Advertisement

Plant Systematics and Evolution

, Volume 304, Issue 6, pp 793–806 | Cite as

Comparative reproductive biology reveals two distinct pollination strategies in Neotropical twig-epiphyte orchids

  • Emerson R. Pansarin
  • Pedro J. Bergamo
  • Lucenilda J. C. Ferraz
  • Silvia R. M. Pedro
  • Alessandro W. C. Ferreira
Original Article

Abstract

Members of Oncidiinae are widely known for their interactions with oil-collecting bees that explore lipophilic secretions on flowers. They may also be pollinated through food deception and the offering of nectar. Although data on breeding systems are available for many Oncidiinae orchids, little is known about the reproductive strategies in Rodriguezia, a neotropical genus of ca. 55 species. In this paper, we explore the reproductive biology of two species of Rodriguezia with distinctive morphologies: R. decora and R. lanceolata. Floral features, spectral reflectance, pollinators and pollination mechanisms, and breeding systems were studied. Both species are scentless and produce nectar as a reward. Floral nectar is secreted by a gland at the base of the labellum and stored into the sepaline spur. Rodriguezia decora reflects mainly in the blue and red regions of the light spectrum, while R. lanceolata reflects in the red region. Rodriguezia decora is exclusively visited and pollinated by butterflies, while Trochilidae hummingbirds are the pollinators of R. lanceolata. Pollinaria attach to the upper third of the proboscis of butterflies (R. decora), and to the bill of hummingbirds (R. lanceolata), during the collection of nectar from the spur. Both Rodriguezia species are self-sterile. Flower features and floral reflectance support the occurrence of psychophily in R. decora and ornithophily in R. lanceolata.

Keywords

Floral biology Nectar Ornithophily Pollination Psychophily Reproductive biology 

Introduction

Pollen and nectar are the commonest reward produced among flowering plants. However, nectar is indubitably the most widespread flower resource, since it is collected by a wider range of biotic vectors (Simpson and Neff 1981). In addition to flies, some groups of moths, hymenopterans, crickets, butterflies and birds are also well known as nectar foragers (e.g., van der Pijl and Dodson 1966; Micheneau et al. 2010). In fact pollination by birds has been recorded in more than 60 families of flowering plants, including orchids (Proctor et al. 1996; Cronk and Ojeda 2008). Along the Americas, the most important group of bird pollinators are the hummingbirds (i.e., Trochilidae; Fleming and Kress 2013). In orchids, pollination by hummingbirds has been recorded in several unrelated groups, such as Spiranthinae (Singer and Sazima 2000), Sobraliinae (Nunes et al. 2013) and Oncidiinae (Rodríguez-Robles et al. 1992), suggesting that this pollination system evolved independently several times along the evolution of this plant family. Although birds have been reported feeding on insects in flowers or eating floral parts, such as petals and anthers, the most widespread flower reward collected by these vertebrates is floral nectar (Cronk and Ojeda 2008). Birds are commonly attracted to flowers by the presence of specific features, such as vibrant colors and secretion of diluted nectar. Furthermore, flowers pollinated by birds (including hummingbirds) are frequently scentless and zygomorphic, with peculiar forms such as brush, gullet or tube (Faegri and van der Pijl 1979). Brush-like and tubular flowers are also frequently visited by Lepidopterans. However, the corolla tube is usually wider in bird-pollinated flowers than in butterfly-pollinated flowers (Cruden and Hermann-Parker 1979). Furthermore, nectar guides in flowers pollinated by butterflies simulate a target instead of lines, which in turn, are more commonly found in bird-pollinated plants (Faegri and van der Pijl 1979). Although some given features are exclusive to a particular syndrome (i.e., euphily), some characteristics of ornithophilous and psychophilous flowers may overlap (i.e., allophily).

Butterflies are known to pollinate several plant groups, including orchids (see Pansarin and Amaral 2008; Pansarin and Ferreira 2015; Pansarin et al. 2015). Psychophilous orchids offer floral nectar or are pollinated by food deception (e.g., van der Pijl and Dodson 1966; Pansarin and Amaral 2008; Pansarin and Ferreira 2015; Pansarin et al. 2015). As in hummingbird-pollinated orchids, psychophily occurs in some unrelated genera, being mainly recorded in members of the subfamily Orchidoideae (e.g., van der Pijl and Dodson 1966; Pansarin and Ferreira 2015; Pansarin et al. 2015). Among epidendroid orchids, pollination by butterflies has been recorded in Comparettia (Pansarin et al. 2015), Oeceoclades (Aguiar et al. 2012), and Epidendrum (e.g., Pansarin 2003; Pansarin and Amaral 2008). In this sense, it seems plausible that both hummingbirds and butterflies have played a role in the evolution of several orchid groups. Characterizing the reproductive biology of additional orchid genera could add evidence to the relevance of specialized pollination systems in the diversification of the richest family of flowering plants.

Nectar is produced by a variety of structures, comprising glands, and secretory epidermis, which can be smooth or bear different types of papillae or trichomes. Furthermore, nectar can be secreted through intercellular spaces between parenchymal cells or through modified stomata (Fahn 1979). In orchids, the offering of floral nectar as a resource has been reported in many groups (van der Pijl and Dodson 1966), including Oncidiinae (e.g., Rodríguez-Robles et al. 1992; Carvalho and Machado 2006; Pansarin et al. 2015), which encompasses more than 1000 species and at least 70 genera (Pridgeon et al. 2009). Members of Oncidiinae are widely known for their relationship with oil-collecting bees that explore lipophilic secretions produced in labellar elaiophores (e.g., Pansarin et al. 2017). Beyond pollination by offering of nectar and lipoidal substances, many Oncidiinae are pollinated through food deception without offering any reward (Chase et al. 2009).

Among Oncidiinae, self-compatibility has been reported for Comparettia, Ionopsis and Oncidium (Charanasri and Kamemoto 1977; Montalvo and Ackerman 1987; Pansarin et al. 2015). Members of this subtribe, however, are more commonly self-incompatible. In fact, self-incompatibility has been reported for Trichocentrum (Parra-Tabla and Magaña-Rueda 2000; Pansarin and Pansarin 2011), Baptistonia (Pansarin and Pansarin 2010), Tolumnia (Vale et al. 2011), as well as for Oncidium, Notylia, and Cohniella (e.g., Charanasri and Kamemoto 1977; Singer and Koehler 2003; Abdala-Roberts et al. 2007; Torretta et al. 2011), and two species of Rodriguezia (Carvalho and Machado 2006; Ospina-Calderón et al. 2015).

Although data on breeding systems are available for many Oncidiinae, little is known about the reproductive strategies in Rodriguezia, a genus with ca. 55 species distributed throughout the neotropics. Based on the evidence that self-incompatibility has been previously recorded for two Rodriguezia species, our main hypothesis is that self-incompatibility also occurs in the remaining species of the genus. Furthermore, it is known that R. bahiensis, R. granadensis and R. venusta offer nectar as a resource (Carvalho and Machado 2006; Leitão and Cortelazzo 2008; Ospina-Calderón et al. 2015), so our second hypothesis is that floral nectar is involved in the rewarding of pollinators in the remaining species of this genus. In order to explore these hypotheses, we have studied the reproductive biology of two species of Rodriguezia (R. decora and R. lanceolata) that belong to two different clades and show distinctive floral features. Specifically, we ask whether the reproductive features of R. decora and R. lanceolata reflect their pollination system.

Materials and methods

Study site and plant material

Data on floral biology and reproduction of Rodriguezia decora were recorded in the natural reserve of Serra do Japi, within the boundaries of the city of Jundiaí (23º11′S, 46º52′W, 900 m a. s. l.), São Paulo state, southeastern Brazil. This region is primarily characterized by mesophytic, semi-deciduous forests on sparse rocky outcrops at medium altitude (Leitão-Filho 1992). Data on reproductive biology of R. lanceolata were collected in Purão dos Pirrós (02º38′S; 45º14′W, ca. 24 m a. s. l.), a small village near the town of Pinheiro, ca. 100 km away from São Luis, Maranhão state, northeastern Brazil. This region is part of the Amazon rainforest (Amazonia Maranhense) (Almeida and Vieira 2010).

To study floral features and breeding system, adult plants of Rodriguezia decora and R. lanceolata were collected from natural populations and cultivated at the LBMBP Orchid House, University of São Paulo (FFCLRP-USP), in Ribeirão Preto (approx. 21º10′S, 47º48′W). According to Köppen (1948), the climate of Jundiaí and Ribeirão Preto is classified as “Cwa” (i.e., mesothermic with dry winter season), while the climate of São Luis is classified as “Aw” (i.e., wet tropical with dry winter season), with the highest rainfall recorded in February and March (average annual rainfall of 1893 mm) and an average annual temperature of 27 °C.

For the experimental treatments, adult plants of Rodriguezia decora (N = 30) and R. lanceolata (N = 10) growing on distinct phorophytes (at least 30 m away from each other) were collected in April 2012 (R. decora) and in November 2015 (R. lanceolata) and planted in individual pots with Pinus bark.

Flowering phenology and floral features

Features of the flowering phenology and flower duration of both Rodriguezia decora and R. lanceolata were recorded by monitoring both cultivated individuals and plants in the field. The studied area in the Serra do Japi was visited monthly, from July 2011 to June 2017, and the frequency of visits was intensified during the flowering period in order to study details of the pollination process. In Pinheiro, the studies were carried out from January to April 2016, during the flowering period, and the observations were intensified in the flowering peak (February).

Flower duration of Rodriguezia decora and R. lanceolata was recorded from plants cultivated in the LBMBP Orchid House, in April 2017 and November 2016, respectively. For each studied species, a total of 30 flowers (five plants; one inflorescence per plant; six flowers per inflorescence) were analyzed. Flowers were monitored every 24 h and photographed with a Nikon D-SLR D800 camera and a Micro Nikkor 105 mm f2.8 lens. Morphological features of fresh flowers (plants cultivated in the LBMBP Orchid House) of both species (n = 30; five plants; one inflorescence per plant; six flowers per inflorescence) were observed under a binocular stereomicroscope. Measurements were taken directly on the floral structures using a vernier caliper. The morphological study considered the format, symmetry, layout and size of flower structures, taking possible intrapopulation variation into account (Faegri and van der Pijl 1979). Flower details were studied and photographed with a stereomicroscope Stereozoom Leica S8 APO attached to a PC employing IM50 image analysis software. The production of floral fragrance was verified daily (day and night), by directly smelling the flowers from blooming to withering. Additionally, fresh flowers were immersed in 1% (w/v) aqueous neutral red for 20 min in order to localize possible secretory tissues (Dafni 1992). Once stained, they were rinsed in tap water and examined.

To characterize labellar secretions, flowers were manually sectioned and tests with Fehling’s reagent were performed to detect reducing sugars (Purvis et al. 1964). Appropriate controls, i.e., unstained cuts mounted in water (blank tests) were run simultaneously to these histochemical tests according to procedures described in Pansarin and Pansarin (2014). Additionally, the labellar secretions were tested with Keto-Diabur-Test 5000 (Roche). To characterize the anatomy of secretory structures, flowers were fixed in formalin-acetic acid-alcohol (FAA 50%) for 48 h (Johansen 1940), left under low vacuum, and stored in 70% ethanol. Secreting regions were dehydrated through an ethanol series and embedded in glycol methacrylate (Historesin Embedding Kit, Leica, Heidelberg, Germany), following the same methodology used in Pansarin et al. (2006). Transversal cuts (10–12 µm thick) were obtained using a rotary microtome and stained with toluidine blue (Sakai 1973), and permanent slides were mounted in synthetic resin. The images were captured with a Leica DM500 optical microscope equipped with a camera Leica ICC50 connected to a PC running IM50 image analysis software.

Nectar volume and concentration of Rodriguezia lanceolata were measured from 120 flowers (four plants; 12 inflorescences; 10 flowers per inflorescence) using a microliter syringe Hamilton 50 µl and a Bellingham & Stanley (series Eclipse) handheld refractometer (Kearns and Inouye, 1993). Measurements were taken only once, between 14:00 and 17:00 h, on flowers from 30 to 48 h from flower opening. Measurements were taken from October 07 to 08, 2015 (relative humidity between 33 and 34%) and from November 04 to 05, 2015 (relative humidity varying between 67 and 70%). Relative air humidity was measured with a thermo-hygrometer Anymetre, model TH603.

The spectral reflectance was measured using a Plug and Play Portable Stand Alone Fiber Optic Miniature Spectrometer model JAZ-EL200 UV/VIS range 200–850 nm coupled with a deuterium-halogen light source JAZ-PX—Pulsed-Xenon module for Jaz Instrumentation Platform. The white standard was measured from a pallet of barium sulfate and the black standard from a black chamber. The sepals, petals, labellum apex and labellum callus were measured from 10 individuals of both species (two flowers per individual). All measurements were taken at the same angle and direction relative to the surface. Since each species is pollinated by groups of pollinators with distinct visual capacities, we could not directly compare each reflectance using a single vision model. To overcome this, we followed procedures in Shrestha et al. (2013). From each reflectance curve, the inflection points were extracted and subtracted from each of the peak sensitivity of the photoreceptors (“AD”—absolute deviation, Shrestha et al. 2013) of butterfly model (peaks at 380, 460, 520 nm; Eguchi et al. 1982) and hummingbird model (peaks at 460, 540, and 600 nm; Shrestha et al. 2013). Then, we averaged these values for each reflectance curve. Since each curve usually produces more than one inflection point, we also averaged all inflection points of each reflectance curve (hereafter “MAD”—mean absolute deviation, Shrestha et al. 2013) to obtain a value for that reflectance in the butterfly vision (MADButterfly) and hummingbird vision (MADHummingbird). Low MAD values indicate that the reflectance matches the peak visual sensitivity of pollinators, and thus, the flower is easily detected and more attractive. The MAD values were compared using an ANOVA, using the pollinator (if the MAD value was calculated based on butterfly or hummingbird peaks), floral part (sepal, petal, labellum apex or labellum callus) and their interactions as fixed factors. A post hoc Tukey test was applied to explore differences among floral parts in the butterfly vs. hummingbird vision. A separate model was conducted for each plant species.

Pollinators and pollination mechanisms

Field visits were performed in order to observe and record the pollination process and the identity of the floral visitors. Focal observations on flowers of Rodriguezia decora were carried out from April 13 to 22, 1999, from May 06 to 12, 2009, and from April 19 to 25, 2010. The daily period of focal observations was from 09:00 to 16:00 h, totaling 168 h. The observations were performed on rock outcrops inside a mesophytic semi-deciduous forest. Additional observations were conducted from May 12 to 17, 2014 in a garden area adjacent to the LBMBP Orchidarium on two adult plants with dozens of inflorescences. The daily period of focal observations was from 09:00 to 16:00 h, totaling 42 h. Floral visitors of R. decora were photographed with a Nikon D-SLR D800 camera and a Micro Nikkor 105 mm f2.8 lens. Focal observations on flowers of R. lanceolata were carried out from 05:30 to 17:30 h (totaling 108 h) in three consecutive periods, (February 5–7, 12–14, 18–20, 2016). The observations were carried out in two patches of orchards 150 meters away from each other, one with Crescentia cujete L. (“Cuieiras,” Bignoniaceae) as main phorophytes, and other with orange trees (Citrus sinensis Osbeck, Rutaceae). Floral visitors of R. lanceolata were photographed with a digital Sony® DSC-HX300 camera, with 20.4 megapixels and 50× zoom.

Hummingbirds were identified using photographs taken during field work and a field guide to Brazilian hummingbirds (Ruschi 1989). Butterflies were identified by André V.L. Freitas (UNICAMP) and bees by SRMP.

Breeding system

The experimental treatments to investigate the breeding system of Rodriguezia decora and R. lanceolata included intact (bagged) flowers to test spontaneous self-pollination, manual self-pollination, manual cross-pollination, and emasculations. The treatments for R. decora were performed in the LBMBP Orchid House (2015 flowering season), with 30 plants (N = 240 flowers; two inflorescences per plant). The treatments for R. lanceolata were performed in both LBMBP Orchid House (2016 flowering season) with four plants (N = 120 flowers; three inflorescences per plant), and in the field (2016 flowering season) with 20 plants (N = 307 flowers; 35 inflorescences; 1–2 inflorescence per plant). For R. decora and R. lanceolata (LBMBP Orchid House), two repetitions of each treatment were carried out on each inflorescence. In the treatments of R. lanceolata performed in the field, the number of repetitions varied among the treatments depending on availability of flowers. All four treatments were applied on each inflorescence, using 1–3 days flowers. Manual cross-pollination was performed using a plant of a different pot (treatments performed in the LBMBP Orchid House) or from a different phorophyte (treatments carried out in the study area; R. lanceolata). Fruit set was recorded from naturally dehiscing capsules. Seeds were counted from fruits obtained through manual pollinations. A sample of 200 fresh seeds per fruit was examined under a light microscope, and those with well-formed embryos were considered as potentially viable. Seeds with rudimentary or no embryos were considered unviable.

Results

Flowering phenology and floral features

The flowering period of Rodriguezia decora extended from April to June. Fruits dehiscence occurred from October to November. The flowering season of R. lanceolata occurred from January to April. Fruits ripened from March to June. The anthesis of both species was diurnal, and, although flowers opened in succession, all the flowers of each inflorescence remained open simultaneously. In the studied populations, intact flowers of R. decora lasted ca. 30 days, while intact flowers of R. lanceolata lasted up to 14 days. The plants of R. lanceolata occurred exclusively as epiphytes, while R. decora can grow as rupicolous or epiphytes.

The inflorescences of both species are racemes (rarely a panicle in R. decora) and lateral. The inflorescence of R. decora is erect, with up to 105 flowers, while the inflorescences of R. lanceolata are pendant with up to 25 flowers.

The flowers of R. decora are resupinate and predominantly whitish with vinous maculae (Fig. 1a). Ovary + pedicel 13–18 × 0.8–1 mm. Ovary green, pedicel white. Dorsal sepal 10.5–16.4 × 4.5–6.5 mm, symmetrically elliptic, free, with an acute apex. Lateral sepals asymmetrically elliptic, fused 2/3 from the base. Synsepal 10.5–16.5 × 5.5–9.5 mm. Sepals fused at the base and the column foot forming a spur (Fig. 1b). Spur 3.2–5.2 × 1.8–2.6 mm, conical and white (Fig. 1b). Petals 12–15.5 × 6.5–8.5 mm, free, asymmetrically elliptic to ovate, with an acuminate apex. Labellum 23.5–32 × 13.5–20 mm, 3-lobed, predominantly white with vinous grooves, articulate with the column foot (Fig. 1a, c). Lateral lobes reduced, rounded to triangular. Apical lobe spathulate, with a conspicuous reniform and white apical portion. Grooves oblique, converging in a single point below the anther and viscidium (Fig. 1a, c). Grooves white, with vinous maculae and vinous or white trichomes. Each groove with a triangular crest. Column ca. 5–5.5 mm long, conical, white with vinous maculae, with two pairs of apical horns, inferior portion with hyaline trichomes (Fig. 1d). Column foot ca. 2 mm long. Inferior horns ca. 1.5–2 mm long, conical, incurved, glabrous. Superior horns ca. 3.5–5 mm long, incurved, vinous, with vinous trichomes. Stigma ca. 1.5 × 1 mm, concave, white (Fig. 1d). Pollinarium ca. 2.5–2.8 mm long. Pollinia 2 (ca 0.8–1 mm long), ovate, waxy, yellow. Stipe ca. 1.8–2 mm, spathulate, white. Viscidium ca. 0.7 mm, ventrally adhesive, brown. Anther cap ca. 3.5–3.8 × 2 mm, pyriform, white (Fig. 1b).
Fig. 1

af Rodriguezia decora. a Detail of a flower showing the oblique grooves (arrows). Note the large apical lobe of the labellum. b Flower in lateral view (petals and apical sepal removed) with the spur formed by the fusion of lateral sepals (arrow): inset box shows a flower in longitudinal section; an arrow indicate the nectary inside the sepaline spur. c Flower in dorsal view (petals and sepals removed) showing the column and labellum. Note the oblique grooves (dashed lines) converging to a single point where the pollinarium is located. d Detail of the column in longitudinal section, with the stigmatic surface (arrowhead) and the columnar horns (arrows) that guide the head of the butterfly inside the floral tube (up) and guide the pollinia inside the stigmatic surface (down). e Flower in longitudinal section showing the secretory tissues (nectary) at the end of labellum, inside the sepaline spur. Note the nectar being secreted (arrows). f Nectary in longitudinal section. Note the secretory cells, with large nuclei and densely stained cytoplasm, the secretory surface (arrowheads) and one idioblast (i). Scale bars: a–c = 5 mm; d–e = 1 mm; f = 50 µm

The basal portion of the labellum of R. decora ends in a glandular nectary covered with unicellular trichomes. Secretory cells possess large nucleus and dense cytoplasm. The central portion of the nectary possesses a large idioblast (Figs. 1e–f).

Rodriguezia lanceolata has resupinate and reddish flowers (Fig. 2a). Ovary + pedicel 8–13 × 1–1.5 mm, creamy to reddish. Dorsal sepal 13–17 × 5–6 mm, symmetrically elliptic to obovate, cymbiform, free, with an acute apex. Lateral sepals asymmetrically elliptic to lanceolate, connate from the base to the apex, apex acute. Synsepal 13–16 × 6–7 mm, incurved. Petals 12.5–14 × 6–7 mm, free, asymmetrically elliptic to ovate, with obtuse apex. Labellum 13–14 × 6.5–7 mm, 3-lobed, spathulate, reddish, with a central yellowish callus (Fig. 2a, b). Lateral lobes rounded, reduced. Apical lobe reniform, red. Column 5.5–6.5 mm long, clavate, white, with two yellowish and incurved horns (Fig. 2b–d). Stigma ca. 1.5 × 1 mm (Fig. 2b–d). Pollinarium ca. 2.2–2.5 mm long, with two round yellowish pollinia (ca. 1 mm). Stipe ca. 1.4–1.5 mm long, spathulate. Viscidium ca. 0.5 mm long, ventrally adhesive, brown (Fig. 2b). Anther ca. 2.9–3.1 × 2 mm, ovate, white (Fig. 2a, b).
Fig. 2

af Rodriguezia lanceolata. a Detail of a flower showing the nectary entrance (arrows). Note the columnar horns that guide the pollinia into the stigmatic surface (arrowheads) and the anther cap (a). b Flower in lateral view (one petal removed) showing the column with the anther cap (arrowhead) and the columnar horns that guide the pollinia into the stigma (arrow). c Detail of a flower in lateral view (one petal and the anther cap removed) showing the stigmatic surface (arrow) and the viscidium of the pollinarium (small dashed circle). Note the synsepal covering the secretory tissue (large dashed circle) and arrowheads showing the nectariferous chamber where the nectar is stored. d Flower in longitudinal section showing the nectary (dashed circle) inside the nectar chamber (c). Note the columnar horn (arrowhead) that guides the pollination to the stigmatic surface (s). e Detail of the nectary. Arrowheads point to the secretory surface. f Nectary in longitudinal section showing the secretory tissue (arrowheads). The detail shows the secretory cells. Note their large nucleus and densely stained cytoplasm. Scale bars: a–d = 1 mm; e–f = 200 µm

The basal portion of the labellum of R. lanceolata possesses a glandular nectary ca. 2 mm long. The nectar gland is covered with unicellular trichomes. The nectary is densely vascularized with longitudinal vascular bundles, which were rich in sieve tube elements (Fig. 2e–f).

Rodriguezia decora and R. lanceolata were scentless to the human nose. Furthermore, tests with neutral red were negative, indicating the absence of osmophores in the flowers. With an air humidity of 33–34%, R. lanceolata produced a nectar column of 1–6 μL (mean 2.7 ± 1.14, n = 60) in the spur. Sugar concentration ranged between 23 and 53% (mean 44% ± 7.5, n = 60). With an air humidity of 67–70%, flowers of R. lanceolata produced a nectar column of 2–12 μL (mean 5.9 ± 2.67, n = 60) in the spur. Sugar concentration ranged between 17 and 38% (mean 28% ± 4.62, n = 60). In flowers of Rodriguezia decora and R. lanceolata, the histochemical procedures with Fehling’s reagent, as well as the tests with the Keto-Diabur-Test, confirmed the presence of sugar.

Rodriguezia decora reflected most on the blue and red wavelengths, while R. lanceolata reflected most on the red wavelength. The reflectance of R. decora produced different MAD values depending on the interaction of pollinator vision and floral part (pollinator effect: F = 520.34, p < 0.001; floral part effect: F = 18.64, p < 0.001; pollinator*floral part effect: F = 322.45, p < 0.001). All floral parts had lower MAD values for butterfly vision than hummingbird vision, except for the labellum callus that had lower values for the hummingbird vision (Table 1). This indicates that R. decora floral reflectance is closer to the peak sensitivity of butterfly vision than to that of hummingbirds. The reflectance of R. lanceolata also produced distinct MAD values depending on the interaction of pollinator vision and floral part (pollinator effect: F = 581.46, p < 0.001; floral part: F = 1.59, p = 0.195; pollinator*floral part effect: F = 607.43, p < 0.001). All floral parts had lower MAD values for hummingbird vision than butterfly vision (Table 1). This indicates that R. lanceolata floral reflectance is closer to the peak sensitivity of hummingbirds than to that of butterflies.
Table 1

Average values for the mean absolute deviation (MAD) butterfly and MAD hummingbird for each floral part of Rodriguezia decora and R. lanceolata. For both species, all floral parts exhibited differences between MAD Butterfly and MAD Hummingbird values after a Tukey post hoc comparison at p < 0.05

 

MAD butterfly

MAD hummingbird

p

Rodriguezia decora

   

Sepal

17.17

62.17

< 0.001

Petal

16.70

63.63

< 0.001

Labellum apex

23.93

56.07

< 0.001

Labellum callus

64.17

33.30

< 0.001

Rodriguezia lanceolata

   

Sepal

47.45

26.92

< 0.001

Petal

48.65

29.85

< 0.001

Labellum apex

49.42

30.65

< 0.001

Labellum callus

55.93

21.06

< 0.001

Pollinators and pollination mechanisms

In our study area, Rodriguezia decora was exclusively visited and pollinated by butterflies (Table 2; Fig. 3a). Flower visits were recorded between 09:45 and 15:00 h. Pollinators visited one or two flowers per inflorescence. Each visit lasted 1–5 s. The total number of pollinaria removals and depositions could not be determined unequivocally because part of the flowering plants was out of sight. Visits by butterflies started with the insect landing on a flower (Fig. 3a) and uncoiling its proboscis, and then inserting it into the flower tube. The longitudinal grooves of the labellum guided the proboscis along the adaxial surface of the labellum to the spur. The horns on the upper side of the column were responsible for the correct positioning of the head of the butterflies in the center of the floral tube (Fig. 3b, c). Pollinarium removal occurred when butterflies contacted the viscidium with the dorsal surface of the proboscis, while they accessed the nectar at the base of the nectary chamber (Fig. 3b–d). The pollinaria were deposited on the upper third of the proboscis (Fig. 3d). The pollinarium was removed without the anther cap. The pollination occurred when the head of a butterfly carrying a pollinarium entered the labellum and the pollinia contacted the stigmatic surface. The horns of the lower side of the column guided the pollinarium to the correct placement into the stigmatic surface.
Table 2

Floral visitors (−) and pollinators (*) recorded on flowers of Rodriguezia decora and R. lanceolata. Animal vectors are grouped by family

Species

R. decora

R. lanceolata

Hesperiidae

  

 Astraptes fulgerator (Walch, 1775)

*

 

 Pompeius pompeius (Latreille, 1824)

 

 Urbanus dorantes (Stoll, 1790)

*

Pieridae

  

 Aphrissa statira (Cramer, 1777)

 

 Ascia monuste (Linnaeus, 1764)

*

Nymphalidae

  

 Dryas iulia (Fabricius, 1775)

*

 

Riodininae

  

 Synargis calyce (C. Felder & R. Felder, 1862)

 

Apidae

  

 Centris (Hemisiella) tarsata Smith, 1854

 

 Melipona (Melikerria) fasciculata Smith, 1854

 

 Trigona pallens (Fabricius, 1798)

 

 Xylocopa (Neoxylocopa) grisescens Lepeletier, 1841

 

Vespidae

  

 Polybia cf. occidentalis Olivier, 1791

 

Trochilidae

  

 Amazilia fimbriata (J.F. Gmelin,1788)

 

*

 Amazilia versicolor (Veillot, 1818)

 

*

Fig. 3

Pollinators of Rodriguezia decora (a–d), and R. lanceolata (e–f). a Ascia monuste butterfly landing on the labellum. b Flower in longitudinal section showing the contact of the proboscis of Urbanus dorantes with the viscidium of the pollinarium (posed). The pollinarium is attached to the upper third of the proboscis (arrow). c Flower in longitudinal section (anther cap removed) showing the contact of the proboscis of Dryas iulia with the pollinarium. Note the pollinarium is attached to the upper third of the proboscis. When removed, the pollinarium is erect (dashed circle). d Detail of a dried pollinarium on the proboscis of Dryas iulia. Note that when dry, the pollinarium is parallel in relation to the proboscis (arrow). e Amazilia vesicolor searching for nectar in a flower. Note the bill inserted in a flower. f Amazilia vesicolor with many pollinaria and stipites on the bill (arrow). Scale bars: a = 2 cm; b–d = 2 mm; e–f = 2 cm

In the study area (Pinheiro—MA), the flowers of Rodriguezia lanceolata were visited by butterflies (four species), bees (four species), one wasp species and two hummingbird species (Trochilidae; Table 2). However, only the hummingbirds (Amazilia fimbriata and A. versicolor; Fig. 3e–f) were effective pollinators of R. lanceolata. Bees, butterflies and wasps acted as nectar robbers. Visits by the hummingbirds were recorded from 05:30 to 17:30 h, while the insects were observed from 06:30 to 17:00 h. Hummingbirds returned to the focal plants every 40 min on average, and visited up to 30 flowers consecutively. Pollinators visited from one to six flowers per inflorescence, and each visit lasted between 2 and 4 s. Visits started with the hummingbird hovering in front of an inflorescence. Then, it inserted the bill into the tube formed by the labellum and column to probe the nectar accumulated in the synsepal (Fig. 3e). Pollinarium removal occurred when the hummingbirds contacted the viscidium with their bill (Fig. 3f). The pollinarium was removed without the anther cap. Pollinia deposition on stigmatic surface occurred when a hummingbird carrying a pollinarium visited another flower of R. lanceolata. As in R. decora, the horns of the lower side of the column guided the pollinarium to the correct placement into the stigmatic surface. After visitation, sometimes the hummingbirds were recorded on the branches of trees, scraping their bill to remove the pollinaria attached to it.

The stingless bee Trigona pallens was observed removing accidentally the pollinarium with the legs, but it remained adhered to the bee only for a short period. In fact, T. pallens was frequently observed on flowers of R. lanceolata, but acting as a nectar robber. Melipona fasciculata, Centris tarsata and the wasp Polybia cf. occidentalis were rarely recorded on flowers.

In both Rodriguezia decora and R. lanceolata, when the pollinarium is removed it is deposited in a perpendicular position in relation to the length of the proboscis of butterflies and the bill of hummingbirds (Fig. 4a–c). Pollination is only possible when the pollinarium is dehydrated (elapsed a period of 12 min) and in parallel position in relation to the proboscis or bill length (Fig. 4d–f).
Fig. 4

af Bending mechanism in one pollinarium of Rodriguezia decora. Soon after being removed, the pollinarium is in an angle of ca. 90° in relation to the butterfly proboscis. After 12 min, the pollinarium assumes a parallel position in relation to the butterfly proboscis. Pictures were taken every two minutes. Scale bars: a–f = 1 mm

Breeding system

Plants of Rodriguezia decora and R. lanceolata were completely self-incompatible. For the treatments performed in the LBMBP Orchid House, the fruit set in cross-pollinated flowers was 98.3% (59 fruits/60 flowers) for R. decora and 100% (30 fruits/30 flowers) for R. lanceolata. For the treatments performed in the natural habitat (R. lanceolata), the manual cross-pollination resulted in a fructification of 58.8%. No fruits were formed in the treatments of manual self-pollination, spontaneous self-pollination, or in emasculated flowers (Table 3). Thus, according to our data, both Rodriguezia decora and R. lanceolata need a biotic pollen vector for pollen transfer, and they must undergo cross-pollination to promote fruit set.
Table 3

Fruit set and potentially viable seeds in Rodriguezia decora and R. lanceolata under different pollination treatments carried out at the LBMBP Orchid Housea or in the fieldb

Species

R. decora a

R. lanceolata a

R. lanceolata b

Treatments

% fr (fr/fl)

% s (pvs/s)

% fr (fr/fl)

% s (pvs/s)

% fr (fr/fl)

% Seeds

Spontaneous self-pollination

0 (0/60)

0 (0/30)

0 (0/155)

Manual self-pollination

0 (0/60)

0 (0/30)

0 (0/88)

Emasculation

0 (0/60)

0 (0/30)

0 (0/30)

Manual cross-pollination

98.3 (59/60)

90.9 (10,361/11,400)

100 (30/30)

58.8 (20/34)

96 (3840/4000)

fr number of fruits, fl number of flowers, s number of seeds, pvs number of potentially viable seeds

The fruits of Rodriguezia decora obtained by cross-pollination yielded 90.9% of potentially viable seeds on average. The fruits of R. lanceolata obtained by cross-pollination yielded 96% of potentially viable seeds on average (Table 3).

Discussion

In the study areas, the flowering period and the floral features are different between R. decora and R. lanceolata. The flowering period of R. decora occurs in the winter, and R. lanceolata flowers in the summer. The inflorescence of R. decora is erect, while R. lanceolata possesses pendant racemes. Furthermore, flowers of R. decora are very distinct from flowers of R. lanceolata. The spur in R. decora is short, parallel in relation to the labellum and formed by the union of the base of sepals. In R. lanceolata, the sepals are completely fused and the synsepal stands in a perpendicular position in relation to the labellum. The structure of the nectary of R. lanceolata is very similar to that reported for R. venusta (Leitão et al. 2014). Both species possesses a tongue-like nectar gland covered with secretory trichomes. As in R. venusta, the nectary of R. lanceolata is an extension of the labellum base, and the nectar accumulates inside the synsepal (Leitão et al. 2014). Conversely, the secretory tissues of R. decora do not have a projection as in R. lanceolata, R. venusta and the remaining Rodriguezia with short rhizomes. In fact, the Rodriguezia with long rhizomes (i.e., R. decora, R. obtusifolia and R. pulcherrima) form a monophyletic group that possesses characteristics not found in the remaining species of this genus (E. R. Pansarin, unpublished data). The nectar volume and concentration of R. lanceolata are similar to that reported for some plant species pollinated by Trochilidae hummingbirds (e.g., Stiles and Feeman 1993). Nectar volume produced in flowers in the hummingbird-pollinated R. lanceolata is higher than in R. decora, whose flowers were visited and pollinated exclusively by butterflies. In fact, flowers pollinated by birds commonly produce higher nectar volumes than those pollinated by butterflies and bees (Stiles 1981). However, the nectar volume and concentration in R. lanceolata vary according to the evaporation rates, in a similar way as previously reported for Oeceoclades maculata (Aguiar and Pansarin 2013).

Although a generalist pollination system has been reported for Rodriguezia (Carvalho and Machado 2006), our data revealed a specialized pollination in R. decora and R. lanceolata, in a similar way as recorded for R. granadensis (Ospina-Calderón et al. 2015). In fact, both species possess flower characteristics that prevent pollen deposition on a floral visitor with incompatible morphology. Although some characteristics of ornithophilous and psychophilous flowers overlap and according to our observations visitation by butterflies also has been recorded on flowers of R. lanceolata, each Rodriguezia species was pollinated by a single functional group of pollinators, i.e., butterflies in R. decora and hummingbirds in R. lanceolata. Although the corolla tube of flowers pollinated by butterflies tends to be narrower than the hummingbird-pollinated flowers (Cruden and Hermann-Parker 1979; Pansarin et al. 2015), the flowers of R. decora possess a relatively wide tube. In this case, the correct placement of the pollinarium on the butterfly proboscis is possible since it is guided by the oblique labellar grooves, which converge to a single point just below the viscidium. Among Oncidiinae, pollination by butterflies has also been recorded in Comparettia (Pansarin et al. 2015), a genus closely related to Rodriguezia. As in R. decora, the pollination of Comparettia coccinea is very specialized, with only two species of Nymphalid butterflies being able to reach the nectar from its incurved spur. The presence of an incurved spur prevents nectar collection by other flower visitors, such as hummingbirds. Furthermore, the existence of two nectar entrances in C. coccinea is adapted to deposit pollinaria on the eyes of the butterflies. In contrast, orchids pollinated by birds have a tubular corolla with a single nectar entrance occupying a central position, since pollinaria are deposited on the bill (see van der Pijl and Dodson 1966; van der Cingel 2001). This is the case of R. lanceolata, whose flowers have a single central nectary entrance and are pollinated exclusively by hummingbirds, although several butterflies and hymenoptera species were recorded as floral visitors. Pollination by bees was recorded in populations of R. granadensis occurring in Colombia (Ospina-Calderón et al. 2015), while the pollination system of R. bahiensis was reported as generalist, with Acroceridae flies as the main pollinators (Carvalho and Machado 2006). In this sense, our results indicate that specialized reproductive strategies are more widespread in the Rodriguezia genus, which might be a key factor promoting the diversification of this group of orchids.

In addition to the peculiar flower morphology, the two Rodriguezia species also show distinct floral reflectance patterns: R. decora matched butterfly vision while R. lanceolata matched hummingbird vision. The white color with pinkish spots on flowers of R. decora results in a pronounced blue reflectance, a color well perceived by Lepidoptera species (Eguchi et al. 1982; Telles et al. 2016). On the other hand, R. lanceolata reflects in the red region, which is often associated with hummingbird pollination (Bergamo et al. 2016). This reinforces the pollination specialization of these species (i.e., psychophily for R. decora and ornithophily for R. lanceolata) and that the floral reflectance targeted its main pollinators (Shrestha et al. 2013). Moreover, this pattern is consistent for all floral parts measured, possibly indicating that the entire flower acts on the visual attraction of the pollinators. The exception is the labellum callus of R. decora, which may not be important in the visual attraction. If flowers of R. lanceolata show a distinct reflectance pattern when compared to flowers of R. decora, why R. lanceolata is also visited by butterflies and hymenopterans? According to Rodríguez-Gironés and Santamaría (2004), the main point is not that insects would not forage on flowers that they poorly detect. It is a question of relative efficiency when searching for flowers. Insects tend to avoid red flowers (presumably more difficult to detect, Bergamo et al. 2016), while hummingbirds consume their nectar.

Although studies on the reproductive strategies involving members of Oncidiinae have been rarely performed, the occurrence of self-incompatibility seems to be the more widespread condition among this diverse subtribe. In fact, self-incompatibility has been recorded in species belonging to many Oncidiinae species (e.g., Parra-Tabla and Magaña-Rueda 2000; Abdala-Roberts et al. 2007; Pansarin and Pansarin 2010, 2011; Pansarin et al. 2017). Although rare, the occurrence of self-compatibility has been recorded for some genera closely related to Rodriguezia, including Ionopsis and Comparettia (Pansarin et al. 2015; J.M.R.B.V. Aguiar and E.R. Pansarin, unpubl. data). However, Brazilian populations of Ionopsis utricularioides were recorded as self-compatible, while self-sterility was found in Caribbean populations (Montalvo and Ackerman 1987). Additionally, both self-compatible and self-sterile individuals were recorded in populations of Gomesa varicosa in southeastern Brazil (Pansarin et al. 2017).

In conclusion, the flowers of both species of Rodriguezia produce nectar as a resource for its main pollinators. The nectar is produced by a gland located at the end of the labellum. Flowers of R. decora are exclusively visited and pollinated by butterflies, while hummingbirds are the pollinators of R. lanceolata. Flower features, such as morphology and floral reflectance, support the occurrence of psychophily in R. decora and ornithophily in R. lanceolata. Pollinaria attach to the upper third of the butterflies’ proboscis (R. decora), and to the bill of hummingbirds (R. lanceolata), during nectar collection from the spur. Despite previous reports on generalist pollination, our comparative reproductive results reveal the presence of at least two specialized reproductive systems in Rodriguezia.

Notes

Acknowledgements

The authors thank André V.L. Freitas (UNICAMP) for the identification of the butterflies. A.W.C.F. thanks FAPEMA for funding this research (grant 0430/2015).

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

References

  1. Abdala-Roberts L, Parra-Tabla V, Navarro J (2007) Is floral longevity influenced by reproductive costs and pollination success in Cohniella ascendens (Orchidaceae)? Ann Bot (Oxford) 100:1367–1371.  https://doi.org/10.1093/aob/mcm219 CrossRefGoogle Scholar
  2. Aguiar JMRBV, Pansarin ER (2013) Does Oeceoclades maculata (Orchidaceae) reabsorbs nectar? Eur J Environm Sci 3:113–118CrossRefGoogle Scholar
  3. Aguiar JMRBV, Pansarin LM, Ackerman JD, Pansarin ER (2012) Biotic versus abiotic pollination in Oeceoclades maculata (Lindl.) Lindl. (Orchidaceae). Pl Spec Biol 27:86–95.  https://doi.org/10.1111/j.1442-1984.2011.00330.x CrossRefGoogle Scholar
  4. Almeida AS, Vieira ICG (2010) Centro de Endemismo Belém: status da vegetação remanescente e desafios para a conservação da biodiversidade e restauração ecológica. Revista Estudos Univ 36:95–111Google Scholar
  5. Bergamo PJ, Rech AR, Brito VLG, Sazima M (2016) Flower colour and visitation rates of Costus arabicus support the “bee-avoidance” hypothesis for red-reflecting hummingbird-pollinated flowers. Funct Ecol 30:710–720.  https://doi.org/10.1111/1365-2435.12537 CrossRefGoogle Scholar
  6. Carvalho R, Machado IC (2006) Rodriguezia bahiensis Rchb.f.: biologia floral, polinizadores e primeiro registro de polinização por moscas Acroceridae em Orchidaceae. Brazil J Bot 29:461–470.  https://doi.org/10.1590/S0100-84042006000300013 Google Scholar
  7. Charanasri U, Kamemoto H (1977) Self-incompatibility in the Oncidium Alliance. Hawaii Orchid J 6:12–15Google Scholar
  8. Chase MW, Williams NH, Faria AD, Neubig KM, Amaral MCE, Whitten WM (2009) Floral convergence in Oncidiinae (Cymbidieae; Orchidaceae): an expanded concept of Gomesa and a new genus Nohawilliamsia. Ann Bot (Oxford) 109:1–16.  https://doi.org/10.1093/aob/mcp067 Google Scholar
  9. Cronk QCB, Ojeda I (2008) Bird-pollinated flowers in an evolutionary and molecular context. J Exp Bot 59:715–727.  https://doi.org/10.1093/jxb/ern009 CrossRefPubMedGoogle Scholar
  10. Cruden RW, Herman-Parker SM (1979) Butterfly pollination of Caesalpinia pulcherrima, with observations on the psychophilous syndrome. J Ecol 67:155–168.  https://doi.org/10.2307/2259342 CrossRefGoogle Scholar
  11. Dafni A (1992) Pollination ecology: a practical approach. Oxford University Press, OxfordGoogle Scholar
  12. Eguchi E, Watanabe K, Hariyama T, Yamamoto K (1982) A comparison of electrophysiologocally determined spectral responses in 35 species of Lepidoptera. J Insect Physiol 28:675–682CrossRefGoogle Scholar
  13. Faegri K, van der Pijl L (1979) The principles of pollination ecology. Pergamon Press, OxfordGoogle Scholar
  14. Fahn A (1979) Secretory tissues in plants. Academic Press, LondonGoogle Scholar
  15. Fleming TH, Kress WJ (2013) The ornaments of life: coevolution and conservation in the tropics. University of Chicago Press, ChicagoCrossRefGoogle Scholar
  16. Johansen DA (1940) Plant microtechnique. McGraw-Hill Book Co, New YorkGoogle Scholar
  17. Kearns CA, Inouye D (1993) Techniques for pollinations biologists. University Press of Colorado, NiwotGoogle Scholar
  18. Köppen W (1948) Climatologia: com um estúdio de los climas de la tierra. Fondo de Cultura Econômica, MéxicoGoogle Scholar
  19. Leitão CAE, Cortelazzo AL (2008) Structural and histochemical characterization of the colleters of Rodriguezia venusta (Orchidaceae). Austral J Bot 56:161–165.  https://doi.org/10.1071/BT07114 CrossRefGoogle Scholar
  20. Leitão CAE, Dolder MAH, Cortelazzo AL (2014) Anatomy and histochemistry of the nectaries of Rodriguezia venusta (Lindl.). Rchb. f. (Orchidaceae). Flora 209:233–243.  https://doi.org/10.1016/j.flora.2014.03.002 Google Scholar
  21. Leitão-Filho HF (1992) A flora arbórea da Serra do Japi. In: Morellato LPC (ed) História natural da Serra do Japi. Editora da Unicamp/Fapesp, Campinas, pp 40–62Google Scholar
  22. Micheneau C, Fournel J, Warren BH, Hugel S, Gauvin-Bialecki A, Pailler T, Strasberg D, Chase MW (2010) Orthoptera, a new order of pollinator. Ann Bot (Oxford) 105:355–364.  https://doi.org/10.1093/aob/mcp299 CrossRefGoogle Scholar
  23. Montalvo AM, Ackerman JD (1987) Limitations to fruit production in Ionopsis utricularioides (Orchidaceae). Biotropica 19:24–31CrossRefGoogle Scholar
  24. Nunes CEP, Castro MM, Galetto L, Sazima M (2013) Anatomy of the floral nectary of ornithophilous Elleanthus brasiliensis (Orchidaceae: Sobralieae). Bot J Linn Soc 171:764–772.  https://doi.org/10.1111/boj.12024 CrossRefGoogle Scholar
  25. Ospina-Calderón NH, Duque-Buitrago CA, Tremblay RL, Tupac-Otero J (2015) Pollination ecology of Rodriguezia granadensis (Orchidaceae). Lankesteriana 15:129–139CrossRefGoogle Scholar
  26. Pansarin ER (2003) Biologia reprodutiva e polinização em Epidendrum paniculatum Ruiz & Pavón (Orchidaceae). Revista Brasil Bot 26:203–211.  https://doi.org/10.1590/S0100-84042003000200008 Google Scholar
  27. Pansarin ER, Amaral MCE (2008) Reproductive biology and pollination mechanisms of Epidendrum secundum (Orchidaceae). Floral variation: a consequence of natural hybridization? Pl Biol 10:211–219.  https://doi.org/10.1111/j.1438-8677.2007.00025.x CrossRefGoogle Scholar
  28. Pansarin ER, Ferreira AWC (2015) Butterfly pollination in Pteroglossa (Orchidaceae, Orchidoideae): a comparative study on the reproductive biology of two species of a Neotropical genus of Spiranthinae. J Pl Res 128:459–468.  https://doi.org/10.1007/s10265-015-0707-x CrossRefGoogle Scholar
  29. Pansarin ER, Pansarin LM (2010) The family Orchidaceae in the Serra do Japi, State of São Paulo, Brazil. Springer, Wien.  https://doi.org/10.1007/978-3-211-99755-0 Google Scholar
  30. Pansarin ER, Pansarin LM (2011) Reproductive biology of Trichocentrum pumilum: an orchid pollinated by oil-collecting bees. Pl Biol 13:576–581.  https://doi.org/10.1111/j.1438-8677.2010.00420.x CrossRefGoogle Scholar
  31. Pansarin ER, Pansarin LM (2014) Floral biology of two Vanilloideae (Orchidaceae) primarily adapted to pollination by euglossine bees. Pl Biol 16:1104–1113.  https://doi.org/10.1111/plb.12160 Google Scholar
  32. Pansarin ER, Bittrich V, Amaral MCE (2006) At daybreak—reproductive biology and isolating mechanisms in Cirrhaea dependens (Orchidaceae). Pl Biol 8:494–502.  https://doi.org/10.1055/s-2006-923800 CrossRefGoogle Scholar
  33. Pansarin ER, Pansarin LM, Santos IA (2015) Floral features, pollination biology, and breeding system of Comparettia coccinea (Orchidaceae: Oncidiinae). Flora 207:57–63.  https://doi.org/10.1016/j.flora.2015.09.008 CrossRefGoogle Scholar
  34. Pansarin ER, Santos IA, Pansarin LM (2017) Comparative reproductive biology and pollinator specificity among sympatric Gomesa (Orchidaceae: Oncidiinae). Pl Biol 19:147–155.  https://doi.org/10.1111/plb.12525 CrossRefGoogle Scholar
  35. Parra-Tabla V, Magaña-Rueda S (2000) Effects of deforestation on the reproductive ecology of Oncidium ascendens (Orchidaceae). Tropical bees: management and diversity. In: Munn, P. (ed), Proceedings of the VI international conference on tropical bees. San José de Costa Rica. IBRA, Cardiff, UKGoogle Scholar
  36. Pridgeon AM, Cribb PJ, Chase MW, Rasmussen FN (2009) Genera Orchidacearum 5: Epidendroideae (part two). Oxford University Press, New YorkGoogle Scholar
  37. Proctor M, Yeo P, Lack A (1996) The natural history of pollination. Timber Press, PortlandGoogle Scholar
  38. Purvis MJ, Collier DC, Walls D (1964) Laboratory techniques in botany. Butterwoths, LondonGoogle Scholar
  39. Rodríguez-Gironés MA, Santamaría L (2004) Why are so many birdflowers red? PLoS Biol 2:e350.  https://doi.org/10.1371/journal.pbio.0020350 CrossRefPubMedPubMedCentralGoogle Scholar
  40. Rodríguez-Robles JA, Meléndez EJ, Ackerman JD (1992) Effects of display size, flowering phenology, and nectar availability on effective visitation frequency in Comparettia falcata (Orchidaceae). Amer J Bot 79:1009–1017CrossRefGoogle Scholar
  41. Ruschi A (1989) Aves do Brasil, Beija-flores. Vol. IV. Editora Expressão e Cultura, Rio de JaneiroGoogle Scholar
  42. Sakai WS (1973) Simple method for differential staining of paraffin embedded plant material using toluidine blue O. Stain Technol 43:247–249CrossRefGoogle Scholar
  43. Shrestha M, Dyer AG, Boyd-Gerny S, Wong BBM, Burd M (2013) Shades of red: bird-pollinated flowers target the specific colour discrimination abilities of avian vision. New Phytol 198:30–310.  https://doi.org/10.1111/nph.12135 CrossRefGoogle Scholar
  44. Simpson BB, Neff JL (1981) Floral rewards: alternatives to pollen and nectar. Ann Missouri Bot Gard 68:301–322.  https://doi.org/10.2307/2398800 CrossRefGoogle Scholar
  45. Singer RB, Koehler S (2003) Notes on the pollination of Notylia nemorosa (Orchidaceae): Do pollinators necessarily promote cross pollination? J Pl Res 116:19–25.  https://doi.org/10.1007/s10265-002-0064-4 Google Scholar
  46. Singer RB, Sazima M (2000) The pollination of Stenorrhynchos lanceolatus (Aublet) L. C. Rich. (Orchidaceae: Spiranthinae) by hummingbirds in southeastern Brazil. Pl Syst Evol 223:221–227CrossRefGoogle Scholar
  47. Stiles FG (1981) Geographical aspects of bird- flower coevolution with particular reference to Central America. Ann Missouri Bot Gard 68:323–351CrossRefGoogle Scholar
  48. Stiles FG, Freeman CE (1993) Patterns in floral nectar characteristics of some bird-visited plant species from Costa Rica. Biotropica 25:191–205CrossRefGoogle Scholar
  49. Telles FJ, Kelber M, Rodríguez-Gironés MA (2016) Wavelength discrimination in the hummingbird hawkmoth Macroglossum stellatarum. J Exp Biol 219:553–560.  https://doi.org/10.1242/jeb.130484 CrossRefPubMedGoogle Scholar
  50. Torretta JP, Gomiz NE, Aliscioni SS, Bello ME (2011) Biología reproductiva de Gomesa bifolia (Orchidaceae, Cymbidieae, Oncidiinae). Darwiniana 49:16–24Google Scholar
  51. Vale A, Navarro L, Rojas D, Álvarez JC (2011) Breeding system and pollination by mimicry of the orchid Tolumnia guibertiana in Western Cuba. Pl Spec Biol 26:163–173.  https://doi.org/10.1111/j.1442-1984.2011.00322.x CrossRefGoogle Scholar
  52. van der Cingel NA (2001) An atlas of orchid pollination. America, Africa, Asia and Australia. Balkema Publishers, RottherdamGoogle Scholar
  53. van der Pijl L, Dodson CH (1966) Orchid flowers: their pollination and evolution. University of Miami, Coral GablesGoogle Scholar

Copyright information

© Springer-Verlag GmbH Austria, part of Springer Nature 2018

Authors and Affiliations

  • Emerson R. Pansarin
    • 1
  • Pedro J. Bergamo
    • 2
  • Lucenilda J. C. Ferraz
    • 3
  • Silvia R. M. Pedro
    • 1
  • Alessandro W. C. Ferreira
    • 4
  1. 1.Departamento de Biologia, Faculdade de Filosofia, Ciências e LetrasUniversidade de São PauloRibeirão PretoBrazil
  2. 2.University of CampinasCampinasBrazil
  3. 3.Centro Universitário de PinheiroUniversidade Federal do MaranhãoPinheiroBrazil
  4. 4.Departamento de BiologiaUniversidade Federal do MaranhãoSão LuísBrazil

Personalised recommendations