Yearly fluctuations of flower landscape in a Mediterranean scrubland: Consequences for floral resource availability
Yearly fluctuations of flower landscape in a Mediterranean scrubland: Consequences for floral resource availability
VÂõctor Flo 1 2
Jordi Bosch 1 2
Xavier Arnan 1 2
Clara Primante 1 2
Ana M. MartÂõn GonzaÂ lez 0 1 2
Helena Barril-Graells 1 2
Anselm Rodrigo 1 2
0 Center for Macroecology , Evolution and Climate, Natural History Museum of Denmark , University of Copenhagen , Universitetsparken, Copenhagen Ø, Denmark, 3 Ecology Unit , Autonomous University of Barcelona , Bellaterra , Spain
1 CREAF, Center for Ecological Research and Forestry Applications , Campus UAB, Bellaterra , Spain
2 Editor: William Oki Wong, Indiana University Bloomington , UNITED STATES
Species flower production and flowering phenology vary from year to year due to extrinsic factors. Inter-annual variability in flowering patterns may have important consequences for attractiveness to pollinators, and ultimately, plant reproductive output. To understand the consequences of flowering pattern variability, a community approach is necessary because pollinator flower choice is highly dependent on flower context. Our objectives were: 1) To quantify yearly variability in flower density and phenology; 2) To evaluate whether changes in flowering patterns result in significant changes in pollen/nectar composition. We monitored weekly flowering patterns in a Mediterranean scrubland community (23 species) over 8 years. Floral resource availability was estimated based on field measures of pollen and nectar production per flower. We analysed inter-annual variation in flowering phenology (duration and date of peak bloom) and flower production, and inter-annual and monthly variability in flower, pollen and nectar species composition. We also investigated potential phylogenetic effects on inter-annual variability of flowering patterns. We found dramatic variation in yearly flower production both at the species and community levels. There was also substantial variation in flowering phenology. Importantly, yearly fluctuations were far from synchronous across species, and resulted in significant changes in floral resources availability and composition at the community level. Changes were especially pronounced late in the season, at a time when flowers are scarce and pollinator visitation rates are particularly high. We discuss the consequences of our findings for pollinator visitation and plant reproductive success in the current scenario of climate change.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
Competing interests: The authors have declared
that no competing interests exist.
Flower production and flowering phenology are species-specific traits that are strongly
constrained by life form and phylogeny [
]. However, these traits are also influenced by extrinsic
factors such as weather variables [3±7] and nutrient availability [8±10], and therefore show a
certain level of variation across years [11±13]. Inter-annual variability in flowering patterns
may have important consequences in terms of attractiveness to pollinators and flower
predators, and ultimately affect plant reproductive output. For example, large floral displays
(number of flowers per individual) (reviewed in ) and high flower densities [15±18] usually
result in increased pollinator visitation rates. Yearly variability in flowering patterns may also
have a strong impact on the reproductive success of flower-visiting insects. In situations in
which a given species does not bloom profusely, flower visitors foraging on this plant take
longer to gather food resources [
], and may be forced to switch to alternative, non-preferred
flower species .
Interest in yearly variability in flowering time has increased in the current scenario of
climate change, especially in relation to potential temporal mismatches between plants and their
pollinators [7,22±27]. On the other hand, much less attention has been given to variation in
flower production [
]. Most studies have focused on individual species and
studies at the community level remain scarce. A community approach is essential to fully
understand the consequences of variation in flowering patterns for interacting species such as
pollinators, because flower choice by pollinators is highly dependent on floral context
(abundance and composition of co-flowering species) [
]. If yearly changes in flowering
patterns are pronounced and, most importantly, if they are not synchronous across species in
the community, flower-visiting insects may encounter a different "flower landscape" from year
to year. Such temporal variability could have important consequences for the structure of
plant-pollinator interactions and pollination function. Alternatively, if changes in flowering
patterns are small and/or all species in the community fluctuate in parallel, flower composition
and interactions with pollinators may remain relatively consistent through time, a scenario
that would favour specialization in plant-pollinator interactions .
In this study, we monitored flower production and flowering phenology in a coastal
scrubland community (23 species) over 8 years. Our specific objectives are: 1) To quantify yearly
variability in flowering patterns in terms of flower density and phenology. We are interested in
establishing whether the various species in the community fluctuate synchronously or not.
Our flower community is strongly seasonal [
]. For this reason, we are also interested in
potential interactions between inter-annual and seasonal (intra-annual) variability; 2) To
evaluate to what extent changes in flowering patterns result in significant changes in flower
resource (pollen and nectar) availability and composition.
The study area is a Mediterranean scrubland located in the Garraf Natural Park, near
Barcelona, NE Spain. Field work was conducted with permission of the Park's administration. The
study plot (~ 1ha; UTM: 409345.0, 4569737.5) is located 340 m above sea level and 1700 m
from the coastline. The vegetation is dominated by Quercus coccifera, Pistacia lentiscus,
Thymus vulgaris and Rosmarinus officinalis. The climate is characterized by warm dry summers
and mild winters. Most precipitation occurs in autumn and spring. Mean annual temperature
is 15.5ÊC and yearly rainfall is around 600 mm.
Data collection took place from March to June of years 2006 to 2014 (except 2010). Bloom
becomes extremely scarce in July due to the severe summer drought. We used 6 permanent 50
x 1 m transects forming a grid. Distance between adjacent parallel transects was 20 m. Once a
2 / 14
week, all open flowers of entomophilous species were counted in these transects. Flower life
span was less than a week in all species. We analyze the data corresponding to the 23 most
abundant species, representing 15 plant families, and amounting to 99.4% of the total number
of flowers counted throughout the study.
We characterized the flowering pattern of each species based on the total number of flowers
produced per hectare, year and transect (flowering density), the duration (in weeks) of the
flowering period (flowering duration) and the date (week; week 1 = first week of March) of
maximum flowering intensity (flowering peak). We do not analyze date of flowering onset
because it was strongly correlated to flowering peak date every year (Pearson's correlation,
n = 23 species, r = 0.67±0.91; all p < 0.001), in addition, peak is a better measure for
phenological comparisons [
]. Although in some years Rosmarinus officinalis started its bloom earlier
than our flower-sampling campaign, we always captured most of the flowering period. Since
sampling dates were not exactly the same each year, we use linear interpolations to work with
the same days of the year (DOY) across the eight years. Nine species did not bloom at all in at
least one of the eight years. For these species, flowering peak and flowering duration values are
calculated excluding years of null flowering.
Nectar and pollen production
To understand whether fluctuations in flowering pattern resulted in temporal changes in floral
resource composition we measured pollen (mm3) and nectar production per flower (mg of
sugar in 24 h) in the 23 plant species. These measurements were taken in 2006 and 2007,
depending on the species. On haphazardly chosen plants, we used nylon bags to enclose
branches with mature buds. After 24 h we extracted and measured the nectar accumulated in
individual flowers by using Drummond micropipettes of 0.25, 0.50 and 1 μl (samples size =
18±144 flowers per species; S1 Table). To measure flower pollen content, we selected between
10 and 15 flower buds per species. Flower buds were removed and kept in vials with 70%
ethanol and taken to the laboratory. Each flower bud was dissected under a stereomicroscope and
the number of anthers was counted. Then, three anthers per flower were removed, suspended
in 2 ml of 70% ethanol and sonicated in a water bath for 2±4 minutes to dislodge pollen grains
and 9 ml of isotonic solution was added. The number of pollen grains in the resulting
suspension was then estimated using an electronic particle counter (Coulter Multisizer) with 200 μm
aperture. From these data we obtained the total number of pollen grains produced per flower.
The use of data from two years for the entire eight-year series assumes that intra-specific pollen
and nectar production per flower was consistent throughout the duration of the study. Both
pollen and nectar production per flower are known to vary at the species level from year to
year [36±38]. However, because in our study inter-specific differences in pollen and nectar
production per flower are as high as 276-fold and 178-fold, respectively (see Results), we
assume that the relatively much smaller intra-specific yearly differences will not significantly
alter our community level results. Flower density data and pollen/nectar production per flower
were combined to obtain estimates of pollen and nectar production per ha.
Variability in flowering patterns. We used Linear Mixed-effect Models (LMMs) to
analyze the effect of year, species and the interaction between year and species on flowering
density, flowering peak and flowering duration. A significant year effect and lack of significant
year x species interaction would indicate that all species in the community fluctuate more or
less in synchrony. Transect was added as a random factor because the same transects were
repeatedly sampled over the years. Eight species did not bloom at all in at least one of the eight
3 / 14
years. For these species/years, flower density was 0 and flowering peak and flower duration
could not be calculated. Consequently, these species were removed from the analyses.
Flowering density was log-transformed and residuals were checked to satisfy normality and
homoscedasticity assumptions. Analyses were performed with the lmer function of lme4 package [
in R (version 3.3.3, R Core Team, Vienna (Austria), 2017, https://www.R-project.org).
To analyze yearly and seasonal variability in overall (sum of all species) flower density, we
conducted a LMM with flower density as the dependent variable and year and month (March,
April, May, June) as factors. The month x year interaction was also included in the model.
Transect was added as a random factor. The same analysis was subsequently repeated with
nectar and pollen production data. Flower, nectar and pollen data were log-transformed to
achieve normality. Nectar and pollen results were very similar. For this reason, only nectar
results are shown.
To analyse yearly and seasonal variation in flower composition we conducted
permutational multivariate analyses of variance (PERMANOVA). The response variable was a
BrayCurtis dissimilarity matrix of flower composition (flower density of each species) between the
different sampling events (combinations of years and months). Year, month and their
interaction were included as predictors, and transects were selected as groups (strata) within which
permutations were restricted. The analysis was executed with 9999 permutations, using Adonis
function of vegan package [
] in R. To facilitate interpretation of the results, non-metric
multidimensional scaling (NMDS) was performed based on flower composition Bray-Curtis
distances using the metaMDS function of vegan package in R. Subsequently, equivalent
PERMANOVA and NMDS analyses were conducted with nectar and pollen composition data.
Again, because pollen and nectar results were very similar, we only show nectar results. All
analyses were conducted without 10 species that had missing values in some transects during
the entire season for the 8 years of sampling.
Phylogenetic constraints. To find out whether inter-annual flowering pattern variability
was constrained by phylogeny, we conducted tests of phylogenetic signal for continuous
variables based on Bloomberg's K assuming Brownian motion character evolution [
]. We first
built a phylogenetic tree of the 23 species with phylomatic V.3 [
]. Polytomies were resolved
manually using timetree databases [
]. Then, we used the phylosig function of phytools R
] to detect phylogenetic signal in flowering density CV, flowering peak SD, and
flowering duration SD. To estimate the significance of the observed phylogenetic signals, K values
obtained with the real data were compared to K values obtained from randomizations (1000
trees in which species where shuffled across the tips of the phylogenetic tree) [
]. We found
that variability in flower density, flowering peak and flowering duration was not constrained
by phylogeny (S2 Table). That is, phylogenetically related species did not show similar
propensity to variation (or lack thereof) in flowering patterns. Therefore, phylogeny was not
accounted for in subsequent analyses.
Variation in flowering phenology and density and consequences for floral
Yearly variation in flower density was strongly species-dependent (Fig 1). Some species (e.g.,
Cistus albidus, Euphorbia flavicoma) had relatively consistent flower density, while others (e.g.,
Scorpiurus muricatus, Galium aparine) fluctuated dramatically (flowering density CV > 1.6; S3
Table). Maximum differences (ratio between highest and lowest years) were 267-fold in Leuzea
conifera and 135-fold in Linum strictum (Fig 1), and eight species did not bloom at all in one
or more years. As for flowering phenology, some species showed important shifts in flowering
4 / 14
Fig 1. Yearly flowering curves of the 23 most abundant plant species of the Garraf community. Ordered by timing of peak bloom. Note different
scales on y-axis.
PLOS ONE | https://doi.org/10.1371/journal.pone.0191268
5 / 14
peak date (e.g., Anagallis arvensis, Cistus salviifolius) while others were much more consistent
(e.g., Linum strictum, Orobanche latisquama) (S3 Table, Fig 1). Different species also showed
different levels of variation in flowering duration. For example, Scorpiurus muricatus and Iris
lutescens were highly variable, while Euphorbia flavicoma and Biscutella laevigata were
relatively consistent (S3 Table, Fig 1). In all species, the CV of flower density was higher than the
CV of flowering duration (S3 Table). Differences across years in flower density, flowering peak
and flowering duration of the 23 species were highly significant (Table 1). Importantly, the
interaction year x species was also highly significant for the three variables (Table 1), indicating
that yearly variation in flowering patterns was strongly asynchronous across species.
The community was largely dominated by two species, Rosmarinus officinalis and Thymus
vulgaris, which together accounted for 91.7 ± 6.5% of the total (8-year) flower density. Because
these two species showed important maximum yearly differences in flower density (10- and
4-fold, respectively), overall flower availability was highly variable across years (Fig 2, Table 2).
In most years, total flower production ranged from 16 to 21 million flowers / ha (Fig 2).
However, in 2007 and 2012 flower production was only ~ 8 million flowers / ha, and in 2009 it was
~ 28 million flowers / ha. The seasonal distribution of flower density also changed noticeably
across years. In most years, flowering peaked in late March or early April, but in 2014 peak
bloom occurred in mid-March, and in 2012 in early May (Fig 2).
As expected given the strong seasonal component of the flower community, overall flower
density also varied dramatically across months (Table 2). Of more interest is the highly
significant year x month interaction, indicating that annual variation is far from uniform across
seasons (Table 2, Fig 3). Importantly, changes in overall flower density had consequences for
nectar availability, which also presented high inter- and intra-annual variability (Fig 3), and a
highly significant year x month interaction (Table 2). Nectar availability was usually highest in
March, but in 2006 maximum availability occurred in April (Fig 3).
Variation in flower composition and consequences for floral resource
Flower composition differed significantly across years (Table 3, Fig 4A), although the factor
year explained only a small part of the observed variability (9.2%). As mentioned, the Garraf
flower community is largely dominated by two species, Rosmarinus officinalis and Thymus
vulgaris that bloom profusely in early spring. Such dominance is likely to mask changes in flower
composition involving non-dominant species. As expected, flower composition differed
strongly across months (Table 3, Fig 4B), explaining 32.6% of the observed variability. More
importantly, the year x month interaction was also strongly significant (Table 3), indicating
Year x Species
Year x Species
Year x Species
Fig 2. Yearly overall (23 species) flowering curves of the Garraf community. Each curve represents the mean value
across transects (n = 6).
that the expected pattern of seasonal variability was not consistent across years. Yearly
variability in flower composition was greatest at the end of the flowering period (June; Fig 4B).
Variability in nectar composition closely paralleled the results obtained with flower composition
(Table 3). Interestingly, consecutive years often had contrasting nectar compositions (e.g.,
2006 vs. 2007; 2013 vs. 2014).
Our study demonstrates that flowering patterns change dramatically from year to year both at
the species and community levels. At the species level, most variability is due to changes in
flower density, rather than flowering phenology. Some studies have addressed the influence of
environmental factors on flowering patterns (see [
] for a novel methodological approach).
In general, these studies have shown flower production to be positively related to rainfall
]. On the other hand, flowering phenology appears to be mostly regulated by temperature,
with warmer temperatures advancing flowering periods [7,45±47].
Importantly, our study also demonstrates that yearly flowering fluctuations are by no
means synchronized across species, resulting in significant yearly changes in flower
composition, especially late in the season (June). Flower seasonal composition is important because
most pollinator species in Garraf have short activity periods in relation to the overall flowering
Year x Month
Year x Month
Fig 3. Monthly distribution of flower density and nectar availability in the Garraf community across eight years.
period of the community, and the the Garraf pollination network is structured in seasonal
]. By June, flower density in the Garraf scrubland dramatically declines, and
visitation rates (visits per flower and time unit) are very high [
]. At that time, it is not
infrequent to see several pollinators foraging simultaneously on the same inflorescence. Two
species blooming in June, Sideritis hirsuta and Galium aparine show marked fluctuations in
flower density (Fig 1), resulting in drastic yearly changes in flower resource availability. Thus,
June-active pollinators are exposed to floral resources that are both scarce and unreliable from
year to year, a situation that is expected to hinder foraging specialization [
Changes in flower density and composition are known to affect pollinator flower choice
and visitation rates, with potentially important consequences for plant reproductive success
Fig 4. Non-metric multidimensional scaling (NMDS) analysis describing yearly and monthly variation in flower
composition. Ellipses correspond to standard deviations of the sampling events of each grouping factor (years or
months). 1: March; 2: April; 3: May; 4: June. Points on figure (A) represents transect±year values. Points on figure (B)
represents each transect±month±year combination. The results for nectar composition were similar and are shown in
[16,49±51]. Changes in flower visitation rates have been shown to affect stigma pollen
deposition, sometimes resulting in changes in seed-set [
]. Plant reproductive success is also
likely to be affected by changes in pollinator composition. Different pollinator species differ in
9 / 14
their pollinating efficiency both in terms of number of pollen grains delivered per visit [54±
57], and the quality of the pollen deposited (e.g. levels of geitonogamy [
]). Changes in
flower composition may also influence indirect interactions among plant species competing
for pollinators or facilitating pollinator visitation [15,60±63]. Because the effects of flower
neighborhood on pollinator visitation are density-dependent (facilitation may turn into
competition as the facilitating species becomes increasingly abundant) [
], shifts in flower
composition may change the direction of these interactions.
Changes in pollen and nectar availability imply that pollinators are confronted with
inconsistent floral resource landscapes from year to year, with potential ecological consequences for
their fitness. In years with low flower densities, pollinators are forced to fly longer distances to
gather pollen/nectar loads. These increased foraging costs result in slow nest provisioning
rates and decreased offspring body size [
], ultimately leading to increased
developmental and wintering mortality [
]. Long provisioning trips are also likely to result in
increased parasitism by cleptoparasites and parasitoids that enter bee nests and lay their eggs
while the nest founder is away foraging [
In the current scenario of climate change, the Mediterranean Basin is predicted to
experience important temperature increases [
]. This is expected to advance the flowering time
of most plant species [7,45±47]. Pollinators may be able to track these phenological changes if,
as suggested by current evidence [26,71±74], they also respond to global warming by
advancing their activity period. We know less about the potential effects of climate change on flower
production. Increased temperature has been shown to have either positive or negative effects
on flower production depending on the species . However, the predicted increase in the
occurrence of drought episodes in the Mediterranean Basin [
] is expected to increase the
frequency of years with low floral resources [
]. Our study shows shifts in flowering
intensity to be as least as important as shifts in flowering time, and to be highly asynchronous across
species. For this reason, and given the irregular occurrence of drought episodes, we expect
shifts in flowering density to have high negative effects on pollinators.
S1 Table. Mean nectar and pollen production per flower.
S2 Table. Results of analyses exploring phylogenetic (Bloomberg's K test) constraints on
flowering pattern variability.
S3 Table. Descriptive statistics of flower density, flowering peak and flowering duration of
the 23 main plant species of the Garraf community. Species ordered by timing of flowering
S1 Fig. Non-metric multidimensional scaling (NMDS) analysis describing yearly and
monthly variation in nectar composition. Ellipses correspond to standard deviations of the
sampling events of each grouping factor (years or months). 1: March; 2: April; 3: May; 4: June.
Points on figure (A) represents transect±year values. Points on figure (B) represents each
S1 Dataset. Database used in the analyses.
10 / 14
We are very grateful to several students that participated in the flower counts. We thank
Roberto Molowny-Horas for his support in statistical analysis. We also thank two anonymous
reviewers for their helpful comments.
Conceptualization: Jordi Bosch, Anselm Rodrigo.
Data curation: VÂõctor Flo, Clara Primante, Ana M. MartÂõn GonzaÂlez, Helena Barril-Graells.
Formal analysis: VÂõctor Flo.
Methodology: Xavier Arnan.
Supervision: Jordi Bosch, Xavier Arnan, Anselm Rodrigo.
Writing ± original draft: VÂõctor Flo.
Writing ± review & editing: Jordi Bosch, Xavier Arnan, Anselm Rodrigo.
11 / 14
12 / 14
13 / 14
1. Kochmer JP , Handel SN . Constraints and competition in the evolution of flowering phenology . Ecol Monogr . 1986 ; 56 : 303 ± 325 .
2. Petanidou T , Ellis WN , Margaris NS , Vokou D. Constraints on flowering phenology in a Phryganic (east Mediterranean shrub) community . Am J Bot . 1995 ; 82 : 607 ± 620 .
3. Peñuelas J , Filella I , Zhang X , Llorens L , Ogaya R , Lloret F , et al. Complex spatiotemporal phenological shifts as a response to rainfall changes . New Phytol . 2004 ; 161 : 837 ± 846 .
4. Prieto P , Peñuelas J , Ogaya R , Estiarte M . Precipitation-dependent flowering of Globularia alypum and Erica multiflora in Mediterranean shrubland under experimental drought and warming, and its interannual variability . Ann Bot . 2008 ; 102 : 275 ± 85 . https://doi.org/10.1093/aob/mcn090 PMID: 18565983
5. Crimmins TM , Crimmins MA , David Bertelsen C. Complex responses to climate drivers in onset of spring flowering across a semi-arid elevation gradient . J Ecol . Blackwell Publishing Ltd; 2010 ; 98 : 1042 ± 1051 . https://doi.org/10.1111/j.1365- 2745 . 2010 . 01696 .x
6. Lesica P , Kittelson PM . Precipitation and temperature are associated with advanced flowering phenology in a semi-arid grassland . J Arid Environ . 2010 ; 74 : 1013 ± 1017 . http://dx.doi.org/10.1016/j.jaridenv. 2010 . 02 .002
7. Gordo O , Sanz JJ . Impact of climate change on plant phenology in Mediterranean ecosystems . Glob Chang Biol . 2010 ; 16 : 1082 ± 1106 . https://doi.org/10.1111/j.1365- 2486 . 2009 . 02084 .x
8. Sardans J , Rodà F , Peñuelas J. Effects of water and a nutrient pulse supply on Rosmarinus officinalis growth, nutrient content and flowering in the field . Environ Exp Bot . 2005 ; 53 : 1± 11 . https://doi.org/10. 1016/j.envexpbot. 2004 . 02 .007
9. Albrectsen BR , Ericson L , Lundberg P . Nutrient addition extends flowering display, which gets tracked by seed predators, but not by their parasitoids . Oikos . Munksgaard; 2008 ; 117 : 473 ± 480 . https://doi. org/10.1111/j. 2008 . 0030 ± 1299 .16381.x
10. Petraglia A , Carbognani M , Tomaselli M. Effects of nutrient amendments on modular growth, flowering effort and reproduction of snowbed plants . Plant Ecol Divers . Taylor & Francis; 2013 ; 6 : 475 ± 486 . https://doi.org/10.1080/17550874. 2013 .795628
11. Arroyo J . Ritmos climaticos y de floracioÂn en matorrales del SW de España. Lagascalia . 1990 ; 16 : 25 ± 50 .
12. Law B , Mackowski C , Schoer L , Tweedie T. Flowering phenology of myrtaceous trees and their relation to climatic, environmental and disturbance variables in northern New South Wales . Austral Ecol . 2000 ; 25 : 160 ± 178 . https://doi.org/10.1046/j.1442- 9993 . 2000 . 01009 .x
13. Chen Y-Y, Satake A , Sun I-F , Kosugi Y , Tani M , Numata S , et al. Species-specific flowering cues among general flowering Shorea species at the Pasoh Research Forest, Malaysia . J Ecol . 2017 ; 1 ± 13 . https://doi.org/10.1111/ 1365 - 2745 . 12836
14. Ohashi K , Yahara T. Behavioral responses of pollinators to variation in floral display size and their influences on the evolution of floral traits . In: Chittka Lars, Thomson James D., editors. Cognitive Ecology of Pollination Animal Behaviour and Floral Evolution . Cambridge University Press; 2001 . pp. 274 ± 296 .
15. Rathcke B. Competition and facilitation among plants for pollination . In: L R, editor. Pollination biology . New York: Academic publication; 1983 . pp. 305 ± 326 .
16. Kunin WE . Population size and density in pollination: effects pollinator foraging and plant reproductive success in experimental arrays of Brassica kaber . J Ecol . 1997 ; 85 : 225 ± 234 . https://doi.org/10.2307/ 2960653
17. Hegland SJ , Boeke L . Relationships between the density and diversity of floral resources and flower visitor activity in a temperate grassland community . Ecol Entomol . 2006 ; 31 : 532 ± 538 . https://doi.org/10. 1111/j.1365- 2311 . 2006 . 00812 .x
18. LaÂzaro A , Lundgren R , Totland Ø . Co-flowering neighbors influence the diversity and identity of pollinator groups visiting plant species . Oikos . Blackwell Publishing Ltd; 2009 ; 118 : 691 ± 702 . https://doi.org/ 10.1111/j.1600- 0706 . 2008 . 17168 .x
19. Bosch J . Production of undersized offspring in a solitary bee . Anim Behav . 2008 ; 75 : 809 ± 816 . https:// doi.org/10.1016/j.anbehav. 2007 . 06 .018
20. Zurbuchen A , Cheesman S , Klaiber J , MuÈller A , Hein S , Dorn S. Long foraging distances impose high costs on offspring production in solitary bees . Anim Ecol . 2010 ; 79 : 674 ± 81 . https://doi.org/10.1111/j. 1365- 2656 . 2010 . 01675 . x PMID : 20233258
21. Mitchell RJ , Flanagan RJ , Brown BJ , Waser NM , Karron JD . New frontiers in competition for pollination . Ann Bot . 2009 ; 103 : 1403 ± 13 . https://doi.org/10.1093/aob/mcp062 PMID: 19304814
22. Inouye DW , Saavedra F , Lee-Yang W. Environmental influences on the phenology and abundance of flowering by Androsace septentrionalis (Primulaceae) . Am J Bot. Botanical Society of America; 2003 ; 90 : 905 ± 910 . https://doi.org/10.3732/ajb.90.6.905 PMID: 21659185
23. Hegland SJ , Nielsen A , LaÂzaro A , Bjerknes A-L , Totland Ø . How does climate warming affect plant-pollinator interactions? Ecol Lett . Blackwell Publishing Ltd; 2009 ; 12 : 184 ± 195 . https://doi.org/10.1111/j. 1461- 0248 . 2008 . 01269 . x PMID : 19049509
24. Forrest J , Miller-Rushing AJ . Toward a synthetic understanding of the role of phenology in ecology and evolution . Philos Trans R Soc B Biol Sci . 2010 ; 365 : 3101 LP±3112. https://doi.org/10.1098/rstb. 2010 . 0145 PMID: 20819806
25. Diez JM , IbaÂñez I , Miller-Rushing AJ , Mazer SJ , Crimmins TM , Crimmins M a , et al. Forecasting phenology: from species variability to community patterns . Ecol Lett . 2012 ; 15 : 545 ± 53 . https://doi.org/10. 1111/j.1461- 0248 . 2012 . 01765 . x PMID : 22433120
26. Bartomeus I , Ascher JS , Wagner D , Danforth BN , Colla S , Kornbluth S , et al. Climate-associated phenological advances in bee pollinators and bee-pollinated plants . Proc Natl Acad Sci . 2011 ; 108 : 20645 ± 20649 . https://doi.org/10.1073/pnas.1115559108 PMID: 22143794
27. Rafferty NE , Caradonna PJ , Burkle L a , Iler AM , Bronstein JL . Phenological overlap of interacting species in a changing climate: an assessment of available approaches . Ecol Evol . 2013 ; 3 : 3183 ± 93 . https://doi.org/10.1002/ece3.668 PMID: 24102003
28. CaraDonna PJ , Iler AM , Inouye DW . Shifts in flowering phenology reshape a subalpine plant community . Proc Natl Acad Sci . 2014 ; 111 : 4916 ± 4921 . https://doi.org/10.1073/pnas.1323073111 PMID: 24639544
29. Aldridge G , Inouye DW , Forrest JRK , Barr WA , Miller-Rushing AJ . Emergence of a mid-season period of low floral resources in a montane meadow ecosystem associated with climate change . J Ecol . Blackwell Publishing Ltd; 2011 ; 99 : 905 ± 913 . https://doi.org/10.1111/j.1365- 2745 . 2011 . 01826 .x
30. LaÂzaro A , Totland O . Population dependence in the interactions with neighbors for pollination: A field experiment with Taraxacum officinale . J Bot . 2010 ; 97 : 760 ±9. https://doi.org/10.3732/ajb.0900263 PMID: 21622442
31. Petanidou T , Kallimanis AS , Sgardelis SP , Mazaris AD , Pantis JD , Waser NM . Variable flowering phenology and pollinator use in a community suggest future phenological mismatch . Acta Oecologica . 2014 ; 59 : 104 ± 111 . https://doi.org/10.1016/j.actao. 2014 . 06 .001
32. Ebeling A , Klein A -M, Tscharntke T. Plant±flower visitor interaction webs: Temporal stability and pollinator specialization increases along an experimental plant diversity gradient . Basic Appl Ecol . 2011 ; 12 : 300 ± 309 . https://doi.org/10.1016/j.baae. 2011 . 04 .005
33. Bosch J , GonzaÂlez AMM , Rodrigo A , Navarro D . Plant-pollinator networks: adding the pollinator's perspective . Ecol Lett . 2009 ; 12 : 409 ± 19 . https://doi.org/10.1111/j.1461- 0248 . 2009 . 01296 . x PMID : 19379135
34. Filella I , Primante C , Llusià J , MartÂõn GonzaÂlez AM , Seco R , FarreÂ-Armengol G , et al. Floral advertisement scent in a changing plant-pollinators market . Sci Rep . 2013 ; 3 : 3434 . https://doi.org/10.1038/ srep03434 PMID: 24305624
35. Miller-Rushing AJ , Inouye DW , Primack RB . How well do first flowering dates measure plant responses to climate change? The effects of population size and sampling frequency . J Ecol . Blackwell Publishing Ltd; 2008 ; 96 : 1289 ± 1296 . https://doi.org/10.1111/j.1365- 2745 . 2008 . 01436 .x
Walker AK , Barnes DK , Furgala B . Genetic and environmental effects on quantity and quality of alfalfa nectar . Crop Sci. Madison , WI: Crop Science Society of America; 1974 ; 14 : 235 ± 238 . https://doi.org/ 10.2135/cropsci1974.0011183X001400020020x
37. Zywiec M , Delibes M , Fedriani JM . Microgeographical, inter -individual, and intra-individual variation in the flower characters of Iberian pear Pyrus bourgaeana (Rosaceae) . Oecologia . Springer; 2012 ; 169 : 713 ± 722 . https://doi.org/10.1007/s00442-011 -2232-5 PMID: 22200854
38. Khanduri VP , Sharma CM , Kumar KS , Ghildiyal SK . Annual variation in flowering phenology, pollination, mating system, and pollen yield in two natural populations of Schima wallichii (DC.) Korth . Sci World J. 2013 ; 2013 : 1 ± 11 . Available: http://dx.doi.org/10.1155/ 2013 /350157
39. Bates D , Maechler M , Bolker B , Steven W. lme4: Linear mixed-effects models using Eigen and S4 . 2014 . Available: http://cran.r-project. org/package=lme4
40. Oksanen AJ , Blanchet FG , Kindt R , Legen- P, Minchin PR , Hara RBO , et al. Package ªvegan .º 2014 .
41. Blomberg SP , Garland T , Ives AR . Testing for phylogenetic signal in comparative data: behavioral traits are more labile . Evolution (N Y) . 2003 ; 57 : 717 ± 45 . https://doi.org/10.1111/j.0014- 3820 . 2003 .tb00285.x
42. Cam W , Donoghue M. Phylomatic V3 . 2012 [cited 7 Oct 2014 ]. Available: http://www.phylodiversity.net/ phylomatic
43. Kumar S , Hedges SB . TimeTree2: species divergence times on the iPhone . Bioinformatics . 2011 ; 27 : 2023 ± 4 . https://doi.org/10.1093/bioinformatics/btr315 PMID: 21622662
44. Revell MLJ . Package ªphytoolsº. 2016 . Available: https://cran.r-project.org/web/packages/phytools/ phytools.pdf
45. Diekmann M. Relationship between flowering phenology of perennial herbs and meteorological data in deciduous forests of Sweden. Can J Bot . NRC Research Press; 1996 ; 74 : 528 ± 537 . https://doi.org/10. 1139/b96- 067
46. Sparks TH , Jeffree EP , Jeffree CE . An examination of the relationship between flowering times and temperature at the national scale using long-term phenological records from the UK . Int J Biometeorol. United States ; 2000 ; 44 : 82 ± 87 . https://doi.org/10.1007/s004840000049 PMID: 10993562
47. Sherry RA , Zhou X , Gu S , Arnone JA , Schimel DS , Verburg PS , et al. Divergence of reproductive phenology under climate warming . Proc Natl Acad Sci . 2007 ; 104 : 198 ± 202 . https://doi.org/10.1073/pnas. 0605642104 PMID: 17182748
48. MartÂõn GonzaÂlez AM , Allesina S , Rodrigo A , Bosch J . Drivers of compartmentalization in a Mediterranean pollination network . Oikos . Blackwell Publishing Ltd; 2012 ; 121: 2001 ± 2013 . https://doi.org/10. 1111/j.1600- 0706 . 2012 . 20279 .x
49. LaÂzaro A , Jakobsson A , Totland Ø . How do pollinator visitation rate and seed set relate to species' floral traits and community context? Oecologia . 2013 ; 173 : 881 ± 93 . https://doi.org/10.1007/s00442-013 - 2652-5 PMID: 23579571
50. Bruckman D , Campbell DR . Floral neighborhood influences pollinator assemblages and effective pollination in a native plant . Oecologia . 2014 ; 176 : 465 ± 476 . https://doi.org/10.1007/s00442-014 -3023-6 PMID: 25047026
51. Schmid B , Nottebrock H , Esler KJ , Pagel J , Pauw A , BoÈhning-Gaese K , et al. Responses of nectarfeeding birds to floral resources at multiple spatial scales . Ecography . Blackwell Publishing Ltd; 2016 ; 39 : 619 ± 629 . https://doi.org/10.1111/ecog.01621
52. Engel EC , Irwin RE . Linking pollinator visitation rate and pollen receipt . Am J Bot . 2003 ; 90 : 1612 ± 1618 . https://doi.org/10.3732/ajb.90.11.1612 PMID: 21653336
53. Karron JD , Mitchell RJ , Bell JM . Multiple pollinator visits to Mimulus ringens (Phrymaceae) flowers increase mate number and seed set within fruits . Am J Bot . 2006 ; 93 : 1306 ± 1312 . https://doi.org/10. 3732/ajb.93.9.1306 PMID: 21642195
54. Herrera CM . Componentes del flujo geÂnico en Lavandula latifolia Medicus: polinizacioÂn y dispersioÂn de semillas . An del JardÂõn BotaÂnico Madrid . 1987 ; 44 : 49 ± 61 .
55. GoÂmez JM , Zamora R . Generalization vs. specialization in the pollination system of Hormathophylla spinosa (Cruciferae) . Ecology. Ecological Society of America; 1999 ; 80 : 796 ± 805 . https://doi.org/10.2307/ 177018
56. Olsen KM . Pollination effectiveness and pollinator importance in a population of Heterotheca subaxillaris (Asteraceae) . Oecologia . Springer; 1997 ; 109 : 114 ± 121 . https://doi.org/10.1007/PL00008811 PMID: 28307601
57. Sahli HF , Conner JK . Characterizing ecological generalization in plant-pollination systems . Oecologia . Springer; 2006 ; 148 : 365 ± 372 . https://doi.org/10.1007/s00442-006 -0396-1 PMID: 16514533
58. Schmitt J . Pollinator foraging behavior and gene dispersal in Senecio (Compositae) . Evolution; 1980 ; 34 : 934 ± 943 . https://doi.org/10.1111/j.1558- 5646 . 1980 .tb04031. x PMID: 28581139
59. Arnan X , Escolà A , Rodrigo A , Bosch J . Female reproductive success in gynodioecious Thymus vulgaris: pollen versus nutrient limitation and pollinator foraging behaviour . Bot J Linn Soc . 2014 ; 175 : 395 ± 408 . https://doi.org/10.1111/boj.12173
60. Schemske DW . Floral convergence and pollinator sharing in two bee-pollinated tropical herbs . Ecology . 1981 ; 62 : 946 ± 954 . https://doi.org/10.2307/1936993
61. Moeller DA . Facilitative interactions among plants via shared pollinators . Ecology . 2004 ; 85 : 3289 ± 3301 . https://doi.org/10. 1890 /03- 0810
62. Bell JM , Karron JD , Mitchell RJ . Interspecific competition for pollination lowers seed production and outcrossing in Mimulus ringens . Ecology . 2005 ; 86 : 762 ± 771 . https://doi.org/10. 1890 /04- 0694
63. Ghazoul J . Floral diversity and the facilitation of pollination . J Ecol . 2006 ; 94 : 295 ± 304 . https://doi.org/ 10.1111/j.1365- 2745 . 2006 . 01098 .x
64. Peterson AJH , Roitberg BD . Impact of resource levels on sex ratio and resource allocation in the solitary bee, Megachile rotundata . Environ Entomol . 2006 ; 35 : 1404 ± 1410 . Available: https://doi.org/10.1093/ ee/35.5. 1404
65. Zurbuchen A , Landert L , Klaiber J , MuÈller A , Hein S , Dorn S. Maximum foraging ranges in solitary bees: only few individuals have the capability to cover long foraging distances . Biol Conserv. Elsevier Ltd ; 2010 ; 143 : 669 ± 676 . https://doi.org/10.1016/j.biocon. 2009 . 12 .003
66. Bosch J , Kemp WP . Effect of pre-wintering and wintering temperature regimes on weight loss, survival, and emergence time in the mason bee Osmia cornuta (Hymenoptera: Megachilidae) . Apidologie . 2004 ; 35 : 469 ± 479 . https://doi.org/10.1051/apido
67. Goodell K. Food availability affects Osmia pumila (Hymenoptera: Megachilidae) foraging, reproduction, and brood parasitism . Oecologia . 2003 ; 134 : 518 ± 27 . https://doi.org/10.1007/s00442-002 -1159-2 PMID: 12647124
68. Seidelmann K. Open-cell parasitism shapes maternal investment patterns in the Red Mason bee Osmia rufa . Behav Ecol . 2006 ; 17 : 839 . https://doi.org/10.1093/beheco/arl017
69. Hoerling M , Eischeid J , Perlwitz J , Quan X , Zhang T , Pegion P. On the increased frequency of Mediterranean drought . J Clim . American Meteorological Society; 2011 ; 25 : 2146 ± 2161 . https://doi.org/10. 1175/ JCLI-D- 11- 00296 . 1
70. Giorgi F , Lionello P. Climate change projections for the Mediterranean region . Glob Planet Change . 2008 ; 63 : 90 ± 104 . https://doi.org/10.1016/j.gloplacha. 2007 . 09 .005
71. Bosch J , Kemp WP . Effect of wintering duration and temperature on survival and emergence time in males of the orchard pollinator Osmia lignaria (Hymenoptera: Megachilidae) . Environ Entomol. Entomological Society of America; 2003 ; 32 : 711 ± 716 . https://doi.org/10.1603/ 0046 -225X -32.4.711
72. Stefanescu C , Peñuelas J , Filella I. Effects of climatic change on the phenology of butterflies in the northwest Mediterranean Basin . Glob Chang Biol. Blackwell Science Ltd; 2003 ; 9 : 1494 ± 1506 . https:// doi.org/10.1046/j.1365- 2486 . 2003 . 00682 .x
73. Forrest J , Inouye DW , Thomson JD . Flowering phenology in subalpine meadows: does climate variation influence community co-flowering patterns? Ecology . 2010 ; 91 : 431 ± 40 . https://doi.org/10. 1890 /09- 0099 .1 PMID: 20392008
74. Sgolastra F , Bosch J , Molowny-Horas R , Maini S , Kemp WP . Effect of temperature regime on diapause intensity in an adult-wintering Hymenopteran with obligate diapause . J Insect Physiol . 2010 ; 56 : 185 ± 194 . https://doi.org/10.1016/j.jinsphys. 2009 . 10 .001 PMID: 19837077
75. Scaven VL , Rafferty NE . Physiological effects of climate warming on flowering plants and insect pollinators and potential consequences for their interactions . Curr Zool . 2013 ; 59 : 418 ± 426 . https://doi.org/10. 1093/czoolo/59.3.418 PMID: 24009624