Multiflora Rose and Rose Rosette Disease

Multiflora rose (MFR) is classified as a noxious weed in numerous states, including Iowa.  The plant can be found throughout Iowa, but is most common in areas where row-crop agriculture does not dominate the landscape.  Thus, MFR is most prevalent in southern and northeastern Iowa.  The species was first introduced into the United States in the 1800’s for use as rootstock for ornamental roses, and later was promoted as a conservation plant for cover and as a food source for animals.  Seed of MFR is spread long-distances by birds, deer and other animals that feed on the rose hips.  The plant also can spread vegetatively by layering of cane tips.

Traditional control programs for MFR rely on herbicides.  Although these products provide effective control of MFR, they have several disadvantages.  First, herbicides are relatively expensive, especially considering that MFR primarily occurs on land that provides landowners relatively small returns.  MFR frequently is found on rough ground where it may be difficult to reach with spray equipment.  Finally, herbicides used for MFR control have the potential to damage sensitive plants that occur in the infested areas.  Because of these factors there is interest by landowners in areas plagued by MFR for alternative control tactics.

Rose rosette disease (RRD) was first reported on MFR in Canada in 1940.  The first report of the disease in the Midwest occurred in 1968 in Nebraska.  There were few additional reports of the disease in the central U.S. until 1976 when the disease was identified  on cultivated roses and multiflora rose in Kansas and Missouri.  A survey conducted in 1999 reported occurrence of the disease in Iowa, Nebraska, Kansas, Colorado, Utah, California, Missouri, Arkansas, Illinois, Indiana, Ohio, West Virginia, Pennsylvania, and Tennessee.

Considerable research with RRD was conducted in the 1980’s and early 90’s, and while much was learned about the disease, wide gaps in our knowledge remain.  RRD is able to infect most species of the Rosa genus, whereas other plants in the Rose family appear to be tolerant to the disease organism.  Within the Rosa genus there is considerable range in susceptibility, with MFR appearing to be most easily infected and most susceptible.  The earliest symptoms of RRD are development of a deep red color on the undersides of leaves.  Affected stems may grow vigorously with a pink to magenta coloration.  Later symptoms include small, distorted leaves with a bright red coloration..  Leaf petioles are shortened and affected stems develop a rosette appearance.  Plants generally survive for two to five years following development of initial symptoms before eventually succumbing to the disease.

The causal agent of RRD has not been identified, but it is thought a virus may be responsible for the disease.  The disease has been successfully transmitted by grafting infected stems onto healthy plants.  In the wild, an eriophyid mite is suspected as being the vector for RRD.  Although there have been few studies evaluating the rate of spread of the disease, the disease spreads most rapidly in dense stands of MFR and in conditions favorable for survival of the mite.  RRD is less prevalent on MFR growing in shade than on plants growing in full sun.  It has been proposed that mites have a preference for locations with greater than eight hours of direct sunlight per day.  The risk of movement of RRD from MFR to cultivated roses has been proposed as low due to the greater tolerance of cultivated varieties to RRD and because cultivated roses generally do not grow in close proximity to MFR.  The risk of infection decreases rapidly as distance from the infected plants increases, with little risk to plants separated by at least 300 feet.

There is continued interest in landowners in southern Iowa to use RRD as a biological control agent for MFR.   Rose gardeners are concerned that the intentional spread of the disease may increase the likelihood of cultivated roses becoming infected.  Current research at Iowa State University is designed to provide a better understanding of the risks and benefits of using RRD as a biocontrol agent for MFR.  Objectives of this research include:  1) documenting the occurrence of RRD in the state, 2) determining distribution of the disease within different habitats, 3) determining the role of eriophyid mites (Phyllocoptes fructiphilus) as a vector of RRD, and 4) evaluating movement of RRD from MFR to cultivated roses.

In 2001 a survey was conduced in three Iowa counties (Appanoose, Taylor and Boone) to determine the natural occurrence of RRD in MFR infestations..  In Appanoose county, sites on property managed by the Army Corp of Engineers infested with MFR were identified by rangers employed at Lake Rathbun.  In Taylor and Boone county, sites were arbitrarily selected by looking for infested areas while driving through the counties.  At all sites, the type of habitat, density of MFR, and percent of plants infested with RRD was recorded.  MFR growing in wooded areas had the lowest incidence of  RRD, with less than 25% of the sites having the disease present (Table 1).  In Appanoose county the disease was found in 3 of 8 wooded habitats, whereas no RRD was identified at these sites in Boone and Taylor county.  RRD was present in all of the savannas surveyed, and in these locations MFR was found predominately under trees.  Nearly 80% of the prairie/pasture sites surveyed were infested with RRD.   The percentage of MFR plants infested with RRD ranged form 10 to 90% at the surveyed sites.  Similar levels of RRD infestation were found in the three counties surveyed.   RRD had been purposely introduced at the Lake Rathbun site in the early 1990’s.  We are unaware of any organized efforts at spreading the disease in the other two counties surveyed.   Similar rates of infection at sites where the disease was introduced earlier compared to sites with no history of introduction suggest that the disease is endemic in the state.

Table 1.  Occurrence of rose rosette disease in Iowa.

In 2002 experiments were conducted to quantify the effects of habitat on infestation of MFR with the mites believed to vector RRD.  Mites were sampled from MFR growing in two Story County Conservation Board Parks (Christiansen Forest Preserve (Site 1) and Dakins Lake (Site 2)).  At each site, 20 branch tips were cut from each of 3 healthy MFR growing in the sun, 3 plants in the shade (under nearby trees), and from 3 MFR growing in the sun with symptoms of RRD.  Plants were sampled 7 times during the summer and fall from June 13 to Nov. 22.

Phyllocoptes fructiphilus mites were found at all three locations at both sites, but were most prevalent on the MFR plants with symptoms of RRD.  Relatively low numbers (<9 mites/g leaf material) were observed on non-symptomatic plants in either the sunny or shady habitats .  The number of mites increased throughout the summer with the highest populations observed during October.  Other mite species were found on MFR branch tips and tentatively identified as cyclamen and spider mites (samples will be sent to the Systematic Entomology Laboratory for positive identification).  Unlike P. fructiphilus these mites were more common on the MFR growing in the shady habitat.

The increased densities of P. fructiphilus on diseased MFR does not necessarily indicate that the mite is responsible for disease transmittal.  The rosette growth habit associated with RRD provides abundant habitat for small mites and insects, like P. fructiphilus, that prefer to feed within the leaf axil.  Healthy MFR plants have fewer leaf axils per branch tip, and MFR pose in the shade tend to have long stems with only 2 leaf axils at the branch tips.  These plant characteristics, which are likely influenced by the amount of sunlight in each habitat and disease symptoms, may be why we observed much higher populations of P. fructiphilus on plants with RRD.

A project was initiated in the spring of 2002 to evaluate the potential movement of RRD occurring on MFR to cultivated roses.  Several plantings of cultivated rose varieties were established in central Iowa.  Each planting included three rose varieties:  1) Rosa ‘Hansa’, hybrid rugos rose; 2) Rosa ‘Nearly Wild’, Floribunda rose; and 3) Rosa ‘BAIpeace, hybrid tea rose.  Each planting included six plants of each species.  The roses were donated by Bailey Nurseries, Newport, Minnesota.  Three plantings were made at locations where no MFR is found within at least one mile.   Four plantings were made at locations where MFR occurs within 50 yards of the cultivated varieties.  At two of these locations the MFR is infected with RRD, whereas at two sites no RRD has been detected.  The roses will be closely monitored for the next several years to determine the presence of RRD.  No disease was found on any roses during the 2002 growing season, however several plants were damaged by deer feeding at two locations.

Summary  The ongoing research is intended to improve our understanding of the threat posed by RRD to cultivated rose plantings in Iowa.  Our survey of MFR infestations determined that the disease is widely distributed in the state, thus minimizing the effect of introducing the disease to uninfected stands of MFR.  However, it would be prudent to avoid introducing RRD to locations with cultivated roses in the immediate vicinity (0.5 mile or less).  The cultivated rose study initiated in 2002 should provide an improved understanding of the movement of the disease from MFR to ornamental varieties.

The project also has provided insight on the utility of RRD as a biological control agent.  It is clear that RRD can provide short-term control of MFR in many habitats; however, it is unlikely that RRD by itself will provide a permanent solution to this weed problem.  The inability of the disease to control plants growing in shaded areas provides a seed source for new infestations.  Fruit of MFR is a preferred food source of birds and other animals which disseminate the seed.  In addition, areas that have been infested with MFR will have a persistent seed bank that provides a source for reinfestation.  Observations by land managers indicate that areas where RRD has cleared MFR infestations frequently are reinfested by the plant fairly quickly.  Thus, RRD should not be viewed as the single solution for managing MFR. 

As with any pest, an integrated management program is essential for developing long-term solutions for MFR.  While RRD may provide an alternative to mechanical or herbicidal control in certain areas, supplemental tactics should be used to ensure long-term control.  Failure to control plants occurring in shaded areas or those plants developing from seed will result in rapid reinfestation of areas where RRD has killed existing plants.

The current MFR project at ISU involves John Obrycki, Professor of Entomology specializing in biological control, Laura Jesse, a graduate research assistant in the Department of Entomology, and Bob Hartzler, Department of Agronomy.

Prepared by Bob Hartzler, extension weed management specialist