Printer Friendly

Does flower color change in aster vimineus cue pollinators?


The flowers of numerous plant species change color throughout their lives. Floral color changes can be localized, influencing only a select part of the flower, or generalized, affecting the entire flower (Weiss, 1995). These specific color changes differ from the simple degeneration of the flower and have been explained as either a pollination-induced or an age-dependent pattern (Gori, 1983; Casper and La Pine, 1984; Eisikowitch and Lazar, 1987; Lamont and Collins, 1988; Delph and Lively, 1989). In either model, the plant retains the postchange flower until natural senescence occurs. The physiological mechanisms responsible for the color change of the flower may include the gain or loss of pigments such as anthocyanin, carotenoid and flavonol, the appearance of betalain or a change in pH (Faegri and van der Pijl, 1979; Scogin, 1983; Weiss, 1995).

Several workers believe that floral color change is an adaptive trait that benefits both the plant and its insect pollinators by cuing the insects to visit the flowers at the optimal reproductive stage and with the greatest reward (Gori, 1983; Kevan, 1983; Kevan and Baker, 1983; Casper and La Pine, 1984; Delph and Lively, 1985; Eisikowitch and Lazar, 1987; Eisikowitch and Rotem, 1987; Lamont and Collins, 1988; Delph and Lively, 1989; Weiss, 1991; Sutherland and Vickery, 1993; Ne'eman and Nesher, 1995). The prechange flower offers more viable pollen and nectar than the postchange flower. The plant essentially advertises its most rewarding flowers to the foraging pollinators through visible color cues. By learning to recognize and interpret these color cues, the pollinator can more efficiently focus on visiting the more reproductively viable flower.

The purpose for retaining the postchange, sexually inviable flower has been questioned in several studies. Pollinators most likely cannot recognize color cues from greater distances, and are initially attracted to the amount of potential reward that they may receive at an individual plant based on the total floral display. After arriving at a plant, pollinators should be able to identify and follow the guidance of the color cues. Therefore, many argue that a plant maintains its postchange flowers to benefit from the advantage of an increased floral display, and the color change serves to guide pollinators to appropriate flowers once they arrive (Gori, 1989; Weiss, 1995; Ne'eman and Nesher, 1995). Others believe that for certain species a larger floral display does not influence the pollinators' foraging decisions (Casper and LaPine, 1984; Delph and Lively, 1989).

Despite the wide prevalence of flower color change and the well-developed hypotheses offered to explain the adaptive nature of this trait (Weiss, 1995), this phenomenon has been experimentally examined in only a few species. We examined variation in flower color in Aster vimineus (Asteraceae) and addressed the following questions: (1) Is variation in flower color due to a systematic change in flower color? (2) Do pollen availability and viability vary with flower color? (3) Do color and/or floral nectar volumes influence visitation rates by pollinators? and (4) Does an increased floral display size attract more pollinators? Our goal was to test the hypothesis that flower color change permits the plant to retain postreproductive flowers in order to increase pollinator visitation from a distance, and then guides pollinators to the most viable flowers once they arrive at a plant.


Species and study site. - Aster vimineus Lam. (Asteraceae) is a small aster, with a head under 1cm wide, narrow, lance-shaped leaves and numerous flower heads with white ray florets. Within a plant there are observable differences in the color of the rioters that comprise the central disk which are either all yellow or all red. Blooming in late summer and autumn, this wildflower is most commonly found along the E coast of North America extending to southern Michigan, and it prospers in a wide variety of soil moisture and sunlight conditions. This study was conducted in two 30-[m.sup.2] patches of A.vimineus in a plot of deciduous forest located at the Graver Arboretum of Muhlenberg College in Bushkill Township, Northampton County, in eastern Pennsylvania. The forest canopy is dominated by Fagus grandifolia, Acer rubrum, Carya tomentosa and Quercas rubra. All experiments were conducted in September and October of 1996.

Flower color change. - To determine whether existing variation in disk color was due to systematized color change, individual flowers were marked and monitored over time. We marked 40 yellow disk asters and 40 red disk asters. After 7 days, the labeled asters were revisited and any color change was recorded.

Pollen availability and viability. - The amount of pollen available to pollinators was quantified in both yellow and red disk asters. For 10 flowers of each color, 10 anthers were removed from disk florets and placed in microcentrifuge tubes with 10 ml basic fuchsin in solution to stain the pollen for viewing with a light microscope (Kearns and Inouye, 1993). The tube was vibrated using a vortex for 60 sec which we had determined to effectively free pollen from the anthers, and the stain-pollen suspension mixture was transferred by pipette into the chambers of a hemocytometer which each hold exactly 1 ml. Pollen grains were counted in each chamber and the number of pollen grains per anther for a given sample was calculated using the average number per chamber, the original volume of the suspension, and dividing by 10 anthers. Means were compared using a nonparametric Wilcoxon signed-rank test and reported as [Mathematical Expression Omitted] SD.

We used the Fluorochromatic (FCR) Procedure (Heslop-Harrison and Heslop-Harrison, 1970; Kearns and Inouye, 1993) to assess pollen viability among the floral color types. Viability was assessed by enzymatic induction of fluorescence using fluorescein diacetate where viable pollen grains fluoresce brightly and visibly when viewed by fluorescence microscopy, while the inviable pollen grains appear faint and indistinct (Shivana and Heslop Harrison, 1981; Kearns and Inouye, 1993). The mean percentage of pollen grains that were viable were compared between flower color types using a non-parametric Wilcoxon signed-rank test and reported as [Mathematical Expression Omitted] standard deviation.

Tests of pollinator preference. - To first determine if pollinators preferentially visited flowers with yellow or red disks, observations were made of unmanipulated flowers. Two patches of the aster with equal numbers of yellow and red disk flowers were simultaneously observed for 25 min, and the total number of visits to each color morph was recorded.

To further explore pollinator preference, pollinators were offered a choice between color morphs using pollinator interview sticks (Thomson, 1981) which were constructed by attaching two pieces of plastic tubing (each approximately 5 cm long) to the end of a stick 1 m long. The peduncles of flowers of comparable shapes and sizes were inserted into each of the two tubes so that flower heads were approximately 2-cm apart (Thomson, 1981). We then offered a variety of potential pollinators occurring on the flowers Aster vimineus an opportunity to select one of the two flowers on our pollinator interview stick. For 63 pollinators selected at random we recorded whether they visited the yellow or red disk aster, and the pollinator type and number of probes the pollinator initiated into the flower. One complete probe was defined as the pollinator landing on and inserting its feeding apparatus into a floret (Sutherland and Vickery, 1993). Avoidances of both offered flowers were not included.

We then tested the influence of nectar in attracting pollinators to the asters. The nectar content of the aster disk was artificially manipulated throughout three choice experiments. Using a micropipette, 0.25 [[micro]liter] of a 30% sucrose solution was injected directly into the disk of the experimental aster. In one nectar experiment, pollinators were randomly selected and offered a choice of either a yellow disk aster or a red disk aster supplemented with the sucrose solution. A second study offered the pollinator a choice of a yellow disk aster or a yellow disk aster enhanced with the sucrose solution. Finally, in a third experiment we presented the pollinators with a choice between a red disk aster and a red disk aster supplemented with the sucrose solution. Again, flower choice, pollinator type and number of probes were recorded. The frequency of visits and the average number of probes per visit at each choice were compared to an expected 50% likelihood of visitation using a chi-square test (Thompson, 1981). Means are reported as [Mathematical Expression Omitted] SD.

Floral display. - The size of floral display anti its effect on pollinator visitation were investigated by measuring the rate of visitation from a distance to natural flower patches of different sizes and to manipulated floral displays. In the first comparison, one large aster patch (3.0 [m.sup.2]) and one small aster patch (1.2 [m.sup.2]) were isolated. Carefully selected plots with equal proportions of red disk and yellow disk flowers were used. The arrival of pollinators at each patch was observed simultaneously for 1 min, and these intervals were repeated seven times in 30 min. The total number of pollinators that arrived at each patch per 1-min interval was recorded. A nonparametric Wilcoxon signed-rank test was used to compare the numbers of arrivals at the large and small patches. In the second experiment, two patches of aster of comparable sizes (approximately 0.45 [m.sup.2]) were isolated and measured. All of the red disk asters were entirely thinned from one of the patches. The second patch was left intact. Pollinator arrival was observed simultaneously again for 1-min intervals and the total number of pollinators arriving at each patch was recorded as described above. A nonparametric Wilcoxon signed-rank test was used to compare at the thinned and unthinned patches. All data are reported as [Mathematical Expression Omitted] SD.


Flower color change. - Of the 40 yellow asters, 33 had central disks that turned red in 1 wk, and seven had disks that remained yellow. None of the 40 labeled red disk asters changed color.
TABLE 1. - Pollinator choice experiments - yellow disk vs. red disk
flowers. Comparisons with other types of pollinators that are
included with all pollinators are not shown because of small sample
size. Chi-square tests differences between observed numbers and an
expectation based on equal numbers of visits

                      Number of visits

Pollinator type       Yellow       Red          Significance

All pollinators         63          0        P [less than] 0.001
Honey bees              32          0        P [less than] 0.001
Bumble bees             26          0        P [less than] 0.001
Flies                    5          0        P [less than] 0.001

Pollen availability and viability. - Yellow disk asters contained significantly more pollen than the red disk asters (signed-rank = 27.50, P [less than] 0.002). We found an average of 1121 [+ or -] 163 (n = 10) pollen grains per anther in the florets of yellow disks. In comparison, we found only an average of 340 [+ or -] 164 (n = 10) pollen grains per anther in the florets of red disks. The pollen viability tests suggest that the yellow disk asters have a larger percentage of viable pollen grains than the red disk asters. The FCR procedure resulted in a significantly larger proportion of brightly fluorescing pollen grains from yellow disk florets (57.5 [+ or -] 9.7 %, N = 10) than those from red disk asters (25.1 [+ or -] 5.8 %, n = 10; Signed-rank = 27.5, P [less than] 0.001).

Tests of pollinator preference. - Insect pollinators included honey bees (Apis mellifera), bumble bees (Bombus sp.), vespid wasps, moths and butterflies (any Lepidoptera) and flies (any Diptera). During the 25 min observation period, all pollinator types visited only the yellow flower morph, and no insects visited the flowers with red central disks.

For the choice experiment, data were analyzed for all pollinators. Honey bees, bumble bees and vespid wasps represented a significant proportion of the visits and were analyzed as subsets of the total assemblage of pollinators in order to examine if different types of pollinators respond differently to the choices offered.

When offered a choice between yellow and red flowers, all 63 pollinators tested chose the yellow disk aster over the red disk aster (Table 1). This was true for all types of pollinators and chi-square tests were all significant.

Addition of sucrose to yellow disk asters did not affect visitation rate or choice over unmanipulated yellow disk asters (Table 2). Overall, pollinators did not spend much time [TABULAR DATA FOR TABLE 2 OMITTED] [TABULAR DATA FOR TABLE 3 OMITTED] probing individual florets after landing on a flower, and the number of probes per visit was the same for manipulated and unmanipulated flowers. Honey bees tended to have higher rates of probing, but the rates were not significantly different (Table 2).

When offered a choice between an unmanipulated red disk aster and a red disk aster with nectar added, significantly more pollinators chose the sucrose enriched red disk aster, but there were no significant differences within pollinator types (Table 3). Although the number of probes per visit was statistically the same for manipulated and unmanipulated flowers, insect visitors tended to probe the flowers without the additional nectar more often (Table 3).

Finally, when offered a choice between an unmodified yellow disk aster and a sucrose enriched red disk aster, pollinators still preferred yellow disk flowers (Table 4). Visitation rates were significantly different for all pollinators, and for bees and wasps separately. Pollinators regularly probed florets of the yellow disk asters, but did not probe the red disk aster (Table 4).

Floral display. - For all intervals of observation, significantly more pollinators arrived at the large patch of asters than at the small patch of asters in the same time period [ILLUSTRATION FOR FIGURE 1 OMITTED]. During the 30-min observation period, the mean rate of arrival at the large patch was 9.85 [+ or -] 1.77 per minute (n = 7), and 4.42 [+ or -] 1.32 per min (n = 7) at the smaller patch. The nonparametric comparison of means indicated that these rates of arrival were significantly [TABULAR DATA FOR TABLE 4 OMITTED] different (Signed-rank = 14, P = 0.008). Similarly, more pollinators arrived at the floral display composed of a mix of yellow and red disk asters than the display where all red disk asters were removed leaving behind a smaller total number of flowers [ILLUSTRATION FOR FIGURE 2 OMITTED]. The mean rate of arrival at the unthinned, multicolored display was 4.57 [+ or -] 1.27 per min (n = 7), and was significantly greater than that at the thinned display where red asters were removed (1.01 [+ or -] 0.82, n = 7, Signed-rank = 14.00, P = 0.008).


The central disk of Aster vimineus changes color from yellow to red, and flowers with yellow disks are more reproductively viable than those with red disks. Yellow disks have florets with more pollen per anther than the red disk asters, and they have a higher proportion of viable pollen as suggested by the FCR test. From this it seems that the color change in Aster vimineus acts as a signal to potential pollinators and serves to cue the pollinators to visit the flowers which offer the most viable pollen (yellow disk asters) and potentially the greatest reward, thus increasing pollination efficiency.

Pollinators clearly respond to this color cue by almost exclusively visiting the yellow disk aster, potentially resulting in increased collection and distribution of viable pollen. These results are similar to those found by Gori (1983), Kevan (1983), Kevan and Baker (1983), Casper and La Pine (1984), Delph and Lively (1985), Eisikowitch and Lazar (1987), Eisikowitch and Rotem (1987), Lamont and Collins (1988), Delph and Lively (1989), Weiss (1991), Sutherland and Vickery (1993), and Ne'eman and Nesher (1995). That color change is a signal to pollinators is further supported by our choice experiment in which we offered pollinators a choice of a yellow disk aster and a red disk aster, all of the pollinators preferred the yellow disk asters to the red disk asters. Apparently, the yellow color is effective in drawing the pollinators to the flowers offering the greatest amount of viable pollen. This action could result in an increase in reproductive efficiency as larger amounts of viable pollen are transferred to neighboring asters (Gori, 1989; Kevan, 1983; Casper and La Pine, 1984; Delph and Lively, 1989).

The fact that the yellow disk asters offer significantly more viable pollen and are visited more often, if not exclusively, by insect pollinators than red disk asters begs the question of why old, inviable, red disk flowers are retained. We hypothesized that the red disk asters are attracting pollinators to a particular plant from a distance by adding to the size of the total floral display, as proposed by Gori (1989), Weiss (1995) and Ne'eman and Nesher (1995). We found that the larger floral display had a significantly greater rate of arrival in comparison to the smaller display [ILLUSTRATION FOR FIGURE 1 OMITTED]. By removing red disk flowers from an equal sized floral display we showed that the retention of red disk asters is crucial in attracting pollinators [ILLUSTRATION FOR FIGURE 2 OMITTED]. For example, if old and infertile flowers were simply to wither and die, the resulting reduction in display size would decrease pollinator visitation to that plant. Others have found that retention of older flowers increases attractiveness to pollinators (Wainwright, 1978; Schaal and Leverich, 1980) or that thinning of flowers reduced visitation (Gori, 1983); but Delph and Lively (1989) argued that retention of old flowers may also be the result of a physiological constraint when senescence or abscission of attractive structures cannot happen quickly. This might be the case if flower abscission results in the loss of the style so that flowers must be retained at least until pollen tube growth past the point of abscission has occurred which in Fuchsia exocorticata took at least 3 days (Delph and Lively, 1989).

Whether older flowers are retained to attract pollinators from a distance or because of a physiological constraint as described above, once a pollinator arrives at a plant, color could serve as a cue to guide pollinators to the most reproductively viable flowers offering the greatest rewards. The color change from yellow to red in older flowers and the preference of pollinators for yellow flowers in Aster vimineus is consistent with patterns of color change and pollination observed in other species (Weiss, 1995) and with the notion that insect pollinators prefer yellow and avoid or are insensitive to red (Kevan and Baker, 1983; Eisikowitch and Lazar, 1987); however, Chittka and Waser (1997) argue that red flowers are not invisible to bee pollinators.

It is clear that yellow flowers offer more viable pollen and thus increase pollinator efficiency by attracting pollinators to these flowers. It is less clear, however, that pollinators are responding to greater levels of reward in yellow flowers or simply to the yellow color. Pollinator preferences for specific character sets should depend largely on the quantity and quality of food that rewards a particular choice (Pleasants and Waser, 1985). In the case of pollen foragers, like bumble bees, the preference for yellow flowers makes sense. There may also be more nectar in yellow flowers but this was difficult to measure in this case; however, we observed that yellow flowers were more aromatic. The consistent choice of yellow over red by all pollinator types suggests that there are greater rewards other than pollen, like nectar, in yellow flowers. However, rates of probing indicated that there were no major differences in nectar rewards and nectar manipulations did not alter attractiveness or the rate of probing.

The consistent movement of pollinators to yellow disk flowers in our choice experiments may also reflect constancy in pollinator behavior. The restriction of pollinator visits to flowers of a single species or morph has been observed in numerous species (Thomson 1981; Waser, 1986, Dukas and Real, 1993); although some have argued that more careful study might reveal that pollinators are more likely to be generalists (Waser and Chittka, 1996). Although pollinators in the studied community do not exclusively visit Aster vimneus, when they are at this species preferences for color morphs appear fixed (sensu Waser, 1986) rather than labile for all pollinator types.

Further investigations into the color change patterns of Aster vimineus and the behaviors of pollinators could augment our understanding of the adaptive nature of this phenominon from the plant and pollinator perspective. As stated above, the color change of A. vimineus could be either pollination induced or age dependent. Additional studies are needed to identify the exact mechanism of color change. Tests conducted by Gori (1983), Casper and La Pine (1984), Eisikowitch and Lazar (1987), Lamont and Collins (1988), and Delph and Lively (1989) on other species of flowers provide support for both theories. Further research could also provide a more complete understanding of how pollinators learn that the yellow disk asters contain more pollen than red disk asters. Perhaps the pollinators simply learn on a trial-and-error basis that they receive the greatest pollen or nectar reward from the yellow disk asters. Or insect visual systems may be more sensitive to yellow. Studies by Weiss (1991) support the associative learning theory of some pollinators.

We found that the retention of flowers with less abundant and less viable pollen is beneficial to individual plants because the resulting increase in floral display size effectively increases the attraction of pollinators from a distance. This trait may be particularly adaptive in the asters where retaining ray petals on otherwise reproductively inviable flowers may significantly contribute to the total floral display and the attraction of pollinators. It is also clear from this study that the color change of the central disk serves to guide pollinators to more reproductively viable flowers once they have arrived at a particular plant increasing the efficiency of pollination. The insect's response to this color cue is likely to be adaptive in that it guides them to flowers that offer more pollen and potentially more nectar increasing the efficiency of their foraging.

Acknowledgments. - Many of the ideas and techniques employed in this work came from discussions during an NSF-sponsored workshop on Plant-Animal interactions held at the Rocky Mountain Biological Laboratory in August 1996. Discussions with Martha Weiss, James Thompson, David Inouye, Carol Kearns, Nick Waser and Mike Messier were particularly helpful and relevant. We also thank Martha Weiss for her input on this work. Elizabeth Dale and Elizabeth McCain provided valuable comments on earlier versions of this paper. Muhlenberg College provided financial and logistical support during all phases of this research. This work could not have taken place without the gracious generosity of Dr. Lee and Virginia Graver.


CASPER. B. B. AND T. R. LA PINE. 1984. Changes in corolla color and other floral characteristics in Cryptantha humilis (Boraginaceae): cues to discourage pollinators? Evolution, 38:128-141.

CHITTKA, L. AND N.M. WASER. 1997. Why red flowers are not invisible to bees. Isr. J. Plant Sci., 45:169-174.

DELPH, L. F. AND C. M. LIVELY. 1985. Pollinator visits to floral colour phases of Fuchsia excorticata. N. Z. J. Zool., 12:599-603.

----- AND -----. 1989. The evolution of floral color change: pollinator attraction versus physiological constraints in Fuchsia excorticata. Evolution, 43:1252-1262.

DUKAS, R. AND L. REAL. 1993. Learning constraints and floral choice behavior in bumble bees. Anita. Behav., 41:637-644.

EISIKOWITCH, D. AND Z. LAZAR. 1987. Flower change in Oenothera drummondii Hooker as a response to pollinators' visits. Bot. J. Linn. Soc., 95:101-111.

----- AND R. ROTEM. 1987. Flower orientation and color change in Quisqualis indica and their possible role in pollinator partitioning. Bot. Gaz., 148:175-179.

FAEGRI, K. AND L. VAN DER PIJL. 1979. The principles of pollination ecology. Pergamon Press, New York, N,Y. 304 p.

GORI, D. F. 1983. Post-pollination phenomena and adaptive floral changes, p. 31-49. In: C. E. Jones and R.J. Little (eds.). Handbook of experimental pollination biology. Scientific and Academic Editions, New York, N.Y.

-----. 1989. Floral color change in Lupinus argenteus (Fabaceae): why should plants advertise the location of unrewarding flowers to pollinators? Evolution, 43:870-881.

HESLOP-HARRISON, J. AND Y. HESLOP-HARRISON. 1970. Evaluation of pollen viability by enzymatically induced florescence: intra-cellular hydrolysis of florescein diacetate. Stain Technol, 45:115-122.

KEARNS, C. A. AND. W. INOUYE. 1993. Techniques for pollination biologists. University Press of Colorado, Niwot. 583 p.

KEVAN, P. G. 1983. Floral colors through the insect eye: what they are and what they mean, p. 3-30. In: C. E. Jones and R.J. Little (eds.). Handbook of experimental pollination biology. Scientific and Academic Editions, New York, N.Y.

-----, AND H. G. BAKER. 1983 Insects as flower visitors and pollinators. Annu. Rev. Entomol., 28:407-453.

LAMONT, B. B. AND B. G. COLLINS. 1988. Flower colour change in Banksia ilicifolia: a signal for pollinators. Aust. J. Ecol., 13:129-135

NE'EMAN G. AND R. NESHER. 1995. Pollination ecology and the significance of floral color change in Lupinus pilosus L. (Fabaceae). Isr. J. Plant Sci., 43:135-145.

PLEASANTS, J. M. AND N.M. WASER. 1985. Bumblebee foraging at a "hummingbird" flower: reward economics and floral choice. Am. Midl. Nat., 114:283-291.

SCHAAL, B. A. AND W. J. LEVERICH, 1980. Pollination and banner markings in Lupinus texensis (Leguminosae). Southwest Nat., 25:280-282.

SCOGIN, R. 1983. Visible floral pigments and pollinators, p. 160-172. In: C. E. Jones and R. J. Little (eds.). Handbook of experimental pollination biology. Scientific and Academic Editions, New York, N.Y.

SHIVANNA, K. R. AND J. HESLOP-HARRISON. 1981. Membrane state and pollen viability. Ann. Bot., 467:759-770.

SUTHERLAND, S. D. AND R. K. VICKERY. 1993. On the relative importance of floral color, shape, and nectar rewards in attracting pollinators to Mimulus. Great Basin Nat., 53:107-117.

THOMSON, J. D. 1981. Field measure of flower constancy in bumblebees. Am. Midi. Nat., 105:377-380.

WASER, N.M. 1986. Flower constancy: definition, cause and measurement. Am. Nat., 127:593-603.

----- AND L. CHITTKA. 1996. Generalization in pollination systems, and why it matters. Ecology, 77:1043-1060.

WAINWRIGHT, C. M. 1978. The floral biology' and pollination ecology of two desert lupines. Bull. Torrey. Bot. Club, 105:24-38.

WEISS, M. R. 1991. Floral colour changes as cues for pollinators. Nature, 354:227-229.

-----. 1995. Floral color change: a widespread functional convergence. Am. J. Bot., 82:167-185.
COPYRIGHT 1999 University of Notre Dame, Department of Biological Sciences
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Niesenbaum, Richard A.; Patselas, Maria G.; Weiner, Samuel D.
Publication:The American Midland Naturalist
Date:Jan 1, 1999
Previous Article:Effects of fire, browsers and gallers on New Jersey tea (Ceanothus herbaceous) growth and reproduction.
Next Article:Growth responses of Carya ovata (Juglandaceae) seedlings to experimental sun patches.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters