T REPRO N DU The International Journal of Plant Reproductive Biology 12(2) July, 2020, pp.128-137 LA C P T I F V O E
B Y
T I
DOI 10.14787/ijprb.2020 12.2. O
E
I L
O C
G O
S I
S
T E
S H Herkogamy variation and alternative pollination modes in an ornithophilous species, T Pyrostegia venusta (Ker Gawl.) Miers (Bignoniaceae)
Priscila Tunes1, 2 and Elza Guimarães2* ¹Graduation Program of Biological Sciences (Botany), Institute of Biosciences, São Paulo State University (Unesp), 18618-970, Botucatu, SP, Brazil ² Laboratory of Ecology and Evolution of plant-animal interactions, Institute of Biosciences, São Paulo State University (Unesp), 18618-689, Botucatu, SP, Brazil. *e-mail : [email protected]. Received : 05. 06. 2020; Revised: 23.06.2020; Accepted and published online: 01.07. 2020 ABSTRACT In species with mixed-mating systems, dichogamy and herkogamy separating male and female functions are commonly present, which may reduce self-pollination. Here, we described variation in herkogamy in Pyrostegia venusta (Bignoniaceae) and investigated its consequences for pollination in this ornithophilous species. Even though this species is described as exhibiting approach herkogamy, we observed intra-plant variations in style length in nature. Therefore, we aimed to answer the following questions: Can P. venusta flowers spontaneously self-pollinate? Can they be pollinated by biotic vectors other than hummingbirds? If so, are the seeds as vigorous as those produced by hummingbird-pollination? Thus, we tested if the flowers could produce fruits by [1] spontaneous self-pollination and by [2] pollination by biotic vectors other than hummingbirds, which we compared to [3] hummingbird-pollination. We also compared the number of seeds/fruit, percentage of seed germination and germination rate among pollination modes. We observed a continuous variation of stigma position in relation to the anthers in the population, and we verified that a single plant may present flowers with different degrees of herkogamy. The probability of fruit set was the same, regardless of pollination mode (p-value=0.69634). The fruits produced by [2] presented more seeds than those produced by [1] and [3] (p-value<0.0001). However, the seeds produced by hummingbird- pollination presented higher viability (p-value=0.0214) and germinated faster (p-value<0.001) than the seeds produced by [1] and [2]. Our results indicate that P. venusta can spontaneously self-pollinate, can be pollinated by small bees, and by hummingbirds. The presence of diverse types of herkogamy within the same plant, suggests the existence of a wider range of possibilities for this species reproduction. Therefore, spontaneous self-pollination and bee-pollination can represent alternative pollination modes and constitute an advantage for this plant species’ reproduction, as they can continue to set fruits and generate new progeny in an unstable pollinator scenario. Keywords : approach herkogamy, hummingbird-pollination, small bee-pollination, spontaneous self-pollination.
The evolutionary transition from outcrossing to self- compatible species (Bartoš et al. 2020), we observe several fertilization is common among Angiosperm species (Stebbins mechanisms related to self-fertilization avoidance in nature. 1974) and combinations of outcrossing and selfing seem to be In self-compatible species, with hermaphrodite flowers, widespread in natural populations (Whitehead et al. 2018). temporal (dichogamy) and spatial (herkogamy) separations Outcrossing in flowering plants is highly dependent on between male and female functions are commonly present, pollinators, and around 94% of angiosperms from tropical which may reduce the occurence of self-pollination (Willmer communities rely on animals for pollen transfer (Ollerton et 2011). Concerning temporal separation, protandry has been al. 2011). Thus, on one hand, pollen limitation (e.g. pollinator commonly described for Bignoniaceae species (Lopes et al. scarcity or lack of compatible mates) has been considered a 2002, Sargent & Otto 2004, Guimarães et al. 2008, Guimarães possible reason for the commonness of mixed mating systems et al. 2015), as well as spatial separation, including ordered (Knight et al. 2005) and for the evolution of selfing (Busch and (approach and reverse) and movement herkogamy (Webb and Delph 2012). On the other hand, pollinators have been Loyd 1986, Guimarães et al. 2008, Milet-Pinheiro et al. 2009, considered as agents that may constrain the evolution of Bittencourt et al. 2011, Ai et al. 2013, Souza et al. 2017). In selfing (Devaux et al. 2014), since most of them have a strong Bignoniaceae flowers, stigmas are usually contacted first, as dependency on floral resources to complete their life cycles, the pollinator enters the floral tube, and pollen is picked up and selfing may reduce flower investment, compromising subsequently, characterizing approach herkogamy; primary and secondary attractants (Spigler and Kalisz 2017). additionally, many Bignoniaceae species have sensitive Despite controversial questions about the evolution of stigmas that are responsible for movement herkogamy (Souza mating systems and variation of crossing rates in self- et al. 2017 and references therein). 2020 Herkogamy variation and alternative pollination modes in an ornithophilous species, Pyrostegia .... 129
Even though cross-pollination can be favoured by the axillary panicles (Pool 2008). Its flowers are zygomorphic, presence of herkogamy and protandry, autonomous self- presenting a long orange tubular corolla with curved-back pollination can represent a potential advantage when petal lobes, with four exerted didynamous stamens, and a pollinators are scarce (Lemos et al. 2020). Additionally, recent syncarpic gynoecium with bilocular ovary, long style and studies show that some plant species may exhibit a continuum bilabiate stigma, with sensitive lobes that close under stimuli in the separation of sex organs, varying from reverse to (Pool 2008). Its flowers are functionally hermaphroditic, approach herkogamy (Lázaro et al. 2020 and references protandrous and self-compatible (Pool 2008). Pollen fertility therein). This variation has been ascribed to contrasting and germination on the stigmatic surface of P. venusta flowers selection exerted by different pollinators, e.g., hawkmoths can be affected by abiotic factors, precisely by temperature selecting for reverse herkogamy and hummingbirds selecting and relative humidity (Singh et al. 2009b). Flowers present a for approach herkogamy (Kulbaba and Worley 2013). nectary disk that produces copious amounts of sucrose-rich Additionally, selection can act reducing herkogamy when nectar, approximately 30 µl, throughout its life, with a pollinators are scarce (de Vos et al. 2012), favouring, for concentration of 28% of sugars (Galetto et al. 1994). We example, flowers without ordered herkogamy, which may deposited a voucher specimen in the Herbarium BOTU “Irina allow spontaneous self-pollination and fruit set in pollinator Delanova de Gemtchujnicov” (voucher number 30788). unstable environments. However, little is known about the Authorization for collection of biological samples registered consequences of herkogamy variation on the reproduction of on Sisgen under #A90A83C. natural populations ( Lázaro et al. 2020). EXPERIMENTAL PROCEDURES Aiming to bring more information about this issue, we investigated a natural population of Pyrostegia venusta (Ker- Herkogamy variation survey—We surveyed P. venusta Gawl.) Miers,, a Bignoniaceae species, which was described plants growing at the edge of a seasonal tropical forest as self-compatible in its original habitat (Gobatto-Rodrigues fragment, in Botucatu municipality, São Paulo state, Brazil. & Stort 1992). However, Singh et al. (2009a) identified self- We sampled 20 to 30 functional flowers from the surveyed incompatible individuals cultivated by vegetative plants (n = 527 flowers from 20 plants) in order to evaluate the reproduction in India. Substantial inter-population variation in frequency of flowers presenting approach herkogamy (stigma mating system is common in nature, with around 60% of the above the anthers), reverse herkogamy (stigma below the studied species showing at least one mixed population anthers) and no ordered herkogamy (stigmas and anthers at the (Whitehead et al. 2018). Thus, besides having populations or same level) in the population. individuals with different (in) compatibility systems, the Floral visitors—We performed approximately 40 hours of species exhibits protandry, approach and movement focal observation to identify the floral visitors of P. venusta herkogamy (Gobatto-Rodrigues and Stort 1992, Pool 2008), and to characterize their behaviour, in order to identify if any and is foraged by several hummingbird species that pollinate visitors other than hummingbirds could end up pollinating its its flowers when searching for nectar as trophic resource flowers in the study population. (Galetto et al. 1994). Even though this species is described as Pollination mode—We tested if P. venusta flowers could exhibiting approach herkogamy, there is an intra-plant produce fruits by [1] spontaneous self-pollination and by [2] variation in style length in cultivated plants (Singh et al. pollination by biotic vectors other than hummingbirds. Then, 2009a) and in natural populations (authors’ personal we compared the frequency of fruit production in both observation). In these cases spontaneous self-pollination may pollination modes to the production of fruits by [3] occur, by the anthers and the stigma touching each other in hummingbird-pollination. If other pollination vectors are flowers without pronounced ordered herkogamy, which in involved in P. venusta sexual reproduction, we expect to find a self-compatible individuals may lead to self-fertilization. higher number of fruits in [2] pollination by biotic vectors Therefore, our aim is to answer the following questions: other than hummingbirds than in [1] spontaneous self- (i) Can P. venusta flowers spontaneously self-pollinate? (ii) pollination. As it was impossible to experimentally exclude Can they be pollinated by biotic vectors other than the possibility of spontaneous self-pollination in test [2], the hummingbirds? (iii) If spontaneous self-pollination and fruits produced by this treatment in fact corresponded to [1] pollination by other biotic vectors are possible, are the seeds as plus the effect of pollination by biotic vectors other than vigorous as those produced by hummingbird-pollination? hummingbirds. To investigate the natural frequency of [1] spontaneous MATERIAL AND METHODS self-pollination, we randomly isolated, with voile bags, 117 Study species—Pyrostegia venusta (Ker-Gawl.) Miers is a floral buds at pre-anthesis stage distributed in six plants. We neotropical vine. Its occurrence ranges from the northeast accompanied them until complete flower abscission or fruit coast of Brazil to the northeast of Argentina (35-58°W and 5- formation in order to evaluate if P. venusta flowers could 30°S) (Pool 2008). Pyrostegia venusta presents terminal or produce fruits and viable seeds without pollen transfer by 130 The International Journal of Plant Reproductive Biology 12(2) July, 2020, pp.128-137
biotic vectors (see Table 1 for flower distribution per plant). To respectively. Finally, we tested if pollination mode affected test for [2] pollination by biotic vectors other than seed germination speed using a generalized linear model hummingbirds, e.g. small bees that visited the flowers (GLM) with Poisson error distribution. We carried out gathering pollen, we randomly isolated 116 pre-anthesis floral statistical analyses in R v. 4.0.0 (R Core Team 2020) with buds from hummingbirds, also using six plants, by installing standard and additional packages: fifer (Fife 2017), ggplot2 cylindrical wire cages (2 cm mesh). We accompanied them (Wickham 2016), lme4 (Bates et al. 2015) and nlme (Pinheiro until complete flower abscission or fruit formation (see Table et al. 2018). 2 for flower distribution per plant). Additionally, to test for [3] RESULTS AND DISCUSSION hummingbird-pollination, we exposed 116 recently-opened flowers exhibiting approach herkogamy in 25 plants to Herkogamy variation survey—Variation in herkogamy was hummingbird visits and accompanied them until complete registered in this natural population of P. venusta, similarly to flower abscission or fruit formation (see Table 3 for flower the observed in other P. venusta individuals (Singh et al. distribution per plant). It is noteworthy that for this test we 2009a) and in Lonicera implexa by Lázaro et al. (2020). We only used flowers showed an evident separation between observed a continuous variation of stigma position in relation stigma and anthers in order to avoid any interference of self- to the anthers, varying from 3.00 cm below than the anthers to pollination or of pollination by other floral visitors. 1.11 cm above them. Within this gradient, we could identify We kept all the produced fruits individually bagged until flowers with clear ordered herkogamy varying from approach their dehiscence. Then, we collected them and counted their (85.58%; Fig. 1A-B) to reverse herkogamy (2.84%; Fig. 1C), seeds. We randomly took nine to ten seeds from each fruit and sensu Willmer (2011), passing through flowers that presented uniformly conditioned them in germination boxes covered the stigma at the same level of the anthers (without ordered with filter paper. We distributed the boxes with the seeds in a herkogamy; 11.58%; Fig. 1D). germination chamber at 30 ± 1°C with artificial white light A single plant may present flowers with different degrees (12-hour photoperiod), and irrigated them daily with distilled of herkogamy (Fig. 2), similarly to the described for cultivated water. We monitored the seeds daily in order to evaluate the individuals of the same species (Singh et al. 2009a). It is percentage and speed of seed germination. When the amount suggested that temperature during flowering might modulate of germinated seeds was stable for five consecutive days, we the frequency in which each type of flower occurs (Singh et al. ended the monitoring. We considered that germination had 2009a). Additionally, we observed that these types of flowers occurred when the radicle protrusion reached 1 mm of length. occurred in variable frequencies within the plants of the Additionally, in case any of the plants involved in the population (χ² = 130.71, df = 38, p-value < 0.001; Fig. 2). The previously described experiments [1] and [2] presented fruits, vast majority of the plants (70%) presented two types of we performed an additional experiment to test for apomixis. flowers, with approach herkogamy and without ordered For that, we emasculated flowers in those plants, isolated them herkogamy; whereas 20% of the plants presented all the three with voile bags and accompanied them until complete flower types of flowers; 5% of the plants presented only flowers abscission or fruit formation. exhibiting approach herkogamy; and 5% of the plants Seed production, viability and vigour—We compared the presented two types of flowers, with approach and reverse seeds produced by both [1] spontaneous self-pollination and herkogamy (Fig. 2). The fact that 90% of the plants presented [2] pollination by biotic vectors other than hummingbirds to flowers without ordered herkogamy allows them to the seeds produced by [3] hummingbird-pollination. For that, spontaneously self-pollinate due to the proximity between we registered the number of seeds per fruit, percentage of seed stigma and anthers (Lázaro et al. 2020). On the other hand, germination and germination rate. reverse herkogamy was rarer and was only present in 25% of Statistical analyses—Pearson’s chi-squared test to compare the plants from the population. These flowers are highly the frequency of flowers with approach herkogamy, with unlikely to be spontaneously self-pollinated, especially reverse herkogamy and without ordered herkogamy within the because flowers are horizontally arranged in the population was performed. In addition, we modelled the inflorescences, in such a way that pollen could not fall from probability of a flower turning into fruit (binary variable) the anthers onto the stigmas. among the three treatments, using GLMM with binomial error Floral visitors—We observed the hummingbirds Phaethornis distribution, considering pollination mode as a fixed factor pretrei (Lesson and Delattre 1839) (Phaethornithinae), and individual plant as a random variable. Additionally, we Chlorostilbon lucidus (Shaw 1812) (Trochilinae) and compared the number of seeds per fruit (integer variable) and Polytmus guainumbi (Pallas 1764) (Trochilinae) (Fig. 3A) compared the proportion of germinated seeds (continuous acting as pollinators in our study system, by legitimately variable) among the three treatments, using generalized linear visiting P. venusta flowers. Hummingbirds can act as models (GLM) with Poisson and Binomial error distribution, pollinators in flowers with approach herkogamy and without 2020 Herkogamy variation and alternative pollination modes in an ornithophilous species, Pyrostegia .... 131
Fig.1–Variation of herkogamy registered in flowers of Pyrostegia venusta (Bignoniaceae) plants growing on edges of a semideciduous forest in Southeastern Brazil. A. Lateral view of a flower with approach herkogamy. Scale bar: 0.97 cm. B. Frontal view of a flower with approach herkogamy. Scale bar: 0.8 cm. C. Lateral view of a flower with reverse herkogamy with gynoecium detached from the flower in order to show the stigma's position in relation to the anthers' (arrow). Scale bar: 1.29 cm. D. Lateral view a flower without ordered herkogamy. Scale bar: 0.98 cm.
30
s 20 r e w o l f
Reverse herkogamy f o
No ordered herkogamy r
e Approach herkogamy b m
u 10 N
0
0 5 10 15 20 Plants Fig. 2 – Number of flowers with reverse herkogamy, without ordered herkogamy and with approach herkogamy registered in Pyrostegia venusta (Bignoniaceae) plants growing on edges of a semideciduous forest in Southeastern Brazil. 132 The International Journal of Plant Reproductive Biology 12(2) July, 2020, pp.128-137
Fig. 3 – Floral visitors of Pyrostegia venusta (Bignoniaceae) plants growing on edges of a semideciduous forest in Southeastern Brazil. A. The hummingbird Polytmus guainumbi (Trochilinae) collecting nectar by legitimately visiting a flower with approach herkogamy, note that the flower’s reproductive structures touch the hummingbird’s forehead. Scale bar: 1.95 cm. B. Apis mellifera scutellata collecting pollen in a recently opened flower, note the abdomen of the bee on the top touching the stigma when it is scraping an anther above it. Scale bar: 0.49 cm. C. Xylocopa sp. robbing nectar by piercing the corolla basis. Scale bar: 0.96 cm. 0 . 1 8 . 0 t e s
t i 6 u . r f 0
f o
y t i l i b a 4 . b 0 o r P 2 . 0 0 . 0
[1] [2] [3] POLLINATION MODE Fig. 4 – Results showing the probability of Pyrostegia venusta (Bignoniaceae) flowers turning into fruit in the tests for [1] spontaneous self- pollination, [2] pollination by biotic vectors other than hummingbirds + spontaneous self-pollination and [3] hummingbird-pollination. 2020 Herkogamy variation and alternative pollination modes in an ornithophilous species, Pyrostegia .... 133 0 0 . 6 1 8 0 . 5 0 ) t i u % r ( f
/ n s 6 o 0 . d i t 4 0 e a e n s i
f m o r
4 r e 0 . e g 3 0
b d m e u e N S 2 0 . 2 A 0 B 0 0 . 1 0
[1] [2] [3] [1] [2] [3] POLLINATION MODE POLLINATION MODE 1 3 0 3 ) s y a d 5 (
2 d e e p 0 s
2 n o i t a 5 n i 1 m r e G 0 1 C 5
[1] [2] [3] POLLINATION MODE Fig. 5- Results comparing viability and vigour of Pyrostegia venusta (Bignoniaceae) seeds formed by [1] spontaneous self-pollination, [2] pollination by biotic vectors other than hummingbirds + spontaneous self-pollination and [3] hummingbird-pollination. A. Number of seeds per fruit. B. Percentage of seed germination. C. Germination speed. ordered herkogamy by contacting the flowers’ reproductive pollination in flowers without ordered herkogamy, as the structures with their foreheads, similar to the described for reproductive structures are close and can be mutually and another ornithophilous Bignoniaceae species (Bittencourt and simultaneously contacted. Especially because every time this Semir 2004). Whereas, in flowers exhibiting reverse bee collected pollen, we could see pollen grains adhered to its herkogamy, the stigma is located deep into the corolla tube, so body, sometimes to its thorax and wings, but mainly to its legs. that the only way in which those flowers could be pollinated Due to the described bee behaviour, these bees tend to perform would be if some pollen grains adhered to the hummingbirds’ pollen transfer mainly within a flower, and considering that P. bill and this portion of the bill touched the stigma of the same venusta is self-compatible (Gobatto-Rodrigues and Stort or another flower. 1992), they can act in self-fertilization. After visiting a flower, Additionally, we observed individuals of Apis mellifera usually the bees flew to a nearby flower in the same plant, scutellata (Lepeletier 1836) actively collecting pollen from which could contribute to geitonogamy (Guimarães et al. the anthers of functional flowers (Fig. 3B). These small bees 2008). It is remarkable that these bees can only act as usually approached the flowers from the front and clung to the pollinators in flowers without ordered herkogamy, whereas in anthers, and then they started to scrape them and collect flowers with reverse and approach herkogamy the bees do not pollen. The bees would move from one anther to another, reach the stigma surface when collecting pollen. Thus, in those sometimes clinging to more than one at a time and walking flowers these bees acted as pollen thieves, as described for over the stigma. Thus, the manoeuvres performed by A. other Bignoniaceae species (Guimarães et al. 2008, Quinalha mellifera scutellata in order to collect pollen could lead to et al. 2017). 134 The International Journal of Plant Reproductive Biology 12(2) July, 2020, pp.128-137
We also observed individuals of Trigona spinipes verified that the probability of a given flower turning into a (Fabricius 1793) and Xylocopa sp. (Latreille 1802) (Fig. 3C) fruit was the same, regardless of pollination mode (Z = 0.390, visiting P. venusta flowers, however, due to their behaviour, p-value = 0.69634; Fig. 4). The results of these experiments these bees were only involved in floral larceny (Inouye 1980). indicate that P. venusta can spontaneously self-pollinate, and Both species landed externally on the corolla and walked can additionally be pollinated by small pollen-collecting bees. towards its basis. Trigona spinipes cut a small role with its Even though these bees are not adjusted to the floral mouth apparatus and inserted its head through the role, dimensions of these ornithophilous species, their behaviour apparently consuming nectar, whereas Xylocopa sp. allows us to infer that they may act as pollinators of P. venusta. perforated the corolla and introduced its mouth apparatus Nevertheless, we expected that these bees will favour mainly through the perforation to obtain nectar (Fig. 3C). Thus, both self-pollination, since they present self-grooming behaviour, species were acting as nectar robbers in the study system, constantly cleaning their head, thorax and legs, gathering the which is common in Bignoniaceae species (Gottsberger and pollen grains and packing them into the corbiculae in their Silberbauer-Gottsberger 2006, Guimarães et al. 2008, hind legs (Jander 1976), where it is no longer suited to be Guimarães et al. 2015, Quinalha et al. 2017, Souza et al. transferred to another flower’s stigma. Thus, bee-pollination 2019). in this scenario is restricted to a narrow window of time before Pollination mode—When testing for [1] spontaneous self- the bees self-groom again. This behaviour reduces the bees’ pollination, we found that two out of 117 bagged flowers role in the cross-fertilization in this system. If these small bees (1.7%) turned into fruits (both in the same plant), with 29 and presented a substantial role in P. venusta pollination, we 30 seeds, being that 60.00% and 66.67% of their seeds would expect that the fruit set from treatment [2] would have germinated, respectively (Table 1). Additionally, when testing been higher than the fruit set by [1] spontaneous self- for [2] pollination by biotic vectors other than hummingbirds, pollination. However, our results do not show a significant we found that five out of the 116 caged flowers (4.3%) turned difference between those treatments. into fruits (all of them in the same plant), with 48.80 ± 3.90 It is remarkable that all plants in the population (100%) seeds per fruit, being that 70.00 ± 12.25% of their seeds presented flowers with approach herkogamy and those germinated (Table 2). When evaluating [3] hummingbird- flowers were the most common type, constituting 85.58% of pollination, we observed that three from the 116 flowers the flowers. The abundance of flowers with approach (2.6%) exposed to hummingbird visitation turned into fruits, herkogamy shows the importance of hummingbird- with 90.00 ± 10.00% of their seeds germinated (Table 3). pollination for the reproduction of this highly specialized After observing that the fruits formed in both pollination system sensu Armbruster (2017). Even though experimental treatments, [1] spontaneous self-pollination and hummingbirds can promote geitonogamy, they may represent [2] pollination by biotic vectors other than hummingbirds, a crucial advantage for the plant species as they favour belonged to the same plant, we tested this plant for apomixis (n outcrossing. This is because hummingbirds present a high = 38 flowers). We performed this test to exclude the possibility metabolic rate, a reduced body size, and rely on sugars for that this plant was able to produce apomictic seeds, ensuring obtaining energy (Suarez and Welch 2017), thus they need to that the result obtained was, in fact, due to spontaneous self- visit a substantial amount of flowers, visiting a plethora of pollination or bee-pollination. As the bagged and caged plants per day to meet their caloric needs. Even though flowers flowers from the other sampled plants produced no fruits, we without ordered herkogamy occur in low frequency within the did not check those plants for apomixis. There were no signs of plants, they are present in 90% of the plants, which means that apomixis in P. venusta, differently from other species of the the majority of the plants would still be able to produce fruits same tribe, Bignonieae (see references in Firetti 2018). Thus, in the absence of hummingbirds, be it by spontaneous self- we can assert that indeed there are cases of spontaneous self- pollination, be it by bee-pollination. Thus, these three pollination and of bee-pollination in P. venusta flowers. These pollination modes can be complementary and guarantee results highlight that in natural populations, of P. venusta reproduction via seeds in diverse environmental conditions. occur self-compatible plants, as shown by Gobatto-Rodrigues Seed production, viability and vigour— We verified that the & Stort (1992). However, this species seems to present a fruits produced by bee-pollination presented more seeds than flexible reproductive system, as shown for other Bignoniaceae those produced by spontaneous self-pollination and by species (Bertin and Sullivan 1988, Bianchi et al. 2005, hummingbird-pollination (F = 52.62, p-value < 0.0001, Fig. Bittencourt and Semir 2006, Guimarães et al. 2008), 5A). However, the seeds produced by hummingbird- presenting a mixture of self-compatible and self-incompatible pollination presented higher viability (F = 8.136, p-value = individuals within natural populations. Thus, the self- 0.0214, Fig. 5B) and germinated faster (F = 14.42, p-value < incompatibility found by Singh et al. (2009a) for cultivated 0.001, Fig. 5C) than the seeds produced by both spontaneous individuals, might be a consequence of clonal reproduction of self-pollination and by hummingbird-pollination. The faster self-incompatible individuals for ornamental purposes. We 2020 Herkogamy variation and alternative pollination modes in an ornithophilous species, Pyrostegia .... 135
seed germination in hummingbird pollination could be due to inbreeding depression and hybrid vigor in the larger pollen loads deposited by hummingbirds Balfourodendron riedelianum (Engl.) Engl.(Rutaceae). (Jσhannsson and Stephenson 1997). Additionally, pollen Conserv. Genet. 21 305–317. transfer among distant plants may increase seed vigour in Ai H, Zhou W, Xu K, Wang H and Li D 2013. The reproductive general (Carpenter and Guard 1950), which could happen in strategy of a pollinator- limited Himalayan plant, hummingbird pollination. On the other hand, selfing and Incarvillea mairei (Bignoniaceae). BMC Plant Biol. mating among relatives, which can result from spontaneous 13(1) 195. self-pollination and bee-pollination, can lead to reduced fertility and seed viability, which are signs of inbreeding Armbruster WS 2017. The specialization continuum in depression and can indicate a greater risk of population pollination systems: diversity of concepts and extinction (Aguiar et al. 2020 and references therein). implications for ecology, evolution and conservation. The presence of herkogamy influences pollen placement Funct. Ecol. 31(1) 88-100. onto the pollinator body and favours its transfer to the stigma Barrett SC 2002. The evolution of plant sexual diversity. Nat. of other flowers (Webb and Lloyd 1986, Barrett 2002). Rev. Genet. 3(4) 274-284. However, the occurrence of herkogamy variation within the same plant, suggests a flexibility in plant reproduction Bartoš M, Janeček Š, Janečková P, Padyšáková E, Tropek R, (Bartoš et al. 2020, Lázaro et al. 2020), as different floral Götzenberger L, … and Jersáková J 2020. Self morphotypes can be associated with pollination by different compatibility and autonomous selfing of plants in vectors (Lázaro et al. 2020). meadow communities. Plant Biol. 22(1) 120-128. In conclusion, the presence of alternative pollination Bates D, Maechler M, Bolker B and Walker S 2015. Fitting modes can constitute an advantage for P. venusta reproduction, Linear Mixed-Effects Models Using lme4. J. Stat. Softw. as they can continue to set fruits and generate new progenies in 67(1) 1-48; doi:10.18637/jss.v067.i01. a scenario of unpredictable environmental disturbances. On the one hand, hummingbird-pollination seems to be more Bertin RI and Sullivan M 1988. Pollen interference and efficient overall, producing seeds with higher viability and cryptic self-fertility in Campsis radicans. Am. J. Bot. 75 seeds that germinate fast, which can be an advantage in the 1140–1147. unstable environments where this species occurs, like forest Bianchi MB, Harris SA, Gibbs PE and Prado DE 2005. A study edges and road margins. On the other hand, bee-pollination can of the mating system in Dolichandra cynanchoides constitute an advantage for this species when hummingbirds (Bignoniaceae): an Argentinian Chaco woodlands liane are scarce, leading to the production of more seeds per fruit, with a late-acting self-incompatibility. Plant Syst. Evol. which translate into more chances of generating new 251 173–181. individuals, even though the seed formed by bee-pollination take longer to germinate and are not so vigorous as the ones Bittencourt Jr NS and Semir J 2004. Pollination biology and produced by hummingbirds. Additionally, spontaneous self- breeding system of Zeyheria montana (Bignoniaceae). pollination can constitute an even greater advantage for this Plant Syst. Evol. 247(3-4) 241-254. species, as this capability ensures that, even in conditions and Bittencourt NS and Semir J 2006. Floral biology and late- places where neither hummingbirds nor bees are present; at acting self-incompatibility in Jacaranda racemosa least a small percentage of the flowers can form fruits in almost (Bignoniaceae). Aust. J. Bot. 54 315–324. all plants in the population. Bittencourt Jr NS, Pereira Jr EJ, de Souza Săo-Thiago P and Acknowledgments—We thank the ‘Coordenaηão de Semir J 2011. The reproductive biology of Cybistax Aperfeiηoamento de Pessoal de Nível Superior - Brasil antisyphilitica (Bignoniaceae), a characteristic tree of the (CAPES)’ - Finance Code 001 for the scholarship granted to P. South American savannah-like “Cerrado” vegetation. Tunes. We thank Dr. F. F. de Oliveira for the identifiçation of Flora. 206(10) 872-886. Apis mellifera scutelatta and Xylocopa sp.. We also thank the students from the laboratory of ‘Ecology and Evolution of Busch JW and Delph LF 2012. The relative importance of Plant-Animal Interactions’ for field support. reproductive assurance and automatic selection as hypotheses for the evolution of self-fertilization. Ann. REFERENCES Bot-London. 109(3) 553-562. Aguiar BI, Freitas ML, Zannato AS, Tambarussi EV, Moraes Carpenter IW and Guard AT 1950. Some effects of cross- ML, Ambrosano MN, ... and Sebbenn AM 2020. The pollination on seed production and hybrid vigor of Tulip effects of pollen dispersal and mating pattern on tree. J. Forest. 48(12) 852-5. 136 The International Journal of Plant Reproductive Biology 12(2) July, 2020, pp.128-137
de Vos JM, Keller B, Isham ST, Kelso S and Conti E 2012. of plant reproduction: pattern and process. Annu. Rev. Reproductive implications of herkogamy in homostylous Ecol. Evol. Syst. 36 467-497. primroses: variation during anthesis and reproductive Kulbaba MW and Worley AC 2013. Selection on Polemonium assurance in alpine environments. Funct. Ecol. 26(4) brandegeei (Polemoniaceae) flowers under 854-865. hummingbird pollination: in opposition, parallel, or Devaux C, Lepers C and Porcher E 2014. Constraints imposed independent of selection by hawkmoths? Evolution. by pollinator behaviour on the ecology and evolution of 67(8) 2194-2206. plant mating systems. J. Evolution. Biol. 27(7) 1413- Lázaro A, Seguí J and Santamaría L 2020. Continuous 1430. variation in herkogamy enhances the reproductive Fife D 2017. fifer: A Biostatisticians Toolbox for Various response of Lonicera implexa to spatial variation in Activities, Including Plotting, Data Cleanup, and Data pollinator assemblages. AoB Plants. 12(1) plz078. Analysis. R package version 1.1. https://CRAN.R- Lemos AL, Moreira MM, Benevides CR, Miranda AS, project.org/package=fifer Rodarte ATA and de Lima HA 2020. Reproductive Firetti F 2018. Apomixis in Neotropical vegetation. Vegetation biology of Prepusa hookeriana (Gentianaceae): an 7 129-149. endangered species of high-altitude grasslands in Brazil. Braz. J. Bot. 43 379–387. Galetto L, Bernardello LM and Juliani HR 1994. Characteristics of secretion of nectar in Pyrostegia Lopes AV, Vogel S and Machado IC 2002. Secretory venusta (Ker Gawl.) Miers (Bignoniaceae). New Phytol. trichomes, a substitutive floral nectar source in Lundia A. 127(3) 465-471. DC. (Bignoniaceae), a genus lacking a functional disc. Ann. Bot-London. 90(2) 169-174. Gobatto-Rodrigues AA and Stort MNS 1992. Biologia floral e Milet-Pinheiro P, Carvalho AT, Kevan PG and Schlindwein C reproduçăo de Pyrostegia venusta (Ker-Gawl) Miers 2009. Permanent stigma closure in Bignoniaceae: (Bignoniaceae). Braz. J. Bot. 15(1) 37-41. mechanism and implications for fruit set in self- Gottsberger G and Silberbauer-Gottsberger I 2006. Life in the incompatible species. Flora. 204(1) 82-88. Cerrado. Reta, Ulm. Ollerton J, Winfree R and Tarrant S 2011. How many Guimarăes E, Di Stasi LC and Maimoni-Rodella RDCS 2008. flowering plants are pollinated by animals?. Oikos. Pollination biology of Jacaranda oxyphylla with an 120(3) 321-326. emphasis on staminode function. Ann. Bot-London. Pinheiro J, Bates D, DebRoy S, Sarkar D and R Core Team 102(5) 699-711. 2018. nlme: Linear and Nonlinear Mixed Effects Models. Guimarăes E, Nogueira A, Santos EMG, Netto CGD and R package version 3.1-137,
(Bignoniaceae) with special reference to seedlessness. J. Stebbins GL 1974. Flowering plants: evolution above the Plant Repro. Biol. 1(1) 87-92. species level. Harvard University Press, Cambridge, Massachusetts. Singh S, Rana A and Chauhan SVS 2009b. Impact of environmental changes on the reproductive biology Suarez RK and Welch KC 2017. Sugar metabolism in in Pyrostegia venusta Presl. J. Environ. Biol. 30(2) 271- hummingbirds and nectar bats. Nutrients. 9(7) 743. 273. Webb CJ and Lloyd DG 1986. The avoidance of interference Souza CV, Salvador MV, Tunes P, Di Stasi LC and between the presentation of pollen and stigmas in Guimarăes E 2019. I’ve been robbed!–Can changes angiosperms II. Herkogamy. New Zeal. J. Bot. 24(1) 163- in floral traits discourage bee pollination?. PloS one. 178. 14(11). Whitehead MR, Lanfear R, Mitchell RJ and Karron JD 2018. Souza CV, Nepi M, Machado SR and Guimarăes E 2017. Plant mating systems often vary widely among Floral biology, nectar secretion pattern and fruit set of a populations. Frontiers in Ecology and Evolution. 6 38. threatened Bignoniaceae tree from Brazilian tropical Wickham H 2016. ggplot2: Elegant Graphics for Data forest. Flora. 227 46-55. Analysis. Springer-Verlag New York. Spigler RB and Kalisz S 2017. Persistent pollinators and Willmer P 2011. Pollination and floral ecology. Princeton the evolution of complete selfing. Am. J. Bot. 104(12) University Press. 1783-1786.