queenright colonies and tracked damage by
multiple replicate colonies that were simul-
taneously exposed to environments differing
in floral resource availability. The 2019 study
made use of the same rooftop employed in 2018
(roof1)aswellasasecondnearbyrooftop
(~200 m from the first). Eight queenright
B. terrestrishives were placed on each roof,
adjacent to focal patches comprising 300 flow-
erless plants (from seven different species; see
supplementary materials for details), which
were replaced after 3 weeks. No plants in flower
were present on roof 1. On roof 2, bees had
access to a preexisting rooftop garden planted
with wildflowers (measuring ~4.5 m by 7 m,
located ~20 m from the focal patch) as well as
a patch of 30 plants in flower (Antirrhinum
sp.) placed adjacent to the focal patch of flow-
erless plants. This experiment was initiated
in late May, and we monitored damage on
150 randomly chosen plants from each focal
patch. At the end of June, the wildflower gar-
den was mown and other plants in flower were
removed from roof 2 (see supplementary ma-
terials for detailed methods). Bees damaged
plants in both focal patches; however, damage
levels were consistently higher on roof 1, where
no floral resources were present. Furthermore,
damage levels on roof 2 increased significantly
in the month after the wildflower garden was
mown (before tapering off on both roofs as hives
approached their reproductive switch points).
These results provide further evidence that dam-
age behavior is influenced by the availability
of floral resources, complementing the find-
ings from our 2018 rooftop study as well as our
pollen-deprivation laboratory experiments.
Taken together, the results of our laboratory
and rooftop experiments show that (i) bumble
bees damage the leaves of flowerless plants in
a way that accelerates flowering time; (ii) the
availability of pollen influences this behavior;
(iii) the behavior persists under seminatural
conditions where bees have the option of
foraging farther afield; and (iv) wild bees, in-
cluding species other thanB. terrestris,dam-
age plant leaves. These findings thus establish
that leaf-damaging behavior is a notable com-
ponent of bumble bee behavior that can influ-
ence the local availability of floral resources.
The current findings do not establish the mech-
anism by which bee damage accelerates flower-
ing or the relationship of the observed response
to previously documented pathways involved
in stress-induced flowering ( 23 ). The fact that
mechanical damage alone does not fully re-
produce the observed effect raises the possi-
bility that some additional bee-derived cue
maybeatplay.
Flower scarcity poses serious challenges for
pollinator populations ( 26 ), and bumble bees
have a particularly pressing need for floral
resources during early spring, when queens
emerging from diapause must establish their
colonies ( 27 ). Damaging plant leaves when
floral resources are scarce might therefore be
an adaptive strategy for accelerating flower
production. Conversely, because pollinators
are powerful agents of plant evolution ( 28 , 29 ),
it may also be adaptive for some plant species
to actively respond to bee-inflicted damage or
associated cues ( 30 ) to mitigate asynchrony.
Furthermore, the demonstration that bee-
inflicted leaf damage can have strong effects
on time to flowering may have important eco-
logical implications, including for the resilience
of plant–pollinator interactions to increases in
phenological asymmetry caused by anthropo-
genic environmental changes.REFERENCES AND NOTES- T. Benton,Bumblebees: The Natural History and Identification
of the Species Found in Britain, vol. 98 ofCollins New Naturalist
Library(Harper UK, 2006). - W. Faganet al.,Theor. Ecol. 7 , 289–297 (2014).
- E. E. Crone, N. M. Williams,Ecol. Lett. 19 , 460–468 (2016).
- D. Goulson, E. Nicholls, C. Botías, E. L. Rotheray,Science 347 ,
1255957 (2015). - M. Schenk, J. Krauss, A. Holzschuh,J. Anim. Ecol. 87 ,139–149 (2018).
- J. T. Kerret al.,Science 349 , 177–180 (2015).
- J. Memmott, P. G. Craze, N. M. Waser, M. V. Price,Ecol. Lett.
10 , 710–717 (2007). - N. E. Miller-Struttmannet al.,Science 349 , 1541–1544 (2015).
9. C. Carvellet al.,Nature 543 , 547–549 (2017).
10. J. D. Thomson,Philos. Trans. R. Soc. London Ser. B 365 ,
3187 – 3199 (2010).
11. J. Settele, J. Bishop, S. G. Potts,Nat. Plants 2 , 16092 (2016).
12. T. H. Roulston, K. Goodell,Annu. Rev. Entomol. 56 ,293–312 (2011).
13. G. Kudo, Y. Nishikawa, T. Kasagi, S. Kosuge,Ecol. Res. 19 ,
255 – 259 (2004).
14. F. Requier, K. K. Jowanowitsch, K. Kallnik, I. Steffan-Dewenter,
Ecology 101 , e02946 (2020).
15. J. Forrest, J. Thomson,Ecol. Monogr. 81 , 469–491 (2011).
16. R. J. Stelzer, L. Chittka, M. Carlton, T. C. Ings,PLOS ONE 5 ,
e9559 (2010).
17. P. Willmer,Curr.Biol. 22 , R131–R132 (2012).
18. I. Bartomeuset al.,Proc. Natl. Acad. Sci. U.S.A. 108 ,
20645 – 20649 (2011).
19. J. E. Ogilvieet al.,Ecol. Lett. 20 , 1507–1515 (2017).
20. C. Parmesan,Annu. Rev. Ecol. Evol. Syst. 37 , 637–669 (2006).
21. M. Mikát, C. Franchino, S. Rehan,Behav. Ecol. Sociobiol. 71 , 135
(2017).
22. L. D. Harder,Oecologia 85 ,41–47 (1990).
23. K. Takeno,J. Exp. Bot. 67 , 4925–4934 (2016).
24. N. P. Lemoine, D. Doublet, J.-P. Salminen, D. E. Burkepile,
J. D. Parker,Ecology 98 , 1817–1828 (2017).
25. P. Smeets, M. Duchateau,Apidologie 34 , 333–337 (2003).
26. P. Williams, J. Osborne,Apidologie 40 , 367–387 (2009).
27. R. Mallinger, J. Gibbs, C. Gratton,Landsc. Ecol. 31 ,1523–1535 (2016).
28. D. D. Gervasi, F. P. Schiestl,Nat. Commun. 8 , 14691 (2017).
29. A. A. Agrawal, A. P. Hastings, M. T. J. Johnson, J. L. Maron,
J.-P. Salminen,Science 338 , 113–116 (2012).
30. M. C. Mescher, C. M. De Moraes,J. Exp. Bot. 66 ,425–433 (2015).
31. F. G. Pashalidou, H. Lambert, T. Peybernes, M. Mescher,
C. De Moraes, Bumble bees damage plant leaves and accelerate
flower production when pollen is scarce, Version 2, Dryad
(2020); https://doi.org/10.5061/dryad.9ghx3ffdv.
ACKNOWLEDGMENTS
The authors thank L. Eyman and G. Ulrich for help with experiments,
G. Losapio for statistical advice, and H. Pulido for photography.
Funding:This research was funded by ETH Zürich, Switzerland.
Author contributions:F.G.P., H.L., M.C.M., and C.M.D.M. designed
the research. F.G.P., H.L., and T.P. carried out the research. F.G.P.,
H.L., T.P., M.C.M., and C.M.D.M. interpreted the data. F.G.P., H.L.,
M.C.M., and C.M.D.M. wrote the paper. All authors gave comments
and final approval for publication.Competing interests:The
authors declare no competing interests.Data and materials
availability:All data are available in the manuscript, the
supplementary materials, or at Dryad ( 31 ).SUPPLEMENTARY MATERIALS
science.sciencemag.org/content/368/6493/881/suppl/DC1
Materials and Methods
Figs. S1 to S5
Tables S1 to S3
References ( 32 , 33 )
Movie S1
17 May 2019; resubmitted 31 January 2020
Accepted 10 April 2020
10.1126/science.aay0496Pashalidouet al.,Science 368 , 881–884 (2020) 22 May 2020 4of4
RESEARCH | REPORT