Functional significance of flower orientation and green marks on tepals in the snowdrop Galanthus nivalis (Linnaeus, 1753)

ABSTRACT

Flower shapes, colors, sizes and fragrances are shaped mostly for pollinator attraction. Flower phenotypes are, however, subjected to conflicting selection directed by both pollinators and non-pollinating agents. We investigated flower attractiveness to a model pollinator in the snowdrop (Galanthus nivalis L.) under laboratory conditions. Naïve bumblebees (Bombus terrestris L.) showed strong, innate preferences for experimentally altered upward positioned flowers, suggesting that the natural, downward orientation did not evolve to attract pollinators. Experimentally treated green marks on inner tepals decreased pollinator attraction compared with flowers expressing intact marks, suggesting that green marks serve to guide/attract pollinators. Attractiveness of green marks was significantly compromised by flower orientation; green marks were attractive only for untreated downward-oriented flowers, but they did not improve the attractiveness of upward-oriented flowers. Our results suggest that downward flowers in snowdrop evolved under conflicting selection directed by biotic and abiotic factors, and that green marks on inner tepals could evolve later to enhance flower attractiveness.

KEYWORDS:

• Bumblebee
• evolutionary trade-off
• flowering
• pollination

Introduction

Entomophilous plants increase their reproductive success by means of pollinator attraction.¹ Pollinator attraction resulted in the evolution of the enormous diversity of flower colors, shapes, sizes, fragrances and rewards.^2,3 Flower phenotypes that are attractive to pollinators are simultaneously subjected to conflicting selection directed by herbivores.^4–7 and abiotic factors such as temperature, heat/water stress, wind action, photoperiod or altitude.^8–10 This suggests that floral morphology represents a trade-off to selection caused by both pollinators and non-pollinating agents.⁸

Floral orientation is an abiotic factor associated with pollinator behavior^11-14 and plant’s reproductive success.^12,15,16 For example, artificially erected flowers experienced lower seed production^15,17 and higher visitation rates.^11,13 Preferences for upward floral orientation are innate in some pollinators.¹⁸ This suggests that downward flowers could have a reduced attractiveness for pollinators.¹² In contrast, downward floral orientation can prevent pollen damage/number/germinability caused by certain abiotic factors such as rainfall or solar radiation,^19–22 and can maintain heat loss.^23,24 The latter factor can be important, particularly in early spring flowering plants when the temperature is low and the number of available pollinators is limited.^16,25

A phylogenetically broad range of angiosperm families possesses conspicuous petal spots on their flowers. Over two centuries ago, Sprengel²⁶ was the first to propose that these prominent features on flowers function to guide pollinators toward the flower’s nectary. Later research largely confirmed his idea; experimental evidence shows that petal spots on flower corollas attract various pollinators.^27–33 As far as we are aware, the relative simultaneous contribution of petal spots and flower orientation to flower attractiveness for pollinators has been rarely investigated. Moreover, the role of petal spots in pollinator attraction remains unstudied for the majority of plant species and many reports are based on anecdotal observations. However, it has been proposed that there are cases that spots on flowers mimic ants as defense from egg laying by butterflies.³⁴

Snowdrop (Galanthus nivalis L., Amaryllidaceae) is a perennial, early flowering plant (February–March) occurring in central Europe, Asia Minor and in the Caucasus.³⁵ It reproduces both vegetatively and sexually. Sexual reproduction is mainly accomplished by xenogamy and less by autogamy, which is costly in terms of reduced reproductive success.³⁶ Snowdrops are pollinated mainly by honey bees and bumblebees.³⁵ Several reports suggested that snowdrops have green patterned inner tepals attracting and guiding pollinators with visual and/or olfactory displays.^36–39 To our best knowledge, however, no study investigated the role of green marks of snowdrop flowers in pollinator attraction experimentally. Moreover, snowdrop has a consistent downward floral orientation, but its significance in pollinator attraction remains unclear. In this study, we experimentally investigated the relative contribution of floral orientation and green-patterned inner tepals to pollinator attraction in the snowdrop under laboratory conditions.

Methods

Experimental conditions

A single captive colony of naïve bumblebees (Bombus terrestris, L.) was procured from Koppert© (Nové Zámky, Slovakia), and was kept at 22–24°C in a room lit by natural light and neon light (370 lx). The bees were connected to a 90 × 50 × 40 cm insectarium by a plastic mesh tube and daily fed with pollen (1/4 honey solution and crushed pollen) and honey solution (water 60%, and honey 40%). Food was removed one-hour prior to testing in order to increase foraging motivation of tested individuals.

The bees could not be tested individually in the insectarium, because a significant proportion of workers (about 30%) resided permanently in the insectarium outside the hive. We, therefore, tested the bees individually by gently capturing individual foragers on the feeding array by using an entomological net and quickly transporting them in a separate terrarium 30 × 20 × 20 cm. All sides of the terrarium were covered with white paper to minimize potential disturbance except for a small 20 × 4 cm opening through which visual observations were carried out. Trials started by insertion of a single bee through a hole (2.5 cm diameter) in the semi-transparent plastic cover positioned on the terrarium. Two freshly collected snowdrop flowers of similar size and tepal opening were placed in glass test-tubes 10 cm apart from each other on the rear part of the terrarium, 6 cm apart from the back wall. For a given day, the placement of the flowers was randomly determined (i.e., left or right). We recorded the time till the bee started to feed on one of the two simultaneously presented flowers to the nearest 1 second following Patiño et al.⁴⁰ Feeding was defined as the presence of the bee inside the flower for at least 3 seconds. If the bee did not choose any of the flowers after 5 min, the trial was terminated and the bee was tested again on one of the following days. Each bee was used only once.

The trials took place 1 week after the colony arrived, between 26 February and 6 March 2019, during 09:00–14:30, indoors at the same conditions as described for keeping the colony. After each trial, the terrarium was washed with water and the paper from the bottom was replaced with a new one. New flowers were used for each trial.

Treatments

To test the influence of flower position on bee preference, one of the two simultaneously presented flower combinations had unmanipulated flower (downward) and one was treated (upward) (N = 32) (Figure 1a,b). Conspicuousness of the green marks on the abaxial surface of the inner tepals was manipulated by painting these marks on one of the two flowers, i.e., the one with downward flowers with white water-color (N = 28) (Figure 1c). The same approach was applied for another treatment, where both flowers were oriented upward, but in one flower we painted the green marks on its abaxial surface of the inner tepals with white color (N = 35). Finally, we examined whether the green color of the adaxial part of the inner tepals could attract bees by painting one of the two upward flowers, while green marks on the abaxial surface of inner tepals were painted in both flowers (N = 28). To control for possible confounding effect of the water-color, an additional treatment in which the abaxial surface of the outer tepals of downward flowers was painted with white colors or left unpainted was performed (N = 13).

Figure 1. Examples of a) intact, downward oriented flower, b) intact, upward oriented flower and c) treated downward oriented flower with abaxial tepals painted with white water-color

Statistical analyses

Bee’s choices between the two flowers were analyzed with a binomial test. Latency time between the beginning of the experiment and the bee’s choice was Box-Cox transformed to achieve normality and comparisons between treatments were then performed with one-way ANOVA and Tukey post-hoc test. Sample sizes of the statistical analyses presented in the Results section refer to successful trials. All statistical tests were performed in SPSS ver. 23.

Results

Bees showed significant preference for upward flowers (20/24, 83%) over downward flowers (4/24) (binomial test, P = .002). The preference for upward flowers was not influenced by green marks on the abaxial part of the inner tepals, because both white painted (12/25, 48%) and intact flowers (13/25, 52%) were visited by bees almost identically (binomial test, P = 1.0). Similarly, green color on the adaxial part of the inside of the inner tepals was not exceptionally attractive for bees, because both painted (9/20, 45%) and unpainted upward flowers (11/20, 55%) were visited at a similar ratio (binomial test, P = .824). Green marks on the abaxial part of the inner tepals of downward flowers were strongly attractive for bees; almost all bees (19/21, 90%) visited unpainted downward flowers, while only two bees visited downward flowers with their green marks painted white (binomial test, P < .0001). The possible influence of the paint’s odors on bee visitation was examined in additional treatments. Half of bees (5/10) visited a flower with its outer tepals painted and the remaining five bees visited intact flowers. This suggests that the visitation of flowers was not influenced by the presence of odor cues produced by the paint (binomial test, P = 1.0).

The mean latency time between the beginning of the experiment on the bee’s choice was 138 sec (SE = 9.76, N = 100). Latency time was significantly different among treatments (ANOVA, F(4, 95) = 3.12, p =.019). A subsequent post-hoc test showed that the bees made their choices significantly sooner in treatment where downward and upward flowers were presented simultaneously, than in a treatment with downward flowers with their abaxial inner tepals of one of two flowers painted (Tukey’s P = .011). Other differences were not significant (all P > .17). Among all five treatments, about 23–28% of the bees did not perform any choice. There were no significant differences in the rate of non-choices with respect to the treatments (Pearson χ2 = 0.29, df = 4, P = .99).

Discussion

Our study demonstrates that green marks on inner tepals of snowdrop attract naïve bumblebees as model pollinators, but floral orientation had a greater importance in pollinator attraction than the green marks on the inner tepals.

Certain studies showed that bumblebees specialize on downward floral orientation.^41,42 Upward floral orientation of herb flowers is more visible for the majority of pollinators flying from above.⁴³ The surface of the inflorescence of upward-oriented flowers is more visible and therefore strongly preferred by a range of competing pollinators such as solitary bees, muscoid flies, hoverflies, and beetles.¹³ Perhaps, the bumblebee’s preference for downward floral orientation in nature is driven by reducing interspecific competition. Our results clearly suggest that naïve bumblebees show innate preferences for upward floral orientation, which supports previous studies on Bombus impatiens Cresson, 1863.¹⁸ A downward flower might help to filter pollinators in the field to reduce the potential waste of flower reward because some pollinators cannot handle a downward-oriented flower.

Evolution of the downward floral orientation of snowdrop could be traded-off between pollination attraction and abiotic factors such as rainfall/snow damage avoidance,^19,21 or perhaps downward-facing flowers can capture convection heat from the ground, because even a small increase in flower temperature of early spring ephemerals may enhance reproductive success.⁴⁴ However, flower temperatures of snowdrops were about 2.7°C cooler than the ambient air and about 1.7°C cooler than the ground which does not support the thermoregulatory hypothesis.⁴⁵ A better understanding of snowdrop ancestors would shed more light into the question of the evolution of downward flower orientation, but the fossil records of Amaryllidaceae are very poor.⁴⁶ Pyramidal shape of snowdrop downward flowers can prevent damage by snow, which can be serious threat, particularly in early spring, when deciduous trees are still defoliated and the understory woody herbs are thus not protected against harsh weather conditions. Indeed, spruce trees with downward hanging branches and narrow crowns are less vulnerable to damage caused by snow than those with rigid horizontal branches and broad crowns.⁴⁷ Further research on the possible risk of nectar dilution by rain,⁴⁸ pollen damage/number²¹ and harm avoidance by snow in experimentally altered flowers is required.

Green marks on abaxial snowdrop tepals were hypothesized to guide/attract pollinators.^38,39 Laboratory tests showed that naïve bumblebees prefer model flowers with contrasting colors over monochromatic models.⁴⁹ This suggests that conspicuous tepal spots may be easier to detect, and thus be effective in attracting pollinators,²⁶ a character that is beneficial to the plant, because spots are placed close to the location of rewards.⁵⁰ It was found that bumblebees show an innate preference for nectar guides at close range, but not from a distance.^51,52 Thus, green marks are probably not attractive to bees from long distance, but are followed by bees only after a closer contact with the flower. If green marks serve as food signals, then the bees are probably able to associate them with a food reward. Conspicuous signals like these green marks may enhance insect learning processes,⁵⁰ and, consequently, increase snowdrop’s visitation rate.

Green marks on both sides of the inner tepals produce odorous substances that are predicted to attract pollinators.³⁹ We suggest that these odor scents help the flower to guide pollinators to their rewards after the pollinator approaches the flower. Odor scents appear to play no significant role in experimentally altered flowers with upward corollas.

Lev-Yadun³⁴ suggests that dark dots and short stripes on many Passiflora flowers visually mimic ants, which may protect flowers from herbivory. It is possible that green marks along with lectins which are involved in defense against pests and pathogens⁵³ protect snowdrop flowers from herbivory, but this idea requires further research.

With respect to other species with downward flowers, the Solomon’s seal (Polygonatum multiflorum, Asparagaceae) has also white flowers with similar green markings and it also occupies similar habitats as the snowdrop. We suggest that the evolution of downward flowers and their coloration evolved independently from the snowdrop, and that this process could also be driven by early flowering. Interestingly, however, late, downward flowering plants from the family Campanulaceae, having pink or violet flowers occupying in the same habitats lack any apparent flower marks. Further, comparative research should involve estimations of pollinator availability, environmental factors and flower visibility of early and late flowering plants in deciduous woods are required.

To conclude, green tepal marks increase flower attraction for pollinators and may help to guide pollinators to their rewards. Downward floral orientation of snowdrop flowers appears not to have evolved only in order to attract pollinators, but it seems to be a result of selection directed by both biotic and abiotic factors. Green marks enhance learning processes of bumblebees and this increases snowdrop visitation rates. Future research should investigate the benefits of downward floral orientation for early spring blooming plants, when the risk of damage by rain or snow is high.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We are extremely grateful to Prof. Simcha Lev-Yadun and an anonymous referee for insightful comments on earlier drafts of this manuscript.

Additional information

Funding

This work was supported by the Scientific Grant Agency VEGA [1/0286/20].