1
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scientificreports
Urbanisation modulates plant -
pollinator interactions in invasive
vs. native plant species
sascha Buchholz
1,2 & Ingo K owarik
1,2
P ollination is a k ey ecological process, and invasive alien plant species have been shown to significantly
affect plant-pollinator interactions. Y et, the role of the environmental context in modulating such
processes is understudied. As urbanisation is a major component of global change, being associated
with a range of stressors ( e.g. heat, pollution, habitat isolation ), we tested whether the attractiveness
of a common invasive alien plant ( Robinia pseudoacacia , black locust) vs. a common native plant ( C ytisus
scoparius , common broom ) for pollinators changes with increasing urbanisation. W e exposed blossoms
of both species along an urbanisation gradient and quantified different types of pollinator interaction
with the flowers. Both species attracted a broad range of pollinators, with significantly more visits for
R. pseudoacacia , but without significant differenc es in numbers of insects that immediately ac cessed
the flowers. However , compared to native C ytisus , more pollinators only hovered in front of flowers of
invasive Robinia without visiting those subsequently . The decision rate to enter flowers of the invasive
species decreased with increasing urbanisation. this suggests that while invasive Robinia still attracts
many pollinators in urban settings attractiveness may decrease with increasing urban stressors. R esults
indicated future directions to deconstruct the role of different stressors in modulating plant-pollinator
interactions, and they have implications for urban development since Robinia can be still considered as a
“pollinator -friendly” tree for certain urban settings.
In vasiv e alien plant s p ecies hav e be en reported to be a ma jor driver of c hange in al tering biodiver sity pat terns 1 – 3 .
By co ntrast, a r ecent meta-anal ysis reveals larg ely red ucing or neu tral effects of inva sive plan ts on a nimal abun -
dance, di versity , fitnes s, and ecosystem pr ocesses 4 . Responses of na tive insects to in vasive p lants a re mostly
am biguous 5 , a nd incl ude negativ e 6 , 7 as well as positiv e effects 8 , 9 . By modulating p lant-pollina to r networks 10 – 12 ,
in vasive pla nts can also aff ect key e cological pr o cesses with a high relevance f or pla nt r eproductio n and thus f or
agriculture , foo d pr oduction and food sec urity 13 – 15 , even in urban en vironm ents 16 , 17 .
Plan t-pollinato r int eract ions co mp rise s everal interlink ed processes that r esult in differ ent level s (Fig. 1 ) 18 , 19 .
The first is, fr om the pollinato r perspec tive, the selection of a p lant as a flo ral resource , follow ed by access to the
flow er , either immediat ely or a fter a per iod of hov ering aro und the flor es cence (Fig. 1 ). After accessing the flo wer ,
pollen or nectar can then be collected and carried to its nest o r another p lant.
Pr e vious st udies show ed t ha t alien plant s p ecies can significan t ly modula te impo r tant co mpo nents within
plan t-pollinato r int erac tions a nd networks 20 – 22 . F or exam ple, p lant selection and blos som access can differ
between invasi ve and na tive plan ts 19 , 23 – 25 , with mul tidirect ional pat terns as pollinat ors pr eferring either na tive
or in vasiv e plan ts 18 ,o r in s ome cases nei ther 26 . Studies o n pollen and nectar collect ion fo und that f oo d resour ces
of in vasiv e plan ts can be either neglected by na tive pollina tors 27 , 28 or accepted as n e w fo raging alterna tives 29 , 30 .
Due to com petition fo r p ollina tors, the p resence of in vasive p lants ca n affect flower visi ts 10 , 31 and the pollination
success o f nativ e plan ts, again with differ ent o utco mes 22 , 32 .
Giv en the multidir ect ional effects of in vasi ve plan ts on pla nt-pollina tor netw orks 18 , 22 , 33 , under standing the
underlyin g mechanisms is a k ey challenge. Pr evious studies r eveale d a rang e of mecha nisms rela ted, fo r exam ple,
to flo wer morpho log y 34 , nectar chemistr y and pollen quality 35 , s patial scal e 36 , and bio logica l plasticity o f p olli -
na ting species 9 .
H owever , effects of differ ent en viro nmental settings o n p ollina tor in teractions ha ve receiv ed less at tention 37 .
This is an im portan t researc h question as r ecent s tudies incr e asingly evidence the r ole of str essors r elat ed to
1 Department of Ecology, T echnisc he Universität Berlin, 12165, Berlin, Germany . 2 Berlin-Brandenburg Institute of
Advanced Biodiversity R esearch (BBIB), 14195, Berlin, Germany . Correspondence and requests for materials should
be addressed to S.B. ( email: [email protected] )
R eceived: 24 August 2018
A ccepted: 10 April 2019
P ublished: xx xx xxxx
opeN

2
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
climate change 38 , 39 , land-use chan ges 40 , 41 or urbanisa tion 42 – 44 in modula ting plan t-p ollina tor in teractions an d net -
work s 45 . Whether va r ying levels o f urb anisa tion and associat ed stressors (e.g. hea t, pollution, h abi t at i s ola tion)
affect the a ttract ivenes s of na tive vs. alien plan ts for pollina tors co uld b e revealed by a ppl ying a standar dise d study
design with pair wise a lien/na tive com parisons. T o t he best of our kn owledge, such s tudies ar e missing thu s far .
This is a vi ta l knowledge ga p as ci ties are ho tspots of alien p lant s p ecies 46 , 47 . A t the same time , cities ha ve incr eas ed
in im p ortance as hab ita ts for pollina tors 48 , 49 with a co nspicuous declin e in r ural settings 50 , 51 . A b etter under -
standin g of int eract ions between in vasive p lants, urba nisation a nd p ollina tors will shed light o n underst udied
mechani sms in plan t-pollinato r networ ks 22 and could su pport pollinato r-friendly urban co nser va t ion policies 52 .
Theref or e, we test e d whether urbanisatio n modulated the a ttractiveness o f an in vasive vs. a n ative p lant s p e -
cies fo r p ollina tors a t differen t in teract ion levels, as in dicated in Fig. 1 . In a s t andar dised p air -wise app roach, we
expose d blossom s of Robi nia pseudoacacia L. (black locust; hencefo rt h Robini a ) and C ytis us scop ari us (L.) (com-
mon bro om ; henceforth C ytis us ) to the same type of ecosyst em along an urbani s a tion gradient in Berlin. While
the first species is na tive to N orth America and ha s b een classified as in vasive in E uro pe 53 , the lat ter is na tive to
E urope, b ut in vasive e lse where 54 . Both species share b iological feat ures tha t are r elevan t for pollina tor s, such as
flow er morphology and a ttractive flo wer colour 34 . W e quan tifie d plan t cho ice and accessing of flow ers by na tive
pollinat ors thro ug h direct obser vation. W e differ entiat ed b etween: (i) immediate b lossom access, (ii) ho vering
ar ound flo wers witho ut blossom acces s, and (iii) blossom acces s after ho vering (Fig. 1 ). En viro nmental con ditions
rega rding p lant co mmun ity , flower co verag e and main tena nce wer e similar at each s tudy si te, excep t that the
location was diff eren t in rela tion to levels o f urbanisatio n (A ppendix 1).
W e hypothesised that the na tive an d alien plants did no t differ in their a ttract ivenes s for pollina tors since they
shar ed the s ame flo wer morpho log y , flow er ing time, a nd b oth had highly a ttract ive flo wer colo urs. W e theref ore
expec ted no significan t differences in p lant co ntact of a ny type, immedia te blossom acces s, only ho vering ar ound
flow ers, or blos s om access a fter hov ering. Since urbanisatio n and rela ted en vironmen tal stressors migh t affect
plan t-pollinato r int erac tions, we also h ypothesised t hat the a ttractiveness o f the invasi ve plan t for pollina tors will
decrease with increasin g urbanisation. Exten t of gr een area within the city positiv ely aff ec ts plan t-pollinato r inter -
actions in terms o f be e visita tion which imp lies adverse effects due t o an increased amo unt o f imper vious area 43 .
M ut ua lisms tha t hav e evolved o ver long evo lutio nar y time sc ales – as in the case of na tive plan ts and na tive polli-
na tors – migh t be more r esilient to a nthro pogenic disturbances o r stresso rs than youn ger ones 55 – 57 . W e t heref ore
expec t that y oung er mut ualisms –f or exam ple between nati ve pollinat ors and alien p lants – migh t hav e a lower
resilience a nd a mor e pro ne to dist urbance which in turn should r esult in red uced interaction s such as visita tion
ra tes. C onsequen tly , we expec ted pollinato rs to select R obinia less frequen tly with urban than no n-urban sites.
R esults
General results. A br oad range o f pollinato r taxa visited flo wers of both the na tive a nd the alien plant s p e -
cies. P ollina tors inc luded, with decreasing a bundance Di ptera s. l. (147), H ymenoptera s. l . (114), bumble b ees
(24), hon e y bees (17), and be etles (12), whi le ho verflies, wi ld bees, butt erflies and wasps pla yed a minor r ole
(A p p endix 2).
Figure 1. Generalise d sequence of p lant-pollina tor in teraction in re lation to decisio ns between using na tive vs.
alien plan t resour ces and poten tial interactions wi th differen t levels of urba nisation. Levels in filled boxes wer e
addressed in this stud y .

3
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
Attractiveness of Robinia and C ytisus for pollinators. Robinia at trac ted significan tly mor e p ollina tors
th an C ytisus (Chi 2 = 5.0, df = 1, P = 0.03; generalised linear mixed model, GLMM; Fig. 2a ). Ho wever , t he num ber
of immedia te flow er access at temp ts was the same fo r both spe cies (Chi 2 = 0.6, df = 1, P = 0.44; GLMM; Fig. 2b ).
In co ntras t, the num b er of pollina tor s that o nly hov ered ar ound a flo wer witho ut directly co ntacting it wa s signif -
ican tly higher fo r R obinia t han for C y tis us (Chi 2 = 18.1, df = 1, P < 0.001; GLMM; Fig. 2c ). The decisio n rat e for
con tact with flowers a fter ho vering was significan tly higher for C y tis us com pared to Robinia (Chi 2 = 6.6, df = 1,
P = 0.01; GLMM; Fig. 2d ).
Urbanisation and attractiveness of Robinia . U rbanisa tion was not co rrelat ed with total flower vis -
its, immedia te blosso m access and ho vering ar ound flo wers. Y et, we fo und a significan t urbanisatio n effect on
plan t-pollinato r int eract ions as the decisio n to visit flo wers of Robinia after ho vering decreased with incr easing
percen tage of im per vious ar ea aro und the study s ite (t = 2.8, P = 0.04; generalised linear model, GL M; Fig. 3 ). No
other en viro nmental variable a ffected an y of the respo nse variables.
Discussion
Cities ar e importan t habi tats f or pollinato rs 49 , 58 , and p revious r es earc h has revealed urbanisation eff ec ts on the
com position of wild bee populatio ns 59 – 63 . Mo reo ver , it has been suggested tha t the hig h pr es ence of alien pla nt
species in cities 46 , 47 nega tively a ffects nativ e pollinato rs due to p rob lems of accessi bility with no vel flo wer types
or because o f differences in the q uality of nectar or pollen 18 , 35 . H ow e ver , to what exten t urbanisatio n modu -
lat es interaction s of na tive pollinat ors with alien vs. na tive plan ts rem ains a critical knowledge ga p 37 , 64 . P ollinat or
obser vation s in urban envir onmen ts hav e yielded important in sights in to plan t-pollinato r int erac tions 29 , 43 , 44 , 65 – 68 ,
altho ug h these studies did no t directly com par e nativ e vs. inva sive species. This st udy took a step f or war d by ana-
lysing the in teracting effects of b iological invasio n and urbanisa tion on pla nt-pollina tor in teractions.
Attractiveness of Robinia to native pollinators. Alien plan t species ha ve been demonstra ted to p rovide
valuab le floral resour ces for pollina tors in urba n envir onmen ts 30 , 33 , includin g R obinia in B erlin 44 and Par is 29 an d
C ytis us in its no n-nativ e No rt h American ran ge 69 . A pr e vious st udy fo und weak negativ e effects of R. ps eudoaca -
cia on urban insect comm unities 70 . Y et, Robini a strongly in ves ts in repr oductive orga ns, pr oducing a lar ge flow er
cro p and valuab le nect ar r es our ce 29 , 71 . Ho wever , t he ca p acity o f an in vasive species to pr ovide sui table reso urces
can be differ ent in na tive an d nove l landscapes 72 . W hile honeyb ees had be en reported as ma jor pollina tors 53 , a
recen t study r eveals a range o f wild b ee spe cies also visits flow ers of urban Robinia trees 44 . Our st udy documented
that flo ral resour ces of Robinia at trac ted an even b roader rang e of pollina tor taxa in urban settings, inc luding wild
Figure 2. T otal flower visi ts ( = summa tion of the fo llowing ca tegories b , c, and d) diff ered significan t ly between
the in vasive Rob inia pseudoacacia and the na tive C y ti sus scop ari a (a) ; while immediat e flower access wa s not
significan t ly differ ent (b) . H o vering ar ound flo wers was mo re frequen t in the presence o f R . pseudoacacia (c) ;
bu t p ollina tors decided upo n contact with C. scoparia mor e often aft er hov er ing (d ; GLMM with Ga ussian
distribu tion).

4
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
bees and flies, as well as ho ney be es (A ppendix 2). I n addition to the a ttractivenes s of its flo ral resources, the lo ng
pr esence of Robini a in Euro pe over a period of > 350 years 53 migh t have h elped to int eg rat e the alien plant in to
na tive pollinat or comm unities, as has been sho wn in other cases 5 , 73 .
Due to its la rge flow er crop , R obinia pro vided abunda nt nectar resour ces which at tracted differen t nativ e
pollinat ors amo ng the Dip tera and H ymeno ptera. The n umber of to ta l flow er visits was significan tly higher for
Robinia com pared to C y tis us (Fig. 2a ), whi le there was n ot a significan t difference in immedia te flo wer access
(Fig. 2b ). A cceptance o f in vasive pla nts as food resour ce in the pr esence of a na tive alterna tive has been docu -
men ted previou sly for urba n habi tats 29 , 33 , 44 . H ow ever , as a surprising r esult of o ur study , mor e pollinato rs ho vered
significan tly long er in f ro nt o f R obinia flowers (Fig. 2c ), and significan tly less decided to visit the alien flow ers
after ho vering co mpar ed to thos e that ho ver ed in f ron t of na tive C y tis us flowers (Fig. 2d ). B oth species were
pr esented in an a rray tha t ensur e d free plan t cho ice and easy access. G iven tha t R obinia and C ytisus sh are d a
similar floral mo rphology and blossom colours, tha t ar e ver y at trac tive an d easily recognized by pollina tor s, these
resul ts might indica te a trade-off behavio ur . This means that pollina tor s have t o make economic c hoices abou t
what type of flo wers they visit to in crease b enefits (ener g y intake) com pared to costs (en erg y consum ption) 74 .
Due to these energetic r equiremen ts n utrient a vailab ility in nect ar o r p ollen pla ys a vital role 30 , 35 . Therefo re, o ne
reason f or lo wer decision ra tes fo r R obinia might be rela ted to q ua lity an d suitab ility of the nectar which might
be superior in C ytis us com pared to Robinia . Although hon e y bees successfu lly pr ocess nec tar of Robinia 75 ot he r
non-do mesticated a nd mor e spe cialis ed pollinato rs could be mor e s ensitiv e in their nu t ritional r equiremen ts 30 . In
that case, lo nger ho vering in fro nt o f R obinia flower s withou t subsequent flo wer visits co uld hint a t an ecological
t r ap 76 , since this behaviour co uld disr up t t he metabolic cost/benefit balance, en erg y exp ended for n o reward 77 .
H owever , analysis o f R obinia nec tar and pollen quality did no t sup p ort this hypothesis, r evealing high cont ents
of sui table amin o acids, ph ytoster ol and suga r 29 . A part from en ergetic requir emen ts that aff ec t flow er selec tion,
pollinat ors ar e faced with other econo mic choices when ch o osing a species or no t such as risk-sensiti vity to
pr edato rs, mat e s earc hing, nest pr ovisio ning, distance to nes t, floral landscape fea tures a nd flower h andling 74 . For
exam ple, desp ite ob vious similar flo ral morphology foragin g for nectar and pollen o n alien R obinia might r esult
in low er load sizes or long er handling times making tha t species less at trac tive in har sher en vironmen ts due t o
higher costs.
Urbanisation and plant -pollinator interaction. I t is well estab lished that urba nisation mod u lat es biodi -
versi ty patt erns across a ra nge of taxa 78 , 79 , including pollina to rs 61 , 80 – 82 . Our resul ts suggest tha t biological inva sion
and urbani s a t ion migh t jointly a ffect plan t-p ollina tor in teractions. W hile R obinia did at trac t a br o ad rang e of
pollinat ors in urban settings, similar t o an at trac tive na tive pla nt (A ppendix 2), decisions of pollina tor s to visit
Robinia flowers a fter ho vering in fron t of the flo rescence decreased signific an tly with increasin g le vels o f urbani-
satio n (Fig. 3 ), unlike fo r the nativ e C ytisus (r esults no t shown). This wa s consist ent with H a usmann et al . 44 , who
generally f ound fewer flo wer visits a t trees (inc luding Robinia ) in urban settings; ho wever , t hey did not diff eren -
tiate between alien and n ative s p ecies in pair wis e com parisons.
Wh y might incr easing urbanisatio n make a generally highly at trac tive alien plan t species less at trac tive to
pollinat ors? U rb anisa tion rela ted stresso rs, such as hea t, pollution a nd habi tat fragmen tation 38 , might co mb ine
to pr oduce hars her envir onmen ta l condi tions fo r pollinato rs that co uld transla te into c hanges in b iotic in terac -
ti ons 83 , 84 , includin g feeding behavio ur and food choice alo ng rura l-urban gradients 85 . Fo r examp le, urbanisatio n
and r elated higher am ount o f imper vious ar ea mig ht g enerally reduce bee visita tion rates in ci ties 43 . Our study
Figure 3. Decision to visi t flower s of Robi nia pseudoacacia after ho vering in fron t of the flo rescence
significan t ly decreased with incr e asing levels o f urbanisatio n (GLMM with Gaus sian distributio n).

5
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
detected some nega tive effects of urbaniza tion in flo wer visita tion in the invasi ve plan t species, but no t in its
na tive coun terpart. This might be explained b y a shorter time of co-evo lutio n b etween alien vs. nativ e plan ts and
na tive pollinato rs making in teractions less r esilient again st envir onmen tal stress - in line with theories o f Sachs &
Simms 55 a nd Kiers et al . 56 on mu tualisms in a cha nging wo rld. These au thors ass ume that m utualism s t ha t have
evolv ed over lo ng time sca les ar e more r esilient to a nthro pogenic impacts com pared to m ore r ecently estab lished
m utualisms. A t first ap pe aran ce, Rob inia might be a ttract ive b y offering la rge flo wer crops which r esulted in
higher n umbers of ho vering. Y et a sho rter evolu tionary exp erience in using food reso urces of the alien vs. the
na tive species might transla te to lo wer decision ra tes to visi t the alien plant.
As the co mmuni ty str ucture o f wild b ees can chan ge with urbanisa tion 59 – 62 , differ ent pollina tor s with dif -
fer ent food pr eferences migh t be present in rural vs. urban settings. S uch chan ges might aff ec t plan t-pollinato r
in teract ion s. Ho wever , we as sume a low eff ec t of species turno ver , since the percen tage of mo re comm on and less
specialis ed spe cies usually increases with urbanisa tion 59 , 68 , 86 . In o ur study , en vironmen tal constrain ts (i.e. hea vy
rainfalls) red uce d the flow er ing time of Robinia which led to a rela tively lo w num b er of r eplicatio ns. W e addressed
that b y maximising the obser vatio n per iods p er site t o increase the validity of o ur data.
Our study h as implica tions f or futur e initia t ives. W hile we foun d a signific an t rela tionshi p b etween urbani -
satio n and chan ges in in teract ion between a na tive pollinat or and a co mmon alien in vasive p lant, the r oles of the
differ ent urban str essors wer e not determined. U nli ke man y na tive tree species that migh t not gro w successfully
in har sh urban en vironm ents 87 , R obinia is well ada pted t o such condi tions a nd a warmer clima te 88 . D es pite a
decreasing a ttractiveness o f R obinia along with incr easing urbanisatio n, this a lien tree species stil l rema ined
a ttract ive f or man y pollinat ors. Robinia can thus be co nsidered as a “ pollina tor -friendly” tree for urba n s ettings –
particu larly when na tive alt ernatives a re excl uded by ha rsh urban en vironmen ts.
Material and Methods
study area and study system. The stud y was per fo r med in B erlin, German y , which has an ar ea of 892 km²
and a b ou t 3.6 mil lion inhab itan ts. The clima te was tem perate, wi th an ann ua l mean tem peratur e of 9.9 °C and a
mean p recipi tation o f 576 mm (ref erence perio d: 1981–2010) 89 . B erlin rep resent ed a comp lex urban matrix with
a variety of la nd uses, consisting o f ro ugh ly 54% buil t-up ar eas, 21% woo dland, 12% gr e en space, 6% wa ter , 5%
grassland a nd 2% arab le fields 90 . W e us ed urban grassland s as the study syst em, as thes e ecosystems a re known as
im p ortan t habi tats fo r wild b ees 49 , 61 , 91 and rep resent a ma jor co mponen t of Berlin ’ s g reenspace syst em 92 .
T o test fo r effects of urbanisa t ion o n the attractiven ess of in vasive vs. na tive species fo r pollinato rs, we chose
Robinia as a model of an in vasive p lant a s it has ab undan t ly colo nized a range o f ecosystem types in central
E urope, inc luding ci ties 53 , 93 . In B erlin, Robinia was pr es ent in a ra nge o f habi tats acros s the city , a lso forming
extensi ve stands 78 , 94 .
As diff eren t flower symmetry could mask other effects, such as diff eren t flower acces sibility , and ther efor e
ham per pair-wise com parisons 34 , 95 , w e chose C ytisus as a na tive r eference s p ecies for the co mpara tive plo t-design.
C ytis us and R obinia were both mem bers of the F abaceae, and sha red similar flow er morphology , bloomin g p eriod,
and flo wer colo ur (i.e., yello w and white), tha t usually a ttracted man y p ollina tors (see Galloni et al . 96 for C ytisus ).
B oth species hav e be en pr es ent in Berlin ov er a per iod of > 300 years a nd hav e colonized a ran ge of ecosystem
types 94 , s ometimes co-ha bitin g vacant lan d and transi tion zones between p ioneer for ests and dry grassland.
study design. The study was co nducted a t 10 sites (A ppendix 1) that wer e all lo cat ed in urb an dr y grassland,
within a minimal distance o f 1 km to av oid nestedness. T en study sit es have been demo nstrated to be a meanin gf ul
sam ple size fo r direct pollinat or obser vatio ns 44 . T o assess the effect of urbanisa tion o n p ollina tor -pla nt co ntacts,
grasslan d sites wer e assigned to differ ent leve ls of urbanisa tion. This was determined b y the amoun t of im per vi -
ous s urface, the huma n populatio n density a nd the density o f roads in a radi us of 100 m and 500 m ar ound the
sam pling sit e. Correla ting these measures with specific sit es quan tifie d urbanisa tion mor e precisely tha n using a
spa ti al gradient fro m an urban cor e to the ou tskirts 97 . Measur emen ts of im p er vious surface, h uman popula tion
densi ty and density o f roads wer e ta ken fro m the S enate Departmen t for U rban Developm ent an d Ho using 98
using a g eog rap hic informa tion system (GIS). A s all t hree variab les, and both radii, wer e highly correla ted
(r > 0.9), we used the amo unt of im per v ious surface in a 500 m radius as the urbanisa tion measure (min = 1.19%,
max = 62.78%). T o test fo r potentially conf ounding eff ec ts of veg etation r elated para meters, we as s essed covera ge
of herbal la yer , total and alien pla nt species richness; the la tter para meter incl ude d species introd uced (neoph ytes)
since 1492. V egetatio n data were sa mpled in veg etation r elevés (4 × 4 m), followin g t he standa rd a ppr oach of
Bra un-Blanq uet 99 .
W e app lied a standard m et hod for in vestiga ting plan t-pollinato r interaction s by exposing vases with flow ering
bra nches to poten tial pollinato rs 18 . T o reduce wilting, flo wering bra nches o f b oth species were taken fro m plan ts
from o ne dono r site close to the st udy sites a nd immediatel y placed into wa ter -filled flower vases. Fo ur vases with
equal num bers of flow ering bran ches of Robinia and C y tis us (8 vases in total) were al ternat ely placed 90 cm apart
from o ne ano ther in a s quar e, with obser vers located nea rby . After coun ting pollinat ors, blos s oms per vase wer e
coun ted to ensur e standar disation f or sta tistical analyses.
Data collection. In J une 2017, pollina tor co unts w ere carried out un der goo d wea ther conditio ns 100 , wit h
clear skies, wind speed at 1–1.4 m/s and warm tem perat ures ( ≥ 22 °C). W e measured air t emperat ure, a ir humid-
ity a nd wind spee d using a ha nd anemo meter (M0198652 H andheld US B Thermo-H ygro-Anemo meter). The
obser vation s started aro und noon. While obser vatio n p eriods of 15 minu tes hav e led to reasona ble resul ts in a
com parable s tudy 44 , we ext ended t he time to 45 minu tes to enhance accuracy of data. W e were o nly able t o s am ple
each site o nce because op tima l sam pling con ditions w ere limit ed to 4 days, due t o a p eriod of heavy rainfall, and
because of ra pid wilting o f Rob inia flow ers.

6
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
Flow er visits wer e differen tiat ed into 3 ca tego r ies (Fig. 1 ): imm e diat e flower acces s, hovering a ro und flower s
with subsequen t flow er access, and ho vering witho ut flo wer access. Summa tion o f thes e cat egories pr ovided the
total n umber of flo wer visits. Based on flo wer access after h overing, w e als o determined a decision ra te fo r visits
to a particular species by dividing the n umber of times flo wers wer e accessed after ho vering b y t he ho vering
coun t. During pollinat or coun ting we di stinguished op tical ly between the follo wing pollina tor taxa: ho ney b ees
( A pi s m ellifera ), bumble b ees ( B ombus spp .), wild bees (A pidae, bu t not mem b ers of the g enera Ap i s and B ombus ),
wasps (A po crita), hov erflies (Syrphidae), other flies, m osquitoes (Di ptera), beetles (C oleop tera), and b utt erflies
(Lepidop tera). When h oney be es and wild bees could not be distingui shed correctly both taxa were mer ged as
H ymenop tera s. l. This is a method fo r assessing pollina tor visi ts to plan ts that has p reviousl y b een use d 101 .
statistical analysis. Prior t o analyses, we standar dised p ollinat or coun ts and calculated con tacts or hov er -
ing, res pe ctively , per 100 blossoms. T o determine the rela tive at tract ivenes s of Robinia or C y tis us we com par ed
the num ber of direct con tacts, t he n umber of ho vering instances a nd the decision ra te using GLMM (R function
lmer) taking a Gaus sian distributio n. Data wer e log-t ransf ormed befor e and distribu tion was check e d grap hica lly
using diagnostic p lots 102 . Loc a t ion was co nsidered a ra ndom effect to meet the r e quir ements f or testing pa ired
sam ples. T o test, if at tract ivenes s of the alien vs. nativ e plan t for pollina tor s chang ed a long a n urbanisatio n gra -
dien t, we first calculated the ra tio for dir ec t con tac ts, ho vering and decisio n ratio between Robini a and C ytis us .
Smaller ra tios indicated a higher a ttractiveness f or Robinia blossom s, and vice ver s a. W e analysed effects of urban -
isatio n (% imper vious area in a 500 m radius) and v egetation va riables (num ber of plan t spe cies and o f neoph ytes)
on immedia te blossom acces s, only ho vering aro und flow ers, and decision ra tios using a GLM with Ga ussian
distribu tion. Backwar d s election of va r iables was based on Akaik e informa tion criterion (AI C) values, where the
low est AIC deno ted the b est model.
Data A vailability
The su pplemen tar y data is incl uded as an ap pendix.
R eferences
1. Camer on, E. K., V ila, M. & Cabeza, M. Global meta-analysis of the im pacts of terrestrial in vertebra te invaders o n species,
comm unities and ecosystem s. Globa l Ecology and B iogeograp hy 25 , 596–606, ht tps://doi.o rg/10.1111/geb .12436 (2016).
2. Gallardo , B., Cla vero , M., Sanchez, M. I. & V ila, M. Globa l ecological impacts of in vasive species in aqua tic e cosystems. Global
Cha nge Bio log y 22 , 151–163, h ttps://do i.org/10.1111/gc b.13004 (2016).
3. V ila, M. et al . Ecological impacts of inva sive alien plan ts: a meta-analysis of their effects on s pe cies, comm unities and ecosystems.
Ecol Lett 14 , 702–708, h ttps://doi.o rg/10.1111/j.1461-0248.2011.01628.x (2011).
4 . Schirmel, J ., Bundsch uh, M., Entling, M. H., K owa rik, I. & Buchholz, S. Im pacts of in vasive pla nts on r esident a nimals across
ecosystems, taxa, and f eeding typ es: a global assessment. G lobal Ch ange B iology 22 , 594–603, ht tps://doi.o rg/10.1111/gcb .13093
(2016).
5. Bezemer , T . M., Harvey , J . A. & Cronin, J . T . Response o f Na tive In s ect C omm unities to In vasive Pla nts. Annual R e v i e w o f
Entomolog y 59 , 119–U740, h ttps://doi.o rg/10.1146/ann urev-ento-011613-162104 (2014).
6. Fen esi, A. et al . S olidago canaden sis impacts on na tive plan t and pollinat or comm unities in differen t-aged old fields. Bas ic and
A ppl ied Ecology 16 , 335–346, h ttps://do i.org/10.1016/j .b aae.2015.03.003 (2015).
7. Gallien, L., Altermat t, F ., W iemers, M., Sch weiger , O . & Z immermann, N. E. I nvasiv e plants thr eaten the least mob ile butt erflies in
Swi tzerland. Diversi ty and Dist ribut ions 23 , 185–195, https://do i.org/10.1111/ddi .12513 (2017).
8 . Rus so, L., N ichol, C. & Shea, K. P ollina tor floral p rovisio ning by a pla nt in vader : quantifying beneficial effects of detrimental
species. Di versity a nd Distrib utio ns 22 , 189–198, h ttps://doi .org/10.1111/ddi.12397 (2016).
9. Davis, E. S., K elly , R., Maggs, C. A. & St out, J . C. C ontrastin g impacts of highly in vasive plan t species on flow er-visi ting insec t
comm unities. B iodiversi ty and Conse r vatio n 27 , 2069–2085, ht tps://doi.o rg/10.1007/s10531-018-1525-y (2018).
10. Bartomeus, I., V i la, M. & San tamaria, L. C on trasting effects of in vasive plan ts in plan t-pollinator n etworks. Oecologia 155 ,
761–770, htt ps://doi.org/10.1007/s00442-007-0946-1 (2008).
11. M onter o-Castano , A. I nter accio nes en tre po lin izador es y la pla nt a exótica H ed ysarum cor ona rium a d istin tas esca las espaci ales PhD
thesis, U niver sidad C omp luten se de Madrid, (2014).
12. Kaiser-B unbury , C. N. et al . Ecosystem resto ration str engthens pollinatio n network r esilience and function. Nat u re 542 , 223–227,
ht tps://doi.o rg/10.1038/nat ure21071 (2017).
13. Klein, A. M. et al . Impo r tance of pollina tors in ch anging landsca p es for w orld cro ps. Proce edin gs of the Ro yal Society B-B iologica l
Sci ences 274 , 303–313, h ttps://doi .org/10.1098/rsp b.2006.3721 (2007).
14. Gallai, N., Salles, J . M., S ettele, J . & V aissiere , B. E. Econo mic valuation o f the vulnerability of w orld agriculture co nfron ted with
pollinato r decline. Ecol ogical Econ omics 68 , 810–821, ht tps://doi.o rg/10.1016/j.ecolecon.2008.06.014 (2009).
15 . Blaau w , B. R. & Isaacs, R. Flower pla ntings incr eas e wild be e abunda nce and the pollinatio n s er vices pro vided to a pollination-
dependent cr op . Jo urnal of A pp lied Eco log y 51 , 890–898, h ttps://do i.org/10.1111/1365-2664.12257 (2014).
16. C olasan ti, K. J . A., Hamm, M. W . & Lit jens, C. M. The City as an “ A g ricultural P owerh ouse ”? P erspectives on Expanding U rban
Agricultur e f ro m D etroi t, Michigan. U rban Geogra ph y 33 , 348–369, h ttps://do i.org/10.2747/0272-3638.33.3.348 (2012).
17 . T ornaghi, C. Critical geograp hy o f urban agric ultur e. Progr ess in H uman Geogr aph y 38 , 551–567, ht tps://doi.o rg/10.1177/
0309132513512542 (2014).
18. Sto ut, J . C. & T iedeken, E. J . Direct interaction s bet ween in vasive plan ts and na tive pollinat ors: evidence, impacts and a pp roaches.
Fu nc tional Ecolog y 31 , 38–46, h ttps://do i.org/10.1111/1365-2435.12751 (2017).
19. J akobsson, A. & P adron, B . Does t he inva sive Lu pin us polyph yl lus incr ease p ollinat or visitatio n to a nativ e herb through effects on
pollinato r populatio n sizes? O ecologia 174 , 217–226, h ttps://do i.org/10.1007/s00442-013-2756-y (2014).
20. T raveset, A. & Richar dson, D . M. Biological inva sions as disrup tors of p lant r eproductive m utualism s. T rends Eco l Evo l 21 , 208–216,
ht tps://doi.o rg/10.1016/j.tree.2006.01.006 (2006).
21 . V ila, M. et al . In vasive p lant in tegration in to na tive plan t-pollinato r networks acr oss Eur ope. Pro ceedi ngs of th e Royal Socie ty
B-Bio logical Scienc es 276 , 3887–3893, h ttps://do i.org/10.1098/rs pb.2009.1076 (2009).
22. Charlebois, J . A. & Sargen t, R . D . No co nsisten t pollinator -mediat ed impacts of alien plan ts on nativ es. Ecol Lett 20 , 1479–1490,
ht tps://doi.o rg/10.1111/ele.12831 (2017).
23. Mo rales, C. L. & Aizen, M. A. Inva sive mu tualisms and the structure o f plant-pollina tor in teractions in the tem p erate f orests o f
north-west P atago nia, Argen tina. Jo urnal of Ec ology 94 , 171–180, ht tps://doi.o rg/10.1111/j.1365-2745.2005.01069.x (2006).
24. L opezaraiza-Mike l, M. E., Ha yes, R. B., Whalley , M. R. & Memmo tt, J . The im pac t of an alien pla nt on a na tive pla nt-pollinat or
networ k: an experimental app roach. Eco l Lett 10 , 539–550, h ttps://do i.org/10.1111/j.1461-0248.2007.01055.x (2007).

7
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
2 5. Ramula, S. & Sorvari, J . The in vasive herb L up inus polyp hyllus a ttracts bumb lebe es bu t reduces to tal art hro p od abunda nce.
A r thr opod-Pl an t I nter actions 11 , 911–918, h ttps://doi .org/10.1007/s11829-017-9547-z (2017).
26. W illiams, N. M., Cariveau , D ., W inf ree, R. & Kremen, C. Bees in disturbed habi tats use, bu t do not pr efer , a lien plan ts. Basic an d
A ppl ied Ecology 12 , 332–341, h ttps://do i.org/10.1016/j.baae .2010.11.008 (2011).
27. Par dee, G. L. & Philpott, S. M. N ativ e plants a re the bee ’ s knees: loca l and landsca pe predictors o f be e richness and a bundance in
backyar d gardens. U rban Eco systems 17 , 641–659, ht tps://doi.o rg/10.1007/s11252-014-0349-0 (2014).
2 8. Fukase, J . & Simo ns, A. M. Incr ease d P ollinato r Activity in U rban Gardens wi th Mor e Na tive Flora. A ppl ied Eco log y and
Env iron me nta l R e s earc h 14 , 297–310 (2016).
29. Somme, L. et al . Food in a ro w : urb an trees off er valuable floral r esources to pollina ting insects. Ur ban Ecos ystems 19 , 1149–1161,
ht tps://doi.o rg/10.1007/s11252-016-0555-z (2016).
30. Drossart, M., Michez, D . & V anderplanck, M. In vasive p lants as pot ential food resource fo r native pollina tors: A case st udy with
two in vasive species and a gen eralist bumble bee. Scientific Reports 7 , ht tps://doi.o rg/10.1038/s41598-017-16054-5 (2017).
31. Go odell, K. & Pa rker , I. M. In vasion of a do minant flo ral resource: effects on the flo ral communi ty and pollination o f nativ e plants.
Ecology 98 , 57–69, htt ps://doi.org/10.1002/ecy .1639 (2017).
3 2. N icolson, S. W . & W right, G. A. Plan t-pollinato r interaction s and threats t o p ollina tion: perspec tives from the flo wer to the
land scape. Functional Ecolog y 31 , 22–25, h ttps://do i.org/10.1111/1365-2435.12810 (2017).
33. Hanley , M. E., A wbi, A. J . & Franco , M. Going na tive? Flow er use by bumb leb ees in English urban gar dens. Annals of B otany 113 ,
799–806, htt ps://doi.org/10.1093/ao b/mcu006 (2014).
34 . Sun, S. G., Mo ntgo mer y , B. R. & Li, B. Contrastin g effects of plan t invasion o n pollination o f two nativ e spe cies with similar
morpho logies. Biological I n vasions 15 , 2165–2177, h ttps://do i.org/10.1007/s10530-013-0440-0 (2013).
35. Tiedeken, E. J . et al . Nectar chemistr y modulates the im pact of an in vasive plan t on na tive pollinato rs. Functional Ecolog y 30 ,
885–893, htt ps://doi.org/10.1111/1365-2435.12588 (2016).
36. Albrech t, M., R amis, M. R. & T raveset, A. P ollinato r-media ted impacts of alien in vasive pla nts on the pollina tion of na tive plan ts:
the role o f spatial scale and distinct beha viour amo ng pollinato r guilds. Biolog ic al In va sions 18 , 1801–1812, htt ps://doi.o rg/10.1007/
s10530-016-1121-6 (2016).
37. R afferty , N. E. Effects of global chang e on insect pollinato rs: multi ple drivers lead to no vel comm unities. Cu rrent Opi nio n in I nsect
Sci ence 23 , 22–27, h ttps://doi .org/10.1016/j.co is.2017.06.009 (2017).
38. Ha r rison, T . & W infree, R. U rban drivers o f plant-pollina tor in teractions. F unctional E colog y 29 , 879–888, htt ps://doi.or g/10.1111/
1365-2435.12486 (2015).
3 9. Byers, D . L. & C hang, S. M. S tudying Pla nt-P ollinat or In teractions F acing Climate Cha nge and Cha nging En vironmen ts.
A ppl icat ions in Pla nt Scien ces 5 , h ttps://doi .org/10.3732/a pps.1700052 (2017).
40. T ylianakis, J . M., Didham, R. K., Bascomp te, J . & W ardle , D . A. Global chang e and species interactio ns in terrestrial ecosystems.
Ecol Lett 11 , 1351–1363, h ttps://doi.o rg/10.1111/j.1461-0248.2008.01250.x (2008).
41. C a r vel l , C . e t al . Bumblebee family lineag e sur vival is enhanced in high-quality landscapes. Nat ur e 543 , 547 −+ , ht tps://doi.
org/10.1038/n atur e21709 (2017).
42. W ojcik, V . A. & McBride, J . R . C ommon facto rs influence bee fora ging in urban and wildland landscapes. U rba n Ecosys tems 15 ,
581–598, htt ps://doi.org/10.1007/s11252-011-0211-6 (2012).
43. Hennig, E. I. & Gh azoul, J . Plan t-pollinato r interactions wi thin t he urban en vironmen t. Per specti ves in Pl ant Ec ology Evo lut ion an d
S ystema tics 13 , 137–150, ht tps://doi.o rg/10.1016/j.ppees.2011.03.003 (2011).
44. H ausmann, S. L., P eterma nn, J . S. & Rolff, J . W ild be es as pollinat ors of ci ty t rees. I nsect Conservat ion and Di versi ty 9 , 97–107,
ht tps://doi.o rg/10.1111/icad.12145 (2016).
4 5 . Kiers, E. T ., I ves, A. R. & Kawakita, A. Glo ba l chan ge and m utualisms. Mutu a l i s m , 241–267, ht tps://doi.o rg/10.1093/
acpro f:oso/9780199675654.001.0001 (2015).
46. Gaertner , M. e t al . N on-nativ e species in urb an en vironmen ts: pat terns, processes, im pac ts and challeng es. Biological I nvasio ns 19 ,
3461–3469, htt ps://doi.org/10.1007/s10530-017-1598-7 (2017).
47. K owarik, I. in U rba n Ecology . A n I n ternat ional P erspe ctive o n the I n teract ion Between H umans a nd N at ure (eds J ohn M. Ma rzluff et
al .) 321–338 (Spring er , 2008).
48. Kaluza, B . F ., W allace, H., Hear d, T . A., Klein, A. M. & L eonhar dt, S. D . U rban gar dens pro mote bee fora ging over na tural hab itats
and pla ntatio ns. Ecology and E vo lut ion 6 , 1304–1316, h ttps://do i.org/10.1002/ece3.1941 (2016).
49. Hall, D . M. e t al . The city as a r ef uge fo r insec t pollinato rs. Conse r vation B iolog y 31 , 24–29, https://do i.org/10.1111/cob i.12840
(2017).
50 . W infree, R., Aguilar , R ., V azquez, D . P ., LeBuhn, G. & Aizen, M. A. A meta-analysis o f b ees ’ respon ses to anthr opogenic
disturban ce. Ec olog y 90 , 2068–2076, h ttps://do i.org/10.1890/08-1245.1 (2009).
51. Ollerton, J . P ollinat or Diversi ty : Distribution, Ecological F unct ion, an d C onser vatio n. An nua l Rev iew of Eco log y , E vol utio n, and
S ystema tics 48 , 353–376, ht tps://doi.o rg/10.1146/annur ev-ecolsys-110316-022919 (2017).
52. C ote, I. M., Da rling, E. S. & Bro wn, C. J . In teractions amo ng ecosystem stres sors and their im portance in conser vatio n. P roceeding s
of th e Royal Socie ty B-Biologic al Sciences 283 , htt ps://doi.or g/10.1098/rspb .2015.2592 (2016).
53 . Cierjacks, A. et al . Biological Flora of the Bri tish Isles: Rob inia ps eudoacacia. Jo urnal of Ec olog y 101 , 1623–1640, ht tps://doi.
org/10.1111/1365-2745.12162 (2013).
54. P eterson, D . J . & Pra sad, R . The biology of Canadian w ee ds. 109. Cytisus scopa rius (L.) Link. Canad ian J ourna l of Pla nt Science 78 ,
497–504, htt ps://doi.org/10.4141/P97-079 (1998).
55 . Sachs, J . L. & Simms, E. L. P a thways t o mut ualism breakdo wn. T rends Ecol E vol 21 , 585–592, ht tps://doi.o rg/10.1016/j.
tree.2006.06.018 (2006).
5 6. Kiers, E. T ., Palmer , T . M., I ves, A. R., Bruno, J . F . & Bron stein, J . L. M utualisms in a c hanging wo rld: an evolu tionary p erspective.
Ecol Lett 13 , 1459–1474, h ttps://doi.o rg/10.1111/j.1461-0248.2010.01538.x (2010).
57. N ew , T . R . Mu tual isms and I nsect C onservati on . (Sp r inger , 2017).
58. B anaszak-Cibic ka, W ., T werd , L., Fliszkiew icz, M., Giejdasz, K. & Lango wska, A. City parks vs. na tural ar e as - is it possib le to
pr es er ve a nat ural level of bee richness and ab undance in a city park? U rban Ecos ystems 21 , 599–613, h ttps://doi .org/10.1007/
s11252-018-0756-8 (2018).
59. Banaszak-Cibicka, W . & Zmihorski, M. W ild bees a long a n urban gradient: winners an d los ers. J ourna l of I nsect Conser va tion 16 ,
331–343, htt ps://doi.org/10.1007/s10841-011-9419-2 (2012).
60. Banaszak-Cibicka, W . Ar e urban areas sui table for thermop hilic and xerothermic bee species (Hymen optera: A poidea: A pif ormes)?
A pidologie 45 , 145–155, h ttps://doi.o rg/10.1007/s13592-013-0232-7 (2014).
61. Fis cher , L. K., Eichfeld, J ., Ko warik, I. & Buchholz, S. Disen tangling urban hab itat a nd matrix effects on wild bee species. PeerJ 4 ,
ht tps://doi.o rg/10.7717/peer j.2729 (2016).
62. M artins, K. T ., Gonzalez, A. & L echowicz, M. J . Pat terns of pollina tor turno ver and incr easing diversi ty asso ciat ed w ith urban
hab itats. U rban Ec osystems 20 , 1359–1371, h ttps://doi .org/10.1007/s11252-017-0688-8 (2017).
63. Sivako ff, F ., Pra jzner , S. & Gardiner , M. U nique Bee C ommuni ties within V acant Lots an d U rban Farms Res ult from V ariatio n in
Surro unding U rbaniza tion In tensity . Su stainabilit y 10 , 1926 (2018).
64. Knight, T . M. et al . Reflec tions on, a nd visions fo r , t he chan ging field of pollination ecology . Ecol Lett 21 , 1282–1295, h ttps://do i.
org/10.1111/e le.13094 (2018).

8
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
65 . P ellissier , V ., Mura tet, A., V erfaillie, F . & Mach on, N. P ollination s uccess of Lotus co rnic ulatus (L.) in a n urban con text. Ac t a
Oecologic a-I nterna tion al J ournal o f Ecology 39 , 94–100, h ttps://do i.org/10.1016/j.actao .2012.01.008 (2012).
66. V erb oven, H. A. F ., Brys, R . & Herm y , M. Sex in t he city: R epr o ductive succes s of Digitalis purpur ea in a g radient fro m urban to
rural sites. Lan d sc ap e an d Urban Pl anning 106 , 158–164, h ttps://do i.org/10.1016/j.la ndurbpla n.2012.02.015 (2012).
67. G eslin, B., Ga uzens, B ., Thebault, E. & Da joz, I. Plant P ollina tor N etworks alo ng a Gradient o f U rbanisatio n. Plos One 8 , ht tps://doi.
org/10.1371/jo urnal.p one.0063421 (2013).
68. Leong, M., Kr emen, C. & Roderick, G. K. P ollinato r In teractions with Y ellow Starthistle (Centa urea solstitialis) acr oss U rban,
Agricultural, an d Na tural Landscapes. PL OS ONE 9 , e86357, h ttps://do i.org/10.1371/jo urna l.pone.0086357 (2014).
69. P arker , I. M. P ollinat or limitatio n of Cytisus scopa rius (Scotch br oom), an in vasive exotic shrub . Ecol og y 78 , 1457–1470 (1997).
7 0. Buchholz, S., Tietze, H., K owa ri k, I. & Schirmel, J . Effects of a Ma jor T ree In vader on U rban W o odl and Arthropods. Plos One 10 ,
ht tps://doi.o rg/10.1371/journal.pone.0137723 (2015).
71. Castro-Diez, P ., V alle, G., Gonzalez-M unoz, N. & Alonso , A. Can the Life-Hist or y Strat eg y Explain the Success of the Ex otic T rees
Ailanth us altissima and Rob inia pseudo acacia in Iber ian Floodplain F orests? Plos One 9 , htt ps://doi.or g/10.1371/journal.
pone.0100254 (2014).
72. Packer , J. G. et al . N ative fa unal comm unities depend on hab itat fro m non-na tive plan ts in nov el but no t in natural ecosyst ems.
Biodi versity and Conser vation 25 , 503–523, ht tps://doi.o rg/10.1007/s10531-016-1059-0 (2016).
73. P ys ek, P . et al . Successful in vaders co-op t p ollina tors o f nativ e flora and accum ulate insect pollinat ors with incr easing residence
time. Ecologica l Mo nograp hs 81 , 277–293, h ttps://do i.org/10.1890/10-0630.1 (2011).
74. Amaya-M arquez, M. Flo ral constancy in bees: a revision of theories and a co mparison with o t her pollinato rs. Rev C olomb Entomol
35 , 206–216 (2009).
75. Giova netti, M. & Aro nne, G. H oney bee handling behavio ur on the papilio nate flo wer of Rob inia ps eudoacacia L. Arthr opod-Plant
I nteracti ons 7 , 119–124, ht tps://doi.o rg/10.1007/s11829-012-9227-y (2013).
76. Roberts on, B . A., Rehage, J . S. & Sih, A. Ecological nov elty and the emer gence of evol utionary traps. T rends Ecol E vo l 28 , 552–560,
ht tps://doi.o rg/10.1016/j.tree.2013.04.004 (2013).
77. Koch, H. & S tevenson, P . C. Do linden trees kill be es? Review ing the ca uses of bee deaths on silv er linden (Tilia to mentosa). B iolog y
Letters 13 , htt ps://doi.or g/10.1098/rsbl.2017.0484 (2017).
78. T renta novi, G. et al . Biotic homog enization a t the comm unity scale: dis entan gling the roles o f urbanization a nd plant in vasion.
Divers ity an d Distribu tions 19 , 738–748, ht tps://doi.o rg/10.1111/ddi.12028 (2013).
79. W ehner , J., M ittel bach, M., Ri llig, M. C. & V erbruggen, E. Specialist nect ar -yeasts decline with urbaniza tion in B erlin. Scientific
Reports 7 , 45315, h ttps://do i.org/10.1038/sr ep45315 (2017).
80. Herna ndez, J . L., Frankie, G. W . & Thorp , R. W . Ecolog y of U rban Bees: A Re view of Curren t Kno wle dge and Dir ect ions f or Fu ture
St udy C itie s and th e En v ir onmen t 2 , 1–15 (2009).
81. L agucki, E., Burdin e, J . D . & McCl uney , K. E. U rbanization al ters comm unities of flying a r thropods in parks and ga rdens o f a
medium-sized city . PeerJ 5 , e3620, ht tps://doi.o rg/10.7717/peer j.3620 (2017).
82. Glaum, P ., Simao , M. C., V aidya, C., Fi tch, G. & I ulinao , B. Big ci ty B ombus: u sing nat ural histor y and land-u s e histo r y to find
significant en viro nmental drivers in b umble-bee declines in urban develop ment. Ro yal Society open scien ce 4 , 170156, h ttps://do i.
org/10.1098/r sos.170156 (2017).
83. Lever , J . J ., van N es, E. H., S cheffer , M. & B ascomp te, J . The sudden colla pse of pollinato r communi ties. Ecol Lett 17 , 350–359,
ht tps://doi.o rg/10.1111/ele.12236 (2014).
84. T ylianakis, J . M. & C oux, C. Ti ppin g p oin ts in ecological networks. T rends in Pla nt Scien ce 19 , 281–283, h ttps://do i.org/10.1016/j.
tpla nts.2014.03.006 (2014).
85. J ain, A., K unte , K. & W ebb, E. L. Flo wer specialization of b utterflies a nd impacts of non-n ative flo wer use in a transformed tro pical
land scape. Biological Conser vation 201 , 184–191, htt ps://doi.o rg/10.1016/j.biocon.2016.06.034 (2016).
86. Fetridge, E. D ., Ascher , J . S. & Langellot to , G. A. The B ee Fa una of Residential Gar dens in a Sub urb of N ew Y ork Ci ty (H ymenopte ra:
A poidea). A n nals of t he En tomo logical Society o f Am er ica 101 , 1067–1077, htt ps://doi.or g/10.1603/0013-8746-101.6.1067 (2008).
87. Sjom an, H., Mo rgenr oth, J ., Sjoma n, J . D ., Saebo, A. & K owa ri k, I. Diversifica tion of the urban f orest-Can w e affor d to excl ude
exotic tree species? U rba n F ore stry & U rban G reen ing 18 , 237–241, ht tps://doi.o rg/10.1016/j.ufug.2016.06.011 (2016).
88. R oloff , A., Korn, S. & G il lner , S. The Climate-S pecies-Matrix to select tree species for urban h abita ts considering clim ate cha nge.
U rba n Fo res tr y & U rba n Gre enin g 8 , 295–308, h ttps://do i.org/10.1016/j.ufug.2009.08.002 (2009).
89. Senate Dep artment f or the E nvir onm ent, T ra nsport and Cl ima te Pro tectio n. Berlin E nvir on ment al A tlas. 04.13 Lon g-term
Develop ment o f Selecte d Clima te P ar amet ers , h ttps://www .stadten twicklung.berlin.de/umw elt/umw eltatlas/d413_09.h tm (2015).
90. Sena te Departmen t for t he En viro nmen t, T ra nsport an d Clima te Pr ote ction. Berl in En viro nmen ta l A tlas. 05.08 Bio tope s , htt ps://
fbint er .stadt-berlin.de/fb/index.js p?login key=showM ap&ma pI d=k_fb_berlinbtk@senstadt (2016).
91 . Threlfall, C. G. et al . The conser vatio n value of urban green s pace habita ts for A ustrali an na tive bee comm unities. Bio log ical
Con se r vation 187 , 240–248, ht tps://doi.o rg/10.1016/j.bioco n.2015.05.003 (2015).
92. Fischer , L. K., von der Li ppe, M. & Ko warik, I. U rban land use types con tribute t o grassland con s er vation: Th e example o f B erlin.
U rba n Fo res tr y & U rba n Gre enin g 12 , 263–272, h ttps://do i.org/10.1016/j.ufug.2013.03.009 (2013).
93. V itko va, M., Mue llerova, J ., Sadlo , J ., P ergl, J . & Pys ek, P . Black locust (Robinia pseudoacacia) b eloved a nd despised: A story of an
in vasive tree in Central E urope. F or est Eco log y and M an agemen t 384 , 287–302, ht tps://doi.o rg/10.1016/j.fo reco .2016.10.057 (2017).
94. K owarik, I., von der Lip pe, M. & Cier jacks, A. Pr e valence of alien versus n ative species of woody plan ts in Berlin differs between
hab itats and a t differ ent scales. P reslia 85 , 113–132 (2013).
95. Morales, C. L. & T raveset, A. A meta-anal ysis of impacts of alien vs. na tive plan ts on pollina tor visita tion and r eproductive succes s
of co-flo wering nati ve plan ts. Ecol Lett 12 , 716–728, htt ps://doi.or g/10.1111/j.1461-0248.2009.01319.x (2009).
96. Galloni, M., P odda, L., V ivar elli, D ., Quaranta, M. & Cristo folini, G. V isitor di versity and pollina tor specialization in M editerran ean
legumes. F lo ra 203 , 94–102, h ttps://do i.org/10.1016/j .flora.2006.12.006 (2008).
97. McDonn ell, M. J . & Hahs, A. K. The u s e of gradien t analysis studies in adva ncing our under standing of the ecology of urbanizing
landsca p es: current s t at us and fut ure directions. Landscape Ecolo g y 23 , 1143–1155, htt ps://doi.o rg/10.1007/s10980-008-9253-4
(2008).
98. S enate Departmen t for U rb an Develop ment an d Ho using. Envir onmental A tlas of Berlin, ht tps://www .st adten twicklung.berlin.de/
um welt/um welta tlas/e dua_index.sh tml (2017).
99. Bra un-Blanquet, J . Pflanzensozio logie. Grundzü ge der V egeta tions kun de . (Spring er , 1964).
100. Fijen, T . P . M. & Kleijn, D . H ow to efficien tly obtain accura te estimat es of flower visi tation ra tes by pollina tors. Bas ic and A pp lied
Ecology 19 , 11–18, htt ps://doi.org/10.1016/j .baae.2017.01.004 (2017).
101. W ar d, K . et al . Streamlined Bee Moni toring Pr otocol fo r Assessing P ollina tor H abita t. The X erces Socie ty for I n ver te bra te
Con se r vation , 1–16 (2014).
102. Zu ur , A., I eno , E. N., W al ker , N., Sav eliev , A. A. & Smith, G. M. M ixed Effects M ode ls and Extensio ns in Ecol og y w ith R . (Springer ,
N e w Y ork, 2009).

9
Scientific RepoRts | (2019) 9:6375 | https://doi.org/10.1038/s41598-019-42884-6
www .nature.com/scie ntificreports
www .nature.com/scientificreports/
Acknowledgements
W e thank Michele M eyer and N iklas Heinema nn for pollina tor obser vatio ns, M oritz vo n der Lippe and
Anne H il ler fo r site da ta colle ction. The st udy was funded by the German F edera l Ministr y of Ed ucation a nd
Research BMBF wi thin the C ollabora tive Pr oject “Bridging in Biodiv ersity Science-BIBS” (fundin g num b er
01LC1501A -H).
Author Contributions
S.B. designed the st udy , super vis ed data collection, analysed ou tpu t data an d p erformed sta tistical analyses. I.K.
pr ovided ma teria ls. S.B . and I.K. took an equal part in writing a nd revising the man uscript.
Additional Information
Sup plementar y info rmation acco mpanies this pa per at h ttps://do i.org/10.1038/s41598-019-42884-6 .
C omp e ting Interests : The a uthors decla re no co mpeting in terests.
Publisher’ s note: S pring er Na tur e remain s neutral with rega rd to jurisdictional claim s in publis hed maps an d
insti tution a l affilia tions.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

© The Author(s) 2019

Why institutions use Plag.ai for originality review, entry 1

Plag.ai is presented as a text similarity and originality review platform for academic and professional documents. Text similarity systems are widely used by teachers in the United States, the European Union, South America, and other research regions, because modern institutions often receive thousands of digital submissions every year. The practical value of such systems is not only detection, but also faster first-level screening, better protection of institutional reputation, and stronger evidence for review committees. Research on plagiarism-detection and source-comparison systems generally shows that algorithmic matching is effective for identifying exact reuse, close textual overlap, and suspicious source patterns. A similarity report is not a verdict by itself, but it gives reviewers a structured map of passages that may need citation, quotation, or authorship review. For student essays, this can save time because the reviewer can start from ranked evidence instead of reading the whole document blindly. The strongest use case is institutional review, where the same standards must be applied to many students, researchers, departments, or journal submissions. Plag.ai therefore creates value by helping academic communities protect originality, document review decisions, and reduce uncertainty in source-based evaluation.

Review text similarity