You are on page 1of 12

Int J Biometeorol (2011) 55:921932

DOI 10.1007/s00484-011-0466-x

ORIGINAL PAPER

Effects of recent warm and cold spells on European plant


phenology
Annette Menzel & Holm Seifert & Nicole Estrella

Received: 6 December 2010 / Revised: 10 June 2011 / Accepted: 10 June 2011 / Published online: 14 July 2011
# ISB 2011

Abstract Climate change is already altering the magnitude harvest of winter cereals were the best indicators of warm/
and/or frequency of extreme events which will in turn affect cold spells in early spring and summer, also being spatially
plant fitness more than any change in the average. Although coherent with the patterns of warm/cold spells.
the fingerprint of anthropogenic warming in recent pheno-
logical records is well understood, the impacts of extreme Keywords Climate change . Extreme events . Heat waves .
events have been largely neglected. Thus, the temperature Temperature response
response of European phenological records to warm and
cold spells was studied using the COST725 database. We
restricted our analysis to the period 19512004 due to Introduction
better spatial coverage. Warm and cold spells were
identified using monthly mean ENSEMBLES temperature Existing or projected consequences of climate change
data on a 0.5 grid for Europe. Their phenological impact include alterations in the frequency, intensity, geographic
was assessed as anomalies from maps displaying mean scale, and location of extreme weather and climate events
onsets for 19301939. Our results clearly exhibit continen- (e.g. Horton et al. 2001, Trenberth et al. 2007). For many
tal cold spells predominating in the period 19511988, phenomena, such as warmer and more frequent hot days on
especially during the growing season, whereas the period land, trends in the twentieth century were significant, the
from 1989 onwards was mainly characterised by warm human contribution is likely and continued trends for the
spells in all seasons. The impacts of these warm/cold spells twenty-first century are virtually certain (Parry et al. 2007).
on the onset of phenological seasons differed strongly Although climate change is strongly linked to more and
depending on species, phase and timing. False phases stronger physical extreme events, not all will necessarily
such as the sowing of winter cereals hardly reacted to translate into extreme impacts since systems may be
summer warm/cold spells; only the sowing of summer naturally resilient or well adapted by management or
cereals mirrored spring temperature warm/cold spells. The special adaptation measures (e.g. Easterling et al. 2000;
heading dates of winter cereals did not reveal any consistent Parmesan et al. 2000; Huynen et al. 2001; Chau et al.
results probably due to fewer warm/cold spells identified in 2009). Consequently, the perspectives on extreme events
the relevant late spring months. Apple flowering and the vary broadly, from statistical definitions of measured
physical attributes of phenomena used by climatologists
A. Menzel (*) : N. Estrella (e.g. Easterling et al. 2000; Kioutsioukis et al. 2010; Katz
Chair of Ecoclimatology, Technische Universitt Mnchen,
2010) to impact related approaches in natural sciences (e.g.
Hans-Carl-von-Carlowitz-Platz 2,
85354 Freising, Germany Parry et al. 2007). Sometimes, both aspects are linked if an
e-mail: amenzel@wzw.tum.de extreme event is seen as a notable, rare or unique event or
otherwise significant, e.g. in terms of its impacts.
H. Seifert
Extremes are commonly defined on the basis of the tails
CAD/GIS-Lab, WZW, Technische Universitt Mnchen,
Weihenstephaner Berg 13, of distributions (e.g. the upper or lower 10th percentiles;
85354 Freising, Germany Beniston 2009), by exceedance of absolute thresholds (e.g.
922 Int J Biometeorol (2011) 55:921932

and winter rye (Secale cereale)


Table 1 Phenological seasons and their key phases according to Schnelles maps of 1965, the best corresponding phenological phases of the COST725 database and the climate triggers studied

Sowing/drilling of winter wheat


(Triticum aestivum L.), winter
Sowing of winter wheat/cereals
Satyamurty et al. 2007) or by measures of variation (e.g.

barley (Hordeum vulgare L.)


Jones et al. 1999; Horton et al. 2001; Katz 2010). Warm and

Later part of full autumn


cold spells are not single extreme events, but can be regarded
as a compound extreme, i.e. as a persistence of weather
conditions, comparable to drought. However, although the
term warm and cold spells is frequently used (e.g. Huynen

September
et al. 2001; Beniston 2005), there is no general definition of

230320

Autumn
October
August
it. Spells and their frequency, duration, or intensity are
defined variously in the literature (Jones et al. 1999; Horton

winter barley (Hordeum


et al. 2001; Shabbar and Bonsal 2003; Satyamurty et al.

Harvest of winter wheat/

(Triticum aestivum L.),

vulgare L.) and winter


Harvest of winter wheat

rye (Secale cereale)


2007; Beniston 2009; Klein Tank et al. 2009; Katz 2010).

End of full summer


Periods of at least six consecutive days with daily mean
temperatures exceeding either a percentile of the distribu-
tion or a baseline temperature have been suggested (e.g.

160250
cereals

Summer
Klein Tank et al. 2009), while other definitions use mean

May
June
July
durations of warm/cold spells longer than 11 days (e.g.
Satyamurty et al. 2007). In general, the concept of using

common oat (Avena sativa L.) L.), both early and late cultivars (Triticum aestivum L.) as
heading of winter wheat

well as of winter barley


Heading of winter wheat/

Beginning and middle of


relative temperature thresholds instead of absolute ones

Middle of early summer

(Hordeum vulgare L.)


seems widely accepted (e.g. McCalla et al. 1978).
In the past, climate change-driven variations in extreme
events were mainly studied in physical systems, and their
impacts on the natural environment were mostly neglected.

120190
cereals

Spring
March
Recent research suggests that extreme events have been

April
May
and, in the course of accelerated anthropogenic climate
change, will be the major driving factor influencing fitness

Sowing/drilling of spring barley First flowers and full flowering


(growth and survival), and thus distributional ranges and
of apple (Malus x domestica
sustainability of ecosystem services. In particular, the
effects of winter warm spells on dehardening and frost
resistance have been studied. However, the observed
Flowering of apple

responses differed with duration, timing and species,


ranging from a dramatic decrease in frost hardiness, with
Full spring

or without subsequent recovery, to no reaction (e.g. Nielsen


February
80170

Spring
March
and Rasmussen 2009). In general, the effects of extreme April
events are likely greater than those resulting from any
change in climate averages (e.g. Easterling et al. 2000;
phase in the COST725 data base (Hordeum vulgare L.) and

Jentsch et al. 2007; Zimmermann et al. 2009).


Sowing of summer cereals

Among ecological traits of interest is phenology, which


Earliest to early spring

tracks annually recurring events in ecosystems such as plant


germination, flowering, growth and fruit maturation. Since
these events are triggered predominantly by temperature,
phenology has emerged as a key asset in identifying current
February
January
50140

Winter
March

fingerprints of climate change in nature, especially since


recent warming is mirrored by significantly advancing
spring events of generally about 25 days decade1 in the
Key phenophase (Schnelle 1965)

northern hemisphere (e.g. Menzel and Fabian 1999; Walther


Mean onset in Europe (DOY)

Triggering mean monthly and


Corresponding phenological

et al. 2002; Root et al. 2003; Matsumoto et al. 2003;


Walther 2004; Menzel et al. 2006a). The climate signal
seasonal temperatures

controlling spring and summer phenology is fairly well


understood: nearly all phenophases correlate with temper-
atures in the preceding 13 months (Sparks et al. 2000;
Sparks and Menzel 2002; Menzel 2003; Cleland et al.
Season

2007). Globally, and at continental scales, the observed


changes can be attributed to human induced global
Int J Biometeorol (2011) 55:921932 923

warming (e.g. Root et al. 2003; Parmesan and Yohe 2003; certainly longer than a few days, and are apt to be reflected
Rosenzweig et al. 2007, 2008). in monthly and seasonal temperature anomalies. To classify
However, current phenological studies have predominantly warm and cold spells, we used quality-controlled, high-
addressed changes in mean or average onset dates and have resolution gridded (0.5) monthly and seasonal (DJF,
analysed their variability and temporal trends in the last five MAM, JJA, SON) temperature means provided by
decades. Only a few more recent publications have described ENSEMBLES (Ensembles-Based Predictions of Climate
the effects of naturally occurring or manipulated extremes on Changes and their Impactsa project funded by the
phenology (e.g. Luterbacher et al. 2007; Rutishauser et al. European Commission: EU FP6, http://www.ensembles-eu.
2008; Jentsch et al. 2009). The study of extreme phenological org, data provided through http://eca.knmi.nl). The geo-
events seems to be important since they may lead to critical graphic domain was Europe west of 40E between 36.25 and
disturbances and mismatches in ecosystems (Parmesan 2006). 70.75N and the study period was 19512006. We used the
In particular, the response of ecosystems is most sensitive to criteria of exceeding +1.5, +3, 1.5 and 3 standard
extreme events, not average conditions, and thus events, not deviations from the long-term mean at the respective grid
trends, matter (Jentsch et al. 2009; Zimmermann et al. 2009). point to classify warm and cold spells into four categories of
This study will not concentrate on whether extreme warm, very warm, cold and very cold, respectively. We
phenological events change in frequency and magnitude, but defined "continental" warm or cold spells as those in which
evaluates whether and to what extent recent warm and cold >40% of the grid points were classified as either warm or
spells in Europe from 1951 to 2006 have resulted in extreme cold, and "regional" warm or cold spells if >1% were
phenological onset dates. This analysis will also include classified as either very warm or very cold.
agricultural events, so-called false phenological phases, which
are driven by farm management and thus may clearly differ in Phenological data
their response from spring/summer events in wild nature
(Menzel et al. 2006b; Estrella et al. 2007). In this study, we used the phenological data collected by the
COST725 Action to assess the impacts of warm and cold spells
on European plant phenology (www.cost725.org). To date, 20
Materials and methods European countries contributed more than 7 million pheno-
logical data values to its database covering 64 species and 22
Classification of warm and cold spells different phases. We restricted our analysis to the time period
19512004 due to the relatively better spatial coverage. A
The warm and cold events likely to coherently influence constraint in the analyses was the lumped density of the
phenological responses at regional to continental scales are almost 8,000 observational sites across countries in Central

a b

Fig. 1 Standardised phenological reference data (mean onset dates in 19301939 according to Schnelle 1965). The example of full spring based on
apple flowering includes a the original published map and b the result from the automatic reclassification procedure developed in GIS
924 Int J Biometeorol (2011) 55:921932

100 20
'06A
Continental warm spells
1989_3 1990_2 1990_3
80 2000_4 2000_10 2002_2 16
'03E
2006_7 2006_9 2006_10
2006_12 1990W
2006A
60 Regional warm spells 12
'53_6 '60_12 '63_5 '72_8
% of grids cells exceeding 1.5 sd

'89_1 '89_4 '92_8 2005A

% of grid cells exceeding 3 sd


1989S 2000A
'01_3 '01_7 '02_1 '02_8
40 2003E 8
'03_6 '03_8 '05_11

20 4
'90W
'72E '89S
'02E
0 0
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
'54W '57W
'52W '77E '93A Regional cold spells
-20 '55S -4
'52_2 '56_2 '57_1
'63_1 '68_12
'74_10 '76_10 '77_7 '77_8
-40 '81_12 '96_3 -8
'63W 1976E
1952A Continental cold spells
1976A
1955S 1962E 1955_4 1956_8 1956_11
-60 1962_3 1986_9 1987_8 -12
% warm (>1.5 sd) [left y-axis] % cold (<-1.5 sd) [left y-axis] 1988_111992_10
1993A 1993_9 1993_11
% warm (>3 sd) [right y-axis] % cold (<-3 sd) [right y-axis]
-80 -16

Fig. 2 Seasonal warm and cold spells in Europe based on 19512006 deviations (negative = cold event, if >40% then named continental
ENSEMBLES temperature data (http://www.ensembles-eu.org, 0.5 warm/cold spell). If more than 1% of the grids cells exceeded 3
grid, limited to 40E). Curves depict the percentage of grid cells where standard deviations, the respective seasons were defined as a regional
the respective seasonal mean temperatures (W winter DJF, S spring warm/cold spell, marked with squares. Boxes additionally list all
MAM, E summer JJA, and A autumn SON) exceeded 1.5 standard monthly (continental and regional) warm and cold spells

and Eastern Europe with data from several countries lacking station density in Central Europe. They display mean starting
completely (e.g. UK, Italy) or being represented only by dates (for 19301939) of five phenological seasons from
stations from the network of the International Phenological early spring to autumn: (1) beginning of sowing of summer
Gardens (Menzel and Fabian 1999; Chmielewski and Rtzer cereals in early spring, (2) flowering of apple in full spring,
2001). In consequence, in those countries where national (3) heading of winter wheat in early summer, (4) harvest of
networks provided not only selected stations but their entire winter wheat at the end of full summer, and (5) sowing of
dataset, a huge concentration of stations was present, e.g. in winter wheat in later full autumn (see Table 1). When data
Germany, Austria, Switzerland, Czech Republic and Estonia. were available only from other periods, they were converted
In order to study the impacts of warm and cold spells on to the 1930s period. In regions or countries were the species/
European phenology, we used onset dates of phenological phases were not available, the maps were based on
seasons, such as early spring and full spring, for the large- corresponding, so-called substitute, phases, e.g. lilac flower-
scale comparison, since species and phases observed in the ing (Syringa vulgaris) instead of apple flowering (Malus
various national networks strongly differ. domestica). The isolines were drawn at 10-day intervals at a
scale of 1:5,000,000, later downsized to 1:7,500,000
Baseline of mean onsets of phenological seasons (Schnelle 1965; see also Fig. 1a).
In order to derive digital information from these maps,
Since the observational periods provided in the COST they were scanned into a rasterformat. For data processing,
database were not identical across countries and sometimes a highly novel procedure in Definiens Developer, accurate-
did not even overlap (e.g. Netherlands 18681921, Belgium ly defined with rulesets, was set up. First, three different
19431977), the baseline of mean onset dates of phenological image object levels were created with related segmentation
seasons was derived from contoured and digitised phenolog- algorithms based on the pixel level. Within these levels,
ical maps averaged between 1930 and 1939 (Schnelle 1965), a
quiescent period in terms of warm and cold spells. In Fig. 3 Examples of impacts of warm spells in Europe on phenological
seasons. a Advances of harvest dates of winter cereals (days) in the
contrast to newer European phenological maps (e.g. Rtzer course of the European heat wave in summer 2003 (temperature
and Chmielewski 2001; Menzel et al. 2005), the old maps by anomalies in C), and b advances in the dates of apple flowering (days)
Schnelle (1965) did not suffer from a limited and lumped after the warm spell in February 1990 (temperature anomalies in C)
Int J Biometeorol (2011) 55:921932 925

a
Crop harvest
-61 - -33
-32,99 - -18
-17,99 - -11
-10,99 - -6
-5,99 - 0
0,01 - 14

Summer
temperature
-2,0
-1,9 - -1,5
-1,4 - -1,0
-0,9 - -0,5
-0,4 - 0,0
0,1 - 0,5
0,6 - 1,0
1,1 - 1,5
1,6 - 2,0
2,1 - 2,5
2,6 - 3,0
3,1 - 3,5
3,6 - 4,0
4,1 - 4,5
4,6 - 5,0
5,1 - 5,5
5,6 - 6,0

b Apple flowering
-34 - -28
-27,9 - -18
-17,9- -12
-11,9 - -5
-4,9 - 13
13,1 - 32

February
temperature
-2,5
-2,4 - -2,0
-1,9 - -1,5
-1,4 - -1,0
-0,9 - -0,5
-0,4 - 0,0
0,1 - 0,5
0,6 - 1,0
1,1 - 1,5
1,6 - 2,0
2,1 - 2,5
2,6 - 3,0
3,1 - 3,5
3,6 - 4,0
4,1 - 4,5
4,6 - 5,0
5,1 - 5,5
5,6 - 6,0
6,1 - 6,5
926 Int J Biometeorol (2011) 55:921932

comprehensive and customised classification algorithms season (1952, 1954, 1957, 1963) and in winter months (e.g.
were used in a logical sequence to clearly separate the December 1968, 1981, January 1957, 1963, February 1952,
different classes. Finally, all background artefacts and 1956). In contrast, only a very few warm spells were
redundant mapping, such as the grid, altitudinal shading, revealed. Key examples of cold spells included February
and annotations, were eliminated. Afterwards, the different 1952 (up to 9C cooler in Western Europe) as well as winter
classes were exported as a rasterformat into an ArcGIS 9.3 1963 (up to 6C cooler in Western and Central Europe).
environment and transformed into vector data. Thus, we In contrast, the (almost) two last decades of our study
received digitised isoline information on phenological onset period (19892006) were mainly characterised by warm
dates as shape files. Finally, the maps were geo-referenced spells in all seasons, predominantly in summer and autumn,
in ArcGIS 9.3 and the onset dates were re-assembled on a and only a few months in autumn exhibited regional cold
0.5 grid. Figure 1b displays the digitised GIS information spells. Key examples included the famous Central European
of 19301939 mean onset dates of full spring. heat wave of summer 2003 (up to 5C warmer in Western and
The best corresponding species and phenological phases Central Europe) as well as February 1990 (up to 6C warmer
of COST725 (see Table 1) were carefully chosen as a in Central Europe and around the Baltic Sea).
compromise between accuracy provided by identical
species and phases and increased spatial and temporal Analysis of the phenological impacts of warm and cold
coverage granted by additional corresponding phases. spells
Triggering months and seasons were defined according to
mean onset dates in Europe. Graphs with isolines of temperature anomalies in warm and
cold spells as well as phenological anomalies from the mean
Assessment of impacts of warm and cold spells onset dates in the standard reference period 19301939 clearly
on phenological seasons demonstrated the close linkage of climate extremes and their
impacts on recurring plant events. Figure 3 displays two key
Departures of annual phenological onsets in 1951 to 2004 examples for warm spells, the advances of harvest dates of
from the 19301939 baselines were plotted against the winter cereals after the European heat wave in summer 2003,
monthly and seasonal temperature anomalies for the and advances in the dates of apple flowering after the warm
corresponding grid points, and expressed as confidence spell in February 1990.
ellipses for bivariate normally distributed data covering 90% We summarised the phenological impacts of warm and cold
of data points (based on the SAS macro ellipses www. spells by confidence ellipses covering 90% of the data points
datavis.ca/sasmac/ellipses.html). The total number of data of phenological impacts against temperature deviations in
points, depending on season and warm/cold spells, ranged warm and cold spells (Fig. 4). It was quite obvious that not all
from 16 to 6,009; however, 58 out of 72 ellipses each warm and cold spells were characterised by significant
represent more than 1,000 observations in Europe for a temperature deviations, shown by ellipses around the y-axis
specific warm/cold spell. Confidence ellipses may be used as indicating no or only small temperature anomalies. The
visual indicators of correlations, since they collapse diago- reason for this was that the revealed warm or cold spells
nally when there is a high correlation between two variables, were situated outside the spatial range covered by pheno-
whereas they are more circular when the variables are logical stations in the COST725 data base.
uncorrelated. We regard the amount of variance explained by Sowing of summer cereals in spring (Fig. 4a) can be
linear regressions of phenological impacts against tempera- regarded as a false phenological phase which mainly reflect
ture deviation (R2) as a measure of spatial coherence. farmers management decisions, and finally may or may not
be reflecting what is happening climatically or biologically
(see Menzel et al. 2006a, b; Estrella et al. 2007). Sowing of
Results summer cereals, however, seemed to successfully mirror
warm and cold spells of January to March and of winter. If
Warm and cold spells in Europe warm spells occurred in the phenological study area of
Fig. 4 Results of the impacts of warm and cold spells on
For the seasons in the study period 19512006, our phenological seasons in Europe (19512004). Graphs depict confi-
screening method identified 6 continental and 6 regional dence ellipses covering 90% of data of phenological anomalies (days)
warm spells as well as 6 continental and 7 regional cold against temperature anomalies (C), bold lines indicate spells that
spells (see Fig. 2). From 19511988, the continental cold occurred in Central and Eastern Europe. The warm/cold spells (c cold,
w warm) are ordered by year, month (1 January, 2 February, ..), and
spells predominated, especially in spring to autumn (e.g. season (13 MAM, 14 JJA, 15 SON, 16 DJF). a Sowing of summer
spring 1955, summer 1962, 1976, autumn 1952, 1976), cereals, b flowering of apple, c heading of winter cereals, d harvest of
accompanied by regional cold spells mainly in the winter winter cereals, e sowing of winter cereals (see Table 1)
Int J Biometeorol (2011) 55:921932 927

a
Phenological anomaly (days)

Temperature anomaly (C)

b
928 Int J Biometeorol (2011) 55:921932

d
Phenological anomaly (days)

Temperature anomaly (C)

Fig. 4 (continued)
Int J Biometeorol (2011) 55:921932 929

Phenological anomaly (days)

Temperature anomaly (C)

Fig. 4 (continued)

Central Europe, sowing was almost exclusively earlier by up also seemed to be a good indicator of climate extremes. The
to 30 days, and cold spells were manifested in later sowing of cold spell in summer 1962 was manifested in predominantly
spring cereals by up to 20 days. Three monthly and seasonal later harvest dates in central Europe by up to 40 days, while
cold spells, February 1956, January 1957 and winter 1957, the summer European heat wave of 2003 as well as June
demonstrated meaningful spatial coherence with observed 2003 were nearly exclusively linked to earlier harvest dates
phenological impacts (R2 of 40, 17 and 59%, respectively). by up to 40 days. However, none of the warm/cold spells
Annual anomalies of flowering dates of apple, a true displayed in Fig. 4d exhibited meaningful spatial coherence
phenological phase predominantly triggered by climate con- between temperature and phenological anomalies.
ditions, did mirror the warm and cold spells quite well Sowing of winter wheat (Fig. 4e) seemed to be the
(Fig. 4b). The cold spring of 1955, as well as cold February classical example of a false phenological phase. Regard-
1956, were linked to later apple flowering dates. Warm less of a warm or cold spell inside or outside Central
February and March 1990, as well as March 1989, Europe, the sowing of winter wheat, barley and rye was
corresponded to earlier flowering dates. Ten out of 11 warm/ up to 50 days earlier than average winter wheat sowing
cold spells studied revealed a negative spatial coherence with dates in the standard reference period (19301939). Only
cooler deviations being linked to later phenology. However, sowing dates of winter cereals in 1976 seemed to
only for the five warm spells of 1989 (March, April, spring) reasonably mirror the regional cold spell of October
and 1990 (February and March) could this spatial coherence 1976 on a spatial scale (linear regression, p< 0.0001, R2 =
be regarded as meaningful with larger R2 from 21 to 52%. 15%), while for all other warm/cold spells there was no
The picture for heading of winter cereals was unclear, spatial relationship between phenological impact and
mostly due to fewer identified warm and cold spells in temperature deviation.
spring months and the spring season (Fig. 4c). Only two
warm spells in March 1989 and 2001 exhibited spatial
coherence (R2 of 59 and 53%, respectively); however, with Discussion and conclusions
oppositely directed relationships.
Most interestingly, the harvest of winter cereals (Fig. 4d), Based on various methods proposed in the literature for
which can also be regarded as a false phenological phase, defining warm and cold spells (e.g. Beniston 2005, 2009),
930 Int J Biometeorol (2011) 55:921932

we defined the minimum length of a potential period in the type of extreme (e.g. earlier onsets linked to warm
which a notable spell could occur as 1 month since shorter spells), (2) the exclusiveness of the signal, and (3) the
periods were less comparable in their effects on phenology spatial coherence, i.e. whether phenological changes also
at an European scale. Despite this coarse temporal spatially mimic the respective warm/cold spell.
resolution, the method presented in this paper of defining Only a few studies on phenological extremes have so far
a continental or regional warm or cold spell via relative reported species' differences in impacts of extreme warm
temperature deviations from a long-term reference (1) spells. Rutishauser et al. (2008), for example, studied the
exceeding 1.5 or 3 standard deviations and (2) covering a effect of the very warm spring (March to May) of 2007 on
considerable spatial extent of more than 40% of grid points 302 different phenophases and revealed that roughly one-
or more than 1% of grid points, proved to adequately third were especially sensitive in Switzerland. Luterbacher
describe longer periods of extreme temperatures. The et al. (2007) also studied the impact of one warm spell in
results clearly revealed that, during the non-stationary autumn and winter of 2006/2007. They reported phenolog-
conditions in recent decades triggered by anthropogenic ical impacts related to this warmth such as unusual partial
warming, the frequencies of warm and cold spells changed second flowering or extended flowering until the beginning
with time. Cold spells predominated until 1988, and of winter. In addition, early flowering species in spring also
afterwards, warm spells prevailed. Nonetheless, these two exhibited distinct earlier flowering after this warm winter of
periods were not completely distinct, as in the first period 2007. Our study equally reveals differences in the sensitiv-
regional warm spells were also noted. For the obviously ity of phenophases/phenological seasons to warm and cold
warmer period of the last two decades, few regional and spells. The major results revealed that the harvest of winter
continental cold spells were identified. cereals, linked to summer temperatures, and flowering of
Our analysis of the impacts of these warm and cold apple, linked to early spring temperatures, were the most
spells on European plant phenology relied on a mix of promising when studying the impacts of warm and cold
different species, phenophases, and datasets, ranging from spells in Europe. Caution should be applied if false
point observational data to digitized maps. Can the results phenological phases are used as indicators for impacts of
be regarded as robust in the face of potential pitfalls extreme events; sometimes they appeared to be suitable, e.
involved in using these datasets? Studying phenological g. harvest of winter cereals, sowing of summer cereals, in
seasons defined by several phenophases is a common, well- contrast to sowing of winter cereals which was largely
accepted procedure (Schwartz 2003). Similarly, the spatial unaffected by any change in autumn extreme temperatures
matching of point and vector data is an accepted practice. (see also Menzel et al. 2006b). Most interestingly, and only
Thus, we believe that the standard reference period of for apple flowering in spring, the warm and cold spells in
19301939 based on maps drawn by Schnelle in 1965, and February to April, as well as in the spring season also
inclusion of other substitute phases (winter cereals) in exhibited a spatial coherence with the observed phenolog-
addition to winter wheat, was not a major constraint in our ical anomalies, i.e. the amount of extreme temperature was
study. However, the results may have suffered from a also mirrored in phenology.
general lack of available data on the European scale and the What about differences in the sensitivity of phenological
spatial concentration of analysed sites to mainly Central data to warm and cold spells? Using very long-term records
and Eastern Europe. This effect can be observed in Fig. 4 of three tree species (300 years), Rutishauser et al. (2008)
where confidence ellipses without significant temperature studied changes in their sensitivity, revealing distinct
anomaly are displayed due to the fact that phenological data periods with stronger sensitivity in the cooler periods and
were lacking in the area where a continental or regional decreased sensitivity during two periods with warming
warm/cold spell was identified. trends (18901950, 19702007). The summer heat wave of
The main topic of this paper, the question of whether and 2003 may be a promising case study for changes in
how these identified warm/cold spells translated into sensitivity. Although phenological impacts of the European
phenological (extreme) impacts can be assessed from two summer heat wave of 2003 have not so far been studied on
directions: firstly, studying which phenological phases and a broader scale, except very recently by Garcia-Herrera et
species were most affected and, in contrast, which seemed al. (2010), grape harvest dates have been used to prove the
to be more resilient according to subsequent change in their uniqueness of this heat wave in a historical context, clearly
onset dates, and secondly, by analysing which key demonstrating a general sensitivity of grape harvest dates to
phenological phases were more suitable to also mirror summer temperatures (Chuine et al. 2004; Menzel 2005).
extreme temperature spells as a proxy for paleo-conditions. However, our confidence ellipses of summer and June 2003
In particular for the latter case, different indicators should temperature against anomalies of harvest of winter wheat
be checked for (1) their sensitivity, i.e. the strength of the indicated a reduced response compared to other events.
observed advances and delays, and their consistency with Further studies are needed to clarify whether a reduced
Int J Biometeorol (2011) 55:921932 931

sensitivity can be observed in this special year (2003); in Jentsch A, Kreyling J, Bttcher-Treschkow J et al (2009) Beyond
other words, the observed harvest dates did not seem to gradual warming - extreme weather events alter flower phenol-
ogy of European grassland and heath species. Glob Change Biol
allow a reconstruction of the exact extent of this extreme 5(4):837849
event (see also Keenan 2007). Jones PD, Horton EB, Folland CK et al (1999) The use of indices to
Little attention has been paid so far to the spatial identify changes in climatic extremes. Clim Change 42:131149
coherence of the phenological signal, i.e. whether their Katz RW (2010) Statistics of extremes in climate change. Clim
Change 100:7176
variations are synchronous over the same spatial domain as Keenan DJ (2007) Grape harvest dates are poor indicators of summer
the temperature variations. This issue of spatial synchrony warmth. Theor Appl Climatol 87(14):255256
can very easily also be studied by the confidence ellipses. Kioutsioukis I, Melas D, Zerefos C (2010) Statistical assessment of
Diagonal ellipses, such as for the response of flowering of changes in climate extremes over Greece (19552002). Int J
Climatol 30(11):17231737
apple to warm spells in February/March 1990 (see Fig. 4b) Klein Tank AMG, Zwiers FW, Zhang X (2009) Climate Data and
or of harvest of winter wheat to the warm spell of June Monitoring: Guidelines on Analysis of extremes in a changing
2003 (Fig. 4d), indicated those cases where a spatial climate in support of informed decisions for adaptation.
reconstruction of the extent and amount of the temperature WCDMP-No. 72, WMO-TD No. 1500
Luterbacher J, Liniger MA, Menzel A et al (2007) Exceptional
extreme should also be possible. European warmth of autumn 2006 and winter 2007: Historical
We conclude that continental cold spells during the context, the underlying dynamics, and its phenological impacts.
growing season were linked post-1989 to variable impacts Geophys Res Lett 34(12):L12704
on different species and phenophases. True and a few false Matsumoto K, Ohta T, Irasawa M et al (2003) Climate change and
extension of the Ginkgo biloba L. growing season in Japan. Glob
phenological phases seem to be suitable indicators for Change Biol 9(11):16341642
warm and cold spells in different seasons, especially apple McCalla RJ, Day EED, Millward HA (1978) The relative concept of
flowering and harvest of winter cereals; however, more warm and cold spells of temperature: methodology and applica-
research is needed to assess the value of these proxies in tion. Arch Met Geoph Biokl, Ser B 25:323336
Menzel A (2003) Plant phenological anomalies in Germany and their
creating an absolute reconstruction of very extreme events. relation to air temperature and NAO. Clim Change 57(3):243263
Menzel A (2005) A 500 year pheno-climatological view on the 2003
heatwave in Europe assessed by grape harvest dates. Meteorol Z
References 14(1):7577
Menzel A, Fabian P (1999) Growing season extended in Europe.
Nature 397:659
Beniston M (2005) Warm winter spells in the Swiss Alps: Strong heat Menzel A, Sparks TH, Estrella N et al (2005) SSW to NNE - NAO
waves in a cold season? A study focusing on climate observations at affects the progress of seasons across Europe. Glob Change Biol
the Saentis high mountain site. Geophys Res Lett 32:L01812 11(6):909918
Beniston M (2009) Decadal-scale changes in the tails of probability Menzel A, Sparks TH, Estrella N et al (2006a) European phenological
distribution functions of climate variables in Switzerland. Int J response to climate change matches the warming pattern. Change
Climatol 29(10):13621368 Biol 12:19691976
Chau PH, Chan KC, Woo J (2009) Hot weather warning might help to Menzel A, Von Vopelius J, Estrella N et al (2006b) Farmers' annual
reduce elderly mortality in Hong Kong. Int J Biometeorol 53 activities are not tracking the speed of climate change. Clim Res
(5):461468 32:201207
Chmielewski FM, Rtzer T (2001) Response of phenology to climate Nielsen CCN, Rasmussen HN (2009) Frost hardening and dehardening in
change across Europe. Agric For Meteorol 108:101112 Abies procera and other conifers under differing temperature
Chuine I, Yiou P, Viovy N et al (2004) Historical phenology: Grape regimes and warm-spell treatments. Forestry 82(1):4359
ripening as a past climate indicator. Nature 432(7015):289290 Parmesan C, Root TL, Willig MR (2000) Impacts of extreme weather and
Cleland EE, Chuine I, Menzel A et al (2007) Shifting plant phenology climate on terrestrial biota. Bull Am Meteorol Soc 81(3):443450
in response to global change. Trends Ecol Evol 22:357365 Parmesan C, Yohe G (2003) A globally coherent fingerprint of climate
Easterling DR, Meehl GA, Parmesan C et al (2000) Climate extremes: change impacts across natural systems. Nature 421:3742
observations, modeling, and impact. Science 289:20682074 Parmesan C (2006) Ecological and evolutionary responses to recent
Estrella N, Sparks TH, Menzel A (2007) Trends and temperature climate change. Annu Rev Ecol Evol Syst 37:637669
response in the phenology of crops in Germany. Glob Change Parry ML, Canziani OF, Palutikof JP (Eds) et al. (2007) IPCC 2007:
Biol 13(8):17371747 Summary for Policymakers. In: Climate Change 2007: Impacts,
Garcia-Herrera R, Diaz J, Ttrigo RM et al (2010) A review of the Adaptation and Vulnerability. Contribution of Working Group II to the
European Summer Heat Wave of 2003. Critical Rev Environ Sci Fourth Assessment Report of the Intergovernmental Panel on Climate
Technol 40(4):267306 Change, Cambridge University Press, Cambridge, UK, pp 22
Horton EB, Folland CK, Parker DE (2001) The changing incidence of Root TL, Price JT, Hall KR et al (2003) Fingerprints of global
extremes in worldwide and central England temperatures to the warming on wild animals and plants. Nature 421:5760
end of the twentieth century. Clim Change 50:267295 Rosenzweig C, Casassa G, Karoly D et al (2007) Assessment of
Huynen MMTE, Martens P, Schram D et al (2001) The impact of heat observed changes in responses in natural and managed systems.
waves and cold spells on mortality rates in the Dutch population. In: Parry ML, Canziani OF, Palutikof Jp et al (eds) Climate
Environ Health Perspect 109(5):463470 Change 2007: Impacts, Adaptation and Vulnerability. Contribu-
Jentsch A, Kreyling J, Beierkuhnlein C (2007) A new generation of tion of Working Group II to the Fourth Assessment Report of the
climate change experiments: events not trends. Frontiers Ecol Intergovernmental Panel on Climate Change. Cambridge Univer-
Environ 5(7):365374 sity Press, Cambridge, pp 79131
932 Int J Biometeorol (2011) 55:921932

Rosenzweig C, Karoly D, Vicarelli M et al (2008) Attributing physical Sparks TH, Jeffree EP, Jeffree CE (2000) An examination on the
and biological impacts to anthropogenic climate change. Nature relationship between flowering times and temperature at the
453:353357 national scale using long-term phenological records from the UK.
Rtzer T, Chmielewski FM (2001) Phenological maps of Europe. Clim Int J Biometeorol 44:8287
Res 18:249257 Sparks TH, Menzel A (2002) Observed changes in seasons: an
Rutishauser T, Luterbacher J, Defila C et al (2008) Swiss spring overview. Int J Climatol 22:17151725
plant phenology 2007: extremes, a multi-century perspective, Trenberth KE, Jones PD, Ambenje P et al. (2007) Observations:
and changes in temperature sensitivity. Geophys Res Lett 35: Surface and Atmospheric Climate Change. In: Climate Change
L05703 2007: The Physical Science Basis. Contribution of Working
Satyamurty P, Teixeira MD, Padilha CK (2007) Warm and dry spells Group I to the Fourth Assessment Report of the Intergovern-
(WDS) in austral winter over central South America. Ann mental Panel on Climate Change [Solomon, S., D. Qin, M.
Geophys 25(5):10491069 Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L.
Schnelle F (1965) Beitrge zur Phnologie Europas I. 5 Miller (eds.)]. Cambridge University Press, Cambridge
Mittelwertskarten, Erstfrhling bis Herbst. Berichte des Walther GR (2004) Plants in a warmer world. Perspect Plant Ecol Evol
Deutschen Wetterdienstes Nr. 101 (Band 14), Offenbach am Syst 6(3):169185
Main, Selbstverlag des Deutschen Wetterdienstes Walther GR, Post E, Convey P et al (2002) Ecological responses to
Schwartz MD (Ed) (2003) Phenology - an Integrative Environmental recent climate change. Nature 416:389395
Science. Kluwer, Dordrecht Zimmermann NE, Yoccoz NG, Edwards TC et al. (2009) Climatic
Shabbar A, Bonsal B (2003) An assessment of changes in winter cold extremes improve predictions of spatial patterns of tree species.
and warm spells over Canada. Nat Hazards 29(2):173188 PNAS 17(106, suppl 2):1972319728