A M E R I C A N J O U R N A L O F B OTA NY 102 ( 9 ): 1453 – 1461 , 2015 ; http://www.amjbot.org/ © 2015 Botanical Society of America • 1453 A M E R I C A N J O U R N A L O F B O T A N Y R E S E A R C H A R T I C L E Fruit colors are traditionally viewed as an adaptation to increase detection of fruits by animals that disperse seeds ( Kerner, 1895 ; Schmidt et al., 2004 ). Th erefore, communication is the most com- monly studied role of fruit colors because they facilitate detection and removal of ripe fruits by diurnal seed dispersers ( Willson and Whelan, 1990 ; Schaefer and Schaefer, 2007 ; Cazetta et al., 2009 ; Valido et al., 2011 ). Diurnal frugivores use fruit color as a selection criteria, selecting, either more conspicuous red and black fruits against the leaf background ( Schaefer and Schaefer, 2007 ; Cazetta et al., 2009 ; Melo et al., 2011 ) or those that reliably signal nutrients or antioxidants as a reward ( Schmidt et al., 2004 ; Schaefer et al., 2008a ; Schaefer and Ruxton, 2011 ). Fleshy-fruited plants generally produce green unripe fruits that are not conspicuous against the green leaf background ( Knight and Siegfried, 1983 ; Schaefer and Ruxton, 2011 ). As ripening pro- gresses, green, unripe fruits change color as chlorophyll degrades, and eventually exhibit the colors of carotenoids and anthocyanins, which result in nongreen-colored, ripe fruits ( Schaefer and Ruxton, 2011 ). Yet, some species produce fruits that are green when ripe. 1 Manuscript received 31 March 2015; revision accepted 7 August 2015. 2 Departamento de Botânica, Laboratório de Fenologia, Grupo de Fenologia e Dispersão de Sementes, Universidade Estadual Paulista, Avenida 24A 1515, CEP 13506-900, Rio Claro, SP, Brazil; fax: 55 19 3526-4201; 3 Department of Evolutionary Biology and Animal Ecology, Faculty of Biology, University of Freiburg, Hauptstr. 1 79104 Freiburg, Germany; 4 Departamento de Botânica, Instituto de Biociências, Universidade Estadual Paulista, Av. 24-A 1515, CEP 13506-900, Rio Claro, SP, Brazil; fax: 55 19 3526-4201; and 5 Departamento de Ciências Biológicas, Universidade Estadual de Santa Cruz, Rodovia Jorge Amado km 16, CEP 45662-900, Ilhéus, BA, Brazil; fax: 55 73 3680 5226 6 Author for correspondence (e-mail: gabicamargo@yahoo.com) doi:10.3732/ajb.1500138 Bicolored display of Miconia albicans fruits: Evaluating visual and physiological functions of fruit colors 1 Maria Gabriela G. de Camargo 2,6 , H. Martin Schaefer 3 , Gustavo Habermann 4 , Eliana Cazetta 5 , Natalia Costa Soares 2 , and Leonor Patrícia C. Morellato 2 PREMISE OF THE STUDY: Most bird-dispersed fruits are green when unripe and become colored and conspicuous when ripe, signaling that fruits are ready to be consumed and dispersed. The color pattern for fruits of Miconia albicans (Melastomataceae), however, is the opposite, with reddish unripe and green ripe fruits. We (1) verifi ed the maintenance over time of its bicolored display, (2) tested the communicative function of unripe fruits, (3) tested the photo- protective role of anthocyanins in unripe fruits, and (4) verifi ed whether green ripe fruits can assimilate carbon. METHODS: Using a paired experiment, we tested whether detection of ripe fruits was higher on infructescences with unripe and ripe fruits compared with infructescences with only ripe fruits. We also measured and compared gas exchange, chlorophyll a fl uorescence, and heat dissipation of covered (to pre- vent anthocyanin synthesis) and uncovered ripe and unripe fruits. KEY RESULTS: Although the bicolored display was maintained over time, unripe fruits had no infl uence on bird detection and removal of ripe fruits. Ripe and unripe fruits did not assimilate CO 2 , but they respired instead. CONCLUSIONS: Since the communicative function of unripe fruits was not confi rmed, seed dispersers are unlikely to select the display with bicolored fruits. Because of the absence of photosynthetic activity in ripe and unripe fruits and enhanced photoprotective mechanisms in ripe fruits rather than in unripe fruits, we could not confi rm the photoprotective role of anthocyanins in unripe fruits. As an alternative hypothesis, we suggest that the bicolored fruit display could be an adaptation to diversify seed dispersal vectors instead of restricting dispersal to birds and that anthocyanins in unripe fruits may have a defense role against pathogens. KEY WORDS anthocyanin photoprotection; frugivory; fruit contrast; fruit phenology; fruit photosynthesis; Melastomataceae; Miconia albicans 1454 • A M E R I C A N J O U R N A L O F B OTA NY Typically, these fruits are large, inconspicuous, and dispersed by large mammals or bats ( Knight and Siegfried, 1983 ; Cipollini and Levey, 1991 ). A green color seems disadvantageous for the attrac- tion of visually guided frugivores, and yet these fruits retain chloro- phyll and may contribute to the carbon requirements needed for plant reproduction, thereby reducing the costs of ripening ( Cipollini and Levey, 1991 ; Aschan and Pfanz, 2003 ). For example, chloro- phyll in fruits contributes to the production costs of developing peaches by fixing 15% of their carbon requirements ( Pavel and DeJong, 1993 ). A long-standing hypothesis predicts that species with relatively inconspicuous ripe fruits can improve the conspicuousness of fruit displays through nongreen secondary structures, such as bracts, capsules and stems ( Schaefer and Schaefer, 2007 ; Schaefer et al., 2007 ). Likewise, colored unripe fruits occurring concomitantly with ripe fruits can increase detection and fruit removal by diurnal frugivores because of the temporally maintained bicolored fruit display ( Stiles, 1982 ; Willson and Th ompson, 1982 ; Fuentes, 1995 ; Cramer et al., 2003 ). Anthocyanins contribute to blue and red to purple-black fruit colors ( Steyn et al., 2002 ; Jaakola, 2013 ). Th is group of pigments has various functions, but few of them have been investigated in reproductive structures. In leaves, anthocyanins can play a role in photoprotection against excessive sunlight and ultraviolet radia- tion ( Li et al., 1993 ; Gould, 2004 ). In the peels of apples and pears ( Steyn et al., 2009 ) and in the red stems of Cornus stolonifera ( Gould et al., 2010 ), anthocyanins prevent a decrease in photo- chemical performance under stress conditions such as low tem- peratures and excessive sunlight. Th is protective role is important for maintaining the photosynthetic capacity in leaves ( Gould, 2004 ) and in other organs such as fruits, pedicels, and bracts ( Cooney et al., 2015 ). However, except for cultivated species, stud- ies focusing on fruit color rarely consider the alternative hypoth- esis of a photosynthetic role of pigments on fruit growth, rather than just the visual function of the colors ( Cipollini and Levey, 1991 ; Aschan and Pfanz, 2003 ). Unlike the typical fruit color pattern of bird-dispersed species, reddish unripe fruits and green ripe fruits are produced by Miconia albicans (Melastomataceae), a small tree from the Brazilian savanna (cerrado). Th ese fruits are produced by obligatory apomixy, like many Melastomataceae species ( Goldenberg and Shepherd, 1998 ), and seed dispersal is the only mechanism for gene fl ow ( Goldenberg and Shepherd, 1998 ; Sales et al., 2013 ). Here we investigate the ecological function of this unusual bicolored fruit display, verifying its infl uence on fruit removal by seed dispersers and possible physi- ological functions of anthocyanin present in unripe fruits. To ac- complish this goal, we (1) described the ripening pattern of M. albicans to verify the maintenance of the bicolored display over time, (2) tested the communicative function of red unripe fruits, (3) tested the photoprotective role of anthocyanins in reddish unripe fruits, and (4) verifi ed whether green ripe fruits are photosyntheti- cally active (carbon assimilation). We modeled the contrasts of ripe and unripe fruits and leaves and tested whether consumption of ripe fruits was higher in infructescences with both unripe and ripe fruits compared with infructescences with only ripe fruits. To pre- vent anthocyanin synthesis in unripe fruits, we covered unripe fruits with aluminum foil at the initial phase of fruit development and compared gas exchange and chlorophyll a fl uorescence of cov- ered and uncovered unripe fruits and of covered and uncovered ripe fruits. If unripe fruits function as a communicative signal to seed dis- persers we expected that birds would detect and remove more fruits from infructescences with unripe and ripe fruits than from those bearing only ripe fruits. Conversely, if anthocyanins have a photo- protective function in unripe fruits, we expected increased heat dissipation and lower light fraction used by the photosynthetic ap- paratus in unripe fruits with anthocyanins than in unripe fruits in which anthocyanin synthesis was prevented. Finally, we expected positive carbon assimilation in green, ripe fruits, or, at least, an in- crease in CO 2 assimilation rates and light fraction used by the pho- tosynthetic apparatus in green, ripe fruits compared with reddish, unripe fruits. MATERIALS AND METHODS Study site — We did this study in a woody cerrado with an area of 260 ha and altitude of 770 m a.s.l. in the municipality of Itirapina, São Paulo State, Brazil (22 ° 10 ′ 31,41 ″ S; 47 ° 52 ′ 26,13 ″ W). Th e vegeta- tion is classifi ed as a cerrado sensu stricto according to Coutinho (1978) , a savanna with discontinuous woody strata reaching 6 to 7 m high or up to 12 m high in denser patches and a continuous her- baceous layer ( Reys et al., 2013 ). The climate corresponds to the Cwa type according to the Köppen system ( Köppen, 1948 ), with a dry winter from April to September and a wet summer from October to March. Average climatic data from 1972 to 2002 shows an annual rainfall of 1524 mm and mean annual temperature of 20.7 ° C ( Camargo et al., 2011a ). Study species — Miconia albicans (Sw.) Triana (Melastomataceae) is a small tree up to 3 m high, distributed from southeastern Brazil to southern Mexico, and it is abundant in diff erent physiognomies of the Brazilian cerrado and recently disturbed areas ( Goldenberg, 2004 ; Allenspach and Dias, 2012 ). Flowers have a white-greenish cream color and a light smell, and are organized in infl orescences. Th ey fl ower at the end of the dry season, peaking in September (M. G. G, Camargo and L. P. C. Morellato, personal observations based on 60 individuals observed monthly for 6 yr). Th ey have poricidal anthers, thus requiring buzz pollination to release pollen grains ( Buchmann, 1983 ; Renner, 1989 ; Goldenberg, 2004 ). In this spe- cies, pollen grains and fl oral visitors are rarely observed; fruits are produced by obligatory apomixy ( Renner, 1989 ; Goldenberg and Shepherd, 1998 ; M/ G. G. Camargo, personal observations). Fruits are small berries, 7.5 mm wide and 5.9 mm long, with an average of 34 seeds per fruit; seeds are 0.72 mm wide and 1.02 mm long (MGG Camargo, UNESP, personal observations). Th e pulp of ripe fruits is mainly composed of water (76.6%), sugars (13.4%) and proteins (2.17%) ( Maruyama et al., 2007 ). Fruits are organized in infructescences. On the basis of fruit size and color, three stages could be distinguished during fruit development: (1) inconspicu- ous, greenish-gray, initially unripe fruits; (2) reddish, unripe fruits at an intermediate stage; (3) and fully expanded, green, ripe fruits. Th e intermediate unripe (hereaft er “unripe”) and ripe fruits have a persistent red calyx (Appendix S1a and b, see Supplemental Data with the online version of this article). Miconia albicans is one of the most common species at the study site ( Reys et al., 2013 ) and is responsible for 30% of the annual production of animal-dispersed fruits in the local woody community (M. G. G. Camargo, unpub- lished manuscript). Ripe fruits are available mainly at the begin- ning of the wet season, peaking in November (based on unpublished S E P T E M B E R 2015 , V O LU M E 102 • C AMA R G O E T A L . — V I S UA L A N D P HYS I O LO G I C A L F U N C T I O N S O F F R U I T CO LO R S • 1455 data from 6 yr of monthly phenological observations of 60 indi- viduals). Th ey are an important food source consumed by diff er- ent bird species, mainly generalists ( Allenspach and Dias, 2012 ), marsupials Gracilinanus spp. ( Pereira et al., 2009 ; Camargo et al., 2011b ), and fallen fruits are also collected by ants ( Christianini et al., 2012 ). As far as we know, there is no information in the lit- erature whether reddish, unripe fruits are consumed by birds. Frugivory — To access the importance of M. albicans fruits as a re- source for birds and to see whether birds consume reddish, unripe fruits, we performed focal observations in 14 plant individuals. We observed bird activity from 05:00 to 18:00 during 5 d between 15 and 30 December 2010, during the peak of fruit production. Fruiting pattern — From September 2010 to March 2011, we quan- tifi ed weekly the number of initial unripe, unripe, and ripe fruits on 30 individuals of M. albicans selected along a trail and at least 5 m from each other. To quantify the phenophases, we counted the number of fruits present on three infructescences in each indi- vidual at diff erent stages and multiplied the mean number by the total number of infructescences present on each individual. To estimate the life span of unripe and ripe fruits, we selected two infructescences on each of the 10 marked individuals that ex- hibited initial unripe fruits and marked 20 to 25 fruits on each one, totaling 445 marked fruits. We observed fruits every other day at the beginning and daily at the peak of the fruiting period. Color and contrast — Using a spectrometer (Ocean Optics USB 4000, Dunedin, Florida, USA), we measured the refl ectance of the fruit skin of 30 unripe and 30 ripe fruits and from 30 leaves col- lected during the fruiting period. We recorded the light refl ected within 300–850 nm, which includes the visual range for all frugivo- rous animals, including birds that are sensitive to UV light. With the mean refl ectance spectrum of each structure, we calculated the chromatic and achromatic contrast between fruits and leaves and between ripe and unripe fruits. For that, we used visual models proposed by Vorobyev et al. (1998) for two kinds of avian visual sys- tems according to their peaks of sensitivity for ultraviolet wave- lengths: around 365 nm (UVS sensitive) or around 405 nm (VS sensitive). We performed these analyses using the R package Pavo ( Maia et al., 2013 ). Fruit detection and removal — To analyze whether the presence of reddish, unripe fruits infl uences the removal of ripe fruits, we performed a paired experiment using two treatments in which each pair had a corresponding number of fruits: Treatment 1: infructes- cences with unripe and ripe fruits according to this proportion: 55% initial unripe fruits (i.e., inconspicuous), 25% unripe fruits, and 20% ripe fruits, according to the mean proportion of each fruit type in nonmanipulated plants (Appendix S1c). Treatment 2: in- fructescences with only ripe fruits (Appendix S1d), equivalent in number to the sum of unripe and ripe fruits present in the paired infructescences of treatment 1. Considering a possible neighborhood eff ect, i.e., an infl uence of the density of fruiting plants in the vicinity and an infl uence of fruit crop size of individual trees on the detection of fruits by fru- givores ( Carlo and Morales, 2008 ; Christianini and Oliveira, 2009 ), we isolated the infructescences by removing and transfer- ring them to a perch. Each perch was fi xed on the top of a 2 m high wooden stake with a collector on the bottom ( Fig. 1B ). We repeatedly used 100 perches distributed along a trail, spaced 30 m apart. On the fi rst day, we distributed the infructescences on the perches, counted the number of ripe fruits in each treatment, and replaced the dry and damaged fruits or infructescences. An in- fructescence was regarded as being detected if at least one ripe fruit was removed. If a perch was detected, it was used again only aft er 4 d, to avoid subsequent visits based on memory rather than fruit detection. We conducted the experiment between December 2012 and January 2013, during the fruiting peak of M. albicans , totaling 30 d that correspond to four to fi ve continuous days of data collection with 1 d between each data collection. In the fi rst day of each data collection, we randomly selected perches and the corresponding treatment using from 40 to 60 perches. All 100 perches were used during the experiment, and 77 were detected at least once. For each perch, we estimated canopy openness, number of M. albicans reproductive individuals present in a radius of 5–10 m from the perch, and their respective tree circumference 3 cm from the soil. Th ese data allowed us to verify the infl uence that the en- vironment and the presence of fruiting M. albicans individuals around the perch could have on fruit removal. To estimate canopy openness, we took photos using a hemispherical lens “fi sheye con- verter” (Nikon FC-E9) coupled to a digital camera (Nikon Coolpix, model 8700) that was positioned above each perch. We transferred these photos to the Gap Light Analyzer 2.0 soft ware ( Frazer et al., 1999 ), which provides data on the percentage of canopy openness according to the proportion of black or white pixels in the hemi- spherical photo. Photosynthesis and anthocyanins — To verify the photoprotective role of anthocyanins, we fi rst covered some unripe fruits with alu- minum foil to prevent anthocyanin synthesis and left some uncov- ered. Aft er that, we compared CO 2 assimilation rate ( A ), chlorophyll a fl uorescence, and heat dissipation (1) between unripe fruits with anthocyanins (uncovered) and unripe fruits without anthocyanins (covered) and, aft er fruit development, (2) between covered and uncovered ripe fruits. Considering that the synthesis of anthocyanins usually re- quires exposure to light ( Mancinelli, 1985 ; Steyn et al., 2009 ; Jaakola, 2013 , Cooney et al., 2015 ), we selected 60 infructescences (on 30 plant individuals) and covered 8 to 10 initially unripe fruits with aluminum foil to block sunlight. We left an aperture on the underside of the aluminum foil to allow fruit gas exchange and prevent fruit abortion. Th e remaining fruits were completely ex- posed (uncovered) to sunlight (Appendix S1e). Aft er fruit devel- opment, which was assessed based on fruit size, we measured the refl ectance spectra of 20 covered and uncovered unripe fruits us- ing a spectrometer (Ocean Optics USB4000). Because anthocya- nins are responsible for the reddish color of unripe fruits, we calculated the ratio of red to green fruits for covered and uncov- ered unripe fruits and used this ratio as an estimation of fruit an- thocyanin content ( Gamon and Surfus, 1999 ; Schaefer and Braun, 2009 ; Gould et al., 2010 ). To verify any photosynthetic activity in ripe and unripe fruits, we evaluated fruit carbon assimilation by measuring the CO 2 as- similation rate ( A ) in response to an artifi cial light inside the cham- ber that provided photosynthetic photon flux density (PPFD) similar to the external environment and also in response to increas- ing the artifi cial PPFD (photosynthetic light curves). For this, we measured A of seven to nine covered and uncovered ripe and 1456 • A M E R I C A N J O U R N A L O F B OTA NY unripe fruits on 3 days during the ripening period. Measurements were performed in the fi eld, using a portable open gas exchange sys- tem (LI-6400xt; LI-COR, Lincoln, Nebraska, USA) coupled to a fluorometer chamber (LCF 6400-40; LI-COR) that was also used for measuring chlorophyll a fl uorescence. For gas exchange measurements, in vivo fruits were accommodated within the cham- ber, and measurements were performed between 09:00 and 12:00, as this interval seems to exhibit mild temperatures and best PPFD for gas exchange ( Feistler and Habermann, 2012 ). Th e measurements were taken between 10 December 2013 and 06 January 2014, with ambient temperature around 23 ° C ( ± 1.6) on the days of data collec- tion. Th e PPFD within the chamber was provided with an artificial red (90%) blue (10%) LED light source (6400-02B, LI-COR; Tennessen et al., 1994 ; Shimazaki et al., 2007 ) held on the top of the chamber. The PPFD was set according to the PPFD observed in the external ambient ( ≈ 600 μmol photons·m −2 ·s −1 ), the CO 2 concentration entering in the leaf chamber was set to 390 μmol·mol −1 , and the chamber fl ow was maintained around 450 μmol·s −1 . Th e mean tem- perature in the chamber was 30 ° C ( ± 2.15). Light curves ( A /PPFD) were obtained by varying the PPFD from 1200 μmol·m −2 ·s −1 to zero ( Habermann et al., 2011 ). Aft er measurements, we estimated the area (cm 2 ) of fruits that comprised the sample within the chamber and recalculated A based on the corrected fruit surface. For this, the diameter of every fruit was registered, and the half sphere surface was calculated because only the upper half of the fruits were exposed to the light coming from the top of the chamber. Th e sum of the half sphere surfaces was used as the corrected area logged in the LI-6400xt data analysis program. To estimate the photoprotective role of anthocyanins, we mea- sured chlorophyll a fl uorescence and calculated the fraction of ab- sorbed light used in photochemical activity [ P = (Fm ′ − Fs)/Fm ′ ], which is equivalent to the eff ective quantum yield of photosystem II ( Φ PSII), the heat dissipation in the light-gathering pigment an- tenna, which collects light and transfers the energy to the reaction center complex ( D = 1 – Fv ′ /Fm ′ ) and the heat dissipation in reac- tion centers of photosystem II, PSII [ E = (1 – qP)(Fv ′ /Fm ′ )] accord- ing to da Veiga and Habermann (2013) . We used saturating light pulses of 7000 μmol photons·m −2 ·s −1 for 0.7 s. For these equations, Fv ′ is the variable fl uorescence between the maximal (Fm ′ ) and minimal (Fo ′ ) fluorescence of light-adapted fruits, and Fs is the steady-state fluorescence of light-adapted fruits. Photochemical quenching (qP) was calculated as [(Fm ′ − Fs)/(Fm ′ − Fo ′ )], as adapted from Maxwell and Johnson (2000) . Chlorophyll fl uores- cence is the re-emitted and nonutilized light in photosynthesis, and it is widely used as a method to estimate the performance of PSII under abiotic stresses ( Maxwell and Johnson, 2000 ). Heat dissipa- tion in the light-gathering antenna ( D ) and reaction centers of PSII ( E ) and the fractions of absorbed light used in photochemical activ- ity ( P ) provide useful information about photoprotection (increased D and E ) or effi ciency of light reactions in photosynthesis (high P ) ( Maxwell and Johnson, 2000 ). FIGURE 1 (A) Miconia albicans infructescence with few initial unripe fruits (indicated by an arrow), reddish, intermediate unripe fruits and green, ripe fruits. Perch used for detection and fruit re- moval experiments (B). Data analyses — To assess the infl uence of the presence of unripe fruits on the number of detected infructescences and fruits re- moved, we used generalized linear models (GLM), specifi cally, bi- nary logistic regressions for infructescences detection and Poisson GLM for the number of fruits removed. We entered the number of ripe fruits at each perch as a covariate to assess whether the variation in the number of ripe fruits infl uenced the likelihood that an infruc- tescence was detected. We also used Poisson GLM to verify the infl u- ence of the canopy openness above the perches and the number and size of reproductive M. albicans individuals around the perches on the detection rate of infructescences and in the number of ripe fruits removed in each perch. All GLM models used are presented in the Appendix S2 (Appendix S2, see online Supplemental Data). To detect possible diff erences in fruit color indicating the syn- thesis of anthocyanins, we compared covered and uncovered unripe fruits refl ectance spectra between 300 to 670 nm and the refl ectance spectra in the red band (between 600 to 670 nm) using Wilcoxon tests (T). We used t tests to compare red: green ratio and CO 2 assimilation rate, fractions of absorbed light used in photo- chemistry (P), heat dissipation in the antenna (D) and in reactions centers (E) of covered and uncovered unripe fruits ( Zar, 1999 ). We performed all analyses in the program R, version 3.1.2 ( R Development Core Team, 2009 ). RESULTS Fruiting pattern — Th e 30 observed individuals produced fruits from 22 October 2010 to 30 March 2011 ( Fig. 2 ). Th e number of initially unripe fruits peaked on 12 November 2010 and were ob- served until 17 February 2011, with a mean of 7209 fruits per indi- vidual. Unripe and ripe fruit production overlapped, peaking on 30 December 2010, but they could be observed from 3 December 2010 to 30 March 2011, with a mean of 500 unripe and 447 ripe fruits per plant. S E P T E M B E R 2015 , V O LU M E 102 • C AMA R G O E T A L . — V I S UA L A N D P HYS I O LO G I C A L F U N C T I O N S O F F R U I T CO LO R S • 1457 From 445 initially unripe fruits that were marked, 382 (85.8%) turned into unripe fruits, and 346 (77.7%) ripened and were moni- tored until the abscission of ripe fruits. Unripe fruits lasted for 7 ± 5.9 (SD) days, and ripe fruits for 4 ± 2 (SD). Infructescence detection and fruit removal — During the 219 h of focal observation, we registered 194 visits of 26 bird species remov- ing ripe fruits of M. albicans (online Appendix S3). Unripe fruits were not consumed or removed. Refl ectance data confi rmed the diff erence in unripe and ripe fruit color previously observed in the fi eld ( Fig. 3 ). Th ere was al- most no overlapping in the refl ectance data of leaves and fruits with the highest chromatic and achromatic contrasts between unripe fruits and leaves and not between ripe and unripe fruits ( Fig. 3 and Table 1 ). Treatments did not infl uence the number of infructescences de- tected ( z = 1.28, p = 0.20). Perches of treatment 1 (infructescences with ripe and unripe fruits) were detected 72 times, and 145 fruits were removed, while perches of treatment 2 (only ripe fruits) were detected 88 times, and 310 fruits were removed. Yet, treatments did infl uence the number of fruits removed ( z = 2.9, p < 0.01), which was higher in treatment 2 with only ripe fruits. We found no infl u- ence of the canopy openness or the number and circumference size of M. albicans reproductive individuals around the perches on the detection of infructescences ( z = 0.59, p = 0.55) or on fruit removal ( z = 0.51, p = 0.60). Photosynthesis and anthocyanins — Covered and uncovered un- ripe fruits did not diff er in the red to green fruits ratio ( t test, t = 0.02, df = 35.4, p = 0.98). Th e CO 2 assimilation rate was negative in covered and uncovered unripe and ripe fruits, indicating that respiration and no photosyn- thetic activity ( Fig. 4 ). Respiration was higher in covered compared to uncovered unripe fruits ( t test, t = 2.47, df = 39.6, p = 0.02); for ripe fruits, respiration diff ered marginally between covered and uncov- ered fruits ( t test, t = 2.00, df = 38.70, p = 0.05). Respiration (negative CO 2 assimilation) did not respond to the variation in PPFD ( A /PPFD curves) either in unripe ( Fig. 5A ) or ripe fruits ( Fig. 5B ). Although we did not check the presence of stomata on fruits, Cortez and Carmello- Guerreiro (2008) described stomata in the pericarp of Miconia albicans fruits. We found positive stomatal conductance for unripe (0.06 ± 0.06 mol·m −2 ·s −1 ) and ripe fruits (0.07 ± 0.5 mol·m −2 ·s −1 ). Th erefore, we must consider that these responses could be related to gas exchange occurring on the fruit’s epidermis. Fluorescence variables ( P , D , and E ) were signifi cantly diff erent between covered and uncovered fruits, regardless of the ripening stage of the fruit ( Fig. 6 ). Relative to covered fruits, uncovered ones had higher fractions of absorbed light used in photochemistry, P ( t test, t = 5.66, df = 41.93, p < 0.01 for unripe and t = 3.58, df = 45.97, FIGURE 2 Fruiting patterns of 30 Miconia albicans individuals observed between September 2010 and March 2011 in wood cerrado–savanna vegetation in southeastern Brazil. FIGURE 3 Mean refl ectance spectra (lines) and standard deviation (shad- ing) of Miconia albicans ripe fruits, unripe fruits, and leaves. TABLE 1. Chromatic and achromatic contrast between ripe fruits, leaves, unripe fruits, and fruit persistent calyx of Miconia albicans according to a bird’s visual system model considering sensitivities for ultraviolet light. Chromatic UVS Chromatic VS Achromatic Stage Ripe fruit Leaf Ripe fruit Leaf Ripe fruit Leaf Ripe fruit — 23.1 — 14.1 — 36.9 Unripe fruit 11.41 26.3 11.81 20.8 5.63 42.5 Notes: UVS = visual bird system with peak sensitivity for ultraviolet wavelengths around 365 nm, VS = visual bird system with peak sensitivity for ultraviolet wavelengths around 405 nm. FIGURE 4 Box plot of CO 2 assimilation rates in uncovered and covered unripe and ripe fruits of Miconia albicans . * Signifi cantly diff erent accord- ing to t test. 1458 • A M E R I C A N J O U R N A L O F B OTA NY p < 0.01 for ripe fruits; Fig. 6A ), and higher heat dissipation in the antenna, D ( t test, t = 4.68, df = 38.32, p < 0.01 for unripe and t = 2.71, df = 43.86, p = 0.01 for ripe fruits; Fig. 6B ). Th e opposite (covered > uncovered) occurred for dissipation in reactions cen- ters, E ( t test, t = −6.24, df = 39.74, p < 0.01 for unripe and t = −3.53, df = 41.91, p < 0.01 for ripe fruits; Fig. 6C ). DISCUSSION We observed the presence of reddish, unripe fruits throughout the fruit season, confi rming the persistence of the bicolored of M. albicans infructescences. However, our results did not support a communicative function of the unripe fruits, because they did not increase ripe fruit contrast, detection, or removal. Similarly, a pho- toprotective role of anthocyanins in fruits could not be confi rmed due to the enhanced photoprotective mechanism observed in green, ripe fruits rather than in reddish, unripe fruits, with in- creased heat dissipation in the light-gathering antenna ( D ) in ripe fruits. Th ough not confi rmed here, our study highlights the impor- tance of considering alternative functions of fruit color beyond plant–animal communication and the role of anthocyanins in non- vegetative plant organs, factors that infl uence fruit color evolution ( Schaefer and Schaefer, 2007 ). Our results showed that different bird species consume M. albicans , mainly generalists, as also observed by Allenspach and Dias (2012) , and the birds we observed fed exclusively on ripe fruits, diff erentiating them from the reddish, unripe ones. In gen- eral, birds have an innate preference for red fruits, which are more common, conspicuous, and preferentially selected over other colors, such as white and green ( Chen et al., 2004 ; Schaefer et al., 2008b ; Schaefer and Ruxton, 2011 ). However, fruit color pre- ferences are inconsistent in birds because color-reward associations FIGURE 5 Light response curve representing CO 2 assimilation rates for diff erent photosynthetic photon fl ux densities (PPFD) of (A) unripe and (B) ripe fruits of Miconia albicans . FIGURE 6 Chlorophyll a fl uorescence variables, (A) absorbed light used in photochemistry, (B) heat dissipation in the antenna, and (C) heat dissipa- tion in reactions centers in uncovered and covered unripe and ripe fruits of Miconia albicans . Values for each variable diff ered signifi cantly be- tween covered and uncovered fruits according to a t test. can be spatially variable, and learning plays an important role in fruit selection ( Schaefer et al., 2008b ; Schaefer and Ruxton, 2011 ). Th e removal of only green, ripe fruits of M. albicans could S E P T E M B E R 2015 , V O LU M E 102 • C AMA R G O E T A L . — V I S UA L A N D P HYS I O LO G I C A L F U N C T I O N S O F F R U I T CO LO R S • 1459 be related to associative learning by birds, based on a larger size or green color of fruit. Although unripe fruits do not directly enhance the contrast of ripe fruits, unripe fruits have a higher contrast than the ripe ones against leaves, increasing the overall contrasts of the fruit display. However, this bicolored display of M. albicans infructescences seems to have no infl uence on fruit detection and removal. Since unripe, reddish fruits did not infl uence the probability that infruc- tescences were detected by birds, the higher fruit removal in infruc- tescences with only ripe fruits is probably related to the higher number of ripe fruits available in this treatment. Instead of serving as a communicative function, the bicolored display of M. albicans could be related to limited resources for synchronous fruit matura- tion or to a sequential anthesis strategy in the infl orescence, confer- ring asynchronous fruit ripening in the infructescence (see also Amsberry and Steff en, 2008 ). In general, there is no consensus on the eff ects of bicolored dis- play of ripe and unripe fruits on fruit detection and removal by birds. For example, higher fruit removal from bicolored displays of black, ripe and red, unripe fruits was reported for Prunus serotina and Phytolacca americana ( Willson and Melampy, 1983 ) and also a higher removal rate for blue Psychotria pittieri ripe fruits when as- sociated with an artifi cial red bract ( Wenny, 2003 ). Similar results are reported for infructescences with only red, unripe fruits associated with infructescences with only green, ripe fruits of Pistacia terebin- thus ( Fuentes, 1995 ). In contrast, fruit removal was higher in in- fructescences with only green, ripe fruits of Pistacia terebinthus compared with treatments that off ered red, unripe and green, ripe fruits together in the same infructescence ( Fuentes, 1995 ). Amsberry and Steff en (2008) found higher fruit removal by birds in infructes- cences with only red, ripe fruits than when combined with white and pink, unripe fruits of Ardisia nigropunctata . Finally, Burns, (2005) also found no preference for experimental bicolored display in Rubus spectabilis . Some of the contrasting results in the selection of bicolored fruit displays could be related to diff erences in the methods used. First, all studies mentioned above based their hypotheses on fruit colors as they are perceived by humans and not by seed dispersers. Sec- ond, only the number of fruits removed was compared between treatments, without considering the detection of infructescences. Th ird, the experimental design used by these studies diff ers among them, with infructescences being paired or randomly distributed, and with variations in the number of fruits of different colors offered in each experimental infructescence (e.g., Willson and Melampy, 1983 ; Amsberry and Steff en, 2008 ). Finally, in most stud- ies fruits that drop beneath the infructescences are oft en not dif- ferentiated from those actually removed by birds. One bias of our study could be the absence of protection against nonvisually guided frugivores, such as marsupials ( Gracilinanus spp.), that consume M. albicans fruits ( Pereira et al., 2009 ). However, during the experi- ment, we did not observe any trace of mammals feeding on experi- mental infructescences. With regard to the photoprotective role of anthocyanins, our re- sults did not show diff erences in anthocyanin content between cov- ered and uncovered unripe fruits. Th erefore, we concluded that sunlight might not be the most important factor inducing anthocy- anin synthesis in M. albicans , as this biosynthesis is also controlled by genetic factors ( Jaakola, 2013 ). Contrasting with our predic- tions, the change from reddish to green during ripening did not cause any increase in photosynthetic performances in ripe fruits. Neither ripe nor unripe fruits assimilated CO 2 , but they respired instead. Th us, in M. albicans , chlorophyll in fruits did not contrib- ute to the maintenance costs of these organs. Even as PPFD in- creased, the respiration rate did not decrease in M. albicans fruits, indicating that respiration is self-controlled rather than environ- mentally controlled in these fruits. Th is result contrasts with previ- ous studies on larger fruits ( Cipollini and Levey, 1991 ), which may be more costly to produce and maintain. As such, we rule out the hypothesis that the adaptive signifi cance of the fruit color shift to- ward green, ripe fruits is increasing their photosynthetic abilities. Uncovered fruits had enhanced light reactions in the photo- chemical “apparatus” because P and D were higher in relation to fruits that were protected from sunlight, indicating that electron transporters are more active (not oxidized; Baker, 2008 ) in the presence of light, suggesting that light does play a role in the func- tioning of the photochemical apparatus of M. albicans fruits. During the ripening of nongreen, ripe fruits, less chlorophyll fl uorescence is generally expected as chlorophyll is degraded, with correspond- ing changes from green to yellow ( Bron et al., 2004 ). For M. albi- cans , the opposite is expected as ripe fruits become green. However, in the current study, P consistently decreased and D increased as fruits ripened, suggesting that the photochemical activities to pro- vide NADPH and ATP to the carboxylation requirement are at- tenuated ( Baker, 2008 ). In addition, the effi ciency of photosystem II decreased as fruits ripened, as evidenced by decreased values of P and increased values of D during ripening, reinforcing a low pho- tochemical activity of green ripe fruits. Th is fi nding strengthens our conclusion that ripe fruits, although green, do not contribute to photosynthates for fruit development. Given that the observed photoprotective mechanism (increased D ) is enhanced in ripe fruits, rather than in unripe fruits that are supposedly rich in antho- cyanins, we could not confi rm the photoprotective role of antho- cyanins in reddish, unripe fruits. Although the unexpected higher P values in reddish, unripe fruits in relation to green, ripe fruits could indicate that some photochemical activity occurs in unripe fruits, this phenomenon can be a consequence of the anthocyanin presence in the unripe fruits, which sometimes can cause an over- estimate of Φ PSII values ( Pietrini and Massacci, 1998 ). In a similar experiment, Cooney et al. (2015) compared Sambucus nigra infruc- tescences with green and red peduncles and did not confi rm a com- municative role of anthocyanins. However, contrary to our results in unripe fruits of M. albicans , anthocyanins exerted a photopro- tective function, which probably is the driver for anthocyanin ac- cumulation in the peduncles of Sambucus nigra ( Cooney et al., 2015 ). Although various roles are frequently described for antho- cyanins in nonreproductive plant organs, it is still necessary to verify possible diff erent roles of anthocyanins in fruits to under- stand evolutionary drivers of fruit color. Taking into account the importance of fruit removal and seed dispersal for gene fl ow in M. albicans due to the obligatory apomic- tic reproductive system, we suggest that the bicolored fruit display may be a strategy toward a diversifi cation of seed disperser vectors, instead of restricting dispersal to birds, improving seed dispersal possibilities and effi ciency. Our assumption is supported by the importance of M. albicans fruits for some cerrado marsupials that are potential seed dispersers of this species ( Pereira et al., 2009 ; Camargo et al., 2011b ). One example is the high frequency of M. albicans in fecal samples of Gracilinanus agilis , in which M. al- bicans , M. ferruginata , and M. pohliana were present in 44% of 186 samples ( Camargo et al., 2011b ). 1460 • A M E R I C A N J O U R N A L O F B OTA NY A possible alternative role for anthocyanins on reddish unripe fruits of M. albicans might be related to fruit defense against patho- gens and herbivores, as already reported for leaves ( Gould, 2004 ; Cooney et al., 2012 ) and nonvegetative structures in plants. Irwin et al. (2003) demonstrated that anthocyanin color of Raphanus sativus fl owers works as a cue to deter herbivores. Anthocyanins in combination with defensive compounds, such as as thiarubrine A, can also indicate the presence of defensive compounds in imma- ture fruits ( Gould, 2004 ) or promote crypsis in red fruits to frugivo- rous insects since most of them are not sensitive to red band wavelengths ( Briscoe and Chittka, 2001 ). Th e alternative hypothe- sis of a defensive role for anthocyanins might be reinforced by fi eld observations of the absence of herbivory of unripe fruits of M. albi- cans and the high ripening success, just 10% failure, even with the frequent presence of fungi and galls ( Viana et al., 2013 ; Appendix S1f and g) in leaves, fl ower buds, and fruits. ACKNOWLEDGEMENTS Th e authors thank the owners of Fazenda São José da Conquista and the Instituto Arruda Botelho (IAB) for allowing us to conduct fi eldwork at the farm. Th e authors thank Dr. Pedro Jordano for valuable suggestions, mainly on the statistical analyses. Th is study was supported by the São Paulo Research Foundation (FAPESP) (grants #2007/59779-6 and #2010/51307-0). MGGC received scholarships from FAPESP (grants #2010/01762-3 and BEPE #2012/19827-0). They acknowledge the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for Research Productivity fellowships to G.H. (308902/2014-9) and L.P.C.M. (306243/2010-5). Th is work is a contribution from the Laboratory of Phenology and the Plant Phenology and Seed Dispersal Group. H.M.S. benefi ted from FAPESP grant #2010/52113-5 and was further supported by CAPES through the Plant Biology Graduation Program of UNESP Rio Claro. L.P.C.M. received CNPq PVE fellowship. L.P.C.M. and G.H. received a grant from FAPESP (#2009/54208-6). LITERATURE CITED Allenspach , N. , and M. M. Dias . 2012 . Frugivory by birds on Miconia albicans (Melastomataceae), in a fragment of cerrado in São Carlos, southeastern Brazil. Brazilian Journal of Biology 72 : 407 – 413 . Amsberry , L. K. , and J. E. Steff en . 2008 . Do contrastingly colored unripe fruits of the neotropical tree Ardisia nigropunctata attract avian seed dispersers? Biotropica 40 : 575 – 580 . Aschan , G. , and H. Pfanz . 2003 . Non-foliar photosynthesis—A strategy of ad- ditional carbon acquisition. Flora 198 : 81 – 97 . Baker , N. R. 2008 . Chlorophyll fl uorescence: A probe of photosynthesis in vivo. Annual Review of Plant Biology 59 : 89 – 113 . Briscoe , A. D. , and L. Chittka . 2001 . Th e evolution of color vision in insects. Annual Review of Entomology 46 : 471 – 510 . Bron , I. U. , R. V. Ribeiro , M. Azzolini , A. P. Jacomino , and E. C. Machado . 2004 . Chlorophyll fl uorescence as a tool to evaluate the ripening of ‘Golden’ pa- paya fruit. Postharvest Biology and Technology 33 : 163 – 173 . Buchmann , S. C. 1983 . Buzz pollination in angiosperms. In C. E. Jones and R. J. Little [eds.], Handbook of experimental pollination biology, 73–113. Van Nostrand & Reinhold, New York, New York, USA. Burns , K. C. 2005 . Eff ects of bi-colored displays on avian fruit color prefer- ences in a color polymorphic plant. Journal of the Torrey Botanical Society 132 : 505 – 509 . Camargo , M. G. G. , R. M. Souza , P. Reys , and L. P. C. Morellato . 2011a . Eff ects of environmental conditions associated to the cardinal orientation on the reproductive phenology of the cerrado savanna tree Xylopia aromatica (Annonaceae). Annals of the Brazilian Academy of Sciences 83 : 1007 – 1020 . Camargo , N. F. , R. M. S. Cruz , J. F. Ribeiro , and E. M. Vieira . 2011b . Frugivoria e potencial dispersão de sementes pelo marsupial Gracilinanus agilis (Didelphidae: Didelphimorphia) em á reas de Cerrado no Brasil Central. Acta Botanica Brasílica 25 : 646 – 656 . Carlo , T. , and J. M. Morales . 2008 . Inequalities in fruit-removal and seed dis- persal: Consequences of bird behaviour, neighbourhood density and land- scape aggregation. Journal of Ecology 96 : 609 – 618 . Cazetta , E. , H. M. Schaefer , and M. Galetti . 2009 . Why are fruits colorful? Th e relative importance of achromatic and chromatic contrasts for detection by birds. Evolutionary Ecology 23 : 233 – 244 . Chen , J. , T. H. Fleming , L. Zhang , H. Wang , and Y. Liu . 2004 . Patterns of fruit traits in a tropical rainforest in Xishuangbanna, SW China. Acta Oecologica 26 : 157 – 164 . Christianini , A. , A. J. Mayhé-Nunes , and P. S. Oliveira . 2012 . Exploitation of fallen diaspores by ants: Are there ant–plant partner choices? Biotropica 44 : 360 – 367 . Christianini , A. V. , and P. S. Oliveira . 2009 . Th e relevance of ants as seed res- cuers of a primarily bird-dispersed tree in the Neotropical cerrado savanna. Oecologia 160 : 735 – 745 . Cipollini , M. L. , and D. J. Levey . 1991 . Why some fruits are green when they are ripe: Carbon balance in fl eshy fruits. Oecologia 88 : 371 – 377 . Cooney , L. J. , J. W. van Klink , N. M. Hughes , N. B. Perry , H. M. Schaefer , I. J. Menzies , and K. S. Gould . 2012 . Red leaf margins indicate increased polygodial content and function as visual signals to reduce herbivory in Pseudowintera colorata. New Phytologist 194 : 488 – 497 . Cooney , L. J. , H. M. Schaefer , B. A. Logan , B. Cox , and K. S. Gould . 2015 . Functional signifi cance of anthocyanins in peduncles of Sambucus nigra . Environmental and Experimental Botany 119 : 18 – 26 . Cortez , P. A. , and S. M. Carmello-Guerreiro . 2008 . Ontogeny and struc- ture of the pericarp and the seed coat of Miconia albicans (Sw.) Triana (Melastomataceae) from “Cerrado”, Brazil. Brazilian Journal of Botany 31 : 71 – 79 . Coutinho , L. M. 1978 . O conceito do cerrado . Revista Brasileira de Botanica. Brazilian Journal of Botany 1 : 17 – 23 . Cramer , J. M. , M. L. Cloud , N. C. Muchhala , A. E. Ware , B. H. Smith , and G. B. Williamson . 2003 . A test of the bicolored fruit display hypothesis: Berry removal with artifi cial fruit fl ags. Journal of the Torrey Botanical Society 130 : 30 – 33 . da Veiga , E. B. , and G. Habermann . 2013 . Instantaneously measured traits may detect non-plastic ecophysiological performances in response to drought, explaining distributions of Styrax species in the Cerrado. Trees (Berlin) 27 : 1737 – 1745 . Feistler , A. M. , and G. Habermann . 2012 . Assessing the role of vertical leaves within the photosynthetic function of Styrax camporum under drought con- ditions. Photosynthetica 50 : 613 – 622 . Frazer , G. W. , C. D. Canham , and K. P. Lertzman . 1999 . Gap Light Analyzer (GLA), version 2.0: Imaging soft ware to extract canopy structure and gap light transmission indices from true-colour fi sheye photographs, users man- ual and program documentation. Simon Fraser University, Burnaby, British Columbia; Institute of Ecosystem Studies, Millbrook, New York, USA. Fuentes , M. 1995 . Th e eff ect of unripe fruits on ripe fruit removal by birds in Pistacia terebinthus : Flag or handicap? Oecologia 101 : 55 – 58 . Gamon , J. A. , and J. S. Surfus . 1999 . Assessing leaf pigment content and activity with a refl ectometer. New Phytologist 143 : 105 – 117 . Goldenberg , R. 2004 . O gênero Miconia (Melastomataceae) no estado do Paran á . Acta Botanica Brasílica 18 : 927 – 947 . Goldenberg , R. , and G. J. Shepherd . 1998 . Studies on the reproductive biol- ogy of Melastomataceae in “cerrado” vegetation. Plant Systematics and Evolution 211 : 13 – 29 . Gould , K. S. 2004 . Nature’s Swiss army knife: Th e diverse protective roles of anthocyanins in leaves. BioMed Research International 5 : 314 – 320 . Gould , K. S. , D. A. Dudle , and H. S. Neufeld . 2010 . Why some stems are red: Cauline anthocyanins shield photosystem II against high light stress. Journal of Experimental Botany 61 : 2707 – 2717 . Habermann , G. , P. F. V. Ellsworth , J. L. Cazoto , A. M. Feistler , L. Silva , D. A. Donatti , and S. R. Machado . 2011 . Leaf paraheliotropism in Styrax S E P T E M B E R 2015 , V O LU M E 102 • C AMA R G O E T A L . — V I S UA L A N D P HYS I O LO G I C A L F U N C T I O N S O F F R U I T CO LO R S • 1461 camporum confers increased light use effi ciency and advantageous pho- tosynthetic responses rather than photoprotection. Environmental and Experimental Botany 71 : 10 – 17 . Irwin , R. E. , S. Y. Strauss , S. Storz , A. Emerson , and G. Guibert . 2003 . Th e role of herbivores in the maintenance of a fl ower color polymorphism in wild radish. Ecology 84 : 1733 – 1743 . Jaakola , L. 2013 . New insights into the regulation of anthocyanin biosynthesis in fruits. Trends in Plant Science 18 : 477 – 483 . Kerner , A. 1895 . Th e natural history of plants: Th eir forms, growth, reproduc- tion and distribution. Holt, London, UK. Knight , R. S. , and W. R. Siegfried . 1983 . Inter-relationships between type, size and colour of fruits and dispersal in southern African trees. Oecologia 56 : 405 – 412 . Köppen , W. 1948 . Climatología . Fondo Cultura Económica, Mexico City, Mexico. Li , J. , T. M. Ou-Lee , R. Raba , R. G. Amundson , and R. L. Last . 1993 . Arabdopsis fl avonoid mutants are hypersensitive to UV-B irradiation. Plant Cell 5 : 171 – 179 . Maia , R. , C. M. Eliason , P. P. Bitton , S. M. Doucet , and M. D. Shawkey . 2013 . Pavo: an R package for the analysis, visualization and organization of spec- tral data . Methods in Ecology and Evolution 4 : 906 – 913 . Mancinelli , A. L. 1985 . Light-dependent anthocyanin synthesis: A model sys- tem for the study of plant morphogenesis. Botanical Review 51 : 107 – 157 . Maruyama , P. K., E. Alves-Silva , and C . Melo . 2007 . Oferta qualitativa e quantitativa de frutos em espécies ornitocóricas do gênero Miconia (Melastomataceae). Revista Brasileira de Biociências 5 : 672 – 674 . Maxwell , K. , and G. N. Johnson . 2000 . Chlorophyll fl uorescence—A practical guide. Journal of Experimental Botany 51 : 659 – 668 . Melo , G. L. , N. C. Penatti , and J. Raizer . 2011 . Fruit of a contrasting colour is more detectable by frugivores. Journal of Tropical Ecology 27 : 319 – 322 . Pavel , E. W. , and T. M. Dejong . 1993 . Estimating the photosynthetic con- tribution of developing peach ( Prunus persica ) fruits to their growth and maintenance carbohydrate requirements. Physiologia Plantarum 88 : 331 – 338 . Pereira , M. S. , M. Passamani , and E. A. A. Silva . 2009 . Germinação de semen- tes de Miconia (Melastomataceae) ingeridas pelo marsupial Gracilinanus microtarsus (Didelphidae). Boletim do Museu de Biologia Mello Leitão 25 : 43 – 51 . Pietrini , F. , and A. Massacci . 1998 . Leaf anthocyanin content changes in Zea mays L. grown at low temperature: Signifi cance for the relationship between the quantum yield of PS II and the apparent quantum yield of CO 2 assimila- tion. Photosynthesis Research 58 : 213 – 219 . R Development Core Team . 2009 . R: A language and environment for statisti- cal computing. R Foundation for Statistical Computing, Vienna, Austria. Website: https://www.R-project.org . Renner , S. S. 1989 . A survey of reproductive biology in Neotropical Melastomataceae and Memecylaceae. Annals of the Missouri Botanical Garden 76 : 496 – 518 . Reys , P. , M. G. G. Camargo , A. P. Teixeira , M. A. Assis , M. T. Grombone- Guaratini , and L. P. C. Morellato . 2013 . Estrutura e composição fl orística entre borda e interior de um cerrado sensu stricto e sua importância para propostas de restauração ecológica. Hoehnea 40 : 437 – 452 . Sales , N. M. , F. Pérez-García , and F. A. O. Silveira . 2013 . Consistent variation in seed germination across an environmental gradient in a Neotropical sa- vanna. South African Journal of Botany 87 : 129 – 133 . Schaefer , H. M. , and J. Braun . 2009 . Reliable cues and signals of fruit quality are contingent on the habitat in black elder ( Sambucus nigra ). Ecology 90 : 1564 – 1573 . Schaefer , H. M. , K. McGraw , and C. Catoni . 2008a . Birds use fruit colour as honest signal of dietary antioxidant rewards. Functional Ecology 22 : 303 – 310 . Schaefer , H. M. , and G. D. Ruxton . 2011 . Plant–animal communication . Oxford University Press, Oxford, UK. Schaefer , H. M. , and V. Schaefer . 2007 . Th e evolution of visual signals: Concepts and constraints. In A. J. Dennis, E. W. Schupp, R. A. Green, and D. A. Westcott [eds.], Seed dispersal: Th eory and its application in a chang- ing world, 59–77, CABI Publishing, Wallingford, UK. Schaefer , H. M. , V. Schaefer , and M. Vorobyev . 2007 . Are fruit colors adapted to consumer vision and birds equally effi cient in detecting colorful signals? American Naturalist 169 : S159 – S169 . Schaefer , H. M. , K. Spitzer , and F. Bairlein . 2008b . Long- term eff ects of previous experience determine nutrient discrimination abilities in birds. Frontiers in Zoology 5 : 4 . Schmidt , V. , H. M. Schaefer , and H. Winkler . 2004 . Conspicuousness, not color as foraging cue in plant–animal interactions. Oikos 106 : 551 – 557 . Shimazaki , K. I. , M. Doi , S. M. Assmann , and T. Kinoshita . 2007 . Light regula- tion of stomatal movement. Annual Review of Plant Biology 58 : 219 – 247 . Steyn , W. J. , S. J. E. Wand , D. M. Holcroft , and G. Jacobs . 2002 . Anthocyanins in vegetative tissues: A proposed unifi ed function in photoprotection. New Phytologist 155 : 349 – 361 . Steyn , W. J. , S. J. E. Wand , G. Jacobs , R. C. Rosecrance , and S. C. Roberts . 2009 . Evidence for a photoprotective function low-temperature-induced antho- cyanin accumulation in apple and pear peel. Physiologia Plantarum 136 : 461 – 472 . Stiles , E. W. 1982 . Fruit flays: Two hypotheses. American Naturalist 120 : 500 – 509 . Tennessen , D. J. , E. L. Singsaas , and T. D. Sharkey . 1994 . Light-emitting di- odes as a light source for photosynthesis research. Photosynthesis Research 39 : 85 – 92 . Valido , A. , H. M. Schaefer , and P. Jordano . 2011 . Colour, design and reward: Phenotypic integration of fleshy fruit displays. Journal of Evolutionary Biology 24 : 751 – 760 . Viana , L. R. , F. A. O. Silveira , J. C. Santos , L. H. Rosa , J. E. Cares , A. C. Café- Filho , and G. W. Fernandes . 2013 . Nematode-induced galls in Miconia al- bicans : eff ect of host plant density and correlations with performance. Plant Species Biology 28 : 63 – 69 . Vorobyev , M. , D. Osorio , A. Bennett , N. Marshall , and I. Cuthill . 1998 . Tetrachromacy, oil droplets and bird plumage colours. Journal of Com- parative Physiology 183 : 621 – 633 . Wenny , D. G. 2003 . Short-term fruit removal from a Neotropical shrub with experimental bicoloured displays . Journal of Tropical Ecology 19 : 469 – 472 . Willson , M. F. , and M. N. Melampy . 1983 . Th e eff ect of bicolored fruit displays on fruit removal by avian frugivores. Oikos 41 : 27 – 31 . Willson , M. F. , and J. N. Th ompson . 1982 . Phenology and ecology of color in bird-dispersed fruits or why some fruits are red when they are “green”. Canadian Journal of Botany 60 : 701 – 713 . Willson , M. F. , and C. J. Whelan . 1990 . Th e evolution of fruit color in fl eshy- fruited plants. American Naturalist 136 : 790 – 809 . Zar , H. J. 1999 . Biostatistical analysis, 4th ed. Prentice-Hall, Upper Saddle River, New Jersey, USA.