Do visual traits honestly signal floral rewards at community level?

1 Do visual traits honestly signal floral rewards at community level? 1 2 Ortiz, P.L.1, Fernández-Díaz, P. 1, Pareja, D. 1, Escudero, M. 1 & Arista, M. 1 3 4 Departamento de Biología Vegetal y Ecología, Universidad de Sevilla, Apdo. 41080. 5 Sevilla, Spain 6 7 8 9 Corresponding author. Pedro L. Ortiz. Departamento de Biología Vegetal y Ecología. 10 Universidad de Sevilla. Apdo. 41080. Sevilla, Spain. E-mail: [email protected] 11 Phone number: + 954557054 12 13 14 ACKNOWLEDGEMENTS 15 This work was supported by the European Regional Development Fund (ERDF) and grants 16 from the Spanish MINECO (CGL2015-63827) to M.A. and P.L.O. and (PGC2018- 17 099608-B-I00) to ME and to Programa Iberoamericano de Ciencia y Tecnología para el 18 Desarrollo RED CYTED-SEPODI (417RT0527). The authors thank two anonymous 19 reviewers and Dr C. Van der Kooi for their valuable comments and suggestions on the 20 manuscript. The authors thank Servicios Generales de Herbario de la Universidad de 21 Sevilla (Herb SEV) were the voucher specimens were deposited. 22 23 24 25 AUTHOR’S CONTRIBUTIONS 26 PLO and MA planned and designed the research. All authors collected the samples and 27 PLO, PFD and DP measured floral traits; ME, MA, PFD and PLO analyzed the data. MA, 28 PLO and ME wrote the first versions of the manuscript that was later edited by all authors. 29 30 31 2 DATA AVAILABILITY STATEMENT 32 Data deposited in the Dryad Digital Repository: http://doi.org/10.5061/dryad.pg4f4qrn4, 33 (Ortiz, Fernández-Díaz, Pareja, Escudero, & Arista, 2020). 34 1 Do visual traits honestly signal floral rewards at community level? 1 2 3 Abstract 4 1. The high variability observed in floral traits has been interpreted as resulting from the 5 adaptation of plants to pollinators, as the latter present innate preferences for specific floral 6 traits and impose selection over them. However, some pollinators such as bees can learn to 7 associate floral signals with rewards, thus increasing floral constancy on more rewarding 8 flowers. The integration of all these rewards and cues is markedly important at community 9 level, where co-flowering species compete for pollinators. 10 2. In order to verify the honesty of the above mentioned signals, we examined the association 11 between floral visual signals (size, colour, symmetry and floral display) and rewards (pollen 12 and nectar) for 98 species in a Mediterranean community. The associations between floral 13 traits were analysed considering the phylogenetic relationship between the different species. 14 3. Flower colour, size, pollen volume, or amount of sucrose exhibited no phylogenetic signal, 15 which suggests an adaptive evolution in response to different conditions in the pollinator 16 community. Flower size was seen to constitute the most honest signal for pollinators, as this 17 was significantly associated with quantities of pollen and nectar. In contrast, nectar 18 concentration was observed to be positively associated with chromatic contrast. We detected 19 no relationship between flower shape and rewards, on the one hand, or between flower 20 display and rewards, on the other. 21 4. Our study unequivocally demonstrates the correlation between rewards and the visual 22 signals perceived by bees, the most effective pollinators in the Mediterranean Basin. In the 23 community studied, bees employed flower size at longer distances and chromatic contrast at 24 shorter distances to predict rewards. The limited number of studies existing in this sense 25 indicates that this kind of association appears to be community specific. 26 2 KEY-WORDS: Flower colour, Flower size, Flower symmetry, Mediterranean, Nectar, 27 Plant-pollinator interaction, Phylogenetic signal 28 29 1. INTRODUCTION 30 The process of natural selection has shaped the evolution of zoophilous flowers by matching 31 floral traits with pollinators (Feinsinger, 1983; Stebbins, 1970). The high degree of 32 variability in floral traits, such as colour, scent, size and shape, has been interpreted as the 33 outcome of the selection imposed by pollinators. It has been suggested that unrelated plants 34 adapted to the same functional group of pollinators exhibit floral convergence (traditional 35 pollination syndrome concept, Faegri & van der Pijl, 1979). Nonetheless, the pollinator 36 assemblage visiting a given plant species is often diverse, varying both spatially and 37 temporarily (Herrera, 1996; Waser, Chittka, Price, Williams, & Ollerton, 1996). Regardless 38 of the degree of generalization in a pollination system, plant-pollinator interactions are 39 mostly mediated by the rewards offered by flowers and advertised by means of visual or 40 olfactory signals. 41 Nectar and pollen are the principal rewards offered to pollinators. Nectar plays a key role in 42 plant reproduction by rewarding floral visitors (Simpson & Neff, 1983) and its 43 characteristics, such as concentration, volume, viscosity and chemical composition, can 44 determine plant-pollinator relationships (Baker & Baker, 1983; Heinrich & Raven, 1972; 45 Heyneman, 1983; Kingsolver & Daniel, 1983; Parachnowitsch, Manson, & Sletvold, 2019). 46 Many studies reveal that pollinators (mainly bees) preferentially visit flowers proffering a 47 greater nectar volume or a higher sugar concentration, which they associate with specific 48 floral traits (e.g. Gómez & Perfectti, 2010; Grüter & Ratnieks, 2011). Most experimental 49 studies focus upon nectar because it strongly attracts most pollinators and the nectar 50 parameter is easily manipulated for research. Pollen is an important protein source, providing 51 3 the greater part of the dietary nitrogen required by most bees and many other insects, and by 52 bats (Law, 1992) or birds (Grant, 1996). Pollen is particularly important for bees, as it 53 represents the main protein source for their larvae. Bees show preferences for pollen sources 54 with a higher protein:lipid ratio (Ruedenauer, Spaethe, van der Kooi, & Leonhardt, 2019; 55 Vaudo, Patch, Mortensen, Tooker, & Grozinger, 2016). However, despite the importance of 56 pollen for bees, the main pollinators in the majority of natural communities, few studies 57 focus on pollen as a reward (Muth, Papaj, & Leonard, 2016; Nicholls & Hempel de Ibarra, 58 2014, 2017). 59 Floral designs likely evolved to advertise concealed or access-limited rewards; thus, floral 60 traits such as colour, size or shape would appear to act as signals enabling plants to announce 61 their rewards (Fenster, Cheely, Dudash, & Reynolds, 2006). For a signal to be reliable for 62 pollinators, it must be honest, i.e. it must be correlated with the quantity or quality of the 63 reward (Armbruster, Antonsen, & Pélabon, 2005; Knauer & Schiestl, 2015). In certain 64 circumstances, there can be a disassociation between floral signal and reward; for example, 65 when a visitor has foraged upon a flower, its nectar can be temporally depleted, and the floral 66 signal therefore becomes dishonest for some time. Accordingly, some pollinators such as 67 honeybees and bumblebees deposit short-lived repellent scents to communicate reward 68 depletion to congeners (Stout & Goulson, 2001); however, many others are unable to do this. 69 In addition, pollinated or old flowers usually stop secreting nectar and their signals can 70 become dishonest if the corolla remains intact for some time. Despite these temporary 71 mismatches between floral signals and rewards, plants providing a more abundant reward 72 will generally possess higher nectar standing crops and/or larger amounts of available pollen 73 (Cohen & Shmida, 1993; Jachuła, Wrzesień, Strzałkowska-Abramek, & Denisow, 2018); 74 this would generally render their distinctive signals honest and would enable pollinators to 75 remember them as providing a reward (Knauer & Schiestl, 2015; McLinn & Stephens, 2006). 76 4 Notwithstanding, some plant species produce little or no rewards, and can benefit from 77 pollinator visits by mimicking the signals of the rewarding plants existing in a community 78 (Schiestl & Johnson, 2013; Vereecken & Schiestl, 2008). However, their mimicry can be 79 considered to be adaptive only if this similarity with regard to signals enhances plant fitness 80 (Roy & Widmer, 1999). 81 Pollinators can show both perceptual bias and innate preference for some floral traits, and 82 they tend to visit these flowers more frequently (Chittka & Raine, 2006; Gómez et al., 2008; 83 Grüter, Moore, Firmin, Helanterä, & Ratnieks, 2011; Gumbert, 2000; K. Lunau & Maier, 84 1995; Rohde, Papiorek, & Lunau, 2013; Schiestl & Johnson, 2013). However, some of the 85 principal pollinators can learn and memorize associations between some floral traits and 86 rewards, thus overriding their innate preferences (Chittka & Raine, 2006; Gómez et al., 87 2008; Grüter & Ratnieks, 2011; Gumbert, 2000; Lunau & Maier, 1995; Rohde et al., 2013). 88 Specifically, bees exhibit a high learning capacity, responding rapidly to an artificially 89 increased reward (e.g. Morse, 1980; Nicholls & Hempel de Ibarra, 2014; Schaffer & Schaffer, 90 1979; Thomson, 1988; Waddington et al., 1981) by increasing floral constancy (Grüter et 91 al., 2011). Thus, foraging pollinators use floral traits as predictors of rewards (Chittka, Dyer, 92 Bock, & Dornhaus, 2003; Giurfa, 2004; Makino & Sakai, 2007). In any case, flower choice 93 by pollinators constitutes a complex mechanism possibly involving many factors, both 94 intrinsic, such as perceptual abilities or innate preferences (Lunau & Maier, 1995; Schiestl 95 & Johnson, 2013), and extrinsic ones, like alternative flower resources or competitors 96 (Inouye, 1978; Lázaro, Lundgren, & Totland, 2009; Mitchell, Flanagan, Brown, Waser, & 97 Karron, 2009). 98 In seasonal communities, which present a high degree of flowering overlap, many plants 99 may compete simultaneously for pollinator services (Herrera, 1988; Petanidou, Ellis, 100 Margaris, & Vokou, 1995). In such a situation, rewarding plant species could be expected to 101 5 produce distinctive signals that can be memorized by pollinators, consequently promoting 102 their floral constancy (Schiestl & Johnson, 2013), which would result in enhanced plant 103 fitness due to an increase in intraspecific pollen receipt and export, and a decrease in 104 heterospecific pollen deposition (Feinsinger, 1987; Wright & Schiestl, 2009). At plant 105 community level, pollinators’ choice of flowers may generate differences in reproductive 106 success among plant species, affecting their population dynamics (Levin & Anderson, 1970; 107 Morales & Traveset, 2009). In turn, pollinators that make correct foraging decisions will be 108 more efficient with regard to acquiring nutrients and increasing their foraging efficiency, 109 thus enhancing fitness levels (Burns & Dyer, 2008; Chittka et al., 2003; Dyer et al., 2012). 110 Briefly, many evolutionary processes are involved in adjusting flowers’ reward signalling to 111 pollinators’ foraging choices (Balamurali, Krishna, & Somanathan, 2015; Schiestl & 112 Johnson, 2013), but assessing these factors falls outside the scope of our study. The present 113 paper attempts to verify the existence of phenotypic associations between floral visual 114 signals and rewards (pollen and nectar) at community level in the Mediterranean Basin in 115 order to assess the honesty of the signals. The association between rewards and floral cues 116 mediates pollinator services at community level (Herrera, 2020). However, there has been 117 scant research on this relationship at the aforementioned level (Chittka, Ings, & Raine, 2004; 118 Kantsa et al., 2017; Shrestha, Garcia, Burd, & Dyer, 2020) and the offer both of nectar and 119 of pollen has never been studied. In the Mediterranean Basin, most species flower in the 120 spring, coinciding with maximum richness of insect taxa (Herrera, 1988). We focused on the 121 sensory capacities of bees because i) they are the main pollinators in the Mediterranean Basin 122 (Petanidou & Ellis, 1993; Potts et al., 2006), which is one of the world’s centres of bee 123 speciation (Michener, 2000), ii) they depend on pollen and nectar for survival and 124 reproduction, iii) they depend on pollen more than any other group of pollinators in the 125 Mediterranean (Petanidou & Ellis, 1993) and iiii) there has been abundant research on their 126 6 sensory capacities for floral trait discrimination (e.g. Chittka & Wells, 2004; Garcia et al., 127 2017). We recorded the floral rewards and flower traits of 98 melitophilous species in a 128 Mediterranean community. To avoid the phylogenetic bias in flower traits and reward 129 association resulting from species composition, we tested the degree of association 130 controlling the non-independence of the species using a phylogeny of the community. The 131 possible association between rewards and floral visual signals, on one hand, and the 132 associative learning by bees in a community context, on the other, can play an important role 133 in the evolution of flower variability in the Mediterranean Basin, one of the world’s 134 biodiversity hotspots (Médail & Quezel, 1997). 135 136 2. MATERIAL AND METHODS 137 2.1. Study site 138 The relationship between flower colour and rewards was studied along the year in a species- 139 rich natural Mediterranean community situated in the Natural Park of Doñana (37o 15’ – 37o 140 20’ N and 6o 30’ – 6o 32’ W). The community consists of assorted scrubland and rich 141 grasslands, with scattered-to-dense woodland, mainly containing Quercus suber and Pinus 142 pinea. The shrub layer mainly comprises species of Cistaceae, Lamiaceae, Fabaceae or 143 Rosaceae. The grasslands, much more diverse, mostly consist of species of Liliaceae, 144 Ranunculaceae, Plantaginaceae, Lamiaceae, Brassicaceae, Fabaceae, Caryophyllaceae or 145 Asteraceae. We sampled a total of 98 native species from 33 families, including trees, shrubs 146 and herbs. These species were selected because they are visited by bees, whereas other 147 flowering species of the community, pollinated by ants (De Vega, Arista, Ortiz, Herrera, & 148 Talavera, 2009) or dipterans (Berjano, Ortiz, Arista, & Talavera, 2009), were not sampled. 149 150 2.2. Floral visual signals 151 7 We recorded symmetry, size and colour of flowers for each species studied; we also recorded 152 floral display size for a subset of these species on a scale ranging from low to high (n = 42). 153 We considered two categories of flower symmetry: “bilateral” for flowers with a single plane 154 of symmetry and “radial” for flowers exhibiting more than one plane. To characterize flower 155 size, we considered the diameter for radial flowers, whereas for bilateral flowers we 156 contemplate the longest axis of flower silhouette as perceived by pollinators on approaching 157 the flowers. These measurements were provided by the Flora Vascular de Andalucía 158 Occidental by Valdés et al. (1987b), considering for each species the middle value of the 159 range shown in the latter Flora. For the Asteraceae species, we considered head size, rather 160 than flower diameter, because in the plants of this family, the head constitutes the visual unit. 161 For each species, we measured the colour of three recently opened flowers from different 162 plants. We obtained spectral reflection functions of the adaxial surface of the petals of each 163 flower at wavelengths from 300 to 700 nm by using an Ocean Optics reflectance 164 spectrometer as well as Spectra Suite software. Our spectrometer device incorporated a 165 probe with a deuterium-tungsten halogen light source (200-2,000 nm) and a black metal 166 probe-holder with a 6mm-diameter opening at a 45° angle, in which the probe was always 167 fully inserted. We used a UV-reflecting white standard provided by the same manufacturer 168 to calibrate the spectrometer. We glued each petal sample as flat as possible on the sticky 169 side of a piece of black electrical insulation tape, attaching it to the basal opening of the 170 probe-holder to prevent any external light from interfering. We first used procspec function 171 in the R package PAVO to zero any negative value in the spectral functions obtained (Maia, 172 Gruson, Endler, & White, 2019). We then averaged the spectral reflection functions for each 173 species and for noise removal, we once again employed the procspec function of the R 174 package PAVO to smooth each averaged spectral function (Maia et al., 2019). For each 175 species we selected the wavelength range to be smoothed, as well as the appropriate 176 14 Among the reward parameters, variability in pollen grains produced per flower was 326 accounted for mainly by phylogeny (λ = 0.726). Some families such as Boraginaceae, 327 Rosaceae or Cistaceae presented a high number of pollen grains per flower, while others 328 such as Ericaceae or Convolvulaceae produced a low number. Nectar presence and sugar 329 concentration also showed a strong phylogenetic signal (λ = 0.922 and λ = 0.878, 330 respectively) with families such as Cistaceae producing highly concentrated nectar. In 331 contrast, pollen volume and sucrose mass exhibited no significant phylogenetic signals 332 (Table 2 and Fig. S2). 333 334 3.3. Phenotypic association among floral traits 335 Based on the optimal models (fitted lambda), positive and significant associations were 336 observed between sugar mass and pollen grain number, on one hand, and between sugar 337 mass and pollen volume, on the other. We also detected significant associations between the 338 following pairs of visual signals: brightness contrast and floral display size, brightness 339 contrast and symmetry, bee-hexagon categories and floral display size, spectral purity and 340 symmetry, and green contrast and flower size (Table 3). However, these associations were 341 not significant after Bonferroni correction and the remaining associations were also not 342 significant (Table S2). 343 344 Multivariate phylogenetic regressions showed that flower size was the only visual signal 345 significantly associated with both rewards (Table 4). Flower size presented a strong, positive 346 and significant association with pollen number and volume and also with sugar mass (Fig. 347 3; Table 4); this occurred even after removing Cistus ladanifer, the species with the largest 348 and most pollen-rewarding flowers, as well as the fourth most nectar-rewarding ones (Table 349 S1). We observed high adjusted r2-values for the relationships between flower size and both 350 15 pollen volume and pollen grains per flower (r 2 = 0.655 and 0.388, respectively; Fig. 3). On 351 the contrary, only chromatic contrast was positively and significantly associated with sugar 352 concentration (Fig. 3, Table 4). Similar results were obtained when multivariate 353 phylogenetic regressions included “floral display” (Table S3), as this trait was not associated 354 with any reward parameter. The only difference was seen to be in the multivariate model for 355 sugar concentration, as all the colour variables and flower size proved to be significantly 356 associated with sugar concentration (Table S3). Despite the big differences in pollen volume 357 and sucrose amount among the colour hexagon categories, these were not significant due to 358 the strong variability within each category (Table S2). Similarly, flower shape (radial or 359 bilateral) showed no significant association with any rewards, although we did detect a 360 tendency in the radial flowers to produce higher pollen volumes. 361 362 4- DISCUSSION 363 Despite the abundant literature on plant-pollinator interactions at community level, little is 364 as yet known about the relationship between floral visual signals and rewards. The wide 365 range of plant species in the present study highlights the association between rewards and 366 certain visual signals that are perceived and learnt by pollinators, specifically by bees. 367 368 4.1. Selection and phylogenetic inertia of floral traits 369 In the community studied, some floral traits presented significant phylogenetic signals, 370 suggesting it is unlikely that pollinators have recently exerted strong selection. Among floral 371 visual cues, symmetry and reflectance amplitude exhibited a strong phylogenetic signal. This 372 phylogenetic signal of reflectance amplitude is surprising, especially because no other colour 373 parameter displayed any phylogenetic inertia. Reflectance amplitude is perceived as the 374 intensity of the stimulus; its role in pollinator attraction is still unclear and some authors 375 16 even suggest that it is unimportant to bees (Ng, Garcia, & Dyer, 2018; van der Kooi et al., 376 2019). We encountered no explanation for the phylogenetic inertia detected in this flower 377 colour trait. However, reflectance amplitude depends both upon pigmentary coloration and 378 on petal structure (van der Kooi et al., 2019) and that structure characterizes some clades 379 (e.g. Ranunculus, van der Kooi et al., 2017). Differences between clades might possibly be 380 due to differences in petal structure. 381 In contrast, a low phylogenetic signal was found in sugar mass and pollen volume per flower 382 (the most indicative parameters of reward value), most colour parameters, and flower size. 383 These traits showed a high variability at community level, which might suggest an adaptive 384 evolution in response to different conditions in the pollinator community. This hypothesis is 385 supported by the results of a meta-analysis that demonstrated that the degree of dependence 386 of a plant species upon insects for pollination correlated with its pollen nutritional content, 387 regardless of phylogeny (Ruedenauer et al., 2019). The lack of a phylogenetic signal in 388 flower colour when using spectral reflectance measures has previously been reported in other 389 communities (Arnold et al., 2009; McEwen & Vamosi, 2010; Shrestha et al., 2014; but see 390 Reverté et al., 2016). It has been suggested that flower colour constitutes an evolutionarily 391 labile trait, frequently subjected to natural selection; indeed, colour changes are frequent 392 within flowering plant clades (e.g. Beardsley, Yen, & Olmstead, 2003; McEwen & Vamosi, 393 2010; van der Kooi & Stavenga, 2019) or even within species (e.g. Imbert et al., 2014; Irwin 394 & Strauss, 2005; Narbona et al., 2018; Turelli et al., 2001; Zufall & Rausher, 2003). Colour 395 changes commonly occur in populations as a consequence of mutations in the anthocyanin 396 biosynthetic pathway, and their maintenance is usually attributed to selection mediated by 397 pollinators (e.g. Hopkins & Rausher, 2012; Irwin & Strauss 2005; Narbona et al., 2018). 398 399 4.2. Flower size: The most honest signal for pollinators 400 17 Our extensive sampling of co-occurring species in a Mediterranean community clearly 401 indicates that flower size, rather than flower colour or other visual signals, is the most honest 402 signal to pollinators. Bigger flowers produced a higher pollen volume, pollen grain number 403 and sucrose amount, even after correcting for the phylogenetic dependency of species traits, 404 and following Bonferroni correction. The relationship between flower size and nectar 405 production has frequently been reported (e.g. Herrera, 1985; Petanidou & Smets, 1995, 406 Petanidou et al., 2000), but few studies refer to the association of flower size with pollen 407 grain number or pollen volume (Chouteau, Barabé, & Gibernaut, 2006; Stanton & Preston, 408 1988). Moreover, these associations were relatively strong (adjusted r2=0.655 for pollen 409 volume and r2=0.388 for pollen grain number) and they could therefore be used as honest 410 signals that are easily learnt by bees or other pollinators (Blarer, Keasar, & Shmida, 2002; 411 Grüter & Ratnieks, 2011). 412 Over long distances, flower size is probably the most important visual cue for floral 413 perception (Hempel de Ibarra, Langridge, & Vorobyev, 2015). The association between 414 flower size and rewards suggests that pollinators can reliably assess flower reward quantity 415 from a distance. If, as has been reported (Blarer et al., 2002; Chittka & Raine, 2006; Gómez 416 et al., 2008; Grüter & Ratnieks, 2011; Gumbert, 2000), bees rapidly learn to associate visual 417 signals with the most rewarding flowers, then a higher floral constancy is to be expected on 418 the bigger flowers, which in turn will enhance flower fitness. Indeed, in a previous study in 419 the same community involving the large flowers of Cistus salvifolius, an experimental 420 reduction of flower size caused a sharp drop in both male and female fitness as a 421 consequence of fewer pollinator visits (Arista & Ortiz, 2007). Similarly, a recent study in 422 another Mediterranean community also revealed a positive relationship between pollinator 423 visitation rates and flower size (Herrera, 2020), thus indicating the importance of floral size 424 in pollinator attraction. 425 18 Flower size also showed a significant association with the green-contrast parameter, 426 although only prior to Bonferroni correction. Green-contrast is the contrast between a visual 427 stimulus and its background mediated solely by the green photoreceptor, and many insects 428 use this to detect flowers from large distances (van der Kooi et al., 2019). It has been 429 suggested that plants with small flowers need to compensate for their low level of 430 conspicuousness by increasing this green-contrast in order to better attract bees (van der 431 Kooi et al., 2019). This hypothesis has not been tested previously, and only a study by 432 Hempel de Ibarra & Vorobyev (2009), based on photographed flowers, has provided any 433 evidence to support it. Nonetheless, our research, based on spectral measurements, did not 434 corroborate this theory, even indicating evidence to the contrary, as bigger flowers also 435 displayed a higher degree of green-contrast, thus becoming much more conspicuous, and 436 increasing their visual attractiveness to bees. 437 438 4.3. Flower colour: A trait weakly associated with reward production 439 It has been hypothesised that co-occurring species with overlapping flowering tend to be 440 more divergent in floral colour than might be expected by chance as a way to ensure floral 441 constancy and avoid competition for pollinators (Botes, Johnson, & Cowling, 2008; Caruso, 442 2000; Chittka, Thomson, & Waser, 1999; Dyer & Chittka, 2004; Spaethe et al., 2001; van 443 der Kooi, Pen, Staal, Stavenga, & Elzenga, 2016). The few studies addressing this issue have 444 provided some evidence in this sense (McEwen & Vamosi, 2010; Shrestha et al., 2014). In 445 our community, floral colour showed a high variability, with most species situated at the 446 blue, blue-UV and green-UV sectors of the hexagon, and many of these species were 447 separated by at least 0.2 hexagon units (Euclidean distance; see Fig. 1a), the minimum 448 threshold with regard to enhancing floral constancy by bees (Dyer & Chittka, 2004; Spaethe 449 et al., 2001). Surprisingly, congeneric plants, very similar in flower size, shape, and colour 450 19 to the human eye, were seen by bees to display different colours, thus indicating that these 451 pollinators can discriminate clearly between them. For example, despite the high similarity 452 in visual cues from a human perspective, Halimium calycinum and Halimium halimifolium 453 (Cistaceae) were situated at UV and Green-UV hexagon sectors respectively, and, for 454 instance, Ononis baetica and Ononis pinnata (Fabaceae) were situated at Blue-Green and 455 Blue sectors, respectively. In closely related species, floral differentiation is probably 456 selected as a way to enhance floral constancy, preventing the formation of maladapted 457 hybrids and maintaining species integrity (Levin, 1972); consequently, differences in colour 458 as perceived by bees might have been selected in these similar co-flowering species. Indeed, 459 in the present paper this hypothesis is supported by the generally weak phylogenetic signal 460 found for the colour parameters. 461 Other species in the community studied, albeit different in colour to the human eye, were 462 situated close together in the bee vision model, at below 0.1 hexagon units (Euclidean 463 distance). Although the Chittka model establishes a Euclidean distance of 0.1 hexagon units 464 as the threshold value for colour discrimination by bees (Chittka, 1992; Chittka & Kevan, 465 2005), recent physiological models have reported a higher discrimination capacity by bees, 466 because colour discrimination obeys sigmoidal-shaped functions (Garcia, Shrestha, & Dyer, 467 2018; Garcia et al., 2017). These new models of bee colour discrimination have been 468 supported by both experimental (Garcia et al., 2018) and empirical data (Jiménez-López, 469 Matas, Arista, & Ortiz, 2019). However, the ability of bees to discriminate between flowers 470 presenting very small hexagon distances depends on context (Dyer, 2006; Dyer & Chittka, 471 2004) and it is therefore difficult to ascertain whether bees can discriminate among these 472 species in the community studied. 473 Although flower colour has traditionally been considered to be one of the most important 474 signals in pollinator attraction, in our community it was significantly associated only with 475 20 the nectar-quality parameter. Chromatic contrast was the most important colour parameter 476 associated with nectar concentration, this association accounting for 7.3% of reward 477 variability. Chromatic contrast has previously been proven to play a significant role at the 478 plant community level (van der Kooi et al., 2019) and it has been suggested that plant species 479 may maximize chromatic contrast as a way to distinguish themselves from their co-flowering 480 neighbours (McEwen & Vamosi, 2010; van der Kooi et al., 2016). Given the importance of 481 chromatic contrast for visual detection by bees (Dyer, Whitney, Arnold, Glover, & Chittka, 482 2007), in bee-pollinated species with overlapping flowering periods, an association between 483 chromatic contrast and reward could be a way to promote flower constancy and to prevent 484 heterospecific pollen deposition. Optimal nectar concentration depends on feeding 485 mechanism and is approximately 35% for flowers pollinated by bees (Kim, Gilet, & Bush, 486 2011). Linking chromatic contrast to nectar concentration, bees are thought to make correct 487 decisions when visiting flowers in order to maximize energy intake rates. However, the 488 compound eyes of bees possess a low resolution for small flowers and consequently, bees 489 only appear to perceive the chromatic contrast of individual flowers when they are very close 490 to them (Hempel de Ibarra et al., 2015). Given that many flowers in our community are small 491 (mean size 18mm), the association between chromatic contrast and nectar concentration is 492 likely to be particularly important at close viewing distances. Unfortunately, no pollen 493 quality parameter was studied in our research, but since the nutritional value of pollen has 494 been reported to be selected by pollinators (Ruedenauer et al., 2019), pollen nutritional 495 quality might also be signalled by any floral colour parameter. 496 To our knowledge, the relationship between nectar offer and colour as perceived by bees has 497 been explored only in five communities. In three of these (including the present one), some 498 degree of association was found between both parameters (Giurfa, Núñez, Chittka, & 499 Menzel, 1995; Kantsa et al., 2017), whereas no relationship was detected in the other two 500 21 (Shrestha et al., 2020). We do not know any other study, aside from the present one, which 501 explores this kind of relationships for pollen offer at community level. Due to the fact that 502 few studies address these relationships and because of the existence of certain methodical 503 differences among them, it is difficult to establish the factors driving inter-community 504 differences, but those relationship would appear to be community specific. 505 506 4.4. Shape is negligible with regard to explaining rewards 507 Lastly, flower shape, another important visual signal in pollinator attraction (Gómez, 508 Perfectti, & Camacho, 2006; Gong & Huang, 2009; Møller & Eriksson, 1995; Møller & 509 Sorci, 1998), exhibited no significant association with any rewards in our community when 510 the regression was controlled by phylogeny. In contrast, corolla shape has been found to 511 constitute an honest signal to pollinators in some species at intraspecific level (Gómez & 512 Perfectti, 2010). In our community, we sampled a large amount of both radial and bilateral 513 species from different families, finding that sample size is probably not the cause of this lack 514 of association. The evolution of bilateral flowers is considered to be a key angiosperm 515 innovation, because they restrict the direction of approach and movement of pollinators on 516 flowers, thus increasing accurate pollen placement on stigmas (Neal, Dafni, & Giurfa, 1998; 517 Sargent, 2004). In contrast, radial flowers do not restrict access to rewards and usually 518 receive visits by many insect orders, some of these inefficient pollinators which act as nectar 519 or pollen thieves (Faegri & van der Pijl, 1979; Neal et al., 1998). This has given rise to the 520 theory that bilateral flowers invest less in producing pollen, which is a very expensive reward 521 (Hargreaves, Harder, & Johnson, 2009; Nicholls & Hempel de Ibarra, 2017), and that they 522 provide more abundant nectar, which is cheaper to produce in terms of energy (Heil, 2011; 523 Simpson & Neff, 1983). However, in the Mediterranean Basin, nectar is also expensive to 524 produce due to water limitations (Petanidou & Smets, 1995), and this could limit nectar 525 22 production. 526 527 Concluding remarks 528 In conclusion, in the community studied naïve bees might visit flowers driven by their innate 529 preferences for colour (Raine & Chittka, 2007), but following a learning process, it is 530 expected that their foraging decisions come to depend on floral signals that are honestly 531 associated with rewards, hence increasing constancy on highly rewarding flowers. Our study 532 clearly shows that, at plant community level, flower size honestly signals the quantity of 533 rewards and likely represents the most important cue for bees at greater distances. In contrast, 534 flower colour indicates a nectar quality parameter, i.e. nectar concentration. At shorter 535 distances, bees might be employing chromatic contrast to predict nectar concentration. 536 The relationship between visual signals and rewards may indicate an evolutionary history of 537 floral adaptation to the most effective pollinators in the Mediterranean Basin. This could be 538 applied to chromatic contrast, a parameter perceived specifically by bees. However, the 539 relationship between flower size and reward could be a simple consequence of either a higher 540 resource allocation in the larger flowers (Cohen & Shmida, 1993) or a variation in the mating 541 system, as predominantly selfing species frequently possess small flowers and produce little 542 pollen or nectar (Sicard & Lenhard, 2011). If the mechanical adjustment between flower and 543 visitor relative sizes implies that larger flowers are more successful when visited by larger 544 visitors, then the relationship between flower size and reward might be adaptive (Cohen & 545 Shmida, 1993; Herrera, 1985). Some studies have suggested that big bees, which have a 546 higher energy expenditure than small ones (Heinrich & Raven, 1972), prefer large flowers, 547 which provide greater rewards (Heinrich & Raven, 1972; Herrera, 1985). Indeed, at the 548 species level, however, some studies show that larger flowers are more successfully 549 pollinated by larger visitors due to a matching between bees and flowers (Galen & Newport, 550 23 1987; Stout, 2000); there is a need to demonstrate this relationship at community level. 551 Whatever may be the case, if the existence of honest associations in the community studied 552 implies an increase in floral constancy by bees, then a lower heterospecific pollen deposition 553 and higher fitness levels can be expected in bigger flowers and in flowers presenting a higher 554 chromatic contrast. However, there is a need to investigate the influence of these associations 555 in maintaining floral constancy and plant fitness at community level. Lastly, our study has 556 only focused upon visual signals, but olfactory signals can also play an important role in 557 plant-pollinator interactions. The integration of flower colour and scent as perceived by bees 558 found in another Mediterranean community (Kantsa et al., 2017), and the association 559 between visual signals and rewards found in the present study, highlight the existence of the 560 floral integration that can be perceived and learnt by bees, and this integration can therefore 561 be subjected to selection. 562 563 30 Harmon, L. J., Weir, J. T., Brock, C. D., Glor, R. E., & Challenger, W. (2008). GEIGER: 714 Investigating evolutionary radiations. Bioinformatics, 24(1), 129–131. doi: 715 10.1093/bioinformatics/btm538 716 Heil, M. (2011). Nectar: Generation, regulation and ecological functions. Trends in Plant 717 Science, 16(4), 191–200. doi: 10.1016/j.tplants.2011.01.003 718 Heinrich, B., & Raven, P. H. (1972). Energetics and pollination ecology. Science, 719 176(991135), 597–602. doi: DOI: 10.1126/science.176.4035.597 720 Hempel de Ibarra, N., Langridge, K. V., & Vorobyev, M. (2015). More than colour 721 attraction: behavioural functions of flower patterns. Current Opinion in Insect 722 Science, 12(Figure 1), 64–70. doi: 10.1016/j.cois.2015.09.005 723 Hempel de Ibarra, N., & Vorobyev, M. (2009). Flower patterns are adapted for detection 724 by bees. Journal of Comparative Physiology A, 195(3), 319–323. doi: 725 10.1007/s00359-009-0412-0 726 Herrera, C. M. (1996). Floral traits and plant adaptation to insect pollinators: a devil’s 727 advocate approach. In Floral Biology. Boston, MA: Springer. doi: 10.1007/978-1- 728 4613-1165-2_3 729 Herrera, C. M. (2020). Flower traits, habitat, and phylogeny as predictors of pollinator 730 service: a plant community perspective. Ecological Monographs, 90(2), 1–27. doi: 731 10.1002/ecm.1402 732 Herrera, J. (1985). Nectar secretion patterns in Southern Spanish Mediterranean 733 scrublands. Israel Journal of Botany, 34(1), 47–58. doi: 734 10.1080/0021213X.1985.10677008 735 Herrera, J. (1988). Pollination Relationships in Southern Spanish Mediterranean 736 Shrublands. Journal of Ecology, 76(1), 274–287. doi: DOI: 10.2307/2260469 737 Heyneman, A. J. (1983). Optimal sugar concentrations of floral nectars-dependence on 738 31 sugar intake efficiency and foraging costs. Oecologia, 60(2), 198–213. doi: 739 10.1007/BF00379522 740 Hopkins, R., & Rausher, M. D. (2012). Pollinator-mediated selection on flower color allele 741 drives reinforcement. Science, 335(6072), 1090–1092. doi: 10.1126/science.1215198 742 Imbert, E., Wang, H., Anderson, B., Hervouet, B., Talavera, M., & Schatz, B. (2014). 743 Reproductive biology and colour polymorphism in the food-deceptive Iris lutescens 744 (Iridaceae). Acta Botanica Gallica, 161(2), 117–127. doi: 745 10.1080/12538078.2014.895419 746 Inouye, D. W. (1978). Resource partitioning in bumblebees: experimental studies of 747 foraging behavior. Ecology, 59(4), 672–678. doi: 10.2307/1938769 748 Irwin, R. E., & Strauss, S. Y. (2005). Flower color microevolution in wild radish: 749 Evolutionary response to pollinator-mediated selection. The American Naturalist, 750 165(2), 225–237. doi: 10.1086/426714 751 Irwin, R. E., Strauss, S. Y., Storz, S., Emerson, A., & Guibert, G. (2003). The role of 752 herbivores in the maintenance of a flower color polymorphism in wild radish. 753 Ecology, 84(7), 1733–1743. doi: 10.1890/0012- 754 9658(2003)084[1733:TROHIT]2.0.CO;2 755 Jachuła, J., Wrzesień, M., Strzałkowska-Abramek, M., & Denisow, B. (2018). The impact 756 of spatio-temporal changes in flora attributes and pollen availability on insect visitors 757 in Lamiaceae species. Acta Botanica Croatica, 77(2), 161–171. doi: 10.2478/botcro- 758 2018-0018 759 Jiménez-López, F. J., Matas, L., Arista, M., & Ortiz, P. L. (2019). Flower colour 760 segregation and flower discrimination under the bee vision model in the polymorphic 761 Lysimachia arvensis. Plant Biosystems, 1–11. doi: 10.1080/11263504.2019.1651776 762 Jin, Y., & Qian, H. (2019). V.PhyloMaker: an R package that can generate very large 763 32 phylogenies for vascular plants. Ecography, 42(8), 1353–1359. doi: 764 10.1111/ecog.04434 765 Kantsa, A., Raguso, R. A., Dyer, A. G., Sgardelis, S. P., Olesen, J. M., & Petanidou, T. 766 (2017). Community-wide integration of floral colour and scent in a Mediterranean 767 scrubland. Nature Ecology and Evolution, 1(10), 1502–1510. doi: 10.1038/s41559- 768 017-0298-0 769 Kim, W., Gilet, T., & Bush, J. W. M. (2011). Optimal concentrations in nectar feeding. 770 Proceedings of the National Academy of Sciences, 108(40), 16618–16621. doi: 771 10.1073/pnas.1108642108 772 Kingsolver, J., & Daniel, T. (1983). Mechanical determinants of nectar feeding strategy in 773 hummingbirds: energetics, tongue morphology, and licking behavior. Oecologia, 774 60(2), 214–226. doi: https://doi.org/10.1007/BF00379523 775 Knauer, A. C., & Schiestl, F. P. (2015). Bees use honest floral signals as indicators of 776 reward when visiting flowers. Ecology Letters, 18(2), 135–143. doi: 777 10.1111/ele.12386 778 Koethe, S., Bossems, J., Dyer, A. G., & Lunau, K. (2016). Colour is more than hue: 779 preferences for compiled colour traits in the stingless bees Melipona mondury and M. 780 quadrifasciata. Journal of Comparative Physiology A, 202(9–10), 615–627. doi: 781 10.1007/s00359-016-1115-y 782 Law, B. S. (1992). Physiological Factors Affecting Pollen Use by Queensland Blossom 783 Bats (Syconycteris australis). Functional Ecology, 6(3), 257. doi: 10.2307/2389515 784 Lázaro, A., Lundgren, R., & Totland, Ø. (2009). Co-flowering neighbors influence the 785 diversity and identity of pollinator groups visiting plant species. Oikos, 118(5), 691– 786 702. doi: 10.1111/j.1600-0706.2008.17168.x 787 Levin, D. A. (1972). The adaptedness of corolla-colour and outline in experimental and 788 33 natural populations of Phlox drummondii. The American Naturalist, 106, 57–70. doi: 789 https://doi.org/10.1086/282751 790 Levin, D. A., & Anderson, W. W. (1970). Competition for pollinators between 791 simultaneously flowering species. The American Naturalist, 104, 455–467. 792 Lunau, K., Konzmann, S., Winter, L., Kamphausen, V., & Ren, Z. X. (2017). Pollen and 793 stamen mimicry: the alpine flora as a case study. Arthropod-Plant Interactions, 11(3), 794 427–447. doi: 10.1007/s11829-017-9525-5 795 Lunau, K., & Maier, E. J. (1995). Innate colour preferences of flower visitors. Journal of 796 Comparative Physiology A, 177(1), 1–19. doi: 10.1007/BF00243394 797 Maia, R., Gruson, H., Endler, J. A., & White, T. E. (2019). pavo 2: New tools for the 798 spectral and spatial analysis of colour in R. Methods in Ecology and Evolution, 10(7), 799 1097–1107. doi: 10.1111/2041-210X.13174 800 Makino, T. T., & Sakai, S. (2007). Experience changes pollinator responses to floral 801 display size: from size-based foraging. Ecology, 21, 854–863. 802 McEwen, J. R., & Vamosi, J. C. (2010). Floral colour versus phylogeny in structuring 803 subalpine flowering communities. Proceedings of the Royal Society B, 277(1696), 804 2957–2965. doi: 10.1098/rspb.2010.0501 805 McLinn, C. M., & Stephens, D. W. (2006). What makes information valuable: signal 806 reliability and environmental uncertainty. Animal Behaviour, 71, 1119–1129. 807 Médail, F., & Quezel, P. (1997). Hot-spots analysis for conservation of plant biodiversity 808 in the Mediterranean Basin. Annals of the Missouri Botanical Garden, 84(1), 112– 809 127. doi: DOI: 10.2307/2399957 810 Michener, C. D. (2000). The bees of the world (Vol. 1). JHU press. 811 Mitchell, R. J., Flanagan, R. J., Brown, B. J., Waser, N. M., & Karron, J. D. (2009). New 812 frontiers in competition for pollination. Annals of Botany, 103(9), 1403–1413. doi: 813 34 10.1093/aob/mcp062 814 Møller, A. M., & Eriksson, M. (1995). Pollinator preference for symmetrical flowers and 815 sexual selection in plants. Oikos, 73(1), 15–22. doi: 10.2307/3545720 816 Møller, A. P., & Sorci, G. (1998). Insect preference for symmetrical artificial flowers. 817 Oecologia, (114), 37–42. doi: https://doi.org/10.1007/s004420050417 818 Morales, C. L., & Traveset, A. (2009). A meta-analysis of impacts of alien vs. native plants 819 on pollinator visitation and reproductive success of co-flowering native plants. 820 Ecology Letters, 12(7), 716–728. doi: 10.1111/j.1461-0248.2009.01319.x 821 Moran, M. D. (2003). Arguments for rejecting the sequential bonferroni in ecological 822 studies. Oikos, 100(2), 403–405. doi: 10.1034/j.1600-0706.2003.12010.x 823 Morse, D. H. (1980). The effect of nectar abundance on foraging patterns of bumble bees. 824 Ecological Entomology, 5(1), 53–59. doi: 10.1111/j.1365-2311.1980.tb01123.x 825 Müller, A., Diener, S., Schnyder, S., Stutz, K., Sedivy, C., & Dorn, S. (2006). Quantitative 826 pollen requirements of solitary bees: Implications for bee conservation and the 827 evolution of bee-flower relationships. Biological Conservation, 130(4), 604–615. doi: 828 10.1016/j.biocon.2006.01.023 829 Muth, F., Papaj, D. R., & Leonard, A. S. (2016). Bees remember flowers for more than one 830 reason: Pollen mediates associative learning. Animal Behaviour, 111, 93–100. doi: 831 10.1016/j.anbehav.2015.09.029 832 Narbona, E., Wang, H., Ortiz, P. L., Arista, M., & Imbert, E. (2018). Flower colour 833 polymorphism in the Mediterranean Basin: occurrence, maintenance and implications 834 for speciation. Plant Biology, 20(1), 8–20. doi: https://doi.org/10.1111/plb.12575 835 Neal, P. R., Dafni, A., & Giurfa, M. (1998). Floral symmetry and its role in plant- 836 pollinator systems: terminology, distribution, and hypotheses. Annual Review of 837 Ecology and Systematics, 29, 345–373. doi: 838 35 https://doi.org/10.1146/annurev.ecolsys.29.1.345 839 Ng, L., Garcia, J. E., & Dyer, A. G. (2018). Why colour is complex: Evidence that bees 840 perceive neither brightness nor green contrast in colour signal processing. Facets, 841 3(1), 800–817. doi: 10.1139/facets-2017-0116 842 Nicholls, E., & Hempel de Ibarra, N. (2014). Bees associate colour cues with differences in 843 pollen rewards. Journal of Experimental Biology, 217(15), 2783–2788. doi: 844 10.1242/jeb.106120 845 Nicholls, E., & Hempel de Ibarra, N. (2017). Assessment of pollen rewards by foraging 846 bees. Functional Ecology, 31(1), 76–87. doi: 10.1111/1365-2435.12778 847 Orme, D., Freckleton, R., Thomas, G., Petzoldt, T., Fritz, S., Isaac, N., & Pearse, W. 848 (2013). The caper package : comparative analysis of phylogenetics and evolution in 849 R. R Package Version 0.5, 2, pp. 1–36. doi: 1 850 Ortiz, P. L. (1991). Melitopalinología en Andalucía Occidental. Tesis Doctoral en 851 Microfichas. Sevilla: Universidad de Sevilla. 852 Ortiz, P.L., Fernández-Díaz, P., Pareja, D., Escudero, M., & Arista, M. (2020). Data from: 853 Do visual traits honestly signal floral rewards at community level? Dryad Digital 854 Repository. http://doi.org/10.5061/dryad.pg4f4qrn4 855 Pagel, M. (1999). Inferring the historical patterns of biological evolution. Nature, 401, 856 877–884. 857 Parachnowitsch, A. L., Manson, J. S., & Sletvold, N. (2019). Evolutionary ecology of 858 nectar. Annals of Botany, 123, 247-261. 859 Petanidou, T., Ellis, W. N., Margaris, N. S., & Vokou, D. (1995). Constraints on flowering 860 phenology in a phryganic (East Mediterranean shrub) community. American Journal 861 of Botany, 82, 607–620. 862 Petanidou, T., Goethals, V., & Smets, E. (2000). Nectary structure of Labiatae in relation 863 36 to their nectar secretion and characteristics in a Mediterranean shrub community - 864 Does flowering time matter? Plant Systematics and Evolution, 225(1–4), 103–118. 865 doi: 10.1007/BF00985461 866 Petanidou, T., & Smets, E. (1995). The potential of marginal lands for bees and apiculture: 867 nectar secretion in Mediterranean shrublands. Apidologie, 26, 39–52. doi: 868 https://doi.org/10.1051/apido:19950106 869 Petanidou, T., & Ellis, W. N. (1993). Pollinating fauna of a phryganic ecosystem: 870 composition and diversity. Biodiversity Letters, 1(1), 9–22. doi: 10.2307/2999643 871 Potts, S. G., Petanidou, T., Roberts, S., O’Toole, C., Hulbert, A., & Willmer, P. (2006). 872 Plant-pollinator biodiversity and pollination services in a complex Mediterranean 873 landscape. Biological Conservation, 129(4), 519–529. doi: 874 10.1016/j.biocon.2005.11.019 875 Potts, S. G., Vulliamy, B., Dafni, A., Ne’eman, G., O’Toole, C., Roberts, S., & Willmer, P. 876 (2003). Response of plant-pollinator communities to fire: Changes in diversity, 877 abundance and floral reward structure. Oikos, 101(1), 103–112. doi: 10.1034/j.1600- 878 0706.2003.12186.x 879 R, C. T. (2014). R: A language and environment for statistical computing. Vienna, Austria: 880 R Foundation for Statistical Computing. Retrieved from http://www.r-project.org/ 881 Raine, N. E., & Chittka, L. (2007). The adaptive significance of sensory bias in a foraging 882 context: Floral colour preferences in the bumblebee Bombus terrestris. PLoS ONE, 883 2(6), 1–8. doi: 10.1371/journal.pone.0000556 884 Revell, L. J. (2012). Phytools: An R package for phylogenetic comparative biology (and 885 other things). Methods in Ecology and Evolution, 3(2), 217–223. doi: 10.1111/j.2041- 886 210X.2011.00169.x 887 Reverté, S., Retana, J., Gómez, J. M., & Bosch, J. (2016). Pollinators show flower colour 888 37 preferences but flowers with similar colours do not attract similar pollinators. Annals 889 of Botany, 118(2), 249–257. doi: 10.1093/aob/mcw103 890 Rohde, K., Papiorek, S., & Lunau, K. (2013). Bumblebees (Bombus terrestris) and 891 honeybees (Apis mellifera) prefer similar colours of higher spectral purity over 892 trained colours. Journal of Comparative Physiology A, 199(3), 197–210. doi: 893 10.1007/s00359-012-0783-5 894 Roy, B. A., & Widmer, A. (1999). Floral mimicry: a fascinating yet poorly understood 895 phenomenon. Trends in Plant Science, 4, 325–330. 896 Ruedenauer, F. A., Spaethe, J., van der Kooi, C. J., & Leonhardt, S. D. (2019). Pollinator 897 or pedigree: which factors determine the evolution of pollen nutrients? Oecologia, 898 191, 349–358. 899 Sargent, R. D. (2004). Floral symmetry affects speciation rates in angiosperms. 900 Proceedings of the Royal Society B, 271(1539), 603–608. doi: 901 10.1098/rspb.2003.2644 902 Schaffer, W. M., & Schaffer, M. V. (1979). The adaptive significance of variations in 903 reproductive habit in the Agavaceae II: Pollinator foraging behavior and selection for 904 increased reproductive expenditure. Ecology, 60(5), 1051–1069. doi: 905 https://doi.org/10.2307/1936872 906 Schiestl, F. P., & Johnson, S. D. (2013). Pollinator-mediated evolution of floral signals. 907 Trends in Ecology and Evolution, 28(5), 307–315. doi: 10.1016/j.tree.2013.01.019 908 Shrestha, M., Dyer, A. G., Bhattarai, P., & Burd, M. (2014). Flower colour and phylogeny 909 along an altitudinal gradient in the Himalayas of Nepal. Journal of Ecology, 102(1), 910 126–135. doi: 10.1111/1365-2745.12185 911 Shrestha, M., Garcia, J. E., Burd, M., & Dyer, A. G. (2020). Australian native flower 912 colours: Does nectar reward drive bee pollinator flower preferences? PLoS ONE, 913 38 15(6), 11–15. doi: 10.1371/journal.pone.0226469 914 Sicard, A., & Lenhard, M. (2011). The selfing syndrome: A model for studying the genetic 915 and evolutionary basis of morphological adaptation in plants. Annals of Botany, 916 107(9), 1433–1443. doi: 10.1093/aob/mcr023 917 Simpson, B. B., & Neff, J. L. (1983). Evolution and diversity of floral rewards. In C. E. 918 Jones & R. J. Little (Eds.), Handbook of Experimental Pollination Biology (pp. 142– 919 159). New York: Van Nostrand Reinold. 920 Smith, S. A., & Brown, J. W. (2018). Constructing a broadly inclusive seed plant 921 phylogeny. American Journal of Botany, 105(3), 302–314. doi: 10.1002/ajb2.1019 922 Spaethe, J., Tautz, J., & Chittka, L. (2001). Visual constraints in foraging bumblebees: 923 Flower size and color affect search time and flight behavior. Proceedings of the 924 National Academy of Sciences, 98(7), 3898–3903. doi: 10.1073/pnas.071053098 925 Stanton, M. L., & Preston, R. E. (1988). Ecological Consequences and Phenotypic 926 Correlates of Petal Size Variation in Wild Radish, Raphanus Sativus (Brassicaceae). 927 American Journal of Botany, 75(4), 528–539. doi: 10.1002/j.1537- 928 2197.1988.tb13471.x 929 Stebbins, G. L. (1970). Adaptive Radiation of Reproductive Characteristics in 930 Angiosperms, I: Pollination Mechanisms. Annual Review of Ecology and Systematics, 931 1(1), 307–326. doi: 10.1146/annurev.es.01.110170.001515 932 Stout, J. C. (2000). Does size matter? Bumblebee behaviour and the pollination of Cytisus 933 scoparius L. (Fabaceae). Apidologie, 31(1), 129–139. doi: 10.1051/apido:2000111 934 Stout, J. C., & Goulson, D. (2001). The use of conspecific and interspecific scent marks by 935 foraging bumblebees and honeybees. Animal Behaviour, 62(1), 183–189. doi: 936 10.1006/anbe.2001.1729 937 Talavera, S., Herrera, J., Arroyo, J., Devesa, J. A., & Ortiz, P. L. (1988). Estudio de la 938 39 flora apícola de Andalucia Occidental. Lagascalia, 15(1), 567–592. 939 Thomson, J. D. (1988). Effects of variation in inflorescence size and floral rewards on the 940 visitation rates of traplining pollinators of Aralia hispida. Evolutionary Ecology, 2(1), 941 65–76. doi: 10.1007/BF02071589 942 Turelli, M., Schemske, D. W., & Bierzychudek, P. (2001). Stable two-allele 943 polymorphisms maintained by fluctuating fitnesses and seed banks: Protecting the 944 blues in Linanthus parryae. Evolution, 55(7), 1283–1298. doi: 10.1111/j.0014- 945 3820.2001.tb00651.x 946 Valdés, B., Díez, M. J., & Fernández, I. (1987). Atlas polínico de Andalucía Occidental. 947 IDR. Universidad de Sevilla. 948 Valdés, B., Talavera, S., & Fernández-Galiano, E. (1987). Flora Vascular de Andalucía 949 Occidental. Barcelona: Ketres Editora. 950 Valido, A., Schaefer, H. M., & Jordano, P. (2011). Colour, design and reward: Phenotypic 951 integration of fleshy fruit displays. Journal of Evolutionary Biology, 24(4), 751–760. 952 doi: 10.1111/j.1420-9101.2010.02206.x 953 van der Kooi, C. J., Dyer, A. G., Kevan, P. G., & Lunau, K. (2019). Functional 954 significance of the optical properties of flowers for visual signalling. Annals of 955 Botany, 123(2), 263–276. doi: 10.1093/aob/mcy119 956 van der Kooi, C. J., Elzenga, J. T. M., Dijksterhuis, J., & Stavenga, D. G. (2017). 957 Functional optics of glossy buttercup flowers. Journal of the Royal Society Interface, 958 14(127), 20160933. doi: 10.1098/rsif.2016.0933 959 van der Kooi, C. J., Pen, I., Staal, M., Stavenga, D. G., & Elzenga, J. T. M. (2016). 960 Competition for pollinators and intra-communal spectral dissimilarity of flowers. 961 Plant Biology, 18(1), 56–62. doi: 10.1111/plb.12328 962 van der Kooi, C. J., & Stavenga, D. G. (2019). Vividly coloured poppy flowers due to 963 46 TABLE 4. Multivariate phylogenetic generalized least squares model between floral traits 1034 and rewards of 98 species pollinated by bees in a Mediterranean community. 1035 1036 Sugar mass: Adjusted r 2= 0.2004, F4,85= 5.326 , P= 0.0007 1037 Trait Estimate Std. Error t value Pr(>|t|) (Intercept) -25.7365 238.1267 -0.1081 0.9142 Flower size 18.4661 4.0693 4.5379 1.854e-05 *** Green contrast -43.2572 628.1678 -0.0689 0.9453 Brightness contrast -56.2499 269.3752 -0.2088 0.8351 Chromatic contrast -441.9439 692.9066 -0.6378 0.5253 1038 Sugar concentration: Adjusted r 2= 0.2351, F4,40 = 3.073, P= 0.026 1039 Trait Estimate Std. Error t value Pr(>|t|) (Intercept) 4.885998 10.511439 0.4648 0.644576 Flower size -0.105261 0.082593 -1.2745 0.209858 Green contrast 11.308046 13.723844 0.8240 0.414842 Brightness contrast 12.151439 8.692147 1.3980 0.169824 Chromatic contrast 67.951162 24.311873 2.7950 0.007933 ** 1040 Pollen grains/flower: Adjusted r 2= 0.4159, F4,69 = 12.26, P= 1.35e-07 1041 Trait Estimate Std. Error t value Pr(>|t|) (Intercept) 26172.49 103701.31 0.2524 0.8015 Flower size 6604.67 980.29 6.7375 3.965e-09 *** Green contrast 50826.09 163847.54 0.3102 0.7573 Brightness contrast -65169.37 77663.35 -0.8391 0.4043 Chromatic contrast 76443.62 198560.01 0.3850 0.7014 1042 Pollen volume: Adjusted r 2= 0.6833, F4,70 = 37.76, P= < 2.2e-16 1043 Trait Estimate Std. Error t value Pr(>|t|) (Intercept) -1.854093 1.128503 -1.6430 0.1049 Flower size 0.228981 0.019152 11.9557 <2e-16 *** Green contrast -2.570704 3.149196 -0.8163 0.4171 Brightness contrast -1.526828 1.365145 -1.1184 0.2672 Chromatic contrast 0.982505 3.272178 0.3003 0.7649 47 1044 Blue Blue-green Green UV-green UV UV-blue 300 400 500 600 700 0 20 40 60 80 Wavelength (nm) Reflectance (%) a" b" Figure'1' Dittrichia viscosa Scolymus hispanicus Galactites tomentosa Campanula lusitanica Campanula rapunculus Campanula erinus Fedia cornucopiae Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Bartsia trixago Parentucellia viscosa Verbena officinalis Scrophularia frutescens Linaria viscosa Misopates orontium Cynoglossum creticum Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Borago officinalis Heliotropium europaeum Centaurium erythraea Blackstonia perfoliata Solanum nigrum Convolvulus althaeoides Convolvulus arvensis Erica arborea Erica umbellata Anagallis arvensis Anagallis monelli Silene colorata Tamarix gallica Ononis baetica Ononis pinnata Ononis viscosa Vicia sativa Vicia faba Vicia lutea Vicia benghalensis Erophaca baetica Ornithopus sativus Anthyllis cytisoides Coronilla juncea Bituminaria bituminosa Retama sphaerocarpa Lupinus luteus Lupinus angustifolius Genista triacanthos Genista hirsuta Cytisus grandiflorus Ulex eriocladus Scorpiurus sulcatus Stauracanthus genistoides Rubus ulmifolius Crataegus monogyna Rhamnus lycioides Bryonia dioica Hypericum perforatum Raphanus raphanistrum Diplotaxis virgata Diplotaxis erucoides Cytinus hypocistis Halimium halimifolium Cistus libanotis Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Halimium calycinum Tuberaria guttata Daphne gnidium Myrtus communis Ranunculus muricatus Ranunculus paludosus Clematis flammula Papaver rhoeas Serapias parviflora Scilla autumnalis Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Leucojum autumnale Leucojum trichophyllum Asphodelus ramosus Asphodelus fistulosus Smilax aspera Dittrichia viscosa Scolymus hispanicus Galactites tomentosa Campanula lusitanica Campanula rapunculus Campanula erinus Fedia cornucopiae Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Bartsia trixago Parentucellia viscosa Verbena officinalis Scrophularia frutescens Linaria viscosa Misopates orontium Cynoglossum creticum Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Borago officinalis Heliotropium europaeum Centaurium erythraea Blackstonia perfoliata Solanum nigrum Convolvulus althaeoides Convolvulus arvensis Erica arborea Erica umbellata Anagallis arvensis Anagallis monelli Silene colorata Tamarix gallica Ononis baetica Ononis pinnata Ononis viscosa Vicia sativa Vicia faba Vicia lutea Vicia benghalensis Erophaca baetica Ornithopus sativus Anthyllis cytisoides Coronilla juncea Bituminaria bituminosa Retama sphaerocarpa Lupinus luteus Lupinus angustifolius Genista triacanthos Genista hirsuta Cytisus grandiflorus Ulex eriocladus Scorpiurus sulcatus Stauracanthus genistoides Rubus ulmifolius Crataegus monogyna Rhamnus lycioides Bryonia dioica Hypericum perforatum Raphanus raphanistrum Diplotaxis virgata Diplotaxis erucoides Cytinus hypocistis Halimium halimifolium Cistus libanotis Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Halimium calycinum Tuberaria guttata Daphne gnidium Myrtus communis Ranunculus muricatus Ranunculus paludosus Clematis flammula Papaver rhoeas Serapias parviflora Scilla autumnalis Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Leucojum autumnale Leucojum trichophyllum Asphodelus ramosus Asphodelus fistulosus Smilax aspera 0.04 0.62trait value length=95.031 Reflectance)amplitude) Figure)2) Symmetry) Dittrichia viscosa Campanula lusitanica Campanula erinus Fedia cornucopiae Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Bartsia trixago Parentucellia viscosa Verbena officinalis Scrophularia frutescens Linaria viscosa Misopates orontium Cynoglossum creticum Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Centaurium erythraea Blackstonia perfoliata Solanum nigrum Convolvulus arvensis Erica umbellata Anagallis arvensis Anagallis monelli Silene colorata Tamarix gallica Ononis baetica Erophaca baetica Anthyllis cytisoides Coronilla juncea Retama sphaerocarpa Lupinus luteus Lupinus angustifolius Genista triacanthos Genista hirsuta Cytisus grandiflorus Ulex eriocladus Scorpiurus sulcatus Stauracanthus genistoides Rubus ulmifolius Crataegus monogyna Rhamnus lycioides Bryonia dioica Raphanus raphanistrum Diplotaxis virgata Diplotaxis erucoides Halimium halimifolium Cistus libanotis Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Halimium calycinum Daphne gnidium Myrtus communis Ranunculus muricatus Ranunculus paludosus Clematis flammula Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Leucojum trichophyllum Asphodelus fistulosus Smilax aspera Dittrichia viscosa Campanula lusitanica Campanula erinus Fedia cornucopiae Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Bartsia trixago Parentucellia viscosa Verbena officinalis Scrophularia frutescens Linaria viscosa Misopates orontium Cynoglossum creticum Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Centaurium erythraea Blackstonia perfoliata Solanum nigrum Convolvulus arvensis Erica umbellata Anagallis arvensis Anagallis monelli Silene colorata Tamarix gallica Ononis baetica Erophaca baetica Anthyllis cytisoides Coronilla juncea Retama sphaerocarpa Lupinus luteus Lupinus angustifolius Genista triacanthos Genista hirsuta Cytisus grandiflorus Ulex eriocladus Scorpiurus sulcatus Stauracanthus genistoides Rubus ulmifolius Crataegus monogyna Rhamnus lycioides Bryonia dioica Raphanus raphanistrum Diplotaxis virgata Diplotaxis erucoides Halimium halimifolium Cistus libanotis Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Halimium calycinum Daphne gnidium Myrtus communis Ranunculus muricatus Ranunculus paludosus Clematis flammula Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Leucojum trichophyllum Asphodelus fistulosus Smilax aspera 119 1330295trait value length=95.031 Pollen)grain)number) Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus 10.6 105trait value length=95.031 Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus 10.6 105trait value length=95.031 Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus 10.6 105trait value length=95.031 Dittrichia viscosa Campanula lusitanica Campanula erinus Fedia cornucopiae Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Bartsia trixago Parentucellia viscosa Verbena officinalis Scrophularia frutescens Linaria viscosa Misopates orontium Cynoglossum creticum Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Centaurium erythraea Blackstonia perfoliata Solanum nigrum Convolvulus arvensis Erica umbellata Anagallis arvensis Anagallis monelli Silene colorata Tamarix gallica Ononis baetica Erophaca baetica Anthyllis cytisoides Coronilla juncea Retama sphaerocarpa Lupinus luteus Lupinus angustifolius Genista triacanthos Genista hirsuta Cytisus grandiflorus Ulex eriocladus Scorpiurus sulcatus Stauracanthus genistoides Rubus ulmifolius Crataegus monogyna Rhamnus lycioides Bryonia dioica Raphanus raphanistrum Diplotaxis virgata Diplotaxis erucoides Halimium halimifolium Cistus libanotis Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Halimium calycinum Daphne gnidium Myrtus communis Ranunculus muricatus Ranunculus paludosus Clematis flammula Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Leucojum trichophyllum Asphodelus fistulosus Smilax aspera Dittrichia viscosa Campanula lusitanica Campanula erinus Fedia cornucopiae Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Bartsia trixago Parentucellia viscosa Verbena officinalis Scrophularia frutescens Linaria viscosa Misopates orontium Cynoglossum creticum Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Centaurium erythraea Blackstonia perfoliata Solanum nigrum Convolvulus arvensis Erica umbellata Anagallis arvensis Anagallis monelli Silene colorata Tamarix gallica Ononis baetica Erophaca baetica Anthyllis cytisoides Coronilla juncea Retama sphaerocarpa Lupinus luteus Lupinus angustifolius Genista triacanthos Genista hirsuta Cytisus grandiflorus Ulex eriocladus Scorpiurus sulcatus Stauracanthus genistoides Rubus ulmifolius Crataegus monogyna Rhamnus lycioides Bryonia dioica Raphanus raphanistrum Diplotaxis virgata Diplotaxis erucoides Halimium halimifolium Cistus libanotis Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Halimium calycinum Daphne gnidium Myrtus communis Ranunculus muricatus Ranunculus paludosus Clematis flammula Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Leucojum trichophyllum Asphodelus fistulosus Smilax aspera 119 1330295trait value length=95.031 Sugar)concentration) Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus Dittrichia viscosa Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Parentucellia viscosa Scrophularia frutescens Linaria viscosa Misopates orontium Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Heliotropium europaeum Convolvulus arvensis Erica arborea Erica umbellata Silene colorata Vicia faba Erophaca baetica Ornithopus sativus Anthyllis cytisoides Bituminaria bituminosa Scorpiurus sulcatus Rubus ulmifolius Crataegus monogyna Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Daphne gnidium Serapias parviflora Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Asphodelus ramosus Asphodelus fistulosus 10.6 105trait value length=95.031 Dittrichia viscosa Scolymus hispanicus Galactites tomentosa Campanula lusitanica Campanula rapunculus Campanula erinus Fedia cornucopiae Lonicera implexa Thymus mastichina Salvia verbenaca Rosmarinus officinalis Lavandula stoechas Teucrium capitatum Teucrium fruticans Phlomis purpurea Stachys ocymastrum Bartsia trixago Parentucellia viscosa Verbena officinalis Scrophularia frutescens Linaria viscosa Misopates orontium Cynoglossum creticum Echium gaditanum Echium plantagineum Cerinthe gymnandra Cerinthe major Anchusa calcarea Anchusa azurea Borago officinalis Heliotropium europaeum Centaurium erythraea Blackstonia perfoliata Solanum nigrum Convolvulus althaeoides Convolvulus arvensis Erica arborea Erica umbellata Anagallis arvensis Anagallis monelli Silene colorata Tamarix gallica Ononis baetica Ononis pinnata Ononis viscosa Vicia sativa Vicia faba Vicia lutea Vicia benghalensis Erophaca baetica Ornithopus sativus Anthyllis cytisoides Coronilla juncea Bituminaria bituminosa Retama sphaerocarpa Lupinus luteus Lupinus angustifolius Genista triacanthos Genista hirsuta Cytisus grandiflorus Ulex eriocladus Scorpiurus sulcatus Stauracanthus genistoides Rubus ulmifolius Crataegus monogyna Rhamnus lycioides Bryonia dioica Hypericum perforatum Raphanus raphanistrum Diplotaxis virgata Diplotaxis erucoides Cytinus hypocistis Halimium halimifolium Cistus libanotis Cistus albidus Cistus monspeliensis Cistus crispus Cistus ladanifer Cistus salviifolius Halimium calycinum Tuberaria guttata Daphne gnidium Myrtus communis Ranunculus muricatus Ranunculus paludosus Clematis flammula Papaver rhoeas Serapias parviflora Scilla autumnalis Ornithogalum baeticum Dipcadi serotinum Drimia maritima Allium roseum Leucojum autumnale Leucojum trichophyllum Asphodelus ramosus Asphodelus fistulosus Smilax aspera ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●●●●●●●●●● ● ● ● ●● ● ●●●●●●●● ● ● ● ● ● ● ● ● ● ● ● R Z Radial) Bilateral) Figure 3