PHYSIOLOGIA PLANTARUM 112: 71–77. 2001 Copyright © Physiologia Plantarum 2001ISSN 0031-9317Printed in Ireland—all rights reser6ed

Effects of climatic warming on cold hardiness of some northern woodyplants assessed from simulation experiments

Erling O8 gren

Department of Forest Genetics and Plant Physiology, Swedish Uni6ersity of Agricultural Sciences, SE-901 83 Umea, SwedenE-mail: erling.ogren@genfys.slu.se

Received 25 August 2000; revised 31 October 2000

Effects of climatic warming on cold hardiness were investigated energy reserves that buffered respiratory expenditure of sugars.for some northern woody plants. In the first experiment, A strong, linear relationship between levels of cold hardiness,

assessed by the electrolyte leakage method, and sugars wasseedlings of Norway spruce (Picea abies [L.] Karst.), Scots pine(Pinus syl7estris L.) and lodgepole pine (Pinus contorta Dougl. found when combining data from this and previous, similar

experiments. In the second experiment, the evergreen dwarfvar. latifolia Engelm.) were exposed to naturally fluctuatingtemperatures averaging −6°C (ambient) and 0°C (elevated) shrub Empetrum hermaphroditum Hagerup was analysed for

leaf cold hardiness, using the electrolyte leakage method, andfor 16 weeks in midwinter before they were thawed andre-saturated with water. In lodgepole pine, needle sugar concen- sugar concentrations in late spring and late autumn during thetrations had decreased by 15%, and the temperature needed to third year of a warming experiment in a subarctic dwarf shrub

community. The objective was to test the hypothesis thatinduce 10% injury to needles in terms of electrolyte leakage hadincreased by 6°C following treatment to elevated as compared warming in the growing season alters hardening/dehardening

cycles by increasing soil nitrogen mineralization and plantwith control temperatures. In contrast, Norway spruce andgrowth. Data found, however, suggested that cold hardening/Scots pine showed no effects. The lack of an effect for Scots

pine was ascribed to seedlings containing unusually large dehardening cycles were unaffected by warming.

tures may delay (Repo et al. 1996) or even prevent (Guak etal. 1998) the completion of cold hardening.

In addition, climatic warming may affect the ability ofplants to maintain maximum levels of cold hardiness inmidwinter. We have demonstrated that coniferous seedlingstreated with moderately raised temperatures in midwintergradually lost cold-protective sugars even at temperaturesbelow 0°C, and, in proportion, cold hardiness (O8 gren 1997).This occurred more rapidly in lodgepole pine than in Scotspine and Norway spruce (O8 gren et al. 1997), which wasexplained by lodgepole pine having the highest respirationrate and hence the highest rate of sugar expenditure (O8 gren2000). In contrast, Scots pine saplings treated to experimen-tal warming during the dormant-season showed no signs ofdehardening in midwinter although spring dehardening wasadvanced (Repo et al. 1996). The apparent conflict betweenresults might reflect the fact that fluctuating temperaturesprevailed in the latter study as opposed to constant temper-atures in our laboratory study. Plants may deharden moreslowly under fluctuating than constant temperatures

Introduction

Global circulation models predict that the global meantemperature will rise by 1.5–4.5°C by the year 2050 becauseof an expected doubling in atmospheric CO2 concentration(Kattenberg et al. 1996). Warming may be more pro-nounced at the higher than the lower latitudes (Cattle andCrossley 1996) and in the winter than in the summer months(Johannesson et al. 1995). As winter temperature minimamay remain unaffected (MacCracken et al. 1991), propercold hardening will remain a decisive factor for the survivalof northern plants.

Climatic warming may upset cold hardening/de-hardeningcycles in plants for several reasons. Because spring budburstis controlled by thermal time, climatic warming may causepremature budburst and thereby increase the risk of frostdamage because of premature de-hardening (Hanninen1991, Heide 1993). Experimental data support this predic-tion (Repo et al. 1996). Although the onset of cold harden-ing in autumn is largely controlled by photoperiod (Heide1974), later stages of cold hardening are driven by loweredtemperature (Repo 1992). Consequently, elevated tempera-

Physiol. Plant. 112, 2001 71

(Leinonen et al. 1997). One objective of the present workwas to explore this possibility by repeating our experimentswith elevated winter temperatures but in the field.

Theoretically, warming during the growing season mayupset cold hardening/dehardening cycles. In cold climatesplants grow slowly, partly because low soil temperatureslimit nitrogen (N) mineralization and, hence, the N supplyto plants (Chapin et al. 1993). Effects of soil warming may,therefore, resemble effects of N fertilization. Applications ofhigh doses of N fertilizers often give rise to delayed growthcessation (and cold hardening) in autumn and advancedbudburst (and deharding) in spring (e.g. Sennerby-Forsseand von Fircks 1987, Power et al. 1998). In addition, areduced capacity for cold hardening has been observed insome species (Hellergren 1981, Junttila et al. 1995). Asecond objective of this study was to evaluate this risk thatrising growing-season temperatures at the northern latitudeswill adversely affect cold hardening/dehardening cycles ofplants as a consequence of an increased N mineralization.To this end, we have utilized a long-term soil warmingexperiment in a subarctic dwarf shrub community whereindeed increased rates of N mineralization and plant growthhave been demonstrated (Hartley et al. 1999).

Materials and methods

Design of winter warming experiment

Seedlings of Norway spruce (Picea abies [L.] Karst., seedorchard progeny ‘Hissjo’ with average clonal origin 64° N,interior of Scandinavia), Scots pine (Pinus syl6estris L., seedorchard progeny ‘O8 stteg’ with average clonal origin 65° N,interior of Scandinavia), and lodgepole pine (Pinus contortaDougl. var. latifolia Engelm., ‘continental variety’, prove-nance ‘Rusty Creek’, 63.5° N 136.5° W) were raised for 2years, the first in 50-ml and the second in 2-l pots, at anoutdoor nursery (64° N, 20.5° E) (O8 gren 1999a).

Treatment started on 10 November 1995, when random-ized sets of 20 plants were transferred to each of twogreenhouses: the control greenhouse, which was unheatedand open to the atmosphere, and the experimental green-house, which was heated and fan-ventilated to maintain atemperature that was typically 5°C warmer than the control.However, the temperature was not allowed to drop below−10°C in the experimental greenhouse. Air temperatures ofthe two greenhouses, measured with thermocouples (0.05mm in diameter) and a data logger (CR10, Campbell Scien-tific Inc., Logan, UT, USA), averaged −6.0 and −0.1 °C,respectively, during the experimental period and varied asshown in Fig. 1b,c. From 22 December 1995 onwards, potswere covered with snow to protect roots from cold injury.Shoots were exposed to near-full daylight as snow wasmanually removed from greenhouses within a day after asnowfall and there were no trees and buildings nearby. Theirradiance above plants, measured with a quantum sensor(LI-189; Li-Cor Inc., Lincoln, NE, USA) and the datalogger, varied as shown in Fig. 1a. On 28 February 1996plants were transferred to a cold room kept at 5°C and anirradiance of 1–2 mmol m−2 s−1 (8 h photoperiod), where

they were stored for 5–6 days to thaw and regain full waterstatus before analysis. At the end of this period, shoot waterpotentials were –0.5590.05 MPa, as measured with aScholander bomb (Scholander et al. 1964). Fifteen plants ofeach category were chosen for analyses and randomized intofive replicate groups of three plants each.

Design of summer warming experiment

A warming experiment was utilized that was set up byHartley and co-workers (1999) in northern Sweden (68.21°N 18.50° E, 380 m a.s.l.) in a subarctic dwarf shrub under-story of an open birch woodland (Betula pubescens Ehrh.ssp. tortuosa Ledeb. Nyman). One of the dominant dwarfshrubs was the evergreen Empetrum hermaphroditumHagerup. The soil temperature of a 4.8×6-m plot wasraised by 5°C above that of an adjacent control plot usingburied heating wires. The warmed-soil plot was divided into20 subplots of 1 m2. Four of these were furnished withopen-top chambers to elevate air temperatures as well. Thepresent study was carried out during the third year of soilwarming and the second year of air warming. Warming wasconfined to the snow-free-season, from late May to earlyOctober during this particular year. Air temperatures, mea-

Fig. 1. Daily light dose (a) and hourly mean air temperature ofcontrol (b) and elevated temperature regimes (c) during the experi-ment with coniferous seedlings.

Physiol. Plant. 112, 200172

Fig. 2. Two-hourly mean temperature of air (a) on soil-warmedplot, and of soil (b) on soil-warmed (upper) and control plots (lowercurve) during the experiment with E. hermaphroditum. Warmingcontinued uninterrupted until early October although some data aremissing.

were immediately transferred to 5°C to thaw. After the lasttubes had thawed, 4 ml of incubation medium containing0.002% Triton X-100 and 10 mM boric acid was added toall tubes. Tubes were shaken at 250 rpm for 18 h and thenanalysed for conductivity as detailed in (O8 gren 1997). Mea-surements were repeated after the tubes had been heated for10 min at 90°C and shaken for an additional 18 h. Indices ofinjury were calculated according to Flint et al. (1967), thusobtaining indices of 0 and 100% for unfrozen controls andheated samples, respectively. Categories were compared atthe 10% index of injury assuming that the maximal, freeze-induced index of injury did not vary much, which wouldinvalidate such comparisons at fixed levels. This assumptionwas tested in retrospect using a more powerful coolingsystem (DC180PU freezing cabinet, Weiss UmwelttechnikGmbH, Lindenstruth, Germany). Needles of winter-hardyand summer-active seedlings, expected to show the largestdifference in the maximal index (e.g. Repo et al. 1996), werefrozen to −80°C at a rate of 5.0°C h−1 (10.0°C h−1

thawing rate). Analyses yielded the following indices ofinjury of winter-hardy/summer-active plants (%): 5194/6892 (Norway spruce), 6892/7792 (Scots pine) and6891/7091 (lodgepole pine; mean9SE values for fourreplicate determinations of ten pooled seedlings each). Theplants had similar pre-history as those used in the mainexperiment, but the winter-hardy plants had been stored at−5°C in winter and the summer-active plants had beenforced to start growing in a greenhouse. Differences werethus relatively small, despite the fact that, first, seasonalextremes were compared and, second, −80°C may not havebeen sufficiently low a temperature to induce maximal injuryin the more winter-hardy species Norway spruce and Scotspine. Thus, the practice in the present study of comparingplants of the same seasonal state at a fixed index of injury of10% seems justified.

Shoots of E. hermaphroditum were prepared for freezetests in the cold room on the day following sampling. Theywere randomized into five replicate groups of 3–6 shootseach. Similar amounts of leaves of the current-year’s (Octo-ber) and previous-year’s (June) flush were sampled fromeach shoot belonging to a replicate. The leaves were ran-domized, cut into halves transversely, washed in distilledwater (2×60 s), before 35 segments were transferred to eachof the 7–8 test tubes representing a temperature series pluscontrol, a procedure which was repeated for all replicates.Procedures for freezing and electrolyte analysis were thoseused for needles but modified as follows: ice formation wasinduced at −4°C by immersing a liquid-nitrogen pre-cooledwire into the water of the tubes (to achieve equilibriumfreezing at temperatures \−10°C, a critical range in sum-mer); a smaller amount of incubation medium (0.8 ml) anda smaller conductivity probe (Unicam microcell type 9554,Unicam Ltd., Cambridge, UK) was used to fit the smallersample size; tubes were shaken less vigorously (150 rpm)because smaller leaves should equilibrate more effectively;and finally, the maximal electrolyte leakage was induced byfreezing in liquid nitrogen to obtain a better estimate of100% injury.

sured with a set of radiation-shielded thermistors, and soiltemperatures, measured with a set of thermistors buried at adepth of 5 cm in between heating wires, varied on controland soil-warmed plots (no air-warming) as shown in Fig. 2.The extent of air-warming achieved by open-top chambersvaried with solar radiation. When measured on a clear day(20 July 1995), daily mean and maximum air temperatureswere elevated by 2.1 and 7.5°C, respectively, in one of thechambers as compared with the ambient (the chamber effecton soil temperature was negligible). Fifteen to thirty termi-nal shoots of E. hermaphroditum were sampled randomlyfrom central parts of all plots and chambers and shipped onice to the laboratory.

Freeze test

The previous year’s flushes of coniferous needles were pre-pared for freeze tests in the cold room: about 1.5 g ofneedles was sampled from each of the three plants belongingto a replicate group. They were cut into segments (1 cmlength), washed in distilled water (2×60 s) before 0.4 g ofthe combined material was transferred to a each of the 8–9test tubes representing a temperature series plus control.After this procedure had been repeated for all replicates,distilled water (0.2 ml) was added to all tubes which werethen transferred, except controls, to the freezing bath wherethey were cooled at a rate of 4.8°C h−1, beginning at 0°C.Temperatures of one of the samples and of the bath,recorded with thermocouples and a data logger, agreed towithin 0.2°C except at temperatures \−10°C where thetemperature of the sample decreased more slowly than thatof the bath. When a test temperature was reached, the tubes

Physiol. Plant. 112, 2001 73

Fig. 3. Index of injury of needles as a function of test temperaturefor seedlings of Norway spruce (a), Scots pine (b) and lodgepolepine (c) that had experienced control (closed) and elevated tempera-tures (open symbols) for 16 weeks in midwinter. Mean9SE valuesare given for five replicate determinations of three pooled individu-als each.

Results

Effects of warming in winter on conifers

Seedlings of lodgepole pine treated to elevated temperaturesaveraging 0°C for 16 weeks in midwinter showed partialde-hardening relative to controls kept at normal ambienttemperatures averaging −6°C, both analysed after restora-tion of full water status (Fig. 3). For example, the testtemperature associated with an index of injury of 10% forneedles increased from −26 to −20°C. In contrast,seedlings of Norway spruce and Scots pine were unaffectedby treatment (Fig. 3). In addition, lodgepole pine wasinherently less cold-hardy than the other species: tempera-tures associated with an index of injury of 10% for controlswere −26, −29 and −44°C for lodgepole pine, Scots pineand Norway spruce, respectively.

Reasons for this variation in cold hardiness was sought byplotting temperatures associated with indices of injury of10% as a function of total needle sugar concentrations (Fig.4, closed symbols). Also included are data of two previousexperiments with elevated but constant temperatures usingthe same species (open symbols). Evidently, most of theoverall variation in needle cold hardiness in winter acrossspecies and treatments at the state of fully hydrated tissuesis accounted for by the same, linear correlation with needlesugar concentration (r2=0.95). The extent of dehardeningshown by lodgepole pine in the present experiment (6°C

Fig. 4. Temperature inducing an index of injury of 10% of needles,plotted as a function of needle sugar concentration for seedlings ofNorway spruce (, �) Scots pine (, ) and lodgepole pine(�, �), combining data of the present (closed) and two previousexperiments with elevated winter temperatures (open symbols;O8 gren 1997, O8 gren et al. 1997). Arrows connect treatments belong-ing to the same experiment in sequences of increasing mean temper-atures starting with the control. Data of critical temperatures forthe present experiment were extracted from Fig. 3. Mean9SEvalues are given for five replicate determinations of 3–6 pooledindividuals each. The line is fitted by linear regression (y=−2.542x+0.67; r2=0.95).

Carbohydrate analyses

The total amount of soluble sugars was analysed forneedles. The same category of needles as used for freezetests was frozen in liquid nitrogen and stored at −60°Cbefore they were freeze-dried and ground to powder in aball mill. Analysis was carried out as detailed elsewhere(O8 gren 1997) using samples of 50 mg. Soluble sugars wereextracted with 80% ethanol and quantified according tothe anthrone method, modified by Jermyn (1975) to yieldidentical colour reactions for the major sugars. Theresidue from extraction was analysed for starch using en-zymatic methods for hydrolysis of starch and detection ofglucose released as detailed in (O8 gren 1997).

The total amount of glucose, fructose and sucrose wasanalysed for E. hermaphroditum leaves. The same cate-gories of leaves as used for freeze tests were prepared asdescribed above for needles. They were extracted andanalysed for sugars using an enzymatic method as de-scribed elsewhere (O8 gren 1999b). Treatment effects onsugar concentrations were tested using a two-factor analy-sis of variance (ANOVA) with time and treatment beingthe two factors.

Physiol. Plant. 112, 200174

decrease in critical temperature) is essentially that predictedby the regression line for the amount of sugars lost (−15%loss; cf. closed triangles, Fig. 4). The lack of an effect forNorway spruce and Scots pine in the present experiment isconsistent with these species maintaining their sugar concen-trations (Fig. 4).

Effects of warming in summer on E. hermaphroditum

On 1 June 1995, about 1 week after soil heating wasresumed after the winter break, and about 1 week beforebud burst, plants from the control, soil-warmed and soil+air warmed plots, showed equal levels of cold hardiness withindices of injury of 50% associated with a temperature of−14°C (Fig. 5a). On 8 October 1995, about 2 months afterthe plants had stopped growing, all categories of plants hadacquired deep cold hardiness: indices of injury of 10% wereassociated with temperatures lower than −30°C through-out. The critical temperature was actually 3°C lower for theplants from the soil-warming plot than the rest but thisdifference at the stage of deep cold hardiness is too small tobe of any significance. Leaf sugar concentrations in springand autumn were unaffected by either air or soil warming(Table 1).

Discussion

Seedlings of lodgpole pine treated to elevated temperaturesin winter showed decreased levels of sugars and cold hardi-ness, whereas seedlings of Scots pine and Norway spruceshowed no effects (Fig. 4). The higher temperature sensitiv-

Table 1. Total leaf concentration of glucose, fructose and sucrosefor the different categories of E. hermaphroditum plants assessed ontwo occasions. Mean9SE values are given for three pooled repli-cates, each representing 5–6 shoots. Differences among treatmentsare not significant whereas those between dates are significant atPB0.0001.

Sugar concentration (% of DW)Treatment

1 June 1995 8 October 1995

Control 6.2490.11 12.590.6Soil-warming 11.990.66.1290.20Soil+air warming 6.5890.25 11.890.3

ity of lodgepole pine was also observed in a previousexperiment and then ascribed to its inherently higher respi-ration rate, and hence higher sugar expenditure rate (O8 grenet al. 1997). Two possible explanations for its higher respira-tion rate have been suggested (O8 gren 2000). First, theparticular subspecies used by us originates from interiorCanada where winter temperatures are generally lower andmore stable than in Scandinavia, the origin of the otherplants studied. Plants from colder regions may rely on lowtemperatures suppressing respiration rates in winter whereasplants native to milder regions may actively down-regulaterates. Alternatively, its higher respiration rate may be asso-ciated with its higher growth rate per unit N (Norgren1996), a trait which may require a higher rate of proteinturnover and, hence, an increased respiration rate.

Spruce and pine showed the same, quantitative relation-ship between cold hardiness levels and sugar concentrations,suggesting a common underlying mechanism (Fig. 4). Mech-anistic studies indicate that cellular sugars protect cells bytwo means. First, by reducing the amount of water diffusingout of cells to growing ice crystals, thereby reducing thedegree of cell shrinkage (Hodge and Weir 1993), and hencemechanical stress to particularly membranes. Second, bystabilizing cellular structures once freeze-induced dehydra-tion occurs (Hirsh 1987, Crowe et al. 1990). Both functionsare likely to stabilize the cell membrane including its activetransport system believed to be the primary site of coldinjury causing release of electrolytes when injured (Palta etal. 1977, Arora and Palta 1991). Thus, the strong inverserelationship found here between levels of cold injury as-sessed from electrolyte leakage and levels of sugar concen-trations (Fig. 4) might well represent a causal relationship.In addition, cold hardiness is likely to vary in winter becauseof weather-induced changes in the state, amount and distri-bution of water, non-physiological factors which were neu-tralized in these studies by reconditioning plants to fullwater status before analyses.

In a previous experiment with similarly elevated but con-stant temperatures, Scots pine seedlings also showed partialde-hardening in midwinter (O8 gren 1997). The seedlings con-taining unusually large energy reserves might explain thelack of an effect in the present experiment for Scots pine.Starch reserves stored in the roots buffers sugar expenditureduring mild winters (O8 gren et al. 1997) and lipid reservesstored in the stem (Glerum and Balatinez 1980) may do thesame. Although neither roots nor stems were analysed here,

Fig. 5. Index of injury of leaves plotted as a function of testtemperature for E. hermaphroditum plants growing on the control(), soil-warmed (�) and soil+air warmed plots ( ) assessed onJune 1, 1995 (a) and October 8, 1995 (b). Mean9SE values aregiven for five replicate determinations of 3–6 pooled shoots each.

Physiol. Plant. 112, 2001 75

needles contained 13-fold higher amounts of starch (1.3versus 0.1% of dry weight, data not shown), and 20% higheramounts of sugars than previously observed (Fig. 4), indi-cating that energy reserves were generally larger. The newsituation was presumably explained by weather conditionsin autumn being exceptionally favourable for photosynthesiswith the largest number of sunny days for the last 75 years(Swedish Meteorological and Hydrological Institute, per-sonal communication).

Thus, the extent of sugar depletion in winter shoulddepend on the overall size of the energy reserve. Nursery-raised seedlings might contain a smaller reserve than field-grown saplings as fertilization and irrigation treatments arelikely to promote growth more than storage. This may offeran explanation for why nursery-raised seedlings of Scotspine used by us have shown de-hardening in response tolong-term temperature elevation in midwinter (O8 gren 1997),whereas field-grown saplings of Scots pine have not (Repoet al. 1996). For similar reasons, understory plants growingin the shade may be more at risk of losing sugars andundergo de-hardening during mild winters than overstoryplants. Indeed, billberry plants growing in the understory ofa northern boreal forest showed progressive losses of sugarsand cold hardiness during an exceptionally mild winter(O8 gren 1996). At these high latitudes, respiratory lossesduring mild winters are not compensated for by concurrentphotosynthesis because daylight intensities are too low inwinter (Fig. 1a). This is in contrast to the situation in theBritish Isles where over-wintering plants may actually gaindry mass (Rutter 1957).

There were no carry-over effects of elevated temperaturesfrom the growing season to subsequent phases of coldhardening and dehardening for the E. hermophroditumplants. This was inferred from the finding that leaf sugarconcentrations and cold hardiness levels in late spring andautumn were unaffected by treatment (Table 1, Fig. 5).Theoretically, carry-over effects might have occurred giventhat N mineralization, and hence N availability had in-creased on the warmed plots (Hartley et. al. 1999), andgiven that N fertilization treatments often alter cold harden-ing/dehardening cycles by altering growth rhythms (Sen-nerby-Forsse and von Fircks 1987, Power et al. 1998). Thus,the fertilization effect of soil warming seems to have beenminor only. This was also concluded from the observationthat plant community structure was unchanged after 5 yearsof soil warming on these plots (Hartley et al. 1999). North-ern perennial plants may be largely independent of growing-season conditions for proper cold hardening: variability ingrowing season conditions at the northern latitudes is highwhilst plants are extremely dependent on successful coldhardening for their survival.

In conclusion, the risk that climate change in Scandinaviawill cause premature de-hardening in midwinter seems smallfor the native species − as long as their energy reserves arenot unduly small − but moderate for the introduced specieslodgepole pine. The risk that warming in the growing seasonwill upset cold hardening/de hardening cycles seems evensmaller.

Acknowledgements – This work was financially supported by theSwedish Council for Forestry and Agricultural Research. I amgrateful to investigators at the Abisko soil-warming experiment forhaving access to their plots. I am especially grateful to Dr RoseCrabtree for providing temperature data and plant material.

ReferencesArora R, Palta JP (1991) A loss in the plasma membrane ATPase

activity and its recovery coincides with incipient freeze-thawinjury and postthaw recovery in onions bulb scale tissue. PlantPhysiol 95: 846–852

Cattle H, Crossley J (1996) Modeling arctic climate change. In:Raschke E, Cattle H, Lemke P, Rossow W (eds) The Arctic andEnvironmental Change. The Royal Society, New York, NY, pp1–13

Chapin FS III, Moilanen L, Kielland K (1993) Preferential use oforganic nitrogen for growth by a nonmycorrhizal arctic sedge.Nature 361: 150–153

Crowe JH, Carpenter JF, Crowe LM, Anchordoguy TJ (1990) Arefreezing and dehydration similar stress vectors? A comparison ofmodes of interaction of stabilizing solutes with biomolecules.Cryobiology 27: 219–231

Flint HL, Boyse BR, Beattie DJ (1967) Index of injury – a usefulexpression of freezing injury to plant tissues as determined bythe electrolytic method. Can J Plant Sci 47: 229–230

Glerum C, Balatinez JJ (1980) Formation and distribution of foodreserves during autumn and their subsequent utilization in jackpine. Can J Bot 58: 40–54

Guak S, Olsyzk DM, Fuchigami LH, Tingey DT (1998) Effects ofelevated CO2 and temperature on cold hardiness and spring budburst and growth in Douglas-fir (Pseudotsuga menziesii ). TreePhysiol 18: 671–679

Hanninen H (1991) Does climatic warming increase the risk of frostdamage in northern trees? Plant Cell Environ 14: 449–454

Hartley AE, Neill C, Melillo JM, Crabtree R, Bowles FP (1999)Plant performance and soil nitrogen mineralization in responseto simulated climate change in subarctic dwarf shrub heath.Oikos 86: 331–343

Heide OM (1974) Growth and dormancy in Norway spruce eco-types (Picea abies). I. Interaction of photoperiod and tempera-ture. Physiol Plant 30: 1–12

Heide OM (1993) Daylength and thermal time responses of bud-burst during dormancy release in some northern deciduous trees.Physiol Plant 88: 531–540

Hellergren J (1981) Frost hardiness development in Pinus sil6estrisseedlings in response to fertilization. Physiol Plant 52: 297–301

Hirsh AG (1987) Vitrification in plants as a natural form ofcryoprotection. Cryobiology 24: 214–228

Hodge GR, Weir RJ (1993) Freezing stress tolerance of hardy andtender families of loblolly pine. Can J For Res 23: 1892–1899

Jermyn MA (1975) Increasing the sensitivity of the anthronemethod for carbohydrates. Anal Biochem 68: 332–335

Johannesson T, Jonsson T, Kallen E, Kaas E (1995) Climate changescenarios for the Nordic countries. Climate Res 5: 181–195

Junttila O, Svenning MM, Rosnes K (1995) Influence of mineralnitrogen source on growth and frost resistance of white clover(Trifolium repens L.) and timothy (Phleum pratense L.) seedlings.Acta Agric Scand Sect B Soil Plant Sci 45: 261–267

Kattenberg A, Giorgi F, Grassl H, Meehl GA, Mitchell JFB,Stouffer RJ, Tokioka T, Weaver AJ, Wigley TML (1996) Cli-mate models – projections of future climate. In: Hougton JT,Meira Filho LG, Callander BA, Harris N, Kattenberg A,Maskell K (eds) Climate Change 1995: The Science of ClimateChange. Cambridge University Press, Cambridge, pp 289–357

Leinonen I, Repo T, Hanninen H (1997) Changing environmentaleffects on frost hardiness of Scots pine during dehardening. AnnBot 79: 133–138

MacCracken M, Cubasch U, Gates WL, Harvey LD, Hunt B, KatzR, Lorenz E, Manabe S, McAvaney B, McFarlane N, Meehl G,Meleshko V, Robock A, Stenchikov G, Stouffer R, Wang W-C,Washington W, Watts R, Zebiak S (1991) A critical appraisal ofmodel simulations. In: Schlesinger MF (ed) Greenhouse-Gas-In-duced Climate Change: A Critical Appraisal of Simulations andObservations, Developments in Atmospheric Science 19. El-sevier, Amsterdam, pp 583–591

Physiol. Plant. 112, 200176

Norgren O (1996) Growth analysis of Scots pine and lodgepole pineseedlings. For Ecol Manage 86: 15–26

O8 gren E (1996) Premature dehardening in Vaccinium myrtillusduring a mild winter: A cause for winter dieback? Funct Ecol10: 724–732

O8 gren E (1997) Relationship between temperature, respiratory lossof sugar and premature dehardening in dormant Scots pineseedlings. Tree Physiol 17: 47–51

O8 gren E (1999a) Fall frost resistance in willows used for biomassproduction. I. Characterization of seasonal and genetic varia-tion. Tree Physiol 19: 749–754

O8 gren E (1999b) Fall frost resistance in willows used for biomassproduction. II. Predictive relationships with sugar concentrationand dry matter content. Tree Physiol 19: 755–760

O8 gren E (2000) Maintenance respiration correlates with sugar butnot nitrogen content in dormant plants. Physiol Plant 108:295–299

O8 gren E, Nilsson T, Sundblad L-G (1997) Relationship betweenrespiratory depletion of sugars and loss of cold hardiness inconiferous seedlings over-wintering at raised temperatures: Indi-cations of different sensitivities of spruce and pine. Plant CellEnviron 20: 247–253

Palta JP, Levitt J, Stadelmann EJ (1977) Freezing injury in onionbulb cells. II. Post-thawing injury or recovery. Plant Physiol 60:398–401

Power SA, Ashmore MR, Cousins DA, Sheppard LJ (1998) Effectsof nitrogen addition on the stress sensitivity of Calluna 6ulgaris.New Phytol 138: 663–673

Repo T (1992) Seasonal changes of frost hardiness in Picea abiesand Pinus syl6estris in Finland. Can J For Res 22: 1949–1957

Repo T, Hanninen H, Kellomaki S (1996) The effects of long-termelevation of air temperature and CO2 on the frost hardiness ofScots pine. Plant Cell Environ 19: 209–216

Rutter AJ (1957) Studies in the growth of young plants of Pinussyl6estris L. I. The annual cycle of assimilation and growth. AnnBot 21: 399–426

Scholander PF, Hammel HT, Hemmingsen EA, Bradstreet ED(1964) Hydrostatic pressure and osmotic potential in leaves ofmangroves and some other plants. Proc Natl Acad Sci USA 52:119–125

Sennerby-Forsse L, von Fircks HA (1987) Ultrastructure of cells inthe cambial region during winter hardening and spring dehard-ening in Salix dasyclados Wim. grown at two nutrient levels.Trees 1: 151–163

Edited by M. Griffith

Physiol. Plant. 112, 2001 77