Content uploaded by Lenka Bartošová
Author content
All content in this area was uploaded by Lenka Bartošová on Jun 17, 2022
Content may be subject to copyright.
Citation: Bartošová, L.; Dížkova, P.;
Bauerová, J.; Hájková, L.; Fischer, M.;
Balek, J.; Bláhová, M.; Možný, M.;
Zahradníˇcek, P.; Štˇepánek, P.; et al.
Phenological Response of Flood Plain
Forest Ecosystem Species to Climate
Change during 1961–2021.
Atmosphere 2022,13, 978. https://
doi.org/10.3390/atmos13060978
Academic Editor: Junhu Dai
Received: 6 May 2022
Accepted: 13 June 2022
Published: 17 June 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
atmosphere
Article
Phenological Response of Flood Plain Forest Ecosystem Species
to Climate Change during 1961–2021
Lenka Bartošová1,*, Petra Dížkova 1,2 , Jana Bauerová2, Lenka Hájková3, Milan Fischer 1,2 , Jan Balek 1,
Monika Bláhová1,2 , Martin Možný3, Pavel Zahradníˇcek 1, Petr Štˇepánek 1, Zdenˇek Žalud 1,2
and Miroslav Trnka 1,2
1Global Change Research Institute AS CR v.v.i., Bˇelidla 986/4b, 60300 Brno, Czech Republic;
petra.dizkova@seznam.cz (P.D.); fischer.m@czechglobe.cz (M.F.); jan.balek.83@seznam.cz (J.B.);
blahova.m@czechglobe.cz (M.B.); zahradnicek.p@czechglobe.cz (P.Z.); stepanek.p@czechglobe.cz (P.Š.);
zdenek.zalud@mendelu.cz (Z.Ž.); mirek_trnka@yahoo.com (M.T.)
2Institute of Agrosystems and Bioclimatology, Mendel University in Brno, Zemˇedˇelská1,
61300 Brno, Czech Republic; janabau39@seznam.cz
3Department of Biometeorological Applications, Czech Hydrometeorological Institute, Na Šabatce 17,
14300 Prague, Czech Republic; lenka.hajkova@chmi.cz (L.H.); martin.mozny@chmi.cz (M.M.)
*Correspondence: lenka.bartosova@mendelu.cz
Abstract:
The present study analyses 61 years of phenological observations (1961–2021) of five herb,
five shrub, four tree, and one bird species representing the prevalent spring species of floodplain
forest ecosystems in the Czech Republic, central Europe. The in situ observations were conducted at
the Vranovice site (48
◦
48
0
N, 16
◦
46
0
E, 170 m above mean sea level) representing the Plaˇcek’ Forest
National Reserve. The observed plants and bird species showed statistically significant (p< 0.05)
shifts in phenological terms to an earlier date of the year, but the rate of the shift among the observed
species differed. The most progressive shifts were detected for the herbs (14 days), followed by
the shrubs (13 days), trees (9 days), and finally by the bird species (8 days). All the phenophases
were significantly correlated with the daily maximum temperature (r= 0.72–0.91). The results also
showed a decline in the correlation for species among the phenophases of the herbs and trees. The
phenophases that were highly correlated in the past were less correlated and had higher variability in
the last decades. We conclude that the phenological response of the ecosystem to warming in the
spring resulted in higher variability and a lower correlation among the observed phenophases mainly
caused by the most expressive phenological shifts of the early herbs.
Keywords: phenology; asynchrony; mismatch; herbs; shrubs; trees; bird species; overlapping
1. Introduction
Phenology, the study of the seasonal timing of recurring biological events, is commonly
known as an indicator of climate change, e.g., [
1
–
3
]. Various authors have already published
that forest ecosystems in Europe are likely to be strongly influenced by climate change
and that many species are already exhibiting phenological shifts, e.g., [
4
]. Several studies
on phenological interactions have brought complex insights into phenological shifts and
responses to climate change for plants [
5
] or animals [
6
]. For example, in the UK, the
phenological climate sensitivity of broad taxonomic groups varies greatly, and primary
producers showed the highest tendency towards “earliness” with seasonal warming [
7
]; the
synchronicity between the phenology of predators and their prey is becoming weaker [
8
],
and a phenological mismatch in the late breeding of Great Tits to the caterpillar food peak
date was evaluated in the Netherlands [9].
Although there are many phenological studies, few of them can rely on long-term
and robust in situ data for the robust group of species (from herbs to shrubs, trees, and
animal representatives) and to analyse how each part of the ecosystem responds to climate
Atmosphere 2022,13, 978. https://doi.org/10.3390/atmos13060978 https://www.mdpi.com/journal/atmosphere
Atmosphere 2022,13, 978 2 of 17
change and the coherence of these responses. Phenological ecosystem studies mainly use
vegetation indices and the start and end of the season or the length of the growing season for
a specific ecosystem, e.g., [
10
], or there are studies regarding the phenology of communities
for a shorter period such as 2008–2010, e.g., [
11
]. One of the major concerns about the
biological impacts of recent anthropogenically driven climate change is that they can alter
interactions between species in ways that could cascade through communities or even
whole ecosystems, e.g., [
8
,
12
–
14
]. The ecosystem of floodplain forests is characterised by a
rich species diversity of plants and animals and is influenced by the level of groundwater,
elevation and time of flooding, and granular soil structure [
15
], and this type of ecosystem
was used within this study as an experimental site.
There is a growing concern that climate change has significant impacts on species’
phenology and seasonal population dynamics, and thus causes interaction (a)synchrony
between species [
16
–
18
]. If individual species change their phenologies, there is the po-
tential for phenological mismatches: species that have historically undergone life history
events on the same seasonal calendar may lose synchrony and therefore lose the ability
to interact as they have in the past [
8
,
16
,
17
,
19
]. Given the match/mismatch hypothesis,
the different extents or directions of the phenological shifts among interacting species may
have significant implications for community structures and dynamics [20].
Here, we use long-term phenological data (61 years) of the flood plain forest ecosystem
(shrubs, herbs, trees, and bird species) from central Europe, with aims to: (1) evaluate the
correlation among the phenological phases of a given species (of the flood plain forest
ecosystem) with meteorological parameters to determine both the change in correlation
in time (using a running correlation) among the phenological phases and the phenolog-
ical phases with the strongest meteorological predictor (that means the meteorological
variable which is correlated with phenological dates the most and influences the onset of
phenophases); and (2) analyse the overlaps among the chosen phenological phases and
their lengthening or shortening trends to indicate the overlap index for the selected species
in 1961–2021. For these goals (1) and (2), we hypothesise that each given group of species
(herbs, shrubs, trees, and bird species) of the ecosystem has advanced to an earlier date
in the year because the meteorological records showed a temperature trend to a warmer
spring. Nevertheless, the warmer conditions in spring (mainly in February and March)
can cause not only advancing phenological phases of early species but also their larger
phenological separation from the latter species. We supposed the latter species, who are
facing lower warmer trends, could advance the terms of the phenological dates to the
earlier time, less than the early species. Subsequently, this situation can cause intense
shifting and phenological differences among all the observed species and different lengths
of phenological overlapping (among all the species).
2. Materials and Methods
The flood plain forest ecosystem study was conducted at the experimental site of
Vranovice, the Plaˇcek’ Forest National Reserve (48
◦
48
0
N, 16
◦
46
0
E, altitude 170 m above
sea level) in the southern part of the Czech Republic, central Europe. Phenological in
situ observations were performed from 1961 to 2021. The experimental site was a forest
ecosystem with remnants of the floodplain forest, which was characterised by a fully
grown, multi aged canopy with high biological diversity (Figure 1). The species diversity
of the flood plain forests is influenced by the level of groundwater, elevation and time of
flooding, and soil granular structure [
16
]. Long-term phenological research was possible
only in locations away from the flooding zone, i.e., with hornbeam–elm–ash growth. This
growth ranged from a slightly elevated plateau (plain) to the edges of the flood plain
forest. Flooding of this area rarely occurs and only for a short time during periods of
higher water levels. This area was also chosen for phenological observations because of the
considerable species diversity and rich herb components. The study site Plaˇcek’ forest was
established in the fifties as an additional experimental site for the UNESCO projects Man
and the Biosphere and the International Biological Program [
15
]. Since 1961, phenological
Atmosphere 2022,13, 978 3 of 17
observations have been performed systematically for selected species, and since 2008, the
observations have been performed by Mendel University in Brno and since 2016 by the
Global Change Research Institute, CAS (Czech Academy of Sciences).
Atmosphere 2022, 13, x FOR PEER REVIEW 3 of 18
and the Biosphere and the International Biological Program [15]. Since 1961, phenological
observations have been performed systematically for selected species, and since 2008, the
observations have been performed by Mendel University in Brno and since 2016 by the
Global Change Research Institute, CAS (Czech Academy of Sciences).
Figure 1. Location of the experimental site (Plaček´ Forest National Reserve) in the Czech Republic
(upper part) and the table in the lower part with all species and mean dates of two phenological
phases for each species. The hatching in the boxes indicates the phenological episodes that overlap
among all others. FLD indicates the date when the first clutch in a given year was initiated at a given
site. MLD indicates the mean initiation date of all first clutches in the population at each site in a
given year.
2.1. Phenological Data
Within this study, 17 species were observed (16 plants and 1 bird species, Figure 1),
and for each species, 2 phenological phases were observed (all during the spring period).
The phenological phases were defined as (the number of species shown in brackets): first
flower + full flowering (5 herbs and 5 shrubs), bud breaking + full foliage (6 trees), and
first laying date (FLD—the date when the first clutch in a given year was initiated at a
given site) and mean laying date (MLD—the mean initiation date of all first clutches in
the population at each site in a given year) (1 bird species) (Figure 1). The observations
were performed by 3 observers during the whole period of observation. The observations
were performed by senior lecturer Zdeněk Bauer for the longest time (from 1961 to 2009),
who established the experimental site. Since 2009, the observations were done performed
by a Ph.D. student from Mendel University in Brno and a junior scientist from the Global
Figure 1.
Location of the experimental site (Plaˇcek’ Forest National Reserve) in the Czech Republic
(upper part) and the table in the lower part with all species and mean dates of two phenological
phases for each species. The hatching in the boxes indicates the phenological episodes that overlap
among all others. FLD indicates the date when the first clutch in a given year was initiated at a given
site. MLD indicates the mean initiation date of all first clutches in the population at each site in a
given year.
2.1. Phenological Data
Within this study, 17 species were observed (16 plants and 1 bird species, Figure 1),
and for each species, 2 phenological phases were observed (all during the spring period).
The phenological phases were defined as (the number of species shown in brackets): first
flower + full flowering (5 herbs and 5 shrubs), bud breaking + full foliage (6 trees), and
first laying date (FLD—the date when the first clutch in a given year was initiated at a
given site) and mean laying date (MLD—the mean initiation date of all first clutches in the
population at each site in a given year) (1 bird species) (Figure 1). The observations were
performed by 3 observers during the whole period of observation. The observations were
performed by senior lecturer Zdenˇek Bauer for the longest time (from 1961 to 2009), who
established the experimental site. Since 2009, the observations were done performed by a
Ph.D. student from Mendel University in Brno and a junior scientist from the Global Change
Research Institute CAS until now (both are co-authors of this study). The methodology of
Atmosphere 2022,13, 978 4 of 17
the observations was a crucial point and was exchanged as well as possible between all
the observers. The methodology also bears on the Czech Hydrometeorological Institute’s
methods and Guidelines for phenological observations [21].
For each species, we knew the phenological phases (e.g., for herbs and shrubs, first
flower and full flowering; for trees, bud breaking and full foliage; and for bird species, FLD
and MLD) and both phases together created a so-called phenological stage. The pheno-
logical stages of the species overlapped with each other and in total created 62 overlaps
(Figure 1). The trends of the 62 overlaps were analysed for pairs of the phenophases of
the selected species, so the length of the overlap was created, e.g., by the phenological
phase of a herb with the phenological phase of a shrub (and for pairs of the overlaps of
herb–tree, herb–bird, shrub–bird, tree–bird, and tree–shrub). Then, the ‘overlap index’
was calculated as the difference between the number of overlap days and the number of
species. A high overlap index (and with a high number of overlapping days) indicates
phenological stages that are long. On the other hand, a low overlap index (years with a
small number of overlapping days) indicates short phenological stages. For the analysis,
all possible overlaps of all the available species were included to see the dynamics and
linkage among all the observed phenophases, not only those which were connected, e.g.,
by a trophic chain.
2.2. Meteorological Data
Meteorological data for the station Plaˇcek’ Forest were retrieved from maps inter-
polated based on 268 climatological stations and 787 precipitation stations in the Czech
Republic for the period 1961–2021. From this map were extracted the necessary values for
the average, maximum, and minimum air temperature and precipitation. Input climatolog-
ical data were quality controlled, homogenised, and did not contain missing values [
22
–
25
],
(Figure 2). The daily data were interpolated by the regression kriging method using geo-
graphical coordinates, elevation, and other terrain characteristics as predictors into gridded
data with a spatial resolution of 500 m. The relationship between the predictors and a
predictand was estimated for each meteorological variable and each day individually, and
only predictors that were significant for the regression were used for the final estimates.
For each of the cases, the best type of semi-variogram was assessed. The calculations were
performed in R by applying the packages rgdal and raster. The weather variables were
separated into various periods for each month; from February to May for the particular
year of the given phenophase occurrence, the months were denoted by F for February, M
for March, A for April, and M for May for the average values of the pairs (FM—February,
March; MA—March, April; and AM—April, May) and subsequently for the longer periods
(FMA—February, March, April; MAM—March, April, May). These meteorological series
for the mean, maximum, and minimum temperature and precipitation were then used for
statistical analysis.
Correlation coefficients (r, we used the parametric Pearson coefficient) were used as
the primary indicators of the strength of the relationship between the given variables. The
trend is representing the slope of the linear regression between the date of the phenological
date and the year. Any significance in the observed trends (for both the climatological
and phenological parameters) was assessed using a t test. Additionally, we used 20-year
subsamples (e.g., 1961–1980, 1962–1981, etc.) for the calculation of the running correlation to
see the variability of the correlation among the phenological dates and climate parameters
(the mean, maximum, and minimum temperature and precipitation). All the tests were
performed with the statistical/programming tool R 3.6.1. [
26
] and with AnClim software
for removing potential outliers in the phenological series [27].
Atmosphere 2022,13, 978 5 of 17
Atmosphere 2022, 13, x FOR PEER REVIEW 5 of 18
Figure 2. Location of climatological and precipitation stations series used for interpolation.
3. Results
The mean annual air temperature showed a significant increase of 1.45 °C for the
period of 61 observational years in the experimental site Plaček´ forest. Statistically signif-
icant warming trends were observed mainly for the maximum air temperature in Febru-
ary, followed by March and April, and for the average values of the daily maximum tem-
perature in the given months (Table 1). Generally, the maximum air temperature showed
the most notable changes, followed by the mean air temperature (with significant warm-
ing trends in the spring period) and minimum air temperature with the lowest changes.
The total precipitation showed a decrease in the data in February, April, and May but with
no trend. The same lack of trend was detected for precipitation in March but with increas-
ing values. (Table 1).
The spring phenological phases of all the investigated species advanced to an earlier
time and each part of the ecosystem had different responses to climate change. The largest
shifts were in the herbs (which had advanced their terms of the phenological phases by
14 days on average since 1961), followed by the shrubs (13 days on average) and trees (9
days on average), and the lowest rate of shifting was exhibited by the bird species (8 days
on average) (Figure 3A). A correlation analysis among the phenological phases and cli-
mate parameters showed the strongest relationship with the maximum temperature ( =
0.7–0.9, = 0.000) (followed by the mean and minimum temperature and with a weak
relationship with the precipitation) for various months (Table 2). Early herbs and shrubs
with average phenological dates in March showed the strongest relationship with the
maximum temperature in the months of February and March ( = 0.8–0.9, = 0.000). The
majority of the phenophases of the observed species proceeded during April, and the
strongest relationship was again identified for the maximum air temperature in the pre-
ceding and current months ( = 0.7–0.8, = 0.000). The latest phenophases were ob-
served in May and all of them were correlated with the maximum air temperature ( =
0.7–0.9, = 0.000) in the preceding months (the average of the months March, April, and
May) (Figure 3B).
A running correlation (a 20-year moving window) among the tree phenophases and
the maximum temperature (in specific months) showed variability in time, specifically for
the phenophases of Tilia cordata, Quercus robur, and Fraxinus excelsior. The correlation de-
creased from 1993–2012 (Figure 4A). Correlations for the 20-year periods for the herb,
Figure 2. Location of climatological and precipitation stations series used for interpolation.
3. Results
The mean annual air temperature showed a significant increase of 1.45
◦
C for the
period of 61 observational years in the experimental site Plaˇcek
´
forest. Statistically signifi-
cant warming trends were observed mainly for the maximum air temperature in February,
followed by March and April, and for the average values of the daily maximum tempera-
ture in the given months (Table 1). Generally, the maximum air temperature showed the
most notable changes, followed by the mean air temperature (with significant warming
trends in the spring period) and minimum air temperature with the lowest changes. The
total precipitation showed a decrease in the data in February, April, and May but with no
trend. The same lack of trend was detected for precipitation in March but with increasing
values (Table 1).
The spring phenological phases of all the investigated species advanced to an earlier
time and each part of the ecosystem had different responses to climate change. The largest
shifts were in the herbs (which had advanced their terms of the phenological phases by
14 days on average since 1961), followed by the shrubs (13 days on average) and trees
(9 days on average), and the lowest rate of shifting was exhibited by the bird species
(8 days on average) (Figure 3A). A correlation analysis among the phenological phases
and climate parameters showed the strongest relationship with the maximum temperature
(r= 0.7–0.9, p= 0.000) (followed by the mean and minimum temperature and with a
weak relationship with the precipitation) for various months (Table 2). Early herbs and
shrubs with average phenological dates in March showed the strongest relationship with
the maximum temperature in the months of February and March (r= 0.8–0.9, p= 0.000).
The majority of the phenophases of the observed species proceeded during April, and
the strongest relationship was again identified for the maximum air temperature in the
preceding and current months (r= 0.7–0.8, p= 0.000). The latest phenophases were observed
in May and all of them were correlated with the maximum air temperature (r= 0.7–0.9,
p= 0.000) in the preceding months (the average of the months March, April, and May)
(Figure 3B).
Atmosphere 2022,13, 978 6 of 17
Table 1.
Average mean air temperature and the sum of precipitation for the experimental site and linear trends during 1961–2021 for mean, maximum and minimum
temperature, and the sum of total precipitation. FM indicates the average for months February and March; MA for March and April; AM for April and May; FMA
for February, March, and April; MAM for March, April, and May; and FMAM for February, March, April, and May.
Nature Reserve Altitude/m Latitude Longitude
Vranovice Plaˇckek’ forest 170 48◦5605400
16
◦
35
0
50
00
Average values
February March April May FM MA AM FMA MAM FMAM Year
Mean daily air temperature (◦C) 0.8 5 10.2 14.8 2.9 7.6 12.5 5.3 10 7.7 9.8
Maximum daily air temperature (◦C) 4.4 9.8 16 20.8 7.1 12.9 18.4 10.1 15.5 12.8 14.7
Minimum daily air temperature (◦C) −2.9 0.3 4.3 8.9 −1.3 2.3 6.6 0.6 4.5 2.7 5.0
Precipitation (mm) 23 26 30 55 49 56 85 78 111 134 473
Linear trends in 1961–2021
February March April May FM MA AM FMA MAM FMAM Year
Mean daily air temperature (◦C) 1.69 1.5 1.05 1.05 1.59 1.27 * 1.05 * 1.42 * 1.2 ** 1.32 ** 1.45 ***
Maximum daily air temperature (◦C) 2.5 * 2.74 * 2.35 ** 1.24 2.62 * 2.55 *** 1.8 ** 2.53 *** 2.11 *** 2.21 *** 1.89 ***
Minimum daily air temperature (◦C) 0.89 0.25 −0.24 0.86 0.57 0.05 0.31 0.3 0.29 0.44 1.0 **
Total precipitation (mm) −3.13 4.18 −10.73 −6.23 1.05 −6.55 −16.96 −9.68 −12.78 −15.91 6.32
*Significant at α= 0.05; ** significant at α= 0.01; *** significant at α= 0.001%.
Atmosphere 2022,13, 978 7 of 17
Atmosphere 2022, 13, x FOR PEER REVIEW 6 of 18
shrub, and bird species with the maximum temperature (in specific months) were stable,
significant, and with no changes during the whole time of observations (Figure 4B); the
average values of the correlation coefficients for all the phenophases of the herb, shrub,
and bird species represent a significant correlation for each phenophase.
Figure 3. (A) Average terms of phenological phases for herbs, shrubs, trees, and birds for the period
of 1961–2021. (B) Relationship between daily air maximum temperature in MAM (March, April,
May) and the terms of phenophases for average values of herbs, shrubs, trees, and birds.
Figure 3.
(
A
) Average terms of phenological phases for herbs, shrubs, trees, and birds for the period
of 1961–2021. (
B
) Relationship between daily air maximum temperature in MAM (March, April, May)
and the terms of phenophases for average values of herbs, shrubs, trees, and birds.
A running correlation (a 20-year moving window) among the tree phenophases and
the maximum temperature (in specific months) showed variability in time, specifically
for the phenophases of Tilia cordata,Quercus robur, and Fraxinus excelsior. The correlation
decreased from 1993–2012 (Figure 4A). Correlations for the 20-year periods for the herb,
shrub, and bird species with the maximum temperature (in specific months) were stable,
significant, and with no changes during the whole time of observations (Figure 4B); the
average values of the correlation coefficients for all the phenophases of the herb, shrub, and
bird species represent a significant correlation for each phenophase.
The tree phenophases also showed a higher variability for the running correlation
among each phenophase of the observed tree species. Only three species (Carpinus betulus,
Acer campestre, and Ulmus laevis) and their phenophase bud breaking showed high and
stable correlations during all 61 years. The other six tree species showed a high and stable
running correlation only until the period 1980–1999; since that period, the correlation
became weaker, especially for species with the phenophases of ash (Fraxinus excelsior). The
Atmosphere 2022,13, 978 8 of 17
remaining five species showed variable correlations during the whole observational period;
the correlation increased until 1982–2001, after which it decreased (Figure 5C).
Table 2.
Linear trends for observed phenological terms (slope of the linear regression line). P1
indicates phenophase ‘first flower’ for herbs and shrubs; bud breaking for trees and FLD (first laying
date) for bird species. P2 indicates phenophase ‘full flowering’ for herbs and shrubs; full foliage
for trees and MLD (mean laying date) for bird species. Correlation with maximum air temperature
(T
max
) was detected for averaged values of months: FM (av. temperature for February and March), M
(av. temp. for March), MA (av. temp. for March and April), FMA (av. temp. for February, March,
and April) and MAM (average temperature for March, April, and May). The value “r” (correlation
coefficient) describes the relationship among terms of phenophases and T
max
(all coefficients are
significant at α= 0.05).
Species
Phenophases
Abbreviation
Days/1961–2021
(Days/Decade)
Temperature
Correlation r
Herbs Common Lungwort
(Pulmonaria officinalis, L.) P1 PO1 12.5 * (2.1) Tmax/FM 0.81
P2 PO2 7.8 * (1.3) Tmax/M 0.79
Yellow Starof-Bethlehem
(Gagea lutea, L.) P1 GL1 20.7 *** (3.5) Tmax/FM 0.77
P2 GL2 19.6 *** (3.3) Tmax/M 0.75
Bulbous Corydalis
(Corydalis cava, L.) P1 CC1 15.1 ** (2.6) Tmax/M 0.79
P2 CC2 14.0 *** (2.4) Tmax/MA 0.74
Yellow Wood Anemone
(Anemone ranunculoides, L)
P1 AR1 13.7 ** (2.3) Tmax/FM 0.82
P2 AR2 11.1 ** (1.9) Tmax/MA 0.78
Herb Paris
(Paris quadrifolia, L.) P1 PQ1 10.4 ** (1.8) Tmax/MA 0.81
P2 PQ2 10.8 *** (1.8) Tmax/MA 0.81
Shrubs European Cornel
(Cornus mas, L.)
P1 CM1 11.5 (1.9) Tmax /FM 0.91
P2 CM2 9 (1.5) Tmax/FM 0.85
Blackthorn
(Prunus spinosa, L.) P1 PS1 10.07 ** (1.7) Tmax/FMA 0.85
P2 PS2 10.1 ** (1.7) Tmax/FMA 0.83
English Hawthorn
(Crataegus oxyacantha, L.) P1 CO1 15.4 *** (2.6) Tmax /MAM 0.86
P2 CO2 15.4 *** (2.6) Tmax/MAM 0.89
Common hawthorn
(Crataegus monogyna, Jacq.) P1 CMo1 13.4 *** (2.3) Tmax/MAM 0.88
P2 CMo2 13.0 *** (2.2) Tmax/MAM 0.91
Common Dogwood
(Cornus sanguinea, L.) P1 CS1 15.4 *** (2.6) Tmax/MAM 0.85
P2 CS2 15.7 *** (2.7) Tmax /AM 0.86
Trees Common Hornbeam
(Carpinus betulus, L.) P1 CB1 9.2 ** (1.6) Tmax/M 0.82
P2 CB2 7.1 * (1.2) Tmax/MA 0.81
European White Elm
(Ulmus laevis, P.)
P1 UL1 8.8 * (1.5) Tmax/M 0.78
P2 UL2 10.7 *** (1.8) Tmax/MAM 0.85
English Oak
(Quercus robur, L.)
P1 QR1 11.9 *** (2.0) Tmax/MA 0.75
P2 QR2 10.4 *** (1.8) Tmax/MAM 0.82
European ash
(Fraxinus excelsior, L.)
P1 FE1 10.3 *** (1.7) Tmax/MA 0.79
P2 FE2 8.6 ** (1.4) Tmax/MAM 0.88
Field maple
(Acer campestre,L.)P1 AC1 5.9 (1.0) Tmax/M 0.83
P2 AC2 7.5 * (1.0) Tmax/MA 0.81
Small-leaved lime
(Tilia cordata,Mill.)
P1 TC1 6.2 (1.0) Tmax/MA 0.76
P2 TC2 6.4 * (1.1) Tmax/MAM 0.72
Birds Great Tit
(Parus major, L.) P1 PM1 7.8 ** (1.3) Tmax/MA 0.75
P2 PM2 8.5 *** (1.4) Tmax/MA 0.82
* Significant at α= 0.05; ** significant at α= 0.01; *** significant at α= 0.001%.
Atmosphere 2022,13, 978 9 of 17
Atmosphere 2022, 13, x FOR PEER REVIEW 9 of 18
Figure 4. Running correlation for tree phenophases with maximum temperature in the 20-year pe-
riods (A). Abbreviations for species and temperature are explained in Table 2. Correlation in the 20-
year periods for average values of phenophases for herbs, shrubs, and bird species of an ecosystem
with maximal temperature in March and April (B).
The tree phenophases also showed a higher variability for the running correlation
among each phenophase of the observed tree species. Only three species (Carpinus betulus,
Acer campestre, and Ulmus laevis) and their phenophase bud breaking showed high and
stable correlations during all 61 years. The other six tree species showed a high and stable
running correlation only until the period 1980–1999; since that period, the correlation be-
came weaker, especially for species with the phenophases of ash (Fraxinus excelsior). The
remaining five species showed variable correlations during the whole observational pe-
riod; the correlation increased until 1982–2001, after which it decreased (Figure 5C).
The running correlation among all the shrub phenophases was stable and did not
change specifically; only the phenophases of the early shrubs (Cornus mas—CM1 and
Prunus spinosa—PS1) with the late phenophase of Cornus sanguinea (CS1) showed a low
correlation but for the whole time of observations (Figure 5B). The running correlation for
the herb phenophases showed a stable and fixed connection among the early herbs (Pul-
monaria officinalis, Gagea lutea, Corydalis cava, and Anemone ranunculoides) for the whole pe-
riod; the same phenophases of the early herbs were similarly and highly correlated with
the late phenophases of Paris quadrifolia until 1974–1993 and since the correlation de-
creased (Figure 5A). The running correlation for pairs of all the species (the average terms
of the herb, shrub, tree, and bird phenophases) of the observed ecosystem showed a stable
correlation from 1960 to 1981–2000. Since 1982–2001, the correlation became weaker and
Figure 4.
Running correlation for tree phenophases with maximum temperature in the 20-year
periods (
A
). Abbreviations for species and temperature are explained in Table 2. Correlation in the
20-year periods for average values of phenophases for herbs, shrubs, and bird species of an ecosystem
with maximal temperature in March and April (B).
The running correlation among all the shrub phenophases was stable and did not
change specifically; only the phenophases of the early shrubs (Cornus mas—CM1 and
Prunus spinosa—PS1) with the late phenophase of Cornus sanguinea (CS1) showed a low
correlation but for the whole time of observations (Figure 5B). The running correlation
for the herb phenophases showed a stable and fixed connection among the early herbs
(Pulmonaria officinalis,Gagea lutea,Corydalis cava, and Anemone ranunculoides) for the whole
period; the same phenophases of the early herbs were similarly and highly correlated
with the late phenophases of Paris quadrifolia until 1974–1993 and since the correlation
decreased (Figure 5A). The running correlation for pairs of all the species (the average
terms of the herb, shrub, tree, and bird phenophases) of the observed ecosystem showed a
stable correlation from 1960 to 1981–2000. Since 1982–2001, the correlation became weaker
and had higher variability. During the last 10 periods (from 1991–2010 until now), the
decline in the correlation coefficients was intensive, and the correlation declined mainly
between the herbs and bird species and the shrubs with trees (Figure 5D).
Atmosphere 2022,13, 978 10 of 17
Atmosphere 2022, 13, x FOR PEER REVIEW 11 of 18
Figure 5. Running correlations for terms of phenological phases among all herbs (A), shrubs (B), and trees (C) and averaged terms of phenophases
for herbs, shrubs, trees, and bird species (D). Abbreviations for each species: PO1—Pulmonaria officinalis; GL1—Gagea lutea; CC1—Corydalis cava;
AR1—Anemone ranunculoides; PQ1—Paris quadrifolia; CM1—Cornus mas; PS1—Prunus spinosa; CO1—Crataegus oxyacantha; CMo1—Crataegus monog-
yna; CS1—Cornus sanguinea; CB1—Carpinus betulus; UL1—Ulmus laevis; QR1—Quercus robur; FE1—Fraxinus excelsior; AC1—Acer campestre; TC1—
Tilia cordata; and PM1—Parus major.
Figure 5.
Running correlations for terms of phenological phases among all herbs (
A
), shrubs (
B
), and trees (
C
) and averaged terms of phenophases for herbs,
shrubs, trees, and bird species (
D
). Abbreviations for each species: PO1—Pulmonaria officinalis; GL1—Gagea lutea; CC1—Corydalis cava; AR1—Anemone ranunculoides;
PQ1—Paris quadrifolia; CM1—Cornus mas; PS1—Prunus spinosa; CO1—Crataegus oxyacantha; CMo1—Crataegus monogyna; CS1—Cornus sanguinea; CB1—Carpinus
betulus; UL1—Ulmus laevis; QR1—Quercus robur; FE1—Fraxinus excelsior; AC1—Acer campestre; TC1—Tilia cordata; and PM1—Parus major.
Atmosphere 2022,13, 978 11 of 17
The analysis of the overlapping phenophases of the chosen species showed that the
length of the overlap was significantly longer in ten cases and significantly shorter in
seven cases from the total number of 62 studied overlaps. The statistically significant
lengthening of the phenophase overlaps was detected for the early blossoming herbs with
other early herbs and one shrub and was moving in the range of 5–12 days (Figure A1).
Next, significant lengthening was calculated for the trees (between each other) and for
the trees with shrubs. On the other hand, the situation when the length of the overlaps
was getting significantly shorter arose for the overlaying of the phenophases of the herbs
(between each other) and for the herbs with trees (Figure A1). The remaining overlaps (41)
of all the phenological stages among various species was getting shorter or longer (from
−5 to +5 days) but with no significance; the last five overlaps did not show change.
The overlap index showed higher variability in the first half of the observed period
(Figure 6). The lowest overlap index was calculated for the years 1969 and 1986 when
the number of overlap days was low, which means that the phenological stages and
overlapping time were short. Conversely, a high index was calculated for 1981, 1990, and
1991 (and 2016) when the phenological stages were long. Nevertheless, the overlap index
did not show any significant trend for all the observed species. However, interactions
mainly among the herbs and trees, for which stages the overlap significantly changed, are
clear. Years with a high overlap index (and with a high number of overlapping days) were
years with average or subnormal temperature conditions and phenological stages that were
long. On the other hand, a low overlap index (years with a small number of overlapping
days) indicates years with temperatures above normal and short phenological stages.
Atmosphere 2022, 13, x FOR PEER REVIEW 12 of 18
Figure 6. Overlap index for all observed overlaps of the studied phenological stages of herbs, shrubs,
trees, and bird species in 1961–2021. The blue colour indicates years with a high number of overlap-
ping days, and the red colour indicates years with a low number of overlapping days.
4. Discussion
The main goal of this study was to verify whether long-term (61 years of in situ ob-
servations) phenological data covering 17 species from floodplain forest ecosystems show
phenological (a)synchrony among all observed species and if there are any changes in
correlation in time between phenological onset and climate (mainly temperature) param-
eters and if there are any changes in overlap among phenological stages.
The timing of the phenological phases (e.g., terms of the given phenophases) moved
to an earlier time, and each part of the studied group of species (herbs, shrubs, trees, and
bird species) moved with different intensities. The largest shifts were observed for the
herbs and shrubs (13–14 days on average for all the observed species) and the lowest rate
of shifting was detected for the tree and bird species (8–9 days on average). Similar out-
puts were also published [2] about expressive shifts of herbs in the United Kingdom. The
authors mentioned that the herbs showed the most significant lifeform-specific phenolog-
ical shift (by 32 days during 1753–2019) compared to the shrubs and trees. Nevertheless,
we conclude that these differences among the given group of species of the observed eco-
system should not be generalised because some species responded specifically. The most
expressive shift was observed for the herb Yellow Star-of-Bethlehem (Gagea lutea) by al-
most 21 days (for the phenophase first flower in population), which was higher than the
rest of the herbs; e.g., Herb Paris (Paris quadrifolia), who advanced its phenophases by 10–
11 days, or the full flowering of the Common Lungwort (Pulmonaria officinalis), which
shifted by only 8 days during the whole period (61 years). Similarly, the lowest rate of
shifting was observed for the field maple tree (Acer campestre) and small-leaved lime (Tilia
cordata) with a change of 5.9–6.2 days for bud breaking, which was again lower than the
other shifts of all the observed trees. In general, the shifting of the phenological phases to
earlier times was published and observed by many authors earlier [28,29] and also in re-
cent publications [2,3]. The reason for the most intensive phenological shift of the herbs in
our study could be described by temperature trends. Linear trends in 1961–2021 showed
the most significant changes in the mean and maximum air temperatures in February and
March (see Table 2). The average date of the first three phenophases of the herbs (the first
flower of Pulmonaria officinalis, Gagea lutea, and Corydalis cava) was at the end of March
(see Table 1), which means that these species had to face the warming with the highest
trend (for the maximum and mean temperature). Regardless, the phenophases of the
shrub Cornus mas were occurring midway and at the end of March and the shifting was
not as intensive as for the herbs (only by 11.5 and 9 days for the first and full flowers).
This means that each species may have different sensitivities to climate conditions and
Figure 6.
Overlap index for all observed overlaps of the studied phenological stages of herbs,
shrubs, trees, and bird species in 1961–2021. The blue colour indicates years with a high number of
overlapping days, and the red colour indicates years with a low number of overlapping days.
4. Discussion
The main goal of this study was to verify whether long-term (61 years of in situ ob-
servations) phenological data covering 17 species from floodplain forest ecosystems show
phenological (a)synchrony among all observed species and if there are any changes in cor-
relation in time between phenological onset and climate (mainly temperature) parameters
and if there are any changes in overlap among phenological stages.
The timing of the phenological phases (e.g., terms of the given phenophases) moved to
an earlier time, and each part of the studied group of species (herbs, shrubs, trees, and bird
species) moved with different intensities. The largest shifts were observed for the herbs and
shrubs (13–14 days on average for all the observed species) and the lowest rate of shifting
was detected for the tree and bird species (8–9 days on average). Similar outputs were
also published [
2
] about expressive shifts of herbs in the United Kingdom. The authors
mentioned that the herbs showed the most significant lifeform-specific phenological shift
(by 32 days during 1753–2019) compared to the shrubs and trees. Nevertheless, we conclude
Atmosphere 2022,13, 978 12 of 17
that these differences among the given group of species of the observed ecosystem should
not be generalised because some species responded specifically. The most expressive shift
was observed for the herb Yellow Star-of-Bethlehem (Gagea lutea) by almost 21 days (for
the phenophase first flower in population), which was higher than the rest of the herbs;
e.g., Herb Paris (Paris quadrifolia), who advanced its phenophases by 10–11 days, or the full
flowering of the Common Lungwort (Pulmonaria officinalis), which shifted by only 8 days
during the whole period (61 years). Similarly, the lowest rate of shifting was observed for
the field maple tree (Acer campestre) and small-leaved lime (Tilia cordata) with a change
of 5.9–6.2 days for bud breaking, which was again lower than the other shifts of all the
observed trees. In general, the shifting of the phenological phases to earlier times was
published and observed by many authors earlier [
28
,
29
] and also in recent publications [
2
,
3
].
The reason for the most intensive phenological shift of the herbs in our study could be
described by temperature trends. Linear trends in 1961–2021 showed the most significant
changes in the mean and maximum air temperatures in February and March (see Table 2).
The average date of the first three phenophases of the herbs (the first flower of Pulmonaria
officinalis,Gagea lutea, and Corydalis cava) was at the end of March (see Table 1), which means
that these species had to face the warming with the highest trend (for the maximum and
mean temperature). Regardless, the phenophases of the shrub Cornus mas were occurring
midway and at the end of March and the shifting was not as intensive as for the herbs (only
by 11.5 and 9 days for the first and full flowers). This means that each species may have
different sensitivities to climate conditions and need different heat sums (the sum of the
daily mean temperatures above a set basic temperature, [
30
]). Overall, there is a concern if
plants continue to flower earlier, and if the frequency, intensity, and duration of climatic
extremes increase further, the functioning and productivity of biological, ecological, and
agricultural systems will be at an unprecedented risk [
2
]. The different timing (mostly
earlier timing) of the phenological phases of plants can affect not only the plant itself and
its species-specific chilling requirements, the likelihood of frost injury, and the demands
on energy and water balance [
4
]. Additionally, other species could be influenced, e.g., the
timing of plant flowering can affect their pollination, especially when insect pollinators are
themselves seasonal and determine the timing of seed ripening and dispersal [
2
]. Plant
flowering also influences animals for which pollen, nectar, fruits, and seeds are important
resources, and earlier flowering implies earlier activity in leaf expansion, root growth, and
nutrient uptake [
4
,
31
]. According to our results and the conditions of flood plain forest
ecosystems, there should also be a positive impact on trees because an earlier and more
vigorous start of the growing season (during the warm spring of 2018) caused a significant
increase in the growth (gross primary production) of the investigated species (oaks, ashes,
and hornbeams—Quercus,Fraxinus, and Carpinus) in the flood plain forest in the vicinity of
our experimental site [32].
A running correlation showed three key outputs. First, herb, shrub, and bird species
were correlated with the maximum temperature during the whole time of observations.
The correlation was stable and high with no specific decline or increase. Next, some of
the tree phenophases showed a decreasing correlation with temperature since a specific
time. According to the first output, we conclude that it is the temperature (in our case the
maximum temperature) that influences the timing of phenophases the most, which is in
agreement with results by other authors; e.g., [
3
] states that the phenological advancing
trends still could be attributed to warming and that temperature is still considered to be
the fundamental prerequisite of climate change detection and attribution for phenological
changes [
7
,
33
]. However, there is also evidence that spring phenological development is
partially influenced by the fulfilment of winter chilling and by the photoperiod [
34
]. The
temperature sensitivity (for four dominant tree species in Europe) of the leafing phenology
decreased during 1981–2013 and it was because the accumulation of the chilling units
was reduced [
35
]. Additionally, other researchers [
36
,
37
] have stated that there are differ-
ent changes in temperature sensitivity and that species-specific increases in temperature
requirements are mainly caused by a decrease in chilling [
36
]. In our case, the running
Atmosphere 2022,13, 978 13 of 17
correlation for three trees and the temperature (same results for the maximum, mean, and
minimum temperature) also showed decreasing temperature sensitivity in the last decades.
It was not the aim of our paper to also incorporate chilling units within the analyses but it
needs to be performed in future studies to see the possible reason for the decreasing tem-
perature sensitivity of trees. Others state that plants must respond not only to temperature
and warming but also to soil water availability [
38
]. Nevertheless, we did not assess soil
water availability in this study; in the case of floodplain forests, the groundwater level is
high year-round and precipitation is not as important for the development of plants. The
Plaˇcek’ Forest hydrological ratio is specific and exceptional. The water regime is mainly
driven by the river Svratka and also by the brook Šatava (which runs in the National
Reserve). Until the seventies (when regulation of the river Svratka was done), the period of
flooding was common and huge [
39
]. During this time, there were some remnants of water
diversions and thanks to gravel sand subsoil, the level of groundwater is still connected
with the river Svratka; nevertheless, the flooding is not as huge as it was before stream
regulations in the seventies [
40
]. However, the rate of flooding and the water regime in
the forest is also driven by precipitation, which is not saturated on a loess that is on the
hillside northward from Plaˇcek’ Forest and brings more water in the hydrological ratio [
39
].
However, it must be stressed that although observations took place in the floodplain forest,
drying oaks have been observed not only at the experimental site but also in other areas of
the floodplain forest in the last 10 years. During the last 30-year period (1991–2020), the
mean maximum temperatures in March, April, and May (MAM) experienced a statistically
significant warming trend according to 1961–2020. Additionally, the MAM temperature
from 1501 to 2020 exhibited the highest temperature in the last 30-year period. Two drought
indices (SPEI—the standard precipitation evapotranspiration index, PDSI—the Palmer
Drought Severity Index) agreed on the driest 30-year interval of 1990–2019 for the spring
period (months MAM) since 1500 [41,42].
Finally, the running correlation analysis among the phenophases of all the species
showed that the given tree and herb phenophases correlated well until a specific time
after that, in which the correlation decreased. The decreasing running correlation was
also noted for the average phenological terms of the herbs, shrubs, trees, and birds. This
means that there were changes in the phenological timing and that the phenophases that
were synchronised during the first decades of observations changed the onset date, which
resulted in higher variability in synchrony and lower values of the correlation coefficients
during the last decades. At the same time, the phenological overlap analysis confirmed that
significant phenological shifts to the earlier time of herbs contributed to the lengthening
and shortening of the overlaps among the given phenophases of mainly the herbs and
trees. Mismatched species should respond differently to environmental cues relevant to
climate change, not only air temperature but also winter chilling or photoperiod as was
mentioned earlier. Different responses can lead to different multi annual trajectories (in
traits linked to climate such as leaf-out times, caterpillar emergence, or egg laying) in
species’ phenophases because the relative importance of each environmental cue might
differ among species—e.g., one species relying more on day length and thus responding
less to temperature increases than the other—and because, even if species are solely driven
by temperature, species might rely on different temperature ranges and/or time intervals to
time their phenophases [
31
]. In our case, the results showed that temperature (specifically
the maximum and mean temperature) is the predictor for the timing of phenophases.
However, two major backup systems to prevent a premature spring development (chilling
and photoperiod) may influence the timing also [
3
]. Specifically, the photoperiod may play
an important role for early herbs and shrubs (although the required heat sum is reached for
the onset of the phenophase, the short photoperiod stops its onset). Additionally, as was
mentioned earlier, for trees that showed a lower temperature sensitivity in the last years,
chilling may be another decelerating mechanism [35].
Nevertheless, there are published articles describing an asynchrony or mismatch
among the observed species with negative impacts on these interactions and the food
Atmosphere 2022,13, 978 14 of 17
chain among those species [
8
,
43
,
44
]; there are also a few clear cases of mismatch due to
the ongoing rapid climate change [
45
–
47
]. Asynchrony in phenological timing should
not always have a negative impact; changes in the onsets could have a neutral [
48
] or, in
some cases, a positive impact [
49
]. Our results showed fast and largest reactions by the
herbs and shrubs, resulting in a lower correlation among the other species in the ecosystem.
Additionally, because early herbs or shrubs are not trophically connected with trees or bird
species, the changes in the onsets may have a neutral impact. However, if the growing
season starts earlier, plants take moisture from the soil for a longer period and can negatively
affect the species that come later. Still, the knowledge about phenological interactions or
plasticity in wild nature can serve as a test baseline and show us the natural reactions of
plants to changing climate conditions. Plasticity plays a key role, which has an important
impact on phenotypic change and is a crucial strategy for coping with temporal and spatial
heterogeneity [
39
]. Plasticity and microevolution are the most important for trophically
linked populations and variations among individuals can indicate what can be expected
in upcoming climate conditions [
2
]. The most expressive shifts of herbs suggests that
directional selection and adaptive evolution can be faster in short-lived plants that exhibit
faster turnover rates, while phenotypic plasticity is more important for long-lived trees and
shrubs [
2
]. Our phenological dataset of species is not robust enough to make conclusions
about adaptive evolution but still brings information about the phenological reactions of
plants and bird species on the specific location of flood plain forest and showed a decreasing
temperature sensitivity for trees and also a decreasing synchrony in phenological timing
among the observed species.
5. Conclusions
The principal finding of this study was that the correlation among the given observed
species of the ecosystem became weaker over time, mainly since 1982–2001. Long-term
phenological data of the flood plain forest ecosystem from 1961 to 2021 showed that
there were changes in the timing of all the observed species from the flood plain forest
ecosystem. Herbs, followed by shrubs, exhibited the most expressive shifts compared to
the tree and bird species, and these changes also caused lengthening or shortening in the
phenological overlaps among specific phenological phases. The running correlation of all
the phenophases showed a decline in correlation from 1991–2010 for the pairs of herbs with
birds and shrubs with trees. The data also showed a decline in correlation among specific
tree and herb phenophases from the period of the 1970s to the 1990s. The most significant
changes in phenological timing were detected for herbs, in which the phenophases started
in the second half of March and were the most correlated with the maximum temperature
in the preceding months (February and March).
Author Contributions:
Conceptualization, L.B. and M.T.; methodology, L.B., M.F. and M.T.; software,
J.B. (Jan Balek); validation, P.Š. and P.Z.; formal analysis, L.B. and P.D.; investigation, J.B. (Jana
Bauerová); resources, M.T. and Z.Ž.; data curation, L.H. and M.M.; writing—original draft preparation,
L.B.; writing—review and editing, L.B.; visualization, M.B.; supervision, M.F., M.T. and Z.Ž.; project
administration, M.T. and Z.Ž.; funding acquisition, M.T. and Z.Ž. All authors have read and agreed
to the published version of the manuscript.
Funding:
This research was funded by Ministry of Education, Youth and Sports of the Czech
Republic the project SustES—Adaptation strategies for the sustainability of ecosystem services and
food security in adverse natural conditions (CZ.02.1.01/0.0/0.0/16_019/0000797).
Data Availability Statement: Not applicable.
Acknowledgments:
L.B., P.D., M.F., J.B. (Jan Balek), M.B., P.Z., P.Š., Z.Ž. and M.T. were supported
by the project SustES—Adaptation strategies for sustainable ecosystem services and food security
under adverse environmental conditions (CZ.02.1.01/0.0/0.0/16_019/0000797). In loving memory
of our colleague Zdenˇek Bauer (*1931
†
2016) who found the experiment and made 80% of all the
phenological observations and together with his wife J.B. (Jana Bauerová) prepared a significant part
of the data from the period 1961–2010. At the same time, he actively participated in the data analysis
Atmosphere 2022,13, 978 15 of 17
until 2016 and significantly contributed to the preparation of the concept of the whole study; its first
part was published 12 years ago. The Global Change Research Institution CAS has permission to use
the historical data.
Conflicts of Interest: The authors declare no conflict of interest.
Appendix A
Atmosphere 2022, 13, x FOR PEER REVIEW 16 of 18
Appendix A
Figure A1. Trend in the length of overlap (days) among 62 phenological stages in total. Asterisks
indicate a significant lengthening of overlaps; points indicate significant shortening of overlaps. The
green colour indicates the lengthening of overlaps, and the red colour indicates the shortening of
overlaps. Positive values indicate the number of days by which the overlaps were extended. Nega-
tive values indicate the number of days by which the overlaps were shortened.
References
1. Beard, K.H.; Kelsey, K.C.; Leffler, A.J.; Welker, J.M. The Missing Angle: Ecosystem Consequences of Phenological Mismatch.
Trends Ecol. Evol. 2019, 34, 885–888. https://doi.org/10.1016/j.tree.2019.07.019.
2. Büntgen, U.; Piermattei, A.; Krusic, P.J.; Esper, J.; Sparks, T.; Crivellaro, A. Plants in the UK flower a month earlier under recent
warming. Proc. R. Soc. B 2022, 289, 20212456. https://doi.org/10.1098/rspb.2021.2456.
3. Menzel, A.; Yuan, Y.; Matiu, M.; Sparks, T.; Scheifinger, H.; Gehrig, R.; Estrella, N. Climate change fingerprints in recent Euro-
pean plant phenology. Glob Chang. Biol. 2020, 26, 2599–2612. https://doi.org/10.1111/gcb.1500O.
4. Piao, S.; Liu, Q.; Chen, A.; Janssens, I.A.; Fu, Y.; Dai, J.; Liu, L.; Lian, X.; Shen, M.; Zhu, X. Plant phenology and global climate
change: Current progresses and challenges. Glob. Chang. Biol. 2019, 25, 1922–1940. https://doi.org/10.1111/gcb.14619.
5. Wolkovich, E.M.; Cook, B.I.; Allen, J.M.; Crimmins, T.M.; Betancourt, J.L.; Travers, S.E.; Pau, S.; Regetz, J.; Davies, T.J.; Kraft,
N.J.B.; et al. Warming experiments underpredict plant phenological responses to climate change. Nature 2012, 485, 494–497.
https://doi.org/10.1038/nature11014.
6. Cohen, J.M.; Lajeunesse, M.J.; Rohr, J.R. A global synthesis of animal phenological responses to climate change. Nat. Clim. Chang.
2018, 8, 224–228. https://doi.org/10.1038/s41558-018-0067-3.
7. Thackeray, S.J.; Henrys, P.A.; Hemming, D.; Bell, J.R.; Botham, M.S.; Burthe, S.; Helaouët, P.; Johns, D.G.; Jones, I.D.; Leech, D.I.;
et al. Phenological sensitivity to climate across taxa and trophic levels. Nature 2016, 535, 241–245. https://doi.org/10.1038/na-
ture18608.
Trend in the lenght of
overlaps/days
Common Lungwort
(Pulmonaria officinalis, L.)
Yellow Star-of -Bethlehem
(Gagea lutea, L.)
Bulbous Corydalis
(Corydalis cava, L.)
Yellow Wood Anemone
(Anemone ranunculo ides, L)
Herb Paris
(Paris quadrifolia , L.)
European Cornel
(Cornus mas , L.)
Blackthorn
(Prunus spinosa , L.)
English Hawthorn
(Crataegus o1yacantha, L.)
Common hawthorn
(Crataegus monogyna , Jacq.)
Common Hornbeam
(Carpinus betul us , L.)
European White Elm