Long Island Horticultural Research Laboratory, Cornell University, Riverhead, NY 11901, USA

Background and objectives
The North American native flowering dogwoods, Cornus florida in the eastern USA and Cornus nuttallii in the western USA and Canada, are hosts of a relatively new anthracnose disease caused by Discula destructiva Redlin [1]. The non-native Cornus kousa, an oriental species grown as an ornamental, is also a host but is less susceptible than native dogwoods. Dogwood anthracnose was first noted in the north-western USA in 1976 and in the east in 1979. Strong public concern about the disease was voiced immediately in the east, because C. florida is a widely planted ornamental that provides a beautiful spring bract display in gardens and parks, while the woodland trees beautify roadsides. The flowering dogwood is a major understorey species in eastern woodlands. C. florida fruits are also a valuable food source for wildlife, including migratory birds. Researchers have been challenged to understand the origin of the pathogen, its impact, and the environmental influences on disease development, with the goal of developing control actions appropriate for forest and ornamental dogwoods.

Results and conclusions
The impact of dogwood anthracnose is severe on both C. florida and C. nuttallii. Leaf lesions may be delimited by purple rims, or may develop as extensive blotches. Often the entire leaf is blighted. Blight may progress down the petiole into the current-season shoot, resulting in death of the twig. Symptoms appear in the lower branches initially. Extensive crown dieback often causes the development of epicormic branches, which are highly subject to infection. D. destructiva sporulates on necrotic portions of leaves and twigs. Inoculum production on the previous season's shoots is the primary source for spring leaf infections. C. florida may be killed within 3 years; seedlings may be killed in one season. Tree mortality is sometimes accelerated by drought and/or root rot. Dogwoods in favourable sites may appear severely diseased for a few seasons and then recover if spring environmental conditions shift. Dogwood population reduction from 112 per hectare to only 13 per hectare was documented over a 4-year period at one woodland site where anthracnose was severe. Data from impact plots in seven southern states supported an estimate of 7 million hectares with diseased dogwoods by 1993.

Open settings with exposure to sunlight (more typical for ornamental plantings) have been found to be much less conducive to disease than the forest understorey. Measurements of evaporative potential (EP) and light were consistently lower in the canopies of understorey trees than the canopies of exposed trees. Low EP and low light levels were associated with higher disease severity. Disease was positively correlated with elevation, northern or eastern aspect and nearness to streams, and negatively correlated with dogwood population density. Analysis of DNA from isolates of the pathogen from east and west coasts has indicated that the fungus is probably introduced, perhaps in two separate introductions to the two coasts. The results of comparisons of dsRNA are consistent with this hypothesis. The nearly simultaneous appearance of the disease in the vicinity of two ports with major arboreta (Portland and New York City), its rapid progression through the dogwood population, and the severity of the disease's impact also suggest exotic pathogen introduction.

1. Daughtrey ML, Hibben CR, Britton KO et al., 1996. Plant Disease 80, 349-358.