See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/230654127 Divergence on floral traits and vertebrate pollinators of two endemic Encholirium bromeliads Article in Plant Biology · August 2012 DOI: 10.1111/j.1438-8677.2012.00649.x · Source: PubMed CITATIONS READS 8 78 3 authors: Alexander V. Christianini Rafaela Forzza 22 PUBLICATIONS 218 CITATIONS 131 PUBLICATIONS 642 CITATIONS Universidade Federal de São Carlos SEE PROFILE Instituto de Pesquisas Jardim Botânico do Ri… SEE PROFILE Silvana Buzato University of São Paulo 19 PUBLICATIONS 665 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Taxonomic revision of Tradescantia L. sect. Austrotradescantia D.R.Hunt (Commelinaceae) View project Systematics of Floscopa Lour. (Commelinaceae) View project All content following this page was uploaded by Alexander V. Christianini on 28 May 2014. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document and are linked to publications on ResearchGate, letting you access and read them immediately. Plant Biology ISSN 1435-8603 RESEARCH PAPER Divergence on floral traits and vertebrate pollinators of two endemic Encholirium bromeliads A. V. Christianini1, R. C. Forzza2 & S. Buzato3 1 Universidade Federal de São Carlos, Campus Sorocaba, Sorocaba, SP, Brazil 2 Jardim Botânico do Rio de Janeiro, Jardim Botânico, Rio de Janeiro, RJ, Brazil 3 Departamento de Ecologia, Instituto de Biociências, Universidade de São Paulo, São Paulo, SP, Brazil Keywords Breeding system; Bromeliaceae; herbivore; hummingbird; nectar-feeding bat; plant– animal interactions; pollination syndromes. Correspondence S. Buzato, Departamento de Ecologia, Instituto de Biociências, Universidade de São Paulo, Rua do Matão, 321, travessa 14, 05508-900 São Paulo, SP, Brazil E-mail: sbuzato@usp.br Editor A. Dafni Received: 5 December 2011; Accepted: 6 June 2012 doi:10.1111/j.1438-8677.2012.00649.x ABSTRACT Shifts in pollen vectors favour diversification of floral traits, and differences in pollination strategies between congeneric sympatric species can contribute to reproductive isolation. Divergence in flowering phenology and selfing could also reduce interspecific crossing between self-compatible species. We investigated floral traits and visitation rates of pollinators of two sympatric Encholirium species on rocky outcrops to evaluate whether prior knowledge of floral characters could indicate actual pollinators. Data on flowering phenology, visitation rates and breeding system were used to evaluate reproductive isolation. Flowering phenology overlapped between species, but there were differences in floral characters, nectar volume and concentration. Several hummingbird species visited flowers of both Encholirium spp., but the endemic bat Lonchophylla bokermanni and an unidentified sphingid only visited E. vogelii. Pollination treatments demonstrated that E. heloisae and E. vogelii were partially self-compatible, with weak pollen limitation to seed set. Herbivores feeding on inflorescences decreased reproductive output of both species, but for E. vogelii the damage was higher. Our results indicate that actual pollinators can be known beforehand through floral traits, in agreement with pollination syndromes stating that a set of floral traits can be associated with the attraction of specific groups of pollinators. Divergence on floral traits and pollinator assemblage indicate that shifts in pollination strategies contribute to reproductive isolation between these Encholirium species, not divergence on flowering phenology or selfing. We suggest that hummingbird pollination might be the ancestral condition in Encholirium and that evolution of bat pollination made a substantial contribution to the diversification of this clade. INTRODUCTION Most of angiosperm diversification is associated with variation in floral traits, which are mainly explained as the result of natural selection imposed by interactions with pollen vectors (Feinsinger 1983; Fenster et al. 2004; Johnson 2006; but see complementary view in Gallen 1999; Cariveau et al. 2004). Studies in Neotropical angiosperms have reported frequent evolutionary shifts between pollination strategies within lineages (Perret et al. 2003; Kay et al. 2005; Duchen & Renner 2010). Such shifts in pollination strategies can play a role in reproductive isolation between congeners when in sympatry (Schemske & Bradshaw 1999; Ramsey et al. 2003). Among pollination strategies, pollination by vertebrates is less frequent than by insects (Bawa 1990; Kress & Beach 1994), but not less important for floral diversification in some angiosperms (Kay et al. 2005; Daniel et al. 2008; Fleming et al. 2009; Duchen & Renner 2010). Within vertebrates, birds play a more frequent role as pollinators than bats (Sazima et al. 1999; Fleming et al. 2009), and results of pollen delivery by birds and bats indicate that evolutionary shifts from bird to bat pollination should be common (Muchhala 360 & Thomson 2010). Information on visitation rates of congeneric plant species could help in understanding floral patterns and the contribution of pollinators to reproductive isolation of sympatric species (Kay & Schemske 2003; Muchhala 2003). A remarkable radiation associated with vertebrate pollination occurs in Bromeliaceae (Benzing 2000). Out of about 100 species studied, 78% are bird-pollinated, 21% are batpollinated and 1% are insect-pollinated. Hummingbird pollination occurs in all subfamilies, while Phyllostomid bats are found as pollinators in Tillandsioideae and Pitcairnioideae (Martinelli 1994; Sazima et al. 1999; Benzing 2000; Buzato et al. 2000; Fleming et al. 2009). Although artificial hybrids can be produced between bromeliad species in greenhouses, natural hybrids are rare under field conditions (Smith & Downs 1974; Benzing 2000; Wendt et al. 2001). Several premating isolating mechanisms are suggested to prevent hybridisation (Smith & Downs 1974; Benzing 2000). In addition, a survey of breeding systems of bromeliads indicated that 75% of species are self-compatible and about 43% set fruit spontaneously in the absence of pollen vectors (Matallana et al. 2010). Considering the possibility of interspecific crosses due to geographic and flowering overlap of related Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands Christianini, Forzza & Buzato species, selfing has been suggested as a method to avoid hybridisation between sympatric bromeliad species (Wendt et al. 2002; Matallana et al. 2010). In the Espinhaço range, a mountain chain in the Brazilian states of Minas Gerais and Bahia that contains one of the richest floras in Brazil (Echternacht et al. 2011), Bromeliaceae presents a high number of related species in sympatry (Versieux et al. 2008). Encholirium (Pitcairnioideae), an endemic genus to Brazil with a centre of diversity in the Espinhaço range, contains 25 species (Forzza & Zappi 2011; Forzza et al. 2011), all of them terrestrial on rocky outcrops (Forzza 2005). Sazima et al. (1989) provided the first field observation of pollination by bats of E. glaziovii Mez (= E. subsecundum (Baker) Mez; Forzza 2005), which also was the first record of bat pollination in the Pitcairnioideae. Using floral characters, they suggested bats as pollinators of E. vogelii Rauh and hummingbirds as pollinators of E. sazimae Rauh (= E. heloisae (L. B. Sm.) Forzza & Wand.; Forzza 2005). Based on floral traits, most Encholirium species can be regarded as chiropterophilous: brush-type, many flowered and tall inflorescence, greenish–pale yellow flowers, stamens and style placed outside a wide corolla and high production of nectar (Forzza 2005). After Sazima et al. (1989), no other investigation into plant– pollinator interaction in Encholirium species was made, and the clade remains an opportunity to explore the correspondence between floral traits and functional groups of pollinators as predicted by pollination syndromes. Because E. heloisae and E. vogelii occur in sympatry, there is also an opportunity to investigate the importance of selfing and pollinators on the reproductive isolation of species on rocky outcrops. The purpose of this study was to document flowering and fruiting phenology, floral traits, the identity and visitation rates of floral visitors and breeding system of two sympatric species of Encholirium (E. heloisae and E. vogelii). Two main questions were explored: (i) can pollinators of Encholirium be predicted based on a set of floral traits as provided in Sazima et al. (1989); and (ii) what is the contribution of flowering phenology, floral traits, pollinator specificity and breeding system to the reproductive isolation of Encholirium species in sympatry? MATERIAL AND METHODS Floral traits and vertebrate pollinators of Encholirium individuals in a given area (Fig. 1A). Although some clonal reproduction is documented for Encholirium species, genetic analyses showed that individuals have unique genotypes, and FST values are compatible with outcrossing mating systems (FST = 0.080–0.160; Cavallari et al. 2006). FST values can be evaluated using the formula FST = Vp ⁄ p(1 – p), where Vp = variance of allele frequencies among populations, standardised relative to the maximum value possible given an observed mean allele frequency (p), which can be interpreted as the proportion of genetic variation distributed among populations (Avise 2004). Flowering and fruiting phenology, and herbivory on reproductive structures From November 1997 to April 1999, individuals of E. heloisae (n = 49) and E. vogelii (n = 48) were labelled and monitored A B C D E F G Study site and species The study site is in the Serra do Cipó, a strip on the southern part of the Espinhaço Mountain chain (Echternacht et al. 2011), about 100 km NW of Belo Horizonte, MG, southeast Brazil (1915¢ S, 4330¢ W). A subtropical seasonality defines the climate, with a dry season during winter and a rainy season during summer (mean annual temperature = 18 C; annual rainfall about 1400 mm; Nimer 1989; Madeira & Fernandes 1999). Soils are dystrophic and derived mainly from quartzitic rocks (Echternacht et al. 2011). The vegetation is comprised mainly of herbs and shrubs, with most endemic species belonging to Asteraceae, Melastomataceae, Gramineae, Eriocaulaceae, Velloziaceae and Xyridaceae (Giulietti et al. 1987; Echternacht et al. 2011). Encholirium heloisae and E. vogelii occur on sandy–rocky soils, about 1100 m a.s.l. (Forzza 2005; Echternacht et al. 2011). Populations are limited to small areas within Serra do Cipó (Echternacht et al. 2011), with very few and scattered Fig. 1. Flowering individuals of Encholirium heloisae on rocky outcrops (white arrow) at Serra do Cipó (MG), southeast Brazil (A); frontal view of flowers of E. heloisae (B) and E. vogelii (C); the hummingbird Augastes scutatus on flower of E. heloisae (D) and E. vogelii (E); nighttime floral visitors of E. vogelii: the bat Lonchophylla bokermanni (F) and an unidentified Sphingid moth (G). Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands 361 Floral traits and vertebrate pollinators of Encholirium in an area of ca. 30 ha. We made visits of 3–7 days, at intervals of 2–3 weeks during the flowering period and at intervals of 2–11 weeks during fruiting period to record individuals with flowers and ⁄ or mature fruits. In addition, former observations indicated that herbivores could cause the collapse of the inflorescence before flower and fruit maturation. Therefore, at the time of the phenology survey, inflorescences with signs of herbivory, due to the presence of small holes or chewed tissue, were marked as damaged by herbivorous insects. The proportion of damaged inflorescences at each time presents a measure of the activity of herbivores on the reproductive structures of the plants. Floral features Flowering individuals of E. heloisae (n = 9) and E. vogelii (n = 8) were randomly sampled, and 30 flowers were collected for each species to investigate interspecific differences in internal length of the corolla tube, corolla diameter at opening and length of the pistil and stamens. An exploratory investigation on the presence of floral scent was made by keeping flowers of each species in closed separate glass vials for detection of fragrance. To evaluate floral display, we measured inflorescence height (from base to tip) and counted the total number of flowers in an inflorescence. Anthesis was followed in E. heloisae (n = 10 flowers from ten individuals) and E. vogelii (n = 10 flowers from four individuals). Nectar that had accumulated in bagged flowers since the bud stage was extracted with a micro-syringe. Each measurement of nectar volume was divided by the number of hours the flower remained bagged. Nectar sugar concentration in sucrose equivalents, percentage sugar = (sugar mass ⁄ total mass) · 100, was measured with a pocket refractometer (Kearns & Inouye 1993; Galetto & Bernadello 2005). Floral visitors In the years 1998 and 1999, from December to February, we recorded the identity of flower visitors on E. heloisae (n = 8 individuals) and E. vogelii (n = 3 individuals). For both species, observation sessions of up to 3 h were made systematically throughout 24 h, as the flower is in anthesis in both day and night. Sampling effort was 44.6 h for E. heloisae (21.3 h day, and 23.3 h night) and 33.3 h for E. vogelii (16.5 h day and 16.8 h night). At the time when flower visitor approached an inflorescence and probed at least one flower, a visit to that inflorescence was recorded. At the time of visits, we also counted the number of visited flowers to estimate mean number of flowers visited per hour. Christianini, Forzza & Buzato To test for the effectiveness of diurnal and nocturnal floral visitors as pollinators in E. vogelli, in January 1999 the following treatments were conducted: diurnal exclusion – flowers were kept in bags from dawn to dusk, allowing only nocturnal visitors to access the flowers; nocturnal exclusion – flowers were kept in bags from dusk to dawn, allowing only diurnal visitors to access the flowers. Before removing bags from each treatment at dawn and dusk, we waited until we saw no further diurnal or nocturnal visitors having foraging activity. Some flowers were never bagged to detect natural levels of seed production when exposed to both diurnal and nocturnal pollinators (Sahley 1996; Young 2002). Pollinator exclusion experiments were restricted to two inflorescences of E. vogelii due to the scarce inflorescence production and difficulties in controlling for the effects of herbivores on inflorescences (see below). Pollinator exclusion experiments were made only in E. vogelii because the floral biology of E. heloisae presented signs of senescence at the beginning of the night: lack of pollen grains in the anthers, darkened stigma and very low nectar volume. In addition, we made several hours of night observation to check whether bats visited the flowers. By considering signs of senescence and no records of bats at night on flowers of E. heloisae, we decided not to concentrate the effort of this kind of experiment on E. heloisae. All individuals received all pollination treatments among their flowers. Flowers in all treatments were bagged until fruit set, except for natural pollinations in the first flowering season that remained unbagged. When fruits were ripe, fruit and its seed were recorded in each treatment. Misshaped seeds were classified as non-viable and not included in the analysis. Statistical analysis Flowering and fruiting patterns were compared between E. heloisae and E. vogelii using the Kolmogorov–Smirnov two-sample test applied to the frequency of individuals bearing flowers or fruits at a given sampling date. Differences in floral traits between species were determined with the Student t-test. Fruit-set from different pollination treatments were compared using G-test. Comparisons of the number of seeds produced in different experimental pollination treatments and in the pollinator exclusion experiments were performed by means of one-way analysis of variance (anova), followed by a posterior HSD test (unequal n) in each species. Data were log-transformed to improve normality and homocedasticity when necessary. All tests followed Sokal & Rohlf (1995). RESULTS Breeding system and pollinator exclusion experiments We performed controlled pollinations in the field to determine the breeding system of both species. Buds were marked and bagged on the evening before anthesis for the following treatments: self-pollination – flowers were hand-pollinated with their own pollen; cross-pollination – flowers were handpollinated with pollen from flowers of other individual at least 30 m away. For natural pollinations, flowers were left in open-pollinated conditions. 362 Flowering and fruiting phenology, and herbivory on reproductive structures Flowering of E. heloisae and E. vogelii occurred mainly from December to April, with a blooming climax between January and February for about 15 days (Fig. 2). Both species had the same flowering pattern during the year (Kolmogorov–Smirnov test: maximum difference = 0.47, P > 0.10). Each plant of E. heloisae had 26.9 ± 1.2 flowers per inflorescence (mean ± SE, n = 35 inflorescences), and bloomed for Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands Christianini, Forzza & Buzato Floral traits and vertebrate pollinators of Encholirium Fig. 2. Flowering and fruiting phenology of Encholirium heloisae and E. vogelii from November 1997 to April 1999, and the proportion of inflorescences damaged by herbivores at Serra do Cipó (MG), southeast Brazil. 5 days (about six flowers opened per day). Encholirium vogelii had 199.6 ± 14.3 flowers per inflorescence (n = 14 inflorescences), and bloomed for 8 days (about 40 flowers opened per day; Table 1). Fruiting occurred mainly from January to April for E. heloisae and from February to July for E. vogelii. Both species had the same fruiting pattern during the year (Kolmogorov–Smirnov test: maximum difference = 0.17, P > 0.10), and the peak of fruiting occurred between February and March (Fig. 2). Many sampled rosettes of the studied species remained sterile (84% for E. heloisae and 91% for E. vogelii), while several others were damaged by herbivores. For E. vogelii, about 50% of fertile individuals had completely destroyed inflorescences (Fig. 2). Heilipodus sp. and other unidentified Cryptorhynchinae (Curculionidae) acted as sapsuckers on inflorescences, while Allorrhina menetriesi (Scarabeidae) fed on buds and flowers. Unidentified Curculionidae larvae fed on ovules and ovary tissues of both species, while grasshoppers (Orthoptera: Acrididae) fed on inflorescences of E. heloisae, destroying about 13%. Table 1. Comparison of Encholirium heloisae and E. vogelii floral traits at Serra do Cipó (MG), southeast Brazil. Data are mean (mm) ± SE, unless otherwise stated. Floral features Both bromeliad species had greenish yellow imbricated petals, forming a short corolla tube (Fig. 1B and C, Table 1). In E. heloisae flowers, the pistil was inside the corolla, while in E. vogelii, it was slightly outside the corolla. Both species had anthers placed above the pistil and outside the corolla tube (Fig. 1B and C, Table 1). Flowers of E. vogelii had a corolla ca. 30% wider and 40% longer than that of E. heloisae (Table 1). While E. heloisae produced few flowers close to the ground (maximum inflorescence height of 0.5 m), the inflorescences of E. vogelii were at least three times taller, and produced seven times more flowers than E. heloisae (Table 1). Anthesis of E. heloisae occurred at dawn (about 06:00 h) and lasted for about 28 h, while anthesis of E. vogelii occurred at dusk (about 17:00 h) and lasted for about 60 h. For both species, stigma receptivity and anther dehiscence occurred simultaneously with flower opening. Anthers and stigmas were spatially close and there was no physical barrier for self-pollen transfer within a flower. Nectar volume in floral trait E. heloisae E. vogelii comparison range of inflorescence height (m) flowers per inflorescence corolla width corolla length stamens length pistil length nectar volume (llÆh)1) sugar concentration (%) 0.2–0.5 (n = 35) 1.4–2.0 (n = 14) – 199.64 ± 14.27 (n = 14) – 26.89 ± 1.17 (n = 35) 6.32 8.05 11.49 7.37 1.6 ± ± ± ± ± 0.18 (n = 30) 0.18 (n = 29) 0.18 (n = 30) 0.36 (n = 27) 0.3 (n = 16) 32.3 ± 3.2 (n = 12) Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands 8.30 11.56 16.95 14.56 4.0 ± ± ± ± ± 0.27 (n = 30) 0.27 (n = 30) 0.33 (n = 30) 0.54 (n = 30) 1.1 (n = 6) 17.6 ± 1.9 (n = 6) t t t t t = = = = = 6.12, P < 0.001 11.06, P < 0.001 15.23, P < 0.001 11.97, P < 0.001 3.41, P < 0.01 t = 3.36, P < 0.01 363 Floral traits and vertebrate pollinators of Encholirium Christianini, Forzza & Buzato E. vogelii was almost three times higher than in E. heloisae (Table 1), while sugar concentration was higher for E. heloisae than E. vogelii (Table 1). Flowers of both species had no scent perceptible to human senses. Table 2. Fruit set related to pollination treatments in Encholirium heloisae and E. vogelii flowers at Serra do Cipó (MG), southeast Brazil. Dash indicates no data available. species treatment Floral visitors E. heloisae natural conditions (1st season)a natural conditions (2nd season)b self-pollination cross-pollination natural conditions (1st season)a natural conditions (2nd season)b self-pollination cross-pollination nocturnal exclusion diurnal exclusion We recorded only hummingbirds as flower visitors of E. heloisae (Fig. 1D). During the visits, the birds introduced their bills into the flower and contacted both stigma and anthers, with pollen grains loaded on the bills. As a whole, hummingbirds performed 2.5 ± 0.5 visits h)1 to a given inflorescence and were observed visiting 18.7 ± 4.5 flowers h)1. Chlorostilbon lucidus (0.3 ± 0.2 visits h)1; 4.7 ± 2.0 flowers h)1), Augastes scutatus (1 ± 0.3 visits h)1; 6.4 ± 2.3 flowers h)1) and Colibri serrirostris (1.2 ± 0.4 visit h)1; 7.6 ± 2.5 flowers h)1) were observed regularly on flowers. Eupetomena macroura and Phaethornis pretrei were seen sporadically. Hummingbirds had higher mean visitation frequency to inflorescences during the morning (06:00–13:00 h; 3.3 ± 0.7 visits h)1) than afternoon (16:30–18:30 h; 1.2 ± 0.2 visits h)1). We recorded one bat species, a sphingid moth and several hummingbird species as flower visitors of E. vogelii (Fig. 1E– G). As a whole, hummingbirds performed 5.4 ± 1.3 visits h)1 to a given inflorescence and were observed visiting 66.1 ± 22.5 flowers h)1. Augastes scutatus (0.7 ± 0.3 visits h)1; 12.6 ± 6.5 flowers h)1), E. macroura (1.7 ± 0.9 visits h)1; 22.9 ± 12.4 flowers h)1), and C. serrirostris (3.0 ± 0.5 visits h)1; 30.6 ± 7.0 flowers h)1) were observed regularly on flowers. Phaethornis pretrei was seen only once. After a visit to a patch of any Encholirium species in flower, C. serrirostris often perched nearby, apparently holding a territory over the patch. The patch-holding bird vocalised frequently, chasing other hummingbirds, but not the large E. macroura. Hummingbird visitation frequency tended to be higher in late afternoon (16:30–18:30 h; 6.1 ± 1.9 visits h)1) than during the morning (06:00–13:00 h; 4.1 ± 1.1 visits h)1). Only one species of bat was recorded visiting flowers of E. vogelii, Lonchophylla bokermanni (Fig. 1F). This bat arrived at nightfall and licked the nectar during short visits. Pollen was deposited on the bat’s snout. As a whole, bats performed 0.7 ± 0.6 visits h)1 to a given inflorescence and were observed visiting 2.6 ± 2.3 flowers h)1. An unidentified sphingid moth also visited the inflorescences of E. vogelii at night (1.0 ± 0.4 visits h)1). The sphingid probed with its tongue within the open flowers while hovering or landing outside the corolla (Fig. 1 G), but we never observed the moth touching the reproductive organs of the plant. Breeding system and pollinator exclusion experiments Encholirium heloisae developed fruits under self-pollination treatment, but fruit set was lower than in cross-pollinated flowers (Table 2; G = 5.21, df = 1, P = 0.023). Seeds produced from self-pollinated flowers were misshaped and smaller than those from cross-pollinated flowers. The mean number of seeds produced per fruit varied among different pollination treatments, but only the open-pollinated flowers (natural condition) in the first flowering season showed a significant decrease in the seed set. Comparison of seed set between cross-pollinated and open-pollinated flowers in the 364 E. vogelii a flower (n) fruit (n) fruit set 381 221 0.58 49 38 0.78 27 6 1575 9 5 940 0.33 0.83 0.60 48 42 0.88 14 – 16 10 8 – 9 6 0.57 – 0.56 0.60 Herbivore access to flowers was not prevented. Herbivore access to flowers was prevented by bagging flowers. b second flowering season indicated no occurrence of pollen limitation (Table 3). Encholirium vogelii developed fruits under self-pollination treatment (Table 2), but similar to E. heloisae, seeds were misshaped. Seed set varied according to the pollination treatment. Cross-pollinated flowers produced few fruits. Fruit set from a diurnal or a nocturnal exclusion pollinator treatment was lower than under natural conditions in the second season (nocturnal exclusion G = 6.50, df = 1, P = 0.011; diurnal exclusion G = 3.69, df = 1, P = 0.055, Table 3). The comparison of fruit set between unbagged flowers in natural conditions (first reproductive season) with bagged flowers in natural conditions (second season) showed that bagged flowers under natural conditions of pollination had higher fruit set than unbagged flowers (Table 2). Assuming that all other things that may influence fruit set were equal between seasons, exclusion of herbivores by bagging flowers after anthesis reduced the loss of fruits (Table 2). DISCUSSION Our study was stimulated by the suggestion of Sazima et al. (1989) that different pollination strategies would be found among Encholirium species because some species have floral traits conforming to bird pollination, while other species conformed to bat pollination syndromes. In this study we provide original information on pollination and mating of two endemic and sympatric species of Encholirium from the southern part of the Espinhaço Mountain chain of Brazil. The ecological significance of these findings is discussed, as well as a possible scenario for diversification of the Encholirium group in the region of the Espinhaço Mountain chain, southeast Brazil. Flowering and fruiting phenology, and herbivory on reproductive structures Species within Encholirium present variability in flowering pattern (Forzza 2005), which could indicate a lack of phylogenetic signs for this trait. On the other hand, E. heloisae Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands Christianini, Forzza & Buzato Floral traits and vertebrate pollinators of Encholirium Table 3. Number of seeds per fruit (mean ± SE) related to pollination treatments in flowers of Encholirium heloisae and E. vogelii at Serra do Cipó (MG), southeast Brazil. E. heloisae treatment natural conditions (1st season) natural conditions (2nd season) cross-pollination nocturnal exclusion diurnal exclusion fruit (n) 29 38 5 – – df = 2 E. vogelii seeds per fruit b 47.83 ± 3.25 63.55 ± 2.19a 71.80 ± 4.72a – – F = 11.06, P < 0.001 fruit (n) seeds per fruit 125 42 – 9 6 df = 3 131.69 ± 3.23a 125.57 ± 1.65a – 105.11 ± 0.77a 93.00 ± 10.67a F = 4.88, P < 0.01 Within columns, results followed by the same superscript are not significantly different (P < 0.05). and E. vogelii have convergence in flowering phenology to the wettest and warmest time of the year (see Madeira & Fernandes 1999 for climate data). In addition, although there is no systematic census of flowering phenology of E. subsecundum, there is evidence that this sympatric species also blooms at the same time of the year as E. heloisae and E. vogelii (Sazima et al. 1989; Forzza 2005). In the absence of phylogenetic constraints, the convergence on flowering pattern among sympatric species of Encholirium could be a result of strong climatic differences between months within a year in this region of Brazil, which could impose limits on flowering time of species (see Madeira & Fernandes 1999 for similar data). Convergence in flowering phenology of sympatric species also seems to be a common pattern for herbs in seasonal areas of the Neotropics, including those on the rocky outcrops (Rathcke & Lacey 1985; Conceição et al. 2007). This fact brings an additional challenge to be solved for species, as convergence on flowering could promote pollinator sharing and interference with outcrossing rates. Several studies provide evidence that co-flowering plants usually share pollinators and, as a result, competition among plants is frequent due to interspecific pollen transfer (Morales & Traveset 2008; Mitchell et al. 2009). On the other hand, sharing pollinators by co-flowering plants can reduce pollen limitation (Moeller 2004). In Encholirium, comparison of seed set between cross-pollinated flowers and open-pollinated flowers showed weak pollen limitation in populations of both species. This could be evidence that partial overlap in pollinator use had already been solved through divergence in floral traits and shifts in pollinator use. Anthesis of E. heloisae occurs at dawn and most hummingbird visits occur in the morning. At the end of the day, the frequency of visits of hummingbirds increases on flowers of E. vogelii, highly coincident with anthesis of this species at dusk. Also, E. vogelii depends on bats for pollen transfer. Although both species occur in a restricted area with a very low number of reproductive individuals, which is further decreased by herbivory on inflorescences, it is hard to believe that selection for floral similarity and pollinator sharing might have occurred (Macior 1971; Schemske 1981). In Encholirium, pale floral colour seems to be among the few conserved traits imposed by factors related to phylogeny, and might have little effect on pollination. Hummingbirds visit food sources regardless of their colour (Schemske & Bradshaw 1999; Buzato et al. 2000), and colour preferences are learned associa- tions with nectar amount (Healey & Hurly 2001; see Herrera et al. 2008 for additional information of colour discrimination in hummingbirds). Insects can use reproductive organs of bromeliads as food or nest sites, which could have negative effects on fecundity and survival of the plants (Ramalho et al. 2004; Winkler et al. 2005). Both studied species lost part of their reproductive effort due to the activity of herbivores, but the damage was higher in E. vogelii. Although we found no differences in the flowering and fruiting patterns between species, the amount of food resources provided by inflorescences to herbivores is higher in E. vogelii than E. heloisae. Hence, we conclude that phenotypic variation between species on this trait may affect the proportion of inflorescence damaged by herbivores. Floral features and floral visitors Based on floral features, Sazima et al. (1989) suggested that hummingbirds would be the pollinators of E. heloisae while bats would pollinate E. vogelii flowers. Our results support their suggestion, as only hummingbirds were observed on flowers of E. heloisae and bats were recorded only on E. vogelii flowers. The comparison of floral traits shows divergence between species on floral display, time of anthesis, corolla size and production of nectar. Experiments with bats and hummingbirds on artificial flowers demonstrated that differences in corolla width between bat and hummingbird flowers could be expected, as a narrow corolla could guide hummingbird bills better than wide corollas, the latter being more appropriate to fit bat snouts (Muchhala 2007). Flower features of E. heloisae cope with specialisation on hummingbirds: anthers and stigma almost completely included in a narrow floral tube, improving precision of pollen placement on the bill. In the hummingbird-pollinated species Salvia haenkei, adjustments to the floral tube, allowing little room for the bill, seem to be important for effective pollen transfer from the pollen sacs to the bird’s bill or feathers (Wester & Claben-Bockhoff 2006). On the other hand, the wider corolla of E. vogelii seems adequate to specialisation on bats, but without excluding hummingbirds. We suspect that bats cannot visit the flowers of E. heloisae, because the narrow corolla should limit their access to the nectar. In addition, flowers of the latter species were presented very close to the ground, probably restricting flight ability of bats. There is a mention Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands 365 Floral traits and vertebrate pollinators of Encholirium of bats visiting flowers at 0.5 m from the ground (Machado et al. 1998), but we suspect that this distance should be at the limit to which hovering bats can undertake manoeuvrable flight to visit flowers. At least three morphological traits are conspicuous in species of Encholirium pollinated by bats (Sazima et al. 1989; this work): wide corollas, exertion of anthers and high number of flowers per inflorescence. These morphological traits seem to be associated with increased nectar production and should be strong candidates under selection. Phylogenetic reconstruction in Bromeliaceae revealed that the divergence of Dyckia and Encholirium from the sister clade Deuterocohnia began 8.5 Ma. Subsequently, dispersal of Dyckia to the Brazilian shield occurred, and divergence of Dyckia and Encholirium was estimated about 2.4 Ma (Givnish et al. 2011). This information indicates that recent and rapid radiation of Encholirium has been in parallel with changes in vertebrate pollination strategies. A possible scenario for the diversification of Encholirium would be the ancestor of the Brazilian shield Encholirium having hummingbirds as pollinators. The occurrence of hummingbirds as pollinators of sister clades, but not bats, supports the idea that the basal pollination strategy in Encholirium should involve hummingbirds (Bernardello et al. 1991; Benzing 2000; Kessler & Krömer 2000; Vosgueritchian & Buzato 2006). Although hummingbirds are common and very active in the area, extended anthesis and nectar production by day and night might have favoured visits of bats in plants with wider corollas and ⁄ or exerted anthers. The occurrence of species with intermediate floral traits between hummingbird and bat syndromes, e.g. E. vogelii, indicates that not all events related to clade diversification have involved pronounced shifts in pollination traits. In addition, the absence of fragrance in flowers of E. vogelii could be a sign of its derived condition from hummingbird ancestors, because for birds, visual cues seem to be more important than senses of smell. Although recent research has shown that some birds can improve olfactory ability under certain ecological conditions (Steiger et al. 2009). Olfaction is considered the primary sense for long-distance detection of flowers by bats (Winter & Helversen 2001), and most bat flowers have a strong smell to humans, a trait reported for the bat-pollinated E. subsecundum (Sazima et al. 1989). Validation of suggested inferences requires additional information on pollination strategies mapped onto a species-level phylogeny. Christianini, Forzza & Buzato populations pattern (Cavallari et al. 2006). Unlike other Bromeliaceae species in inselbergs (Wendt et al. 2001), there is no occurrence of hybrid swarms between Encholirium species in sympatry. We consider that this result is associated with specialisation for pollination by either hummingbirds or bats. However, as pollen transfer is also possible between pairs of species due to partial overlap in pollinator use, post-pollination reproductive isolation might also be present (Yost & Kay 2009). Natural history and ecology of Encholirium: a candidate clade for studies of plant diversification on rocky outcrops The occurrence of Encholirium species mainly associated with the Espinhaço Mountain chain, their differences in range distribution and recent radiation make them a strong candidate to be used as a study system to examine process involved in plant speciation within this mega-diversity area of Brazil (Echternacht et al. 2011). Therefore, the construction of a species-level phylogeny and a plan to obtain additional information on the natural history of Encholirium species are highly recommended. Based on data presented here, we anticipate that biotic interactions contributed to adaptive divergence and reproductive isolation in this genus. In addition, an overlap of information on topology and climate oscillations on species distribution could suggest the importance of historical processes on range shifts and diversification of the species. Data on floral visitors of sympatric species (Sazima et al. 1989; this study) indicate that specialisation on pollination strategies could have contributed to the occurrence of reproductive isolation among Encholirium species, instead of selfing (but see Wendt et al. 2002; Matallana et al. 2010 for other results of Bromeliaceae on rocky outcrops). Frequency of visits indicates the importance of hummingbirds and bats to pollen dispersal differs among species. There are no visits of bats to E. heloisae flowers, and hummingbirds do not contact anthers and stigmas when visiting E. subsecundum flowers (Sazima et al. 1989). In E. vogelii, temporal differences in use of flowers by hummingbirds might contribute to assortative pollen delivery between E. heloisae and E. vogelii. Based on a description of floral traits of other species (Forzza 2005), we consider that the evolution of bat pollination has made a substantial contribution to diversification of the Encholirium lineage. Breeding system and pollinator exclusion experiments ACKNOWLEDGEMENTS Selfing is widespread among Pitcairnioideae bromeliads (Martinelli 1994; Bush & Beach 1995; Wendt et al. 2001, 2002), and self-pollination in Encholirium is facilitated by the position of the anthers and stigma. However, despite fruit development in both Encholirium species under self-pollination, post-fertilisation barriers might be in action due to the abortion of seeds. As reported for other Encholirium species (Cavallari et al. 2006), further studies are needed to confirm that gene flow between individuals within a given population predominates, and E. heloisae and E. vogelii might have outcrossing mating systems. In addition, spatial isolation among populations might have contributed to the genetic structured We thank S.A. Vanin for the identification of beetles, and the Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio) for allowing us to work in areas under its responsibility. We dedicate this paper to the late Prof. Dr. W. Rauh for his studies in Encholirium and to the late Mrs. A. Bendazoli for her relentless dedication in working for the protection of Brazilian wildlife. We thank three anonymous reviewers for helpful comments on the manuscript. This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP 97 ⁄ 10582-2, 97 ⁄ 13341-6). Rafaela C. Forzza is a research fellow supported by Conselho Nacional de Desenvolvimento Cientı́fico e Tecnológico (CNPq). 366 Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands Christianini, Forzza & Buzato REFERENCES Avise J.C. (2004) Molecular markers, natural history, and evolution. Sinauer Associates, Sunderland, MA, USA: 684 pp. Bawa K.S. (1990) Plant–pollinator interactions in tropical rain forests. Annual Review of Ecology and Systematics, 21, 399–422. Benzing D.H. (2000) Bromeliaceae: profile of an adaptive radiation. Cambridge University Press, London, UK, 690 pp. Bernardello L., Galetto L., Juliani H.R. (1991) Floral nectary structure and pollinators in some Argentinian Bromeliaceae. Annals of Botany, 67, 401–411. Bush S.P., Beach J.H. (1995) Breeding systems of epiphytes in a tropical montane wet forest. Selbyana, 16, 155–158. Buzato S., Sazima M., Sazima I. (2000) Hummingbird-pollinated floras at three Atlantic forest sites. Biotropica, 32, 824–841. Cariveau D., Irwin R.E., Brody A.K., Garcia-Mayeya L.S., von der Ohe A. (2004) Direct and indirect effects of pollinators and seed predators to selection on plant and floral traits. Oikos, 104, 15–26. Cavallari M.M., Forzza R.C., Veasey E.A., Zucchi M.I., Oliveira G.C.X. (2006) Genetic variation in three endangered species of Encholirium (Bromeliaceae) from Cadeia do Espinhaço, Brazil, detected using RAPD markers. Biodiversity and Conservation, 15, 4357–4373. Conceição A.A., Funch L.S., Pirani J.R. (2007) Reproductive phenology, pollination and seed dispersal syndromes on sandstone outcrop vegetation in the ‘‘Chapada Diamantina’’, northeastern Brazil: population and community analyses. Revista Brasileira de Botânica, 30, 475–485. Daniel T.F., McDade L.A., Manktelow M., Kiel C.A. (2008) The ‘‘Tetramerium Lineage’’ (Acanthaceae: Justicieae): delimitation and intra-lineage relationships based on cp and nrITS sequence data. Systematic Botany, 33, 416–436. Duchen P., Renner S.S. (2010) The evolution of Cayaponia (Cucurbitaceae): repeated shifts from bat to bee pollination and long-distance dispersal to Africa 2–5 million years ago. American Journal of Botany, 97, 1129–1141. Echternacht L., Trovo M., Oliveira C.T., Pirani J.R. (2011) Areas of endemism in the Espinhaço Range in Minas Gerais, Brazil. Flora, 206, 782–791. Feinsinger P. (1983) Coevolution and pollination. In: Futuyma D.J., Slatkin M. (Eds), Coevolution. Sinauer, Sunderland, MA, USA, pp 282–310. Fenster C.B., Armbruster W.S., Wilson P., Dudash M.R., Thomson J.D. (2004) Pollination syndromes and floral specialization. Annual Review of Ecology and Systematics, 35, 375–403. Fleming T.H., Geiselman C., Kress W.J. (2009) The evolution of bat pollination: a phylogenetic perspective. Annals of Botany, 104, 1017–1043. Forzza R.C. (2005) Revisão taxonômica de Encholirium Mart. ex Schult. & Schult. f. (Pitcairnioideae – Bromeliaceae). Boletim de Botânica da Universidade de São Paulo, 23, 1–49. Forzza R.C., Zappi D. (2011) Side by side: two remarkable new species of Encholirium Mart. Ex Schult. & Schult. F. (Bromeliaceae) found in the Cadeia do Espinhaço, Minas Gerais, Brazil. Kew Bulletin, 66, 281–287. Floral traits and vertebrate pollinators of Encholirium Forzza R.C., Costa A., Siqueira Filho J.A., Martinelli G. (2011) Bromeliaceae. In: Forzza R.C., Leitman P.M., Costa A.F., de Carvalho A.A. Jr, Peixoto A.L., Bruno Machado Teles Walter B.M.T., Carlos Bicudo C., Moura C.W.N., Zappi D., da Costa D.P., Lleras E., Martinelli G., de Lima H.C., Prado J., Baumgratz J.F.A., Pirani J.R., Sylvestre L.S., Maia L.C., Lohmann L.G., Paganucci L., Alves M.V.S., Silveira M., Mamede M.C.H., Bastos M.N.C., Morim M.P., Barbosa M.R., Menezes M., Hopkins M., Evangelista P.H.L., Goldenberg R., Secco R., Rodrigues R.S., Cavalcanti T., Souza V.C. (Eds), Lista de Espécies da Flora do Brasil. Jardim Botânico do Rio de Janeiro, Rio de Janeiro, Brazil (http://floradobrasil.jbrj.gov.br/2011/FB006086). Galetto L., Bernadello G. (2005) Rewards in flowers – Nectar. In: Dafni A., Kevan P.G., Husband B.C. (Eds), Practical pollination biology. Enviroquest, Ontario, Canada, pp 261–313. Gallen C. (1999) Why do flowers vary? BioScience, 49, 631–640. Giulietti A.M., Menezes N.L., Pirani J.R., Meguro M., Wanderley M.G.L. (1987) Flora da Serra do Cipó, Minas Gerais: caracterização e lista de espécies. Boletim de Botânica da Universidade de São Paulo, 9, 1–151. Givnish T.J., Barfuss M.H.J., Benjamin E., Riina R., Schulte K., Horres R., Gonsiska P.A., Jabaily R.S., Crayn D.M., Smith J.A.C., Winter K., Brown G.K., Evans T.M., Holst B.K., Luther H., Till W., Zizka G., Berry P.E., Sytsma K.J. (2011) Phylogeny, adaptive radiation, and historical biogeography in Bromeliaceae: insights from an eight-locus plastid phylogeny. American Journal of Botany, 98, 872–895. Healey S.D., Hurly T.A. (2001) Foraging and spatial learning in hummingbirds. In: Chittka L., Thomson J.D. (Eds), Cognitive ecology of pollination: animal behavior and floral evolution. Cambridge University Press, Cambridge, UK, pp 127–147. Herrera G., Zagal J.C., Diaz M., Fernández M.J., Vielma A., Cure M., Martinez J., Bozinovic F., Palacios A.G. (2008) Spectral sensitivities of photoreceptors and their role in colour discrimination in the green-backed firecrown hummingbird (Sephanoides sephanoides). Journal of Comparative Physiology, 194, 785–794. Johnson S.D. (2006) Pollinator-driven speciation in plants. In: Harder L.D., Barrett S.C.H. (Eds), The ecology and evolution of flowers. Oxford University Press, Oxford, UK, pp 295–310. Kay K.M., Schemske D.W. (2003) Pollinator assemblages and visitation rates for 11 species of neotropical Costus (Costaceae). Biotropica, 35, 198–207. Kay K.M., Reeves P.A., Olmstead R.G., Schemske D.W. (2005) Rapid speciation and the evolution of hummingbird pollination in neotropical Costus subgenus Costus (Costaceae): evidence from NRDNA, ITS and ETS sequences. American Journal of Botany, 92, 1899–1910. Kearns C.A., Inouye D.W. (1993) Techniques for pollination biologists. University of Colorado Press, Niwot, CO, USA, 583 pp. Kessler M., Krömer T. (2000) Patterns and ecological correlates of pollination modes among bromeliad communities of Andean forests in Bolivia. Plant Biology, 2, 659–669. Kress W.J., Beach J.H. (1994) Flowering plant reproductive systems. In: McDade L.A., Bawa K.S., Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands Hespenheide H.A., Hartshorn G.S. (Eds), La Selva: ecology and natural history of neotropical rain forest. University of Chicago Press, Chicago, IL, USA, pp 161–182. Machado I.C.S., Sazima I., Sazima M. (1998) Bat pollination of the terrestrial herb Irlbachia alata (Gentianaceae) in northeastern Brazil. Plant Systematics and Evolution, 209, 231–237. Macior L.W. (1971) Co-evolution of plants and animals – systematic insights from plant–insect interactions. Taxon, 20, 17–28. Madeira J.A., Fernandes G.W. (1999) Reproductive phenology of sympatric taxa of Chamaecrista (Leguminosae) in Serra do Cipó, Brazil. Journal of Tropical Ecology, 15, 463–479. Martinelli G. (1994) Reproductive biology of Bromeliaceae in the Atlantic Rainforest of southeastern Brazil. PhD Thesis, University of St. Andrews, Scotland, UK: 197 pp. Matallana G., Godinho M.A.S., Guilherme F.A.G., Belisario M., Coser T.S., Wendt T. (2010) Breeding systems of Bromeliaceae species: evolution of selfing in the context of sympatric occurrence. Plant Systematics and Evolution, 289, 57–65. Mitchell R.J., Flanagan R.J., Brown B.J., Waser N.M., Karron J.D. (2009) New frontiers in competition for pollination. Annals of Botany, 103, 1403–1413. Moeller D.A. (2004) Facilitative interactions among plants via shared pollinators. Ecology, 85, 3289–3301. Morales C.L., Traveset A. (2008) Interspecific pollen transfer: magnitude, prevalence and consequences for plant fitness. Critical Reviews in Plant Sciences, 27, 221–238. Muchhala N. (2003) Exploring the boundary between pollination syndromes: bats and hummingbirds as pollinators of Burmeistera cyclostigmata and B. tenuiflora (Campanulaceae). Oecologia, 134, 373–380. Muchhala N. (2007) Adaptive trade-off in floral morphology mediates specialization for flowers pollinated by bats and hummingbirds. American Naturalist, 169, 494–504. Muchhala N., Thomson J.D. (2010) Fur versus feathers: pollen delivery by bats and hummingbirds and consequences for pollen production. American Naturalist, 175, 717–726. Nimer E. (1989) Climatologia do Brasil. Instituto Brasileiro de Geografia e Estatı́stica, Rio de Janeiro, Brazil: 422 pp. Perret M., Chautems A., Spichiger R., Kite G., Savolainen V. (2003) Systematics and evolution of Tribe Sinningieae (Gesneriaceae): evidence from phylogenetic analyses of six plastid DNA regions and nuclear NCPGS. American Journal of Botany, 90, 445–460. Ramalho M., Batista M.A., Silva M. (2004) Xylocopa (Monoxylocopa) abbreviata Hurd & Moure (Hymenoptera: Apidae) e Encholirium spectabile (Bromeliaceae): uma associação estreita no semiárido do Brasil tropical. Neotropical Entomology, 33, 417–425. Ramsey J., Bradshaw H.D. Jr, Schemske D.W. (2003) Components of reproductive isolation between the monkeyflowers Mimulus lewisii and M. cardinalis (Phrymaceae). Evolution, 57, 1520–1534. Rathcke B., Lacey E.P. (1985) Phenological patterns of terrestrial plants. Annual Review of Ecology and Systematics, 16, 179–214. Sahley C.T. (1996) Bat and hummingbird pollination of an autotetraploid columnar cactus, Weberbauer- 367 Floral traits and vertebrate pollinators of Encholirium ocereus weberbaueri (Cactaceae). American Jounal of Botany, 83, 1329–1336. Sazima I., Vogel S., Sazima M. (1989) Bat pollination of Encholirium glaziovii, a terrestrial bromeliad. Plant Systematics and Evolution, 168, 167–179. Sazima M., Buzato S., Sazima I. (1999) Bat-pollinated flower assemblage and bat visitors at two Atlantic forest sites in Brazil. Annals of Botany, 83, 705–712. Schemske D.W. (1981) Floral convergence and pollinator sharing in two bee-pollinated tropical herbs. Ecology, 62, 946–954. Schemske D.W., Bradshaw H.D. Jr (1999) Pollinator preference and the evolution of floral traits in monkeyflowers (Mimulus). Proceedings of the National Academy of Sciences, USA, 96, 11910– 11915. Smith L.B., Dows R.J. (1974) Pitcairnioideae (Bromeliaceae). Flora Neotropica, 14, 1–662. Sokal R.R., Rohlf F.J. (1995) Biometry. WH Freeman & Co., New York, NY, USA, pp. 887. 368 View publication stats Steiger S.S., Fidler A.E., Kempenaers B. (2009) Evidence for increased olfactory receptor gene repertoire size in two nocturnal bird species with welldeveloped olfactory ability. BMC Evolutionary Biology, 9, 117–127. Versieux L.M., Wendt T., Louzada R.B., Wanderley M.G.L. (2008) Bromeliaceae da Cadeia do Espinhaço. Megadiversidade, 4, 126–138. Vosgueritchian S.B., Buzato S. (2006) Reprodução sexuada de Dyckia tuberosa (Vell.) Beer (Bromeliaceae, Pitcairnioideae) e interação planta–animal. Revista Brasileira de Botânica, 29, 433–442. Wendt T., Canela M.B.F., Faria A.P.G., Rios R.I. (2001) Reproductive biology and natural hybridization between two endemic species of Pitcairnia (Bromeliaceae). American Journal of Botany, 88, 1760–1767. Wendt T., Canela M.B.F., Klein D.E., Rios R.I. (2002) Selfing facilitates reproductive isolation among three sympatric species of Pitcairnia (Bromeliaceae). Plant Systematics and Evolution, 232, 201–212. Christianini, Forzza & Buzato Wester P., Claben-Bockhoff R. (2006) Hummingbird pollination in Salvia haenkei (Lamiaceae) lacking the typical lever mechanism. Plant Systematics and Evolution, 257, 133–146. Winkler M., Hülber K., Mehltreter K., Franco J.G., Hietz P. (2005) Herbivory in epiphytic bromeliads, orchids and ferns in a Mexican montane forest. Journal of Tropical Ecology, 21, 147–154. Winter Y., Helversen O. von (2001) Bats as pollinators: foraging energetics and floral adaptations. In: Chittka L., Thomson J.D. (Eds), Cognitive ecology of pollination: animal behavior and floral evolution. Cambridge University Press, Cambridge, UK, pp 148–170. Yost J.M., Kay K.M. (2009) The evolution of postpollination reproductive isolation in Costus. Sexual Plant Reproduction, 22, 247–255. Young H.J. (2002) Diurnal and nocturnal pollination of Silene alba (Caryophyllaceae). American Journal of Botany, 89, 433–440. Plant Biology 15 (2013) 360–368 ª 2012 German Botanical Society and The Royal Botanical Society of the Netherlands