David Robinson / Other articles and papers / Scientific Papers and Lectures / Cold Hardiness

Previous / Next: Juania australis / Plant Hardiness

Cold hardiness and Climatic Zone 9 plants

[Paper prepared for the International Magnolia Society 2001 Annual Meeting held in Dublin, Ireland. © Karen Foley 2019

See also the additional notes made for this conference on plant hardiness.]

Although my garden at Earlscliffe, Baily, Co Dublin is situated at 53º north latitude it has a favourable microclimate and contains many climatic zone 9 plants. The harsh 2000/2001 winter, the most severe since I started taking records in 1969, provided a good opportunity to record the effect of low temperatures on marginally hardy plants.

The 2000/01 winter

Autumn 2000 was mild and wet with temperatures well above average in early December. However, on the night of December 26/27, 2000 the air temperature fell to - 6º and to - 4º, - 7º, - 5.5º and - 4.5º C on the following four nights. For the three days between the 28th and 30th the temperature never rose above  - 3º or - 4º C during the day. At the end of the cold spell (December 30), the temperature rose 7.5 C º during the night and was +3º by morning, so that the speed of thawing was rapid.

After a relatively mild January, another cold spell began on February 23/24, 2001 and the minimum air temperatures recorded on this and the following 9 days were - 2º, - 2º, - 1.5º, - 2.5º - 2.5º, - 5.5º,   - 5º, - 5º, - 5º and - 1º C. A severe northerly gale blew up on the night of February 25/26 and the snow that fell that night persisted in shaded places for a full week. For an area that had seen little snow in the last 30 years, these conditions were exceptional.

Apart from temperature records, the response of many plants confirms the severity of the winter. Young plants of the banana Musa basjoo were killed to ground level, although they had survived all previous winters with stems intact since this species was first planted in 1973. Erica canaliculata the most vigorous and one of the most hardy of the South African heaths growing at Earlscliffe, survives most winters unscathed or, in a cold winter, a few cm of unripened shoot tips may die back. The 2000/01 winter killed back previous year's growth on some plants by over 50 cm.  

As one of my interests is growing trees and shrubs that are close to their climatic limit, it is inevitable that a cold winter will kill many of the more tender plants. I do not regard plants killed in this way a tragic loss but rather as an educational experience, an opportunity to gather fresh information and a chance to plant up again with more frost tender plants.

To obtain an accurate picture of plant hardiness, no fleece or other artificial means of protecting plants is used no matter how severe the frost and no plants are overwintered under cover.

Factors influencing plant hardiness

Plant hardiness is mainly influenced by the absolute minimum temperature but this is only one of many factors involved. Others include the duration of freezing temperatures, the speed of drop, the hardening up process, the speed of thaw, the moisture content of the plant, soil and air, and the soil type.

Healthy, well grown plants are better able to cope with cold than plants in a debilitated condition. A correlation between the development of frost hardiness and an increase in sugar content has been demonstrated for a large number but not for all plants. The exact role of sugar is still debated but it has been suggested that high sugar decreases the freezing point by accumulating in the vacuoles and decreasing the amount of ice formed. Others believe that both sugars and amino acids can protect specific sensitive proteins and enzymes from the effect of freezing.

Because of the link between plant well being and tolerance of low temperature, a wide range of cultural factors are relevant, including nutrition, freedom from pests and diseases and general husbandry. The morphology of the plants also plays a part with some plants having especially susceptible stem, buds or roots. In addition, other climatic factors such as wind chill and depth of snow cover may play a negative or positive role.

Previously it has been suggested that high fertility predisposes plants to low temperature injury. With food crops this has been shown to be false. Plants fertilised for optimum growth or yield are not unduly susceptible to low temperatures; in contrast nutrient deficient plants and those overly fertilised so that yield is depressed are susceptible. It seems likely that these results are also linked to sugar levels within the plant. 

Hardening up process

With the approach of winter, important differences are evident between many climatic zone 8 plants and those in zone 9. Zone 8 plants undergo a hardening up process under the influence of photoperiod and falling temperatures. With temperatures in the range of 1 and 5 º C plant growth gradually ceases, metabolic changes take place in plant proteins, soft growing tips lose moisture and become more lignified. In contrast many zone 9 plants harden up only slowly or not at all and many, including a large number of Eucalyptus species continue to grow slowly during a mild winter. Consequently they are susceptible to damage by sub-zero temperatures.

Low temperatures can kill plants in different ways. In highly susceptible plants, the contents of the cell may freeze and death occurs rapidly due to disruption of protoplasm. In other plants, however, ice crystals are first formed outside the cells in the walls and intercellular space as a result of the antifreeze effect of the cell contents. Further cooling leads to the growth of these crystals with the water being drawn from the cells. When this occurs the plant may die slowly as a result of desiccation and this is now believed to be a more common cause of plant death than cell disruption by ice crystals (Levitt, 1972).

Contrary to previous widely held beliefs, it is now thought that the formation of ice in the intercellular spaces of itself is not responsible for plant death (Kramer and Kozlowski, 1979). The view that plants tolerate low temperatures because the freezing point of protoplasm is lowered as water moved out of the cell has also now been largely discredited.

When plants are killed immediately due to intracellular freezing, e.g. many succulents such as Aeonium tabuliforme, the plant collapses quickly as mushy pulp. Within a few weeks the remains of many non-lignified plants killed in this way have virtually disappeared.

Plants that are killed by cell desiccation, e.g. many plants with small to medium sized, firm leaves such as Metrosideros excelsa, Psoralea pinnata and Eucalyptus ficifolia, do not show the effects of low temperature injury for many weeks. Their leaves gradually wilt and turn brown and the plant shows typical symptoms of drought. At sub lethal temperatures, some plants with large fleshy leaves, e.g. the Giant Borage, Echium pininana, wilt repeatedly when the temperature drops at night below approximately - 3º C and regain turgidity when the temperature rises again during the day, presumably as water moves in and out of cells into intercellular space. With experience it is possible to gauge approximately the temperature between -2 º C and - 8 º C by the extent of the flaccidity of the Echium leaves. 

Because so many plant processes are involved when a plant is severely damaged by frost and many different parts of the plant may be affected, it is not possible to be certain whether a particular plant will live or die under specific temperature conditions. Evergreen plants that are damaged by low temperatures may absciss their leaves cleanly or, in cases of more severe injury, may hold onto the dead foliage for many months. In general, plants that absciss their damaged leaves appear to have a better chance of survival than those that hold onto them in a dead condition.

Leaf abscission is controlled by plant hormones and it is generally believed that auxins prevent abscission as long as the leaf blade is healthy and growing (Leopold and Kriedemann, 1975). In the case of deciduous species in the autumn and evergreen plants with damaged leaves, auxin flow diminishes and the plants produce ethylene, which triggers leaf fall.

Where death or severe injury occurs suddenly, the plant's hormonal control mechanism is destroyed and so foliage dies but does not fall. In these cases the chances of recovery by resprouting from the above ground part of the plant are very remote. However many plants affected in this way can resprout strongly from soil level. The likelihood of regrowth from adventitious buds on stems and branches appears to be greater where leaf abscission has occurred.  

Effects of the 2000/2001 winter on Zone 9 plants at Earlscliffe

Zone 9 plants at Earlscliffe reacted in many different ways to the 2000/01 winter. Many fleshy succulents, that contain much water, were especially susceptible to freezing conditions and collapsed quickly during the cold spell in December. Plants included in this group were:- Aeonium tabuliforme, A. arboreum 'Atropurpureum' and many other Aeonium species. Aeonium balsamiferum, although damaged, appeared to be the most hardy of the species tested and many plants survived. Another Canary Island succulent, Greenovia aurea, was more tolerant of freezing conditions than Aeonium spp, although its response varied from no discernible effect to severe blackening of foliage and death in different parts of the garden. Solanum laciniatum, the soft subshrub Solanum quitoense (z 10) and young plant of Musa sikkimensis and M. basjoo were killed back to ground level.

Not all plants with a high moisture content were sensitive to low temperature damage. A number of plants with fleshy leaves, including some Agave species, Aloe striatula, A.aristata, and Beschorneria yuccoides, all survived completely unharmed. Among the Agaves, A parryi, A.ferox and A. americana were uninjured but all the outside leaves of A sisalana were desiccated, although the youngest leaves in the heart of the plant remained green. Furcraea longaeva, another succulent related to Agave, reacted differently; the older leaves were undamaged but some of the younger leaves turned temporarily chlorotic with a few dead patches but the plant quickly recovered when temperatures rose in the spring.

The foliage of a 4 metre high plant of Brugmansia sanguinea collapsed immediately the first severe frost arrived. However, although the entire top growth was killed, much of the root system remained alive during the winter and, as after previous hard winters, the plant began to resprout strongly from the base in mid-June 2001. In contrast, plants that were slow to die but are unlikely to recover include the blue flowering South African tree Psoralea pinnata,  Corynocarpus laevegata from New Zealand .

Frost hardiness and sugar content

The link between frost hardiness and sugar content was demonstrated by the variable response of seedlings of Echium pininana, the Giant Borage from the Canary Islands, naturalised in a Betula woodland. This plant is hardy in a normal winter but in 2000/01 many plants were killed in the middle the woodland but survived in the woodland fringe. It seems likely that sugar levels were low in plants in the shade inside the wood and plants died even though the canopy of branches would have given extra protection from frost. The higher sugar levels of the plants grown in the open and in the woodland fringe probably accounts for their greater tolerance.

Differences within genera

As a result of their evolutionary history, big differences often occur in the susceptibility of species within a genus. Euryops virgineus started flowering in December and continued until June at least without showing any signs of damage. Some plants of E. pectinatus suffered moderate leaf scorching and dieback but survived while E. chrysanthemoides died quickly. The roots of this plant appeared dead in January and recovery was not expected. Metrosideros robusta was not affected in any way by low temperatures but plants of M. excelsa were killed back to ground level. Among the tree ferns, Dicksonia antarctica, D. squarosa and Cyathea dealbata showed only slight signs of damage but C. cooperi was killed.   

As would be expected from a genus covering a vast geographical area, different species of Eucalyptus responded differently to the low temperatures. All Eucalyptus ficifolia saplings raised from seed collected from the wild in Western Australia in January 1996, were killed back to ground level. Several hundred plants survived the mild winters between 1996/97 and 1999/2000 in good condition but were unable to withstand temperatures of -6º and -7º after Christmas 2000. About 6% of these saplings showed some regrowth from their base in June 2001 although their survival is far from assured.  A few plants raised from seed collected from higher elevations, have been severely damaged but appear to be somewhat less susceptible to low temperature injury than the progeny of seed collected in the plant's areas of natural distribution. E. curtisii ffrom the Brisbane area of Queensland was also severely scorched and is unlikely to live. Approximately 70 other species of Eucalyptus, mainly from New South Wales, Victoria and Tasmania, survived in good condition. These include Eucalyptus crenulata from Victoria which continued to grow throughout the winter without any significant injury to the expanding young foliage. 

Leaf abscission and plant recovery.

With few exceptions, no regrowth from branches occurred on evergreen plants that held onto their dead foliage after the cold spells in December and February. Plants in this group included:- Casuarina equisetifolia, Erica x hiemalis, Erica versicolor, Agathosma ciliata, Cineraria 'Purple picotee' and Psoralea pinnata. However, many plants which had their top growth killed, resprouted strongly from the base such as Pelargonium papillionaceum and Sonchus arboreus. Nerium oleander was an exception insofar as epicormic growth occurred in June although most of the dead foliage was still attached to the plant.

Other normally evergreen plants which abscissed all their foliage cleanly after low temperature injury included Itoa orientalis, Tetrapanax papyfera and Grevillea robusta.. Plants in this group have produced leaves from adventitious buds on branches although those of Grevillea robusta are still weak in late June.

Root system

The location of roots in the soil is important both with regard to plant health and also freezing injury. Examination of the root system of many plants growing at Earlscliffe shows that a large proportion have most of their roots close to the soil surface. This is due to the absence of severe drought, relatively shallow soil and the routine use of herbicides to control weeds so that the soil surface is not disturbed.

Stem tissues are usually less sensitive to low temperature injury than root tissue, but roots of trees and shrubs are injured less in winter because the soil (and snow) cover protects them to some extent from exposure to freezing temperatures. In theory plants with their roots close to the soil surface, should be especially susceptible but there was no evidence of this under conditions at Earlscliffe. Nevertheless when soil freezes, roots are often damaged, especially the small physiologically important ones.

A large number of roots were examined on February 18, 2001 and as far as could be ascertained root damage on many plants was minimal compared with the severe injury to the above ground portion. Despite extensive damage to foliage, stems and branches, the roots of many plants appeared to be turgid, moist and fresh. Plants in this category which eventually recovered included:- Brugmansia sanquineum, Tetrapanax papyfera, Sonchus arboreus, Strobilanthes pentstemonoides, Metrosideros excelsus and a small number of saplings of Eucalyptus ficifolia. Plants with apparently healthy roots in February which failed to recover included:-, Cyathea cooperi, Psoralea pinnata and many saplings of Eucalyptus ficifolia. With some plants it was clear that both top and root were severely damaged e.g. Solanum quitoense, Europys chrysanthemoides and Aeonium spp and that recovery was unlikely.

It is clear from the way so many plants resprout strongly from the base, six months or more after the entire top growth has been killed that, in many plants, the root system is less susceptible to low temperature injury than the above ground part of the plant.

The fact that a large part of the root system of most plants at Earlscliffe is close to the soil surface may be an overall advantage rather than a disadvantage. Roots growing in the surface soil layers seem to be especially important to plants, due partly to the fact that this region is richer in nutrients and oxygen. It has also been shown that young roots are the main production sites in plants for some essential growth hormones such as gibberellins and cytokinins (Scott Russell, 1978).  Gibberellins promote stem elongation by stimulating cell division. Cytokinins stimulate cell enlargement as well as cell division and so promote leaf expansion. On balance, the horticultural advantages of allowing roots to exploit the soil surface seem to outweigh any possible disadvantage due to increased risk of low temperature injury. 


Although large numbers of plants have been killed or severely injured, even larger numbers of other zone 9 plants have apparently survived unharmed or with slight damage only. All palm trees appear to have come through the winter well. These include Trachycarpus fortunei, Chamaerops humilis,, Phoenix canariensis, Jubaea chilensis and young plants of Washingtonia robusta, W. filifera, Brahea armata and Butia capitata. The rare palm from the Juan Fernandez Islands Juania australis showed no signs of low temperature injury and continues to grow strongly. 

Many tender conifers have also survived in good condition, such as the Rimu, Dacrydium cupressinum, D. franklinii, Callitris rhomboidea and C. oblonga. Other non coniferous survivors include:- Pseudopanax arboreus, A. ferox and A. laetus, Telopea speciosissima, Bromeliads including Ochagavea carnea and Fascicularia bicolor and many species of Puya. The aromatic tree Doryophora sassifras is also in good condition. This plant has assumed a special interest in recent years as, in its native New South Wales, it grows alongside the 'fossil tree' the Wollemi pine, Wollemia nobilis,   

Protea lacticolor showed slight scorching only of unripened shoot tips and the King Protea, P. cynaroides has also survived although with some leaf damage Banksia spinulosa flowered in November 2000 and was little affected by winter cold but the combination of a wet blanket of snow and gale force winds on the night of February 26/27, split off several main branches. The grass tree, Xanthorrhoea preissii appeared to be in as good condition in spring as it was when it entered the winter. It was planted in August 1995 and has been slow to recover from transplanting shock.

Increase in hardiness of Echium pininana by natural selection

Although Echium pininana, the Giant Borage is an endangered species in its native Canary Islands, it self seeds freely in many coastal areas of Ireland. Plants can grow from over 1 m to a height of 6 m between February and early June. The plant flowers in its second or third year and, being monocarpic, dies after flowering. E. pininana is cross pollinated and produces seeds prolifically (in excess of 200,000/plant), which germinate readily and dense carpets of Echium seedlings are commonplace. Plants developing from these seedlings survive most winters but, in hard winters many are killed by cold. On two occasions (1978/79 and 1986/87) during the last 24 years all plants were killed by low temperatures of -6º C in January. After these severe winters new seedlings develop from the seed bank in the soil.  

It seems possible that natural selection from cold stress is producing a more cold hardy strain of E. pininana, as in two recent cold winters (1995/6 and 2000/01) when temperatures fell to -7 º C, many seedling survived.


It is obvious that a large number of climatic, physiological, genetic and cultural factors are involved in plant tolerance to low temperature.  In addition, marked differences in microclimate can occur in a small area, even on a garden scale. Consequently, it would appear that the only way to find out if a zone 9 plant is hardy in a warm temperate areas is to plant it and see what happens.

The response of some tender plants to the severe 2000/2001 winter.



Severely injured

Likely to recover

Injured. Rapid recovery likely Uninjured
Erica x hiemalis Musa sikimensis Tetrapanax papifer Metrosideros robustus
Vireya rhododendron 'Tuba' Sonchus arboreus Dodonea viscosa 'Purpurea' Dacrydium cupressinum
Eucalptus ficifolia (a few may recover) Metrosideros excelsus    
Many Echium wildpretii Brugmansia sanguinea Euryops pectinatus Juania australis
Solanum quitoense Tetraclinus articulatus Roldana petasites Jubaea chilensis
Solanum laciniatum Itoa orientalis Cyathea dealbata  
Corynocarpum laevigata Grevillea robusta Correa backhouseana Euryops virgineus
Psoralea pinnata Protea cynaroides Protea lacticolor Pseudopanax ferox
Psoralea affine Agave sisalana   Taiwania crytomeriodes
Euryops chrysanthemoides     Metapanax davidii
Casuarina equisetifolia     Dosryophora sassifras
Casuarina equisetifolia     Cordyline indivisa
Cyathea cooperi     Dicksonia antarctica
Leucodendron argentea     Beschorneria yuccoides
Aloe tabuliforme     Elegia capensis
Agathosma ciliata     Agave ferox
      Agave americana
      Agave parryi
      Aloe aristata
      Aloe striatula
      Many Puya spp

Further reading

  • Kramer, P.J. and Kozlowski, T.T. (1979). Physiology of woody plants. Academic Press, New York
  • Leopold, A.C. and Kriedemann, P.E. (1975). 'Plant growth and development.' Mcgraw-Hill, New York
  • Pellett, H.M., (1981). Effect of Nutritional factors on cold hardiness of plants. Horticultural Reviews. AVI Publishing Company Inc. 3, 144 -171.
  • Scott Russell, R. (1978). Root growth and function. 1978 Master's Memorial Lecture. Journal of the Royal Horticultural Society. 325 - 331.

[See also the additional notes made for this conference on plant hardiness.]

Back to top

Other pages related to David Robinson

Articles for the Irish Garden 

This page was last updated on 27-Jul-2023 .