Effects of nitrogen fertilization and temperature on frost hardiness of Aleppo pine (Pinus halepensis Mill.) seedlings assessed by chlorophyll fluorescence

In a 14-week study, 1-year-old Aleppo pine seedlings were grownin two growth chambers. Seedlings were artificially hardenedby decreasing photoperiod and temperature. In each chamber halfof the seedlings were fertilized with nitrogen (8.4 mg seedling^–1).In order to determine the relative importance of the hardeningenvironment versus fertilization, each chamber was programmedto decrease night temperatures down to a low of 8 or 4°C.Chlorophyll fluorescence and frost hardiness was measured fivetimes during the experiment. A sample of seedlings from eachtreatment was exposed to an artificial frost at –5°Cand the freezing effects were assessed by measurements of chlorophyllfluorescence and visual evaluation of needle damage. Seedlingsincreased their frost hardiness during the experiment in allthe treatments but the ratio of variable to maximal chlorophyllfluorescence (F_v/F_m) measured before freezing did not vary duringthe experiment. This indicates that Aleppo pine maintains itsphotosynthetic ability during hardening in contrast to otherconiferous species from colder climates. The effect of nitrogenfertilization on frost hardiness was small in comparison withchamber effect. Nitrogen fertilization slightly delayed theacquisition of hardening in the coldest chamber. Seedlings inthe warmest chamber did not become fully resistant to –5°C,but in the coldest chamber, where night temperature reached4°C, all the seedlings were resistant to the frost. Severedamage caused by frost could be related to a rapid rise of minimalfluorescence (F₀) but the best index of damage was the dropof F_v/F_m after freezing.     

Frost hardiness gradients in shoots and roots of picea mariana seedlings

Frost hardiness of tissues along the length of the stem and the root was investigated in first‐year black spruce (Picea mariana (Mill.) B.S.P.) seedlings. Frost hardiness of 1 cm long stem and root segments was evaluated based on Index of Injury, calculated from post‐freezing electrolyte leakage. Frost hardiness was tested approximately weekly beginning seven weeks after seedlings were transferred from an 18 to a 10 h photoperiod, both at day/night temperatures of 26°C/16°C. Trees were transferred to temperatures of 10°C day and 5°C night at a 10 h photoperiod after a further 18 days. Frost hardiness was greater at the terminal bud and least at the root tips. Although shoots were generally more frost hardy than roots, differences in hardiness along the stem and root axes were gradual, rather than abruptly differing at the shoot‐root interface. All tissues, including root tips, increased in frost hardiness after conditioning for 18 days under short photoperiods (10 h) and warm temperatures (26?C/16°C, day/night). Under cold temperatures (10°C/5°C, day/night) all tissues, excepting the root tips, tolerated — 16°C with little subsequent electrolyte leakage.     

Seasonal Changes in the Frost Hardiness of Provenances of Picea sitchensis in Scotland

Changes in the natural level of frost hardiness of shoots offour provenances of Picea sitchensis were monitored over twogrowing seasons by detaching shoots from 7 to 10-year-old treesgrowing in a nursery in Scotland, and subjecting them to freezingtemperatures under conditions which simulated night frosts. Six seasonal phases of frost hardiness were identified (Fig.3).

1. During each autumn, killing temperatures (the level of hardiness)decreased from –5°C to below –20°C, beginningseveral weeks after shoot elongation ceased. Alaskan provenanceshardened in September, apparently in response to shorteningday lengths alone, whereas an Oregon provenance did not hardenuntil November, after repeated frosts. Queen Charlotte Islandsprovenances were intermediate.
2. From November to March allprovenances were hardy to below –20°C,which is adequateto prevent direct freezing injury at mostplantation sites.
3. In March-April, several weeks before bud-burst, old shootsdehardenedto killing temperatures of about –10°Cin responseto warm temperatures, and southerly provenancesdid so beforenortherly ones.
4. During bud-burst the newly-emergingshoots were hardy to only–3°C to –5°C untilthey were about 3.5 cmlong. All provenances burst bud at thesame time and were equallyfrost susceptible at this time.
5. DuringMay-July the elongating shoots fluctuated in hardinessbetween–5°C and –10°C apparently in responsetofluctuating ambient temperatures.
6. In August 1980 there wasa period of late summer dehardeningto killing temperaturesof about –3°C.

Seasonal changes in hardiness are discussed in relation to changesin shoot growth and environmental factors. The main opportunitiesfor selecting frost hardy genotypes seem to be in the rate ofautumn hardening, the time of pre-bud burst dehardening, andthe time of bud-burst.     

Frost Hardiness of Subalpine Eucalypts in Britain

Laboratory freezing tests were used to determine seasonal changesin the frost hardiness of detached shoots of young trees ofEucalyptus gunnii (from central Tasmania), and E. niphophilaand E. debeuzevillei (‘snow gums’ from the SnowyMountains, Australian Capital Territory). The trees were growingat the Bush Estate. No difference was found between the speciesor between seedlots, all of which were from high altitudes nearthe tree line. In midwinter (February) the shoots tolerated –16°Cwithout suffering damage, and many shoots survived temperaturesas low as –18°C to –22°C. This result agreedwith Evans' (1986) observation that some trees within theseseedlots survived temperatures in the range –19°Cto –23°C in field planting during the winter 1981/82.In their native habitats the trees rarely experience temperaturesbelow –20°C. The shoots did not harden appreciably before they experiencedfrosts (in late October/early November) and so may be proneto autumn frost damage. By contrast, they were slow to dehardenin late winter and spring and did not seem prone to spring frostdamage. There were no killing air frosts during the period of this study(winter 1985/6), but many trees died, possibly as a result ofground freezing, producing root injury and/or shoot desiccation.In subalpine regions of Tasmania and the Snowy Mountains theground is covered by snow throughout the winter. Variation in frost hardiness within these hardy seedlots couldbe exploited.     

Morphological and physiological differences in Scots pine seedlings of six seed origins

One-year-old seedlings of Scots pine (Pinus sylvestris L.) offour native seed origins (Loch Maree Islands, Glengarry/GlenMorriston, Glen Affric and Abernethy), a commercial Britishseedlot, and a seedlot from Hedesunda, in middle Sweden, werecompared at monthly intervals from October 1993 to April 1994.Seedling morphology, root condition, root frost hardiness andbud dry matter were determined at each date. There were clear morphological differences among seed origins.Seedlings raised from the commercial seedlot (A70) were largerbut had a poorer root:shoot ratio than the other seed origins.Of the native pines tested, the Loch Maree Islands origin allocateda larger proportion of its photosynthate to fine roots and needlesand smaller proportion to woody structures. Seedlings raised from the commercial British seedlot tendedto have poorer bud lignification than the other origins andalso, in autumn, higher electrolyte leakage rates from its fineroots. During winter, the Swedish origin had the lowest fineroot electrolyte leakage. Seedlings of all origins showed aprogressive increase in fine root hardiness towards mid-winterwith maximum hardiness (–7°C) in January. Dehardeningoccurred over subsequent months reaching –3°C in April.Differences among origins were evident. The Swedish seedlotdeveloped greater frost resistance than the other origins, hardeningbegan earlier in autumn and dehardening began later in spring.The commercial seedlot hardened later than the other originsbut reached a similar level of frost hardiness by January. Ofthe native pines, seedlings of the Loch Maree Islands originwere slowest to develop root hardiness.     

Autumn Frost Damage on Young Picea sitchensis 1. Occurrence of Autumn Frosts in Scotland compared with Western North America

Dates of first autumn air frosts (in Stevenson screens) of –2.5°Cand 4.5°C were obtained for 42 meteorological stations innorthern Britain with runs of 18 to 116 years. Frequency distributionsof first frost dates were approximately normal. Altitude, distancefrom the sea, latitude and distance from the NWSE axis of Britaintogether accounted for 75% and 81% of the variation in datesof first –2.5°C and –4.5°C frosts, respectively,at the 42 stations. The variance in dates of first frosts decreasedfrom lowland coastal to upland inland sites. Multiple regressionswere used to produce maps of first frost dates in 20 x 20 kmgrid squares of Scotland. Dates of first 28°F (–2.2°C) and 24°F (–4.4°C)frosts were obtained for 20 sites in western North America,spanning the natural range of P. sitchensis. Mean dates wereearlier, and variances decreased, from south to north. Upland sites like Eskdalemuir and Kielder Castle experience–2.5°C and –4.5°C frosts earlier in theautumn than all coastal stations in western North America southof about latitude 58°N (between Cordova and Sitka), andabout 4 weeks earlier than at Masset on the Queen CharlotteIslands. Thus, P. sitchensis from Q.C.I. and further south mayoften experience autumn frosts in Scotland before the treeshave experienced the cool/short days that they require to inducefrost hardening.     

Shading Reduces both Visible and Invisible Frost Damage to Norway Spruce Seedlings in the Field

Two-year-old cuttings of Norway spruce were subjected to nightfrosts in spring on an exposed site in southern Sweden. Shadingwas used to assess the influence of sunlight on the extent ofdamage resulting from night frost. Chlorophyll fluorescencewas measured in needles in flushing shoots, and in shoots atthe stage of bud burst. The F_v:F_m ratio was significantly lowerfor plants exposed to light, compared with shaded plants onthe days following the night frost (minimum temperature –6°C).The effect was similar both in 1-year-old and current year needles.The low F_v:F_m ratios indicate damage to photosystem H, causedby an interaction between sub freezing temperatures and highlight intensity. Shading also increased the survival of flushingshoots. It is suggested that regeneration of Norway spruce onsites exposed to frost should be carried out in partial shade,for example under a shelterwood.     

Frost Hardiness of Nothofagus procera and Nothofagus obliqua in Britain

The frost hardiness of four seedlots of Nothofagus proccra andsix seedlots of Nothofagus obliqua was measured experimentallyduring three winters. Shoots were taken from saplings growingat the Bush Estate in Scotland. All seedlots set buds in lateSeptember, hardened very slowly in the autumn, were damagedto some extent by temperatures below –14°C in mid-winter,and dehardened during frosty weather in February/March priorto budburst in mid to late April. The range of mid-winter temperaturesgiving 0%, 50% and 100% kill (LT₀, LT₅₀ and LT₁₀₀) were 8–14°C,13%20°C and 14–22°C, respectively. By contrast,British Fagus sylvatica hardened off rapidly in September, wasundamaged by frosts well below –20°C in mid-winter,and did not deharden until late April, prior to budburst atthe end of May. Nothofagus seedlots from Nuble in Chile (the most Equatorialsource) were the most frost susceptible: unfortunately, seedof this origin was supplied to many British nurseries between1976 and 1W9. Seedlots from Neuquen in Argentina, and from maturetrees of Malleco (Chile) origin growing in Britain, were themost host hardy. N. procera tended to be hardier than N. obliqua,and the trees became hardier with age. Past temperature records for Britain suggested that all theNothofagus seedlots had a high risk of suffering severe frostdamage at least once during a timkr rotation in all but mildcoastal regions. Spring and autumn frosts may be more damagingthan winter frosts. However, if it were possible to select individualswithin populations that were 3–6°C more frost hardythan the population means, such trees would be sufficientlyhardy to avoid frost damage in most lowland regions.     

Spring Frost Damage on Young Picea sitchensis 1. Occurrence of Damaging Frosts in Scotland Compared with Western North America

Air frosts (in Stevenson screens) of –2.5°C or belowin Scotland were judged to be potentially damaging to the newlyemerging shoots on young trees of Picea sitchensis at the timeof budburst. This was based on a knowledge of the killing tissuetemperature and the relationship between air and grass minimain May. Dates of last spring air frosts of –2.5°C and –4.5°Cwere obtained for 42 meteorological stations in northern Britainwith runs of 19 to 116 years. Frequency distributions of lastfrost dates were approximately normal. About 80% of the variationbetween stations could be attribuoted to altitude, distancefrom the sea and latitude. The variance in last dates of –2.5°Cfrosts decreased from mild to cold sites. Multiple regressionswere used to produce maps of last frost dates in 20x20 km gridsquares in Scotland. Dates of last 28° F (–2.2°C) and 24°F (–4.4°C)frosts were obtained for 20 sites in western North America,spanning the natural range of P. sitchensis. Mean dates increased,and variances decreased, from south to north. Most Scottish upland plantation sites (e.g. Glentress and Eskdalemuir)experience –2.5°C air frosts until later in the yearthan all coastal stations in western North America south ofCordova, Alaska. On average, Eskdalemuir (242 m altitude) experiences–2.5°C air frosts about 4 weeks later in the springthan Masset (3 m) on the coast of the Queen Charlotte Islands,which has been the source of much of the P. sitchensis plantedin Scotland. Masset has a spring frost climate somewhat milderthan Durham.     

Effects of simulated thaw on xylem cavitation,residual embolism,spring dieback and shoot growth in yellow birch

Yellow birch seedlings (Betula alleghaniensis Britt.) that had lost more than 90% of their stem hydraulic conductivity during ambient winter temperatures were exposed to 0 and 20 days of a simulated winter thaw followed by a 48-h freezing treatment at 0, -5, -10, -20 and -30 degrees C. After measuring freezing injury to shoots and roots, the seedlings were placed in a greenhouse where recovery of xylem conductivity and new growth were measured. Shoot xylem cavitation was measured as percent loss of hydraulic conductivity. Shoot freezing injury was assessed by electrolyte leakage (EL) and root freezing injury was assessed by EL and triphenyl tetrazolium chloride reduction. Seedlings pretreated with thaw had higher stem water contents and suffered more freezing damage to roots and shoots (at -20 and -30 degrees C, respectively) than unthawed seedlings. After 3 weeks in a greenhouse, seedlings from the 0, -5 and -10 degrees C freezing treatments showed complete recovery of xylem conductivity, with substantially increased stem water contents. Poor recovery of hydraulic conductivity was observed only in seedlings that were subjected to freezing treatments at -20 and -30 degrees C, regardless of thaw treatment. Of these embolized seedlings, however, only those not previously thawed showed recovery of hydraulic conductivity or regained stem water content after 9 weeks in the greenhouse. Shoot dieback, bud burst and length of new shoots were significantly related to the extent of stem xylem cavitation and freezing injury. We conclude that (1) the simulated winter thaw predisposed yellow birch seedlings to freezing damage in shoots and roots by dehardening tissues and increasing their water content; (2) root freezing damage in turn affected the seedlings' ability to refill embolized stem xylem, resulting in considerable residual xylem embolism after spring refilling; (3) further recovery of stem xylem conductivity was attributable to growth of new vessels; (4) and the permanent residual embolism, together with root and shoot freezing injury, caused increased dieback, bud mortality and reduced growth of new shoots.     

The use of field and artificial freezing studies to assess frost tolerance in natural populations of Pinus oocarpa

The susceptibility of Pinus oocarpa to freezing temperatures limits the commercial deployment of the highly productive Pinus patula × P. oocarpa hybrid in South Africa. Identifying P. oocarpa germplasm with increased frost tolerance is important. Twenty-three P. oocarpa provenances, originating from Mexico, Honduras, Guatemala and Nicaragua, were therefore assessed for their tolerance to freezing conditions by analysing field survival after frost events, subjecting needles to freezing temperatures and assessing damage using the electrolyte leakage test, and exposing young plants to freezing temperatures in a semi-controlled environment and scoring tissue damage based on a visual assessment. The performance of many of the provenances represented in the field and artificial freezing studies were similar and there was a strong correlation between provenance ranking in the whole plant freezing and electrolyte leakage test. We therefore support the use of these techniques as a means to assess cold tolerance in P. oocarpa at the provenance level. Provenances from north-western Mexico demonstrated more frost tolerance than those from southern Mexico. Provenances representing Honduras and Guatemala appear to be highly susceptible to frost.     

The effect of sub-zero temperatures in the light and dark on cold-hardened,dehardened and newly flushed white spruce (Picea glauca [Moench.] Voss) seedlings

Cold hardened, dehardened, and newly flushed foliage of one year old white spruce (Picea glauca [Moench.] Voss) seedlings were exposed to various sub-zero temperatures (--2 to --22.5°C) either in the dark or light. The freezing treatment had no significant effect on the variable fluorescence to maximal fluorescence ratio (F_v/F_m) of hardened seedlings, either in the light or dark. Also, no visible damage or increase in electrolyte leakage were evident in either the light or the dark treated seedlings. Both dehardened and newly flushed foliage were significantly affected by the freezing treatment, and light enhanced the effect. A decline in F_v/F_mincreased electrolyte leakage and visible damage were observed at warmer temperatures in newly flushed needles than in dehardened needles. Seedlings exposed to sub-zero treatments in the light also had lower F_v/F_m, increased electrolyte leakage and showed more visible damage than seedlings exposed to the same sub-zero treatments in the dark. The temperature where 50% of the needles were damaged (LT₅₀) as estimated from visible damage data was --10.8°C in the light and --12.1°C in the dark for dehardened, one year old needles. The LT₅₀in newly flushed needles was --4.8°C in the light and --6.2°C in the dark. Recovery of F_v/F_mvalues 3 days after freezing exposure was only evident in treatments where little visible damage was present. Both F_v/F_mand electrolyte leakage were strongly correlated with visible damage.     

The Infection of Corsican Pine by Crumenula sororia Karst

For germination ascospores of Crumenula sororia require freemoisture. A temperature range of 20–25 °C was optimal.In these conditions rate of germination was increased by extractsof lodgepole pine bark and extracts of Corsican pine bark andwood. Irrespective of whether spores or hyphae were used, C. sororiaonly Successfully invaded Corsican pine through wounds madewith a scalpel or by removing needle fascicles. Lesions didnot develop when frozen tissues were inoculated. In field inoculation experiments different isolates of the fungusdiffered significantly in pathogenicity. Infections developedmore severely on 4-year-old than on 1-year-old branches, andthis agrees with the incidence of natural branch-infections.     

Overwinter Storability of Conifer Planting Stock: Operational Testing of Fall Frost Hardiness

Operational stock-testing facilities that estimate overwinter storability of seedlings (ability to survive and grow after storage) need a reliable method that provides fast results to forest nurseries. We compared three methods using container-grown seedlings of Douglas-fir, interior spruce, lodgepole pine, and western larch from forest nurseries in British Columbia. On three to nine dates in autumn, frost hardiness at −18°C was estimated using visible injury of foliage or stems (VI), electrolyte leakage from needles or stems (EL), and chlorophyll fluorescence of shoots (CF). Seedlings were placed into overwinter cold storage (−2°C). In the spring, stored seedlings were planted in nursery beds; survival and growth were assessed after one growing season. There were close correlations (r ≥ 0.93) between the assessment methods. Seedlings lifted after they reached thresholds of 69% or higher for CF and 25% or lower for EL and VI had over 90% survival at harvest and doubled shoot dry weight compared with seedlings lifted earlier. Measuring CF was the fastest and most easily replicated method to estimate successful storability, and reduced testing time by 6 days relative to VI tests.     

Some Effects of Afforestation with Lodgepole Pine on Rates of Nitrogen Mineralization in Peat

Rates of mineral nitrogen production and carbon dioxide evolutionin incubated samples from the upper 300mm of peat beneath lodgepolepine (Pinus contorta Dougl.) have been compared with those foradjacent unplanted areas at each of six sites in the North ofScotland. Under both aerobic (moist) and anaerobic (water logged)conditions, rates of mineral nitrogen production at 30°Care strongly influenced by peatland type, sampling depth andafforestation. During the early stages of the incubation underaerobic conditions, samples of planted peat showed a more rapidaccumulation of mineral nitrogen than did samples from unplantedareas, the amounts after 17 days being 170ppm and 46ppm mineralN, respectively; after 62 days however, the difference was nolonger significant. The mean rate of CO₂ production averaged446µg CO₂ g^–1 day^–1 in planted as against728µg in unplanted peat. Under anaerobic conditions, amountsof mineral nitrogen accumulated were similar in planted andunplanted sites but a difference in accumulation between the0–150 and 150–300mm horizons in unplanted peat wassignificantly reduced beneath the trees.     

Assessing freeze damage in loblolly pine seedlings: A comparison of ethane production to electrolyte leakage

Ethane production was evaluated as a method for assessing freeze damage to loblolly pine (Pinus taeda L.) seedlings by comparing it to the widely used electrolyte leakage method. Paired measurements, first ethane production and then electrolyte leakage, were conducted on the pooled needle samples at temperatures between 0° and –12°C. Ethane production rates increased in a linear fashion with decreasing temperatures between 0° and –12°C for both Virginia Coastal Plain (R²=0.80) and Marion County, Florida (R²=0.87) seed sources. The Florida seedlings were consistently 2° to 4°C higher than the Virginia seedlings at a given ethane level. Electrolyte leakage expressed as Index of Injury initially increased with decreasing temperatures, but then leveled off at or decreased below –8°C. The log-log linear regression of ethane production against Index of Injury indicated good correspondence for both seed sources (Virginia – R²=0.81; Florida – R²=0.91). Ethane production appears more rapid and to require less sampling than does electrolyte leakage while producing comparable results to the electrolyte leakage method.     

Impregnation of Norway spruce (Picea abies L. Karst.) wood by hydrophobic oil and dispersion patterns in different tissues

Wood from Norway spruce (Picea abies L. Karst.) is biologicallydegraded in exposed conditions. It also has anatomical featuresthat make it difficult to impregnate with preservatives by currentlyavailable industrial processes. In the study reported here,we used the new Linotech process to impregnate Norway sprucewood with hydrophobic linseed oil and then quantified its uptakeand dispersal in anatomically distinct wood tissues. We alsoinvestigated the effects of the wood moisture content on theresults of the impregnation. Samples (500 x 25 x 25 mm) weretaken from 15 trees in a coniferous forest in northern Sweden(64° 10' N, 160–320 m a.s.l.). The parameters forthe Linotech process were 2–3 h treatment time at 0.8–1.4MPa and 60–140°C. To determine the level of uptake,the linseed oil was extracted from the impregnated wood usingmethyl-tertiary-butyl-ether. The uptake was quantitatively analysedby comparing X-ray microdensitometry values obtained followingimpregnation both before and after oil removal. In earlywood,initial moisture content had an obvious effect on the impregnationresult. Six times more oil was taken up when the moisture contentwas greater than ~150 per cent than when it was less than 30per cent. Theoretical calculations, based on density levels,suggest that the water-filled porosity of the wood (water volumedivided by porosity volume) was positively correlated with thelinseed oil uptake, and more strongly correlated in earlywoodthan in latewood. There were also significant differences inuptake between different wood tissues; heartwood/mature woodand heartwood/juvenile wood showed 10–20 per cent weightincreases due to linseed oil uptake, compared with 30–50per cent in sapwood/mature wood. Examination by scanning electronmicroscopy confirmed these uptake patterns. The moisture contentafter impregnation was about 5 per cent, irrespective of theLinotech process parameters, tissue type and initial moisturecontent. In conclusion, the impregnation process used here resultsin high levels of well-dispersed linseed oil uptake and shouldfacilitate drying.     

Winter Frost Hardiness of two Chilean Provenances of Nothofagus procera in Scotland

The frost hardiness of the shoots of individual trees withintwo Chilean provenances of Nothofagus procera (Poepp & Endl.)Oerst. was measured once in each of the months January, February,November and December 1989 and January and February 1990. Therewere significant (P<0.05) differences of frost hardinessbetween provenances but only one tree could be shown to be significantlymore frost hardy than the others within the same provenance.During the winter of 1989/90 both provenances were hardy toabout –14°C (temperature killing 50 per cent of shoots)in December, but the shoots dehardened to about –9°Cin January before hardening again in February. This patternof alternate hardening and dehardening seemed to mirror changesin air temperature and could render N. procera liable to frostdamage where (as happened in 1988/9 in the UK) mild spells occurin winter followed by severe frosts.     

Frost Hardiness of Red Alder (Alnus rubra) Provenances in Britain

The phenology and frost hardiness of shoots of 15 provenancesof Alnus rubra growing in Scotland were measured over one autumn,winter and spring. Dates of budset (in September) and the onsetof rapid frost hardening (in October-November) occurred about2 days earlier for each degree latitude of origin northwards,except for an Idaho provenance. However, all provenances dehardenedat about the same time in March and burst their buds between8 and 14 April. Assuming that rapid frost hardening in the autumnwas triggered primarily by shortening daylengths, Alaskan provenancesof A. rubra seemed better adapted to British conditions thansouthern British Columbian provenances, which have been mostcommonly planted. However, even Alaskan provenances are proneto spring frost damage. Scottish A. glutinosa and Alaskan A.sinuata set buds and frost hardened 1–2 weeks before eventhe Alaskan A. rubra, and burst their buds 2–3 weeks laterin April-May. All three species were hardy to below –30°Cfrom December to mid-March.     

Autumn Frost Damage on Young Picea sitchensis 2. Shoot Frost Hardening, and the Probability of Frost Damage in Scotland

The natural increase in frost hardiness of detached shoots ofPicea sltchensis during August to November was measured usinga programmable freezing chamber. Oregon, Queen Charlotte Islandsand Alaskan provenances were compared, and the effects on hardeningof long days, warm temperatures and frosts were determined.A computer model was constructed to mimic the observed patternsof autumn frost hardening, as functions of air minimum temperatures,daylengths and the occurrence of frosts. The model was used(a) to describe the pattern of autumn frost hardening at differentsites in northern Britain, using past meteorological records,and hence (b) to determine when frosts occurred that might havedamaged young trees. The model accurately predicted known instancesof autumn frost damage at Kirroughtree and Carnwath. The predicted probability of autumn frost damage on young treesof P. sitchensis in upland areas of Scotland was much lowerthan that previously predicted for spring frost damage. Theestimated return time for autumn frost damage to an Oregon provenanceat Eskdalemuir was 8.3 years, and the return time for a Q.C.I.provenance was longer than 10 years. Most damaging frosts occurredin October, but frosts like those on 13–15 October 1971,which followed warm weather and caused wide spread damage inScotland, have been quite rare. Alaskan provenances would rarelybe damaged by autumn frosts, nor would trees of Q.C.I. provenancegrowing in lowland areas of Scotland, or at Masset on the QueenCharlotte Islands.