The habits and floral attributes of the species examined in the present study are described in Table 2. Among the floral attributes evaluated, only four were identified as being characteristic of the level of specialization of the flower and presented restrictions to accessing floral resources: floral shape; the type of floral resource offered (nectar, pollen, or oil); the spatial separation of the anthers and stigmas; temporal differences in pollen availability and stigma receptivity.
Table 2
Habit and floral attributes of the plant species studied.
Plant species studied
|
Plant habit
|
Floral shape
|
General color
|
Nectar guides or pollen
|
Attractiveness (days)
|
Odors
|
Ultraviolet light reflections
|
Reward
|
Separations of the anthers and stigmas
|
Spatial
|
Temporal
|
Bignoniaceae
|
|
|
|
|
|
|
|
|
|
|
|
Arrabidaea chica
|
Liana
|
Gullet
|
Purple
|
Yes
|
1
|
Yes
|
No
|
Nectar and pollen
|
No
|
No
|
Arrabidaea florida
|
Liana
|
Gullet
|
White
|
Yes
|
1
|
Yes
|
Yes
|
Nectar and pollen
|
No
|
No
|
Cuspidaria convoluta
|
Liana
|
Gullet
|
Pink
|
Yes
|
1
|
Yes
|
Yes
|
Nectar and pollen
|
No
|
No
|
Adenocalymma bracteatum
|
Liana
|
Gullet
|
Yellow
|
Yes
|
1
|
Yes
|
Yes
|
Nectar
|
No
|
No
|
Pyrostegia venusta
|
Liana
|
Tubular
|
Red
|
No
|
1
|
Yes
|
Yes
|
Nectar and pollen
|
Yes
|
No
|
Malpighiaceae
|
|
|
|
|
|
|
|
|
|
|
|
Diplopterys pubipetala
|
Liana
|
Dish
|
Yellow
|
Yes
|
3
|
Yes
|
Yes
|
Oil
|
No
|
No
|
Byrsonima intermedia
|
Shrub
|
Dish
|
Yellow
|
No
|
3
|
Yes
|
Yes
|
Oil and pollen
|
No
|
No
|
Banisteriopsis cf. campestris
|
Liana
|
Dish
|
Pink
|
Yes
|
1
|
Yes
|
Yes
|
Oil and pollen
|
No
|
No
|
Banisteriopsis laevifolia
|
Liana
|
Dish
|
Yellow
|
No
|
2
|
Yes
|
Yes
|
Oil and pollen
|
No
|
No
|
Asteraceae
|
|
|
|
|
|
|
|
|
|
|
|
Trixis antimenorrhoea
|
Herb
|
Capitula
|
White
|
Yes
|
7
|
Yes
|
Yes
|
Nectar and pollen
|
Yes a
|
No b
|
Eupatorium maximalianii
|
Herb
|
Capitula
|
Purple
|
No
|
7
|
Yes
|
Yes
|
Nectar and pollen
|
Yes a
|
No b
|
Eupatorium cf. dimorpholepis
|
Shrub
|
Capitula
|
White
|
No
|
6
|
Yes
|
Yes
|
Nectar and pollen
|
Yes a
|
No b
|
Fabaceae
|
|
|
|
|
|
|
|
|
|
|
|
Senna obtusifolia
|
Shrub
|
Dish
|
Yellow
|
No
|
2
|
Yes
|
No
|
Pollen
|
Yes
|
No
|
Senegalia sp.
|
Shrub
|
Brush
|
White
|
No
|
2
|
Yes
|
Yes
|
Nectar and pollen
|
No
|
No
|
Senegalia polyphylla
|
Tree
|
Brush
|
White
|
No
|
2
|
Yes
|
Yes
|
Nectar
|
No
|
No
|
Sapindaceae
|
|
|
|
|
|
|
|
|
|
|
|
Serjania caracasana
|
Liana
|
Dish-polypetalous
|
White
|
Yes
|
2
|
Yes
|
Yes
|
Nectar and pollen
|
Yes
|
No
|
Matayba guianensis
|
Tree
|
Dish
|
Green
|
No
|
3
|
Yes
|
No
|
Nectar and pollen
|
Yes
|
No
|
Lamiaceae
|
|
|
|
|
|
|
|
|
|
|
|
Aegiphilla sellowiana
|
Shrub
|
Tubular
|
White
|
No
|
1
|
Yes
|
No
|
Nectar and pollen
|
Yes
|
No
|
Rhamniaceae
|
|
|
|
|
|
|
|
|
|
|
|
Gouania cf. latifolia
|
Liana
|
Dish
|
White
|
No
|
3
|
Yes
|
No
|
Nectar and pollen
|
Yes
|
No
|
a = With spatial separation between the anthers and stigmas in the same capitulum, but not in the same flower. |
b = No temporal difference between stigma receptivity and the desiccation of the anthers in the capitula, although temporal differences persist in the flowers. |
Five body regions of the floral visitors were identified as being involved in pollen transfer to the stigmas:
1. Dorsal region of the thorax - nototribic (Table 3; Online Resource – Fig. 1). The gullet-shaped flowers of Arrabidaea chica, Arrabidaea florida, Cuspidaria convoluta, and Adenocalymma bracteatum (Bignoniaceae) were pollinated by pollen from this body region of the insects. In order for pollination to occur, the floral visitor has to be collecting nectar during legitimate floral visits (entering through the corolla mouth to gather nectar by suction), and must have sufficient body size to allow contact between the dorsal region of its thorax and the reproductive organs of the plant. To reach the nectar, the visiting insect must penetrate the corolla tube, and the dorsal region of its thorax must simultaneously touch the anthers and the stigmas. After foraging on various flowers, the dorsal region of the insect’s body can be seen to be stippled with pollen (Online Resource – Fig. 1d). The fruits of groups of plants of the same species will demonstrate high proportions of seeds originating from cross-pollination, reflecting the high numbers of pollinating insects foraging on numerous different plants during a single outing. Visitors with body diameters smaller than the corolla tube opening, on the other hand, can retrieve nectar without touching the floral reproductive structures – characterizing nectar theft (Tabela 3). Nectar robbing also frequently occurs, especially by Oxaea flavescens (Apidae).
2. Ventral region of the thorax and abdomen (Table 3, Online Resource – Fig. 2). This type of pollination was especially observed among Malpighiaceae species (Diplopterys pubipetala, Byrsonima intermedia, Banisteriopsis cf. campestris, and Banisteriopsis laevifolia), as they evidenced typically open floral forms; it was also seen in species with brush -shaped flowers [Senegalia polyphylla and Senegalia sp. (Fabaceae)]. The ventral surfaces of the thoraxes and abdomens of the bees will be strongly pressed against the anthers and stigmas of Malpighiaceae species as they attempt to extract oils with their anterior and median legs. As the stigmas are located very near the anthers, pollen is frequently deposited on the bees during foraging. Cross-pollination is therefore very common in those species, as insect pollinators were observed visiting numerous different plants before completing their oil caches and returning to their nests. Among plants that produce brush-shaped flowers, on the other hand, the simple contact of the ventral portions of the thorax or abdomen with the anthers and stigmas and can result in successful pollination. The production of large numbers of flowers by those two species will tend to favor endogamic pollination.
3. Frontal region of the head (Table 3, Online Resource – Figs. 3 and 4). We did not observe a basic pattern of floral shape associated with pollination effected by visitors that use their mouthpieces to extract floral resources. Pollination occurs in this way in open, brush-shaped, tubular, polypetalous, and capitular-type flowers. When the resource collected is nectar (Online Resource – Fig. 3), the frontal portion of the pollinators’ heads constantly enter into contact with the anthers, and pollen grains will adhere to that part of their bodies. During foraging on different flowers, the pollen-impregnated body structures of those insects will enter into contact with the stigma and deposit part of the pollen load onto those reproductive organs, thus effecting pollination. If the resource collected is pollen (Online Resource – Fig. 4), the pollen-impregnated heads of those bees will enter into direct contact with the stigmas of the flowers. That type of pollination is possible because some pollen collection techniques using mouth parts can deposit material onto other body parts (such as the tibias of their hind legs), and the quantities of pollen transported in these cases will be related to the size of those agents. Large bees tend to forage on two or more plants to complete their pollen loads – thus promoting cross-pollination. Small bees and other functional pollinator groups, however, carry lesser volumes of resources and their foraging is usually limited to a single plant.
4. Tibia (Table 3, Online Resource – Fig. 5). This type of pollination has exclusively been developed by bees of the Apidae family that deposit the collected pollen in corbicula (scopa). There is therefore no requirement for the anthers and/or stigmas to be located near each other, because large quantities of pollen are naturally deposited on the stigma as those bees scramble over the reproductive organs while transporting their pollen loads.
5. Dorsal region of the abdomen (Table 3, Online Resource – Fig. 6). Bees with the ability to buzz the floral anthers have their entire bodies covered with pollen grains – a method of pollen extraction that was only observed here with Senna obtusifolia (Fabaceae). The upper region of the abdomen of those bees then generally come into contact with the stigma of the flower. As those plants produce few flowers, and in light of the elevated demand for that resource (as those bees are quite large – especially the genera Bombus and Xylocopa, with total lengths > 14 mm and thorax diameters of 6 mm [Roubik 1989]), cross-pollination probably occurs much more frequently in S. obtusifolia than in any other plant species examined in the present study
Table 3
Floral adaptations of the plant species selected here in terms of their levels of pollinator specialization. Pollen transfer behavior to the stigmas of the flowers described in the first column refer exclusively to the functional insect group most adapted to pollinating each type of flower (description in italics in the final column).
Reward | Regions of pollen transfer to the stigma
|
Floral shape
|
(Number: degrees of floral specialization) Plant species studied
|
Functional groups (N. of species) | Behavior | N. of foraging
|
Nectar and pollen | Frontal region of the head (Casual pollinator)
|
Dish
|
I. Matayba guianensis
|
Long-tongued bees (7) | Casual pollinator | 1210
Short-tongued bees (5) | Casual pollinator | 20
Other Hymenoptera (25) | Casual pollinator | 123
Diptera (10) | Casual pollinator | 294
Lepidoptera (8) | Casual pollinator | 317
Coleoptera (7) | Casual pollinator | 15
|
Nectar | Frontal region of the head (Casual pollinator)
|
Tubular
|
I. Aegiphilla sellowiana
|
Nectar and pollen | Frontal region of the head (Casual pollinator)
Nectar and pollen | Ventral region of the thorax and abdomen (Casual pollinator)
Pollen | Tibia (Casual pollinator)
|
Brush
|
I. Senegalia polyphylla, Senegalia sp.
|
Nectar and pollen | Frontal region of the head (Non-efficient pollinator)
|
Capitula
|
II. Eupatorium maximalianii, Eupatorium cf. dimorpholepis
|
Long-tongued bees (3) | Non-efficient pollinator | 404
Short-tongued bees (2) | Non-efficient pollinator | 4
Diptera (8) | Nectar thief | 103
Lepidoptera (18) | Nectar thief | 159
|
Nectar (Aves) | Frontal region of the head (Efficient pollinator)
Pollen | Tibia (long-tongued bees) (Efficient pollinator)
|
Tubular
|
III. Pyrostegia venusta
|
Long-tongued bees (3) | Efficient pollinator | 169
Short-tongued bees (3) | Nectar and pollen thief | 10
Aves (2) | Efficient pollinator | Not quantified
|
Nectar and pollen | Frontal region of the head (Efficient pollinator)
|
Dish
|
III. Gouania cf. latifolia
|
Long-tongued bees (3) | Efficient pollinator | 1105
Short-tongued bees (3) | Efficient pollinator | 28
Other Hymenoptera (7) | Efficient pollinator | 21
Diptera (9) | Nectar thief | 317
Lepidoptera (1) | Nectar thief | 9
|
Dish-polypetalous
|
III. Serjania caracasana
|
Nectar | Frontal region of the head (Efficient pollinator)
Pollen | Tibia (Efficient pollinator)
|
Capitula
|
III. Trixis antimenorrhoea
|
Long-tongued bees (1) | Efficient pollinator | 163
Lepidoptera (1) | Non-efficient pollinator | 3
|
Nectar | Nototribic (Efficient pollinator)
|
Gullet
|
IV. Arrabidaea chica, Adenocalymma bracteatum, Arrabidaea florida, Cuspidaria convoluta
|
Long-tongued bees (16) | Efficient pollinator | 309
Short-tongued bees (11) | Nectar thief or robber | 193
Other Hymenoptera (5) | Nectar thief | 29
Diptera (3) | Nectar thief | 120
Lepidoptera (8) | Nectar thief | 43
|
Pollen | Dorsal region of the abdomen (Efficient pollinator)
|
Dish-Poricidal anthers
|
V. Senna obtusifolia
|
Long-tongued bees (3) | Efficient pollinator | 22
Short-tongued bees (1) | Pollen robber | 14
|
Oil | Ventral region of the thorax and abdomen (Casual pollinator) (Efficient pollinator)
|
Dish
|
V. Diplopterys pubipetala, Banisteriopsis cf. campestris, Banisteriopsis laevifolia, Byrsonima intermedia
|
Long-tongued bees (14) | Efficient pollinator | 511
Short-tongued bees (1) | Casual pollinator | 4
|
I. Highly generalist; II. Generalists; III. Intermediate; IV. Specialized; V. Highly specialized |
Among the various insect functional groups examined here, the largest number of foraging events were carried out by long-tongued bees. That same functional group was also responsible for the greatest success of pollen transport to the floral stigmas. The other foraging behaviors were characterized as robbing or theft of floral resources (nectar, pollen, or oils) (Table 3, Online Resource – Figs. 7 and 8).
Four floral specialization levels were observed among most of the plant species classified as evidencing highly specialized pollination; generalist systems were observed in only two plant species (Table 3). The flowers of generalist and highly generalist plants were pollinated by 45 functional insect groups considered casual or inefficient pollinators (Table 3). The plant species Matayba guianensis (Sapindaceae), for example, was foraged on by 28 different visitor species, with all of them being classified as casual pollen vectors. Only small quantities of pollen remained adhered to their bodies, and when they perchance contacted the floral stigmas, only very reduced volumes of pollen grains became deposited. Although generalist flowers evidence low levels of pollination efficiency, it was generally possible to characterize an insect as a pollinating species, even though they did not demonstrate any notable efficiency at transferring pollen to the stigmas (Table 3).
Specialist and highly specialized plant species, on the other hand, evidenced visitation by only a few pollinators that shared very similar body shapes. The efficient deposition of pollen grains on the stigmas of Malpighiaceae species and on S. obtusifolia flowers evolved with only one type of pollinator, and has established as a single unique behavioral technique (Table 3). The flowers of B. laevifolia, for example, were only pollinated by four bee species of the Centridini and Tapinotaspidini tribes. Visitors to those flowers were bees belonging to the Centridini, Tapinotaspidini, and Tetrapediini tribes.