FO-06594     1997

To Order

Roses for the North

Cold Hardiness

---

The Minnesota Landscape Arboretum is located in USDA Hardiness Zone 4a, ⁶ where minimum winter temperatures are commonly between -25° and -30° F (-32° to -34° C). Most rose cultivars lack sufficient cold hardiness to survive these low temperatures.

The roses in the Arboretum's Shrub Rose Garden are not given any winter protection except for a wood chip mulch that is replenished to a 4- to 6-inch depth each year. When present, snowcover provides additional winter protection, insulating canes beneath the snowline and buffering soil temperatures. But the main factor affecting plant survival is a cultivar's inherited ability to withstand harsh winter conditions.

What Cold Hardiness Is

A plant's cold hardiness has three components: acclimation, mid-winter hardiness, and deacclimation. The acclimation process is triggered late in the growing season by decreasing photoperiods as daylengths shorten, and as temperatures decline. These environmental cues induce physiological and biochemical changes in the plant that result in greater cold tolerance.

Mid-winter hardiness is the lowest temperature a plant can survive without injury after it has gone through the acclimation process and has reached its maximum hardiness level.

Deacclimation refers to a decrease in the hardiness of plant tissues in response to warming temperatures in late winter and early spring.

Mid-winter hardiness levels and the rates at which acclimation and deacclimation occur for a particular cultivar or species can vary from year to year. These annual differences are caused by changes in plant health and by annual variations in temperature patterns.

The suitability of a cultivar for a particular climate depends not only on its maximum mid-winter hardiness level, but also on the timing and rates of acclimation and deacclimation in response to environmental cues. A plant that acclimates too slowly can be injured by low temperatures in early winter. The timing and rate of acclimation have been found to be the limiting factors for the winter survivability of seven rose cultivars in Minnesota during previous research. ⁷

Plant Hardiness and Vulnerability to Winter Injury
	When a plant's winter hardiness ratings are below the minimum temperatures which can be expected for a location, the plant's canes will not be injured (Plant A). When hardiness ratings intersect the minimum temperature line, cane injury occurs (Plant B).8

Plants can also be injured when mid-winter temperatures drop below a plant's maximum hardiness level. Rapid deacclimation in response to warm temperatures in late winter will also leave plants vulnerable to injury by subsequent freezing conditions.

Observations of Injury

Field observations on winter injury were conducted at the Minnesota Landscape Arboretum during early spring of 1989, 1990, and 1991. A variety of winter injury patterns were seen among the roses. Some species and cultivars exhibit no injury at all. Others show "tip" injury, where a small amount of dieback occurs at the ends of canes.

Snow serves as a natural insulator of plants, and many rose cultivars die back to the snowline. Those parts of the canes exposed to the cold air above the snow die, while the lower parts of canes covered with snow survive.

There are also cultivars that show variation in injury pattern from cane to cane. Within one plant, canes with no winter injury can be found next to canes that are either partially or completely killed.

Other cultivars behave much like herbaceous perennials, dying back to the ground every winter.

Winter injury observations for the winters of 1988-89, 1989-90, and 1990-91 are listed in Table 4. The extent of cane injury observed on the roses was categorized as none, tip, snow, <½ (less than half), >½ (greater than half), or base.

 

Definitions and descriptions of winter injury
None: No injury is seen on these plants. Canes are hardy to the tip.	Tip: Indicates injury to cane tips that resulted in dieback of 10 percent or less of the plant.
	Snow: Indicates dieback of all canes to the snowline. Cane material below the snowline remained alive.
>½ (greater than half): <½ (less than half):	Closeup of dieback in the above panel shows this variable injury pattern.
Mixed injury to canes; respectively indicating dieback of 50 percent or less of the plant, or over 50 percent of the plant. Cane injury within a single plant can vary from none to complete kill to the ground.
	Base: Every cane is killed to the ground. In spring, new canes grow from plant's crown.

Effects of Snowcover

The importance of snowcover as insulation is demonstrated by comparing winter injury observations for the winter of 1989-90 to those of 1988-89 and 1990-91. The minimum air temperatures at the Arboretum and average snowcover ⁹ on a weekly basis for the winters during which injury was recorded can be seen in the figures below. Snowcover was minimal or absent during 1989-90. Many of the cultivars which typically exhibit the "snowline" pattern of hardiness died back to the ground in the absence of snowcover during that winter.

This pattern of dieback to the ground was very common among the Old Garden Roses in the winter of 1989-90. Because these roses bloom on previous years' wood, there was no bloom during spring of 1990. Cultivars of other rose classes (Floribundas, Hybrid Musks, Hybrid Perpetuals, Hybrid Rugosas, Hybrid Suffultas, Kordesiis) also died to the ground. Because these plants bloom on current year's canes as well as on previous years' wood, their flowering was less affected during the spring of 1990.

Winter Injury During Acclimation

A reduced level or a complete absence of winter injury was also observed on roses after the winter of 1988-89, in comparison to the injury sustained during the following two winters. This was seen on many cultivars among the Old Garden Roses (Albas, Bourbons, Centifolias, Gallicas, Hybrid Perpetuals, Mosses) as well as some of the Hybrid Caninas, Hybrid Moyesiis, Hybrid Spinosissimas, and Kordesiis.

If differences in minimum winter temperatures are the only comparisons made to explain annual differences in injury levels, the reduced injury level after the winter of 1988-89 seems unusual. The smallest amount of injury was observed after the coldest winter (-28° F, -33° C), while more severe injury was observed after the warmer winters of 1989-90 (-26° F, -32° C) and 1990-91 (-24° F, -31° C).

Minimum air temperatures (line graph) at the Minnesota Landscape Arboretum and the average snowcover (bar graph) on a weekly basis, for the winters during which injury was recorded-1988-89 (top), 1989-90 (middle), and 1990-91 (bottom).

Temperature comparisons of late fall and early winter over the three years offer a more feasible explanation for the differences in winter injury patterns. Temperature declines were very gradual during late 1988. The low temperature for that winter did not occur until February 3, 1989 and the lowest temperature prior to that was -18° F (-28° C) on January 9. This gradual temperature decline may have allowed plants to escape injury during the acclimation phase.

In contrast, the next two years had faster, more severe temperature declines during late fall and early winter. The minimum temperatures during the winters of 1989-90 and 1990-91 occurred in mid- to late December rather than in early February. The higher levels of injury observed for these two years may have occurred because plants acclimated too slowly for the rapid temperature declines in early winter. When early temperature declines outpaced increases in plant hardiness, canes were injured.

Quantifying Winter Hardiness in the Laboratory

Attempts were made as part of this research to quantify mid-winter hardiness levels for 85 hardy rose cultivars and species using an ultralow-temperature freezer in a laboratory. Cane samples from the current year's growth were harvested, placed in plastic bags, and held outdoors at ambient temperature until processed. Because of the insulating property of snow, only canes above the snowline were sampled. All the sampling occurred between January 11 and February 4 during 1990, 1991, 1993, or 1994. Cane section shows discoloration caused by freezing injury.

Laboratory processing involved cutting the harvested canes into 1½-inch (4 cm) lengths after discarding the terminal 4 inches (10 cm) of the shoot. Ten polyethylene bags were prepared containing four cuttings each of 10 to 12 cultivars in contact with moist paper towels, which promoted ice formation at freezing temperatures. A copper-constantan thermocouple was inserted into the center of one stem section in each bag monitor stem temperature. The bags were closed and placed in the ultralow-temperature freezer. During this processing, no cane material was exposed to room temperature for longer than five minutes.

The samples were held overnight in the freezer at a temperature approximating the previous night's outdoor minimum temperature. An eleventh bag of cuttings was held overnight under refrigeration at 36° F (2° C) and served as a control.

The following day, the temperature in the freezer was dropped at a rate of 10° F (5.5° C) per hour. This temperature drop was monitored through the thermocouples, which connected samples in the freezer to temperature-recording equipment outside the freezer. A temperature range was selected to bracket the estimated temperature at which cane injury would occur. As the temperature in the freezer dropped through this selected range, one bag was removed at 3° or 5° F (2° or 3°C) intervals.

The samples removed from the freezer were thawed in a refrigerator at 36° F (2° C) for 24 hours, followed by incubation at room temperature (72 ± 4° F; 22 ± 2° C) for 7 to 10 days. Stem sections were then cut lengthwise and evaluated visually for injury using a dissecting microscope. Brown discoloration of the xylem or cambium was considered fatal (see photo). The lowest sampling temperature at which 50 percent of the stem sections were uninjured was interpreted as the mid-winter hardiness level for each cultivar or species.

Hardiness Determinations in Containers

Mid-winter hardiness determinations are made on cane tissue above the snowline on field-grown plants. However, many of the roses in the Arboretum's Shrub Rose Garden die back to the snowline by mid-winter each year, making mid-winter hardiness determinations of these plants impossible.

To circumvent the cane dieback problem, container-grown plants of some of these cultivars were placed in an overwintering greenhouse. The air temperature in the house was regulated so that roses could reach their mid-winter hardiness levels without being exposed to temperatures normally lethal to canes.

The container-grown roses were moved into the greenhouse during the first week of November after hardening under natural outdoor conditions. Inside the greenhouse, containers were completely covered with wood chips to a depth of 4 inches (10 cm) above the top rim to prevent cold injury to roots. The greenhouse was covered with an inflated double layer of white polyethylene. Interior air temperature was regulated with thermostatically controlled electric heaters and ventilation fans. The greenhouse was ventilated during the day to keep the temperatures inside similar to what was outside. Minimum nighttime temperatures were maintained at or above 20° F (-6° C) from November 5 through December 14, and at or above -4° F (-20° C) for the remainder of the winter. Mid-winter hardiness levels were then measured in January and February using the ultralow freezer in the laboratory.

Mid-winter hardiness levels, when determined, are listed in Table 4, along with the year that cultivars or species were sampled. Table 5 lists winter injury observations from the winter of 1992-93 along with 1993 determinations of mid-winter hardiness levels for recently planted Species Roses that were not included in the main evaluation study.

Winter Hardiness Comparisons of Field and Container-Grown Roses

Comparisons of mid-winter hardiness levels of container-grown plants in the overwintering greenhouse and field-grown plants for each of five cultivars and one species were attempted. Canes were sampled in January of 1991. As expected from previous years' observations, the field-grown plant of R. primula was uninjured while field-grown plants of each cultivar were already injured at the time of sampling. The canes of the container-grown R. primula were found to be uninjured, and their hardiness level was identical to that of the canes of the field-grown plant. Canes of each of the five container-grown cultivars were also uninjured. Hardiness levels among the container-grown cultivars ranged from -30° F (-34° C) to -34° F (-37° C). This indicated that all five cultivars possessed adequate mid-winter hardiness to survive the low temperature for the winter of 1990-91 (-24° F, -31° C), which had already occurred on December 26, 1990. The injury to canes of the field-grown cultivars probably occurred when plants acclimated too slowly for the early temperature declines of 1990.

Table 6. Comparison of mid-winter hardiness levels between field-grown and container-grown plants of six selected rose taxa. Field injury descriptions are defined above. The designation "(I)" for the Fahrenheit (°F) and Celsius (°C) columns under "field plant" indicates that the canes were injured in the field prior to collection.
Cultivar	Class	Field Injury	Hardiness Rating
			Field Plant	Container Plant
Mme. Legras de St. Germain	Alba	snow	(I)	-32°F (-36°C)
Léda	Damask	snow	(I)	-30°F (-34°C)
Mme. Hardy	Damask	<½	(I)	-32°F (-36°C)
Sir Thomas Lipton	Hybrid Rugosa	<½	(I)	-34°F (-37°C)
Salet	Moss	snow	(I)	-30°F (-34°C)
R. primula	Species	none	-33°F (-37°C)	-33°F (-37°C)

Hardiest Selections

When selecting roses for gardens in cold climates, some winter injury should be expected. Three years of winter injury observations show that even among the hardiest of roses, the number of cultivars or species at the Minnesota Landscape Arboretum which survive winters with little or no cane injury is very limited. Many of these are cultivars ('Alika', 'Betty Bland', 'Harison's Yellow', 'Aylsham', 'Metis', 'Suzanne', 'A. MacKenzie', 'Haidee', 'J. P. Connell', 'Prairie Wren', and 'Prairie Youth') or species that have only one season of bloom or very slight levels of rebloom.

The Hybrid Musk 'Ballerina' suffers extensive cane injury every winter but comes back vigorously to bloom profusely. The single flowers are borne in large panicles.	R. setigera, the Prairie Rose, blooms later than other roses, displaying its deep pink flowers in mid-July. Because it produces long winter-hardy canes, the Prairie Rose can be used as a climbing rose.	A portion of the Gallica section of the Shrub Rose Garden at the Minnesota Landscape Arboretum.

As a group, the Hybrid Rugosa cultivars best combine cane hardiness and moderate or heavy levels of repeat bloom, although winter injury among the Hybrid Rugosas at the Minnesota Landscape Arboretum was more severe than expected. During the evaluation period, the Hybrid Rugosas suffered severe insect infestations of a cynipid wasp gall (Diplolepsis spinosa) and the rose stem borer (Agrilus aurichalceus), both of which cause galls that stress plants. Reduced hardiness levels may have resulted from these stresses.

Three Kordesii cultivars, 'Henry Kelsey', 'John Davis', and 'William Baffin', also combine cane hardiness with the ability to repeat bloom.

Among other cultivars with moderate to high levels of rebloom, there are many that suffer winter injury, often to the snowline or to the base of the plant, and then regrow vigorously in spring. Because these cultivars bloom on the current year's growth as well as on previous years' wood, they are still attractive flowering plants in the landscape. Plant size is affected rather than flowering ability. Some of these cultivars are 'Chuckles', 'Dapple Dawn', 'Eutin', 'Nearly Wild', 'Redcoat', 'Earth Song', 'Ballerina', 'Cuthbert Grant', 'John Cabot', 'Karlsruhe', 'Parkdirektor Riggers', 'Adelaide Hoodless', 'Amiga Mia', 'Applejack', 'Carefree Beauty', 'Champlain', 'Country Dancer', 'Golden Wings', 'John Franklin', 'Morden Amorette', 'Morden Centennial', 'Morden Ruby', 'Prairie Dawn', 'Prairie Princess', 'Robusta', and 'Summer Wind'.

Severe winter injury is more detrimental to the performance of Old Garden Rose and Shrub Rose cultivars that bloom once in spring on previous years' wood. If winter dieback occurs to the base of the plant, these cultivars will not flower in the spring.

Dieback to the snowline is also detrimental to the largest of these one-time-blooming cultivars since the flowers that occur on the previous years' wood are often "buried" beneath the vigorous growth of new canes in spring. The floral display is better on those large cultivars whose winter injury pattern varies from cane to cane within the crown (<½ and >½ in Table 4). Flowers on the canes that suffered little dieback are still beautifully displayed as the surrounding canes that were injured regrow vigorously from below.

The floral display on the shortest of the one-time-blooming cultivars, such as the Gallicas, is not as severely affected as that of the larger cultivars. During winters with constant snowcover, a smaller proportion of their crowns are lost to winter injury, allowing mature heights to be regained in the following growing season without burying the floral display.

6	A hardiness zone map covering the United States and much of Canada is included on page 76, in the "Cultivation Tips" chapter.
7	L.A. Minsk. Acclimation of Selected Rosa Taxa. Unpublished Thesis. University of Minnesota, St. Paul. 1993.
8	Adapted from Minsk, Acclimation of Selected Rosa Taxa.
9	Climatological Data, Minnesota. National Oceanic and Atmospheric Administration, Environmental Data and Information Service, National Climate Center. Asheville, North Carolina.

---

« Previous Section | Table of Contents | Next Section »

---

Produced by Communication and Educational Technology Services, University of Minnesota Extension.

In accordance with the Americans with Disabilities Act, this material is available in alternative formats upon request. Please contact your University of Minnesota Extension office or the Extension Store at (800) 876-8636.