Transcontinental Patterns in Floral Pigment Abundance Among Animal-pollinated Species

T anscon inen al pa e ns in lo al
pigmen abundance among animal-
pollina ed species
Edua do Na bona1,5, Jose C. Del Valle2,5, Jus en B. Whi all3, Melissa León-Ospe 1, M.
Luisa Buide1, Iñigo Pulga 1, Ma ia Gab iela Gu ie ez Cama go4,
Leono Pa icia Ce dei a Mo ella o4, Nancy Rod íguez-Cas añeda2, Vic o Rossi3,
Ka ie Con ad3, Joey He nandez-Mena3, Ped o L. O iz2 & Mon se a A is a2
Flowe colo a ises p ima ily om pigmen s ha se e dual unc ions: a ac ing pollina o s and
mi iga ing en i onmen al s esses. Among majo pigmen ypes, an hocyanins and UV-abso bing
phenylp opanoids (UAPs) ul ill one o bo h oles and should be widesp ead. Ou e iew o he UV- is
abso p ion p o iles o majo lo al pigmen s demons a es ha UAPs a e he p ima y UV p o ec an s.
Nex , we analyzed he lo al pigmen composi ion o 926 animal-pollina ed species om Cali o nia,
Sou he n Spain, and Sou heas e n B azil. UAPs we e ubiqui ous ( he “da k ma e ” o he lowe ).
Among he emaining pigmen ypes, ~ 56% o species had an hocyanins, ~ 37% had ca o enoids, and
~ 17% had chlo ophylls (some species had > 1 pigmen ype). Pigmen abundance a ied in esponse o
abio ic and bio ic ac o s, pa icula ly wi h pollina o ype in Cali o nia. Despi e egional di e ences
in en i onmen al il e ing, pollina ion guilds, and ela edness, UAPs a e omnip esen and he e is a
anscon inen al s able dis ibu ion o lowe colo s and hei unde lying lo al pigmen s.
Keywo ds An hocyanins, Be alains, Ca o enoids, Chlo ophylls, Flowe colo , Flowe pigmen s, UV-
abso bing phenylp opanoids, UV- is abso p ion capaci y
The as majo i y o lowe ing plan s depend on pollina o s o ep oduc ion (~ 90%1), , ye plan s p oduce
di e en ypes o lo al signals o a ac di e en pollina o s and ensu e seed p oduc ion. Flowe colo is one
o he mos impo an isual cues pe cei ed by pollina o s2 and is con e ed by lo al pigmen s3,4. Pollina o s
can a ec lowe colo s a wo scales (1) as impo an selec i e agen s in popula ions5 and (2) as ecological
il e s in communi ies6–8. The e o e, we expec lo al pigmen composi ion o be igh ly linked o he egion,
i s pollina o s and he local en i onmen al ac o s ha may be selec i e agen s o ecological il e s o lo al
pigmen s9,10. None heless, he ela i e equency o di e en ypes o pigmen s in ela ion o egion, pollina o s
and en i onmen al ac o s emains la gely unexplo ed a a communi y and lo is ic scale.
In addi ion o hei p ima y ole in a ac ing pollina o s, lo al pigmen s may also con e s ess ole ance11–14.
All majo ypes o pigmen s and hei unc ional in e media es and side b anches, such as la onoids,
ca o enoids, chlo ophylls and be alains, possess an ioxidan p ope ies o some ex en 15–18. The UV-abso bing
phenylp opanoids (UAPs he ea e ) a e no ewo hy o hei ene gy di using, UV p o ec i e, and ROS sca enging
capabili ies19–21. This ca ego y encompasses UV-abso bing la onoids, p ima ily la onols and la ones, as well as
hyd oxycinnamic acids, a subg oup o phenylp opanoids. UAPs and an hocyanins in ege a i e issues mi iga e
many plan s esses including ex eme empe a u es, d ough , pa hogens, and he bi o es, as demons a ed19,22.
UAPs a e also ecognized o hei abili y o p e en UV-A and UV-B damage20,21. In ege a i e o gans, UAPs play
a c i ical ole in p o ec ing agains UV adia ion du ing plan s’ ansi ion om wa e o land21,23. Howe e , he
p esence, amoun and unc ion o UAPs in lowe s has ecei ed limi ed a en ion3,24 and hen, only in es iga ed
in a ew lineages (e.g., B assicaceae and Pe unieae25,26). On he con a y, he abundance o an hocyanins a
communi y o egional scales has ecei ed mo e a en ion, bu is mos o en deduced om he e lec ed ligh o
e en mo e gene ally based on human pe cei ed lowe colo ca ego ies, a he han di ec ly h ough pigmen
1Á ea de Bo ánica, Depa amen o de Biología Molecula e Ingenie ía Bioquímica, Uni e sidad Pablo de Ola ide,
Se illa, Spain. 2Depa amen o de Biología Vege al y Ecología, Facul ad de Biología, Uni e sidad de Se illa, Se illa,
Spain. 3Depa men o Biology, San a Cla a Uni e si y, San a Cla a, CA, USA. 4Cen e o Resea ch on Biodi e si y
Dynamics and Clima e Change and Depa men o Biodi e si y, Phenology Lab, UNESP - São Paulo S a e Uni e si y,
Biosciences Ins i u e, Rio Cla o, São Paulo, B azil. 5Edua do Na bona and Jose C. Del Valle con ibu ed equally o
his wo k. email: [email p o ec ed]
OPEN
Scien i ic Repo s | (2025) 15:15927 1
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s
analysis (e.g., B i ish Isles, Aus alia7,27). The concu en accumula ion o an hocyanins and UAPs in lowe s may
no only enhance he an ioxidan capaci y o hese pigmen g oups bu also p o ide a biochemical oolki o
pollina o a ac ion as a main e ec o a byp oduc o pigmen p oduc ion elsewhe e16,28. A c i ical challenge
in his con ex is de e mining how p e alen UAPs and an hocyanins a e in lowe s ac oss a di e si y o plan
lineages and communi ies om di e en egions o he Ea h.
Gi en ha UV i adiance is a signi ican s esso o e es ial plan s29, and pigmen s can mi iga e his
s ess16,19, we i s e iew he UV-Vis abso p ion spec a o he mos common lo al pigmen ypes. In pa icula ,
we compa e each pigmen ’s abili y o abso b damaging UV-A/B ligh (280–400nm). We hen analyse lo al
pigmen s in 936 animal-pollina ed species spanning 115 amilies and 486 gene a om h ee geog aphic egions:
Cali o nia, sou he n Spain, and sou heas e n B azil. We p edic ha i he accumula ion o lo al UAPs is
essen ial o coping wi h many en i onmen al s esses21,29,30 hen hey should be he mos common pigmen s in
angiospe m lowe s whe e hey migh “double dip” as UV-abso p i e guides o pollina o s31,32. In compa ing
lo al pigmen s ac oss hese egions, we also examined each pigmen ’s abundance wi h ega ds o ligh
en i onmen s ( o es shade s. open) and p ima y pollina o (insec s. bi d). Pollina o a ac ion is in luenced
by he wa eleng hs o ligh p esen in a gi en en i onmen , which can a y ac oss di e en habi a s33,34. We
p edic ha lowe s in o es shade should be yellow o yellow-g een o maximize b igh ness and/o ach oma ic
con as when ha is he p ima y wa eleng h o ligh pene a ing a dense canopy33. These lowe colo s a e usually
con e ed by ca o enoids, au one-chalcones, and/o chlo ophylls. Finally, bi d-pollina ed lowe s a e usually ed
(as pe cei ed by humans) and s ongly UV-abso p i e o simul aneously a ac bi ds and be less conspicuous
o bees35–37. These bi d pollina ed lowe s should be mo e likely o combine UAPs wi h ed an hocyanins (o en
complemen ed wi h ca o enoids) han insec pollina ed lowe s9,37. We e eal ha ega dless o he lowe colo
and i espec i e o which pigmen s a e p esen in he lowe , UAPs a e ubiqui ous ( he “da k ma e ” o he
lowe ). Flo al pigmen composi ion was su p isingly consis en ac oss he h ee geog aphic egions s udied (a
anscon inen al s able dis ibu ion). Regional a ia ion was p ima ily in luenced by di e ences in pollina ion
sys ems and, o a lesse ex en , by ligh en i onmen .
Resul s and discussion
Abso p ion p ope ies o majo g oups o lo al pigmen s
We e iewed he li e a u e o ind ou which pa s o he ligh spec um is abso bed by he ou main classes o
lo al pigmen s, namely phenylp opanoids, ca o enoids, chlo ophylls and be alains, and hei de i a i es (Fig.
1). Wi hin he phenylp opanoids, hyd oxycinnama es (aka hyd oxycinnamic acids), la ones and la onols
exclusi ely abso b in he UV ange. Speci ically, hyd oxycinnama es abso b mainly in he UV-B egion o
he spec um (peaks ≈ 280–330 nm) and la ones and la onols in he UV-A egion (≈ 310–390 nm38,39).
Au ones and chalcones ( ea ed oge he he ein) abso b in he UV-blue egion (peaks ≈ 350–430nm), whe eas
an hocyanins mainly abso b in he g een-blue egion (≈ 475–560nm)39–41. Flowe s may con ain o he g oups
o la onoids such as la anones, iso la ones, ca echins o epica echins, bu hey a e ela i ely a e compa ed o
he a o emen ioned g oups41,42. Al e na i ely, ca o enoids abso b mainly in he blue-g een spec al egion o
he isible ligh (peaks ≈ 400–530nm) and chlo ophylls abso b in he blue and ed egions (≈ 440 and 660nm,
espec i ely); al hough chlo ophyll a has a ela i ely mode a e abso p ion abili y in he UV-A egion39,43,44.
Las ly, be alains show abso p ion spec a simila o an hocyanins16,39, bu hey a e es ic ed o 20 amilies
wi hin he o de Ca yophyllales45. In summa y, he only pigmen s wi h a conside able capaci y o abso b ligh
in he UV egion o he spec um a e he UAPs (i.e., hyd oxycinnama es, la onols and la ones) and au ones-
chalcones. This would explain why in ege a i e issues UV-B exposu e gene ally p omo es he biosyn hesis o
hyd oxycinnama es, while UV-A exposu e s imula es he p oduc ion o la onols and la ones20,46,47 (see also48).
Al hough e idence in lowe s is mo e limi ed, a ailable s udies sugges a simila end14,49.
UV-abso bing phenylp opanoids a e ubiqui ous in lowe s
We pe o med a biochemical analysis o lowe s o 926 animal-pollina ed species om di e se habi a s o
Cali o nia (442 species), sou he n Spain (381), and sou h-eas e n B azil (103) (Supplemen a y Da a 1). Using a
di e en ial ex ac ion me hod ollowed by an analysis o abso bance spec a (see de ails in Me hods), we we e
able o iden i y six majo g oups o pigmen s: UAPs (hyd oxycinnamic acids, la ones and la onols), au ones-
chalcones, an hocyanins, chlo ophylls, ca o enoids, and be alains (Fig. 2A). No ably, he p esence o UAPs was
almos ubiqui ous in species om he h ee egions (> 99.8%; Fig. 2B). Only one species om Cali o nia, Geum
mac ophyllum (Rosaceae), lacked UAPs in he lowe s, bu i s ill exhibi ed an abso bance peak below 280nm,
howe e his was below ou cu -o o de ining UAPs. An hocyanins we e he second mos abundan pigmen
g oup, p esen in 55.9% o species ac oss he h ee s udy a eas, while ca o enoids appea ed in 37.2% o species
(some species con ain mul iple pigmen s). Chlo ophylls anged om 12 o 21% (mean 17.2% ac oss he h ee
egions), while au one-chalcone pigmen s we e in equen , ound in less han 4% o all species. We also con i m
ha be alains a e uncommon pigmen s, p esen in only ou sampled species om Cali o nia (0.9%; Fig. 2B).
Recen ly, sepa a e s udies in he B assicaceae, O chidaceae, and Solanaceae sugges ha UAPs accumula e in
lowe s ega dless o isible colo and i espec i e o he p esence o absence o lo al guides25,26,50. Ou esul s
clea ly show ha lo al UAPs a e omnip esen in he species s udied and, p esumably, mos angiospe ms. Ou
esul s align wi h he ubiqui y o UAPs epo ed om ege a i e issues whe e hey se e mul iple p o ec i e
unc ions19,21. Fo ins ance, lo al UAPs may coun e ac oxida i e damage induced by UV adia ion o d ough
in pe als51,52 o help o main ain cellula u go h ough suga signaling53. Howe e , he exac unc ion(s) o
UAPs in pe al issues is la gely unknown and he e o e, we e e o i as he “da k ma e ” o he lowe , o now.
Al hough we ound ha lowe s can accumula e up o ou ypes o pigmen g oups, mos species con ained
only wo pigmen ypes (57% species wi h wo ypes in Cali o nia & B azil and 68% o species in Spain) (Fig. 2C
and Supplemen a y Fig.1). Because o he omnip esence o UAPs in lowe s, he mos common combina ion
Scien i ic Repo s | (2025) 15:15927 2
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/
o he wo pigmen s was UAPs + an hocyanins (o e all 58–66% o species; Supplemen a y Fig.S2). In pe al
cells, an hocyanins ypically accumula e in acuoles54, whe e hey can be ound alone o bonded o UAPs
by copigmen a ion o o he molecula in e ac ions o s abilize and/o in ensi y he colo 41,55. Flowe colo
can also appea mo e in ense when pigmen s a e concen a ed on he isible side o when ligh sca e ed by
he unde lying unpigmen ed laye passes h ough he pigmen ed laye wice56. The associa ion o UAPs
wi h an hocyanin de i a i es is one ac o con ibu ing o he eno mous ange o colo s and colo pa e ns
in angiospe m lowe s41,42,57. In addi ion o he e ec on lo al colo pe cep ible o humans and pollina o s,
he associa ion be ween UAPs and an hocyanins esul s in an hocyanins gaining a subs an ial ligh abso p ion
capaci y in he UV-A and/o UV-B egion20,58,59. Thus, accumula ion o UAPs and an hocyanins in lowe s may
be ad an ageous o he plan and can be shaped by bo h bio ic and abio ic ac o s16,60.
Fig. 1. Ul a iole (UV) and isible abso p ion spec a o majo pigmen g oups con ibu ing o lowe
colo a ion. Due o hei s uc u al di e si y and dis inc con ibu ions o lo al colo a ion, phenylp opanoids
we e ca ego ized in o hyd oxycinnama es, UV-abso bing la onoids, au ones-chalcones, and an hocyanins.
The spec al egion co esponding o each pigmen ype is colo -coded based on he human pe cep ion
o e lec ed ligh om pe als. The iole -shaded a ea highligh s he UV-A and UV-B egions o he ligh
spec um (400–315nm and 315–280nm, espec i ely) and shows ha hyd oxycinnama es and UV-abso bing
la onoids a e he pigmen s wi h he highes ligh abso p ion capaci y in his ange. Fo each g oup o
pigmen s, a no malized abso p ion spec um o an example compound is shown (see Me hods). The s uc u es
o p-couma ic acid (hyd oxycinnama e), que ce in (UV-abso bing la onoid), cyanidin (an hocyanin),
be acyanin (be alain), lu ein (ca o enoid), and chlo ophyll a (chlo ophyll) a e depic ed.
Scien i ic Repo s | (2025) 15:15927 3
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/
UV pa e ning in pe als is a e
Mos species accumula ed UAPs ela i ely e enly h oughou hei pe als (80.7, 86.9 and 92.3% in Cali o nia,
S Spain and SE B azil, espec i ely). The emaining species had UV–pa e ning wi hin hei lowe s including
bullseyes, eins, ays, spo s, and/o con as ing pe als/ epals (e.g., lowe s o Fabaceae o O chidaceae, Fig. 3,
Supplemen a y Da a 1; see also61). F om ou UV digi al images, a ew species appea ed ha e UV- e lec i e
co ollas (see o he examples in32,62,63); howe e , ou abso bance analysis con i med he p esence o UAPs, e en
in hese mos ly UV e lec i e species. This appa en con adic ion may be explained by: (1) he highe accu acy
o abso bance analysis, which de ec s UAPs e en a low concen a ions; (2) he limi a ions o UV digi al images,
which do no cap u e he ull UV-A and UV-B ange used in he abso bance analysis; and (3) he possibili y ha
UAPs accumula e exclusi ely on he ou e sides o pe als63,64. No ably, UAPs may p o ec lo al issues e en when
Scien i ic Repo s | (2025) 15:15927 4
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/
con ined o lo al guides. Fo ins ance, UAPs accumula ion in bullseyes may p o ec ep oduc i e s uc u es
om UV adia ion, con ibu e o wa ming he gynoecium, o enhance pe al esis ance o desicca ion14,65–67.
Abundance o lowe pigmen s a e consis en ac oss egions
We obse ed s iking simila i ies in he abundance o lo al pigmen ac oss he h ee con inen s (Fig. 2B and
Supplemen a y Fig. 3). A su ey o he B i ish lo a using 1249 species, which ca ego ized pigmen s in o b oad
ca ego ies based on human pe cei ed colo —such as pink, ed, blue, iole , and pu ple (an hocyanins) and
yellow (ca o enoids)—simila ly epo ed a highe equency o an hocyanins compa ed o ca o enoids (36
and 26%, espec i ely)27, which is lowe han wha we epo he ein. This disc epancy likely s ems om hei
me hod, which conside ed only he p edominan pigmen esponsible o he main lowe colo , whe eas ou
app oach accoun ed o all pigmen s p esen in he lowe s. Using he same lowe colo ca ego iza ion as Wa en
and MacKenzie27, a highe equency o an hocyanins compa ed o ca o enoids has been consis en ly obse ed
ac oss a ious egions, including F ance, Scandina ia, Canada, opical A ica, Aus alia, and Ja a68,69, as well as
on a global scale70. Phylogene ic simila i y can be uled ou as possible explana ion o he consis en abundance
o pigmen wo ldwide gi en he dis inc lo as o hese egions. In ac , he equency o sha ed amilies in ou
s udy was low: 30.6% (38 o 124 amilies) be ween Cali o nia and Spain, 15.9% (14 o 88) be ween Spain and
B azil, 14.0% (14 o 100) be ween Cali o nia and B azil, and only 7.6% (12 o 157) o amilies a e ound in
all h ee egions. Howe e , he consis en equencies obse ed ac oss egions wo ldwide a e likely d i en by
mul iple ac o s, such as pollina o p e e ences, e olu iona y cons ains and/o en i onmen al il e ing6,8,71–74.
Pollina o s a e conside ed impo an selec i e agen s on lowe colo 5–7. Hymenop e ans show di e ences in
nai e p e e ences among species and ha e complex lea ning mechanisms75, bu in gene al hey p e e blue-
iole and yellow o e o he lowe colo s, wi h a bias owa ds blue- iole 76. Thus, he highe abundance o
an hocyanins compa ed o ca o enoids obse ed ac oss h ee s udy egions may be explained by he lo al
colo p e e ences o hymenop e ans, which a e he mos equen pollina o s in each egion. E olu iona y
cons ain s can also shape pigmen abundance, as he biochemical pa hways esponsible o he p oduc ion
o majo lo al pigmen g oups (e.g., an hocyanins, ca o enoids, and chlo ophylls) a e highly conse ed ac oss
angiospe ms16,17,22,23. Las ly, en i onmen al il e ing may con ibu e o he highe equency o an hocyanins
compa ed o ca o enoids i en i onmen al s esso s a e simila ac oss egions. I is well-documen ed ha
abio ic ac o s such as empe a u e, p ecipi a ion, and sola adia ion a e globally p e alen and can inc ease he
equency o species ha accumula e lo al an hocyanins and/o UAPs30,77–79.
Pigmen composi ion is in luenced by bo h abio ic and bio ic ac o s
We es ed whe he he abundance o majo lo al pigmen g oups is in luenced by di e en ligh en i onmen s—
shaded ( o es , ipa ian) s. exposed (g asslands, coas al, ocky, e c.)—in Cali o nia and S Spain (sample sizes
pe ligh en i onmen we e oo small o SE B azil). In bo h egions, mos sampled species belong o he
exposed ligh en i onmen (82% and 88.8%, espec i ely). We ound highe equency o chlo ophylls in shaded
en i onmen s compa ed o exposed en i onmen s in S Spain (Fig. 4 and Supplemen a y Table 1). Howe e ,
only 13% o hese species showed en i ely g een lowe s, while he emaining species accumula ed chlo ophylls
alongside o he pigmen s (mainly an hocyanins), esul ing in di e se colo pa e ns. This in e ac ion be ween
chlo ophylls and o he pigmen s has been shown o enhance a ac ion o insec pollina o s80,81. Thus, ou indings
do no suppo Endle ’s p edic ion o a highe equency o yellow-g een lowe s in shaded en i onmen s33.
Simila ly, a compa a i e s udy conduc ed in Ge man g asslands, which analyzed bo h ch oma ic and ach oma ic
Fig. 2. Pigmen iden i ica ion and equency in he s udied egions. (A) Examples o abso bance spec a
o acidi ied me hanol and ace one ex ac s (black and g ay lines, espec i ely) om lowe s o di e en
colo s, caused by a combina ion o pigmen classes. The di e en ial sol en ex ac ion and he cha ac e is ic
abso p ion peaks o each pigmen ype allow iden i ica ion o he pigmen ypes p esen in he c ude ex ac .
Colo ed d ops a e used o show he iden i iable peaks o he pigmen ypes. In he me hanol ex ac s, iole
d ops a e used o UV-abso bing phenylp opanoids (UAPs), g ey d ops o au ones-chalcones, ed d ops
o an hocyanins, and g een d ops o chlo ophylls. Be alains no included due o hei a i y. In lowe s
ha accumula e se e al kinds o pigmen s, some pigmen peaks may o e lap, bu he cha ac e is ic ypes
o each pigmen may allow hem o be di e en ia ed (see me hods). Ace one ex ac s a e used o iden i y
dis inc i e h ee-peaked ca o enoids and a e ma ked by a yellow d op. Be alains show a simila beha io o
hose o an hocyanins bu a e no ep esen ed because o hei a i y. An example o a species showing a single
pigmen class (Calys egia sepium, Con ol ulaceae, Spain; whi e lowe s wi h UAPs). Examples o species wi h
wo pigmen s classes: P osa es hooke i (Liliaceae, Cali o nia; g een lowe s wi h UAPs and chlo ophylls),
Na cissus assoanus (Ama yllidaceae, Spain; yellow lowe s wi h UAPs and ca o enoids), Lina ia iscosa
(Plan aginaceae, Spain; yellow lowe s wi h UAPs and au ones-chacoles), Cen au ium e y h aea (Gen ianaceae,
Spain; pink lowe s wi h UAPs and an hocyanins), Commelina e ec a (Commelinaceae, B azil; blue lowe s
wi h UAPs and an hocyanins). An example o species wi h h ee pigmen s (Bella dia ixago, O obanchaceae,
Spain; yellow- ed lowe s wi h UAPs, an hocyanins and ca o enoids), and wi h ou pigmen s (Cas illeja
minia a, O obanchaceae, Cali o nia; ed lowe s wi h UAPs, an hocyanins, chlo ophylls and ca o enoids). (B)
Pe cen age o species con aining each ype o lo al pigmen in Cali o nia, S Spain, and SE B azil (N = 442, 381
and 103, espec i ely). No e ha in a lowe each pigmen ype may appea alone o coexis wi h o he pigmen
ypes. (C) F equency o he numbe o lo al pigmen ypes p esen in species om Cali o nia, S Spain, and SE
B azil (N = 442, 381 and 103, espec i ely).
◂
Scien i ic Repo s | (2025) 15:15927 5
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/

componen s o lowe colo s using he honeybee colo ision model, ound no signi ican di e ences in lowe
colo s be ween closed o es and open ligh en i onmen s82.
We also examined i insec - and hummingbi d-pollina ed species di e ed o lo al pigmen dis ibu ion.
Ou da abase includes 9.7% o species ca ego ized as hummingbi d-pollina ed in Cali o nia and 12.8% in
sou heas e n SE B azil (hummingbi ds a e no p esen in S Spain). In Cali o nia, he abundance o an hocyanins
and ca o enoids in bi d-pollina ed lowe s was nea ly double ha o insec -pollina ed lowe s, while chlo ophyll
showed he opposi e end (Fig. 4 and Supplemen a y Table 2). The ed colo is a de ining cha ac e is ic o he
Fig. 3. Visible and UV digi al images e eal he homogeneous accumula ion o UAPs in pe als o hei spa ial
dis ibu ion c ea ing pa e ns. Visible (le ) and UV ( igh ) digi al images indica e UV-abso bing a eas in he
co olla caused by he accumula ion o UAPs. Scale ba s = 10mm. (A) Aquilegia o mosa (Rannunculaceae,
Cali o nia) showing homogeneous UAPs accumula ion in bo h pe als (spu ) and sepals (pe al like) a e UV
abso bing. (B) Symphyo ichum spa hula um (As e aceae, Cali o nia) homogeneous UAPs accumula ion in
bo h disk and ay lo e s. (C) Ononis pubescens (Fabaceae, Spain) showing isible and UV eins, and keel and
wings wi h high UV abso p ion. (D) Madia elegans (As e aceae, Cali o nia) showing a highly UV-abso bing
bullseye in disk lo e s and base o ay lo e s. (E) T i eleia ixioides (Aspa agaceae, Cali o nia) showing UV-
abso bing ays on he epals and co ona, in his case is accompanied by chlo ophylls causing g eenish pu ple
ays in isible pho og aphy. (F) Mal a syl es is (Mal aceae, Spain) showing UAPs nec a guides in addi ion
o an hocyanins in he emainde o he pe al. (G) Pa kinsonia lo ida (Fabaceae, Cali o nia) showing isible
do s and banne wi h high UV abso p ion. (H) Adelinia g andis (Bo aginaceae, Cali o nia) showing a UV-
abso bing bullseye in he co olla.
Scien i ic Repo s | (2025) 15:15927 6
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/
hummingbi d pollina ion synd ome83. This colo a ion esul s om he accumula ion o pink- ed an hocyanins,
speci ically cyanidin and pela gonidin de i a i es, o , mo e commonly, om a combina ion o hese an hocyanins
wi h yellow ca o enoids9,37. Se e al s udies ha e demons a ed ha he p esence o hese pigmen s, along wi h
UAPs, enhances lowe conspicuousness o hummingbi ds while educing de ec abili y o bees35,37,84. Ou
indings con i m he ubiqui y o hese pigmen s ac oss 16 plan amilies in Cali o nia (Supplemen a y Da a
1) and sugges ha hummingbi ds a e d i ing he e olu ion o lo al pigmen a ion in his egion. A highe
abundance o an hocyanins and ca o enoids was obse ed in hummingbi d-pollina ed lowe s om SE B azil,
bu di e ences we e no s a is ically signi ican likely due o he limi ed sample size. No ably, he numbe o plan
Fig. 4. Pe cen age o species accumula ing lo al pigmen ypes ac oss di e en ligh en i onmen s and
pollina ion sys ems. (A) Compa ison o he pe cen age o species wi h he h ee main pigmen g oups be ween
exposed (ligh b own) and shaded (g een) ligh en i onmen s in Cali o nia (N = 328 and 72, espec i ely). (B)
Compa ison o he pe cen age o species wi h h ee main pigmen g oups be ween exposed (ligh b own) and
shaded (g een) ligh en i onmen s in S Spain (N = 309 and 39). (C) Compa ison o he pe cen age o species
wi h he h ee main pigmen g oups be ween insec (ligh o ange) and hummingbi d ( ed) pollina ion sys ems
in Cali o nia (N = 393 and 45, espec i ely). (D) Compa ison o he pe cen age o species wi h h ee main
pigmen g oups be ween insec (ligh o ange) and hummingbi d ( ed) pollina ion sys ems in SE B azil (N = 94
and 8). Symbols on he igh show he esul s o pe mu a ion es s o compa e he equency o each pigmen
g oup be ween ligh en i onmen s and pollina ion sys ems (see s a is ical esul s in Supplemen a y Tables 1
and 2). n.s. no signi ican , **p < 0.01, ***p < 0.001. Images o ligh en i onmen s and pollina ion sys ems we e
gene a ed by AI image gene a o DALL-E.
Scien i ic Repo s | (2025) 15:15927 7
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/
species exhibi ing he hummingbi d pollina ion synd ome in he B azilian Ce ado is ela i ely low compa ed
o o he egions o he Ame icas85–87.
Conclusions
Ou lo al biochemical esul s highligh he ubiqui ous na u e o UAP compounds in animal-pollina ed
lowe s ac oss h ee con inen s, di e se habi a s and pollina ion synd omes. We a gue ha UAPs may ha e
a dual ole in lowe s, a ac ing pollina o s and p o ec ing agains en i onmen al s ess16. UAPs a e p esen
in ea ly land plan s, allowing hem o cope wi h UV adia ion and he mal s ess23,29, which we e likely co-
op ed by angiospe m pe al-speci ic egula o y modules o p oduce he di e si y o lowe colo s we know oday.
Thus, we conclude ha he ances al (and p ima y) unc ion o UAPs in lo al issues is o p o ec cells om
en i onmen al s esso s and he ole o pollina o a ac ion could hen be an exap a ion88. The omnip esence
o lo al UAPs may be explained by hei o igin du ing he e es ializa ion o plan s, ha ing been e ained in
all plan s due o essen ial na u e o su i al and a o ed due o hei e sa ili y as en i onmen al p o ec o s89.
Ne e heless, UAPs can be ega ded as he “da k ma e ” o lowe s, as hei p esence is demons able, ye hei
p ecise unc ions emain unce ain.
We ound ha abundance o lo al pigmen s a e consis en ac oss he h ee s udied egions, a pa e n ha
likely ex ends globally. Ou esul s o e a mo e comp ehensi e unde s anding o lo al pigmen abundance
and u he s eng hen p e ious s udies based on lowe colo o heo e ical models71,77,78. The causes o hese
s able dis ibu ions a e no well unde s ood bu likely e lec he lowe colo p e e ences o insec s— he mos
equen pollina o s—as well as unde lying biochemical and molecula ac o s, wa an ing u he in es iga ion.
The a ia ion in p esence o pigmen s be ween exposed and shaded en i onmen s, as well as be ween insec - and
hummingbi d-pollina ed lowe s, sugges s ha eco-e olu iona y p ocesses a e ac ing a egional scales amids a
ela i ely s able global dis ibu ion o lowe colo 7,77,90.
Me hods
UV- is abso p ion spec a o main lowe pigmen g oups
We e iewed e e ences p esen ing spec al da a o pigmen s ex ac ed in me hanol, as i is one o he mos
common sol en s used o plan pigmen s31,91. Fo each main pigmen g oup, we selec ed he mos ep esen a i e
o equen subg oups ound in lowe s41,42,57,92,93 and p esen a no malized abso p ion spec um o an example
compound dilu ed in me hanol in (Fig. 1). The sou ces used in d awing hese a e age abso p ion spec a a e
shown in (Supplemen a y Table 3). Al hough he e may be a ia ion in he shape o he cu es and maximum
abso bance wa eleng h be ween di e en pigmen ypes wi hin he same biochemical g oup, his a ia ion is
usually qui e small compa ed o he di e ences be ween pigmen g oups31,39,92.
Sampling o he lo al pigmen composi ion analyses
In Cali o nia (Cali o nia Flo is ic P o ince), we collec ed lowe s om a o al o 442 na i e species belonging
o 249 gene a and 69 amilies, and in sou he n Spain, we collec ed a o al o 381 na i e species belonging o 198
gene a and 56 amilies (Supplemen a y Da a 1). In sou heas e n B azil (Na ional Pa k o Se a do Cipó, Minas
Ge ais, B azil), we collec ed lowe s om 103 species belonging o 72 gene a and 32 amilies. Samples we e
collec ed om 2019 o 2023 in Cali o nia and Spain and in 2019 in B azil.
Collec ion o plan ma e ial complied wi h all ele an na ional and in e na ional guidelines and legisla ion.
Necessa y pe missions we e ob ained o collec samples in he h ee s udied egions. Vouche specimens we e
deposi ed in he he ba ia o he Uni e si y o Se ille (He ba io SEV), Uni e si y Pablo de Ola ide (UPOS),
Ins i u o de Biociências a he Uni e sidade Es adual Paulis a in Rio Cla o (HRCB), and San Jose S a e Uni e si y
(SJSU). Collec ions ha e accession numbe s a he SJSU and SEV he ba ia, whe eas in UPOS and HRCB, hey
a e unnumbe ed bu s o ed in sepa a e boxes. Plan specimens we e iden i ied by Jus en B. Whi all (Cali o nia),
Edua do Na bona, Ma isa Buide, Mon se A is a, and Ped o L. O iz (Spain), and Pa icia Mo ella o (B azil)
using specialized e e ences, including he Jepson Manual eFlo a (h ps://ucjeps.be keley.edu/e lo a) and Flo a
Ibe ica (h p://www. lo aibe ica.es), as well as he ba ium collec ions and p e ious s udies conduc ed a he same
s udy si es35.
Bo h Cali o nia and S Spain ha e a Medi e anean clima e wi h ypically ho , d y summe s and cool, we
win e s94. In SE B azil (Minas Ge ais s a e), he clima e is opical, wi h a cool d y season, a wa m we season,
and equen i es and s ong winds du ing he d y- o-we ansi ion95. In Cali o nia and S Spain, we sampled
b oadly ac oss habi a s by collec ing in g asslands, sh ublands, o es s, ipa ian, ocky, we lands, coas al,
moun ain and dese s communi ies h oughou he yea . In SE B azil, we su eyed he ocky g asslands a high
al i udes (> 900m a.s.l., Campo upes e sensu s ic o) and he woody sa anna ege a ion (Ce ado sensu s ic o).
The main pollina o s in he h ee s udy egions we e insec s (p ima ily hymenop e ans, bu also dip e ans,
lepidop e ans, coleop e ans, among o he s35,87,94). Howe e , in bo h Cali o nia and SE B azil, hummingbi ds
se e as pollina o s, wi h app oxima ely 5–12% o he lo a exhibi ing he ypical “hummingbi d-pollina ion
synd ome”35,87,96. We analyzed he pigmen con en o lo al pa s wi h he highes ad e isemen display,
ypically pe als o epals, a an hesis.
In all cases, we picked lowe s o in lo escences a an hesis o se e al indi iduals o accoun o be ween-
indi idual a ia ion in pigmen composi ion. Addi ionally, 56 species we e collec ed in wo di e en popula ions
o es o pigmen a ia ion be ween popula ions; in all cases, he quali a i e pigmen p o iles we e iden ical
and, hus, we used he da a o he i s popula ion analyzed.
We sampled “a ac ion uni s”97, which in mos species coincide wi h indi idual lowe s bu in some species,
we conside ed he en i e in lo escence (e.g., spadix o A aceae spp. o compound in lo escences o As e aceae
spp.). We analyzed pigmen con en o he lo al piece ha gene a es he highes ad e ising display, usually
Scien i ic Repo s | (2025) 15:15927 8
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/
pe als o epals, bu lo al b ac s we e also examined in some species (e.g., A is olochiaceae spp., Eupho biaceae
spp., Cas illeja spp., o Rhynchospo a spp.).
Ex ac ion, iden i ica ion, and quan i ica ion o majo pigmen g oups
Collec ions we e s o ed a 4°C un il pigmen ex ac ion was ca ied ou , ypically wi hin 24–36h. We used wo
sol en s o ex ac and sepa a e he p incipal pigmen classes in each sample: me hanol wi h 1% HCl ( : ) and
pu e ace one39,98. The acidi ied me hanol solu ion is pa icula ly e ec i e in ex ac ing UAPs, an hocyanins,
be alains, and chlo ophylls, whe eas ace one mainly ex ac s ca o enoids and chlo ophylls39,91,99. We used
me hanol as a sol en ins ead o e hanol o aqueous me hanol solu ions due o i s supe io e iciency o ex ac
pola compounds100. Acidi ica ion o me hanol wi h HCl con ibu es o he s abiliza ion o he an hocyanins in
pa icula 99. We weighed (5–25mg o esh weigh ) and subme ged he lo al issue in wo mic o ubes con aining
1.5ml o each sol en , which was s o ed a -20°C un il u he analysis. In Cali o nia, we used 5–10 silica beads
o issue homogeniza ion o 2–3min ollowed by 5min o cen i uga ion a maximum speed. In Spain and
B azil, homogeniza ion was employed solely in cases in ol ing hick epals o lo al b ac s101.
We ob ained he abso bance spec a o he samples in each sol en by means o wo ul a iole - isible (UV-
is) spec opho ome e s. We used a Mul iskan GO mic opla e spec opho ome e (The mo Fishe Scien i ic
Inc., MA, USA) and Spec aMax M3 (Molecula De ices, San Jose, CA, USA) wi h ace one-compa ible
polyp opylene 96-well mic opla es, and a D awell DU-8800DS double beam UV/Vis spec opho ome e
(Chongqing D awell Ins umen Co., L d, China) wi h 1-cm qua z cu e es, espec i ely. P e ious s udies
showed ha pigmen quan i ica ion using plas ic pla es s. single-sample cu e es p o ides nea ly iden ical
esul s102. We se he scan mode om 280 o 700nm wi h 1–2nm s eps a a cons an empe a u e o 22°C (no
shaking). We selec ed wa eleng hs abo e 280nm since below ha all phenolic compounds a e cha ac e ized
by a UV band II ha peaks a 240–275nm making i useless o di e en ia ion39. We used 150µl and 500µl
pe sample in 96-well mic opla es and 1-cm qua z cu e es, espec i ely. Following he echnical speci ica ions
o he spec opho ome e s, concen a ed pigmen ex ac s we e dilu ed o ob ain abso bance alues (op ical
densi y) below 2.0 abso bance uni s (AU) o gua an ee eliable measu emen s.
Since ou objec i e was o iden i y majo pigmen classes, we pe o med spec opho ome ic analyses o
ob ain he abso p ion spec a o he sample ex ac s. Al hough spec opho ome e s ha e less abili y o dis inguish
biochemical a ia ion wi hin a pigmen class, his app oach is a eliable, as , and inexpensi e al e na i e o
HPLC sepa a ion wi h mass o NMR spec ome y99,103. Since he majo pigmen classes ha e a cha ac e is ic
ligh abso p ion spec um wi h dis inc i e peaks, we we e able o iden i y he p esence o main pigmen ypes
om aw lo al ex ac s28,39,103. Thus, ca o enoids show h ee dis inc i e peaks be ween 400 and 530nm wi h a
majo peak a ound 450nm, whe eas chlo ophylls show wo main peaks a 415–460nm and 650–665nm43,44. All
lo al samples wi h chlo ophylls showed a peak a ~ 418nm, indica i e o chlo ophyll a44. We dis inguish h ee
majo g oups o phenylp opanoids: an hocyanins, au ones-chalcones, and UV-abso bing phenylp opanoids
(UAPs). The la onoid an hocyanins show a cha ac e is ic peak a ound 475–560nm, whe eas he au ones-
chalcones exhibi a peak a 350–430nm39,41. UAPs include non- isibly pigmen ed la onoids such as la anones,
la ones, and la onols wi h p incipal peaks be ween 280 and 360nm, and some g oups o hyd oxycinnama es,
such as cinnamic, ca eic, e ulic, p-couma ic, and sinapic acids which ha e a dis inguishing peak a 280–
330nm38–41. The di e en g oups o UAPs show dis inc i e peaks, bu mos o hem o e lap, which p ecludes
hei di e en ia ion using his me hodology104. Spec opho ome ic iden i ica ion does no allow he de ec ion
o o he g oups o la onoids such as iso la ones o ca echins ha show hei main peaks below 280nm39,104.
Wi h espec o be alains pigmen s, pink- ed be acyanins we e dis inguished om an hocyanins due o a highe
wa eleng h o abso p ion peak (532–554nm), and yellow be axan hins we e dis inguished om ca o enoids
by he ex ac ion in me hanol sol en and he p esence o only one peak a 450–500nm39. In addi ion, we
con i med ha he sample was in a plan amily ha was p e iously desc ibed o p oduce be alains45. We ha e
no quan i ied concen a ions o each pigmen g oup because ou me hods do no allow o he iden i ica ion o
speci ic compounds (i.e. ype o an hocyanins, class o ca o enoids, mo e speci ic iden i ica ion o UAPs, e c.)
p esen in each species. In species om Spain wi h whi ish, c eam, pale-yellow, and yellow ex ac s, in which we
obse e an abso p ion peak a ound 350–450nm, we pe o med wo addi ional es s o co obo a e he p esence
o au ones-chalcones s. ca o enoids: colo eac ion in me hanolic HCL ex ac s and di e en ial sepa a ion wi h
wa e and dichlo ome hane105.
We ound some species showed abso bance maxima ha did no i any o he majo pigmen s p e iously
men ioned. We pe o med a bibliog aphic sea ch o ind ou i he e we e biochemical da a on he lo al pigmen
o hese species o hei ela i es. Quinones a e a a e g oup o pigmen s ha may be ound in some lowe s, mainly
an h aquinones o quinochalcones106. In ou s udy, Dipcadi se o inum p esen ed a compound wi h only one peak
a 460nm ha was ex ac ed in bo h me hanol and ace one solu ions; his was cong uen wi h a quinone39, ye
yellow an h aquinones ha e been ound in o he species o Aspa agaceae106. Simila ly, he compound peaking a
500nm in all species o Xy is was also cong uen wi h p e iously desc ibed an h aquinones o his genus107. To
calcula e he equency o he main ypes o pigmen s, quinones we e included in he an hocyanin g oup since
hey sha e he ea ly s eps o he biosyn he ic pa hway106. Xan hones is a a e g oup o la onoids in lowe s ha
show abso bance peaks simila o iso la ones, la ones, and la onols39,106. In lowe s o I is spp. and Hype icum
spp., xan hones ha e been p e iously desc ibed108,109, hus, hey may be p esen , along wi h o he UAPs, in ou
me hanol ex ac s. Finally, Adonis mac oca pa showed a compound wi h a single peak a 480nm appea ed in
bo h me hanol and ace one solu ions, which ag eed wi h he ed ca o enoid as axan hin p e iously epo ed in
A. aes i alis110.
In species wi h obse ed UV o isible colo pa e ns (see UAPs loca ion in pe als sec ion), sepa a e samples
we e analyzed (e.g., di e en lo al pieces o O chidaceae spp., apical and basal po ions o ligula e/ ay lowe s o
As e aceae). In he Fabaceae amily, we sepa a ely analyzed he banne , keels, and wings, excep o species wi h
Scien i ic Repo s | (2025) 15:15927 9
| h ps://doi.o g/10.1038/s41598-025-94709-4
www.na u e.com/scien i ic epo s/