Rose Rosette Disease

Rosa Scarlet Meidiland showing signs of rose rosette disease infection.

In the 1940s a new disease of roses was first noticed in Manitoba, Canada, and in Wyoming on Rosa multiflora. This was quite strange as both regions are far from the rose centers of the universe, such as California and Texas. By the mid-1960s, it was noticed in parts of the Midwest, particularly in Nebraska, and dubbed rose rosette disease (RRD). By 1999 the disease had progressed to Iowa, Kansas, Colorado, Utah, Missouri, Arkansas, Illinois, Indiana, Ohio, West Virginia, Tennessee, Pennsylvania and an isolated outbreak in California.

RRD and its natural vector are of North American origin, and in most cases, initial discoveries of the problem were with R. multiflora. In more recent years it has begun to be problematic on other roses, especially those with Asian origin. As of this writing, RRD has been found to attack hybrid tea roses, floribundas, grandifloras and old-fashioned varieties.

In many cases, and in particular with respect to R. multiflora, it is deadly. Infection of R. multiflora culminates in the death of a typical plant in two to five years. Casual field observations shows R. multiflora to die after a two-year infection.

Fortunately, many of the native species of rose in North America are not generally affected by RRD. R. woodsii, R. setigera, R. arkansana and R. blanda are only slightly affected. R. californica, R. spinoissima, R. carolina and R. palustris are not affected by RRD. R. bracteata is not affected by transmission from natural vectors but can be infected by grafting. Hybrid tea roses vary considerably as to which ones are affected. Peace and Chrysler Imperial are minimally affected by RRD, but R. var. bonica is not susceptible from natural vectors but can be infected via grafting.
 

General Biology of RRD
RRD is thought to be a virus. Researchers have shown that the causal agent of infection can be transmitted partially to Nicotiana sylvestris under laboratory conditions by mechanical means, which tend to implicate a virus.

Knock Out rose with RRD compared to normal growth.

In the natural environment, the viroid is transmitted to rose plants via an eriophyid mite, Phyllocoptes fructiphilus. Phyllocoptes is host specific and only feeds on species of roses. Other members of the rose family such as Prunus, Amelanchier, Crateagus, Malus and Pyrus are not affected by RRD. R. bractea and R. var. bonica are susceptible to RRD via grafting, but are not susceptible to Phyllocoptes invasion. The disease only stays alive in Phyllocoptes for about 10 days, and then the mite has to be re-infected with the pathogen for transmission to occur. Aphids and thrips have not been shown to be vectors of RRD.

In the wild, RRD manifests itself with significant formations of witches brooms on infected plants. In the case of R. multiflora and some other roses, a preliminary symptom is bright-red pigmentation of the new growth, followed by the aberration of the witches brooms. In the case of hybrid tea roses and related hybrids, the new growth will often occur as lime green instead of bright red.

The root systems of infected plants are not under direct influence of RRD. But since the viral mechanism disrupts the carbohydrate levels in the infected plants, the whole plant essentially starves to death. It is generally considered that only the arboreal portions of the plants are directly infected.
 

Flowering disruption
I've observed that flowering on R. multiflora is disrupted by RRD and results in severe die back of flowering stems. Flowering also exposes susceptible roses to invasion by the vector mite.

In the case of hybrid garden roses such as Scarlet Meidiland, flowering does occur after infection, but the flowers are severely affected with both changes in form and in color. Flowers that are normally pure red can be mottled and streaked with white or pink. It also seems that in the case of Scarlet Meidiland, the new growth can be tinted with the same red coloration as in R. multiflora, but it is initially muted and will progress to a dark red if the infection becomes full blown. Whether the muted pigmentation denotes a partial infection is not clear. However, stems of Scarlet Meidiland that appear to be partially affected do bloom with blossoms that are irregular but not severely distorted. One clue to infection is an increase in numbers of flower buds and flowers per unit measure of stem compared to normal shoots. In addition, those blossoms that appear to be under the influence of a partial infection will set seed, whereas blossoms of R. multiflora that are infected will not set seed. But when a full-blown infection occurs on Scarlet Meidiland, blossoms will not set seed and follow the same pattern as exhibited by R. multiflora.

Another facet of RRD is increased susceptibility of both leaves and flowering shoots to typical rose diseases such as powdery mildew and black spot. While most new roses varieties are fairly immune to these problems, infection with RRD will allow an otherwise immune plant to develop the secondary disease problems.
 

Left: Death is the ultimate fate of RRD on R. multiflora. Right: Flowering shoots of R. multiflora showing RRD symptoms.

 
RRD control
Since the causal agent appears to be a virus or a mycoplasma, chemical control once an infection occurs is not feasible. Cultural preventive control should be exercised by judicious application of Cygon 2E, which will control the vector mite (P. fructiphilus). Since this is a eriophyid mite, the more common miticides are ineffectual.

Researchers have shown that the disease can be spread by mechanical means to roses that are not normally affected by the vector mite. Extrapolation would suggest that pruners, saws and other instruments that might carry plant matter or sap from an infected plant to a non-infected plant could result in further infections down the line. In this case, sterilization of cutting tools with a disinfectant or alcohol between uses on different plants would be prudent.

It is generally thought that only the arboreal portions of the infected plants are susceptible to the problem and it does seem prudent that judicious pruning far below the point of infection may eliminate the problem, at least on a temporary basis. If a plant is severely infected, complete removal is recommended along with spraying the general area with Cygon to stop transportation of the accompanying mites. This particular eriophyid mite does not fly and can only move about by wind or hitchhiking. Further control is helped by the removal of all invasive volunteer R. multiflora in the immediate area, even if they are not infected. Burning infected plants is strongly recommended to eliminate both the pathogen and the vector mites.

In Canada, a thornless variety of R. multifora is in use as an understock for grafting, and some Canadian nurseries maintain stock blocks to provide seed for this desirable strain. RRD may be a serious problem in these circumstances, and extra care should be exercised to scout for the existence of this pathogen as well as volunteer R. multiflora in the surrounding area. If RRD is discovered, replanting of the seed stock blocks in new uninfected areas is strongly recommended.

RRD could pose a significant problem to nurseries that have large rose-production programs. While the incidence for infection of commercially available roses in the landscape is limited, it is occurring with increasing frequency. The occurrence of this pathogen in commercial production should be cause for increased diligence and concern. During a tour of a production nursery with a large quantity of roses in containers, no RRD was found. However, when walking the perimeter of the nursery, significant amounts of RRD was found in the weedy populations of R. multiflora. The nursery responded by destroying all of the populations of R. multiflora along the perimeter and with a prophylactic spraying of Cygon to eliminate residual mites. This is a good management practice, but it will have to be repeated at least yearly to discourage the re-importation of RRD into the nursery.

H. William Barnes is owner of Barnes Horticultural Services LLC in Warrington, Pa.; bhs16@verizon.net.