Phenological patterns of flowering across biogeographical regions of Europe

Abstract

Long-term changes of plant phenological phases determined by complex interactions of environmental factors are in the focus of recent climate impact research. There is a lack of studies on the comparison of biogeographical regions in Europe in terms of plant responses to climate. We examined the flowering phenology of plant species to identify the spatio-temporal patterns in their responses to environmental variables over the period 1970–2010. Data were collected from 12 countries along a 3000-km-long, North–South transect from northern to eastern Central Europe.

Full text

1 23
International Journal of
Biometeorology
ISSN 0020-7128
Int J Biometeorol
DOI 10.1007/s00484-017-1312-6
Phenological patterns of flowering across
biogeographical regions of Europe
NS-Pheno Team

1 23
Your article is protected by copyright and
all rights are held exclusively by ISB. This e-
offprint is for personal use only and shall not
be self-archived in electronic repositories. If
you wish to self-archive your article, please
use the accepted manuscript version for
posting on your own website. You may
further deposit the accepted manuscript
version in any repository, provided it is only
made publicly available 12 months after
official publication or later and provided
acknowledgement is given to the original
source of publication and a link is inserted
to the published article on Springer's
website. The link must be accompanied by
the following text: "The final publication is
available at link.springer.com”.

Int J Biometeorol
DOI 10.1007/s00484-017-1312-6
ORIGINAL PAPER
Phenological patterns of flowering across biogeographical
regions of Europe
Barbara Templ1·Matthias Templ2·Peter Filzmoser3·Annam´
aria Lehoczky4·
Eugenija Bakˇ
sien`
e5·Stefan Fleck6·Hilppa Gregow7·Sabina Hodzic8·
Gunta Kalvane9·Eero Kubin10 ·Vello Palm11 ·Danuta Romanovskaja5·
Viˇ
snja Vuˇ
ceti´
c12 ·Ana ˇ
Zust13 ·B´
alint Cz´
ucz14,15 ·NS-Pheno Team
Received: 28 October 2016 / Revised: 18 January 2017 / Accepted: 18 January 2017
©ISB2017
Abstract Long-term changes of plant phenological phases
determined by complex interactions of environmental fac-
tors are in the focus of recent climate impact research. There
is a lack of studies on the comparison of biogeographical
regions in Europe in terms of plant responses to climate. We
examined the flowering phenology of plant species to iden-
tify the spatio-temporal patterns in their responses to envi-
ronmental variables over the period 1970–2010. Data were
collected from 12 countries along a 3000-km-long, North–
South transect from northern to eastern Central Europe.
!Barbara Templ
barbara.a.templ@gmail.com
1Department of Plant Systematics, Ecology and Theoretical
Biology, E¨
otv¨
os Lor´
and University, 1117, Budapest, Hungary
2Institute of Data Analysis and Process Design, Zurich
University of Applied Sciences, 8401, Winterthur, Switzerland
3Institute of Statistics and Mathematical Methods
in Economics, Vienna University of Technology,
1040, Vienna, Austria
4Centre for Climate Change, University Rovira i Virgili,
43500, Tortosa, Spain
5Vo k e B r a n c h o f t h e L i t h u a n i a n R e s e a r c h , C e n t r e
for Agriculture and Forestry, 02232, Vilnius, Lithuania
6Statistics Austria, 1110, Vienna, Austria
7Climate Service Centre, Finnish Meteorological Institute,
00101, Helsinki, Finnland
8Sector for Applied Meteorology, Federal Hydrometeorological
Institute of Federation of Bosnia and Herzegovina, 71000,
Sarajevo, Bosnia and Herzegovina
Biogeographical regions of Europe were covered from
Finland to Macedonia. Robust statistical methods were
used to determine the most influential factors driving the
changes of the beginning of flowering dates. Significant
species-specific advancements in plant flowering onsets
within the Continental (3 to 8.3 days), Alpine (2 to 3.8 days)
and by highest magnitude in the Boreal biogeographical
regions (2.2 to 9.6 days per decades) were found, while less
pronounced responses were detected in the Pannonian and
Mediterranean regions. While most of the other studies only
9Faculty of Geography and Earth Sciences,
University of Latvia, 1010, Riga, Latvia
10 Natural Resources and Bioproduction, Natural Resources
Institute Finland, 00790, Helsinki, Finnland
11 Institute of Ecology and Earth Science, University of Tartu,
51014, Tartu, Estonia
12 Meteorological Research and Development Division,
Meteorological and Hydrological Service,
10000, Zagreb, Croatia
13 Agrometeorological Department, Environmental Agency
of the Republic of Slovenia, 1000, Ljubljana, Slovenia
14 European Topic Centre on Biological Diversity,
French National Museum of Natural History,
75231, Paris, France
15 Institute of Ecology and Botany, MTA Centre
for Ecological Research, 2163, V´
acr´
at´
ot, Hungary
Author's personal copy

Int J Biometeorol
use mean temperature in the models, we show that also the
distribution of minimum and maximum temperatures are
reasonable to consider as explanatory variable. Not just
local (e.g. temperature) but large scale (e.g. North Atlantic
Oscillation) climate factors, as well as altitude and latitude
play significant role in the timing of flowering across
biogeographical regions of Europe. Our analysis gave evi-
dences that species show a delay in the timing of flowering
with an increase in latitude (between the geographical coor-
dinates of 40.9 and 67.9), and an advance with changing
climate. The woody species (black locust and small-leaved
lime) showed stronger advancements in their timing of
flowering than the herbaceous species (dandelion, lily of
the valley). In later decades (1991–2010), more pronounced
phenological change was detected than during the earlier
years (1970–1990), which indicates the increased influence
of human induced higher spring temperatures in the late
twentieth century.
Keywords Beginning of flowering ·Biogeographical
regions ·Climate change ·Europe ·Robust regression ·
Shifting trend
Introduction
The scientific understanding of the causes of observed
changes in the climate system has been increasing accord-
ing to the report of the Intergovernmental Panel on Cli-
mate Change (Stocker et al. 2013). Climate model projec-
tions indicate that, regarding temperature and precipitation
changes, similar tendencies are likely to continue over the
coming century; however, future changes will vary across
regions (Stocker et al. 2013). These evidences also call
ecologists’s attention to phenomena in natural ecosystems’s
shifting in time related to global warming (Walther et al.
2002;Parmesan2006;IPCC2007; Franks 2015).
Phenology is the study of periodically repeating stages in
the life cycle of animals and plants as influenced by envi-
ronmental conditions (Demar`
ee and Rutishauser 2009). The
likelihood of species occurrence in a certain area depends
on survival and reproduction, which are both depending on
the species’ phenology and thus intimately linked to climate
(Cleland et al. 2007). Observational (Menzel et al. 2006;
Koch et al. 2009;Schleipetal.2009), field experimental
(Wolkovich et al. 2012), predicted (Aguilera et al. 2015)and
remotely sensed (White et al. 2005) data suggest that the
timing of several plant phenological phases advance and/or
delay across the globe, from the Northern (Schwartz et al.
2006) to the Southern Hemisphere (Chambers et al. 2013)
due to climatic changes. Several studies demonstrate signif-
icant advancements in phenological phases of plants across
Europe (Menzel and Fabian 1999; Chmielewski and Rotzer
2001;Schleipetal.2009). These changes in central Eastern
Europe have so far been documented to be less marked than
in western and central Europe (Askeyev et al. 2010).
Climate factors, phenophases and their timing play the
most important role in such changes. The main causes
depend on the climatic region(s) from Mediterranean to
high latitudes. The most influential variables are tempera-
ture (Rutishauser et al. 2009), precipitation (Penuelas et al.
2004), photoperiod (K¨
orner and Basler 2010), the North
Atlantic Oscillation (Scheifinger et al. 2002), as well as cold
or warm spells (Menzel et al. 2011) and edaphic factors
(Wielgolaski 2001).
Biogeographical regions are useful geographical refer-
ence units when describing habitat types and species living
under similar conditions in different countries (Roekaerts
2002). The establishment of plant phenology across regions
of Europe is a first important step towards providing a gen-
eral overview, still covering a wide spatial window. The
purpose of this study was (i) to compare different biogeo-
graphical regions (Boreal, Continental, Alpine, Pannonian
and Mediterranean), and test whether the areas experienced
any trends in flowering time, (ii) to evaluate the possible fac-
tors that influence phenological shifts and (iii) to discover
Fig. 1 Locations (dots) of the 963 phenological stations in certain bio-
geographical regions of Europe along a North–South transect where
phenological records have been collected
Author's personal copy

Int J Biometeorol
Table 1 Phenological networks along a North to South transect in Europe, maintaining observations and provided data by the members of the
NS-Pheno Team
Country Observational network Reference
Finland National Phenological Network Kubin et al. (2007)
Estonia Estonian Naturalists Society, Estonian Environment Agency Ahas and Aasa (2006)
Latvia Volunteer collected sites Grisule and Briede (2008)
Lithuania Vo k e B r a n c h o f t h e L i t h u a n i a n Romanovskaja and Baksiene (2008)
Research Centre for Agriculture and Forestry
Poland Institute of Meteorology and Water Management Nied`
zwied`
zandJatczak(2008)
Slovakia Slovak Hydrometeorological Institute Remisov`
a and Nejedlik (2008)
Hungary Hungarian Meteorological Service Szalai et al. (2008)
Slovenia Environmental Agency of the Republic of Slovenia Crepinsek et al. (2008)
Croatia Meteorological and Hydrological Service Vuc e ti c e t al . ( 2008)
Bosnia and Hercegovina National Phenological Network Hodzic and Voljevica (2008)
Montenegro National Phenological Network Popovic and Drljevic (2008)
Macedonia National Phenological Network Nekov`
ar et al. (2008)
phenological patterns along various latitudes and periods
(1970–1980, 1981–1990, 1991–2000 vs 2001–2010).
Materials and methods
Phenological data of plants
Phenological data we analysed were collected from 12
countries (Finland, Estonia, Latvia, Lithuania, Poland, Slo-
vakia, Hungary, Slovenia, Croatia, Bosnia and Herzegovina,
Montenegro, Macedonia) in northern to eastern Central
Europe for the period 1970–2010 (Fig. 1). The data com-
prise phenological records on the first flowering date of
six plant species: lily of the valley (Convallaria majalis
L.), common dandelion (Taraxacum officinale L.), common
lilac (Syringa vulgaris L.), black elder (Sambucus nigra
L.), black locust (Robinia pseudoacaica L.) and small-
leaved lime (Tilia cordata Mill.). Even though the datasets
include observations originating from different phenologi-
cal networks (Table 1), the studied beginning of flowering
(BF) event was consequently defined as “the appearance
of the first flowers producing pollen on at least 10 %
of the observed plants visible”. This phenophase equals
the event 61, according to the BBCH (Biologische Bun-
desanstalt, Bundessortenamt and Chemical Industry) code
(see Meier 2001). The observations provided coverage for
twelve North- and East-Central European countries along
the geographical coordinates of 40.9–67.9 latitudes rang-
ing to the 13.6–32.1 longitudes (Fig. 1). The aim of the
phenological site selection was to provide the best temporal
and spatial coverage as possible. To reach this, the selec-
tion criteria were as follows: (1) the site has at least 10
years of continuous records; (2) there are at least five sites
within one biogeographical region. This criteria set resulted
in the North–South phenological (NS-Pheno) database (see
also Templ et al. 2016) that included different numbers of
observations per biogeographical region.
The indicative map of European Biogeographical
Regions was first defined in practice of the conservation
of natural habitats, wild fauna and flora (Roekaerts 2002;
ETCBD 2006). The dataset of biogeographical regions was
taken from the European Environment Agency web page.1
We merged these data sets with the phenological time
series in order to compare the following biogeographical
macroregions: Boreal, Continental, Alpine, Pannonian and
Mediterranean.
The Boreal region is the largest biogeographical region
of Europe. Its climate is cool and mainly continental,
its vegetation is dominated by coniferous forests, while
the biodiversity is relatively low. The Continental region
is characterised by clear continental climate, especially
across the central and eastern parts. Widespread grass-
lands are decreasing due to intensification of agriculture
and afforestation. The region shows increasing fragmenta-
tion of habitats due to dense and increasing infrastructure
within urban areas. The Alpine region is determined by
1http://www.eea.europa.eu/data-and-maps/data/biogeographical-
regions-europe-1
Author's personal copy

Int J Biometeorol
vertical zonality induced by the exposition of mountain
slopes and advecting air masses. In this way, different
ecological conditions are represented at different altitudes
resulting in various vegetation types. The Pannonian region,
situated in the lowland areas of the Carpathian Basin, used
to be dominated by a mosaic of deciduous forests and for-
est steppes, which are mostly turned into agricultural fields
by now. The Mediterranean region is characterised by a
climate where warm, moist winters alternate with hot, dry
summers. The region is dominated by evergreen forests and
shrublands.
Environmental data
Climatic variables based on air temperature (daily arith-
metic mean, minimum, maximum), precipitation and the
indices of the North Atlantic Oscillation (NAO) were
obtained from different databases. Daily data (January to
May) of air temperature and precipitation were used from
the E-OBS high-resolution gridded dataset developed by
the ENSEMBLES EU-FP6 project2with a 0.25◦spatial
resolution (Haylock et al. 2008).
As a descriptor of the frequency distribution, the quar-
tiles at 0.25 (Q.25), 0.5 (median), 0.75 (Q.75)levelandthe
skewness of climate data series were determined. Addition-
ally to raw climate data, the motivation behind using such
characteristics of predictors was to also take into account
the spread of values, in terms of the interquartile distance.
This is often done using the classical standard deviation.
However, since squared distances to the mean are taken into
account, outliers have a large influence on this estimate.
We created a set of environmental predictors (a) for
monthly temperature (namely, the 0.25, 0.5, 0.75 quartiles
and skewness of the minimum-, mean-, maximum- tem-
perature datasets) and (b) for precipitation (0.25, 0.5, 0.75
quartiles and skewness of the monthly precipitation).
Additionally, monthly indices (January to May) of the
NAO (Hurrel 1995) were used from the database provided
by the Climatic Research Unit (CRU) of the University of
East Anglia.
Furthermore, metadata information regarding the loca-
tions, namely latitude, longitude and altitude of phenolog-
ical sites, were also used in the models to consider spatial
differences.
Data analysis
Data pre-processing Dates of the phenological observations—
flowering data—were converted to days of the year (doy),
starting with first of January and considering leap years.
2http://www.ecad.eu/download/ensembles/ensembles.php
Each phenological station (shown in Fig. 1)wasassignedto
the closest grid cell.
Calculation of trends The obtained time series were
assigned to five biogeographical regions (Fig. 1), based
on the code list of the European Environmental Agency.
Accordingly, trend analyses were carried out on long-term
(1970–2010) data series of (a) monthly climate data (Fig. 2)
and (b) flowering onset (Fig. 3) for each biogeographical
region. We found that the data contain outliers, therefore a
robust regression method, namely MM-type estimators for
linear regression (see Maronna et al. 2006) were applied
to calculate trends. The reason for using this method was
that least squares estimates for regression models are highly
sensitive to outliers. Outliers are observations which do
not follow the pattern of the other observations. Robust
techniques reduce the influence of outliers (without remov-
ing them from the data series), but approximately give the
same results as if no outliers were presented in the dataset
(see more details in Todorov and Filzmoser 2009). Finally,
significant trends were found at the level of significance
p<0.05 using the Mann Kendall test (Mann 1945).
Comparison of decades In order to compare various
decades, phenological time series were divided into four
decadal-long periods: 1970–1980, 1981–1990, 1991–2000
and 2001–2010. As it was found to be a highly influential
factor, the differences in flowering onset dates (station-
wise) according to latitudes (N) were analysed. Trends in the
timing of flowering dates were illustrated with regression
lines using locally weighted scatterplot smoothing (loess)
(Cleveland 1979) for all decades (Fig. 4).
Influence of environmental variables To describe the
influence of the environmental variables on the timing of
flowering, we again used robust MM-type estimators for
linear regression (see Maronna et al. 2006)oneachplant
species for each biogeographical region. The difference
between the models was only given by the applied explana-
tory variables. Namely, the climatological data sets preced-
ing the timing of flowering and metadata information about
the station locations were included in the models, fitted by
the lmrob function of the R package robustbase (Rousseeuw
et al. 2015). For better interpretability, the predictors were
standardized to zero mean and unit variance. The predictor
expressing the first quartile of precipitation was excluded
from the models because these first quartiles were mostly
zero. The estimated regression coefficients obtained with
robust methods were visualized on heatmaps (Fig. 5). On
the heatmaps, we distinguish between cells including values
of corresponding regression coefficients and empty (white
or grey) cells. The colour key of the heatmaps expresses
the values of the regression coefficients. Namely, the darker
Author's personal copy

Int J Biometeorol
the colour the stronger the effect, which is either negative
or positive. Naturally, the maximum and minimum of the
coefficients vary depending on each heatmap. For better com-
parison, the colour range was restricted to −1and1,thus
any coefficient larger or smaller than this range was assigned
to black colour. Non-significant regression coefficients were
suppressed to reduce the amount of information to gain a
better overview about the important values. Thus, for any
empty white-coloured cell, the null hypothesis (regression
coefficient equals zero) cannot be rejected (no effect). The
empty grey-coloured cells report that no data were available
in some biogeographical regions for certain species.
All statistical analyses were performed using R (R Devel-
opment Core Team 2016).
Results
Trends in climatic variables
Regarding the climatological variables, we found that the
monthly mean (Fig. 2), minimum, maximum tempera-
tures preceding the flowering onset dates showed signif-
icant warming trends (1970–2010) across the Alpine and
Pannonian
Continental Mediterranean
Alpine Boreal
1970 1980 1990 2000 2010
1970 1980 1990 2000 2010 1970 1980 1990 2000 2010
1970 1980 1990 2000 2010 1970 1980 1990 2000 2010
−10
0
10
−10
0
10
−10
0
10
mean temperature [ °C]
Month jan feb mar apr may Significance p < 0.05 p > 0.05
Fig. 2 Annual variation and trends (1970–2010) of monthly mean temperature in different biogeographical regions of Europe. Solid lines report
significant trends
Author's personal copy

Int J Biometeorol
Continental regions calculated by the Mann-Kendall trend
test. Temperature has been increased significantly during
April (1970–2010) across the Mediterranean and the Pan-
nonian region. Over the studied 41 years, the Boreal region
did not show significant changes in temperature. Further-
more, we did not detect any significant long-term changes
in case of precipitation and NAO.
Temporal characteristics of flowering
According to climatological trends, this section provides
an overview about flowering trends (1970–2010) over
biogeographical regions of Europe, with special inter-
est on the North–South transect, drawn by latitude. As
expected, the flowering time starts earlier across the warmer
Mediterranean and Pannonian regions, i.e. along the lower
latitudes, while it starts later across the cooler Boreal and
Alpine regions. From 23 studied cases, 17 showed signif-
icant flowering trends (Fig. 3). All of these phenological
changes were related to earlier appearance—indicated by
negative regression coefficients (Table 2). Most species
showed significant trends in the Continental and Alpine
regions (Table 2), according to significant temperature
increase (Fig. 2). Less significant phenological shifts were
found across the Pannonian region. However, data availabil-
ity does not allow us to give such general statements about
phenological changes in the Mediterranean region. Still it
is noticeable that all coefficients were negative (except for
S. vulgaris in the Boreal region and C. majalis in the Pan-
nonian region) indicating advancements in flowering time
Syringa vulgaris Taraxacum officinale Tilia cordata
Convallaria majalis Robinia pseudoacacia Sambucus nigra
1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010
100
150
200
100
150
200
flowering date [doy]
Alpine Boreal Continental Mediterranean Pannonian significant trend p < 0.05 p > 0.05
Fig. 3 Inter-annual variation in the timing of flowering onset in different biogeographical regions of Europe (1970–2010). Solid lines report
significant trends
Author's personal copy

Int J Biometeorol
(Table 2). The strongest advancement was found for R.
pseudoacacia (3.8–9.6 days earlier shift per decade) and T.
cordata (3.2–9.5 days per decade), while less pronounced
responses were given by the herbaceous T. officinale and C.
majalis (Table 2).
Furthermore, we detected differences in the mean flow-
ering onset date between decades (Fig. 4). Differences were
detected in mean flowering date (doy) along a North–South
transect as we evaluated the time series over different time
periods. Accordingly, the flowering of different species
generally starts earlier in the latest period (1991–2010)
compared to the earlier years (1970–1990). This applies
especially to R. pseudoacacia, S. vulgaris and S. nigra
over the whole range of latitudes. For other species, like
T. officinale, this is only true along some ranges of lati-
tudes, especially in the southern part of Northern Europe.
However, because of missing data problems (no values for
the North during the period 2001–2010) not much can be
concluded from the current database for T. cordata.
Spatial patterns in flowering phenology across Europe
In order to explain the causes of phenological changes, the
effect of climatic variables and geographical information
on flowering dates were analysed (1970–2010). In Fig. 5,
we illustrate the robust regression coefficients for the six
species in each biogeographical region of Europe using
heatmaps. In most cases, the effects of latitude and altitude
were significantly positive. Thus, the species living in north-
ern or higher habitats were characterised by later dates of
flowering onset (see Fig. 4).
On the contrary, the effect of longitude was rather nega-
tive or non-explainable (70 % of the cases). This indicates
that the species living in more eastern parts of Europe were
characterised by earlier dates of flowering.
Regarding climatological variables, the clearest pattern
arises from the effect of NAO. All species revealed a signif-
icantly negative relationship to the index of NAO (Fig. 5).
From our study, the effect of mean temperature seems
Syringa vulgaris Taraxacum officinale Tilia cordata
Convallaria majalis Robinia pseudoacacia Sambucus nigra
40 50 60 40 50 60 40 50 60
100
150
200
100
150
200
latitude
flowering date [doy]
1970−1980 1981−1990 1991−2000 2001−2010
Fig. 4 Flowering onset dates (station-wise) against latitude over different decades (1970–1980, 1981–1990, 1991–2000, 2001–2010) and
corresponding regression lines (loess)
Author's personal copy

Int J Biometeorol
0.5 0.4
0.5 0.2
−0.1
−0.1 −0.1
1
0.1 0.1
0.3
0.4 −0.6
0.1 −0.4
0.1
−0.4 0.4
1
−0.1
−0.2 −0.3
0.1 −0.5
0.4 0.5 0.8 0.4 0.8
0.6 0.1 0.9 0.7 −1
−0.1 −0.2 −1
−0.1 −0.1 −0.1
−1 1
0.1
0.1
−0.1
−0.2 0.2 0.6 0.1 −0.7
0.4 0.3 −0.4 −0.2
0.1 −0.2 −0.8 −0.2 −0.1
0.1 0.1
0.1 −0.6 0.2 0.5
−0.4 −0.3 0.2 0.4
−0.3 0.3 0.4 0.4 0.2
−0.1 −0.1 0.1 0.1
0.3 −0.1 −0.2
0.1 0.4 0.2
0.2 −0.1 −0.3 −0.3 −0.2
0.1 0.1
0.4 0.6 0.5
0.3 0.1 0.2
−0.1 0.1
−0.1 −0.1 −0.1
−1
0.1
0.1 0.1 0.1
−0.1
0.2 −0.2
0.4 −0.1 −0.7
0.1
−0.1
−0.2 0.5
−0.4 −0.3 0.2
−0.1 0.1 0.6
−0.1 −0.1 0.1
0.2 −0.1 −0.4
0.1 0.5
−0.1 −0.7
0.3 0.3 0.1 0.2 0.6
10.50.91 −1
−0.1 −0.1 −1
−0.1 −0.1 −0.1
11
−0.1 0.2 −0.1 −0.9
0.1 −0.4 −0.3 −0.3
−0.2 −0.2 −0.1 0.1
0.1 0.1 0.1 0.2 0.7
−0.1 0.1 0.2 0.4
−0.2 0.6 0.4
0.1
0.2 −0.1 −0.2 −0.3
−0.3 −0.3 0.2 0.1
0.2 −0.2 −0.2
0.2 0.1 0.5 0.3
0.8 0.6 0.9 0.9
−0.2 −0.1
−0.1 −0.1 −0.1
0.7
0.1
0.1 0.1 0.1
−0.1
0.1 0.1
−0.3 −0.8 0.3 −0.3
0.4 −0.2 −0.1 0.2
−0.1
−0.2 0.4 0.1 0.2
0.3 0.6 −0.1 0.1
−0.2 0.8 −0.1 0.2
0.1 0.1
−0.2 −0.4 −0.2 −0.2
−0.1 −0.2 0.1
0.1 −0.4 0.1 −0.2
−0.1 −0.1
0.4 0.6 0.2 0.5
0.8 0.6 0.9 0.6
−0.1 −0.1 −0.2
−0.1
1−1
0.1 0.1
−0.1 0.1 0.5 −0.1
0.3 −0.1 0.2 −0.4
0.1 −0.2 −0.3 0.1
0.1 0.1 −0.2 0.2
−0.3 −0.2 −0.1 0.4
−0.4 0.2
−0.1 0.1
0.1 −0.1 −0.1 −0.1
−0.1 0.3 0.1 0.1
0.1 −0.4 −0.2
0.1
Convallaria m. Robinia p. Sambucus n. Syringa v. Taraxacum o. Tilia c.
Alpine
Pannonian
Boreal
Continental
Mediterranean
Alpine
Pannonian
Boreal
Continental
Mediterranean
Alpine
Pannonian
Boreal
Continental
Mediterranean
Alpine
Pannonian
Boreal
Continental
Mediterranean
Alpine
Pannonian
Boreal
Continental
Mediterranean
Alpine
Pannonian
Boreal
Continental
Mediterranean
altitude
latitude (N)
longitude (E)
NAO
precipitation (Q.25)
precipitation (median)
precipitation (Q.75)
precipitation (skewness)
Tmax (Q.25)
Tmax (median)
Tmax (Q.75)
Tmax (skewness)
Tmean (Q.25)
Tmean (median)
Tmean (Q.75)
Tmean (skewness)
Tmin (Q.25)
Tmin (median)
Tmin (Q.75)
Tmin (skewness)
−1.0−0.5 0.0 0.5 1.0
Regression coefficients
Fig. 5 Regression coefficients given for each plant and explanatory
variable within di fferent biogeographical regions of Europe. Cases of
grey cells without values report that no data were available in those
biogeographical regions. White cells indicate that non-significant
influence was found. Negative values indicate negative influence
on flowering time (i.e. advancement), while positive values express
positive effect
to give the most vague information. Earlier flowering in
response to increased temperatures is better visible when
looking at the minimum and maximum temperatures and
thus it is easier to interpret the flowering dates with the
distribution of minimum and maximum temperatures. It can
be seen that in most cases, the quartiles of the minimum
temperature have negative effect on the timing of flowering.
Namely, the higher the minimum temperature the earlier the
flowering time—except in the Alpine region. The second
(median) and third quartiles of the maximum temperature
Table 2 Coefficients of robust
linear regression between
flowering onset and years
(1970–2010) for the studied
species and biogeographical
regions in Europe
Species Alpine Boreal Cont. Mediter. Pannonian
Convallaria majalis −0.20 0.07
Robinia pseudoacacia −0.38 −0.96 −0.69 −0.17 −0.18
Sambucus nigra −0.35 −0.30 −0.18
Syringa vulgaris −0.39 0.03 −0.55 −0.24 −0.08
Taraxacum officinale −0.16 −0.22 −0.83 −0.16
Tilia cordata −0.32 −0.95 −0.52 −0.01
Negative values: indicate advancement in the beginning of flowering. Significant relationships (p<0.05)
are visualized in bold
Author's personal copy

Int J Biometeorol
distribution also show negative effects (again except for the
Alpine region). The effect of precipitation did not show a
general pattern among species and biogeographical regions,
but about half of the the studied cases indicated significant
influence.
Discussion
The term of phenological pattern has mainly been associ-
ated with plant communities (e.g. Pilar and Gabriel 1998;
Mart`
ınkov`
aetal.2002); in other cases, areas at different
scales were compared to describe phenological patterns of
areas. Studies have described phenological changes in tim-
ing of various spring plant phenophases across hemispheres
(Schwartz et al. 2006;Chambersetal.2013), continents
(Menzel and Fabian 1999), along countries (Ahas and Aasa
2006; Kalvane et al. 2009;Szab
`
oetal.2016) and zones
(Karyieva et al. 2012)relatedtoclimatedrivenmechanisms
and recent human induced climatic changes. Phenological
events of plants across biogeographical regions are par-
ticularly poorly documented, except the efforts done by
Rodriguez-Galiano et al. (2015)andTempletal.(2016).
We aimed to discover phenological patterns across bio-
geographical regions of Europe between various time win-
dows of the period 1970–2010. Besides the established
NS-Pheno database, the novelty in our study is that we
described flowering patterns along a 3000-km-long North–
South transect of Europe. Our NS-Pheno database allowed
us to test, if the species living in the northern latitudes show
more pronounced response to climate change than those
living in southern biogeographical regions. In the Boreal
region, the intensity of the phenological response to temper-
ature increases from South to North across Finland (Pudas
et al. 2008). The reason is that the observed (1847-2013)
warming in Finland is almost twice as high as the global
temperature increase (Mikkonen et al. 2015). Our results
seem to be contradictory to the findings of Mikkonen et al.
(2015), because we did not show significant temperature
increase in this region (see Fig. 2.). In contrast to Mikkonen
et al. (2015), we studied shorter time series (1970–2010)
and applied robust methods, which both may have influ-
ence on the results. Similarly to Lappalainen et al. (2008),
we experienced the most pronounced advancing flowering
trends (Table 1) at a species level within this subarctic
climate zone. There is also a known phenological sensitiv-
ity of species in other boreal countries such as in Estonia
(Ahas and Aasa 2006), Latvia and Lithuania (Kalvane et al.
2009). Furthermore, our results confirm the observed cli-
mate variability patterns and trends in the Alpine region
(Auer et al. 2007;Gobietetal.2014). Namely, the region
has been facing a significant temperature increase (Fig. 2),
which resulted in significant phenological shifts in the area
(see Table 2). As we move from the northern areas to tem-
perate and cool climate zones, phenological responses of
plants to warmer environment are strong (Menzel et al.
2006;JatczakandWalawender2009). The territory of Hun-
gary covers 80–85 % of the drier Pannonian region. For this
region, Szab`
oetal.(2016) already showed that plant species
advanced their flowering time (1952–2000) by 1.9–4.4 days
per decade. This tendency is confirmed in our study, but
only for two out of five species significantly (Table 2),
which can originate from the different lengths of the study
periods.
It is known that the annual pattern of phenological sea-
sons across Europe is related to the North Atlantic Oscilla-
tion (Menzel et al. 2005). Similarly to most of the studies,
we have also found temperature to be an influential deter-
minant for the timing of flowering (Stocker et al. 2013).
Our results highlighted that not just the mean temperature
but the distribution of minimum and maximum temperatures
are reasonable to consider as explanatory variables when
explaining flowering times. The importance of rainfall and
water availability is pronounced by Penuelas et al. (2004)as
complex drivers of phenological shifts. We showed (Fig. 5)
a significant influence of precipitation on the beginning of
flowering in approximately half of the studied cases.
Our main focus was not only to test the effect of climatic
variables, but also others such as latitude. We addressed
the question: Which patterns can we draw when comparing
northern biogeographical regions to southern ones? Are they
similiar to the patterns shown for the territory of China (Ge
et al. 2015) and the findings of a meta-analysis conducted
by Root et al. (2003)?
It is known that over the past half century the temperature
along higher latitudes has increased more than along lower
latitudes (Stocker et al. 2013). Accordingly, Root et al.
(2003) showed that the estimated phenological shifts from
32.0 N to 49.9 N latitude are smaller than between the 50.0
N to 72.0 N latitude band. Our observations confirm these
findings for Europe, since we noticed the most significant
plant responses within the Boreal biogeographical region
(approximately between 54.0 N and 67.0 N), which was fol-
lowed by the Continental and Alpine regions (from 40.0 N
in 55.0 N). But, is only the latitude responsible for this pat-
tern? Ge et al. (2015) investigated the 20.0–50.0 latitudes
in China and found significant phenological advancements;
however, they could only explain 9% of the overall vari-
ance in spring phenological trends. Previously, Estrella et al.
(2009) stated that the geographic coordinates (latitude and
longitude) have only a modest influence on the mean onset
of the groups of phenophases; however, inclusion of alti-
tude can improve models for some cases. In our study, not
only the effects of latitude (Fig. 4), but also altitude were
found to have a significantly positive effect on the beginning
of flowering (Fig. 5). These findings indicate that although
Author's personal copy

Int J Biometeorol
we experience a similar pattern (stronger response at higher
latitudes) among continents, the drivers of these changes
cannot be described simply. We showed that among bio-
geographical regions of Europe, the effect of longitude was
mostly non-significant. This can be explained by the lon-
gitudinal extent of our study window, which probably was
too narrow (13.6–32.1 longitudes) to show any West–East
oriented flowering pattern. Therefore, our results cannot
support the findings of Askeyev et al. (2010) who demon-
strated less marked phenological changes at the eastern edge
of Europe.
According to our results, more pronounced phenological
changes occur in the latest (1991–2010) than in the earliest
(1970–1990) study periods, as the effect of climate change
is more and more influential since the industrial era (Stocker
et al. 2013).
Acknowledgments We ac k no wle d ge th e E -O B S dat a set f rom t h e
EU-FP6 project ENSEMBLES (http://ensembles-eu.metoffice.com)
and the data providers in the ECA&D project (http://www.ecad.eu).
We are very grateful to all institutes and scientists who provided
data for the North–South phenological database. We would especially
like to emphasize our gratefulness to those data contributors who
did not participate as authors in the writing of this manuscript: K.
Jatczak (Centre for Poland`
s Climate Monitoring), P. Nejedlik (Slovak
Hydrometeorological Institute), T. Nied`
zwied`
z (University of Silesia),
T. Popovic (Hydrometeorological Institute of Montenegro), H. Simola
(Finnish Meteorological Institute), Z. Snopkov`
a(SlovakHydrometeo-
rological Institute) and S. Stevkova (Hydrometeorological Institute of
Macedonia). Additionally, we would like to pay respect to J. Terhivuo
(Finnish Museum of Natural History) who unfortunately could not
see these results published. And finally, thanks to F. Szentkir´
alyi for
inspiritaion during the project planning.
References
Aguilera F, Fornaciari M, Ruiz-Valenzuela L, Gal`
an C, Msallem M,
Dhiab AB, D`
ıaz-de la Guardia C, del Mar Trigo M, Bonofiglio
T, Orlandi F (2015) Phenological models to predict the main
flowering phases of olive (Olea europaea L.) along a latitudinal
and longitudinal gradient across the Mediterranean region. Int J
Biometeorol 95:629–641
Ahas R, Aasa A (2006) The effects of climate change on the phe-
nology of selected Estonian plant, bird and fish populations. Int J
Biometeorol 51:17–26
Askeyev OV, Sparks TH, Askeyev IV, Tishin DV, Tryjanowski P
(2010) East versus West: contrasts in phenological patterns? Glob
Ecol Biogeogr 19:783–793
Auer I, B¨
ohm R, Jurkovic A, Lipa W, Orlik A, Potzmann R, Sch¨
oner
W, Ungersb¨
ock M, Matulla C, Briffa K, Jones P, Efthymiadis D,
Brunetti M, Nanni T, Maugeri M, Mercalli L, Mestre O, Moisselin
JM, Begert M, M¨
uller-Westermeier G, Kveton V, Bochnicek O,
Stastny P, Lapin M, Szalai S, Szentimrey T, Cegnar T, Dolinar M,
Gajic-Capka M, Zaninovic K, Majstorovic Z, Nieplova E (2007)
Histalp—historical instrumental climatological surface time series
of the Greater Alpine Region. Int J Climatol 27:17–46
Chambers LE, Altwegg R, Barbraud C, Barnard P, Beaumont LJ,
Crawford RJM, Durant JM, Hughes L, Keatley MR, Low M,
Morellato PC, Poloczanska ES, Ruoppolo V, Vanstreels RET,
Woe h ler E J , Wol f aar d t AC ( 201 3) Ph e nol ogi c al ch a ng e s in th e
southern hemisphere. PLoS ONE e75514:8
Chmielewski FM, Rotzer T (2001) Response of tree phenology to
climate change across Europe. Agric For Meteorol 108:101–112
Cleland E, Chuine I, Menzel A, Mooney HA, Schwartz MD (2007)
Shifting plant phenology in response to global change. Trends
Ecol Evol 22:357–365
Cleveland W (1979) Robust locally weighted regression and smooth-
ing scatterplots. J Am Stat Assoc 74:829–836
Crepinsek Z, Zrnec C, Susnik A, Zust A (2008) Slovenian phenolog-
ical observations. In: Nekovar J, Koch E, Kubin E, Nejedlik P,
Sparks T, Wielgolaski FE (eds) COST Action 725—the history
and current status of plant phenology in Europe. COST Office,
Brussels
Demar`
ee GR, Rutishauser T (2009) Origins of the word “Phenology”.
EOS Trans Am Geophys Union 90:291–291
Estrella N, Sparks T, Menzel A (2009) Effects of temperature, phase
type and timing, location, and human density on plant phenologi-
cal responses in Europe. Clim Res 39:235–248
ETCBD (2006) The indicative map of European biogeographical
regions: methodology and development. Technical report Euro-
pean Topic Centre on Biological Diversity. National Museum of
Natural History, France
Franks SJ (2015) The unique and multifaceted importance of the
timing of flowering. Am J Bot 102:1401–1402
Ge Q, Wang H, Rutishauser T, Dai J (2015) Phenological response
to climate change in China: a meta-analysis. Glob Chang Biol
21:265–274
Gobiet A, Kotlarski K, Beniston M, Heinrich G, Rajczak J, Stoffel
M(2014)21stcenturyclimatechangeintheEuropeanAlps—a
review. Sci Total Environ 493:1138–1151
Grisule G, Briede E (2008) The history and current status of phenolog-
ical recording in Latvia. In: Nekovar J, Koch E, Kubin E, Nejedlik
P, Sparks T, Wielgolaski FE (eds) COST Action 725 – the history
and current status of plant phenology in Europe. COST Office,
Brussels
Haylock MR, Hofstra N, Klein Tank AMG, Klok EJ, Jones PD, New
M(2008)AEuropeandailyhigh-resolutiongriddeddatasetof
surface temperature and precipitation for 1950–2006. J Geophys
Res - Atmos 113, doi:10.1029/2008JD010201
Hodzic S, Voljevica A (2008) Phenology in Bosnia and Hercegovina.
In: Nekovar J, Koch E, Kubin E, Nejedlik P, Sparks T, Wielgolaski
FE (eds) COST Action 725—the history and current status of plant
phenology in Europe. COST Office, Brussels
Hurrel JW (1995) Decadal trends in the North Atlantic Oscillation
regional temperatures and precipitation. Science 269:676–679
IPCC (2007) Technical summary. Cambridge University Press,
Cambridge
Jatczak K, Walawender J (2009) Average rate of phenological
changes in Poland according to climatic changes. Adv Sci Res 3:
127–131
Kalvane G, Romanovskaja D, Briede A, Baksiene E (2009) Influence
of climate change on phenological phases in Latvia and Lithuania.
Clim Res 39:209–210
Karyieva J, van Leeuven W, Woodhouse C (2012) Impacts of climate
gradients on the vegetation phenology of major land use types in
Central Asia (1981–2008). Front Earth Sci 6:206–225
Author's personal copy

Int J Biometeorol
Koch E, Dittmann E, Lippa W, Menzel A, Nekovar J, Sparks T, van
Vliet AJH (2009) COST725—Establishing a European phenolog-
ical data platform for climatological applications: major results.
Adv Sci Res 3:119–122
K¨
orner C, Basler D (2010) Phenology under global warming. Science
327:1461–1462
Kubin E, Kotilainen E, Poikolainen J, Hokkanen T, Nevalainen
S, Pouttu A, Karhu J, Pasanen J (2007) Monitoring instruc-
tions of the finnish national phenological network. Technical
report Finnish Forest Research Institute, Muhos Research Unit,
Metla
Lappalainen H, Linkosalo T, Venalainen A (2008) Long-term trends
in spring phenology in a boreal forest in central Finland. Boreal
Environ Res 13:303–318
Mann HB (1945) Nonparametric tests against trend. Econometrica
13:245–259
Maronna R, Martin R, Yohai V (2006) Robust statistics: theory and
methods. Wiley, New York. ISBN:978-0-470-01092-1
Mart`
ınkov`
aJ,SmilauerP,MihulkaS(2002)Phenologicalpatternof
grassland species: relation to the ecological and morphological
traits. Flora 197:290–302
Meier U (2001) BBCH-Monograph: growth stages of mono- and
dicotyledonous plants. Technical Report, 2 Edn. Federal Biologi-
cal Research Centre for Agriculture and Forestry
Menzel A, Fabian P (1999) Growing season extended in Europe.
Nature 397:659
Menzel A, Seifert H, Estrella N (2005) SSW to NNE—North Atlantic
Oscillation affects the progress of seasons across Europe. Glob
Chang Biol 11:909–918
Menzel A, Seifert H, Estrella N (2011) Effects of recent warm and cold
spells on European plant phenology. Int J Biometeorol 55:921–
932
Menzel A, Sparks TH, Estrella N, Koch E, Aasa A, Ahas R,
Alm-Kuebler K, Bissollii P, Braslavska O, Briede A, Chmielewski
FM, Crepinsek Z, Curnel Y, Dahl A, Defila C, Donelly A, Filella
Y, Jatczak K, Mage F, Mestre A, Nordli A, Penuelas J, Pirinen P,
Remisova V, Scheifinger H, Striz M, Susnik A, Van Vliet AJH,
Wiel golaski FE, Zac h S, Zu st A (2006) Europ ean p henologi cal
response to climate change matches the warming pattern. Glob
Chang Biol 12:1969–1976
Mikkonen S, Laine M, M¨
akel¨
aH,GregowH,TuomenvirtaH,
Lahtinen M, Laaksonen A (2015) Trends in the average temper-
ature in Finland, 1847–2013. Stoch Env Res Risk A 29:1521–
1529
Nekov`
ar J, Koch E, Kubin E, Nejedlik P, Sparks T, Wielgolaski
FE (2008) COST ACtion 725—The history and current status
of plant phenology in Europe Technical report. COST Office,
Brussels
Nied`
zwied`
zT,JatczakK(2008)HistoryofphenologyinPoland.In:
Nekovar J, Koch E, Kubin E, Nejedlik P, Sparks T, Wielgolaski
FE (eds) COST Action 725—the history and current status of plant
phenology in Europe. COST Office, Brussels
Parmesan C (2006) Ecological and evolutionary responses to recent
climate change. Annual review of ecology. Evol Syst 37:637–
669
Penuelas J, Filella I, Zhang X, Llorens L, Ogaya R, Lloret F, Comas
P, Estiarte M, Terradas J (2004) Complex spatiotemporal pheno-
logical shifts as a response to rainfall changes. New Phytol 161:
837–846
Pilar C, Gabriel M (1998) Phenological pattern of fifteen Mediter-
ranean phanaerophytes from Quercus ilex communities of NE-
Spain. Plant Ecol 139:103–112
Popovic T, Drljevic M (2008) History and current status of phenol-
ogy in Montenegro. In: Nekovar J, Koch E, Kubin E, Nejedlik P,
Sparks T, Wielgolaski FE (eds) COST Action 725—the history
and current status of plant phenology in Europe. COST Office,
Brussels
Pudas E, Leppala M, Tolvanen A, Poikolainen J, Venalainen A, Kubin
E(2008)TrendsinphenologyofBetulapubescensacrossthe
boreal zone in Finland. Int J Biometeorol 52:251–259
RDevelopmentCoreTeam(2016)R:alanguageandenvironment
for statistical computing r foundation for statistical computing.
Vie nn a, A ustria. http://www.R-project.org
Remisov`
aT,NejedlikP(2008)Historyandpresentobservationsin
Slovak plant phenology. In: Nekovar J, Koch E, Kubin E, Nejedlik
P, Sparks T, Wielgolaski FE (eds) COST Action 725—the history
and current status of plant phenology in Europe. COST Office,
Brussels
Rodriguez-Galiano V, Dash J, Atkinson P (2015) Characterising the
land surface phenology of Europe using decadal MERIS data.
Remote Sens 7:9390–9409
Roekaerts M (2002) The biogeographical regions map of Europe:
basic principles of its creation and overview of its development
Tec hni ca l r ep ort . Eur op ean To pi c C en tre N atu re P rot ect io n and
Biodiversity, European Environment Agency
Romanovskaja D, Baksiene ENekovar J, Koch E, Kubin E, Nejedlik P,
Sparks T, Wielgolaski FE (eds) (2008) Phenological investigations
in Lithuania. COST Office, Brussels
Root TL, Price JT, Hall KR, Schneider SH, Rosenzweig C, Pounds JA
(2003) Fingerprints of global warming on wild animals and plants.
Nature 421:57–60
Rousseeuw P, Croux C, Todorov V, Ruckstuhl A, Salibian-Barrera
M, Verbeke T, Koller M, Maechler M (2015) Robustbase: basic
Robust Statistics. R Foundation for Statistical Computing, pack-
age version 0.92-3
Rutishauser T, Schleip C, Sparks TH, Nordli O, Menzel A, Wanner H,
Jeanneret F, Luterbacher J (2009) Temperature sensitivity of Swiss
and British plant phenology from 1753 to 1958. Clim Res 39:
179–190
Scheifinger H, Menzel A, Koch E, Peter C, Ahas R (2002) Atmo-
spheric mechanisms governing the spatial and temporal variability
of phenological phases in Central Europe. Int J Climatol 22:1739–
1755
Schleip C, Sparks TH, Estrella N, Menzel A (2009) Spatial variation
in onset dates and trends in phenology across Europe. Clim Res
39:249–260
Schwartz MD, Ahas R, Aasa A (2006) Onset of spring starting earlier
across the Northern Hemisphere. Glob Chang Biol 12:343–351
Stocker T, Qin D, Plattner GK, Tignor M, Allen SK, Boschung J,
Nauels A, Xia Y, Bex V, Midgley PM (2013) Climate Change
2013: The Physical Science Basis. Contribution of Working Group
ItotheFifthAssessmentReportoftheIntergovernmentalPanel
on Climate Change. Cambridge University Press, Cambridge
Szab`
o B, Vincze E, Cz`
ucz B(2016)Floweringphenologicalchangesin
relation to climate change in Hungary. Int J Biometeorol 60:1347–
1356
Szalai S, Bella S, N`
emeth A, Dunay S (2008) History of Hungar-
ian phenological observations. In: Nekovar J, Koch E, Kubin E,
Nejedlik P, Sparks T, Wielgolaski F (eds) COST Action 725—
the history and current status of plant phenology in Europe. COST
Office, Brussels
Tem pl B, F le c k s, Te mpl M ( 201 6) C han ge of p la n t ph eno pha se s
explained by survival modelling. Int J Biometeorol, doi:10.1007/
s00484-016-1267-z
Author's personal copy

Int J Biometeorol
Todorov V, Filzmoser P (2009) Multivariate robust statistics:
methods and computation. S¨
udwestdeutscher Verlag f ¨
ur
Hochschulschriften, Saarbr¨
ucken
Vuc e ti c V , Vuc e ti c M , Lo n ca r Z ( 200 8) Hi sto ry an d pre sen t o bs e rv a-
tions in Croatian plant phenology. In: Nekov `
ar J, Koch E, Kubin E,
Nejedlik P, Sparks T, Wielgolaski FE (eds) COST Action 725—
the history and current status of plant phenology in Europe. COST
Office, Brussels
Wal t her G R , Po s t E, Co nv ey P, M e nze l A, Pa rme s an C, B e eb e e TJC ,
Fromentin JM, Hoegh-Guldberg O, Bairlein F (2002) Ecological
responses to recent climate change. Nature 416:389–395
White MA, Hoffman F, Hargrove WW, Nemani RR (2005) A global
framework for monitoring phenological responses to climate
change. Geophys Res Lett 32:L04705
Wielgolaski F (2001) Phenological modifications in plants by various
edaphic factors. Int J Biometeorol 45:196–202
Wolkovich E, Cook B, Allen JM, Crimmins TM, Betancourt JL,
Travers SE, Pau S, R egetz J, Davies TJ, Kraft N JB, A ult
TR, Bolmgren K, Mazer SJ, McCabe GJ, McGill BJ, Parme-
san C, Salamin N, Schwartz MD, Cleland EE (2012) Warming
experiments underpredict plant phenological responses to climate
change. Nature 485:494–497
Author's personal copy