2016 Northeastern Naturalist Notes Vol. 23, No. 3 N12 B. Heinrich A Note on Iris Flower Anthesis: Mechanism and Meaning of Sudden Flower Opening Bernd Heinrich* Abstract - Experiments with Iris pseudacorus (Yellow Iris) reveal the mechanics of a sudden-release phenomenon in these flowers caused by slow bending of the outer tepals while their top-margins cling together, leading to stored mechanical energy that is released explosively. In this note, I discuss the possible adaptive significance of this flower behavior in terms o f pollinator learning-behavior. Flower opening and closing (Meeuse 1961) is affected by temperature and light conditions, circadian rhythms, and the mechanics of elongation as governed by differential growth processes (van Doorn and Meeteren 2003). Similar to leaf movements (Volkov et al. 2010), flower opening (Bielski et al. 2000) is dependent on physical processes involving the edges of the petals and changes in turgor pressure that lead to contraction and expansion of flower tissues (Liang and Mahadevan 2011). Iris flower opening and closure require chemical factors that affect growth (Celikel and van Doorn 2012), and elongation of the ovary and flower pedicel are thought to “allow tepals to move laterally” for the flower to open (van Doorn and Kamdee 2014, van Doorn et al. 2013). However, the nearly “instant” (within a second) flower-opening in Iris pseudacorus L. (Yellow Iris; Heinrich 2015) cannot be explained by these processes alone, and I report here that my observations combined with surgical manipulation experiments elucidate the explosive mechanis m. In a way analogous to an inflorescence consisting of multiple flowers, each with its own ovary, for this discussion, I consider an iris flower as an inflorescence of 3 identical florets (Fig. 1A, B), each of which consists of the same 4 parts (Fig. 1C, D) which share a 3-carpellate compound ovary. In the flower-bud stage, iris florets lack sepals that enclose a corolla comprised of petals. Instead, each opened floret has a large, showy, hanging outer tepal on which the pollinator lands, an inner erect one (the banner), and a stamen and an adherent, flattened, robust style and stigma hidden within the flower. The 3 florets’ reproductive parts face in 3 different directions, separated by 120°. In contrast to the morphology of most inflorescences, Yellow Iris florets in the bud stage are morphologically coordinated into a single unit; in the early flower-bud stage, the outer tepals wrap the 3 florets together in a counterclockwise-twisted cochleate pattern. In this paper, I refer to anthesis as when 1 or all 3 of the erect outer tepals extend downward. In this process, the tepals’ previously erect inside surface is exposed to serve as a pollinating bees’ landing platform and as the channel leading to the floret’s hidden nectar past the anther and style. As observed previously (Heinrich 2015) and as shown by high-speed photography (Hartmann 2015), the explosive flower opening that presents the bees’ landing platform and the route to hidden nectar in Yellow Iris occurs simultaneously for the 3 florets, or with first 1 and then the other 2 together. In a sample of 14 Yellow Iris inflorescences in Hartmann’s (2015) video, the opening time of all 42 florets was less than 1 sec. In this study, I observed Yellow Iris flowers to document their explosive opening, understand the mechanism by which it occurs, and gain insight on its adaptive significance. I conducted the study of Yellow Iris flowers from 15 June through 6 July 2015, near Weld in western Maine in a group of 9 flower stems. They held an average of 7.8 flowers per *Box 153, Weld, ME 04285; bheinrich153@gmail.com. Manuscript Editor: Jill Weber Notes of the Northeastern Naturalist, Issue 23/3, 2016 N13 2016 Northeastern Naturalist Notes Vol. 23, No. 3 B. Heinrich day (1.2 flowers per stalk) throughout that 21-day period. Sixteen of 18 Yellow Iris flowers were observed to open in the afternoon, and were available for pollination only through the following day before senescing. For some comparative observations, I also observed nearby Iris sanguinea Donn ex. Hornem. (Japanese Iris) and I. versicolor L. (Blueflag Iris) (Fig. 1), whose flowering overlapped with Yellow Iris and were observed blooming from dawn to early morning (14 of 20, and 15 of 20 flowers, respectively). Throughout the 2 days before opening, the flower parts remained erect in the 3 species, creating a tall (6–8 cm), pointed, conical bud. In the early-bud stage (Fig. 2), the bud cross-sections had 7–8 layers of tepal tissue wrapped around each other; i.e., each tepal was wrapped ~3 times all the way around. All 3 species appeared to have the same arrangement. In all flower buds in which I examined chirality (26 Yellow Iris, 24 Japanese Iris, and 15 Blueflag Iris), it was always counter-clockwise. Flower Behavior. On the exterior of the flowers, the sepals started to slowly unfurl and the buds began to swell several hours before flower-bud opening (Fig. 2). Simultaneously, inside the flowers, the thickly robust style (with its closely associated stamen) of all 3 florets began to extend laterally and downward, possibly applying pressure from within the bud to spread the tepals apart. To test for a possible role of the flower bud’s internal structure in the explosive floret-opening mechanism, I surgically removed the styles, stamens, and inner tepals from later-stage flower buds by inserting the point of a small scissor laterally between Figure. 1. Diagrams of Iris flowers. A to D = Iris pseudacorus (Yellow Iris). A, B = lateral and top views of triple-floret inflorescence, respectively. C, D = lateral and front view of single floret, respectively. E = I. sanguinea (Japanese Iris). F = I. versicolar (Blueflag Iris). ot = outer tepal, it = inner tepal (banner), a =anther, sy = style, and si = stigma. 2016 Northeastern Naturalist Notes Vol. 23, No. 3 N14 B. Heinrich 2 outer tepals and removing each structure. (Fig. 3A, B). The outer tepals retained their normal explosive flower-opening behavior after this operation, suggesting that the internal flower parts were not responsible for the explosive flower-opening. Therefore, I hypothesized that the outer tepals may provide the power for the explosive flower-opening. In a second treatment, to test if the tepals bend by themselves, I cut through 5 entire flower buds directly above the bud pedicel (Fig. 3C, D, and E). After this procedure, the hafts (blade) of the outer tepals in swollen buds immediately bent laterally, whereas in young flower buds the hafts bent only slightly or not at all at first, but then continued to bend slowly for several hours on approximately the same timeline as normal bud swelling. However, all of these flower buds failed to open. The tepals had merely partially uncurled at their bases while remaining lightly curled around each other at their tips. In a third experiment, I tested the possibility that the outer tepal tips had prevented anthesis. Swollen flower buds that I judged to be within minutes of opening could be induced to open suddenly by manually separating the outer tepals. I chose 5 younger flower buds that I judged to be within several hours of opening, and manually separated the 3 outer tepals from one-another. They remained upright, but each spontaneously re-curled around itself rather than reverting to the arrangement of wrapping around each other as observed in unaltered buds. The tepals of these 5 flowers eventually bent downward and uncurled, but only so slowly that the movement was not perceptible as it took place; the typical explosive floweropening did not occur. Apparently, the expansion of the flower buds is caused by the downward-bending of the outer tepals. The tepal tips remain curled together, thus holding each other in place and resisting the pressure of the outward bending until the force is great enough to trigger the release. As long as the 3 outer tepals remain held together at and by their tips, the increasing angle of the bend between the broad blade of the petal and its elongate base acts like a spring. When the force of the spring of the 3 folded hafts becomes strong enough, the outer Figure. 2. Diagrams of 1-, 2-, and 3-day-old Yellow Iris buds, and cross sections of flower buds on day of opening at 3 sections of the bud indicated by A, B, and C, where the outside lines show how the 3 tepals wound around each other. N15 2016 Northeastern Naturalist Notes Vol. 23, No. 3 B. Heinrich tepals slip apart entirely to then flip down into the stable position. As soon as the flowers are fully open, they can receive pollinators. For comparison, the adjacent Japanese and Blueflag Irises appeared similar to Yellow Iris except for their enlarged inner banner tepal, which in Yellow Iris is vestigial. My observations of the former species were casual. Anthesis in these species was less explosive, possibly because the outer tepal tips were sometimes tangled with the erect inner tepals, suggesting that part of the adaptation for explosive opening resides in inner-tepal minimization. Alternately, some of the flowers were found to contain fly larvae that may have destroyed parts of the flower structure and prevented normal mov ement. Adaptive significance. Mechanical catapulting of seeds from fruits (Fahn and Werker 1972) for their dispersal, and of pollen from anthers of Cornus canadensis L. (Bunchberry) (Whitaker et al. 2007) for promoting wind dispersal, rely on release of stored energy. The Yellow Iris employs the same principle. However, the adaptive significance of the explosive change from flower bud into functional flower is not known. In the following discussion, preliminary observations point to a possible function relating to signal reliability in reward advertising (Cohen and Shmida 1993), contingent on previous work on the foraging behavior of pollinators. Naïve Bombus spp. (bumble bees) starting their foraging recognize a wide variety of flower signals, such as color (Heinrich et al. 1977), shape, and location (Jin et al. 2014, Figure. 3. Diagrammatic representation of 2 surgical manipulations indicating the effect of the tepals’ role in the instant flower-opening of fully formed buds. A and B show excision of style–anther units and inner tepals. C, D, and E indicate ef fect of severing the tepals from the flower base. 2016 Northeastern Naturalist Notes Vol. 23, No. 3 N16 B. Heinrich Ney-Nifle et al. 2001), but they learn to focus on flowers from which they can reliably harvest the most food (Heinrich 1975a, 1975b; Heinrich 1976; Keasar et al. 2013). As expected, anecdotal observations during this study confirm that bumble bees visit iris, and Bombus vagans Smith (Half-black Bumble Bee) repeatedly visited the Yellow Iris plant throughout flowering. The (presumed first) visitors to the plant repeatedly landed on and probed both flower buds and senescent flowers. However, even on the same foraging trip, these initially naïve bees soon preferentially landed on the flowers’ unfurled outer tepals’ nectar guides, and crawled into the flower to take nectar. In several subsequent visits they disregarded both senescent flowers and flower buds. Nectar-rewards motivate bees to both search widely for other plants identified by the same flower signal(s) and also return to them (Muth et al. 2015). In irises, the flower bud is large and not encased (hidden) by green sepals, and thus for at least 2 days before flower opening, the plant displays conspicuous, tall, yellow spikes, which may act as a distracting attractant or false advertisement from afar, that the plant must counter-balance with a reliable food reward to keep attracted potential pollinators from straying to a variety of other flowers advertised by other signals. In the competition between plants for pollinators and cross-pollination, the timing of food availability of flowers becomes critical (Heinrich 1975); revisits to the same plant, and visits to other plants of the same kind for cross-pollination, depend on reliability of finding a food reward. When a potentially false (unrewarding) signal lasts longer than the flower itself, any pollinator that has been attracted to the location of the plant has inducement to search elsewhere and sample other potential food sources. But as indicated in the present results, in iris plants that provide flowers sequentially for several weeks, large and colorful flowers are almost always next to flower buds and spent flowers. Naïve uncommitted bees that visit an iris bud or a spent flower could quickly switch over to identify the intact flower with the reward because the signal of the flowers’ geometry is not only highly conspicuous and distinctive, but also a more honest advertisement than its yellow color alone, thereby facilitating the learning process (Hammer and Menzel 1995). The quick senescence of the flower after it is pollinated is associated with a change of color and shape that signals any new potential pollinator to be preferentially attracted to the signal provided by the shape and size of the more-colorful new flowers. Experienced pollinators at the plant would likely avoid wasted visits to those non-rewarding flowers. Similarly, in the huge iris-flower buds that are not encased (hidden) by green sepals, but instead display the showy outer tepals with their nectar guide, the signal of the openflower shape with its nectar guide is a distinct and honest advertisement of reward, not a camouflage of it. All of the Yellow Iris flowers I observed were in a large-bud stage for 2 days, available to pollinators for only ~1day, and then wilted and senesced but remained on the plant for many days. If a bee did not distinguish between the 3 stages, but instead visited all flowers randomly, it would reduce the amount of food it received per flower by at least 10 times. My observations of Yellow Iris flowers documented the process of their explosive flower opening, and the outcomes of my manipulations provided information on the mechanism by which it occurs and led me to suggest the possible significance of the phenomenon. It appears that the explosive opening of the complex Yellow Iris flower may sharpen the signal by enhancing the difference between a rewarding and a non-rewarding flower because there are no intermediate forms. The phenomenon may therefore increase the probability that new bees attracted to the plant identify and learn the sig nal that indicates the presence of food. On the other hand, individuals that have previous experience at the plant may have N17 2016 Northeastern Naturalist Notes Vol. 23, No. 3 B. Heinrich an increased bonus due to reduced travel time and acquired skill at accessing nectar in these morphologically cryptic flowers. Correct choice, enhanced by accentuated differences, may therefore induce bees to search for other rewarding flowers of the same kind and also to return to the same ones after having visited others. Such a scenario would enhance the plants’ cross-pollination and indicate an adaptive significance for expl osive flower opening. Acknowledgments. I thank Glen Mittelhauser, David Barrington, and Daniel M. Keppie for comments and suggestions that greatly improved the manuscript. Literature Cited Bieleski, R., L. Elgar, and J. Heyes. 2000. Mechanical aspects of rapid flower-opening in Asiatic Lily. Annals of Botany 86:1175–1183. Celikel, F.G., and W.G. van Doorn. 2012. Endogenous ethylene does not regulate opening of unstressed iris flowers but strongly inhibits it in water-stressed flowers. Journal of Plant Physiology 169:1425–1429. Cohen, D., and A. Shmida. 1993. The evolution of flower display and reward. Evolutionary Biology 27:197–243. Fahn, A., and E. Werker. 1972. Anatomical mechanisms of seed dispersal. Pp.152–222, In T.T. Kozlowski (Ed.). Seed Biology, Volume 1. Academic Press, Inc. New York, NY. 430 pp. Hammer, M., and R. Menzel. 1995. Learning and memory in the honeybee. Journal of Neuroscience 15:1617–1630. Hartmann, R. 2015. Die Sumpfschwertlilie, Iris pseudacorus. Available online at https://www.youtube. com/watch?v=q4M9gqNHZ0g. Accessed February 2016. Heinrich, B. 1975a. Bee flowers: A hypothesis on flower variety and blooming times. Evolution 29:325–334. Heinrich, B. 1975b. Energetics of pollination. Annual Review of Ecology and Systematics 6:139–170. Heinrich, B. 1976. Foraging specializations of individual bumblebees. Ecological Monographs 46:105–128. Heinrich, B. 2015. Rapid flower opening in Iris pseudacoris. Northeastern Naturalist 22:N11–N14. Heinrich, B., P. Mudge, and P. Deringis. 1977. A laboratory analysis of flower constancy in foraging bumblebees: Bombus ternarius and B. terricola. Behavioral Ecology and Sociobiology 2:247–266. Jin, N., T. Landgraf, S. Klein, and R. Menzel. 2014. Walking bumblebees memorize panorama and local cues in a laboratory test of navigation. Animal Behaviour 97:13–23. Keasar,T., U. Motro, and A. Shmida. 2013. Temporal reward-variability promotes sampling of a new flower type by bumblebees. Animal Behaviour 86:747–753. Liang, H., and L. Mahaderan. 2011. Growth, geometry, and mechanics of a blooming lily. Proceedings of the American Academy of Sciences US 108(4):5516–5521. Meeuse, J.D. 1961. The Story of Pollination. Ronald Press, New York, NY. 243 pp. Muth, F., T. Keasar, and A. Dornhaus. 2015. Trading off short-term costs for long-term gains: How do bumblebees decide to learn morphologically complex flowers? Animal Behavior 101:191–199. Ney-Nifle, M., T. Keasar, and A. Shmida. 2001. Location- and color-learning in bumblebees in a twophase conditioning experiment. Journal of Insect Behavior 14(5) :697–711. van Doorn, W.G., and C. Kamdee. 2014. Flower opening and closure: An update. Journal of Experimental Botany 65:5749–5757. van Doorn, W.G., and U. van Meetteran. 2003. Flower opening and closure: A review. Journal of Experimental Botany 54:1801–1812. van Doorn, W.G., I. Dole, F.G. Celikel, and H. Harkema. 2013. Opening of Iris flowers is regulated by endogenous auxins. Journal of Plant Physiology 170:161–164. Volkov, A.G., J.C. Foster, K.D. Baker, and V.S. Markin. 2010. Mechanical and electrical anisotropy in Mimosa pudica pulvini. Plant Signaling and Behavior 5:1211–1221. Whitaker, D.L., L.A. Webster, and J. Edwards. 2007. The biomechanics of Cornus canadansis stamens are ideal for catapulting pollen vertically. Functional Ecology 21: 219–225.