Int. J. Plant Sci. 179(4):287–295. 2018. q 2018 by The University of Chicago. All rights reserved. 1058-5893/2018/17904-0003$15.00 DOI: 10.1086/696824 DYNAMICS OF POLLEN RELEASE AND STIGMATIC EXPOSURE IN NEOTROPICAL PIPER SPECIES: A POSSIBLE PATTERN FOR THE GENUS Adriano Valentin-Silva,* Marco Antonio Batalha,† and Elza Guimarães1,‡ *Graduate Program of Biological Sciences (Botany), Universidade Estadual Paulista, 18618-689 Botucatu, São Paulo State, Brazil; †Department of Botany, Universidade Federal de São Carlos, 13565-905 São Carlos, São Paulo State, Brazil; and ‡Department of Botany, Institute of Biosciences, Universidade Estadual Paulista, Botucatu, 18618-689 Botucatu, São Paulo State, Brazil Editor: Janette Steets 1 Autho Manuscript electronical Premise of research. Dichogamy is a common feature in early-divergent angiosperms, as in Piper, and is re- lated to the effectiveness of pollination. To better understand the sexual reproduction in Piper species, we deter- mined whether there is a pattern of pollen release and stigmatic exposure in Neotropical species of this genus, independent of flower sexuality and the number of stamens and carpels. Methodology. We studied 16Neotropical Piper species, which have flowers with different numbers of stamens and carpels. We analyzed the floral events, in the field and in the laboratory, from bud to senescent flowers. Pivotal results. Twelve species had only bisexual flowers, and four species had bisexual plus staminate flowers. The bisexual flowers of all the analyzed species showed incomplete protogyny: the pistillate phase of flowers lasted from 2 to 7 d, and the bisexual phase lasted from 1 to 9 d. The stigmas were long-lived (4–16 d), and the stigmatic papillae were exposed sequentially and gradually in a basipetal direction, regardless of the number of carpels in the flower. Anther dehiscence was asynchronous and sequential, occurring in one stamen at a time, regardless of the number of stamens and flower sexuality. Conclusions. The dynamic of floral events (sequential and gradual exposure and senescence of stigmas in a basipetal direction, and asynchronous and sequential pollen release) suggests a pattern for Neotropical Piper spe- cies, considering that the analyzed species belong to different clades of the genus. In addition, these characteristics may also represent a pattern for the genus as a whole, as the floral development and morphology of bisexual and unisexual flowers are similar. Keywords: asynchronous pollen release, dichogamy, incomplete protogyny, stigmatic longevity. Introduction Dichogamy, mainly protogyny, is a common feature in early- divergent angiosperms, is related to the effectiveness of pollina- tion (Endress 2010), andmay influence the population genetic di- versity (Li et al. 2016). Additionally, protogyny may act as an effective mechanism to prevent autogamy in self-compatible spe- cies (Lloyd and Webb 1986; Bertin and Newman 1993). Never- theless, protogyny does not prevent geitonogamy if flowers in distinct phases occur in a plant (Pang and Saunders 2014 and references therein). Because Piper species are pollinated by small insects that search for pollen, and dichogamy is a common fea- ture in this genus (Figueiredo and Sazima 2000; Valentin-Silva 2017), it is essential to know the dynamic offloral events to better understand the sexual reproduction in these species, especially the period when pollen is available to pollinators and the period during which pollen can be deposited on a receptive stigma. r for correspondence; e-mail: eguimaraes@ibb.unesp.br. received October 2017; revised manuscript received October 2017; ly published March 23, 2018. 287 This content downloaded from 186. All use subject to University of Chicago Press Terms The genus Piper has pantropical distribution, and species dif- fer in the sexuality of their flowers: Neotropical species have bi- sexual flowers, and Paleotropical species have unisexual flowers (Jaramillo et al. 2008). However, functionally unisexual flowers (staminate) were observed in different populations of four Neo- tropical Piper species (Figueiredo and Sazima 2000; Valentin- Silva 2017). Despite this, the bisexual and unisexual flowers are morphologically similar: perianthless, protected by a bract, with 1–10 stamens and 3–4 carpels (Jaramillo and Manos 2001; Valentin-Silva 2017). Most studied species of this genus are protogynous, and stig- mas are long-lived (Menon 1949; Martin and Gregory 1962; Figueiredo and Sazima 2000; Valentin-Silva et al. 2015). Infor- mation about the exposuremode of stigmatic papillae is restricted to the study of Valentin-Silva et al. (2015), which showed that the stigmatic papillae are exposed sequentially and gradually in a basipetal direction inP. vicosanumYunck.Ontogenetic studies inPiper species with four and six stamens showed that the lateral pairs of stamens develop simultaneously while the median pair develops sequentially, with the anterior-median stamen being formed before the posterior-median stamen (Tucker 1982). Pol- len release alsooccurs asynchronously, in one stamenat a time, as 217.236.064 on May 31, 2019 09:38:31 AM and Conditions (http://www.journals.uchicago.edu/t-and-c). 288 INTERNATIONAL JOURNAL OF PLANT SCIENCES described by Valentin-Silva et al. (2015) in P. vicosanum, a spe- cies with four stamens. However, the pollen release of the first two stamens is not simultaneous, differing from the sequence of stamen development proposed by Tucker (1982); the other two stamens release pollen sequentially on subsequent days. Thus, we aimed to determine whether there was a pattern of pollen release and stigmatic exposure in Neotropical Piper spe- cies.Wedescribed these floral events in 16Piper species that have different numbers of stamens and carpels, and we also described the floral events in different floral types (bisexual and staminate flowers). Since bisexual and unisexual flowers ofPiper species are similar in floral development andmorphology (Tucker 1982; Lei and Liang 1998; Jaramillo et al. 2008), we expected that pollen release and stigmatic exposure would be similar among species and, in specieswith twofloral types, the dynamicof pollen release would be similar between bisexual and staminate flowers. Material and Methods We carried out this study fromApril 2014 to December 2015 in the Station of Research, Environmental Training, and Educa- tionMatadoParaíso, a forest reserve inViçosa (lat. 2074702480S, long. 4275002520W), Minas Gerais State, in southeastern Brazil. This reserve has 194 ha, from 690 to 870 m above sea level, and its vegetation is classified as seasonal semideciduous montane forest (Veloso et al. 1991). We studied 16 Piper species (table 1). Most species have in- florescence of the spike type, with bisexual flowers, sessile, an- droecium with four stamens, three-carpellate gynoecium (the style is usually absent), with superior ovary, which is uniovular (Yuncker 1972, 1973). The studied species belong to different clades of the genus (table 1), representing all clades of Neotrop- ical speciesof thegenusexcept for theEnckeaclade,whichhas spe- cies with restricted distribution in Central America (Jaramillo et al. 2008). Despite this, the floral morphology of these species is similar to that of the Ottonia clade species (Jaramillo et al. 2008). We deposited voucher specimens in BOTU (table 1). This content downloaded from 186.2 All use subject to University of Chicago Press Terms We used four transects in the area (total of 5 km) and marked the individuals that had reproductive structures (inflorescences with floral buds, flowers, or fruits) or inflorescence scars, as sug- gestedbyValentin-Silva andVieira (2015).Whenpossible,we ran- domly sampled 15 individuals from all marked plants of each spe- cies (table 1) using the sample function of R (R Core Team 2016) to analyze theirflowers in the fieldwith a hand lens (#60 increase) and in the laboratory under a stereomicroscope (fresh or stored in 70% ethanol), to verify the presence of different floral types and characterize the dynamics of pollen release and stigmatic exposure. We made daily observations during the flowering period of each species (based on phenology data in Valentin-Silva 2017) in two inflorescences per individual (totaling 640 h of field observation, or about 40 h per species). Monitoring the exposure of stigmas, which have papillae only on their ventral surfaces, serves as an indication of stigma receptivity (Valentin-Silva et al. 2015). This information, when associated with the period of pollen release, may help in determining the type of dichogamy, when present. The floral events of all 16 species were investigated and char- acterized in the field. As the sequence of events was the same in all of the 16 studied species (which had from two to six stamens), we decided to use the species with the highest number of stamens among the analyzed species (P. amplum; see table 1) to illustrate the sequence of events that happens in all of them. For this, we collected samples of P. amplum from flower buds to senescent flowers.We fixed the samples in FAA (formaldehyde, acetic acid, and 50% ethanol, 1∶1∶18 v/v/v) for 48 h and then stored them in 70% ethanol (Johansen 1940). Thereafter, we dehydrated the samples in an ethanol series and subjected them to critical-point drying using CO2 and metal deposition with gold (Robards 1978). We examined and captured images with a scanning elec- tron microscope at the Electron Microscopy Center of the Insti- tute of Biosciences of Botucatu, Universidade Estadual Paulista. Results Twelve species had only bisexual flowers, and four species had bisexual plus staminate flowers. The dynamic of floral events was Table 1 Piper Species Studied and Clade to Which They Belong, Type of Inflorescence, Number of Stamens and Stigmas in Flower, and Number of Sampled Individuals Species Clade Inflorescence No. stamens 17.236.064 on M and Conditions ( No. stigmas ay 31, 2019 09:3 http://www.journ No. individuals 8:31 AM als.uchicago.edu/t- Voucher specimens P. aduncum L. Radula Spike 4 3 13 32,180; 32,198 P. amplum Kunth Radula Spike 6 3 11 32,187; 32,194 P. anisum (Spreng.) Angely Ottonia Raceme 4 4 12 32,188 P. arboreum Aubl. Macrostachys Spike 4 3 10 32,181 P. caldense C. DC. Peltobryon Spike 4 3 15 32,195; 32,201 P. cernuum Vell. Macrostachys Spike 4 3 14 32,172; 32,200 P. chimonanthifolium Kunth Radula Spike 4 3 10 32,196; 32,199 P. corcovadensis (Miq.) C. DC. Ottonia Raceme 4 4 15 32,191 P. crassinervium Kunth Radula Spike 4 3 11 32,175; 32,177 P. gaudichaudianum Kunth Radula Spike 4 3 12 32,175; 32,177 P. hispidum Sw. Radula Spike 4 3 15 32,174; 32,184 P. lucaeanum Kunth Schilleria Spike 3 3 15 32,182; 32,193 P. malacophyllum (C. Presl) C. DC. Radula Spike 4 3 10 32,178; 32,185 P. mollicomum Kunth Radula Spike 4 3 8 32,179; 32,190 P. pubisubmarginalum Yunck. Schilleria Spike 3 3 15 32,192 P. umbellatum L. Pothomorphe Umbel of spikes 2 3 15 32,176 and-c). VALENTIN-SILVA ET AL.—DYNAMIC OF FLORAL EVENTS IN PIPER 289 similar in bisexualflowers of all the 16 studied species, regardless of the number of stamens and carpels in the flower (fig. 1). Bisexual flowers showed incomplete protogyny: the pistillate phase of flow- ers lasted 2–7 d (table 2) and varied among flowers of the same in- dividual, and the bisexual phase of flowers lasted 1–9 d (table 2). The stigmas were long-lived (4–16 d), and the stigmatic papillae were exposed sequentially and gradually on a basipetal direction, both in species with three stigmas and in those with four. Anther dehiscence was asynchronous and sequential, occurring in one stamen at a time; the pollen release sequence was similar among species, regardless of the number of stamens in the flower (fig. 1). The onset of stigmatic senescence varied, among both species and inflorescences of the same species. We noted that the stigmatic senescence, characterized by darkening of the stigmas, could start at the end of the pistillate phase or during the bisexual phase of the flowers in all 16 investigated species (fig. 2). Therefore, it is possible that intraspecific and interspecific variations occur in the duration of the analyzed floral events but not in their dy- namics. We observed two floral types (bisexual and staminate flowers) inPiper arboreum, P. caldense, P. cernuum, and P. chimonanthi- folium. The functionally unisexual flowersweremorphologically similar to bisexual flowers: they had a bract, four stamens with dithecous anthers, and a pistillode (reduced and nonfunctional pistil). In these flowers, the anther dehiscence was also asynchro- nous and sequential, occurring in one stamen at a time, and the period of pollen release was similar to that observed for the bisex- This content downloaded from 186. All use subject to University of Chicago Press Terms ual flowers (table 2). We observed no changes in the pistillode throughout the anthesis period, and these flowers did not bear fruit. In bisexual flowers of P. amplum, the bract fully covered the androecium and gynoecium during the flower bud stage. As the bract began to open, the apical portion of the anthers of three stamens—stamens 1, 2 (the posterior-lateral pair), and 3 (the anterior-median stamen, located between the first two; fig. 3A)— became visible. At that time, stamen 4 (the posterior-median sta- men, located between the bract and the ovary) was still hidden by the bract; stamens 5 and 6 (the anterior-lateral pair) were also not visible. On average, 7 d after the beginning of bract opening, the juxtaposed and upright stigmas became visible among the stamens (fig. 3B). Stamen 4 became visible, as well as stamens 5 and 6 (fig. 3B), due to the bract opening. Anthesis began with the receptivity of the stigmas, character- ized by exposure of turgid papillae at the distal portion (located on the ventral surface of stigmas; fig. 3C), while the median and proximal portions of the stigmas remained juxtaposed. Anthers of all stamens remained indehiscent, so the species was proto- gynous due to the onset of stigmatic receptivity before pollen re- lease. The pistillate phase of flowers lasted from 2 to 7 d and varied among flowers of the same individual. Pollen release began in flowers that still showed receptive stig- mas, indicating incomplete protogyny and characterizing the bi- sexual phase. Pollen was released during the hottest hours of the day, usually between 10:00 a.m. and 2:00 p.m., forming clumps on the anther, which could remain on it for more than 24 h. Stamens 1 and 2 were the first ones to release pollen, but asyn- chronously. The stamen that released pollen became higher than the other stamen, due to the total exposure of filament and an- ther; both stamens 1 and 2 can be the first to release pollen. These stamens released pollen on the same day but at different times, with at least 1 h of difference (figs. 3D, 4). Another observation was the pollen release on consecutive days, usually in the morn- ing, about 24 h apart (figs. 3E–3H, 4). On that occasion, the stig- mas were farther apart from each other, exposing the papillae of the distal and median portions of the stigmas. Besides, the stig- matic papillae located in the distal portion of the stigmas could be darkened (fig. 3E–3H), indicating the onset of stigmatic senes- cence on this portion. One day after pollen release by stamens 1 and 2, the abscission of the anther and part of the filament of these stamens generally 217.23 and C Table 2 Duration of Pistillate and Bisexual Phases in the Bisexual Flowers, According to the Number of Stamens in the Flower, and Duration of Anthesis in the Staminate Flowers No. stamens 6.064 on May 31, 201 onditions (http://www Pistillate phase 9 09:38:31 AM .journals.uchicago.edu/ Bisexual phase 6 2–7 4–9 4 2–5 3–6 4 . . . 3–6 3 2–4 2–4 2 3–4 1–2 Note. Underlining is used to differentiate staminate flow- ers from bisexual flowers. Fig. 1 Pollen release sequence in 16 Piper species with six stamens (P. amplum), four stamens (P. aduncum, P. anisum, P. arboreum, P. cal- dense, P. cernuum, P. chimonanthifolium, P. corcovadensis, P. crassiner- vium, P. gaudichaudianum, P. hispidum, P. malacophyllum, and P. mol- licomum), three stamens (P. lucaeanum and P. pubisubmarginalum), and two stamens (P. umbellatum). Stamens with two numbers have simulta- neous development (as referenced in Tucker 1982), forming a pair; de- spite this, pollen release is asynchronous, and any of the stamens in the pair can be the first to release pollen. 1/2p lateral pair (posterior-lateral pair in species with six stamens); 3 p anterior-median stamen; 4 p posterior-median stamen; 5/6 p anterior-lateral pair; trianglep gynoe- cium; curved line p bract. t-and-c). 290 INTERNATIONAL JOURNAL OF PLANT SCIENCES occurred (fig. 3I). On that occasion, stamen 3 was more exposed and could release pollen (fig. 5A) or remain indehiscent and re- lease pollen on the subsequent day (fig. 4). The stigmas were al- most completely exposed, and the papillae of the distal andmedian portions were already darkened in part of the flowers in the inflo- rescence. At this stage, the extent of the stigmas darkened was variable among the flowers. One day after pollen release by stamen 3, three stamens with indehiscent anthers remained in the flower (fig. 5B). On that oc- casion, stamen 4 could release pollen (fig. 5C) or remain indehis- cent and release pollen on the subsequent day, presenting a var- iation similar to that observed for stamen3 (fig. 4). Stamens 5 and 6 were more visible at this time (fig. 5C) due to the abscission of stamens that already released pollen, and the stigmas usually had nondarkened papillae on just the proximal portion. One day after pollen release by stamen4, stamens 5 and6were more exposed and could release pollen (fig. 5D). The variation in pollen release was similar to that observed in stamens 1 and 2 (figs. 4, 5E–5I). At that time,most of the flowers had stigmas that were totally darkened, indicating the end of stigmatic receptivity; however, a few flowers could still have receptive stigmas. Anther dehiscence asynchrony led to variations in the extension of the pollen release period among the stamens, whichmay last 4– 9 d (fig. 4). This asynchrony was related to the pollen release by one stamen at a time, on different days. Besides this, there was the possibility of stamens releasing pollen on consecutive days or possibly having an interval of 1 d until the next stamen released pollen. The combination of these different possibilities is shown in figure 4. After the abscission of stamens 5 and 6, the stigmas were This content downloaded from 186. All use subject to University of Chicago Press Terms collapsed. Then, ovary size increased during fruiting (fig. 4J) if the flower was fertilized. Discussion Incomplete protogyny, as observed here, can favor cross- pollination during the pistillate phase of flowers but does not prevent the occurrence of self-pollination during the bisexual phase of flowers (Lloyd and Webb 1986; Bertin and Newman 1993; Endress 2010; Li et al. 2016). Thus, other reproductive mechanisms may be associated with dichogamy, preventing au- togamy—self-incompatibility, for example. On the one hand, this combination of incomplete protogyny and self-incompatibility was observed in 13 Piper species in southeastern Brazil (Figuei- redo and Sazima 2000; Valentin-Silva 2017). On the other hand, the combination of incomplete protogyny and self-compatibility is also common (11 species; Figueiredo and Sazima 2000; Valentin-Silva 2017). In this case, autogamy or geitonogamy may occur, depending on the visiting behavior of the pollina- tors (Lloyd and Schoen 1992). Even if dichogamy may vary depending on the environmental conditions and geographical limits of each species (Lloyd and Bawa 1984), the dynamic of floral events can occur in a similar way in these Piper species, as well as in a possible adichogamous species such as P. regnellii (Miq.) C. DC. (Figueiredo and Sazima 2000). Gradual and sequential exposure of stigmatic papillae is re- lated both to the incomplete protogyny and to the longevity of the stigmas (Valentin-Silva et al. 2015), characteristics ob- served in all analyzed species. Long-lived stigmas are common Fig. 2 Stigmatic senescence inPiper chimonantifolium.A, Flower in the pistillate phase;B, onset of stigmatic senescence on distal portion;C, flower in the bisexual phase with partially senescent stigmas; observe an anther releasing pollen (arrow). brp bract; stp stamen. Scale bars p 500 mm. 217.236.064 on May 31, 2019 09:38:31 AM and Conditions (http://www.journals.uchicago.edu/t-and-c). Fig. 3 Flowers of Piper amplum as buds, in the pistillate phase and at the beginning of the bisexual phase. A, Onset of stamens’ exposure; B, floral bud with stigmas among the anthers (arrow); C, partial exposure of the six stamens and the papillae on the distal portion of stigmas; D, pollen release by stamens 1 and 2 on the same day; E, F, pollen release by stamens 1 and 2 on subsequent days, starting with stamen 1; G, H, pollen release by stamens 1 and 2 on subsequent days, starting with stamen 2; I, flower after pollen release by stamens 1 and 2. The bract apex is directed toward the distal region of the spike. br p bract; fs p filament scar; s1–s6 p stamens 1–6. Scale bars p 200 mm. This content downloaded from 186.217.236.064 on May 31, 2019 09:38:31 AM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 292 This content downloaded from 186.217.236.064 on May 31, 2019 09:38:31 AM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). t Fig. 4 Diagrams of Piper amplum flowers with all the possible pollen-release sequences from the six stamens. Open circles p indehiscen anther; filled circles p pollen-releasing anther; dots p filament scar; triangles p gynoecium; curved lines p bract. Fig. 5 Flowers of Piper amplum at the end of the bisexual phase. A, Pollen release by stamen 3 and stigmas almost completely exposed; B, flower after pollen release by stamen 3; C, pollen release by stamen 4 and turgid papillae on the proximal portion of stigmas; D, flower after pollen release by stamen 4; E, pollen release by stamens 5 and 6 on the same day; F, G, pollen release by stamens 5 and 6 on subsequent days, starting with stamen 5; H, I, pollen release by stamens 5 and 6 on subsequent days, starting with stamen 6 (observe the collapsed stigmas); J, immature fruit. The bract apex is directed toward the distal region of the spike. br p bract; fs p filament scar; ov p ovary; s3–s6 p stamens 3– 6. Scale bars p 200 mm. 293 This content downloaded from 186.217.236.064 on May 31, 2019 09:38:31 AM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 294 INTERNATIONAL JOURNAL OF PLANT SCIENCES in Piper species (Figueiredo and Sazima 2000; Valentin-Silva et al. 2015), including the Paleotropical species P. nigrum L. (Menon 1949; Martin and Gregory 1962). This characteristic has been related to the greater probability of stigmas receiving enough pollen grains (Cruden 2000), which may favor sexual reproduction. The mode of exposure for stigmatic papillae also seems to be common in the species of this genus, as demonstrated herein. It was observed in species with four stigmas, the ancestral condition in Piperaceae (Jaramillo et al. 2004), and in species with three stigmas. Based on this, the mode of stigmatic papillae exposuremay be phylogenetically conserved because it was inde- pendent of the number of carpels in the flower. Furthermore, it is possible that this mode of stigmatic exposure also occurs in the pistillate flowers of Paleotropical Piper species, which have gynoe- cium development similar to that described for the bisexual flowers of Neotropical species (Lei and Liang 1998). Due to the sequential and gradual exposure of stigmas, a set of receptive papillae is exposed in a basipetal direction, suggesting that the stigmas are functionally fragmented (Valentin-Silva et al. 2015). Moreover, there seems to be a functional separation be- tween portions of the stigmas. The stigmatic papillae of the distal portion should be associatedwith pollen received during the pistil- late phase of the flower, so coming from other flowers, favoring xenogamyor geitonogamy, dependingon the dynamics of opening of flowers in the inflorescence and in the plant. After the onset of the bisexual phase of the flower and the beginning of the stigmatic senescence, which is gradual, themedian and proximal portions of the stigmas can receive, in addition to pollen from other plants, their own pollen, featuring self-pollination. Stigmatic senescence occurred similarly to the exposure of stig- matic papillae, sequentially and gradually in a basipetal direction, as described by Valentin-Silva et al. (2015). The darkening of the stigmas observed in the analyzed species may be related to the ac- cumulation of phenolic compounds, as recorded in P. vicosanum (Valentin-Silva et al. 2015). The variation in the time of onset of stigmatic senescence may be related to the pollinator visits to the flowers of each species, considering that stigmas not pollinated may remain receptive for a longer time than pollinated stigmas (Devlin and Stephenson 1984; Richardson and Stephenson 1989). Pollen releasewas asynchronous, as it occurred in one stamen at a time, even in the lateral pairs of simultaneously developed stamens (Tucker 1982), andwas sequential, coinciding with the bilateral initiation sequence of stamens. Although the number This content downloaded from 186.2 All use subject to University of Chicago Press Terms of stamens was homoplastic in Piperales (Jaramillo et al. 2004), the asynchronous and sequential pollen release was observed in all species, similar to that recorded in P. vicosanum (Valentin- Silva et al. 2015), andwas independent of the number of stamens in the flower. Nevertheless, the greater the number of stamens on a flower, the longer the pollen release period. Thus, the offering period of floral resource (pollen) is increased, for example, in spe- cies with six stamens, which can increase the chances of cross- pollination (Li et al. 2016). Furthermore, pollen release in the four species with staminate flowers was also asynchronous and sequential. This indicates that the asynchrony in pollen release is independent of the pres- ence of the pistil and can also occur in staminate flowers of Pa- leotropical Piper species, which have androecium development similar to that described for the bisexual flowers of Neotropical species (Lei and Liang 1998). The characteristics observed here (sequential and gradual ex- posure and senescence of stigmas in a basipetal direction, and asynchronous and sequential pollen release) suggest a pattern for Neotropical Piper species, considering that the analyzed species belong to different clades of the genus. Even though we observed variations in the duration of floral events, the 16 stud- ied species showed no variations in the dynamics of floral events, reinforcing the idea that it represents a pattern for these clades. There are no similar studies with Paleotropical Piper species, so it is not possible to do direct comparisons with these species. Despite this, as the floral development and morphology of bi- sexual and unisexual flowers are similar between Neotropical and Paleotropical Piper species, the characteristics described herein may represent a pattern for the genus as a whole. Acknowledgments We thank the Coordination for the Improvement of Higher Education Personnel and the BrazilianNational Council for Sci- entific and Technological Development for the scholarships for a doctor’s degree and research productivity grant (305912/2013- 5), respectively, to A. Valentin-Silva andM. A. Batalha. We also thank the Electron Microscopy Center of the Institute of Bio- sciences of Botucatu at the Universidade Estadual Paulista. We are indebted to Elsie Franklin Guimarães and Micheline Carvalho-Silva for their assistance with plant identification. Literature Cited Bertin RI, CM Newman 1993 Dichogamy in angiosperms. Bot Rev 59:112–152. Cruden RW 2000 Pollen grains: why so many? Plant Syst Evol 222: 143–165. Devlin B, AG Stephenson 1984 Factors that influence the duration of the staminate and pistillate phases of Lobelia cardinalis flowers. Bot Gaz 145:323–328. Endress PK 2010 The evolution of floral biology in basal angiosperms. Philos Trans R Soc B 365:411–421. Figueiredo RA, M Sazima 2000 Pollination biology of Piperaceae species in southeastern Brazil. Ann Bot 85:455–460. Jaramillo MA, R Callejas, C Davidson, JF Smith, AC Stevens, E Tepe 2008 A phylogeny of the tropical genus Piper using ITS and the chloroplast intron psbJ-petA. Syst Bot 33:647–660. Jaramillo MA, P Manos 2001 Phylogeny and patterns of floral di- versity in the genus Piper (Piperaceae). Am J Bot 88:706–716. Jaramillo MA, PS Manos, EA Zimmer 2004 Phylogenetic relation- ships of the perianthless Piperales: reconstructing the evolution of floral development. Int J Plant Sci 165:403–416. Johansen DA 1940 Plant microtechnique. McGraw-Hill, New York. Lei LG, HX Liang 1998 Floral development of dioecious species and trends of floral evolution in Piper sensu lato. Bot J Linn Soc 127: 225–237. Li PS, CC Pang, RMK Saunders 2016 Contrasting pollination ecology of Disepalum species (Annonaceae): evolutionary loss of the floral chamber and partial breakdown of protogyny associated with a shift in pollination system. Bot J Linn Soc 182:708–718. 17.236.064 on May 31, 2019 09:38:31 AM and Conditions (http://www.journals.uchicago.edu/t-and-c). https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1006%2Fanbo.1999.1087&citationId=p_7 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&pmid=11302858&crossref=10.2307%2F2657072&citationId=p_9 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1111%2Fboj.12468&citationId=p_13 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1007%2FBF00984100&citationId=p_4 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&pmid=20047868&crossref=10.1098%2Frstb.2009.0228&citationId=p_6 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&system=10.1086%2F382803&citationId=p_10 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1600%2F036364408786500244&citationId=p_8 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1007%2FBF02856676&citationId=p_3 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&system=10.1086%2F337462&citationId=p_5 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&system=10.1086%2F337462&citationId=p_5 VALENTIN-SILVA ET AL.—DYNAMIC OF FLORAL EVENTS IN PIPER 295 Lloyd DG, KS Bawa 1984 Modification of the gender of seed plants in varying conditions. Evol Biol 17:255–338. Lloyd DG, DJ Schoen 1992 Self- and cross-fertilization in plants. I. Functional dimensions. Int J Plant Sci 153:358–369. Lloyd DG, CJ Webb 1986 The avoidance of interference between the presentation of pollen stigmas in angiosperms I. Dichogamy. N Z J Bot 24:135–162. Martin FW, LE Gregory 1962 Mode of pollination and factors af- fecting fruit set in Piper nigrum L. in Puerto Rico. Crop Sci 2: 295–299. Menon KK 1949 The survey of pollu and root diseases of pepper. Indian J Agric Sci 19:89–136. Pang CC, RMK Saunders 2014 The evolution of alternative mech- anisms that promote outcrossing in Annonaceae, a self-compatible family of early-divergent angiosperms. Bot J Linn Soc 174:93– 109. RCore Team 2016 R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vienna. https://www .R-project.org/. Richardson TE, AG Stephenson 1989 Pollen removal and pollen de- position affect the duration of the staminate and pistillate phases in Campanula rapunculoides. Am J Bot 76:532–538. This content downloaded from 186. All use subject to University of Chicago Press Terms Robards AW 1978 An introduction to techniques for scanning elec- tronmicroscopy of plant cells. Pages 343–403 in JLHall, ed. Electron microscopy and cytochemistry of plant cells. Elsevier, New York. Tucker SC 1982 Inflorescence and flower development in the Pipera- ceae. III. Floral ontogeny of Piper. Am J Bot 69:1389–1401. Valentin-Silva A 2017 Fenologia e biologia reprodutiva de espécies de Piper (Piperaceae) em floresta estacional semidecídua: aspectos ecológicos e filogenéticos. PhD diss. Universidade Estadual Paulista, Botucatu, São Paulo. Valentin-Silva A, VPM Coelho, MC Ventrella, MF Vieira 2015 Tim- ing of pollen release and stigma receptivity period of Piper vicosanum (Piperaceae): new insights into sexual reproduction of the genus. Am J Bot 102:626–633. Valentin-Silva A, MF Vieira 2015 Phenology of two co-occurring Piper (Piperaceae) species in Brazil. Aust J Bot 63:581–589. Veloso HP, ALR Rangel-Filho, JCA Lima 1991 Classificação da ve- getaçãobrasileira, adaptada a umsistemauniversal. InstitutoBrasileiro de Geografia e Estatística, Rio de Janeiro. Yuncker TG 1972 The Piperaceae of Brazil. I. Piper: group I, II, III, IV. Hoehnea 2:19–366. ——— 1973 The Piperaceae of Brazil. II. Piper: group V; Ottonia; Pothomorphe; Sarcorhachis. Hoehnea 3:29–284. 217.236.064 on May 31, 2019 09:38:31 AM and Conditions (http://www.journals.uchicago.edu/t-and-c). https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1002%2Fj.1537-2197.1989.tb11344.x&citationId=p_21 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1002%2Fj.1537-2197.1982.tb13387.x&citationId=p_23 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.3732%2Fajb.1400419&citationId=p_25 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&system=10.1086%2F297040&citationId=p_15 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.3732%2Fajb.1400419&citationId=p_25 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.2135%2Fcropsci1962.0011183X000200040007x&citationId=p_17 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1111%2Fboj.12118&citationId=p_19 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1007%2F978-1-4615-6974-9_6&citationId=p_14 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1071%2FBT14332&citationId=p_26 https://www.journals.uchicago.edu/action/showLinks?doi=10.1086%2F696824&crossref=10.1080%2F0028825X.1986.10409725&citationId=p_16