ISSN 1188-603X

No. 194 May 30, Victoria, B.C.
Dr. A. Ceska, P.O.Box 8546, Victoria, B.C. Canada V8W 3S2


From: Ted Mosquin, Box 279, Lanark, Ontario KOG I K0

Abridged from: The Canadian Field-Naturalist, Volume 99, Number I, Jan.-Mar. 1985

[Since I read Dr. Mosquin's article published in The Canadian Field-Naturalist, I have been poking bunchberry flowers at every opportunity. I asked Dr. Mosquin to adapt his article for BEN. I hope that you will enjoy playing with bunchberry flowers as much as I do. - AC]

This is the time of year to go out and take a look at the floral pollination method of Cornus canadensis. All parts of the flower are synchronized to explode in a split second to affect pollination. The only other species in the genus, C. suecica (of Eurasia) possess an identical mechanism. The basic elements of the mechanism are: a sensitive antenna-like structure projecting from one petal of the unopened flower bud, reflexive petals (on a tensile 'spring'), and stamen filaments also possessing "elbow springs" which act to catapult the pollen in the anthers upwards toward the top of the flower. In comparison to all other 'rapid movement" pollination mechanisms in the Angiosperms, this explosive mechanism is singularly unique.

Explosive or other rapid-movement floral mechanisms related to pollination are rare in the plant kingdom. In known examples where rapid movements do occur, it is the anthers or the stamens which move more or less alone or in consort with restraining petals. Examples described in the literature include the "explosive anthers" of Pilea microphylla Liebm. ( Taylor 1942, p. 608), often referred to as the "artillery plant" and Urtica (H.G. Baker, pers com.) both members of the Urticaceae. Another example occurs in Kalmia angustifolia L. (Marie Victorin 1942, p. 466) a member of the Ericaceae where the anthers are partially embedded in the petals and are simply released at maturity catapulting their pollen in the direction of the stigma (and at any pollinating insect) but with the petals playing a stationary role. The examples of Medicago (alfalfa) and Sarothamnus (Scotch Broom) both in the Leguminosae have been widely reported in the literature (e.g. Meeuse 1961). In the genus Lopezia (Onagraceae) and in the genus Hyptis (Labiatae) the stamen, when touched, snaps upward and deposits pollen on the underside of an insect visiting the flower (P. Raven, pers. com). Some other examples brought to my attention by H. G. Baker (pers com) are Stylidium (Stylidiaceae) Mucuna and Ilex (Leguminosae), Odontonema (Acanthaceae) and Ravenala (Musaceae). A number of less rapid floral movements related to pollination are described by Meeuse (1961).

I first learned about the pollination mechanism in C. canadensis in 1968 while conducting observations and experiments on the reproductive biology of native plants in Banff National Park, Alberta. It was not until 10 years later that I was to find out that mine was not the first observation or recording of this explosive mechanism. The first and, to my knowledge, the only previous reference is contained in a one-line note by Marie-Victorin (1942) in Flore Laurentienne. In his description of Cornus canadensis L. he noted that "Les vrais petales et les etamines vent elastiques et reagissent lorsque un insecte les touche." He did not record any observations on an "antenna" nor did he comment on the presence of a similar mechanism in Cornus suecica L. although the latter species was also included in his flora.

The inflorescence of Cornus canadensis is comprised of four white showy involucral bracts surrounding a few to several dozen small, relatively inconspicuous flowers. The true petals are a very light greenish yellow to nearly white in colour. The filaments and anthers are also very light coloured. However, the ovary, style and stigma are a very dark purple and provide striking miniature contrast points in the inflorescence. In the Banff area the number of flowers in each inflorescence varied from 8 to 15.

The pollination mechanism of C. canadensis is so intrinsically interesting as well as unique in the plant kingdom that I feel it would be worthwhile to record here a brief description of how the discovery was made. This account is based on plants growing near Altrude Creek about one mile south of Mt. Eisenhower Junction in Banff National Park.

Lying stomach down on the forest floor looking through a microscope, I began to examine the plants searching for a series of flowers in various stages of maturation. Normally, it is relatively easy to assemble such a sequence ranging from young buds to flowers that are very old and in a state of senescence. But each flower of C. canadensis was either still in the bud stage or completely open with the petals very strongly reflexed outwardly or downward. In all open flowers the anthers extended upward, well above the tip of the stigma, and were empty of pollen. While the absence of flowers in the process of opening was puzzling, I attributed it at first to local environmental factors and continued to examine more inflorescences. The absence of pollen in the open flowers also seemed unusual but as the area was frequented by many species of pollen-feeding flies (Syrphidae), I speculated that perhaps the pollen had been collected by these colourful flower visitors.

Then with dissecting needles I began opening one of the buds, only to discover that it seemed to transform itself in a fraction of a second into a fully open flower. I turned to a second bud, opened it and found four normal-looking, fully developed, undehisced anthers. I tried a third and was again surprised by what appeared to be a tiny explosion and what seemed like a small amount of pollen flying in all directions. I realized then that I might be looking at a unique phenomenon-perhaps never before witnessed by humans and perhaps undescribed. It was then that I began to pay more attention to another unusual characteristic of each flower. On the abaxial side and near the tip of one of the four petals of each unopened flower and projecting upward was a miniature "antenna" just over one mm long. It did not take long to establish that even the slightest touch of the dissecting needle to the antenna of any "ready to pop" flower would trigger the explosive mechanism; the petals would reflex, the anthers would spring out simultaneously like four tiny catapults and shoot their entire pollen loads into the air above the inflorescence.

The mode of attachment of the anther to the filament, its position, and the timing of its development within the bud cast some light on the morphological basis for the popping action. In the young bud, that is, one that is not mature enough to be tripped by mechanical means, the four anthers are undehisced. As the bud matures, anthers dehisce fully while still in the bud (but do not release their pollen). Once dehiscence is complete the pop mechanism seems to become activated and the slightest touch to the antenna will cause the flower to burst open.

The attachment of the anther to the filament is basal but in the enclosed bud the anther itself is pointed downward so that in younger buds the attachment of the filament appears to be terminal. This means that the abaxial (outside) surface of each anther is appressed tightly against the upper part of the pistil.

As the bud and the pollen mature, slits are formed on the abaxial surfaces of the anthers. At this stage both the reflexed tip of the filament as well as the petals come under a powerful tension which is released only when the pop mechanism is triggered. During the "popping" the arched tips of the filaments act as elbow springs and the four anthers snap upward. The "popping" of the flowers occurs so rapidly as to be scarcely perceived by the naked eye even through the microscope. After popping, the stamens assume a more or less vertical position and appear to be empty of pollen. Pollen grains are light yellow in colour and slightly sticky; they are too large and heavy to be carried away by the wind.

Aside from the popping mechanism, which seems to be a device favouring cross pollination, little is known about the breeding system of plants of this genus. When a flower pops, some pollen is deposited on its own stigma so automatic self-pollination is possible. It would be interesting to find out whether an incompatibility system is associated with the popping mechanism.

While at Banff, I failed to record a single insect visitor on flowers of this species. However, Sadlier and Sadlier (1977, p. 100) published a photograph showing a wild species of the leafcutter bee genus, Megachile, visiting an inflorescence in which about two-thirds of the flowers had already popped. The leafcutter bees are known to be major collectors of pollen (Hobbs and Lilly 1954; Krombein 1967; Rank 1982) which they use in quantity to provision their nest cells. My observations of Cornus flowers showed a complete lack of nectar and it is likely that various Megachile species, which occur throughout the North American range of Cornus (Ivanochko 1980) are the principal pollinators. Pollen eating flies (Syrphidae) may also be effective pollinators. However, in view of the sensitivity of the antenna of any "ready to pop" flower, it would appear that even very small insects such as woodland midges could act as pollinators should they happen by chance to fly from one inflorescence to another. In view of the seeming force of the popping mechanism and the presence of the needle-like antenna, it would be surprising if the fully mature buds did not pose some threat to life and limb of the smaller and more fragile of the woodland insects.

The colour of the involucral bracts is creamy white and it is known that the bracts strongly absorb ultraviolet light (UV photo in original article). This means that the higher groups of pollinating insects such as the bees would be able to distinguish the bracts from the adjacent and background objects quite clearly ( Mazokhin-Porshnyakov 1969). Thus it is not the flowers but rather the involucral bracts which make it possible for bees to effect cross pollination.

The conclusion that C. suecica has an explosive mechanism very similar to C. canadensis is based upon examination of herbarium specimens at Agriculture Canada (DAO) and the National Museum of Natural Sciences (CAN). The flowers of these two species are essentially identical except that the floral parts of C. suecica, including pistil, stamens, petals and the tiny sepals are all deep purple in colour.

An understanding of the reproductive biology of a taxonomic group will strengthen the foundation upon which taxonomic judgements are made (Ornduff 1969). Cornus canadensis is usually included with the genus Cornus (Marie-Victorin 1942; Harrington 1954; Scoggan 1957; Moss 1959; Munz and Keck 1968), a north-temperate genus of shrubs and trees with some 45 species (Lawrence 1951). Some modern European floras recognize C.canadensis and C. suecica as a distinct genus: Chamaepericlymenum (Shiskin 1951; Clapham et al. 1962) although all authors with the exception of Marie-Victorin (1942) were apparently unaware of the unique nature of the floral mechanism. The reproductive characteristics described herein provide strong added argument, I feel, for separate generic status for these two species. [In the Pacific Northwest we have yet another species, Cornus unalaschkensis Ledeb. that originated as an allotetraploid from C. canadensis L. x C. suecica L. - cf. Bain & Denford, Bot. Not. 132(1979):121-129. - AC]

To make certain that any species in the genus Cornus did not contain the antenna feature so characteristic of the pop flowers of these species of Cornus, I examined petals of flowering specimens of all species of Cornus in the extensive herbarium collections at Agriculture Canada (DAO) and the National Museum of Natural Sciences (CAN) at Ottawa; all had petals with smooth abaxial surfaces; petals lacked any semblance of pollination structures described here.

The widely used but mundane common name of this species is "Bunchberry." A much more interesting, meaningful and dynamic name would be "Pop Flower."

Literature Cited

Clapham, A. R., T. G. Tutin, and E. F. Warburg. 1962.
Flora of the British Isles. Cambridge University Press. 1269 p.
Harrington, H. D. 1954.
Manual of the Plants of Colorado. Sage Books, Denver. 666 p.
Hobbs, G. A., and C. E. Lilly. 1954.
Ecology of Species of Megachile Latreille in the Mixed Prairie Region of Southern Alberta with Special Reference to Pollination of Alfalfa. Ecology 35(4): 453-462.
Ivanochko, M. 1980.
Taxonomy, Biology and Alfalfa Pollinating Potential of Canadian Leaf-Cutter Bees-Genus Megachile Latreille (Hymenoptera: Megachilidae). M.Sc. thesis, Macdonald College Library, McGill University, Montreal. 378 p.
Krombein, Karl V. 1967.
Trap-Nesting Wasps and Bees: Life Histories, Nests and Associates. Smithsonian Press, Washington, D.C. 570 p.
Lawrence, A. G. 1951.
Taxonomy of Vascular Plants. The Macmillan Co., New York. 823 p.
Marie-Victorin, Frere. 1942.
Flore Laurentienne. Les Freres des ecoles Chretiennes, Montreal. 916 p.
Mazokbin-Porshnykov, G. A. 1969.
Insect Vision. Plenum Press, New York [Translated from Russian by R. L. Masironi and T. H. Goldsmith]. 306 p.
Meeuse, B. J. D. 1961.
The Story of Pollination. The Ronald Press Co., New York. 243 p.
Moss, E. H. 1959.
Flora of Alberta. University of Toronto Press. 546 p.
Munz, P. A., and D. D. Keck. 1968.
A California Flora. University of California Press, Berkeley and Los Angeles. 1681 p.
Ornduff, R. 1969.
Reproductive Biology in Relation to Systematics. Taxon 18: 121-133.
Rank, G. H. 1982.
First International Symposium on Alfalfa Leafcutting Bee Management. University of Saskatchewan Press, Saskatoon. 281 p.
Sadlier, Ruth, and Paul Sadlier. 1977.
Short Walks along the Maine Coast. The Pequot Press, Chester, Connecticutt. 131 p.
Scoggan, H. J. 1957.
Flora of Manitoba. National Museum of Canada, Ottawa, Bulletin 140. 619 pp.
Taylor, N. 1942.
The Practical Encyclopedia of Gardening. Garden City Publishing Co., New York. 888 p.

Submissions, subscriptions, etc.:
BEN is archived at