Macroevolutionary trade-offs among floral pigments shape angiosperm color diversity

Macroevolutionary trade-offs among floral pigments shape angiosperm color diversity | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 29 October 2025 V1 Latest version Share on Macroevolutionary trade-offs among floral pigments shape angiosperm color diversity Authors : Montserrat Arista , Marcial Escudero , Justen Whittall , Jose C. Del Valle , Melissa León-Osper , M. Luisa Buide , Maria Gabriela Gutierrez Camargo , … Show All … , Leonor Patricia Cerdeira Morellato , Nancy Rodríguez-Castañeda , Victor Rossi , Katie Conrad , Josephine Hernandez-Mena , Pedro L. Ortiz , and Eduardo Narbona 0000-0003-1790-6821 [email protected] Show Fewer Authors Info & Affiliations https://doi.org/10.22541/au.176175426.68257877/v1 269 views 174 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract The remarkable diversity of floral color arises primarily from three biosynthetically independent pigment groups: anthocyanins, carotenoids and chlorophylls. Because each group aids both pollinator attraction and stress tolerance, we hypothesized that metabolic costs favor minimizing redundancy, leading most species to rely on a single dominant group. Using a dataset of 1,133 animal-pollinated species spanning ~27 % of angiosperm families, we reconstructed the macroevolution of anthocyanins, carotenoids, chlorophylls and aurones. Pigment evolution followed a stepwise trajectory with a strong trend toward single-group dominance. Anthocyanins exhibited balanced gains and losses, while carotenoids, chlorophylls and aurones were more often lost. All groups showed significant phylogenetic signal, strongest in anthocyanins and carotenoids. Negative phylogenetic associations between carotenoids and both anthocyanins and chlorophylls reveal biosynthetic trade-offs limiting pigment combinations. These results reveal that most angiosperm lineages achieve floral color diversity through the predominance of a single pigment group, with specific pigments concentrated in distinct clades. Macroevolutionary trade-offs among floral pigments shape angiosperm color diversity Montserrat Arista 1 , Marcial Escudero 1 * , Justen B. Whittall 2 , Jose C. Del Valle 1 , Melissa León-Osper 2 , M. Luisa Buide 3 , Maria Gabriela Gutierrez Camargo 4 , Leonor Patricia Cerdeira Morellato 4 , Nancy Rodríguez-Castañeda 1 , Victor Rossi 2 , Katie Conrad 2 , Josephine Hernandez-Mena 2 , Pedro L. Ortiz 1 , Eduardo Narbona 3 * 1 , Departamento de Biología Vegetal y Ecología, Facultad de Biología, Universidad de Sevilla, Sevilla, Spain. 2 , Department of Biology, Santa Clara University, Santa Clara, California, United States of America. 3 , Departamento de Biología Molecular e Ingeniería Bioquímica, Universidad Pablo de Olavide, Sevilla, Spain. 4 , Center for Research on Biodiversity Dynamics and Climate Change and Department of Biodiversity, Phenology Lab, UNESP - São Paulo State University, Biosciences Institute, São Paulo, Rio Claro, Brazil. *Author for correspondence. Montserrat Arista 1 ( [email protected] ) Marcial Escudero 1 * ( [email protected] ) Justen B. Whittall 2 ( [email protected] ) Jose C. Del Valle 1 ( [email protected] ) Melissa León-Osper 2 ( [email protected] ) M. Luisa Buide 3 ( [email protected] ) Maria Gabriela Gutierrez Camargo 4 ( [email protected] ) Leonor Patricia Cerdeira Morellato 4 ( [email protected] ) Nancy Rodríguez-Castañeda 1 ( [email protected] ) Victor Rossi 2 ( [email protected] ) Katie Conrad 2 ( [email protected] ) Josephine Hernandez-Mena 2 ( [email protected] ) Pedro L. Ortiz 1 ( [email protected] ) Eduardo Narbona 3 ( [email protected] ) Running title: Floral pigment evolution in angiosperms Keywords: anthocyanins, aurones, carotenoids, chlorophylls, evolutionary transitions, floral pigments, flower color, phylogenetic association, plants, trait evolution. Type of article: Letter Words in the abstract: 150 Words in the main text: 5524 Words in the text bok: NA Number of references: 90 Number of figures: 6 Number of tables: 0 Number of text boxes: 0 Author for correspondence: Eduardo Narbona; Universidad Pablo de Olavide, Crrt de Utrera S/N, 41013, Seville, Spain; [email protected] ; tfl: +34954 977 546; fax: +34954 349 813 Author contributions MA, EN, ME, JBW, MBR, LPCM and PLO conceived the study and established the methodology. EN, JBW, JCD, MLO, MLB, IP, MGGC, LPCM, NRC, VR, KC, JHM obtained floral pigment data. ME performed phylogenetic inference and correlated evolution between pigments. MA, EN, JBW and PLO analyzed the frequency of pigment in major angiosperm lineages. MA, EN, ME and JBW drafted the manuscript. All authors contributed to the final manuscript and approved the submitted version. Data Availability Statement All data used in this analysis are publicly accessible. The data and code supporting the findings of this study are available on GitHub (https://github.com/amesclir/PigmentEvolution). Abstract The remarkable diversity of floral color arises primarily from three biosynthetically independent pigment groups: anthocyanins, carotenoids and chlorophylls. Because each group aids both pollinator attraction and stress tolerance, we hypothesized that metabolic costs favor minimizing redundancy, leading most species to rely on a single dominant group. Using a dataset of 1,133 animal-pollinated species spanning macroevolution of anthocyanins, carotenoids, chlorophylls and aurones. Pigment evolution followed a stepwise trajectory with a strong trend toward single-group dominance. Anthocyanins exhibited balanced gains and losses, while carotenoids, chlorophylls and aurones were more often lost. All groups showed significant phylogenetic signal, strongest in anthocyanins and carotenoids. Negative phylogenetic associations between carotenoids and both anthocyanins and chlorophylls reveal biosynthetic trade-offs limiting pigment combinations. These results reveal that most angiosperm lineages achieve floral color diversity through the predominance of a single pigment group, with specific pigments concentrated in distinct clades. 1. Introduction With approximately 300,000 extant species, angiosperms represent by far the most diverse group of land plants (Benton et al. 2022). Angiosperms are characterized by the presence of fruits and distinctive angiosperm-type flowers, a suite of key innovations with critical ecological functions and substantial evolutionary consequences for plant reproduction and speciation (Baum and Hileman 2006). Flowers have therefore been conserved throughout angiosperm evolutionary trajectory while also exhibiting enormous morphological diversity (Endress 2010; Benton et al. 2022; López-Martínez et al. 2024). Recent studies have begun to resolve key questions about angiosperm evolution, including the timing of their origin and diversification (Zuntini et al. 2024) and the role of insect pollination during early angiosperm evolution (Peña-Kairath et al. 2023; Stephens et al. 2023; Peris and Condamine 2024). In ancestral angiosperms, olfactory and thermal signals appear to have been predominant cues in the interaction with pollinators (Thien et al. 2009; Peris and Condamine 2024), with flower color being probably a secondary signal to attract pollinators (Rudall 2020). However, the growing significance of visual floral signals may have played a central role in driving rapid angiosperm diversification, and today color serves as a primary cue for pollinator attraction (Thien et al. 2009; van der Kooi and Ollerton 2020; Dorin et al. 2023). Flower color diversity is a key distinguishing feature of angiosperms, in sharp contrast to the relatively monochromatic reproductive structures in non-flower bearing seed plants (i.e. cones of gymnosperms; Rudall 2020; Wiens and Emberts 2024). In most plant species, flower color is primarily determined by pigments (Kay et al. 1981; van der Kooi et al. 2016), and genes involved in pigment biosynthesis are considered targets of pollinator-mediated selection (Bradshaw and Schemnske, 2003; Wessinger and Rausher 2012; Yuan 2019). The striking diversity of floral colors is primarily produced by three major pigment classes: flavonoids, carotenoids and chlorophylls (Davies 2004). Flavonoids comprise three main groups: UV-absorbing flavonoids, aurones–chalcones and anthocyanins (Iwashina 2015; Narbona et al. 2025a). Another class of pigments, betalains, are rare in angiosperms and tend to be concentrated in specific families within the Caryophyllales (Timoneda et al. 2019). Chlorophylls, carotenoids, and flavonoids were likely present in the vegetative tissues of the common ancestor of angiosperms, as these pigments also occur in extant gymnosperms (Rudall 2020; Davies et al. 2022). In these lineages, their primary roles were probably unrelated to pollinator attraction and instead included functions such as photosynthesis, antioxidant activity, and protection against UVB radiation, among others (Davies et al. 2022; 2024). In the extant members of the early branching angiosperm lineages, the ANA-grade, flowers exhibit considerable color variability, including white, green, yellow, pink, purple and red (Endress 2001; Thien et al., 2009). This suggests that all pigment groups, and their biosynthetic machinery, were present in flowers from the origin of angiosperms, and that flowering plants may have co-opted pigments originally present in vegetative tissues or in reproductive structures such as cones for use in floral displays to attract pollinators (Rudall 2020; Kellenberger and Glover 2023; Wiens and Emberts 2024). However, understanding how these major pigment groups have evolved over time remains largely unexplored, primarily because of a lack of comprehensive pigment characterization across the breadth of angiosperm diversity. In a transcontinental study analyzing the major pigment groups present in petals of animal-pollinated species, UV-absorbing flavonoids and other phenylpropanoids (hereafter referred as UAPs) were nearly ubiquitous across angiosperms (Narbona et al. 2025a). These compounds contribute to both pollinator attraction and resistance to environmental stressors (Borgi et al. 2019; Agati et al. 2024; Narbona et al. 2021a, 2025b). However, flowers that exhibit only UAPs (typically white flowers) show low chromatic contrast, reducing their detectability to diverse pollinator groups compared with flowers containing additional pigment classes (Narbona et al. 2021a). Most species accumulated UAPs alongside one or more additional pigments (85% of sampled species)—most commonly anthocyanins (56%), secondarily carotenoids (37%) and more rarely chlorophylls (17%) (Narbona et al. 2025a). Because pigment biosynthesis incurs metabolic costs (Galen 1999; Blanchard and Holeski 2024), the co-occurrence of functionally redundant pigments—those producing similar color phenotypes or serving equivalent ecological roles—may be selectively disfavored (Grotewold 2006; but see León-Osper et al. 2025). Indeed, 61% of species accumulated UAPs with a single additional pigment group, suggesting an evolutionary strategy that economizes pigment investment. In animals, resource trade-offs associated with pigments are well established: a single compound, such as melanin or carotenoids, can simultaneously mediate signaling and physiological functions (e.g. antioxidant activity, thermoregulation; McGraw 2005; Koch and Hill 2018; Ma et al. 2024). In plants, different pigment groups can likewise provide overlapping signaling and protective functions, implying that trade-offs should emerge among pigment groups (Agrawal et al. 2010). Yet these predicted floral pigment dynamics remain untested in a macroevolutionary context. Because floral pigments serve both biotic and abiotic functions (Strauss and Whittall 2006), their evolutionary rates may differ. Essential pigments for physiological or ecological functions might be more conserved, while others evolve more dynamically (Tanaka et al. 2008). Most information about pigment evolution comes from the study of anthocyanins in some specific clades (Perret et al., 2003; Whittall et al., 2006; Sobel and Streisfeld, 2013; Smith and Goldber 2015; Ho and Smith 2016; Ng and Smith 2016a,b; Berardi et al. 2021; Wheeler et al. 2023) as they are the most common flower pigments (Marin-Recinos and Pucker 2024; Narbona et al. 2025a). However, other pigment groups such as carotenoids or aurones‐chalcones (hereafter aurones for simplicity) generate higher color conspicuousness for many groups of pollinators (i.e. hymenopterans, dipterans, lepidopterans and birds; Narbona et al. 2021a). Although various evolutionary processes shape floral pigment composition, it remains unclear whether different pigment groups evolve at different rates, are concentrated in certain lineages and rare in other lineages, or whether positive or negative evolutionary associations exist among pigment groups. In this study, we analyze the evolution of floral pigments across angiosperms using a large dataset of pigment composition and phylogenetic information. We reconstructed the evolution of floral pigmentation—including pigment number, gains and losses of major groups (anthocyanins, carotenoids, chlorophylls and aurones), and evolutionary correlations—across 1,133 animal-pollinated species from three continents, spanning 39 orders and 113 families and representing 61% of recognized angiosperm orders and 27% of families (APG IV, 2016). We hypothesize that pigment loss is more frequent than gain, given the larger mutational target for loss-of-function mutations, which more readily disrupt existing pathways than repair dysfunctional pathways (Whittall et al. 2006; Wessinger and Rausher 2012; Sobel and Streisfeld 2013). We further predict that flowers containing a single pigment group (excluding UAPs) represent the predominant state across angiosperms, reflecting the avoidance of redundancy given the metabolic cost of pigment production (Grotewold 2006; Koch and Hill 2018). Finally, we expect chlorophylls to be relatively common in early-diverging angiosperm lineages and to be progressively replaced by pigments that enhance pollinator attraction, such as carotenoids and anthocyanins (Narbona et al. 2021a; Dorin et al. 2023). 2. Methods 2.1. Flower pigment collection We collected floral pigment data for 1133 animal-pollinated species belonging to 601 genera and 113 families ( Supplementary Dataset 1 ). Most of the data (924 species) were obtained from Narbona et al. (2025a), in which flower pigments were characterized for 436 species from western North America (California Floristic Province), 385 from southern Spain (Mediterranean vegetation), and 103 from southeastern Brazil (Cerrado vegetation). In addition, we obtained new biochemical data for 21 ornamental species growing in Spain. We checked that the floral color of the ornamental species was the same color as the species in its native area using the information available in www.tropicos.org database. Floral pigment composition in ornamental species was obtained using the same method as that applied to the species analyzed in Narbona et al. (2025a), namely a differential extraction method followed by analysis of main distinctive peaks of the absorbance spectra of methanol and acetone extracts. This method allows for the identification of six major groups of pigments (Narbona et al. 2021a, 2025a; León-Osper et al. 2025): UAPs (hydroxycinnamic acids, flavones and flavonols), aurones, anthocyanins, chlorophylls, carotenoids, and betalains. We analyzed the pigment content of the floral structures contributing most prominently to the visual display, usually the petals or tepals (details in Supplementary Information ), We also incorporated data from 188 species obtained via a literature review (see Supplementary Information ). We analyzed the evolutionary history of all pigment groups except UAPs and betalains. UAPs were excluded from the analysis because they were present in all species, except for a single species (Narbona et al. 2025a), preventing meaningful phylogenetic inference of their evolutionary history within angiosperms. Betalains were excluded from the analysis because their synthesis is confined to certain Caryophyllales lineages (Timoneda et al. 2019; Davies et al. 2022), and our dataset included too few betalain-producing species for a robust evolutionary analysis. 2.2. Frequency of pigment groups in major angiosperm lineages We calculated the frequencies of each major pigment group across the principal angiosperm lineages, using the Zuntini et al. (2024) classification as a taxonomic framework. The early-diverging angiosperms group includes species from the ANA-grade (three species of Nymphaeales) and magnoliids (five species each from Laurales, Piperales, and Magnoliales; Supplementary Dataset 1 ). Differences in pigment frequency among phylogenetic groups were assessed using permutation tests ( Supplementary Information ). 2.3. Phylogenic inference We built a phylogenetic tree of all species included in the present study using the function phylo.maker implemented in the r package V.PhyloMaker2 (Jin and Qian, 2022). Species names and the taxonomic framework (genera and families) were set to fit with the ones used by v.phylomaker (World Plant database https://www.worldplants.de) using U.Taxonstand (Zhang and Qian, 2023). We used the following options: tree = GBOTB.extended.WP, nodes = nodes.info.1.WP (the genus- or family-level largest cluster’s root and basal node information was extracted from the mega-tree) and scenarios = ‘S2’ (new species tips are connected to genus- or family-level randomly). The mega-tree GBOTB.extended.WP is a corrected combination of GBOTB for seed plants (Smith and Brown 2018) and Zanne et al. (2014) phylogeny for pteridophytes, and includes 479 families, 10,587 genera and 74,533 species of vascular plants. 2.4. Evolution and phylogenetic signal of pigment and number of pigments For the number of pigments, we used Markov models for multiple state traits. We used generalized Hidden Markov models, as implemented in the function corHMM in the R package corHMM v.2.8 (Boyko and Beaulieu 2021) to estimate the transition rates between trait states (0, 1, 2 and 3 pigments) and reconstruction of number of pigments in the phylogeny. We ran four Markov models: (1) the ’equal rates’ model (1R.ER), which includes a single parameter for all state transitions; (2) the ’symmetric rates’ model (1R.SYM), with six parameters modeling symmetric transition rates among the four possible states; (3) the ’all rates differ’ model (1R.ARD), with twelve parameters allowing asymmetric transition rates among states; and (4) the ’stepwise model’ (1R.STEP), which includes six parameters modeling asymmetric transitions but restricts changes to a single step—either a gain or a loss of one pigment group. For instance, transitions from two pigments to one or three pigments are permitted, but not directly to zero pigments. We applied each of the four models under both single-rate and two-rate category frameworks (in the latter case the number of parameters would be double, as there are two rate categories R1 and R2, and two additional parameters for transitions between R1 and R2: 2R.ER, 2R.SYM, 2R.ARD and 2R.STEP). As noted above, UAPs were excluded from the count of floral pigment groups because they occurred in nearly all species; therefore, only the other four major groups (anthocyanins, carotenoids, aurones, and chlorophylls) were considered in our analyses. For the evolution of individual pigment groups (presence or absence of one of the four major pigments—anthocyanins, carotenoids, aurones, and chlorophylls), we employed generalized Hidden Markov Models using the function corHMM (R package corHMM v.2.8; Boyko and Beaulieu 2021) to estimate transition rates between trait states and reconstruct ancestral states across the phylogeny. We tested ’equal rates’ and ’all rates differ’ Markov models separately with one (R1), two (R1 and R2) and three (R1, R2 and R3) transition rates for absence vs. presence of anthocyanins, carotenoids, chlorophylls and aurones (1R.ER, 1R.ARD, 2R.ER, 2R.ARD, 3R.ER and 3R.ARD). Akaike information criterion (AIC) was used to select the best fitting model, and ancestral states in the phylogeny were inferred marginal likelihoods and stochastic mapping. To estimate phylogenetic signal, we calculated the parameter lambda as implemented in the function fitDiscrete from the R package geiger (Pennell et al. 2014). 2.5. Pagel-like models of correlated evolution between pairs of pigments We tested for correlated evolution between all possible pigment pairings, including anthocyanins vs. carotenoids, anthocyanins vs. chlorophylls, anthocyanins vs. aurones, carotenoids vs. chlorophylls, carotenoids vs. aurones, and chlorophylls vs. aurones. Pairwise comparisons among pigment groups were prioritized because most species exhibited a single pigment. Two-pigment combinations were less common, three-pigment cases were rare, and no species contained all four pigments ( Supplementary Dataset 1 ). We used generalized Hidden Markov models, as implemented in the function corHMM in the R package corHMM v.2.8 (Boyko and Beaulieu 2021). We used models with independent vs. dependent evolution ( Supplementary Information ). 3. Results 3.1. Evolutionary transitions of the number of pigments In addition to the ubiquitous UAPs, most species (62.0%) exhibited only a single visible pigment group in their flowers—among the four considered in this study: anthocyanins, carotenoids, chlorophylls, and aurones—followed by species with two (19.7%), zero (13.6%), or three (4.7%) pigment groups ( Supplementary Dataset 1 ). The model that best explained the evolution of the number of floral pigments was the stepwise model with two transition rates categories (2R.STEP, Table S1 ). Stochastic models indicated that, as expected, most species exhibited a single floral pigment state across the phylogeny. Of the total evolutionary time, 25.8% was assigned to rate category 1 (R1) and 32.1% to rate category 2 (R2) ( Figure 1 ; see Figure S1 for ancestral state reconstruction using marginal likelihood). Much less frequent were species showing two pigments (9.2% in R1 and 11.9% in R2) and zero pigments (10.9% in R1 and 5.8% in R2). Three pigments were extremely rare across the phylogeny (2.8% in R1 and 1.7% in R2). Evolutionary transition rates confirmed this result; in the more frequent R1 category, the transition rates for pigment loss were ~3x and ~5x higher than gain for lineages with two and three pigments, respectively, and ~4x higher for pigment gain in lineages with zero pigments ( Figure 1C , R1 model). Evolutionary transitions in the R2 category were rarer, showing in general low transition rates. In this case, the pattern of transitions was broadly similar, except that the gain of two pigments was more frequent than the loss from two pigments to one (0.001 vs. 1x10 -9 , respectively; Figure 1C , R2 model). In general, angiosperms showed evidence for phylogenetic signal for the number of pigments (lambda 0.71 AIC = 1927.42; both, the Brownian Motion model, AIC = 1999.15, and lambda = 0 model, AIC = 1991.14, were rejected at p < 0.05). Species with a higher number of pigments were more frequently found in early-diverging angiosperms (a group that includes ANA-grade and magnoliid species), monocots (particularly in Iridaceae, Liliaceae and Orchidaceae) and a few scattered lineages of eudicots such as the Euphorbiaceae, Scrophulariaceae and Orobanchaceae ( Figure 1A,B; Figure S1 ; see for more details Figure S2) . Flowers with zero pigments (i.e. white flowers accumulating only UAPs) were concentrated in the Rosaceae in fabids and Apiaceae in campanulids. 3.2. Occurrence of pigment groups in major angiosperm lineages The frequency of each group of pigment differed significantly among angiosperm groups ( Figure 2A, see permutation test for anthocyanins, carotenoids, chlorophylls, and aurones in Figure S3 ). Anthocyanins were more frequent in early-diverging angiosperms and monocots (77.8% and 70.5%, respectively) than in eudicots (from 31.8% of campanulids to 51.7% other superasteris grade), except for lamiids (78.5%). Carotenoids showed marked phylogenetic variation, being common in the other eudicot grade and campanulids (57.7% and 55.9%, respectively) but rare in the other superasterid grade and lamiids (11.8% and 27.4%, respectively). They were especially overrepresented in Asteraceae yet largely absent from Lamiaceae and Boraginaceae ( Figure 1A ). Chlorophylls were 2–4 times more prevalent in flowers of early-diverging angiosperms and monocots (47.1% and 43.3%, respectively) than in eudicot lineages (from 10.7% of campanulids to 23.9% other eudicot grade), except for other superrosids grade (37.9%). Notably, aurones were present exclusively in recently diverging lineages of angiosperms, particularly in malvids, fabids and superasterids, with campanulids exhibiting the highest frequency (12.0%; Figure 2A ). 3.3. Evolution of major pigment groups across angiosperms Evolutionary rates of the main pigment groups varied over time; models using two transition rate categories were selected over models using a single transition rate category or three transition rate categories ( Table S2 ). Anthocyanins showed equal rates of gain and loss (2R.ER), whereas carotenoids, chlorophylls and aurones exhibited asymmetric rates of gain and loss (2R.ARD). The evolution of anthocyanins showed two radically different scenarios: 100 gains and losses per million years (Myr) in the fast R1 category and almost none (1x10 -9 gains and losses per Myr) transitions in the slow R2 category ( Figure 2B ). The transition rates from R1 to R2 and vice versa were moderate. Anthocyanins have been gained and lost a similar number of times in both R1 and R2 categories ( Figure S4). Stochastic mapping indicated that the very slow R2 category was two times more frequent than the fast R1 category across evolutionary time (66% vs 34%; Figure 3A,C ). There was uncertainty about the presence of anthocyanins in ancestral nodes of the phylogeny, but some major lineages such as lamiids (core eudicots) and Asparagales in the monocots showed the presence of anthocyanins as the ancestral state (>0.95; Figure 3B ; see Figure S5 for ancestral reconstruction using marginal likelihood). Carotenoids showed asymmetric rates of fast gain and loss and two different scenarios: more than 100 times higher losses than gains in the fast R1 category and 10 times higher gains than losses in the slow R2 category ( Figure 2B ). The transition rates from R2 to R1 were moderate and faster than from R1 to R2. Carotenoids have been gained and lost a similar number of times in both R1 and R2 ( Figure S4 ). Stochastic mapping indicated that the presence of pigment and slow R1 were two times more frequent than the presence of pigment and the fast R2 across evolutionary time (67% vs 33%; Figure 4A,C ). Thus, carotenoid gains mostly occurred throughout slow R2, for example in highly diverse clades such as subfamily Asteroideae (Asteraceae, Asterids), Brassicaceae (malvids), Ranunculaceae (other eudicots grade), Orobanchaceae and Phrymaceae (lamiids) ( Figure 4A,B ; see Figure S6 for marginal likelihood). Asymmetric transitions rates were also found in chlorophylls, with losses two times higher than gains in moderately fast R1 and 10 -5 times higher in the very slow R2 category ( Figure 2B ). The transition rates from R1 to R2 and from R1 to R2 were moderately slow. Similar to carotenoids, chlorophylls have been gained and lost a similar number of times in the R1 category ( Figure S4 ). Stochastic mapping indicated that the presence of chlorophylls and very low R2 were less frequent than the presence of chlorophylls and the moderately fast R1 across evolutionary time (37% vs 63%; Figure S7A,C ). As noted above, floral chlorophylls were common in early-diverging angiosperms, some monocots (Iridaceae, Orchidaceae, Liliaceae), and two superrosid families (Crassulaceae and Grossulariaceae), where chlorophyll gain mostly occurred in the slow R2 category ( Figure S7A,B ; see Figure S8 for marginal likelihood). However, most of the chlorophyll gains in other eudicots groups occurred in the fast R1 category (except in Euphorbiaceae). Finally, aurones show asymmetric transition rates, with losses at a higher rate than gains in the fast R1 and in the very slow R2 category (4-fold and 37-fold, respectively; Figure 2B ). The transition rates from R1 to R2 were slow and higher than the very slow transition rates from R2 to R1. Again, aurones have been gained and lost a similar number of times, but in this case, losses were slightly higher than gains in R2 ( Figure S4 ). Stochastic mapping indicated that the presence of aurones and fast R1 were much less frequent than the presence of pigment and the very slow R2 across evolutionary time (90% vs 10%; Figure S9A,C ; see Figure S10 for marginal likelihood). Evolutionary transitions were mostly concentrated in Plantaginaceae (lamids) and Asteraceae (campanulids) lineages, occurring in the very slow R2 and fast R1, respectively ( Figure S9B ). Strong and significant phylogenetic signal were found for anthocyanins (estimated lambda 0.92 AIC = 1206.20; Brownian Motion model AIC 1289.03; lambda 0 AIC = 1510.86) and carotenoids (estimated lambda 0.95 AIC = 1139.70; Brownian Motion model AIC = 1197.73; lambda 0 AIC = 1448.94). In contrast, chlorophylls showed a moderately lower phylogenetic signal, representing a reduction of 0.07 and 0.10 compared to anthocyanins and carotenoids, respectively (estimated lambda 0.85 AIC = 846.29; Brownian Motion model AIC = 900.34; lambda 0 AIC = 945.04). Aurones displayed an even lower signal, with a reduction of 0.15 relative to chlorophylls (estimated lambda = 0.70; AIC = 838.27). Although the phylogenetic signal for aurones was still significant, the Brownian motion model could not be rejected (Brownian Motion model AIC = 851.94; lambda = 0 AIC = 867.78). 3.4. Evolutionary relationships between pigment groups Evolutionary transitions in anthocyanins depended on carotenoid state but not the other way around (dependence model with two evolutionary rates categories; Table S3 ). In the R1 category, when carotenoids are present, the rate of anthocyanin loss (0.86 events per Myr) is approximately three times higher than the rate of anthocyanin gain (0.30 events per Myr; Figure 5C ). Conversely, when carotenoids are absent, the rate of anthocyanin gain (100.00 events per Myr) is about three times higher than the rate of loss (38.36 events per Myr). This pattern appears to occur in scattered lineages across the angiosperm phylogeny ( Figure 5A,B ). Furthermore, when both anthocyanins and carotenoids are present, anthocyanins are 8.6 times more likely to be lost (0.86 events per Myr) than carotenoids (0.10 events per Myr). In the R2 category, transitions were generally much slower, and the tendency remained toward anthocyanin loss when carotenoids were present (loss rate = 0.011 vs. gain rate = 0.004 events per Myr). However, when carotenoids were absent, anthocyanins were also lost, with a loss rate of 5×10⁻⁵ compared to a much lower gain rate of 1×10⁻⁹. Chlorophyll evolution also appears to be influenced by the presence of carotenoids ( Table S3 ). In the R1 category, there was a tendency to lose chlorophylls both in the absence (loss rate = 0.014 vs. gain rate = 0.0001 events per Myr) and presence of carotenoids (loss rate = 0.231 vs. gain rate = 0.028 events per Myr), although the loss rate was approximately ten times lower when carotenoids were present ( Fig. 6C ). In the less frequent R2 category, a strong trend toward chlorophyll loss was observed in the absence of carotenoids (loss rate = 34.21 vs. gain rate = 1×10⁻⁹ events per Myr). However, when carotenoids were present, the pattern reversed, with a strong tendency toward chlorophyll gain (loss rate = 2×10⁻⁹ vs. gain rate = 97.40 events per Myr). This gain occurred in several lineages, including Iridaceae, Fabaceae, and Asteraceae ( Figure 6A,B ). Other pigment groups showed distinct evolutionary associations: anthocyanins and chlorophylls followed a dependent-switching model; anthocyanins depended on aurones; and carotenoids–aurones and aurones–chlorophylls fit an all-rates-different model (see Supporting Information ; Table S3; Figure S11 ). 4. Discussion In addition to the near-ubiquity of UAPs, which can serve dual roles in pollinator attraction and environmental protection, our analyses reveal a pronounced evolutionary trend toward the predominance of a single floral pigment group per species. These findings suggest that pairing UAPs with a single visible pigment is generally sufficient to meet the physiological and ecological demands of floral pigmentation (Rusman et al. 2019; Davies et al. 2022; Narbona et al. 2025a) while minimizing metabolic cost. Pigment production entails significant metabolic costs, including the activation of biosynthetic pathways, enzymatic systems, and specialized organelles (Tanaka et al. 2008; Zhao et al. 2022). Assuming that pigment groups compete for limiting resources, plants may optimize resource allocation by selectively synthesizing those pigments that maximize ecological functionality (Agrawal et al. 2010; He et al. 2022). Supporting this interpretation, we observed elevated transition rates from flowers containing three pigment groups—characteristic of early-diverging angiosperms and many monocots—to those with only two pigment groups with heterogeneous rates of evolution across the phylogeny (two-rate categories models), consistent with a resource allocation trade-off (Koch and Hill 2018; Ma et al. 2024). Thus, as suggested at the species level (Briggs and Anderson 2025; Narbona et al. 2025a), our findings confirm that, at a macroevolutionary scale, the production of UAPs combined with a single additional pigment group may be sufficient to fulfill dual functions: attracting pollinators and providing protection against abiotic and biotic stressors (Strauss and Whittall 2006). The evolutionary rates of primary floral pigment groups exhibited variation across clades and time, with models incorporating two transition rate categories providing a better fit than those assuming either a single or three rates. Ancestral state reconstructions revealed no consistent pattern in pigment distribution across major angiosperm lineages. However, chlorophylls were nearly twice as prevalent in the flowers of early-diverging angiosperms and monocots compared to those of the eudicot clade ( Figure 2A ). Notably, chlorophylls appear to represent the ancestral state within the Polemoniaceae (Landis et al. 2018). All pigment groups showed significant phylogenetic signal, strongest for anthocyanins and carotenoids. By contrast, studies analysing floral colour without direct pigment data often report weak phylogenetic signal (e.g. Smith et al. 2008; Gómez et al. 2015; Ortiz et al. 2021). This discrepancy likely arises because (1) a single pigment group can yield a wide range of hues (Grotewold 2006; Tanaka et al. 2008; Narbona et al. 2021a; Davies et al. 2022), and (2) some species combine multiple pigment groups to produce distinct colours (Ng and Smith 2016a; Narbona et al. 2021b; Yeo and Moyroud 2025). Our results suggest that species accumulating each major pigment group are more closely related than expected under a random model. Indeed, the evolution of biochemical pathways responsible for floral pigmentation appears to be strongly influenced by phylogenetic history, as illustrated by anthocyanins and carotenoids observed in the Solanaceae (Ng and Smith, 2016b). Taken together, these findings suggest that the evolutionary dynamics of floral pigment groups are lineage-specific, supporting previous studies that analyzed multiple pigments at the family or lower taxonomic levels (Ng and Smith 2018; Landis et al. 2018). One of our key findings is that anthocyanins represent the most widespread floral pigment group among angiosperms. Anthocyanins have been gained and lost at the same rate throughout angiosperm evolution, reflecting their evolutionary lability (Smith and Goldberg 2015; Wessinger et al. 2023; Marin-Recinos and Pucker 2024). Our findings contrast with both prior empirical studies and theoretical expectations—such as those derived from Dollo’s Law (Gould, 1970)—which generally report higher rates of anthocyanin loss relative to gain (Whittall et al., 2006; Wessinger and Rausher, 2012; Sobel and Streisfeld, 2013; but see Smith and Goldberg 2015 for an opposing trend). This discrepancy is due to differences in the methods used. In our study pigment presence was recorded even when it was confined to very localized floral regions, whereas in other studies, only predominant petal color was considered. Several factors may explain the widespread prevalence of anthocyanins throughout angiosperm evolutionary history. Anthocyanins are synthesized via the general flavonoid pathway and require only a few additional genes (Marin-Recinos and Pucker 2024). Although mutations disrupting the anthocyanin biosynthetic pathway (ABP) are common, they usually affect a few conserved regulatory loci (Sobel and Streisfeld 2013). Regain is therefore likely when structural ABP genes remain intact (often preserved by duplicates) and regulatory elements from other tissues are co-opted to restore floral expression (Wessinger and Rausher 2012; Marin-Recinos and Pucker 2024). The widespread conservation of ABP genes across angiosperms thus suggests that anthocyanin loss is often reversible (Ho and Smith 2016). Moreover, pigmentation gains may be favored by natural selection, as anthocyanins not only attract pollinators but also act as antioxidants that mitigate oxidative stress and enhance tolerance to drought, UV radiation, herbivory, and pathogens (Gould 2004; Davies et al. 2022). While these functions are well-documented in vegetative tissues, increasing evidence suggests they also benefit petals (Johnson et al. 2008; Dalrymple et al. 2020; Dellinger et al. 2025). Finally, anthocyanins produce a broad range of colors readily perceived by key pollinators, including bees, flies, butterflies, and birds (Grotewold 2006; Ng and Smith 2016a; Ogutcen et al. 2020; Narbona et al. 2021a). The predominance of anthocyanins in the angiosperm evolutionary history may be partially explained by their versatility (Davies et al. 2022), allowing flowers to remain attractive across different pollination systems (Narbona et al. 2021a). Carotenoids were the second most prevalent floral pigment group in an evolutionary context ( Figure 2A ; Narbona et al. 2025a). Their evolutionary history was marked by elevated rates of loss; nevertheless, carotenoids remained particularly common in certain lineages—such as Asteraceae, Brassicaceae and Ranunculaceae—where they were especially prevalent under the slow R2 category. Notably, the presence of carotenoids in a lineage negatively affected the presence of anthocyanins, suggesting competitive exclusion between these pigment groups. This pattern was evident across multiple taxonomic levels: high-carotenoid lineages (e.g. the other eudicot grade, campanulids) and families (e.g. Asteraceae, Fabaceae) exhibited reduced prevalence of anthocyanins, or anthocyanins were restricted to particular clades. A similar trade-off has been suggested in Antirrhineae, Cistaceae and Polemoniaceae, in which the presence of both pigments is extremely rare (Guzmán and Vargas 2005; Ellis and Field 2016; Landis et al. 2018). While the underlying causes of this pattern remain unclear, our results allow us to exclude certain alternative hypotheses. Specifically, biosynthesis of carotenoids and anthocyanins are virtually independent as they do not compete for shared metabolic intermediates, nor are they regulated by the same genetic pathways (Stanley et al. 2020; Liang et al. 2023). In addition, although anthocyanins and carotenoids can coexist to produce novel petal colors (e.g., Ojeda et al. 2013; Narbona et al. 2021b), their combined presence does not necessarily confer a fitness advantage (Narbona et al. 2021a; Tenhumberg et al. 2023) and, consequently, may not be favored by natural selection (Sobel et al. 2010; Wenzell et al. 2025). Consistently, microevolutionary patterns observed in polymorphic species bearing both pigments often indicate a trend toward the fixation of a single pigment, as flower color morphs producing both are typically much rarer than those expressing only one (e.g., Whibley et al. 2006; Irwin and Strauss 2005; Berman et al. 2016). An exception to the negative phylogenetic association between carotenoids and anthocyanins occurs in some red-flowered species with the hummingbird pollination syndrome. These species often combine yellow carotenoids with blue to pink anthocyanins to produce red floral coloration (Bradshaw and Schemske 2003; Ng and Smith 2016a,b; León-Osper et al. 2025). Such combinations have evolved repeatedly across the eudicots, with a particularly high incidence among lamiids (e.g., Phrymaceae, Orobanchaceae, Scrophulariaceae; Figure 5 ). Hummingbird pollination is generally derived from bee pollination (Ng and Smith 2018; Wessinger et al. 2019, 2023; but see Stephens et al. 2023; Barreto et al. 2024). In Californian red-flowered species, red coloration results from anthocyanins alone or, more often, from combined anthocyanin–carotenoid accumulation (33% vs 67%; León-Osper et al. 2025). Flowers combining anthocyanins and carotenoids contained only half the anthocyanin concentration of flowers relying solely on anthocyanins, suggesting resource-based constraints on the simultaneous synthesis of both pigment classes (León-Osper et al. 2025). In summary, most sampled angiosperms produce a single pigment group per flower, following a stepwise model of pigment evolution with anthocyanins as the most common pigment . Chlorophylls are relatively common in the flowers of early-diverging angiosperms and monocots. However, in more recently evolved lineages, floral pigmentation has shifted toward the production of either anthocyanins or carotenoids, which provide stronger visual signals to primary pollinator groups than chlorophylls (Narbona et al. 2021a). Most studies reporting a lack of phylogenetic signal have focused on the tips of the phylogeny—typically at the family or genus level—within hyperdiverse lineages where floral color is often linked to relatively evolutionarily recent speciation events ( Smith et al. 2008; Gómez et al. 2015; Ortiz et al. 2021) . In contrast, our macroevolutionary analysis offers a broader perspective, revealing strong phylogenetic signal in nearly all pigment groups and significant negative associations between anthocyanins and both carotenoids and chlorophylls. These patterns have given rise to lineages dominated by anthocyanins (e.g. monocots, lamiids) and others dominated by carotenoids (e.g. campanulids, including many asterids). Angiosperms are a megadiverse clade with a complex evolutionary history (Doyle 2012; Benton et al. 2022; Zuntini et al. 2024), and their flowers have undergone extraordinary diversification in both form and function, playing central roles in reproductive success and speciation (Endress 2010; Sauquet et al. 2017; Stephens et al. 2023; Khojayori et al. 2024; López-Martínez et al. 2024). Despite the complexity of pigment evolution, our analyses reveal clear macroevolutionary trends underpinning the diversification of floral color. These emergent patterns provide a foundation for future research which should combine broader taxonomic sampling with hypothesis-driven approaches. Integrating new insights from pigment biochemistry and the genetic architecture of major pigment pathways (Wheeler et al. 2023; Lin et al. 2024; Marin-Recinos and Pucker 2024 ) will be essential for refining our understanding of the evolutionary dynamics shaping floral color diversity. Acknowledgements We thank Elaine Meslow, Conso Barciela, Pilar Fernandez-Díaz and Julia Fernandez-Boraita for technical support, and the General Herbarium of the Universidad de Sevilla (CITIUS) for its logistical support. This study was supported by the projects PID2020-116222GB-I00 and PID2024-161815NB-I00 funding by the Spanish government MICIU/AEI/ 10.13039/501100011033. We also want to thank grants of the Andalusian Regional Ministry of Economy, Knowledge, Business and University (PREDOC-00336 and PAIDI BIO-305, Spain), the São Paulo Research Foundation (FAPESP, Brazil) (grants #2013/50155-0, #2010/51307-0, #2009/54208-6; #2021/10639-5), the National Council for Scientific and Technological Development (CNPq, Brazil) (grants #400717/2013-1 and 306563/2022-3), and the Coordination for the Improvement of Higher Education Personnel (CAPES, Brazil) (Finance code 1 and CAPES-Print 88887.374156/2019-00). MGGC received CNPq-PDJ (#161293/2015-8) and FAPESP scholarships (#2015/10754-8, #2018/21646-0). JBW received Santa Clara University’s WAVE grant (California, USA), and VR received an undergraduate research scholarship from the Northern California Botanists and TriBeta. JHM received support from Santa Clara University’s LEAD program. ME was supported by DiversiChrom project, funded by AEI-MICINN (PID2021-122715NB-I00). References Agati, G., Brunetti, C., dos Santos Nascimento, L. B., Gori, A., Piccolo, E. L., and Tattini, M. (2024). Antioxidants by nature: An ancient feature at the heart of flavonoids’ multifunctionality. The New Phytologist, 245, 11. Agrawal, A. A., Conner, J. K., and Rasmann, S. (2010). Tradeoffs and negative correlations in evolutionary ecology. In M. Bell, W. Eanes, D. Futuyma, and J. Levinton (Eds.), Evolution Since Darwin: The First 150 Years (pp. 243–268). Sinauer Associates. The Angiosperm Phylogeny Group et al. (2016). An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG IV. Botanical Journal of the Linnean Society, 181, 1–20. Barreto, E., Boehm, M. M., Ogutcen, E., Abrahamczyk, S., Kessler, M., Bascompte, J., Dellinger, A. S., Bello, C., Dehling, D. M., Duchenne, F., and Kaehler, M. (2024). Macroevolution of the plant–hummingbird pollination system. Biological Reviews, 99, 1831–1847. Baum, D. A., and Hileman, L. C. (2006). A developmental genetic model for the origin of the flower. In A. M. Ainsworth (Ed.), Flowering and its Manipulation (Vol. 20, pp. 1–27). Blackwell Publishing. Benton, M. J., Wilf, P., and Sauquet, H. (2022). The Angiosperm Terrestrial Revolution and the origins of modern biodiversity. The New Phytologist, 233, 2017–2035. Berardi, A. E., Esfeld, K., Jäggi, L., Mandel, T., Cannarozzi, G. M., and Kuhlemeier, C. (2021). Complex evolution of novel red floral color in Petunia. The Plant Cell, 33, 2273–2295. Berman, J., Sheng, Y., Gómez Gómez, L., Veiga, T., Ni, X., Farre, G., Capell, T., Guitian, J., Guitian, P., Sandmann, G., Christou, P., and Zhu, C. (2016). Red anthocyanins and yellow carotenoids form the color of orange-flower gentian ( Gentiana lutea L. var . aurantiaca ). PLoS One, 11, e0162410. Blanchard, M., and Holeski, L. M. (2024). Consequences and costs of chemical complexity: The evolutionary ecology of direct phytochemical defense against herbivores. International Journal of Plant Sciences, 185, 3–14. Borghi, M., Perez de Souza, L., Yoshida, T., and Fernie, A. R. (2019). Flowers and climate change: A metabolic perspective. The New Phytologist, 224, 1425–1441. Boyko, J. D., and Beaulieu, J. M. (2021). Generalized hidden Markov models for phylogenetic comparative datasets. Methods in Ecology and Evolution, 12, 468–478. Bradshaw, H. D., Jr., and Schemske, D. W. (2003). Allele substitution at a flower colour locus produces a pollinator shift in monkeyflowers. Nature, 426, 176–178. Day Briggs, S., and Anderson, J. T. (2025). The effect of global change on the expression and evolution of floral traits. Annals of Botany, 135, 9-24. Davies, K. M. ( 2004). An introduction to plant pigments in biology and commerce. In: Plant Pigments and their Manipulation. (ed Davies, K. M.) Annual Plant Review, Blackwell Publishing Ltd, 1–22. Davies, K. M., Andre, C. M., Kulshrestha, S., Zhou, Y., Schwinn, K. E., Albert, N. W., Chagné, D., van Klink, J. W., Landi, M., and Bowman, J. L. (2024). The evolution of flavonoid biosynthesis. Philosophical Transactions of the Royal Society B: Biological Sciences, 379, 20230361. Davies, K. M., Landi, M., van Klink, J. W., Schwinn, K. E., Brummell, D. A., Albert, N. W., Chagné, D., Jibran, R., Kulshrestha, S., Zhou, Y., and Bowman, J. L. (2022). Evolution and function of red pigmentation in land plants. Annals of Botany, 130, 613–636. Dellinger, A. S., Meier, L., Smith, S., and Sinnott‐Armstrong, M. (2024). Does the abiotic environment influence the distribution of flower and fruit colors? American Journal of Botany, e70044. Dorin, A., Shrestha, M., Garcia, J. E., Burd, M., and Dyer, A. G. (2023). Ancient insect vision tuned for flight among rocks and plants underpins natural flower colour diversity. Proceedings of the Royal Society B, 290, 20232018. Ellis, T. J., and Field, D. L. (2016). Repeated gains in yellow and anthocyanin pigmentation in flower colour transitions in the Antirrhineae. Annals of Botany, 117, 1133–1140. Endress, P. K. (2001). The flowers in extant basal angiosperms and inferences on ancestral flowers. International Journal of Plant Sciences, 162, 1111–1140. Endress, P. K. (2010). Flower structure and trends of evolution in eudicots and their major subclades. Annals of the Missouri Botanical Garden, 97, 541–583. Escribano-Bailon, M. T., and Santos-Buelga, C. (2012). Anthocyanin copigmentation: Evaluation, mechanisms and implications for the colour of red wines. Current Organic Chemistry, 16, 715–723. Galen, C. (1999). Why do flowers vary? The functional ecology of variation in flower size and form within natural plant populations. BioScience, 49, 631–640. Gómez, J. M., Perfectti, F., and Lorite, J. (2015). The role of pollinators in floral diversification in a clade of generalist flowers. Evolution, 69, 863-878. Gould, S. J. (1970). Dollo on Dollo’s law: Irreversibility and the status of evolutionary laws. Journal of the History of Biology, 3, 189–212. Gould, K. S. (2004). Nature’s Swiss army knife: The diverse protective roles of anthocyanins in leaves. Journal of Biomedicine and Biotechnology, 5, 314. Grotewold, E. (2006). The genetics and biochemistry of floral pigments. Annual Review of Plant Biology, 57, 761–780. Guzmán, B., and Vargas, P. (2005). Systematics, character evolution, and biogeography of Cistus L. (Cistaceae) based on ITS, trnL-trnF, and matK sequences. Molecular Phylogenetics and Evolution, 37, 644–660. He, Z., Webster, S., and He, S. Y. (2022). Growth–defense trade-offs in plants. Current Biology, 32, R634–R639. Ho, W. W., and Smith, S. D. (2016). Molecular evolution of anthocyanin pigmentation genes following losses of flower color. BMC Evolutionary Biology, 16, 98. Iwashina, T. (2015). Contribution to flower colors of flavonoids including anthocyanins: A review. Natural Products and Communications, 10, 529–544 Irwin, R. E., and Strauss, S. Y. (2005). Flower color microevolution in wild radish: evolutionary response to pollinator-mediated selection. The American Naturalist, 165, 225-237. Jin, Y., and Qian, H. (2022). V. PhyloMaker2: An updated and enlarged R package that can generate very large phylogenies for vascular plants. Plant Diversity, 44, 335–339. Johnson, E. T., Berhow, M. A., and Dowd, P. F. (2008). Colored and white sectors from star-patterned petunia flowers display differential resistance to corn earworm and cabbage looper larvae. Journal of Chemical Ecology, 34, 757–765. Kay, Q. O. N., Daoud, H. S., and Stirton, C. H. (1981). Pigment distribution, light reflection and cell structure in petals. Botanical Journal of the Linnean Society, 83, 57–83. Kellenberger, R. T., and Glover, B. J. (2023). The evolution of flower colour. Current Biology, 33, R484–R488. Khojayori, F., Ponraj, U., Buch, K., Zhao, Y., Herrera-Ubaldo, H., and Glover, B. (2024). Evolution and development of complex floral displays. Development 151, dev203027. Koch, R. E., and Hill, G. E. (2018). Do carotenoid‐based ornaments entail resource trade‐offs? An evaluation of theory and data. Functional Ecology, 32, 1908–1920. LaFountain, A. M., and Yuan, Y. W. (2021). Repressors of anthocyanin biosynthesis. New Phytologist, 231, 933–949. Landis, J. B., Bell, C. D., Hernandez, M., Zenil-Ferguson, R., McCarthy, E. W., Soltis, D. E., and Soltis, P. S. (2018). Evolution of floral traits and impact of reproductive mode on diversification in the phlox family (Polemoniaceae). Molecular Phylogenetics and Evolution, 127, 878–890. León-Osper, M., Rossi, V., Conrad, K., Mena, J. H., Meslow, E., Fuller, A., Narbona, E., and Whittall, J. B. (2025). California red hummingbird flowers: Color convergence across four biochemical categories. Madroño, 72, 68–95. Liang, M., Chen, W., LaFountain, A. M., Liu, Y., Peng, F., Xia, R., Bradshaw, H. D., and Yuan, Y. W. (2023). Taxon-specific, phased siRNAs underlie a speciation locus in monkeyflowers. Science, 379, 576–582. Lin, R. C., Ferreira, B. T., and Yuan, Y. W. (2024). The molecular basis of phenotypic evolution: Beyond the usual suspects. Trends in Genetics, 40, 668–680. López‐Martínez, A. M., Magallón, S., von Balthazar, M., Schönenberger, J., Sauquet, H., and Chartier, M. (2024). Angiosperm flowers reached their highest morphological diversity early in their evolutionary history. The New Phytologist, 241, 1348–1360. Ma, R., Jiang, Y., and Wan, D. (2024). Multiple ornaments: Trade-offs and redundancy in signalling functions in male and female varied tits ( Sittiparus varius ). Biological Journal of the Linnean Society, 143, blad150. Marin-Recinos, M. F., and Pucker, B. (2024). Genetic factors explaining anthocyanin pigmentation differences. BMC Plant Biology, 24, 627. McGraw, K. J. (2005). The antioxidant function of many animal pigments: are there consistent health benefits of sexually selected colourants?. Animal Behaviour, 69, 757–764. Narbona, E., del Valle, J. C., Arista, M., Buide, M. L., and Ortiz, P. L. (2021a). Major flower pigments originate different colour signals to pollinators. Frontiers in Ecology and Evolution, 9, 743850. Narbona, E., del Valle, J. C., and Whittall, J. B. (2021b). Painting the green canvas: how pigments produce flower colours. The Biochemist, 43, 6–12. Narbona, E., del Valle, J. C., Whittall, J. B., León-Osper, M., Buide, M. L., Pulgar, I., Camargo, M. G. G., Morellato, L. P. C., Rodríguez-Castañeda, N., Rossi, V., and Conrad, K. (2025a). Transcontinental patterns in floral pigment abundance among animal-pollinated species. Scientific Reports, 15, 15927. Narbona, E., Perfectti, F., González-Megías, A., et al. (2025b). Heat drastically alters floral color and pigment composition without affecting flower conspicuousness. American Journal of Botany, In Press. Ng, J., and Smith, S. D. (2016a). How to make a red flower: The combinatorial effect of pigments. AoB Plants, 8, plw055. Ng, J., and Smith, S. D. (2016b). Widespread flower color convergence in Solanaceae via alternate biochemical pathways. The New Phytologist, 209, 407–417. Ng, J., and Smith, S. D. (2018). Why are red flowers so rare? Testing the macroevolutionary causes of tippiness. Journal of Evolutionary Biology, 31, 1863–1875. Ogutcen, E., Durand, K., Wolowski, M., Clavijo, L., Graham, C., Glauser, G., and Perret, M. (2020). Chemical basis of floral color signals in Gesneriaceae: The effect of alternative anthocyanin pathways. Frontiers in Plant Science, 11, 604389. Ojeda, D. I., Santos-Guerra, A., Oliva-Tejera, F., Valido, A., Xue, X., Marrero, A., Caujapé-Castells, J., and Cronk, Q. (2013). Bird-pollinated Macaronesian Lotus (Leguminosae) evolved within a group of entomophilous ancestors with post-anthesis flower color change. Perspectives in Plant Ecology, Evolution and Systematics, 15, 193–204. Ortiz, P. L., Fernández‐Díaz, P., Pareja, D., Escudero, M., and Arista, M. (2021). Do visual traits honestly signal floral rewards at community level?. Functional Ecology, 35, 369-383. Peña-Kairath, C., Delclòs, X., Álvarez-Parra, S., Peñalver, E., Engel, M. S., Ollerton, J., and Peris, D. (2023). Insect pollination in deep time. Trends in Ecology and Evolution, 38, 749–759. Pennell, M. W., Eastman, J. M., Slater, G. J., Brown, J. W., Uyeda, J. C., FitzJohn, R. G., … and Harmon, L. J. (2014). geiger v2.0: An expanded suite of methods for fitting macroevolutionary models to phylogenetic trees. Bioinformatics, 30, 2216–2218. Peris, D., and Condamine, F. L. (2024). The angiosperm radiation played a dual role in the diversification of insects and insect pollinators. Nature Communications, 15, 552. Perret, M., Chautems, A., Spichiger, R., Kite, G., and Savolainen, V. (2003). Systematics and evolution of tribe Sinningieae (Gesneriaceae): Evidence from phylogenetic analysis of six plastid DNA regions and nuclear ncpGS. American Journal of Botany, 90, 445–460. R Core Team. (2024). R: A language and environment for statistical computing. R Foundation for Statistical Computing. Rudall, P. J. (2020). Colourful cones: How did flower colour first evolve? Journal of Experimental Botany, 71, 759–767. Rusman, Q., Poelman, E. H., Nowrin, F., Polder, G., and Lucas‐Barbosa, D. (2019). Floral plasticity: Herbivore‐species‐specific‐induced changes in flower traits with contrasting effects on pollinator visitation. Plant, Cell and Environment, 42, 1882–1896. Sauquet, H., Von Balthazar, M., Magallón, S., Doyle, J. A., Endress, P. K., Bailes, E. J., Barroso de Morais, E., Bull-Hereñu, K., Carrive, L., Chartier, M., and Chomicki, G. (2017). The ancestral flower of angiosperms and its early diversification. Nature Communications, 8, 1–10. Smith, S. A., and Brown, J. W. (2018). Constructing a broadly inclusive seed plant phylogeny. American Journal of Botany, 105, 302–314. Smith, S. D., Ané, C., and Baum, D. A. (2008). The role of pollinator shifts in the floral diversification of Iochroma (Solanaceae). Evolution, 62, 793-806. Smith, S. D., and Goldberg, E. E. (2015). Tempo and mode of flower color evolution. American Journal of Botany, 102, 1014–1025. Sobel, J. M., and Streisfeld, M. A. (2013). Flower color as a model system for studies of plant evo-devo. Frontiers in Plant Science, 4, 321. Sobel, J. M., Chen, G. F., Watt, L. R., and Schemske, D. W. (2010). The biology of speciation. Evolution, 64, 295–315. Stanley, L. E., Ding, B., Sun, W., Mou, F., Hill, C., Chen, S., and Yuan, Y. W. (2020). A tetratricopeptide repeat protein regulates carotenoid biosynthesis and chromoplast development in monkeyflowers (Mimulus). The Plant Cell, 32, 1536–1555. Stephens, R. E., Gallagher, R. V., Dun, L., Cornwell, W., and Sauquet, H. (2023). Insect pollination for most of angiosperm evolutionary history. New Phytologist, 240: 880–891 Strauss, S. Y., and Whittall, J. B. (2006). Non-pollinator agents of selection on floral traits. In L. D. Harder and H. S. C. Barret (Eds.), Ecology and evolution of flowers (pp. 120–138). Oxford University Press. Tanaka, Y., Sasaki, N., and Ohmiya, A. (2008). Biosynthesis of plant pigments: Anthocyanins, betalains and carotenoids. Plant Journal, 54, 733–749. Tenhumberg, B., Dellinger, A. S., and Smith, S. D. (2023). Modelling pollinator and nonpollinator selection on flower colour variation. Journal of Ecology, 111, 746–760. Thien, L. B., Bernhardt, P., Devall, M. S., Chen, Z. D., Luo, Y. B., Fan, J. H., Yuan, L. C., and Williams, J. H. (2009). Pollination biology of basal angiosperms (ANITA grade). American Journal of Botany, 96, 166–182. Timoneda, A., Feng, T., Sheehan, H., Walker‐Hale, N., Pucker, B., Lopez‐Nieves, S., Gou, R., and Brockington, S. (2019). The evolution of betalain biosynthesis in Caryophyllales. New Phytologist, 224, 71–85. van der Kooi, C. J., and Ollerton, J. (2020). The origins of flowering plants and pollinators. Science, 368, 1306–1308. Wessinger, C. A. (2024). How the switch to hummingbird pollination has greatly contributed to our understanding of evolutionary processes. The New Phytologist, 241, 59-64. Wessinger, C. A., and Rausher, M. D. (2012). Lessons from flower colour evolution on targets of selection. Journal of Experimental Botany, 63, 5741–5749. Wheeler, L. C., Dunbar-Wallis, A., Schutz, K., and Smith, S. D. (2023). Evolutionary walks through flower colour space driven by gene expression in Petunia and allies (Petunieae). Proceedings of the Royal Society B, 290, 20230275. Whibley, A. C., Langlade, N. B., Andalo, C., Hanna, A. I., Bangham, A., Thébaud, C., and Coen, E. (2006). Evolutionary paths underlying flower color variation in Antirrhinum . Science, 313, 963–966. Whittall, J. B., Voelckel, C., Kliebenstein, D. J., and Hodges, S. A. (2006). Convergence, constraint and the role of gene expression during adaptive radiation: Floral anthocyanins in Aquilegia . Molecular Ecology, 15, 4645–4657. Wiens, J. J., and Emberts, Z. (2024). How life became colourful: Colour vision, aposematism, sexual selection, flowers, and fruits. Biological Reviews, 100, 308-326. Yeo, M. T., and Moyroud, E. (2025). Under the rainbow: novel insights on the mechanisms driving the development and evolution of petal pigmentation. Current Opinion in Plant Biology, 86, 102743. Yuan, Y. W. (2019). Monkeyflowers (Mimulus): New model for plant developmental genetics and evo‐devo. The New Phytologist, 222, 694–700. Zanne, A. E., Tank, D. C., Cornwell, W. K., Eastman, J. M., Smith, S. A., FitzJohn, R. G., … and Beaulieu, J. M. (2014). Three keys to the radiation of angiosperms into freezing environments. Nature, 506, 89–92. Zhang, J., and Qian, H. (2023). U. Taxonstand: An R package for standardizing scientific names of plants and animals. Plant Diversity, 45, 1–5. Zhao, X., Zhang, Y., Long, T., Wang, S., and Yang, J. (2022). Regulation mechanism of plant pigments biosynthesis: Anthocyanins, carotenoids, and betalains. Metabolites, 12, 871. Zuntini, A. R., Carruthers, T., Maurin, O., Bailey, P. C., Leempoel, K., Brewer, G. E., Epitawalage, N., Françoso, E., Gallego-Paramo, B., McGinnie, C., and Negrão, R. (2024). Phylogenomics and the rise of the angiosperms. Nature, 629, 843–850. FIGURE LEGENDS Figure 1. Macroevolutionary patterns displaying the number of pigments per flower across angiosperms from the four-state step model with two rates of evolution. Tree nodes are colored as their probability of the presence zero, one, two or three pigments per flower ( A ), and classification into transition rate categories R1 or R2 ( B ). ( C ) Evolutionary transition rates of the number of pigments per flower in R1 or R2; the thickness of the arrows represents transition rates, with thicker arrows indicating higher rates and dashed arrows indicating extremely low rates (1×10⁻⁹). Ancestral state reconstruction was performed using stochastic character mapping. Outer circles in each tree represent the presence of floral pigment groups using violet for anthocyanins, orange for carotenoids, green for chlorophylls, and yellow for aurones. Grey indicates absence of each pigment group. All the species showed UAPs in flowers (innermost ring in cream color). Figure 2. ( A ) Frequency of floral pigment groups across major lineages of flowering plants. Within each pigment group, asterisks indicate whether the observed frequency of pigment in a lineage differs from the randomized number using 1000 random permutations and two-tailed p-values (see Supplementary Figure 3). The number of species sampled in each major lineage is shown in brackets. Early-diverging angiosperms group includes ANA-grade and magnoliid species. Phylogenetic relationships according to Zuntini et al. (2024). ( B ) Evolutionary transition rates (events per Myr) between presence and absence of floral pigments with two transition rates categories (R1 and R2). The thickness of the arrows represents transition rates, with thicker arrows indicating higher rates and dashed arrows indicating extremely low rates (1×10⁻⁹ events per Myr). Figure 3. Ancestral state reconstruction of anthocyanins using stochastic mapping tree with two equal rates model of evolution. Tree nodes are color-coded to represent the probability of the presence or absence of anthocyanins ( A ), and classification into transition rate categories R1 or R2 ( B ). Percentage of mean total time spent in each state are showed in ( C ). Outer circles in each tree represent the presence of floral pigment groups using violet for anthocyanins, orange for carotenoids, green for chlorophylls, and yellow for aurones. Grey indicates absence of each pigment group. All the species showed UAPs in flowers (innermost ring in cream color). Figure 4. Ancestral state reconstruction of carotenoids using stochastic mapping tree with two equal rates model of evolution. Tree nodes are color-coded to represent the probability of the presence or absence of carotenoids ( A ), and classification into transition rate categories R1 or R2 ( B ). Percentage of mean total time spent in each state are showed in ( C ). Outer circles in each tree represent the presence of floral pigment groups using violet for anthocyanins, orange for carotenoids, green for chlorophylls, and yellow for aurones. Grey indicates absence of each pigment group. All the species showed UAPs in flowers (innermost ring in cream color). Figure 5. Phylogenetic associations between anthocyanins and carotenoids using stochastic mapping tree with dependent model of evolution with two transition rate categories. Tree nodes are color-coded to represent the posterior probability of the presence or absence of anthocyanins and carotenoids ( A ), and classification into transition rate categories R1 or R2 ( B ). Percentage of mean total time spent in each state using Pagel-like models ( C ). Outer circles in each tree represent the presence of floral pigment groups using violet for anthocyanins, orange for carotenoids, green for chlorophylls, and yellow for aurones. Grey indicates absence of each pigment group. All the species showed UAPs in flowers (innermost ring in cream color). Figure 6. Phylogenetic associations between carotenoids and chlorophylls using stochastic mapping tree with dependent model of evolution with two transition rate categories. Tree nodes are color-coded to represent the posterior probability of the presence or absence of anthocyanins and carotenoids ( A ), and classification into transition rate categories R1 or R2 ( B ). Percentage of mean total time spent in each state using Pagel-like models ( C ). Outer circles in each tree represent the presence of floral pigment groups using violet for anthocyanins, orange for carotenoids, green for chlorophylls, and yellow for aurones. Grey indicates absence of each pigment group. All the species showed UAPs in flowers (innermost ring in cream color). Information & Authors Information Version history V1 Version 1 29 October 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords anthocyanins aurones carotenoids chlorophylls evolutionary transitions floral pigments flower color phylogenetic association plants trait evolution Authors Affiliations Montserrat Arista Universidad de Sevilla Facultad de Biologia View all articles by this author Marcial Escudero Universidad de Sevilla Facultad de Biologia View all articles by this author Justen Whittall Santa Clara University Department of Biology View all articles by this author Jose C. Del Valle Universidad de Sevilla Facultad de Biologia View all articles by this author Melissa León-Osper Universidad Pablo de Olavide Departamento de Biologia Molecular e Ingenieria Bioquimica View all articles by this author M. Luisa Buide Universidad Pablo de Olavide Departamento de Biologia Molecular e Ingenieria Bioquimica View all articles by this author Maria Gabriela Gutierrez Camargo Universidade Estadual Paulista Julio de Mesquita Filho - Campus de Rio Claro View all articles by this author Leonor Patricia Cerdeira Morellato Universidade Estadual Paulista Julio de Mesquita Filho - Campus de Rio Claro View all articles by this author Nancy Rodríguez-Castañeda Universidad de Sevilla Facultad de Biologia View all articles by this author Victor Rossi Santa Clara University Department of Biology View all articles by this author Katie Conrad Santa Clara University Department of Biology View all articles by this author Josephine Hernandez-Mena Santa Clara University Department of Biology View all articles by this author Pedro L. Ortiz Universidad de Sevilla Facultad de Biologia View all articles by this author Eduardo Narbona 0000-0003-1790-6821 [email protected] Universidad Pablo de Olavide Departamento de Biologia Molecular e Ingenieria Bioquimica View all articles by this author Metrics & Citations Metrics Article Usage 269 views 174 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Montserrat Arista, Marcial Escudero, Justen Whittall, et al. Macroevolutionary trade-offs among floral pigments shape angiosperm color diversity. Authorea . 29 October 2025. DOI: https://doi.org/10.22541/au.176175426.68257877/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.176175426.68257877/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fed93964d6852ad',t:'MTc3OTMwMzU0NQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();