New Forests 12 : 243-259, 1996 .© 1996 Kluwer Academic Publishers . Printed in the Netherlands .

Frost hardiness, bud phenology and growth ofcontainerized Picea mariana seedlings grown at threenitrogen levels and three temperature regimes

F.J. BIGRASI ,*, A. GONZALEZ', A.L. D'AOUST' and C. HEBERT2Natural Resources Canada, Canadian Forest Service - Quebec, P.O. Box 3800 Sainte-Foy,Quebec G1V 4C7, Canada 'Research scientist ; 2 Statistician (*Author to whom allcorrespondence should be addressed. Telephone: (418) 648-2528; Fax: (418) 648-5849;e-mail: BIGRAS@CFL.FORESTRY.CA

Received 15 December 1994; accepted in revised form 1 March 1996

Key words: hardening, dehardening, mineral nutrition, nutrient content, black spruce

Application. As luxury consumption of macroelements by plants is generally considered toreduce hardiness, tree seedlings must reach an optimum N concentration for hardening . Blackspruce seedlings at the end of their second growing season with 1 .28% N (shoot dry mass basis)harden more than seedlings containing sub-optimal levels of 0.64% or 0.87% N. Fertilizationduring the second growing period should produce black spruce seedlings containing at least1 .28% N in shoot tissues to ensure hardening during the first stage of cold acclimation .

Abstract. We studied the influence of temperature and near- and sub- optimal mineral nutritionof black spruce seedlings (Picea mariana [Mill .] B .S .P.) during their second growing period onbud set, bud development, growth, mineral content and cold tolerance . Bud break and growthafter bud break were also studied . Seedlings were grown for 106 d in growth chambers underthree temperature regimes in combination with three concentrations of a fertilizer . They werethen cold hardened for 56 d and dehardened for 66 d .

Under these near- and sub-optimal N levels, bud formation occurred during the growingseason. Bud formation was accelerated with decreasing fertilization, but was not affectedby temperature treatments . Needles from seedlings with 0.64% N (dry mass basis) beforehardening did not harden . Those with 0.87% N showed a lesser degree of hardiness than thosewith 1.28% N. Stem diameter increased at the beginning of the hardening period . During thisacclimation period, shoot dry mass decreased with time at a constant rate and at the same rateover time for all treatments whereas root dry mass was more variable . Total number of needleprimordia was low and no difference was observed among growing conditions . Bud breakwas similar in all treatments . Following bud break, shoot height and stem diameter increaseswere small but their magnitude varied with the nutritional regimes applied during the previousgrowing period. During hardening, nitrogen concentration of shoot tissues first increased andthen decreased ; phosphorus concentration first increased and then remained stable ; potassiumconcentration remained stable . Concentration of these three elements generally decreased inthe roots during this hardening .

Introduction

Cold hardening of conifer seedlings intended for reforestation is a majorconcern among forestry nursery growers . In temperate regions, seedlings are

244

lost each year during fall and spring as a consequence of incomplete hardeningand premature dehardening (Levitt 1980) . The influence of cultural practicesduring the growing season has a considerable effect on these physiologicalprocesses (Sandvik 1980) . As black spruce is the most widely used speciesfor reforestation in Eastern Canada (Margolis and Brand 1990), millions ofseedlings could be lost each year if cultural practices, particularly mineralnutrition and growing temperature, are inadequately managed .

Taken as a single factor, the influence of mineral nutrition on hardeningand dehardening of woody plants has not been studied frequently for coniferseedlings (Pellett and Carter 1981 ; Bigras 1987; van den Driessche 1991) . Nfertilization is critical : too low or too high an N content in the seedlings couldimpair the cold hardening process . It is generally recognized that there existsan optimum N concentration for hardening . For most species, the optimal, sub-optimal and luxury values for nitrogen content have not been determined inrelation to cold hardening . Furthermore, the influence of sub-optimal N valueson hardening has not been clearly demonstrated experimentally. Mineral nutri-tion has also been studied in relation to bud break and growth after bud break(Benzian et al. 1974; van den Driessche 1991) . The influence of temperatureon growth processes, hardening and the accumulation of heat units for budbreak are well documented (Owston and Kowloski 1981 ; Dormling et al .1968). The joint influence of temperature and fertilization during the grow-ing period on the subsequent hardening and dehardening period of coniferseedlings is less well documented .

The objectives of this experiment were : (i) to evaluate the joint influenceof low mineral nutrition and growing temperature of black spruce (Piceamariana [Mill .] B.S .P.) seedlings at the end of the second growing seasonon bud set, bud development, growth, mineral content, and cold tolerance ;(ii) their effect on subsequent bud break and growth after bud break ; (iii) tomake recommendations for an optimal N content for hardening in relationto the growing temperature. Test temperatures were chosen to correspondto a range of degree-days observed in forestry nurseries in Quebec . Mineralnutrition levels were chosen to provide seedlings with a range of nutrientconcentrations from sub-optimal to near optimal .

Materials and methods

Plant material

Black spruce seedlings (Picea mariana [Mill .] B.S.P.) (provenance: 49'10'N,70°05' W) were grown in multicellular containers (45 cells per container,110 cm3 per cell ; IPL, Saint-Damien, Bellechasse, QC) in a peat moss-

vermiculite substrate (4:1, v/v) at a commercial forest tree nursery (47°02' N,70°55' W) under unheated polyhouses and normal nursery conditions forcontainer seedling production in Quebec (Anonymous 1984, 1990) . Seedswere sown in May 1988, and liquid fertilizer was applied from 27 June until26 September (total of 8 .2 mg N, 5.1 mg P, 4 .6 mg K per container cell) . InOctober, seedlings were placed outside for the winter. In the following spring(1 May 1989), 72 containers of 45 seedlings (shoot height, 6.8 cm; diameter,1 .02 mm; shoot dry mass, 140.6 mg; root dry mass, 61 .52 mg) were shippedto the Laurentian Forestry Centre, Sainte-Foy, QC .

Experimental conditions

Experimental treatments were applied to the seedlings during their secondgrowing season. Detailed experimental conditions during the growing, hard-ening, and dehardening periods are shown in Table 1 . During the growingperiod (days 0 to 106), seedlings were exposed to three day :night temperatureregimes, 17 :12° (1522 degree-days, DD), 19 :14° (1732 DD), and 21 :16 °C(1942 DD) under three concentrations of fertilizer, 12, 24, and 48 mg N percontainer cell hereafter referred to as low, intermediate and high fertilizationlevels. After this growing period, seedlings were transferred from growthchambers to a greenhouse (days 107 to 112) and irrigated frequently to elimi-nate fertilizer from the substrate until the electrical conductivity of the leachedwater was near zero . This procedure was used to eliminate the influence offertilization during the cold hardening process ; our goal was to cold-acclimatethe seedlings at their tissue nutrient concentration at the end of the growingseason. From day 113 to 168, seedlings were hardened and from day 169to 234 they were dehardened . No fertilizer was added during hardening anddehardening . As of day 113, average seedling N content was 0 .64, 0.87, and1 .28% for seedlings under low, intermediate and high fertilization regimes,respectively .

Bud formation

On day 86, the number of seedlings with formed apical buds was noted . Budswere considered formed when pale brown or straw yellow scales were visible .

Freezing tests and viability tests

Frost tolerance of excised needles was determined on days 113, 120, 127, and134; failure of the freezing equipment prevented further tests . Needles wereexcised on the smallest portion of the stem located 2 cm below the apicalmeristem of the seedlings, wrapped in damp paper (Keen-White) and placed

245

246

Table 1. Experimental conditions during the growing, hardening, and dehardening periods .

' GC, growth chambers. The photon flux density (PAR) in the growth chambers was250 µmol.m-2 •s-' at the top of the plants (fluorescent : incandescent ratio, 16 :12), therelative humidity was around 60% ; seedlings were watered with deionized water tomaintain a total mass (container + seedlings + substrate) of 3000 g ; Gr, greenhouse .2 20-20-20 fertilizer applied on days 2, 33, 64, and 94 for a total of 12, 24, or 48 mgper container cell for the growing period . In addition, each seedling received 0 .25 mgmagnesium (MgSO4 .7H20) .3 Day:night temperature during the growing period was on a 12 h :12 h basis; 17 :12°,19 :14°, and 21 :16° correspond to 1522, 1732, and 1942 DD respectively.4 Min-max during the day: min-max during the night.

in 13-mm test tubes, at the bottom of which was placed a wet polystyrenefoam stopper to help induce ice crystal nucleation . Test tubes were placed ina modified freezer (model E271, W.C. Wood Co. Ltd., Guelph, ON, Canada)with a programmable controller (model 2010, LFE Corp ., Clinton, MA,U.S.A.) . Temperatures were monitored inside the test tubes with thermo-couples connected to a datalogger (model 21X, Campbell Scientific, Logan,UT, U.S .A .). Air temperature in the freezer was maintained at 0 °C for 15 h .Afterwards, it was lowered at a rate of 3 °C per hour, followed by 1-h plateausat each test temperature . Each freezing test included 10 temperatures. Afterremoval from the freezer, test tubes were kept at 2 ° C for 2-3 d for slow thaw-ing. Needles were then placed on damp filter paper in petri dishes at roomtemperature for 10-12 d . Needle damage was evaluated using the followingcategories: 0, no damage ; 1, less than one-quarter of the length of needlesshowing gray-green, brown or red color ; 2, one-quarter to less than one-half

Condition Location' Day Photoperiod(h)

Day:nighttemperature (°C)

Growing, GC 0-106 16 21:16319 :1417:12

Gr 107-112 16 26.3-35 .0 :17 .9-21 .04

Hardening GC 113-119 8 15:10GC 120-126 8 10:5GC 127-133 8 5:5GC 134-140 0 2GC 141-147 0 0GC 14S-168 0 -2

Dehardening GC 169-203 16 20:15Gr 204-234 16 21.6-24.2 :

14.4-20 .24

247

affected; 3, one-half to less than three-quarters affected ; and 4, three-quartersaffected to entirely damaged. The Spearman-Karber method (Bittenbenderand Howell 1974) was used to estimate the temperature (T50) that damaged50% of the total length of the needles and its variance .

Morphology, phenology, and mineral concentrations

Shoot height was measured on day 113 . During the hardening period fromday 113 to day 169, stem diameter, shoot and root dry mass and the number ofprimordia were observed every 7 d . During dehardening from day 169 to day234, stem diameter was measured on days 178, 210, and 234 and shoot elonga-tion on day 234. Bud opening was recorded during dehardening from day 169until buds were broken; buds were considered broken when scales were partedwith green color visible between bud scales. Shoot elongation was measuredwith a ruler to the nearest millimeter, and stem diameter was measured 1 cmabove the substrate level using a digital caliper (0 .01-200 mm, Digimatic,Mitutoyo Corp., Japan) . Shoots and roots were dried in an oven at 70 °C for24 h and weighed. Primordia were counted following the method describedin Templeton et al. (1991) . Mineral concentrations were determined on days113,141, and 169 during hardening, and on day 234 at the end of dehardeningon a dry matter basis for both shoots and roots . Nitrogen concentration wasdetermined by the Kjeldahl method (Bremner and Mulvaney 1982). Phos-phorus was determined by colorimetry of phospho-molybdic complex (Olsenand Sommers 1982) and potassium, by atomic absorption spectrophotometry(Perkin-Elmer, model 2380) (Chapman and Pratt 1961) .

Experimental design and statistical analyses

For most variables, the experimental design was a split-split-plot . The threetemperature regimes (17:12°, 19 :14°, 21 :16 °C; 12h: 12h) were completelyrandomized among six growth chambers, the main experimental units, withtwo growth chambers per temperature regime. Fertilization levels (total of 12,24, 48 mg N per container cell) were randomized among 12 containers withineach growth chamber, there were four 45-cavity containers per fertilizationlevel in each growth chamber. Temperature effects (T), fertilization effects(F), time effects (Ti), and interactions thereof were considered fixed . Interestlies in the fixed effects, but it is advantageous to first reduce the randompart of the models if some of the random effects can confidently be assumednegligible; this provides simpler models, more powerful tests of significance,and avoids zero or negative variance components (Milliken and Johnson1984, p . 262) .

248

Temperature, fertilization, and time sums of squares were partitioned intoorthogonal polynomials; interactions were partitioned accordingly . The struc-ture of the analyses of variance is shown in Tables 2, 3, and 4 but in orderto keep these tables as simple as possible, only main effects are presented .Results of polynomial contrasts are presented in the Results section only.The data from mineral analyses (N, P, K) in the hardening and dehardeningperiods were analyzed within a single model, and non-orthogonal contrastswere constructed to investigate the effect of time within the hardening period,and to compare the mineral concentrations at the end of the hardening period(day 169) with those at the end of dehardening (day 234) .

Polynomial response surfaces were adjusted to the least squares means .The polynomials were functions of time, fertilization level, and temperatureregime. These terms, powers thereof, products and products of powers wereincluded in the regression model if the corresponding contrast in the analysisof variance was significant, and if the overall interaction or main effect,of which these contrasts were a component, was also significant . Repeatedmeasurements of stem diameters during hardening were analyzed followingthe method described by Rowell and Walters (1976), and similarly for stemdiameters during dehardening .

Transformations were required to stabilize the residual error variance ofseveral variables . The number of apical buds formed on day 86 among thoseof 38 seedlings was transformed to its empirical logit and weighted by theinverse of the estimated variance of the logit (Cox 1970, pp . 33-34) . T50's

were weighted by the inverse of their estimated variance . Shoot dry mass,shoot height on day 113, root dry mass, and days-to-bud-break requireda transformation to their natural logarithm while the number of primordiarequired a square root transformation . All means and response surfaces werecomputed on the transformed scales and back-transformed for presentationin the figures .

Sampling

Bud formation was noted in a group of 38 seedlings for each fertilizationlevel within each growth chamber. For the freezing tests, 25 needles wereremoved from each of 10 randomly selected seedlings per fertilization levelper growth chamber on each sampling date . Needles from the 10 seedlingswere mixed and 15 randomly selected needles were allocated to each of10 test temperatures (total 150 needles) . During hardening, stem diameterwas measured repeatedly on 8 randomly selected seedlings per fertilizationlevel per growth chamber, shoot and root dry mass were measured on 6 newseedlings per fertilization per growth chamber on each date, and primordiawere counted on 8 new seedlings from each fertilization level and each

Tabl

e 2

.An

alys

is o

f va

rian

ce o

f fr

ost

tole

ranc

e an

d mo

rpho

logi

cal

vari

able

s du

ring

hardening of black spruce seedlings

.

1An asterisk indicates an interaction

.2Dashes appear when a random effect was removed from

the

mode

l in

the

red

ucti

onprocess.

Degr

ees

of f

reed

om(D

.F.),

mean squarest(M.S

.),and significance levels (P>F)

Dry

mass

Sour

ce o

f va

riat

ion

Bud formation

T50

Stem diameter

Shoot

Root

Prim

ordi

a

D.F

.M

.S. P>F

D.F

.M

.S.P>F

D.F

.M

.S.

P>F

D.F

.M

.S.

P>F

D.F

.M

.S.

P>F

D.F

.M

.S.

P>F

Temperature (T)

25.07

0.2841

23.

20.8

968

21.

593

0.2

478

20.003

0.9870

20.413

0.24

492

45.3

80.7

769

Error a

--

--

--

--

-3

0.736

--

316

4.9

3

Fert

iliz

atio

n (F

)2

14.83

0.0163

22735

.70.0001

2112.654

0.0001

235

.002

0.00

012

11.4

390.

0001

268

.13

0.1

893

T*F

40.92

0.7966

419

5.5

0.0

004

41 .

550

0.0001

40.

342

0.00

014

0.11

60.698

54

10.5

30.8

397

Erro

r b

93.

61-

--

--

--

90.309

630

.61

Time (D)

312

81.0

0.0001

80.988

0.0001

80.115

0.00

018

1.150

0.00

018

127.4

60.0

001

T*D

623

.20.5

737

160.053

0.0004

160.

028

0.96

1716

0.059

0.71

0316

12.0

30.5

568

F*D

6343

.30.0001

160.3

210.0001

160.053

0.4164

160.120

0.09

7216

11.2

10.6

308

T*F*D

1212

8.1

0.0

003

320.118

0.0001

320.058

0.70

6532

0.063

0.78

1332

15.64

0.2

236

Error c

3428

.971

0.123

780.053

720.

070

--

-

250

Table 3. Analysis of variance of bud break, shoot elongation, and stem diameter duringdehardening of black spruce seedlings .

' An asterisk indicates an interaction .2 Dashes appear when a random effect was removed from the model in the reduction process .

growth chamber on each occasion; there were 9 measurement dates for stemdiameter, dry mass, and primordia . During dehardening, stem diameter of 20seedlings per fertilization level per growth chamber was measured repeatedlyon 3 dates, and elongation of their shoot was measured on one date (day234) . Bud break was also observed on 20 seedlings; it was expressed in daysbetween the beginning of dehardening and a fixed stage of bud opening .Mineral analysis was performed on shoots and roots of one seedling perfertilization per growth chamber on each of 4 sampling dates, three duringhardening and one during dehardening .

Results

Bud formation

The percentage of formed apical buds increased linearly (P = 0.007 for Flinear) with decreasing levels of fertilization with averages of 87, 75, and64% at the 12, 24, and 48 mg N per cell levels of fertilization, respectively.Growing temperature had no apparent effect on this variable (P = 0.284,Table 2) . At the beginning of the hardening period on day 113, all buds wereformed .

Degrees of freedom (D.F.), mean squarest (M.S .)and significance levels (P>F)

Bud break Shoot elongation Stem diameterSource of variation' D.F. M.S. P>F D.F. M.S. P>F D.F. M.S . P>F

Temperature (T) 2 0.001 0.9913 2 494 0.6996 2 0.150 0.4988Error a - - - 3 1232 - - -Fertilization (F) 2 0.139 0.2860 2 2782 0.0001 2 115.294 0.0001T*F 4 0.056 0.6824 4 109 0.3255 4 0.957 0.0785Error b 9 0.097 - - - -Time (D) 2 8.823 0.0001T*D 4 0.209 0.0849F*D 4 0.571 0.0005T*F*D 8 0.112 0.3941Error c 18 0.152

Tabl

e 4.

Anal

ysis

of

vari

ance

of

mine

ral

conc

entr

atio

ns d

urin

g ha

rden

ing

and

deha

rden

ing

.

1An asterisk indicates an interaction

.2

Dash

es a

ppea

r wh

en a

ran

dom

effe

ct w

as r

emov

ed f

rom the model in the reduction process

.

Degrees

of f

reed

om(D

.F.),

mean squarest(M

.S.),

and

sign

ific

ance

lev

els

(P>F

)Shoot

Root

Sour

ce o

f va

riat

ion'

NP

KN

PK

D.F. M.

S.P>F

D.F. M

.S.P>F

D.F. M

.S.P>F

D.F. M

.S.

P>F

D.F. M.

S.P>F

D.F.

M.S.

P>F

Temp

erat

ure

(T)

20.

023

0.49

422

0.001

0.98

092

0.215

0.81

392

0.009

0.41

322

0.046

0.68

062

0.411

0.8411

Error a

30.

026

--

30.

973

--

-3

0.104

32.

360

Fertilization (F)

21.

801

0.00

012

1.05

40.

0001

23.

167

0.00

012

1.689

0.00

012

1.535

0.0001

21.735

0.00

07T*

F4

0.012

0.37

134

0.02

50.

6141

40.

277

0.25

564

0.004

0.79

824

0.030

0.29

084

0.180

0.4461

Erro

r b

--

--

--

--

--

--

--

--

--

Time

(D)

30 .

502

0.00

013

2.345

0.0001

36.

319

0.0001

30.

529

0.00

013

4.796

0.0001

356

.990

0.0001

T*D

60.

011

0.41

506

0.026

0.63

556

0.150

0.60

876

0.023

0.04

646

0.040

0.14

506

0.098

0.78

77F*

D6

0.048

0.00

166

0.09

30.

0367

60.

698

0.00

846

0.054

0.00

036

0.243

0.0001

60.

876

0.00

15T*F*D

120.

011

0.45

8012

0.051

0.21

0012

0.156

0.65

8912

0.01

10.

3764

120 .

016

0.7591

120.

109

0.84

15Error c

330.

010

360.

037

330.

198

360.

010

330.

023

330.

188

252

v0nH

-4

-p)

Figure 1 . Frost tolerance (Tso) of excised needles during hardening of black spruce seedlingsexposed to day :night temperature regimes of 17 :12°, 19:14°, and 21 :16 °C in combinationwith mineral concentrations of 12 (0), 24 (0), and 48 (A) mg N per container cell duringtheir growing period . Each symbol is the mean of 2 observations for the T50 .

Frost tolerance

Needles showed a slight frost tolerance of -6° to -8 °C before hardening(Figure 1). Needle hardening evolved differently over time depending ontemperature and fertilization level (P < 0 .001 for the interaction between T,F and D, Table 2 and Figure 1) . Under growing temperature regimes of 19:14or 21 :16 °C, needles from seedlings grown at the low fertilization level didnot harden while those from seedlings grown under high fertilization showedthe highest level of frost tolerance ; needles from seedlings that received theintermediate fertilization level also had intermediate frost tolerance . Underthe 17:12 °C temperature regime, needles from the more fertilized seedlingsdid not seem to have an advantage over those from seedlings that receivedthe intermediate level of fertilization . Maximal cold tolerance was obtainedwith the highest growing temperature (21 :16 °C) combined with 48 mg Nper container cell . Our results are for the early acclimation period only anddo not show the full hardening potential of this species .

Morphology and phenology

HardeningFertilization and time had a significant effect on the variables of morphologyand phenology while growing temperature had no effect on these variables(P > 0.245, Tables 2 and 3) .

At the beginning of hardening (day 113), average shoot height was 132,164, and 188 mm for 12, 24, and 48 mg N per container cell fertilizationlevels, respectively .

Over the nine weeks of hardening, stem diameter increased from day 113to day 127 and remained stable for most of the hardening period thereafter

17:12'C

19:14'C

21:16'CO

a

4

113

120

127

134

113

120

07

134

113

120

127

134Days since start of experiment

EE

Eay

E0V1

0 .4

B

0.9-

90-~

~o0.8- v v 80-

0.7- °

- _o

a -omss

E 70-

0.6 -

d so-e0.5 - -', 110

50-

s

-a-r_aa

oo

0

--o--e--4--0----0

0

0

0

253

0

40113

127

141

155

169

113

127

141

155

169

Days since start of experiment

Figure 2 . (A) Stem diameter, (B) shoot dry mass, (C) root dry mass, (D) and needle primordiaformation during hardening of black spruce seedlings exposed to day: night temperatureregimesof 17 :12 ° , 19 :14°, and 21 :16 °C in combination with mineral concentrations of 12 (O), 24(o), and 48 (o) mg N per container cell during their growing period . Each symbol is the meanof 16 observations for stem diameter, 12 for shoot dry mass, 36 for root dry mass and 144 forneedle primordia . Only significant effects are presented .

(P < 0.001 for Ti*F, Table 2, Figure 2A). Shoot dry mass decreased linearlyover time (P < 0.001) at a constant rate for all fertilization (P = 0.416 forTi*F, Table 2, Figure 2B) . Root dry mass first increased from day 113 to day127, it then declined until day 162, and finally increased again slightly inthe last week of hardening (P < 0.001 for Ti, Table 2). At 24 and 48 mg Nper container cell, root dry mass was about 1 .5 times higher than at 12 mgN per container cell (P < 0 .001 for F, Table 2, Figure 2C). The number ofprimordia increased mainly during the first half of the hardening period andremained relatively stable thereafter (P < 0 .001 for Ti, Table 2, Figure 2D) .There was no evidence that fertilization regime had any effect on the numberof primordia (P = 0.189) .

DehardeningDuring dehardening, bud break occurred at approximately the same timewhatever the fertilization level (P = 0 .286 for F, Table 3) . Shoot elonga-

4.0- Aa

2.0-

a a 1 .6 -3.5-E 1.4-

1.2-3.0 - w 1.0-

0

o~.o'-4- -a-~-V--oeo 0H

0.8 -

2 .5 r I

I

I

I

1.0- C 100 -

254

tion after bud break increased from 14 .7 mm to 24.1 mm when fertilizationincreased from 12 to 24 mg N per container cell, and remained stable whenN was further increased to 48 mg per container cell (P < 0 .001 for F, Table3). After bud break, stem diameter was larger in the more fertilized seedlings(P < 0.001 for F, Table 3) with averages of 4 .1, 3.5, and 3 .0 on day 234 for12, 24, and 48 mg N per container cell respectively .

Mineral concentrationDuring hardening and dehardening, temperature had a negligible effect onmineral concentration (P > 0.411 for the effect of T on N, P and K of shootand root, Table 4 and Figure 3) . During hardening from day 113 to 169,nitrogen concentrations in shoot tissues first increased until day 141 and thendecreased (P < 0.001 for Ti quadratic). Phosphorus in shoots also increasedfrom day 113 to day 141 but remained high during the rest of the hardeningperiod (P < 0 .001 for Ti linear and quadratic) . Potassium concentration inshoots remained stable during hardening (P = 0 .330 for Ti). During the sameperiod, nitrogen, phosphorus, and potassium concentrations decreased in roottissues (P < 0 .001 for Ti) . At the end of the experiment on day 234, adecrease in mineral concentration was observed in both shoot and root tissuesas compared with day 169 (P < 0.001 for all contrasts between day 169 andday 234).

Discussion

The low range of temperature used in this experiment has probably eliminatedthe overall influence of temperature on the measured variables . The influenceof fertilization and time were the most important on these variables . Therefore,the significant interactions encountered were mainly attributed to the influenceof fertilization and time .

At the beginning of hardening, black spruce seedlings showed low nutri-ent shoot content ranging from 0.64 to 1 .28% N, 0.14 to 0 .19% P, and 0 .55to 0.65% K. Swan (1960) argued that N content lower than 1 .3% should beconsidered deficient for 14-week-old black spruce seedlings . In containerizedblack spruce nurseries, Langlois (1990) reported that the mineral concentra-tions for whole seedlings ranged from 1 .60 to 1 .82% for N, 0 .28 to 0.31 % forP and 0.77 to 0.87% for K at the end of the second growing season ; slightylower concentrations should be expected for the shoot parts alone . Therefore,seedlings used in this experiment could be considered as a low limit value at1 .28% N and deficient at lower values .

Within the low range of N content, our results show that a deficient foliarconcentration of nitrogen (0.64% shoot dry mass basis) in 17-month-old black

.- . 1.2 -

at

Z

. . 0.25-

It

CL

1 .6 -

0.0-

0 .4

0 .35-

0.15-

0 .05

0 .e

0 .7

0.5

0•'

4

Shoot

- --a- -''

0

00

-e

11

Root

D

255

9

se

0.3

,

,

,

, ,113

141

16e

234

113

141

16e

234

Days since start of experiment

Figure 3 . Mineral concentration of shoot and root tissues of black spruce seedlings duringhardening and dehardening after 106 days of growth under day :night temperature regimesof 17 :12°, 19:14°, and 21 :16 °C in combination with mineral concentrations of 12 (0), 24(o), and 48 (o) mg N per container cell . Each symbol is the mean of 6 observations . Onlysignificant effects are presented .

spruce seedlings limits the hardening process during the first stage of accli-mation while a 1 .28% nitrogen concentration offers, under our conditions,a maximal level of hardiness. Seedlings grown at the intermediate level offertilization had 0 .87% N content at the beginning of hardening ; their levelof frost hardiness was also intermediate. Since freezing tests were performedfor only four weeks, it is possible that further hardening could have beenobtained. Low N content could have delayed achievement of frost toleranceor decreased the maximal level attainable . It is generally recognized that lowlevels of N are related to poor frost hardiness (Levitt 1956), but the range oflimiting concentrations is not known for most species . Christersson (1975)reported no hardening problems with concentrations of 0 .75 and 0.79% Nfor 6-month-old Norway spruce (Picea abies [L .] Karst.). With Juniperuschinensis L. "Hetzi", Pellett and White (1969) reported no differences inshoot frost tolerance with N concentration between 0 .8 and 1 .8%. On the

256

other hand, a high level of nitrogen has been reported as reducing hardi-ness (van den Driessche 1991) . In our experiment, seedling N concentrationsnever exceeded 1.28% and thus no recommendations can be made on theupper limit of N concentration in relation to cold hardening . In nurseries,low levels of N could clearly limit cold hardening of black spruce foliagein the early acclimation period, which is a particularly vulnerable stage forfrost damage in nurseries since spells of freezing temperatures can occur atthat time . Fertilization programs should thus ensure that a concentration of atleast 1 .28% N is obtained in foliage tissues of black spruce seedlings at thebeginning of their second hardening period .

The influence of phosphorus or potassium on cold hardiness of coniferseedlings showed conflicting results . However, after reviewing the literature,van den Driessche (1991) reported that it is unlikely that P has much effecton winter cold hardiness and that little correlation between tissue K concen-trations and cold hardiness has been found in a number of experiments .

Needles show a slight frost tolerance of -6° to -8 °C before hardening,probably as a consequence of bud formation during the growing period .Bigras and D'Aoust (1992) have reported a frost tolerance of -5°C forneedles of 4-month-old black spruce seedlings before hardening . The degreeof hardiness is however not related to bud formation since buds at the lowestN concentration show the least tolerant foliage even if their buds formedearlier than those of seedlings at higher fertilization levels .

Bud formation occurred during the growing period. This could beexplained by low mineral content of the seedlings . Our results corroboratethose of Macey and Arnott (1986) who showed that periodic moisture stressand nutrient withdrawal were effective in reducing shoot length and induc-ing bud formation in white spruce (Picea glauca [Moench] Voss) seedlings .Young and Hanover (1977) also reported that low nitrogen concentrationsstimulated bud set on Picea pungens Engelm. However, since low mineralcontent limits cold hardiness, it cannot be recommended to stimulate bud setfor black spruce seedlings .

There was no evidence that fertilization or growing temperature had anyeffect on the number of primordia . In our experiment, the total number ofabout 90 needle primordia was low and was probably the consequence of lowmineral content during their formation . Colombo et al . (1982) reported thatover 200 needle primordia can be formed on black spruce seedlings underoptimal conditions . On the other hand, Pollard (1974, for white spruce) andYoung and Hanover (1977, for blue spruce, Picea pungens) reported that thenumber of formed primordia is related to seedling size; the taller the seedlings,the greater the number of needle primordia . In our experiment, this relationcould not be verified because the range of seedling size was too narrow.

257

During hardening, shoot elongation stopped when terminal buds wereformed; stem diameter continued to increase over the first two weeks of hard-ening however. Root dry mass increased in the first three weeks of hardeningand decreased afterwards while shoot dry mass started to decrease as soonas seedlings were submitted to hardening conditions . The decrease in rootand shoot dry mass could be due to the low levels of nutrient content atthe beginning of the hardening period, with leaching of the substrate havingfurther contributed to the absence of absorption . In nurseries, where optimallevels of mineral nutrition are practiced, diameter and root dry mass increaseduring the fall (Langlois 1990) . Also, in our experiment, photosynthesisgradually became less efficient because temperatures decreased (and lightswere turned off) while respiration continued ; the conjunction of these twophenomena may have caused a decrease in shoot and root dry mass . McGregorand Kramer (1963) observed that from September until the end of February,photosynthesis expressed per unit of foliage decreased to zero on loblollypine (Pinus taeda L.) and white pine (Pinus strobus L.) . During that time,respiration remained constant for white pine and increased for loblolly pine .Their observation lends support to our conjecture .

In our experiment, bud break occurred at the same time despite differencesin mineral concentrations that resulted in a fairly wide range of N foliarconcentration (0 .64 to 1 .31%) before bud break . Benzian et al. (1974) andvan den Driessche (1991) reported that high nitrogen content accelerates budbreak; however in these experiments, seedlings had a higher level of N thanin our test.

Nutritional regimes during the growing period in our experiment influ-enced shoot elongation after bud break ; seedlings treated with 24 and 48 mgN per container cell had higher elongations than those treated with 12 mg N percontainer cell. However, even the greater elongations were small, and couldreflect the low nutrient content of the seedlings before bud break . Benzian etal. (1974) showed that N application in the nursery has an influence on thegrowth of the seedlings after plantation . However, species differed in theirresponse: Sitka spruce (Picea sitchensis [Bong.] Carr.) increased up to 18%in height when N concentration went from 0 .81 to 1 .40% whereas Norwayspruce and Western hemlock (Tsuga heterophylla (Raf.) Sarg.) showed onlya small response to an increase in N fertilization in the nursery . There was noevidence of any interaction between the effects of N fertilization and growingtemperature .

258

Conclusions

Black spruce seedlings with 1 .28% N at the beginning of cold acclimationin the second growing season : (i) were more frost tolerant during the earlyacclimation period; (ii) had larger stem diameters during hardening and (iii)had higher shoot elongations after dehardening compared with seedlingswith lower N content. Forest nurseries should be aware of the influence ofthe fertilization regime used during the growing period on the cold hardeningand subsequent growth of the seedlings .

Acknowledgments

The authors wish to express their gratitude to Mr. Yves Dubuc, Mr. ReneTurcotte, and Mrs . Diane Trudel for their technical support during theexperimentation . We also thank Mrs . Michele Bernier-Cardou for statisticalsupervision .

References

Anonymous 1984. Deuxii me atelier de travail sur la culture des semis en recipients . Servicecanadien des forets . Ministi re de l'Energie et des Ressources du Quebec . Ordre desingenieurs forestiers du Quebec. Sainte-Foy (QC), 29-30 nov. 139 pp.

Anonymous 1990 . Troisii me atelier quebecois sur la culture des plants forestiers en recipients .Gouv. du Quebec, ministi re de l'Energie et des Ressources (forets). Direction de larecherche et du developpement. Service du transfert de technologie . 341 pp .

Benzian, B ., Brown, R.M. and Freeman, S .C.R. 1974. Effect of late-season top-dressings ofN (and K) applied to conifer transplants in the nursery on their survival and growth onBritish forest sites . Forestry 47: 153-184.

Bigras, F.J . 1987. Influence de la temperature, de la photoperiode et de la nutrition minerale surl'acclimatation et la tolerance au gel de Juniperus chinensis "Pfitzerana" et de Juniperuschinensis "Pfitzerana aurea" . Ph .D. thesis . Universite Laval . Quebec (QC) . 153 pp.

Bigras, FJ . and D'Aoust, A .L. 1992. Hardening and dehardening of shoots and roots ofcontainerized black spruce and white spruce seedlings under short and long days. Can . J .For. Res . 22 : 388-396 .

Bittenbender, H .C. and Howell, G .S . Jr. 1974. Adaptation of the Spearman-Karber method forestimating the T5o of cold stressed flower buds. J. Am. Soc. Hort. Sci. 99 : 187-190 .

Bremner, J.M. and Mulvaney, C.S . 1982 . Nitrogen-total, pp . 595-624 . In : Page, A.L ., Miller,R.H. and Keeney, D .R. (Eds) Methods of Soil Analysis, Part 2. Chemical and Microbio-logical Properties, Vol . 9. Agron . Monogr. (Madison) .

Chapman, H.D. and Pratt, P.E. 1961 . Methods of Analysis for Soils, Plants and Waters .University of California, Division of Agricultural Science, Riverside, Calif. 309 pp .

Christersson, L. 1975 . Frost hardiness development in rapid- and slow-growing Norway spruceseedlings . Can . J . For. Res . 5 : 340-343 .

Colombo, S .J ., Webb, D .P. and Glerum, C . 1982 . Cold hardiness and bud development undershort days in black spruce and white spruce seedlings, pp . 171-176. In: Scarratt, J.B .,Glerum, C . and Plexman, C .A. (Eds) Proceedings, Canadian Containerized Tree Seedling

259

Symposium, Canadian Forest Service, Great Lakes Forestry Centre, COJFRC Symp . Proc .O-P-10.

Cox, D .R. 1970. Analysis of Binary Data . Methuen, London, 142 pp .Dormling, I. and Gustafsson, A . and von Wettstein, D . 1968 . The experimental control of the

life cycle in Picea abies (L .) Karst . I . Some basic experiments on the vegetative cycle .Silvae Genet. 17 : 11 64.

Langlois, C.-G . 1990. Le dynamisme de la croissance des plants cultiv8s dans le recipient45-110 et les relations entre les variables, pp . 1-24. In: Troisieme atelier qu6b6cois sur laculture des plants forestiers en recipients . Gouv. du Quebec, minist8re de 1'Energie et desRessources (forets) . Direction de la recherche et du ddveloppement. Service du transfertde technologie .

Levitt, J . 1956. The Hardiness of Plants . Academic Press, N.Y. 278 pp .Levitt, J . 1980 . Responses of Plants to Environmental Stresses . Vol. 1 . Chilling, Freezing, and

High Temperature Stresses . Academic Press, New York. 497 pp .Macey, D .E. and Amott, J.T. 1986 . The effect of moderate moisture and nutrient stress on bud

formation and growth of container-grown white spruce seedlings . Can. J . For. Res . 16 :949-954.

Margolis, H .A. and Brand, D.G. 1990 . An ecophysiological basis for understanding plantationestablishment. Can. J . For. Res . 20 : 375-390 .

McGregor, W.H.D. and Kramer, P.J . 1963 . Seasonal trends in rates of photosynthesis andrespiration of loblolly pine . Am . J . Bot . 50: 760-765 .

Milliken, G.A. and Johnson, D .E . 1984 . Analysis of messy data . Vol . 1 : Designed experiments .Van Nostrand Reinhold, New York. 473 pp .

Olsen, S .R. and Sommers, L .E . 1982 . Phosphorus, pp . 403-430 . In: Page, A.L., Miller, R .H .and Keeney, D .R. (Eds) Methods of Soil Analysis, Part 2 . Chemical and MicrobiologicalProperties, Vol . 9 . Agron . Monogr. (Madison) .

Owston, P.W. and Kozlowski, T.T. 1981 . Growth and cold hardiness of container-grownDouglas-fir, noble fir, and Sitka spruce seedlings in simulated greenhouses regimes . Can .J . For. Res . 4 : 97-100 .

Pellett, H.M. and Carter, LV 1981 . Effect of nutritional factors on cold hardiness of plants .Hort. Rev . 3 : 144-171 .

Pellett, N .E . and White, D .B . 1969 . Effect of soil nitrogen and soil moisture levels on the coldacclimation of container grown Juniperus chinensis "Hetzi". J. Am. Soc. Hort. Sci . 94:457-459.

Pollard, D .F.W. 1974. Seedling size and age as factors of morphogenesis in white spruce Piceaglauca (Moench) Voss buds . Can. J . For. Res . 4: 97-100 .

Rowell, J.G. and Walters, D .E. 1976. Analysing data with repeated observations on eachexperimental unit. J . Agric . Sci . 87 : 423-432 .

Sandvik, M . 1980 . Environmental control of winter stress tolerance and growth potential inseedlings of Picea abies (L.) Karst . Part B . Use of controlled environment facilities . N .Z .J . For. Sci . 10 : 97-104.

SAS Institute, Inc . 1990 . SAS/Stat User's Manual . Volume 2 . Carey, North Carolina .Swan, H .S.D. 1965 . Studies of the mineral nutrition of Canadian pulpwood species. Phase II.

Fertilizer pellet field trials . 1959-1963 . Final report. Pulp and Paper Research Institute ofCanada . Tech . Rep., 1965, 405 .

Templeton, C.W.G ., Odlum, K.D. and Colombo, S .J . 1991 . How to Dissect Spruce Buds.Ontario Ministry of Natural Resources, Sault Ste . Marie, Ontario .

van den Driessche, R . 1991 . Effects of nutrients on stock performance in the forest, pp . 229-260. In : van den Driessche, R. (Ed) Mineral Nutrition of Conifer Seedlings . CRC Press,Boca Raton, Florida .

Young, E. and Hanover, J .W. 1977. Development of the shoot apex of blue spruce (Piceapungens) . Can . J . For. Res . 7: 614-620.