Next Page

Rose rosette disease (RRD) is spreading through much of the wild rose population of the Midwest, South and Eastern United States.  From infected wild roses it moves to cultivated roses.  Recognition of the symptoms of RRD is necessary to interrupt its spread within rose gardens.

Infection starts with one mite on one cane on one rose bush. Recognition of aberrant growth EARLY is the key to prudent reaction. RRD differs from rose diseases in which the infected parts of the rose die as a direct result of expression of the disease symptoms.  Instead, RRD is characterized by vigorous growth that is both rampant and contorted.  Death of affected canes comes later.

The strongest evidence supporting diagnosis of RRD is confirmed rose rosette disease in ‘nearby’ roses.  Nearby can mean in the same bed, the same city block, the same county, or any place where a wind burst can deposit a dislodged infection-carrying microscopic mite.  Once “nearby” roses are known to harbor RRD, the prudent rosarian must become and remain vigilant. Unfortunately, no one set of symptoms can be used to diagnose RRD especially in the early stages.  Symptoms differ for roses of different parentage, in different seasons and at different stages of growth.

There are no laboratory tests to confirm a diagnosis of RRD; graft transmission tests can confirm the disease when the recipient plant (R. multiflora) develops RRD. A lack of transmission does not prove lack of the disease, just lack of transmission.  Dr. Jim Amrine (personnal communication) suggests that  graft transmission tests are more often successful in spring than in other seasons.  This, to me, corresponds to  the time that multiflora puts on most of its new growth (when its growth isn't affected by RRD).


The overwhelming majority of photographs of RRD illustrate kinds of witches’ broom:  leaves and stems are reddish purple, and so contorted as to be indistinguishable one from another ( Fig. 1).  Witches' broom is a common symptom of many diseases of sick plants.  On roses closely spaced leaves form a rosette of leaves growing very close together on a stem at the end of a cane or at a leaf axil . A variation is where too many stems emerge from a single site; the worse I ever counted was in Canada where eleven short stems emerged from a single leaf axil on 'Snowqueen'. That sort of growth happens only after the rose has been infected for several months.  In addition, some roses with RRD will never show those colors nor will you ever see that degree of stem distortion (Fig. 1, Chapter 1). Photographs of RRD in its different stages are rare, and expectedly so.  Photographing sick roses does not rate high among hobby choices, and once the rose is suspect, removing it quickly is prudent.

Click on Pictures to Enlarge
Fig. 1 Reddish purple growth of severely infected new vigorous basal breaks on a floribunda.   This rose showed only normal leaves a month earlier.   This color is common on fall RRD growth on some HTs and floribundas but healthy roses' reddish growth changes to green quickly. Fig. 2  Subtle  pastel colors featuring reddish pink growth on contorted, subized leaves.  These leaves will never grow to normal size.  Note the normal leaf in the background from before the RRD infection.

My husband and I have had the opportunity to watch RRD infected roses in gardens, fields and along public rights of way. The roses at Lincoln Memorial University (LMU) Rose Garden in Harrogate Tennessee are downwind from acres of infected ‘R. multiflora’.  In retrospect, we think that some garden roses there showed symptoms in summer of 1999, but we lacked the knowledge to recognize them.  There is no doubt that infection was being spread from the infected wild roses to the cultivated roses. Doubtless the cultivated roses were sharing it among themselves as well.

We watched other roses in Union, Grainger, Monroe, and Claiborne Counties, TN, all in eastern Tennessee, - big mature bushes, of several types, sicken with RRD.   There were subtle symptoms of rose development that we noticed as being different from the ordinary,  and we watched those symptoms develop and evolve, before the appearance of the more common RRD symptoms.  On other roses the typical symptoms never appeared, but the roses were  undeniably sick.


Garish reddish purple: This color is common on fall growth on many healthy roses, on which the garish red ‘disappears’ with age. On RRD infected roses, the garish color persists (Fig. 1).

Reddish-pink: A less garish shade appears along the margins of greenish leaves and on the undersides
of new leaves (Fig. 2).  

Bronze tints grading to garish lime green: Unexpected shades of green on stems and leaves of infected 'R. multiflora’ and other roses  (Fig. 3)  resemble some mineral deficiencies and chlorotic conditions. No nutrients can return the roses to normal colors.

Fig. 3 Classic Witches' broom on an unknown rambler.
Roses with RRD can continue to bloom on healthy or recently infected canes, although overall bloom production is reduced and the blooms are often deformed or missing parts.
Fig. 4  Enlarged caliper (stem diameter) of RRD sickened  cane on 'Scarlet Meidiland'.  The cane from which these thorny canes emerged is much thinner than the thorny cane. This fall growth lacked leaves and had only thorns for eight weeks.

Gray: In east Tennessee some plants developed the look of powdery mildew long before other roses in the same garden (even those more susceptible to pm).  Many of these plants have been followed and found to develop symptoms of RRD (Fig. 10).

Black and Purple/Black: Some big ramblers have stems and canes that died quickly from RRD and are black to purple black in color while canes that died of other causes have a grayish-brown color (Fig. 5).


The development of  the stems and canes depends on the timing of the infection.   Rose canes may grow in a spiral pattern or may appear normal. The spiral growth pattern may be caused by differential excessive growth on one side - the infected side - of the cane.   A cane that is totally infected may be thicker than the cane from which it emerged (Fig. 4).  Stems that emerge from the canes  may form the classic witches' broom (Figs. 3, 5& 6) or they may appear to be a mass of twigs from a single growth apex or leaf node.  There may be three or more stems from a single bud eye, and some may appear to emerge from ‘slits’ in the cane as if they grew from  the cambium without bud eyes but were coming from dormant bud eyes so rapidly that the surface of the cane splits.

Fig.5 at left- Dead black canes on an unknown RRD infected rambler. (See Fig. 3 for this cane in May of the same year.) These stems died much faster than most RRD stems.

Fig 6 below- Small witches’ broom on ‘Dr. Huey’


The development of thorns is variable.  (In botanical nomenclature, roses are said not to have thorns, but to have prickles, instead. Thorns have vascular, conducting tissue. I follow use of the words as used by most rose growers. If it's big and draws a lot of blood when it snags me, it's a thorn.  When it's thin and tends to puncture, it's a prickle.)  RRD affected hyper thorny new growth may consist of thorns identical to those on healthy canes of the same bush or they may be different (Fig. 8).  The excessive growth of prickles and thorns (and the canes they are on) happens fast and they remain soft, almost elastic longer than thorns on a non-RRD cane of the same age.  They appear translucent; a sick four foot long cane may have soft translucent thorns from base to the tip. (These prickles and thorns will harden with time.)  In my garden in healthy roses, when a basal break is six inches long, the thorns near its base will be starting to harden. When I've seen newly grown RRD canes, all the thorns along a three foot cane may still be flexible.
Fig 7 (below) Recently RRD infected ‘Crystalline’
with pebbly leaf surfaces and elongated sepals
and shortened bloom petals.

Fig. 8 Hyperthorny RRD infected growth on HT ‘First Prize’ .  Compare hyperthorny cane on right to normal ones in  center and on left of same picture. Also, note buds and leaves that are bending backwards.

Hyper thorniness is not a dependable indication of RRD, it does alert rosarians to be vigilant. It appears on some HTs, but sick OGRs and roses related to 'R. multiflora' may have  thorns no denser than ususal. Further confusing diagnoses,  many classes of old garden roses are extremely thorny  and no one could look at a sick rugosa, damask, or spinossissima  and declare it’s hyper thorniness to be aberrant. Likewise basal breaks on some healthy HTs appear thornier closer to the bud union, but become less thorny with increasing healthy growth.  This year, I have also seen a characteristic of some roses from cold hardy breeding programs to have denser thorns near the base; these roses have a built in rodent deterrent that I had not noticed until a question from New Brunswick about the possibility of RRD in that part of Canada.


A very early symptom is the development of an irregular texture on an otherwise healthy leaf. These normally colored and shaped leaves, (Figs. 7& 9) have distinctive irregularly undulating  surfaces.  On fall growth RRD leaves the ‘pebbly’ surface is also seen on reddish purple leaves.  

Powdery Mildew results in a similar leaf distortion but on RRD leaves the distortion is not always accompanied by the grayish mycelia associated with severe pm. Some plants that will develop other RRD symptoms first show powdery mildew when other, more pm susceptible roses are pm free (Fig. 10).

 Fig. 9 at left - The pebbly surface on RRD infected ‘Dr. Huey’  typical of  powdery mildew.
Fig. 10 Powdery Mildew on leaves of an infected 'Crystalline'.  This Powdery mildew did not correspond to the time of appearance of PM on roses some  might call mildew magnets. Also, note the new elongated discolored leaves.                                      

Red veining on green leaves, a so-called mosaic pattern is another symptom (Fig.11).  The term mosaic pattern for this discoloration was illustrated in color in Amrine and Hindal (1988, p2). Mosaic is a plant pathology term for a specific symptom in which the boundaries of the two colors are sharply defined.  (Mottle refers to symptoms where the boundaries are not sharp).  This is not "Rose Mosaic Virus" (Fig. 11b) which is the name for a group of  non-contagious and non-fatal rose diseases that are spread only by grafting. (The possibility of pollen transmission or root to root  graft transmission of these latter viruses is beyond the subject of this e-book.)

Fig. 11 below - So-called mosaic pattern on the leaves of an unidentified RRD infected Large Flowered Climber in fall.
The above picture is of RRD
Fig. 11b below - 'Love & Peace' ,a new introduction by Week's Roses, with Rose Mosaic Virus. RMV is the name of a disease while "mosaic" is a symptom. RMV is not the same as Rose Rosette!

Fig.12 above- An RRD infected rose coming back. Yellow green subsized chlorotic leaves that look like a zinc deficiency coming up from remaining root segments. If RRD is seen on only one or two canes of a bush and diagnosed as  zinc deficiency, one must ask why the entire bush isn't short of zinc or why other roses in the bed aren't also zinc deficient.

The shapes of the leaves also varies. Some appear normal but others are narrow or barely resemble leaves (Figs. 2&9).  Horst’s Compendium of rose diseases illustrates Zinc deficiency (plate 79) and this is a look that some RRD leaves have (Fig. 12).

Most rose leaves are symmetric. When there is a pronounced asymmetry (if you fold the leaf along the stem, both sides should meet exactly on healthy leaves), RRD is affecting the growth of the leaf, to its detriment. (Please see the additional pictures in Chapter 11, for a further discussion).

The spacing of leaves on a stem or cane is also a tool for diagnosis: a short internodal distance,  the distance between leaf axils, is often characteristic of RRD (Fig. 12). The ultimate short internodal distance results in a rosette of rose leaves.  In addition, the leaves may remain folded along the midveins and appear in the contorted patterns seen elsewhere. ( Once you start looking closely at roses you'll see a common thread of closer leaf axil spaces on the bottom two inches of large canes; with RRD those basal will be even closer spaced and the leaves farther out will be more closely spaced as well.)

Epinastic leaf distortion is a futher symptom.  This is a distortion when the leaves appear to bend over backwards (Fig. 8).


A stem may produce some normal growth, only to terminate in an area where  new leaves are  reduced in size and where  the bud is deformed and  may have elongated sepals (Figs. 4,9,13).  These sepals are very  thin, narrow and different from  the enlarged and elongated sepals of roses such as ‘Uncle Joe’.

Other blooms may develop enlarged receptacles, exhibiting the kind of swelling that occurs after successful pollination, except these blooms were never fertilized.  These are parthenocarpic hips. Sometimes the sepals will open slightly and petal color can be seen, other times the sepals are partly fused along their margins.

We have seen three examples of blooms where there were no petals, no recetacles, only sepals that are aberrant.  In one case the blooms resembles small brussels sprouts with hollow centers.

Some RRD infected roses will bloom, but the resulting blooms are aberrant. The petal count may be  reduced to 10 or 15 petals. Seldom is the petal color the correct color for that cultivar (Fig.4) The petal color is often the first characteristic that’s noticed.  Rose petals with clear uniform color may appear  mottled  and  less dense. For example, red roses will produce mottled pink blooms and pink roses will produce mottled  lighter pink blooms. White roses will produce white blooms but the petal count and shape differ from  healthy blooms.

Bud shapes and sizes will even vary within a bush. On a bush with RRD, what will pass for blooms may develop much more rapidly than on normal bushes, but the blooms will be small and incomplete.

RRD interrupts development of the bloom in many different ways; we do not know if the differences are due to differences in concentration of the viroids or to cultural differences or even to synergy with other viruses in the roses compounded with seasonal differences.

The Consulting Rosarian Manual cites Phyllody as a symptom of RRD; phyllody refers to leaf-like flower petals similar to proliferation.  I have not seen that yet, but I do see some roses that continue to bloom, albeit badly (Figs. 9,13).

Fig 13. Hyperthorny cane of ‘Fourth of July’ with leaves with shortened internodal distances and  enlarged caliper farther from the bud  union.  

Fig. 14 on right-  At the top, a bud of ‘First Prize’ with RRD shows spiral distorted sepals, garish red-purple in color. Lower the dense RRD thorns contrast with the normal spaced thorns at left.  There are three buds in this rosette of aberrant growth.

Fig. 13b Enlargement of Fig. 13 above showing Hyperthornyness on swollen cane of Fourth of July. These are massed prickles unlike the vicious thorns that FoJ is known for.  


Winter is an excellent time to look for aberrant growth: canes with too many thorns should be checked to see if the internodal distances are normal or shortened.  Growth patterns of branches should also be checked...three or more branches from a single stem (when other roses of similar heritage have only one) is another reason to worry.  General contortions show better when there are neither leaves nor flowers to distract.

Spring is a time of vigorous growth for all roses; most rose growers are familiar with blind shoots, but a rose that produces all blind shoots should be immediately suspect.  Some spring new growth will feature colors that set it apart from healthy canes on the same bush.  On healthy canes the colors will soon merge with the rest of the colors of the gardens, on RRD plants, aberrant growth never corrects itself.

Summer is usually a time of quiescence in Tennessee.  Growth slows, even for roses with rampant RRD growth and many of the colors of infection are muted.  Summers I used to see very little reddish purple RRD growth, although ramblers continued to grow with slowly spreading infections. (Summer 2003 has caused a change in this.  2003 was a very wet summer and RRD sick multiflora retained its reddish look through summer and into fall. This suggests a color-pigment-rainfall link.)

Fall follows the winds of spring and summer, the winds that redistributed the mites. The damage done by the mites emerges as new and nasty growth: often bright reddish purple, often contorted. Fall freezes often damage the reddish purple growth before the older, normal growth.


A single symptom does not make  a confirmation of RRD.  When the growth character of a cultivar is known, appearance of multiple RRD symptoms will alert the rose owner to possible danger.   

Some symptoms of RRD resemble herbicide damage; a thorough search of the area around the roses for signs of herbicide use must allow for long distance herbicide drift - especially in warm weather. A rose bush outgrows herbicide damage, RRD infections persist from year to year until the death.

Roses likewise outgrow ‘blind shoots’ but the blind shoot looking growth on a rose in Georgia became more symptomatic of RRD two months later (Fig. 6).  

Weather damage in spring or fall can also cause contorted new growth, but comparison with growth on other roses in the same garden can be used evaluate the problem. If the damage was weather caused, the rose will out grow it.  

Because ot the chance of false diagnosis, each rosarian needs to think about the problem and decide in advance a course of action. Then when RRD enters a garden, the hard decisions have already been made.

If  RRD has not been reported nearby, and there are symptoms that suggest it, you may choose to reduce the mite population that transmits the infection.

Digging up and isolating the questionable rose should also be considered.  In addition, I would (and have had to) spray the bush in question and the surrounding roses at ten day intervals with Cygon 2E (following  label precautions) while waiting for confirmation of RRD.  Cygon 2E is the only spray that will kill the eriophyte mites.  They are different from spider mites.  Most chemicals approved for treatment of spider mites are not effective for the mites that transmit RRD.  At present (2007) Cygon 2E is not readily available. There have been no tests to see if other miticides kill P. fructiphilus.  Avid is approved for general use against eriophyid mites, but its efficacy in fighting RRD vectors is unknown.  

Scientists who study the mite populations do so in their laboratories.  One way of capturing the insects and mites is to spray the plant with high holding hairspray and take it to the lab for specimen counts and identification.  I have recommended the hairspray to keep the mites on a RRD sick plant as it is being bagged and removed from a garden.  Two recent papers from Iowa State indicate that there are large populations of the mites on RRD sick roses.  Stopping their movement should be an imperative for a gardener who is about to remove a RRD sick rose.  

There are no ‘organic’ controls other than an indigenous beneficial insect/mite/etc. population.

There are no lab tests to confirm RRD.  A check with Agdia confirms that none are under development, Mike Tiffany, Agdia, (Pers. com.  Dec. 8, 2000).  Diagnosis is  visual examination by someone experienced in the identification of Rose Rosette.

Meanwhile, we wait and we watch very closely.                                                       Next Page