree cores.MM, PJ,OV,RB, VČ, VT,MSvo, JSS,andAB
conceptualized and calculated the disturbance parameters. JP, MSvi,
andOVanalysedthe data.JPand TANwrote the draft of themanu-
Data supporting the findings of this study are available at the Dr yad
Jakob Pavlin
Thomas A. Nagel
Ondřej Vostarek
Radek Bače
Bigler,C.(2016)Trade-offsbetweengrowth rate,treesizeand lifespan
ofMountainPine(Pinus montana)intheSwissNationalPark. PLoS
Bigler, C. & Vebl en, T.T. (2009) Inc reased ear ly growth r ates decre ase
longevities of conifers in subalpine forest s. Oikos,118 ,1130–1138.
Biondi, F. (1992) Development of a tree-ring network for the Italian
Peninsula. TreeRing Bulletin,52,15–29.
Black, B.A., Colbert,J.J. & Pederson, N. (2008) Relationships between
radial growth rates and lifespan within North Americantree spe-
cies. Écoscience,15,349–357.
Bouriaud,O.,Bréda,N.,LeMoguédec,G.&Nepveu,G.(200 4)Modelling
variability of wood density in beech as affected by ring age, r adial
growth and climate. Trees - Structure and Function,18,264–276.
Briene n, R.J.W., Cald well, L., Du chesne, L ., Voelker, S., Barich ivich, J.,
growth-lifespantrade-offs.Nature Communications,11,1–10.
Brooks, M.E., Kristensen, K., van Benthem, K.J., Magnusson, A., Berg,
C.W.,Nielsen, A.etal.(2017)glmmTMBbalancesspeed andflex-
ibilit y among pack ages for zero-inf lated gener alized linear m ixed
modeling. R Journal,9,378–400.
Bugmann,H.& Bigler,C. (2011)WilltheCO2fertilization effect infor-
ests be offset by reduced tree longevity? Oecologia,165,533–544.
V.S. et al. (2019) Limi ted capaci ty of tree gr owth to miti gate the
global greenhouse effect under predicted warming. Nature
Communications, 10, 2171.
Čada, V., Trotsiuk,V.,Janda, P.,Mikoláš, M., Bače, R.,Nagel, T.A.et al.
(2020) Qu antifying n atural dist urbances usi ng a large-sc ale den-
drochronological reconstruction to guide forest management.
Ecological Applications,30,1–13.
Cailleret,M.,Jansen,S.,Rober t,E.M.R.,Desoto,L.,Aakala,T.,Antos,J.A .
etal. (2017) A synthesisofradialgrowthpatterns preceding tree
mort ality. Global Change Biology,23,1675–1690.
terns of old Norway spruce trees in Trillemarka forest, Norway.
Scandinavian Journal of Forest Research,28,232–240.
Di Filip po, A., Bio ndi, F., Čufar, K., D e Luis, M., G rabner, M., Ma ugeri,
M. et al. (2007) Bioclimatology of beech (Fagus sylvatica L.) in
the Eastern Alps: Spatial and altitudinal climatic signals identi-
fied through a tree-ring network. Journal of Biogeography, 34,
Bioclimate and grow th history af fect beech lifespan in the Italian
AlpsandApennines.Global Change Biology,18,960–972.
metrics for assessing old-growth forest naturalness. Journal of
Applied Ecology,54,737–749.
PAVLIN et AL.    
 11 of 12
Journal of Vegetation Science
K. et al. (2015) The longevity of broadleaf deciduous trees in
Nort hern Hemisph ere temperate f orests: insi ghts from tree -rin g
series. Frontiers in Ecology and Evolution,3,1–15.
Diaconu, D., Wassenberg, M. & Spiecker, H. (2016) Variability of
European beech wood density as influenced by interactions be-
tweentree-ringgrowthandaspect.Forest Ecosystems,3(6),1–9.
on increm ent cores in kahik atea (Dacrycarpus dacrydioides). New
Zealand Natural Sciences,16,31–37.
Fick, S.E. & Hijmans, R.J. (2017) WorldClim2:new 1km spatial resolu-
tion climate surf aces for global land areas. International Journal of
Fox, J. & Weisberg,S. (2019) An R companion to applied regression, 3rd
primar y forests in Central Europe. Forest Ecology and Management,
479, 11852 2.
Frelich, L .E. & Lorimer, C.G. (1991) Natural disturbance regimes in
hemlock-hardwood forests of the upper great lakes region.
Ecological Monographs,61,145–164.
tatsfo re pi phyticliche ns an db r yophytesonbeech(Fagus sylvatica).
Biological Conservation,143,1008–1016.
Herms, D.A. & Mattson,W.J.(1992)Thedilemma of plants: Togrow
or defend . Journal of Chemical Information and Modeling, 67,
2 8 3 – 3 3 5 .
Holmes,R.(1983) Computer-assistedqualitycontrol intree-ringdating
and measurement. Tree- Ring Bulletin,43,69–77.
Ireland,K.B., Moore,M.M.,Fulé, P.Z., Zegler,T.J.&Keane,R.E.(2014)
Slow lifelong grow th predisposes Populus tremuloides trees to
mort ality. Oecologia,175,847–859.
Janda, P.,Tepley,A.J., Schurman,J.S.,Brabec,M., Nagel,T.A.,Bače, R.
et al. (2019) Driv ers of basal area v ariation acros s primary late -
successional Picea abies forests of the Carpathian Mountains.
Forest Ecology and Management,435,196–204.
Janda, P., Trotsiuk, V., Mikoláš, M., Bače, R., Nagel, T.A ., Seidl, R.
spruce forests in the Western Carpathians and it s influence on
current forest structure and composition. Forest Ecology and
Johnson, S.E. & Abrams, M.D.(2009) Age class,longevity and growth
rate relationships: Protracted growth increases in old trees in the
eastern United States. Tree Physiology,29,1317–1328.
Kahl, T.,Arnstadt,T.,Baber,K.,Bässler,C.,Bauhus,J.,Borken, W.et al.
(2017)Wooddecay ratesof 13temperatetree speciesinrelation
to wood properties, enzyme activities and organismic diversities.
Forest Ecology and Management,391,86–95.
Korpel', Š. (1995) Die Urwälder der Westkarpaten. Stuttgar t, Jena, New
Kozák, D., Mikoláš, M.,Svitok,M., Bače,R., Paillet,Y.,Larrieu, L. etal.
(2018)Profile of tree-related microhabitats in European primary
beech-dominated forests. Forest Ecology and Management, 429,
Lanner,R.M.(2002)Whydotreeslivesolong?Ageing Research Reviews,
Larson,D.W.(2001)Theparadoxofgreatlongevit yinashortlivedtree
species. Experimental Gerontology,36,651–673.
Larson,D.W.,Matthes,U., Gerrath,J.A.,Larson,N.W.K., Gerrath,J.M.,
Nekola, J.C.et al. (2000) Evidenceforthewidespread occurrence
of ancient forest s on clif fs. Journal of Biogeography,27,319–331.
Larsson,L.A.(2003)CDendr o: Cybis Dendro datin g program.Saltsjöbaden,
recruitment probabilities. Journal of Vegetation Science,6,191–204.
Leuschner,C.&Ellenberg,H.(2017)Ecology of Central European Forests.
Springer International Publishing.
species:Traitspectra, functionaltrade-offs, and ecologicalclassi-
fication of adult trees. Perspectives in Plant Ecology, Evolution and
Linden mayer, D.B. & Lauran ce, W.F. (2017) The e cology, distrib ution,
conser vation and management of large old trees. Biological Reviews,
Loehle, C. (1987) Tree life history strategies: the role of defenses.
Canadian Journal of Forest Research,18,209–222.
Loehle,C.(2000)Strategy spaceandthedisturbance spectrum:Alife-
history model for tree species coexistence. American Naturalist,
disturbance frequency andintensityin densetemperatefor- ests.
Canadian Journal of Forest Research,19,651–663.
Lusk,C.&Smith,B.(1998)Lifehistorydifferencesandtreespecies co-
existen ce in an old-gr owth New Zeal and rain for est. Ecology, 79,
Martin-Benito, D., Pederson, N.,Férriz, M. & Gea-Izquierdo,G. (2021)
longevity of aboveground carbon. Science of the Total Environment,
765, 142737.
Meigs, G.W., Morrissey, R.C., Bače, R., Chaskovsk yy, O., Čada, V.,
Després,T.etal. (2017) More waysthanone:Mixed-severitydis-
mental pathways. Forest Ecology and Management,406,410–426.
Mikoláš, M., Ujházy, K., Jasík, M.,Wiezik, M., Gallay,I., Polák, P.et al.
(2019)Primaryforest distribution and representation in a Central
Europea n landscape : Results of a lar ge-s cale field-based census .
Forest Ecology and Management, 449, 117466.
Z.etal .(2011)Towa rdadefinitionofth erangeofvariabilityofcen-
tralEuropeanmixedFagus-abies-piceaforests:Thenearly steady-
state forestofLom(Bosnia andHerzegovina).Canadian Journal of
Forest Research,41,1871–1884.
Nagel, T.A ., Mikac, S., Dolinar, M., Klopcic, M., Keren, S., Svoboda,
M. et al. (2 017) The natura l disturb ance regime i n forests of t he
Dinari c Mountains : A syn thesis of evidenc e. Forest Ecology and
Nagel, T.A.,Svoboda,M.& Diaci,J.(2006) Regenerationpatternsafter
intermediate wind disturbance in an old-growth Fagus-Abies for-
est in southeastern Slovenia. Forest Ecology and Management, 226,
and replacementpatterns in an old-growth Fagus-Abies forest of
Bosnia-Herzegovina.Plant Ecology,208,307–318.
Nagel, T.A ., Svoboda, M . & Kobal, M. (2014) Dist urbance, life h istory
traits, and dynamics in an old-growth forest landscape of south-
eastern Europe. Ecological Applications,24,663–679.
Nakagaw a, S., Johns on, P.C.D. & Schielze th, H. (2017) Th e coeffic ient
of determination R2 and intra-class correlation coefficient from
genera lized linear mixe d-ef fects mo dels revisited a nd expanded .
Journal of the Royal Society Interface, 14 .
ous fores ts. Trends in Ecology and Evolution,16,205–210.
Orwig,D. A. &Abrams,M.D.(1994)Contrastingradial growthandcan-
opy recruitment patterns in Liriodendron tulipifera and Nyssa syl-
vatica:Gap-obligate versusgap-facultativetree species.Canadian
Journal of Forest Research,24,2141–2149.
Petritan,A .M.,VonLüpke,B.&Petritan,I.C.(2007)Effectsofshadeon
growth andmortalityofmaple(Acer pseudoplatanus),ash(Fraxinus
excelsior)an d beec h(Fagus sylvatica)saplings.Forestry,80, 397–412 .
12 of 12 
Journal of Vegetation Science
Piispan en, R., Hei nonen, J., Val konen, S., Mä kinen, H. , Lundqvis t, S.O.
&Saranpää,P.(2014)Wooddensityof Norwayspruce inuneven-
aged stands. Canadian Journal of Forest Research,44,136–144.
growth allowsfor maximumlongevityin Europeanbeech.Ecology,
etal.(2019)Treegrowthpatternsassociatedwithex tremelongev-
ity: Implications for the ecology and conservation of primeval t rees
in Mediterranean mountains. Anthropocene, 26, 100199.
Structure,dynamicsand dendroecologyofan old-growthFagusfor-
estintheApennines.Journal of Vegetation Science,16,13–28.
Roskilly, B., Keeling, E., Hood, S., Giuggiola, A. & Sala, A. (2019)
Conflicting functional effectsof xylem pit structure relate to the
growth-longevity trade-offina coniferspecies.Proceedings of the
National Academy of Sciences of the United States of America, 116,
ogy and site conditions on longevity of Norway spruce in the north-
ernSwissAlps.European Journal of Forest Research,131,1117–1125.
Sabatini, F.M., Burrascano, S., Keeton, W.S., Levers, C ., Lindner, M.,
Diversity and Distributions,24,1426–1439.
(2019) The climat ic drivers of p rimary Pi cea forest gr owth along
the Carpathian arc are changing under rising temperatures. Global
Change Biology,25,3136–3150.
Schurman,J.S., Trotsiuk,V.,Bače,R.,Čada, V.,Fraver,S.,Janda, P.etal.
of primar y Picea abies forest s. Global Change Biology,2 4,21 69–21 81.
Seidl, R., Thom, D., Kaut z, M., Martin-Benito, D., Peltoniemi, M.,
Vacchiano, G. et al. (2017) Forest disturbances under climate
change. Nature Climate Change,7,395–402.
Splechtna,B.E.,Dobrý,J.&Klinka,K.(200 0)Tree-ringcharacteristicsof
subalpinefir(Abies lasiocarpa(Hook.)Nutt.)inrelationtoelevation
and climatic fluctuations. Annals of Forest Science,57,89–100.
Stancioiu, P.T.& O’hara, K.L. (2006) Regeneration growth in different
lightenvironmentsofmixed species, multiaged, mountainous for-
ests of Romania. European Journal of Forest Research,125,151–162.
Standovár,T.&Kenderes,K. (2003)Areview on natural standdynam-
ics in Beechwoods of East Central Europe. Applied Ecology and
Environmental Research,1,19–46.
Svoboda, M.,Janda,P.,Bače,R., Fraver,S.,Nagel, T.A ., Rejzek,J. et al.
mary Picea abies mount ain forests of the Eastern Carpathians,
Romania. Journal of Vegetation Science,25,386–401.
forest s in the Cross Timbers of Osage Count y, Oklahoma. Jour nal of
(2014) A mixed seve rity dist urbance regi me in the prim ary Picea
abies(L.)Karst.forestsoftheUkrainianC arpathians.Fores t Ecology
and Management,334,144–153.
VanPelt,R.(2007)Identifying mature and old forests in western Washington.
forest s of the central Rockies. Ecology,67,644–649.
fers. Science,119,883–884.
Wickham,H.(2016)ggplot2: Elegant Gra phics for Data Ana lysis.NewYork:
Wong, C.M. & Lertzman, K.P.(2001)Errorsinestimating tree age: im-
plications for studies of s tand dynamic s. Canadian Journal of Forest
Yamaguchi, D.K. (1991) A simple method for cross-dating increment
cores from living trees. Canadian Journal of Forest Research, 21,
Zuur,A .F.,Ieno, E.N., Walker,N.J.,Saveliev, A. A. & Smith, G.M.(2009)
Mixed Effec ts models and extensions in ecology with R.NewYork,NY:
Springer Science & Business Media.
Additional supporting information may be found online in the
Supporting Information section.
How to cite this article:Pavlin,J.,Nagel,T.A .,Svitok,M.,
history is a key driver of tree life span in temperate primary
forests. Journal of Vegetation Science, 32:e013069. ht t p s: //
doi .org /10.1111/j vs .13069
... Nevertheless, as far as we know, no previous research has aimed at studying tree-ring patterns and disturbance histories to describe wind mortality following extreme windstorms in any type of forest ecosystems. Pavlin et al. [32] quantified the relationship between disturbance history and tree-growth patterns and tree lifespan across primeval forests of Carpathian Mountains, but they did not provide clear evidence that prolonged periods of suppression of tree in understory contributes to its greater mechanical resistance during the strong winds. ...
... A root system weakened by pathogens might be more prone to wind disturbance. Finally, frequent and severe windstorms may not allow trees to grow to exceptional age [32]. ...
Full-text available
The driving forces of tree mortality following wind disturbances of mountain mixed European temperate forests belongs among issues not comprehensively resolved. Hence, we aimed to elucidate the key factors of tree resistance to historical severe disturbance events in the Boubínský Primeval Forest, one of the oldest forest reserves in the Czech Republic. By using spatially explicit tree census, dendrochronological and soil data, we study spatial and temporal patterns of past disturbances and mathematically compared selected characteristics of neighboring trees that were killed by a severe storm in 2017 and those that remained undisturbed. The tendency of trees toward falling was primarily driven edaphically, limiting severe events non-randomly to previously disturbed sites occupied by hydromorphic soils and promoting the existence of two spatially-separated disturbance regimes. While disturbed trees usually recruited in gaps and experienced only one severe release event, surviving trees characteristically regenerated under the canopy and were repeatedly released. Despite the fact that disturbed trees tended to reach both lower ages and dimensions than survivors, they experienced significantly higher growth rates. Our study indicates that slow growth with several suppression periods emerged as the most effective tree strategy for withstanding severe windstorms, dying of senescence in overaged life stage. Despite the selective impact of the Herwart storm on conifer population, we did not find any difference in species sensitivity for most characteristics studied. We conclude that the presence of such ancient, high-density wood trees contributes significantly to the resistance of an entire stand to severe storms.
... Extensive studies of Norway spruce growth-climate relationships have shown high sensitivity of Carpathian high-elevation spruce forests to growing season temperature (Bouriaud and Popa, 2009;Büntgen et al., 2007;Nagavciuc et al., 2019;Sedmáková et al., 2019;Svobodová et al., 2019). In addition to climatic drivers, natural disturbances have an important role in regulating forest structure, species composition, and developmental processes across spatial scales (Čada et al., 2020;Pavlin et al., 2021;Schurman et al., 2018;Turner 1987). Reconstructions of disturbance histories have shown that most Carpathian forests are driven by frequent low-to-moderate severity disturbances, with windstorms, and subsequent bark beetle outbreaks, as the main disturbance agents (Čada et al. 2013;Svoboda et al., 2014;Trotsiuk et al., 2014;Vacek et al., 2015;Zielonka et al., 2010). ...
Full-text available
Tree radial growth is influenced by climatic and various non-climatic factors, which can complicate the extraction of climate signals from tree rings. We investigated the influence of disturbance on tree-ring width (RW) and latewood blue intensity (BI) chronologies of Norway spruce from the Carpathian Mountains to explore the extent to which disturbance can affect temperature signals in tree rings. Overall, ∼15000 high-elevation Norway spruce tree cores from 34 sites grouped into four regions (Slovakia, Ukraine, North and South Romania) were analyzed. The curve intervention detection (CID) method was applied to detect and correct identified disturbance trends. RW chronology structural comparisons were performed among disturbance-affected and disturbance-corrected chronologies for various spatial (regional / site) scales and sampling subsets. Structural comparisons were also performed for RW and BI chronologies developed from separate groups of series (i.e., disturbed, and undisturbed) for five sites exhibiting clear disturbance trends. Temperature sensitivity was assessed for all chronology variants of both parameters. We found that disturbance trends only affected RW chronologies at the site/subset scale with relatively small series replication and were not detected at the regional scale. Unlike RW, BI chronologies were generally unaffected by disturbance. BI data also contained much stronger growing season temperature signals, which appeared to be both spatially and temporally more coherent. Whereas highly replicated and spatially extensive datasets can help minimize or eliminate disturbance trends in RW chronologies, this potential influence should be considered when interpreting climatic signals in tree rings and reconstructing historical climate in weakly replicated periods. On the other hand, BI is a promising alternative tree ring parameter with stronger and more stable growing season temperature signals, whose seemingly disturbance-free chronology structure does not appear to suffer from this ecological bias, and therefore represents a more suitable parameter for dendroclimatological research.
... Although growth-mortality tradeoff is a well-conserved phenomenon (e.g., Bigler & Veblen, 2009;Black et al., 2008;Bošela et al., 2018;Brienen et al., 2016;Johnson & Abrams, 2009;Piovesan et al., 2019), additional regional factors should be considered before strong conclusions are made regarding the implications of the growth-mortality tradeoff for emergent forest properties such as productivity, carbon storage, and LOT abundance. The character of a region's natural disturbance regime plays an important role in dictating forest age structure (Fraver et al., 2009;Meigs et al., 2017;Pavlin et al., 2021;Svoboda et al., 2014;Turner, 2010) and the competitive interactions between trees . In closed-canopy forests, many trees will have recruited in shade and undergone repeated cycles of suppression and releases, whereas others will have recruited in canopy openings following a severe disturbance (Canham et al., 1990), thus shaping the conservativeness of a tree's growth history and, potentially, longevity. ...
In a world of accelerating changes in environmental conditions driving tree growth, tradeoffs between tree growth rate and longevity could curtail the abundance of large, old trees (LOTs), with potentially dire consequences for biodiversity and carbon storage. However, the influence of tree‐level tradeoffs on forest structure at landscape scales will also depend on disturbances, which shape tree size and age distribution, and on whether LOTs can benefit from improved growing conditions due to climate warming. We analyzed temporal and spatial variation in radial growth patterns from ~ 5000 Norway spruce (Picea abies (L.) H. Karst) live and dead trees from the Western Carpathian primary spruce forest stands. We applied mixed‐linear modeling to quantify the importance of LOT growth histories and stand dynamics (i.e. competition and disturbance factors) on lifespan. Finally, we assessed regional synchronization in radial growth variability over the 20th century, and modelled the effects of stand dynamics and climate on LOTs recent growth trends. Tree age varied considerably among forest stands, implying an important role of disturbance as an age constraint. Slow juvenile growth and longer period of suppressed growth prolonged tree lifespan, while increasing disturbance severity and shorter time since last disturbance decreased it. The highest age was not achieved only by trees with continuous slow growth, but those with slow juvenile growth followed by subsequent growth releases. Growth trend analysis demonstrated an increase in absolute growth rates in response to climate warming, with late summer temperatures driving the recent growth trend. Contrary to our expectation that LOTs would eventually exhibit declining growth rates, the oldest LOTs (> 400 years) continuously increase growth throughout their lives, indicating a high phenotypic plasticity of LOTs for increasing biomass, and a strong carbon sink role of primary spruce forests under rising temperatures, intensifying droughts, and increasing bark beetle outbreaks.
... Tree diameter is only roughly correlated with age, as it is highly dependent on the tree species and environmental factors such as the fertility of the site, climate or level of competition. Some latesuccessional tree species, such as silver fir (Abies alba) and European beech (Fagus silvatica), can go through a very long-lasting stagnation stage, sometimes for more than one century, when dbh increases only slightly (Pantic et al., 2015;Pavlin et al., 2021). As dbh is much easier to record than age, it is often used in studies based on longitudinal monitoring of individual trees (e.g. ...
Tree-related microhabitats (TreMs) have been identified as key features for forest-dwelling taxa and are often employed as measures for biodiversity conservation in integrative forest management. However, managing forests to ensure an uninterrupted resource supply for TreM-dwelling taxa is challenging since TreMs are structures with a limited availability, some of which are triggered by stochastic events or require a long time to develop. At the tree scale, the role of tree species, diameter at breast height (dbh) and status (i.e. living vs standing dead) for favouring TreM occurrence has been quantified and modelled in several studies, since these tree features are routinely recorded in the field. However, TreM occurrence remains difficult to predict, hampering the elaboration of applicable management strategies that consider TreMs. Using an international database encompassing 110,000 trees, we quantified the explanatory power of tree species, dbh, status, time since last harvest and plot context for predicting TreM occurrence at the tree level. Plot context is so far a “black box” that combines local environmental conditions, past and current management legacies, with local biotic features that have high explanatory power for predicting TreM occurrence. Then, based on the literature, we established a set of 21 factors related to site, stand and tree features for which there is a strong assumption that they play a key role in TreM formation. Finally, we identified a sub-set of nine features that should be recorded in the future to provide additional information to enable better prediction of the occurrence of particular TreMs: (i) at plot level: slope, exposure, altitude and presence of cliffs; and (ii) at tree level: bark features, phyllotaxis and compartmentalization capacity of the tree species, plus ontogenic stage and physiological state of the individual tree sampled.
Hotter droughts have become important drivers of change in the structure of forest canopy, and the role of structural loss of overstory trees is gaining attention as an important factor that would trigger changes in understory behavior. The projection of understory individuals, allows us to analyze the potential future of the forests. Through a dendroecological approach, we evaluated growth responses in understory Austrocedrus chilensis trees growing underneath canopy that experienced tree mortality as a consequence of severe droughts occurred in Patagonia Argentina. We analyze the climatic response, tree growth patterns and drought resilience of understory components in three A. chilensis mixed forests. Tree growth was significantly reduced by drought, highlighting the climatic sensitivity of understory individuals; but indirectly, tree growth also showed releases associated with the openings due to canopy mortality. Thus, we found that understory tree growth increased over time. Our results demonstrated that growth performance of understory A. chilensis trees become largely modeled by a combination of the releases associated with changes of the canopy and their ability to withstand adverse weather conditions. We were able to show a clear capacity of understory A. chilensis to cope with drought events by increasing its growth, taking advantage of the disturbance in the upper layers of the canopy, while resisting successive dry years. However, the shadier environment underneath canopy did not buffered drought conditions, and understory A. chilensis trees suffers the effects of droughts. Considering an increase in drought frequency and intensity in climate predictions, with the consequent openings due to overstory tree mortality, understory species behavior will result as a complex interaction among a potential increase in understory vulnerability to more severe droughts, and the opportunity of canopy ascension.
Full-text available
Ancient Pu’er tea trees (Camellia sinensis var. assamica (J. W. Mast.) Kitam.) are an important ecological resource with high economic value. Knowledge of the environmental variables shaping the original distribution and the effects of climate change on the future potential distribution of these trees, as well as the identification of sustainable management approaches, is essential for ensuring their future health and production. Here, we used 28 current environmental variables and the future climate data to model the suitable areas for ancient Pu’er tea trees. We also compared the health of these ancient trees in areas under different local management strategies. The results suggested the general distribution is likely to remain stable, but there are environmentally suitable areas outside its current habitats. To achieve more sustainable management, the main areas in which the management of poorly-managed trees can be improved include learning from managers of well-managed trees and following the common technical management regulations stipulated by the local government. The suitable value ranges for environmental factors, potentially suitable areas under climate change, and assessment of management approaches will aid the future cultivation and transplantation of ancient Pu’er tea trees. The methodology includes management-level analysis and provides practical insights that could be applied to regions outside the most suitable areas identified.
Full-text available
Forest typology is yet to get sufficiently integrated into related ecological and geographical sciences. The succession of plant phytocenoses necessitates constant adjustments to forest types. The discussed studies have been conducted to improve the classification of forest types in the Ukrainian Carpathians and bring the description closer to the actual distribution of vegetation within the studied area. The paper provides a detailed analysis of forest typology research for the Ukrainian Carpathians area during late 20 th and early 21 st century. The forest fund areas of the Carpathian region, which are subordinated to the State Agency of Forest Resources of Ukraine, have been classified by forest vegetation types. Total area of subor and coniferous forest types is 1,493.1 ha and 28,910.2 ha, respectively. The study involved establishing of permanent sample plots on the territory of nature protection institutions and forest management enterprises. According to the findings, it is proposed to complement the classification of subor and coniferous forest type as defined by Z.Yu. Herushynskyi on the territory of the Ukrainian Carpathians with the following types: fresh pine subor forest type, wet pine subor forest type, and wet pine coniferous type. The paper defines the main diagnostic features of the suggested forest types. These subor and coniferous forest types can be clearly distinguished from other forest types by soil and hydrological conditions, and can be used to describe the corresponding forest vegetation types. The correctness of definition of new forest types is confirmed with a set of plant indicator species that have been identified within the relevant areas. The findings provide a better understanding of forest ecology and make a significant contribution to forest typology studies on the territory of the Ukrainian Carpathians. Another step towards researching the patterns of the establishment of plant complexes in the Ukrainian Carpathians has been taken. Forest formations of the Carpathians are presented in more detail in forest typological science
Full-text available
Land vegetation is currently taking up large amounts of atmospheric CO 2 , possibly due to tree growth stimulation. Extant models predict that this growth stimulation will continue to cause a net carbon uptake this century. However, there are indications that increased growth rates may shorten trees′ lifespan and thus recent increases in forest carbon stocks may be transient due to lagged increases in mortality. Here we show that growth-lifespan trade-offs are indeed near universal, occurring across almost all species and climates. This trade-off is directly linked to faster growth reducing tree lifespan, and not due to covariance with climate or environment. Thus, current tree growth stimulation will, inevitably, result in a lagged increase in canopy tree mortality, as is indeed widely observed, and eventually neutralise carbon gains due to growth stimulation. Results from a strongly data-based forest simulator confirm these expectations. Extant Earth system model projections of global forest carbon sink persistence are likely too optimistic, increasing the need to curb greenhouse gas emissions.
Full-text available
Consistent with a ubiquitous life history trade-off, trees exhibit a negative relationship between growth and longevity both among and within species. However, the mechanistic basis of this life history trade-off is not well understood. In addition to resource allocation conflicts among multiple traits, functional conflicts arising from individual morphological traits may also contribute to life history trade-offs. We hypothesized that conflicting functional effects of xylem structural traits contribute to the growth-longevity trade-off in trees. We tested this hypothesis by examining the extent to which xylem morphological traits (i.e., wood density, tracheid diameters, and pit structure) relate to growth rates and longevity in two natural populations of the conifer species Pinus ponderosa. Hydraulic constraints arise as trees grow larger and xylem anatomical traits adjust to compensate. We disentangled the effects of size through ontogeny in individual trees and growth rates among trees on xylem traits by sampling each tree at multiple trunk diameters. We found that the oldest trees had slower lifetime growth rates compared with younger trees in the studied populations, indicating a growth-longevity trade-off. We further provide evidence that a single xylem trait, pit structure, with conflicting effects on xylem function (hydraulic safety and efficiency) relates to the growth-longevity trade-off in a conifer species. This study highlights that, in addition to trade-offs among multiple traits, functional constraints based on individual morphological traits like that of pit structure provide mech-anistic insight into how and when life history trade-offs arise.
Full-text available
It is generally accepted that animal heartbeat and lifespan are often inversely correlated, however, the relationship between productivity and longevity has not yet been described for trees growing under industrial and pre-industrial climates. Using 1768 annually resolved and absolutely dated ring width measurement series from living and dead conifers that grew in undisturbed, high-elevation sites in the Spanish Pyrenees and the Russian Altai over the past 2000 years, we test the hypothesis of grow fast-die young. We find maximum tree ages are significantly correlated with slow juvenile growth rates. We conclude, the interdependence between higher stem productivity, faster tree turnover, and shorter carbon residence time, reduces the capacity of forest ecosystems to store carbon under a climate warming-induced stimulation of tree growth at policy-relevant timescales.
Natural disturbances are key factors in the formation of forest ecosystem structure. Evaluation of the spatial and temporal extent of disturbance regimes is critical for understanding forest dynamics, forest structural hetero-geneity, and biodiversity habitats. Quantifying disturbance regimes is therefore imperative for appropriate management of forests and protected areas. However, natural disturbance regimes have rarely been assessed using dendrochronological methods on a regional scale across primary mixed beech-fir forest stands-one of the dominant forest vegetation types in Europe. To study the natural disturbance regimes of beech-dominated mixed-forest stands, we established 42 permanent study plots with an area of 0.1 ha across three primary forest stands in the Western Carpathians, a region that still contains large areas of primary forest. We reconstructed each stand-level disturbance history using a tree-ring based approach. The temporal synchronicity of disturbance events was then evaluated by delineating stand-level disturbance events using a kernel density function, and through the detection of plot-level disturbances with severities greater than 10 percent. The results obtained from the chronologies showed substantial variability in time and space, especially in the mid-19th century. Low-and moderate-severity plot-level disturbance events were most common, but high-and extremely high-severity plot-level disturbance events also occurred. The observed spatial and temporal variability suggests that the beech-dominated forests were primarily driven by mixed-severity disturbance regimes, with windstorms as the main disturbance agent. This reconstruction of the disturbance regime provided unique insight into the scale of mortality processes in these beech-dominated mixed forests. This information can help guide ecological forestry in areas where both wood production and biodiversity preservation are simultaneous goals, such as by employing more spatio-temporally-complex silvicultural systems that resemble natural disturbance patterns and facilitate heterogeneous forest structures.
Most information on the ecology of oak-dominated forests in Europe comes from forests altered for centuries because remnants of old-growth forests are rare. Disturbance and recruitment regimes in old-growth forests provide information on forest dynamics and their effects on long-term carbon storage. In an old-growth Quercus petraea forest in northwestern Spain, we inventoried three plots and extracted cores from 166 live and dead trees across canopy classes (DBH ≥ 5 cm). We reconstructed disturbance dynamics for the last 500 years from tree-ring widths. We also reconstructed past dynamics of above ground biomass (AGB) and recent AGB accumulation rates at stand level using allometric equations. From these data, we present a new tree-ring-based approach to estimate the age of carbon stored in AGB. The oldest tree was at least 568 years, making it the oldest known precisely-dated oak to date and one of the oldest broadleaved trees in the Northern Hemisphere. All plots contained trees over 400 years old. The disturbance regime was dominated by small, frequent releases with just a few more intense disturbances that affected ≤20% of trees. Oak recruitment was variable but rather continuous for 500 years. Carbon turnover times ranged between 153 and 229 years and mean carbon ages between 108 and 167 years. Over 50% of AGB (150 Mg·ha⁻¹) persisted ≥100 years and up to 21% of AGB (77 Mg·ha⁻¹) ≥300 years. Low disturbance rates and low productivity maintained current canopy oak dominance. Absence of management or stand-replacing disturbances over the last 500 years resulted in high forest stability, long carbon turnover times and long mean carbon ages. Observed dynamics and the absence of shade-tolerant species suggest that oak dominance could continue in the future. Our estimations of long-term carbon storage at centennial scales in unmanaged old-growth forests highlights the importance of management and natural disturbances for the global carbon cycle.
Estimates of historical disturbance patterns are essential to guide forest management aimed at ensuring the sustainability of ecosystem functions and biodiversity. However, quantitative estimates of various disturbance characteristics required in management applications are rare in longer‐term historical studies. Thus, our objectives were to: (1) quantify past disturbance severity, patch size, and stand proportion disturbed, and (2) test for temporal and sub‐regional differences in these characteristics. We developed a comprehensive dendrochronological method to evaluate an approximately two‐century‐long disturbance record in the remaining Central and Eastern European primary mountain spruce forests, where wind and bark beetles are the predominant disturbance agents. We used an unprecedented large‐scale nested design dataset of 541 plots located within 44 stands and 6 sub‐regions. To quantify individual disturbance events, we used tree‐ring proxies, which were aggregated at plot and stand levels by smoothing and detecting peaks in their distributions. The spatial aggregation of disturbance events was used to estimate patch sizes. Data exhibited continuous gradients from low‐ to high‐severity and small‐ to large‐size disturbance events. In addition to the importance of small disturbance events, moderate‐scale (25‐75% of the stand disturbed, >10 ha patch size) and moderate‐severity (25‐75% of canopy disturbed) events were also common. Moderate disturbances represented more than 50% of the total disturbed area and their rotation periods ranged from one to several hundred years, which is within the lifespan of local tree species. Disturbance severities differed among sub‐regions, whereas the stand proportion disturbed varied significantly over time. This indicates partially independent variations among disturbance characteristics. Our quantitative estimates of disturbance severity, patch size, stand proportion disturbed, and associated rotation periods provide rigorous baseline data for future ecological research, decisions within biodiversity conservation, and silviculture intended to maintain native biodiversity and ecosystem functions. These results highlight a need for sufficiently large and adequately connected networks of strict reserves, more complex silvicultural treatments that emulate the natural disturbance spectrum in harvest rotation times, sizes, and intensities, and higher levels of tree and structural legacy retention.
Given the global intensification of forest management and climate change, protecting and studying forests that develop free of direct human intervention-also known as primary forests-are becoming increasingly important. Yet, most countries still lack data regarding primary forest distribution. Previous studies have tested remote sensing approaches as a promising tool for identifying primary forests. However, their precision is highly dependent on data quality and resolution, which vary considerably. This has led to underestimation of primary forest abundance and distribution in some regions, such as the temperate zone of Europe. Field-based inventories of primary forests and methodologies to conduct these assessments are inconsistent; incomplete or inaccurate mapping increases the vulnerability of primary forest systems to continued loss from clearing and land-use change. We developed a comprehensive methodological approach for identifying primary forests, and tested it within one of Europe's hotspots of primary forest abundance: the Carpathian Mountains. From 2009 to 2015, we conducted the first national-scale primary forest census covering the entire 49,036 km 2 area of the Slovak Republic. We analyzed primary forest distribution patterns and the representativeness of potential vegetation types within primary forest remnants. We further evaluated the conservation status and extent of primary forest loss. Remaining primary forests are small, fragmented, and often do not represent the potential natural vegetation. We identified 261 primary forest localities. However, they represent only 0.47% of the total forested area, which is 0.21% of the country's land area. The spatial pattern of primary forests was clustered. Primary forests have tended to escape anthropogenic disturbance on sites with higher elevations, steeper slopes, rugged terrain, and greater distances from roads and settlements. Primary forest stands of montane mixed and subalpine spruce forests are more abundant compared to broadleaved forests. Notably, several habitat types are completely missing within primary forests (e.g., floodplain forests). More than 30% of the remaining primary forests are not strictly protected, and harvesting occurred at 32 primary forest localities within the study period. Almost all logging of primary forests was conducted inside of protected areas, underscoring the critical status of primary forest distribution in this part of Europe. Effective conservation strategies are urgently needed to stop the rapid loss and fragmentation of the remaining primary forests. Our approach based on precise, field-based surveys is widely applicable and transferrable to other fragmented forest landscapes.
Climatic constraints on tree growth mediate an important link between terrestrial and atmospheric carbon pools. Tree rings provide valuable information on climate‐driven growth patterns, but existing data tend to be biased towards older trees on climatically extreme sites. Understanding climate change responses of biogeographic regions requires data that integrate spatial variability in growing conditions and forest structure. We analyzed both temporal (c. 1901‐2010) and spatial variation in radial growth patterns in 9 876 trees from fragments of primary Picea abies forests spanning the latitudinal and altitudinal extent of the Carpathian arc. Growth was positively correlated with summer temperatures and spring moisture availability throughout the entire region. However, important seasonal variation in climate responses occurred along geospatial gradients. At northern sites, winter precipitation and October temperatures of the year preceding ring formation were positively correlated with ring width. In contrast, trees at the southern extent of the Carpathians responded negatively to warm and dry conditions in autumn of the year preceding ring formation. An assessment of regional synchronization in radial growth variability showed temporal fluctuations throughout the 20th century linked to the onset of moisture limitation in southern landscapes. Since the beginning of the study period, differences between high and low elevations in the temperature sensitivity of tree growth generally declined, while moisture sensitivity increased at lower elevations. Growth trend analyses demonstrated changes in absolute tree growth rates linked to climatic change, with basal area increments in northern landscapes and lower altitudes responding positively to recent warming. Tree growth has predominantly increased with rising temperatures in the Carpathians, accompanied by early indicators that portions of the mountain range are transitioning from temperature to moisture limitation. Continued warming will alleviate large‐scale temperature constraints on tree growth, giving increasing weight to local drivers that are more challenging to predict. This article is protected by copyright. All rights reserved.