Skip booklet index and go to page content

COSEWIC Assessment and Status Report on the Eastern Flowering Dogwood in Canada


Cornus florida is an understory and edge tree of mid-age to mature forests. Its flowers (made conspicuous by large floral bracts) are insect-pollinated and its clusters of red fruit are mostly bird-dispersed.

Life Cycle and Reproduction

This species is a long-lived, slow growing, woody perennial (typically living to 125 years, Strobl & Bland, 2000). Before the arrival of dogwood anthracnose (Discula destructiva), the exotic fungal disease that is causing a major decline in the species, individuals with trunk diameters of 20-30 cm were occasionally encountered in southern Ontario. The smallest individuals observed flowering were in the range of 4-7 cm, diameter breast height (dbh). Flowers are perfect (hermaphrodite) and occur in mid-spring as the leaves are beginning to open. In 2004, flowering in Ontario occurred between May 12 (floral bracts open but flowers not quite at anthesis (flowers open and ready to release pollen)) and May 27 (the last of the floral bracts present, just past anthesis). Fruits mature and turn red by late summer (Soper & Heimburger, 1982).

Individuals are cross-pollinated by insects (Mayor et al., 1999). There is evidence of partial self-incompatibility (Craddock et al., 1997). Hybridization has not been reported in natural stands of Cornus florida in Ontario or elsewhere; however, in cultivation it has been hybridized with the western Cornus nuttallii and the oriental C. kousa and there are hybrids of these species in the horticultural trade.

The mature fruit are dispersed by numerous birds and a few mammals. Germination and seedling establishment occur in forest shade but intermediate light levels promote growth (Strobl & Bland, 2000; McLemore, 1990). Although most of the individuals are well separated, suggesting an origin from dispersed seeds, some reproduction occurs by vegetative means, primarily layering of low lying branches. This seems to be most prominent in smaller plants where large clusters of apparently once connected stems can be found. On examination, some near-surface connections can be found between upright stems. In some cases of clustered young stems, the stems may also have arisen from stump sprouts of trees that have died.  Prior to the arrival of dogwood anthracnose, occurrences of clonal growth were not commonly observed by the writer. From personal observation in the field, there was some suggestion that dogwood anthracnose may be stimulating limited clonal growth in the form of multiple trunks slightly beyond the common base plus the spread of low clonal growth at ground level.

Seeds germinate in the spring a year or two after dispersal (Strobl & Bland, 2000) after having experienced a winter on the forest floor. The timing for germination can be accelerated by first providing warm stratification for 60 days followed by a cold stratification (just above freezing) for 90-120 days (Kock, 1998).


Deer and rabbits are known to browse on this species (Strobl & Bland, 2000). From field observations, this appears to be a problem only for seedlings and saplings, possibly promoting the proliferation of low vegetative shoots.


No special physiological requirements or adaptations are known for this species, other than the previously mentioned pre-treatments for seed germination.


Numerous birds and mammal species have been reported to feed on the fruit (Tirmenstein, 1991); no doubt many of these are effective dispersal agents. Birds may take the intact fruits beyond the immediate area of the parent tree and later separate the seed from the flesh and drop the seed, consume the fruit whole and later pass the seed through their digestive system at a greater distance, or cache them for later consumption. Gravity also plays a role, but these seeds would move little beyond the parent tree unless picked up by mammals. Populations tend to be clusters of individuals within a larger forest habitat, suggesting that most of the effective dispersal is over a short distance. Seeds carried longer distances are less likely to fall in a suitable habitat.

Dispersal across the Canada-US boundary is possible, for example, by migratory birds. With the original occurrence of dogwood anthracnose in the US, populations there are also diminished and likely the remaining trees are producing little fruit, as is the condition in Canada.

Forest fragmentation in the Carolinian Zone reduces the possibility of effective long distance dispersal, such as by birds from an occupied habitat to an unoccupied, suitable habitat.

Interspecific Interactions

Cornus florida depends on pollinating insects for seed set and frugivores for dispersal.

Browsing animals, such as deer and rabbits, negatively impact populations but are only likely a serious problem during times of high population density of the browsers.  Animals that feed on flowering dogwood fruit are impacted by the loss of mature trees and reduction of fruiting in survivors (Rossell et al., 2001).

The recent arrival of the dogwood anthracnose fungusin Ontario (NRC, 2004) represents the most serious interspecific interaction, putting the survival of Cornus florida in Canada in jeopardy. Symptoms are first apparent as leaf spots and later some trees show cankers on twigs and branches. In the final stages of the disease, branch death occurs and clusters of epicormic shoots develop on the trunk; these are shoots that grow from dormant buds on a stem after a wound or other stress. Infection is thought to occur during cool wet conditions (NRC, 2004).


This species can survive periodic openings in its forest habitat and such openings provide increased light that can promote seedling/sapling growth. However, excessive light also promotes open site species that would out-compete flowering dogwood. Although it tends to grow on well-drained sites, periods of extremes in drought conditions likely limit recruitment.

Cornus florida is commercially cultivated and widely available, though most large landscape nurseries likely bring in stock from the United States which would not be suitable for re-introduction into natural habitats. Stock propagated from known local Ontario sources would be better adapted and more likely to do well; local stock is available from some native plant nurseries.